Hypoglycaemia in Clinical

Hypoglycaemia in Clinical Diabetes

Edited by Brian M. Frier, MD, FRCP(Edin), FRCP(Glas) Honorary Professor of Diabetes British Heart Foundation Centre for Cardiovascular Science The Queen’s Medical Research Institute University of Edinburgh; Formerly, Consultant Diabetologist Royal Infirmary of Edinburgh Edinburgh, UK Simon R. Heller, DM, FRCP Professor of Clinical Diabetes and Hon Consultant Physician Director of Research and Development Sheffield Teaching Hospitals Foundation Trust; Department of Human Metabolism University of Sheffield Sheffield, UK Rory J. McCrimmon, MD, FRCP(Edin) Professor of Experimental Diabetes and Metabolism Cardiovascular and Diabetes Medicine Medical Research Institute University of Dundee Dundee, UK

Third Edition This edition first published 2014 © 2014 by John Wiley & Sons, Ltd. Registered office: John Wiley & Sons, Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK Editorial offices: 9600 Garsington Road, Oxford, OX4 2DQ, UK The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK 111 River Street, Hoboken, NJ 07030-5774, USA For details of our global editorial offices, for customer services and for information about how to apply for permission to reuse the copyright material in this book please see our website at www.wiley.com/wiley-blackwell The right of the author to be identified as the author of this work has been asserted in accordance with the UK Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher. Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The publisher is not associated with any product or vendor mentioned in this book. This publication is designed to provide accurate and authoritative information in regard to the subject matter covered. It is sold on the understanding that the publisher is not engaged in rendering professional services. If professional advice or other expert assistance is required, the services of a competent professional should be sought. The contents of this work are intended to further general scientific research, understanding, and discussion only and are not intended and should not be relied upon as recommending or promoting a specific method, diagnosis, or treatment by physicians for any particular patient. The publisher and the author make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of fitness for a particular purpose. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of medicines, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each medicine, equipment, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. Readers should consult with a specialist where appropriate. The fact that an organization or Website is referred to in this work as a citation and/or a potential source of further information does not mean that the author or the publisher endorses the information the organization or Website may provide or recommendations it may make. Further, readers should be aware that Internet Websites listed in this work may have changed or disappeared between when this work was written and when it is read. No warranty may be created or extended by any promotional statements for this work. Neither the publisher nor the author shall be liable for any damages arising herefrom. Library of Congress Cataloging-in-Publication Data Hypoglycaemia in clinical diabetes / edited by Brian M. Frier, Simon R. Heller, Rory J. McCrimmon. – Third edition. p. ; cm. Includes bibliographical references and index. ISBN 978-0-470-67200-6 (cloth : alk. paper) – ISBN 978-1-118-69543-2 (ebook) – ISBN 978-1-118-69590-6 (epdf) – ISBN 978-1-118-69787-0 (epub) – ISBN 978-1-118-69790-0 I. Frier, Brian M., editor of compilation. II. Heller, Simon, editor of compilation. III. McCrimmon, Rory J., editor of compilation. [DNLM: 1. Hypoglycemia–complications. 2. Hypoglycemia–physiopathology. 3. Diabetes Complications. 4. –adverse effects. WK 880] RC662.2 616.4'66–dc23 2013024992 A catalogue record for this book is available from the British Library. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic books. Cover images: Top left image: iStock File #19373251 © Willie B. Thomas. Top right image: iStock File #19332350 © Willie B. Thomas. Bottom right image: iStock File #13348540 © MorePixels. Cover design by Steve Thompson Set in 10/12 pt Times Roman by Toppan Best-set Premedia Limited

1 2014 Contents

List of Contributors vii Preface x

1 Normal Glucose Metabolism and Responses to Hypoglycaemia 1 Ian A. Macdonald and Paromita King

2 Symptoms of Hypoglycaemia and Effects on Mental Performance and Emotions 23 Ian J. Deary and Nicola N. Zammitt

3 Counterregulatory Deficiencies in Diabetes 46 Rory J. McCrimmon

4 Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 63 Mark W.J. Strachan

5 Nocturnal Hypoglycaemia 96 Elaine Chow and Simon R. Heller

6 Impaired Awareness of Hypoglycaemia 114 Brian M. Frier

7 Risks of Intensive Therapy 145 Stephanie A. Amiel

8 Management of Acute and Recurrent Hypoglycaemia 165 Rory J. McCrimmon

9 Technology for Hypoglycaemia: CSII and CGM 184 Josefine E. Schopman and J. Hans DeVries

10 Hypoglycaemia in Children with Diabetes 197 Krystyna A. Matyka

11 Hypoglycaemia During Pregnancy in Women with Pregestational Diabetes 218 Lene Ringholm, Peter Damm and Elisabeth R. Mathiesen vi Contents

12 Hypoglycaemia in Type 2 Diabetes and in Elderly People 230 Nicola N. Zammitt and Brian M. Frier

13 Mortality, Cardiovascular Morbidity and Possible Effects of Hypoglycaemia on Diabetic Complications 263 Elaine Chow, Miles Fisher and Simon R. Heller

14 Long-Term Effects of Hypoglycaemia on Cognitive Function in Diabetes 285 Christopher M. Ryan

15 Neurological Sequelae of Hypoglycaemia 305 Petros Perros and Marit R. Bjørgaas

16 Psychological Effects of Hypoglycaemia 323 Frank J. Snoek, Tibor R.S. Hajos and Stefanie M.P.A. Rondags

17 Exercise Management and Hypoglycaemia in Type 1 Diabetes 334 Ian W. Gallen and Alistair Lumb

18 Living with Hypoglycaemia 347 Brian M. Frier

Index 369 List of Contributors

Stephanie A. Amiel, MD, FRCP J. Hans DeVries, MD, PhD RD Lawrence Professor of Diabetic Consultant Physician and Endocrinologist Medicine Principal Investigator King’s College London Department of Endocrinology London, UK Academic Medical Center Amsterdam, The Netherlands Marit R. Bjørgaas, MD, PhD Adjunct Associate Professor and Senior Miles Fisher, MD, FRCP(Edin), Consultant FRCP(Glas) St. Olavs Hospital Professor Trondheim University Hospital Consultant Physician Trondheim, Norway Glasgow Royal Infirmary; Honorary Professor Elaine Chow, MSc, MBChB, University of Glasgow MRCP(UK) Glasgow, UK Clinical Research Fellow and Honorary Diabetes Registrar Brian M Frier, MD, FRCP(Edin), University of Sheffield FRCP(Glas) Sheffield, UK Honorary Professor of Diabetes British Heart Foundation Centre for Peter Damm, MD, DMSc Cardiovascular Science Professor The Queen’s Medical Research Institute Center for Pregnant Women with Diabetes University of Edinburgh; Department of Obstetrics Formerly, Consultant Diabetologist Rigshospitalet Royal Infirmary of Edinburgh University of Copenhagen Edinburgh, UK Copenhagen, Denmark Ian W. Gallen, MD, FRCP, FRCP(Edin) Ian J. Deary, PhD, FRCP(Edin), Consultant Physician and Endocrinologist FRCPsych, FMedSci Diabetes Centre Professor of Differential Psychology and Royal Berkshire NHS Trust Director of the Centre for Reading, UK Cognitive Ageing and Cognitive Epidemiology University of Edinburgh Edinburgh, UK viii List of Contributors

Tibor R.S. Hajos, PhD Rory J. McCrimmon MD, FRCP(Edin) Researcher Professor of Experimental Diabetes and Department of Medical Psychology Metabolism Diabetes Psychology Research Group Cardiovascular and Diabetes Medicine VU University Medical Centre Medical Research Institute Amsterdam, The Netherlands University of Dundee Dundee, UK Simon R. Heller, DM, FRCP Professor of Clinical Diabetes and Hon Petros Perros, MD, FRCP Consultant Physician Consultant Physician Director of Research and Development Department of Endocrinology Sheffield Teaching Hospitals Foundation Royal Victoria Infirmary Trust; Newcastle upon Tyne, UK Department of Human Metabolism University of Sheffield Lene Ringholm, MD, PhD Sheffield, UK Specialist Registrar Center for Pregnant Women with Diabetes Paromita King, DM, FRCP, Department of Endocrinology FRCP(Edin) Rigshospitalet Consultant Physician University of Copenhagen Derby Hospitals NHS Foundation Trust Denmark Derby, UK Stefanie M.P.A. Rondags, MSc Alistair Lumb, PhD, MRCP(UK) Junior Researcher Specialist Registrar in Diabetes/ Department of Medical Psychology Endocrinology Diabetes Psychology Research Group Buckinghamshire Healthcare Trust VU University Medical Centre Buckinghamshire, UK Amsterdam, The Netherlands

Ian A. Macdonald, PhD, FSB, FAfN Christopher M. Ryan, PhD Professor of Metabolic Physiology Professor University of Nottingham Medical School Department of Psychiatry Nottingham, UK University of Pittsburgh School of Medicine Elisabeth R. Mathiesen, MD, DMSc Pittsburgh, PA, USA Professor Center for Pregnant Women with Diabetes Josefine E. Schopman, MD, PhD Department of Endocrinology Department of Internal Medicine Rigshospitalet Academic Medical Center University of Copenhagen Amsterdam, The Netherlands Copenhagen, Denmark Frank J. Snoek, PhD Krystyna A. Matyka, MD, MRCPCH Professor Associate Clinical Professor Department of Medical Psychology Division of Metabolic and Vascular Health Diabetes Psychology Research Group Warwick University VU University Medical Centre Coventry, UK Amsterdam, The Netherlands List of Contributors ix

Mark W.J. Strachan, MD, FRCP(Edin) Nicola N. Zammitt, MD, FRCP(Edin) Honorary Professor Consultant Physician and Consultant in Diabetes and Endocrinology Honorary Clinical Senior Lecturer Metabolic Unit Royal Infirmary of Edinburgh Western General Hospital; Edinburgh, UK University of Edinburgh Edinburgh, UK Preface

Every field in medicine evolves continuously with the incremental acquisition of scientific knowledge leading to developments in clinical management, and hypoglycaemia is no exception. Since the second edition of this book was published in 2007, the interest of clinicians in hypoglycaemia has escalated. This has been stimulated in part by the unex- pected and disconcerting results of the ACCORD trial that was conducted in people with type 2 diabetes and the association between severe hypoglycaemia and mortality was con- firmed in the ADVANCE and VADT studies. These highlighted the potential hazards associated with strict glycaemic control particularly when other co-morbidities such as coronary heart disease are present, and have focused attention on the need for individualisa- tion of glycaemic targets, especially in vulnerable groups such as the frail and elderly. Most clinicians already recognise that hypoglycaemia is potentially dangerous and even life- threatening, but few appreciated until recently that this side-effect of treatment is actually very common even in type 2 diabetes, and can provoke substantial morbidity (and possibly sudden death), in high risk groups such as those at the extremes of age and in specific situ- ations such as during pregnancy or when undertaking physical exercise. These recent events highlight the importance of physicians and allied health professionals having an in-depth knowledge of hypoglycemia, because an increased understanding of the causes and conse- quences of hypoglycemia are essential to improving the care of people with diabetes and in helping to shape therapeutic strategies and policies. The rapid expansion in scientific knowledge about the effects of hypoglycaemia in humans, with research developments on, for example, how hypoglycaemia affects the brain, has direct relevance to people with diabetes, particularly those receiving treatments that increase the risk of hypoglycaemia. As in previous editions of this book, the emphasis has been maintained on the clinical importance of hypoglycaemia and its effects to the indi- vidual patient in different age-groups and situations, and the topics that are reviewed should be of direct value to all health care professionals who are involved with the management of people with diabetes. In addition to a change in the team of co-editors for this edition, new authors have been enlisted to provide their expertise and detailed knowledge of dif- ferent aspects of hypoglycaemia that are highly relevant to everyday practice. Chapters have been added on exercise, psychological effects and the longterm risk of hypoglycaemia to cognitive function, while the role and application of new technologies that may help to identify and avoid hypoglycaemia are reviewed and updated. We remain indebted to all of the authors for their continuing help, expertise and support to update the content of the book in what is a rapidly changing field. We hope that the clinical information and advice about managing and avoiding hypoglycaemia that is contained in the third edition of this book will assist clinicians with the overall Preface xi management of diabetes, and particularly increase their awareness, not only of the risks associated with hypoglycaemia, but also how this serious problem affects patients and their families.

Brian M. Frier Simon R. Heller Rory J. McCrimmon

1 Normal Glucose Metabolism and Responses to Hypoglycaemia Ian A. Macdonald1 and Paromita King2 1 University of Nottingham Medical School, Nottingham, UK 2 Derby Hospitals NHS Foundation Trust, Derby, UK

NORMAL GLUCOSE HOMEOSTASIS

Humans evolved as hunter-gatherers, and, unlike people today, did not consume regular meals. Mechanisms therefore evolved for the body to store food when it was in abundance, and to use these stores to provide an adequate supply of energy, in particular in the form of glucose when food was scarce. Cahill (1971) originally described the ‘rules of the meta- bolic game’ which man had to follow to ensure his survival. These were modified by Tat- tersall (personal communication) and are:

1. Maintain glucose within very narrow limits. 2. Maintain an emergency energy source (glycogen) that can be tapped quickly in time of need; ‘Fight or Flight’ response. 3. Waste not, want not, i.e. store (fat and protein) in times of plenty. 4. Use every trick in the book to maintain protein reserves.

Insulin and glucagon are the two key hormones controlling glucose homeostasis, and are therefore critical to the mechanisms enabling these ‘rules’ to be followed. The most important processes governed by these hormones are:

• Glycogen synthesis and breakdown (glycogenolysis): Glycogen, a carbohydrate, is the most readily accessible energy store and is mostly found in liver and skeletal muscle. Liver glycogen is broken down to provide glucose for all tissues, whereas the break- down of muscle glycogen results in lactate formation. • Gluconeogenesis: This is the production of glucose in the liver from precursors: glyc- erol, lactate and amino acids (in particular alanine). The process can also occur in the kidneys, but this site is not important under most physiological conditions. • Glucose uptake and metabolism (glycolysis) by skeletal muscle and adipose tissue.

The actions of insulin and glucagon are summarised in Boxes 1.1 and 1.2, respectively. Insulin is an anabolic hormone, reducing glucose output by the liver (hepatic glucose

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 2 Hypoglycaemia in Clinical Diabetes

Box 1.1 Actions of insulin

Liver ↑ glycogen synthesis (↑ glycogen synthetase activity) ↑ glycolysis ↑ lipid formation ↑ protein formation ↓ glycogenolysis (↓ phosphorylase activity) ↓ gluconeogenesis ↓ ketone formation Muscle ↑ uptake of glucose ↑ uptake of amino acids ↑ uptake of ketone ↑ uptake of potassium ↑ glycolysis ↑ synthesis of glycogen ↑ synthesis of protein ↓ protein catabolism ↓ release of amino acids Adipose tissue ↑ uptake of glucose ↑ uptake of potassium ↑ storage of triglyceride

Box 1.2 Actions of glucagon

Liver ↑ glycogenolysis ↑ gluconeogenesis ↑ extraction of alanine ↑ ketogenesis No significant peripheral action

output), increasing uptake of glucose by muscle and adipose tissue (increasing peripheral uptake) and increasing protein and fat formation. Glucagon opposes the actions of insulin in the liver. Thus insulin tends to reduce, and glucagon increase blood glucose concentrations. The metabolic effects of insulin and glucagon and their relationship to glucose home- ostasis are best considered in relationship to fasting and the postprandial state (Siegal and Kreisberg 1975). In both these situations it is the relative and not absolute concentrations of these hormones that are important.

Fasting (Figure 1.1a) During fasting, insulin concentrations are reduced and glucagon increased, which maintains blood glucose concentrations in accordance with rule 1 above. The net effect is to reduce peripheral glucose utilisation, increase hepatic glucose production and to provide non- Chapter 1: Normal Glucose Metabolism and Responses to Hypoglycaemia 3 glucose fuels for tissues not entirely dependent on glucose. After a short (e.g. overnight) fast, glucose production needs to be 5–6 g/h to maintain blood glucose concentrations, with the brain using 80% of this. Glycogenolysis provides 60–80% and gluconeogenesis 20–40% of the required glucose. In prolonged fasts, glycogen becomes depleted and glucose produc- tion is primarily from gluconeogenesis, with an increasing proportion from the kidney as opposed to the liver. In extreme situations renal gluconeogenesis can contribute as much as 45% of glucose production. Thus glycogen is the short-term or ‘emergency’ fuel source (rule 2), with gluconeogenesis predominating in more prolonged fasts. The following metabolic alterations enable this increase in glucose production to occur.

• Muscle: Glucose uptake and oxidative metabolism are reduced and fatty acid oxidation increased. Amino acids are released. • Adipose tissue: There are reductions in glucose uptake and triglyceride storage. The increase in the activity of the enzyme hormone-sensitive lipase, results in hydrolysis of triglyceride to glycerol (a gluconeogenic precursor) and fatty acids, which can be metabolised. • Liver: Increased cAMP concentrations result in increased glycogenolysis and gluco- neogenesis thus increasing hepatic glucose output. The uptake of gluconeogenic precur- sors (i.e. amino acids, glycerol, lactate and pyruvate) is also increased. Ketone bodies are produced in the liver from fatty acids. This process is normally inhibited by insulin and stimulated by glucagon, thus the hormonal changes during fasting lead to an increase in ketone production. Fatty acids are also a metabolic fuel used by the liver as a source of energy needed for the reactions involved in gluconeogenesis.

The reduced insulin:glucagon ratio favours a catabolic state, but the effect on fat metabo- lism is greater than protein, and thus muscle is relatively preserved (rule 4). These adapta- tions meant that not only did the hunter-gatherer have sufficient muscle power to pursue his next meal, but that brain function was optimally maintained to help him do this.

Fed state (Figure 1.1b) In the fed state, in accordance with the rules of the metabolic game, excess food is stored as glycogen, protein and fat (rule 3). Rising glucose concentrations after a meal result in an increase in insulin and reduction in glucagon secretion. This balance favours glucose utilisation, reduction of glucose production and increases glycogen, triglyceride and protein formation. The following changes enable these processes to occur:

• Muscle: Insulin increases glucose transport, oxidative metabolism and glycogen syn- thesis. Amino acid release is inhibited and protein synthesis increased. • Adipose tissue: In the fat cells, glucose transport is increased, while lipolysis is inhib- ited. At the same time the enzyme lipoprotein lipase, located in the capillaries, is activated and causes triglyceride to be broken down to fatty acids and glycerol. The fatty acids are taken up into the fat cells and re-esterified to triglyceride (using glycerol phosphate derived from glucose) before being stored. • Liver: Glucose uptake is increased in proportion to plasma glucose, a process that does not need insulin. However, insulin does decrease cAMP concentrations, which result in an increase in glycogen synthesis and the inhibition of glycogenolysis and gluconeo- genesis. These effects ‘retain’ excess glucose as glycogen in the liver. 4 Hypoglycaemia in Clinical Diabetes

Fasting Post-prandial glucagon, insulin Insulin, glucagon

Glucose Glycogen lactate Glycogen

Amino acids Protein Amino acids Protein

Muscle

TG Glucose

FFA Glycerolphosphate FFA Glycerolphosphate

Free FFA Free FFA Glycerol Triglyceride Triglyceride

Adipose tissue

Amino acids lactate

Ketones Ketones Glycogen Glycogen

Glucose FFA Glucose FFA Glycerol FFA, TG, Glycerol lipoprotein

Liver FFA, TG, Hepatic glucose Hepatic glucose output output (a) (b)

Figure 1.1 Metabolic pathways in glucose homeostasis in muscle, adipose tissue and liver. (a) Fasting; (b) post-prandial. FFA, free fatty acids; TG, triglyceride.

While insulin and glucagon are the key hormones involved in glucose homeostasis in the fed state, there are a number of other glucoregulatory hormones released in response to an oral glucose load including Gastric Inhibitory Peptide, Glucagon-like peptide 1 (GLP- 1), cholecystokinin Peptide Y and Ghrelin. GLP-1, for example is made in the L-cells of the distal gut as well as in the brain. Peripheral GLP-1 is produced in response to a glucose load. Through vagally-mediated central and peripheral mechanisms, GLP-1 augments glucose-stimulated insulin produc- tion, reduces glucagon secretion, slows gastric emptying and promotes satiety. A detailed discussion of the role of these peptides in glucose homeostasis is beyond the scope of this chapter, but has been reviewed by Heijboer et al. (2006), Drucker (2007) and Chapter 1: Normal Glucose Metabolism and Responses to Hypoglycaemia 5

Maggs et al. (2008). Some recently introduced antidiabetes drugs act through the GLP-1 receptor, either through a direct action (GLP-1 agonism) or through prevention of endog- enous GLP-1 break-down (dipeptidyl peptidase 4 inhibitors), and via this action stimulate insulin release. Concerns that this might lead to problems with glucoregulation during exercise do not seem to be well founded, as Khoo et al. (2010) showed no detrimental effects of the GLP-1 agonist, exenatide, on glycaemia during exercise. This complex interplay between insulin and glucagon as well as gut peptides maintains euglycaemia and enables the rules of the metabolic game to be followed, ensuring not only the survival of the hunter-gatherer, but also of modern humans.

EFFECTS OF GLUCOSE DEPRIVATION ON CENTRAL NERVOUS SYSTEM METABOLISM

The brain constitutes only 2% of body weight, but consumes 20% of the body’s oxygen and receives 15% of its cardiac output (Sokaloff 1989). It is almost totally dependent on carbohydrate as a fuel and since it cannot store or synthesise glucose, depends on a con- tinuous supply from the blood. The brain contains the enzymes needed to metabolise fuels other than glucose such as lactate, ketones and amino acids, but under physiological condi- tions their use is limited by insufficient quantities in the blood or slow rates of transport across the blood–brain barrier. When arterial blood glucose falls below 3 mmol/l, cerebral metabolism and function decline. Metabolism of glucose by the brain releases energy, and also generates neurotransmitters such as gamma amino butyric acid (GABA) and acetylcholine, together with phospholipids needed for cell membrane synthesis. When blood glucose concentration falls, changes in the synthesis of these products may occur within minutes because of reduced glucose metabolism, which can alter cerebral function. This is likely to be a factor in producing the subtle changes in cerebral function detectable at blood glucose concentrations of 3 mmol/l, a degree of hypoglycaemia that is actually not sufficiently low to cause a major depletion in ATP or creatine phosphate, the brain’s two main sources of energy (McCall 1993). Isotope techniques and Positron Emission Tomography (PET) allow the study of metab- olism in different parts of the brain and show regional variations in metabolism during hypoglycaemia. The neocortex, hippocampus, hypothalamus and cerebellum are most sensitive to hypoglycaemia, whereas metabolism is relatively preserved in the thalamus and brainstem. Changes in cerebral function are initially reversible, but during prolonged severe hypoglycaemia, general energy failure (due to the depletion of ATP and creatine phosphate) can cause permanent cerebral damage. Pathologically this is caused by selective neuronal necrosis most likely due to ‘excitotoxin’ damage. Local energy failure induces the intrasynaptic release of glutamate or aspartate, and failure of reuptake of the neuro- transmitters increases their concentrations. This leads to the activation of N-methyl- D-aspartate (NMDA) receptors causing cerebral damage. One study in rats has shown that an experimental compound called AP7, which blocks the NMDA receptor, can prevent 90% of the cerebral damage associated with severe hypoglycaemia (Wieloch 1985). In humans with fatal hypoglycaemia, protracted neuroglycopenia causes laminar necrosis in the cer- ebral cortex and diffuse demyelination. Regional differences in neuronal necrosis are seen, with the basal ganglia and hippocampus being affected, but the hypothalamus and cerebel- lum relatively spared (Auer and Siesjö 1988; Sieber and Traysman 1992). 6 Hypoglycaemia in Clinical Diabetes

Transport of glucose across blood-brain barrier

Capillary

Glucose

GLUT–1 GLUT–3

Neurone

Figure 1.2 Transport of glucose into the brain across the blood–brain barrier.

The brain is very sensitive to acute hypoglycaemia, but can adapt to chronic fuel dep- rivation. For example, during starvation, it can metabolise ketones for up to 60% of its energy requirements (Owen et al. 1967). Glucose transport can also be increased in the face of hypoglycaemia. Normally, glucose is transported into tissues using proteins called glucose transporters (GLUT) (Bell et al. 1990). This transport occurs down a concentration gradient faster than it would by simple diffusion and does not require energy (facilitated diffusion). There are several of these transporters, with GLUT 1 being responsible for transporting glucose across the blood–brain barrier and GLUT 3 into neurones (Figure 1.2). Chronic hypoglycaemia in animals (McCall et al. 1986) and in humans (Boyle et al. 1995) increases global cerebral glucose uptake, which is thought to be due to an increase in the production and action of GLUT 1 protein. It has not been established whether this adapta- tion is of major benefit in protecting brain function during hypoglycaemia or whether this adaptation also occurs in response to repeated, rather than just chronic (up to 3 days), hypoglycaemia.

COUNTERREGULATION DURING HYPOGLYCAEMIA

The potentially serious effects of hypoglycaemia on cerebral function mean that not only are stable blood glucose concentrations maintained under physiological conditions, but if hypoglycaemia occurs, mechanisms have developed to combat it. In clinical practice, the principal causes of hypoglycaemia are iatrogenic (as side-effects of insulin and sulfonylu- reas used to treat diabetes) and excessive alcohol consumption. Insulin-secreting tumours (such as insulinoma) are rare. The mechanisms that correct hypoglycaemia are called counterregulation, because the hormones involved oppose the action of insulin and are thus the counterregulatory hormones. The processes of counterregulation were identified in the mid-1970s and early 1980s, using either a bolus injection or continuous infusion of insulin to induce hypoglycaemia (Cryer 1981; Gerich 1988). The response to the bolus injection of 0.1 U/kg insulin in a normal subject is shown in Figure 1.3. Blood glucose concentrations Chapter 1: Normal Glucose Metabolism and Responses to Hypoglycaemia 7

Glucose 6 100

4 75 50 mmol/l 2 mg/dl 25

0 0 –240 –60 0 30 60 90 120 (a) Time (min)

Insulin 7200 1000

720 100 μU/ml pmol/l

72 10

0 0 –60 0 30 60 90 120 (b) Time (min)

Figure 1.3 Glucose (a) and insulin (b) concentrations after intravenous injection of insulin 0.1 U/kg at time 0. (Source: Garber AJ 1976. Reproduced with permission of the American Society for Clinical Investigation). decline within minutes of the administration of insulin and reach a nadir after 20–30 minutes, then gradually rise to near normal by 2 hours after the insulin was administered. The fact that blood glucose starts to rise when plasma insulin concentrations are still 10 times the baseline values means that it is not simply the reduction in insulin that reverses hypoglycaemia, but active counterregulation must also occur. Many hormones are released when blood glucose is lowered (see below), but glucagon, the catecholamines, growth hormone and cortisol are regarded as being the most important. Several studies have determined the relative importance of these hormones by producing isolated deficiencies of each hormone (by blocking its release or action) and assessing the subsequent response to administration of insulin. These studies assessed the relative impor- tance of glucagon, adrenaline (epinephrine) and growth hormone in the counterregulation of short-term hypoglycaemia. Somatostatin infusion was used to block glucagon and growth hormone secretion and significantly impaired glucose recovery. When growth hormone was replaced in the same model to produce isolated glucagon deficiency or glu- cagon replaced to produce isolated growth hormone deficiency, it is clear that it was glucagon and not growth hormone that was responsible for acute counterregulation. Com- bined alpha- and beta-adrenoceptor blockade using phentolamine and propranolol infusions or studies of patients who had undergone an adrenalectomy were used to evaluate the role of the catecholamines. Overall these studies demonstrate that glucagon is the most important counterregulatory hormone while catecholamines provide a backup if glucagon 8 Hypoglycaemia in Clinical Diabetes is deficient (e.g. in type 1 diabetes, Chapters 3 and 6). Cortisol and growth hormone are probably more important to prolonged hypoglycaemia. Therefore, if glucagon and catecholamines are both deficient, as in longstanding type 1 diabetes, counterregulation is seriously compromised, and the individual is nearly defenceless against acute hypoglycae- mia (Cryer 1981). Glucagon and catecholamines increase glycogenolysis and stimulate gluconeogenesis. Catecholamines also reduce glucose utilisation peripherally and inhibit insulin secretion. Cortisol and growth hormone increase gluconeogenesis and reduce glucose utilisation. The role of the other hormones (see below) in counterregulation is unclear, but they are unlikely to make a significant contribution. Finally, there is evidence that in profound hypoglycae- mia (blood glucose below 1.7 mmol/l), hepatic glucose output is stimulated directly, although the mechanism is unknown. This is termed hepatic autoregulation. The depth, as well as the duration of hypoglycaemia is important in determining the magnitude of the counterregulatory hormone response. Studies using ‘hyperinsulinaemic clamps’ show a hierarchical response of hormone production. In this technique, insulin is infused at a constant rate and a glucose infusion rate varied to maintain blood glucose concentrations within ±0.2 mmol/l of target concentrations. This permits the controlled evaluation of the counterregulatory hormone response at varying degrees of hypoglycae- mia. It also demonstrates that in human subjects without diabetes, glucagon, catecho- lamines and growth hormone start to be produced at a blood glucose concentration of 3.5–3.7 mmol/l, with cortisol produced at a lower glucose of 3.0 mmol/l (Mitrakou et al. 1991). The counterregulatory response is initiated before impairment in cerebral function is first evident, usually at a glucose concentration of approximately 3.0 mmol/l (Heller and Macdonald 1996). The magnitude of the hormonal response also depends on the length of the hypoglycaem­ ­ic episode. The counterregulatory hormonal response commences up to 20 minutes after hypoglycaemia is achieved and continues to rise for 60 minutes (Kerr et al. 1989). In contrast, this response is attenuated if it occurs within 2–3 days of a previous episode of hypoglycaemia (Heller and Macdonald 1996) or if it follows prolonged exercise the day before hypoglycaemia is induced. Galassetti et al. (2001) showed in non-diabetic subjects that 3 hours of moderate intensity exercise the previous day markedly decreased the coun- terregulatory response to hypoglycaemia induced by the infusion of insulin, and that the reduced counterregulatory response was more marked in men than in women. It is also now apparent that antecedent bouts of exercise can affect the counterregulatory response to subsequent hypoglycaemia (Sandoval et al. 2006). While the primary role of the counterregulatory hormones is on glucose metabolism, any effects on fatty acid utilisation can have an indirect effect on blood glucose. Thus, the increase in plasma adrenaline (epinephrine) (and activation of the sympathetic nervous system) that occurs in hypoglycaemia can stimulate lipolysis of triglyceride in adipose tissue and muscle and release fatty acids which can be used as an alternative fuel to glucose, making more glucose available for the central nervous system (CNS). Enoksson et al. (2003) demonstrated that adults with type 1 diabetes who had lower plasma adrenaline responses to hypoglycaemia than non-diabetic controls also had reduced rates of lipolysis in adipose tissue and skeletal muscle, making them more dependent on glucose as a fuel and therefore at risk of developing more severe hypoglycaemia. The complex counterregulatory and homeostatic mechanisms described above are thought to be mostly under CNS control. Evidence for this comes from studies in dogs, where glucose was infused into the carotid and vertebral arteries to maintain euglycaemia Chapter 1: Normal Glucose Metabolism and Responses to Hypoglycaemia 9 in the brain. Despite peripheral hypoglycaemia, glucagon did not increase, and responses of the other counterregulatory hormones were blunted. This, and other studies in rats, led to the hypothesis that the ventromedial nucleus of the hypothalamus (VMH) acts as the primary glucose sensor and coordinates counterregulation during hypoglycaemia (Borg et al. 1997). However, other parts of the brain are also involved in mediating counterregula- tion, and it is likely that the VMH is part of an integrated network of specialised glucose- sensing brain regions. It is now clear that glucose-sensing neurones use glucokinase and ATP-sensitive K+ channels to transduce the glucose signal into changes in neuronal activity (Levin et al. 2004). In rats, the VMH has ATP-sensitive K+ channels that seem to be involved in the counterregulatory responses to hypoglycaemia, as injection of the sulfonylurea, glib- enclamide, directly into the VMH, suppressed hormonal responses to systemic hypogly- caemia (Evans et al. 2004), whereas diazoxide amplified hormonal responses (McCrimmon et al. 2005). The existence of hepatic autoregulation suggests that some peripheral control also exists. Studies producing central euglycaemia and hepatic-portalvenous hypoglycaemia in dogs have provided evidence for hepatic glucose sensors and suggest that these sensors, as well as those in the brain, are important in the regulation of glucose (Hamilton-Wessler et al. 1994). However, this topic is somewhat controversial and more recent studies on dogs have failed to demonstrate an effect of hepatic sensory nerves on the responses to hypoglycaemia (Jackson et al. 2000). Moreover, studies in human subjects by Heptulla et al. (2001) showed that providing glucose orally rather than intravenously during a hypoglycaemic hyperin- sulinaemic clamp actually enhanced the counterregulatory hormone responses rather than reduced them.

HORMONAL CHANGES DURING HYPOGLYCAEMIA

Hypoglycaemia induces a change in various hormones some of which also contribute to symptom generation (Chapter 2), counterregulation, and many of the physiological changes that occur as a consequence of lowering blood glucose. The stimulation of the autonomic nervous system is central to many of these changes.

Activation of the autonomic nervous system The autonomic nervous system comprises sympathetic and parasympathetic components (Figure 1.4). Fibres from the sympathetic division leave the spinal cord with the ventral roots from the first thoracic to the third or fourth lumbar nerves to synapse in the sympa- thetic chain or visceral ganglia, and the long postganglionic fibres are incorporated in somatic nerves. The parasympathetic pathways originate in the nuclei of cranial nerves III, VII, IX and X and travel with the vagus nerve. A second component, the sacral outflow, supplies the pelvic viscera via the pelvic branches of the second to fourth spinal nerves. The ganglia in both cases are located near the organs supplied, and the postganglionic neurones are therefore short. Activation of both components of the autonomic system occurs during hypoglycaemia. The sympathetic nervous system in particular is responsible for many of the physiological changes during hypoglycaemia, and the evidence for its activation can be obtained indi- rectly by observing functional changes such as cardiovascular responses (considered 10 Hypoglycaemia in Clinical Diabetes

Cranial nerves – From spinal cord, III, VII, IX, X medulla, hypothalamus Brain Post stem Pre Spinal nerves Pre T1 – L3 Post Cranial outflow Viscus To blood vessels, sweat glands Spinal nerves Sympathetic ganglion S2 – S4

Gray ramus Pre White communicans RC Sacral outflow Post Viscus Collateral ganglion

Sympathetic division Parasympathetic division

Parasympathetic Sympathetic

III

VII IX X

T 1 2 3 Arm 4 5 Heart 6 7 Viscera Heart 8 9 10 11 12 1L Adrenal 2 medulla Bowel (preganlionic supply) 2S 3

Leg Sympathetic Terminal chain ganglion (coccygeal)

Figure 1.4 Anatomy of the autonomic nervous system. Pre, preganglionic neurones; post, postgan- glionic neurones; RC, ramus communicans. below), measuring plasma catecholamines, which gives a general index of sympathetic activation, or by directly recording sympathetic activity. Direct recordings are possible from sympathetic nerves supplying skeletal muscle and skin. Sympathetic neural activity in skeletal muscle involves vasoconstrictor fibres that innervate blood vessels and are involved in controlling blood pressure. During hypogly- caemia (induced by insulin), the frequency and amplitude of muscle sympathetic activity are increased as blood glucose falls, with an increase in activity 8 minutes after insulin is Chapter 1: Normal Glucose Metabolism and Responses to Hypoglycaemia 11 injected intravenously, peaking at 25–30 minutes coincident with the glucose nadir, and persisting for 90 minutes after euglycaemia is restored (Fagius et al. 1986). During hypogly- caemia, a sudden increase in skin sympathetic activity is seen, which coincides with the onset of sweating. This sweating leads to vasodilatation of skin blood vessels, which is also contributed to by a reduction in sympathetic stimulation of the vasoconstrictor com- ponents of skin arterio-venous anastomoses (Berne and Fagius 1986). These effects (at least initially) increase total skin blood flow and promote heat loss from the body. Activation of both muscle and skin sympathetic nerve activity is thought to be centrally mediated. Tissue neuroglycopenia can be produced by 2-deoxy-D-glucose, a glucose ana- logue, without increasing insulin. Infusion of this analogue causes stimulation of muscle and skin sympathetic activity demonstrating that it is hypoglycaemia and not the insulin used to induce it that is responsible for the sympathetic activation (Fagius and Berne 1989). The activation of the parasympathetic nervous system (vagus nerve) during hypogly- caemia cannot be measured directly. The most useful index of parasympathetic function is the measurement of plasma pancreatic polypeptide (PP), the peptide hormone secreted by the PP cells of the pancreas, which is released in response to vagal stimulation.

Neuroendocrine activation (Box 1.3) Insulin-induced hypoglycaemia was used to study pituitary function as early as the 1940s. The glucose response to insulin was used to measure insulin resistance, and thereby evalu- ate conditions such as Cushing’s syndrome and hypopituitarism. The development of assays for adrenocorticotrophic hormone (ACTH) and growth hormone (GH) allowed the direct measurement of pituitary function during hypoglycaemia in the 1960s, and many of the processes governing these changes were unravelled before elucidation of the counterregulat­ ory system. The studies are comparable to those evaluating counterregulation, in that potential regulatory factors are blocked to measure the hormonal response to hypoglycae- mia with and without the regulating factor.

Hypothalamus and anterior pituitary gland ACTH, GH and prolactin concentrations increase during hypoglycaemia, but there is no change in thyrotrophin or gonadotrophin secretion. The secretion of these pituitary hor- mones is controlled by releasing factors that are produced in the median eminence of the hypothalamus and secreted into the hypophyseal portal vessels and then pass to the pituitary gland (Figure 1.5). The mechanisms regulating the releasing factors are incompletely understood but may involve the VMH and other hypothalamic regions where brain glucose sensors are situated (Fish et al. 1986).

• ACTH: Secretion is governed by release of corticotrophin releasing hormone (CRH) from the hypothalamus; alpha adrenoceptors stimulate CRH release, and beta adreno- ceptors have an inhibitory action. A variety of neurotransmitters control the release of CRH into the portal vessels, including serotonin and acetylcholine, which are stimulatory, and GABA, which is inhibitory. The increase in ACTH causes cortisol to be secreted from the cortices of the adrenal glands. • Beta endorphins are derived from the same precursors as ACTH and are co-secreted with it. The role of endorphins in counterregulation is uncertain, but they may influence the secretion of the other pituitary hormones during hypoglycaemia. 12 Hypoglycaemia in Clinical Diabetes

Box 1.3 Neuroendocrine activation

Hypothalamus ↑ CRH ↑ GHRH Ant Pituitary ↑ ACTH ↑ β endorphin ↑ GH ↑ Prolactin ↔ TSH ↔ Gonadotrophins Post Pituitary ↑ Vasopressin ↑ Oxytocin Pancreas ↑ Glucagon ↑ Somatostatin-28 ↑ Pancreatic polypeptide ↓ Insulin Adrenal ↑ Cortisol ↑ Adrenaline (epinephrine) ↑ Aldosterone Others ↑ PTH ↑ Gastrin

Hypothalamus

Optic chiasm Mammillary body Artery

Median eminence Primary capillary plexus

Hypothalamic-hypophysial portal vessels

Sinuses Posterior pituitary gland Anterior pituitary gland Vein

Figure 1.5 Anatomy of the hypothalamus and pituitary. Chapter 1: Normal Glucose Metabolism and Responses to Hypoglycaemia 13 • GH: Growth hormone secretion is governed by two hypothalamic hormones: growth hormone releasing hormone (GHRH), which stimulates GH secretion, and somatosta- tin, which is inhibitory. GHRH secretion is stimulated by dopamine, GABA, opiates and through alpha adrenoceptors, while it is inhibited by serotonin and beta adrenocep- tors. A study in rats showed that bio-assayable GH and GHRH are depleted in the pituitary and hypothalamus respectively after insulin-induced hypoglycaemia (Katz et al. 1967). • Prolactin: The mechanisms are not established. Prolactin secretion is normally under inhibitory control of dopamine, but there is also evidence for releasing factors during hypoglycaemia. Prolactin does not contribute to counterregulation.

Posterior pituitary gland Vasopressin and oxytocin both increase during hypoglycaemia (Fisher et al. 1987). Their secretion is under hormonal and neurotransmitter control in a similar way to the hypotha- lamic hormones. Vasopressin has glycolytic actions and oxytocin increases hepatic glucose output in dogs, but their contribution to glucose counterregulation is uncertain.

Pancreas • Glucagon: The mechanisms of glucagon secretion during hypoglycaemia are still not fully understood. Although activation of the autonomic nervous system stimulates its release, this pathway has been shown to be less important in humans. A reduction in glucose concentrations may have a direct effect on the glucagon-secreting pancreatic alpha cells, or the reduced beta cell activity (reduced insulin and/or cofactors such as zinc or GABA secretion), which also occurs with low blood glucose, may release the tonic inhibition of glucagon secretion. In type 1 diabetes, where hypoglycaemia is normally associated with high plasma insulin levels, this mechanism is disturbed, and the failure of insulin to suppress (‘switch-off’) may contribute to the selective defect in glucagon secretion that is characteristic of C-peptide-negative type 1 diabetes. However, this remains to be proven (Cryer 2012). • Somatostatin: This is thought of as a pancreatic hormone produced from D cells of the islets of Langerhans but it is also secreted in other parts of the gastrointestinal tract. There are a number of structurally different polypeptides derived from prosomatostatin: the somatostatin-14 peptide is secreted from D cells, and somatostatin-28 from the gastrointestinal tract. The plasma concentration of somatostatin-28 increases during hypoglycaemia (Francis and Ensinck 1987). The normal action of somatostatin is to inhibit the secretion both of insulin and glucagon, but somatostatin-28 inhibits insulin 10 times more effectively than glucagon, and thus may have a role in counterregulation by suppressing insulin release. • Pancreatic polypeptide: This peptide has no role in counterregulation, but its release during hypoglycaemia is stimulated by cholinergic fibres through muscarinic receptors and is a useful marker of parasympathetic activity.

Adrenal gland and renin – angiotensin system The processes governing the increase in cortisol during hypoglycaemia are discussed above. A rise in catecholamines, in particular adrenaline from the adrenal medulla that 14 Hypoglycaemia in Clinical Diabetes occurs when blood glucose is lowered, is controlled by sympathetic fibres in the splanchnic nerve. The increase in renin, angiotensin and aldosterone during hypoglycaemia, results primarily from the intra-renal actions of catecholamines, mediated through beta adrenocep- tors, although the hypoglycaemia-induced increase in ACTH and hypokalaemia may con- tribute (Trovati et al. 1988; Jungman et al. 1989). These changes do not have a significant role in counterregulation, although angiotensin II has glycolytic actions in vitro.

PHYSIOLOGICAL RESPONSES

Haemodynamic changes (Box 1.4) The haemodynamic changes during hypoglycaemia (Hilsted 1993) are mostly caused by the activation of the sympathetic nervous system and an increase in circulating adrenaline. An increase in heart rate (tachycardia) and cardiac output occurs, which is mediated through beta1 adrenoceptors, but increasing vagal tone counteracts this so the increase is transient. Peripheral resistance, estimated from mean arterial pressure divided by cardiac output, is reduced. A combination of increased cardiac output and reduced peripheral resistance results in an increase in systolic and a small decrease in diastolic pressure, in other words, widening of pulse pressure. However, central blood pressure falls during hypoglycaemia accompanied by an increase in arterial elasticity, as demonstrated by changes in the Aug- mentation Index (Sommerfield et al. 2007).

(Changes in regional blood flow (Box 1.5 and Figure 1.6 • Cerebral blood flow: Early work produced conflicting results, but these studies were in subjects receiving insulin shock therapy, and the varying effects of convulsions and altered level of consciousness may have influenced the outcome. Subsequent studies have consistently shown an increase in cerebral blood flow during hypoglycaemia

Box 1.4 Haemodynamic changes

↑ Heart rate ↑ Blood pressure (peripheral systolic); ↓ central blood pressure ↑ Cardiac output ↓ Peripheral resistance

Box 1.5 Changes in regional blood flow

↑ cerebral flow ↑ total splanchnic flow ↑ muscle flow skin flow variable ↓ renal flow ↓ splenic flow Chapter 1: Normal Glucose Metabolism and Responses to Hypoglycaemia 15

Brain Blood flow reduced

Blood flow increased Eye

Heart Skin

Stomach

Emptying Liver

Spleen Motility Gut Skeletal muscle

Kidney

Figure 1.6 Changes in regional blood flow during hypoglycaemia.

despite the use of different methods of measurement (isotopic, single photon emission computed tomography (SPECT), and Doppler ultrasound). In most of the studies blood glucose concentration was less than 2 mmol/l before a change was observed. In the last decade the use of PET and Magnetic Resonance Imaging (MRI) has allowed a more detailed study of cerebral blood flow, showing regional changes even with minimal changes in glucose concentration. Using PET, Teves et al. (2004) demonstrated a 6.8% reduction in cerebral blood flow in the brainstem, cerebellum and hemispheres, and an increase in the thalamus, globus pallidus and medial frontal cortex at glucose concen- trations of 3 mmol/l. The subsequent differences in substrate delivery may explain why certain areas of the brain such as the cerebellum are more vulnerable to hypoglycaemia- induced cerebral damage. More recently, Page et al. (2009) used MRI to demonstrate an increase in hypothalamic blood flow when glucose was reduced from 5.3 to 4.3 mmol/l. In animals, hypoglycaemia is associated with loss of cerebral autoregula- tion (the ability of the brain to maintain cerebral blood flow despite variability in cardiac output) through beta adrenoceptor stimulation, but the exact mechanisms are unknown (Bryan 1990; Sieber and Traysman 1992). • Gastrointestinal system: Total splanchnic blood flow (that supplying the intestines, liver, spleen and stomach) is increased and splanchnic vascular resistance reduced as assessed by the bromosulphthalein extraction technique (Bearn et al. 1952). Superior 16 Hypoglycaemia in Clinical Diabetes

mesenteric artery blood flow measured using Doppler ultrasound increases during hypoglycaemia due to beta adrenoceptor stimulation (Braatvedt et al. 1993). Radioiso- tope scanning has demonstrated a reduction in splenic activity during hypoglycaemia (Fisher et al. 1990), which is thought to be due to alpha adrenoceptor-mediated reduc- tion in blood flow. All these changes would all be expected to increase hepatic blood flow, although this has not been confirmed experimentally. • Skin: The control of blood flow to the skin is complex and different mechanisms pre- dominate in different areas. Studies of the effect of hypoglycaemia on skin blood flow are inconsistent partly because different methods have been used for blood flow meas- urement and induction of hypoglycaemia, as well as differences in the part of the body studied. Definitive conclusions are therefore not possible. Studies using the dorsum of the foot, cheek and forehead have consistently shown an initial vasodilatation and increase in blood flow followed by later vasoconstriction at a blood glucose of 2.5 mmol/l (Maggs et al. 1994a). These findings are consistent with the clinical picture of initial flushing and later pallor, with an early rise in skin blood flow and a later fall. • Muscle blood flow: A variety of techniques have been used to study muscle blood flow (including venous occlusion plethysmography, isotopic clearance techniques and the use of thermal conductivity meters). All studies have consistently shown an increase in muscle blood flow during hypoglycaemia irrespective of skin blood flow. This change

is mediated by beta2 adrenoceptors (Allwood et al. 1959; Abramson et al. 1966). • Kidney: Inulin and sodium hippurate clearance can be used to estimate glomerular filtration rate and renal blood flow, respectively. Both decrease during hypoglycaemia (Patrick et al. 1989) and catecholamines and renin are implicated in initiating the changes.

The changes in blood flow in various organs, like the haemodynamic changes, are mostly mediated by the activation of the sympathetic nervous system or circulating adrenaline. The majority either protect against hypoglycaemia or increase substrate delivery to vital organs. The increase in cerebral blood flow increases substrate delivery to the brain. Increasing muscle flow enhances the release and washout of gluconeogenic precursors. The increase in splanchnic blood flow and reduction in splenic blood flow serve to increase hepatic blood flow to maximise hepatic glucose production. Meanwhile blood is diverted away from organs such as the kidney that are not required in the acute response to the metabolic stress.

Functional changes (Box 1.6) • Sweating: Sweating is mediated by sympathetic cholinergic nerves, although other neurotransmitters such as vasoactive intestinal peptide and bradykinin may also be involved. The activation of the sympathetic innervation of the skin as described above results in the sudden onset of sweating. Sweating is one of the first physiological responses to occur during hypoglycaemia and can be demonstrated within 10 minutes of achieving a blood glucose of 2.5 mmol/l (Maggs et al. 1994b). It coincides with the onset of other measures of autonomic activation, such as an increase in heart rate and tremor (Figure 1.7). • Tremor: Trembling and shaking are characteristic features of hypoglycaemia and result from an increase in physiological tremor. The rise in cardiac output and vasodilatation occurring during hypoglycaemia increase the level of physiological tremor and this is Chapter 1: Normal Glucose Metabolism and Responses to Hypoglycaemia 17

Box 1.6 Functional changes

↑ sweating (sudden onset) ↑ tremor ↓ core temperature ↓ intraocular pressure ↑ jejunal activity ↑ gastric emptying

80

70

60 Heartrate 50 R (beats per minute) per (beats

40

4000

3000 O) 2

2000 (PPmV H

Sweat production Sweat 1000

0 R

80

60 ) –2 40 Tremor (cm s 20

0 R

B 0 + 20 R R + 30 R + 60 Time (min)

Figure 1.7 Sudden onset of sweating, tremor and increase in heart rate during the induction of hypoglycaemia. 18 Hypoglycaemia in Clinical Diabetes

exacerbated by beta adrenoceptor stimulation associated with increased adrenaline concentrations (Kerr et al. 1990). Since adrenalectomy does not entirely abolish tremor, other components such as the activation of muscle sympathetic activity must be involved. • Temperature: Despite a beta adrenoceptor-mediated increase in metabolic rate, core temperature falls during hypoglycaemia. The mechanisms by which this occurs depend on whether the environment is warm or cold. In a warm environment, heat is lost because of sweating and increased heat conduction from vasodilatation. Hypoglycae- mia reduces core temperature by 0.3°C and skin temperature by up to 2°C (depending on the part of the body measured) after 60 minutes (Maggs et al. 1994b). In the cold, shivering is reduced, and this together with vasodilatation and sweating causes a sub- stantial reduction in core temperature (Gale et al. 1983). In rats, mortality was increased in animals whose core temperature was prevented from falling during hypoglycaemia (Buchanan et al. 1991). In humans there is anecdotal evidence from subjects undergo- ing insulin shock therapy that those who had a rise in body temperature showed delayed neurological recovery (Ramos et al. 1968). These findings support the hypothesis that the fall in core temperature reduces metabolic rate, allowing hypoglycaemia to be better tolerated, and thus the changes in body temperature are of survival value. The beneficial effects are likely to be limited, particularly in the cold, where the impairment of cerebral function means subjects may not realise they are cold, causing them to be at risk of severe hypothermia. • Other functional changes include a reduction in intra-ocular pressure, greater jejunal but not gastric motility and inconsistent abnormalities of liver function tests. An increase in gastric emptying occurs during hypoglycaemia (Schvarcz et al. 1995), which may be protective in that carbohydrate delivery to the intestine is increased, enabling faster glucose absorption and reversal of hypoglycaemia.

CONCLUSIONS • Homeostatic mechanisms exist to maintain glucose concentration within narrow limits despite a wide variety of circumstances. • The dependence of the central nervous system on glucose has led to a complex series of biochemical, functional and haemodynamic changes aimed at restoring glucose concentrations, generating symptoms and protecting the body in general and central nervous system in particular against the effects of a low blood glucose (Figure 1.8). • Many symptoms of hypoglycaemia result from the activation of the autonomic nervous system and help to warn the individual that their blood glucose is low. This encourages the ingestion of carbohydrate, so helping to restore glucose concentrations in addition to counterregulation. • Faster gastric emptying and the changes in regional blood flow, which also occur as a result of the activation of the autonomic nervous system, increase substrate delivery. • The greater cerebral blood flow increases glucose delivery to the brain (although loss of autoregulation is undesirable), and the increased splanchnic flow results in a greater delivery of gluconeogenic precursors to the liver. • Activation of the autonomic nervous system also increases sweating, and this, together with the inhibition of shivering predisposes to hypothermia, which may be neuroprotective. Chapter 1: Normal Glucose Metabolism and Responses to Hypoglycaemia 19

Normal glucose homeostasis

Insulin : glucagon ratio

Excess insulin Hypoglycaemia

Detection by brain glucose sensors

Activation of

Pancreas Pituitary Adrenal autonomic nervous system

Glucagon (GH and cortisol) Catecholamines

Gastric emptying Counterregulation and Sweating

changes in blood flow

(Increased substrate Symptoms Cerebral Splanchnic production) (autonomic) blood flow blood flow

Hypothermia (Eating)

Restore blood Increase substrate Protection of

glucose delivery vital organs

Figure 1.8 Glucose homeostasis and the correction of hypoglycaemia. 20 Hypoglycaemia in Clinical Diabetes

ACKNOWLEDGEMENT

We would like to thank Professor Robert Tattersall for his helpful suggestions.

REFERENCES

Abramson, E.A., Arky, R.A. & Waeber, K.A. (1966) Effects of propranolol on the hormonal and metabolic responses to insulin induced hypoglycaemia. Lancet, ii, 1386–1389. Allwood, M.J., Hensel, H. & Papenberg, J. (1959) Muscle and skin blood flow in the human forearm during insulin hypoglycaemia. The Journal of Physiology, 147, 269–273. Auer, R.N. & Siesjö, B.K. (1988) Biological differences between ischaemia, hypoglycaemia and epilepsy. Annals of Neurology, 24, 699–707. Bearn, A.G., Bulling, B.H. & Sherlock, S. (1952) The response of the liver to insulin in normal subjects and in diabetes mellitus: hepatic vein catheterisation studies. Clinical Science, 11, 151–164. Bell, G.I., Karyano, T., Buse, J.B. et al. (1990) Molecular biology of mammalian glucose transporters. Diabetes Care, 13, 198–208. Berne, C. & Fagius, J. (1986) Skin sympathetic activity during insulin induced hypoglycemia. Diabetolo- gia, 29, 855–860. Borg, M.A., Sherwin, R.S., Borg, W.P., Tamborlane, W.V. & Shulman, G.I. (1997) Local ventromedial hypothalamus glucose perfusion blocks counterregulation during systemic hypoglycemia in awake rats. The Journal of Clinical Investigation, 99, 361–365. Boyle, P.J., Kempers, S.E., O’Conner, A.M. & Nagy, R.J. (1995) Brain glucose uptake and unawareness of hypoglycemia in patients with insulin – dependent diabetes mellitus. New England Journal of Medi- cine, 333, 1726–1731. Braatvedt, G.D., Flynn, M.D., Stanners, A., Halliwell, M. & Corrall, R.J.M. (1993) Splanchnic blood flow in man: evidence for mediation via a β-adrenergic mechanism. Clinical Science, 84, 201–207. Bryan, R.M. (1990) Cerebral blood flow and energy metabolism during stress. The American Journal of Physiology, 259, H269–H280. Buchanan, T.A., Cane, P., Eng, C.C., Sipos, G.F. & Lee, C. (1991) Hypothermia is critical for survival during prolonged insulin induced hypoglycemia in rats. Metabolism: Clinical and Experimental, 40, 330–334. Cahill, G.F. (1971) Physiology of insulin in man. Diabetes, 20, 785–799. Cryer, P.E. (1981) Glucose counterregulation in man. Diabetes, 30, 261–264. Cryer, P.E. (2012) Glucagon in the pathogenesis of hypoglycemia and hyperglycemia in diabetes. Endo- crinology, 153, 1039–1048. Drucker, D.J. (2007) The role of gut hormones in glucose homeostasis. The Journal of Clinical Investiga- tion, 117, 24–32. Enoksson, S., Caprio, S.K., Rife, F., Shulman, G.I., Tamborlane, W.V. & Sherwin, R.S. (2003) Defective activation of skeletal muscle and adipose tissue lipolysis in type 1 diabetes mellitus during hypoglycemia. Journal of Clinical Endocrinology and Metabolism, 88, 1503–1511. Evans, M.L., McCrimmon, R.J., Flanagan, D.E. et al. (2004) Hypothalamic ATP-sensitive K+ channels play a key role in sensing hypoglycemia and triggering counterregulatory epinephrine and glucagon responses. Diabetes, 53, 2542–2551. Fagius, J. & Berne, C. (1989) Changes in sympathetic nerve activity induced by 2-deoxy-D-glucose infu- sion in humans. The American Journal of Physiology, 256, E714–E721. Fagius, J., Niklasson, F. & Berne, C. (1986) Sympathetic outflow in human muscle nerves increases during hypoglycemia. Diabetes, 35, 1124–1129. Fish, H.R., Chernow, B. & O’Brian, J.T. (1986) Endocrine and neurophysiologic responses of the pituitary of insulin-induced hypoglyaemia: a review. Metabolism: Clinical and Experimental, 35, 763–780. Fisher, B.M., Baylis, P.H. & Frier, B.M. (1987) Plasma oxytocin, arginine vasopressin and atrial natriuretic peptide response to insulin-induced hypoglycaemia in man. Clinical Endocrinology, 26, 179–185. Fisher, B.M., Gillen, G., Hepburn, D.A., Dargie, H.J., Barnett, E. & Frier, B.M. (1990) Splenic responses to acute insulin-induced hypoglycaemia in humans. Clinical Science, 78, 469–474. Chapter 1: Normal Glucose Metabolism and Responses to Hypoglycaemia 21

Francis, B.H. & Ensinck, J.W. (1987) Differential alterations of the circulating prosomatostatin-derived peptides during insulin induced hypoglycemia in man. Journal of Clinical Endocrinology and Metabo- lism, 65, 880–884. Galassetti, P., Neill, A.R., Tate, D., Ertl, A.C., Wasserman, D.H. & Davis, S.N. (2001) Sexual dimorphism in counterregulatory responses to hypoglycemia after antecedent exercise. Journal of Clinical Endo- crinology and Metabolism, 86, 3516–3524. Gale, E.A.M., Bennet, J., Macdonald, I.A., Holst, J.J. & Mathews, J.A. (1983) The physiological effects of insulin-induced hypoglycaemia in man: responses at differing levels of blood glucose. Clinical Science, 65, 262–271. Garber, A.J., Cryer, P.E., Santiago, J.V., Hammond, M.W., Pagliara, A.S. & Kipnis, D.M. (1976) The role of adrenergic mechanisms in the substrate and hormonal response to insulin-induced hypoglycemia in man. The Journal of Clinical Investigation, 58, 7–15. Gerich, J.E. (1988) Glucose counterregulation and its impact in diabetes mellitus. Diabetes, 37, 1608–1617. Hamilton-Wessler, M., Bergman, R.N., Halter, J.B., Watanabe, R.M. & Donovan, C.M. (1994) The role of liver glucosensors in the integrated sympathetic response induced by deep hypoglycemia in dogs. Dia- betes, 43, 1052–1060. Heijboer, A.C., Pijl, H., Van den Hoek, A.M., Havekes, L.M., Romijn, J.A. & Corssmit, E.P.M. (2006) Gut-brain axis: regulation of glucose metabolism. Journal of Neuroendocrinology, 18, 883–894. Heller, S.R. & Macdonald, I.A. (1996) The measurement of cognitive function during acute hypoglycaemia: experimental limitations and their effect on the study of hypoglycaemia unawareness. Diabetic Medicine, 13, 607–615. Heptulla, R.A., Tamborlane, W.V., Ma, T.Y.-Z., Rife, F. & Sherwin, R.S. (2001) Oral glucose augments the counterregulatory hormone response during insulin-induced hypoglycemia in humans. Journal of Clinical Endocrinology and Metabolism, 86, 645–648. Hilsted, J. (1993) Cardiovascular changes during hypoglycaemia. Clinical Physiology, 13, 1–10. Jackson, P.A., Cardin, S., Coffey, C.S. et al. (2000) Effect of hepatic denervation on the counterregulatory response to insulin-induced hypoglycemia in the dog. American Journal of Physiology. Endocrinology and Metabolism, 279, E1249–E1257. Jungman, E., Konzog, C., Holl, E., Fassbinder, W. & Schoffling, K. (1989) Effect of a human atrial natriu- retic peptide on blood glucose concentrations and hormone stimulation during insulin-induced hypogly- caemia in healthy man. European Journal of Clinical Pharmacology, 36, 593–597. Katz, S.H., Dhariwal, A.P.S. & McCann, S.M. (1967) Effects of hypoglycemia on the content of pituitary growth hormone (GH) and hypothalamic growth hormone releasing factor (GNRH) in the rat. Endo- crinology, 81, 333–339. Kerr, D., Macdonald, I.A. & Tattersall, R.B. (1989) Influence of duration of hypoglycemia on the hormonal counterregulatory response in normal subjects. Journal of Clinical Endocrinology and Metabolism, 68, 118–122. Kerr, D., Macdonald, I.A., Heller, S.R. & Tattersall, R.B. (1990) A randomised double-blind placebo con- trolled trial of the effects of Metoprolol CR, Atenolol and Propranolol LA on the physiological responses to hypoglycaemia in the non-diabetic. British Journal of Clinical Pharmacology, 29, 685–694. Khoo, E., Wallis, J., Tsintzas, O.K., Macdonald, I. & Mansell, P. (2010) Effects of exenatide on circulating glucose, insulin, glucagon, cortisol and catecholamines in healthy volunteers during exercise. Diabeto- logia, 53, 139–143. Levin, B.E., Routh, V.H., Kang, L., Sanders, N.M. & Dunn-Meynell, A.A. (2004) Neuronal glucosensing: what do we know after 50 years? Diabetes, 53, 2521–2528. Maggs, D., Macdonald, I. & Nauck, M.A. (2008) Glucose homeostasis and the gastrointestinal tract: insights into the treatment of diabetes. Diabetes, Obesity and Metabolism, 10, 18–33. Maggs, D.G., Macdonald, I.A. & Tattersall, R.B. (1994a) Thermoregulatory responses to hyperinsulinaemic hypoglycaemia and euglycaemia in insulin dependent diabetes mellitus. Diabetologia, 37, 689–696. Maggs, D.G., Scott, A.R. & Macdonald, I.A. (1994b) Thermoregulatory responses to hyperinsulinemic hypoglycemia and euglycemia in humans. American Journal of Physiology – Regulatory Integrative and Comparative Physiology, 267, R1266–R1272. McCall, A.L. (1993) Effects of glucose deprivation on glucose metabolism in the central nervous system. In: B.M. Frier & B.M. Fisher (eds), Hypoglycaemia and Diabetes, Clinical and Physiological Aspects, pp. 56–71. Edward Arnold, London. 22 Hypoglycaemia in Clinical Diabetes

McCall, A.L., Fixman, L.B., Fleming, N., Tornheim, K., Chick, W. & Ruderman, N.D. (1986) Chronic hypoglycemia increases brain glucose transport. American Journal of Physiology. Endocrinology and Metabolism, 251, E442–E445. McCrimmon, R.J., Fan, X., Cheng, H. et al. (2005) Activation of ATP-sensitive K+ channels in the ven- tromedial hypothalamus amplifies counterregulatory hormone responses to hypoglycemia in normal and recurrently hypoglycemic rats. Diabetes, 54, 3169–3174. Mitrakou, A., Ryan, C., Veneman, T. et al. (1991) Hierarchy of glycemic thresholds for counterregulatory hormone secretion, symptoms and cerebral dysfunction. American Journal of Physiology. Endocrinology and Metabolism, 266, E67–E74. Owen, O.E., Morgan, A.P., Kemp, H.G., Sullivan, J.M., Herrara, M.G. & Cahill, G.F., Jr. (1967) Brain metabolism during fasting. The Journal of Clinical Investigation, 46, 1589–1595. Page, K.A., Arora, J., Qiu, M., Relwani, R., Constable, R.T. & Sherwin, R.S. (2009) Small decrements in systemic glucose provoke increases in hypothalamic blood flow prior to the release of counterregulatory hormones. Diabetes, 58, 448–452. Patrick, A.W., Hepburn, D.A., Craig, K.J., Thomson, I., Swainson, C.P. & Frier, B.M. (1989) The effects of acute insulin-induced hypoglycaemia on renal function in normal human subjects. Diabetic Medic­ine, 6, 703–708. Ramos, E., Zorilla, E. & Hadley, W.B. (1968) Fever as a manifestation of hypoglycemia. Journal of the American Medical Association, 205, 590–592. Sandoval, D.A., Guy, D.L., Richardson, M.A., Ertl, A.C. & Davis, S.N. (2006) Acute, same-day effects of antecedent exercise on counterregulatory responses to subsequent hypoglycemia in type 1 diabetes mel- litus. American Journal of Physiology. Endocrinology and Metabolism, 290, E1331–E1338. Schvarcz, E., Palmér, M., Åman, J. & Berne, C. (1995) Hypoglycemia increases the gastric emptying rate in healthy subjects. Diabetes Care, 18, 674–676. Sieber, F.E. & Traysman, R.J. (1992) Special issues: glucose and the brain. Critical Care Medicine, 20, 104–114. Siegal, A.M. & Kreisberg, R.A. (1975) Metabolic homeostasis: insulin-glucagon interactions. In: K.E. Sussman & R.J.S. Metz (eds), Diabetes Mellitus, 4th edn, pp. 29–35. American Diabetes Association, New York. Sokaloff, L. (1989) Circulation and energy metabolism of the brain. In: G. Siegel, B. Agranoff, R.W. Albers & P. Molinoff (eds), Basic Neurochemistry, pp. 565–590. Raven Press, New York. Sommerfield, A.J., Wilkinson, I.B., Webb, D.J. & Frier, B.M. (2007) Vessel wall stiffness in type 1 diabetes and the central hemodynamic effects of acute hypoglycemia. American Journal of Physiology. Endo- crinology and Metabolism, 293, E1274–E1279. Teves, D., Videen, T.O., Cryer, P. & Powers, W.J. (2004) Activation of human medial prefrontal cortex during autonomic responses to hypoglycaemia. Proceedings of the National Academy of Sciences of the of America, 101, 6217–6622. Trovati, M., Massucco, P., Mularoni, E. et al. (1988) Insulin-induced hypoglycaemia increases plasma concentrations of angiotensin II and does not modify atrial naturetic polypeptide secretion in man. Diabetologia, 31, 816–820. Wieloch, T. (1985) Hypoglycaemia induced neuronal damage prevented by an N-Methyl D-Aspartate antagonist. Science, 230, 681–683. 2 Symptoms of Hypoglycaemia and Effects on Mental Performance and Emotions Ian J. Deary1 and Nicola N. Zammitt2 1 University of Edinburgh, Edinburgh, UK 2 Royal Infirmary of Edinburgh, Edinburgh, UK

INTRODUCTION

This chapter describes the symptoms that are perceived during acute hypoglycaemia and the changes in mental functions and emotions that occur during this metabolic state. The most obvious benefit to a person of knowing about the symptoms of hypoglycaemia is the ability to recognise its onset as early as possible. This is of key importance in inform- ing and educating people with diabetes. Moreover, if a person with diabetes understands which mental functions are affected by hypoglycaemia, he or she can judge which activities may be most threatened in this state.

SYMPTOMS OF HYPOGLYCAEMIA

Identifying the symptoms The physiological responses to hypoglycaemia are described in Chapter 1. The response to hypoglycaemia results in physical symptoms, which raises several questions (McAulay et al. 2001a). Can we compile a comprehensive list of symptoms of hypoglycaemia? Which are the more common symptoms of hypoglycaemia? Are there early warning symptoms of hypoglycaemia? Do people differ in how quickly and accurately they detect or recognise hypoglycaemia? Do people differ in the set of symptoms of hypoglycaemia they experi- ence? How can individuals distinguish the symptoms of hypoglycaemia from other bodily changes?

The total symptom complex The most basic question is: what symptoms do people report when they develop hypoglycaemia? In humans the symptoms associated with hypoglycaemia were first recorded when insulin became available for the treatment of diabetes (Fletcher and Campbell 1922). A list of characteristic symptoms was described (Table 2.1). It was noted that:

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 24 Hypoglycaemia in Clinical Diabetes

Table 2.1 Common symptoms associated with hypoglycaemia

Symptoms Percentage (minimum– Symptoms of Symptoms (and % [to associated with maximum) of people hypoglycaemia nearest 5%]) of people hypoglycaemia as reporting the given as noted by endorsing symptoms as derived from symptom as Fletcher and associated with population studies associated with Campbell hypoglycaemia; (after (after Hepburn hypoglycaemia (after (1922) Cox et al. 1993a, 1993) Hepburn 1993) Figure 2)

Sweating 47–84 Sweating Sweating (80) Trembling 32–78 Tremulousness Trembling (65) Weakness 28–71 Weakness Fatigue/weak (70) Visual disturbance 24–60 Diplopia Blurred vision (20) Hunger 39–49 Excessive hunger Hunger (60) Pounding heart 8–62 Change in pulse Pounding heart (55) rate Difficulty with 7–41 Dysarthria, Slurred speech (40) speaking sensory and motor aphasia Tingling around the 10–39 – Numb lips (50) mouth Dizziness 11–41 Vertigo, faintness, Light-headed/dizzy (60) syncope Headache 24–36 Headache (30) Anxiety 10–44 Nervousness, Nervous/tense (65) anxiety, excitement, emotional upset Nausea 5–20 – – Difficulty 31–75 – Difficulty concentrating concentrating (80) Tiredness 38–46 – Drowsiness 16–33 – Drowsy–sleepy (40) Confusion 13–53 Confusion Disorientation ‘Goneness’ Pallor Incoordination Uncoordinated (75) Feeling hot or Cold sweats (40) cold Emotional instability Slowed thinking (70)

• some symptoms appeared before others during hypoglycaemia; • the blood glucose level at which subjects became aware of hypoglycaemia was char- acteristic for the individual; • there were large individual differences in the levels of blood glucose at which aware- ness of hypoglycaemia commenced; • the preceding blood glucose concentration could affect the onset of symptoms.

Lists of common symptoms of hypoglycaemia have been compiled from more recent research. Hepburn (1993) summarised eight population studies of the symptoms of hypogly- Chapter 2: Symptoms of Hypoglycaemia and Effects on Mental Performance and Emotions 25 caemia experienced by adults and children with insulin-treated diabetes, and Cox et al. (1993a) also produced a list of symptoms (Table 2.1). It is evident that the three lists of symptoms do not differ greatly, and that Fletcher and Campbell’s early report (1922), had captured many of the symptoms found in subsequent, more structured investigations. However, their report omitted to mention some symptoms such as tiredness, drowsiness and difficulty concentrating, though it did include others – such as pallor (a sign rather than a symptom), incoordination and feelings of temperature change – that are emphasised by other researchers as regularly perceived symptoms. Table 2.1 establishes a useful group of symptoms that are commonly reported in hypoglycaemia. The way we ask people to describe their symptoms of hypoglycaemia can alter what they tell us. The rank order of symptoms alters considerably if patients are asked to indicate the relevance of each symptom rather than merely to identify that the symptom is associ- ated with hypoglycaemia (Cox et al. 1993a). With regard to the criterion of relevance, the most useful symptoms in detecting hypoglycaemia are as follows:

• sweating; • trembling; • difficulty concentrating; • nervousness, tenseness; • light-headedness, dizziness.

The initial symptoms Another important question is which hypoglycaemic symptoms appear early during an episode? The symptoms of hypoglycaemia that appear first and offer early warning of the onset of hypoglycaemia (Hepburn 1993) are as follows:

• trembling; • sweating; • tiredness; • difficulty concentrating; • hunger.

This knowledge is obviously useful for the prompt detection and treatment of hypoglycaemia.

The validity of symptom beliefs The individuality of hypoglycaemic symptom clusters A great deal of the interest in symptoms of hypoglycaemia has been stimulated by concerns about patient education. It is helpful to let patients know the range of symptoms found in hypoglycaemia and to inform them of the common early warning symptoms, much as we all know the range of symptoms that are experienced with the common cold. Many surveys and laboratory studies have shown that people differ considerably in the symptoms of hypoglycaemia they experience (Cox et al. 1993a). In addition to learning the generally reported symptoms, individuals with diabetes should be encouraged to learn about their own typical symptoms of hypoglycaemia. 26 Hypoglycaemia in Clinical Diabetes

The extent to which an individual’s symptoms vary from one episode of hypoglycae- mia to another has been examined recently. A statistical method for quantifying variability in symptom reporting has demonstrated that most people exhibit high levels of intra- individual and between-episode variability, with female gender being the only factor that is associated with greater variability (Zammitt et al. 2011). This has an important message for patient education. Although people with diabetes should be familiar with their own symptom complex, they should appreciate that these symptoms can vary between episodes and they must not therefore be too rigid in what they perceive to be a typical episode of hypoglycaemia.

Correctly interpreting symptoms as representing hypoglycaemia Symptoms of hypoglycaemia do not appear on top of the bodily equivalent of a blank sheet of paper. The alert person with diabetes who is on the lookout for hypoglycaemia must make two sorts of decisions. First, symptoms of hypoglycaemia must be detected and cor- rectly identified. It would be dangerous for a patient to ignore symptoms of hypoglycaemia because he or she thought they were related to something else. Second, symptoms that have nothing to do with hypoglycaemia must be excluded. Unwanted hyperglycaemia could occur if patients treated themselves for hypoglycaemia when the symptoms had another cause. These two main types of error are a failure to treat hypoglycaemia when blood glucose is low, and inappropriate treatment when blood glucose is acceptable or high (Cox et al. 1985, 1993a) (Figure 2.1).

Blood glucose concentration – symptom report correlations Do patients’ symptom reports bear any relation to their concurrent blood glucose concentra- tions? After all, the principal aim of educating people to be aware of symptoms of hypogly- caemia is that they become alert to low and potentially dangerous levels of blood glucose. To answer the above question some researchers have employed a field study approach where people list any symptoms that they experience and then measure and record their blood glucose concentration several times a day for weeks. As a result of these studies it is known that each person has some symptoms that are most reliably associated with their

True symptom status

Symptom of Non-hypoglycaemic hypoglycaemia symptom

Hypoglycaemic Appropriate Inappropriate anti- Perception of action hypoglycaemia action symptom by person Non- Inappropriate Appropriate inaction hypoglycaemic inaction (danger of (with respect to severe hypoglycaemia) hypoglycaemia)

Figure 2.1 Consequences of correct and incorrect perception of hypoglycaemic symptoms. Chapter 2: Symptoms of Hypoglycaemia and Effects on Mental Performance and Emotions 27 actual blood glucose concentrations (Pennebaker et al. 1981). Some of the symptoms that people report during hypoglycaemia are more closely related to their actual blood glucose concentrations than are others, and if we can identify each individual’s most informative symptoms, we can instruct people to pay more attention to them. The following symptoms are most consistently associated with actual blood glucose concentrations (Pennebaker et al. 1981):

• hunger (in 53% of people); • trembling (in 33%); • weakness (in 27%); • light-headedness (in 20%); • pounding heart and rapid heart rate (both 17%).

The same symptoms are not informative for everyone. There were 27% of people for whom weakness was significantly associated with hypoglycaemia, but in 7% it was a good symptom of hyperglycaemia! Most people reported more than three symptoms that were strongly associated with the measured blood glucose concentration. It is evident that an individual’s symptoms are idiosyncratic. If we can help a patient to identify the symptoms of hypoglycaemia peculiar to him or her, which relate to actual blood glucose concentrations, then, by attending to these symptoms, the person should be espe- cially accurate in recognising hypoglycaemia. People who have one or more reliable symptom(s) of hypoglycaemia correctly recognise half of their episodes of hypoglycaemia (defined as a blood glucose less than 3.95 mmol/l (Cox et al. 1993a)). Those who have four or more reliable symptoms recognise a blood glucose below 3.95 mmol/l three-quarters of the time. The field study method has suggested that attention to the following symptoms was particularly useful in detecting actual low blood glucose concentrations:

• nervousness/tenseness; • slowed thinking; • trembling; • light-headedness/dizziness; • difficulty concentrating; • pounding heart; • lack of coordination.

Classifying symptoms of hypoglycaemia Until now the symptoms of hypoglycaemia have been treated as a homogeneous whole. Can these symptoms be divided into different groups? Hypoglycaemia has effects on more than one part of the body. First, the direct effects of a low blood glucose concentration on the brain – especially the cerebral cortex – cause neuroglycopenic symptoms. Second, autonomic symptoms result from activation of parts of the autonomic nervous system. Finally, there may be some non-specific symptoms that are not directly generated by either of these two mechanisms. It is only relatively recently that scientific investigations have taken place to confirm the idea that these separable groups of hypoglycaemic symptoms exist. As suggested above, at least two distinct groups of symptoms develop during the body’s reaction to hypoglycaemia (Hepburn et al. 1991): 28 Hypoglycaemia in Clinical Diabetes • Autonomic, with symptoms such as trembling, anxiety, sweating and warmness. • Neuroglycopenic, with symptoms such as drowsiness, confusion, tiredness, inability to concentrate and difficulty speaking.

This information can assist with patient education by supplying evidence for separable groups of symptoms, and by indicating which symptoms belong to each group. Some neuroglycopenic symptoms, such as the inability to concentrate, weakness and drowsiness, are among the earliest detectable symptoms, but patients tend to rely more on autonomic symptoms when detecting the onset of hypoglycaemia. Paying more attention to the poten- tially useful, early neuroglycopenic symptoms could help with the early detection of hypoglycaemia. Similar groups of symptoms of hypoglycaemia have been discovered by asking people to recall the symptoms they typically noticed during hypoglycaemia. However, in addition to the two groups described above, a general feeling of malaise is added (Deary et al. 1993):

• Autonomic: e.g. sweating, palpitations, shaking and hunger. • Neuroglycopenic: e.g. confusion, drowsiness, odd behaviour, speech difficulty and incoordination. • General malaise: e.g. headache and nausea.

These 11 symptoms are so reliably reported by people and so clearly separable into these three groups, that they are used as the ‘Edinburgh Hypoglycaemia Scale’ (Deary et al. 1993). Table 2.2 shows how different researchers have found similar groups of autonomic and hypoglycaemic symptoms. Physiological studies have also confirmed that the symptoms of hypoglycaemia can be divided into autonomic and neuroglycopenic groups. Symptoms such as sweating, hunger, pounding heart, tingling, nervousness and feeling shaky/tremulous (autonomic symptoms) can be reduced or even prevented by drugs that block neurotransmission within the auto- nomic nervous system (Towler et al. 1993), confirming that these symptoms are caused by the autonomic response to hypoglycaemia. Symptoms such as warmth, weakness, difficulty thinking/confusion, feeling tired/drowsy, feeling faint, difficulty speaking, dizziness and blurred vision (neuroglycopenic symptoms) are not prevented by drugs that block the autonomic nervous system. Therefore, neuroglycopenic symptoms are not mediated via the

Table 2.2 Different authors’ lists of autonomic and neuroglycopenic symptoms of hypoglycaemia

Autonomic Neuroglycopenic

Deary et al. Towler et al. Weinger et al. Deary et al. Towler et al. (1993) (1993) (1993) (1995) (1993) Sweating Sweaty Sweating Confusion Difficulty thinking/ confused Palpitation Heart pounding Pounding heart, Drowsiness Tired/drowsy fast pulse Shaking Shaky/tremulous Trembling Odd behaviour Hunger Hungry Speech difficulty Difficulty speaking Tingling Incoordination Nervous/anxious Tense Weak Breathing hard Warm Chapter 2: Symptoms of Hypoglycaemia and Effects on Mental Performance and Emotions 29

Table 2.3 Symptom groupings of the Edinburgh Hypoglycaemia Scale during experimentally- induced hypoglycaemia. The scale can also be used to grade symptom intensity on a Likert scale from 1 to 7, where 1 indicates that a symptom is not present while 7 indicates that the symptom is intense. The symptoms are divided into Neuroglycopenic (N), Autonomic (A) and Malaise (M) groups

Instructions: Please score the extent to which you experience the following symptoms during a typical daytime hypoglycaemic episode.

Not present Present a great deal (N) Confusion 1 2 3 4 5 6 7 (A) Sweating 1 2 3 4 5 6 7 (N) Drowsiness 1 2 3 4 5 6 7 (N) Incoordination 1 2 3 4 5 6 7 (N) Speech difficulty 1 2 3 4 5 6 7 (A) Palpitation 1 2 3 4 5 6 7 (N) Odd behaviour 1 2 3 4 5 6 7 (A) Hunger 1 2 3 4 5 6 7 (M) Nausea 1 2 3 4 5 6 7 (A) Shaking 1 2 3 4 5 6 7 (M) Headache 1 2 3 4 5 6 7

autonomic nervous system and are thought to be caused by the direct effect of glucose deprivation on the brain. This type of research has also observed that people tend to rely on autonomic symptoms to detect hypoglycaemia, even when neuroglycopenic symptoms are just as prominent (Towler et al. 1993). Once more, this suggests that more emphasis should be placed on education of the potential importance of neuroglycopenic symptoms for the early warning of hypoglycaemia. Symptoms might gather into slightly different groupings depending on the situation. The symptom groupings in the Edinburgh Hypoglycaemia Scale were developed from retrospec- tive reports by adults with diabetes. However, when people are asked to rate the same group of symptoms during acute, experimentally-induced moderate hypoglycaemia, a slightly different pattern emerges (McCrimmon et al. 2003). In Table 2.3 there is an autonomic grouping (labelled A), and the single neuroglycopenia group has been divided into two symptom groups: one with mostly cognitive symptoms (labelled N to denote neuroglyco- penia) and the other with more general symptoms (labelled M to denote malaise). This division probably arose because the subjects in the studies used to form Table 2.3 were engaged in cognitive tasks during the period of hypoglycaemia. Therefore, they would be especially aware of cognitive shortcomings, making this group of symptoms more promi- nent and coherent.

Symptoms in children and older people Children often have difficulty in recognising symptoms of hypoglycaemia, and they show marked variability in symptoms between episodes of hypoglycaemia (Macfarlane and Smith 1988). Trembling and sweating are often the first symptoms recognised by children. From interviews with the parents of children (aged up to 16 years) with type 1 diabetes, and with some of their children, more is known about the frequency of symptoms of hypoglycaemia in children (McCrimmon et al. 1995; Ross et al. 1998) (Table 2.4). The most frequently reported sign that parents observed was pallor (noted by 88%). The parents frequently reported symptoms of behavioural disturbance such as irritability, argumentativeness and 30 Hypoglycaemia in Clinical Diabetes

Table 2.4 Symptoms of hypoglycaemia in children (Derived from Ross et al. 1998)

Frequency of rating (%)

Symptom Parents’ reports Children’s reports

Tearful 73 47 Headache 73 65 Irritable 85 65 Uncoordinated 62 56 Naughty 47 31 Weak 79 83 Aggressive 75 62 Trembling 79 88 Sleepiness 63 69 Nightmares 33 19 Sweating 76 73 Slurred speech 53 45 Blurred vision 52 55 Tummy pain 67 41 Feeling sick 63 53 Hungry 74 84 Yawning 48 45 Odd behaviour 65 50 Warmness 57 68 Restless 61 57 Daydreaming 70 48 Argumentative 64 50 Pounding heart 21 44 Confused 75 70 Tingling lips 20 24 Dizziness 66 87 Tired 83 76 Feeling awful 92 79

aggression. This latter group of symptoms is not prominent in adults, although the Edin- burgh Hypoglycaemia Scale includes ‘odd behaviour’ as an adult neuroglycopenic symptom. Others had previously noted the prominence of symptoms such as irritability, aggression and disobedience in the parents’ reports of their children’s symptoms of hypoglycaemia (Mac- farlane and Smith 1988; Macfarlane et al. 1989). Parents tend to under-report the subjective symptoms of hypoglycaemia, such as weakness and dizziness, but generally there is good agreement between parents and their children about the most prominent symptoms of child- hood hypoglycaemia (McCrimmon et al. 1995; Ross et al. 1998). Separate groups of autonomic and neuroglycopenic symptoms were not found in chil- dren with type 1 diabetes (McCrimmon et al. 1995; Ross et al. 1998). These symptoms are reported together by children and are not distinguished as separate groups, whereas the group of symptoms related to behavioural disturbance is clearly reported as a distinct group. In a refinement of the earlier study by McCrimmon et al. (1995), Ross et al. (1998) found that parents could distinguish between autonomic and neuroglycopenic symptoms. People with insulin-treated type 2 diabetes report symptoms during hypoglycaemia that separate into autonomic and neuroglycopenic groups (Henderson et al. 2003). Older people with insulin-treated type 2 diabetes commonly report neurological symptoms of hypogly- caemia, which may be misinterpreted as features of cerebrovascular disease, such as tran- Chapter 2: Symptoms of Hypoglycaemia and Effects on Mental Performance and Emotions 31

Table 2.5 Classification of symptoms of hypoglycaemia using Principal Components Analysis in patients with insulin-treated diabetes depending on age group

Children (pre pubertal) Adults Elderly

Autonomic/neuroglycopenic Autonomic Autonomic Neuroglycopenic Neuroglycopenic Behavioural Non-specific malaise Neurological

sient ischaemic attacks (Jaap et al. 1998). The features of hypoglycaemia in the elderly are discussed in more detail in Chapter 12. The age-specific differences in the groups of hypoglycaemic symptoms, classified using statistical techniques (Principal Component Analysis), are shown in Table 2.5. Health professionals and carers who are involved in the treatment and education of diabetic patients should be aware of which symptoms are common at either end of the age spectrum.

From symptom perception to action People with diabetes are better at estimating their blood glucose in natural, everyday situ- ations, as opposed to laboratory settings (Cox et al. 1985). In some ways this is surprising as non-experimental hypoglycaemia often occurs unexpectedly. In this situation, attention toward symptoms will not be as actively directed toward detection as in the laboratory setting where it is usually anticipated. Furthermore, hypoglycaemia in everyday life occurs on the background of other bodily feelings and must be separated from other causes of the same symptoms. For example, exercise and various acute illnesses can provoke sweating in people with diabetes, independently of their association with hypoglycaemia. In a real-life situation a person must first detect symptoms of hypoglycaemia and then interpret them. Failure to detect symptoms can lead to a failure to treat hypoglycaemia, but detection without the correct interpretation of the cause of the symptoms is equally danger- ous. Furthermore, it is obvious that someone who does interpret symptoms correctly as being caused by hypoglycaemia, but who does not take action to treat the low blood glucose, will be at the same risk. These and other steps toward the avoidance of severe hypoglycaemia demonstrate the key role of education about symptoms in people with diabetes (Gonder- Frederick et al. 1997). The psycho-educational programmes of Blood Glucose Awareness Training (BGAT; Schachinger et al. 2005) and Hypoglycaemia Anticipation, Awareness and Treatment Training (HAATT; Cox et al. 2004) have led to better recognition of hypogly- caemic states and reduced frequency of hypoglycaemia (see Chapters 6 and 16).

Symptom generation Table 2.6 outlines the stages that intervene between low blood glucose occurring in an individual and the implementation of effective treatment, along with the factors that can affect each stage (Gonder-Frederick et al. 1997). It is interesting to note the importance of behavioural factors in generating states of low blood glucose. Episodes of low blood glucose are most likely to come about because of changes in routine aspects of diabetes management, such as taking extra insulin, eating less food or taking more exercise (Clarke et al. 1997). These factors predict more than 85% of episodes of hypoglycaemia in people with diabetes. 32 Hypoglycaemia in Clinical Diabetes

Table 2.6 A model for the occurrence and avoidance of severe hypoglycaemia. Note on the right-hand side of each stage the factors that affect its occurrence. (Date from Gonder-Frederick L 1997)

Stage in recognition and prevention Factors affecting occurrence of of severe hypoglycaemia each stage

Generation of physiological response Degree of hypoglycaemia Metabolic control Antecedent hypoglycaemia Integrity of hormonal counterregulation Generation of physical symptoms Amount of adrenaline secreted Brain glucose level Medications Caffeine Detection of symptoms Attention Distraction Salience Relevant activity Detection of low blood glucose Knowledge Symptom beliefs Competing explanations Denial Impaired consciousness Decision to treat low blood glucose Perceived risk Cost-benefit analysis Tolerance of hypoglycaemia Personality, coping Initiation of appropriate self-treatment Availability of carbohydrate Too neuroglycopenic to respond

In the presence of an intact physiological response to low blood glucose, autonomic and neuroglycopenic symptoms and symptoms of general malaise are generated (Table 2.6). The degree of hypoglycaemia, the person’s quality of glycaemic control and any recent episodes of hypoglycaemia may all affect the magnitude of the body’s physiological response. Recent, preceding hypoglycaemia can reduce the symptomatic and counterregulatory hor- monal responses to subsequent hypoglycaemia, resulting in a diminished awareness of symptoms. This effect of ‘antecedent’ hypoglycaemia is described in Chapter 6. Gender does not appear to influence the symptomatic response to hypoglycaemia (Geddes et al. 2006). In the second row in Table 2.6 comes the actual generation of physical symptoms. Among the variables that can influence this stage is the prior ingestion of caffeine, which enhances the intensity of the autonomic and neuroglycopenic symptoms of experimentally induced hypoglycaemia (Debrah et al. 1996) (see Chapter 8). Caffeine may act by increas- ing the intensity of symptoms of hypoglycaemia to perceptible levels, much as a magnify- ing glass enables one to read otherwise too-small print.

Symptom detection The occurrence of physiological changes in the body does not guarantee that a person will detect symptoms (Table 2.6) (Gonder-Frederick et al. 1997). If attention is directed to physical changes, people are more likely to detect symptoms than if their attention is held Chapter 2: Symptoms of Hypoglycaemia and Effects on Mental Performance and Emotions 33 elsewhere. Everyone has had the experience of feeling less discomfort when being dis- tracted by something diverting. The personal relevance of the symptom may affect detec- tion; for example, a person with heart disease may be very likely to detect palpitations (Cox et al. 1993a). The activity of the person at the time of the physiological change is obviously important. Hypoglycaemic symptoms will be more obvious to the person engaged in active mental effort (McCrimmon et al. 2003), such as sitting an examination, than to the person relaxing and watching television. A doctor engaged in microsurgery may be very sensitive to the onset of tremor. In one laboratory study the people who had higher anxiety levels were better at detecting symptoms of hypoglycaemia (Ryan et al. 2002). The autonomic symptoms of hypoglycaemia are often emphasised in the detection of hypoglycaemia. However, a strong case can be made for an equal emphasis on neuroglyco- penic symptoms (Gonder-Frederick et al. 1997; McCrimmon et al. 2003) because:

• performance on mental tasks deteriorates during hypoglycaemia, and subjective aware- ness of this decrement begins at very mild levels of hypoglycaemia; • the difference in glycaemic thresholds for the onset of autonomic and neuroglycopenic symptoms is so small that it is unlikely to be detected when blood glucose declines rapidly; • neuroglycopenic symptoms are as strongly related to actual blood glucose concentra- tions as are autonomic symptoms (Cox et al. 1993a); • people with insulin-treated diabetes cite autonomic and neuroglycopenic symptoms with equal frequency as the primary warning symptoms of hypoglycaemia (Hepburn et al. 1991).

Low blood glucose detection – symptom interpretation The correct detection of a symptom of hypoglycaemia does not always lead to correct inter- pretation (Table 2.6). After correctly detecting relevant symptoms, people fail to detect about 26% of episodes of low blood glucose (Gonder-Frederick et al. 1997). Several factors could break a perfect relationship between detection and recognition. However, it should be appre- ciated that symptom detection (internal cues) is not mandatory for the detection of low blood glucose. Self-testing of blood glucose or the information of family members (external cues) can aid recognition of hypoglycaemia without symptom perception (see Box 2.1). Correct knowledge about symptoms of hypoglycaemia is necessary for the detection of low blood glucose. The lack of such knowledge among elderly people with diabetes in particular gives cause for concern (Mutch and Dingwall-Fordyce 1985). Of 161 diabetic people aged between 60 and 87, all of whom were injecting insulin or taking a sulfonylurea, only 22% had ever been told the symptoms of hypoglycaemia and 9% knew no symptoms

Box 2.1 Identification of hypoglycaemia • Internal cues – autonomic, neuroglycopenic and non-specific symptoms. • External cues – relationship of insulin injection to meals, exercise and experience of self-management. • Blood glucose testing – conventional capillary blood glucose measurements and Continuous Glucose Monitoring. • Information from observers (e.g. relatives, friends, colleagues). 34 Hypoglycaemia in Clinical Diabetes at all. The percentages of the insulin- treated diabetic patients who knew that the following symptoms were associated with hypoglycaemia were as follows:

• sweating (82%); • palpitations (62%); • confusion (53%); • hunger (51%); • inability to concentrate (50%); • speech problems (41%); • sleepiness (33%).

In the midst of this ignorance, much hypoglycaemia may not be treated because of a lack of knowledge of the symptoms of hypoglycaemia, which would aid their recognition. Most, if not all, of the symptoms of hypoglycaemia can be explained by other physical conditions. Therefore, correct symptom detection may be usurped by incorrect attribution of the cause. For example, having completed some strenuous activity, an athlete may attribute the symptoms of sweating and palpitations to physical exertion. An obvious problem in detecting a low blood glucose is the fact that the organ responsible for the detection and interpretation of symptoms – the brain, especially the cerebral cortex – is impaired. Thus, impaired concentration and lowered consciousness levels can beget even more severe hypoglycaemia.

Symptom scoring systems The controversy about the effect of human insulin on symptom awareness (Chapter 6) stimulated the development of scoring systems for hypoglycaemia to allow comparative studies between insulin species. This produced scoring systems such as the Edinburgh Hypoglycaemia Scale (Deary et al. 1993), and any such system must be validated for research application. It is important to note that the nature and intensity of individual symp- toms are as important as, if not more important than, the number of symptoms generated by hypoglycaemia. The concepts involved are discussed in detail by Hepburn (1993). More information on the symptoms of hypoglycaemia is provided by McAulay et al. (2001a). Although the initial studies to define these symptom groupings simply asked patients to report the presence or absence of each symptom, it was recommended that a seven-point visual analogue scale could be used to grade symptom intensity (Deary et al. 1993). Since then, the Edinburgh Hypoglycaemia Scale has been used routinely to grade the presence and intensity of symptoms during hypoglycaemia studies (e.g. Warren et al. 2007; Geddes et al. 2008; Zammitt et al. 2008; Wright et al. 2009). It has also been used in studies exam- ining the frequency of hypoglycaemia (UK Hypoglycaemia Study Group 2007) and the intra-individual between-episode variability in symptom reporting (Zammitt et al. 2011).

ACUTE HYPOGLYCAEMIA AND COGNITIVE FUNCTIONING

Symptoms are subjective reports of bodily sensations. With respect to hypoglycaemia some of these reports – especially neuroglycopenic symptoms – pertain to altered cognitive (mental ability) functioning. Do reports of ‘confusion’ and ‘difficulty thinking’ (Table 2.2) concur with objective mental test performance in hypoglycaemia? Before experimental Chapter 2: Symptoms of Hypoglycaemia and Effects on Mental Performance and Emotions 35 hypoglycaemia became an accepted investigative tool in diabetes, expert clinical observers noted impairments of cognitive functions despite clear consciousness during hypoglycae- mia (Fletcher and Campbell 1922; Wilder 1943). Cognitive functions include the following sorts of mental activity: orientation and attention, perception, memory (verbal and non- verbal), language, construction, reasoning, executive function and motor performance. Early studies (Russell and Rix-Trot 1975) established that the following abilities become disrupted below blood glucose levels of about 3.0 mmol/l:

• fine motor coordination; • mental speed; • concentration; • some memory functions.

The hyperinsulinaemic glucose clamp technique allows more controlled experiments of acute hypo- and hyperglycaemia. However, although this technique is used in most studies of cognitive function in hypoglycaemia, it does not mimic the physiological or temporal characteristics of ‘natural’ or intercurrent episodes of hypoglycaemia experienced by people with type 1 diabetes. From laboratory experiments using the glucose clamp technique it was found that blood glucose concentrations between 3.1 and 3.4 mmol/l caused the fol- lowing effects (Holmes 1987; Deary et al. 1993):

• slowed reaction times (hypoglycaemia had more effect on reaction times when the reaction involved making a decision); • slowed mental arithmetic; • impaired verbal fluency (in this test one has to think of words beginning with a given letter, probably involving the frontal lobes of the brain); • impaired performance in parts of the Stroop test (in this test one has to read aloud a series of ink colours when words are printed in a different colour from that of the name, e.g. the word RED printed in green ink).

Some mental functions were spared during hypoglycaemia, for example:

• simple motor (like the speed of tapping) and sensory skills; • the speed of reading words aloud.

By 1993 over 16 studies had investigated cognitive functions during acute and mild– moderate hypoglycaemia (Deary et al. 1993). The levels of blood glucose ranged from 2.0 to 3.7 mmol/l. The way that hypoglycaemia was induced varied among studies, as did the methods of blood sampling (e.g. arterialised or venous blood). Moreover, the ability levels of the people in different studies varied, and there was much heterogeneity in the test bat- teries used to assess mental performance. An authoritative statement as to the mental func- tions disrupted during hypoglycaemia is still not possible. However, in at least one or more of the studies a number of tests were significantly impaired during hypoglycaemia (Box 2.2). Few areas of mental function are preserved at normal levels during acute hypoglycae- mia. There is a general dampening of many abilities that involve conscious mental effort. In the face of so many deleterious effects, what mental functions remain intact during acute hypoglycaemia? At blood glucose concentrations similar to those indicated above, the fol- lowing mental tests are not significantly impaired: 36 Hypoglycaemia in Clinical Diabetes

Box 2.2 Cognitive function tests impaired during acute hypoglycaemia • Trail making (involving visual scanning and mental flexibility). • Digit symbol (speed of replacing a list of numbers with abstract codes). • Reaction time (especially involving a decision). • Mental arithmetic. • Verbal fluency. • Stroop test. • Grooved pegboard (a test of fine manual dexterity). • Pursuit rotor (a test of eye–hand co-ordination). • Letter cancellation (striking out occurrences of a given letter in a page of letters). • Delayed verbal memory, story and word recall, prospective memory, working memory. • Spatial abilities (hidden patterns, paper folding, card rotations, maze tracing). • Executive function: coordinating initiation, sequencing and monitoring of complex behaviour.

• finger tapping; • forward digit span (repeating back a list of numbers in the same order); • simple reaction time; • elementary sensory processing. Thus tests which involve speeded responses and which are more cognitively complex and attention-demanding tend to show impairment during hypoglycaemia (Deary et al. 1993). Heller and Macdonald (1996) have concluded that: • even quite severe degrees of hypoglycaemia do not impair simple motor functions; • choice reaction time (where a mental decision is needed before reacting to a stimulus) is affected at higher blood glucose concentrations more than simple reaction time; • speed of responding is sometimes slowed in a task in which accuracy is preserved; • many aspects of mental performance become impaired when blood glucose falls below about 3.0 mmol/l; • there are important individual differences; some people’s mental performance is already impaired above a blood glucose of 3.0 mmol/l, whereas others continue to function well at lower levels; • the speed of response of the brain in making decisions slows down during hypoglycae- mia (Jones et al. 1990; Tallroth et al. 1990); • it can take as long as 40–90 minutes after blood glucose returns to normal for the brain to recover fully (Blackman et al. 1992; Lindgren et al. 1996; Evans et al. 2000; Lobmann et al. 2000; Zammitt et al. 2008).

Determining whether mental performance is impaired at all during mild to moderate hypoglycaemia, while a person is still fully conscious, is only the beginning of this line of investigation. The next question to ask is whether some particular functions are more sus- ceptible and some less so? Figure 2.2 encapsulates this problem and illustrates three other important questions about the cognitive effects of acute hypoglycaemia:

1. What factors affect the degree of cognitive impairment during hypoglycaemia, other than the level of blood glucose? Chapter 2: Symptoms of Hypoglycaemia and Effects on Mental Performance and Emotions 37

Job sample tasks Driving

Practical applications Factors affecting degree of Psychometric tests cognitive dysfunction Memory Fluid intelligence e.g. gender, IQ level, Information processing Attention glycaemic control, impaired Psychomotor speed Dexterity hypoglycaemia Language Spatial ability awareness, previous hypoglycaemia Investigation of basic brain processes

Cognitive studies: Brain imaging: e.g. e.g. working memory, PET, SPET, MRI implicit learning

Psychophysics: Psychophysiology: e.g. visual and e.g. evoked auditory processing potentials

Figure 2.2 Cognitive effects of hypoglycaemia. (Source: Deary IJ 1998. Reproduced with permission from Elsevier).

2. Do impairments in laboratory cognitive tasks have a bearing on mental performance in real life? 3. Which basic brain functions are disturbed during acute hypoglycaemia?

Influences on the degree of, and threshold for, cognitive dysfunction During acute hypoglycaemia Although, on average, impairment of mental performance is worse during hypoglycaemia, some people do not change or may even improve (Pramming et al. 1986; Hoffman et al. 1989; Zammitt et al. 2008). It is not yet certain whether such individual differences in responses are stable (Gonder-Frederick et al. 1994; Driesen et al. 1995). The following factors might increase a person’s degree of cognitive impairment during acute hypoglycaemia: • male gender (Draelos et al. 1995; but this is disputed for people with type 2 diabetes by Bremer et al. 2006); • impaired hypoglycaemia awareness has been associated with a worsening of cognitive impairment in some studies (Gold et al. 1995b) but preservation of cognitive function in others (Zammitt et al. 2008); • type 1 diabetes (Wirsen et al. 1992); • high IQ (Gold et al. 1995a). 38 Hypoglycaemia in Clinical Diabetes

Does glycaemic control affect the cognitive impact of hypoglycaemia? People with type 1 diabetes on intensified insulin therapy attain glycaemic control that is nearer to normal than most people treated with conventional insulin treatment. As a result, the frequency of severe hypoglycaemia is increased and is associated with a greater risk of impaired hypogly- caemia awareness. Neuroendocrine responses to hypoglycaemia are reduced in magnitude and begin at lower absolute blood glucose concentrations than in people with less strict glycaemic control. However, it is uncertain whether the hypoglycaemic threshold for cogni- tive dysfunction changes in a similar fashion. People with type 1 diabetes on intensive insulin therapy reported autonomic and neuroglycopenic symptoms at blood glucose con- centrations of about 2.4 and 2.3 mmol/l, respectively, whereas in those with less strict glycaemic control and in non-diabetic individuals, these symptoms commenced at between 2.8 and 3.0 mmol/l. However, in all three groups, the accuracy and speed in a reaction time test deteriorated significantly at blood glucose concentrations between 2.8 and 3.0 mmol/l (Amiel et al. 1991; Maran et al. 1995). Therefore, those with insulin-treated diabetes who have strict glycaemic control have the misfortune that their mental performance starts to deteriorate before the onset of warning symptoms of hypoglycaemia. By contrast, neuro- physiological responses (P300 event-related potentials), which have been linked to various measures of cognitive function, occur at lower blood glucose concentrations, suggesting that cerebral adaptation has occurred (Ziegler et al. 1992). The effects of quality of gly- caemic control, antecedent hypoglycaemia and impaired awareness of hypoglycaemia on the mental performance responses to hypoglycaemia, and the relation of these responses to the perceptions of the symptoms of hypoglycaemia, are important topics still under study (see Chapters 6 and 7). The interrelation of these factors makes the field complex, and progress is further hampered by the lack of consensus agreement on a validated battery of cognitive tests for use in hypoglycaemia.

Are the cognitive changes during acute hypoglycaemia important and valid? Do the impairments of mental test performance actually have implications for real-life functions? In addition, are the mental changes during hypoglycaemia a result of impair- ments in basic brain functions? One common, important and potentially dangerous area of real-life functioning is driving (see Chapter 18), which involves many cognitive abilities including psycho-motor control and divided attention. Cox et al. (1993b, 2000) employed a sophisticated driving simulator to study driving performance during controlled hypoglycaemia using a glucose clamp technique. With very mild hypoglycaemia (arterialised blood glucose below 3.8 mmol/l) drivers with type 1 diabetes committed significant driving errors, and during hypoglycaemia often drove very slowly, possibly using a compensatory mechanism to avoid making errors. Despite this, more global errors of driving were committed and about half of the participants, despite demonstrating a seriously impaired ability to drive, said they felt competent to drive irrespective of their low blood glucose. It cannot be stated with certainty that the findings obtained in a driving simulator will apply to real-life driving but they go some way towards assessing the effect of hypoglycaemia on daily life. It is relevant to ask whether all drivers with insulin-treated diabetes are equally vulner- able to the disruptive effects of hypoglycaemia. Thirty eight drivers with type 1 diabetes who underwent simulator testing during hypoglycaemia and euglycaemia were subdivided on the basis of a past history (n = 16) or no past history (n = 22) of experiencing recurrent Chapter 2: Symptoms of Hypoglycaemia and Effects on Mental Performance and Emotions 39 hypoglycaemia-related driving mishaps. During hypoglycaemia, drivers with a past history of mishaps secreted less epinephrine (adrenaline) (P = 0.02) and exhibited greater driving impairments (P = 0.03) than drivers with no past history of mishaps, suggesting that a subgroup of drivers may have an increased susceptibility to hypoglycaemia-related driving mishaps (Cox et al. 2010). Other studies that do not involve simulators have examined domains of cognitive func- tion relevant to driving. For example, hypoglycaemia affects aspects of psycho-motor function, including hand-eye coordination and fine motor skills (as measured by the pursuit rotor test) and choice reaction time (Geddes et al. 2008). Reaction time tests measure a range of cognitive and non-cognitive processes. When the effects of hypoglycaemia on choice reaction time were examined in greater detail, it was apparent that hypoglycaemia affects central processing with no effect on either the assessment of the evidence that was necessary to make a decision nor the motor processes involved in reacting to a stimulus (Geddes et al. 2010). Choice reaction time remains impaired for up to 75 minutes after euglycaemia has been restored (Zammitt et al. 2008), a finding that supports the current advice of the Driver and Vehicle Licensing Agency in the UK to wait for at least 45 minutes after correcting hypoglycaemia before resuming driving a vehicle. Spatial ability (Wright et al. 2009) and executive function; the coordination of planning, initiation, sequencing and monitoring of complex behaviours (Graveling et al. 2013), are also skills relevant to driving which deteriorate during hypoglycaemia. Studies that examine the practical cognitive effects of hypoglycaemia are invaluable and more are required. Just as more studies that examine the practical cognitive aspects of hypoglycaemia would be useful, so would more studies of the brain’s processing efficiency. Cognitive tests typically involve a melange of inseparable mental processes, and yet very specific aspects of the human brain’s activities can be measured in the clinical laboratory (Massaro 1993). Studies of the cognitive effects of hypoglycaemia have thus begun to address the impair- ments to various cognitive domains in more detail. Basic, specific aspects of visual and auditory processing have been examined during acute hypoglycaemia in non-diabetic humans. Standard tests of visual acuity – those that are measured by an optometrist – are not affected by hypoglycaemia, but other aspects of vision are affected (McCrimmon et al. 1996; Ewing et al. 1998). These include: • contrast sensitivity (the ability to discriminate faint patterns); • inspection time (the ability to see what is in a pattern when it is shown for a very brief period of time); • visual change detection (the ability to spot a small, quick change in a pattern); • visual movement detection (the ability to spot brief movement in a pattern). This means that the ability to see the environment changes in important ways during hypoglycaemia. Visual acuity is preserved, as tested by the ability to read black letters on a white background. However, most visual activity is not like that; many of the things we see happen quickly and in relatively poor light. When the level of contrast falls, or dis- criminations must be made under pressure of time, visual processing is impaired during hypoglycaemia. These changes are relevant to driving in poorly lit conditions or at dusk. Interestingly, at about the same degree of hypoglycaemia, the ability to distinguish one colour from another is not impaired (Hardy et al. 1995). Speed of auditory processing also appears to be impaired by hypoglycaemia, and the ability to discriminate the loudness of two tones is disrupted (McCrimmon et al. 1997; Strachan et al. 2003). This suggests that 40 Hypoglycaemia in Clinical Diabetes the ability to understand language may be compromised during hypoglycaemia. However, despite there being disruption to central nervous system processing during hypoglycaemia, no disturbance has been detected in peripheral nerve conduction (Strachan et al. 2001). If basic information processing provides a fundamental limitation to how well the brain is operating, then at a higher level of function, attention is important in carrying out a number of cognitive functions. A detailed study of a number of different aspects of atten- tion during hypoglycaemia found that the abilities to attend selectively and to switch atten- tion as necessary both deteriorated (McAulay et al. 2001b, 2005). In turn, attention is necessary in order to learn and form new memories. Detailed studies have been made of different aspects of memory during acute, insulin-induced hypoglycae- mia (Deary et al. 2003; Sommerfield et al. 2003a, 2003b). Most memory systems are disrupted during hypoglycaemia. However, some are especially badly affected: long-term memory, which is the ability to retain new information after many minutes and much dis- traction; working memory, which is the ability to retain and manipulate information at the same time; and prospective memory, which is the ability to remember to do things (as in a shopping list) (Warren et al. 2007). Indeed, the ability to perform one tricky working memory task was obliterated during hypoglycaemia (Deary et al. 2003). That is, no matter how good the person was at performing the task during euglycaemia, the same task could not be done during moderate hypoglycaemia. Hypoglycaemia and cognitive function has been reviewed (Warren and Frier 2005; Inkster and Frier 2012).

ACUTE HYPOGLYCAEMIA AND EMOTIONS

Mood change is part of the experience of hypoglycaemia. Moods are emotion-like experi- ences that are quite general rather than applied to specific situations. Psychologists recog- nise three basic moods:

• energetic arousal (a tendency to feel lively and active rather than tired and sluggish); • tense arousal (a tendency to feel anxious and nervous versus relaxed and calm); • hedonic tone (a tendency to feel happy versus sad).

When people are asked to rate their mood states during hypoglycaemia induced in the laboratory, changes occur in all of these basic mood states. People feel less energetic, more tense and less happy (Gold et al. 1995c; McCrimmon et al. 1999a; Hermanns et al. 2003). During hypoglycaemia the emotional arousal in response to stimuli becomes more intense (Hermanns et al. 2003). In addition, some people become more irritable and have angry feelings during hypoglycaemia (Merbis et al. 1996; McCrimmon et al. 1999b). The feeling of low energy takes over 30 minutes to be restored to normal, whereas the feelings of tense- ness and unhappiness disappear when blood glucose returns to normal. The prolonged feeling of low energy after hypoglycaemia may affect work performance, so that when hypoglycaem­ ­ia has been treated, an immediate return to the normal state should not be expected. Hypoglycaemia also alters the way some people look at their life problems. When junior doctors were asked to assess their career prospects during controlled hypoglycaemia, they were more pessimistic (McCrimmon et al. 1995), and if a general state of pessimism is common during hypoglycaemia, it would be a poor state from which to make personal decisions. It is possible that the change in mood states during hypoglycaemia is one of the Chapter 2: Symptoms of Hypoglycaemia and Effects on Mental Performance and Emotions 41 causes of adults admitting to more ‘odd behaviour’ (Deary et al. 1993). Altered mood may also account in part for symptoms of behavioural disturbance that are so prominent in the responses to hypoglycaemia of children with diabetes (McCrimmon et al. 1995). In addition to emotional responses as a result of hypoglycaemia, some people have emotional responses in anticipation of hypoglycaemia. In Edinburgh one young man with insulin-treated diabetes developed a phobic anxiety state; his phobia related to becoming comatose as a result of hypoglycaemia (Gold et al. 1997a). Such a case is exceptional, but many people with diabetes are afraid of hypoglycaemia (see Chapters 16 and 18). The Hypoglycaemia Fear Survey (HFS), which comes in two parts, measures this tendency (Cox et al. 1987). The first part (Worry Subscale) asks people several questions concerning how much they worry about hypoglycaemia (e.g. ‘Do you worry if you have no one around you during a [hypoglycaemic] reaction?’). The second part (Behaviour Subscale) asks several questions about what people do to avoid hypoglycaemia (e.g. ‘Do you eat large snacks at bedtime?’). People with greater fear of hypoglycaemia (Irvine et al. 1992; Polon- sky et al. 1992; Hepburn et al. 1994; ter Braak and Appleman 2000): • have more anxious personalities in general; • are more likely to confuse symptoms of anxiety for those of hypoglycaemia; • report having had more episodes of hypoglycaemia. It is not yet known whether people who experience more hypoglycaemia become wor- riers about it, or whether people who are worriers in general just worry more about hypogly- caemia as well. Perhaps both are true. However, it does seem likely that the experience of more severe hypoglycaemia in the past and the development of impaired awareness of hypoglycaemia lead to increased worry about subsequent hypoglycaemia (Gold et al. 1997b). It should also be noted that increased anxiety levels can impair the ability to detect hypoglycaemia and thus increase the risk of future episodes (Ryan et al. 2002). However, the Blood Glucose Awareness Training programme, which is specifically designed for people with diabetes, has been shown to reduce both fear and frequency of hypoglycaemia (Cox et al. 2001). Fear of hypoglycaemia in diabetes has been reviewed by Wild et al. (2007).

CONCLUSIONS • Because people with diabetes are closely involved in their own treatment it is important that they and their educators know about the main side effects and sequelae of the disorder and its treatments. • Accurate knowledge of the symptoms of hypoglycaemia may be used to avoid the dangers of hypoglycaemia. • The progressively more serious impairment in cognitive function that occurs as blood glucose declines provides knowledge about the brain’s compromised state during hypoglycaemia: basic functions such as visual processing deteriorate and driving becomes dangerously error prone. • Performance on a host of mental tests becomes worse during hypoglycaemia. • Some of the neuroglycopenic symptoms of hypoglycaemia are thought to be subjective impressions of impaired cognitive function: these impressions are fully supported by the results of objective cognitive testing. 42 Hypoglycaemia in Clinical Diabetes • Mild hypoglycaemia may induce a state of anxious tension, unhappiness and low energy, and even irritability and anger. Thus hypoglycaemia importantly touches the emotions as well as inducing bodily symptoms and affecting mental performance.

REFERENCES

Amiel, S.A., Pottinger, R.C., Archibald, H.R. et al. (1991) Effect of antecedent glucose control on cerebral function during hypoglycemia. Diabetes Care, 14, 109–118. Blackman, J.D., Towle, V.L., Sturis, J., Lewis, G.F., Spire, J.-P. & Polonsky, K.S. (1992) Hypoglycemic thresholds for cognitive dysfunction in IDDM. Diabetes, 41, 392–399. Bremer, J.P., Baron, M., Peters, H. et al. (2006) Hormonal, subjective and neurocognitive responses to brief hypoglycemia in postmenopausal women and age-matched men with type 2 diabetes mellitus. Metabolism: Clinical and Experimental, 55, 331–338. Clarke, W.L., Cox, D.J., Gonder-Frederick, L.A., Julian, D., Schlundt, D. & Polonsky, W. (1997) The relationship between nonroutine use of insulin, food, and exercise and the occurrence of hypoglycemia in adults with IDDM and varying degrees of hypoglycemia awareness and metabolic control. Diabetes Education, 23, 55–58. Cox, D.J., Clarke, W.L., Gonder-Frederick, L. et al. (1985) Accuracy of perceiving blood glucose in IDDM. Diabetes Care, 8, 529–536. Cox, D.J., Irvine, A., Gonder-Frederick, L., Nowacek, G. & Butterfield, J. (1987) Fear of hypoglycemia: quantification, validation and utilization. Diabetes Care, 10, 617–621. Cox, D.J., Gonder-Frederick, L., Antoun, B., Cryer, P.E. & Clarke, W.L. (1993a) Perceived symptoms in the recognition of hypoglycemia. Diabetes Care, 16, 519–527. Cox, D.J., Gonder-Frederick, L. & Clarke, W. (1993b) Driving decrements in type 1 diabetes during moder- ate hypoglycemia. Diabetes, 42, 239–243. Cox, D.J., Gonder-Frederick, L.A., Kovatchev, B.P., Julian, D.M. & Clarke, W.L. (2000) Progressive hypoglycemia’s impact on driving simulation performance. Occurrence, awareness and correction. Dia- betes Care, 23, 163–170. Cox, D.J., Gonder-Frederick, L., Polonsky, W., Schlundt, D., Kovatchev, B. & Clarke, W. (2001) Blood glucose awareness training (BGAT-2): long-term benefits. Diabetes Care, 24, 637–642. Cox, D.L., Kovatchev, B., Koev, D. et al. (2004) Hypoglycemia anticipation, awareness and treatment training (HAAT) reduces occurrence of severe hypoglycemia among adults with type 1 diabetes. Inter- national Journal of Behavioral Medicine, 11, 212–218. Cox, D.J., Kovatchev, B.P., Anderson, S.M., Clarke, W.L. & Gonder-Frederick, L.A. (2010) Type 1 diabetic drivers with and without a history of recurrent hypoglycemia-related driving mishaps. Physiological and performance differences during euglycemia and the induction of hypoglycemia. Diabetes Care, 33, 2430–2435. Deary, I.J. (1998) The effects of diabetes on cognitive function. Diabetes Annual, 11, 97–118. Deary, I.J., Hepburn, D.A., MacLeod, K.M. & Frier, B.M. (1993) Partitioning the symptoms of hypogly- caemia using multi-sample confirmatory factor analysis. Diabetologia, 36, 771–777. Deary, I.J., Sommerfield, A.J., McAulay, V. & Frier, B.M. (2003) Moderate hypoglycaemia obliterates working memory in humans with and without insulin treated diabetes. Journal of Neurology, Neurosur- gery, and Psychiatry, 74, 277–282. Debrah, T., Sherwin, R.S., Murphy, J. & Kerr, D. (1996) Effect of caffeine on recognition of and physiologi- cal responses to hypoglycaemia in insulin-dependent diabetes. Lancet, 347, 19–24. Draelos, M.T., Jacobson, A.M., Weinger, K. et al. (1995) Cognitive function in patients with insulin- dependent diabetes mellitus during hyperglycemia and hypoglycemia. The American Journal of Medi- cine, 98, 135–144. Driesen, N.R., Cox, D.J., Gonder-Frederick, L. & Clarke, W. (1995) Reaction time impairment in insulin- dependent diabetes: task complexity, blood glucose levels, and individual differences. Neuropsychology, 9, 246–254. Evans, M.L., Pernet, A., Lomas, J., Jones, J. & Amiel, S.A. (2000) Delay in onset of awareness of acute hypoglycemia and of restoration of cognitive performance during recovery. Diabetes Care, 23, 893–897. Chapter 2: Symptoms of Hypoglycaemia and Effects on Mental Performance and Emotions 43

Ewing, F.M.E., Deary, I.J., McCrimmon, R.J., Strachan, M.W.J. & Frier, B.M. (1998) Effect of acute hypoglycemia on visual information processing in adults with type 1 diabetes mellitus. Physiology and Behavior, 64, 653–660. Fletcher, A.A. & Campbell, W.R. (1922) The blood sugar following insulin administration and the symptom complex-hypoglycemia. Journal of Metabolic Research, 2, 637–649. Geddes, J., Warren, R.E., Sommerfield, A.J.et al. (2006) Absence of sexual dimorphism in the symptomatic responses to hypoglycemia in adults with and without type 1 diabetes. Diabetes Care, 29, 1667–1669. Geddes, J., Deary, I.J. & Frier, B.M. (2008) Effects of acute insulin-induced hypoglycaemia on psychomo- tor function: people with type 1 diabetes are less affected than non-diabetic adults. Diabetologia, 51, 1814–1821. Geddes, J., Ratcliff, R., Allerhand, M. et al. (2010) Modeling the effects of hypoglycemia on a two-choice task in adult humans. Neuropsychology, 24, 652–660. Gold, A.E., Deary, I.J., MacLeod, K.M. & Frier, B.M. (1995a) The effect of IQ level on the degree of cognitive deterioration experienced during acute hypoglycemia in normal humans. Intelligence, 20, 267–290. Gold, A.E., MacLeod, K.M., Deary, I.J. & Frier, B.M. (1995b) Hypoglycemia-induced cognitive dysfunction in diabetes mellitus: effect of hypoglycemia unawareness. Physiology and Behavior, 58, 501–511. Gold, A.E., MacLeod, K.M., Frier, B.M. & Deary, I.J. (1995c) Changes in mood during acute hypoglycemia in healthy subjects. Journal of Personality and Social Psychology, 68, 498–504. Gold, A.E., Deary, I.J. & Frier, B.M. (1997a) Hypoglycaemia and non-cognitive aspects of psychological function in insulin-dependent (type 1) diabetes mellitus (IDDM). Diabetic Medicine, 14, 111–118. Gold, A.E., Frier, B.M., MacLeod, K.M. & Deary, I.J. (1997b) A structural equation model for predictors of severe hypoglycaemia in patients with insulin-dependent diabetes mellitus. Diabetic Medicine, 14, 309–315. Gonder-Frederick, L., Cox, D., Driesen, N.R., Ryan, C.M. & Clarke, W. (1994) Individual differences in neurobehavioral disruption during mild and moderate hypoglycemia in adults with IDDM. Diabetes, 43, 1407–1412. Gonder-Frederick, L., Cox, D., Kovatchev, B., Schlundt, D. & Clarke, W. (1997) A biopsychobehavioral model of risk of severe hypoglycemia. Diabetes Care, 20, 661–669. Graveling, A.J., Deary, I.J., Frier, B.M. (In press, 2013). Acute hypoglycemia impairs executive cognitive function in adults with and without type 1 diabetes. Diabetes Care. DOI: 10.2337/dc13-0194 Hardy, K.J., Scase, M.O., Foster, D.H. & Scarpello, J.H. (1995) Effect of short term changes in blood glucose on visual pathway function in insulin dependent diabetes. The British Journal of Ophthalmology, 79, 38–41. Heller, S.R. & Macdonald, I.A. (1996) The measurement of cognitive function during acute hypoglycaemia: experimental limitations and their effects on the study of hypoglycaemia unawareness. Diabetic Medi- cine, 13, 607–615. Henderson, J.N., Allen, K.V., Deary, I.J. & Frier, B.M. (2003) Hypoglycaemia in insulin-treated type 2 diabetes: frequency, symptoms and impaired awareness. Diabetic Medicine, 20, 1016–1021. Hepburn, D.A. (1993) Symptoms of hypoglycaemia. In: B.M. Frier & M. Fisher (eds), Hypoglycaemia and Diabetes: Clinical and Physiological Aspects, pp. 93–103. Edward Arnold, London. Hepburn, D.A., Deary, I.J., Frier, B.M., Patrick, A.W., Quinn, J.D. & Fisher, M. (1991) Symptoms of acute insulin-induced hypoglycemia in humans with and without IDDM. Factor-analysis approach. Diabetes Care, 14, 949–957. Hepburn, D.A., Deary, I.J., MacLeod, K.M. & Frier, B.M. (1994) Structural equation modeling of symp- toms, awareness and fear of hypoglycemia, and personality in patients with insulin-treated diabetes. Diabetes Care, 17, 1273–1280. Hermanns, N., Kubiak, T., Kulzer, B. & Haak, T. (2003) Emotional changes during experimentally-induced hypoglycaemia in type 1 diabetes. Biological Psychology, 63, 15–44. Hoffman, R.G., Speelman, D.J., Hinnen, D.A., Conley, K.L., Guthrie, R.A. & Knapp, R.K. (1989) Changes in cortical functioning with acute hypoglycemia and hyperglycemia in type 1 diabetes. Diabetes Care, 12, 193–197. Holmes, C.S. (1987) Metabolic control and auditory information processing at altered glucose levels in insulin dependent diabetes. Brain and Cognition, 6, 161–174. Inkster, B. & Frier, B.M. (2012) The effects of acute hypoglycaemia on cognitive function in type 1 dia- betes. British Journal of Diabetes and Vascular Disease, 12, 221–226. 44 Hypoglycaemia in Clinical Diabetes

Irvine, A.A., Cox, D. & Gonder-Frederick, L. (1992) Fear of hypoglycemia: relationship to physical and psychological symptoms in patients with insulin-dependent diabetes mellitus. Health Psychology, 11, 135–138. Jaap, A.J., Jones, G.C., McCrimmon, R.J., Deary, I.J. & Frier, B.M. (1998) Perceived symptoms of hypogly- caemia in elderly type 2 diabetic patients treated with insulin. Diabetic Medicine, 15, 398–401. Jones, T.W., McCarthy, G., Tamborlane, W.V. et al. (1990) Mild hypoglycemia and impairment of brainstem and cortical evoked potentials in healthy subjects. Diabetes, 39, 1550–1555. Lindgren, M., Eckert, B., Stenberg, G. & Agardh, C.-D. (1996) Restitution of neurophysiological functions, performance, and subjective symptoms after moderate insulin-induced hypoglycaemia in non-diabetic men. Diabetic Medicine, 13, 218–225. Lobmann, R., Smid, H.G.O.M., Pottag, G., Wagner, K., Heinze, H.J. & Lehnert, H. (2000) Impairment and recovery of elementary cognitive function induced by hypoglycemia in type-1 diabetic patients and healthy controls. The Journal of Clinical Endocrinology and Metabolism, 85, 2758–2766. Macfarlane, P.I. & Smith, C.S. (1988) Perceptions of hypoglycaemia in childhood diabetes mellitus: a questionnaire study. Practical Diabetes, 5, 56–58. Macfarlane, P.I., Walters, M., Stutchfield, P. & Smith, C.S. (1989) A prospective study of symptomatic hypoglycaemia in childhood diabetes. Diabetic Medicine, 6, 627–630. Maran, A., Lomas, J., Macdonald, I.A. & Amiel, S.A. (1995) Lack of preservation of higher brain function during hypoglycaemia in patients with intensively-treated IDDM. Diabetologia, 38, 1412–1418. Massaro, D.W. (1993) Information processing models: microscopes of the mind. Annual Review of Psychol- ogy, 44, 383–425. McAulay, V., Deary, I.J. & Frier, B.M. (2001a) Symptoms of hypoglycaemia in people with diabetes. Diabetic Medicine, 18, 690–705. McAulay, V., Deary, I.J., Ferguson, S.C. & Frier, B.M. (2001b) Acute hypoglycemia in humans causes attentional dysfunction while nonverbal intelligence is preserved. Diabetes Care, 24, 1745–1750. McAulay, V., Deary, I.J., Sommerfield, A.J. & Frier, B.M. (2005) Attentional functioning is impaired during acute hypoglycaemia in people with type 1 diabetes. Diabetic Medicine, 23, 26–31. McCrimmon, R.J., Gold, A.E., Deary, I.J., Kelnar, C.J.H. & Frier, B.M. (1995) Symptoms of hypoglycemia in children with IDDM. Diabetes Care, 18, 858–861. McCrimmon, R.J., Deary, I.J., Huntly, B.J.P., MacLeod, K.J. & Frier, B.M. (1996) Visual information processing during controlled hypoglycaemia in humans. Brain, 119, 1277–1287. McCrimmon, R.J., Deary, I.J. & Frier, B.M. (1997) Auditory information processing during acute insulin induced hypoglycaemia in non-diabetic human subjects. Neuropsychologia, 35, 1547–1553. McCrimmon, R.J., Deary, I.J. & Frier, B.M. (1999a) Appraisal of mood and personality during hypogly- caemia. Physiology and Behavior, 67, 27–33. McCrimmon, R.J., Ewing, F.M.E., Frier, B.M. & Deary, I.J. (1999b) Anger-state during acute insulin- induced hypoglycaemia. Physiology and Behavior, 67, 35–39. McCrimmon, R.J., Deary, I.J., Gold, A.E. et al. (2003) Symptoms reported during experimental hypogly- caemia: effect of method of induction of hypoglycaemia and of diabetes per se. Diabetic Medicine, 20, 507–509. Merbis, M.A.E., Snoek, F.J., Kanc, K. & Heine, R.J. (1996) Hypoglycaemia induces emotional disruption. Patient Education and Counseling, 29, 117–122. Mutch, W.J. & Dingwall-Fordyce, I. (1985) Is it a hypo? Knowledge of symptoms of hypoglycaemia in elderly diabetic patients. Diabetic Medicine, 2, 54–56. Pennebaker, J.W., Cox, D.J., Gonder-Frederick, L., Wunsch, M.G., Evans, W.S. & Pohl, S. (1981) Physical symptoms related to blood glucose in insulin-dependent diabetics. Psychosomatic Medicine, 43, 489–500. Polonsky, W.H., Davis, C.L., Jacobson, A.M. & Anderson, B.J. (1992) Correlates of hypoglycemic fear in type 1 and type 2 diabetes mellitus. Health Psychology, 11, 199–202. Pramming, S., Thorsteinsson, B., Theilgaard, A., Pinner, E.M. & Binder, C. (1986) Cognitive function during hypoglycaemia in type 1 diabetes mellitus. British Medical Journal, 292, 647–650. Ross, L.A., McCrimmon, R.J., Frier, B.M., Kelnar, C.J.H. & Deary, I.J. (1998) Hypoglycaemic symptoms reported by children with type 1 diabetes mellitus and by their parents. Diabetic Medicine, 15, 836–843. Russell, P.N. & Rix-Trot, H.M. (1975) An exploratory study of some behavioural consequences of insulin induced hypoglycaemia. The New Zealand Medical Journal, 81, 337–340. Chapter 2: Symptoms of Hypoglycaemia and Effects on Mental Performance and Emotions 45

Ryan, C.M., Dulay, D., Suprasongsin, C. & Becker, D.J. (2002) Detection of symptoms by adolescents and young adults with type 1 diabetes during experimental induction of mild hypoglycemia: role of hormonal and psychological variables. Diabetes Care, 25, 852–858. Schachinger, H., Hegar, K., Hermanns, N. et al. (2005) Randomized controlled clinical trial of Blood Glucose Awareness Training (BGAT III) in Switzerland and Germany. Journal of Behavioral Medicine, 28, 587–594. Sommerfield, A.J., Deary, I.J., McAulay, V. & Frier, B.M. (2003a) Moderate hypoglycemia impairs multiple memory functions in healthy adults. Neuropsychology, 17, 125–132. Sommerfield, A.J., Deary, I.J., McAulay, V. & Frier, B.M. (2003b) Short-term, delayed, and working memory are impaired during hypoglycemia in individuals with type 1 diabetes. Diabetes Care, 26, 390–396. Strachan, M.W.J., Deary, I.J., Ewing, F.M.E., Ferguson, S.C., Young, M.J. & Frier, B.M. (2001) Acute hypoglycemia impairs the functioning of the central but not peripheral nervous system. Physiology & Behavior, 72, 83–92. Strachan, M.W.J., Ewing, F.M.E., Frier, B.M., McCrimmon, R.J. & Deary, I.J. (2003) Effects of acute hypoglycaemia on auditory information processing in adults with type 1 diabetes. Diabetologia, 46, 97–105. Tallroth, G., Lindgren, M., Stenberg, G., Rosen, I. & Agardh, C.-D. (1990) Neurophysiological changes during insulin-induced hypoglycaemia and in the recovery period following glucose infusion in type 1 (insulin-dependent) diabetes mellitus and in normal man. Diabetologia, 33, 319–323. ter Braak, E.W.M.T., Appleman, A.M.M.E., van de Laak, M.E., Stolk, R.P., van Haeften, T.W. & Erkelens, D.W. (2000) Clinical characteristics of type 1 diabetic patients with and without severe hypoglycemia. Diabetes Care, 23, 1467–1471. Towler, D.A., Havlin, C.E., Craft, S. & Cryer, P. (1993) Mechanism of awareness of hypoglycemia: percep- tion of neurogenic (predominantly cholinergic) rather than neuroglycopenic symptoms. Diabetes, 42, 1791–1798. UK Hypoglycaemia Study Group (2007) Risks of hypoglycaemia in types 1 and 2 diabetes: effects of treatment modalities and their duration. Diabetologia, 50, 1140–1147. Warren, R.E. & Frier, B.M. (2005) Hypoglycaemia and cognitive function. Diabetes, Obesity and Metabo- lism, 7, 493–503. Warren, R.E., Zammitt, N.N., Deary, I.J. & Frier, B.M. (2007) The effects of acute hypoglycaemia on memory acquisition and recall and prospective memory in type 1 diabetes. Diabetologia, 50, 178–185. Weinger, K., Jacobson, A.M., Draelos, M.T., Finkelstein, D.M. & Simonson, D.C. (1995) Blood glucose estimation and symptoms during hyperglycemia and hypoglycemia in patients with insulin-dependent diabetes mellitus. The American Journal of Medicine, 98, 22–31. Wild, D., van Maltzahn, R., Brohan, E., Christensen, T., Clauson, P. & Gonder-Frederick, L. (2007) A critical review of the literature on fear of hypoglycaemia in diabetes: implications for diabetes manage- ment and patient education. Patient Education and Counselling, 26, 10–15. Wilder, J. (1943) Psychological problems in hypoglycemia. The American Journal of Digestive Diseases, 10, 428–435. Wirsen, A., Tallroth, G., Lindgren, M. & Agardh, C.-D. (1992) Neuropsychological performance differs between type 1 diabetic and normal men during insulin-induced hypoglycaemia. Diabetic Medicine, 9, 156–165. Wright, R., Frier, B.M. & Deary, I.J. (2009) Effects of acute insulin-induced hypoglycemia on spatial abilities in adults with type 1 diabetes. Diabetes Care, 32, 1503–1506. Zammitt, N., Warren, R.E., Deary, I.J. & Frier, B.M. (2008) Delayed recovery of cognitive function fol- lowing hypoglycemia in adults with type 1 diabetes: effects of impaired awareness of hypoglycemia. Diabetes, 3, 732–736. Zammitt, N.N., Streftaris, G., Gibson, G.J., Deary, I.J. & Frier, B.M. (2011) Modeling the consistency of hypoglycemic symptoms: high variability in diabetes. Diabetes Technology and Therapeutics, 13, 571–578. Ziegler, D., Hubinger, A., Muhlen, H. & Gries, F.A. (1992) Effects of previous glycaemic control on the onset and magnitude of cognitive dysfunction during hypoglycaemia in type 1 (insulin-dependent) dia- betic patients. Diabetologia, 35, 828–834. 3 Counterregulatory Deficiencies in Diabetes Rory J. McCrimmon University of Dundee, Dundee, UK

INTRODUCTION

It is now well recognised that good glycaemic control in both type 1 diabetes (The Diabetes Control and Complications Trial Research Group 1993) and type 2 diabetes (UK Prospec- tive Diabetes Study (UKPDS) Group 1998a) reduces both the incidence and progression of long-term microvascular complications. However, in clinical practice it has been difficult to achieve near-normalisation of glucose levels in people with diabetes. In the Diabetes Control and Complications Trial (DCCT) for instance, <5% of subjects in the intensively treated arm maintained normal haemoglobin A1c levels (<6.1%; 43 mmol/mol), and the median HbA1c was 7.2% (55 mmol/mol), which is significantly above the non-diabetic range (The Diabetes Control and Complications Trial Research Group 1993). It became apparent in the DCCT that euglycaemia could not be safely achieved in these individuals with type 1 diabetes because of a markedly increased risk of severe hypoglycaemia. Sub- jects who underwent intensive insulin therapy overall had a 3-fold increased risk of severe hypoglycaemia, and this risk increased the stricter the glycaemic control (Figure 3.1) (The Diabetes Control and Complications Trial Research Group 1997). Similarly in the United Kingdom Prospective Diabetes Study (UKPDS), hypoglycaemia became limiting in the intensive therapy arm over time, and median haemoglobin A1c levels rose to 8.1% (65 mmol/ mol), a point at which microvascular complications are more liable to develop (UK Pro- spective Diabetes Study (UKPDS) Group 1998b). In both type 1 diabetes treated with insulin or type 2 diabetes treated with sulfonylureas and/or insulin, aggressive management was associated with a significantly increased risk of severe hypoglycaemia. It is now appreciated that hypoglycaemia in diabetes results from more than just the inadequacies of modern insulin replacement therapy. Despite the introduction of insulin analogues, continuous glucose sensors and structured education programmes, the overall rates of severe hypoglycaemia have remained unchanged over the last 20 years (Bulsara et al. 2004; Heller and Kerr 2007). People with type 1 diabetes and longer duration type 2 diabetes have clear defects in their physiological responses to hypoglycaemia. These deficiencies in the counterregulatory response markedly increase the risk of severe hypogly- caemia and ensure that this limits the ability to achieve good glycaemic control. Figure 3.2 illustrates the profound disturbance of glucose homeostasis observed in individuals with

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. Chapter 3: Counterregulatory Deficiencies in Diabetes 47

Severe Hypoglycaemia Microvascular complications 120

100

80

60

40

20

0 4.5 5.5 6.5 7.5 8.5 9.5 10.5 Rate of SHG (per 100-patient years) of SHG (per 100-patient Rate Mean HbA1c (%)

Figure 3.1 The Diabetes Care and Complications Trial showed that as HbA1c was lowered towards the non-diabetic range the incidence and progression of microvascular disease was much lower (squares). However, this was accompanied by a concomitant rise in the risk of severe hypoglycaemia (circles). (Source: Adapted from the Diabetes Control and Complications Trial Research Group 1993. Reproduced with permission from John Wiley & Sons Ltd).

300

200

140 100 70 60

0 12:00a 2:00a 4:00a 6:00a 8:00a 10:00a 12:00p 2:00p 4:00p 6:00p 8:00p 10:00p 12:00a (a) Sensor data (mg/dl) from an individual without diabetes

300

200

140 100 70 60

0 12:00a 2:00a 4:00a 6:00a 8:00a 10:00a 12:00p 2:00p 4:00p 6:00p 8:00p 10:00p 12:00a (b) Sensor data (mg/dl) from an individual with type 1 diabetes

Figure 3.2 Abnormal glucose homeostasis in type 1 diabetes. (a) Sensor data collected using a con- tinuous subcutaneous glucose monitor in an individual without diabetes over a 7-day period. Each line represents readings from a single day and glucose levels are ordinarily maintained within a narrow physiological range (shaded bar). (b) The same technique applied to an individual with type 1 diabetes treated using an insulin pump showing that glucose homeostasis is profoundly affected with prolonged periods of both hyper- and hypoglycaemia. 48 Hypoglycaemia in Clinical Diabetes type 1 diabetes despite the technological advances in insulin replacement therapy. In this chapter the nature of this counterregulatory deficiency in diabetes and the impact this has on hypoglycaemia risk are discussed.

NORMAL GLUCOSE COUNTERREGULATION

Under most circumstances the brain relies on glucose to fuel its energy needs. Other organs can use alternate fuels such as fatty acids to provide energy substrates, but despite represent- ing about 1/40th of body weight the brain consumes 1/4 of the energy supplied by glucose. At the same time the brain has no substantial energy stores other than a small amount of glycogen and therefore is reliant on a steady supply of glucose via the circulation. Glucose levels in the brain are significantly lower than in the systemic circulation (Gru- etter et al. 1992) and range between 0.8 and 1.5 mmol/l (Abi-Saab et al. 2002) (Figure 3.3). Furthermore, the transport of glucose into the brain across the blood–brain barrier (BBB) is through facilitative diffusion, which means it is dependent on the concentration of glucose in the arteries. When blood glucose concentrations are in the normal range, suf- ficient glucose is transported across the BBB to meet the energy needs of the brain, but when systemic blood glucose levels decline below normal (around 3.9–6.1 mmol/l), glucose levels in the brain fall rapidly and transport of glucose becomes limiting to normal brain function and therefore survival. Glucose levels in the brain under hypoglycaemic conditions reach very low levels (Abi-Saab et al. 2002), which is why, given the critical importance of glucose to normal brain functioning, humans have evolved complex physiological mechanisms that prevent or correct hypoglycaemia rapidly (see Chapter 1). These responses to hypoglycaemia are triggered when low glucose is sensed by special- ised glucose-sensing neurones that are found both in the brain and periphery (McCrimmon and Sherwin 2010). Fluctuations in peripheral glucose are detected by glucose sensing

14

12

10

8 Plasma Glucose 6 Brain ECF Gucose

Glucose (mmol/l) 4

2

0 Baseline Hyperglycaemia Hypoglycaemia

Figure 3.3 Brain extra-cellular fluid (ECF) levels are significantly lower than plasma glucose levels. Using microdialysis in human subjects undergoing surgery for intractable epilepsy, Abi-Saab and coll­ eagues demonstrated that under basal, hyperglycaemic and hypoglycaemic conditions, brain ECF glucose levels (light purple bars) were markedly lower than levels measured simultaneously in plasma (dark purple bars) (Source: Abi-Saab WM 2002. Reproduced with permission from Nature Publishing Group). Chapter 3: Counterregulatory Deficiencies in Diabetes 49

Table 3.1 Hierarchy of glucose counterregulatory responses to acute hypoglycaemia

Counterregulatory Change during Glycaemic Major physiological response hypoglycaemia threshold action (mmol/l)

Insulin ↓ 4.4–4.7 ↑Ra (↓Rd insulin sensitive tissues) Glucagon ↑ 3.6–3.9 ↑Ra (Liver) Adrenaline ↑ 3.6–3.9 ↑Ra (liver and kidney) ↓Rd (Epinephrine) Cortisol / Growth ↑ 3.6–3.9 ↑Ra ↓Rd (chronic) Hormone Symptoms ↑ 2.8–3.0 ↓ Food Intake Cognition ↓ <2.8 Impaired behavioural responses

Ra and Rd, rate of appearance and disappearance, respectively, of glucose in the systemic circulation. neurones in the oral cavity, gut, portal/mesenteric vein (PMV) and carotid body. PMV neurones detect changes in blood glucose prior to entry into the liver from the gut. This information is then relayed through the vagus nerve and spinal cord to the hindbrain and from there to higher brain centres such as the hypothalamus. The hypothalamus (and other forebrain centres) also detects a falling glucose directly and is probably the most important glucose-sensing region. These specialised neurones detect hypoglycaemia using mecha- nisms that appear to be identical to the pancreatic β-cell in that glucokinase, AMP-activated protein kinase and the ATP-sensitive potassium channel are critical steps in the translation of a change in extra-cellular glucose to a change in neuronal activation (McCrimmon and Sherwin 2010). The sequence of physiological responses that develop following activation of glucose- sensing neurones by hypoglycaemia are called the counterregulatory response and are detailed in Table 3.1, along with the glucose threshold at which they are usually initiated. The threshold for activation was established in many studies undertaken some years ago using the hyperinsulinaemic hypoglycaemic clamp. With this technique, individuals are infused intravenously with insulin at a relatively high and constant dose and at the same time receive intravenous dextrose to prevent severe hypoglycaemia. The dextrose infusion is then adjusted according to bedside ‘arterialised’ blood glucose readings obtained from a cannula inserted into a hand vein. The hand is heated to create arterio-venous shunting so that the glucose concentration approximately measures the blood glucose level reaching the brain. The aim is to achieve a series of predetermined levels of hypoglycaemia. During each glycaemic plateau, hormonal responses are measured and the glucose level at which there is both a significant increase in counterregulatory hormone or symptoms and deterio- ration in cognitive function can be estimated. This approach has shown that in healthy individuals, insulin secretion from pancreatic β-cells is inhibited as blood glucose levels fall to below approximately 4.4–4.7 mmol/l. This leads to the loss of a tonic inhibitory effect of insulin (as well as the inhibitory neu- rotransmitter GABA and zinc, which are co-secreted with insulin) on pancreatic α-cell function (Wendt et al. 2004; Zhou et al. 2007), rapidly increasing glucagon secretion (Ishihara et al. 2003). Low insulin levels also favour increased hepatic glucose production (gluconeogenesis) as well as reduced uptake of glucose in insulin-sensitive tissue such as 50 Hypoglycaemia in Clinical Diabetes muscle, while increased glucagon secretion stimulates liver glycogen breakdown (hepatic glycogenolysis) as well as gluconeogenesis. If glucose levels fall further (∼3.6–3.9 mmol/l), adrenaline and noradrenaline are released both from the adrenal glands and sympathetic nerve terminals and this exerts a more tonic influence on glucose levels by further suppressing insulin secretion, increasing glucagon secretion and decreasing peripheral glucose utilisation in muscle and increasing lipolysis in fat (Avogaro et al. 1993). Additional hormones released during hypoglycaemia include growth hormone (GH) and cortisol, secreted when blood glucose falls below ∼3.6–3.9 mmol/l glucose (Schwartz et al. 1987; Mitrakou et al. 1991). These act to stimulate lipolysis, ketogenesis and glu­ coneogenesis. However, the responses initiated by GH and cortisol are slower in onset (several hours) and are perhaps more important as initiators of the adaptive response to hypoglycaemia. Overall, these counterregulatory responses provide a number of ‘fail safes’ ensuring that in healthy non-diabetic individuals hypoglycaemia is rare, but occurs occasionally during starvation or ultra-endurance sports. Furthermore, there is a clear hierarchy in the relative importance of each counterregulatory hormone as discussed below. Briefly, suppression of insulin secretion and release of glucagon from the alpha-cell are critical to the ability to respond to and recover acutely from hypoglycaemia. Hypoglycaemia will not develop in the absence of relative hyperinsulinaemia and the loss of glucagon secretion markedly impairs our ability to respond to and recover from hypoglycaemia. In the absence of glu- cagon, the catecholaminergic response to hypoglycaemia becomes critical leading to pro- found and prolonged hypoglycaemia even in the fasted state. GH and cortisol appear less critical to acute hypoglycaemia or even the prevention of hypoglycaemia after an overnight fast (Boyle and Cryer 1991).

AGE AND GLUCOSE COUNTERREGULATION

In children with type 1 diabetes, the glucagon response to hypoglycaemia is markedly attenuated compared to non-diabetic individuals but this is usually compensated by vigor- ous secretion of other counterregulatory hormones, particularly adrenaline, with the peak adrenaline responses being almost 2-fold higher than in adults (Amiel et al. 1987). The total sympathoadrenal responses to hypoglycaemia are also influenced by pubertal stage (Ross et al. 2005). Furthermore, it appears that the glycaemic thresholds for the secretion of adrenaline and growth hormone are set at a higher blood glucose level in non-diabetic children compared to adults (Jones et al. 1991). In children with type 1 diabetes, the secretion of adrenaline in response to hypoglycaemia commences at an even higher glucose level, although this may reflect adaptation to recent blood glucose control. In children who have markedly elevated HbA1c values, there is a shift of the blood glucose threshold to an even higher level for the release of counterregulatory hormones. Advanced age, in otherwise healthy people, does not appear to diminish or delay coun- terregulatory responses to hypoglycaemia (Brierley et al. 1995), although the magnitude of responses of adrenaline and glucagon is lower at milder hypoglycaemic levels (around 3.4 mmol/l) compared to younger non-diabetic subjects but is much more comparable with a more profound hypoglycaemic stimulus (2.8 mmol/l) (Ortiz-Alonso et al. 1994) (see Chapter 12). The magnitude of counterregulatory responses to low blood glucose levels Chapter 3: Counterregulatory Deficiencies in Diabetes 51 following preceding hypoglycaemia also appears to depend on the gender of experimental subjects, with men having lower responses compared to women (Davis et al. 2000a).

DEFECTIVE HORMONAL GLUCOSE COUNTERREGULATION

The counterregulatory response to hypoglycaemia is impaired in all individuals with estab- lished (C-peptide negative) type 1 diabetes (Mokan et al. 1994) and in many individuals with longer duration type 2 diabetes (Segel et al. 2002). The most important defects in counterregulation are listed below. The defects in glucose counterregulation are not ‘all or nothing’ and are influenced by a number of factors. Some of the defects maybe reversible.

• absolute or relative therapeutic insulin excess; • absolute insulin deficiency (C-peptide negative) leading to failure to suppress insulin during hypoglycaemia; • hypoglycaemia-specific loss of glucagon counterregulatory response; • reduced magnitude and raised threshold (lower blood glucose required) for adrenaline release following repeated hypoglycaemia; • reduced symptomatic awareness of hypoglycaemia.

Taken together these defects ensure that hypoglycaemia is far more likely to develop in people with type 1 and longer duration type 2 diabetes. Figure 3.4 illustrates how the ‘window of opportunity’, during which many people with type 1 diabetes may be able to recognise and treat incipient hypoglycaemia adequately, is significantly shorter than that for non-diabetic individuals who have an intact counterregulatory response. Some of the mechanisms that contribute to each of these defects are discussed in the following sections.

Glucose (mmol/l)

4

Release of hormones

3 Warning Symptoms Cognitive impairment

2

Individuals without Individuals with type 1 diabetes diabetes

Figure 3.4 The ‘window of opportunity’ experienced by an individual that allows them to take the appropriate corrective action on recognition of incipient hypoglycaemia and before cognitive impairment develops is significantly shorter in type 1 diabetes partly because of profound defects in the counterregu- latory response during acute hypoglycaemia. 52 Hypoglycaemia in Clinical Diabetes

Insulin The first important counterregulatory defect in diabetes is unregulated hyperinsulinaemia caused either by injected insulin or by ingestion of oral hypoglycaemic agents that stimulate insulin secretion by a non-glucose dependent mechanism (e.g. sulfonylureas). Hyperinsuli- naemia provides a continuous glucose-lowering stimulus by increasing glucose uptake into muscle, fat and liver. Insulin limits gluconeogenesis by repressing gluconeogenic genes, such as PEPCK and G6Pase, and also inhibits lipolysis, thus preventing the release of alternate fuel stores. In addition, insulin suppresses α-cell glucagon secretion either through direct action (Cooperberg and Cryer 2010) or indirectly by central mechanisms (Paranjape et al. 2010).

Glucagon Independent of hyperinsulinaemia, glucagon secretion in response to hypoglycaemia is markedly diminished with increasing duration of diabetes for poorly understood reasons and this occurs in both type 1 (Gerich et al. 1973) and advanced type 2 diabetes (Segel et al. 2002). Loss of a glucagon response to hypoglycaemia means the action of insulin on the liver is unopposed, reducing the drive to hepatic glycogenolysis and gluconeogenesis. This results in deeper and more prolonged hypoglycaemia (Figure 3.5). Interestingly, glu- cagon responses to other stimuli such as amino acids and exercise persist in type 1 diabetes, and adrenaline-induced glucagon secretion is enhanced, indicating that α-cells are still functional and that this is a defect that is specific to hypoglycaemia (Gerich et al. 1973).

* * 200 500 180 450 160 400 140 350 120 300 100 250 80 200 60 150 40 100 20 50 0 0 Basal Hypoglycaemia Basal Hypoglycaemia

(a) Glucose (mg/dl) (b) Glucagon (pg/ml)

Figure 3.5 Impaired glucagon secretory response to hypoglycaemia in type 1 diabetes. Hypoglycae- mia was induced using a bolus of insulin in 6 individuals with type 1 diabetes and 15 control subjects. (a) Despite the presence of fasting hyperglycaemia, both type 1 diabetic and non-diabetic individuals reached the same glucose nadir at 60 minutes after injection. (b) However, while hypoglycaemia induced an approximate 4-fold increase in glucagon secretion in the non-diabetic subjects, no significant rise in glucagon occurred in those with type 1 diabetes. Non-diabetic subjects are shown by light purple bars and type 1 diabetic subjects with dark purple bars. Data are presented as mean and standard error of the mean (SEM). *, P < 0.05. (Data from Gerich JE 1973). Chapter 3: Counterregulatory Deficiencies in Diabetes 53

It is thought that impaired glucagon secretion develops because of a defect in the islet fol- lowing autoimmune β-cell destruction. This is because loss of glucagon secretion closely follows a decline in C-peptide levels (Fukuda et al. 1988). This has led to the hypothesis that the inability to suppress insulin during hypoglycaemia in type 1 diabetes prevents glucagon release – the ‘switch-off’ hypothesis. Two studies, using separate methods to reduce intra-islet insulin concentration during hypoglycaemia (diazoxide and somatostatin), observed reduced glucagon responses on the induction of hypoglycaemia (Gosmanov et al. 2005; Raju and Cryer 2005), consistent with this hypothesis. However, a number of additional factors such as zinc can also restrain glucagon release during hypoglycaemia (Zhou et al. 2007), and glucagon can also be regulated by central CNS mechanisms, includ- ing neurones located in the hypothalamus (McCrimmon et al. 2006a; Tong et al. 2007). Overall, most studies suggest that an intra-islet defect plays the most important role, but the recognition that other mechanisms can influence glucagon release might offer up the opportunity for future therapeutic intervention. It is important to note that differences between species in terms of islet substructure and innervation mean that data derived from animal studies may not translate into the human field.

Catecholamines The second major abnormality in type 1 diabetes and advanced type 2 diabetes is an attenu- ated adrenaline response to hypoglycaemia. This appears to be reduced in magnitude and activated at a lower glucose concentration (Segel et al. 2002). An attenuated adrenaline response (from the adrenal medulla) to falling blood glucose levels, together with an absent glucagon response, markedly increases the risk of severe hypoglycaemia in individuals with type 1 diabetes. It is unknown how common adrenaline counterregulatory failure occurs in type 1 diabetes, but it appears related both to the duration of diabetes and the prevailing quality of glycaemic control. These defects are relatively common; one study reported that around 55% of people with type 1 diabetes of long duration are affected (Mokan et al. 1994). The defect in the catecholaminergic response to hypoglycaemia is particularly important because it is closely associated with impaired hypoglycaemia aware- ness, and increases the risk of severe hypoglycaemia by 25-fold (White et al. 1983). Similar to glucagon, the adrenaline response to other stimuli such as exercise persists (Bjorkman et al. 1988), although the response is diminished in poorly controlled diabetes. Importantly, control of hepatic glucose output during hypoglycaemia (as well as during exercise) is shifted toward catecholamines when the glucagon responses to hypoglycaemia are poor (Popp et al. 1982; Bjorkman et al. 1988). This means that maintenance of the catecholamine response to hypoglycaemia is critical in diabetes. As with the pancreatic α-cell, the adrenal glands are regulated by the autonomic nervous system and this plays a major role in the response of the adrenals to a variety of physiologi- cal and pathophysiological stimuli. However, in contrast to the pancreatic α-cell, the attenuated adrenaline response observed in individuals with type 1 diabetes and longer duration type 2 diabetes is thought to arise primarily through altered sympathetic drive to the adrenals rather than an intrinsic defect in catecholamine release. Following the seminal work of Heller and Cryer (1991), studies in a variety of human and animal models have demonstrated that a single episode of hypoglycaemia is sufficient for defects to be observed in adrenaline (and in glucagon) secretion during a subsequent episode of hypoglycaemia. This phenomenon was subsequently demonstrated in humans with type 1 diabetes (Davis et al. 1992) as well as in the rat (Powell et al. 1993) and mouse (Jacobson et al. 2006) 54 Hypoglycaemia in Clinical Diabetes indicating that this response to hypoglycaemic stress is probably conserved through evolu- tion. Moreover, both the depth (Davis et al. 1997b), duration (Davis et al. 2000b) and frequency (Davis and Shamoon 1991; Davis et al. 1992) of antecedent hypoglycaemia correlates with the magnitude of the ensuing counterregulatory deficit, and the defect resolves over time if the hypoglycaemic stimulus is not repeated (George et al. 1995, 1997). The change in threshold for adrenaline release is usually associated with a reduction in symptom responses to hypoglycaemia, and taken together this clinical condition has been called Hypoglycaemia-Associated Autonomic Failure (HAAF) (Cryer 1992). This name is to some extent misleading as there is no failure of the other components of the autonomic system, and the changes evoked by recurrent hypoglycaemia may be adaptive, not mala­ daptive (see below). However, the model encapsulates the important physiological features and reflects the whole body changes that develop following exposure to repeated hypoglycaemia.

Cortisol and growth hormone Growth hormone (GH) and cortisol are thought to become important glucose-raising hor- mones only after hypoglycaemia has been prolonged for more than one hour. However, defects in cortisol and GH release can cause profound and prolonged hypoglycaemia because of a reduction in hepatic glucose production and, to a lesser extent, by exaggera- tion of insulin-stimulated glucose uptake by muscle. In addition, cortisol provides an additional stimulus to adrenaline production in the adrenal glands and may contribute to reduced adrenaline release following repeated hypoglycaemia. Abnormalities in growth hormone and cortisol secretion in response to hypoglycaemia are characteristic of long-standing type 1 diabetes, affecting up to a quarter of patients who have had diabetes for more than 10 years (Mokan et al. 1994). In rare cases, coexistent endocrine failure as in Addison’s disease or hypopituitarism also predisposes to severe hypoglycaemia. Pituitary failure, although uncommonly associated with type 1 diabetes, occasionally devel- ops in young women as a consequence of ante-partum pituitary infarction. As an intact hypothalamic–pituitary–adrenal axis is important for adequate counterregulation, some investigators have advocated formal assessment of this axis in any individual with brittle diabetes presenting with unexplained, recurrent hypoglycaemia (Hardy et al. 1994).

MECHANISMS OF COUNTERREGULATORY FAILURE

As noted previously, at the onset of type 1 diabetes, hormonal counterregulation is usually normal but within 5 years of diagnosis, glucagon responses to hypoglycaemia become markedly impaired or even absent, although a glucagon response can occur if the hypogly- caemic stimulus is sufficiently profound (Frier et al. 1988). After 10 years of diabetes, patients usually have a suboptimal adrenaline response to compound the absent glucagon response to a fall in blood glucose (Mokan et al. 1994). Thus, patients with type 1 diabetes of long duration are at risk of severe and prolonged neuroglycopenia during hypoglycaemia as a direct consequence of inadequate glucose counterregulation. Although attenuated growth hormone and cortisol responses are less common, they are also late manifestations in terms of diabetes duration (Mokan et al. 1994). Importantly, these defects in glucose counterregulation are not ‘all or nothing’ changes but can be influenced by the prevailing standard of glycaemic control and by the frequency Chapter 3: Counterregulatory Deficiencies in Diabetes 55

Capillary

Glucose KATP UCN3/GC GLUT-1 GLUT-3 Glucose GK GABA Glycogen AMPK Lactate MCT ? Astrocyte Glucose-sensing neurone

Figure 3.6 Potential mechanisms through which abnormal glucose sensing following recurrent hypoglycaemia may develop in type 1 diabetes. Recurrent hypoglycaemia could increase glucose trans- port (GLUT-3) into the neurone or could result in the more rapid metabolism of glucose via glucokinase (GK) or AMP-activated protein kinase (AMPK) producing more ATP. Alternatively, there may be in increase in glycogen stores in astrocytes, releasing more lactate to be used by the neurone during hypoglycaemia. Other factors also released by the astrocyte might be important. Another possibility is that external factors released systemically (e.g. glucocorticoids – GC) or locally such as urocortin 3 (UCN3) or γ-aminobutyric acid (GABA) may suppress the firing of glucose-sensing neurones. Given the complex nature ofthe physiological response to acute and repeated hypoglycaemia it seems likely that no one single mecha- nism is responsible for this alteration in glucose sensing. of hypoglycaemic episodes. Both intensive insulin therapy (Amiel et al. 1988) and anteced- ent hypoglycaemia (Heller and Cryer 1991) increase the glycaemic threshold for adrenaline release (i.e. secretion is stimulated at a lower glucose level). Given that, in the absence of glucagon, adrenaline is critical to the control of hepatic glucose output during hypoglycae- mia, most research has focused on understanding why antecedent hypoglycaemia results in suppressed adrenaline responses to subsequent hypoglycaemia. Various theories have been proposed although as yet there is no convincing evidence in support of any one mechanism (Figure 3.6). Moreover, it remains likely given the complexity of the physio- logical response to hypoglycaemia that a number of mechanisms together contribute to the development of defective counterregulation.

Systemic mediator The systemic mediator theory suggests that a substance is released in response to hypogly- caemia that attenuates subsequent autonomic responses to further episodes of hypoglycae- mia. Cortisol was proposed as the initial candidate, based on two observations: firstly, the attenuating effect of antecedent hypoglycaemia on later autonomic responses is absent in patients with primary adrenocortical failure; secondly, in healthy volunteers, following infusions of cortisol (to supraphysiological levels) during euglycaemia, adrenomedullary adrenaline secretion and muscle sympathetic neural activity were reduced during subse- quent hypoglycaemia (Davis et al. 1996, 1997a). However, this effect of cortisol is lost if infusion rates are reduced to reflect cortisol levels observed during hypoglycaemia (Raju et al. 2003). The defect also cannot be replicated when metyrapone is used to block endog- enous cortisol release during antecedent hypoglycaemia (Goldberg et al. 2006). 56 Hypoglycaemia in Clinical Diabetes

While systemic glucocorticoids do not appear to induce defective counterregulation, at least in short-term models used to study this phenomenon, it is still possible that the related corticotrophin releasing hormone (CRH) family of neuropeptides in the brain may be involved. The CRH family of neuropeptides includes CRH and urocortins (UCN) 1–3, and they act via CRH-R1 and R-2 receptor subtypes. These neuropeptides modulate cellular and whole system responses to a number of physiological and behavioural stressors. Data from transgenic CRH-R1 and CRH-R2 null mouse models implicate CHR/CRH-R1 as stress amplifying and UCN 1–3/CRH-R2 as stress suppressing mechanisms (Timpl et al. 1998; Bale et al. 2000). In the hypothalamus, UCN3 acting through CRH-R2 suppresses (McCrimmon et al. 2006b), while CRH acting through CRH-R1 amplifies (Cheng et al. 2007), the counterregulatory response to hypoglycaemia. Neuronal tracing studies show that a number of forebrain hypoglycaemia-sensing regions are innervated by UCN3 neurons (Zhou et al. 2010). This suggests the presence of a neural network regulating these critical hypoglycaemia-sensing brain regions and a disruption in this stress regulatory mechanism might also contribute to the development of altered glucose sensing following recurrent hypoglycaemia.

Brain fuel transport This mechanism is based on the hypothesis that following antecedent hypoglycaemia, glucose transport from blood into brain tissue is increased. In animals this occurs by increasing GLUT-1 transport across the brain microvasculature (Kumagai et al. 1995). Boyle and colleagues (Boyle et al. 1995) examined this by studying individuals with type 1 diabetes who had very strict glycaemic control. Such individuals develop impaired aware- ness of hypoglycaemia and are at increased risk of seizures and coma. They hypothesised that during hypoglycaemia, subjects whose HbA1c was at near-normal levels would have normal glucose uptake in the brain and consequently that autonomic activation would not occur, resulting in impaired awareness of hypoglycaemia. They did report that there was no significant change in the uptake of glucose in the brain among the participants with type

1 diabetes who had the lowest HbA1c levels. Conversely, glucose uptake in the brain fell in people with less well-controlled type 1 diabetes. The responses of plasma adrenaline and pancreatic polypeptide and the frequency of symptoms of hypoglycaemia were also lowest in the group with the lowest HbA1c values. They concluded that during hypogly- caemia, people with nearly normal HbA1c values have normal glucose uptake in the brain, preserving cerebral metabolism, reducing the responses of counterregulatory hormones, and causing impaired awareness of hypoglycaemia (Boyle et al. 1995). However, these findings occurred after days of prolonged hypoglycaemia, which may not be typical of the clinical situation where hypoglycaemia is usually intermittent and short-lived. More recently, studies using positron emission tomography have found no change in blood-to-brain glucose transport 24 hours after an episode of hypoglycaemia and no dif- ferences between individuals with, and without, hypoglycaemia awareness (Segel et al. 2001; Bingham et al. 2005). It remains possible, however, that there are changes in the transport of alternative cerebral fuels following antecedent hypoglycaemia. It has been shown that there is increased uptake of acetate in intensively-treated subjects with type 1 diabetes (Mason et al. 2006), while medium chain fatty acids prevent cognitive dysfunc- tion during acute hypoglycaemia (Page et al. 2009), although in both of these studies pharmacological levels of alternate fuels were required to demonstrate changes in trans- port. In the clinical situation, therapeutic insulin concentrations might be expected to Chapter 3: Counterregulatory Deficiencies in Diabetes 57 suppress alternate fuel generation in peripheral tissues. Overall, the data supporting an increased uptake of glucose and/or alternate fuel causing defective hormonal counterregu- lation are inconclusive.

Brain metabolism It has been hypothesised that brain metabolism per se is altered following an episode of hypoglycaemia. Most research in this area has focused on the ventromedial nucleus of the hypothalamus (VMH). Glucose deprivation in the VMH (by administration of 2- deoxyglucose) activates the autonomic nervous system and increases glucagon secretion, whereas local perfusion of the area with glucose suppresses these responses during systemic hypoglycaemia (Borg et al. 1995 1997). The mechanisms involved are unknown but may be a consequence of increased glucokinase activity to enhance glucose metabolism in neurones in the region (Levin et al. 2008) or an alteration in AMP-activated protein kinase inducing a switch to increased mitochondrial oxidation of free fatty acids (McCrimmon et al. 2006a). Finally, the brain glycogen supercompensation hypothesis suggests that after a single episode of hypoglycaemia, there is a rebound increase in glycogen formation in brain astrocytes to provide additional substrates (e.g. lactate) for brain metabolism (Choi et al. 2003). However, animal studies where glycogen levels in the brain can be directly measured find no evidence of glycogen supercompensation after 24 hours (Herzog et al. 2006). Additionally, glycogen supercompensation could not be demonstrated in individuals with strictly controlled type 1 diabetes using nuclear magnetic resonance (NMR) spectroscopy (Oz et al. 2012).

Hypoglycaemia tolerance The depth (Davis et al. 1997b), duration (Davis et al. 2000b) and frequency (Davis and Shamoon 1991; Davis et al. 1992) of antecedent hypoglycaemia correlate with the magni- tude of the ensuing counterregulatory deficit, and the defect resolves over periods of up to one week if the hypoglycaemic stimulus is not repeated (George et al. 1995, 1997). This is remarkably similar to the classical description of ischaemic tolerance that develops in response to ischaemic pre-conditioning (for detailed review of mechanisms see Gidday 2006). Pre-conditioning to induce tolerance occurs when a non-damaging stressful stimu- lus is presented to cells, tissues or organisms to promote a transient adaptive response so that injury resulting from subsequent exposure to the harmful stimulus is reduced. Even without pre-conditioning, cells respond to stressful stimuli by mobilising a host of defen- sive and counteractive responses to mitigate injury and prevent death, but adaptive pre- conditioning ensures that the cell is better able to survive the stress. The degree of ischaemic tolerance is also dependent on the intensity, duration and frequency of the pre-conditioning stimulus (Gidday 2006), shows cross-tolerance with other stressors such as exercise (Endres et al. 2003) and temperature (Chopp et al. 1989), and the effect declines over time unless the pre-conditioning stimulus is repeated (Gidday 2006). Thus, the parallels with the impact on glucose counterregulation of antecedent hypoglycaemia are clear and support the concept that this phenomenon represents a model of hypoglycaemia tolerance (or stress habituation). The obvious next question is why should an adaptive response designed to be neuro- protective increase an individual’s susceptibility to severe hypoglycaemia? The likely 58 Hypoglycaemia in Clinical Diabetes

Hypoglycaemia

“Stress” “Starvation”

Pre-conditioning Alternate Fuels

Neuroprotective

Hypoglycaemia Tolerance

Figure 3.7 Hypoglycaemia when experienced by the brain induces a stress response that acts to restore blood glucose levels and increase delivery of available glucose to the brain. Hypothetically, this stimulus will evoke two primary adaptive responses: pre-conditioning a response that ensures the neurone will be better able to withstand future hypoglycaemic stress, and a starvation response that enables the neurone to adapt to use alternate fuels. While this response is adaptive, hypoglycaemia in diabetes develops in the unusual ‘non-physiological’ context of high insulin levels and a specific glucagon defect. This allows significant hypoglycaemia to develop quickly because the counterregulatory response is impaired and access to alternate fuels is suppressed.

answer is that the adaptation in itself does not do this, but in the presence of high insulin levels this adaptation becomes maladaptive. In insulin-induced hypoglycaemia, as opposed to starvation where insulin levels are very low, high insulin suppresses peripheral lipolysis and the delivery of gluconeogenic substrates to the liver and therefore reduces the supply of alternate substrates to the brain. Simultaneously, the combination of unregulated hyper- insulinaemia and a hypoglycaemia-specific defect in glucagon secretion ensures a more rapid and severe hypoglycaemia (Figure 3.7). Thus, in type 1 diabetes, hypoglycaemia tolerance develops in the non-physiological context of hyperinsulinaemia, which prevents appropriate counterregulation and induces severe hypoglycaemia where at some point cell­ ular defence mechanisms will be overwhelmed and cellular death will ensue.

CONCLUSIONS • In non-diabetic individuals, clinically significant hypoglycaemia is an extremely rare event because of effective glucose counterregulation. This includes suppression of endogenous pancreatic insulin secretion, and release of glucagon, catecholamines, cortisol and growth hormone. • The brain is the critical organ for coordination of the physiological responses to low blood glucose levels. • People with diabetes almost inevitably lose their ability to release glucagon in response to a fall in blood glucose within 5 years of diagnosis. After 10 years, a significant proportion of patients also have deficient adrenaline responses and are at increased risk of more protracted hypoglycaemia and neuroglycopenia. • Repeated exposure to hypoglycaemia further attenuates the adrenaline response to hypoglycaemia and is often associated with reduced symptomatic awareness (impaired hypoglycaemia awareness), markedly increasing severe hypoglycaemia risk (by 25-fold). Chapter 3: Counterregulatory Deficiencies in Diabetes 59 • The overall defect in catecholamine release may reflect an adaptive response to the repeated stress of hypoglycaemia, which becomes maladaptive because of inappropri- ately high insulin concentrations that occur during subcutaneous insulin therapy. • Modern treatment of type 1 diabetes, including intensive education and treatment regi- mens utilising new technologies, has failed to eradicate the problem of recurrent hypoglycaemia.

REFERENCES

Abi-Saab, W.M., Maggs, D.G., Jones, T. et al. (2002) Striking differences in glucose and lactate levels between brain extracellular fluid and plasma in conscious human subjects: effects of hyperglycemia and hypoglycemia. Journal of Cerebral Blood Flow and Metabolism, 22, 271–279. Amiel, S.A., Simonson, D.C., Sherwin, R.S., Lauritano, A.A. & Tamborlane, W.V. (1987) Exaggerated epinephrine responses to hypoglycemia in normal and insulin-dependent diabetic children. Journal of Pediatrics, 110, 832–837. Amiel, S.A., Sherwin, R.S., Simonson, D.C. & Tamborlane, W.V. (1988) Effect of intensive insulin therapy on glycemic thresholds for counterregulatory hormone release. Diabetes, 37, 901–907. Avogaro, A., Gnudi, L., Valerio, A. et al. (1993) Effects of different plasma glucose concentrations on lipolytic and ketogenic responsiveness to epinephrine in type I (insulin-dependent) diabetic subjects. The Journal of Clinical Endocrinology and Metabolism, 76, 845–850. Bale, T.L., Contarino, A., Smith, G.W. et al. (2000) Mice deficient for corticotropin-releasing hormone receptor-2 display anxiety-like behaviour and are hypersensitive to stress. Nature Genetics, 24, 410–414. Bingham, E.M., Dunn, J.T., Smith, D. et al. (2005) Differential changes in brain glucose metabolism during hypoglycaemia accompany loss of hypoglycaemia awareness in men with type 1 diabetes mellitus. An [11C]-3-O-methyl-D-glucose PET study. Diabetologia, 48, 2080–2089. Bjorkman, O., Miles, P., Wasserman, D., Lickley, L. & Vranic, M. (1988) Regulation of glucose turnover during exercise in pancreatectomized, totally insulin-deficient dogs. Effects of beta-adrenergic blockade. The Journal of Clinical Investigation, 81, 1759–1767. Borg, M.A., Sherwin, R.S., Borg, W.P., Tamborlane, W.V. & Shulman, G.I. (1997) Local ventromedial hypothalamus glucose perfusion blocks counterregulation during systemic hypoglycemia in awake rats. The Journal of Clinical Investigation, 99, 361–365. Borg, W.P., Sherwin, R.S., During, M.J., Borg, M.A. & Shulman, G.I. (1995) Local ventromedial hypotha- lamus glucopenia triggers counterregulatory hormone release. Diabetes, 44, 180–184. Boyle, P.J. & Cryer, P.E. (1991) Growth hormone, cortisol, or both are involved in defense against, but are not critical to recovery from, hypoglycemia. The American Journal of Physiology, Endocrinology and Metabolism 260, E395–E402. Boyle, P.J., Kempers, S.F., O’Connor, A.M. & Nagy, R.J. (1995) Brain glucose uptake and unawareness of hypoglycemia in patients with insulin-dependent diabetes mellitus. New England Journal of Medicine, 333, 1726–1731. Brierley, E.J., Broughton, D.L., James, O.F. & Alberti, K.G.M.M. (1995) Reduced awareness of hypogly- caemia in the elderly despite an intact counter-regulatory response. Quarterly Journal of Medicine, 88, 439–445. Bulsara, M.K., Holman, C.D., Davis, E.A. & Jones, T.W. (2004) The impact of a decade of changing treat- ment on rates of severe hypoglycemia in a population-based cohort of children with type 1 diabetes. Diabetes Care, 27, 2293–2298. Cheng, H., Zhou, L., Zhu, W. et al. (2007) Type 1 corticotropin-releasing factor receptors in the ventro- medial hypothalamus promote hypoglycemia-induced hormonal counterregulation. American Journal of Physiology. Endocrinology and Metabolism, 293, E705–E712. Choi, I.Y., Seaquist, E.R. & Gruetter, R. (2003) Effect of hypoglycemia on brain glycogen metabolism in vivo. Journal of Neuroscience Research, 72, 25–32. Chopp, M., Chen, H., Ho, K.L. et al. (1989) Transient hyperthermia protects against subsequent forebrain ischemic cell damage in the rat. Neurology, 39, 1396–1398. Cooperberg, B.A. & Cryer, P.E. (2010) Insulin reciprocally regulates glucagon secretion in humans. Diab­ etes, 59, 2936–2940. 60 Hypoglycaemia in Clinical Diabetes

Cryer, P.E. (1992) Iatrogenic hypoglycemia as a cause of hypoglycemia-associated autonomic failure in IDDM. A vicious circle. Diabetes, 41, 255–260. Davis, M.R. & Shamoon, H. (1991) Counterregulatory adaptation to recurrent hypoglycemia in normal humans. The Journal of Clinical Endocrinology and Metabolism, 73, 995–1001. Davis, M.R., Mellman, M. & Shamoon, H. (1992) Further defects in counterregulatory responses induced by recurrent hypoglycemia in IDDM. Diabetes, 41, 1335–1340. Davis, S.N., Shavers, C., Costa, F. & Mosqueda-Garcia, R. (1996) Role of cortisol in the pathogenesis of deficient counterregulation after antecedent hypoglycemia in normal humans. The Journal of Clinical Investigation, 98, 680–691. Davis, S.N., Shavers, C., Davis, B. & Costa, F. (1997a) Prevention of an increase in plasma cortisol during hypoglycemia preserves subsequent counterregulatory responses. The Journal of Clinical Investigation, 100, 429–438. Davis, S.N., Shavers, C., Mosqueda-Garcia, R. & Costa, F. (1997b) Effects of differing antecedent hypogly- cemia on subsequent counterregulation in normal humans. Diabetes, 46, 1328–1335. Davis, S.N., Fowler, S. & Costa, F. (2000a) Hypoglycemic counterregulatory responses differ between men and women with type 1 diabetes. Diabetes, 49, 65–72. Davis, S.N., Mann, S., Galassetti, P. et al. (2000b) Effects of differing durations of antecedent hypoglycemia on counterregulatory responses to subsequent hypoglycemia in normal humans. Diabetes, 49, 1897–1903. Endres, M., Gertz, K., Lindauer, U. et al. (2003) Mechanisms of stroke protection by physical activity. Annals of Neurology, 54, 582–590. Frier, B.M., Fisher, B.M., Gray, C.E. & Beastall, G.H. (1988) Counterregulatory hormonal responses to hypoglycaemia in type 1 (insulin-dependent) diabetes: evidence for diminished hypothalamic-pituitary hormonal secretion. Diabetologia, 31, 421–429. Fukuda, M., Tanaka, A., Tahara, Y. et al. (1988) Correlation between minimal secretory capacity of pan- creatic beta-cells and stability of diabetic control. Diabetes, 37, 81–88. George, E., Harris, N., Bedford, C., Macdonald, I.A., Hardisty, C.A. & Heller, S.R. (1995) Prolonged but partial impairment of the hypoglycaemic physiological response following short-term hypoglycaemia in normal subjects. Diabetologia, 38, 1183–1190. George, E., Marques, J.L., Harris, N.D., Macdonald, I.A., Hardisty, C.A. & Heller, S.R. (1997) Preservation of physiological responses to hypoglycemia 2 days after antecedent hypoglycemia in patients with IDDM. Diabetes Care, 20, 1293–1298. Gerich, J.E., Langlois, M., Noacco, C., Karam, J.H. & Forsham, P.H. (1973) Lack of glucagon response to hypoglycemia in diabetes: evidence for an intrinsic pancreatic alpha cell defect. Science, 182, 171–173. Gidday, J.M. (2006) Cerebral preconditioning and ischaemic tolerance. Nature Reviews. Neuroscience, 7, 437–448. Goldberg, P.A., Weiss, R., McCrimmon, R.J., Hintz, E.V., Dziura, J.D. & Sherwin, R.S. (2006) Antecedent hypercortisolemia is not primarily responsible for generating hypoglycemia-associated autonomic failure. Diabetes, 55, 1121–1126. Gosmanov, N.R., Szoke, E., Israelian, Z. et al. (2005) Role of the decrement in intraislet insulin for the glucagon response to hypoglycemia in humans. Diabetes Care, 28, 1124–1131. Gruetter, R., Novotny, E.J., Boulware, S.D. et al. (1992) Direct measurement of brain glucose concentra- tions in humans by 13C NMR spectroscopy. Proceedings of the National Academy of Sciences of the United States of America, 89, 1109–1112. Hardy, K.J., Burge, M.R., Boyle, P.J., Scarpello, J.H. (1994) A treatable cause of recurrent severe hypogly- cemia. Diabetes Care, 17, 722–724. Heller, S.R. & Cryer, P.E. (1991) Reduced neuroendocrine and symptomatic responses to subsequent hypoglycemia after 1 episode of hypoglycemia in nondiabetic humans. Diabetes, 40, 223–226. Heller, S.R. & Kerr, D. (2007) Risk of hypoglycaemia in types 1 and 2 diabetes: effects of treatment modalities and their duration. Diabetologia, 50, 1140–1147. Herzog, R., McNay, E.C., McCrimmon, R.J. et al. (2006) Effect of acute and recurrent hypoglycemia on regional changes in brain glycogen concentration. Diabetes, 55 (Suppl. 1), 626–P(abstract). Ishihara, H., Maechler, P., Gjinovci, A., Herrera, P.L. & Wollheim, C.B. (2003) Islet beta-cell secretion determines glucagon release from neighbouring alpha-cells. Nature Cell Biology, 5, 330–335. Jacobson, L., Ansari, T. & McGuinness, O.P. (2006) Counterregulatory deficits occur within 24 h of a single hypoglycemic episode in conscious, unrestrained, chronically cannulated mice. American Journal of Physiology. Endocrinology and Metabolism, 290, E678–E684. Chapter 3: Counterregulatory Deficiencies in Diabetes 61

Jones, T.W., Boulware, S.D., Kraemer, D.T., Caprio, S., Sherwin, R.S. & Tamborlane, W.V. (1991) Inde- pendent effects of youth and poor diabetes control on responses to hypoglycemia in children. Diabetes, 40, 358–363. Kumagai, A.K., Kang, Y.S., Boado, R.J. & Pardridge, W.M. (1995) Upregulation of blood-brain barrier GLUT1 glucose transporter protein and mRNA in experimental chronic hypoglycemia. Diabetes, 44, 1399–1404. Levin, B.E., Becker, T.C., Eiki, J., Zhang, B.B. & Dunn-Meynell, A.A. (2008) Ventromedial hypothalamic glucokinase is an important mediator of the counterregulatory response to insulin-induced hypoglycemia. Diabetes, 57, 1371–1379. Mason, G.F., Petersen, K.F., Lebon, V., Rothman, D.L. & Shulman, G.I. (2006) Increased brain monocar- boxylic acid transport and utilization in type 1 diabetes. Diabetes, 55, 929–934. McCrimmon, R.J. & Sherwin, R.S. (2010) Hypoglycemia in type 1 diabetes. Diabetes, 59, 2333–2339. McCrimmon, R.J., Fan, X., Cheng, H. et al. (2006a) Activation of AMP-activated protein kinase within the ventromedial hypothalamus amplifies counterregulatory hormone responses in rats with defective counterregulation. Diabetes, 55, 1755–1760. McCrimmon, R.J., Song, Z., Cheng, H. et al. (2006b) Corticotrophin-releasing factor receptors within the ventromedial hypothalamus regulate hypoglycemia-induced hormonal counterregulation. The Journal of Clinical Investigation, 116, 1723–1730. Mitrakou, A., Ryan, C., Veneman, T. et al. (1991) Hierarchy of glycemic thresholds for counterregulatory hormone secretion, symptoms, and cerebral dysfunction. The American Journal of Physiology, Endo­ crinology and Metabolism, 260, E67–E74. Mokan, M., Mitrakou, A., Veneman, T. et al. (1994) Hypoglycemia unawareness in IDDM. Diabetes Care, 17, 1397–1403. Ortiz-Alonso, F.J., Galecki, A., Herman, W.H., Smith, M.J., Jacquez, J.A., Halter, J.B. (1994) Hypoglycemia counterregulation in elderly humans: relationship to glucose levels. American Journal of Physiology, 267, E497–506. Oz, G., Tesfaye, N., Kumar, A., Deelchand, D.K., Eberly, L.E. & Seaquist, E.R. (2012) Brain glycogen content and metabolism in subjects with type 1 diabetes and hypoglycemia unawareness. Journal of Cerebral Blood Flow and Metabolism, 32, 256–263. Page, K.A., Williamson, A., Yu, N. et al. (2009) Medium-chain fatty acids improve cognitive function in intensively treated type 1 diabetic patients and support in vitro synaptic transmission during acute hypoglycemia. Diabetes, 58, 1237–1244. Paranjape, S.A., Chan, O., Zhu, W. et al. (2010) Influence of insulin in the ventromedial hypothalamus on pancreatic glucagon secretion in vivo. Diabetes, 59, 1521–1527. Popp, D.A., Shah, S.D. & Cryer, P.E. (1982) Role of epinephrine-mediated beta-adrenergic mechanisms in hypoglycemic glucose counterregulation and posthypoglycemic hyperglycemia in insulin-dependent diabetes mellitus. The Journal of Clinical Investigation, 69, 315–326. Powell, A.M., Sherwin, R.S. & Shulman, G.I. (1993) Impaired hormonal responses to hypoglycemia in spontaneously diabetic and recurrently hypoglycemic rats. Reversibility and stimulus specificity of the deficits. The Journal of Clinical Investigation, 92, 2667–2674. Raju, B. & Cryer, P.E. (2005) Loss of the decrement in intraislet insulin plausibly explains loss of the glucagon response to hypoglycemia in insulin-deficient diabetes: documentation of the intraislet insulin hypothesis in humans. Diabetes, 54, 757–764. Raju, B., McGregor, V.P. & Cryer, P.E. (2003) Cortisol elevations comparable to those that occur during hypoglycemia do not cause hypoglycemia-associated autonomic failure. Diabetes, 52, 2083–2089. Ross, L.A., Warren, R.E., Kelnar, C.J. & Frier, B.M. (2005) Pubertal stage and hypoglycaemia counter- regulation in type 1 diabetes. Archives of Disease in Childhood, 90, 190–194. Schwartz, N.S., Clutter, W.E., Shah, S.D. & Cryer, P.E. (1987) Glycemic thresholds for activation of glucose counterregulatory systems are higher than the threshold for symptoms. The Journal of Clinical Investiga­ tion, 79, 777–781. Segel, S.A., Fanelli, C.G., Dence, C.S. et al. (2001) Blood-to-brain glucose transport, cerebral glucose metabolism, and cerebral blood flow are not increased after hypoglycemia. Diabetes, 50, 1911–1917. Segel, S.A., Paramore, D.S. & Cryer, P.E. (2002) Hypoglycemia-associated autonomic failure in advanced type 2 diabetes. Diabetes, 51, 724–733. The Diabetes Control and Complications Trial Research Group (1993) The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. The New England Journal of Medicine, 329, 977–986. 62 Hypoglycaemia in Clinical Diabetes

The Diabetes Control and Complications Trial Research Group (1997) Hypoglycemia in the Diabetes Control and Complications Trial. Diabetes, 46, 271–286. Timpl, P., Spanagel, R., Sillaber, I. et al. (1998) Impaired stress response and reduced anxiety in mice lacking a functional corticotrophin releasing hormone receptor. Nature Genetics, 19, 162–166. Tong, Q., Ye, C., McCrimmon, R.J. et al. (2007) Synaptic glutamate release by ventromedial hypothalamic neurons is part of the neurocircuitry that prevents hypoglycemia. Cell Metabolism, 5, 383–393. UK Prospective Diabetes Study (UKPDS) Group (1998a) Intensive blood-glucose control with sulphony- lureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). Lancet, 352, 837–853. UK Prospective Diabetes Study (UKPDS) Group (1998b) United Kingdom Prospective Diabetes Study 24: a 6-year, randomized, controlled trial comparing sulfonylurea, insulin, and metformin therapy in patients with newly diagnosed type 2 diabetes that could not be controlled with diet therapy. Annals of Internal Medicine, 128, 165–175. Wendt, A., Birnir, B., Buschard, K. et al. (2004) Glucose inhibition of glucagon secretion from rat alpha- cells is mediated by GABA released from neighboring beta-cells. Diabetes, 53, 1038–1045. White, N.H., Skor, D.A., Cryer, P.E., Levandoski, L.A., Bier, D.M. & Santiago, J.V. (1983) Identification of type I diabetic patients at increased risk for hypoglycemia during intensive therapy. The New England Journal of Medicine, 308, 485–491. Zhou, H., Zhang, T., Harmon, J.S., Bryan, J. & Robertson, R.P. (2007) Zinc, not insulin, regulates the rat alpha-cell response to hypoglycemia in vivo. Diabetes, 56, 1107–1112. Zhou, L., Podolsky, N., Sang, Z., Ding, Y., Fan, X., Tong, Q. et al. (2010) The medial amygdalar nucleus: a novel glucose-sensing region that modulates the counterregulatory responses to hypoglycemia. Dia­ betes, 59, 2646–2652. 4 Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes Mark W.J. Strachan Western General Hospital, Edinburgh, UK

INTRODUCTION

Hypoglycaemia was first described in humans in the early years of the 20th century, but did not become firmly established as a pathophysiological entity until the introduction of insulin in 1922. Despite the substantial advances in insulin therapy and blood glucose monitoring that have occurred in the subsequent 90 years, hypoglycaemia remains the most common complication of type 1 diabetes (The Diabetes Control and Complications Trial Research Group 1993) and generates as much anxiety in patients as the threat of advanced diabetic complications, such as blindness or renal failure (Pramming et al. 1991). Few people with type 1 diabetes escape intermittent exposure and, as a result, hypoglycaemia is the principal limiting factor in achieving good glycaemic control (Cryer et al. 2003). The magnitude of the psychological and physical consequences of hypoglycaemia should not be underestimated and is considered in detail in other chapters. In this chapter the frequency of hypoglycaemia is described in people with type 1 diabetes, along with its underlying causes and risk factors.

DEFINITIONS OF HYPOGLYCAEMIA

Any attempt to consider critically the frequency of hypoglycaemia in clinical practice, requires definitive criteria for what constitutes an episode of hypoglycaemia. This poses an immediate difficulty because researchers of the epidemiology of hypoglycaemia have not employed common definitions with shared specifications.

Biochemical definitions of hypoglycaemia At first glance, it would seem sensible to employ a biochemical definition ofhypogly­ caemia, specifying a given blood glucose concentration, below which hypoglycaemia would be deemed to occur. However, it is not possible to provide such a precise bio­ chemical criterion for the diagnosis of hypoglycaemia (Service 1995). As blood glucose concentrations decline, a hierarchy of events takes place, each occurring at individual

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 64 Hypoglycaemia in Clinical Diabetes glycaemic thresholds, commencing with counterregulation (arterialised blood glucose around 3.8 mmol/l), impairment of cognitive functions (between 3.2 and 2.6 mmol/l) and the onset of symptoms and neurophysiological changes (between 3.2–2.4 mmol/l). In clinical practice, however, it is usual for venous or capillary blood glucose levels to be measured, and these are lower than the concomitant estimation of arterialised blood glucose concentrations, which are usually measured in research studies of hypoglycaemia (Heller and Macdonald 1996). In the non-diabetic individual, venous blood glucose concentrations below 3.0 mmol/l may occur following an overnight fast or during the course of a prolonged oral glucose tolerance test (Service 1995). Moreover, as is discussed later, the blood glucose thresholds for the onset of symptoms and counterregulation in people with type 1 diabetes may vary depending on the preceding or prevailing glycaemic control. Thus, patients with poor glycaemic control may experience symptoms of hypoglycaemia at venous plasma glucose concentrations substantially higher than 3.0 mmol/l (Boyle et al. 1988). Patients with pre­ ceding strict glycaemic control may not experience the onset of symptoms of hypoglycae­ mia until venous plasma glucose concentrations have declined to below 2.0 mmol/l (Boyle et al. 1995). On a pragmatic basis, in routine clinical practice, Diabetes UK has recom­ mended that individuals with diabetes should try to ensure that their blood glucose con­ centrations do not fall below 4.0 mmol/l (O’Neill 1997), but this practical recommendation does not define hypoglycaemia. The American Diabetes Association has proposed a blood glucose concentration of 3.9 mmol/l (70 mg/dl) as the value that represents hypoglycaemia (American Diabetes Association Workgroup on Hypoglycemia 2005), but this has been challenged as being too high. If the definition of biochemical hypoglycaemia is set at too high a cut-off, many episodes that have no clinical relevance will be captured in clinical trials of therapeutic agents (Swinnen et al. 2009).

Clinical definitions of hypoglycaemia The inability to obtain international consensus to agree on a biochemical definition for hypoglycaemia requires instead the application of clinical criteria (Box 4.1). Because the symptoms of hypoglycaemia are not specific, and vary between individuals (see Chapter 2), the use of symptomatology alone may be unreliable and may result in the inclusion of episodes that are not true hypoglycaemia. In one prospective study, where capillary blood glucose was measured whenever a patient had symptoms suggestive of hypoglycaemia,

Box 4.1 Clinical definitions of hypoglycaemia • Asymptomatic hypoglycaemia – low blood glucose identified on routine blood test, with no associated symptoms. • Mild symptomatic hypoglycaemia* – symptoms suggestive of hypoglycaemia; episode successfully treated by the patient alone. • Severe hypoglycaemia* – assistance from a third party is required to effect treatment. • Profound hypoglycaemia* – associated with permanent neurological deficits or death.

* Contemporaneous demonstration of a low blood glucose concentration is not mandatory but, particularly in the case of mild episodes, if it is available it does make identification more definite. Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 65 only 29% of such episodes were accompanied by evidence of biochemical hypoglycaemia (blood glucose <3.0 mmol/l) (Pramming et al. 1990). For that reason, hypoglycaemia is usually defined as an episode in which typical symptoms occur and the symptoms are reversed by appropriate treatment to raise blood glucose. Ideally, corroborative evidence should be provided by the contemporaneous demonstration of a low blood glucose con­ centration but, as this may be neither available nor possible, it need not be regarded as an absolute requirement to identify hypoglycaemia. The definition ofsevere hypoglycaemia used in the Diabetes Control and Complications Trial (DCCT) has gained widespread acceptance and is now regarded as the standard defi­ nition for clinical research (The Diabetes Control and Complications Trial Research Group 1997). Severe hypoglycaemia was defined as an episode of hypoglycaemia in which assist­ ance from a third party was required to effect treatment and recovery. The clear advantage of this definition is its inherent simplicity, which allows it to be used in a variety of differ­ ent clinical and research settings. The disadvantages of the definition are that it does not specify a threshold level below which the blood glucose concentration must fall before an episode of hypoglycaemia can be considered to be severe, nor does it specify any particular method of treating the low blood glucose. There is a substantial difference in the degree of neuroglycopenia during an episode of hypoglycaemia that has resolved following the oral administration of carbohydrate, and one that has resulted in coma and required the administration of parenteral glucagon or glucose. Mild hypoglycaemia is considered to occur when an episode is self-treated by the affected individual. Although the recording of episodes of severe hypoglycaemia is gener­ ally regarded as being a fairly robust measure, mild hypoglycaemia is a much ‘softer’ endpoint. Symptom perception by the patient may be difficult and the glycaemic thresholds at which individuals develop symptoms of hypoglycaemia are very variable. In many clini­ cal research studies, it is not uncommon to include a composite definition of mild hypogly­ caemia, which includes symptomatic episodes that respond to self-treatment (with or without a confirmatory blood test) andasymptomatic episodes below an arbitrary biochemi­ cal blood glucose value (e.g. 3.5 mmol/l) that have been identified through routine blood glucose monitoring (Anderson et al. 1997). The introduction by some investigators of a further category of ‘moderate’ hypoglycaemia is not helpful and confuses the distinction between mild and severe episodes.

FREQUENCY OF HYPOGLYCAEMIA

The true frequency of hypoglycaemia in people with type 1 diabetes is difficult to estimate accurately and not simply because of differences in the definitions of hypoglycaemia. Most episodes occur at home, at work or during leisure activities, without involving medical, nursing or paramedical staff, and the subsequent recall of such episodes by patients is generally poor, particularly with regard to mild hypoglycaemia. Prospective studies therefore have a strong advantage over retrospective studies in their respective abilities to document the frequency of hypoglycaemia with accuracy. Another crucially important feature is the nature of the population of people with type 1 diabetes under study; factors such as the quality of glycaemic control (The Diabetes Control and Com­ plications Trial Research Group 1997) and the presence of impaired awareness of hypogly­ caemia (Gold et al. 1997) have a major influence on the frequency of hypoglycaemia. Participants in clinical interventional studies are often not representative of the wider 66 Hypoglycaemia in Clinical Diabetes population with type 1 diabetes. For example, in the DCCT (The Diabetes Control and Complications Trial Research Group 1993) those recruited were relatively young, well motivated and received substantial professional support (particularly those in the intensive group). They were pre-selected depending on their history of hypoglycaemia. Thus, indi­ viduals were excluded if, in the previous 2 years, they had experienced more than one episode of severe hypoglycaemia causing neurological impairment, without experiencing warning symptoms, or more than two episodes of seizure or coma, regardless of attributed cause. Such exclusions inevitably influenced the baseline frequency of severe hypogly­ caemia in the study groups. Small studies that attempt to estimate the frequency of hypoglycaemia are open to bias because chance variations in the prevalence of risk factors for hypoglycaemia can magnify (or diminish) the frequency of hypoglycaemic events to a much greater extent than in larger investigations. There is also likely to be a major period effect in studies of the prevalence of hypoglycaemia. In the post-DCCT era, it might be anticipated that the incidence of hypoglycaemia would rise as patients and diabetes healthcare professionals sought to tighten glycaemic control (Johnson et al. 2002). This trend may be counterbalanced by the increased use of home blood glucose monitoring, improved educational programmes and the greater use of insulin analogues and continuous subcutaneous insulin infusion (CSII) therapy (Chase et al. 2001). Thus, estimates of the frequency of hypoglycaemia must be considered in relation to the definitions employed and the characteristics of the patients studied. Comparisons between present-day and earlier studies must be made in the knowledge that treatment targets, methods of monitoring blood glucose and insulin therapy have all changed greatly in recent years.

Frequency of mild, symptomatic hypoglycaemia The features of several studies that have examined the frequency of mild hypoglycaemia, either prospectively or retrospectively, in adults with type 1 diabetes are shown in Table 4.1. Rates of mild hypoglycaemia vary substantially, ranging from 8 to 160 episodes per patient per year. However, direct comparisons between individual studies are extremely difficult because of differences in methodology and patient characteristics.

Retrospective studies In two separate retrospective studies in Denmark (Pedersen-Bjergaard et al. 2001, 2004) the patient characteristics were very similar, with most administering four or more injec­ tions of insulin per day and having moderate glycaemic control. On being asked to recall the number of episodes of mild, symptomatic hypoglycaemia they had experienced in the preceding week, the subjects reported an average frequency of two episodes per patient per week. The strength of these studies lies in the large numbers of patients examined and the short period of recall. This should have minimised inaccuracy, but may have been an unrepresentative time frame.

Prospective studies Prospective studies offer the potential to provide more convincing data on frequency of mild hypoglycaemia, but substantial differences in prevalence have again been Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 67

or d

8 94 88 42 35 29 160 104 104 (number/pt/ yr) Frequency hypoglycaemia

of

and/or

mmol/l

awareness

mmol/l

3.0

< 3.5

<

BG impaired

Symptomatic Symptomatic Symptomatic Symptomatic Symptomatic Symptomatic BG Symptomatic Definition of hypoglycaemia with

e.g.

(%) c 1c 8.7 7.2 8.6 8.4 8.6 8.3 9.1 7.3 7.8 patients,

HbA of

b subsets daily

inj

f 4 injections

twice multiple multiple S&I multiple ≥ 2

multiple ≥ insulin.

particular

in Treatment 78% All 86% 87% 72% 49% 80% All

person.

per

isophane

(P) (P) a

(P) (P) rate

(P) (R) (R)

months specified.

bedtime

hypoglycaemia retrospective.

week weeks week months week weeks months not annual

R, diabetes

and 1 6 1 1 4 9–12

mild an 12 12 Follow-up 1

was

of

insulin type

rates frequency;

e

represent

injections with years)

on

to

T2) years)

of

5 15 analogue) < >

(27 ( ( focus

adults

(or to

31 94 46 54 adjusted in

411 201 170 243 frequency hypoglycaemia 1076 Number of patients

of been soluble

but

tended

of

has

which (inj)

insulin.

(2007)

study

estimation

diabetes. study.

hypoglycaemia

2

studies each

human

injections Group in

isophane mild

type (2001) (2003a) (2004) under

concentration. . . . early

of prospective

versus and

had

Study P,

et al et al et al

subjects

glucose include

study (1991)

preprandial

soluble (2005) . the porcine

(2000) not to .

this

. hypoglycaemia Frequency (2005)

on for

blood

follow-up:

. in et al

of

1c

using

et al does

of

et al

refers

were

et al HbA

Hypoglycaemia

table capillary Patients

Duration Multiple Frequency 27 Mean

Patients

Table 4.1 Table Authors Pramming Janssen Pedersen-Bjergaard Pedersen-Bjergaard Pedersen-Bjergaard Donnelly Leckie U.K. This individuals a b c d e f BG, 68 Hypoglycaemia in Clinical Diabetes reported. In an earlier study of 441 patients with type 1 diabetes, managed principally with a twice-daily insulin regimen containing soluble and isophane insulins, the weekly average was 1.8 episodes of mild symptomatic hypoglycaemia (Pramming et al. 1991). These patients had moderate glycaemic control and the period of assessment was one week. This study may be of less relevance today in view of the current use of more intensive insulin therapy and insulin analogues, yet it is interesting that the rate of mild hypoglycaemia has remained unchanged. Another Danish prospective study (Pedersen-Bjergaard et al. 2003a) included patients with similar characteristics to those in the two retrospective studies by the same group (Pedersen-Bjergaard et al. 2001, 2004). Hypoglycaemia was recorded monthly with epi­ sodes of mild symptomatic hypoglycaemia being reported for the preceding week. Mild hypoglycaemia occurred on average 1.7 times per patient per week. Subjects were also asked to perform a monthly five-point blood glucose profile and to record in addition any blood glucose value below 3.0 mmol/l. Measurements demonstrating biochemical hypogly­ caemia represented 3.7% of all blood glucose readings. In a community-based study in Tayside, Scotland, 94 adults with type 1 diabetes were selected at random from a regional diabetes database and were asked to record episodes of hypoglycaemia prospectively over one month (Donnelly et al. 2005). Their median age was 40 years, median duration of diabetes was 18 years and median HbA1c was 8.3%. Biphasic insulin was used by 35% of participants, and 49% used a combination of inter­ mediate and short-acting insulins (although the frequency of injections was not reported). A total of 325 episodes of mild hypoglycaemia occurred, representing a rate of 41.5 epi­ sodes per person per year, i.e. approximately half the rate reported by Pramming et al. (1991) and Pedersen-Bjergaard et al. (2003a). The study has weaknesses; it was relatively small, and data on frequency of blood glucose monitoring were limited. The precise criteria for defining mild hypoglycaemia were not clearly described and, in particular, the role of contemporaneous monitoring data was not specified. Nevertheless, the subjects were prob­ ably very representative of the local population with type 1 diabetes. Similar data were recorded in a multicentre study from the UK (U.K. Hypoglycaemia Study Group 2007). The primary aim was to compare the frequencies of hypoglycaemia in adults with different types and durations of diabetes, receiving different treatment modalities. Two of the groups had type 1 diabetes: 50 with a duration of less than 5 years and 57 with diabetes for more than 15 years. All were taking two or more injections of insulin per day and had good glycaemic control (mean HbA1c 7.6%). The participants were followed for between 9 and 12 months (mean 10 months) and reported all episodes of symptomatic, self-treated hypoglycaemia and episodes where blood glucose was less than 3.0 mmol/l, regardless of symptomatology. Forms were used to document such episodes and the participants were encouraged to record simultaneous blood glucose levels. To maximise compliance, subjects were required to submit completed forms every month, whether or not hypoglycaemia episodes had occurred. They were contacted by phone if no forms were received. Using this robust methodology, mean rates of hypoglycaemia of 35 and 29 episodes per person per year were reported for the short and long duration groups, respectively. The distribution of episodes was much skewed, with some reporting no events and others in excess of 200 per year; overall, around 85% of individuals experienced at least one episode of mild hypoglycaemia. The relative proportion of symptomatic and asymptomatic episodes was not reported, nor the proportion of symptomatic episodes that had corroborative blood measurements, but these informative data indicate that duration Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 69 of diabetes has little impact on the frequency of mild hypoglycaemia, consistent with previ­ ous observations (Pedersen-Bjergaard et al. 2004). In a 12-month prospective study of 243 insulin-treated adults, most of whom had type 1 diabetes, Leckie and colleagues reported that mild, symptomatic hypoglycaemia occurred with a frequency of only eight episodes per patient per year (Leckie et al. 2005). This is one of the lowest rates to be reported in a large group of adults of working age. However, the participants had suboptimal glycaemic control (mean HbA1c 9.1%), included some people with insulin-treated type 2 diabetes, and the prevalence of impaired awareness of hypoglycaemia, a recognised risk factor for hypoglycaemia (see Chapter 6), was exception­ ally low in this cohort, at 3%. The apparent strength of the study, namely a long period of follow-up, may have been an inadvertent weakness by inducing ‘patient fatigue’ in that participants may have been less assiduous in recording episodes of mild hypoglycaemia as the study progressed. Finally, this was primarily a study of the impact of hypoglycaemia occurring in the workplace. Thus, all of the participants were in fulltime employment and so represented an atypical group of individuals who may have adopted strategies to reduce the frequency of hypoglycaemia because of the potentially adverse effects that this could have on their jobs.

Frequency of asymptomatic, biochemical hypoglycaemia Aside from the problem of specifying a biochemical threshold for hypoglycaemia, there can be little doubt that asymptomatic hypoglycaemia is even more common than sympto­ matic, mild hypoglycaemia. However, a major determinant of the frequency of documented asymptomatic hypoglycaemia is, necessarily, the frequency with which blood glucose is measured. Traditionally, two approaches have been used to investigate this phenomenon: (a) taking multiple sequential blood samples in a controlled, experimental setting, over a pre-specified time period; (b) inviting patients in the community to perform capillary blood tests at multiple, pre-set time points. Providing the sampling frame is sufficiently frequent, the advantage of the former strategy is that it will capture all episodes of biochemical hypogly­ caemia, however brief, during the time period under study. The disadvantage is that it is labour intensive for investigators and only a small number of subjects can be studied over a relatively short time period, such as 12 or 24 hours. By contrast, the use of home blood glucose monitoring, particularly with meters that store the results electronically, allows larger numbers of subjects to be studied over prolonged time periods. Typically, subjects are asked to perform periodic seven-point profiles, that is, blood glucose estimations before each meal, 2 hours after food and during the night. The obvious disadvantages of this mode of investigation are that subjects may forget to perform the relevant monitoring, or do so at the wrong times, and that episodes of asymptomatic hypoglycaemia may occur outside the sampling time frames and are therefore missed. Thorsteinsson et al. (1986) examined seven-point capillary blood glucose profiles in 99 adults with type 1 diabetes and demonstrated that the frequency of biochemical hypogly­ caemia was inversely related, in a curvilinear manner, to the median blood glucose con­ centration. Thus, for example, in patients who had a median blood glucose concentration of 5.0 mmol/l, 10% of blood glucose levels were less than 3.0 mmol/l. By contrast, only 2.5% of blood glucose levels were below 3.0 mmol/l, in patients whose median blood glucose concentration was 10 mmol/l (Figure 4.1). 70 Hypoglycaemia in Clinical Diabetes

100

50

25

10 BG < 4 5

2.5 BG < 3

1

BG < 2.5 Frequency of low blood glucose concentration (%) blood of low Frequency

BG < 2 2 4 6 8 10 12 Median blood glucose (mmol/l)

Figure 4.1 Correlation between the median blood glucose concentration and the frequencies of blood glucose concentrations below 4.0, 3.0, 2.5 and 2.0 mmol/l in adults with type 1 diabetes. Solid lines represent data from 70 adults on twice daily insulin therapy and dotted lines represent data from 20 adults treated with continuous subcutaneous insulin. Approximately 10% of readings were below 3.0 mmol/l, when median blood glucose concentration was 5.0 mmol/l. (Source: Thorsteinsson B 1986. Reproduced with permission from John Wiley & Sons, Ltd).

In a different study, nocturnal blood glucose profiles were examined in 31 patients with type 1 diabetes using multiple injection therapy with soluble and isophane (NPH) insulin

(mean HbA1c 8.6%). Venous blood samples were taken every 30 minutes from 23:00 h until 07:30 h. Nocturnal hypoglycaemia (blood glucose less than 3.0 mmol/l) occurred on 29% of occasions, and 67% of these episodes were asymptomatic (Vervoort et al. 1996). When six individuals were studied on two separate nights, the blood glucose profiles showed considerable intra-individual variation. Janssen et al. (2000) have studied, prospectively, the frequency of biochemical hypogly­ caemia over a 6-week period in 31 people with type 1 diabetes with a mean HbA1c of 7.2% (all had an HbA1c ≤ 8.3%). Subjects were all using a multiple injection regimen with soluble and isophane insulins and performed one seven-point blood glucose profile and six four- point profiles each week. No overnight readings were performed. Patients completed a mean of 82% of the required monitoring schedule and overall experienced a mean of 18.7 episodes of hypoglycaemia (i.e. approximately 160 episodes per year). The range was wide, ranging from 0 to 41 episodes per individual during the period of the study. Continuous glucose monitoring (CGM) has become widely available over the last 10 years (see Chapter 9) and has an obvious advantage over intermittent blood glucose moni­ toring in identifying the frequency of biochemically low glucose. The main ‘problem’ with CGM is that it measures interstitial ‘tissue’ glucose, rather than blood glucose and so by Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 71 definition does not detect hypoglycaemia. There are also important ‘quality control’ issues that must be considered and in particular artefacts caused by sensor failure can be misclas­ sified as low interstitial glucose. Data from the UK Hypoglycaemia Study Group suggest that a sensor glucose of <3.0 mmol/l provides optimum sensitivity and specificity for detecting true hypoglycaemia (Choudhary et al. 2011). Using higher interstitial glucose levels increases sensitivity but reduces specificity for true hypoglycaemia, whereas the converse is true for lower cut-offs. In a post hoc analysis by the UK Hypoglycaemia Study Group, CGM and capillary blood glucose data were collected from 95 people with type 1 diabetes (Choudhary et al. 2010). Seventy-four had normal hypoglycaemia awareness and 21 had impaired awareness of hypoglycaemia. CGM was undertaken for at least 96 hours (mean 5 days) and patients were asked to perform a weekly four-point capillary blood glucose profile and to record prospectively episodes of symptomatic hypoglycaemia over a 9–12 month period. The mean monthly rates of biochemical hypoglycaemia in the normal and impaired awareness groups were 0.8 and 1.3 respectively. The weekly rates for daytime ‘CGM hypoglycaemia’ (interstitial glucose <3.0 mmol/l) were 2.5 and 3.7, respectively, and the nocturnal ‘CGM hypoglycaemia’ rates were 1.9 and 2.2, respectively. In a post hoc analysis from a randomised trial, nocturnal CGM data from 176 children and adults (representing some 36 467 nights of recordings) were analysed (Juvenile Dia­ betes Research Foundation Continuous Glucose Monitoring Study Group 2010). In this study, the mean HbA1c of participants was between 7.1 and 7.6%, and participants were asked to wear a CGM device on a daily basis for 12 months; low glucose was defined using a 3.3 mmol/l interstitial glucose cut-off. The median percentage of nights with low glucose per subject was 7.4%, which equated to approximately two episodes per month per subject. Only six individuals (3%) had no low nocturnal glucose over the period of the study.

Frequency of severe hypoglycaemia As with mild hypoglycaemia, direct comparisons between individual studies on the fre­ quency of severe hypoglycaemia may be difficult. However, this is a more robust endpoint than mild hypoglycaemia and, because episodes typically have a more profound effect on individuals, retrospective recall is much more reliable. At the conclusion of their prospec­ tive study, Pedersen-Bjergaard et al. (2003b) showed that 90% of subjects correctly recalled their experience of severe hypoglycaemia over the preceding year (Figure 4.2). Subjects who had experienced a high incidence of severe hypoglycaemia (prospectively recorded), retrospectively underestimated the overall rate by around 15%. Table 4.2 lists some of the large surveys (each in excess of 100 participants) that have examined the frequency of severe hypoglycaemia. The table focuses primarily on studies examining unselected groups of individuals with type 1 diabetes, and so excludes interven­ tion trials (The Diabetes Control and Complications Trial Research Group 1993; Reichard and Pihl 1994; MacLeod et al. 1995) and studies that examined particular subgroups of patients, such as people with impaired awareness of hypoglycaemia (Gold et al. 1994; MacLeod et al. 1994; Clarke et al. 1995), differing durations of diabetes (U.K. Hypogly­ caemia Study Group 2007) or subjects who have received intensive therapy or education (Muhlhauser et al. 1985; Pampanelli et al. 1996; Bott et al. 1997). A study that examined a mixed group of children, adolescents and adults (Allen et al. 2001) has also been excluded. The frequency of severe hypoglycaemia (defined as episodes requiring third- party assistance) in the studies listed in Table 4.2 is remarkably consistent at 1.0–1.6 epi­ sodes per patient per year. 72 Hypoglycaemia in Clinical Diabetes

15

10

5 (episodes per patient-year) Recalled severe hypoglycaemia severe Recalled

0

0 5 10 15 Prospectively recorded severe hypoglycaemia (episodes per patient-year)

Figure 4.2 Correlation between prospectively recorded and retrospectively recalled rate of severe hypoglycaemia over the same one year period in 230 people with type 1 diabetes. Marker sizes are weighted by the number of cases. R2 = 0.66; P < 0.001. (Source: Pedersen-Bjergaard U 2003b. Repro- duced with permission from John Wiley & Sons, Ltd).

However, it is important to appreciate that the frequency of severe hypoglycaemia in unselected populations does not follow a Gaussian distribution (Figure 4.3). The distribu­ tion is heavily skewed such that the majority of individuals do not experience any severe hypoglycaemia in a given year, while a small number of individuals have recurrent epi­ sodes. In the studies in Table 4.2, between 30 and 40% of individuals experienced at least one episode of severe hypoglycaemia over the period in question. The proportion affected in the study by Pramming et al. (1991) was much lower, but the period of follow-up was only one week. This skewed distribution serves to emphasise further the importance of patient selection in ascertaining the frequency of severe hypoglycaemia with accuracy, as the exclusion of a relatively small number of people at high risk would substantially reduce the overall risk. In three of the studies of unselected adults with type 1 diabetes, severe hypoglycaemia was further subdivided to examine episodes associated with more significant neuroglyco­ penia, i.e. those resulting in coma and/or seizures (Table 4.2) (ter Braak et al. 2000; Pedersen-Bjergaard et al. 2003a, 2004). Furthermore, in the study by Muhlhauser et al. (1998) only episodes treated with intra-muscular glucagon or intravenous glucose were reviewed. Predictably, such events were rarer and represented about one-quarter of all episodes of severe hypoglycaemia. Emergency and hospital services will occasionally be involved in the management of severe hypoglycaemia and data exist on the use of such agencies. This has to be inter­ preted with caution, as most hypoglycaemic events are managed in the community, without involvement of (para) clinical staff. Individuals admitted to hospital with Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 73

3 subsets

29 34 22 46 41 39 37 13 NR Proportion Affected (%) particular

in d

1.4 1.6 1.1 0.98 1.1 3.2 1.5 0.4 1.1 0.3 1.3 0.35 0.21 (episodes/ pt/yr) Frequency hypoglycaemia

severe

of

Coma/ Coma/ Coma/

rates

party party party party party party party party on

seizure seizure seizure Definition of hypoglycaemia Third Third Third Third Third Third Third Third Glucagon/glucose focus

to

(%) tended c

(A1) 1c

8.7 8.0 7.8 8.6 8.4 8.6 9.1 7.3 7.8 which

insulin. HbA 10.7

insulin.

studies

person. human

b daily daily per early

isophane

inj inj inj inj

versus

2 4 4 4 injections rate

twice twice intensive multiple > ≥ ≥ ≥ 2

include ≥

bedtime

retrospective. porcine

diabetes

Treatment 78% 76% 70% 82% 86% 87% 72% 80% All not annual

R, 1

and

an

using does

type

(R) (R) (R) (R) (P) (R) (P) a insulin

and

(P)

frequency; with

represent months

to individuals

week months months months months months months months . subjects

or 1

analogue) 1 9–12 adults

12 12 12 24 12 12 12 Follow-up in 100 (or HbA

adjusted is

of

hypoglycaemia

of been soluble

e e

figure excess

years)

of

hypoglycaemia

T2) T2) has years)

in

A1,

of 5 15

(inj) < >

(56 (27 ( (

study hypoglycaemia estimation

study; 46 54

411 600 684 195 243 Number of patients 207 170 1076 each examining

injections awareness in

severe diabetes.

under

. . . 2

of prospective

studies

P,

type et al et al et al Study

subjects impaired (1998)

. (1991) preprandial with

(2000) (1993)

. the

with . . to

considers

hypoglycaemia Frequency (2005)

et al for

reported.

follow-up:

. et al

of e.g. 1c

et al et al

of

only

patients

not refers

et al HbA

table were data Hypoglycaemia

Braak

patients, (2001) (2003a) (2004) Group

Duration Multiple Frequency 56 Mean

Table 4.2 Table Authors Pramming MacLeod Muhlhauser ter Pedersen-Bjergaard Pedersen-Bjergaard Pedersen-Bjergaard Leckie UK This of a b c d e NR, 74 Hypoglycaemia in Clinical Diabetes

90

80

70

60

50

40 % of patients 30

20

10

0 0 2 4 6 8 10 Severe hypoglycaemia, episodes per year

Figure 4.3 Distribution of self-reported number of episodes of severe hypoglycaemia during the pre- ceding year in 1049 unselected patients with type 1 diabetes (light bars) and 209 patients selected by criteria to mimic the characteristics of the DCCT cohort (dark bars). (Source: Pedersen-Bjergaard U 2004. Reproduced with permission from John Wiley & Sons, Ltd). hypoglycaemia are probably atypical and have a higher prevalence of alcohol dependence and mental illness (Hart and Frier 1998). An early study from Australia reported that, over one year, 3.5% of people attending an urban diabetes clinic had an episode of hypoglycaemia severe enough to warrant referral to hospital (Moses et al. 1985). More recent data from Tayside, Scotland, demonstrated that over one year 7.1% of people with type 1 diabetes had an episode of hypoglycaemia that required treatment by emergency medical staff, with an incidence rate of 0.12 per patient per year (Leese et al. 2003). Thus, only about one in 10 of all episodes of severe hypoglycaemia result in contact with emergency services.

Frequency of hypoglycaemia: summary and conclusions Published studies on frequency of hypoglycaemia are heterogeneous and inconsistent, and there are considerable methodological limitations in the ability to ascertain accurately the frequency of hypoglycaemia. The definitive study, a prolonged, prospective and compre­ hensive evaluation of a large number of unselected people with type 1 diabetes, outside ‘trial’ conditions, has yet to be performed. Existing data on severe hypoglycaemia are probably fairly accurate, but estimates of mild hypoglycaemia should be regarded with caution. However, the ‘average’ patient with type 1 diabetes will probably experience about 1–2 episodes of mild, symptomatic hypoglycaemia per week and be exposed to several thousand episodes over a lifetime with diabetes. CGM data suggest that asymptomatic, biochemical hypoglycaemia is even more common and, over the course of a year, probably occurs in almost all patients who have at least moderate glycaemic control. Overall, severe hypoglycaemia may be expected to occur once or twice each year, but the distribution is Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 75 heavily skewed, such that most individuals will be unaffected while a small number will have multiple episodes.

CAUSES OF HYPOGLYCAEMIA

Exogenous administration of insulin is still a relatively imprecise way of managing type 1 diabetes. The time-action profiles of modern insulins do not mimic the physiological changes in plasma insulin concentrations that occur in non-diabetic individuals (Figure 4.4) and, crucially, plasma concentrations of exogenously administered insulin cannot respond to changes in blood glucose concentration. Therefore, at its most fundamental level, hypoglycaemia in people with type 1 diabetes is the result of an imbalance between insulin- mediated glucose efflux from the circulation and the amount of glucose entering the blood from ingested carbohydrate and from the liver and kidneys. Cryer et al. (2003) have grouped the causes of hypoglycaemia in type 1 diabetes into six categories (Box 4.2) depending on their relative effects on insulin concentrations or sensitivity and on glucose

Glucose homeostasis 8 0.08 6 4 2 0.04

0 Glucose (mmol/L) Insulin (U/L)

0

08·00 13·00 16·00 19·00 Time of day (h)

Figure 4.4 Mean 24-hour plasma glucose and insulin profiles in 12 healthy non-diabetic individuals. Shaded areas represent 95% confidence intervals. Glucose levels remain within tight limits, while there is considerable variation in insulin concentrations, particularly around mealtimes. (Source: Owens DR 2001. Reproduced with permission from Elsevier).

Box 4.2 Causes of hypoglycaemia in type 1 diabetes 1. Inappropriate insulin injection – e.g. excessive dose, inappropriate time, inappropriate insulin formulation. 2. Inadequate exogenous carbohydrate – e.g. missed meal or snack, overnight fast. 3. Increased carbohydrate utilisation – e.g. exercise. 4. Decreased endogenous glucose production – e.g. excessive alcohol consumption. 5. Increased insulin sensitivity – e.g. night time, exercise, weight loss. 6. Decreased insulin clearance – e.g. renal failure.

Modified from Cryer et al. (2003). 76 Hypoglycaemia in Clinical Diabetes entry into the circulation. Although this classification is imperfect, it is a useful way of considering possible causes of an episode of hypoglycaemia.

Patient error It is common after an episode of hypoglycaemia for the person with diabetes or a member of the diabetes team to try to ascertain why the episode occurred. Although there is always a risk of ‘spurious attribution’, a cause can often be identified and may be the result of an error of judgement. Carbohydrate intake may have been inadequate because a meal was missed or delayed, or simply contained an insufficient content of carbohydrate. Alterna­ tively, the individual may have injected too much insulin relative to the amount of carbo­ hydrate in the meal. It is also not uncommon for insulin to be administered at an inappropriate time or for the ‘wrong’ insulin to be injected accidentally, e.g. a rapid-acting insulin ana­ logue has been administered at a time when basal insulin should have been given. An often unrecognised problem is the inadequate resuspension of isophane (or lente) insulins or of fixed mixtures of insulin. If these insulins are not fully resuspended prior to subcutaneous injection, the insulin may be absorbed at variable rates resulting in unpredictable insulin levels and, thus, an increased risk of hypoglycaemia (Owens et al. 2001). Deliberate over­ dose of insulin is rare, but may result in protracted hypoglycaemia.

Alcohol The relationship of alcohol to hypoglycaemia is considered in detail in Chapter 8. Surveys based on patient interviews have implicated alcohol in up to one-fifth of episodes of severe hypoglycaemia requiring hospital admission (Potter et al. 1982; Moses et al. 1985; Feher et al. 1989; Hart and Frier 1998). In a recent study of 141 people treated for severe hypoglycaemia in three emergency centres in Copenhagen, alcohol was detected in the blood of 17% of the subjects (Pedersen-Bjergaard et al. 2005). The median blood alcohol concentration was 11 mmol/l. Alcohol inhibits gluconeogenesis and may directly contrib­ ute to the development of hypoglycaemia. In addition, there is some evidence to suggest that alcohol attenuates the counterregulatory response to hypoglycaemia (Avogaro et al. 1993; Turner et al. 2001). However, alcohol also impairs awareness of hypoglycaemia and so hinders the ability of individuals to take appropriate corrective therapy (Kerr et al. 1990). Thus, an individual under the influence of alcohol may not recognise that he or she is developing hypoglycaemia, and even if the symptoms are recognised, the person may not have the capacity to self-treat. Hypoglycaemia may therefore be allowed to progress rapidly in severity. Moreover, friends, colleagues or bystanders may presume that the neuroglycopenic symptoms and signs exhibited by the individual are a consequence of alcohol intoxication and appropriate treatment and prompt medical help may not then be provided. Alcohol has been implicated in many instances of profound and protracted insulin-induced hypoglycaemia associated with permanent neurological damage (Arky et al. 1968).

Exercise Exercise is recommended for people with type 1 diabetes because of its positive physi­ ological and psychological effects. However, exercise can increase the risk of hypogly­ caemia both during the physical activity itself and in the recovery period (see Chapters Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 77

17 and 18). Moderate exercise in non-diabetic individuals causes a fall in plasma insulin to 40–50% of pre-exercise levels (Galassetti et al. 2003). This normal physiological decline cannot occur when insulin is being administered exogenously, unless the dose is reduced before exercise commences (Sonnenberg et al. 1990). An acute increase in insulin sensitivity following exercise also increases the risk of hypoglycaemia (Sonnenberg et al. 1990). The metabolic and counterregulatory hormonal responses to acute hypoglycaemia and exercise are qualitatively very similar. Antecedent exercise blunts the counterregulatory response to subsequent acute hypoglycaemia (Galassetti et al. 2001a, 2001b). The inverse situation also applies, i.e. antecedent hypoglycaemia diminishes the counterregulatory response to subsequent exercise (Galassetti et al. 2003). Thus, at the very time that meta­ bolic substrate requirements are increasing, there is the potential for an acute failure of endogenous glucose production. Impaired counterregulatory responses may be an impor­ tant mechanism promoting exercise-related hypoglycaemia in type 1 diabetes.

RISK FACTORS FOR SEVERE HYPOGLYCAEMIA

In a relatively high proportion of cases – in some series as high as 40% (Feher et al. 1989) – it is not possible to identify the precipitating cause of an episode of hypoglycaemia. Indeed, this figure may be even higher, because the phenomenon of ‘spurious attribution’ means that in some instances the perceived precipitant of a given episode may have been an innocent bystander. It has, therefore, been increasingly recognised that healthcare profes­ sionals must look beyond conventional precipitating factors and consider other phenomena which may be associated with an increased risk of hypoglycaemia (Table 4.3). Many of these are discussed in more detail in other chapters.

Intensive insulin therapy Randomised trials, most notably the DCCT, have provided substantial data on the epide­ miology of hypoglycaemia in adults with type 1 diabetes and, in particular, on the impact of intensive insulin therapy.

The Diabetes Control and Complications Trial This was a landmark study and provided the proof that strict glycaemic control limits the incidence and severity of microvascular complications in people with type 1 diabetes. A total of 1441 patients with type 1 diabetes were randomly assigned either to intensive or to conventional insulin therapy (The Diabetes Control and Complications Trial Research Group 1993). In the conventional group, patients did not generally perform home blood glucose monitoring, clinical reviews were undertaken every 3 months and patients were not informed about their HbA1c result unless it was in excess of 13%. By contrast, in the intensive therapy group, subjects performed frequent home blood glucose monitoring, had monthly consultations with the specialist team and frequent telephone contact to achieve as strict glycaemic control as possible. Over a mean follow-up of 6.5 years, the average

HbA1c concentration in the intensive group was approximately 7.0% and in the conven­ tional group approximately 8.8% (The Diabetes Control and Complications Trial Research Group 1993). 78 Hypoglycaemia in Clinical Diabetes - + + hypogly

ACE ACE activity mild

women. − − + − − +

in

Insulin dose predominantly − + − − − − + − +

common

Age hypoglycaemia.

more

severe

examining F − − − + − −

M and

Gender Studies

factor

hypoglycaemia

risk

+ + − + − diabetes.

severe 1

= C-peptide negative

F between

type

with men;

in

+ + + + + − + + + + + + association adults

in no

Impaired awareness

= common

more

examined

− + + − − + + + − + − + + + * Duration hypoglycaemia; commonly diabetes

1

hypoglycaemia

+ + + + + most

severe

type

Previous episode severe been

and

= with

M 1c

have

factor − − + + − − + − − − − − −

+ * that adults Low HbA

risk

in

included.

not

between − − + − − + −

hypoglycaemia Intensive therapy

hypoglycaemia association

severe

hypoglycaemia

for of

severe

positive Group

(2001) (2003a) (2004)

=

for . . . factors

+

Complications Research and

Study risk awareness

et al et al et al

if (1994) (1985) (1987) (1998)

the (1996)

factors Trial

. . .

. IDDM (2000) (1993) (1994) (1994)

included.

(2008) . . . Pihl Control

factor (1995)

. Risk (2005)

et al et al et al (2003)

not . et al (1994) (1997)

.

.

(1997) . . et al et al et al risk

(1997) and

. examines

et al Group

a

et al

were et al

et al

et al et al

Hypoglycaemia

et al Diabetes EURODIAB

table

only

Braak

Study Complications Group (2007) =

Table 4.3 Table Study Muhlhauser Muhlhauser MacLeod Gold The MacLeod Reichard Clarke Pampanelli Bott The Gold Muhlhauser Pedersen-Bjergaard ter Leese Pedersen-Bjergaard Pedersen-Bjergaard Leckie U.K. Geddes This caemia * Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 79

As noted above, the subjects recruited to this study were not typical of the wider popula­ tion of people with type 1 diabetes. In particular, individuals were not permitted to take part if, in the preceding 2 years, they had experienced more than one episode of severe neurological impairment without warning symptoms of hypoglycaemia, or more than two episodes of seizure or coma, regardless of attributed cause. Moreover, during the study itself, the occurrence of an episode of severe hypoglycaemia in an individual patient prompted a review of conventional risk factors and, in instances where probable causes were identified, corrective actions such as re-educating the patient were undertaken (The Diabetes Control and Complications Trial Research Group 1997). Despite all these factors, 3788 episodes of severe hypoglycaemia occurred in the 1441 patients over the course of the study and, of these, 1027 were associated with coma and/ or seizure (The Diabetes Control and Complications Trial Research Group 1997). The rate of severe hypoglycaemia in the intensively-treated patients was 0.61 per patient per year, while that in the conventionally-treated group was 0.19 per patient per year, i.e. a 3-fold difference. Over the mean period of follow-up (6.5 years), 65% of the intensive group experienced at least one episode of severe hypoglycaemia, compared with 35% of the individuals in the conventional group. The DCCT was a multicentre study; 27 out of the 29 centres reported that intensive therapy was associated with an increased risk of severe hypoglycaemia, but in two centres this was not observed (The Diabetes Control and Complications Trial Research Group 1997). This led some investigators to claim that with appropriate education and training intensive insulin therapy need not necessarily be associated with an increased risk of hypoglycaemia (Plank et al. 2004). However, more recently DAFNE investigators reported both a significant reduction in the incidence of severe hypoglycaemia from 1.7 ± 8.5 to 0.6 ± 3.7 episodes per person per year and that hypoglycaemia recognition improved in 43% of those reporting unawareness in a 1-year follow-up of enrolled patients (Hopkins et al. 2012).

The Stockholm Diabetes Intervention Study (SDIS) A much smaller study than the DCCT, it had similar aims and objectives. A group of 102 patients with type 1 diabetes were recruited and 89 remained after 7.5 years of follow-up

(Reichard and Pihl 1994). The mean HbA1c was 7.1% in the intensive group and 8.5% in the conventional group. Severe hypoglycaemia, defined as episodes requiring external assist­ ance, occurred in 80% of subjects in the intensive group and 58% of those in the conven­ tional group over the follow-up period. Overall, the incidence of severe hypoglycaemia was 1.1 episodes per patient per year in the intensive group and 0.4 episodes per patient per year in the conventional group. Other risk factors for severe hypoglycaemia were not reported.

The Bucharest-Dusseldorf Study A total of 300 individuals with type 1 diabetes were randomised to: (a) conventional ther­ apy for one year and then to one year of intensive therapy; (b) 2 years of intensive therapy; or (c) a 4-day inpatient group teaching programme with conventional insulin therapy for one year (Muhlhauser et al. 1987). Glycated haemoglobin (HbA1) remained unchanged at 12–13% during conventional therapy but fell to around 9.5% during intensive therapy. In the first year of the study, severe hypoglycaemia occurred in 6% of the intensively-treated patients and 12% of the conventionally-treated patients, and in year two the proportion of patients experiencing severe hypoglycaemia in both intensive groups fell to 3–4%. 80 Hypoglycaemia in Clinical Diabetes

Observational data The Dusseldorf team reported observational data on the impact of intensive therapy on the frequency of severe hypoglycaemia (Bott et al. 1997). A group of 636 people with type 1 diabetes who had participated in a structured 5-day inpatient treatment and teaching pro­ gramme for intensification of insulin therapy in one of ten different hospitals in Germany, was re-examined at intervals over 6 years. The mean HbA1c fell from 8.3 to 7.6%. Severe hypoglycaemia, defined as episodes treated with intramuscular glucagon or intravenous glucose, decreased from 0.28 episodes per patient per year in the year preceding the pro­ gramme to 0.17 episodes per patient per year afterwards (Bott et al. 1997). The variation in incidence of severe hypoglycaemia between different centres ranged from 0.05 to 0.27 episodes per patient per year. In Perugia in Italy, Pampanelli et al. (1996) reported retrospective data on 112 individu­ als who had been commenced on intensive insulin therapy (preprandial soluble insulin and bedtime isophane insulin) at diagnosis of diabetes and who were attending clinic at least four times per year. Mean HbA1c was 7.2% and mean duration of diabetes was 7.8 years. Frequency of mild hypoglycaemia was estimated by a review of the patients’ blood glucose monitoring diaries in the 9 months prior to study, and the overall rate was 35.6 episodes per patient per year. Severe hypoglycaemia, defined as episodes requiring third-party assist­ ance, was assessed retrospectively over the duration of diabetes and was recalled by only six patients (representing a tiny overall rate of 0.001 episodes per patient per year).

Structured education and new technologies The structured education course for people with type 1 diabetes, Dose Adjustment for Normal Eating (DAFNE), was developed on the back of the above observational data and is now used widely. A trial of this 5-day programme was performed in secondary diabetes clinics in three English Health Authorities (DAFNE Study Group 2002). A total of 169 adults with moderate to poorly controlled type 1 diabetes were randomised to an immediate or delayed DAFNE course. At 6 months, glycaemic control was significantly better in the immediate versus the delayed group (8.4% vs 9.4%; P < 0.0001), with commensurate improvements in general well-being and treatment satisfaction. No change was observed in the frequency of severe hypoglycaemia. Clearly this study did not support the hypothesis that structured education reduces the frequency of severe hypoglycaemia, but it did not result in an increased frequency of severe hypoglycaemia, which might have been antici­ pated given the improvement in HbA1c. No further formal trials of DAFNE have been reported to date, but an audit of 145 individuals who had completed a DAFNE course in Australia, reported a lower frequency of severe hypoglycaemia, despite a modest improve­ ment in HbA1c (McIntyre et al. 2010). In this study, 28% of individuals self-reported fewer episodes of hypoglycaemia over the 12 months after the course, while 8% reported more frequent episodes; the overwhelming majority of individuals reporting no change in severe hypoglycaemia history had not experienced any episodes in the year preceding training. Advances in technologies in insulin delivery using continuous subcutaneous insulin infusion and continuous glucose monitoring have been advocated as a means of improving glycaemic control while reducing the risk of hypoglycaemia. A recent systematic review and meta-analysis identified 33 randomised trials that compared CSII with multiple daily injection (MDI) regimes or CGM with capillary blood glucose monitoring (Yeh et al. 2012). In these trials, no difference was reported between MDI and CSII in terms of severe Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 81 hypoglycaemia, though in adults with type 1 diabetes CSII was associated with lower

HbA1c (although this outcome was heavily influenced by the results of one study). CGM was associated with better glycaemic control (HbA1c was 0.26% lower) compared to self- monitoring of capillary blood glucose. Sensor-augmented CSII was associated with an

HbA1c reduction of 0.68% compared with MDI and capillary blood glucose monitoring. In both instances no difference occurred in the risk of severe hypoglycaemia.

Summary Although intensive insulin therapy has been associated with an increased risk of severe hypoglycaemia, in other studies the incidence of severe hypoglycaemia declined when intensive therapy was combined with structured education. If educational programmes are of high quality, this can lower the risk of severe hypoglycaemia. However, the importance of patient selection cannot be overemphasised. In their study of 1076 adults with type 1 diabetes from the UK and Denmark, Pedersen-Bjergaard et al. (2004) examined a subset of 230 individuals whose clinical characteristics were similar to patients enrolled in the DCCT. This subgroup accounted for only 5.4% of all reported episodes of severe hypogly­ caemia, with an overall rate of 0.35 episodes per patient per year, i.e. approximately one- quarter that of the entire group. Risk factors for severe hypoglycaemia in this group (impaired awareness and retinopathy) differed from that of the study population as a whole (see Table 4.3). The DCCT participants were not representative of all people with type 1 diabetes, whereas the education programme undertaken in Dusseldorf was not widely replicated in mainstream diabetes practice until the advent of DAFNE programmes. New technologies, particularly sensor-augmented CSII, may permit intensification of glycaemic control, without increasing the risk of severe hypoglycaemia. In conclusion, although most specialists would accept that severe hypoglycaemia is more common in patients receiving intensive insulin therapy, it is not inevitable and better patient education and use of new technologies may mitigate the risk.

Strict glycaemic control Strict glycaemic control, as demonstrated by glycated haemoglobin concentrations that approach the upper end of the non-diabetic range, is closely linked to intensive insulin therapy. After the DCCT, glycated haemoglobin concentrations at or around this level have become the usual target for many people with type 1 diabetes. In the DCCT, there was a quadratic relationship between HbA1c and risk of severe hypoglycaemia (Figure 4.5), with the risk of hypoglycaemia increasing as HbA1c decreased (The Diabetes Control and Com­ plications Trial Research Group 1997). However, the attained HbA1c did not account for all of the difference in risk of severe hypoglycaemia between the two arms of the study, as subjects in the intensive group still had an excess risk of severe hypoglycaemia after statistical adjustment for HbA1c concentration. Indeed, in the intensive group, only about 5% of the variation in frequency of severe hypoglycaemia could be explained by the gly­ cated haemoglobin concentration (The Diabetes Control and Complications Trial Research Group 1997). Other studies have reported variable relationships between glycaemic control and fre­ quency of severe hypoglycaemia. In the study by Bott et al. (1997) a lower mean HbA1c was associated with severe hypoglycaemia, but there was no linear or quadratic relationship between HbA1c and severe hypoglycaemia. In the EURODIAB IDDM Complications 82 Hypoglycaemia in Clinical Diabetes

DCCT: Hypoglycaemia and glycaemic control 100

80

60

40

(per 100 100 (perpatient-years) 20 Rate of of Rateseverehypoglycaemia 0 5.05.5 6.06.5 7.07.5 8.08.5 9.0 10.09.5 10.5

HbA 1c (%) Hypoglycaemia rate

Figure 4.5 Risk of severe hypoglycaemia as a function of HbA1c in the Diabetes Control and Complic­ ations Trial. Based on data reported in the DCCT (Source: Adapted from The Diabetes Control and Complications Trial Research Group 1997).

Study, severe hypoglycaemia (defined as episodes requiring external assistance) occurred in 32% of individuals over 12 months. A clear relationship to glycated haemoglobin was evident, in that 40% of individuals with an HbA1c < 5.4% were affected compared with 24% of individuals with an HbA1c > 7.8% (The EURODIAB IDDM Complications Study Group 1994). In the study of workplace hypoglycaemia, recurrent severe hypoglycaemia was associated with strict glycaemic control, but no link was found in individuals who had experienced only one episode (Leckie et al. 2005). In several other studies, no relationship was observed between severe hypoglycaemia and glycated haemoglobin (Muhlhauser et al. 1985; Muhlhauser et al. 1987; MacLeod et al. 1993; Gold et al. 1997; Muhlhauser et al. 1998; ter Braak et al. 2000; Pedersen-Bjergaard et al. 2001; Leese et al. 2003; Pedersen-Bjergaard et al. 2003a; Pedersen-Bjergaard et al. 2004) after adjustment for other risk factors. Glycated haemoglobin does not provide the entire picture about an individual’s glycae­ mic control and although HbA1c may not predict risk of hypoglycaemia, low mean home blood glucose concentrations and excessive variability in blood glucose can identify indi­ viduals more prone to hypoglycaemia (Cox et al. 1994; Janssen et al. 2000). Thus, a straightforward relationship does not exist between severe hypoglycaemia and glycaemic control. For an episode of mild hypoglycaemia to progress to one that causes significant neuroglycopenia and impairs consciousness, other factors must operate, which negate the normal symptomatic and hormonal responses to hypoglycaemia.

Acquired hypoglycaemia syndromes Within each treatment group of the DCCT, the number of previous episodes of severe hypoglycaemia was the strongest predictor of risk of future episodes (The Diabetes Control Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 83 and Complications Trial Research Group 1997). Moreover, approximately 30% of patients in each group experienced a second episode of severe hypoglycaemia within 4 months following an initial episode. The importance of a previous history of severe hypoglycaemia in predicting future risk has been replicated in several other studies (MacLeod et al. 1993; Bott et al. 1997; Gold et al. 1997; Muhlhauser et al. 1998). Moreover, as demonstrated in Table 4.3, many studies have also linked increased duration of diabetes with an increased risk of severe hypoglycaemia. Cryer has suggested that the integrity of the glucose counterregulatory system may be a pivotal factor in determining whether the relative or absolute hyperinsulinism that fre­ quently occurs in insulin-treated diabetes ultimately results in the development of hypogly­ caemia (Cryer et al. 2003). Acquired hypoglycaemia syndromes are associated with an increased risk of severe hypoglycaemia in people with type 1 diabetes and are discussed in Chapters 3 and 6. However, it should be emphasised that mild hypoglycaemia is as important, if not more so, than relatively infrequent severe events, in the pathogenesis of these acquired syndromes, and recurrent exposure to mild hypoglycaemia can induce counterregulatory deficiencies (Heller and Cryer 1991; Widom and Simonson 1992) and impaired awareness (Veneman et al. 1993).

Counterregulatory hormonal deficiencies Hypoglycaemia-induced secretion of glucagon declines in most patients within 5 years of developing type 1 diabetes (Bolli et al. 1983). A defective epinephrine response to hypogly­ caemia may develop some years later (Bolli et al. 1983; Hirsch and Shamoon 1987; Dagogo-Jack et al. 1993). As with the glucagon response, the impaired epinephrine (adren­ aline) response is hypoglycaemia-specific but, in contrast to glucagon, it exhibits a thresh­ old effect – i.e. an epinephrine response can still be elicited by hypoglycaemia, but only at a lower blood glucose concentration (Dagogo-Jack et al. 1993). If hypoglycaemia develops in patients who have this combined counterregulatory hormonal deficiency, glucose recovery may be severely compromised (see Chapter 3). Subjecting such patients to intensified insulin therapy greatly increases the risk of severe hypoglycaemia (White et al. 1983).

Impaired awareness of hypoglycaemia In many patients with insulin-treated diabetes, the hypoglycaemia symptom profile alters with time, resulting in impaired perception of the onset of hypoglycaemia (see Chapter 6). Commonly, autonomic warning symptoms are diminished and neuroglycopenic symptoms predominate. Around 20–25% of people with type 1 diabetes develop impaired awareness of hypoglycaemia and the prevalence of this problem increases with the duration of insulin treatment (Pramming et al. 1991; Geddes et al. 2008; Graveling and Frier 2010). Prospec­ tive studies have demonstrated that the frequency of severe hypoglycaemia is increased up to 6-fold in patients with impaired awareness compared to those with normal hypoglycae­ mia awareness (Gold et al. 1994; Clarke et al. 1995; Geddes et al. 2008).

Hypoglycaemia-associated central autonomic failure The above acquired hypoglycaemia syndromes tend to segregate together clinically. People with glycated haemoglobin concentrations close to the non-diabetic range are at 84 Hypoglycaemia in Clinical Diabetes greater risk of developing impaired awareness (Boyle et al. 1995; Kinsley et al. 1995; Pampanelli et al. 1996), while the glycaemic thresholds for the onset of symptoms and responses are altered in people with impaired awareness (Graveling and Frier 2010). Cryer has suggested that these acquired abnormalities represent a form of central Hypoglycae­ mia-Associated Autonomic Failure (HAAF) in type 1 diabetes, speculating that recurrent severe hypoglycaemia is the primary cause (Cryer et al. 2003). If hypoglycaemia is the precipitant, then it is possible to see how a vicious cycle may become established with the development of the acquired hypoglycaemia syndromes promoting further episodes of severe hypoglycaemia.

Summary There is a well-known adage that ‘hypoglycaemia begets hypoglycaemia’. People who have experienced one episode of severe hypoglycaemia are much more likely to develop further episodes, and the greatest risk occurs in the weeks and months after the index event. Severe hypoglycaemia becomes more common in people with increasing duration of type 1 dia­ betes (U.K. Hypoglycaemia Study Group 2007). This is the legacy of the burden of hypoglycaemia that such individuals have experienced over many years with diabetes. The impaired symptomatic and counterregulatory responses to hypoglycaemia dramatically increase the likelihood that an episode of mild hypoglycaemia will progress to a more severe event.

Genetic predisposition to hypoglycaemia Most of the precipitants and risk factors for hypoglycaemia have been known about for many years. In 2001, Pedersen-Bjergaard et al. (2001) reported a new risk factor for severe hypoglycaemia in adults with type 1 diabetes and raised the notion that some individuals may have an inherent genetic susceptibility to hypoglycaemia. The Danish investigators noted the similarity between endurance exercise and hypoglycaemia in that both are states of limited metabolic fuel availability. Previous studies had linked exercise performance to a particular polymorphism of the angiotensin-converting enzyme (ACE) gene. Specifically, the insertion (I) allele, which resulted in low serum ACE activity, was associated with superior performance capacity compared with the deletion (D) allele. In an initial retrospec­ tive survey of 207 adults with type 1 diabetes, patients with the DD genotype had a 3.2-fold increased risk of severe hypoglycaemia in the preceding 2 years, compared with individuals with the II genotype. A significant relationship also existed between serum ACE activity, with a 1.4 increment in risk of severe hypoglycaemia for every 10 U/l rise in serum ACE concentration (Figure 4.6). The serum ACE activity was directly linked to ACE genotype, and it remained a significant risk factor even after adjustment for conventional risk factors. Moreover, serum ACE activity was a stronger risk factor for severe hypoglycaemia in C-peptide negative individuals with impaired awareness, than in other groups (relative risk 1.7 per 10 U/l; Figure 4.7). No significant relationship was observed between serum ACE activity or genotype and frequency of mild hypoglycaemia. These findings were replicated in a prospective study for one year in 107 adults (Pedersen-Bjergaard et al. 2003a). Serum ACE activity in the fourth quartile was associated with a 2.7-fold increased risk of severe hypoglycaemia compared to activity in the lowest quartile. Compared to subjects with the II genotype, individuals with the DD genotype had a 1.8-fold increased risk of severe hypoglycaemia, although this did not reach statistical Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 85

10

8

6

4 per patient-year 2

0 Episodes of severe hypoglycaemia Episodes of severe 0 20 40 60 80 100 120 Serum ACE activity (U/L)

Figure 4.6 Risk of severe hypoglycaemia according to serum ACE activity in 207 patients with type 1 diabetes, untreated with ACE inhibitors or angiotensin-2 receptor antagonists. Broken lines represent 95% confidence intervals. (Source: Pedersen-Bjergaard U 2001. Reproduced with permission from Elsevier).

C-peptide-negative, impaired awareness (n=47) C-peptide-positive, impaired awareness (n=63) C-peptide-negative, normal awareness (n=32) C-peptide-positive, normal awareness (n=60) 10

8

6

4

per patient-year 2

0

Episodes of severe hypoglycaemia Episodes of severe 0 20 40 60 80 100 120 Serum ACE activity (U/L)

Figure 4.7 Association between severe hypoglycaemia and serum ACE activity according to C-peptide status and self-estimated awareness of hypoglycaemia. (Source: Pedersen-Bjergaard U 2001. Repro- duced with permission from Elsevier). significance. Higher serum ACE concentrations were also associated with an increased risk of severe hypoglycaemia in Swedish children and adolescents with type 1 diabetes (Nor­ dfeldt and Samuelsson 2003). More recently, data from a Danish registry of 1000 children with diabetes found an association between severe hypoglycaemia and higher serum ACE levels in girls, but not boys, and reported no association between severe hypoglycaemia and ACE genotype (Johannesen et al. 2011). The Danish group put forward a number of possible mechanisms to explain their obser­ vations (Pedersen-Bjergaard et al. 2001, 2003a). They speculated that low levels of serum ACE may be associated with less cognitive deterioration during acute hypoglycaemia, thereby increasing the likelihood that remedial action to correct hypoglycaemia is taken. 86 Hypoglycaemia in Clinical Diabetes

Alternatively, low serum ACE activity might enhance counterregulation or reduce produc­ tion of toxic substances, e.g. reactive oxygen species, during hypoglycaemia. All these putative mechanisms remain highly speculative, but the authors also raise one other intriguing possibility: namely that ACE inhibition might reduce the frequency of hypogly­ caemia. This may seem counterintuitive, because previous population-based studies sug­ gested an association between severe hypoglycaemia and use of ACE inhibitors (Herings et al. 1996; Morris et al. 1997). However, these population-based studies can be criticised and the role of ACE inhibitors in ameliorating the impact of hypoglycaemia warrants further examination. However, other studies have been less supportive. Data from a Scottish centre suggested that the link between serum ACE and severe hypoglycaemia was weak in adults with type 1 diabetes (Zammitt et al. 2007), while a large study in Australia of children and adoles­ cents with type 1 diabetes could find no association between ACE genotype and severe hypoglycaemia (Bulsara et al. 2007). This should serve as a reminder that genetic suscep­ tibility to any biological variable may differ considerably between populations and rein­ forces the need for the Danish observations to be examined in other countries and other ethnic groups. Other data have linked a genetic variant of the beta2-adrenergic receptor (which results in reduced sensitivity of the receptor) to impaired awareness of hypoglycaemia in type 1 diabetes (Schouwenberg et al. 2008). Although no data have been provided yet on the association between this variant and frequency of hypoglycaemia, these data and the Danish studies raise the intriguing prospect that there may be a range of genetic factors that influ­ ence susceptibility to hypoglycaemia. Their identification may help to stratify risk in individuals with type 1 diabetes and may ultimately lead to the development of novel therapeutic interventions to prevent or ameliorate the impact of hypoglycaemia.

Absent endogenous insulin secretion Several studies have demonstrated the importance of endogenous insulin secretion in defin­ ing risk of hypoglycaemia. Individuals who are C-peptide negative, i.e. who have no endogenous insulin production, have an approximately 2 to 4-fold increased risk of severe hypoglycaemia compared to people with detectable C-peptide (Bott et al. 1997; The Dia­ betes Control and Complications Trial Research Group 1997; Muhlhauser et al. 1998; Pedersen-Bjergaard et al. 2001). These data mirror the clinical experience that severe hypoglycaemia is rare in the 12 months after the diagnosis of type 1 diabetes (Davis et al. 1997), when significant measurable concentrations of endogenous insulin are present. The presumption is that the ability of endogenous insulin levels to be suppressed in the face of a declining blood glucose concentration provides an additional measure of protection in the defence against hypoglycaemia and thus reduces overall risk. In addition, at this stage of early type 1 diabetes, glucagon secretion in response to hypoglycaemia is still intact and will also protect against progression to severe hypoglycaemia.

Dose and type of insulin Senior diabetes specialists will not forget the controversy in the mid-1980s that surrounded the introduction of human insulin. A vocal minority of people with type 1 diabetes claimed that the change from animal-derived to human insulin was associated with a loss of warning symptoms of hypoglycaemia and an increased risk of severe hypoglycaemia. Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 87

Their claims were subjected to careful scientific scrutiny, and ultimately a systematic review of the evidence found no evidence to support the premise that treatment with human, as opposed to animal, insulin was associated with an increased risk of hypoglycae­ mia (Airey et al. 2000). Since that time, several insulin analogues have been developed and their introduction into clinical practice has been accompanied by the publication of studies that purport to demonstrate that use of the insulin analogues is associated with a lower frequency of hypoglycaemia, in the face of stable or improved glycaemic control (Anderson et al. 1997; Garg et al. 2004; Hermansen et al. 2004). However, systematic reviews have suggested that neither short-acting (Siebenhofer et al. 2004) nor long-acting (Warren et al. 2004) insulin analogues are associated with lower rates of hypoglycaemia that are clinically significant. The usual caveats of such clinical trials apply in terms of patient characteristics and recording of hypoglycaemia. In one observational study from Colorado, the frequency of severe hypoglycaemia rose in clinic patients in the immediate aftermath of the DCCT as efforts to intensify insulin therapy were instituted, but from 1996 onwards, the rates of severe hypoglycaemia declined and the authors linked this to the introduction of insulin lispro (Chase et al. 2001). This study was subject to the effects of numerous confounders, but further data from routine clinical practice are required to help clarify the impact of insulin analogues on risk of hypoglycaemia. Higher doses of insulin have also been associated with an increased risk of hypoglycae­ mia in some studies (The Diabetes Control and Complications Trial Research Group 1997; ter Braak et al. 2000) but not in others (Table 4.3). Several possible explanations may account for this relationship – high insulin doses may be a sign of less endogenous insulin production, or of efforts to achieve strict glycaemic control. It may also reflect suboptimal compliance with insulin therapy and so indicate a pattern of behaviour that predisposes to marked fluctuations in glucose control.

Sleep Nocturnal hypoglycaemia is very common (Allen and Frier 2003). In the DCCT, 43% of episodes of severe hypoglycaemia occurred between midnight and 08:00 h, and 55% of episodes occurred when individuals were asleep (The DCCT Research Group 1991). The potential for nocturnal hypoglycaemia engenders significant anxiety among people taking insulin, particularly in those who live alone. As a result, many maintain higher blood glucose concentrations at bedtime to reduce the risk of nocturnal hypoglycaemia. Hypogly­ caemia appears to be more common at night because counterregulatory hormone responses are blunted during sleep in people with type 1 diabetes (Jones et al. 1998; Banarer and Cryer 2003). Moreover, hypoglycaemia awareness is reduced during sleep (Banarer and Cryer 2003), and so ultimately sleep impairs both the physiological and behavioural responses to hypoglycaemia. Unsurprisingly, considerable effort has been directed at devel­ oping strategies to reduce the risk of nocturnal hypoglycaemia (see Chapter 5).

Microvascular complications In a retrospective study of 44 patients with type 1 diabetes with impaired renal function (serum creatinine ≥ 133 umol/l and proteinuria), the incidence of severe hypoglycaemia was five times higher than in matched subjects with normal renal function (Muhlhauser et al. 1991). In other studies, severe hypoglycaemia was linked with nephropathy (ter Braak et al. 88 Hypoglycaemia in Clinical Diabetes

2000), peripheral neuropathy and retinopathy (Pedersen-Bjergaard et al. 2004). Although it is generally recognised that insulin requirement declines in advanced renal disease, with reduced clearance of insulin, this association between hypoglycaemia and nephropathy (and other microvascular complications) could be confounded by other factors such as concomi­ tant drug therapy (ter Braak et al. 2000) and the coexistence of acquired hypoglycaemia syndromes which, like microvascular complications, are linked with increasing duration of diabetes. However, renal impairment is a recognised risk factor for severe hypoglycaemia. The role of peripheral autonomic neuropathy in increasing the risk of severe hypogly­ caemia has been considered in several studies and deserves mention. In the EURODIAB IDDM Complications Study, the presence of abnormal cardiovascular reflexes was associ­ ated with a 1.7-fold increased risk of severe hypoglycaemia (Stephenson et al. 1996). Gold et al. (1997) also demonstrated that autonomic neuropathy was associated with a small increased risk of severe hypoglycaemia in 60 patients with type 1 diabetes, but no such relationship was demonstrated in the DCCT (The DCCT Research Group 1991). The mechanism underlying this association remains to be fully elucidated. Peripheral autonomic neuropathy often coexists with impaired hypoglycaemia awareness in patients with type 1 diabetes, presumably because both conditions are associated with diabetes of long duration (Hepburn et al. 1990), but impaired awareness can readily occur in the absence of periph­ eral autonomic neuropathy (Hepburn et al. 1990; Ryder et al. 1990; Bacatselos et al. 1995). Severe autonomic neuropathy is associated with gastroparesis, which can cause marked swings in blood glucose concentration. Although severe gastroparesis is now rare, delayed gastric emptying is relatively common (Kong et al. 1996) and may explain at least part of the association between hypoglycaemia and autonomic dysfunction.

Social and psychological factors Psychological factors clearly play a crucial role in determining an individual’s likelihood of developing severe hypoglycaemia (see Chapter 16). Low mood (Gonder-Frederick and Cox 1997), poor emotional coping (Bott et al. 1997) and low socio-economic status (Muhl­ hauser et al. 1998) have been linked to risk of severe hypoglycaemia and so have other more straightforward behavioural factors such an individual’s propensity to carry a supply of carbohydrate for emergency use (Bott et al. 1997) and their determination to achieve normoglycaemia (Muhlhauser et al. 1998). Cox and his colleagues have explored the psychological impact of hypoglycaemia on people with type 1 diabetes. They developed a Fear of Hypoglycaemia scale, which quanti­ fies the anxieties that people with type 1 diabetes have with respect to hypoglycaemia and the extent to which individuals take steps to avoid experiencing such episodes (Cox et al. 1987). Unsurprisingly, there is a close association between fear of hypoglycaemia and perceived risk of future severe hypoglycaemia (Gonder-Frederick et al. 1997). In many instances this is appropriate, in that ‘fear’ ratings are often high in people who have impaired awareness of hypoglycaemia and/or have experienced multiple episodes in the past (Gold et al. 1997). However, in other people, fear of hypoglycaemia may be high while absolute risk is low – such individuals often display high levels of trait anxiety or have had a traumatic previous experience of hypoglycaemia. It is often extremely difficult to persuade such individuals to maintain strict glycaemic control. Conversely, there are people who have a very low fear of hypoglycaemia, despite a propensity to recurrent epi­ sodes. Such individuals may fail to take appropriate precautionary measures, thereby putting themselves and others at risk if hypoglycaemia occurs. Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 89

Endocrinopathies Endocrine disorders, such as Addison’s disease and hypopituitarism, which are associated with a deficiency of counterregulatory hormones, can be associated with an increased risk of hypoglycaemia in adults with type 1 diabetes. These conditions are uncommon in every­ day diabetes practice, but clinicians should maintain a high index of suspicion particularly in the patient who simultaneously demonstrates a marked and otherwise unexplained decline in insulin requirements.

Other risk factors Other possible risk factors for hypoglycaemia include smoking as a relatively novel marker of increased risk (Pedersen-Bjergaard et al. 2004), although this may be confounded by other diabetes-related and lifestyle factors, e.g. regular use of alcohol (ter Braak et al. 2000). In the DCCT, men and adolescents were at higher risk (The Diabetes Control and Complications Trial Research Group 1997), but these associations have not been replicated with any consistency in other studies (Table 4.3). Advanced age has also been associated with a higher frequency of severe hypoglycaemia (Schütt et al. 2012), but clearly there are numerous potential confounders in that older people tend to have a longer duration of diabetes, more complications and co-morbidities and lower usage of MDI and CSII thera­ pies. Risks associated with pregnancy are addressed in Chapter 11.

Causes and risk factors for hypoglycaemia: summary and conclusions Many of the risk factors described above are interrelated and any given individual may have more than one, which clearly will increase their overall risk. However, at its most fundamental level, hypoglycaemia in adults with type 1 diabetes is a consequence of the inability of exogenous insulin levels to fall in response to a declining blood glucose con­ centration. In people who have had type 1 diabetes for several years, the situation is exacerbated because of a failure of the normal physiological counterregulatory defence mechanisms, which in non-diabetic individuals serve to increase exogenous glucose pro­ duction and generate warning symptoms. This counterregulatory failure worsens with increasing duration of diabetes, particularly if glycaemic control is strict and the indi­ vidual has experienced recurrent episodes of severe hypoglycaemia. Other factors may come in to play, for example particular behavioural patterns and the time-action profiles of the exogenous insulin. There may be inherent genetic susceptibility to hypoglycaemia, but this needs to be affirmed in larger populations, and the underlying mechanisms more clearly dissected. Novel strategies to reduce the overall risk of hypoglycaemia are urgently required.

CONCLUSIONS • There are no definitive criteria for what constitutes hypoglycaemia, but most specialists distinguish mild from severe episodes depending on whether or not the individual is able to self-treat. 90 Hypoglycaemia in Clinical Diabetes • The ‘average’ adult with type 1 diabetes will experience many thousands of episodes of mild, symptomatic hypoglycaemia over a lifetime, with a typical frequency of one to two episodes per week. Asymptomatic hypoglycaemia is even more common. • Severe hypoglycaemia is less common and on average occurs once or twice every year, with an annual prevalence of around 30%. However, the distribution is heavily skewed, such that many individuals are unaffected, while a small number experience recurrent episodes. • Severe hypoglycaemia requiring treatment with intra-muscular glucagon and/or intra­ venous glucose is even less common, and the majority of all episodes of hypoglycaemia are managed in the community by the patient and/or relatives and friends, without recourse to emergency services. • Hypoglycaemia occurs when there is an imbalance between insulin-mediated glucose disposal and glucose influx into the circulation from the liver, and exogenous car­ bohydrate. Typical precipitants include patient error in insulin dosage, alcohol and exercise. • Plasma concentrations of exogenous insulin cannot decline in response to falling blood glucose and the time-action profiles of current insulins do not accurately mimic the normal physiological variation in insulin. In a high proportion of cases, no underlying precipitant of a given episode of hypoglycaemia can be identified. • The DCCT and other intervention studies have provided substantial information of the epidemiology of severe hypoglycaemia, but individuals participating in such studies may not be representative of the wider population who have type 1 diabetes. • A major determinant of increased risk of severe hypoglycaemia is central failure of normal responses to a falling blood glucose, which may be a direct consequence of exposure to recurrent episodes of hypoglycaemia. This is a feature of individuals with long-standing diabetes, intensive insulin therapy and strict glycaemic control. • There may be specific genetic factors that predispose to an increased riskof hypoglycaemia.

REFERENCES

Airey, C.M., Williams, D.R.R., Martin, P.G., Bennett, C.M.T. & Spoor, P.A. (2000) Hypoglycaemia induced by exogenous insulin – ‘human’ and animal insulin compared. Diabetic Medicine, 17, 416–432. Allen, C., LeClaire, T., Palta, M., Daniels, K., Meredith, M., D’Alessio, D.J. & for the Wisconsin Diabetes Registry Project (2001) Risk factors for frequent and severe hypoglycemia in type 1 diabetes. Diabetes Care, 24, 1878–1881. Allen, K.V. & Frier, B.M. (2003) Nocturnal hypoglycemia: clinical manifestations and therapeutic strate­ gies toward prevention. Endocrine Practice, 9, 530–543. American Diabetes Association Workgroup on Hypoglycemia (2005) Defining and reporting hypoglycemia in diabetes. Diabetes Care, 28, 1245–1249. Anderson, J.H., Brunelle, R.L., Koivisto, V.A. et al.; The Multicenter Insulin Lispro Study Group (1997) Reduction of postprandial hyperglycemia and frequency of hypoglycemia in IDDM patients on insulin- analog treatment. Diabetes, 46, 265–270. Arky, R.A., Veverbrants, E. & Abramson, E.A. (1968) Irreversible hypoglycemia. A complication of alcohol and insulin. Journal of the American Medical Association, 206, 575–578. Avogaro, A., Beltramello, P., Gnudi, L. et al. (1993) Alcohol intake impairs glucose counterregulation during acute insulin-induced hypoglycemia in IDDM patients. Evidence for a critical role of free fatty acids. Diabetes, 42, 1626–1634. Bacatselos, S.O., Karamitsos, D.T., Kourtoglou, G.I., Zambulis, C.X., Yovos, J.G. & Vyzantiadis, A.T. (1995) Hypoglycaemia unawareness in type 1 diabetic patients under conventional insulin treatment. Diabetes, Nutrition and Metabolism, 8, 267–275. Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 91

Banarer, S. & Cryer, P.E. (2003) Sleep-related hypoglycemia-associated autonomic failure in type 1 dia­ betes. Diabetes, 52, 1195–1203. Bolli, G., De Feo, P., Compagnucci, P. et al. (1983) Abnormal glucose counterregulation in insulin- dependent diabetes mellitus. Interaction of anti-insulin antibodies and impaired glucagon and epine­ phrine secretion. Diabetes, 32, 134–141. Bott, S., Bott, U., Berger, M. & Muhlhauser, I. (1997) Intensified insulin therapy and the risk of severe hypoglycaemia. Diabetologia, 40, 926–932. Boyle, P.J., Schwartz, N.S., Shah, S.D., Clutter, W.E. & Cryer, P.E. (1988) Plasma glucose concentrations at the onset of hypoglycemic symptoms in patients with poorly controlled diabetes and in non-diabetics. New England Journal of Medicine, 318, 1487–1492. Boyle, P.J., Kempers, S.F., O’Connor, A.M. & Nagy, R.J. (1995) Brain glucose uptake and unawareness of hypoglycemia in patients with insulin-dependent diabetes mellitus. New England Journal of Medicine, 333, 1726–1731. Bulsara, M.K., Holman, C.D.J., van Bockxmeer, F.M. et al. (2007) The relationship between ACE genotype and risk of severe hypoglycemia in a large population-based cohort of children and adolescents with type 1 diabetes. Diabetologia, 50, 965–971. Chase, H.P., Lockspeiser, T., Peery, B. et al. (2001) The impact of the Diabetes Control and Complications Trial and Humalog insulin on glycohemoglobin levels and severe hypoglycemia in type 1 diabetes. Diabetes Care, 24, 430–434. Choudhary, P., Davies, C., Emery, C., Freeman, J. & Heller, S.R. (2006) Low interstitial glucose detected by CGMS correlates well with clinically significant hypoglycaemia. Diabetic Medicine, 23, S12 (abstract). Choudhary, P., Geddes, J., Freeman, J.V., Emery, C.J., Heller, S.R. & Frier, B.M. (2010) Frequency of biochemical hypoglycaemia in adults with Type 1 diabetes with and without impaired awareness of hypoglycaemia: no identifiable differences using continuous glucose monitoring.Diabetic Medicine, 27, 666–672. Clarke, W.L., Cox, D.J., Gonder-Frederick, L.A., Julian, D., Schlundt, D. & Polonsky, W. (1995) Reduced awareness of hypoglycemia in adults with IDDM. A prospective study of hypoglycemic frequency and associated symptoms. Diabetes Care, 18, 517–522. Cox, D.J., Irvine, A., Gonder-Frederick, L., Nowacek, G. & Butterfield, J. (1987) Fear of hypoglycemia: quantification, validation and utilization. Diabetes Care, 10, 617–621. Cox, D.J., Kovatchev, B.P., Julian, D.M. et al. (1994) Frequency of severe hypoglycemia in insulin- dependent diabetes mellitus can be predicted from self-monitoring blood glucose data. Journal of Clini- cal Endocrinology and Metabolism, 79, 1659–1662. Cryer, P.E., Davis, S.N. & Shamoon, H. (2003) Hypoglycemia in diabetes. Diabetes Care, 26, 1902–1912. Dagogo-Jack, S.E., Craft, S. & Cryer, P.E. (1993) Hypoglycemia-associated autonomic failure in insulin- dependent diabetes mellitus. Recent antecedent hypoglycemia reduces autonomic responses to, symp­ toms of, and defense against subsequent hypoglycemia. The Journal of Clinical Investigation, 91, 819–828. Davis, E.A., Keating, B., Byrne, G.C., Russell, M. & Jones, T.W. (1997) Hypoglycemia: incidence and clinical predictors in a large population-based sample of children and adolescents with IDDM. Diabetes Care, 20, 22–25. DAFNE Study Group (2002) Training in flexible, intensive insulin management to enable dietary freedom in people with type 1 diabetes: dose adjustment for normal eating (DAFNE) randomised controlled trial. British Medical Journal, 325, 746. Donnelly, L.A., Morris, A.D., Frier, B.M. et al.; for the DARTS/MEMO Collaboration (2005) Frequency and predictors of hypoglycaemia in type 1 and insulin-treated type 2 diabetes: a population-based study. Diabetic Medicine, 22, 749–755. Feher, M.D., Grout, P., Kennedy, A., Elkeles, R.S. & Touquet, R. (1989) Hypoglycaemia in an inner-city accident and emergency department: a 12-month survey. Archives of Emergency Medicine, 6, 183–188. Galassetti, P., Mann, S., Tate, D. et al. (2001a) Effects of antecedent prolonged exercise on subsequent counterregulatory responses to hypoglycemia. American Journal of Endocrinology, 280, E908–E917. Galassetti, P., Neill, A.R., Tate, D., Ertl, A.C., Wasserman, D.H. & Davis, S.N. (2001b) Sexual dimorphism in counterregulatory responses to hypoglycemia after antecedent exercise. Journal of Clinical Endo- crinology and Metabolism, 86, 3516–3524. Galassetti, P., Tate, D., Neill, R.A., Morrey, S., Wasserman, D.H. & Davis, S.N. (2003) Effect of antecedent hypoglycemia on counterregulatory responses to subsequent euglycemic exercise in type 1 diabetes. Diabetes, 52, 1761–1769. 92 Hypoglycaemia in Clinical Diabetes

Garg, S.K., Gottlieb, P.A., Hisatomi, M.E. et al. (2004) Improved glycemic control without an increase in severe hypoglycemic episodes in intensively treated patients with type 1 diabetes receiving morning, evening, or split dose insulin glargine. Diabetes Research and Clinical Practice, 66, 49–56. Geddes, J., Schopman, J.E., Zammitt, N.N. & Frier, B.M. (2008) Prevalence of impaired awareness of hypoglycaemia in adults with Type 1 diabetes. Diabetic Medicine, 25, 501–504. Gold, A.E., MacLeod, K.M. & Frier, B.M. (1994) Frequency of severe hypoglycemia in patients with type 1 diabetes with impaired awareness of hypoglycemia. Diabetes Care, 17, 697–703. Gold, A.E., Frier, B.M., MacLeod, K.M. & Deary, I.J. (1997) A structural equation model for predictors of severe hypoglycaemia in patients with insulin-dependent diabetes mellitus. Diabetic Medicine, 14, 309–315. Gonder-Frederick, L., Clarke, W.L. & Cox, D.J. (1997) The emotional, social, and behavioral implications of insulin-induced hypoglycemia. Seminars in Clinical Neuropsychiatry, 2, 57–65. Gonder-Frederick, L.A. & Cox, D. (1997) The relationship between psychological depression and fre­ quency of severe hypoglycemia (SH) in IDDM. Diabetes, 46, 268A. (Abstract). Graveling, A.J. & Frier, B.M. (2010) Impaired awareness of hypoglycaemia: a review. Diabetes and Metabolism, 36 (Suppl. 2), S64–S74. Hart, S.P. & Frier, B.M. (1998) Causes, management and morbidity of acute hypoglycaemia in adults requiring hospital admission. The Quarterly Journal of Medicine, 91, 505–510. Heller, S.R. & Cryer, P.E. (1991) Reduced neuroendocrine and symptomatic responses to subsequent hypoglycemia after 1 episode of hypoglycemia in non-diabetic humans. Diabetes, 40, 223–226. Heller, S.R. & Macdonald, I.A. (1996) The measurement of cognitive function during acute hypoglycaemia: experimental limitations and their effect on the study of hypoglycaemia unawareness. Diabetic Medicine, 13, 607–615. Hepburn, D.A., Patrick, A.W., Eadington, D.W., Ewing, D.J. & Frier, B.M. (1990) Unawareness of hypogly­ caemia in insulin-treated diabetic patients: prevalence and relationship to autonomic neuropathy. Dia- betic Medicine, 7, 711–717. Herings, R.M.C., de Boer, A., Stricker, B.H.C., Leufkens, H.G.M. & Porsius, A. (1996) Hypoglyc­ emia associated with the use of inhibitors of angiotensin-converting enzyme. Lancet, 345, 1195– 1198. Hermansen, K., Fontaine, P., Kukolja, K.K., Peterkova, V., Leth, G. & Gall, M.-A. (2004) Insulin analogues (insulin detemir and insulin aspart) versus traditional human insulins (NPH insulin and regular human insulin) in basal-bolus therapy for patients with type 1 diabetes. Diabetologia, 47, 622–629. Hirsch, B.R. & Shamoon, H. (1987) Defective epinephrine and growth hormone responses in type 1 dia­ betes are stimulus specific. Diabetes, 36, 20–26. Hopkins, D., Lawrence, I., Mansell, P., Thompson, G., Amiel, S., Campbell, M. et al. (2012) Improved biomedical and psychological outcomes 1 year after structured education in flexible insulin therapy for people with type 1 diabetes: the U.K. DAFNE experience. Diabetes Care, 35, 1638–1642. Janssen, M.M.J., Snoek, F.J., de Jongh, R.T., Casteleijn, S., Deville, W. & Heine, R.J. (2000) Biological and behavioural determinants of the frequency of mild, biochemical hypoglycaemia in patients with Type 1 diabetes on multiple insulin injection therapy. Diabetes/Metabolism Research and Reviews, 16, 157–163. Johannesen, J., Svensson, J., Berqholdt, R. et al. (2011) Hypoglycemia, S-ACE and ACE genotypes in a Danish nationwide population of children and adolescents with Type 1 diabetes. Pediatric Diabetes, 12, 100–106. Johnson, E.S., Koepsell, T.D., Reiber, G., Stergachis, A. & Platt, R. (2002) Increasing incidence of serious hypoglycemia in insulin users. Journal of Clinical Epidemiology, 55, 253–259. Jones, T.W., Porter, P., Sherwin, R.S. et al. (1998) Decreased epinephrine responses to hypoglycemia during sleep. New England Journal of Medicine, 338, 1657–1662. Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group (2010) Prolonged nocturnal hypoglycemia is common during 12 months of continuous glucose monitoring in children and adults with Type 1 diabetes. Diabetes Care, 33, 1004–1008. Kerr, D., Macdonald, I.A., Heller, S.R. & Tattersall, R.B. (1990) Alcohol causes hypoglycaemic unaware­ ness in healthy volunteers and patients with type 1 (insulin-dependent) diabetes. Diabetologia, 33, 216–221. Kinsley, B.T., Widom, B. & Simonson, D.C. (1995) Differential regulation of counterregulatory hormone secretion and symptoms during hypoglycemia in IDDM. Diabetes Care, 18, 17–26. Kong, M.-F., Macdonald, I.A. & Tattersall, R.B. (1996) Gastric emptying in diabetes. Diabetic Medicine, 13, 112–119. Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 93

Leckie, A.M., Graham, M.K., Grant, J.B., Ritchie, P.J. & Frier, B.M. (2005) Frequency, severity, and morbidity of hypoglycemia occurring in the workplace in people with insulin-treated diabetes. Diabetes Care, 28, 1333–1338. Leese, G.P., Wang, J., Broomhall, J. et al.; for the DARTS/MEMO Collaboration (2003) Frequency of severe hypoglycemia requiring emergency department treatment in type 1 and type 2 diabetes. A population-based study of health service resource use. Diabetes Care, 26, 1176–1180. MacLeod, K.M., Hepburn, D.A. & Frier, B.M. (1993) Frequency and morbidity of severe hypoglycaemia in insulin-treated diabetic patients. Diabetic Medicine, 10, 238–245. MacLeod, K.M., Deary, I.J., Graham, K.S., Hepburn, D.A. & Frier, B.M. (1994) Hypoglycaemia unaware­ ness in adult patients with type 1 diabetes: relationship to severe hypoglycaemia and cognitive impair­ ment. Diabetes, Nutrition and Metabolism, 7, 205–211. MacLeod, K.M., Gold, A.E. & Frier, B.M. (1995) Frequency, severity and symptomatology of hypogly­ caemia: a comparative trial of human and porcine insulins in type 1 diabetic patients. Diabetic Medicine, 12, 134–141. McIntyre, H.D., Knight, B.A., Harvey, D.M., Noud, M.N., Hagger, V.L. & Gilshenan, K.S. (2010) Dose adjustment for normal eating (DAFNE) – an audit of outcomes in Australia. The Medical Journal of Australia, 192, 637–640. Morris, A.D., Boyle, D.I.R., McMahon, A.D. et al.; for the DARTS/MEMO Collaboration (1997) ACE inhibitor use is associated with hospitalization for severe hypoglycemia in patients with diabetes. Dia- betes Care, 20, 1363–1367. Moses, R.G., Hubert, P.A. & Lewis-Driver, D.J. (1985) Severe hypoglycaemia. A one-year prospective survey in Wollongong. The Medical Journal of Australia, 142, 294–296. Muhlhauser, I., Berger, M., Sonnenberg, G. et al. (1985) Incidence and management of severe hypoglyc­ emia in 434 adults with insulin-dependent diabetes mellitus. Diabetes Care, 8, 268–273. Muhlhauser, I., Bruckner, I., Berger, M. et al. (1987) Evaluation of an intensified insulin treatment and teaching programme as routine management of Type 1 (insulin-dependent) diabetes. The Bucharest- Düsseldorf Study. Diabetologia, 30, 681–690. Muhlhauser, I., Toth, G., Sawicki, P.T. & Berger, M. (1991) Severe hypoglycemia in type 1 diabetic patients with impaired kidney function. Diabetes Care, 14, 344–346. Muhlhauser, I., Overmann, H., Bender, R., Bott, U. & Berger, M. (1998) Risk factors for severe hypogly­ caemia in adult patients with type 1 diabetes. Diabetologia, 41, 1274–1282. Nordfeldt, S. & Samuelsson, U. (2003) Serum ACE predicts severe hypoglycemia in children and adoles­ cents with type 1 diabetes. Diabetes Care, 26, 274–278. O’Neill, S. (1997) Make four the floor. Balance (Diabetes UK), Jan/Feb, 20–23. Owens, D.R., Zinman, B. & Bolli, G.B. (2001) Insulins today and beyond. Lancet, 358, 739–746. Pampanelli, S., Fanelli, C., Lalli, C. et al. (1996) Long-term intensive insulin therapy in IDDM: effects on

HbA1c, risk for severe and mild hypoglycaemia, status of counterregulation and awareness of hypogly­ caemia. Diabetologia, 39, 677–686. Pedersen-Bjergaard, U., Agerholm-Larsen, B., Pramming, S., Hougaard, P. & Thorsteinsson, B. (2001) Activity of angiotensin-converting enzyme and risk of severe hypoglycaemia in type 1 diabetes mellitus. Lancet, 357, 1248–1253. Pedersen-Bjergaard, U., Agerholm-Larsen, B., Pramming, S., Hougaard, P. & Thorsteinsson, B. (2003a) Prediction of severe hypoglycaemia by angiotensin-converting enzyme activity and genotype in type 1 diabetes. Diabetologia, 46, 89–96. Pedersen-Bjergaard, U., Pramming, S. & Thorsteinsson, B. (2003b) Recall of severe hypoglycaemia and self-estimated state of awareness in type 1 diabetes. Diabetes/Metabolism Research and Reviews, 19, 232–240. Pedersen-Bjergaard, U., Pramming, S., Heller, S.R. et al. (2004) Severe hypoglycaemia in 1076 adult patients with type 1 diabetes: influence of risk markers and selection. Diabetes/Metabolism Research and Reviews, 20, 479–486. Pedersen-Bjergaard, U., Reubsaet, J.L.E., Nielsen, S.L. et al. (2005) Psychoactive drugs, alcohol, and severe hypoglycemia in insulin-treated diabetes: analysis of 141 cases. The American Journal of Medi- cine, 118, 307–310. Plank, J., Kohler, G., Rakovac, I. et al. (2004) Long-term evaluation of a structured out-patient education programme for intensified insulin therapy in patients with type 1 diabetes: a 12 year follow-up. Diabeto- logia, 47, 1370–1375. Potter, J., Clarke, P., Gale, E.A.M., Dave, S.H. & Tattersall, R.B. (1982) Insulin-induced hypoglycaemia in an accident and emergency department: the tip of an iceberg? Br Med J, 285, 1180–1182. 94 Hypoglycaemia in Clinical Diabetes

Pramming, S., Thorsteinsson, B., Bendtson, I. & Binder, C. (1990) The relationship between symptomatic and biochemical hypoglycaemia in insulin-dependent diabetic patients. Journal of Internal Medicine, 228, 641–646. Pramming, S., Thorsteinsson, B., Bendtson, I. & Binder, C. (1991) Symptomatic hypoglycaemia in 411 type 1 diabetic patients. Diabetic Medicine, 8, 217–222. Reichard, P. & Pihl, M. (1994) Mortality and treatment side-effects during long-term intensified conventional insulin treatment in the Stockholm Diabetes Intervention Study. Diabetes, 43, 313– 317. Ryder, R.E.J., Owens, D.R., Hayes, T.M., Ghatei, M.A. & Bloom, S.R. (1990) Unawareness of hypogly­ caemia and inadequate hypoglycaemic counterregulation: no causal relationship with diabetic autonomic neuropathy. British Medical Journal, 301, 783–787. Schouwenberg, B.J., Veldman, B.A., Spiering, W. et al. (2008) The Arg16Gly variant of the beta2- adrenergic receptor predisposes to hypoglycemia unawareness in type 1 diabetes mellitus. Pharmaco- genetics and Genomics, 18, 369–372. Schütt, M., Fach, E.-M., Seufert, J. et al. (2012) Multiple complications and frequent severe hypoglycaemia in ‘elderly’ and ‘old’ patients with Type 1 diabetes. Diabetic Medicine, 29, e176–e179. Service, F.J. (1995) Hypoglycemic disorders. New England Journal of Medicine, 332, 1144–1152. Siebenhofer, A., Plank, J., Berghold, A. et al. (2004) Meta-analysis of short-acting insulin analogues in adult patients with type 1 diabetes: continuous insulin infusion versus injection therapy. Diabetologia, 47, 1895–1905. Sonnenberg, G.E., Kemmer, F.W. & Berger, M. (1990) Exercise in type 1 (insulin-dependent) diabetic patients treated with continuous subcutaneous insulin infusion. Prevention of exercise induced hypogly­ caemia. Diabetologia, 33, 696–703. Stephenson, J.M., Kempler, P., Cavallo Perin, P., Fuller, J.H. & the EURODIAB IDDM Complications Study Group (1996) Is autonomic neuropathy a risk factor for severe hypoglycaemia? The EURODIAB IDDM Complications Study. Diabetologia, 39, 1372–1376. Swinnen, S.G.H.A., Mullins, P., Miller, M., Hoekstra, J.B.L. & Holleman, F. (2009) Changing the glucose cut-off values that define hypoglycaemia has a major effect on reported frequencies of hypoglycaemia. Diabetologia, 52, 38–41. ter Braak, E.W.M.T., Appelman, A.M.M.F., van de Laak, M.F., Stolk, R.P., Van Haeften, T.W. & Erkelens, D.W. (2000) Clinical characteristics of type 1 diabetic patients with and without severe hypoglycemia. Diabetes Care, 23, 1467–1471. The DCCT Research Group (1991) Epidemiology of severe hypoglycemia in the Diabetes Control and Complications Trial. The American Journal of Medicine, 90, 450–459. The Diabetes Control and Complications Trial Research Group (1993) The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. New England Journal of Medicine, 329, 977–986. The Diabetes Control and Complications Trial Research Group (1997) Hypoglycemia in the Diabetes Control and Complications Trial. Diabetes, 46, 271–286. The EURODIAB IDDM Complications Study Group (1994) Microvascular and acute complications in IDDM patients: the EURODIAB IDDM Complications Study. Diabetologia, 37, 278–285. Thorsteinsson, B., Pramming, S., Lauritzen, T. & Binder, C. (1986) Frequency of daytime biochemical hypoglycaemia in insulin-treated diabetic patients: relation to daily median blood glucose concentrations. Diabetic Medicine, 3, 147–151. Turner, B.C., Jenkins, E., Kerr, D., Sherwin, R.S. & Cavan, D.A. (2001) The effect of evening alcohol consumption on next-morning glucose control in type 1 diabetes. Diabetes Care, 24, 1888–1893. U.K. Hypoglycaemia Study Group (2007) Risk of hypoglycaemia in types 1 and 2 diabetes: effects of treatment modalities and their duration. Diabetologia, 50, 1140–1147. Veneman, T., Mitrakou, A., Mokan, M., Cryer, P. & Gerich, J. (1993) Induction of hypoglycemia unaware­ ness by asymptomatic nocturnal hypoglycemia. Diabetes, 42, 1233–1237. Vervoort, G., Goldschmidt, H.M.G. & van Doorn, L.G. (1996) Nocturnal blood glucose profiles in patients with type 1 diabetes mellitus on multiple (>4) daily insulin injection regimens. Diabetic Medicine, 13, 794–799. Warren, E., Weatherley-Jones, E., Chilcott, J. & Beverley, C. (2004) Systematic review and economic evaluation of a long-acting insulin analogue, insulin glargine. Health Technology Assessment, 8 (45), iii, 1–57. Chapter 4: Frequency, Causes and Risk Factors for Hypoglycaemia in Type 1 Diabetes 95

White, N.H., Skor, D.A., Cryer, P.E., Levandoski, L.A., Bier, D.M. & Santiago, J.V. (1983) Identification of type 1 diabetic patients at increased risk for hypoglycemia during intensive therapy. New England Journal of Medicine, 308, 485–491. Widom, B. & Simonson, D.C. (1992) Intermittent hypoglycemia impairs glucose counterregulation. Dia- betes, 41, 1335–1340. Yeh, H.C., Brown, T.T., Maruthur, N. et al. (2012) Comparative effectiveness and safety of methods of insulin delivery and glucose monitoring for diabetes mellitus: a systematic review and meta-analysis. Archives of Internal Medicine, 157, 336–347. Zammitt, N.N., Geddes, J., Warren, R.E., Marioni, R., Ashby, J.P. & Frier, B.M. (2007) Serum angiotensin- converting enzyme and frequency of severe hypoglycaemia in type 1 diabetes: does a relationship exist? Diabetic Medicine, 24, 1449–1454. 5 Nocturnal Hypoglycaemia Elaine Chow and Simon R. Heller University of Sheffield, Sheffield, UK

INTRODUCTION

People with type 1 diabetes worry as much about hypoglycaemia as they do about the prospect of developing serious diabetic complications, and episodes occurring during sleep are feared more than at any other time of day. Several studies have confirmed how com- monly hypoglycaemia occurs during sleep, and the risk of severe nocturnal hypoglycaemia can drive a parent to spend every night in their child’s bedroom for many years. Nocturnal hypoglycaemia may also have consequences beyond a domestic emergency. Asymptomatic episodes may contribute to the development of impaired hypoglycaemia awareness and deficient counterregulation, while recurrent nocturnal hypoglycaemia may be associated with subsequent cognitive impairment and is also implicated in the ‘dead in bed’ syndrome (see Chapter 13) with sudden death occurring during sleep in young people with type 1 diabetes. In this chapter, the reasons why nocturnal hypoglycaemia occurs so frequently in type 1 diabetes are discussed and its frequent occurrence in insulin-treated type 2 diabetes is highlighted, particularly in those who have counterregulatory deficiencies (Chico et al. 2003; Weber et al. 2007). Clinical approaches are described that may help to reduce its frequency and severity and some controversial topics, such as the Somogyi phenomenon, are discussed.

EPIDEMIOLOGY – HOW COMMON IS NOCTURNAL HYPOGLYCAEMIA?

Nocturnal hypoglycaemia has always been of major concern to individuals with diabetes, with many experiencing severe episodes. In the Diabetes Control and Complications Trial, half the severe hypoglycaemic episodes occurred during sleep (The Diabetes Control and Complications Trial Research Group 1991). The frequency of nocturnal hypoglycaemia is difficult to estimate given that a large proportion of episodes are asymptomatic. Early studies required direct observation in a clinical research unit or in a person’s home. In one of the earliest studies, people with type 1 diabetes on routine insulin therapy were admitted

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. Chapter 5: Nocturnal Hypoglycaemia 97 for overnight monitoring using intermittent venous blood sampling (Gale and Tattersall 1979). Nocturnal hypoglycaemia was reported in 22 of 39 adults who were being treated with one or two injections of insulin each day. Subsequent reports have estimated that the prevalence of those experiencing a blood glucose below 3 mmol/l is between 14 and 56% (Table 5.1). The studies that reported the lowest rates of hypoglycaemia sampled blood less frequently, so some asymptomatic episodes may not have been identified. Early studies generally included hospitalised patients who were often in poor glycaemic control. It seems likely that the move to more intensive insulin regimens and targets may have increased the rates of nocturnal hypoglycaemia, particularly in children. In one study of children treated with multiple injection therapy, rates of nocturnal hypoglycaemia exceeded an alarming 50% (Beregszaszi et al. 1997). Children have been directly observed to sleep through most nocturnal episodes despite having blood glucose values <2 mmol/l (Matyka et al. 1999a) (Figure 5.1). The introduction of continuous glucose monitoring (CGM) with glucose measurements every few minutes has permitted more frequent recording in individuals during sleep. These approaches have suggested that hypoglycaemic events may be even more common than was appreciated previously (Table 5.1) although the accuracy of CGM is uncertain. In one large study, 167 children and adults with type 1 diabetes on multiple daily injections or insulin pump therapy, were monitored for a year (Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group 2010). Virtually all participants experienced nocturnal low glucose values (defined as an interstitial tissue glucose< 3.3 mmol). Hypogly- caemia occurred, on average, during 8% of all nights and each patient was hypoglycaemic on 7.4% of nights. However, it has been proposed that CGM recording overestimates the overall frequency (McGowan et al. 2002). The technology measures glucose in the extra- cellular space and not in the blood vessels, and the relationship between concentrations in these two sites is unclear (Kulcu et al. 2003). These and other issues make it somewhat uncertain to what degree the low glucose values identified by CGM reflect true hypogly- caemia (Wentholt et al. 2005). Nevertheless, however nocturnal hypoglycaemia is identi- fied, it is common and often prolonged, with some episodes lasting for 3 hours or more (Gale and Tattersall 1979; Matyka et al. 1999a).

CAUSES OF NOCTURNAL HYPOGLYCAEMIA

The limitations of therapeutic insulin delivery The ability of people with insulin-treated diabetes to maintain strict glycaemic targets and prevent long-term tissue damage is compromised by the deficiencies of currently available methods of insulin delivery. In non-diabetic individuals, endogenous secretion of insulin precisely meets demand. When food is eaten, insulin secretion increases rapidly to match nutrient intake, particularly when the ingested food contains carbohydrate. Between meals, basal insulin secretion falls to a low but consistent level to maintain basal metabolism, without the risk of hypoglycaemia, even during prolonged fasting that lasts for hours or days. In contrast, the limitations of delivering insulin by subcutaneous injection to patients who can no longer produce endogenous insulin, leads not only to inadequate plasma insulin concentrations during eating and the immediate postprandial period, but also to inappro- priately raised plasma insulin levels in the post-absorptive phase (Rizza et al. 1980). This results in high postprandial blood glucose concentrations in the hour or so after eating and 98 Hypoglycaemia in Clinical Diabetes

1c 8.6 8.8 8.6 7.0 8.3 9.5 9.1 8.4 7.4 8.6 7.8 8.7 10.3 Mean HbA (%) 9 49 67 66 57 27 21 64 100 100 100 100 Proportion asymptomatic 8 2 9 9 12 11 27 56 2.0 7 9 27 70 13 47 36 22 69 29 21 32 2.5 35 45 17 29 14 44 29 34 10 30 97 56 % patients with nocturnal glucose below: (mmol/l) 3.0 3 2 1 1 1 1 1 1 1 3 2 2 1 2–3 365 Study nights per participant CGM CGM BG BG BG BG BG BG BG BG BG Glucose measurement CGM CGM CGM CGM BG MIT/CSII MIT/CSII BD BD BD/MIT Basal bolus Conventional CSII Conventional BD insulin OD Insulin MDI BD Treatment MDI/CSII MDI CSII MIT MDI BD A A A A A A A C C C C C C C C A or C Adults (A) or children (C) 47 56 29 61 31 25 58 71 23 82 51 57 25 39 n 150 135 176 Frequency Frequency of nocturnal hypoglycaemia in insulin-treated diabetes

Foundation Foundation Continuous Glucose Monitoring Study Group 2010 Kaufman Kaufman et al. 2002 Wiltshire et al. 2006 Boland et al. 2001 Matyka et al. 1999a Porter Porter et al. 1997 Beregszaszi et al. 1997 Vervoort Vervoort et al. 1996 Shalwitz et al. 1990 Pramming Pramming et al. 1985 Whincup and Milner 1987 Bendtson et al. 1988 Juvenile Diabetes Research Dornan et al. 1981 Table 5.1 Table Study injection multiple MIT, insulin; of injections daily multiple MDI, infusion; insulin subcutaneous continuous CSII, monitoring; glucose continuous CGM, glucose; blood BG, daily; twice BD, therapy; once OD, daily Wentholt Wentholt et al. 2007 Woodward et al. 2009 Gale and Tattersall Gale 1979 and Tattersall Chapter 5: Nocturnal Hypoglycaemia 99

7

6

5

4

3 Glucose (mmol/l) 2

1

0 21:00 23:00 01:00 03:00 05:00 07:00 Time (hours)

Figure 5.1 Overnight glucose profiles of one child with type 1 diabetes who was hypoglycaemic on both study nights. The shaded area represents the range of glucose values from the overnight profiles of the children without diabetes. (Source: Matyka et al. 1999a. Reproduced with permission from the BMJ).

Box 5.1 Factors contributing to the development of nocturnal hypoglycaemia 1. Long period between meals (especially in children). 2. Inconsistency of conventional subcutaneous basal insulin delivery – nocturnal hyperinsulinaemia. 3. Unawareness of early symptoms of hypoglycaemia when asleep. 4. Diminished counterregulatory hormone release and symptomatic response in supine posture and effect of sleep per se. a tendency to cause hypoglycaemia in the period before the next meal. Individuals are particularly likely to be affected during the night, when the interval between ingestion of food may be several hours (Box 5.1). As discussed below, developments in insulin pharm­ acokinetics with insulin analogues or in insulin delivery with continuous subcutaneous insulin infusion (CSII) offer some benefit over insulins of animal or human species, although these approaches mitigate rather than prevent nocturnal hypoglycaemia.

Impaired counterregulatory responses to hypoglycaemia during nocturnal hypoglycaemia Various mechanisms contribute to the additional risk of developing hypoglycaemia during the night. Although an earlier study suggested that hormonal responses to hypoglycaemia might be increased during nocturnal episodes (Bendtson et al. 1993), subsequent work has indicated that counterregulatory defences are generally impaired. Matyka et al. (1999a) studied 29 prepubertal children overnight in their own homes on two separate occasions. 100 Hypoglycaemia in Clinical Diabetes

They confirmed that not only was nocturnal hypoglycaemia common, but also that pro- longed episodes of hypoglycaemia were not accompanied by increased secretion of epine- phrine (adrenaline) and other counterregulatory hormones. Protective counterregulatory responses reduce the severity of hypoglycaemia by increasing hepatic glucose output and production and by reducing peripheral glucose uptake as well as initiating some of the warning autonomic symptoms. Thus, diminished responses during sleep may make a major contribution to increasing the risk of nocturnal hypoglycaemia. It appears that two factors in particular explain why nocturnal hypoglycaemia provokes different physiological responses to those occurring during the day – a supine posture and sleep per se.

Supine posture Two studies have investigated the effect of posture on counterregulatory hormonal defences to hypoglycaemia. Hirsch et al. (1991) compared the physiological responses to hypogly- caemia induced using a hyperinsulinaemic clamp in young people with type 1 diabetes, who were either in an upright or a horizontal position. Increments in hypoglycaemic symptom scores were more than 50% lower when patients remained supine compared to the increment observed when they were standing. These findings have been confirmed by others, who reported that plasma epinephrine levels were around three times lower during hypoglycaemia when subjects were supine rather than erect (Robinson et al. 1994). What- ever the cause, these observations suggest that patients are less likely to identify hypogly- caemia when lying down because of lower symptomatic and hormonal responses.

Sleep Most studies have reported suppression of physiological protective mechanisms when hypoglycaemia occurs during sleep. Jones et al. (1998) lowered blood glucose to 2.8 mmol/l both during the daytime and at night, and demonstrated diminished epinephrine responses in adolescents with and without type 1 diabetes while they were asleep, compared to brisk responses while they were awake, whether during daytime hours or during the night (Figure 5.2). Banarer and Cryer (2003) confirmed these observations in adults with type 1 diabetes but, interestingly, in non-diabetic subjects no temporal differences in the epine- phrine response to hypoglycaemia were observed. Diminished counterregulatory responses, including epinephrine, have also been demonstrated during spontaneous nocturnal hypo­ glycaemic episodes in children with diabetes (Matyka et al. 1999b). Therefore, sleep appears to be associated with diminished catecholamine and symptomatic responses to hypoglycaemia with a lesser prospect of wakening during a hypoglycaemic episode. The diminished response may occur because the glycaemic thresholds for activation of these responses have been reset to a lower glucose level (Gais et al. 2003), so that more profound hypoglycaemia is necessary to provoke a similar response to that observed in subjects when they are awake. As the investigators involved in these studies have pointed out, this increases the risk of a severe hypoglycaemic episode occurring at night. However, the mechanisms that disturb the physiological responses to hypoglycaemia during sleep remain unknown. Sleep is not a unitary process (Oswald 1987), with a pattern of dominance by Slow Wave Sleep (SWS) during the first third of the night, and by the cyclical appearance of Rapid Eye Movement (REM) sleep during the later two-thirds. Autonomic activity during Chapter 5: Nocturnal Hypoglycaemia 101

Image not available in this digital edition

Figure 5.2 Mean (±standard error) plasma epinephrine concentrations in eight patients with type 1 diabetes and six normal subjects during periods of hypoglycaemia when they were awake during the day, awake at night and asleep at night. (To convert plasma epinephrine values to picomoles per litre, multiply by 5.458. The zero on the x-axis indicates the beginning of the hypoglycaemic period.) (Source: Jones TW 1998. Copyright 1998 Massachusetts Medical Society).

SWS is relatively steady but in REM sleep, (when there is desynchronisation of the EEG with absence of activity in the antigravity and periodic eye movements) modulations in respiratory and cardiovascular events occur with other changes in the autonomic nervous system. These differences in autonomic activity between SWS and REM sleep suggest that the effect of hypoglycaemia on autonomic responses may vary depending on which stage of sleep is being experienced. Disturbed sleep architecture has been described in adults with type 1 diabetes, with prolongation of REM cycles (Bendtson et al. 1991). One study compared the counterregulatory responses during hypoglycaemia in the early and later parts of the night (Jauch-Chara et al. 2007a), and observed that the catecholamine, cortisol and growth hormone responses were weaker during late nocturnal hypoglycaemia when REM sleep predominated.

CONSEQUENCES OF NOCTURNAL HYPOGLYCAEMIA

Impaired awareness of hypoglycaemia The demonstration in the early 1990s that repeated exposure to hypoglycaemia leads to impaired physiological defences to subsequent episodes and a reduction in the intensity of symptoms (Heller and Cryer 1991; Dagogo-Jack et al. 1993) identified a mechanism that explains why some individuals lose their hypoglycaemia warning symptoms (see Chapter 6). Further work has suggested that such episodes do not need to be symptomatic to affect physiological responses. Veneman et al. (1993) induced hypoglycaemia in ten non-diabetic subjects overnight and tested their physiological responses to hypoglycaemia the following morning. They reported lower symptomatic and hormonal responses when compared to a non-hypoglycaemia control night. These relatively mild levels of hypoglycaemia are unre- ported by patients because they remain asleep but may contribute to the development of impaired awareness of hypoglycaemia. 102 Hypoglycaemia in Clinical Diabetes

The ‘dead in bed’ syndrome The possible contribution of hypoglycaemia to cardiac arrhythmias and sudden death is discussed in detail in Chapter 13. However, since it has been proposed that nocturnal hypoglycaemia may be a specific precipitant, it merits a brief mention here. A detailed survey of unexpected deaths in young people with type 1 diabetes in the early 1990s high- lighted a rare but distinct situation in which young people were found dead, lying in an undisturbed bed. Autopsy failed to reveal a structural cause but circumstantial evidence implicated nocturnal hypoglycaemia as a precipitant. Not all affected individuals had strict glycaemic control, but many were known to have been susceptible to nocturnal hypogly- caemia. This type of sudden and unexpected death has since been confirmed in epidemio- logical surveys in other countries, possibly accounting for between 5 and 10% of all deaths under the age of 40 in people with diabetes. Hypoglycaemia causes lengthening of the QT interval, and one possible explanation is that an episode of nocturnal hypoglycaemia trig- gers a fatal ventricular arrhythmia in a susceptible individual (Robinson et al. 2004; Gill et al. 2009). Further work is required to establish whether this plausible hypothesis is the underlying cause of these deaths.

Neurological consequences of nocturnal hypoglycaemia on cerebral function The possibility that recurrent exposure to hypoglycaemia, particularly occurring during sleep, might insidiously damage cerebral function and cause permanent cognitive impair- ment was raised by the finding that children who had developed type 1 diabetes before the age of 5 years, exhibited cognitive impairment when compared with non-diabetic controls (Ryan et al. 1985). This observation was replicated in different studies using a range of cognitive tests (Bjorgaas et al. 1997; Rovet and Ehrlich 1999). However, the relationship to hypoglycaemia, particularly when occurring during the night, was unclear (Golden et al. 1989); hypoglycaemia-induced convulsions have also been implicated, and it is pos- sible that other factors associated with early-onset diabetes may contribute to cognitive impairment. Several studies have explored the extent to which nocturnal episodes may affect per- formance on the following day. Bendtson et al. (1992) found no difference in cognitive performance among adults with type 1 diabetes when tested on the morning after an episode of nocturnal hypoglycaemia in comparison with a night with no hypoglycaemia. Similar findings have been reported after the induction of experimental hypoglycaemia during the night (King et al. 1998), although subjects were more fatigued when asked to exercise the following morning and both mood change and symptoms resembling a ‘hang­ over’ are common. Children appear to be relatively unaffected by a single episode of nocturnal hypoglycaemia. Matyka et al. (1999a) tested cognitive function of pre-pubertal children after episodes of prolonged, spontaneously occurring, hypoglycaemia that occurred during sleep, many of which lasted several hours, and found no deleterious effect the following morning, although mood was affected adversely. By contrast, in adults, hypo­ glycaemia during sleep has been shown to affect memory consolidation (Jauch-Chara et al. 2007b). In summary, while hypoglycaemia during sleep may have transient sequelae the follow- ing day, and it seems plausible that recurrent nocturnal hypoglycaemia might contribute to cognitive decline, on the available evidence the verdict remains unproven. Chapter 5: Nocturnal Hypoglycaemia 103

CAN NOCTURNAL HYPOGLYCAEMIA BE PREDICTED?

The ability to predict whether a nocturnal hypoglycaemic episode is likely, by measuring blood glucose concentrations at bedtime, has generally been tested in studies using inter- mittent blood glucose sampling to detect hypoglycaemia. Such a design will fail to identify all episodes. Asymptomatic nocturnal hypoglycaemia as detected by continuous glucose monitoring has been reported in approximately 50–60% of type 1 and type 2 diabetes patients (Chico et al. 2003).The results of two early studies that observed patients who were receiving one or two daily injections of insulin suggested that a bedtime glucose concentration of 6.0 mmol/l or less indicated an 80% chance of a period of biochemical hypoglycaemia occurring during the night (Pramming et al. 1985; Whincup and Milner 1987). Furthermore, these early studies were generally undertaken among patients who were taking a combination of conventional insulins twice daily. Studies of children have suggested that blood glucose measurements at bedtime are less predictive of hypoglycae- mia in the first half of the night, although a value of< 7.5 mmol/l does indicate an increased risk (Matyka et al. 1999a). That study and others have also shown that a low fasting blood glucose is a strong indicator that hypoglycaemia has occurred in the latter half of the night (Matyka 2002). A previous history of severe hypoglycaemia, impaired awareness of hypoglycaemia or strict glycaemic control may enhance the risk of a severe episode but are a poor guide as to whether a severe nocturnal episode will occur on any particular night.

THE SOMOGYI PHENOMENON: THE MYTH OF REBOUND HYPERGLYCAEMIA

In the late 1930s, a Hungarian biochemist, Michael Somogyi, working in St Louis, USA, suggested that nocturnal hypoglycaemia might provoke rebound hyperglycaemia on the following morning, and he supported his hypothesis with a demonstration that reducing evening doses of insulin led to a reduction in fasting glycosuria (Somogyi 1959). He proposed that nocturnal hypoglycaemia provokes a counterregulatory response with secre- tion of plasma epinephrine, cortisol and growth hormone resulting in the release of glucose from the liver and inhibition of the effects of insulin over the subsequent few hours. The logical conclusion from his hypothesis was that this ‘rebound’ elevated fasting blood glucose in the morning should be treated, not by an increase in the evening dose of insulin, but paradoxically by a reduction. The idea of ‘rebound hyperglycaemia’ following nocturnal hypoglycaemia, (also known as the ‘Somogyi phenomenon’), as an explanation for a high fasting blood glucose in insulin-treated patients, has proven attractive to many diabetes healthcare professionals who still firmly believe in its existence. The conse- quences are important as patients are often advised to reduce their evening insulin dose, particularly if they complain of nocturnal hypoglycaemia. However, its clinical relevance was challenged over 20 years ago, and repeated studies have established that fasting hyperglycaemia is largely a result of falling plasma insulin concentrations during the night, with dissipation of the subcutaneous depot of insulin that was injected the previous evening. Gale et al. (1980) demonstrated that periods of hypoglycaemia during the night were often prolonged and were not accompanied by a large rise in counterregulatory hormones. Although fasting glucose concentrations were frequently high in the patients studied, this 104 Hypoglycaemia in Clinical Diabetes

1.0

0.8

0.6

0.4

0.2 Probability of hypoglycaemic episodes in the night of hypoglycaemic Probability

2 3 4 5 6 7 8 91011 12 13 14 15 16 17 18 19 20 21 Morning blood glucose (mmol/I)

Figure 5.3 Risk of nocturnal hypoglycaemia according to fasting morning blood glucose (95% Cl) in 594 nights. Dark purple bars = hypoglycaemic nights; light purple bars = possibly hypoglycaemic nights. (Source: Hoi-Hansen T 2005. Reproduced with kind permission from Springer Science and Business Media). was related directly to a waning of circulating plasma insulin concentrations. Some inves- tigators have demonstrated that when hypoglycaemia is experimentally-induced during the night, this can raise blood glucose on the following morning, even if circulating plasma insulin concentrations are maintained (Perriello et al. 1988). However, the additional increase in the fasting blood glucose concentration is modest (around 2.0 mmol/l) and its clinical relevance is questionable. Other researchers have found no effect on daytime con- centrations of blood glucose after lowering blood glucose to hypoglycaemic levels during the night (Hirsch et al. 1990). Careful analysis of data collected both by self-monitoring of blood glucose (Havlin and Cryer 1987) and by continuous glucose monitoring (Hoi- Hansen et al. 2005) (Figure 5.3) during everyday activities, has also shown that nocturnal hypoglycaemia does not provoke rebound fasting hyperglycaemia. Many people with insulin-treated diabetes experience high fasting blood glucose levels, but this common clinical problem is mainly a consequence of inadequate overnight basal insulin replacement. The important mechanisms that contribute to fasting hyperglycaemia appear to be a combination of waning plasma insulin concentrations and increased glucose release from the liver secondary to nocturnal spikes of growth hormone secretion (Camp- bell et al. 1985); this physiological process is termed the ‘dawn phenomenon’. High blood glucose levels following symptomatic nocturnal hypoglycaemia may also result from excessive intake of oral carbohydrate, ingested as treatment of the hypoglycaemia, rather than from a powerful counterregulatory response, which is usually obtunded during sleep (Jones et al. 1998). The clinical message is therefore clear: fasting hyperglycaemia indi- cates a need for adjustment of basal insulin in terms of type or timing rather than reduction of the dose. Some useful clinical steps to be undertaken in patients presenting with this problem are listed in Box 5.2. Chapter 5: Nocturnal Hypoglycaemia 105

Box 5.2 Clinical approach to high fasting blood glucose complicated by nocturnal hypoglycaemia 1. Measure blood glucose at 2–3 am over a few days. 2. If nocturnal hypoglycaemia is present, ensure that basal insulin is taken at bedtime (i.e. split pre-mixed evening insulin). 3. Progressively increase bedtime long-acting insulin in doses of 2–4 units while checking with 3 am blood glucose measurements that this is not precipitating nocturnal hypoglycaemia. 4. Use a long-acting insulin analogue and administer before breakfast and not at bedtime. 5. Teach patients to take an appropriate (but not excessive) quantity of a high energy glucose drink, orange juice or glucose as sweets or tablets to treat nocturnal hypogly- caemic episodes. 6. When available, consider obtaining a continuous glucose monitoring profile.

Box 5.3 Potential remedies for problematical nocturnal hypoglycaemia

1. Long and rapid-acting insulin analogues 2. β-agonists (terbutaline or salbutamol) 3. Appropriate snacks: – slowly absorbed carbohydrate: uncooked cornstarch; cereal; porridge – high protein foods 4. CSII 5. Real-time continuous glucose monitoring 6. Closed loop insulin delivery

CLINICAL SOLUTIONS (BOX 5.3)

Dietary measures A time-honoured approach to reducing the frequency of nocturnal hypoglycaemia has been to counteract the effect of insulin by ensuring that patients eat a bedtime snack. This seemed particularly important for children who go to bed early and sleep for many hours between their evening insulin dose and breakfast the following day. It clearly makes sense for all individuals taking insulin to measure their blood glucose at bedtime and to take additional food if their blood glucose is low. However, the extent to which protective eating can prevent nocturnal hypoglycaemia in patients taking intermittent injections of insulin is limited. The value of different strategies has generally been assessed by measuring the effective- ness of different types of snack in preventing biochemical or symptomatic episodes of hypoglycaemia. Before the advent of continuous glucose monitoring, such studies demanded regular blood sampling for measurement of glucose, and since this generally required admission to hospital, these failed to simulate the clinical situation. Furthermore, the number of subjects studied has generally been relatively small, and although sufficient 106 Hypoglycaemia in Clinical Diabetes to measure changes in blood glucose concentration or frequency of biochemical hypoglycaemia, statistical power was inadequate to determine effects on the frequency of severe episodes. Studies have examined the potential of carbohydrate foods that are absorbed slowly and thus able to counter the hypoglycaemic effect of insulin over longer periods. Uncooked cornstarch, which has a very low glycaemic index, has been studied intensively, particularly because it is used successfully to prevent severe hypoglycaemia in glycogen storage dis- eases (Goldberg and Slonim 1993). In one study, when young people were given uncooked cornstarch incorporated into a normal bedtime snack, the incidence of symptomatic and biochemical nocturnal hypoglycaemia was 3-fold lower (Kaufman and Devgan 1996). A blinded randomised trial reported a similarly lower frequency of nocturnal episodes (Kaufman et al. 1997). Unfortunately, these preparations of uncooked cornstarch are mostly unpalatable and unlikely to be tolerated in the long term. Other work has examined the effect of snacks that are rich in protein or fat. Different types of snack were compared against placebo in a trial using a crossover design in 15 adults with type 1 diabetes (Kalergis et al. 2003). When patients retired to bed with blood glucose values greater than 10.0 mmol/l, nocturnal episodes were not observed. At a pre-bedtime glucose <7.0 mmol/l, a high protein snack prevented any episode of nocturnal hypoglycaemia by contrast with either a standard snack or one containing cornstarch. An alternative approach to dietary supplements is to reduce the rate of absorption of carbohydrates using a disaccharidase inhibitor such as acarbose. Three studies of these agents have examined their effects in patients with type 1 diabetes. McCulloch et al. (1983) studied the effect of acarbose on the risk of overnight hypoglycaemia, and found that the risk of symptomatic nocturnal hypoglycaemia was lowered by 39%. Taira et al. (2000) reported similar benefits using voglibose. However, a recent study reported no benefit of acarbose over placebo when both pharmaceutical and snacking interventions were com- pared with respect to their effect on preventing hypoglycaemia (Raju et al. 2006). In the light of these limited and conflicting data it seems unlikely that acarbose will ever be widely used for this purpose, particularly as sucrose-containing products may be relatively inef- fective when treating hypoglycaemia in conjunction with acarbose therapy.

Pharmaceutical interventions Pharmacological treatments to prevent hypoglycaemia are further discussed in Chapter 8. There are indications that β-agonists may have some use in reducing the risk of nocturnal hypoglycaemia. For some years inhaled terbutaline has been proposed as a method of elevating blood glucose. In the early 1990s, Wiethop and Cryer (1993) demonstrated that its oral or subcutaneous delivery following induced hypoglycaemia, led to a rise in blood glucose compared to placebo, an effect that lasted for some hours. Wright and Wales (2003) reported that children with type 1 diabetes who were receiving treatment for asthma had fewer episodes of nocturnal hypoglycaemia when compared to a non-asthmatic group of children in a survey lasting 3 months, implicating the β-agonist therapy as having a benefi- cial effect. Raju et al. (2006) also compared the effect of inhaled terbutaline with other therapeutic remedies such as cornstarch, standard snacks and acarbose on nocturnal blood glucose levels in patients with type 1 diabetes. They found that terbutaline prevented noc- turnal hypoglycaemia in all 15 subjects. However, although this treatment offered the greatest protection against hypoglycaemia at night, it also led to the highest fasting blood Chapter 5: Nocturnal Hypoglycaemia 107 glucose among the different remedies, diminishing the effect of this treatment as a realistic therapeutic option.

Timing and type of insulin, including insulin analogues The introduction of insulin analogues with pharmacokinetic properties that bear more resemblance to physiological insulin profiles in non-diabetic individuals highlighted the potential of such preparations to lower the risk of hypoglycaemia. Over a full 24 hours, the overall frequency of symptomatic hypoglycaemia in trials of rapid-acting insulin ana- logues has been modestly lower or no different from conventional short-acting human insulins, but a consistent finding has been a lower rate of nocturnal hypoglycaemia. It appears that the tendency of conventional soluble insulin to self-associate into hexamers at therapeutic concentrations leads to increasing plasma insulin levels with repeated injec- tion. Since rapid-acting insulin analogues separate into single molecules much more readily, accumulation of insulin is less likely and the risk of nocturnal hypoglycaemia is subse- quently lower. A trial with the short-acting analogue lispro, when given with the evening meal, was associated with lower rates of late nocturnal hypoglycaemia because of its faster peak action (Mohn et al. 1999). The frequency of nocturnal hypoglycaemia observed in clinical trials has been variable. Rates of nocturnal hypoglycaemia have been over 50% lower in some trials involving people with type 1 diabetes with strict glycaemic control (Heller et al. 1999, 2004), and although this has generally been demonstrated for sympto- matic episodes, these data might reflect a lower likelihood of severe hypoglycaemia (Holleman et al. 1997). The long-acting insulin analogues – insulins glargine and detemir – which provide basal insulin replacement also appear to be associated with a lower risk of nocturnal hypoglycaemia. They have a longer duration of action when compared to isophane (NPH) insulin, together with less of a peak in their time-action profile, less variability and a more con­sistent duration of action (Mohn et al. 2000; Barnett 2003). These properties probably contribute to the lower rates of nocturnal hypoglycaemia that have been observed during clinical trials. A reduction in risk has been reported for long-acting preparations in trials involving people with type 1 (Pieber et al. 2000; De Leeuw et al. 2005) and with type 2 diabetes (Yki-Jarvinen et al. 2000; Hermansen et al. 2006), but none of these newer insu- lins eradicate nocturnal hypoglycaemia. In a meta-analysis, insulin analogues were associ- ated with around half the risk of nocturnal hypoglycaemia compared with isophane insulin in type 2 diabetes patients (Monami et al. 2008). The combined use of both rapid and long-acting insulin analogues might be expected to have a greater impact in lowering the risk of nocturnal hypoglycaemia. In the few studies that have compared combinations of insulin analogues to conventional human insulins this does appear to be true. Hermansen et al. (2004) reported a 55% lower rate of symptomatic nocturnal hypoglycaemia when using an insulin detemir/aspart combination as basal-bolus therapy in patients with type

1 diabetes, which was accompanied by a very modest but significant fall in 1c HbA of 0.2%. Ashwell et al. (2006) observed a fall in HbA1c of 0.5% using insulin glargine and lispro in a basal-bolus regimen in patients with type 1 diabetes, while nocturnal hypogly- caemia was 44% lower in frequency. However, nocturnal hypoglycaemia was not eradi- cated, leading to the conclusion that although insulin analogues may help to reduce the side effects of insulin therapy, they do not solve the limitations of current therapeutic insulin delivery. 108 Hypoglycaemia in Clinical Diabetes

Continuous Subcutaneous Insulin Infusion (CSII), Real-time Continuous Glucose Monitoring Systems (RT-CGMS) and closed loop insulin delivery Since nocturnal hypoglycaemia is largely the result of inadequate basal insulin replacement, one would expect that the most effective method of basal insulin delivery currently avail- able, CSII, could limit the frequency of nocturnal hypoglycaemia (Pickup and Keen 2002). However, early studies reported surprisingly little effect on limiting hypoglycaemia, perhaps because patients were not trained in the essential related skills of adjusting carbohydrate­ and insulin dosage. More recent work has indicated that the use of CSII can result in lower overall rates of hypoglycaemia (Bode et al. 1996; Kanc et al. 1998; Boland et al. 1999). In a systematic review of randomised trials comparing CSII and multiple daily injections, no significant difference was observed in rates of nocturnal hypoglycaemia (Fatourechi et al. 2009). However, overall event rates were low and trials often excluded people who had a history of severe hypoglycaemia or impaired hypoglycaemia awareness. Real-time continuous glucose monitoring systems (RT-CGMS) can potentially prevent nocturnal hypoglycaemia by functioning as an alarm when glucose falls below a threshold value. However, the success of this approach has been limited. In a study involving children and adolescents, only a third awoke when the hypoglycaemia alarm went off (Buckingham et al. 2005). Further, a high proportion of the alarms were false positives due to inaccuracy of interstitial glucose monitoring. This not only disrupts sleep, it can also lead to lower response rates over time because of alarm fatigue. Sensor-augmented pumps integrate a real-time continuous glucose system with CSII. More recently, some sensor-augmented pumps have included a low glucose-suspend func- tion where basal insulin delivery is automatically suspended for several hours when CGM detects impending hypoglycaemia. In an early trial, the low glucose-suspend function reduced the duration of nocturnal hypoglycaemia in those at highest risk (Choudhary et al. 2011). However, insulin suspension could lead to subsequent hyperglycaemia or even mild ketosis (Buckingham et al. 2010), particularly if the suspension was initiated by a false positive low glucose. Advances have been made in closed loop delivery systems which contain automatic algorithms to calculate rates of CSII insulin delivery based on glucose values detected by CGM. These have been tested in overnight settings and have the potential to reduce the frequency of nocturnal hypoglycaemia. Early trials have been promising. In the first ran- domised controlled trial of closed loop versus CSII in children and adolescents, the time spent in hypoglycaemia was reduced from 4% to 2% (Hovorka et al. 2010). In a randomised crossover trial in adults, closed loop delivery also eliminated glucose concentrations below 3.0 mmol after midnight (Hovorka et al. 2011). Although closed loop systems have been tested in a variety of simulated situations (such as after exercise or following an evening meal when alcohol has been ingested), larger trials of fully automated systems conducted in domestic settings are awaited. It will be important to evaluate such systems in patients who are at high risk of developing nocturnal hypoglycaemia.

CONCLUSIONS • Nocturnal hypoglycaemia remains an unresolved clinical side-effect of insulin therapy preventing the attainment of strict glycaemic control for many people. It contributes to morbidity, and perhaps mortality, in people with type 1 diabetes. Chapter 5: Nocturnal Hypoglycaemia 109 • Nocturnal hypoglycaemia is caused chiefly by the limitations of current methods of insulin delivery. The inability of regimens using subcutaneous insulin injections, particularly with conventional basal human insulins, to maintain low-level stable insulin concentrations overnight, leads both to nocturnal hypoglycaemia and fasting hyperglycaemia. • In addition to the limitations of insulin delivery, hypoglycaemia is also more common as a result of diminished counterregulatory responses overnight, associated in part with sleep, which has a specific inhibitory effect on physiological defences to hypoglycae- mia, and a supine posture, which diminishes autonomic responses. • Nocturnal hypoglycaemia is a particular problem in children, partly as a consequence of the long period of fasting between their evening meal and breakfast the following morning. • The risk of nocturnal hypoglycaemia is greatest in those whose bedtime glucose is below 7.0 mmol/l. Bedtime snacks can reduce the risk during the early part of the night. • Rebound hyperglycaemia (the ‘Somogyi phenomenon’) is rarely caused by counterregu- latory hormone release provoked by an overnight hypoglycaemic episode, since hormo- nal secretion is generally obtunded during sleep. Fasting hyperglycaemia is mainly a consequence of waning of plasma insulin levels and should be treated by adjustment of the timing and type of insulin used rather than a reduction in insulin dose. • The problem of nocturnal hypoglycaemia may respond to the use of insulin analogues (both rapid and long-acting) or CSII therapy. Sensor-augmented pumps which integrate low glucose insulin suspension and closed loop insulin delivery systems hold promise for the future.

REFERENCES

Ashwell, S.G., Amiel, S.A., Bilous, R.W. et al. (2006) Improved glycaemic control with insulin glargine plus insulin lispro: a multicentre, randomized, cross-over trial in people with type 1 diabetes. Diabetic Medicine, 23, 285–292. Banarer, S. & Cryer, P.E. (2003) Sleep-related hypoglycemia-associated autonomic failure in type 1 dia- betes: reduced awakening from sleep during hypoglycemia. Diabetes, 52, 1195–1203. Barnett, A.H. (2003) A review of basal insulins. Diabetic Medicine, 20, 873–885. Bendtson, I., Kverneland, A., Pramming, S. & Binder, C. (1988) Incidence of nocturnal hypoglycaemia in insulin-dependent diabetes. Acta Endocrinologica, 223, 543–548. Bendtson, I., Gade, J., Rosenfalck, A.M., Thomsen, C.E., Wildschiodtz, G. & Binder, C. (1991) Nocturnal electroencephalogram registrations in Type 1 (insulin-dependent) diabetic patients with hypoglycaemia. Diabetologia, 34, 750–756. Bendtson, I., Gade, J., Theilgaard, A. & Binder, C. (1992) Cognitive function in type 1 (insulin-dependent) diabetic patients after nocturnal hypoglycaemia. Diabetologia, 35, 898–903. Bendtson, I., Rosenfalck, A.M. & Binder, C. (1993) Nocturnal versus diurnal hormonal counterregulation to hypoglycemia in type 1 (insulin-dependent) diabetic patients. Acta Endocrinologica, 128, 109–115. Beregszaszi, M., Tubiana-Rufi, N., Benali, K., Noel, M., Bloch, J. & Czernichow, P. (1997) Nocturnal hypoglycemia in children and adolescents with insulin-dependent diabetes mellitus: prevalence and risk factors. Journal of Pediatrics, 131, 27–33. Bjorgaas, M., Gimse, R., Vik, T. & Sand, T. (1997) Cognitive function in type 1 diabetic children with and without episodes of severe hypoglycaemia. Acta Paediatrica, 86, 148–153. Bode, B.W., Steed, R.D. & Davidson, P.C. (1996) Reduction in severe hypoglycemia with long-term con- tinuous subcutaneous insulin infusion in type I diabetes. Diabetes Care, 19, 324–327. Boland, E., Monsod, T., Delucia, M., Brandt, C.A., Fernando, S. & Tamborlane, W.V. (2001) Limitations of conventional methods of self-monitoring of blood glucose: lessons learned from 3 days of continuous glucose sensing in pediatric patients with type 1 diabetes. Diabetes Care, 24, 1858–1862. 110 Hypoglycaemia in Clinical Diabetes

Boland, E.A., Grey, M., Oesterle, A., Fredrickson, L. & Tamborlane, W.V. (1999) Continuous subcutaneous insulin infusion. A new way to lower risk of severe hypoglycemia, improve metabolic control, and enhance coping in adolescents with type 1 diabetes. Diabetes Care, 22, 1779–1784. Buckingham, B., Block, J., Burdick, J. et al. (2005) Response to nocturnal alarms using a real-time glucose sensor. Diabetes Technology and Therapeutics, 7, 440–447. Buckingham, B., Chase, H.P., Dassau, E. et al. (2010) Prevention of nocturnal hypoglycemia using predic- tive alarm algorithms and insulin pump suspension. Diabetes Care, 33, 1013–1017. Campbell, P.J., Bolli, G.B., Cryer, P.E. & Gerich, J.E. (1985) Pathogenesis of the dawn phenomenon in patients with insulin-dependent diabetes mellitus. Accelerated glucose production and impaired glucose utilization due to nocturnal surges in growth hormone secretion. New England Journal of Medicine, 312, 1473–1479. Chico, A., Vidal-Rios, P., Subira, A. & Novials, A. (2003) The continuous glucose monitoring system is useful for detecting unrecognized hypoglycemias in patients with type 1 and type 2 diabetes but is not better than frequent capillary glucose measurements for improving metabolic control. Diabetes Care, 26, 1153–1157. Choudhary, P., Shin, J., Wang, Y. et al. (2011) Insulin pump therapy with automated insulin suspension in response to hypoglycemia: reduction in nocturnal hypoglycemia in those at greatest risk. Diabetes Care, 34, 2023–2025. Dagogo-Jack, S.E., Craft, S. & Cryer, P.E. (1993) Hypoglycemia-associated autonomic failure in insulin- dependent diabetes mellitus. Journal of Clinical Investigation, 91, 819–828. De Leeuw, I., Vague, P., Selam, J.L. et al. (2005) Insulin detemir used in basal-bolus therapy in people with type 1 diabetes is associated with a lower risk of nocturnal hypoglycaemia and less weight gain over 12 months in comparison to NPH insulin. Diabetes, Obesity and Metabolism, 7, 73–82. Dornan, T.L., OrdePeckar, C., Mayon-White, V.A. et al. (1981) Unsuspected hypoglycaemia, haemoglobin A1 and diabetic control. Quarterly Journal of Medicine, 197, 31–38. Fatourechi, M.M., Kudva, Y.C., Murad, M.H., Elamin, M.B., Tabini, C. & Montori, V.M. (2009) Hypogly- cemia with intensive insulin therapy: a systematic review and meta-analyses of randomized trials of continuous subcutaneous insulin infusion versus multiple daily injections. Journal of Clinical Endo­ crinology & Metabolism, 94, 729–740. Gais, S., Born, J., Peters, A. et al. (2003) Hypoglycemia counterregulation during sleep. Sleep, 26, 55–59. Gale, E.A. & Tattersall, R.B. (1979) Unrecognised nocturnal hypoglycaemia in insulin-treated diabetics. Lancet, 1, 1049–1052. Gale, E.A., Kurtz, A.B. & Tattersall, R.B. (1980) In search of the Somogyi effect. Lancet, 2, 279–282. Gill, G.V., Woodward, A., Casson, I.F. & Weston, P.J. (2009) Cardiac arrhythmia and nocturnal hypogly- caemia in type 1 diabetes–the ‘dead in bed’ syndrome revisited. Diabetologia, 52, 42–45. Goldberg, T. & Slonim, A.E. (1993) Nutrition therapy for hepatic glycogen storage diseases. Journal of the American Dietetic Association, 93, 1423–1430. Golden, M.P., Ingersoll, G.M., Brack, C.J., Russell, B.A., Wright, J.C. & Huberty, T.J. (1989) Longitudinal relationship of asymptomatic hypoglycemia to cognitive function in IDDM. Diabetes Care, 12, 89–93. Havlin, C.E. & Cryer, P.E. (1987) Nocturnal hypoglycemia does not commonly result in major morning hyperglycemia in patients with diabetes mellitus. Diabetes Care, 10, 141–147. Heller, S.R. & Cryer, P.E. (1991) Reduced neuroendocrine and symptomatic responses to subsequent hypoglycemia after 1 episode of hypoglycemia in non-diabetic humans. Diabetes, 40, 223–226. Heller, S.R., Amiel, S.A. & Mansell, P. (1999) Effect of the fast-acting insulin analog lispro on the risk of nocturnal hypoglycemia during intensified insulin therapy. Diabetes Care, 22, 1607–1611. Heller, S.R., Colagiuri, S., Vaaler, S. et al. (2004) Hypoglycaemia with insulin aspart: a double-blind, randomised, crossover trial in subjects with type 1 diabetes. Diabetic Medicine, 21, 769–775. Hermansen, K., Fontaine, P., Kukolja, K.K., Peterkova, V., Leth, G. & Gall, M.A. (2004) Insulin analogues (insulin detemir and insulin aspart) versus traditional human insulins (NPH insulin and regular human insulin) in basal-bolus therapy for patients with type 1 diabetes. Diabetologia, 47, 622–629. Hermansen, K., Davies, M., Derezinski, T., Martinez Ravn, G., Clauson, P. & Home, P. (2006) A 26-week, randomized, parallel, treat-to-target trial comparing insulin detemir with NPH insulin as add-on therapy to oral glucose-lowering drugs in insulin-naive people with type 2 diabetes. Diabetes Care, 29, 1269–1274. Hirsch, I.B., Smith, L.J., Havlin, C.E., Shah, S.D., Clutter, W.E. & Cryer, P.E. (1990) Failure of nocturnal hypoglycemia to cause daytime hyperglycemia in patients with IDDM. Diabetes Care, 13, 133–142. Chapter 5: Nocturnal Hypoglycaemia 111

Hirsch, I.B., Heller, S.R. & Cryer, P.E. (1991) Increased symptoms of hypoglycaemia in the standing posi- tion in insulin-dependent diabetes mellitus. Clinical Science, 80, 583–586. Hoi-Hansen, T., Pedersen-Bjergaard, U. & Thorsteinsson, B. (2005) The Somogyi phenomenon revisited using continuous glucose monitoring in daily life. Diabetologia, 48, 2437–2438. Holleman, F., Schmitt, H., Rottiers, R. et al.; The Benelux-UK Insulin Lispro Study Group (1997) Reduced frequency of severe hypoglycemia and coma in well-controlled IDDM patients treated with insulin lispro. Diabetes Care, 20, 1827–1832. Hovorka, R., Allen, J.M., Elleri, D. et al. (2010) Manual closed-loop insulin delivery in children and adolescents with type 1 diabetes: a phase 2 randomised crossover trial. Lancet, 375, 743–751. Hovorka, R., Kumareswaran, K., Harris, J. et al. (2011) Overnight closed loop insulin delivery (artificial pancreas) in adults with type 1 diabetes: crossover randomised controlled studies. British Medical Journal, 342, d1855. Jauch-Chara, K., Hallschmid, M., Gais, S. et al. (2007a) Awakening and counterregulatory response to hypoglycemia during early and late sleep. Diabetes, 56, 1938–1942. Jauch-Chara, K., Hallschmid, M., Gais, S. et al. (2007b) Hypoglycemia during sleep impairs consolidation of declarative memory in type 1 diabetic and healthy humans. Diabetes Care, 30, 2040–2045. Jones, T.W., Porter, P., Sherwin, R.S. et al. (1998) Decreased epinephrine responses to hypoglycemia during sleep. New England Journal of Medicine, 338, 1657–1662. Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group (2010) Prolonged nocturnal hypoglycemia is common during 12 months of continuous glucose monitoring in children and adults with type 1 diabetes. Diabetes Care, 33, 1004–1008. Kalergis, M., Schiffrin, A., Gougeon, R., Jones, P.J. & Yale, J.F. (2003) Impact of bedtime snack composi- tion on prevention of nocturnal hypoglycemia in adults with type 1 diabetes undergoing intensive insulin management using lispro insulin before meals: a randomized, placebo-controlled, crossover trial. Diab­ etes Care, 26, 9–15. Kanc, K., Janssen, M.M., Keulen, E.T. et al. (1998) Substitution of night-time continuous subcutaneous insulin infusion therapy for bedtime NPH insulin in a multiple injection regimen improves counterregula- tory hormonal responses and warning symptoms of hypoglycaemia in IDDM. Diabetologia, 41, 322–329. Kaufman, F.R. & Devgan, S. (1996) Use of uncooked cornstarch to avert nocturnal hypoglycemia in chil- dren and adolescents with type 1 diabetes. Journal of Diabetes and Its Complications, 10, 84–87. Kaufman, F.R., Halvorson, M. & Kaufman, N.D. (1997) Evaluation of a snack bar containing uncooked cornstarch in subjects with diabetes. Diabetes Research and Clinical Practice, 35, 27–33. Kaufman, F.R., Austin, J., Neinstein, A. et al. (2002) Nocturnal hypoglycemia detected with the Continuous Glucose Monitoring System in pediatric patients with type 1 diabetes. Journal of Pediatrics, 141, 625–630. King, P., Kong, M.F., Parkin, H., Macdonald, I.A. & Tattersall, R.B. (1998) Well-being, cerebral function, and physical fatigue after nocturnal hypoglycemia in IDDM. Diabetes Care, 21, 341–345. Kulcu, E., Tamada, J.A., Reach, G., Potts, R.O. & Lesho, M.J. (2003) Physiological differences between interstitial glucose and blood glucose measured in human subjects. Diabetes Care, 26, 2405–2409. Matyka, K.A. (2002) Sweet dreams? Nocturnal hypoglycemia in children with type 1 diabetes. Pediatric Diabetes, 3, 74–81. Matyka, K.A., Wigg, L., Pramming, S., Stores, G. & Dunger, D.B. (1999a) Cognitive function and mood after profound nocturnal hypoglycaemia in prepubertal children with conventional insulin treatment for diabetes. Archives of Disease in Childhood, 81, 138–142. Matyka, K.A., Crowne, E.C., Havel, P.J., Macdonald, I.A., Matthews, D. & Dunger, D.B. (1999b) Counter- regulation during spontaneous nocturnal hypoglycemia in prepubertal children with type 1 diabetes. Diabetes Care, 22, 1144–1150. McCulloch, D.K., Kurtz, A.B. & Tattersall, R.B. (1983) A new approach to the treatment of nocturnal hypoglycemia using alpha-glucosidase inhibition. Diabetes Care, 6, 483–487. McGowan, K., Thomas, W. & Moran, A. (2002) Spurious reporting of nocturnal hypoglycemia by CGMS in patients with tightly controlled type 1 diabetes. Diabetes Care, 25, 1499–1503. Mohn, A., Matyka, K.A., Edge, J.A., Harris, D.A. & Dunger, D.B. (1999) Lispro or regular insulin for multliple injection therapy in adolescence. Differences in free insulin and glucose levels overnight. Diabetes Care, 22, 27–32. Mohn, A., Strang, S., Wernicke-Panten, K., Lang, A.M., Edge, J.A. & Dunger, D.B. (2000) Nocturnal glucose control and free insulin levels in children with type 1 diabetes by use of the long-acting insulin HOE 901 as part of a three-injection regimen. Diabetes Care, 23, 557–559. 112 Hypoglycaemia in Clinical Diabetes

Monami, M., Marchionni, N. & Mannucci, E. (2008) Long-acting insulin analogues versus NPH human insulin in type 2 diabetes: a meta-analysis. Diabetes Research and Clinical Practice, 81, 184–189. Oswald, I. (1987) The normal record of sleep. In: A.M. Halliday, S.R. Butler & R. Paul (eds), A Textbook of Clinical Neurophysiology, pp. 173–185. John Wiley & Sons Inc., New York. Perriello, G., De Feo, P., Torlone, E. et al. (1988) The effect of asymptomatic nocturnal hypoglycemia on glycemic control in diabetes mellitus. New England Journal of Medicine, 319, 1233–1239. Pickup, J. & Keen, H. (2002) Continuous subcutaneous insulin infusion at 25 years: evidence base for the expanding use of insulin pump therapy in type 1 diabetes. Diabetes Care, 25, 593–598. Pieber, T.R., Eugene-Jolchine, I. & Derobert, E. (2000) Efficacy and safety of HOE 901 versus NPH insulin in patients with type 1 diabetes. The European Study Group of HOE 901 in type 1 diabetes. Diabetes Care, 23, 157–162. Porter, P.A., Keating, B., Byrne, G. & Jones, T.W. (1997) Incidence and predictive criteria of nocturnal hypoglycemia in young children with insulin-dependent diabetes mellitus. Journal of Pediatrics, 130, 366–372. Pramming, S., Thorsteinsson, B., Bendtson, I., Ronn, B. & Binder, C. (1985) Nocturnal hypoglycaemia in patients receiving conventional treatment with insulin. British Medical Journal, 291, 376–379. Raju, B., Arbelaez, A.M., Breckenridge, S.M. & Cryer, P.E. (2006) Nocturnal hypoglycemia in type 1 diabetes: an assessment of preventive bedtime treatments. Journal of Clinical Endocrinology and Metab­ olism, 91, 2087–2092. Rizza, R.A., Gerich, J.E., Haymond, M.W. et al. (1980) Control of blood sugar in insulin-dependent dia- betes: comparison of an artificial endocrine pancreas, continuous subcutaneous insulin infusion and intensified conventional insulin therapy. New England Journal of Medicine, 303, 1313–1318. Robinson, A.M., Parkin, H.M., Macdonald, I.A. & Tattersall, R.B. (1994) Physiological response to postural change during mild hypoglycaemia in patients with IDDM. Diabetologia, 37, 1241–1250. Robinson, R., Harris, N., Ireland, R., Macdonald, I. & Heller, S. (2004) Changes in cardiac repolarization during clinical episodes of nocturnal hypoglycaemia in adults with Type 1 diabetes. Diabetologia, 47, 312–315. Rovet, J.F. & Ehrlich, R.M. (1999) The effect of hypoglycemic seizures on cognitive function in children with diabetes: a 7-year prospective study. Journal of Pediatrics, 134, 503–506. Ryan, C., Vega, A. & Drash, A. (1985) Cognitive deficits in adolescents who developed diabetes early in life. Pediatrics, 75, 921–927. Shalwitz, R.A., Farkas-Hirsch, R., White, N.H. & Santiago, J.V. (1990) Prevalence and consequences of nocturnal hypoglycemia among conventionally treated children with diabetes mellitus. Journal of Pedi­ atrics, 116, 685–689. Somogyi, M. (1959) Exacerbation of diabetes by excess insulin action. American Journal of Medicine, 26, 169–191. Taira, M., Takasu, N., Komiya, I., Taira, T. & Tanaka, H. (2000) Voglibose administration before the evening meal improves nocturnal hypoglycemia in insulin-dependent diabetic patients with intensive insulin therapy. Metabolism: Clinical and Experimental, 49, 440–443. The DCCT Research Group (1991) Epidemiology of severe hypoglycemia in the Diabetes Control and Complications Trial. American Journal of Medicine, 90, 450–459. Veneman, T., Mitrakou, A., Mokan, M., Cryer, P. & Gerich, J. (1993) Induction of hypoglycemia unaware- ness by asymptomatic nocturnal hypoglycemia. Diabetes, 42, 1233–1237. Vervoort, G., Goldschmidt, H.M. & van Doorn, L.G. (1996) Nocturnal blood glucose profiles in patients with type 1 diabetes mellitus on multiple (> or = 4) daily insulin injection regimens. Diabetic Medicine, 13, 794–799. Weber, K.K., Lohmann, T., Busch, K., Donati-Hirsch, I. & Riel, R. (2007) High frequency of unrecognized hypoglycaemias in patients with Type 2 diabetes is discovered by continuous glucose monitoring. Experimental and Clinical Endocrinology and Diabetes, 115, 491–494. Wentholt, I.M., Vollebregt, M.A., Hart, A.A., Hoekstra, J.B. & DeVries, J.H. (2005) Comparison of a needle-type and a microdialysis continuous glucose monitor in type 1 diabetic patients. Diabetes Care, 28, 2871–2876. Wentholt, I.M., Maran, A., Masurel, N., Heine, R.J., Hoekstra, J.B. & Devries, J.H. (2007) Nocturnal hypoglycaemia in Type 1 diabetic patients, assessed with continuous glucose monitoring: frequency, duration and associations. Diabetic Medicine, 24, 527–532. Whincup, G. & Milner, R.D. (1987) Prediction and management of nocturnal hypoglycaemia in diabetes. Archives of Disease in Childhood, 62, 333–337. Chapter 5: Nocturnal Hypoglycaemia 113

Wiethop, B.V. & Cryer, P.E. (1993) Alanine and terbutaline in treatment of hypoglycemia in IDDM. Dia­ betes Care, 16, 1131–1136. Wiltshire, E.J., Newton, K. & McTavish, L. (2006) Unrecognised hypoglycaemia in children and adoles- cents with type 1 diabetes using the continuous glucose monitoring system: prevalence and contributors. Journal of Paediatric and Child Health, 42, 758–763. Woodward, A., Weston, P., Casson, I.F. & Gill, G.V. (2009) Nocturnal hypoglycaemia in type 1 diabetes– frequency and predictive factors. Quarterly Journal of Medicine, 102, 603–607. Wright, N.P. & Wales, J.K. (2003) The incidence of hypoglycaemia in children with type 1 diabetes and treated asthma. Archives of Disease in Childhood, 88, 155–156. Yki-Jarvinen, H., Dressler, A. & Ziemen, M. (2000) Less nocturnal hypoglycemia and better post-dinner glucose control with bedtime insulin glargine compared with bedtime NPH insulin during insulin com- bination therapy in type 2 diabetes. HOE 901/3002 Study Group. Diabetes Care, 23, 1130–1136. 6 Impaired Awareness of Hypoglycaemia Brian M. Frier University of Edinburgh, Edinburgh, UK

Dangerous hypoglycaemia may occur without warning symptoms – E.P. Joslin et al. (1922)

INTRODUCTION

The generation of symptoms in response to hypoglycaemia provides a fundamental defence for the brain, by alerting the affected individual to the imminent development of neuro- glycopenia (Chapter 2). This should provoke an appropriate response – obtaining and ingesting some form of carbohydrate to reverse the low blood glucose. If these warning symptoms fail to occur, are ignored or are delayed until the blood glucose has fallen to a level that causes disabling cognitive impairment, serious consequences may ensue. If no avoiding action is taken and blood glucose continues to fall, this may lead to profound neuroglycopenia, with the risk of progression to confusion, altered consciousness and even- tual coma. An inadequate symptomatic warning often occurs in people with insulin-treated diabetes, in various circumstances and with differing causes, and if a persisting problem is described as ‘impaired awareness of hypoglycaemia’ or ‘hypoglycaemia unawareness’. This is an acquired abnormality that is effectively a complication of insulin therapy; it should be ranked alongside the microvascular complications of diabetes such as retinopathy, neuropa- thy or nephropathy, because its morbidity can be just as serious and disabling.

NORMAL RESPONSES TO HYPOGLYCAEMIA

In humans the ventromedial thalamus (VMH) is a key glucose-sensing region, which is involved in the detection of hypoglycaemia and initiation of counterregulatory responses to protect the brain from the effects of neuroglycopenia (McCrimmon 2009). Glucose sensors are also located outside the brain – in the portal vein, intestine, carotid body and in glucokinase-sensing pancreatic beta cells, but their role in moderating the counterregula- tory response to hypoglycaemia is as yet unclear (McCrimmon 2008). In people who develop impaired awareness of hypoglycaemia (IAH), it is not known if the glucose sensors

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. Chapter 6: Impaired Awareness of Hypoglycaemia 115

4.6 mmol/L 5.0 Inhibition of endogenous insulin secretion

3.8 mmol/L 4.0 Counterregulatory hormone release • Glucagon 3.2–2.8 mmol/L • Epinephrine Onset of symptoms 3.0 3.0–2.4 mmol/L 3.0 mmol/L 2.8 mmol/L • Autonomic Neurophysiological Cognitive • Neuroglycopenic dysfunction Widespread EEG changes dysfunction • Evoked responses • Inability to perform complex 2.0 tasks

< 1.5 mmol/L Severe 1.0 neuroglycopenia • Reduced conscious level • Convulsions Arterialised venous blood glucose concentration (mmol/L) • Coma 0

Figure 6.1 Hierarchy of endocrine, symptomatic and neurological responses to acute hypoglycaemia in non-diabetic subjects. Glycaemic thresholds are based on glucose concentrations in arterialised venous blood. (Source: Modified from Textbook of Diabetes, 2nd edition (1997) (eds J. Pickup and G. Williams). Reproduced with permission from John Wiley & Sons, Ltd).

in the brain are simply malfunctioning (probably potentially reversible in the short term) or whether in the long term they become irreversibly damaged. Acute hypoglycaemia induces a series of changes – hormonal, neurophysiological, symptomatic and cognitive – which occur at different and defined blood glucose concentra- tions (Figure 6.1). The thresholds at which these changes are triggered have been described in non-diabetic humans, most occurring within a relatively narrow range of blood glucose concentrations. In people with diabetes these glycaemic thresholds are not static and per- manent, but are dynamic and display plasticity, altering in response to external influences such as changes in glycaemic control and exposure to extremes of blood glucose. Thus the blood glucose level at which symptoms are generated can be modified through the ability of the brain to adapt to environmental change, that is, its exposure to prevailing blood glucose concentrations. Depriving the brain of glucose causes it to malfunction, and cognitive impairment quickly becomes evident as an overt manifestation of neuroglycopenia. Some of these features are relatively subtle and may not be detected immediately. A fall in blood glucose triggers activation of the hypothalamic autonomic centres within the brain, which conse- quently stimulates the peripheral sympathoadrenal system. This promotes typical physio- logical responses including sweating, an increase in rate and contractility of the heart (sensed as a pounding heart), and tremor, these being some of the classic features of the acute autonomic reaction (Figure 6.2). These symptoms (Chapter 2) result from end-organ stimulation via autonomic innervation and not (as is sometimes incorrectly stated) by the direct effect of catecholamines. Although epinephrine (adrenaline) is secreted in large quantities from the adrenal medullae, its contribution is to heighten the magnitude and 116 Hypoglycaemia in Clinical Diabetes

Cerebral cortex

Neuroglycopenic symptoms

Hypoglycaemia

Hypothalamus

Parasympathetic Sympathetic nervous system nerves

Adrenal medulla

Adrenaline

Tremor Heart Sudomotor eccrine sweat glands

Pounding heart Autonomic Sweating symptoms Perception, interpretation and action

Figure 6.2 Generation of neuroglycopenic and autonomic symptoms in response to hypoglycaemia. Autonomic activation and the involvement of the sympathoadrenal system in the stimulation of repre- sentative end-organs associated with common autonomic symptoms of hypoglycaemia. (Source: Hypoglycaemia and Diabetes (eds B.M. Frier and B.M. Fisher), © 1993 Edward Arnold. Reproduced with permission from Edward Arnold Publishers Ltd).

intensity of the symptomatic responses. The early literature on hypoglycaemia and diabetes provides accurate descriptions of the autonomic features of acute hypoglycaemia, and patients and physicians alike commonly discussed hypoglycaemic ‘reactions’, a term that regrettably is now seldom used. It emphatically describes the sudden, and often florid, onset of the autonomic features of hypoglycaemia, which drive the individual to seek assistance or obtain a supply of glucose to relieve these unpleasant symptoms.

‘Awareness’ of hypoglycaemia Awareness of the early onset of hypoglycaemia may not be definable as the detection of specific symptoms per se. In his Nobel lecture in 1925, within a few years of insulin first Chapter 6: Impaired Awareness of Hypoglycaemia 117 being used to treat diabetes, observed: ‘The first warning of hypogly- caemia was an unaccountable anxiety and a feeling of impending trouble associated with restlessness’. Many people describe a vague sensation of apprehension or loss of well-being before developing florid symptoms of hypoglycaemia. This subjective feeling that some- thing is not right, which precedes the onset of autonomic or neuroglycopenic features, is described by many people who experience hypoglycaemia while treated with insulin. When typical physiological responses to hypoglycaemia are generated, they are per- ceived by the brain following sensory feedback, and after central processing, an appropriate motor response is made. Some commentators have emphasised the predominant importance of autonomic symptoms in the detection of the onset of hypoglycaemia. This premise is based partly on the laboratory-based observation during experimental hypoglycaemia in non-diabetic subjects that autonomic symptoms commence at a higher blood glucose concentration (∼0.5 mmol/l) than neuroglycopenic symptoms (Mitrakou et al. 1991). In everyday experience reported by people with insulin-treated diabetes, a rapid decline in blood glucose seldom permits a subjective distinction to be made between these different thresholds for the development of autonomic and neuroglycopenic symptoms, and people treated with insulin identify both types with equal frequency as their initial warning symptoms (Hepburn et al. 1992). It has been assumed that because neuroglycopenia may interfere with cognitive func- tion, this will affect the individual’s ability to perceive and interpret neuroglycopenic cues such as the inability to concentrate, drowsiness or difficulty with mentation. This may be true when a falling blood glucose is not treated and is allowed to drop to a level associated with severe neuroglycopenia, but most patients detect (and often rely upon) neurogly­ copenic symptoms during early hypoglycaemia, and rate these as important as autonomic symptoms in providing a warning. It is the initial perception of any symptom of hypo­ glycaemia, irrespective of whether this is autonomic, neuroglycopenic or simply a vague sensation of apprehension or loss of well-being (a commonly described early feature) that constitutes ‘awareness’ of hypoglycaemia. Only the initial warning symptoms are important in this respect, and not the total spectrum or absolute number of symptoms, some of which occur too late to have any value in alerting the patient to the impending risk of a falling blood glucose. A major difference between the autonomic and the neuro- glycopenic symptomatic response is that, once triggered, the autonomic response quickly reaches a maximum intensity which then gradually declines with time, whereas the neuroglycopenic response becomes more profound the further the blood glucose falls. This qualitative difference in response becomes important if early cues are ignored or are not detected, as progressive neuroglycopenia will eventually interfere with the individual’s ability to identify and self-treat the low blood glucose. Symptom perception is also discussed in Chapter 16. When a person is fully awake, alert and on guard against possible hypoglycaemia, this symptomatic warning system generally operates very effectively (Chapter 2). However, there are many times in everyday life when the symptoms may either be diminished or are disregarded. This is particularly so during sleep when symptoms are seldom detected, or they may be ignored if a person is distracted by other activities, such as watching an inter- esting programme on television, participating in sport or concentrating on a challenging task. Circumstances can modify the value of specific warning symptoms, making them difficult to interpret as features of hypoglycaemia. Examples include sweating on a hot day, shivering when the weather is cold or feeling drowsy during a boring meeting! All of these may represent early hypoglycaemia but are attributed to other causes by the affected person. 118 Hypoglycaemia in Clinical Diabetes

Box 6.1 Factors influencing normal awareness of hypoglycaemia

Internal External

Physiological Drugs Recent glycaemic control Beta-adrenoceptor blockers (non-selective) Degree of neuroglycopenia Hypnotics, tranquillisers Symptom intensity/sensitivity Alcohol Psychological Environmental Focused attention Posture Congruence; denial Distraction Competing explanations Education Knowledge Symptom belief

A list of the factors that influence normal awareness of hypoglycaemia is shown in Box 6.1. The nature and intensity of symptoms can vary considerably and the value of individual symptoms as warning cues may not be constant in any single individual. This is often not appreciated in the assessment of research findings, and it is difficult to extrapo- late the careful measurement of symptomatic responses to hypoglycaemia in studies per- formed in a laboratory setting, to the hurly-burly of everyday life. Warning symptoms provide internal cues, but most people with insulin-treated diabetes also rely on external cues based on their experience of the timing of insulin administra- tion in relation to food, the effect of delaying meals or the amount of food ingested, the effect of exercise on blood glucose and many other factors that can influence short-term glycaemic control. These cues are supplemented by blood glucose monitoring, which gives an exact and objective measure of prevailing glycaemia. Further useful feedback may be obtained from observers such as relatives or friends, many of whom become adept at noticing early neuroglycopenia before the onset of the person’s subjective warning symptoms. ‘Awareness’ of hypoglycaemia is therefore distilled from a combination of sources, and has to be learned by people with newly diagnosed diabetes commencing insulin therapy. They have no previous experience of symptoms of hypoglycaemia and must receive appropriate education on the potential range of symptoms. Through experi- ence they will recognise the cluster of symptoms peculiar to themselves, because symp- toms are idiosyncratic. Awareness of hypoglycaemia therefore assists in protecting the individual from the risk of an unexpected fall in blood glucose. When awareness of hypoglycaemia becomes impaired or is absent while a person is awake, the individual becomes progressively vulnerable to the development of severe hypoglycaemia (requiring help for recovery).

IMPAIRED AWARENESS OF HYPOGLYCAEMIA (IAH)

Definition No satisfactory or comprehensive definition of impaired hypoglycaemia awareness has been suggested or agreed to date. Some laboratory-based studies of experimental hypogly- Chapter 6: Impaired Awareness of Hypoglycaemia 119 caemia have used arbitrary definitions based on witnessed observations of subjects who fail to develop classic features of hypoglycaemia, or the failure of physiological or hormo- nal responses to exceed twice the standard deviation from mean basal levels. These are statistical devices, which take no account of subjective reality, require the application of sophisticated and unphysiological glucose clamp procedures, and have little direct applica- tion to clinical management. Many research publications of IAH do not define what this is or how it has been assessed. Astonishingly, most comparative research trials of different insulins have completely ignored the state of awareness of hypoglycaemia of the partici- pants. No international consensus currently exists regarding a practical definition of IAH, which may mean different things to different people. The semantics associated with ‘aware- ness’ of hypoglycaemia are difficult to comprehend and potentially confusing, which compounds the problem. The dictionary definition of being ‘aware’ includes ‘being informed, having knowledge, being cognisant and being conscious’ of something. The last is probably most relevant to ascertaining the state of awareness of hypoglycaemia. The present author would propose that the clinical definition of IAH should be based primarily upon a diminished ability to perceive the onset of acute hypoglycaemia. However, the syndrome that constitutes IAH defies precise definition because it may also include an altered symptom profile, reduced intensity and number of symptoms and in some cases a failure of the affected individual to interpret symptoms. All of these abnormalities should be sought when attempting to identify and characterise this problem. Because very few people have total loss of symptoms of hypoglycaemia, impaired awareness of hypoglycae- mia is a preferable terminology to ‘hypoglycaemia unawareness’.

Classification and natural history Hepburn et al. (1990) subdivided hypoglycaemia awareness into three categories: normal, partial and absent awareness. These were defined as follows:

• Normal awareness: the individual is always aware of the onset of hypoglycaemia. • Partial awareness: the symptom profile has changed with a reduction either inthe intensity or in the number of symptoms or in both. The individual may be aware of some episodes of hypoglycaemia but not of others. • Absent awareness: the individual is not aware of any episode of hypoglycaemia.

Although a subdivision into partial and absent awareness may be artificial, it reflects the natural history of this clinical problem, illustrating the gradual progression of this disability, and emphasising that in some patients the abnormality is severe (absent aware- ness) although total absence of clinical manifestations of hypoglycaemia (particularly the neuroglycopenic features) is exceedingly rare (Gold et al. 1994; Clarke et al. 1995). The problem may not be simply an inability to generate symptoms, but rather that the time during which one or more warning symptoms appears and can be detected is extremely short, affording a very limited opportunity for the affected individual to take avoiding action and self-treat. Some patients describe how the onset of hypoglycaemia has become much more rapid and progresses quickly to overt neuroglycopenia in contrast to their pre­ vious experience. However, impaired awareness does not necessarily evolve into total unawareness of hypoglycaemia, and may vary in severity over time, presumably because of major influences of environmental factors on the generation and perception of symptoms. 120 Hypoglycaemia in Clinical Diabetes

Clinical identification To ascertain the state of hypoglycaemia awareness the person must be in a physical state in which recognition of the onset of hypoglycaemia is possible, i.e. fully awake. If their conscious level is reduced (as when asleep, intoxicated, inebriated or sedated) they are not able to perceive the subjective symptoms of hypoglycaemia. They must also have had previous experience of hypoglycaemia, and to assess their present awareness status, prefer- ably within a relatively recent time period such as the preceding year. Comparison can then be made with earlier experience of symptomatic hypoglycaemia. A diagnosis of IAH cannot be entertained if a patient has never been exposed to acute hypoglycaemia or their first experience of hypoglycaemia has been very recent. Because hypoglycaemia awareness and its impairment is a continuum ranging from normality to complete inability to detect the onset of hypoglycaemia, an evaluation of this condition should consider alteration in symptom intensity in addition to the subjective perception of hypoglycaemia. In clinical practice a careful history is essential in determining whether reduced symp- tomatic warning of hypoglycaemia is a significant problem, and whether this is occurring consistently. Patients who assert that they have a problem with perceiving the onset of symptoms of hypoglycaemia are generally correct in holding this belief (Clarke et al. 1995), so that the identification of impaired awareness of hypoglycaemia can reliably be based on clinical history. Validated scoring systems to assess awareness of hypoglycaemia have been described by Gold et al. (1994) and Clarke et al. (1995), which have good concordance in adults (Geddes et al. 2007). Detailed questioning of a patient about his or her ability to detect the onset of hypoglycaemic symptoms may need to be supplemented by questioning close relatives, who often report a much higher rate of severe hypoglycaemia (Heller et al. 1995; Jorgensen et al. 2003). This will provide a witnessed description of how hypogly- caemia develops in an individual patient, with information on its true frequency and sever- ity. People with IAH often underestimate the frequency of severe hypoglycaemia, partly because of post-hypoglycaemia amnesia. Supportive information can be derived from simultaneous inspection of an individual’s blood glucose results. In people with type 1 diabetes with IAH, asymptomatic biochemical hypoglycaemia (capillary blood glucose <3.5 mmol/l) occurs more frequently (ranging from 2 to 7-fold) during routine blood glucose monitoring (Gold et al. 1994; Clarke et al. 1995; Choudhary et al. 2010; Schopman et al. 2011), which may alert the clinician to the possibility that an individual is developing this problem. Continuous glucose monitoring (CGM) has demonstrated that hypoglycaemia is often undetected and studies have sug- gested that asymptomatic biochemical events are 4-fold higher in people with IAH com- pared to those with normal awareness (Kubiak et al. 2004; Giménez et al. 2009). However, in a different study of people with type 1 diabetes, retrospective analysis of CGM records for up to 72 hours did not identify those with IAH, who had a similar frequency and dura- tion of biochemical hypoglycaemia as those with normal awareness despite having evi- dence of more asymptomatic biochemical hypoglycaemia during prospective self-monitoring of capillary blood glucose (Choudhary et al. 2010).

PREVALENCE OF IMPAIRED AWARENESS OF HYPOGLYCAEMIA

IAH is common in people treated with insulin. Although the chronic form of this acquired condition mainly affects those with type 1 diabetes, an identical syndrome does eventually Chapter 6: Impaired Awareness of Hypoglycaemia 121

Table 6.1 Prevalence of impaired awareness of hypoglycaemia in population studies of type 1 diabetes

Number of Impaired awareness Country patients of hypoglycaemia (%) Reference

Scotland 302 23 Hepburn et al. (1990) Germany 523 25 Mühlhauser et al. (1991) Denmark 411 27 Pramming et al. (1991) USA 628 20 Orchard et al. (1991) Scotland 518 19.5 Geddes et al. (2008)

emerge in people with type 2 diabetes who have been treated with insulin for several years (Hepburn et al. 1993a). Prevalences of 8–10% have been reported in people with insulin- treated type 2 diabetes (Henderson et al. 2003; Schopman et al. 2010). People with type 1 diabetes and IAH were significantly older and had a longer duration of diabetes than those with normal awareness (Geddes et al. 2008). IAH is more commonly associated with type 1 diabetes because many patients with type 2 diabetes who require insulin therapy do not survive for a sufficiently long period to permit this complication to develop. It is not known whether IAH occurs in people with type 2 diabetes treated with oral anti- diabetes agents, such as the insulin secretagogues. It appears to be primarily an acquired abnormality that is associated with insulin therapy. Reduced warning symptoms of hypoglycaemia (of varying severity) occur in approximately one-quarter of all adults with type 1 diabetes. Cross-sectional population surveys in different European and North Amer- ican populations with type 1 diabetes, using similar methods of assessment, have given remarkably consistent estimates (Table 6.1). IAH becomes more common with increasing duration of insulin-treated diabetes (Hepburn et al. 1990), and almost 50% of patients describe hypoglycaemia without warning symptoms after 25 years or more of treatment (Pramming et al. 1991) (Figure 6.3). IAH is described in children with type 1 diabetes (Barkai et al. 1998; Ly et al. 2009) but is probably lower than in adults and is difficult to define or to evaluate because very young children are unable to describe symptoms of hypoglycaemia.

Frequency of associated severe hypoglycaemia It is apparent that impaired awareness of hypoglycaemia is a major risk factor for severe hypoglycaemia. In the Diabetes Control and Complications Trial (DCCT), 36% of all episodes of severe hypoglycaemia occurred with no warning symptoms in patients while they were awake (The DCCT Research Group 1991). In a population study in Edinburgh, retrospective assessment of the frequency of severe hypoglycaemia revealed that 90% of patients with IAH had experienced severe hypoglycaemia in the preceding year, compared to 18% in a comparable group who had retained normal awareness (Hepburn et al. 1990). Prospective studies have confirmed the increase in frequency of mild and severe hypogly- caemia associated with IAH (Gold et al. 1994; Clarke et al. 1995; Geddes et al. 2008), with a 6-fold higher frequency of severe hypoglycaemia being documented in people with impaired awareness (Gold et al. 1994; Geddes et al. 2008) (Figure 6.4). 122 Hypoglycaemia in Clinical Diabetes

100

50 Patients (%) Patients

0 0–9 10–19 20–29 30–39 ≥40 (a) Diabetes duration (years)

100

50 Patients (%) Patients

0 0–9 10–19 20–29 30–39 ≥40 (b) Diabetes duration (years)

100

50 Patients (%) Patients

0 0–9 10–19 20–29 30–39 ≥40 (c) Diabetes duration (years)

Figure 6.3 Comparisons between the duration of diabetes and the percentage of 411 type 1 diabetic patients reporting (a) changes in symptoms of hypoglycaemia, (b) sweating and/or tremor as one of the two cardinal autonomic symptoms of hypoglycaemia, and (c) severe hypoglycaemic episodes without warning symptoms. Values are medians; shaded areas show 95% confidence limits. (Source: Pramming S 1991. Reproduced with permission from John Wiley & Sons, Ltd). Chapter 6: Impaired Awareness of Hypoglycaemia 123

100 2.36 2.5 Normal awareness of hypoglycaemia 80 Impaired awareness 2.0 of hypoglycaemia 60 1.5 50.5%

40 1.0 20.1% 20 0.38 0.5 Percentage of subjects of Percentage Number of events per year Number of events 0 0.0 Percentage Events

Figure 6.4 Prevalence and incidence of severe hypoglycaemia in 518 adults with type 1 diabetes with normal or impaired awareness of hypoglycaemia. (Source: Geddes J 2008. Reproduced with permission from John Wiley & Sons, Ltd).

Box 6.2 Impaired awareness of hypoglycaemia: possible mechanisms CNS adaptation Chronic exposure to low blood glucose • glucose clamp (2.9 mmol/l) for 56 hours in non-diabetic subjects • insulinoma in non-diabetic patients • strict glycaemic control in people with diabetes Recurrent transient exposure to low blood glucose • antecedent hypoglycaemia

CNS glucoregulatory failure • counterregulatory deficiency (hypothalamic defect?) • hypoglycaemia associated (central) autonomic failure (HAAF)

Peripheral nervous system dysfunction • peripheral autonomic neuropathy • reduced peripheral adrenoceptor sensitivity

PATHOGENESIS OF IMPAIRED AWARENESS OF HYPOGLYCAEMIA

Possible mechanisms underlying impaired awareness of hypoglycaemia may be multifacto- rial and are listed in Box 6.2.

Altered glycaemic threshold for initiation of symptoms Symptoms of hypoglycaemia commence when the blood glucose reaches a specific level, and although this threshold may differ between individuals, it is usually constant and reproducible in the non-diabetic state (Vea et al. 1992). This blood glucose threshold for 124 Hypoglycaemia in Clinical Diabetes symptoms can be modified by protracted hypoglycaemia (Boyle et al. 1994) and is not fixed in people with diabetes who are treated with insulin, with its dynamic nature being demonstrated in various situations. In clinical practice, it has long been recognised that people with insulin-treated diabetes who have poor glycaemic control experience symp- toms of hypoglycaemia when their blood glucose declines within a hyperglycaemic range (Maddock and Krall 1953) and this has been shown to be associated with the onset of hypoglycaemic symptoms at a significantly higher blood glucose (4.3 mmol/l) compared to non-diabetic subjects (2.9 mmol/l) (Boyle et al. 1988). Conversely, strict glycaemic control modifies the glycaemic threshold for the onset of symptoms, which do not com- mence until blood glucose has declined to a lower level than that required in less well- controlled patients to initiate a symptomatic response (see Chapter 7). The terminology that is used in relation to a change in the glycaemic threshold is poten- tially confusing. When a lower blood glucose is required to initiate a response, whether symptomatic, physiological or counterregulatory, the glycaemic threshold is said to be raised or elevated; that is, a more profound hypoglycaemic stimulus is necessary to trigger the relevant response. Thus, strict glycaemic control raises the glycaemic threshold for the onset of symptoms, which do not occur until blood glucose has declined to a much lower concentration than would be observed in non-diabetic subjects. For many years, clinicians have recognised that the glycaemic threshold for the onset of hypoglycaemic symptoms is higher in patients with a long duration of type 1 diabetes who require a much lower blood glucose to provoke a symptomatic response. Lawrence (1941) wrote that ‘as years of insulin life go on, sometimes only after 5–10 years, I find it almost the rule that the type of insulin reactions change, the premonitory autonomic symp- toms are missed out and the patient proceeds directly to the more serious manifestations affecting the central nervous system’. He astutely suggested that ‘the tissues may become attuned to a lower sugar concentration’. Recent studies in animals and humans have shown that the brain does adapt to chronic exposure to low blood glucose (see below) but this may not be beneficial to the individual with diabetes who is treated with insulin, i.e. it is a maladaptive response. An early study by Sussman et al. (1963) – revisited and extended by Hepburn et al. (1991) – showed that diabetic patients who had self-reported unawareness of hypoglycae- mia did mount a sympathoadrenal response to acute hypoglycaemia, but that this occurred at a lower blood glucose concentration than comparable diabetic subjects who had normal symptomatic awareness (Figure 6.5). However, the autonomic response was preceded by the development of overt neuroglycopenia, which interfered with perception of the auto- nomic warning symptoms when they did eventually occur. This sequence of responses disrupts the ability of the individual subject to take appropriate action to self-treat low blood glucose. Similar findings have been reported by others (Grimaldiet al. 1990; Mokan et al. 1994; Bacatselos et al. 1995). In these studies, the counterregulatory hormonal responses to hypoglycaemia were delayed (Grimaldi et al. 1990; Hepburn et al. 1991) and their glycaemic thresholds had also shifted to occur at lower blood glucose concentrations (Mokan et al. 1994). This is consist- ent with the reported observation that IAH co-segregates with counterregulatory hormonal deficiency in people with longstanding type 1 diabetes (Ryder et al. 1990). In addition, Mokan et al. (1994) reported that cognitive dysfunction and neuroglycopenic symptoms in people with IAH occurred at lower blood glucose levels than in people with type 1 diabetes who had normal awareness. This suggested that people with impaired awareness can func- tion effectively at very low blood glucose concentrations, when symptoms and cognitive impairment would be evident in non-diabetic and aware diabetic subjects. This was clearly Chapter 6: Impaired Awareness of Hypoglycaemia 125

3.0

2.5

2.0

1.5

1.0 Plasma glucose (mmol/l)

0.5

0 Non-diabetic Aware Unaware Neuropathic subjects diabetic diabetic diabetic patients patients patients

Figure 6.5 Venous blood glucose concentrations for the onset of the autonomic reaction in response to insulin-induced hypoglycaemia in individual non-diabetic control subjects, and in type 1 diabetic patients with normal and impaired awareness of hypoglycaemia and with autonomic neuropathy. Mean + standard error (SE) are shown for each group. (Source: Data derived from Hepburn et al. (1991) and reproduced from Hypoglycaemia and Diabetes (eds B.M. Frier and B.M. Fisher), © 1991 Edward Arnold, by permission of Edward Arnold Publishers Ltd). demonstrated by Zammitt et al. (2008) who showed that subjects with type 1 diabetes with IAH performed better on several cognitive tasks during hypoglycaemia than matched sub- jects with normal awareness, and that they recovered more quickly. Some people with insulin-treated diabetes remain lucid, with no evidence of impaired cognitive function, when their blood glucose is low (often well below 3.0 mmol/l). The altered glycaemic threshold delays the onset of warning symptoms and cognitive dysfunction until the blood glucose falls to a dangerously low level, which is extremely undesirable when striving for safe clinical management of insulin-treated diabetes. The potential risk of this cerebral adapta- tion at low blood glucose levels is apparent: it is akin to walking along the edge of a cliff on a dark night where the margin for error is dangerously small. With such a narrow gly- caemic warning zone the propensity to rapidly develop severe neuroglycopenia is high. The results of these laboratory-based experimental studies of diabetic patients who have established IAH are consistent with clinical observations of people with this acquired problem. At one moment they appear to be cerebrating normally (despite their blood glucose being low), then they rapidly become confused or drowsy, often with a vacant or dazed appearance and an inertia to seek some form of carbohydrate to reverse the neuro- glycopenia. They may have to rely on relatives, friends or colleagues to identify the hypoglycaemia and provide treatment. This becomes a serious emergency if the patient is alone or if the insidious, but often rapid, onset of neuroglycopenia goes unobserved. This explains the increased risk of developing severe hypoglycaemia, and the higher rates reported in people with IAH. 126 Hypoglycaemia in Clinical Diabetes

Studies examining the effects of strict glycaemic control on symptomatic and counter- regulatory responses to hypoglycaemia have also demonstrated a similar shift in glycaemic threshold for autonomic symptoms and an acute sympathoadrenal response. However, the effect on glycaemic thresholds for neuroglycopenic symptoms and cognitive dysfunction remains controversial (see Chapter 7).

Peripheral autonomic neuropathy For many years, peripheral autonomic neuropathy was thought to be the principal cause of impaired awareness of hypoglycaemia (Hoeldtke et al. 1982). This was based on the assump- tion that the diminished secretion of epinephrine in response to hypoglycaemia (Hilsted et al. 1981; Bottini et al. 1997) would either prevent the generation of autonomic symptoms (such as sweating or a pounding heart) or reduce their intensity, resulting in an inability to perceive the onset of hypoglycaemia. Thus, it was assumed that autonomic neuropathy would interfere with the normal physiological responses stimulated by autonomic activation. There are various reasons why this hypothesis is unlikely:

• Although epinephrine can augment the intensity of a few autonomic symptoms of hypoglycaemia, it has a very limited role in their generation, which is modulated by sympathetic neural activation; the reduced secretory response of epinephrine in auto- nomic neuropathy is compensated for by an increase in sensitivity of peripheral beta- adrenoceptors (Hilsted et al. 1987). • Diabetic subjects with autonomic neuropathy have normal physiological responses and experience typical autonomic symptoms during hypoglycaemia (Hilsted et al. 1981; Hepburn et al. 1993b), and no relationship has been found between autonomic dysfunc- tion and hypoglycaemic symptoms (Berlin et al. 1987). • Impaired awareness of hypoglycaemia co-segregates with deficient counterregulatory hormonal responses and not with autonomic neuropathy (Ryder et al. 1990). • The prevalence of autonomic neuropathy is similar in patients with type 1 diabetes of long duration (more than 15 years), whether or not they have IAH (Hepburn et al. 1990). • Although IAH is a major risk factor for severe hypoglycaemia, the latter is either no more common in type 1 diabetic patients with autonomic neuropathy (Bjork et al. 1990; The DCCT Research Group 1991), or is only modestly increased (Stephenson et al. 1996). • Autonomic neuropathy is not a determinant of whether glycaemic thresholds for auto- nomic (including symptomatic) responses to hypoglycaemia are affected by antecedent hypoglycaemia (Dagogo-Jack et al. 1993).

Both IAH and peripheral autonomic neuropathy are common in people with type 1 diabetes of long duration, and frequently coexist. This does not prove a causal relationship, and it would appear that peripheral autonomic dysfunction does not have a prominent role in the pathogenesis of this syndrome. Maladaptation of tissue sensitivity to catecholamines may contribute to the development of IAH even though autonomic neuropathy is not present. Reduced sensitivity of cardiac beta-adrenoceptors to catecholamines has been observed in patients with type 1 diabetes who have IAH (Berlin et al. 1987). Hypoglycaemia per se reduces beta-adrenergic sensitiv- ity in type 1 diabetes (Fritsche et al. 1998), and this sensitivity is increased after avoidance of hypoglycaemia for 4 months in people who have impaired awareness (Fritsche et al. 2001). The improved beta-adrenergic sensitivity correlated with a rise in autonomic Chapter 6: Impaired Awareness of Hypoglycaemia 127

Box 6.3 Acquired syndromes associated with hypoglycaemia in type 1 diabetes • Counterregulatory deficiency • Impaired awareness of hypoglycaemia • Altered glycaemic thresholds for counterregulatory and symptomatic responses symptom scores. Although it has been suggested that people with IAH may have reduced sensitivity to beta agonists, another study has shown that beta-adrenoceptor sensitivity is preserved (de Galan et al. 2006).

Hypoglycaemia-associated autonomic failure The co-segregation of impaired hypoglycaemia awareness with counterregulatory defi- ciency suggests that they share a common underlying pathogenetic mechanism. These acquired abnormalities associated with hypoglycaemia in type 1 diabetes (Box 6.3) are characterised by a high frequency of severe hypoglycaemia and a common pathophysio- logical feature, namely the elevated glycaemic thresholds (or lower blood glucose concen- trations) that are required to trigger symptomatic and hormonal secretory responses. In other words, more profound hypoglycaemia is necessary to produce the usual symptomatic and counterregulatory responses to acute hypoglycaemia. Cryer (1992) has designated this group of abnormalities as a form of hypoglycaemia-associated autonomic failure (HAAF), and has speculated that recurrent severe hypoglycaemia may be the primary problem which establishes a vicious circle (Figure 6.6). It seems likely that this defect resides within the central nervous system. The possible mechanisms underlying HAAF and related hypogly- caemia syndromes in diabetes have been reviewed in detail (Cryer 2005).

Central nervous system adaptation to hypoglycaemia Although the human brain is dependent on a continuous supply of glucose for normal function, it can adapt to prolonged exposure to hypoglycaemia. This adaptation process takes at least several hours and possibly a few days to occur. Short-term exposure to acute hypoglycaemia (blood glucose 2.5 mmol/l) for 60 minutes in non-diabetic subjects showed no improvement in cognitive function and no reduction in symptom scores during this brief time interval (Gold et al. 1995a). However, when non-diabetic subjects were subjected to chronic hypoglycaemia (blood glucose 2.9 mmol/l) for 56 hours, using a glucose clamp, significant cerebral adaptation did occur (Boyle et al. 1994). The responses to acute hypoglycaemia (blood glucose 2.5 mmol/l) were compared before and after the period of chronic hypoglycaemia. Brain glucose uptake was initially reduced when blood glucose was below 3.6 mmol/l, but after a period of chronic hypoglycaemia uptake was preserved and cerebral function was maintained (Figure 6.7), demonstrating an effect of cerebral adaptation to chronic neuroglycopenia. The glycaemic thresholds for the onset of symp- toms, counterregulatory hormonal secretion and cognitive dysfunction, were all modified and occurred at much lower blood glucose concentrations. A similar phenomenon has been observed in non-diabetic patients who had an insulinoma causing chronic hypoglycaemia; symptomatic responses to acute hypoglycaemia were blunted and counterregulatory hor- monal responses were impaired but cognitive function was unaffected (Mitrakou et al. 128 Hypoglycaemia in Clinical Diabetes

Type 1 diabetes

Decreased Insulin Insulin deficiency glucagon therapy response

Strict glycaemic Recurrent severe hypoglycaemia control

Hypoglycaemia Counterregulatory unawareness deficiency

Central (CNS) autonomic failure

Decreased Decreased symptom adrenaline response response

Figure 6.6 Schematic diagram of the concept of hypoglycaemia-associated autonomic failure (HAAF). (Source: Data derived from Cryer 1992).

1993). Surgical removal of the insulin-secreting tumour reversed these abnormalities, indicating that they had resulted from cerebral adaptation to chronic hypoglycaemia. IAH can be associated with strict glycaemic control (see Chapter 7), but significant modification of the symptomatic response to hypoglycaemia does not usually occur until the glycated haemoglobin concentration is maintained within the non-diabetic range (Boyle et al. 1995; Kinsley et al. 1995; Pampanelli et al. 1996), where exposure to hypoglycaemia (often asymptomatic) becomes much more likely using conventional insulin regimens. This is less of a problem with CSII when biochemical hypoglycaemia is diminished and is eradicated completely following whole pancreas or islet cell transplantation. In practice, only a small proportion of (usually) highly motivated people with insulin-treated diabetes can sustain super-optimal glycaemic control for protracted periods, so this ‘acute’ form of IAH is probably relatively uncommon. Strict glycaemic control that increases exposure to intermittent hypoglycaemia alters the glycaemic thresholds for the development of counterregulatory hormones and symptoms (Chapter 7), so that a lower blood glucose concentration is required to trigger these responses. The median daily blood glucose in these individuals is relatively low, and it was demonstrated many years ago that the frequency of biochemical (and symptomatic) hypoglycaemia is significantly greater than in insulin-treated people with less good overall control who have a higher daily median blood glucose (Thorsteinsson et al. 1986). Boyle et al. (1995) showed that patients who had near normal HbA1c values can maintain their uptake of glucose by the brain during hypoglycaemia, so preserving cerebral metabolism, reducing the counterregulatory responses to hypoglycaemia and diminishing symptomatic awareness. Although this capacity to maintain, and even increase, cerebral blood glucose Chapter 6: Impaired Awareness of Hypoglycaemia 129

50 ) –1

40 · min –1 * 30

(μmol · 100 g (μmol · 100 20

6000

4000

2000 *

Adrenaline (pmol/l)Adrenaline Brain glucose uptake *

0

30

20

* 10 * Total symptom score symptom Total

0 4.72 4.16 3.61 3.05 2.50 Nominal glucose step (mmol/l)

Figure 6.7 Rates of brain glucose uptake, epinephrine (adrenaline) concentration and total symptoms of hypoglycaemia in non-diabetic subjects before and after prolonged hypoglycaemia. Initial day of investigation (hatched); after 56 hours of hypoglycaemia (solid). *Significant difference from baseline for each of the 2 days. (Source: Boyle PJ 1997. Reproduced with kind permission from Springer Science+Business Media). uptake during hypoglycaemia protects the brain in these circumstances, it suppresses the normal symptomatic warning of responses and may risk the development of much more profound neuroglycopenia. During hypoglycaemia, glucose transport into the brain becomes rate-limiting, and brain energy metabolism deteriorates. In rodents, the transport of glucose across the blood–brain barrier is increased after several days of chronic hypoglycaemia (McCall et al. 1986). Further studies in rats of glucose transport activity across the blood–brain barrier have shown that when blood glucose was kept below 2.0 mmol/l for several days, changes in expression of the glucose transporter, GLUT-1, in brain microvasculature occurred in response to the chronic hypoglycaemia (Kumagai et al. 1995). This increase in GLUT-1 activity was responsible for the compensatory increase in glucose transport across the blood–brain barrier. Studies such as these are not possible in humans, and the results on 130 Hypoglycaemia in Clinical Diabetes glucose uptake and brain metabolism are mixed; while initial studies suggested recurrent exposure to hypoglycaemia might lead to an increase in glucose uptake and metabolism, subsequent studies have not been able to confirm this finding.

Antecedent (episodic) hypoglycaemia It has been recognised for many years that severe hypoglycaemia is a marker for further episodes of severe hypoglycaemia, and that one episode can influence the clinical manifesta- tions of another occurring soon afterwards (Severinghaus 1926). In recent years, several studies have shown that the symptomatic and counterregulatory responses to an episode of acute hypoglycaemia are diminished if a preceding (or antecedent) episode of hypoglycaemia has occurred within the previous 24 hours. Several studies have been performed in non- diabetic subjects (Table 6.2) and in people with insulin-treated diabetes (Table 6.3). Although these studies differ considerably in design and methods of inducing hypoglycaemia, in general it appears that antecedent hypoglycaemia of a duration between one and two hours has a significant influence on the magnitude of the symptomatic and counterregulatory responses to subsequent hypoglycaemia occurring within the following 24–48 hours (Figure 6.8). The glycaemic thresholds for symptomatic and counterregulatory hormonal responses are altered by antecedent hypoglycaemia, and the degree to which subsequent responses are blunted are determined by the duration and depth of antecedent hypoglycaemia (Davis et al. 1997). Some of the physiological responses (e.g. sweating) may be blunted for longer than other responses following antecedent hypoglycaemia (George et al. 1995). Recurrent, short- lived (15 minutes) episodes of hypoglycaemia on four consecutive days had no effect on counterregulatory and symptomatic responses in non-diabetic subjects (Peters et al. 1995), so transient reductions in blood glucose may not produce this effect. Davis et al. (2000a) have observed that a short duration of antecedent hypoglycaemia (20 minutes to lower and raise blood glucose from 3.9 to 2.9 mmol/l with the blood glucose being maintained for 5 minutes at 2.9 mmol/l) did not affect symptomatic awareness of hypoglycaemia but did blunt the counterregulatory hormonal responses. Antecedent hypoglycaemia also reduces the counterregulatory responses to exercise on the following day, both in non-diabetic (Davis et al. 2000b) and type 1 diabetic subjects (Galassetti et al. 2003) and influences the metabolic responses, particularly diminishing endogenous glucose production in response to exercise. This may promote exercise-induced hypoglycaemia in type 1 diabetes. Some studies have examined the effect of antecedent hypoglycaemia on cognitive func- tion, but in many the methods of assessment were inadequate and insufficient to provide definitive evidence of a change in cognitive response. Although some maintain that the glycaemic threshold for cognitive dysfunction is not altered by hypoglycaemia (see Chapter 7), an increasing number of studies have suggested that this does shift to a lower blood glucose concentration in the same manner as the thresholds for autonomic and counterregu- latory responses (Veneman et al. 1993; Fanelli et al. 1998; Ovalle et al. 1998). Consistent with these observations, a study from Germany in non-diabetic men has shown that after a single episode of antecedent hypoglycaemia, subsequent hypoglycaemia had less effect on auditory-evoked brain potentials and short-term memory (Fruehwald-Schultes et al. 2000), demonstrating cerebral adaptation that preserves cognitive function. Nocturnal (episodic) hypoglycaemia, which is frequently not identified by patients, has been proposed asa mechanism for the induction of hypoglycaemia unawareness in people who give no history of recurrent hypoglycaemia (Veneman et al. 1993). The possible mechanisms of cerebral adaptation causing impaired awareness of hypoglycaemia are summarised in Box 6.4. Chapter 6: Impaired Awareness of Hypoglycaemia 131 Attenuated metabolic responses (reduced endogenous glucose production) symptoms and CR responses elevated CR responses days to depth of AH Symptom scores unaffected by short duration hypoglycaemia short-term memory loss affected CR responses week No effect on symptoms or CR responses Blunting of CR responses to exercise. Reduced symptoms and CR responses Nocturnal hypo: BG thresholds for Elevated BG thresholds for symptoms and Reduced adrenaline and sweating at 6 Magnitude of reduced CR response related Reduced CR responses in all studies. Auditory-evoked brain potentials and Elevated BG thresholds for symptoms and Physiological responses still reduced after 1 Effect of AH on subsequent responses to hypoglycaemia 4

×

min)

2.8) Clamp (2.8) Stepped clamp (2.3) Stepped clamp (2.8) Clamp (2.5) Clamp (2.9) Clamp (2.9) Exercise (90 Clamp (3.1) Stepped clamp (2.2) bolus iv. ( < Clamp (2.5) Test hypo: Test method and BG nadir (mmol/l)

h

h

h

h h and 6 days h h

daily hypos daily hypos h

24 18 4 consecutive 7 1.5 24 4 consecutive 2 and 5 days < 24 24 18–24 Interval before test hypoglycaemia h

min h h h h h h

min

15); 30 16); 5 10) min

h

= = =

h h

15 2 2 ( n ( n ( n 2.5 insulin ( < 2.8) < Clamp (3.0) 2 Clamp (2.2–2.8) 1 Clamp (3.2) 2 Clamp (3.0) 2 Clamp (2.9) 2 Clamp (2.9) 2 Clamp (2.9) 2 Stepped clamp (2.6) iv. bolus iv. injection of Method of induction, nadir BG (mmol/l) and duration of AH iv. infusion iv. (2.2–2.5) Stepped clamp (2.9) 8 9 8 9 10 30 10 10 31 16 10 No. of subjects Studies of antecedent hypoglycaemia in non-diabetic humans et

et al . et al . et al . et al . et al . et al . et al . et al . (1995) (1995) Schultes al . (2000) Simonson (1992) (1994) (1995) (2000a) (2000b) (1993) (1997) Cryer (1991) George Fruehwald- Widom and Mellman Robinson Peters Davis Davis Table 6.2 Table References AH, antecedent hypoglycaemia; BG, blood glucose; CR, counterregulatory (hormonal); intravenous iv, Veneman Veneman Davis Heller and 132 Hypoglycaemia in Clinical Diabetes increase in glucose need output autonomic responses dysfunction. Fewer symptomatic episodes of clinical hypoglycaemia. nocturnal hypoglycaemia (versus euglycaemia). Elevated BG threshold for cognitive dysfunction days neurophysiological responses Blunting of CR responses to exercise. 3-fold Reduced CR responses and hepatic glucose Elevated BG thresholds for symptoms and Reduced symptoms, CR and cognitive Less deterioration in cognitive function after Physiological responses diminished for 2 Reduced symptoms, CR and Effect of AH on responses to hypoglycaemia min

Clamp (3.0) Stepped clamp (2.6) Stepped clamp (2.5) Stepped clamp (2.5) Exercise for 90 Stepped clamp (2.2) Stepped clamp (1.7) Test hypo: Test method and BG nadir (mmol/l)

h

min h h h

60 15 2 days 48 5–10 24 3 days Interval before test hypoglycaemia h,

h h h h, twice h

× 3

mins weekly for 1 month nocturnal insulin ( < 2.2) Clamp (3.0) 2 Clamp (2.7) 2 Clamp (2.8) 2 Clamp (2.8) 2 Clamp (2.6) 3.5 Clamp (2.9) 2 iv. bolus iv. injection of Method of induction, nadir BG (mmol/l) and duration of AH (non- diabetic) 6 8 26 ( ± AN) 12 18 15 16 No. of subjects 13 Studies of antecedent hypoglycaemia in people with insulin-treated diabetes

et et al . et al . et al . et al . et al . (1993) et al . (1993) (1998) al . (2003) (1998) (1997) (1992) Dagogo-Jack Lingenfelser Fanelli Galassetti Table 6.3 Table Reference AH, antecedent hypoglycaemia; AN, autonomic neuropathy; BG, blood glucose; CR, counterregulatory (hormonal); intravenous iv, Ovalle George Davis Chapter 6: Impaired Awareness of Hypoglycaemia 133

Symptoms

Neuroendocrine Magnitude of response

Hypo 1 Hypo 2 Time (hours – days)

Figure 6.8 Schematic representation of the effect of antecedent hypoglycaemia on the neuroendocrine and symptomatic responses to subsequent hypoglycaemia.

Box 6.4 Mechanisms of cerebral adaptation causing impaired awareness of hypoglycaemia Symptomatic and Neuroendocrine responses to hypoglycaemia in insulin-treated diabetes are diminished in association with:

• strict glycaemic control (HbA1c in non-diabetic range) • antecedent (episodic) hypoglycaemia • chronic (protracted) hypoglycaemia

They may be restored by:

• relaxation of glycaemic control • scrupulous avoidance of hypoglycaemia

Although antecedent hypoglycaemia may induce transient IAH, it is unclear how this mechanism would induce chronic or prolonged loss of symptomatic perception. Although frequent, recurrent hypoglycaemia may have a contributory effect to inducing IAH; pre- sumably the hypoglycaemia has to be relatively protracted to induce prolonged cerebral adaptation, and the phenomenon is not limited to patients who have strict glycaemic control. The problem remains of explaining the induction of protracted or chronic hypogly- caemia unawareness, which often appears to be a permanent defect. Repetitive exposure to hypoglycaemia presumably ‘downregulates’ central glucose sensors and blunts the mechanisms that activate the counterregulatory responses within the hypothalamus. There is evidence of a permanent redistribution of regional cerebral blood flow in diabetic patients with a history of recurrent severe hypoglycaemia (MacLeod et al. 1994a) with, in particular, a relative increase in blood flow to the frontal lobes. This may represent a chronic adaptive response to protect vulnerable areas of the brain from recurrent neuroglycopenia. However, the changes in regional cerebral blood flow in response to controlled hypoglycaemia in patients with type 1 diabetes appear to occur independently of the state of awareness of hypoglycaemia (MacLeod et al. 1996). The EEG changes associated with modest hypogly- caemia are more pronounced in patients with type 1 diabetes who have IAH (Tribl et al. 1996). Most studies suggest that a diffuse functional abnormality is present in the anterior part of the brain in diabetes, and this may be implicated in the impaired perception of 134 Hypoglycaemia in Clinical Diabetes hypoglycaemia. The pre-frontal areas of the cortex are closely connected to sub-cortical areas, and localised dysfunction could theoretically reduce the ability of the brain to per- ceive symptomatic hypoglycaemia.

Neuroimaging studies Further insight may be provided by the application of neuroimaging. A study using positron emission tomography (PET) compared changes in global and regional brain glucose metabo- lism during euglycaemia and hypoglycaemia in 12 men with type 1 diabetes, six of whom had IAH (Bingham et al. 2005). Global brain glucose content fell during acute hypoglycae- mia with no difference apparent between those with normal and those with impaired aware- ness, implying that global glucose extraction is not enhanced by antecedent hypoglycaemia. Global neuronal activation was observed with hypoglycaemia in the aware patients, but was absent in the unaware, suggesting that cortical activation is a necessary correlate of hypogly- caemia awareness. Studies of regional brain activation that have focused on key areas involved in glucose homeostasis have provided further clues about the possible pathogenesis of IAH. Glucose uptake by the hypothalamus is reduced in people with IAH (Cranston et al. 2001). Antecedent hypoglycaemia was induced in non-diabetic adults to induce a state of counterregulatory failure; subsequent hypoglycaemia resulted in an increase in activity of the dorsal midline thalamus, a region that is thought to have an inhibitory role on counter- regulatory responses following antecedent hypoglycaemia (Arbelaez et al. 2008). This antecedent hypoglycaemia also showed a relative reduction in glucose metabolism in cortical areas that are involved in symptom perception, similar to an observation that had been made in people with IAH (Bingham et al. 2005). Activation of the amygdala is thought to be an unpleasant subjective experience associated with fear and anxiety. During acute hypoglycae- mia, [18F]-fluorodeoxyglucose PET scanning showed greater activation in the amygdala in people with normal awareness of hypoglycaemia compared to those with IAH (Dunn et al. 2007). By contrast, during hypoglycaemia in people with IAH a relative increase in activation was observed in the lateral orbitofrontal cortex. Activation of this area is thought to reduce appetite and limit an appreciation of danger that should be associated with hypoglycaemia (Dunn et al. 2007). These regional changes in people with IAH may be an example of ‘stress sensitisation’, whereby repeated exposure to a stress leads to a diminished response.

IMPAIRED AWARENESS OF HYPOGLYCAEMIA AND LONG- TERM EFFECT ON COGNITIVE FUNCTION

IAH is a major risk factor for severe hypoglycaemia, and patients with the chronic form of this condition have a 6-fold higher frequency (Gold et al. 1994; Geddes et al. 2008). It is possible therefore that IAH may be associated with evidence of a decline in cognitive function. Hepburn et al. (1991) noted that people with diabetes who had a history of IAH performed less well than those with normal awareness of hypoglycaemia on limited cogni- tive function testing, both during normoglycaemia and during hypoglycaemia. This sug- gested that an acquired cognitive impairment may have been superimposed upon an increased susceptibility to neuroglycopenia. A modest, but insignificant, decline in intel- lectual function was noted with progressive loss of hypoglycaemia awareness in a popula- tion study (MacLeod et al. 1994b). In one study, formal measurement of cognitive function during controlled hypoglycae- mia (blood glucose 2.5 mmol/l) showed that people with type 1 diabetes who had impaired Chapter 6: Impaired Awareness of Hypoglycaemia 135 hypoglycaemia awareness exhibited more profound cognitive dysfunction during acute hypoglycaemia than those with normal awareness, and that this persisted for longer following recovery of blood glucose (Gold et al. 1995b). By contrast, a later study of people with type 1 diabetes, which examined the rate of recovery of differing domains of cognitive function after hypoglycaemia, showed that during hypoglycaemia (blood glucose 2.5 mmol/l) cognitive function was much less affected in those with IAH, suggesting that cerebral adaptation to hypoglycaemia had occurred in these patients (Zammitt et al. 2008). These apparently discrepant results leave this issue unresolved. However, animal studies have shown that antecedent moderate hypoglycaemia can protect the brain against subse- quent severe hypoglycaemia with evidence of less neuronal damage (Puente et al. 2010). The relevance of these observations to humans is unknown.

HUMAN INSULIN

For 60 years after its discovery, insulin for therapeutic use was obtained from the pancreata of cattle and pigs. With the development of recombinant DNA technology it was possible to ‘genetically engineer’ molecules, and insulin was the first protein to be made in this way, becoming available for therapeutic use in the 1980s. Several of the animal insulin formula- tions were withdrawn, principally for commercial reasons, and human insulin rapidly became the most commonly prescribed form of insulin. The structure of human insulin differs from porcine insulin by a single amino acid and from bovine insulin by three amino acids. In initial trials it was not expected that human insulin would differ substantially in potency from animal insulin, but because human insulin was slightly purer than some of the animal insulins, patients were advised to reduce the dose by around 10% when convert- ing from animal to human insulin. Detailed pharmacokinetic studies comparing human and animal insulins did not demonstrate any major differences, but human insulin has a slightly faster onset of action, a slightly shorter duration of action, and is less immunogenic than equivalent animal insulins. Most clinical studies, conducted on a worldwide scale, showed no significant differences between human and animal insulins in their clinical application. In Switzerland, however, one group of clinicians reported encountering serious clinical problems with the use of human insulin in patients with type 1 diabetes (Teuscher and Berger 1987). In particular, they claimed that patients experienced more frequent hypogly- caemia and that warning symptoms were modified with human insulin, as a result of which many people were unable to perceive the onset of hypoglycaemia. A pathologist in the UK then claimed that the number of deaths from severe hypoglycaemia had increased since the introduction of human insulin (see Chapter 13). The evidence for this irresponsible state- ment did not withstand scrutiny, but anecdotal reports emerged in the UK of problems experienced by patients with human insulin, and a legal action was pursued against the insulin manufacturers, alleging that human insulin gave less warning of hypoglycaemia. Additional claims included allegations that human insulin may have caused personality changes in individuals and even other disease states such as multiple sclerosis. Litigation was abandoned in 1993 because of the lack of robust scientific evidence for these claims. However, this issue generated much controversy and heated debate and stimulated several studies comparing human with animal insulins, which are not reviewed here. A systematic review of the extensive literature related to this topic examined whether pub- lished evidence supported a difference in the frequency and awareness of hypoglycaemia induced by human and animal insulins (Airey et al. 2000). A total of 52 randomised con- trolled trials were identified, 37 of which were of double-blind design, whereas others 136 Hypoglycaemia in Clinical Diabetes reported hypoglycaemic outcomes as a secondary or incidental outcome during compara- tive investigations of efficacy or immunogenicity. Seven of the double-blind studies reported differences in frequency of hypoglycaemia or of symptomatic awareness, and four of the unblinded trials reported differences in hypoglycaemia. None of the four population time trend studies found any relationship between the increasing use of human insulin and hospital admission for hypoglycaemia or unexplained death among people with insulin- treated diabetes. The authors observed that the least rigorous scientific studies gave the greatest support to the premise that treatment with human insulin influences the frequency, severity or symptoms of hypoglycaemia. The report concluded that the published evidence did not support the contention that human insulin is responsible either for the alleged problems with IAH or for a higher risk of severe hypoglycaemia, but were unable to decide whether some of the phenomena associated with the use of human insulin resulted from stricter glycaemic control (Airey et al. 2000). In clinical practice there are undoubtedly a small number of people with insulin-treated diabetes in whom the use of human insulin has not been satisfactory, being associated with frequent and unpredictable hypoglycaemia and a diminished sense of well-being. Whether this is related to the different pharmacokinetics of human insulin or represents an idiosyn- cratic response in individuals is unclear, but such patients clearly wish to retain the option to use animal species of insulin, and the availability of the animal species of insulins should be maintained. No problem with symptomatic awareness of hypoglycaemia has been reported with the newer insulin analogues.

TREATMENT STRATEGIES

When IAH is therapy-related, that is, resulting from strict glycaemic control, the approach to management is relatively simple. The total insulin dose should be reduced, attention paid to the appropriateness of the insulin regimen, and overall glycaemic control should be relaxed. Liu et al. (1996) reported an improvement in symptomatic and counterregulatory hormonal responses to hypoglycaemia after 3 months of less strict glycaemic control in a small group of insulin-treated patients, in whom the mean HbA1c rose from 6.9 to 8.0%. Intensive insulin therapy is contraindicated in patients who have IAH and treatment goals have to be considered individually. It has been claimed that IAH (and to some extent counterregulatory hormonal defi- ciency) can be reversed by scrupulous avoidance of hypoglycaemia through meticulous attention to diabetic management (Cranston et al. 1994; Dagogo-Jack et al. 1994; Fanelli et al. 1994). The effect that this had on glycaemic thresholds for cognitive dysfunction and the recovery of counterregulatory hormonal secretion to hypoglycaemia differed between these studies, but all demonstrated an improved symptomatic response following avoidance of hypoglycaemia for periods varying from 3 weeks to one year. However, the studies can be criticised for the following reasons: • Only a small number of patients were studied. • The definition of IAH was based on an increased frequency of asymptomatic biochemi- cal hypoglycaemia, and with the exception of the study by Dagogo-Jack et al. (1994), was not based on having a history of hypoglycaemia unawareness. • In all studies there was a small but definite rise in glycated haemoglobin, suggesting that the improved symptomatic awareness was related primarily to relaxation of gly- caemic control. Chapter 6: Impaired Awareness of Hypoglycaemia 137

Although the scrupulous avoidance of hypoglycaemia is clearly desirable, and may help to reduce the severity of IAH, it is very difficult to achieve in clinical practice because it is extremely time-consuming and labour intensive both for patients and healthcare profes- sionals. The use of continuous subcutaneous insulin infusion (CSII) overnight instead of isophane (NPH) insulin at bedtime has been shown to be beneficial in people with IAH, improving warning symptoms and counterregulatory responses to hypoglycaemia, presum- ably by reducing exposure to nocturnal hypoglycaemia (Kanc et al. 1998). CSII was used for 24 months in a group of patients with IAH who had experienced two or more episodes of severe hypoglycaemia in the preceding 2 years. Fewer episodes of severe hypoglycaemia were recorded, although glycaemic control was unchanged and the participants reported an improved quality of life, while the symptomatic response to experimentally-induced hypoglycaemia also improved (Giménez et al. 2010). IAH can also benefit from islet cell transplantation. In one study a marked decline in prevalence occurred post-transplantation and the blood glucose level that was required to generate symptoms of hypoglycaemia rose from 2.3 to 3.2 mmol/l (Leitão et al. 2008). Various drugs have been tried to enhance or restore symptomatic warning in people with IAH. The ingestion of caffeine uncouples the relationship between cerebral blood flow and glucose utilisation via antagonism of adenosine receptors, causing relative neuroglycopenia and earlier release of counterregulatory hormones during moderate hypoglycaemia. The prior consumption of caffeine augments the symptomatic and counterregulatory hormonal responses to a modest reduction of blood glucose in non-diabetic subjects (Kerr et al. 1993), and a similar phenomenon occurs in people with type 1 diabetes following the ingestion of a dose of caffeine equivalent to two or three cups of coffee (Debrah et al. 1996). The reduction in cerebral blood flow is sustained, the counterregulatory response is augmented (Figure 6.9) and greater awareness of hypoglycaemia occurs (see Chapter 8). Functional MRI has shown that caffeine can restore regional brain activation that is usually lost during acute hypoglycaemia (Rosenthal et al. 2007). However, the large daily doses of caffeine that are required do not make this a practical proposition in treating IAH. The adenosine- receptor antagonist, theophylline, stimulates the secretion of catecholamines and reduces cerebral blood flow, and a single intravenous dose has been shown to enhance counterregu- latory hormone responses to hypoglycaemia, partially restore perception of hypoglycaemic symptoms and normalise glycaemic thresholds for haemodynamic and symptomatic responses in patients with type 1 diabetes with IAH (de Galan et al. 2002). It is not known whether the effects can be sustained with oral theophylline, as the development of tolerance to these drugs is common. Beta agonists (e.g. terbutaline) have been shown to significantly reduce nocturnal hypoglycaemia but at the cost of inducing morning hyperglycaemia (Raju et al. 2006; Cooperberg et al. 2008). Modafanil, a drug that reduces gaba-aminobutyric acid levels, has been shown to increase the symptoms of hypoglycaemia (Smith et al. 2004) but in clinical practice has not been used to treat IAH. The HypoCOMPaSS trial is a prospective, multi-center, randomised controlled trial in the UK, the primary aim of which was to compare the ability to avoid hypoglycaemia and reverse IAH using either an optimised subcutaneous insulin analogue multiple injection regimen or insulin pump therapy, with or without adjunctive real-time CGM (Little et al. 2012). Preliminary data suggest that reducing exposure to hypoglycaemia per se can help to reverse IAH and defective counterregulation (Leelarathna et al. 2013, in press). It is clearly desirable to avoid severe hypoglycaemia at all costs, and treatment strategies should be adopted to achieve this aim (Box 6.5). Frequent blood glucose monitoring is essential in affected patients, and may require occasional nocturnal measurements to detect 138 Hypoglycaemia in Clinical Diabetes

8000 800

400 2000 200

Adrenaline (pmol/l) Adrenaline 1000 hypoglycaemia (mm) hypoglycaemia

Symptom response to response Symptom 100

(a) (b) Caffeine Caffeine No caffeine No caffeine

Figure 6.9 Augmentation of the normal secretory response of epinephrine (adrenaline) to, and symp- tomatic response to acute hypoglycaemia (blood glucose 2.8 mmol/l) by the prior ingestion of caffeine in insulin-treated diabetic patients. Symptom intensity scores were estimated by visual analogue assess- ment. (Source: Derived from data in Debrah K et al., 1996).

Box 6.5 Treatment strategies for patients with impaired awareness of hypoglycaemia • Frequent blood glucose monitoring (including nocturnal measurements). • Avoid blood glucose values <4.0 mmol/l. • Set target range of blood glucose higher than for ‘aware’ patients (e.g. preprandial between 6.0–12.0 mmol/l; bedtime >8.0 mmol/l).

• Avoid HbA1c in the non-diabetic range. • Use predominantly short-acting insulins (e.g. basal-bolus regimen; insulin analogues). • Use Continuous Subcutaneous Insulin Infusion (CSII) with an insulin pump. • Regular snacks between meals and at bedtime, containing unrefined carbohydrate. • Appropriate additional carbohydrate consumption and/or insulin dose adjustment for premeditated exercise. • Learn to identify subtle neuroglycopenic cues to low blood glucose. low blood glucose during the night. CGM should have a role when the detection of low blood glucose becomes sufficiently reliable. A short pilot study in adolescents with type 1 diabetes and IAH showed an enhanced epinephrine response to experimental hypoglycae- mia in those using real-time CGM for 4 weeks, but it was not reported whether this reduced the frequency of hypoglycaemia or improved symptomatic awareness (Ly et al. 2011). Blood glucose awareness training (BGAT) was developed in the USA, with re-education of affected patients to recognise neuroglycopenic cues (Cox et al. 1995), and this psycho- educational programme has also been adapted and utilised in some European countries. It does improve the ability to detect hypoglycaemia in people with IAH but requires facilities and resources that are not available in most centres. However, training can now be delivered over the Internet (Cox et al. 2008). There is some evidence that the attitudes and behavioural responses of people with IAH are abnormal. It has been observed that many affected patients fail to modify their behav- Chapter 6: Impaired Awareness of Hypoglycaemia 139 iour to prevent, avoid or correct hypoglycaemia (Gold et al. 1994; Hepburn et al. 1994) and are reluctant to comply with recommended therapeutic measures to minimise hypogly- caemia (Smith et al. 2009). This may represent habituation to recurrent stress (in the form of hypoglycaemia) and psychological counselling may be of value for some patients to modify this potentially harmful behaviour that promotes hypoglycaemia risk. The avoid- ance of severe hypoglycaemia is paramount as this may exacerbate the problem, and the use of mostly short-acting insulin (and possibly insulin analogues) in basal-bolus regimens may be particularly useful in avoiding biochemical and symptomatic hypoglycaemia without compromising overall glycaemic control.

CONCLUSIONS • An inadequate symptomatic warning of hypoglycaemia is common in people with insulin-treated diabetes and is described as impaired awareness of hypoglycaemia or hypoglycaemia unawareness. It increases in prevalence with duration of insulin-treated diabetes. • In people who report IAH, asymptomatic hypoglycaemia occurs more frequently during routine blood glucose monitoring. This may alert the clinician to the possibility that an individual is developing this problem. • IAH may be associated with strict glycaemic control; significant modification of the symptomatic response occurs when the HbA1c concentration is within the non-diabetic range. • The mechanisms underlying IAH may be multifactorial. Possible mechanisms include chronic exposure to a low blood glucose, antecedent hypoglycaemia, and central auto- nomic and glucoregulatory failure. • Antecedent hypoglycaemia has a significant influence on the magnitude of the symp- tomatic and counterregulatory responses to subsequent hypoglycaemia occurring within the following 48 hours. • When IAH results from strict glycaemic control, the total insulin dose should be reduced, attention paid to the suitability of the insulin regimen, and overall glycaemic control should be relaxed. • IAH, and to some extent counterregulatory hormonal deficiency, can probably be reversed by scrupulous avoidance of hypoglycaemia through meticulous attention to diabetic management. • Intensive insulin therapy is contraindicated in patients who have IAH. The use of mostly short-acting insulins (including insulin analogues) in basal-bolus regimens may be particularly useful in avoiding biochemical and symptomatic hypoglycaemia without compromising overall glycaemic control. CSII may be valuable for some patients.

REFERENCES

Airey, C.M., Williams, D.R.R., Martin, P.G., Bennett, C.M.T. & Spoor, P.A. (2000) Hypoglycaemia induced by exogenous insulin – ‘human’ and animal insulin compared. Diabetic Medicine, 17, 416–432. Arbelaez, A.M., Powers, W.J., Videen, T.O., Price, J.L. & Cryer, P.E. (2008) Attenuation of counterregula- tory responses to recurrent hypoglycemia by active thalamic inhibition. A mechanism for hypoglycemia- associated autonomic failure. Diabetes, 57, 470–475. 140 Hypoglycaemia in Clinical Diabetes

Bacatselos, S.O., Karamitsos, D.T., Kourtoglou, G.I., Zambulis, C.X., Yovos, J.G. & Vyzantiadis, A.T. (1995) Hypoglycaemia unawareness in type 1 diabetic patients under conventional insulin treatment. Diabetes, Nutrition and Metabolism, 8, 267–275. Barkai, L., Vamosi, I. & Lukacs, K. (1998) Prospective assessment of severe hypoglycaemia in diabetic children and adolescents with impaired and normal awareness of hypoglycaemia. Diabetologia, 41, 898–903. Berlin, I., Grimaldi, A., Payan, C. et al. (1987) Hypoglycemic symptoms and decreased B-adrenergic sensitivity in insulin-dependent diabetic patients. Diabetes Care, 10, 742–747. Bingham, E.M., Dunn, J.T., Smith, D. et al. (2005) Differential changes in brain glucose metabolism during hypoglycaemia accompany loss of hypoglycaemia awareness in men with type 1 diabetes mellitus. An (11C)-3-O-methyl-D-glucose PET study. Diabetologia, 48, 2080–2089. Bjork, E., Palmer, M., Schvarcz, E. & Berne, C. (1990) Incidence of severe hypoglycaemia in an unselected population of patients with insulin-treated diabetes mellitus, with special reference to autonomic neu- ropathy. Diabetes, Nutrition and Metabolism, 4, 303–309. Bottini, P., Boschetti, E., Pampanelli, S. et al. (1997) Contribution of autonomic neuropathy to reduced plasma adrenaline responses to hypoglycemia in IDDM. Evidence for a nonselective defect. Diabetes, 46, 814–823. Boyle, P.J. (1997) Alteration in brain glucose metabolism induced by hypoglycaemia in man. Diabetologia, 40 (Suppl.), S69–S74. Boyle, P.J., Schwartz, N.S., Shah, S.D., Clutter, W.E. & Cryer, P.E. (1988) Plasma glucose concentrations at the onset of hypoglycemic symptoms in patients with poorly controlled diabetes and in non-diabetics. New England Journal of Medicine, 318, 1487–1492. Boyle, P.J., Nagy, R.J., O’Connor, A.M., Kempers, S.F., Yeo, R.A. & Qualls, C. (1994) Adaptation in brain glucose uptake following recurrent hypoglycemia. Proceedings of the National Academy of Sciences of the United States of America, 91, 9352–9356. Boyle, P.J., Kempers, S.F., O’Connor, A.M. & Nagy, R.J. (1995) Brain glucose uptake and unawareness of hypoglycemia in patients with insulin-dependent diabetes mellitus. New England Journal of Medicine, 333, 1726–1731. Choudhary, P., Geddes, J., Freeman, J.J.V., Emery, C.J., Heller, S.R. & Frier, B.M. (2010) Frequency of biochemical hypoglycaemia in adults with Type 1 diabetes with and without impaired awareness of hypoglycaemia: no identifiable differences using continuous glucose monitoring.Diabetic Medicine, 27, 666–672. Clarke, W.L., Cox, D.J., Gonder-Frederick, L.A., Julian, D., Schlundt, D. & Polonsky, W. (1995) Reduced awareness of hypoglycemia in adults with IDDM. A prospective study of hypoglycemic frequency and associated symptoms. Diabetes Care, 18, 517–522. Cooperberg, B.A., Breckenridge, S.M., Arbelaez, A.M. & Cryer, P.E. (2008) Terbutaline and the prevention of nocturnal hypoglycemia in type 1 diabetes. Diabetes Care, 31, 2271–2272. Cox, D., Gonder-Frederick, L., Polonsky, W., Schlundt, D., Julian, D. & Clarke, W. (1995) A multicenter evaluation of blood glucose awareness training – II. Diabetes Care, 18, 523–528. Cox, D., Ritterband, L., Magee, J., Clarke, W. & Gonder-Frederick, L. (2008) Blood glucose awareness training delivered over the internet. Diabetes Care, 31, 1527–1528. Cranston, I., Lomas, J., Maran, A., Macdonald, I. & Amiel, S.A. (1994) Restoration of hypoglycaemia awareness in patients with long-duration insulin-dependent diabetes. Lancet, 344, 283–287. Cranston, I., Reed, L.J., Marsden, P.K. & Amiel, S.A. (2001) Changes in regional brain (18) F-fluorodeoxyglucose uptake at hypoglycemia in type 1 diabetic men associated with hypoglycemia unawareness and counter-regulatory failure. Diabetes, 50, 2329–2336. Cryer, P.E. (1992) Iatrogenic hypoglycemia as a cause of hypoglycemia-associated autonomic failure in IDDM. A vicious circle. Diabetes, 41, 255–260. Cryer, P.E. (2005) Mechanisms of hypoglycemia-associated autonomic failure and its component syn- dromes in diabetes. Diabetes, 54, 3592–3601. Dagogo-Jack, S., Rattarasarn, C. & Cryer, P.E. (1994) Reversal of hypoglycemia unawareness, but not defective glucose counterregulation, in IDDM. Diabetes, 43, 1426–1434. Dagogo-Jack, S.E., Craft, S. & Cryer, P.E. (1993) Hypoglycemia-associated autonomic failure in insulin- dependent diabetes mellitus. Journal of Clinical Investigation, 91, 819–828. Davis, M.R., Mellman, M. & Shamoon, H. (1992) Further defects in counterregulatory responses induced by recurrent hypoglycemia in IDDM. Diabetes, 41, 1335–1340. Davis, S.N., Shevers, C., Mosqueda-Garcia, R. & Costa, F. (1997) Effects of differing antecedent hypogly- cemia on subsequent counterregulation in normal humans. Diabetes, 46, 1328–1335. Chapter 6: Impaired Awareness of Hypoglycaemia 141

Davis, S.N., Mann, S., Galassetti, P. et al. (2000a) Effects of differing durations of antecedent hypogly- cemia on counterregulatory responses to subsequent hypoglycemia in normal humans. Diabetes, 49, 1897–1903. Davis, S.N., Galassetti, P., Wasserman, D.H. & Tate, D. (2000b) Effects of antecedent hypoglycemia on subsequent counterregulatory responses to exercise. Diabetes, 49, 73–81. Debrah, K., Sherwin, R.S., Murphy, J. & Kerr, D. (1996) Effect of caffeine on recognition of and physi- ological responses to hypoglycaemia in insulin-dependent diabetes. Lancet, 347, 19–24. Dunn, J.T., Cranston, I., Marsden, P.K., Amiel, S.A. & Reed, L.J. (2007) Attenuation of amygdala and frontal cortical responses to low blood glucose concentration in asymptomatic hypoglycemia in type 1 diabetes. A new player in hypoglycemia unawareness? Diabetes, 56, 2766–2773. Fanelli, C., Pampanelli, S., Epifano, L. et al. (1994) Long-term recovery from unawareness, deficient counterregulation and lack of cognitive dysfunction during hypoglycaemia, following institution of rational, intensive insulin therapy in IDDM. Diabetologia, 37, 1265–1276. Fanelli, C.G., Paramore, D.S., Hershey, T. et al. (1998) Impact of nocturnal hypoglycemia on hypoglycemic cognitive dysfunction in type 1 diabetes. Diabetes, 47, 1920–1927. Fritsche, A., Stumvoll, M., Grub, M. et al. (1998) Effect of hypoglycemia on beta-adrenergic sensitivity in normal and type 1 diabetic subjects. Diabetes Care, 21, 1505–1510. Fritsche, A., Stefan, N., Haring, H., Gerich, J. & Stumvoll, M. (2001) Avoidance of hypoglycemia restores hypoglycemia awareness by increasing β-adrenergic sensitivity in type 1 diabetes. Annals of Internal Medicine, 134, 729–736. Fruehwald-Schultes, B., Born, J., Kern, W., Peters, A. & Fehm, H.L. (2000) Adaptation of cognitive func- tion to hypoglycemia in healthy men. Diabetes Care, 23, 1059–1066. de Galan, B., de Mol, P., Wennekes, L., Schouwenberg, B. & Smits, P. (2006) Preserved sensitivity to beta2-adrenergic receptor agonists in patients with type 1 diabetes mellitus and hypoglycemia unaware- ness. Journal of Clinical Endocrinology and Metabolism, 91, 2878–2881. de Galan, B.E., Tack, C.J., Lenders, J.W. et al. (2002) Theophylline improves hypoglycemia unawareness in type 1 diabetes. Diabetes, 51, 790–796. Galassetti, P., Tate, D., Neill, R.A., Morrey, S., Wasserman, D.H. & Davis, S.N. (2003) Effect of antecedent hypoglycemia on counterregulatory responses to subsequent euglycemic exercise in type 1 diabetes. Diabetes, 52, 1761–1769. Geddes, J., Wright, R.J., Zammitt, N.N., Deary, I.J. & Frier, B.M. (2007) An evaluation of methods of assessing impaired awareness of hypoglycemia in Type 1 diabetes. Diabetes Care, 30, 1868–1870. Geddes, J., Schopman, J.E., Zammitt, N.N. & Frier, B.M. (2008) Prevalence of impaired awareness of hypoglycaemia in adults with Type 1 diabetes. Diabetic Medicine, 25, 501–504. George, E., Harris, N., Bedford, C., Macdonald, I.A., Hardisty, C.A. & Heller, S.R. (1995) Prolonged but partial impairment of the hypoglycaemic physiological response following short-term hypoglycaemia in normal subjects. Diabetologia, 38, 1183–1190. George, E., Marques, J.L., Harris, N.D., Macdonald, I.A., Hardisty, C.A. & Heller, S.R. (1997) Preservation of physiological responses to hypoglycemia 2 days after antecedent hypoglycemia in patients with IDDM. Diabetes Care, 20, 1293–1298. Giménez, M., Lara, M., Jiménez, A. & Conget, I. (2009) Glycaemic profile characteristics and frequency of impaired awareness of hypoglycaemia in subjects with T1D and repeated hypoglycaemic events. Acta Diabetologica, 46, 291–293. Giménez, M., Lara, M. & Conget, I. (2010) Sustained efficacy of continuous subcutaneous insulin infusion in type 1 diabetes subjects with recurrent non-severe and severe hypoglycemia and hypoglycemia una- wareness: a pilot study. Diabetes Technology and Therapeutics, 12, 517–521. Gold, A.E., MacLeod, K.M. & Frier, B.M. (1994) Frequency of severe hypoglycemia in patients with type 1 diabetes with impaired awareness of hypoglycemia. Diabetes Care, 17, 697–703. Gold, A.E., Deary, I.J., MacLeod, K.M., Thomson, K.J. & Frier, B.M. (1995a) Cognitive function during insulin-induced hypoglycemia in humans: short-term cerebral adaptation does not occur. Psychophar- macology, 119, 325–333. Gold, A.E., MacLeod, K.M., Deary, I.J. & Frier, B.M. (1995b) Hypoglycemia-induced cognitive dysfunction in diabetes mellitus: effect of hypoglycemia unawareness. Physiology and Behavior, 58, 501–511. Grimaldi, A., Bosquet, F., Davidoff, P. et al. (1990) Unawareness of hypoglycemia by insulin-dependent diabetics. Hormone and Metabolic Research, 22, 90–95. Heller, S.R. & Cryer, P.E. (1991) Reduced neuroendocrine and symptomatic responses to subsequent hypoglycemia after 1 episode of hypoglycemia in non-diabetic humans. Diabetes, 40, 223–226. 142 Hypoglycaemia in Clinical Diabetes

Heller, S.R., Chapman, J., McCloud, J. & Ward, J. (1995) Unreliability of reports of hypoglycaemia by diabetic patients. British Medical Journal (Clinical Research Ed.), 310, 440. Henderson, J.N., Allen, K.V., Deary, I.J. & Frier, B.M. (2003) Hypoglycaemia in insulin-treated type 2 diabetes: frequency, symptoms and impaired awareness. Diabetic Medicine, 20, 1016–1021. Hepburn, D.A., Patrick, A.W., Eadington, D.W., Ewing, D.J. & Frier, B.M. (1990) Unawareness of hypogly- caemia in insulin-treated diabetic patients: prevalence and relationship to autonomic neuropathy. Dia- betic Medicine, 7, 711–717. Hepburn, D.A., Patrick, A.W., Brash, H.M., Thomson, I. & Frier, B.M. (1991) Hypoglycaemia unawareness in type 1 diabetes: a lower plasma glucose is required to stimulate sympatho-adrenal activation. Diabetic Medicine, 8, 934–945. Hepburn, D.A., Deary, I.J. & Frier, B.M. (1992) Classification of symptoms of hypoglycaemia in insulin- treated diabetic patients using factor analysis: relationship to hypoglycaemia unawareness. Diabetic Medicine, 9, 70–75. Hepburn, D.A., MacLeod, K.M., Pell, A.C.H., Scougal, I.J. & Frier, B.M. (1993a) Frequency and symptoms of hypoglycaemia experienced by patients with type 2 diabetes treated with insulin. Diabetic Medicine, 10, 231–237. Hepburn, D.A., MacLeod, K.M. & Frier, B.M. (1993b) Physiological, symptomatic and hormonal responses to acute hypoglycaemia in type 1 diabetic patients with autonomic neuropathy. Diabetic Medicine, 10, 940–949. Hepburn, D.A., Deary, I.J., MacLeod, K.M. & Frier, B.M. (1994) Structural equation modeling of symp- toms, awareness and fear of hypoglycemia, and personality in patients with insulin-treated diabetes. Diabetes Care, 17, 1273–1280. Hilsted, J., Madsbad, S., Krarup, T. et al. (1981) Hormonal, metabolic, and cardiovascular responses to hypoglycemia in diabetic autonomic neuropathy. Diabetes, 30, 626–633. Hilsted, J., Richter, E., Madsbad, S. et al. (1987) Metabolic and cardiovascular responses to epinephrine in diabetic autonomic neuropathy. New England Journal of Medicine, 317, 421–426. Hoeldtke, R.D., Boden, G., Shuman, C.R. & Owen, O.E. (1982) Reduced epinephrine secretion and hypoglycemia unawareness in diabetic autonomic neuropathy. Annals of Internal Medicine, 96, 459–462. Jorgensen, H.V., Pedersen-Bjergaard, U., Rasmussen, A.K. & Borch-Johnsen, K. (2003) The impact of severe hypoglycemia and impaired awareness of hypoglycemia on relatives of patients with type 1 diabetes. Diabetes Care, 26, 1106–1109. Joslin, E.P., Gray, H. & Root, H.F. (1922) Insulin in hospital and home. Journal of Metabolic Research, 2, 651–699. Kanc, K., Janssen, M.M.J., Keulen, E.T.P. et al. (1998) Substitution of night-time continuous subcutaneous insulin infusion therapy for bedtime NPH insulin in a multiple injection regimen improves counterregula- tory hormonal responses and warning symptoms of hypoglycaemia in IDDM. Diabetologia, 41, 322–329. Kerr, D., Sherwin, R.S., Pavalkis, F. et al. (1993) Effect of caffeine on the recognition of and responses to hypoglycemia in humans. Annals of Internal Medicine, 119, 799–804. Kinsley, B.T., Widom, B. & Simonson, D.C. (1995) Differential regulation of counterregulatory hormone secretion and symptoms during hypoglycemia in IDDM. Effect of glycemic control. Diabetes Care, 18, 17–26. Kubiak, T., Hermanns, N., Schreckling, H.J., Kulzer, B. & Haak, T. (2004) Assessment of hypoglycaemia awareness using continuous glucose monitoring. Diabetic Medicine, 21, 487–490. Kumagai, A.K., Kang, Y.-S., Boado, R.J. & Pardridge, W.M. (1995) Upregulation of blood–brain barrier GLUT 1 glucose transporter protein and mRNA in experimental chronic hypoglycemia. Diabetes, 44, 1399–1404. Lawrence, R.D. (1941) Insulin hypoglycaemia. Changes in nervous manifestations. Lancet, ii, 602. Leelarathna, L., Little S.A., Walkinshaw, E. et al. (In press) Restoration of self-awareness of hypoglycemia in adults with long-standing type 1 diabetes: hyperinsulinemic hypoglycemic clamp sub-study results from the HypoCOMPaSS trial. Diabetes Care. Leitão, C.B., Tharavanij, T., Cure, P. et al. (2008) Restoration of hypoglycemia awareness after islet trans- plantation. Diabetes Care, 31, 2113–2115. Lingenfelser, T., Renn, W., Sommerwerck, U. et al. (1993) Compromised hormonal counterregulation, symptom awareness, and neurophysiological function after recurrent short-term episodes of insulin- induced hypoglycemia in IDDM patients. Diabetes, 42, 610–618. Chapter 6: Impaired Awareness of Hypoglycaemia 143

Little, S., Chadwick, T., Choudhary, P., et al. (2012) Comparison of optimised MDI versus pumps with or without sensors in severe hypoglycaemia (the Hypo COMPaSS trial). British Medical Journal, Endo- crine Disorders, 12, 33. Liu, D., McManus, R.M. & Ryan, E.A. (1996) Improved counter-regulatory hormonal and symptomatic responses to hypoglycemia in patients with insulin-dependent diabetes mellitus after 3 months of less strict glycemic control. Clinical and Investigative Medicine, 19, 71–82. Ly, T.T., Gallego, P.H., Davis, E.A. & Jones, T.W. (2009) Impaired awareness of hypoglycemia in a population-based sample of children and adolescents with type 1 diabetes. Diabetes Care, 32, 1802–1806. Ly, T.T., Hewitt, J., Davey, R.J., Lim, E.M., Davis, E.A. & Jones, T.W. (2011) Improving epinephrine responses in hypoglycemia unawareness with real-time continuous glucose monitoring in adolescents with type 1 diabetes. Diabetes Care, 34, 50–52. MacLeod, K.M., Hepburn, D.A., Deary, I.J. et al. (1994a) Regional cerebral blood flow in IDDM patients: effects of diabetes and of recurrent severe hypoglycaemia. Diabetologia, 37, 257–263. MacLeod, K.M., Deary, I.J., Graham, K.S., Hepburn, D.A. & Frier, B.M. (1994b) Hypoglycaemia unaware- ness in adult patients with type 1 diabetes: relationship to severe hypoglycaemia and cognitive impair- ment. Diabetes, Nutrition and Metabolism, 7, 205–212. MacLeod, K.M., Gold, A.E., Ebmeier, K.P. et al. (1996) The effects of acute hypoglycemia on relative cerebral blood flow distribution in patients with type 1 (insulin-dependent) diabetes and impaired hypoglycemia awareness. Metabolism: Clinical and Experimental, 45, 974–980. Maddock, R.K. & Krall, L.P. (1953) Insulin reactions. Manifestations and need for recognition of long- acting insulin reactions. Archives of Internal Medicine, 91, 695–703. McCall, A.L., Fixman, L.B., Fleming, N., Tornheim, K., Chick, W. & Ruderman, N.B. (1986) Chronic hypoglycemia increases brain glucose transport. The American Journal of Physiology, 251, E442–E447. McCrimmon, R. (2008) The mechanisms that underlie glucose sensing during hypoglycaemia in diabetes. Diabetic Medicine, 25, 513–522. McCrimmon, R. (2009) Glucose sensing during hypoglycemia: lessons from the lab. Diabetes Care, 32, 1357–1363. Mellman, M.J., Davis, M.R., Brisman, M. & Shamoon, H. (1994) Effect of antecedent hypoglycemia on cognitive function and on glycemic thresholds for counterregulatory hormone secretion in healthy humans. Diabetes Care, 17, 183–188. Mitrakou, A., Ryan, C., Veneman, T. et al. (1991) Hierarchy of glycemic thresholds for counterregulatory hormone secretion, symptoms and cerebral dysfunction. The American Journal of Physiology, 260, E67–E74. Mitrakou, A., Fanelli, C., Veneman, T. et al. (1993) Reversibility of unawareness of hypoglycemia in patients with insulinomas. New England Journal of Medicine, 329, 834–839. Mokan, M., Mitrakou, A., Veneman, T. et al. (1994) Hypoglycemia unawareness in IDDM. Diabetes Care, 17, 1397–1403. Mühlhauser, I., Heirnemann, L., Fritsche, E., von Lennep, K. & Berger, M. (1991) Hypoglycemic symp- toms and frequency of severe hypoglycemia in patients treated with human and animal insulin prepara- tions. Diabetes Care, 14, 745–749. Orchard, T.J., Maser, R.E., Becker, D.J., Dorman, J.S. & Drash, A.L. (1991) Human insulin use and hypoglycaemia: insights from the Pittsburgh Epidemiology of Diabetes Complications Study. Diabetic Medicine, 8, 469–474. Ovalle, F., Fanelli, C.G., Poramore, D.S., Hersley, T., Craft, S. & Cryer, P.E. (1998) Brief twice-weekly episodes of hypoglycemia reduce detection of clinical hypoglycemia in type 1 diabetes mellitus. Diabe- tes, 47, 1472–1479. Pampanelli, S., Fanelli, C., Lalli, C. et al. (1996) Long-term intensive insulin therapy in IDDM: effects on

HbA1c, risk for severe and mild hypoglycaemia, status of counterregulation and awareness of hypogly- caemia. Diabetologia, 39, 677–686. Peters, A., Rohloff, F. & Kerner, W. (1995) Preserved counterregulatory hormone release and symptoms after short term hypoglycemic episodes in normal men. Journal of Clinical Endocrinology and Metabo- lism, 80, 2894–2898. Pramming, S., Thorsteinsson, B., Bendtson, I. & Binder, C. (1991) Symptomatic hypoglycaemia in 411 type 1 diabetic patients. Diabetic Medicine, 8, 217–222. Puente, E.C., Silverstein, J., Bree, A.J. et al. (2010) Recurrent moderate hypoglycemia ameliorates brain damage and cognitive dysfunction induced by severe hypoglycemia. Diabetes, 59, 1055–1062. 144 Hypoglycaemia in Clinical Diabetes

Raju, B., Arbelaez, A., Breckenridge, S. & Cryer, P. (2006) Nocturnal hypoglycemia in type 1 diabetes: an assessment of preventive bedtime treatments. Journal of Clinical Endocrinology and Metabolism, 91, 2087–2092. Robinson, A.M., Parkin, H.M., Macdonald, I.A. & Tattersall, R.B. (1995) Antecedent hypoglycaemia in nondiabetic subjects reduces the adrenaline response for 6 days but does not affect the catecholamine response to other stimuli. Clinical Science, 89, 359–366. Rosenthal, M., Smith, D., Yaguez, L. et al. (2007) Caffeine restores regional brain activation in acute hypoglycaemia in healthy volunteers. Diabetic Medicine, 24, 720–727. Ryder, R.E.J., Owens, D.R., Hayes, T.M., Ghatei, M.A. & Bloom, S.R. (1990) Unawareness of hypogly- caemia and inadequate hypoglycaemic counterregulation: no causal relation with diabetic autonomic neuropathy. British Medical Journal, 301, 783–787. Schopman, J.E., Geddes, J. & Frier, B.M. (2010) Prevalence of impaired awareness of hypoglycaemia and frequency of hypoglycaemia in insulin-treated Type 2 diabetes. Diabetes Research and Clinical Practice, 87, 64–68. Schopman, J.E., Geddes, J. & Frier, B.M. (2011) Frequency of symptomatic and asymptomatic hypogly- caemia in Type 1 diabetes: effect of impaired awareness of hypoglycaemia. Diabetic Medicine, 28, 352–355. Severinghaus, E.L. (1926) Hypoglycemic coma due to repeated insulin overdosage. American Journal of the Medical Sciences, 172, 573–580. Smith, C.B., Choudhary, P., Pernet, A., Hopkins, D. & Amiel, S.A. (2009) Hypoglycemia unawareness is associated with reduced adherence to therapeutic decisions in patients with type 1 diabetes. Diabetes Care, 32, 1196–1198. Smith, D., Pernet, A., Rosenthal, J.M. et al. (2004) The effect of modafanil on counter-regulatory and cognitive responses to hypoglycaemia. Diabetologia, 47, 1704–1711. Stephenson, J.M., Kempler, P., Cavallo Peria, P., Fuller, J.H. & EURODIAB IDDM Complications Study Group (1996) Is autonomic neuropathy a risk factor for severe hypoglycaemia? The EURODIAB IDDM Complications Study. Diabetologia, 39, 1372–1376. Sussman, K.E., Crout, J.R. & Marble, A. (1963) Failure of warning in insulin-induced hypoglycemic reac- tions. Diabetes, 12, 38–45. Teuscher, A. & Berger, W.G. (1987) Hypoglycaemia unawareness in diabetics transferred from beef/porcine insulin to human insulin. Lancet, ii, 382–385. The DCCT Research Group (1991) Epidemiology of severe hypoglycemia in the Diabetes Control and Complications Trial. American Journal of Medicine, 90, 450–459. Thorsteinsson, B., Pramming, S., Lauritzen, T. & Binder, C. (1986) Frequency of daytime biochemical hypoglycaemia in insulin-treated diabetic patients: relation to daily median blood glucose concentrations. Diabetic Medicine, 3, 147–151. Tribl, G., Howorka, K., Heger, G., Anderer, P., Thoma, H. & Zeitlhofer, J. (1996) EEG topography during insulin-induced hypoglycemia in patients with insulin-dependent diabetes mellitus. European Neurology, 36, 303–309. Vea, H., Jorde, R., Sager, G., Vaaler, S. & Sundsfjord, J. (1992) Reproducibility of glycaemic thresholds for activation of counterregulatory hormones and hypoglycaemic symptoms in healthy subjects. Diabe- tologia, 35, 958–961. Veneman, T., Mitrakou, A., Mokan, M., Cryer, P. & Gerich, J. (1993) Induction of hypoglycemia unaware- ness by asymptomatic nocturnal hypoglycemia. Diabetes, 42, 1233–1237. Widom, B. & Simonson, D.C. (1992) Intermittent hypoglycemia impairs glucose counterregulation. Dia- betes, 41, 1335–1340. Zammitt, N.N., Warren, R.E., Deary, I.J. & Frier, B.M. (2008) Delayed recovery of cognitive function following hypoglycemia in adults with Type 1 diabetes. Effect of impaired awareness of hypoglycemia. Diabetes, 57, 732–736. 7 Risks of Intensive Therapy Stephanie A. Amiel King’s College Hospital, London, UK

INTRODUCTION

The benefit of strict glycaemic control in diminishing the risks of long-term complications of diabetes is beyond doubt, but the negative aspects of such therapies need to be consid- ered, and their risks identified, understood and minimised. Modern intensified insulin management need not necessarily increase the risk of iatrogenic problems and can deliver better glycaemic control more safely than in the past, although scope remains for improve- ment. New drugs for type 2 diabetes may offer greater opportunities to achieve near- normoglycaemia but may also bring new risks. These risks need to be explained carefully to every patient, who can then make an individual, informed choice about the management of their diabetes.

THE NEED FOR INTENSIVE THERAPY

Acute hyperglycaemia can cause symptoms and in extreme forms can be lethal. Once the acute situation is treated, the risks of long-term exposure to high blood glucose – and to other effects of insulin deficiency and insulin resistance – lead to the chronic complications of diabetes. Microvascular disease is unique to diabetes (and also premature and accelerated macrovascular disease) and makes it the commonest cause of blindness, kidney failure and lower limb amputation in the population aged under 60, at least in the developed world. Most deaths in people with diabetes are cardiovascular, with renal failure a close second. Coexisting, but potentially treatable, hypertension and dyslipidaemia add to the disease burden, particularly in people with type 2 diabetes. Increasingly, other co-morbidities with a shared pathogenesis are being recognised in diabetes and are beginning to be considered as diabetes complications: several cancers have an increased prevalence in diabetes, and obesity and dementia occur more commonly in association with insulin-resistant states. While the exact pathogenesis of vascular complications remains controversial, there is no doubt that maintaining strict glycaemic control reduces the prevalence of both micro­ vascular and, over time, macrovascular disease. In type 1 diabetes, the landmark Diabetes Control and Complications Trial (DCCT), in which people with type 1 diabetes were

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 146 Hypoglycaemia in Clinical Diabetes

Box 7.1 Risks of intensive glycaemic management • Theoretical risks of intensive glycaemic therapy include increased risk of hypoglycaem­ ­ia, weight gain with its attendant increase in cardiovascular risk and acute deteriora­ tion of established microvascular disease. • Modern methods of establishing intensive insulin therapy by transferring evidence- based skills in insulin usage to the patient is associated with less risk of severe hypoglycaemia and can be weight neutral. • People embarking upon intensive diabetes therapy who have existing significant retinopathy should have this observed very closely and/or treated with laser therapy before starting. • People with long-standing type 2 diabetes and existing macrovascular disease should

use caution in reducing HbA1c, especially if this requires new insulin therapy and/or is not relatively easy to achieve.

randomly allocated to intensive or conventional insulin therapy, was ended early because of the lower rate of progression of retinopathy observed with intensive therapy (The Dia- betes Control and Complications Trial Research Group 1993). Years later, despite a dete- rioration in glycaemic control in those who had previously been allocated to intensive treatment and an improvement in those randomised to standard therapy, a cumulative benefit in terms of reduced macrovascular disease and mortality became apparent in the former (The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interven- tions and Complications (DCCT/EDIC) Study Research Group 2005; Nathan et al. 2003). Similar results emerged from the United Kingdom Prospective Diabetes Study, which recruited people with newly-diagnosed type 2 diabetes to a randomised controlled trial of both intensive versus conventional glycaemic and blood pressure management: the long- term follow-up of that cohort also showed a growing discrepancy in rates of cardiovascular disease and death in favour of those who had been randomised to the intensive control arm (Holman et al. 2008). The ability of intensive diabetes treatment to reduce cancer or dementia risk is still being investigated. In this chapter, the risks of intensive glycaemic management are considered (see Box 7.1), although allusion will be made to the risks of treating co-morbidities such as hyper- tension and dyslipidaemia. Type 1 diabetes is considered first, before examining the evi- dence in type 2 diabetes.

RISKS OF INTENSIVE INSULIN THERAPY

For type 1 diabetes, and advanced, insulin-deficient, type 2 diabetes, intensive therapy means intensified insulin therapy, and the risks largely reflect those of insulin itself. Thus the major side effects are weight gain (The Diabetes Control and Complications Trial Research Group 1988; Purnell et al. 1998) and hypoglycaemia (The Diabetes Control and Complications Trial Research Group 1993, 1995a, 1997). For patients with a long history of poor control, and existing significant microangiopathy in the form of advanced back- ground or pre-proliferative retinopathy, there is an additional initial risk of accelerated deterioration of retinopathy (Agardh et al. 1992), although in the longer term, microvas- cular disease risk is reduced by good glycaemic control (The Kroc Collaborative Study Chapter 7: Risks of Intensive Therapy 147

Group 1984, 1988). Insulin allergies and local skin reactions to insulin such as lipohyper- trophy and, less commonly, lipoatrophy are not related to the intensity of the therapy.

HYPOGLYCAEMIA IN INTENSIVE INSULIN THERAPY

Intensive insulin therapy may be defined as the use of multiple daily injections of insulin (often abbreviated to MDI) and continuous subcutaneous insulin infusion (CSII or insulin pump therapy) and frequent self-testing of blood glucose to achieve and maintain near- normoglycaemia. The aims of the therapy also include minimising hypoglycaemic risk and treatment targets should (but often do not) include lower limits.

DEFINITION OF HYPOGLYCAEMIA

It is difficult to determine a frequency of hypoglycaemia without first defining what is meant by ‘hypoglycaemia’. In many studies, hypoglycaemia is documented by self- reporting, which may be unreliable (Heller et al. 1995). Retrospective analyses suffer from problems of recall, and accurate documentation is obtained only in prospective research studies requiring biochemical verification (see Chapter 4). Until recently, hypoglycaemia was defined only in terms of clinical episodes and blood glucose records. Each episode can be categorised by its symptomatology and its severity, but no real consensus exists. ‘Mild’ hypoglycaemia is usually defined as an episode that a person recognises and treats them- selves; ‘severe’ hypoglycaemia is an episode where the patient has become so disabled that assistance is required from another person and may be subcategorised by the requirement for parenteral treatment (intramuscular glucagon or intravenous dextrose), with or without hospital admission, or by the development of coma. A category of ‘moderate’ hypoglycae- mia, in which an individual requires external assistance but which falls short of requiring parenteral therapy or developing a coma has also been used (Limbert et al. 1993), but this term has also been applied to symptomatic, self-treated episodes that involved significant disruption of lifestyle. The American Diabetes Association (ADA) produced a useful clinical categorisation of hypoglycaemic episodes, but its biochemical definition of hypoglycaemia as a plasma glucose of below 3.9 mmol/l (70 mg/dl) is controversial (American Diabetes Association Workgroup on Hypoglycemia 2005). Until 2012, the European Medicines Agency (2002) used 3.0 mmol/l (54 mg/dl) as their definition of hypoglycaemia to be considered relevant to the assessment of risk of diabetes therapies, presumably to ensure when assessing new drugs, that only episodes of unquestionable clinical significance are considered (European Medicines Agency 2002). This view is supported by evidence that a fall in plasma glucose below 3.0 mmol/l for a period of time can reduce the symptomatic responses to a further episode of hypoglycaemia occurring within the following 24 hours (Heller and Cryer 1991), so-called antecedent hypoglycaemia (Chapter 6) and that continuous avoidance of low glucose concentrations (below 3.0 mmol/l) can restore, at least in part, hormonal and symp- tomatic responses to hypoglycaemia (Fanelli et al. 1993; Cranston et al. 1994). Further- more, in experimentally-induced controlled hypoglycaemia, onset of detectable cognitive dysfunction (impaired reaction times) is reliably detected when plasma glucose is reduced to 3.0 mmol/l or less, irrespective of the glucose concentration associated with onset of sympathetic activation and catecholamine release, and of such factors as type of diabetes, 148 Hypoglycaemia in Clinical Diabetes degree of hypoglycaemia awareness and advancing age (Maran et al. 1995; Matyka et al. 1997; Heller et al. 2002; Choudhary et al. 2009), reinforcing the premise that this is an important degree of hypoglycaemia to determine – and avoid. The evidence that informed the definition of the ADA Workgroup was that some defects in early counterregulatory responses to hypoglycaemia may be induced by exposure to modest antecedent hypogly- caemia (Davis et al. 1997), although a blood glucose of 3.3 mmol/l was required to blunt the catecholamine and metabolic responses. These data were used supported the ADA’s repetition of 3.9 mmol/l or below as defining biochemical hypoglycaemia, and in an unwel- come revision, the European Medicines Agency have accepted the ADA position without discussion. This definition has been retained in a recent revision by the ADA, although confusingly the same glucose concentration is now also recognised as an ‘alert value’ – an indication that hypoglycaemia risk is high, which is less controversial (Seaquist et al. 2013). Continuous monitoring of interstitial tissue glucose using continuous glucose monitor- ing (CGM), calibrated to equivalent plasma glucose, is now available as a way of assessing exposure to hypoglycaemia. It allows objective measurement of duration of episodes, particularly at night, but the clinical significance of a low sensor glucose remains to be elucidated. The techniques may be useful in comparative studies (UK Hypoglycaemia Study Group 2007). However, there is no correlation between hypoglycaemic exposure measured by short-term (5-day) CGM and impaired awareness of hypoglycaemia (Choud- hary et al. 2010), perhaps because of the relatively short time period over which CGM was recorded and because the exact relationship between interstitial glucose measurement and plasma glucose remains to be determined, especially in terms of duration. It is probably more appropriate to differentiate between alert values and targets for adjusting therapy, which may properly be 3.9 mmol/l and above, and which may be used as an indication for action to prevent hypoglycaemia per se, which is a pathological event that requires intervention and for which values below 3.5 mmol/l may be more appropriate. The former are blood glucose concentrations which, if recurrent, indicate a need to adjust therapy to prevent hypoglycaemia occurring in the future and may even be set slightly higher than 4.0 mmol/l (The DAFNE Study Group 2002).

CONTRIBUTORS TO INCREASED RISK OF SEVERE HYPOGLYCAEMIA IN PATIENTS UNDERTAKING INTENSIFIED INSULIN THERAPY

The relationship between impaired awareness of hypoglycaemia and an increased rate of severe hypoglycaemia is well established (Hepburn et al. 1990; Gold et al. 1994; Clarke et al. 1995), although those studied were not subject to strict glycaemic control. The asso- ciation between counterregulatory failure and increased risk of severe hypoglycaemia is also well recognised (Ryder et al. 1990). Indeed, counterregulatory failure was proposed as a predictor of risk of severe hypoglycaemia in the subsequent application of intensified therapy (White et al. 1983), and it was not until later that the ability of intensified therapy to cause counterregulatory failure was suggested (Simonson et al. 1985a). It is important to appreciate that neither asymptomatic nor severe episodes are restricted to people under- taking intensified therapy. Effectively, all people using insulin secretagogues or exogenous insulin are at risk of occasional and unanticipated severe hypoglycaemia, due to a combina- tion of artificially elevated plasma-free insulin concentrations and impaired counterregula- tory responses. The risk is highest in those most deficient in endogenous insulin, as Chapter 7: Risks of Intensive Therapy 149 demonstrated by the relationship between C-peptide responses and risk of severe hypogly- caemia (Mühlhauser et al. 1998). The degree of insulin deficiency may be the best predictor of risk for severe hypogly- caemia in intensified diabetes therapy, apart from a previous history of severe hypoglycae- mia itself. Absence of C-peptide has long been associated with increased risk (Mühlhauser et al. 1998) with additional evidence from the DCCT, demonstrating that significant residual insulin secretion was associated with less severe hypoglycaemia (Steffes et al. 2003). Other factors include glycaemic control prior to intensification and a determination to reach glycaemic targets (Bott et al. 1994; Mühlhauser et al. 1998). Rate of loss of endogenous insulin secretory capacity is affected by many factors but may be slowed by immediate institution of strict glycaemic control at diagnosis (Shah et al. 1989; Linn et al. 1996; The Diabetes Control and Complications Trial Research Group 1998). The mechanisms whereby residual endogenous insulin secretory capacity defends against severe hypoglycaemia are not known but may include: the ability to reduce portal insulin concentrations during hypoglycaemia; the preservation of glucagon responses to hypoglycaemia, which are driven in part by residual beta cell responses to hypoglycaemia (Peacey et al. 1997); or some protective action of C-peptide per se (Wahren et al. 2012). Other factors increasing the risk of severe hypoglycaemia may include social class (Mühlhauser et al. 1998) and genetics. A Danish study attributed some hypoglycaemic risk to ACE genotype (Pedersen-Bjergaard et al. 2003), although this has not been confirmed by others (see Chapter 4).

The effects of intensified insulin therapy on risk of severe hypoglycaemia In the DCCT, a 3-fold higher rate of severe hypoglycaemia was recorded by the people treated intensively compared to those on standard therapy (The DCCT Research Group 1991; The Diabetes Control and Complications Trial Research Group 1993, 1997). Further- more, the risk of severe hypoglycaemia was higher for any given HbA1c, for the people receiving intensive treatment. This phenomenon has not been adequately explained but has been attributed to the known ability of exposure to antecedent hypoglycaemia to induce defects in counterregulation and loss of subjective awareness of hypoglycaemia (Heller and Cryer 1991; George et al. 1995, 1997; Davis et al. 1997). It has been assumed that intensive therapy exposes the patient to a greater frequency of mild hypoglycaemia sufficient to induce such defects and thereby increase the risk of severe hypoglycaemia. In a longitudinal study of intensification of insulin therapy in people with type 1 diabetes, the change was associated with impaired counterregulation and a shift downwards of the blood glucose concentrations required to initiate the stress responses (Amiel et al. 1987, 1988). A series of studies has shown that hypoglycaemia awareness and counterregulatory hormone responses can be restored, at least in part, in affected individuals with type 1 diabetes by strict avoidance of blood glucose concentrations below 3.0 mmol/l in daily life, confirming the circular link between hypoglycaemia exposure and impaired awareness of hypoglycaemia (Fanelli et al. 1993; Cranston et al. 1994) (Figure 7.1). This does not imply, and should not be taken as, an acceptance of more hyperglycaemia in the treatment of problematic hypoglycaemia; in effect it means equal emphasis should be placed on remaining above the lower limits of the advised treatment targets as on remaining below the higher targets. While it remains true that severe hypoglycaemia rates remain higher in the intensive arm of all randomised controlled trials of intensive versus conventional management of both main types of diabetes, methods for delivering 150 Hypoglycaemia in Clinical Diabetes

Group A: good control Group B: poor control

6.0 Before intervention Before intervention 5.5 After intervention After intervention 5.0 4.5 4.0 3.5 3.0 2.5 2.0 1.5 1 5000 4500 4000 3500 3000 2500 2000 1500 1000 Adrenallne (pmol/L)Adrenallne Plasma glucose (mmol/L) 500 0 40 35 30 25 20 15 10 ∆ Symptom scores ∆ Symptom 5 0 –40 40 120 200 280 –40 40 120 200 280 0 80 160 240 0 80 160 240 Time (mins)

Figure 7.1 Plasma glucose profiles (top), adrenaline (middle) and symptom score responses to induced hypoglycaemia before (dotted line) and after (solid line) therapeutic adjustment to avoid exposure to a plasma glucose of less than 3 mmol/l in people with type 1 diabetes and recurrent severe and asymp­ tomatic hypoglycaemia with glycated haemoglobin in or near target (left) and high (right), showing restoration of severely deficient counterregulatory responses. (Source: Cranston I 1994. Reproduced with permission from Elsevier). intensified diabetes therapy have improved since the DCCT was undertaken. Intype1 diabetes, modern methods that focus on transferring skills of insulin adjustment to the patients themselves using structured education are reported to achieve improvements in glycated haemoglobin, using multiple daily injection therapy regimens (often with conven- tional insulins), obtaining a parallel reduction in glycated haemoglobin and in rates of hypoglycaemia (Jorgens et al. 1993; Plank et al. 2004; Samann et al. 2005; Rogers et al. 2009; Hopkins et al. 2012). Reduction in severe hypoglycaemia is a more durable outcome of many of these programmes, lasting sometimes as long as 12 years, by which time some of the initial improvement in mean glycated haemoglobin, seen after 3 years, may be lost (Figure 7.2) (Plank et al. 2004). Chapter 7: Risks of Intensive Therapy 151

0.6 8

0.5 7.5 0.4 7 c (%) 0.3 1

6.5 HbA 0.2 Hypoglycaemia Hypoglycaemia

0.1 6

0 5.5 Pre-course 3 years 6 years 12 years

Figure 7.2 In this study of the 12 year follow-up of a structured education programme, severe hypogly­ caemia rates (shown by dark purple bars as episodes per patient per year and by light purple bars as percent of patients experiencing one or more episode in a year) fell at the same time as the HbA1c (line with solid triangles) was reduced. The impact of the education on hypoglycaemia persisted beyond the effect on HbA1c. (Source: Plank J 2004. Reproduced with kind permission from Springer Science+Business Media).

These programmes promote maintenance of near-normal glucose, as measured by capil- lary blood glucose values before meals and at bedtime (4.5–7.5 mmol/l; 80–140 mg/dl), with explicitly separate adjustment of basal insulin replacement, provided by twice-daily injections of intermediate-acting insulins, and of soluble or fast-acting analogue insulins before meals to control the post-prandial glucose rise. Patients are shown how to adjust the meal-related doses, both to cover the intended meal and to correct any deviation from the pre-meal target glucose. They are also advised to examine their recent glucose readings periodically, in order to make prospective adjustments to either the basal or the meal doses, thus adapting algorithms for insulin replacement that are based on insulin pharmacodynam- ics, to their own, constantly changing, situation. Parenthetically, patients tend to be more effective in maintaining immediate dose adjustment for meals than making prospective insulin regimen changes based on periodic review of their blood glucose results (Lawton et al. 2012). Reduced exposure to antecedent hypoglycaemia probably contributes to reduced hypoglycaemic risk as suggested by the ability of such structured education pro- grammes to reduce both severe hypoglycaemia rates and also restoring hypoglycaemia awareness, which was achieved in around half of those with problems of impaired aware- ness in one such programme (Hopkins et al. 2012). Why the other half continued to have impaired hypoglycaemia awareness is unknown. While some have suggested that a sub- group of patients may develop irreversible defects in hypoglycaemia defences, others have explored the possibility that the problem lies in an inability of these patients to make thera- peutic changes that will result in less exposure to hypoglycaemia. Neuroimaging studies have shown altered responses in brain reward and aversion networks during acute hypogly- caemia in people with impaired hypoglycaemia awareness (Dunn et al. 2007), and clinical records show reduced adherence to agreed therapeutic change in people with type 1 dia- betes with impaired awareness compared to those with normal awareness (Smith et al. 2009). Some people with impaired awareness fail to perceive the problem as a serious threat and to help these people improve their hypoglycaemia defences, it may be as impor- tant to tackle their health beliefs as providing technological aids (Rogers et al. 2012). The benefits of structured education programmes on both glycated haemoglobin and problematic hypoglycaemia demonstrate the importance of patients with diabetes acquiring 152 Hypoglycaemia in Clinical Diabetes an understanding of insulin pharmacodynamics and interactions from appropriately expe- rienced educators. The benefits are achievable using conventional insulins, and how the insulin is used is as important as the type of insulin prescribed. Nevertheless, in type 1 diabetes, judicious use of insulin analogues and CSII in intensified regimens may be associ- ated with slightly less risk of hypoglycaemia (Ashwell et al. 2006). Use of CSII has been used successfully as treatment in some patients with problematic hypoglycaemia (Pickup et al. 2002) and in the context of clinical trials (Hoogma et al. 2006). A 4-fold reduction in severe hypoglycaemia has been reported (Pickup and Sutton 2008), and progression to CSII is logical when MDI fails to prevent hypoglycaemia. An alternative approach is needed for those in whom the use of this technology also fails, which would include con- sideration of islet cell transplantation (Srinivasan et al. 2007). The benefit of adding real-time glucose sensing to intensive insulin regimens in type 1 diabetes is becoming more clear. In the largest trial of so-called sensor-augmented pump therapy, the patients using the full system achieved lower glycated haemoglobin values when they used the devices most of the time without any increase in severe hypoglycaemia (The Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group 2008). In a subsequent study in which CGM was offered to subjects of the earlier trial who had been in the control arm, a decline in the rate of severe hypoglycaemia did not achieve significance (The Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group 2010). Rates of severe hypoglycaemia were low in a trial comparing sensor- augmented pump therapy with MDI (in terms of achieving glycated haemoglobin targets), and while the rate did not decline any further with the sensor-pump regimen, the glycated haemoglobin fell significantly in the intervention group without any corresponding increase in the rate of severe hypoglycaemia (Bergenstal et al. 2010). It appears that if the risk of hypoglycaemia is to be removed completely using exogenous insulin, an insulin delivery system that is responsive to the plasma glucose will be required. Using mathematical algorithms that adjust the insulin delivery from a pump system based on CGM has shown promising early results for basal insulin replacement (Hovorka et al. 2011; Nimri et al. 2012), although automated predictions of meal insulin requirements remain challenging. A very early study of a clinically available system, which included an automated 2-hour suspension of the insulin infusion if the patient made no obvious response to sensor- detected hypoglycaemia, decreased the time spent in hypoglycaemia in those people who at baseline had experienced the most exposure to hypoglycaemia (Choudhary et al. 2011). Hypoglycaemia can be prevented – and near-normal glycated haemoglobin maintained – for several years after transplantation of the whole pancreas or islets, but such therapies are currently used as a last resort because of the very limited supply of islets for transplan- tation and the risks associated with immunosuppression (CITR Research Group. 2009). At present the duration of benefit is limited with both forms of transplantation (Sutherland and Gruessner 2007; Barton et al. 2012). Much less information is available for type 2 diabetes, although evidence is accumulat- ing that in insulin-treated type 2 diabetes of long duration, the prevalence of severe hypoglycaemia does not differ much from type 1 diabetes (see Chapter 12). The early institution of insulin for the treatment of type 2 diabetes, at a stage when insulin deficiency is not advanced, may minimise the overall risk of hypoglycaemia in people with insulin- treated type 2 diabetes. Studies using bedtime basal insulin as the first line of treatment intensification for people with type 2 diabetes who are not achieving glycaemic targets, have reported a low risk of severe hypoglycaemia, even when conventional human insulins are used (Yki-Jarvinen et al. 2006). However, caution is necessary when patients have Chapter 7: Risks of Intensive Therapy 153 progressed to the stage of requiring full insulin therapy; more advanced beta cell failure is likely to be associated with a higher risk of severe hypoglycaemia, similar to that observed in type 1 diabetes. Rates of severe hypoglycaemia in people with type 2 diabetes are much higher in those who have been treated with insulin for 5 years or more than in the first 2 years of therapy (UK Hypoglycaemia Study Group 2007). Severe hypoglycaemia rates are also increased when treatment is intensified with other antidiabetes therapies (other than insulin); few studies as yet have shown the potential to achieve lower glycated haemoglobin concentrations while simultaneously reducing hypoglycaemia risk. In a small study of poorly-controlled patients treated with oral medica- tion, in which responses to hypoglycaemia were measured before and after improving glycaemic control with insulin, counterregulatory responses and the blood glucose thre­ sholds at which these were initiated were modified, as occurs in type 1 diabetes (Korzon- Burakowska et al. 1998). Higher frequencies of severe hypoglycaemia have been documented in several randomly controlled trials of intensified therapy involving people with long-standing type 2 diabetes and cardiovascular disease. (The Action to Control Cardiovascular Risk in Diabetes Study Group (ACCORD) 2008; The ADVANCE Collabo- rative Group 2008; Duckworth et al. 2009), with higher mortality being reported in one (ACCORD), although hypoglycaemia could not be proven to be the direct cause (Miller et al. 2010; Bonds et al. 2010).

OTHER RISKS OF INTENSIFIED INSULIN THERAPY

Weight gain Weight gain with insulin therapy is multifactorial, but when intensive therapy is applied to a patient who had poor control, a major contributor will be the resolution of caloric loss in glycosuria (Carlson and Campbell 1993). Thus as glycaemic control improves, a failure to compensate for reduced energy loss by reducing calorie intake will inevitably cause weight gain. This is theoretically responsive to dietary strategies, and it is noticeable that structured education programmes around flexible insulin are generally weight neutral (DAFNE Study Group 2002; Rogers et al. 2009). These can even be associated with weight loss (McIntyre et al. 2010), as patients can use the dietary flexibility encouraged by such programmes to control their blood glucose around the weight-reducing diet of their choice. Diminished experience of hypoglycaemia – or perhaps only reduced fear of hypoglycaemia – may also help, as the patient adjusts the insulin regimen with greater confidence to produce the expected results and ceases to eat to keep up with the effects of insulin. However, insulin does cause lipogenesis and can cause fluid retention, the latter especially when therapy is intensified on a background of previous significant underinsulinisation, both of which contribute to weight gain. Anecdotally patients report that insulin increases appetite. Healthy men (but not women) respond to nasal insulin administration with reduced appetite (Benedict et al. 2008). In a study of people with type 1 diabetes of both genders, use of the analogue insulin, detemir, which has been reported to be associated with less weight gain than other insulins, was associated with lower weight and less consumption of food compared to a conventional insulin regimen (Zachariah et al. 2011). The mechanism for this remains to be elucidated, but peripherally infused insulin alters activation of brain regions involved in appetite control and satiety (Bingham et al. 2002). 154 Hypoglycaemia in Clinical Diabetes

Diabetic ketoacidosis and hyperinsulinaemia When insulin pump therapy first moved into clinical practice, the rate of diabetic ketoaci- dosis in pump users increased (Knight et al. 1985). This was thought to relate to the absence of any intermediate-acting or background insulin in the event of pump failure. Because the basal insulin in a CSII regimen is delivered as an infusion of very low volume, in the absence of a subcutaneous depot, an interruption in delivery can rapidly lead to hypergly- caemia and even ketosis, especially if the patient’s blood glucose is already elevated (Castilloa et al. 1996). This can occur as a result of disconnection of the pump, air in the delivery system, blockage in the tubing or more rarely, mechanical failure of the pump. The apparently high risk of diabetic ketoacidosis (DKA) with insulin pump therapy was not entirely confined to lack of clinical experience (Knight et al. 1986). In 1997, a meta- analysis of trials of CSII indicated that the rate of DKA was significantly higher (Egger et al. 1997) and the rate of development of DKA was also slightly greater in intensively- treated patients in the DCCT, although many of those patients used multiple injections of insulin to improve their glycaemic control (The Diabetes Control and Complications Trial Research Group 1995b). However, with reliable modern pumps and intensive treatment regimens that focus more on teaching skills of flexible insulin dose adjustment, including regular blood glucose self-monitoring, DKA rates are not higher. Indeed, some centres have utilised insulin pump therapy to help people avoid recurrent DKA (Rodrigues et al. 2005), and a meta-analysis of published pump studies show reduced, rather than higher, rates of DKA with pump use. Nevertheless, the risk is worth reiterating, as it may be evident when pump technologies are applied in inexperienced settings. Achieving adequate plasma concentrations of insulin in the hepatic circulation is always likely to be at the cost of hyperinsulinaemia in the systemic circulation, as insulin is deliv- ered subcutaneously. Peripheral overinsulinisation may contribute to the risk of hypogly- caemia; when insulin is delivered into the portal system, as with intraperitoneal infusion systems, hypoglycaemia is less frequent at any given blood glucose level. (Lassmann- Vague et al. 1996; Dunn et al. 1997) Hyperinsulinaemia has also been implicated in increased cardiovascular and cancer risks. These risks may be lower during intensified therapy, as appropriate temporal distribution of the action of insulin in a MDI or CSII regimen usually results in lower overall insulin requirement. Improved glycaemic control in type 1 diabetes also improves insulin sensitivity (Simonson et al. 1985b), reducing the dose of exogenous insulin required.

Psychological impact of intensified therapies In type 1 diabetes, patients participating in trials generally report reduced anxiety and distress during intensified treatment. In the DCCT, where patients were randomised, those allocated to the intensive treatment arm showed overall improvement in their subjective feelings of control and well-being, although it was offset by a greater fear of hypoglycaemia (The Diabetes Control and Complications Trial Research Group 1996a). However, this balance may be particularly positive in people who actively choose to use intensified thera- pies. Patients electing to undertake flexible insulin therapy report ‘feeling better’ despite the exigencies of the regimens, perhaps in part because of the potential for more flexible living but also due to greater confidence in undertaking self-management. Thus they may feel able to reach more predictable and stable glucose outcomes using intensified insulin therapy through structured training, which transfers insulin management skills from the Chapter 7: Risks of Intensive Therapy 155

6 2 P < 0.001 P < 0.00001 P < 0.006 P < 0.04 1.5 Average Impact of 5 1 weighted diabetes on 0.5 index QOL 4 0 –0.5 Present 3 QOL Soore –1 2

Mean impact –1.5 –2 1 –2.5 P < 0.005 P < 0.00004 –3 0 Satisfaction BG unacceptably BG unacceptably with treatment high recently low recently

Figure 7.3 Stuctured education in flexible insulin use chosen by patients with type 1 diabetes is associ­ ated with improvement in measures of patient well-being one year later. The figures show improvement in diabetes related quality of life (QOL), with reduced negative impact of diabetes on quality of life and improvement, improvement in present quality of life and improved treatment satisfaction. The dark purple bars show measures prior to a structured education programme, the light purple bars the measures one year later. (Source: Rogers H 2009. Reproduced with permission from John Wiley & Sons, Ltd). healthcare professional to the patient (Rankin et al. 2011). In the original DAFNE Trial, all participants had selected the intensive programme of treatment, and significant and apparently lasting benefits in quality of life (QOL) measures were demonstrated using the intensified management strategy (DAFNE Study Group 2002), which persisted nearly 4 years later (Speight et al. 2010). Subsequent studies have confirmed these benefits (Figure 7.3) in other populations choosing intensive therapy (Rogers et al. 2009; McIntyre et al. 2010). ‘Refresher’ training for people with established type 1 diabetes who fail to achieve glycaemic targets despite structured education in intensive insulin self-management, has been associated with improved self-efficacy, and less severe hypoglycaemia and DKA rates, albeit without improving HbA1c (Bott et al. 2000). Individual patients who strive for near normal glucose profiles frequently experience difficulty and frustration when trying to eliminate blood glucose readings outside the normal range. Nevertheless, when cohorts of patients are studied rather than individuals, the potential psychological risks of intensified insulin therapies do not emerge as a problem, perhaps because the most vulnerable choose not to participate.

Risks of deterioration of microvascular disease The Kroc study, a randomised controlled trial of intensified insulin therapy that was origi- nally designed as a pilot for the DCCT, was the first to report deterioration of retinopathy with stricter glycaemic control. Over time, retinopathy improved to a greater extent in the intensively treated patients who finished with less retinal disease at the end of the trial (The Kroc Collaborative Study Group 1984, 1988). Since then there have been reports of dete- rioration in retinal status including visual loss, especially in those with advanced back- ground or pre-proliferative retinopathy at the beginning of intensive therapy and a background of a long history of hyperglycaemia (Moskalets et al. 1994). It is thought that the sudden reduction in the previously high rate of retinal blood flow that accompanies improved glucose concentrations in those who had poor glycaemic control may be a factor, and a sudden increase in IGF-1 or other cytokines has also been implicated (Chantelau and Frystyk 2005; Klein et al. 2009). Individuals at high risk need careful observation as their

HbA1c falls. Initial laser stabilisation of active sight-threatening eye disease should be considered before embarking on intensive therapy. 156 Hypoglycaemia in Clinical Diabetes

The advent of acute painful neuropathy with the sudden institution of strict glycaemic control is also well described (Dabby et al. 2009; Gibbons and Freeman 2010). It may share a common aetiology with the retinopathy described above, and the term ‘insulin neuritis’ has been used. It is often associated with insulin-induced oedema and is a particu- lar risk where the patient has been previously poorly insulinised. It has been described in both type 1 and type 2 diabetes and in the former at least has been associated with preced- ing anorexia as well as poor glycaemic control. Management includes aggressive analgesia including use of specific agents targeting neuropathy such as gabapentin, and pain dimin- ishes slowly if good glycaemic control is maintained (Gibbons and Freeman 2010).

PATIENTS UNSUITABLE FOR STRICT CONTROL

The DCCT demonstrated that any improvement of glycated haemoglobin is associated with fewer microvascular complications over time, and the benefits are greater with initially higher HbA1c concentrations (The Diabetes Control and Complications Trial Research Group 1996b). A cross-sectional study that suggested that the risk reduction for nephropa- thy is near-maximal at a glycated haemoglobin of 8% (Krolewski et al. 1995) cannot be extrapolated to other microvascular complications, because in the DCCT no glycaemic threshold (estimated by glycated haemoglobin) for the development of retinopathy was demonstrated in a patient group whose average HbA1c was 7%. The long-term follow-up of the DCCT cohort has confirmed that intensive therapy benefits macrovascular as well as microvascular risk and that these effects are sustained (The Diabetes Control and Com- plications Trial/Epidemiology of Diabetes Interventions and Complications Research Group 2002). Thus, unless an individual already has a near normal glycated haemoglobin with no problematic hypoglycaemia, no one who has diabetes is unsuitable for attempts to improve their glycaemic control, especially with modern approaches that deliver improved control with less hypoglycaemia risk. In practice, however, there are patients in whom attempts to achieve near normal glucose levels are not appropriate (Box 7.2). In patients with advanced complications, especially

Box 7.2 Application of strict glycaemic control in type 1 diabetes Caution required:

• Long duration of insulin-treated diabetes (counterregulatory deficiencies) • Previous history of severe hypoglycaemia • Established impaired awareness of hypoglycaemia • History of epilepsy • Patient unwilling to do home blood glucose monitoring

Contraindicated:

• Extremes of age • Active Ischaemic heart disease • Unstable diabetic retinopathy (can be instituted after treatment) • Advanced diabetic complications • Limited life expectancy (e.g. serious co-existing disease) Chapter 7: Risks of Intensive Therapy 157 retinopathy, which has not been shown to benefit, a sudden improvement in glycaemic control may cause acceleration in severity of pre-proliferative or early proliferative retin- opathy (Hanssen et al. 1986). Although some authorities claim that this should not be a contraindication to improving glycaemic control (Chantelau and Kohner 1997), as yet there is no evidence for benefit in advanced cases and the retinopathy should be treated appro- priately before glycaemic control is intensified. Similarly, in those with established renal impairment and severe macrovascular disease, attempts to treat elevated blood pressure and plasma lipids and to encourage patients to stop smoking may be more beneficial than targeting glycaemic control alone. As intensive insulin therapy is aimed at achieving benefit over a period of 5 to 10 years or more, patients with a life expectancy of less than 5 years should not be exposed to the risks associated with this treatment regimen. This applies also to elderly patients who are frail and physically inactive. Poor glycaemic control should not be encouraged in children, as growth may be jeop- ardised, and there is evidence that pre-pubertal glycaemic control may influence the later risk of complications (Donaghue et al. 1997; Holl et al. 1998). However, as small children, who are insulin-sensitive, may be at risk of intellectual damage if exposed to recurrent severe hypoglycaemia (see Chapters 10 and 15), it is important that attempts at intensive glycaemic management are done expertly.

Risks of intensive glycaemic therapy in people with type 2 diabetes People with type 2 diabetes present special challenges for intensive glycaemic manage- ment. Many have diabetes complications at diagnosis and others acquire complications with time. They currently receive a wide range of therapies including insulin sensitisers and secretagogues, and agents provoking glycosuria and enhancing the incretin pathways; thus the risks of such therapies need to be added to the risks of strict glycaemic control besides that of insulin therapy alone. Two things emerge from current evidence: (1) Establishing good glycaemic control from diagnosis of type 2 diabetes reduces long- term risk of micro- and macrovascular complications (Holman et al. 2008). (2) Multimodal intensive risk factor management can improve outcomes (Gaede et al. 2003). However, considerable controversy exists about the risks of intensive glycaemic man- agement particularly in those with long duration type 2 diabetes and established complica- tions. These potential risks and their potential relationship to hypoglycaemia are discussed in detail in Chapter 12. However, it is important to know that the participants in the VADT who had lesser degrees of coronary calcification sustained lower rates of myocardial infarc- tion over time in the intensively managed group (Reaven et al. 2009). The risk–benefit ratio of intensified treatment will be more favourable in people who are now being diag- nosed at younger ages and probably also earlier in the progression of the disease than in the recent trials of people with advanced disease. What seems likely is that intensive gly- caemic management for type 2 diabetes will have major advantages for the patient provided it is offered sufficiently early when the patient is at relatively low risk.

Other risks of intensive therapy in type 2 diabetes The risks of weight gain in type 2 diabetes, where weight is already a major problem, have already been alluded to and have been previously reviewed by Heller (2004). Newer 158 Hypoglycaemia in Clinical Diabetes therapies based on the incretin axis offer potential for improving glycaemic control with no weight gain, and with the injectable GLP-1 receptor analogue therapies achieve genuine weight reduction (Ryder et al. 2010). Deterioration in retinopathy after tightening glycaemic control is described in type 2 diabetes as well as in type 1 diabetes (Hendrickson et al. 1995), as is insulin neuritis (Dabby et al. 2009). Another problem for the type 2 patient is the psychological issues that surround intensification of therapy when this requires either the use of home blood glucose monitor- ing or converting to insulin therapy. The value of self-blood glucose monitoring in type 2 diabetes is beyond the scope of this chapter but when used purely for monitoring, without obvious need or training to adjust therapy immediately around the result, the effects are often negative, including a decline in the patient’s psychological well-being (Clar et al. 2010). There are no reliable data on how monitoring is perceived when it is part of a structured programme of intensifying control. There is convincing evidence of the psycho- logical effects of recommending insulin therapy to people with a long history of manage- ment on other agents (Peyrot et al. 2005). Once again, where insulin therapy is started early in the course of the disease as part of a managed use of strict glycaemic control the issues may be very different, but there is a lack of good data.

Deciding when to use intensive therapy in diabetes In diabetes therapy, the patients themselves determine the degree of glycaemic control that they feel is worth the effort. People with uncomplicated diabetes should be encouraged to pursue strict glycaemic control, especially those with diabetes of short duration. Modern methods of structured education that allow patients to control their diabetes around a flex- ible lifestyle appear to offer real benefit. Those experiencing problematic hypoglycaemia should not be excluded from undertaking intensive therapy training as there is evidence that structured education in flexible insulin therapy, for example, provides not only a lower

HbA1c but also less risk of experiencing severe hypoglycaemia. However, people who cannot commit themselves to regular self-monitoring of blood glucose, with attention to the timing of injection and adjustment of dosage of insulin, cannot safely undertake meas- ures to achieve near-normoglycaemia. The risk–benefit ratio may be less clear in people with long-standing diabetes and advanced vascular complications. A compromise must be reached after a full discussion of the risks. Thus, it is the informed patient who must deter- mine their own therapeutic aims at any given time. The doctor’s role is to try to ensure that the patient has the knowledge to make appropriate decisions and to provide the tools and training to achieve these aims.

CONCLUSIONS • The principal risks of intensive insulin therapy are hypoglycaemia and weight gain. • When intensive insulin therapy is applied using structured education to transfer evidence-based training of using exogenous insulin therapy competently, these risks are minimised. Indeed, modern intensive insulin therapy using such programmes reduces hypoglycaemia rates and reverses impaired awareness of hypoglycaemia in many patients. • It is likely that exposure to hypoglycaemia during any diabetes therapy impairs the counterregulatory responses to hypoglycaemia and symptomatic awareness and Chapter 7: Risks of Intensive Therapy 159

increases risk of severe hypoglycaemia; this may be reversible with properly delivered intensive therapy. • Avoidance of hypoglycaemia can restore the symptomatic response to hypoglycaemia. • Achieving such benefits of intensive therapy once problematic hypoglycaemia has been established may be demanding and time consuming, both for patients and healthcare professionals. Nevertheless, the long-term benefits of good glycaemic control in diabe- tes are unequivocal and current technologies should help more patients to achieve this. • It is essential that all definitions of good glycaemic control include an absence of severe hypoglycaemia as well as near-normal glycated haemoglobin. Intensified treatment regimens should be adjusted to incorporate both.

REFERENCES

Agardh, C.D., Eckert, B. & Agardh, E. (1992) Irreversible progression of severe retinopathy in young type-1 insulin-dependent diabetes mellitus patients after improved metabolic control. Journal of Diabetes and its Complications, 6, 96–100. American Diabetes Association Workgroup on Hypoglycemia (2005) Defining and reporting hypoglycemia in diabetes: a report from the American Diabetes Association Workgroup on Hypoglycemia. Diabetes Care, 28, 1245–1249. Amiel, S.A., Tamborlane, W.V., Simonson, D.C. & Sherwin, R.S. (1987) Defective glucose counterregula- tion after strict glycemic control of insulin-dependent diabetes mellitus. New England Journal of Medi- cine, 316, 1376–1383. Amiel, S.A., Sherwin, R.S., Simonson, D.C. & Tamborlane, W.V. (1988) Effect of intensive insulin therapy on glycemic thresholds for counterregulatory hormone release. Diabetes, 37, 901–907. Ashwell, S.G., Amiel, S.A., Bilous, R.W. et al. (2006) Improved glycaemic control with insulin glargine plus insulin lispro: a multicentre, randomized, cross-over trial in people with type 1 diabetes. Diabetic Medicine, 23, 285–292. Barton, F.B., Rickels, M.R., Alejandro, R. et al. (2012) Improvement in outcomes of clinical islet trans- plantation: 1999–2010. Diabetes Care, 35, 1436–1445. Benedict, C., Kern, W., Schultes, B., Born, J. & Hallschmid, M. (2008) Differential sensitivity of men and women to anorexigenic and memory-improving effects of intranasal insulin. Journal of Clinical Endo- crinology and Metabolism, 93, 1339–1344. Bergenstal, R.M., Tamborlane, W.V., Ahmann, A. et al.; STAR 3 Study Group (2010) Effectiveness of sensor-augmented insulin-pump therapy in type 1 diabetes. New England Journal of Medicine, 363, 311–320. Bingham, E.M., Hopkins, D., Smith, D. et al. (2002) The role of insulin in human brain glucose metabolism: an 18Fluoro-Deoxyglucose positron emission tomography study. Diabetes, 51, 3384–3390. Bonds, D.E., Miller, M.E., Bergenstal, R.M. et al. (2010) The association between symptomatic, severe hypoglycaemia and mortality in type 2 diabetes: retrospective epidemiological analysis of the ACCORD study. British Medical Journal, 340, b4909. Bott, U., Jorgens, V., Grusser, M., Bender, R., Muhlhauser, I. & Berger, M. (1994) Predictors of glycaemic control in type 1 diabetic patients after participation in an intensified treatment and teaching programme. Diabetic Medicine, 11, 362–371. Bott, U., Bott, S., Hemmann, D. & Berger, M. (2000) Evaluation of a holistic treatment and teaching programme for patients with type 1 diabetes who failed to achieve their therapeutic goals under intensi- fied insulin therapy. Diabetic Medicine, 17, 635–643. Carlson, M.G. & Campbell, P.J. (1993) Intensive insulin therapy and weight gain in IDDM. Diabetes, 42, 1700–1707. Castilloa, M.J., Sheen, A.J. & Lefebvre, P.J. (1996) The degree/rapidity of the metabolic deterioration fol- lowing interruption of a continuous subcutaneous insulin infusion is influenced by the prevailing blood glucose level. Journal of Clinical Endocrinology and Metabolism, 81, 1975–1978. Chantelau, E. & Frystyk, J. (2005) Progression of diabetic retinopathy during improved metabolic control may be treated with reduced insulin dosage and/or somatostatin analogue administration – a case report. Growth Hormone & IGF Research, 15, 130–135. 160 Hypoglycaemia in Clinical Diabetes

Chantelau, E. & Kohner, E.M. (1997) Why some cases of retinopathy worsen when diabetic control improves. British Medical Journal, 315, 1105–1106. Choudhary, P., Lonnen, K., Emery, C.J. et al. (2009) Comparing hormonal and symptomatic responses to experimental hypoglycaemia in insulin- and sulphonylurea-treated type 2 diabetes. Diabetic Medicine, 26, 665–672. Choudhary, P., Geddes, J., Freeman, J.V., Emery, C.J., Heller, S.R. & Frier, B.M. (2010) Frequency of biochemical hypoglycaemia in adults with type 1 diabetes with and without impaired awareness of hypoglycaemia: no identifiable differences using continuous glucose monitoring.Diabetic Medicine, 27, 666–672. Choudhary, P., Shin, J., Wang, Y. et al. (2011) Insulin pump therapy with automated insulin suspension in response to hypoglycemia: reduction in nocturnal hypoglycemia in those at greatest risk. Diabetes Care, 34, 2023–2025. CITR Research Group (2009) 2007 update on allogeneic islet transplantation from the Collaborative Islet Transplant Registry (CITR). Cell Transplantation, 18, 753–767. Clar, C., Barnard, K., Cummins, E., Royle, P., Waugh, N. & Aberdeen Health Technology Assessment Group (2010) Self-monitoring of blood glucose in type 2 diabetes: systematic review. Health Technology Assessment, 14, 1–140. Clarke, W.L., Cox, D.J., Gonder-Frederick, L.A., Julian, D., Schlundt, D. & Polonsky, W. (1995) Reduced awareness of hypoglycemia in adults with IDDM; a prospective study of hypoglycemic frequency and associated symptoms. Diabetes Care, 18, 517–522. Cranston, I., Lomas, J., Maran, A., Macdonald, I.A. & Amiel, S.A. (1994) Restoration of hypoglycaemia awareness in patients with long-duration insulin-dependent diabetes. Lancet, 344, 283–287. Dabby, R., Sadeh, M., Lampl, Y., Gilad, R. & Watemberg, N. (2009) Acute painful neuropathy induced by rapid correction of serum glucose levels in diabetic patients. Biomedicine and Pharmacotherapy, 63, 707–709. Davis, S.N., Shavers, C., Mosqueda-Garcia, R. & Costa, F. (1997) Effects of differing antecedent hypogly- cemia on subsequent counterregulation in normal humans. Diabetes, 46, 1328–1335. DAFNE Study Group (2002) Training in flexible, intensive insulin management to enable dietary freedom in people with type 1 diabetes: Dose Adjustment For Normal Eating (DAFNE) randomised controlled trial. British Medical Journal, 325, 746–749. Donaghue, K.C., Fung, A.T., Hing, S. et al. (1997) The effect of prepubertal diabetes duration on diabetes microvascular complications in early and late adolescence. Diabetes Care, 20, 77–80. Duckworth, W., Abraira, C., Moritz, T. et al. (2009) Glucose control and vascular complications in veterans with type 2 diabetes. New England Journal of Medicine, 360, 129–139. Dunn, F.L., Nathan, D.M., Scavini, M., Selam, J.L. & Wingrove, T.G. (1997) Long term therapy of IDDM with an implantable insulin pump. The Plantable Insulin Pump Trial Study Group. Diabetes Care, 20, 59–63. Dunn, J.T., Cranston, I., Marsden, P.K., Amiel, S.A. & Reed, L.J. (2007) Attenuation of amydgala and frontal cortical responses to low blood glucose concentration in asymptomatic hypoglycemia in type 1 diabetes: a new player in hypoglycemia unawareness? Diabetes, 56, 2766–2773. Egger, M., Davey Smith, G., Stettler, C. & Diem, P. (1997) Risk of adverse effects of intensified treatment in insulin-dependent diabetes mellitus: a meta-analysis. Diabetic Medicine, 14, 919–928. European Medicines Agency (2002). Note for guidance on clinical investigation of medicinal products in the treatment of diabetes mellitus [WWW document]. URL http://www.ema.europa.eu/docs/en_GB/ document_library/Scientific_guideline/2009/09/WC500003262.pdf [accessed on 16 December 2011]. Fanelli, C.G., Epifano, L., Rambotti, A.M. et al. (1993) Meticulous prevention of hypoglycemia normalizes the glycemic thresholds and magnitude of most of neuroendocrine responses to, symptoms of, and cogni- tive function during hypoglycemia in intensively treated patients with short-term IDDM. Diabetes, 42, 1683–1689. Gaede, P., Vedel, P., Larsen, N., Jensen, G.V., Parving, H.H. & Pedersen, O. (2003) Multifactorial interven- tion and cardiovascular disease in patients with type 2 diabetes. New England Journal of Medicine, 348, 383–393. George, E., Harris, N., Bedford, C., Macdonald, I.A., Hardisty, C.A. & Heller, S.R. (1995) Prolonged but partial impairment of the hypoglycaemic physiological response following short-term hypoglycaemia in normal subjects. Diabetologia, 38, 1183–1190. George, E., Marques, J.L., Harris, N.D., Macdonald, I.A., Hardisty, C.A. & Heller, S.R. (1997) Preservation of physiological responses to hypoglycemia 2 days after antecedent hypoglycemia in patients with IDDM. Diabetes Care, 20, 1293–1298. Chapter 7: Risks of Intensive Therapy 161

Gibbons, C.H. & Freeman, R. (2010) Treatment-induced diabetic neuropathy: a reversible painful auto- nomic neuropathy. Annals of Neurology, 67, 534–541. Gold, A.E., MacLeod, K.M. & Frier, B.M. (1994) Frequency of severe hypoglycemia in patients with type 1 diabetes with impaired awareness of hypoglycemia. Diabetes Care, 17, 697–703. Hanssen, K.F., Dahl-Jorgensen, K., Lauritzen, T., Feldt-Rasmussen, B., Brinchmann-Hansen, O. & Deckert, T. (1986) Diabetic control and microvascular complications: the near-normoglycaemic experience. Diabetologia, 29, 677–684. Heller, S. (2004) Weight gain during insulin therapy in type 2 diabetes mellitus. Diabetes Research and Clinical Practice, 65 (Suppl 1), S23–S27. Heller, S., Chapman, J., McLoud, J. & Ward, J. (1995) Unreliability of reports of hypoglycaemia by diabetic patients. British Medical Journal, 310, 440. Heller, S.R. & Cryer, P.E. (1991) Reduced neuroendocrine and symptomatic responses to subsequent hypoglycemia after 1 episode of hypoglycemia in non-diabetic humans. Diabetes, 40, 223–226. Heller, S.R., Amiel, S.A., Evans, M.L., Kong, M.F., Macdonald, I.A. & Peacey, S.R. (2002) Does insulin lispro preserve the physiological defences to hypoglycaemia during intensive insulin therapy with a conventional basal bolus regimen? Diabetes, Obesity and Metabolism, 4, 106–112. Hendrickson, M., Janzon, L. & Groop, L. (1995) Progression of retinopathy after change of treatment from oral antihyperglycemic agents to insulin in patients with NIDDM. Diabetes Care, 18, 1571–1576. Hepburn, D.A., Patrick, A.W., Eadington, D.W., Ewing, D.J. & Frier, B.M. (1990) Unawareness of hypogly- caemia in insulin-treated diabetic patients: prevalence and relationship to autonomic neuropathy. Dia- betic Medicine, 7, 711–717. Holl, R.W., Lang, G.E., Grabert, M., Heinze, E., Lang, G.K. & Debatin, K.M. (1998) Diabetic retinopathy in pediatric patients with type 1 diabetes: effect of diabetes duration, prepubertal and pubertal onset of diabetes, and metabolic control. Journal of Pediatrics, 132, 790–794. Holman, R., Paul, S.K., Angelyn Bethel, M., Matthews, D. & Neil, A. (2008) 10-year follow-up of intensive glucose control in type 2 diabetes. New England Journal of Medicine, 359, 1577–1589. Hoogma, R.P., Hammond, P.J., Gomis, R. et al. (2006) Comparison of the effects of continuous subcutane- ous insulin infusion (CSII) and NPH-based multiple daily insulin injections (MDI) on glycaemic control and quality of life: results of the 5-nations trial. Diabetic Medicine, 23, 141–147. Hopkins, D., Lawrence, I., Mansell, P. et al. (2012) Improved biomedical and psychological outcomes 1 year after structured education in flexible insulin therapy for people with type 1 diabetes: the UK DAFNE experience. Diabetes Care, 35, 1638–1642. Hovorka, R., Kumareswaran, K., Harris, J. et al. (2011) Overnight closed loop insulin delivery (artificial pancreas) in adults with type 1 diabetes: crossover randomised controlled studies. British Medical Journal, 342, d1855. doi: 10.1136/bmj.d1855 Jorgens, V., Grusser, M., Bott, U., Mulhauser, I. & Berger, M. (1993) Effective and safe translation of intensified insulin therapy to general internal medicine departments. Diabetologia, 36, 99–105. Klein, B.E., Knudtson, M.D., Tsai, M.Y. & Klein, R. (2009) The relation of markers of inflammation and endothelial dysfunction to the prevalence and progression of diabetic retinopathy: Wisconsin epidemio- logic study of diabetic retinopathy. Archives of Ophthalmology, 127, 1175–1182. Knight, G., Jennings, A.M., Boulton, A.J., Tomlinson, S. & Ward, J.D. (1985) Severe hyperkalaemia and ketoacidosis during routine treatment with an insulin pump. British Medical Journal, 291, 371–372. Knight, G., Boulton, A.J., Drury, J. & Ward, J.D. (1986) Long term glycaemic control by alternative regimens in a feasibility study of continuous subcutaneous insulin infusion. Diabetes Research, 3, 335–338. Korzon-Burakowska, A., Hopkins, D., Matyka, K. et al. (1998) Effects of glycemic control on protective responses against hypoglycemia in type 2 diabetes. Diabetes Care, 21, 283–290. Krolewski, A.S., Laffel, L.M., Krolewski, M., Quinn, M. & Warram, J.H. (1995) Glycosylated hemoglobin and the risk of microalbuminuria in patients with insulin-dependent diabetes mellitus. New England Journal of Medicine, 332, 1251–1255. Lassmann-Vague, V., Belicar, P., Alessis, C., Raccah, D., Vialettes, B. & Vague, P. (1996) Insulin kinetics in type I diabetic patients treated by continuous intraperitoneal insulin infusion: influence of anti-insulin antibodies. Diabetic Medicine, 13, 1051–1055. Lawton, J, Rankin, D, Cooke, D, Elliott, J, Amiel, S & Heller S; for the UK NIHR DAFNE Study Group (2012) Patients′ experiences of adjusting insulin doses when implementing flexible intensive insulin therapy: a longitudinal, qualitative investigation. Diabetes Research and Clinical Practice, 98, 236–242. 162 Hypoglycaemia in Clinical Diabetes

Limbert, C., Schwingshandl, J., Haas, J., Roth, R. & Borkenstein, M. (1993) Severe hypoglycemia in children and adolescents with IDDM: frequency and associated factors. Journal of Diabetes and its Complications, 7, 216–220. Linn, T., Ortac, K., Laube, H. & Federlin, K. (1996) Intensive therapy in adult insulin-dependent diabetes mellitus is associated with improved insulin sensitivity and reserve: a randomized, controlled, prospec- tive study over 5 years in newly diagnosed patients. Metabolism: Clinical and Experimental, 45, 1508–1513. Maran, A., Lomas, J., Macdonald, I.A. & Amiel, S.A. (1995) Lack of preservation of higher brain function during hypoglycaemia in patients with intensively-treated IDDM. Diabetologia, 38, 1412–1418. Matyka, K., Evans, M., Lomas, J., Cranston, I., Macdonald, I. & Amiel, S.A. (1997) Altered hierarchy of protective responses against severe hypoglycemia in normal aging in healthy men. Diabetes Care, 20, 135–141. McIntyre, H.D., Knight, B.A., Harvey, D.M., Noud, M.N., Hagger, V.L. & Gilshenan, K.S. (2010) Dose adjustment for normal eating (DAFNE) - an audit of outcomes in Australia. Medical Journal of Australia, 192, 637–640. Miller, M.E., Bonds, D.E., Gerstein, H.C. et al. (2010) The effects of baseline characteristics, glycaemia treatment approach, and glycated haemoglobin concentration on the risk of severe hypoglycaemia: post hoc epidemiological analysis of the ACCORD study. British Medical Journal, 340, b5444. Moskalets, E., Galstyan, G., Starostina, E., Antsiferov, M. & Chantelau, E. (1994) Association of blindness to intensification of glycemic control in insulin-dependent diabetes mellitus. Journal of Diabetes and its Complications, 8, 45–50. Mühlhauser, I., Overmann, H., Bender, R., Bott, U. & Berger, M. (1998) Risk factors of severe hypogly- caemia in adult patients with type 1 diabetes – a prospective population based study. Diabetologia, 41, 1274–1282. Nathan, D.M., Lachin, J., Cleary, P. et al. (2003) Intensive diabetes therapy and carotid intima-media thickness in type 1 diabetes mellitus. New England Journal of Medicine, 348, 2294–2303. Nimri, R., Atlas, E., Ajzensztejn, M., Miller, S., Oron, T. & Phillip, M. (2012) Feasibility study of automated overnight closed-loop glucose control under MD-logic artificial pancreas in patients with type 1 diabetes: the DREAM Project. Diabetes Technology and Therapeutics, 14, 728–735. Peacey, S.R., Rostami-Hodjegan, A., George, E., Tucker, G.T. & Heller, S.R. (1997) The use of tolbutamide- induced hypoglycemia to examine the intraislet role of insulin in mediating glucagon release in normal humans. Journal of Clinical Endocrinology and Metabolism, 82, 1458–1461. Pedersen-Bjergaard, U., Agerholm-Larsen, B., Pramming, S., Hougaard, P. & Thorsteinsson, B. (2003) Prediction of severe hypoglycaemia by angiotensin-converting enzyme activity and genotype in type 1 diabetes. Diabetologia, 46, 89–96. Peyrot, M., Rubin, R.R., Lauritzen, T. et al. (2005) Resistance to insulin therapy among patients and pro- viders: results of the cross-national Diabetes Attitudes, Wishes, and Needs (DAWN) study. Diabetes Care, 28, 2673–2679. Pickup, J., Mattock, M. & Kerry, S. (2002) Glycaemic control with continuous subcutaneous insulin infu- sion compared with intensive insulin injections in patients with type 1 diabetes: meta-analysis of ran- domised controlled trials. British Medical Journal, 324, 705–708. Pickup, J.C. & Sutton, A.J. (2008) Severe hypoglycaemia and glycaemic control in Type 1 diabetes: meta- analysis of multiple daily insulin injections compared with continuous subcutaneous insulin infusion. Diabetic Medicine, 25, 765–774. Plank, J., Kohler, G., Rakovac, I. et al. (2004) Long-term evaluation of a structured outpatient education programme for intensified insulin therapy in patients with type 1 diabetes: a 12-year follow-up.Diabeto - logia, 47, 1370–1375. Purnell, J.Q., Hokanson, J.E., Marcovina, S.M., Steffes, M.W., Cleary, P.A. & Brunzell, J.D. (1998) Effect of excessive weight gain with intensive therapy of type 1 diabetes on lipid levels and blood pressure: results from the Diabetes Control and Complications Trial. Journal of the American Medical Association, 280, 140–146. Rankin, D., Cooke, D.D., Clark, M. et al. (2011) How and why do patients with Type 1 diabetes sustain their use of flexible intensive insulin therapy? A qualitative longitudinal investigation of patients′ self- management practices following attendance at a Dose Adjustment for Normal Eating (DAFNE) course. Diabetic Medicine, 28, 532–538. Reaven, P.D., Moritz, T.E., Schwenke, D.C. et al. (2009) Intensive glucose-lowering therapy reduces cardiovascular disease events in Veterans Affairs Diabetes Trial participants with lower calcified coronary atherosclerosis. Diabetes, 58, 2642–2648. Chapter 7: Risks of Intensive Therapy 163

Rodrigues, I., Reid, H.A., Ismail, K. & Amiel, S.A. (2005) Indications and efficacy of continuous subcu- taneous insulin infusion (CSII) therapy in type 1 diabetes mellitus: a clinical audit in a specialist service. Diabetic Medicine, 22, 842–849. Rogers, H., Turner, E., Thompson, G., Hopkins, D. & Amiel, S.A. (2009) Hub-and-spoke model for a 5-day structured patient education programme for people with Type 1 diabetes. Diabetic Medicine, 26, 915–920. Rogers, H.A., de Zoysa, N. & Amiel, S.A. (2012) Patient experience of hypoglycaemia unawareness in Type 1 diabetes: are patients appropriately concerned? Diabetic Medicine, 29, 321–327. Ryder, R.E., Owens, D.R., Hayes, T.M., Ghatei, M.A. & Bloom, S.R. (1990) Unawareness of hypoglycaem­ ­ia and inadequate hypoglycaemic counterregulation: no causal relation with diabetic autonomic neuropa- thy. British Medical Journal, 301, 783–787. Ryder, R.E.J., Thong, K.Y., Cull, M.L., Mills, A.P., Walton, C. & Winocour, P.H. (2010) The Association of British Clinical Diabetologists (ABCD) nationwide exenatide audit. Practical Diabetes International, 27, 352–357. Samann, A., Muhlhauser, I., Bender, R., Kloos, C. & Muller, U.A. (2005) Glycaemic control and severe hypoglycaemia following training in flexible, intensive insulin therapy to enable dietary freedom in people with type 1 diabetes: a prospective implementation study. Diabetologia, 48, 1965–1970. Seaquist, E.R., Anderson, J., Childs, B, et al. (2013) Hypoglycemia and diabetes: A report of a Workgroup of the American Diabetes Association and the Endocrine Society. Diabetes Care, 36, 1384–1395. Shah, S.C., Malone, J.L. & Simpson, N.E. (1989) A randomized trial of intensive insulin therapy in newly diagnosed insulin-dependent diabetes mellitus. New England Journal of Medicine, 320, 550– 554. Simonson, D.C., Tamborlane, W.V., DeFronzo, R.A. & Sherwin, R.S. (1985a) Intensive insulin therapy reduces counterregulatory hormone responses to hypoglycemia in patients with type 1 diabetes. Annals of Internal Medicine, 103, 184–190. Simonson, D.C., Tamborlane, W.V., Sherwin, R.S., Smith, J.D. & DeFronzo, R.A. (1985b) Improved insulin sensitivity in patients with type 1 diabetes mellitus after CSII. Diabetes, 34 (Suppl. 3), 80–86. Smith, C.B., Choudhary, P., Pernet, A., Hopkins, D. & Amiel, S.A. (2009) Hypoglycemia unawareness is associated with reduced adherence to therapeutic decisions in patients with type 1 diabetes: evidence from a clinical audit. Diabetes Care, 32, 1196–1198. Speight, J., Amiel, S.A., Bradley, C. et al. (2010) Long-term biomedical and psychosocial outcomes fol- lowing DAFNE (Dose Adjustment For Normal Eating) structured education to promote intensive insulin therapy in adults with sub-optimally controlled Type 1 diabetes. Diabetes Research and Clinical Prac- tice, 89, 22–29. Srinivasan, P., Huang, G.C., Amiel, S.A. & Heaton, N.D. (2007) Islet cell transplantation. Postgraduate Medical Journal, 83, 224–229. Steffes, M.W., Sibley, S., Jackson, M. & Thomas, W. (2003) Beta-cell function and the development of diabetes-related complications in the Diabetes Control and Complications Trial. Diabetes Care, 26, 832–836. Sutherland, D.E. & Gruessner, A.C. (2007) Long-term results after pancreas transplantation. Transplanta- tion Proceedings, 39, 2323–2325. The Action to Control Cardiovascular Risk in Diabetes Study Group (2008) Effects of intensive glucose lowering in type 2 diabetes. New England Journal of Medicine, 358, 2545–2559. The ADVANCE Collaborative Group (2008) Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. New England Journal of Medicine, 358, 2560–2572. The Diabetes Control and Complications Trial Research Group (1991) Epidemiology of severe hypoglyc- emia in the Diabetes Control and Complications Trial. American Journal of Medicine, 90, 450–459. The Diabetes Control and Complications Trial Research Group (1988) Weight gain associated with inten- sive therapy in the Diabetes Control and Complications Trial. Diabetes Care, 11, 567–573. The Diabetes Control and Complications Trial Research Group (1993) The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. New England Journal of Medicine, 329, 977–986. The Diabetes Control and Complications Trial Research Group (1995a) Adverse events and their associa- tion with treatment regimens in the Diabetes Control and Complications Trial. Diabetes Care, 18, 1415–1427. The Diabetes Control and Complications Trial Research Group (1995b) Implementation of treatment pro- tocols in the Diabetes Control and Complications Trial. Diabetes Care, 18, 361–376. 164 Hypoglycaemia in Clinical Diabetes

The Diabetes Control and Complications Trial Research Group (1996a) Influence of intensive diabetes treatment on quality-of-life outcomes in the Diabetes Control and Complications Trial. Diabetes Care, 19, 195–203. The Diabetes Control and Complications Trial Research Group (1996b) The absence of a glycemic thresh- old for the development of long-term complications: the perspective of the Diabetes Control and Com- plications Trial. Diabetes, 45, 1289–1298. The Diabetes Control and Complications Trial Research Group (1997) Hypoglycemia in the Diabetes Control and Complications Trial. Diabetes, 46, 271–286. The Diabetes Control and Complications Trial Research Group (1998) Effect of intensive therapy on residual beta-cell function in patients with type 1 diabetes in the Diabetes Control and Complications Trial. A randomized, controlled trial. Annals of Internal Medicine, 128, 517–523. The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications Research Group (2002) Effect of intensive therapy on the microvascular complications of type 1 diabetes mellitus. Journal of the American Medical Association, 287, 2563–2569. The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Study Research Group (2005) Intensive Diabetes Treatment and Cardiovascular Disease in Patients with Type 1 Diabetes. New England Journal of Medicine, 353, 2643–2653. The Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group (2008) Continu- ous glucose monitoring and intensive treatment of type 1 diabetes. New England Journal of Medicine, 359, 1464–1476. The Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group (2010) Effec- tiveness of continuous glucose monitoring in a clinical care environment: evidence from the Juvenile Diabetes Research Foundation continuous glucose monitoring (JDRF-CGM) trial. Diabetes Care, 33, 17–22. The Kroc Collaborative Study Group (1984) Blood glucose control and the evolution of diabetic retinopathy and albuminuria — a preliminary multicenter trial. New England Journal of Medicine, 311, 365–372. The Kroc Collaborative Study Group (1988) Diabetic retinopathy after two years of intensified insulin treatment. Follow-up of the Kroc Collaborative Study. Journal of the American Medical Association, 260, 37–41. UK Hypoglycaemia Study Group (2007) Risk of hypoglycaemia in types 1 and 2 diabetes: effects of treat- ment modalities and their duration. Diabetologia, 50, 1140–1147. Wahren, J., Kallas, A. & Sima, A.A. (2012) The clinical potential of C-peptide replacement in type 1 diabetes. Diabetes, 61, 761–772. White, N.H., Skor, D.A., Cryer, P.E., Levandoski, L.A., Bier, D.M. & Santiago, J.V. (1983) Identification of type 1 diabetic patients at increased risk for hypoglycemia during intensive therapy. New England Journal of Medicine, 308, 485–491. Yki-Jarvinen, H., Kauppinen-Makelin, R., Tiikkainen, M. et al. (2006) Insulin glargine or NPH combined with metformin in type 2 diabetes: the LANMET study. Diabetologia, 49, 442–451. Zachariah, S., Sheldon, B., Shojaee-Moradie, F. et al. (2011) Insulin detemir reduces weight gain as a result of reduced food intake in patients with type 1 diabetes. Diabetes Care, 34, 1487–1491. 8 Management of Acute and Recurrent Hypoglycaemia Rory J. McCrimmon University of Dundee, Dundee, UK

INTRODUCTION

Despite advances in insulin pharmacology and its delivery and in patient education, the frequency of symptomatic hypoglycaemia remains substantial, with the average patient likely to experience many thousands of episodes over the course of their lifetime with insulin-treated diabetes. Even the rates of severe hypoglycaemia have remained unchanged over the last two decades. Some individuals are at greater risk than others of experiencing severe, disabling hypoglycaemia, which is a frightening experience for all with insulin- treated diabetes as well as being a major source of concern for their relatives and carers. This chapter reviews lifestyle moderators, factors and co-morbidities that influence risk of developing hypoglycaemia and describes the current approaches to the management of both acute and recurrent hypoglycaemia. The approaches used to identify the pattern of hypogly- caemia in any one individual are examined and how they may be used to consider the additional strategies and approaches that could reduce hypoglycaemia frequency, particu- larly of severe episodes.

MODERATORS OF HYPOGLYCAEMIA RISK

Major risk factors for hypoglycaemia in type 1 diabetes are discussed in detail in Chapter 4 and listed in Table 8.1, bearing in mind that the main cause is the limitation of therapeutic subcutaneous insulin delivery. It is often said that insulin is given at the ‘wrong’ time, in the ‘wrong’ place and in the ‘wrong’ dose. Unsurprisingly, a mismatch between insulin administration and carbohydrate absorption frequently occurs, leading to overinsulinis­ ation at some times of day and the potential risk of hypoglycaemia. When combined with deficient counterregulatory hormone responses (see Chapter 3) this ensures that glucose homeostasis is profoundly disrupted in type 1 diabetes and in type 2 diabetes of longer duration. The Diabetes Control and Complications Trial (DCCT) highlighted the risks of intensive insulin therapy by showing that improving glycaemic control and lowering HbA1c was associated with a trebling of risk for severe hypoglycaemia (The Diabetes Control and

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 166 Hypoglycaemia in Clinical Diabetes

Table 8.1 Causes of hypoglycaemia in people with diabetes

Altered insulin: Change in insulin Change in insulin carbohydrate Other related sensitivity pharmacodynamics ratio conditions

‘Honeymoon period’ Change in injection Unplanned Endocrine dysfunction in newly-diagnosed site exercise (Addison’s disease, type 1 diabetes hypopituitarism, hypothyroidism) Post-partum Change in insulin Breastfeeding Psychological formulation Menstruation Change in temperature Gastroparesis Alcohol Lipohypertrophy Malabsorption e.g. Coeliac Disease Renal failure Exercise Puberty, advanced age

Complications Trial Research Group 1997). However, blood glucose control per se does not entirely predict the occurrence of severe hypoglycaemia. In the intensively treated group in the DCCT, HbA1c only accounted for 60% of the risk of severe episodes (The Diabetes Control and Complications Trial Research Group 1997). Additional factors con- tributing to hypoglycaemia risk were also of importance and need to be considered when reviewing an individual patient. These may include simple factors such as those that influence insulin absorption after subcutaneous injection: • depth of injection; • site of injection; • localised lipohypertrophy; • phase of the menstrual cycle in women; • exercise; • ambient temperature; • medicines that affect skin blood flow; • pre-mixed insulin preparations – whether the insulin has been shaken adequately before injecting.

Similarly, the matching of insulin injection into the systemic circulation to carbohydrate delivery is not just dependent on insulin dose and timing. The following are important variables to note: • Food absorption from the gastrointestinal tract can be highly variable within individuals. • Food absorption is affected by meal size, constituents and the speed with which the meal is eaten. • Gastric emptying is influenced by meal content (especially fat), prevailing blood glucose levels, and the integrity of the autonomic nervous system (Feldman and Schiller 1983; Vinik et al. 2003). Chapter 8: Management of Acute and Recurrent Hypoglycaemia 167

In addition, a number of important lifestyle moderators can predispose towards hypogly- caemia. Some, such as pregnancy (Chapter 11) and exercise (Chapter 17), are discussed elsewhere. However, a major moderator, which is particularly relevant in westernised socie- ties, is alcohol.

Alcohol and hypoglycaemia Alcohol is an important predictor of hypoglycaemia for individuals treated with insulin, with estimates suggesting that up to 20% of severe events may be attributable to its use (Potter et al. 1982; Nilsson et al. 1988). However, there is nothing to suggest that (in general terms) people with type 1 diabetes adopt a different approach to their use of alcohol than the rest of the population. Alcohol is associated with hypoglycaemia in several ways. It:

• impairs recognition of hypoglycaemia symptoms; • may lead to a diagnosis of intoxication rather than hypoglycaemia; • may impair counterregulatory responses; • can increase cognitive dysfunction during hypoglycaemia (Cheyne and Kerr 2002; Cheyne et al. 2002).

The biochemistry underlying the effect of alcohol on glucose homeostasis is complex and not completely understood. More than 90% of an ethanol load is metabolised by the liver, being catalysed by alcohol dehydrogenase into acetate. This reaction requires nico- tinamide di-nucleotide (NAD+), which in turn is dependent on glycogen stores in the liver. Hence, when glycogen levels are depleted (e.g. in malnourished individuals without dia- betes and after prolonged exercise), NAD+ levels are insufficient to adequately metabolise ingested alcohol and alcohol can then suppress hepatic gluconeogenesis, leading to a fall in blood glucose. In contrast, when liver glycogen levels are adequate, for instance follow- ing a meal, alcohol does not usually cause hypoglycaemia (Kerr et al. 1990; Avogaro et al. 1993). Ethanol appears to have little effect on hepatic glucose output in well-fed subjects with diabetes (Trojan et al. 1999) or without (Gin et al. 1992). However, alcohol can suppress lipolysis acutely (Avogaro et al. 1993), which may contribute to hypoglycaemia risk in type 1 diabetes, particularly overnight when glucose homeostasis is more dependent upon free fatty acid production (Avogaro et al. 1993) (Hagstrom-Toft et al. 1997). The alcohol-induced suppression of lipolysis may therefore predispose to hypoglycaemia the next morning. Investigators have used the hyperinsulinaemic glucose clamp technique (see Chapter 3) to examine the effect of alcohol on counterregulatory responses to hypoglycaemia in people with and without diabetes. Glucagon release has been shown to be suppressed by alcohol in some studies (Rasmussen et al. 2001), but not in others (Kerr et al. 1990), while in type 1 diabetes, the ingestion of modest amounts of alcohol attenuated the usual growth hormone response to mild hypoglycaemia (Kerr et al. 2007). As alcohol has a prolonged impact on hypoglycaemia following ingestion it seems more likely to implicate either cortisol or growth hormone responses to hypoglycaemia, although direct evidence for this is currently lacking. In a clinical context the main concern for insulin-treated individuals is that moderate alcohol intake (6–9 units) can acutely diminish hypoglycaemia awareness, despite a normal or near normal symptomatic response to hypoglycaemia (Moriarty et al. 1993) and may 168 Hypoglycaemia in Clinical Diabetes

4-Choice reaction time

Alcohol and hypoglycaemia

Hypoglycaemia

Alcohol

–120 –100 –80 –60 –40 –20 0 Time (ms)

Figure 8.1 Effect of alcohol on cognitive function during hypoglycaemia. When studied using the hyperinsulinaemic glucose clamp technique, alcohol ingestion was found to significantly prolong mean 4-choice reaction time in subjects with type 1 diabetes under both euglycaemic and hypoglycaemic conditions (Alcohol). Acute hypoglycaemia (Hypoglycaemia) also resulted in a prolongation of the reaction time and the effects of alcohol plus hypoglycaemia (Alcohol and Hypoglycaemia) were additive resulting in a marked reduction in cognitive performance. (Source: Adapted from Cheyne et al. 2004. Reproduced with permission of John Wiley & Sons Ltd.) exacerbate the effect of hypoglycemia to impair cognitive function (Cheyne et al. 2004) (see Figure 8.1). Guidance on alcohol and diabetes should stress certain key points:

• Alcohol is best taken with, or shortly before food, and carbohydrate should not be omitted from the meal. • Patients should be advised that the risk of hypoglycaemia may extend for ‘several hours’ after drinking. • Alcohol, even at levels within current statutory limits for driving in the UK, can impair awareness of hypoglycaemia and can further impair cognitive function (Cheyne et al. 2004). • Insulin-treated individuals have several reasons to avoid alcohol if planning to drive.

Ingestion of alcohol with an evening meal can also increase the risk of hypoglycaemia the following morning (Turner et al. 2001). A recent ambulatory study of people with type 1 diabetes using continuous glucose monitoring (CGM) confirmed a predisposition to delayed hypoglycaemia (Richardson et al. 2005b). In this study, alcohol was shown to reduce average interstitial tissue glucose compared to a non-alcoholic placebo drink and increase the risk of hypoglycaemia over the following 24 hours, with participants reporting more than twice as many hypoglycaemic episodes per day. In another study, the average interstitial glucose level was 1–2 mmol/l lower following alcohol, but the rate of hypogly- caemia the next day depended on the prevailing levels of glucose overnight. People who had lower baseline glucose at the beginning of the night or in the morning were at a greater risk of developing alcohol-induced hypoglycaemia than those with preceding hyperglycae- mia. This increased risk persisted for nearly a full day after alcohol ingestion (Richardson et al. 2005b). Chapter 8: Management of Acute and Recurrent Hypoglycaemia 169

Sensor profile – before pump therapy Sensor profile – after pump therapy 25.0 25.0

20.0 20.0

15.0 15.0

10.0 10.0 sensor value 5.0 5.0 Sensor value (mmol/l)

0.0 0.0 0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00 Time Time

Figure 8.2 CGM profile before and after conversion to treatment with continuous subcutaneous insulin infusion in a patient with >15 years of poorly controlled type 1 diabetes.

MONITORING FOR HYPOGLYCAEMIA

Capillary blood glucose monitoring A detailed clinical history is invaluable, but it can be difficult to ascertain the true frequency of hypoglycaemia experienced by any one individual with this alone. Many people with type 1 diabetes of long duration or who experience repeated episodes have a reduced ability to recognise hypoglycaemia sufficiently early to take appropriate action. In particular, they may fail to develop or recognise the warning symptoms of hypoglycaemia, or they may recognise the symptoms but be unable to take appropriate action because of neuroglyco- penia. Memory impairment is also associated with hypoglycaemia (see Chapter 2) and thus individual recall may underestimate the true frequency. Furthermore, the definition of hypoglycaemia needs to be agreed by the patient, their relatives and their health profes- sionals, especially for more modest events that are often considered to be an inevitable consequence of insulin treatment. It is therefore important that patients have additional methods of detecting low blood glucose. Thus usually involves ambulatory measurements using capillary samples and glucose meters. These can provide both an index of the frequency of biochemical hypogly- caemia and an estimate of how low the glucose can fall before hypoglycaemia is detected. However, this method has problems related to: • poor technique of blood sampling, e.g. inadequate volume; • technical limitations of the meters; • limitation in the number of readings that can be made during the day and night. Other methods for measuring glucose levels, using sensors for continuous glucose monitoring have been introduced and, while these devices still have limitations, they have highlighted the fact that hypoglycaemia remains a common problem in the management of type 1 diabetes and that many episodes are unrecognised (Cheyne and Kerr 2002; Cheyne et al. 2002). These methods can, however, be a useful aid, allowing patients (and their healthcare professionals) to determine the modulating influences of a number of factors on glycaemic control and insulin use (Figures 8.2 and 8.3).

Continuous Glucose Monitoring (CGM) The well-recognised limitations of traditional blood glucose measurements led to research into the development of new methods for the detection of low blood glucose levels. The most widely used technology is continuous glucose monitoring (Hoi-Hansen et al. 2005). 170 Hypoglycaemia in Clinical Diabetes

20.0

15.0

10.0

5.0

0.0 Glucose concentration(mmol/L) Glucose 0:00 4:00 8:00 12:00 16:00 20:00 24:00 Time

Figure 8.3 CGM profile indicating the extent of postprandial hyperglycaemia in a patient with type 1 diabetes. Individual capillary blood glucose measurements are shown as square boxes.

The first commercially available device for CGM, the Medtronic Minimed (Medtronic, Minneapolis, USA), was designed to monitor glucose levels continuously in tissue interstitial fluid (the extracellular fluid surrounding cells in the human body). The system comprised a disposable glucose sensor that is placed just under the skin (i.e. subcutaneous). This is linked to a non-implanted transmitter that communicates to a radio receiver. The receiver is worn like a pager and either stores or displays glucose levels throughout the day. The sensor measures interstitial glucose levels every 10 seconds and averages them over a 5-minute period, viz. 288 measurements each day, and provides measurements of glucose within the range of 2.2–22 mmol/l. The first models provided retrospective data. The latest generation of sensors transfer data to an onscreen display in real time which allows the patient to adjust his or her own medication (Bode et al. 2004). Several technical considerations that influence interstitial glucose monitoring need to be considered in evaluating the data that is produced. In the context of hypoglycaemia, specific limitations include:

• Glucose levels in interstitial fluid lag temporally behind glucose levels. This can be anywhere between 5 and 15 minutes. • It is unclear as to the significance of changes at the level of interstitial fluid compared to fluctuations in blood glucose level. • It is unclear whether changes in interstitial glucose levels, measured in the anterior abdominal wall, mirror those occurring at the level of glucose sensing neurones in the hypothalamus and elsewhere. • The recordings from CGM are markedly influenced by the accuracy, frequency and timing of the results of finger-stick blood testing for calibration purposes. Thus some systems require glucose to be calibrated when glucose is at steady state and is not rising or falling.

There are other limitations (not least of which is cost) in the use of CGM. The technique is in its infancy and some have challenged its ability to detect ‘true’ hypoglycaemia (McGowan et al. 2002). This is because CGM appears less sensitive during hypoglycaemia Chapter 8: Management of Acute and Recurrent Hypoglycaemia 171

Box 8.1 Guidelines for interpreting interstitial hypoglycaemia using CGM as used by the UK Hypoglycaemia Study Group (2007)

Defining an episode of hypoglycaemia

• There must be four consecutive readings of 3.5 mmol/l or lower for an episode to be classified as hypoglycaemia. • The first of the readings (of 3.5 mmol/l or less) signifies the start of hypoglycaemia. • The first reading of 3.5 mmol/l or above signifies the end of hypoglycaemia. • Hypoglycaemia ends fully when there are four or more consecutive readings above 3.5 mmol/l. • If during hypoglycaemia the sensor value is 3.5 mmol/l or higher for 1–3 readings and then goes back down below 3.5 mmol/l, then the whole episode is counted as one hypoglycaemic event and the short period above 3.5 mmol/l is included in the duration of hypoglycaemia. • To be labelled as moderate hypoglycaemia, the sensor has to read 3.0 mmol/l or below for at least four consecutive readings. • If during mild hypoglycaemia the reading falls to 3.0 mmol/l or below for four or more consecutive readings, the whole hypoglycaemic episode will be considered as moderate hypoglycaemia.

Prolonged hypoglycaemia is defined as lasting >2 hours. Published in the Department for Transport Road Safety Research Report No. 61 (2006)

where values are at the limit of the recording range of the monitor. In addition, differences between blood, interstitial and intra-cellular glucose as well as between different regions of the body (e.g. brain and anterior abdominal wall) can also influence data interpretation. Moreover, this relationship can vary depending on whether glucose is rising or falling (Caplin et al. 2003). There is also the potential for artefact, and hence a ‘flat-line’ may represent either low glucose, body position (such as lying on the sensor while asleep) or another technical problem. Together, these issues have raised uncertainty about how to define hypoglycaemia with CGM. Some groups have defined interstitial hypoglycaemia according to the level marking activation of the counterregulatory hormone cascade and onset of neuroglycopenic symptoms (Fanelli et al. 1994) (Box 8.1). These technical limitations need to be considered when interpreting CGM readings in clinical practice. CGM can overestimate the duration of hypoglycaemia (Cheyne and Kerr 2002; Cheyne et al. 2002) and both the frequency and duration of nocturnal hypoglycaemia in patients with strictly controlled type 1 diabetes (McGowan et al. 2002). Trials have shown poor concordance between a patient’s perception of hypoglycaemia and glucose values recorded using CGM, and even conflicting results when an individual wears two devices simultaneously (Larsen et al. 2004). However, even taking into account the uncer- tainty around CGM-identified hypoglycaemia, rates of low glucose are comparable whether measured by prospective collection of self-reported episodes or by continuous glucose monitoring. Overall therefore, it remains important in clinical practice to use CGM along with more traditional methods of recording hypoglycaemia events; namely relying on patients’ self-reports as well as standard blood glucose monitoring techniques. Taken together this information can be extremely valuable in the management of individuals who are prone to recurrent severe hypoglycaemia. 172 Hypoglycaemia in Clinical Diabetes

MANAGEMENT OF HYPOGLYCAEMIA

Acute hypoglycaemia Prevention is clearly better than cure and individuals with sulfonylurea and insulin-treated diabetes should be thoroughly and continuously educated about the potential risks of hypoglycaemia. In defining hypoglycaemia, Diabetes UK have recommended a policy of ‘4 is the floor’ as the capillary blood glucose level to alert a person that intervention is required to avoid hypoglycaemia, while the American Diabetes Association advocate a biochemical definition of 3.9 mmol/l as representing the onset of hypoglycaemia. However, these are clinical recommendations and represent glucose levels that many investigators consider to be set too high for epidemiological studies. In non-diabetic individuals and in people with type 1 diabetes with normal awareness and reasonable glycaemic control, the onset of symptomatic awareness of hypoglycaemia does not usually commence until the arterialised venous blood approaches 3.2 mmol/l, a level above that for cognitive dysfunc- tion. A capillary blood glucose level of 3.5 mmol/l has been suggested to define the onset of hypoglycaemia, which would be an acceptable threshold while remaining safe for clini- cal practice (Frier 2009). The treatment of acute hypoglycaemia depends on the spectrum of severity and ranges from treatment with the simple ingestion of oral carbohydrate at one end and at the other extreme, the need for acute medical resuscitation in an intensive care unit (Table 8.2). Adults who can recognise the early warning symptoms should be advised to eat 15–20 g of carbohydrate that is palatable, concentrated and portable (Figure 8.4). Examples include:

Table 8.2 Therapeutic options for the management of acute hypoglycaemia. (Source: Hypoglycaemia and Diabetes (eds B.M. Frier and B.M. Fisher), © 1993 Edward Arnold. Reproduced with permission from Edward Arnold Publishers Ltd)

Duration of hypoglycaemia

Ongoing management Initial Management (minutes) (hours)

By patient By family By Paramedics In hospital In intensive care emergency department Oral carbohydrate Oral carbohydrate Glucagon 1 mg Glucagon 1 mg Mannitol (20%, (15–20 g) (liquid/solid) IM or IV IM or IV 20 ml) Glucogel ®/ 75–80 ml 20% 75–80 ml 20% Dexamethasone Dextrogel® Dextrose over Dextrose over 10–15 minutes 10–15 minutes Glucagon 1 mg Dextrose/insulin IM infusion Oxygen Anticonvulsants Sedation

IM, intramuscular IV, intravenous Chapter 8: Management of Acute and Recurrent Hypoglycaemia 173

Acute hypoglycaemia in a person with type 1 diabetes in the community

Conscious Unconscious/ uncooperative

15–20gms CHO Glucogel®/ ® No* Dextrogel or BG>3.5?(mmol/l) IM glucagon 1mg No***

Yes BG>3.5?(mmol/l)

Long-acting CHO/ Yes** normal meal

Figure 8.4 Management of acute hypoglycaemia in the community. * can be repeated every 10–15 minutes and if remains unresponsive seek emergency medical help; ** consider urgent review by diabetes team as person is at increased risk of recurrent severe hypoglycaemia; *** call emergency services; BG, blood glucose; CHO, carbohydrate.

• 150–200 ml pure fruit juice; • 90–120 ml glucose drink, e.g. Lucozade®; • 5–7 Dextrosol ® tablets (4–5 Glucotabs®); • 3–4 heaped teaspoonfuls of sugar dissolved in water.

This can then be repeated up to three times if capillary blood glucose readings, taken every 10–15 minutes, do not rise above 3.5–4 mmol/l. It is important to follow this by consumption of long-acting complex carbohydrate. This can be in the form of bread (one slice) or biscuits/cookies (×2) or a 200–300 ml glass of milk, or a normal meal. The next level of treatment is when the patient is clearly hypoglycaemic but cannot or will not take oral fast-acting carbohydrate. Note that people on insulin may deny that they are hypoglycaemic and may not cooperate, or even react adversely to attempts to give them oral carbohydrate. Liquid glucose solutions are often unsatisfactory because the patient can spit them out. It is better to use a commercially-available glucose gel such as GlucoGel (Diabetic Bio-diagnostics) or Dextrogel®, which can be squeezed like toothpaste into the mouth and is absorbed through the buccal mucosa. Although some doubts have been expressed about the effectiveness of oral glucose gels, relatives often prefer this to injecting glucagon. Even jam (jelly) or honey can be used in an emergency. Glucogels should be avoided in semi-comatose patients or those at risk of aspirating. In the absence of glucogels an alternative is the intramuscular injection of glucagon (1 mg). Glucagon promotes hepatic glycogenolysis and the glycaemic response to a dose of 1 mg is essentially the same whether it is injected subcutaneously, intramuscularly or intravenously (Mühlhauser et al. 1985a). In all cases, glucagon should be followed up with 174 Hypoglycaemia in Clinical Diabetes complex carbohydrate as above. The advantage of glucagon is that it can be given by rela- tives or friends after some training. Paramedics can also use it at the patient’s home or in an ambulance. The disadvantages are that it has to be prepared for injection, takes longer (approximately 10 minutes) than intravenous glucose to restore consciousness and does not work in patients who have deficient or absent hepatic glycogen stores (alcoholics or people with cachexia). Unfortunately, even where glucagon is available, relatives or friends may not use it. In one study (Mühlhauser et al. 1985b), 53 of 123 episodes of severe hypogly- caemia were treated by relatives or friends with glucagon, 30 by attending doctors and 44 required treatment in hospital. When glucagon was available but not used, it was because those who knew how to use it were not present or were too anxious to do so. In one study involving children, (Daneman et al. 1989) glucagon was used in only a third of households in which it was available – presumably because relatives were either too scared or had been inadequately educated about its administration. The limited shelf life is a further limitation; Ward and colleagues found that nearly three-quarters of patients knew about glucagon but only 20% had a supply that was in date (Ward et al. 1990). Its effect is less certain if coma has been prolonged. In a review of 100 patients with hypoglycaemic coma treated in a hospital emergency department, glucagon was immediately effective in only 41% (Mac- Cuish et al. 1970). An essential part of education in the management of hypoglycaemia therefore is to ensure the subjects have in-date glucagon for injection at home and that family members are capable of administering it. For people with type 1 diabetes who are unconscious, and/or having seizures, or are conscious and aggressive, intravenous glucose may be required. Within 5 minutes the traditional dose of 50 ml of a 50% glucose (dextrose) solution raises blood glucose from below 1 to >12 mmol/l (Collier et al. 1987). However, this dose is now considered to be unnecessarily large and requires a very concentrated solution of glucose. Current recommendations to treat severe hypoglycaemia suggest injecting 75–80 ml of 20% glucose, given over 10–15 minutes. The main problem is the difficulty of giving an intravenous injection to an uncooperative patient who may be refusing or resisting treatment, or who has to be physically restrained to receive an intravenous injection. Extravasation of a very concentrated glucose solution outside the vein can cause painful phlebitis and localised tissue damage, and, because of its hypertonicity, even an intra- vascular injection can cause phlebitis or thrombosis. Intravenous treatment is not there- fore used much outside hospital, and few GPs carry glass vials of dextrose to treat severe hypoglycaemia. When a patient known to have type 1 diabetes is admitted to hospital in hypoglycaemic coma, but fails to recover after being given intravenous glucose, other causes of coma such as excessive ingestion of alcohol, self-poisoning with opiates or other drugs, acute vascular events such as subarachnoid haemorrhage or stroke, head injury or another intrac- ranial catastrophe must be excluded. Cerebral oedema is a recognised sequel of severe hypoglycaemia, and urgent neuroimaging is required to establish if it is present. Treatment options include glucocorticoid and mannitol infusion in addition to glucose normalisation (Table 8.2), and this complication is usually treated in an intensive care unit as it has a high mortality. The most difficult decision is to know for how long to continue treatment. Patients who have made a full recovery after being unconscious for several days may (anecdotally) appear subsequently to have significant cognitive impairment or (rarely) permanent brain damage. It is beyond the scope of this book to discuss the investigation and management of hypoglycaemia in non-diabetic individuals. Chapter 8: Management of Acute and Recurrent Hypoglycaemia 175

Recurrent hypoglycaemia The question for those responsible for the care of individuals with type 1 diabetes and insulin-treated type 2 diabetes who are experiencing recurrent hypoglycaemia is how best to manage their care without simply relaxing glycaemic control and thereby potentially increasing vascular risk. The first step should be to ensure that there are no significant co-morbidities contribut- ing to the hypoglycaemia risk. It is important to determine whether the patient has a malab­sorptive problem such as coeliac disease, or a previously undisclosed eating disor- der. Associated endocrine disorders such as Addison’s disease and hypopituitarism should be excluded, or disorders affecting insulin clearance (liver disease, renal failure, and hypothyroidism) or glucose production (alcohol-related disease). In most cases, however, recurrent hypoglycaemia is caused by a mismatch between insulin requirements and deliv- ery. An initial exploration of meal patterns, exercise and alcohol intake (as discussed above) and insulin injection routine is important and may reveal areas where patient edu- cation is necessary. This can be combined with nocturnal blood glucose testing or, if available, CGM to identify patterns in hypoglycaemia presentation or undetected noctur- nal hypoglycaemia.

Structured education Structured education programmes have been developed that are aimed to provide indi­ viduals with the information and skills to successfully manage intensive insulin therapy. The driver for this was the Düsseldorf education and training for dietary flexibility and insulin adjustment programme (Samann et al. 2005). This 5-day inpatient programme, which was provided in many centres across Germany, was the first to really demonstrate the effectiveness of a structured approach to diabetes care. In a 1-year evaluation of 9583 subjects with type 1 diabetes (from 96 participating diabetes centres) who had enrolled in the course, it was shown that mean baseline HbA1c had fallen from 8.1 to 7.3%, and yet despite this the incidence of severe hypoglycaemia actually decreased from 0.37 to 0.14 events per patient per year and the beneficial effects were most obvious in those patients in the lowest quartile of HbA1c. A related programme in the UK, called Dose Adjustment for Normal Eating (DAFNE), resulted in improved HbA1c and quality of life, although no significant reduction was shown in the incidence of severe hypoglycaemia in patients with poorly controlled diabetes (DAFNE Study Group 2002). However, more recently DAFNE investigators reported both a significant reduction in the incidence of severe hypoglycaemia from 1.7 ± 8.5 to 0.6 ± 3.7 episodes per person per year and that hypogly- caemia recognition improved in 43% of those reporting unawareness in a 1-year follow-up of enrolled patients (Hopkins et al. 2012). Other behavioural approaches based on symptom recognition include Blood Glucose Awareness Training and Hypoglycaemia Awareness and Avoidance (Amiel 2009). It should be noted that none of these approaches has as yet been shown to reverse impaired hypoglycaemia awareness in randomised controlled trials, and severe hypoglycaemia remains a risk even following successful implementation of such programmes.

Insulin regimens The experience of most clinicians is that intensive insulin therapy is best achieved using flexible insulin regimens that can more closely mimic normal physiology as well as adapt 176 Hypoglycaemia in Clinical Diabetes to the patient’s general lifestyle. In modern practice, this is most commonly achieved through the use of multi-injection therapy with insulin analogues or using continuous subcutaneous insulin infusion (CSII; insulin pump therapy). The data supporting the use of rapid-acting insulin analogues to limit severe hypoglycaemia risk is not impressive. However, basal analogues do seem effective in lowering the risk of nocturnal hypoglycae- mia. A recent meta-analysis of 20 randomised controlled trials (RCTs) with a duration of over 12 weeks that compared long-acting insulin analogues with isophane or Neutral Pro- tamine Hagedorn (NPH) human insulin found there was a small beneficial effect associated with long-acting analogue use in terms of HbA1c reduction (odds ratio [OR, 95% confidence interval] −0.07[−0.13; −0.01]%; P = 0.026) and risk of nocturnal and severe hypoglycae- mia (OR 0.69 [0.55; 0.86], and OR 0.73 [0.60; 0.89], respectively; all P < 0.01) (Monami et al. 2009). The data supporting CSII use to reduce risk of severe hypoglycaemia are limited, and despite earlier studies advocating this approach the most recent meta-analysis of 33 RCTs reported no overall benefit of CSII versus multiple daily injections (MDI) in severe hypoglycaemia frequency (Yeh et al. 2012).

CGM Although still a developing technique, CGM may be helpful clinically in identifying periods of increased hypoglycaemia risk and also in training individuals to prevent hypogly- caemia. A 26-week study of real-time continuous glucose monitoring systems (RT-CGMS) in 129 adults and children with intensively-treated (HbA1c < 7.0%) type 1 diabetes, found that regular use of RT-CGMS was associated with a small improvement in glucose control and a reduction in glucose variability (less time spent outside the target glucose range of ≤70 or >180 mg/dl; 3.9–10.0 mmol/l) (Beck et al. 2009). Although no overall effect was seen in the frequency of either biochemical or severe hypoglycaemia, the short duration of the trial may have been insufficient to demonstrate these outcomes. In a more recent trial, CGM use was shown to partially restore symptomatic and adrenaline responses to hypoglycaemia in a small cohort of subjects with type 1 diabetes (Ly et al. 2011). However, large trials may be relatively unhelpful in determining the precise benefit of this and other technology.

The artificial pancreas The artificial pancreas is a term that describes a number of different approaches tothe management of type 1 diabetes, all of which involve using technology to help people with diabetes to automatically control their blood glucose level by providing an artificial sub- stitute for the healthy pancreas. The technical approach to the development of an artificial pancreas is to combine CGM with an insulin pump. Using wireless technology and complex algorithms, real-time data from the CGM can then be delivered directly to the insulin pump to ‘close the loop’. Clinical application of this technology is very limited at present although it is an area of intensive research activity. In a small study of eight subjects with type 1 diabetes, an automated, portable closed loop system incorporating a Medtronic Paradigm Veo ® pump and the Medtronic Portable Glucose Control System ® was used to control blood glucose remotely throughout the night. It kept significantly more blood glucose levels within the range of 3.9–8.0 mmol/l compared with an open loop system (84.5% vs 46.7% of readings) and significantly reduced nocturnal hypoglycaemia (O’Grady et al. 2012). Chapter 8: Management of Acute and Recurrent Hypoglycaemia 177

More recently, a sensor-augmented insulin pump with a low glucose suspend feature was shown to reduce nocturnal hypoglycaemia in a short-term trial in subjects with type 1 diabetes (Choudhary et al. 2011).

Pancreas transplantation A few individuals develop impaired hypoglycaemia awareness and associated severe recur- rent episodes that can devastate their lives and those of their families. For them, referral to certain specialised centres offering pancreas transplantation, either whole organ or islet cells, is a therapeutic option. The criteria for referral for transplantation vary between countries, but in most cases severe recurrent hypoglycaemia and hypoglycaemia unaware- ness are major indications for transplantation. Pancreatic transplantation can improve both symptomatic and adrenaline responses to experimental hypoglycaemia in individuals with long-standing type 1 diabetes (Kendall et al. 1997) and can restore hypoglycaemia aware- ness (Leitao et al. 2008).

FUTURE POTENTIAL THERAPEUTIC APPROACHES

Amino acids Glucagon is the most potent counterregulatory hormone but the glucagon response to hypoglycaemia is lost in almost all C-peptide negative individuals with type 1 diabetes (see Chapters 3 and 4). Restoration of this response would help to reduce the risk of severe hypoglycaemia. In this context, amino acids have been used experimentally as they stimulate glucagon release from pancreatic α cells (Kuhara et al. 1991). Given orally, amino acids enhanced glucagon responses (P = 0.016) during hyperinsulinaemic hypogly- caemia in humans with type 1 diabetes and preserved some aspects of cognitive function (Rossetti et al. 2008). However, oral amino acids also stimulated glucagon release during euglycaemia and would therefore be anticipated to have deleterious effects on overall glycaemic control.

γ-aminobutyric acid (GABA) GABA is the major inhibitory neurotransmitter, and in rodent models, GABA levels decrease during systemic hypoglycaemia (Beverly et al. 2001). Pharmacological reduction of hypothalamic GABA has been shown to enhance the neuroendocrine responses to hypoglycaemia in rodent models (Chan et al. 2006). In non-diabetic humans, Modafinil, an agent used in narcolepsy that reduces GABA activity, was shown to produce a moderate increase in autonomic symptoms and heart rate, but had no effect on the counterregulatory responses to hypoglycaemia (Smith et al. 2004). Whether this modest effect could have any useful clinical benefit has still to be determined in clinical trials.

Beta agonists Rodent studies have shown that noradrenaline (norepinephrine), an excitatory monoamine neurotransmitter, is released into the ventromedial hypothalamus (VMH) (a key brain 178 Hypoglycaemia in Clinical Diabetes region involved in hypoglycaemia sensing) during hypoglycaemia (Beverly et al. 2001), and acting through β2 adrenergic receptors (B2AR), increases both adrenaline and gluca- gon responses to hypoglycaemia (Szepietowska et al. 2011). In a human trial based on these observations it was shown that night-time administration of oral terbutaline (5 mg), an oral B2AR agonist, reduced the percentage of blood glucose values below 3.9 mmol/l from 27% to 1% of the measured values. However, night-time terbutaline also increased morning blood glucose values (37% higher) and increased heart rate and blood lactate levels (Raju et al. 2006). Further work is needed in extended clinical trials both to define the optimum dose and establish whether such a treatment could be useful in clinical practice.

Selective Serotonin Reuptake Inhibitors (SSRI) Serotonergic neurones have a key role in the regulation of neuroendocrine function via both sympathoadrenal and hypothalamo-pituitary adrenal pathways. Although SSRIs have been associated with hypoglycaemia, the hypothesis that blocking serotonin uptake might actually increase sympathetic outflow was tested in a 6-week trial of high dose fluoxetine (60–80 mg) in individuals with type 1 diabetes. Fluoxetine was reported to significantly increase adrenaline (by 90%) and noradrenaline responses, as wellas increasing endogenous glucose production and lipolysis (all P < 0.05) (Briscoe et al. 2008). However, fluoxetine did not produce an increase in hypoglycaemia symptom scores, suggesting that the serotonergic pathways are not important for symptom genera- tion. This, together with the psychotropic effects of such therapy, may limit its clinical usefulness.

Theophylline Theophylline is an adenosine-receptor antagonist, which reduces cerebral blood flow and stimulates release of catecholamines. Acute intravenous theophylline can raise the glucose concentration for an increase in symptoms and adrenaline release during insulin-induced hypoglycaemia in those with type 1 diabetes and impaired hypoglycaemia awareness (de Galan et al. 2002). However, when oral theophylline was given over 2 weeks, despite an increase in symptomatic awareness, there was no change in the magnitude of adrenaline release (de Galan et al. 2003). Interestingly, the dissociation between counterregulatory hormone release during hypoglycaemia and symptom generation has been noted in a number of studies, suggesting that the central mechanisms involved in their generation may be different.

Fructose Glucokinase (GK) is the rate-limiting step in the phosphorylation of glucose and this directly regulates the magnitude of insulin release. GK also plays an important role in the sensing of hypoglycaemia by the brain (Levin et al. 2008). Fructose-6-phosphate (F6P) inhibits glucokinase, and fructose can be phosphorylated into F6P in the glucose- sensing neurones. Based on this, it was possible to demonstrate in subjects with type 1 diabetes that fructose delivered systemically produced a significant potentiation of the adrenaline response to hypoglycaemia with both an upward shift in its glycaemic thresh- Chapter 8: Management of Acute and Recurrent Hypoglycaemia 179 old as well as increased magnitude of secretion. Fructose also augmented endogenous glucose production, which aided improved recovery from hypoglycaemia (Gabriely and Shamoon 2005). However, these initial observations have yet to be examined in large clinical trials, particularly to measure the potential effect on lowering risk of severe hypoglycaemia.

Uncooked corn-starch The limitation of most bedtime snacks is that they are effective only over the first half of the night (peak effect of 2 hours), limiting their ability to prevent nocturnal hypoglycaemia (see Chapter 5). Uncooked corn-starch, a slowly released carbohydrate, has a prolonged action (peak action at 4 hours) and can stabilise blood glucose levels for up to 7 hours. When given at night to subjects with type 1 diabetes, blood glucose was found to be con- sistently 1.9 mmol/l higher at 03:00 h compared to placebo (P < 0.01), with a significantly higher fasting blood glucose the following morning (+1.1 mmol/l) (Axelsen et al. 1999). However, in another study where carbohydrate content was matched in the control arm, uncooked corn-starch did not significantly reduce the frequency of nocturnal hypoglycae- mia, although the rate of decline in blood glucose overnight was slower (Ververs et al. 1993).

Caffeine Caffeine ingestion decreases cerebral blood flow (Debrah et al. 1996) and also causes secretion of adrenaline (Nehlig et al. 1992). Ingestion of 250 mg (Debrah et al. 1996) or 400 mg (Kerr et al. 1993) of caffeine immediately before the induction of hypoglycaemia was associated with greater symptomatic awareness, and higher levels of plasma adrena- line. However, a randomised placebo-controlled double-blind study (Watson et al. 2000) with 200 mg caffeine (approximately two cups of drip-brewed coffee) for 3 months, actually resulted in an increase in the number of symptomatic hypoglycaemia episodes (1.3 vs 0.9 episodes/week, P < 0.03), and subjects experienced more intense warning symptoms (total symptom score: 29 vs 26; P < 0.05) during a subsequent hypoglycaemia clamp study. More recently, a double-blind study (Richardson et al. 2005a) reported that caffeine was associated with a significant reduction in duration of nocturnal hypoglycaemia (mean 49 vs 132 mins; P = 0.035) when assessed using CGM.

CONCLUSIONS • Hypoglycaemia is associated with multiple risk factors. • Lifestyle factors such as exercise and alcohol can further moderate hypoglycaemia risk. • Continuous glucose monitoring is a useful adjunct to capillary glucose monitoring in the diagnosis of hypoglycaemia and as an educational tool. • Treatment of acute hypoglycaemia can be regarded as a spectrum of increasing thera- peutic complexity depending on the severity of the hypoglycaemia and the clinical status of the patient. • There needs to be regular educational review in the clinic about the management of hypoglycaemia at home by patients and their relatives. 180 Hypoglycaemia in Clinical Diabetes • Prevention of hypoglycaemia focuses primarily on a review of insulin replacement regimens and a structured education approach to diabetes management, including car- bohydrate counting. • Pancreatic transplantation may be valuable for treating individuals with disabling recur- rent severe hypoglycaemia. • Pharmacological interventions or use of closed loop systems to prevent hypoglycaemia remain experimental.

REFERENCES

Amiel, S.A. (2009) Hypoglycemia: from the laboratory to the clinic. Diabetes Care, 32, 1364–1371. Avogaro, A., Beltramello, P., Gnudi, L. et al. (1993) Alcohol intake impairs glucose counterregulation during acute insulin-induced hypoglycemia in IDDM patients. Evidence for a critical role of free fatty acids. Diabetes, 42, 1626–1634. Axelsen, M., Wesslau, C., Lonnroth, P., Arvidsson Lenner, R., Smith, U. (1999) Bedtime uncooked corn- starch supplement prevents nocturnal hypoglycaemia in intensively treated type 1 diabetes subjects. Journal of Internal Medicine, 245, 229–236. Beck, R.W., Hirsch, I.B., Laffel, L., Tamborlane, W.V., Bode, B.W., Buckingham, B. et al. (2009) The effect of continuous glucose monitoring in well-controlled type 1 diabetes. Diabetes Care, 32, 1378–1383. Beverly, J.L., De Vries, M.G., Bouman, S.D., Arseneau, L.M. (2001) Noradrenergic and GABAergic systems in the medial hypothalamus are activated during hypoglycemia. American Journal of Physiology. Regulatory, Integrative and Comparative Physiology, 280, R563–R569. Bode, B., Gross, K., Rikalo, N. et al. (2004) Alarms based on real-time sensor glucose values alert patients to hypo- and hyperglycemia: the Guardian continuous monitoring system. Diabetes Technology and Therapeutics, 6, 105–113. Briscoe, V.J., Ertl, A.C., Tate, D.B., Davis, S.N. (2008) Effects of the selective serotonin reuptake inhibitor fluoxetine on counterregulatory responses to hypoglycemia in individuals with type 1 diabetes.Diabetes , 57, 3315–3322. Caplin, N.J., O’Leary, P., Bulsara, M., Davis, E.A. & Jones, T.W. (2003) Subcutaneous glucose sensor values closely parallel blood glucose during insulin-induced hypoglycaemia. Diabetic Medicine, 20, 238–241. Chan, O., Zhu, W., Ding, Y., McCrimmon, R.J. & Sherwin, R.S. (2006) Blockade of GABA(A) receptors in the ventromedial hypothalamus further stimulates glucagon and sympathoadrenal but not the hypothalamo-pituitary-adrenal response to hypoglycemia. Diabetes, 55, 1080–1087. Cheyne, E. & Kerr, D. (2002) Making ‘sense’ of diabetes: using a continuous glucose sensor in clinical practice. Diabetes/Metabolism Research and Reviews, 18 (Suppl. 1), S43–S48. Cheyne, E.H., Cavan, D.A. & Kerr, D. (2002) Performance of a continuous glucose monitoring system during controlled hypoglycemia in healthy volunteers. Diabetes Technology and Therapeutics, 4, 607–613. Cheyne, E.H., Sherwin, R.S., Lunt, M.J., Cavan, D.A., Thomas, P.W. & Kerr, D. (2004) Influence of alcohol on cognitive performance during mild hypoglycaemia; implications for type 1 diabetes. Diabetic Medi- cine, 21, 230–237. Choudhary, P., Shin, J., Wang, Y., Evans, M.L., Hammond, P.J., Kerr, D. et al. (2011) Insulin pump therapy with automated insulin suspension in response to hypoglycemia: reduction in nocturnal hypoglycemia in those at greatest risk. Diabetes Care, 34, 2023–2025. Collier, A., Steedman, D.J., Patrick, A.W. et al. (1987) Comparison of intravenous glucagon and dextrose in treatment of severe hypoglycemia in an Accident and Emergency Department. Diabetes Care, 10, 712–715. DAFNE Study Group (2002) Training in flexible, intensive insulin management to enable dietary freedom in people with type 1 diabetes: Dose Adjustment For Normal Eating (DAFNE) randomised controlled trial. British Medical Journal, 325, 746. Daneman, D., Frank, M., Perlman, K., Tamm, J. & Ehrlich, R. (1989) Severe hypoglycemia in children with insulin-dependent diabetes mellitus: frequency and predisposing factors. Journal of Pediatrics, 115, 681–685. Chapter 8: Management of Acute and Recurrent Hypoglycaemia 181

Debrah, K., Sherwin, R.S., Murphy, J. & Kerr, D. (1996) Effect of caffeine on recognition of and physi- ological responses to hypoglycaemia in insulin-dependent diabetes. Lancet, 347, 19–24. Department for Transport Road Safety Research Report No. 61 (2006) Stratifying hypoglycaemic event risk in insulin-treated diabetes. 68–69. de Galan, B.E., Tack, C.J., Lenders, J.W., Pasman, J.W., Elving, L.D., Russel, F.G. et al. (2002) Theophyl- line improves hypoglycemia unawareness in type 1 diabetes. Diabetes, 51, 790–796. de Galan, B.E., Rietjens, S.J., Tack, C.J., van der Werf, S.P., Sweep, C.G., Lenders, J.W. et al. (2003) Antecedent adrenaline attenuates the responsiveness to but not the release of counterregulatory hormones during subsequent hypoglycemia. Journal of Clinical Endocrinology and Metabolism, 88, 5462–5467. Fanelli, C., Pampanelli, S., Epifano, L. et al. (1994) Relative roles of insulin and hypoglycaemia on induc- tion of neuroendocrine responses to, symptoms of, and deterioration of cognitive function in hypogly- caemia in male and female humans. Diabetologia, 37, 797–807. Feldman, M. & Schiller, L.R. (1983) Disorders of gastrointestinal motility associated with diabetes mellitus. Annals of Internal Medicine, 98, 378–384. Frier, B.M. (2009) Defining hypoglycaemia: what level has clinical relevance? Diabetologia, 52, 31–34. Gabriely, I. & Shamoon, H. (2005) Fructose normalizes specific counterregulatory responses to hypogly- cemia in patients with type 1 diabetes. Diabetes, 54, 609–616. Gin, H., Morlat, P., Ragnaud, J.M. & Aubertin, J. (1992) Short-term effect of red wine (consumed during meals) on insulin requirement and glucose tolerance in diabetic patients. Diabetes Care, 15, 546–548. Hagstrom-Toft, E., Bolinder, J., Ungerstedt, U. & Arner, P. (1997) A circadian rhythm in lipid mobilization which is altered in IDDM. Diabetologia, 40, 1070–1078. Hoi-Hansen, T., Pedersen-Bjergaard, U. & Thorsteinsson, B. (2005) Reproducibility and reliability of hypoglycaemic episodes recorded with continuous glucose monitoring system (CGMS) in daily life. Diabetic Medicine, 7, 858–862. Hopkins, D., Lawrence, I., Mansell, P., Thompson, G., Amiel, S., Campbell, M. et al. (2012) Improved biomedical and psychological outcomes 1 year after structured education in flexible insulin therapy for people with type 1 diabetes: the U.K. DAFNE experience. Diabetes Care, 35, 1638–1642. Kendall, D.M., Teuscher, A.U., Robertson, R.P. (1997) Defective glucagon secretion during sustained hypoglycemia following successful islet allo- and autotransplantation in humans. Diabetes, 46, 23–27. Kerr, D., Macdonald, I.A., Heller, S.R. & Tattersall, R.B. (1990) Alcohol causes hypoglycaemic unaware- ness in healthy volunteers and patients with Type 1 (insulin-dependent) diabetes. Diabetologia, 33, 216–221. Kerr, D., Sherwin, R.S., Pavalkis, F. et al. (1993) Effect of caffeine on the recognition of and responses to hypoglycemia in humans. Annals of Internal Medicine, 119, 799–804. Kerr, D., Cheyne, E.H., Thomas, P. & Sherwin, R.S. (2007) Influence of acute alcohol ingestion on the hormonal responses to modest hypoglycaemia in patients with Type 1 diabetes. Diabetic Medicine, 24, 312–316. Kuhara, T., Ikeda, S., Ohneda, A. & Sasaki, Y. (1991) Effects of intravenous infusion of 17 amino acids on the secretion of GH, glucagon, and insulin in sheep. American Journal of Physiology, Endocrinology and Metabolism, 260, E21–E26. Larsen, J., Ford, T., Lyden, E., Colling, C., Mack-Shipman, L., Lane, J. (2004) What is hypoglycemia in patients with well-controlled type 1 diabetes treated by subcutaneous insulin pump with use of the continuous glucose monitoring system? Endocrine Practice, 10, 324–329. Leitao, C.B., Tharavanij, T., Cure, P., Pileggi, A., Baidal, D.A., Ricordi, C. et al. (2008) Restoration of hypoglycemia awareness after islet transplantation. Diabetes Care, 31, 2113–2115. Levin, B.E., Becker, T.C., Eiki, J., Zhang, B.B. & Dunn-Meynell, A.A. (2008) Ventromedial hypothalamic glucokinase is an important mediator of the counterregulatory response to insulin-induced hypoglycemia. Diabetes, 57, 1371–1379. Ly, T.T., Hewitt, J., Davey, R.J., Lim, E.M., Davis, E.A. & Jones, T.W. (2011) Improving epinephrine responses in hypoglycemia unawareness with real-time continuous glucose monitoring in adolescents with type 1 diabetes. Diabetes Care, 34, 50–52. MacCuish, A.C., Munro, J.F. & Duncan, L.J.P. (1970) Treatment of hypoglycaemic coma with glucagon, intravenous dextrose, and mannitol infusion in a hundred diabetics. Lancet, ii, 946–949. MacCuish, A.C. (1993) Treatment of hypoglycaemia. In: B.M. Frier & B.M. Fisher (eds), Hypoglycaemia and Diabetes: Clinical and Physiological Aspects, Edward Arnold, London, pp. 212–212.. 182 Hypoglycaemia in Clinical Diabetes

McGowan, K., Thomas, W. & Moran, A. (2002) Spurious reporting of nocturnal hypoglycemia by CGMS in patients with tightly controlled type 1 diabetes. Diabetes Care, 25, 1499–1503. Monami, M., Marchionni, N. & Mannucci, E. (2009) Long-acting insulin analogues vs. NPH human insulin in type 1 diabetes. A meta-analysis. Diabetes, Obesity & Metabolism, 11, 372–378. Moriarty, K.T., Maggs, D.G., Macdonald, I.A. & Tattersall, R.B. (1993) Does ethanol cause hypoglycaemia in overnight fasted patients with type 1 diabetes? Diabetic Medicine, 10, 61–65. Mühlhauser, I., Koch, J. & Berger, M. (1985a) Pharmacokinetics and bioavailability of injected glucagon: differences between intramuscular, subcutaneous, and intravenous administration. Diabetes Care, 8, 39–42. Mühlhauser, I., Berger, M., Sonnenberg, G. et al. (1985b) Incidence and management of severe hypogly- cemia in 434 adults with insulin-dependent diabetes mellitus. Diabetes Care, 8, 268–273. Nehlig, A., Daval, J.L. & Debry, G. (1992) Caffeine and the central nervous system: mechanisms of action, biochemical, metabolic and psychostimulant effects. Brain Research Reviews, 17, 139–170. Nilsson, A., Tideholm, B., Kalen, J. & Katzman, P. (1988) Incidence of severe hypoglycaemia and its causes in insulin-treated diabetics. Acta Medica Scandinavica, 224, 257–2. O’Grady, M.J., Retterath, A.J., Keenan, D.B., Kurtz, N., Cantwell, M., Spital, G. et al. (2012) The use of an automated, portable glucose control system for overnight glucose control in adolescents and young adults with type 1 diabetes. Diabetes Care, 35, 2182–2187. Potter, J., Clarke, P., Gale, E.A., Dave, S.H. & Tattersall, R.B. (1982) Insulin-induced hypoglycaemia in an Accident and Emergency Department: the tip of an iceberg? British Medical Journal, 285, 1180–1182. Raju, B., Arbelaez, A.M., Breckenridge, S.M., Cryer, P.E. (2006) Nocturnal hypoglycemia in type 1 dia- betes: an assessment of preventive bedtime treatments. Journal of Clinical Endocrinology and Metabo- lism, 91, 2087–2092. Rasmussen, B.M., Orskov, L., Schmitz, O. & Hermansen, K. (2001) Alcohol and glucose counterregulation during acute insulin-induced hypoglycemia in type 2 diabetic subjects. Metabolism: Clinical and Experi- mental, 50, 451–457. Richardson, T., Thomas, P., Ryder, J. & Kerr, D. (2005a) Influence of caffeine on frequency of hypoglyc- emia detected by continuous interstitial glucose monitoring system in patients with long-standing type 1 diabetes. Diabetes Care, 28, 1316–1320. Richardson, T., Weiss, M., Thomas, P. & Kerr, D. (2005b) Day after the night before: influence of evening alcohol on risk of hypoglycemia in patients with type 1 diabetes. Diabetes Care, 28, 1801–1802. Rossetti, P., Porcellati, F., Busciantella Ricci, N., Candeloro, P., Cioli, P., Nair, K.S. et al. (2008) Effect of oral amino acids on counterregulatory responses and cognitive function during insulin-induced hypogly- cemia in nondiabetic and type 1 diabetic people. Diabetes, 57, 1905–1917. Samann, A., Muhlhauser, I., Bender, R., Kloos, C. & Muller, U.A. (2005) Glycaemic control and severe hypoglycaemia following training in flexible, intensive insulin therapy to enable dietary freedom in people with type 1 diabetes: a prospective implementation study. Diabetologia, 48, 1965–1970. Smith, D., Pernet, A., Rosenthal, J.M., Bingham, E.M., Reid, H. & Macdonald, I.A. et al. (2004) The effect of modafinil on counter-regulatory and cognitive responses to hypoglycaemia. Diabetologia, 47, 1704–1711. Szepietowska, B., Zhu, W., Chan, O., Horblitt, A., Dziura, J., Sherwin, R.S. (2011) Modulation of beta- adrenergic receptors in the ventromedial hypothalamus influences counterregulatory responses to hypoglycemia. Diabetes, 60, 3154–3158. The Diabetes Control and Complications Trial Research Group (1997) Hypoglycemia in the Diabetes Control and Complications Trial. Diabetes, 46, 271–286. Trojan, N., Pavan, P., Iori, E. et al. (1999) Effect of different times of administration of a single ethanol dose on insulin action, insulin secretion and redox state. Diabetic Medicine, 16, 400–407. Turner, B.C., Jenkins, E., Kerr, D., Sherwin, R.S. & Cavan, D.A. (2001) The effect of evening alcohol consumption on next-morning glucose control in type 1 diabetes. Diabetes Care, 24, 1888–1893. UK Hypoglycaemia Study Group (2007) Risk of hypoglycaemia in types 1 and 2 diabetes: effects of treat- ment modalities and their duration. Diabetologia, 50, 1140–1147. Ververs, M.T., Rouwe, C. & Smit, G.P. (1993) Complex carbohydrates in the prevention of nocturnal hypoglycaemia in diabetic children. European Journal of Clinical Nutrition, 47, 268–273. Chapter 8: Management of Acute and Recurrent Hypoglycaemia 183

Vinik, A.I., Maser, R.E., Mitchell, B.D. & Freeman, R. (2003) Diabetic autonomic neuropathy. Diabetes Care, 26, 1553–1579. Ward, C.M., Stewart, A.W. & Cutfield, R.G. (1990) Hypoglycaemia in insulin dependent diabetic patients attending an outpatients’ clinic. New Zealand Medical Journal, 25, 339–341. Watson, J.M., Jenkins, E.J., Hamilton, P., Lunt, M.J. & Kerr, D. (2000) Influence of caffeine on the fre- quency and perception of hypoglycemia in free-living patients with type 1 diabetes. Diabetes Care, 23, 455–459. Yeh, H.C., Brown, T.T., Maruther, N. et al. (2012) Comparative effectiveness and safety of methods of insulin delivery and glucose monitoring for diabetes mellitus: a systematic review and meta-analysis. Annals of Internal Medicine, 157, 336–347. 9 Technology for Hypoglycaemia: CSII and CGM Josefine E. Schopman and J. Hans DeVries Academic Medical Center, Amsterdam, The Netherlands

INTRODUCTION

This chapter describes advances in technology that have the potential to prevent or limit exposure to hypoglycaemia, including continuous subcutaneous insulin infusion (CSII) and continuous glucose monitoring (CGM) systems. An overview of their impact on the occur- rence of hypoglycaemic events compared to standard treatment in people with type 1 diabetes is given. The reasons why CGM is apparently unable to prevent severe hypogly- caemia are discussed. This is followed by a section on cost and reimbursement status and ends with future expectations.

INSULIN PUMP THERAPY

Insulin pump therapy or CSII emerged in the late 1970s. It is now a common and accepted way of administering insulin. An insulin pump continuously infuses rapid-acting insulin subcutaneously at variable rates. In addition, the user can administer additional doses at mealtimes to cover ingestion of food or at any time to correct a high glucose concentration. The difference between insulin delivered by CSII and multiple daily injections (MDI) lies in the ability of the pump to dif- ferentiate the supply of basal insulin throughout the full 24 hours of the day. This better mimics the normal physiological situation than giving intermittent injections of long-acting insulin. Also, insulin delivery can be decreased or, in the event of unanticipated exercise, even temporarily stopped – a situation that can be dealt with only by ingesting food when using MDI treatment. The most important indication for CSII is the desire for better glycae- mic control in those with suboptimal glucose levels. The patient’s motivation, educational level and interaction with the treatment team are important determinants of success. Adverse events that may arise with CSII are pump malfunction, infection at the injection site, irrita- tion or discomfort (Plotnick et al. 2003; Richardson and Kerr 2003; Guilhem et al. 2006). Meta-analyses of studies comparing CSII and MDI treatment have demonstrated improved glycaemic control with the use of CSII in patients with type 1 diabetes mellitus (Pickup et al. 2002; Retnakaran et al. 2004). In addition, they have also shown that insulin

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. Chapter 9: Technology for Hypoglycaemia: CSII and CGM 185 pump therapy is associated with a lower risk of severe hypoglycaemia, with a rate ratio of CSII compared with MDI of 2.89 (95% CI, 1.45 to 5.76) for randomised controlled trials. The greatest risk reduction was observed in those whose initial rates of severe hypogly- caemia were highest on MDI (Pickup and Sutton 2008). Another meta-analysis showed an odds ratio for severe hypoglycaemia of 0.48 in favour of CSII (Fatourechi et al. 2009).

MONITORING

The current recommendations of the American Diabetes Association suggest that type 1 diabetes patients should undertake home blood glucose monitoring (HBGM) three or more times daily (American Diabetes Association 2011). This provides an incomplete picture of blood glucose fluctuations. In addition, finger stick measurements are painful and regular testing requires great devotion and commitment from the participants. It is estimated that 46% of people with insulin-treated diabetes self-monitor their blood glucose less than three times each day (Karter et al. 2001).

Continuous glucose monitoring CGM devices date from the 1990s. CGM systems measure glucose via the glucose-oxidase reaction in interstitial fluid and translate this via a calibration into a value resembling blood glucose. This semi-continuous information identifies fluctuations that would not be identi- fied with self-monitoring. CGM is considered to be particularly useful for children, people with poorly-controlled diabetes, women during pregnancy and those with impaired aware- ness of hypoglycaemia. Currently, the use of CGM is not common practice, and reimburse- ment in Europe is patchy (see section on Cost and Reimbursement). Most CGM systems use a small needle type sensor inserted in the subcutaneous adipose tissue. Two types of CGM systems can be discriminated:

• systems that measure the glucose concentration over a number of days: the information is stored in a monitor and can be downloaded later; • real-time systems that continuously provide the actual glucose concentration on a display for longer periods of time.

There are four real-time CGM systems, which are commercially available that display glucose values every 1–5 minutes and feature an alarm function for hypo- and hyperglycaemia:

• the Freestyle Navigator (Abbot Diabetes Care, Alameda, CA, USA); • the Guardian Real-Time (Medtronic MiniMed, Northridge, CA, USA); • the Dexcom SEVEN (Dexcom, San Diego, CA, USA); • the GlucoDay (Menarini Diagnostics, Florence, Italy).

Although HbA1c is lowered more when real-time CGM is used continuously, it is often used intermittently (e.g. a couple of days per month or at intervals of 5–7 days) to restrain costs. The current evidence of CGM for the prevention of severe, mild and nocturnal hypoglycaemia and hypoglycaemia prevention in patients with impaired awareness is discussed. 186 Hypoglycaemia in Clinical Diabetes

Severe hypoglycaemia Early expectations were that CGM would prevent severe hypoglycaemia, but evidence to this effect is still lacking (Table 9.1). In the STAR-1 Trial, there were significantly more severe hypoglycaemic events in the CGM arm than in the control arm (Hirsch et al. 2008). Meta-analysis of CGM intervention trials in type 1 diabetes for the occurrence of severe hypoglycaemia shows a non-significant odds ratio of 1.37 (95% CI; 0.83, 2.25) with CGM use compared to HBGM in a random effects model (Figure 9.1). There are at least six possible explanations for the inability of CGM to prevent severe hypoglycaemia:

• improved glycaemic control; • long learning phase; • CGM inaccuracy; • inadequate response to hypoglycaemic alarm; • intolerance of the CGM device; • insufficient study design.

These are discussed in more detail below.

Improved glycaemic control

In a recent meta-analysis an overall difference of 0.30% (3 mmol/mol) in HbA1c was shown in favour of CGM compared to HBGM, and that was statistically significant (Pickup et al. 2011). The improvement in glycaemic control resulting from CGM use may lead to an increase in the incidence of hypoglycaemia. This might explain why the incidence of hypoglycaemia remains unchanged or is perhaps slightly increasing with CGM use (see the meta-analysis in Figure 9.1: 73 vs 56 severe hypoglycaemic events with CGM and HBGM, respectively).

Long learning phase In an observational follow-up study from the Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group (2008), extending CGM use from 6 to 12 months was associated with a HbA1c that remained within the target range combined with a rate of severe hypoglycaemia that was lowered by almost 70% compared with the former 6-month monitoring period (Bode et al. 2009). This suggests that a longer user experience is required to learn to avoid severe hypoglycaemia with CGM. However, this is inconsistent with an observational study in 16 subjects with type 1 diabetes on MDI treatment (Ryan and Germsheid 2009). All of the subjects included had experienced at least one episode of severe hypoglycaemia in the previous year. After the baseline (HBGM) month they had one month to get used to the CGM device, thereafter the study month (CGM) followed. The number of severe hypoglycaemic episodes dropped from 16 in the month with HBGM to 3 in the month when CGM was used, P = 0.064. The overall number of hypoglycaemic episodes (<3.0 mmol/l) declined from 8.6 to 4.7 (P = 0.01). Participants expressed less fear of hypoglycaemia when they used CGM. Therefore it seems sensible to investigate the value of CGM in preventing severe hypoglycaemia in patients at high risk. Table 9.1 CGM intervention trials in type 1 diabetes

Pre-study Severe Inclusion Treatment hypoglycaemia

Study Population HbA1c (n) Duration Comparison Drop out Outcome HbA1c (n)

Battelino et al. 2011 120 adults and <7.5% CSII (81); 26 weeks 1) CGM; 1) n = 10; 1 vs 2: −0.4%, P < 0.008 1) 0; children MDI (39) 2) HBGM 2) n = 9 2) 0 Bergenstal et al. 2010 156 children 7.4–9.5% CSII (0); 1 year 1) CGM and 1) n = 19; adults 1 vs 2: −0.6%, P < 0.001; 1) 32; (STAR-3 Trial) 329 adults MDI (385) CSII; 2) n = 13 children 1 vs 2: −0.5%, P < 0.001 2) 27 2) HBGM Deiss et al. 2006 81 children ≥8.1% CSII (78); 3 months 1) CGM 1) n = 4; 1 vs 3: −0.6%, P = 0.003 1) 1; 81 adults MDI (84) continuously; 2) n = 1; 2 vs 3: NR, P = NS 2) 1; 2) CGM 3) n = 0 1 vs 2: NR, P = NS 3) 0 biweekly; 3) HBGM Hermanides et al. 2011 83 adults ≥8.2% CSII (0); 6 months 1) CGM and 1) n = 1; 1 vs 2: −1.2%, P < 0.001 1) 4; (Eurythmics Trial) MDI (83) CSII; 2) n = 4 2) 1 2) HBGM Hirsch et al. 2008 146 adults and ≥7.5% CSII (146); 6 months 1) CGM; 1) n = 6; 1 vs 2: −0.11, P = 0.37 1) 11*; (STAR-1 Trial) children MDI (0) 2) HBGM 2) n = 2 2) 3 Juvenile Diabetes Research 114 children 7.0–10% CSII (256); 6 months 1) CGM; 1) n = 3; adults 1 vs 2: −0.5%, P < 0.001; 1) 14; Foundation Continuous 110 adolescents MDI (66) 2) HBGM 2) n = 2 adolescents 1 vs 2: 0.08%, 2) 11 Glucose Monitoring 98 adults P = 0.52; Study Group 2008 children 1 vs 2: −0.1%, P = 0.29 Juvenile Diabetes Research 29 children <7.0% CSII (111); 6 months 1) CGM; 1) n = 1; 1 vs 2: −0.3%, P < 0.001 1) 9; Foundation Continuous 33 adolescents MDI (18) 2) HBGM 2) n = 1 2) 10 Glucose Monitoring 67 adults Study Group 2009 Kordonouri et al. 2010 160 children ≥8.0% CSII (160); 12 months 1) CGM and 1) n = 4; 1 vs 2: −0.2%, P = NS 1) 0; MDI (0) CSII; 2) n = 2 2) 4 2) HBGM O’Connell et al. 2009 62 children and ≤8.5% CSII (62); 3 months 1) CGM; 1) n = 5; 1 vs 2: −0.4%, P = 0.009 1) 0; adults MDI (0) 2) HBGM 2) n = 2 2) 0 Raccah et al. 2009 51 children ≥8.0% CSII (0); 6 months 1) CGM and 1) n = 14; 1 vs 2: −0.2%, P = 0.09; 1) 1†; (RealTrend Study) 77 adults MDI (128) CSII; 2) n = 6 Per protocol analyses: 2) 0 2) HBGM and 1 vs 2: −0.4%, P = 0.004 CSII

*P < 0.05. †Number of total severe hypoglycaemic episodes per group not given, only episode with seizure/coma. NR, not reported. NS, not significant, CGM, continuous glucose monitoring HBGM, home blood glucose monitoring, CSII, continuous subcutaneous insulin infusion, MDI, multiple daily injections 188 Hypoglycaemia in Clinical Diabetes

CGM HBGM Odds Ratio Odds Ratio Study SH Total TotalSH Weight M-H, Random, 95% CI M-H, Random, 95% CI Battelino 2011 0 62 0 58 Not estimatable Bergenstal 2010 32 247 27 248 35.3% 1.22 [0.71, 2.10] Deis 2006 2 103 0 53 2.5% 2.64 [0.12, 55.89] Hermanides 2011 4 43 1 35 4.6% 3.49 [0.37, 32.72] Hirsch 2008 11 66 3 72 11.4% 4.60 [1.22, 17.30] JDRF CGM Study Group 2008 14 162 11 155 22.9% 1.24 [0.54, 2.82] JDRF CGM Study Group 2009 9 67 10 62 18.2% 0.81 [0.30, 2.14] Kordonouri 2010 0 76 4 78 2.7% 0.11 [0.01, 2.04] O'Connell 2009 0 29 0 29 Not estimable Raccah 2009 1 30 0 53 2.3% 5.44 [0.21, 137.82]

Total (95% CI) 885 843 100.0% 1.37 [0.83, 2.25] Total SH 73 56 Heterogeneity: Tau2 = 0.10; Chi2 = 8.93, df = 7 (P = 0.26); I2 = 22% 0.002 0.1 1 10 500 Test for overall effect: Z = 1.24 (P = 0.22) Favours CGM Favours HBGM

Figure 9.1 Forest plot of the incidence of severe hypoglycaemia in CGM intervention trials. CGM: continuous glucose monitoring; HBGM: home blood glucose monitoring; SH: severe hypoglycaemia; Total: number of subjects.

CGM inaccuracy The accuracy of CGM remains the subject of debate. There are some technical issues to consider:

• lag between interstitial glucose and blood glucose; • instrumental delay; • necessity of finger stick blood testing for calibration purposes; • potential error in recording values at the limit of the detection range.

Interstitial glucose values are partially determined by the rate of glucose diffusion from plasma to the interstitial fluid and the rate of glucose uptake by subcutaneous tissue cells, but it is also influenced by many other factors (Koschinsky and Heinemann 2001). The relationship between blood and interstitial glucose is not well understood. The physiological delay that occurs in the equilibration of blood with interstitial tissue glucose may be increased when blood glucose falls rapidly (Rebrin and Steil 2000). During experimental hypoglycaemia in people with type 1 diabetes, CGM underestimated plasma glucose concentrations and thus potentially overestimated the frequency of hypoglycaemia (Caplin et al. 2003). Interestingly, comparison of glucose values detected by CGM and HBGM in a crossover study revealed that in the hypoglycaemic range, CGM values over- estimated blood glucose levels by an average of 0.83 mmol/l (Davey et al. 2010). Superimposed on the physiological delay in the equilibration of blood glucose and interstitial glucose, if any (Wentholt et al. 2007a), there is also an instrumental delay, inher- ent to the current real-time CGM systems. Both are probably contributors to the inaccuracy of the CGM devices. With the current CGM devices it is essential to perform frequent calibration using capil- lary blood samples. Recordings by the CGM are markedly influenced when this is not done correctly. When compared with actual blood glucose values, CGM devices have an inac- curacy of approximately 15% (expressed as mean absolute difference, MAD [sensor value – blood glucose / blood glucose]). The performance of these devices is worse during hypoglycaemia, when inaccuracy may increase up to 25%. This indicates that the accuracy Chapter 9: Technology for Hypoglycaemia: CSII and CGM 189 of CGM requires improvement, especially in hypoglycaemia, which is the clinically most important range for adequate sensor performance (Wentholt et al. 2006).

Inadequate response to hypoglycaemic alarm When the sensor alerts the patient to a hypoglycaemic event it is not guaranteed that the patient is warned timely enough to take appropriate action before neuroglycopenia has occurred. Cognitive function declines during severe hypoglycaemia leading to a less ade- quate response to acoustic or vibration alarms (Cryer et al. 2003). During the night, people may not hear alarms when they are asleep, especially when the device is covered with blan- kets. It is reported that patients sleep through half of all alarms (Buckingham et al. 2005).

Intolerance of the CGM device There are many patients who do not tolerate CGM devices. This is illustrated by the higher drop-out rates in the CGM arms of the randomised controlled trials (RCTs) (Table 9.1). In many trials, patients were already exposed to (blinded) CGM systems before inclusion or randomisation to obtain a baseline CGM measurement for all (Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group 2008; Raccah et al. 2009; Her- manides et al. 2011). This led to a premature drop-out of 4 of 87, 23 of 345 and 4 of 132 patients before randomisation in the Eurythmics, JDRF trial and RealTrend study, respec- tively, probably of patients that did not tolerate the device. Questionnaire studies also support this premise. A questionnaire study to determine barriers to CGM showed that 27 of 54 subjects stopped using CGM (Halford and Harris 2010). A third reported adhesive issues with the device and that the device was too ponderous. Another study identified perceived barriers to CGM in about 30% of 624 responders (Tansey et al. 2011). Commonly reported problems were insertion pain, (false) system alarms, problems with insertion sites and the size of the transmitter or receiver. Hirsch et al. (2008), as mentioned above, reported 11 severe hypoglycaemic events in the CGM group and three in the HBGM group (Table 9.1). Six of eleven events occurred when patients were not wearing, or not using, the CGM device. The remaining five events occurred during use of the device. The following was apparent: subjects ignored alarms that warned of low sensor readings and participants based their treatment decisions on the sensor readings alone, without confirming the capillary blood glucose with a measurement.

Insufficient study design It is known that the incidence of severe hypoglycaemia is skewed (Pedersen-Bjergaard et al. 2004), meaning that most patients experience no severe hypoglycaemia, while others experience very high numbers of severe hypoglycaemic episodes. By excluding the patients with previous severe hypoglycaemic episodes, the incidence of severe hypoglycaemia may have been artificially low in some CGM trials (O’Connell et al. 2009; Bergenstal et al. 2010). The studies performed so far might have been insufficiently powered to detect a difference in hypoglycaemia rates between treatment groups. Table 9.1 largely contains trials designed for patients that used CSII before randomisa- tion or who were switched to CSII during the trial. Consequently, these data have to be interpreted with care when the results of patients using CGM are extrapolated to those using it in combination with MDI treatment. 190 Hypoglycaemia in Clinical Diabetes

CGM HBGM Mean Difference Mean Difference Study Mean SD Total MeanSD Total Weight IV, Random, 95% CI IV, Random, 95% CI Battelino 2011 2 1.7 62 4 3.16 58 47.3% –2.00 [–2.92, –1.08] Hermanides 2011 2.7 3.4 43 2.5 3.6 35 38.2% 0.20 [–1.37, 1.77] O'Connell 2009 9.2 8.7 29 9.1 6.9 29 14.4% 0.10 [–3.94, 4.14]

Total (95% CI) 134 122 100.0% –0.86 [–2.65, 0.94] 2 2 2 Heterogeneity: Tau = 1.55; Chi = 6.18, df = 2 (P = 0.05); I = 68% –4 –2 0 2 4 Test for overall effect: Z = 0.94 (P = 0.35) Favours CGM Favours HBGM

Figure 9.2 Forest plot of time (%) spent in the hypoglycaemic range in CGM intervention trials. CGM: continuous glucose monitoring; HBGM: home blood glucose monitoring; Total: number of subjects.

Mild hypoglycaemia Mild hypoglycaemia detected with CGM is often represented as ‘time spent in the hypogly- caemic range’ or as ‘area under the curve’, which makes comparison with hypoglycaemia detected by HBGM difficult. Frequently it is not clear whether the reported episodes of hypoglycaemia were symptomatic or asymptomatic, and only a few studies report whether patients have impaired hypoglycaemia awareness. Many CGM intervention studies do not provide enough detail to allow inclusion in a meta-analysis. To date it appears that there is no reduction in time spent in the hypogly- caemic range with CGM (Figure 9.2), although one trial reported a reduction in people with good glycaemic control (Battelino et al. 2011).

Nocturnal hypoglycaemia Nocturnal hypoglycaemia is a common problem for people with type 1 diabetes (see Chapter 5). It is estimated that up to 75% of hypoglycaemic episodes associated with coma or seizures occur at night when the counterregulatory response is impaired (The DCCT Research Group 1991; Bolli et al. 1993; Davis et al. 1997; Matyka et al. 1999). The reported prevalence of nocturnal hypoglycaemia has been studied previously with HBGM in outpatient and inpatient settings and ranges from 10 to 56% (Shalwitz et al. 1990; Beregszaszi et al. 1997; Kaufman et al. 2002; Bober and Buyukgebiz 2005; Raju et al. 2006). It is believed that the real prevalence of nocturnal hypoglycaemia might be higher, because testing with HBGM is intermittent and is dependent on the initiative of the patient. The introduction of CGM provided a method of measuring glucose levels every few minutes and led to increased monitoring. The prevalence of nocturnal hypoglycaemia as assessed using CGM devices is up to 68% (Boland et al. 2001; Sachedina and Pickup 2003; Yates et al. 2006; Guillod et al. 2007; Wentholt et al. 2007b; Ahmet et al. 2011). However, the higher prevalence based on the utilisation of CGM has been debated. The CGM devices may have a tendency to report lower glucose values at night (McGowan et al. 2002; Wentholt et al. 2005). CGM as a preventative measure for nocturnal hypoglycaemia is desirable. Until now the evidence supporting the usefulness of CGM in the detection and treatment of nocturnal hypoglycaemia is very sparse. Garg et al. (2006) studied 91 insulin-dependent patients with type 1 (n = 75) and type 2 (n = 16) diabetes, all of whom used a CGM device. The subjects were assigned to a ‘display’ group, who could see the values, and a control group in whom Chapter 9: Technology for Hypoglycaemia: CSII and CGM 191

CGM data were not visible. Nocturnal hypoglycaemia was 38% lower in the display group compared with the control group. The role of CGM to minimise nocturnal hypoglycaemia has yet to be established.

Impaired awareness of hypoglycaemia Recurrent episodes of mild hypoglycaemia, even in the absence of symptoms, are not benign as they can induce defects in counterregulatory hormonal responses and modifica- tion of warning symptoms, thereby increasing the risk of severe hypoglycaemia (Davis et al. 1997). It is assumed that if episodes of asymptomatic hypoglycaemia are detected and treated, severe hypoglycaemia will be prevented. In theory those with impaired aware- ness of hypoglycaemia could benefit from continuous glucose monitoring. Clinical practice recommendations suggest that CGM is useful in patients with impaired awareness of hypoglycaemia and/or frequent episodes of hypoglycaemia (American Diabetes Associa- tion 2011). However, a hypoglycaemia preventive effect of CGM has not been established, with no randomised controlled trials so far with this specific group of patients. One short-term study included a substantial proportion (43%) of patients with impaired awareness of hypoglycaemia (Bode et al. 2004). The subjects were randomised into groups either with real-time CGM or a blinded sensor. The group with real-time CGM demon- strated a significant decrease in duration of hypoglycaemic excursions compared tothe control group with a blinded sensor, but this was accompanied with an increase in hyper- glycaemic excursions. A few other studies provide more in-depth information about impaired awareness of hypoglycaemia with the use of CGM devices. Kubiak et al. (2004) investigated hypogly- caemia awareness using CGM over 72 hours. Patients with impaired awareness of hypogly- caemia showed a significantly higher total number of hypoglycaemic episodes, number of undetected hypoglycaemic episodes, and mean glucose levels. However, these results were not confirmed by Choudharyet al. (2010). The British group performed a post hoc analysis of data collected as part of the survey by UK Hypoglycaemia Study Group (2007), and after the 5 days of CGM utilisation no significant differences were observed either in fre- quency, duration or severity of hypoglycaemia between those with normal and those with impaired awareness of hypoglycaemia. When hypoglycaemia exposure was assessed pro- spectively over 9–12 months using weekly 4-point HBGM, the subjects with impaired awareness had a significantly higher risk both for mild and severe hypoglycaemia. The latter has been reported in previous studies (Gold et al. 1994; Clarke et al. 1995). The dif- ference between the two studies might be that Kubiak et al. selected patients with severely impaired awareness of hypoglycaemia based on patient interviews and a history of severe hypoglycaemia, while Choudhary et al. had used a simple question to elicit their state of awareness (Gold et al. 1994). It is possible therefore that the two study cohorts may have differed in their degree of severity or their prevalence of impaired awareness of hypoglycaemia. Ly et al. (2011) used real-time CGM for 4 weeks in six subjects to prevent hypoglycae- mic episodes compared to five subjects with standard treatment. All patients had impaired awareness of hypoglycaemia and underwent hyperinsulinaemic hypoglycaemic clamp studies at baseline and after 4 weeks. The group with real-time CGM reported a greater percentage change and peak epinephrine response to induced hypoglycaemia than the subjects in the control group. In addition, the subjects in the CGM group reported higher autonomic symptom scores. No change occurred in the glucagon, cortisol and growth 192 Hypoglycaemia in Clinical Diabetes hormone responses during hypoglycaemia for both groups. The greater epinephrine response and partial restoration of autonomic symptoms suggest that real-time CGM might be useful for the restoration of the symptomatic and counterregulatory response to hypoglycaemia in people with impaired awareness of hypoglycaemia. Further research is required to elu- cidate this.

COST AND REIMBURSEMENT

Insulin pump therapy The main cost of CSII is for consumables, such as tubing and cannulae – about €4000 per year. The cost of the pump, assuming a 4-year life, adds another €950 per annum. It is assumed that insulin pump therapy compared to MDI is cost-effective in adults and children with type 1 diabetes in Europe and the USA (Scuffham and Carr 2003; Roze et al. 2005; St Charles et al. 2009). This is calculated on the basis of an improvement in glycaemic control, fewer problems with hypoglycaemia and better quality of life. Insulin pump therapy is reimbursed in most European countries and in the USA.

Continuous glucose monitoring Treatment with CGM costs about €3800 per person per year (based on use of the device 80% of the time) compared with €1625 for HBGM (based on 4-times daily testing). It can be assumed that CGM would be cost-effective in people with poorly controlled type 1 diabetes because of the gain in long-term health benefits, secondary to lowering HbA1c. However, the cost-effectiveness of CGM in other patient groups or in preventing hypogly- caemia is hard to assess because of the lack of evidence. This is only partially reflected in the current reimbursement status of CGM devices. In Europe, real-time CGM is only reimbursed in Estonia, Ireland, the Netherlands, Slovakia, Slovenia, Sweden and Switzerland. CSII-using patients in Sweden with two or more severe hypoglycaemic episodes per year, patients with HbA1c > 10% while receiving intensive insulin therapy and children who require at least 10 HBGM tests per day are eligible for reimbursement. If CGM is not having its desired effect after 3 months, it should be discontinued. Reimbursement in Slovenia is possible for children with type 1 diabetes until the age of 7, patients with impaired hypoglycaemia awareness with a history of severe hypoglycaemia, and pregnant patients with either type 1 or type 2 diabetes on intensive insulin treatment. Selected hospitals in the Netherlands get a specific amount reimbursed from third-party payers for the treatment of each patient with diabetes. Physicians then determine which patient gets a CGM device. It is likely that only a limited number of patients will be selected. In the USA, health plans reimburse real-time CGM for people with type 1 diabetes who are not meeting the American Diabetes Association HbA1c targets or who experience severe hypoglycaemic events. In Israel, real-time CGM is included in the National Health Basket and is reimbursed by the Sickness Funds. Children (aged 6–18 years) with type 1 diabetes and severely impaired hypoglycaemia awareness, who have experienced two severe episodes of hypogly- caemia in the past 12 months (requiring ambulance assistance or emergency ward treat- ment), can apply for reimbursement. Chapter 9: Technology for Hypoglycaemia: CSII and CGM 193

FUTURE EXPECTATIONS

Current developments include overnight and hybrid closed loop (artificial pancreas) insulin delivery systems, dual hormone delivery and other modifications to existing therapies that offer the potential to reduce the risk of hypoglycaemia for people with diabetes. Two groups have reported that closed-loop compared to open-loop treatment may reduce risks of noc- turnal hypoglycaemia (Hovorka et al. 2010; Kovatchev et al. 2010). Whether continuous glucose monitoring is useful in patients with impaired awareness of hypoglycaemia and/or frequent episodes of hypoglycaemia need to be elucidated. One such multicentre trial is underway.

CONCLUSIONS • CSII is effective in reducing the frequency of severe hypoglycaemia. • CGM lowers HbA1c without an increase in the incidence of severe hypoglycaemic episodes. • Against expectations, CGM has not been able to lower the incidence of severe hypoglycaemia. • CGM may be a supplemental tool to home blood glucose monitoring in those with impaired awareness of hypoglycaemia and/or frequent severe hypoglycaemic episodes, but this needs confirmation in trials targeting this population.

REFERENCES

Ahmet, A., Dagenais, S., Barrowman, N.J., Collins, C.J. & Lawson, M.L. (2011) Prevalence of nocturnal hypoglycemia in pediatric type 1 diabetes: a pilot study using continuous glucose monitoring. Journal of Pediatrics, 159, 297–302. American Diabetes Association (2011) Standards of medical care in diabetes. Diabetes Care, 34 (Suppl. 1), S11–S61. Battelino, T., Phillip, M., Bratina, N., Nimri, R., Oskarsson, P. & Bolinder, J. (2011) Effect of continuous glucose monitoring on hypoglycemia in type 1 diabetes. Diabetes Care, 34, 795–800. Beregszaszi, M., Tubiana-Rufi, N., Benali, K., Noel, M., Bloch, J. & Czernichow, P. (1997) Nocturnal hypoglycemia in children and adolescents with insulin-dependent diabetes mellitus: prevalence and risk factors. Journal of Pediatrics, 131, 27–33. Bergenstal, R.M., Tamborlane, W.V., Ahmann, A. et al. (2010) Effectiveness of sensor-augmented insulin- pump therapy in type 1 diabetes. New England Journal of Medicine, 363, 311–320. Bober, E. & Buyukgebiz, A. (2005) Hypoglycemia and its effects on the brain in children with type 1 diabetes mellitus. Pediatric Endocrinological Reviews, 2, 378–382. Bode, B., Gross, K., Rikalo, N. et al. (2004) Alarms based on real-time sensor glucose values alert patients to hypo- and hyperglycemia: the Guardian Continuous Monitoing System. Diabetes Technology and Therapeutics, 6, 105–113. Bode, B., Beck, R.W., Xing, D. et al. (2009) Sustained benefit of continuous glucose monitoring on A1C, glucose profiles, and hypoglycemia in adults with type 1 diabetes. Diabetes Care, 32, 2047–2049. Boland, E., Monsod, T., Delucia, M., Brandt, C.A., Fernando, S. & Tamborlane, W.V. (2001) Limitations of conventional methods of self-monitoring of blood glucose: lessons learned from 3 days of continuous glucose sensing in pediatric patients with type 1 diabetes. Diabetes Care, 24, 1858–1862. Bolli, G.B., Perriello, G., Fanelli, C.G. & De Feo, P. (1993) Nocturnal blood glucose control in type I diabetes mellitus. Diabetes Care, 16 (Suppl. 3), 71–89. Buckingham, B., Block, J., Burdick, J. et al. (2005) Response to nocturnal alarms using a real-time glucose sensor. Diabetes Technology and Therapeutics, 7, 440–447. 194 Hypoglycaemia in Clinical Diabetes

Caplin, N.J., O’Leary, P., Bulsara, M., Davis, E.A. & Jones, T.W. (2003) Subcutaneous glucose sensor values closely parallel blood glucose during insulin-induced hypoglycaemia. Diabetic Medicine, 20, 238–241. Choudhary, P., Geddes, J., Freeman, J.V., Emery, C.J., Heller, S.R. & Frier, B.M. (2010) Frequency of biochemical hypoglycaemia in adults with Type 1 diabetes with and without impaired awareness of hypoglycaemia: no identifiable differences using continuous glucose monitoring.Diabetic Medicine, 27, 666–672. Clarke, W.L., Cox, D.J., Gonder-Frederick, L.A., Julian, D., Schlundt, D. & Polonsky, W. (1995) Reduced awareness of hypoglycemia in adults with IDDM. A prospective study of hypoglycemic frequency and associated symptoms. Diabetes Care, 18, 517–522. Cryer, P.E., Davis, S.N. & Shamoon, H. (2003) Hypoglycemia in diabetes. Diabetes Care, 26, 1902– 1912. Davey, R.J., Jones, T.W. & Fournier, P.A. (2010) Effect of short-term use of a continuous glucose monitor- ing system with a real-time glucose display and a low glucose alarm on incidence and duration of hypoglycemia in a home setting in type 1 diabetes mellitus. Journal of Diabetes Science and Technology, 4, 1457–1464. Davis, E.A., Keating, B., Byrne, G.C., Russell, M. & Jones, T.W. (1997) Hypoglycemia: incidence and clinical predictors in a large population-based sample of children and adolescents with IDDM. Diabetes Care, 20, 22–25. Davis, S.N., Shavers, C., Mosqueda-Garcia, R. & Costa, F. (1997) Effects of differing antecedent hypogly- cemia on subsequent counterregulation in normal humans. Diabetes, 46, 1328–1335. Deiss, D., Bolinder, J., Riveline, J.P. et al. (2006) Improved glycemic control in poorly controlled patients with type 1 diabetes using real-time continuous glucose monitoring. Diabetes Care, 29, 2730– 2732. Fatourechi, M.M., Kudva, Y.C., Murad, M.H., Elamin, M.B., Tabini, C.C. & Montori, V.M. (2009) Clinical review: hypoglycemia with intensive insulin therapy: a systematic review and meta-analyses of rand- omized trials of continuous subcutaneous insulin infusion versus multiple daily injections. Journal of Clinical Endocrinology and Metabolism, 94, 729–740. Garg, S., Zisser, H., Schwartz, S. et al. (2006) Improvement in glycemic excursions with a transcutaneous, real-time continuous glucose sensor: a randomized controlled trial. Diabetes Care, 29, 44–50. Gold, A.E., MacLeod, K.M. & Frier, B.M. (1994) Frequency of severe hypoglycemia in patients with type 1 diabetes with impaired awareness of hypoglycemia. Diabetes Care, 17, 697–703. Guilhem, I., Leguerrier, A.M., Lecordier, F., Poirier, J.Y. & Maugendre, D. (2006) Technical risks with subcutaneous insulin infusion. Diabetes & Metabolism, 32, 279–284. Guillod, L., Comte-Perret, S., Monbaron, D., Gaillard, R.C. & Ruiz, J. (2007) Nocturnal hypoglycaemias in type 1 diabetic patients: what can we learn with continuous glucose monitoring? Diabetes & Metabo- lism, 33, 360–365. Halford, J. & Harris, C. (2010) Determining clinical and psychological benefits and barriers with continu- ous glucose monitoring therapy. Diabetes Technology and Therapeutics, 12, 201–205. Hermanides, J., Norgaard, K., Bruttomesso, D. et al. (2011) Sensor-augmented pump therapy lowers HbA1c in suboptimally controlled Type 1 diabetes; a randomized controlled trial. Diabetic Medicine, 28, 1158–1167. Hirsch, I.B., Abelseth, J., Bode, B.W. et al. (2008) Sensor-augmented insulin pump therapy: results of the first randomized treat-to-target study. Diabetes Technology and Therapeutics, 10, 377–383. Hovorka, R., Allen, J.M., Elleri, D. et al. (2010) Manual closed-loop insulin delivery in children and adolescents with type 1 diabetes: a phase 2 randomised crossover trial. Lancet, 375 (9716), 743–751. Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group (2008) Continuous glucose monitoring and intensive treatment of type 1 diabetes. New England Journal of Medicine, 359, 1464–1476. Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group (2009) The effect of continuous glucose monitoring in well-controlled type 1 diabetes. Diabetes Care, 32, 1378–1383. Karter, A.J., Ackerson, L.M., Darbinian, J.A. et al. (2001) Self-monitoring of blood glucose levels and glycemic control: the Northern California Kaiser Permanente Diabetes registry. American Journal of Medicine, 111, 1–9. Kaufman, F.R., Austin, J., Neinstein, A. et al. (2002) Nocturnal hypoglycemia detected with the Continuous Glucose Monitoring System in pediatric patients with type 1 diabetes. Journal of Pediatrics, 141, 625–630. Chapter 9: Technology for Hypoglycaemia: CSII and CGM 195

Kordonouri, O., Pankowska, E., Rami, B. et al. (2010) Sensor-augmented pump therapy from the diagnosis of childhood type 1 diabetes: results of the Paediatric Onset Study (ONSET) after 12 months of treat- ment. Diabetologia, 53, 2487–2495. Koschinsky, T. & Heinemann, L. (2001) Sensors for glucose monitoring: technical and clinical aspects. Diabetes/Metabolism Research and Reviews, 17, 113–123. Kovatchev, B., Cobelli, C., Renard, E. et al. (2010) Multinational study of subcutaneous model-predictive closed-loop control in type 1 diabetes mellitus: summary of the results. Journal of Diabetes Science and Technology, 4, 1374–1381. Kubiak, T., Hermanns, N., Schreckling, H.J., Kulzer, B. & Haak, T. (2004) Assessment of hypoglycaemia awareness using continuous glucose monitoring. Diabetic Medicine, 21, 487–490. Ly, T.T., Hewitt, J., Davey, R.J., Lim, E.M., Davis, E.A. & Jones, T.W. (2011) Improving epinephrine responses in hypoglycemia unawareness with real-time continuous glucose monitoring in adolescents with type 1 diabetes. Diabetes Care, 34, 50–52. Matyka, K.A., Crowne, E.C., Havel, P.J., Macdonald, I.A., Matthews, D. & Dunger, D.B. (1999) Counter- regulation during spontaneous nocturnal hypoglycemia in prepubertal children with type 1 diabetes. Diabetes Care, 22, 1144–1150. McGowan, K., Thomas, W. & Moran, A. (2002) Spurious reporting of nocturnal hypoglycemia by CGMS in patients with tightly controlled type 1 diabetes. Diabetes Care, 25, 1499–1503. O’Connell, M.A., Donath, S., O’Neal, D.N. et al. (2009) Glycaemic impact of patient-led use of sensor- guided pump therapy in type 1 diabetes: a randomised controlled trial. Diabetologia, 52, 1250–1257. Pedersen-Bjergaard, U., Pramming, S., Heller, S.R. et al. (2004) Severe hypoglycaemia in 1076 adult patients with type 1 diabetes: influence of risk markers and selection. Diabetes/Metabolism Research and Reviews, 20, 479–486. Pickup, J.C. & Sutton, A.J. (2008) Severe hypoglycaemia and glycaemic control in Type 1 diabetes: meta- analysis of multiple daily insulin injections compared with continuous subcutaneous insulin infusion. Diabetic Medicine, 25, 765–774. Pickup, J., Mattock, M. & Kerry, S. (2002) Glycaemic control with continuous subcutaneous insulin infu- sion compared with intensive insulin injections in patients with type 1 diabetes: meta-analysis of ran- domised controlled trials. British Medical Journal, 324 (7339), 705. Pickup, J.C., Freeman, S.C. & Sutton, A.J. (2011) Glycaemic control in type 1 diabetes during real time continuous glucose monitoring compared with self monitoring of blood glucose: meta-analysis of ran- domised controlled trials using individual patient data. British Medical Journal, 343, d3805. Plotnick, L.P., Clark, L.M., Brancati, F.L. & Erlinger, T. (2003) Safety and effectiveness of insulin pump therapy in children and adolescents with type 1 diabetes. Diabetes Care, 26, 1142–1146. Raccah, D., Sulmont, V., Reznik, Y. et al. (2009) Incremental value of continuous glucose monitoring when starting pump therapy in patients with poorly controlled type 1 diabetes: the RealTrend study. Diabetes Care, 32, 2245–2250. Raju, B., Arbelaez, A.M., Breckenridge, S.M. & Cryer, P.E. (2006) Nocturnal hypoglycemia in type 1 diabetes: an assessment of preventive bedtime treatments. Journal of Clinical Endocrinology and Metab- olism, 91, 2087–2092. Rebrin, K. & Steil, G.M. (2000) Can interstitial glucose assessment replace blood glucose measurements? Diabetes Technology and Therapeutics, 2, 461–472. Retnakaran, R., Hochman, J., DeVries, J.H. et al. (2004) Continuous subcutaneous insulin infusion versus multiple daily injections: the impact of baseline A1c. Diabetes Care, 27, 2590–2596. Richardson, T. & Kerr, D. (2003) Skin-related complications of insulin therapy: epidemiology and emerging management strategies. American Journal of Clinical Dermatology, 4, 661–667. Roze, S., Valentine, W.J., Zakrzewska, K.E. & Palmer, A.J. (2005) Health-economic comparison of continu- ous subcutaneous insulin infusion with multiple daily injection for the treatment of Type 1 diabetes in the UK. Diabetic Medicine, 22, 1239–1245. Ryan, E.A. & Germsheid, J. (2009) Use of continuous glucose monitoring system in the management of severe hypoglycemia. Diabetes Technology and Therapeutics, 11, 635–639. Sachedina, N. & Pickup, J.C. (2003) Performance assessment of the Medtronic-MiniMed Continuous Glucose Monitoring System and its use for measurement of glycaemic control in Type 1 diabetic subjects. Diabetic Medicine, 20, 1012–1015. Scuffham, P. & Carr, L. (2003) The cost-effectiveness of continuous subcutaneous insulin infusion compared with multiple daily injections for the management of diabetes. Diabetic Medicine, 20, 586–593. 196 Hypoglycaemia in Clinical Diabetes

Shalwitz, R.A., Farkas-Hirsch, R., White, N.H. & Santiago, J.V. (1990) Prevalence and consequences of nocturnal hypoglycemia among conventionally treated children with diabetes mellitus. Journal of Pedi- atrics, 116, 685–689. St Charles, M., Lynch, P., Graham, C. & Minshall, M.E. (2009) A cost-effectiveness analysis of continuous subcutaneous insulin injection versus multiple daily injections in type 1 diabetes patients: a third-party US payer perspective. Value in Health, 12, 674–686. Tansey, M., Laffel, L., Cheng, J. et al. (2011) Satisfaction with continuous glucose monitoring in adults and youths with Type 1 diabetes. Diabetic Medicine, 28, 1118–1122. The DCCT Research Group (1991) Epidemiology of severe hypoglycemia in the Diabetes Control and Complications Trial. American Journal of Medicine, 90, 450–459. UK Hypoglycaemia Study Group (2007) Risk of hypoglycaemia in types 1 and 2 diabetes: effects of treat- ment modalities and their duration. Diabetologia, 50, 1140–1147. Wentholt, I.M., Vollebregt, M.A., Hart, A.A., Hoekstra, J.B. & DeVries, J.H. (2005) Comparison of a needle-type and a microdialysis continuous glucose monitor in type 1 diabetic patients. Diabetes Care, 28, 2871–2876. Wentholt, I.M., Hoekstra, J.B. & DeVries, J.H. (2006) A critical appraisal of the continuous glucose-error grid analysis. Diabetes Care, 29, 1805–1811. Wentholt, I.M., Hart, A.A., Hoekstra, J.B. & DeVries, J.H. (2007a) Relationship between interstitial and blood glucose in type 1 diabetes patients: delay and the push-pull phenomenon revisited. Diabetes Technology and Therapeutics, 9, 169–175. Wentholt, I.M., Maran, A., Masurel, N., Heine, R.J., Hoekstra, J.B. & DeVries, J.H. (2007b) Nocturnal hypoglycaemia in Type 1 diabetic patients, assessed with continuous glucose monitoring: frequency, duration and associations. Diabetic Medicine, 24, 527–532. Yates, K., Hasnat, M.A., Dear, K. & Ambler, G. (2006) Continuous glucose monitoring-guided insulin adjustment in children and adolescents on near-physiological insulin regimens: a randomized controlled trial. Diabetes Care, 29, 1512–1517. 10 Hypoglycaemia in Children with Diabetes Krystyna A. Matyka University of Warwick, Coventry, UK

INTRODUCTION

Suboptimal care of children with type 1 diabetes mellitus carries devastating consequences. Young children are at significant risk of developing microvascular complications that can present in adolescence (Marcovecchio et al. 2009). Yet they are also at risk of detrimental neuropsychological sequelae, which are thought to be related to recurrent episodes of hypoglycaemia that may accompany intensification of insulin therapy aimed at decreasing the risk of these microvascular complications. The exaggerated metabolic demands of the growing child, combined with a lifestyle that is unpredictable even on a day-to-day basis, make children very vulnerable both to repeated and severe episodes of hypoglycaemia (Allen et al. 2001). This chapter examines the aetiology, physiology, consequences and management of episodes of hypoglycaemia during this dynamic time of life.

DEFINITION OF HYPOGLYCAEMIA IN CHILDHOOD

Defining hypoglycaemia in childhood has proven to be extremely controversial. It has been suggested that children can tolerate lower levels of blood glucose, especially as the devel- oping brain can use alternative substrates for cerebral metabolism, thus limiting potential damage (Prins 2008). Yet this is a difficult area of study in view of the ethical difficulties of performing studies in young children of glucose homeostasis during fasting that could be used to define the limits of normality of blood glucose concentrations. Fasting glucose requirements will, in part, be determined by brain glucose uptake, and the central nervous system has a pivotal role in carbohydrate metabolism throughout life. Stable isotope studies of cerebral glucose metabolism indicate that the brain of infants and children can use glucose at a rate of 3–5 mg/kg/min, which is equivalent to almost all endogenous glucose production, and a linear correlation also exists between glucose pro- duction and estimated brain size, from premature infants to adult life (Bier et al. 1977). There is a marked change in the correlation between body weight and hepatic glucose production corresponding to mid-puberty (approximately 40 kg body weight) indicating that brain growth is by then virtually complete (Bier et al. 1977). However, the developing

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 198 Hypoglycaemia in Clinical Diabetes

Table 10.1 Summary of studies of glucose homeostasis during fasting in children. Data are presented in order of age

Duration of Glucose at end of Author Number Age (yrs) fast (hrs) fast (mmol/l)*

Chaussain (1973) 6 1–6 24 3.0 ± 0.8 Lamers et al. (1985a) 13 3–5 24 3.5 (median) Chaussain et al. (1974) 6 2–6 24 3.1 ± 0.3 Kerr et al. (1983) 8 2–8 36 2.6 ± 0.4 Haymond et al. (1982) 15 5–7 30 2.9 ± 0.2 Saudubray et al. (1981) 19 0.3–13 24 3.3 ± 0.7 Chaussain et al. (1977) 28 2–17 24 3.1 ± 0.1 Lamers et al. (1985b) 72 3–15 24 4.3 ± 0.1 Lamers et al. (1985a) 58 6–15 40 3.5 (median) Kerr et al. (1983) 6 10–18 36 3.7 ± 0.7

* Mean (± standard deviation).

brain also has the ability to use alternative substrates for cerebral metabolism, and during fasting young children have higher plasma concentrations of ketones and lactate when compared to adults (Haymond et al. 1982). A number of studies have examined metabolic parameters, including blood glucose concentrations, following fasts of varying duration in children. On the whole these studies have been used to provide normative data for use in the clinical evaluation of children with potential disorders of metabolism, and as a result the fasts have been pro- longed and sampling infrequent. There is little detailed information on metabolic varia- bles during a short fast as is experienced by children with diabetes, for example when they go to sleep. One study of nocturnal glucose homeostasis in 39 children, subse- quently found to have constitutional short stature, demonstrated a cyclical variation in blood glucose concentrations during the night with a periodicity of 80–120 minutes (Stir- ling et al. 1991). At some stage of the night blood glucose levels fell transiently to below 3 mmol/l in 5% of the children, but it is difficult to know if these children are representa- tive of this population. Other studies have examined glucose and metabolite profiles intermittently during pro- longed fasts of 24–40 hour duration (Chaussain 1973; Chaussain et al. 1974, 1977; Sau- dubray et al. 1981; Haymond et al. 1982; Kerr et al. 1983; Lamers et al. 1985a, 1985b) (Table 10.1). Fasting glucose concentrations varied depending on the duration of the fast and age of child studied. After 24 hours, blood glucose was found to range from 3.0–3.5 mmol/l. Some of these studies suggested a positive correlation between fasting blood glucose concentrations and age (Chaussain et al. 1977; Saudubray et al. 1981; Lamers et al. 1985a).

Definition of hypoglycaemia in childhood diabetes Although there has been great controversy regarding the biochemical definition of hypogly- caemia for the diagnosis of pathological states in the paediatric population (Koh et al. 1988), these arguments should not apply to type 1 diabetes. The management of type 1 Chapter 10: Hypoglycaemia in Children with Diabetes 199 diabetes involves attempting to maintain blood glucose concentrations that are well within the physiological range rather than merely accepting those that are outside the pathological range. The only study that has examined blood glucose concentrations after a period of fasting that would be representative of normal daily life (14 hours) recorded a fasting value of 4.3 ± 0.1 mmol/l in children aged 3–15 years (Lamers et al. 1985b). Guidelines now suggest that diabetes management should aim to keep blood glucose above 3.9 mmol/l and this recommendation is probably also reasonable for children (Working Group on Hypogly- caemia, American Diabetes Association [ADA] 2005; Clarke et al. 2009). Both in research studies and in clinical care, hypoglycaemia is often subdivided into degrees of severity based on the intervention required. The International Society for Pae- diatric and Adolescent Diabetologists (ISPAD) currently recommends that hypoglycaemia in childhood is classified as either moderate or severe (IDF 2011). This is in recognition of the fact that the usual adult definition of mild (i.e. self-treated) hypoglycaemia cannot be applied in young children who rely on their adult carers for every aspect of their diabetes management, because they are unlikely to be able to self-treat any episode.

PREVALENCE OF HYPOGLYCAEMIA

As the definition of hypoglycaemia has been controversial, studies of prevalence have often used variable definitions, both for daytime hypoglycaemia and for that occurring during sleep. Hypoglycaemia is notoriously under-reported, as many episodes, particularly mild ones, are quickly forgotten. A number of studies have examined the prevalence of severe hypoglycaemia in the paediatric population (Table 10.2). Some studies included only episodes of coma/convulsion, whereas others also included those events in which neurological impairment was severe enough to require intervention. Hypoglycaemia has been shown to be a significant problem, but, given the methodological complexities, reported prevalence rates of severe hypogly- caemia have varied greatly from 3.1 to 85.7 episodes per 100 patient-years. It is likely that less severe episodes are much more common and under-reported.

Nocturnal hypoglycaemia The prevalence of nocturnal hypoglycaemia has also been studied, using a biochemical rather than a symptomatic definition of hypoglycaemia, which has varied from 3.0 to 3.8 mmol/l. The prevalence of hypoglycaemia has ranged from 10 to 55% (Beregszaszi et al. 1997; Lopez et al. 1997; Porter et al. 1997; Matyka et al. 1999; Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group 2010). Most of these episodes do not waken the patient and, in everyday life, a great number of episodes of nocturnal hypoglycaemia are completely undetected.

SIGNS AND SYMPTOMS OF HYPOGLYCAEMIA

Daytime hypoglycaemia The classic symptoms of hypoglycaemia, as described in adults (Chapter 2), are classified into three distinct groups: autonomic, neuroglycopenic and non-specific (Dearyet al. 1993). The symptoms of hypoglycaemia in childhood differ from adults. One study examined the 200 Hypoglycaemia in Clinical Diabetes

of years day / 10 duration ratio > glucose visits

duration duration

duration

dose dose onset

Complications

insulin clinic blood at 1c 1c 1c 1c 1c 1c 1c 1c 1c 1c 1c

injections

diabetes duration of age age age age age age age age and

disease diabetes diabetes

insulin insulin experience regimen age

HbA HbA age HbA HbA HbA HbA HbA HbA HbA HbA HbA age apparent bolus exercise daily

evaluated insulin measurements > 0.8 3 Control Correlations Lower Younger Younger Lower Lower Longer Lower Younger Lower Younger Lower Lower More More Lower Higher Younger Lower Number Lower Older Lower Younger Young Increased Lower Underinsurance Younger Males > Longer Not None Younger Longer Higher Insulin Centre Diabetes Basal

and Diabetes yrs

4 yrs the

6.5 since diagnosis

patients after of after 4% No. Episodes / 100 patient years 7% 17.9 17.6 9.4 22 15–19 10.1–12.6 4.8 13.1 7.8 15.4 45 3.1 25 62 19 36 22.6 85.7 14.9 population

collection paediatric sectional sectional

the prospective data

in Study type Cross-sectional Prospective Prospective Prospective Cross-sectional Prospective Retrospective Prospective Prospective Prospective Cross Cross Prospective 7.5-year Prospective Retrospective Prospective

or or

of or or

therapy and injection injection

assistance assistance assistance assistance assistance assistance state requiring requiring requiring requiring

hypoglycaemia seizures seizures seizures seizures seizures seizures seizures seizures seizures seizures seizures

seizure seizure ± ± ± ± ± ± ± ± ± ± ±

mmol/l Requiring assistance Requiring assistance glucagon glucagon outside parenteral outside outside altered consciousness severe

2.8 Definition Coma, Coma, Coma Coma Requiring Coma Requiring Coma Requiring Coma Coma Coma Episodes Coma Coma Coma Coma Episodes Episodes Episodes < Coma of

268 195 195 872 187 146 657 709 376 415 300 prevalence

2807 2579 1243 1190 3560 6309 Number of subjects examining

6.6

± 0–18 1–19 3–26 1–18 0–18 0–18 1–19 1–24 0–34 7–16 0–19 1–16 1–17 0–9 studies 13–17 11–18

Age (years) 13.9 of

1993

in

.

(1997) (2005) Summary .

(1997) Control (2011) (2002) (1998) (2005) .

(1998)

. (2001)

(2002)

. (1997) (1998) .

. . (2001) . .

et al . (2003) . .

and . et al (1994) published

.

et al

et al

et al et al et al et al Research et al et al Complications

et al

et al et al

et al

was et al Diabetes

and Trial Group (1997) Ludvigsson (1999) Table 10.2 Table Trial Study The Bognetti Mortensen Nordfeldt Davis Davis Rosilio Tupola Thomsett Allen Levine Rewers Craig Holl Vanelli Wagner Blasetti Chapter 10: Hypoglycaemia in Children with Diabetes 201

of

years

day

/ 10

duration ratio

> glucose visits duration duration duration dose dose onset

Complications insulin clinic blood at 1c 1c 1c 1c 1c 1c 1c 1c 1c 1c 1c injections diabetes

duration of age age age age age age age age and

diabetes diabetes disease insulin insulin experience regimen age

HbA HbA age HbA HbA HbA HbA HbA HbA HbA HbA HbA age

apparent

bolus exercise daily

evaluated

insulin measurements > 0.8 3 Control Lower Younger Younger Lower Lower Longer Lower Younger Lower Younger Lower Lower More More Lower Higher Younger Lower Number Lower Older Lower Younger Young Increased Lower Underinsurance Younger Males > Longer Not None Younger Longer Higher Insulin Centre Diabetes Basal Correlations

and Diabetes

yrs

4 yrs

the 6.5 since diagnosis

patients after

of

after 4% 7% 17.9 17.6 9.4 22 15–19 10.1–12.6 4.8 13.1 7.8 15.4 45 3.1 25 62 19 36 22.6 85.7 14.9 No. Episodes / 100 patient years population

collection paediatric sectional sectional the prospective data in Cross-sectional Prospective Prospective Prospective Cross-sectional Prospective Retrospective Prospective Prospective Prospective Cross Cross Prospective 7.5-year Prospective Retrospective Prospective Study type or or

of or or

therapy and injection injection assistance assistance assistance assistance assistance assistance state requiring requiring requiring requiring hypoglycaemia seizures seizures seizures seizures seizures seizures seizures seizures seizures seizures seizures

seizure seizure ± ± ± ± ± ± ± ± ± ± ±

mmol/l

Requiring assistance Requiring assistance glucagon glucagon outside parenteral outside outside altered consciousness severe 2.8 Coma, Coma, Coma Coma Requiring Coma Requiring Coma Requiring Coma Coma Coma Episodes Coma Coma Coma Coma Episodes Episodes Episodes < Coma Definition of 268 195 195 872 187 146 657 709 376 415 300 prevalence 2807 2579 1243 1190 3560 6309 Number of subjects examining 6.6

±

0–18 1–19 3–26 1–18 0–18 0–18 1–19 1–24 0–34 7–16 0–19 1–16 1–17 0–9 studies 13–17 11–18 Age (years) 13.9 of 1993 in . (1997) (2005) Summary . (1997) Control

(2011) (2002) (1998) (2005) . (1998) . (2001)

(2002) . (1997) (1998) .

. . (2001) .

. et al

. (2003) . . and . et al (1994) published . et al et al

et al et al et al et al Research et al et al

Complications

et al et al et al

et al

was et al Diabetes and Trial Group (1997) Ludvigsson (1999) Table 10.2 Table Trial Study The Bognetti Mortensen Nordfeldt Davis Davis Rosilio Tupola Thomsett Allen Levine Rewers Craig Holl Vanelli Wagner Blasetti 202 Hypoglycaemia in Clinical Diabetes

Table 10.3 Reported symptoms during 221 episodes of mild hypoglycaemia (Source: Tupola S 1998a. Reproduced with permission from John Wiley & Sons, Ltd))

Symptom Prevalence (%)

Autonomic Tremor 22 Hunger 14 Sweating 4 Neuroglycopenic Drowsiness 34 Irritability/aggressiveness 4 Dizziness 1 Poor concentration 1 Blurred vision 1 Non-specific Weakness 7 Nausea 7 Abdominal pain 2 Headache 2 Tearfulness 1

frequency of hypoglycaemia in a group of children and adolescents using a 3-month diary (Tupola and Rajantie 1998). Episodes of hypoglycaemia (defined as a blood glucose ≤3 mmol/l) were documented along with symptoms. Of the patients (aged 2.5–21 years), 52% had a total of 287 episodes of hypoglycaemia, the majority of which (77%) were mild. The most common presenting symptoms were weakness, tremor, hunger and drowsiness; 39% of symptoms were classified as ‘adrenergic’ (autonomic) and 61% as neuroglycopenic or non-specific behavioural (Table 10.3). The dominant symptoms were different in differ- ent age groups of children. The children less than 6 years of age had fewer autonomic symptoms than adolescents, with the commonest symptom being drowsiness in young children and tremor in the older children (Tupola and Rajantie 1998).

Nocturnal hypoglycaemia The majority of episodes of nocturnal hypoglycaemia do not waken the child from sleep and thus go undetected, even in those who have normal awareness of hypoglycaemia during waking hours (Gale and Tattersall 1979; Bendtson et al. 1993; Porter et al. 1996; Bereg- szaszi et al. 1997; Lopez et al. 1997; Matyka et al. 1999). The possible reasons for lack of awareness of nocturnal episodes of hypoglycaemia are discussed in Chapter 5. Counter- regulatory responses are diminished during sleep in adults and adolescents, which may be related to the absence of symptoms during the night, and it is possible that normal hypotha- lamic activation in response to a falling blood glucose does not occur during sleep (Bendt- son et al. 1993; Jones et al. 1998; Diabetes Research in Children Network (DirecNet) Study Group 2007).

Impaired awareness of hypoglycaemia In one study of hypoglycaemia awareness that was performed in children (Ly et al. 2009), 79% of a paediatric clinic population of over 605 children and young people completed a validated questionnaire. This suggested that impaired awareness was present in 29% of patients, who had an earlier onset of diabetes, were younger and had lower glycated hae- moglobin values than those with intact awareness. Those individuals with impaired aware- Chapter 10: Hypoglycaemia in Children with Diabetes 203 ness had a higher rate of severe hypoglycaemia (37.1 vs 19.3 episodes per 100 patient-years, P < 0.001) (Ly et al. 2009). Another study that examined the ability of children aged 6–11 years and their parents to detect hypoglycaemia (Gonder-Frederick et al. 2008) showed that parents failed to recognise more than 50% of readings in the hypoglycaemic range (<3 mmol/l). Children fared better but still failed to detect more than 40% of episodes. This lack of detection may be a significant risk factor for severe hypoglycaemia in these young children (Gonder-Frederick et al. 2008) but may not represent true impaired awareness of hypoglycaemia. A recent study identified impaired awareness in 22% of children with type 1 diabetes, but very few below 9 years of age were able to self-assess their state of aware- ness (Graveling et al. in press).

RISK FACTORS FOR HYPOGLYCAEMIA

The discussion of risk factors naturally overlaps with that of prevention of hypoglycaemia, and a fuller review of specific interventions is provided in the section entitled ‘Prevention’. Table 10.2 presents the results of a number of studies that have been performed to examine the prevalence of severe hypoglycaemia in childhood since the Diabetes Control and Com- plications Trial (DCCT) was published (The Diabetes Control and Complications Trial Research Group 1993). Although a number of studies have suggested correlations between younger age and strict glycaemic control, it is important to note that many individual epi- sodes of hypoglycaemia may be explained by missed meals or unplanned exercise, but this would not have been addressed in epidemiological studies.

Glycaemic control Insulin requirements vary with age and are approximately 0.7–1 U/kg/day before puberty and 1.5–2 U/kg/day during adolescence, reflecting the insulin resistance that is present during this period of rapid growth and development (Dunger 1992). Despite numerous developments in terms of novel insulin preparations and modes of delivery, people with type 1 diabetes still experience varying states of insulin deficiency or excess that are difficult to control and predict. This is probably most evident in adolescents with type 1 diabetes in whom peripheral hyperinsulinaemia is achieved in an attempt to replace adequate levels of insulin in the portal circulation during the pubertal growth spurt (Dunger 1992). The DCCT highlighted the dilemma faced by all people with type 1 diabetes in relation to risk of complications versus risk of severe hypoglycaemia. A group of 195 adolescents, aged between 13 and 17 years, took part in this trial (The Diabetes Control and Complica- tions Trial Research Group 1994). Although the benefits of improved glycaemic control in terms of microvascular complications were still significant, the adolescents found it more difficult to achieve the low HbA1c concentration than adults (8.06 ± 0.13 vs 7.12 ± 0.03%; 65 ± 1.19 vs 54 ± 0.09 mmol/mol P < 0.001). Despite this, adolescents had a greater ten- dency towards experiencing severe hypoglycaemia: 85.7 events per 100 patient-years versus 56.9 events in the adult cohort (The Diabetes Control and Complications Trial Research Group 1994). However, in a European-wide clinical audit, which was designed to examine metabolic control in children and adolescents, it was found that severe hypogly- caemia was as common in those centres where metabolic control was poor, as in those centres that achieved better control as judged by HbA1c, indicating that research does not always reflect clinical experience (Mortensen et al. 1997). 204 Hypoglycaemia in Clinical Diabetes

Since 1993, there has been ample opportunity to refine the approaches to intensive insulin therapy and to improve education both for patients and physicians. Recent data suggest that lower levels of glycated haemoglobin are not inevitably accompanied by an increased frequency of hypoglycaemia, and treatment with insulin pumps may be particu- larly useful in achieving this (Cummins et al. 2010).

Nocturnal insulin requirements The mismatch of insulin delivery and insulin requirements with standard insulin regimens is particularly evident during the night, and many episodes of severe hypoglycaemia occur during sleep (Edge et al. 1990a; The Diabetes Control and Complications Trial Research Group 1997). Current insulin replacement regimens tend to result in hyperinsulinaemia in the early part of the night, when physiological insulin requirement is at its nadir (between 24:00–03:00 h) and so exacerbates the risk of hypoglycaemia at this time (Matyka et al. 1999; Mohn et al. 1999; Ford-Adams et al. 2003). Insulin requirements then peak between 04:00–08:00 h and a ‘dawn phenomenon’ occurs that can lead to fasting hyperglycaemia (Bolli and Gerich 1984; Edge et al. 1990b) and is thought to result from GH secretion during the later part of the night (De Feo et al. 1990; Edge et al. 1990b) and also, in part, to secretion of cortisol (Trumper et al. 1995). It has been shown that the overnight period is the time of greatest hypoglycaemia risk (The Diabetes Control and Complications Trial Research Group 1997). These overnight variations in insulin requirements can be better matched using continuous subcutaneous insulin infusion (CSII) and data suggest that hypoglycaemia can be reduced with insulin pump usage without compromising glycaemic control (Cummins et al. 2010).

Intensive insulin regimens Few studies have examined the impact of insulin regimen on the risk of hypoglycaemia in children. A number of studies examining prevalence of hypoglycaemia have found an inverse correlation between hypoglycaemia risk and glycated haemoglobin (Table 10.2). The Hvidore Study Group has formed a collaboration between 21 international paediatric centres from 18 countries (Holl et al. 2003). The group surveyed paediatric diabetes man- agement of 2873 children aged up to 18 years in 1995 and restudied 872 of these children in 1998. Although the use of multiple injection regimens increased from 42% to 71%, this did not lead to an improvement in glycaemic control as judged by glycated haemoglobin concentrations. Although there was a tendency towards an increase in the frequency of severe hypoglycaemia in the group of children/adolescents in whom the insulin regimen had been intensified, this did not reach statistical significance, perhaps because of the low number of events recorded (Holl et al. 2003). Another study of more than 6000 children has suggested that injection regimen and centre experience, as judged by the size of the clinic, may be significant risk factors for severe hypoglycaemia (Wagner et al. 2005). In this study of children aged up to 9 years, an increased risk of hypoglycaemia was observed in those children taking four insulin injections daily or on insulin pump therapy, compared to those children taking one to three injections daily. Small studies examining the benefits of insulin analogues suggest that the risk of hypoglycaemia may be lower with analogues without compromising glycaemic control (see later section on hypoglycaemia prevention). Chapter 10: Hypoglycaemia in Children with Diabetes 205

Diet Children with type 1 diabetes have the same nutritional requirements as children without diabetes. However, meals and snacks containing a high proportion of carbohydrate, have to be regularly distributed throughout the day to avoid the extremes of hypo- and hyper­ glycaemia (Magrath et al. 1993). This can be an issue for children who do not want to be different from their peers and do not want to eat at times when their friends are at play. Toddlers can present a special problem as many do not eat regular meals but graze during the day. Surprisingly, there has been little systematic study of the role of both quantity and quality of dietary components on the risk of hypoglycaemia. However, some studies of the preva- lence of hypoglycaemia have found that a number of episodes can be attributed to missed meals (Daneman et al. 1989; Davis et al. 1997; Tupola et al. 1998). The impact of dietary interventions on the avoidance of hypoglycaemia has been examined, mainly during sleep, and this is discussed later, in the section on hypoglycaemia prevention.

Physical activity As early as 1926, it was found that exercise could potentiate the hypoglycaemic effect of insulin in patients with type 1 diabetes (Lawrence 1926). Until recently there has been little systematic study of the impact of exercise on glucose homeostasis in childhood. One study used continuous glucose monitoring to study a standardised exercise protocol in a group of children who were using CSII. Glucose profiles were examined both during, and after, exercise on a cycle ergometer with the infusion pump either switched on or off (Admon et al. 2005). Hypoglycaemia was more common after, than during exercise, and this was true whether CSII was on or off. All subjects had one to three episodes of symptomatic hypoglycaemia within 2.5 to 12 hours after exercise, and four subjects had asymptomatic hypoglycaemia during exercise, only one of whom had consumed extra carbohydrate because their pre-exercise blood glucose had been below 5.5 mmol/l. Another study exam- ined the impact of daytime exercise on overnight blood glucose profiles in 50 subjects aged 10–18 years on intensive insulin regimens (Tsalikian et al. 2005). On one occasion they were studied during a day of afternoon exercise, involving four periods of 15 minutes each on a treadmill at a heart rate estimated to be 55% of maximum effort for this age group, and on a separate occasion during a rest day. In this study, 22% of subjects developed hypoglycaemia during exercise; overnight hypoglycaemia was more common during the night after the afternoon exercise than during a night after the rest day (Tsalikian et al. 2005). A glucose clamp study examined glucose variability during and after a standardised bout of exercise on a cycle ergometer (McMahon et al. 2007). Blood glucose concentrations were maintained at euglycaemia by using a fixed dose of intravenous insulin and a variable dose of intravenous glucose. Glucose infusion rates had to be increased during and shortly after exercise and then increased again from 7 to 11 hours after exercise. This study high- lights that patients are at risk of hypoglycaemia during and shortly after exercise but are also at risk of ‘delayed’ hypoglycaemia from 7 to 11 hours after exercise, often during sleep (McMahon et al. 2007).

Age Studies of prevalence of hypoglycaemia have consistently found that younger children, especially those under the age of 5 years, are at increased risk of hypoglycaemia. This may 206 Hypoglycaemia in Clinical Diabetes be a consequence of increased insulin sensitivity, irregular eating patterns or difficulty in identifying hypoglycaemia.

COUNTERREGULATION IN CHILDHOOD

The physiology of counterregulation is described in Chapters 1 and 3, but a brief overview of studies of counterregulation in childhood is presented here. A small number of studies have examined experimentally-induced hypoglycaemia, either using an insulin infusion method or a hyperinsulinaemic, hypoglycaemic glucose clamp technique (Amiel et al. 1987; Brambilla et al. 1987; Singer-Granick et al. 1988; Hoffman et al. 1991; Jones et al. 1991; Bjorgaas et al. 1997; Ross et al. 2005). As in adults, the glucagon response during hypoglycaemia is lost in children with dia- betes (Amiel et al. 1987; Jones et al. 1991; Ross et al. 2005). This is also the case in tod- dlers, aged 18–57 months old, who have a very short duration of diabetes (Brambilla et al. 1987). The majority of studies do suggest that children have exaggerated epinephrine (adrenaline) responses to hypoglycaemia, with peak values that are 2-fold higher than those found in adults (Amiel et al. 1987). Data from pre-pubertal children have been analysed separately from pubertal children, and no significant differences were found in epinephrine responses (Amiel et al. 1987; Ross et al. 2005). One study, using an insulin infusion as opposed to a hyperinsulinaemic clamp, did suggest that epinephrine responses were blunted in a group of poorly-controlled adolescents (Bjorgaas et al. 1997). The reason for the dis- crepant results is not clear. Glucose thresholds for counterregulatory responses have received very little attention.

In a study of very poorly-controlled adolescents (average total HbA1 = 15%; 140 mmol/ mol), glucose thresholds for epinephrine secretion were significantly different, with the poorly-controlled adolescents releasing epinephrine at a plasma glucose concentration of 4.7 mmol/l compared to 3.9 mmol/l in non-diabetic adolescents (Jones et al. 1991).

Effect of sleep stage on counterregulation Prolonged nocturnal hypoglycaemia is common in childhood (Matyka et al. 1999; Diabetes Research in Children Network (DirecNet) Study Group 2007). Studies of nocturnal hypoglycaemia both in adults and adolescents indicate that counterregulatory responses are blunted during sleep (see Chapter 5). A study of experimentally-induced hypoglycaemia during the night in adolescents with diabetes and non-diabetic controls suggested that epinephrine responses during hypoglycaemia were blunted when hypoglycaemia was induced during slow wave sleep compared to when hypoglycaemia was induced when subjects were awake during the day or during the night (Jones et al. 1998).

CONSEQUENCES OF HYPOGLYCAEMIA

Cognitive impairment Severe hypoglycaemia from a variety of causes can cause catastrophic cerebral damage when it is profound and prolonged in very young children (Lucas et al. 1988). Glucose is critical not only as the major fuel for cerebral metabolism but also as a precursor of essential Chapter 10: Hypoglycaemia in Children with Diabetes 207 substrates, which are essential for normal brain development (Glazer and Weber 1971). Concerns have been raised that recurrent hypoglycaemia could affect long-term academic achievement in children with type 1 diabetes, from the effects of hypoglycaemia disrupting school performance to the possibility of cerebral damage accumulating over time.

Acute effects Few studies have studied the impact of acute hypoglycaemia on cognitive performance in children or adolescents. One study of the effects of experimentally-induced mild hypogly- caemia (3.1–3.6 mmol/l) found decrements in tests of mental flexibility and on measures that required planning and decision-making and a rapid response, although the results were variable within subjects (Ryan et al. 1990). The learning ability of children, who spend much of their day at school assimilating information, could be seriously compromised if they experience frequent episodes of even mild hypoglycaemia during their time in class. Children can also be affected if they miss lessons because of severe hypoglycaemic events and could be compromised by asymptomatic episodes of hypoglycaemia that go undetec- ted and therefore untreated. There is some evidence to suggest that children may be partic- ularly susceptible to mild episodes of hypoglycaemia – studies examining P300 evoked potentials and EEG changes in response to hypoglycaemia have found that abnormalities commence at a higher blood glucose level in children than in adults (Jones et al. 1995; Bjorgaas et al. 1998). The effect of nocturnal hypoglycaemia on cognitive function has received little attention. Cognitive performance could be directly affected by hypoglycaemia, but also by sleep disturbance. The few studies of nocturnal hypoglycaemia in children and adolescents with type 1 diabetes have not shown any deleterious effect on overall sleep physiology (Bendt- son et al. 1992; Porter et al. 1996; Matyka et al. 2000; Pillar et al. 2003) or on cognitive function the following morning.

Long-term effects A comprehensive overview of this topic is not possible here, but a systematic review can be recommended (Gaudieri et al. 2008) and see Chapter 14. Studies suggest that children with diabetes have inferior academic achievements compared to non-diabetic peers (Dahl- quist and Kallen 2007), and perform more poorly on measures of intelligence and tests of attention, executive function, cognitive flexibility, speed of information processing, psy- chomotor efficiency and visual perception (Brandset al. 2005). Developing diabetes within the first 5 years of life appears to be a particularly vulnerable period of development and is associated with clinically significant cognitive deficits (Ryanet al. 1985; Lin et al. 2010). A recent meta-analysis revealed learning and memory, and attention and executive function skills, to be more likely to be affected and concluded that cognitive effects are most pro- nounced and pervasive for children who develop diabetes at a young age (Gaudieri et al. 2008). The reasons for this vulnerability are not clear but it is unlikely to be exclusively a consequence of hypoglycaemia (either symptomatic or asymptomatic) or persistently poor control (Ryan 1997; Lin et al. 2010). It is possible that the brains of young children are more prone to the deleterious effects of glycaemic variability, or there may be other meta- bolic disturbances that occur that have yet to be elucidated (Lin et al. 2010). One of the only longitudinal studies has been performed in Melbourne in Australia. Cognitive performance was assessed in a group of children (123 at baseline) with 208 Hypoglycaemia in Clinical Diabetes newly-diagnosed diabetes, aged 3–14 years, who were compared to healthy controls at 3 months, 2 years, 6 years and 12 years following diagnosis (Lin et al. 2010). At 12 years, 85% (106) of children with diabetes and 81% (75) of the original controls were re-studied. Of the children with diabetes, 40 had diabetes of early onset, 47 had at least one episode of severe hypoglycaemia and 64 had poor metabolic control; 48 children had two to three risk factors. The authors found that children with diabetes performed less well on working memory when compared to their non-diabetic peers. Children experiencing severe hypogly- caemia had lower verbal ability, working memory, and non-verbal processing speed com- pared to both children without severe hypoglycaemia and healthy controls (Lin et al. 2010). The authors point out that these decrements were subtle and that group performances were within the average range, yet it is not clear what the impact of these changes may be on an individual’s ability to self- manage their diabetes. There have been few studies using Magnetic Resonance Imaging (MRI) to assess structural CNS changes in children with type 1 diabetes, and these have shown conflicting results. Studies suggest that there is a high incidence of CNS abnormalities, particularly mesial temporal sclerosis suggesting hippocampal damage, irrespective of a history of severe hypoglycaemia or episodes of diabetic ketoacidosis (Wessels et al. 2006). Other studies have found an impact of severe hypoglycaemia with reduced grey matter in the left superior temporal region, with more extreme hypoglycaemia being associated with volume loss in the cuneus and precuneus (Perantie et al. 2007). Children who have been followed up for 12 years after developing type 1 diabetes have loss of grey matter bilater- ally in the thalamus, in the right parahippocampal region and insula, as well as loss of white matter in the bilateral parahippocampi, left temporal lobe and middle frontal area (Northam et al. 2009).

Hypoglycaemic hemiplegia This is a rare complication of acute hypoglycaemia in which the patient recovers from the hypoglycaemia with a transient hemiparesis lasting no more than 24 hours. When subsequent neuroimaging is performed it is rare to find an any residual abnormality. There is no evidence that this neurological manifestation of severe neuroglycopenia causes any permanent disability (Pocecco and Ronfani 1998).

Fear of hypoglycaemia A recent systematic review has highlighted a lack of data for the prevalence and effects of fear of hypoglycaemia in the paediatric population, although clinical experience would suggest that it is a significant problem (Barnard et al. 2010). Children with diabetes and their parents worry about the prospect of a severe episode of hypoglycaemia (Gold et al. 1997; Gonder-Frederick et al. 1997; Marrero et al. 1997; Clarke et al. 1998; Nordfeldt and Ludvigsson 2005) (see Chapter 16). In one study, severe hypoglycaemia caused more fear than the prospect of an episode of diabetic ketoacidosis (Nordfeldt and Ludvigsson 2005). Severity of hypoglycaemia was related more to fear than frequency of hypoglycaemia (Patton et al. 2008), and those children with a severe episode of hypoglycaemia did have a higher percentage of self-monitoring blood glucose values above the desired target range compared to those who had no severe episodes of hypoglycaemia (Marrero et al. 1997). No studies appear to have found a link between fear of hypoglycaemia and glycated hae- moglobin (Barnard et al. 2010). Chapter 10: Hypoglycaemia in Children with Diabetes 209

MANAGEMENT OF HYPOGLYCAEMIA

The management of acute hypoglycaemia will depend on the severity of the episode. ISPAD has provided guidelines for the management of acute episodes of hypoglycaemia based on severity (International Diabetes Federation, 2011) (Figure 10.1). Blood glucose measurement is the only way to confirm hypoglycaemia if the diagnosis is uncertain and also to confirm that blood glucose has returned to normal after treatment. Figure 10.1 shows the flow diagram for the management of hypoglycaemia based on ISPAD guidelines.

Prevention When a child is experiencing recurrent episodes of hypoglycaemia, a detailed history should be obtained (from the patient and family) regarding the timing of hypoglycaemia,

Check blood glucose – Glucose meter measurement and formal laboratory glucose if possible

Hypoglycaemia

Moderate Severe

Unconscious

Oral glucose to raise blood glucose to 5.6mmol/l I.M. Glucagon Estimate that 0.3g/kg of carbohydrate < 12 yrs 0.5mg will increase glucose by 3–4 mmol/l- > 12 yrs 1.0mg Recheck blood glucose after 10–15 minutes and repeat as necessary

If no response in 15 mins give 1–2 mls / kg of 10% dextrose I.V. Repeat until there is a clinical response.

As symptoms improve or normoglycaemia is restored give oral complex carbohydrate e.g. biscuit, bread or alternatives. If unable to tolerate oral carbohydrate may need a glucose infusion e.g. at a rate of 2–5mg/kg/minute

Figure 10.1 Management of moderate and severe hypoglycaemia in children. I.M., intramuscular; I.V., intravenous. (Source: Clarke W 2009. Reproduced with permission from John Wiley & Sons, Ltd). 210 Hypoglycaemia in Clinical Diabetes insulin regimen, dietary intake and the relation to periods of physical activity. This will enable an assessment to be made of possible risk factors and inform how these may be avoided. If no obvious cause is found then other pathology should be sought, such as coincidental coeliac disease or the possibility of Addison’s disease, although these are rela- tively rare causes of recurrent hypoglycaemia (see Chapter 4). When contemplating preven- tive management the following aspects should be considered.

Education Great importance is placed on education in the management of type 1 diabetes, and struc- tured education programmes are now an essential part of any diabetes service provision. Despite this, little validation has been made of the use of structured educational pro- grammes in children. One study from Scandinavia has shown the benefits of a focused education package (Nordfeldt et al. 2003). In this study families were given both written and video information on diabetes. One group was given general information on diabetes management, and the other was given material to educate about the importance and means of hypoglycaemia avoidance. Although no differences in glycated haemoglobin were found between the two groups, those who had the targeted intervention had a significantly reduced rate of severe hypoglycaemia (Nordfeldt et al. 2003).

Insulin regimen As noted earlier, the DCCT suggested that attempts to intensify insulin regimens may increase the risk of severe hypoglycaemia. However, few studies have been designed specifically to examine the benefits of different insulin regimens on the risk of hypogly- caemia (Siebenhofer et al. 2006). One study that compared insulin lispro with soluble (short-acting) insulin as part of a basal bolus regimen showed small but statistically sig- nificant reductions in the prevalence of hypoglycaemia over a 30-day period when insulin lispro was being used (Holcombe et al. 2002). Other studies have found benefits of insulin analogue regimens on nocturnal hypoglycaemia. One randomised crossover study in ado- lescents compared insulin lispro and glargine, as part of a daily multiple injection regimen, to human soluble and isophane insulins. Nocturnal hypoglycaemia was 43% lower with the analogue regimen, although no difference was observed in self-reported symptomatic hypoglycaemia (Murphy et al. 2003). A large study of almost 350 children, aged 6–17 years, comparing insulin detemir versus NPH insulin as part of a basal bolus regimen with insulin aspart, found a beneficial effect of detemir, with symptomatic nocturnal hypogly- caemia being 26% lower, but with no effect on daytime hypoglycaemia (Robertson et al. 2007). Although not every patient is suited to using an insulin pump, data suggest that this is an effective form of treatment for hypoglycaemia avoidance (Cummins et al. 2010). In the UK, where prescribing of insulin pump therapy is restricted by national guidance, hypogly- caemia and fear of hypoglycaemia are recognised indications for pump therapy (National Institute of Clinical Health and Excellence 2008). Most importantly, hypoglycaemia avoid- ance can be achieved without a significant increase in glycated haemoglobin on pump therapy, although very few well-designed studies to have examined the benefits of insulin pump therapy versus multiple daily injections (Cummins et al. 2010). Chapter 10: Hypoglycaemia in Children with Diabetes 211

Continuous glucose monitoring systems It is now possible to continuously monitor glucose levels (see Chapter 9), although the use of continuous glucose monitoring (CGM) is not widespread in routine clinical practice and still has limitations. Studies of the benefits of CGM on diabetes control have shown mixed results. In a Juvenile Diabetes Research Foundation (JDRF) study, 322 patients, all on intensive insulin therapies, and including 114 children aged 8–14 years, were ran- domly assigned to CGM versus home blood glucose monitoring (Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group 2008). There were no significant group differences in glycated haemoglobin among the paediatric population, which may have been related to low use of CGM. The frequency of severe hypoglycaemia was low and there were no beneficial effects of CGM on rates of severe or biochemical hypoglycaemia (Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group 2008). The STAR 3 study of 485 patients, 156 of whom were children aged 7–18 years, compared sensor-augmented pump therapy with multiple daily injections and blinded continuous glucose monitoring at predetermined time points. In this study glycated haemoglobin was significantly reduced in the sensor-augmented pump group but without any detrimental effects on rates of hypoglycaemia (Bergenstal et al. 2010). Closed loop systems are also now being developed and have been used to control nocturnal glycaemia. A study of 40 adolescents and adults aged 12–39 years showed that algorithms designed to trigger insulin pump suspension could prevent hypoglycaemia in 75% of nights when it would otherwise have been predicted to occur (Buckingham et al. 2010).

Diet Few studies have examined the impact of dietary interventions on hypoglycaemia risk except for nocturnal hypoglycaemia. The major dietary modification has been that of the introduction of a larger proportion of starch, in the form of long-acting carbohydrate, as part of the evening snack (Ververs et al. 1993; Kaufman et al. 1995; Detlofson et al. 1999; Matyka et al. 1999). In these studies the benefits of starch have been inconsistent. One study found a lower frequency of nocturnal hypoglycaemia, although capillary sampling was performed only intermittently during the night and some episodes of hypoglycaemia may have been undetected (Kaufman et al. 1995). Others found no beneficial effect of cornstarch on the prevention of nocturnal hypoglycaemia, although blood glucose concen- trations fell more slowly, but in one study this occurred at the expense of promoting hyperglycaemia (Ververs et al. 1993; Matyka et al. 1999). Although not designed to examine the impact of diet on hypoglycaemia, a study of a low glycaemic index diet has been shown to improve glycaemic control without an increase in rate of hypoglycaemia, and appeared to have enhanced quality of life (Gilbertson et al. 2001).

Exercise The American Diabetes Association (ADA) has provided guidelines for the management of diabetes and exercise; however, this is very much a consensus statement rather than an evidence-based document (ADA 2000). ISPAD have also produced guidelines (Robertson et al. 2009), recommending that careful monitoring of blood glucose is essential and food and insulin should be matched to the intensity of exercise and that a reduction of insulin dose should be considered. Additional slowly absorbed carbohydrate will be necessary, 212 Hypoglycaemia in Clinical Diabetes especially at bedtime, if exercise has been performed in the afternoon or early evening (see Chapters 5 and 17). Taplin et al. have studied 16 adolescents on insulin pump therapy and compared the effect of oral terbutaline (2.5 mg) versus a 20% reduction in the overnight basal rate of insulin after a standardised 60 minute exercise session (Taplin et al. 2010). Terbutaline did result in hypoglycaemia being avoided overnight but caused significantly more fasting hyperglycaemia compared to reducing basal rate insulin (blood glucose >13.8 mmol/l). Blood glucose profiles were better overnight with the basal rate reduction but more hyperglycaemia occurred than was clinically acceptable (Taplin et al. 2010). Given the lack of a robust database it is important to work with the child and family to provide individually-tailored recommendations that can then be tried and tested. The man- agement of unpredictable physical activity is likely to remain a problem until a cure for diabetes is found.

CONCLUSIONS • Hypoglycaemia is a common problem for children with type 1 diabetes, especially young children, and for their families. • Behavioural symptoms are more common in childhood than typical autonomic symp- toms. Most episodes of nocturnal hypoglycaemia are asymptomatic. • Episodes of hypoglycaemia remain a significant barrier when striving for a degree of glycaemic control that will delay or prevent the development of the microvascular complications of diabetes. • Glucagon secretion during hypoglycaemia is lost early in the course of type 1 diabetes. Epinephrine responses to hypoglycaemia during sleep are blunted and may contribute to the absence of symptoms of hypoglycaemia and the failure to waken during the night. • Concerns remain about the possible long-term implications of hypoglycaemia in terms of cognitive dysfunction, although hyperglycaemia may also contribute. • New insulin analogues and insulin pump therapy may enable improvements in glycaemic control to be achieved without a concomitant increase in the risk of hypoglycaemia. • All children with diabetes should have an individual action plan for hypoglycaemia avoidance.

REFERENCES

Admon, G., Weinstein, Y., Falk, B. et al. (2005) Exercise with and without an insulin pump among children and adolescents with type 1 diabetes mellitus. Pediatrics, 116, 348–355. Allen, C., LeCaire, T., Palta, M., Daniles, K., Meredith, M., D’Alessio, D.J. & for the Wisconsin Diabetes Registry Project (2001) Risk factors for frequent and severe hypoglycemia in type 1 diabetes. Diabetes Care, 24, 1878–1881. American Diabetes Association (2000) Diabetes mellitus and exercise. Diabetes Care, 23, s50–s54. American Diabetes Association Workgroup on Hypoglycemia (2005) Defining and reporting hypoglycemia in diabetes: a report from the American Diabetes Association Workgroup on Hypoglycemia. Diabetes Care, 28, 1245–1249. Amiel, S.A., Simonson, D.C., Sherwin, R.S., Lauritano, A.A. & Tamborlane, W.V. (1987) Exaggerated epinephrine responses to hypoglycemia in normal and insulin-dependent diabetic children. The Journal of Pediatrics, 110, 832–837. Chapter 10: Hypoglycaemia in Children with Diabetes 213

Barnard, K., Thomas, S., Royle, P., Noyes, K. & Waugh, N. (2010) Fear of hypoglycaemia in parents of young children with type 1 diabetes: a systematic review. BMC Pediatrics, 10, 50. Bendtson, I., Gade, J., Thomsen, C.E., Rosenfalck, A. & Wildschiodtz, G. (1992) Sleep disturbances in IDDM patients with nocturnal hypoglycemia. Sleep, 15, 74–81. Bendtson, I., Rosenfalck, A.M. & Binder, C. (1993) Nocturnal versus diurnal counterregulation to hypogly- cemia in type 1 (insulin-dependent) diabetic patients. Acta Endocrinologica, 128, 109–115. Beregszaszi, M., Tubiana-Rufi, N., Benali, K., Noel, M., Bloch, J. & Czernichow, P. (1997) Nocturnal hypoglycemia in children and adolescents with insulin-dependent diabetes mellitus: prevalence and risk factors. The Journal of Pediatrics, 131, 27–33. Bergenstal, R.M., Tamborlane, W.V., Ahmann, A. et al.; for the STAR 3 Study Group (2010) Effectiveness of sensor-augmented insulin-pump therapy in type 1 diabetes. New England Journal of Medicine, 363, 311–320. Bier, D.M., Leake, R.D., Haymond, M.W. et al. (1977) Measurement of ‘true’ glucose production rates in infancy and childhood with 6,6-dideuteroglucose. Diabetes, 26, 1016–1023. Bjorgaas, M., Vik, T., Sand, T. et al. (1997) Counterregulatory and symptom responses to hypoglycaemia in diabetic children. Diabetic Medicine, 14, 433–441. Bjorgaas, M., Sand, T., Vik, T. & Jorde, R. (1998) Quantitative EEG during controlled hypoglycaemia in diabetic and non-diabetic children. Diabetic Medicine, 15, 30–37. Blasetti, A., Di Giulio, C., Tocco, A.M. et al. (2011) Variables associated with severe hypoglycemia in children and adolescents with type 1 diabets: a population based study. Pediatric Diabetes, 12, 4–10. Bognetti, F., Brunelli, A., Meschi, F., Viscardi, M., Bonfanti, R. & Chiumello, G. (1997) Frequency and correlates of severe hypoglycaemia in children and adolescents with diabetes mellitus. European Journal of Paediatrics, 156, 589–591. Bolli, G. & Gerich, J. (1984) The dawn phenomenon – a common occurrence in both non-insulin dependent and insulin dependent diabetes mellitus. New England Journal of Medicine, 310, 746–750. Brambilla, P., Bougneres, P.F., Santiago, J.V., Chaussin, J.L., Pouplard, A. & Castano, L. (1987) Glucose counterregulation in pre-school-age children with recurrent hypoglycemia during conventional treat- ment. Diabetes, 36, 300–304. Brands, A.M., Biessels, G.J., de Haan, E.H., Kappelle, L.J. & Kessels, R.P. (2005) The effects of type 1 diabetes on cognitive performance: a meta-analysis. Diabetes Care, 28, 726–735. Buckingham, B., Chase, H.P., Dassau, E. et al. (2010) Prevention of nocturnal hypoglycemia using predic- tive alarm algorithms and insulin pump suspension. Diabetes Care, 33, 1013–1017. Chaussain, J.-L. (1973) Glycemic response to 24-hour fast in normal children and children with ketotic hypoglycemia. The Journal of Pediatrics, 82, 438–443. Chaussain, J.-L., Georges, P., Olive, G. & Job, J.-C. (1974) Glycemic response to 24-hour fast in normal children and children with ketotic hypoglycemia: II Hormonal and metabolic changes. The Journal of Pediatrics, 85, 776–781. Chaussain, J.-L., Georges, P., Calzada, L. & Job, J.-C. (1977) Glycemic response to 24-hour fast in normal children: III Influence of age. The Journal of Pediatrics, 91, 711–714. Clarke, W., Jones, T., Rewers, A., Dunger, D. & Kligensmith, G.J. (2009) Assessment and management of hypoglycemia in children and adolescents with diabetes. Pediatric Diabetes, 10 (Suppl. 12), 134–145. Clarke, W.L., Gonder-Frederick, A., Snyder, A.L. & Cox, D.J. (1998) Maternal fear of hypoglycemia in their children with insulin dependent diabetes mellitus. The Journal of Pediatric Endocrinology and Metabolism, 11, 189–194. Craig, M.E., Handelsman, P., Donaghue, K.C. et al.; NSW/ACT HbA(1c) Study Group (2002) Predictors of glycaemic control and hypoglycaemia in children and adolescents with type 1 diabetes from NSW and ACT. The Medical Journal of Australia, 177, 235–238. Cummins, E., Royle, P., Snaith, A. et al. (2010) Clinical effectiveness and cost-effectiveness of continuous subcutaneous insulin infusion for diabetes: systematic review and economic evaluation. Health Technol- ogy Assessment, 14, iii–iv, xi–xvi, 1–181. Dahlquist, G. & Kallen, B. (2007) School performance in children with type 1 diabetes - a population-based register study. Diabetologia, 50, 957–964. Daneman, D., Frank, M., Perlman, K., Tamm, J. & Ehrlich, R. (1989) Severe hypoglycemia in children with insulin-dependent diabetes mellitus: frequency and predisposing factors. The Journal of Pediatrics, 5, 681–685. 214 Hypoglycaemia in Clinical Diabetes

Davis, E.A., Keating, B., Byrne, G.C., Russell, M. & Jones, T.W. (1997) Hypoglycemia: incidence and clinical predictors in a large population based sample of children and adolescents with IDDM. Diabetes Care, 20, 20–25. Davis, E.A., Keating, B., Byrne, G.C., Russell, M. & Jones, T.W. (1998) Impact of improved glycaemic control on rates of hypoglycaemia in IDDM. Archives of Disease in Childhood, 78, 111–115. De Feo, P., Torlone, E., Fanelli, C., Santeusanio, F., Brunetti, P. & Bolli, G.B. (1990) Nocturnal spikes of growth hormone secretion cause the dawn phenomenon in type 1 (insulin dependent) diabetes mellitus by decreasing hepatic (and extrahepatic) sensitivity to insulin in the absence of insulin waning. Diabeto- logia, 33, 52–59. Deary, I.J., Hepburn, D.A., MacLeod, K.M. & Frier, B.M. (1993) Partitioning the symptoms of hypogly- caemia using multi-sample confirmatory factor analysis. Diabetologia, 36, 771–777. Detlofson, I., Kroon, M. & Aman, J. (1999) Oral bedtime cornstarch supplementation reduces the risk for nocturnal hypoglycaemia in young children with type 1 diabetes. Acta Paediatrica, 88, 595–597. Diabetes Research in Children Network (DirecNet) Study Group (2007) Impaired overnight counterregula- tory hormone responses to spontaneous hypoglycemia in children with Type 1 diabetes. Pediatric Diabetes, 8, 199–205. Dunger, D.B. (1992) Diabetes in puberty. Archives of Disease in Childhood, 67, 569–570. Edge, J.A., Matthews, D.R. & Dunger, D.B. (1990a) Failure of current insulin regimes to meet the overnight requirements of adolescents with insulin-dependent diabetes. Diabetes Research, 15, 109–112. Edge, J.A., Matthews, D.R. & Dunger, D.B. (1990b) The dawn phenomenon is related to overnight growth hormone release in adolescent diabetics. Clinical Endocrinology, 33, 729–737. Ford-Adams, M.E., Murphy, N.P., Moore, E.J. et al. (2003) Insulin lispro: a potential role in preventing nocturnal hypoglycaemia in young children with diabetes mellitus. Diabetic Medicine, 20, 656–660. Gale, E.A.M. & Tattersall, R.B. (1979) Unrecognised nocturnal hypoglycaemia in insulin-treated diabetics. Lancet, i, 1049–1052. Gaudieri, P.A., Chen, R., Greer, T.F. & Holmes, C.S. (2008) Cognitive function in children with type 1 diabetes: a meta-analysis. Diabetes Care, 31, 1892–1897. Gilbertson, H.R., Brand-Miller, J.C., Thorburn, A.W., Evans, S., Chondros, P. & Werther, G.A. (2001) The effect of flexible low glycemic index dietary advice versus measured carbohydrate exchange diets on glycemic control in children with type 1 diabetes. Diabetes Care, 24, 1137–1143. Glazer, R.I. & Weber, G. (1971) Incorporation of [6-3H] glucose into lipid, protein, RNA and DNA slices of differentiating rat cerebral cortex. Journal of Neurochemistry, 18, 1569–1576. Gold, A.E., Deary, I.J. & Frier, B.M. (1997) Hypoglycaemia and non-cognitive aspects of psychological function in insulin-dependent (type 1) diabetes mellitus (IDDM). Diabetic Medicine, 14, 111–118. Gonder-Frederick, L., Cox, D., Kovatchev, B., Julian, D. & Clarke, W. (1997) The psychosocial impact of severe hypoglycemic episodes on spouses of patients with IDDM. Diabetes Care, 20, 1543–1546. Gonder-Frederick, L., Zrebiec, J., Dauchowitz, A. et al. (2008) Detection of hypoglycemia by children with type 1 diabetes 6 to 11 years of age and their parents: a field study. Pediatrics, 121, e489–e495. Graveling, A.J., Noyes, K.J., Allerhand, M.H., Wright, R.J., Bath, L.E., Deary, I.J., Frier, B.M. (in press) Prevalence of impaired awareness of hypoglycemia and identification of predictive symptoms in children and adolescents with type 1 diabetes. Pediatric Diabetes. Haymond, M.W., Karl, I.E., Clarke, W.L., Pagliara, A.S. & Santiago, J.V. (1982) Differences in circulating gluconeogenic substrates during short-term fasting in men, women and children. Metabolism: Clinical and Experimental, 31, 33–42. Hoffman, R.P., Singer-Granick, C., Drash, A.L. & Becker, D.J. (1991) Plasma catecholamine responses to hypoglycemia in children and adolescents with IDDM. Diabetes Care, 14, 81–88. Holcombe, J.H., Zalani, S., Arora, V.K. & Mast, C.J. (2002) Lispro in Adolescents Study Group. Compari- son of insulin lispro with regular human insulin for the treatment of type 1 diabetes in adolescents. Clinical Therapeutics, 24, 629–638. Holl, R.W., Swift, P.G., Mortensen, H.B. et al.; on behalf of the Hvidore Study Group (2003) Insulin injec- tion regimens and metabolic control in an international survey of adolescents with type 1 diabetes over 3 years: results from the Hvidore Study Group. European Journal of Paediatrics, 162, 22–29. International Diabetes Federation (2011) Chapter 11. Global IDF / ISPAD Guideline for Diabetes in Child- hood and Adolescence. International Diabetes Federation (IDF), Brussels. Jones, T.W., Boulware, S.D., Kraemer, D.T., Caprio, S., Sherwin, R.S. & Tamborlane, W.V. (1991) Inde- pendent effects of youth and poor diabetes control on responses to hypoglycemia in children. Diabetes, 40, 358–363. Chapter 10: Hypoglycaemia in Children with Diabetes 215

Jones, T.W., Borg, W.P., Boulware, S.D., McCarthy, G., Sherwin, R.S. & Tamborlane, W.V. (1995) Enhanced adrenomedullary response and increased susceptibility to neuroglycopenia: mechanisms underlying the adverse effects of sugar ingestion in healthy children. The Journal of Pediatrics, 126, 171–177. Jones, T.W., Porter, P., Sherwin, R.S. et al. (1998) Decreased epinephrine responses to hypoglycemia during sleep. New England Journal of Medicine, 338, 1657–1662. Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group (2008) Continuous Glucose Monitoring and intensive treatment of type 1 diabetes. New England Journal of Medicine, 359, 1464–1476. Juvenile Diabetes Research Foundation Continuous Glucose Monitoring Study Group (2010) Prolonged nocturnal hypoglycemia is common during 12 months of continuous glucose monitoring of children and adults with type 1 diabetes. Diabetes Care, 33, 1004–1008. Kaufman, F.R., Halvorson, M. & Kaufman, N.D. (1995) A randomized, blinded trial of uncooked cornstarch to diminish nocturnal hypoglycemia at diabetes camp. Diabetes Research and Clinical Practice, 30, 205–209. Kerr, D.S., Hansen, I.L. & Levy, M.M. (1983) Metabolic and hormonal responses of children and adoles- cents to fasting and 2-deoxyglucose. Metabolism: Clinical and Experimental, 22, 951–959. Koh, T.H.H.G., Eyre, J.A. & Aynsley-Green, A. (1988) Neonatal hypoglycaemia – the controversy regard- ing definition. Archives of Disease in Childhood, 63, 1386–1388. Lamers, K.J.B., Doesburg, W.H., Gabreels, F.J.M. et al. (1985a) The concentration of blood components related to fuel metabolism during prolonged fasting in children. Clinica Chimica Acta, 152, 155–163. Lamers, K.J.B., Doesburg, W.H., Gabreels, F.J.M. et al. (1985b) Reference values of blood components related to fuel metabolism in children after an overnight fast. Clinica Chimica Acta, 145, 17–26. Lawrence, R.D. (1926) The effects of exercise on insulin action in diabetes. British Medical Journal, 1, 648–652. Levine, B., Anderson, B.J., Butler, D.A., Antisdel, J.E., Brackett, J. & Laffel, L.M.B. (2001) Predictors of glycemic control and short-term adverse outcomes in youths with type 1 diabetes. The Journal of Pedi- atrics, 139, 197–203. Lin, A., Northam, E.A., Rankins, D., Werther, G.A. & Cameron, F.J. (2010) Neuropsychological profiles of young people with Type 1 diabetes 12 yr after disease onset. Pediatric Diabetes, 11, 235–243. Lopez, M.J., Oyarzabal, M., Barrio, R. et al. (1997) Nocturnal hypoglycaemia in IDDM patients younger than 18 years. Diabetic Medicine, 14, 772–777. Lucas, A., Morley, R. & Cole, T.J. (1988) Adverse neurodevelopmental outcome of moderate neonatal hypoglycaemia. British Medical Journal, 297, 1304–1308. Ly, T.T., Gallego, P.H., Davis, E.A. & Jones, T.W. (2009) Impaired awareness of hypoglycemia in a population based sample of children and adolescents with type 1 diabetes. Diabetes Care, 32, 1802–1806. Magrath, G., Hartland, B.V. & the Nutrition Subcommittee of the British Diabetic Association’s Profes- sional Advisory Committee (1993) Dietary recommendations for children with diabetes: an implementa- tion paper. Journal of Human Nutrition and Dietetics, 6, 491–507. Marcovecchio, M.L., Dalton, R.N., Prevost, A.T. et al. (2009) Prevalence of abnormal lipid profiles and the relationship with the development of microalbuminuria in adolescents with type 1 diabetes. Diabetes Care, 32, 658–663. Marrero, D.G., Guare, J.C., Vandagriff, J.L. & Fineberg, N.S. (1997) Fear of hypoglycemia in the parents of children and adolescents with diabetes: maladaptive or healthy response? Diabetes Educator, 23, 281–286. Matyka, K.A., Wiggs, L., Pramming, S., Stores, G. & Dunger, D.B. (1999) Cognitive function and mood following profound nocturnal hypoglycaemia during conventional therapy of diabetes. Archives of Disease in Childhood, 81, 138–142. Matyka, K.A., Crawford, C., Wiggs, L., Dunger, D.B. & Stores, G. (2000) Alterations in sleep physiology in young children with insulin dependent diabetes mellitus: relationship to nocturnal hypoglycemia. The Journal of Pediatrics, 137, 233–238. McMahon, S.K., Ferreira, L.D., Ratnam, N. et al. (2007) Glucose requirements to maintain euglycemia after moderate-intensity afternoon exercise in adolescents with type 1 diabetes are increased in a biphasic manner. The Journal of Clinical Endocrinology and Metabolism, 92, 963–968. Mohn, A., Matyka, K.A., Harris, D.A., Ross, K.M., Edge, J.A. & Dunger, D.B. (1999) Lispro or regular insulin for multiple injection therapy in adolescence. Differences in free insulin and glucose levels overnight. Diabetes Care, 22, 27–32. 216 Hypoglycaemia in Clinical Diabetes

Mortensen, H.B., Hougaard, P. & for the Hvidore Study Group on Childhood Diabetes (1997) Comparison of metabolic control in a cross-sectional study of 2,873 children and adolescents with IDDM from 18 countries. Diabetes Care, 20, 714–720. Murphy, N.P., Keane, S.M., Ong, K.K. et al. (2003) Randomized cross-over trial of insulin glargine plus lispro or NPH insulin plus regular human insulin in adolescents with type 1 diabetes on intensive insulin regimens. Diabetes Care, 26, 799–804. National Institute of Clinical Health and Excellence (2008) TA 151 Diabetes – insulin pump therapy: guid- ance National Institute of Clinical Health and Excellence, UK. Nordfeldt, S. & Ludvigsson, J. (1997) Severe hypoglycemia in children with IDDM: a prospective popula- tion study 1992–1994. Diabetes Care, 20, 497–503. Nordfeldt, S. & Ludvigsson, J. (2005) Fear and other disturbances of severe hypoglycemia in children and adolescents with type 1 diabetes. The Journal of Pediatric Endocrinology, 18, 83–91. Nordfeldt, S., Johansson, C., Carlsson, E. & Hammersjo, J.-A. (2003) Prevention of severe hypoglycaemia in type 1 diabetes: a randomised controlled population study. Archives of Disease in Childhood, 88, 240–245. Northam, E.A., Rankins, D., Lin, A. et al. (2009) Central nervous system function in youth with type 1 diabetes 12 years after disease onset. Diabetes Care, 32, 445–450. Patton, S.R., Dolan, L.M., Henry, R. & Powers, S.W. (2008) Fear of hypoglycaemia in parents of children with type 1 diabetes mellitus. Journal of Clinical Psychology in Medical Settings, 15, 252–259. Perantie, D.C., Wu, J., Koller, J.M. et al. (2007) Regional brain volume differences associated with hyper- glycemia and severe hypoglycemia in youth with type 1 diabetes. Diabetes Care, 30, 2331–2337. Pillar, G., Schuscheim, G., Weiss, R. et al. (2003) Interactions between hypoglycemia and sleep architecture in children with type 1 diabetes mellitus. The Journal of Pediatrics, 142, 163–168. Pocecco, M. & Ronfani, L. (1998) Transient focal neurologic deficits associated with hypoglycaemia in children with insulin-dependent diabetes mellitus. Italian Collaborative Paediatric Diabetologic Group. Acta Paediatrica, 87, 542–544. Porter, P., Byrne, G., Stick, S. & Jones, T.W. (1996) Nocturnal hypoglycaemia and sleep disturbances in young teenagers with insulin dependent diabetes mellitus. Archives of Disease in Childhood, 75, 120–123. Porter, P.A., Keating, B., Byrne, G. & Jones, T.W. (1997) Incidence and predictive criteria of nocturnal hypoglycemia in young children with insulin-dependent diabetes mellitus. The Journal of Pediatrics, 130, 366–372. Prins, M.L. (2008) Cerebral metabolic adaptation and ketone metabolism after brain injury. Journal of Cerebral Blood Flow and Metabolism, 28, 1–16. Rewers, A., Chase, H.P., Mackenzie, T. et al. (2002) Predictors of acute complications in children with type 1 diabetes. Journal of the American Medical Association, 287, 2511–2518. Robertson, K., Adolfsson, P., Scheiner, G., Hanas, R. & Riddell, M.C. (2009) Exercise in children and adolescents with diabetes. Pediatric Diabetes, 10, 154–168. Robertson, K.J., Schoenle, E., Gucev, Z., Mordhorst, L., Gall, M.-A. & Ludvigsson, J. (2007) Insulin detemir compared with NPH insulin in children and adolescents with Type 1 diabetes. Diabetic Medicine, 24, 27–34. Rosilio, M.R., Cotton, J.B., Wieliczko, M.C. et al.; on behalf of the French Pediatric Diabetes Group (1998) Factors associated with glycemic control. Diabetes Care, 21, 1146–1153. Ross, L.A., Warren, R.E., Kelnar, C.J.H. & Frier, B.M. (2005) Pubertal stage and hypoglycaemia counter- regulation in type 1 diabetes. Archives of Disease in Childhood, 90, 190–194. Ryan, C., Vega, A. & Drash, A. (1985) Cognitive deficits in adolescents who developed diabetes early in life. Pediatrics, 75, 921–927. Ryan, C.M. (1997) Effects of diabetes mellitus on neuropsychological functioning: a lifespan perspective. Seminars in Clinical Neuropsychiatry, 2, 4–14. Ryan, C.M., Atchison, J., Puczynski, S., Puczynski, M., Arslanian, S. & Becker, D. (1990) Mild hypogly- cemia associated with deterioration of mental efficiency in children with insulin-dependent diabetes mellitus. The Journal of Pediatrics, 117, 32–38. Saudubray, J.M., Marsac, C., Limal, J.M. et al. (1981) Variation in plasma ketone bodies during a 24-hour fast in normal and hypoglycemic children: relationship to age. The Journal of Pediatrics, 98, 904–908. Siebenhofer, A., Plank, J., Berghold, A., Narath, M., Gfrerer, R. & Pieber, T.R. (2006) Short acting insulin analogues versus regular human insulin in patients with diabetes mellitus. Cochrane Database of Sys- tematic Reviews, (2), CD003287. doi: 10.1002/14651858.CD003287.pub4 Chapter 10: Hypoglycaemia in Children with Diabetes 217

Singer-Granick, C., Hoffman, R.P., Kerensky, K., Drash, A.L. & Becker, D.J. (1988) Glucagon responses to hypoglycemia in children and adolescents with IDDM. Diabetes Care, 11, 643–649. Stirling, H.F., Darling, J.A.B. & Kelnar, C.J.H. (1991) Nocturnal glucose homeostasis in normal children. Hormone Research, 35, 54. (abstract). Taplin, C.E., Cobry, E., Messer, L., McFann, K., Chase, H.P. & Fiallo-Scharer, R. (2010) Preventing post- exercise nocturnal hypoglycemia in children with type 1 diabetes. The Journal of Pediatrics, 157, 784–788. The Diabetes Control and Complications Trial Research Group (1993) The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. New England Journal of Medicine, 329, 977–986. The Diabetes Control and Complications Trial Research Group (1994) Effect of intensive diabetes treatment on the development and progression of long-term complications in adolescents with insulin-dependent diabetes mellitus: Diabetes Control and Complications Trial. The Journal of Pediatrics, 125, 177–188. The Diabetes Control and Complications Trial Research Group (1997) Hypoglycemia in the Diabetes Control and Complications Trial. Diabetes, 46, 271–286. Thomsett, M., Shield, G., Batch, J. & Cotterill, A. (1999) How well are we doing? Metabolic control in patients with diabetes. Journal of Paediatrics and Child Health, 35, 479–482. Trumper, B.G., Reschke, K. & Molling, J. (1995) Circadian variation of insulin requirement in IDDM: the relationship between circadian change in insulin demand and diurnal patterns of growth hormone, cor- tisol and glucagon during euglycaemia. Hormone and Metabolic Research, 27, 141–147. Tsalikian, E., Mauras, N., Beck, R.W. et al.; Diabetes Research In Children Network DirecNet Study Group (2005) Impact of exercise on overnight glycemic control in children with type 1 diabetes. The Journal of Pediatrics, 147, 528–534. Tupola, S. & Rajantie, J. (1998) Documented symptomatic hypoglycaemia in children and adolescents using multiple daily injection therapy. Diabetic Medicine, 15, 492–496. Tupola, S., Rajantie, J. & Maenpaa, J. (1998) Severe hypoglycaemia in children and adolescents during multiple-dose insulin therapy. Diabetic Medicine, 15, 695–699. Vanelli, M., Cerutti, F., Chiarelli, F., Lorini, R., Meschi, F. & MCDC–Italy Group (2005) Nationwide cross-sectional survey of 3560 children and adolescents with diabetes in Italy. Journal of Endocrinologi- cal Investigation, 28, 692–699. Ververs, M.T., Rouwe, C. & Smit, G.P. (1993) Complex carbohydrates in the prevention of nocturnal hypoglycaemia in diabetic children. European Journal of Clinical Nutrition, 47, 268–273. Wagner, V.M., Grabert, M. & Holl, R.W. (2005) Severe hypoglycaemia, metabolic control and diabetes management in children with type 1 diabetes in the decade after the Diabetes Control and Complications Trial – a large-scale multicentre study. European Journal of Paediatrics, 164, 73–79. Wessels, A.M., Rombouts, S.A., Simsek, S. et al. (2006) Microvascular disease in type 1 diabetes alters brain activation: a functional magnetic resonance imaging study. Diabetes, 55, 334–340. 11 Hypoglycaemia During Pregnancy in Women with Pregestational Diabetes Lene Ringholm, Peter Damm and Elisabeth R. Mathiesen University of Copenhagen, Copenhagen, Denmark

INTRODUCTION

Pregnancy in patients with known diabetes is particularly challenging for both the woman and the fetus since there is an increased risk of preterm delivery, pre-eclampsia, perinatal mortality, macrosomia, and congenital malformations (Lauenborg et al. 2003; Penney et al. 2003; Evers et al. 2004; Clausen et al. 2005; Ekbom et al. 2008; Lapolla et al. 2008). The current outcomes of pregnancies in women with diabetes still fail to meet the goals of the St Vincent declaration of 1989, which advocated that pregnancy outcomes in women with diabetes should approximate to those in women without diabetes (Saint Vincent Dec- laration 1990).

Higher glycated haemoglobin (HbA1c) levels in early (Lauenborg et al. 2003; Penney et al. 2003; Evers et al. 2004; Lapolla et al. 2008) and late (Ekbom et al. 2008) pregnancy are associated with the risks noted above. There is strong evidence that strict metabolic control both before and during pregnancy is crucial to prevent these adverse outcomes. Striving for near-normoglycaemia increases the risk of severe hypoglycaemia (The Diabetes Control and Complications Trial Research Group 1997), defined as an episode requiring help from another person to restore the blood glucose to normal (American Dia- betes Association 2005). Severe episodes are the main obstacle to good metabolic control and limit the extent to which glycaemic control can be optimised during pregnancy in women with diabetes (Rosenn et al. 1995).

FREQUENCY OF HYPOGLYCAEMIA DURING PREGNANCY

In women with type 1 diabetes, in most studies, severe hypoglycaemia is three to five times more frequent in early pregnancy as in the non-pregnant state (Kimmerle et al. 1992; Evers et al. 2002; Nielsen et al. 2008), whereas in the third trimester, the incidence is lower than in the year preceding pregnancy (Nielsen et al. 2008). In the third trimester, strict metabolic control with a median HbA1c of 5.9% is obtained in most women using an insulin dose around twice that in early pregnancy where median HbA1c was 6.6%

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. This chapter was based on a previous review published in Ringholm, L., Pedersen-Bjergaard, U., Thorsteinsson, B., Damm, P. and Mathiesen, E. R. (2012), Hypoglycaemia during pregnancy in women with Type 1 diabetes. Diabetic Medicine, 29: 558–566. Chapter 11: Hypoglycaemia During Pregnancy in Women with Pregestational Diabetes 219

Number of events 16

14

12

10

8

6

4

2

4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 Gestational week

Figure 11.1 Number of severe hypoglycaemic events per week during pregnancy in women with type 1 diabetes. The first event usually occurs before 20 weeks and 80% of all events are recorded before 20 weeks (Source: Nielsen LR 2008. Reproduced with permission from the American Diabetes Association).

(Nielsen et al. 2008). Up to 45% of women with type 1 diabetes experience severe hypoglycaemia during pregnancy (Kimmerle et al. 1992; Evers et al. 2002; Nielsen et al. 2008). The first event usually occurs before 20 weeks (Rosenn et al. 1995; Nielsen et al. 2008) and up to 84% of all events are recorded before 20 weeks (Kimmerle et al. 1992; Nielsen et al. 2008) (Figure 11.1). The frequency distribution of severe hypoglycaemia during pregnancy is highly skewed (Figure 11.2). Sixty per cent of severe events are accounted for by only 10% of women with type 1 diabetes, despite levels of glycaemic control that are similar to those of women not experiencing severe episodes (Nielsen et al. 2008; Ringholm Nielsen et al. 2008). The subset of women with repeated severe hypoglycaemia presents a major clinical challenge with individual factors making an important contribution in the pregnant (Nielsen et al. 2008; Ringholm Nielsen et al. 2008) as well as the non-pregnant state (Pedersen-Bjergaard et al. 2004). However, research in this population is scarce and there is no consensus on a definition of ‘repeated severe hypoglycaemia’. Despite the higher incidence of severe hypoglycaemia in pregnancy (Evers et al. 2002; Nielsen et al. 2008), the proportion of those episodes occurring at night is generally similar (Mathiesen et al. 2007) to the non-pregnant state (The Diabetes Control and Complications Trial Research Group 1991). A pre-bedtime glucose value below 6.0 mmol/l predicts noc- turnal biochemical hypoglycaemia in the first trimester among pregnant diabetic women (Hellmuth et al. 2000). Interestingly, most nocturnal hypoglycaemic episodes are asymp- tomatic (Hellmuth et al. 2000) suggesting that the nocturnal severe hypoglycaemia events reported by patients (Nielsen et al. 2008) represent the tip of the iceberg of the nocturnal hypoglycaemic burden. 220 Hypoglycaemia in Clinical Diabetes

Number of patients 59

16 15

3 4 4 1 1 1 1 1 1 1

131211109876543210 31 Number of events

Figure 11.2 Frequency distribution of severe hypoglycaemia during pregnancy in women with type 1 diabetes is highly skewed with 10% of the population accounting for 60% of all events (Source: Ringholm Nielsen L 2008. Reproduced with permission from John Wiley & Sons, Ltd).

Evaluated prospectively, mild hypoglycaemia (defined as an event with symptoms rec- ognised by the woman as hypoglycaemia and managed by herself (Pedersen-Bjergaard et al. 2003)), occurs most frequently in early pregnancy (Nielsen et al. 2008). Rates of biochemical hypoglycaemia (blood glucose value ≤3.9 mmol/l (American Diabetes Asso- ciation 2005)) are similar in women with and without severe episodes during pregnancy (Nielsen et al. 2008). Data reporting the incidence of hypoglycaemia in pregnant women with type 2 diabetes are sparse. In a prospective study of 27 women with insulin-treated type 2 diabetes included in early pregnancy, severe episodes were reported by 19% of women during pregnancy (personal communication). Since women with type 2 diabetes maintain both beta cell func- tion and glucose counterregulation, this might offer partial protection against falling glucose levels. However, long-term type 2 diabetes in non-pregnant subjects on insulin therapy was associated with a substantial risk of severe hypoglycaemia during a one-year observation period (Akram et al. 2009).

HORMONAL COUNTERREGULATION

In women with type 1 diabetes, pregnancy per se may impair counterregulatory hormonal responses (Diamond et al. 1992; Rosenn et al. 1996). Hypoglycaemia induced by a hyper- insulinaemic hypoglycaemic clamp failed to elicit an increase in glucagon levels, while the adrenaline (epinephrine) response was markedly lower and the threshold for adrenaline release was decreased compared with that in non-diabetic controls (Diamond et al. 1992). This may render pregnant women with type 1 diabetes more dependent on growth hormone and cortisol (ter Braak et al. 2002). However, only a few studies have investigated hormonal counterregulation to hypoglycaemia during pregnancy in women with type 1 diabetes; all Chapter 11: Hypoglycaemia During Pregnancy in Women with Pregestational Diabetes 221 studies were performed in late pregnancy in a small number of patients (Diamond et al. 1992; Nisell et al. 1994; Rosenn et al. 1996; Bjorklund et al. 1998). No studies have examined the counterregulatory hormonal response with a standardised hyperinsulinaemic hypoglycaemic clamp technique in the first trimester when the frequency of severe hypogly- caemia peaks in women with type 1 diabetes (Evers et al. 2002; Nielsen et al. 2008). However, such studies would raise major ethical concerns. Growth hormone has well-established hyperglycaemic and insulin-antagonistic proper- ties, and may exert its glycaemic effects directly or indirectly through insulin-like growth factor-I (IGF-I) (Caufriez et al. 1990). In young children with growth hormone deficiency, symptomatic hypoglycaemia occurs frequently and disappears after the onset of growth hormone replacement therapy (Bougneres et al. 1985). In pregnant women with type 1 diabetes who have repeated severe hypoglycaemia, lower levels of IGF-I are present throughout pregnancy (Ringholm Nielsen et al. 2008).

BETA CELL FUNCTION

Persisting beta cell function reduces the dependency upon exogenous insulin injection and confers some protection as the suppression of insulin secretion during hypoglycaemia is an important protective mechanism (The Diabetes Control and Complications Trial Research Group 1997). Together with a related preservation of glucagon secretion during hypogly- caemia this may help to mitigate the severity of hypoglycaemia (Madsbad et al. 1982). The loss of residual endogenous insulin increases hypoglycaemic risk in the non-pregnant state (Pedersen-Bjergaard et al. 2001). In non-pregnant individuals with undetectable C-peptide levels, the incidence of severe hypoglycaemia was approximately 2 to 3-fold higher com- pared with patients with detectable C-peptide levels (The Diabetes Control and Complica- tions Trial Research Group 1997; Pedersen-Bjergaard et al. 2001). A pregnancy-induced increase in C-peptide concentrations has been demonstrated in women with long-term type 1 diabetes and good glycaemic control, even in women whose C-peptide concentrations are undetectable in early pregnancy (Pirttiaho et al. 1987; Ilic et al. 2000; Nielsen et al. 2009a). This may be a consequence of rejuvenated beta cell function due to pregnancy-induced growth promoting factors and suppression of the immune system (Ilic et al. 2000), but improved metabolic control may also play a role (Madsbad et al. 1981; Novak et al. 2005; Nielsen et al. 2009a). However, it is unclear whether improved beta cell function in late pregnancy offers protection against hypoglycaemia.

RISK FACTORS FOR SEVERE HYPOGLYCAEMIA DURING PREGNANCY

A history of severe hypoglycaemia in the year preceding pregnancy (Kimmerle et al. 1992; Evers et al. 2002; Nielsen et al. 2008), self-estimated impaired hypoglycaemia awareness, and particularly the combination of these two factors (Nielsen et al. 2008) are important predictors of severe hypoglycaemia (Nielsen et al. 2008), although preserved awareness does not guarantee protection against severe episodes in pregnancy. One prospective study showed that the lower incidence of severe hypoglycaemia in the second part of pregnancy compared with early pregnancy was not explained by changes in women’s perception of hypoglycaemia awareness during pregnancy (Nielsen et al., 2008). 222 Hypoglycaemia in Clinical Diabetes

Pregnancy-induced nausea and vomiting have been suggested as contributing to severe hypoglycaemia in early pregnancy in some studies (Rosenn et al. 1995; Evers et al. 2002). However, nausea and vomiting, when evaluated prospectively five times during pregnancy, did not occur more frequently in women with severe hypoglycaemia during pregnancy (Nielsen et al. 2008). A long duration of diabetes (Rosenn et al. 1995; Evers et al. 2002; Robertson et al.

2009), intensive insulin treatment resulting in lower HbA1c in early pregnancy (Evers et al. 2002), and fluctuating plasma glucose values (Rosenn et al. 1995; Nielsen et al. 2008) appear to contribute to hypoglycaemic risk. However, only 3% of severe events were pre- ceded by recurrent mild hypoglycaemia in one study (Nielsen et al. 2008). No association between severe and mild hypoglycaemia was found (Nielsen et al. 2008), consistent with studies in non-pregnant patients with type 1 diabetes (Pedersen-Bjergaard et al. 2004). Excessive administration of additional insulin between main meals may promote severe hypoglycaemia and has been reported in 14% of severe hypoglycaemic events (Nielsen et al. 2008). In 23%, insufficient carbohydrate intake and/or a postponed meal were identified as possible causes (see Figure 11.3 and Box 11.1) (Nielsen et al. 2008). This suggests that clinicians and patients should ensure adequate food intake and be alert to the risk of sup- plementary fast-acting insulin during pregnancy.

Excessive Insufficient supplementary carbohydrate insulin 14% intake 13%

Postponed meal 10%

Vomiting 2% Many hypos 3% Physical activity 2%

Unknown 56%

Figure 11.3 Causes of severe hypoglycaemia during pregnancy in women with type 1 diabetes. (Source: Data from Ringholm L 2012a).

Box 11.1 Risk factors for severe hypoglycaemia during pregnancy • History of severe hypoglycaemia in the year preceding pregnancy. • Self-estimated impaired hypoglycaemia awareness. • A longer duration of diabetes. • A lower glycated haemoglobin A1c (HbA1c) in early pregnancy. • Fluctuating plasma glucose values. • Excessive supplementary insulin between meals. Chapter 11: Hypoglycaemia During Pregnancy in Women with Pregestational Diabetes 223

MATERNAL CONSEQUENCES OF SEVERE HYPOGLYCAEMIA

Severe hypoglycaemia is unsurprisingly feared in non-pregnant individuals with diabetes (Pramming et al. 1991) and their relatives (Jorgensen et al. 2003). Fear of severe hypogly- caemia has not been evaluated in pregnant women with diabetes, but clinical experience suggests this may be more pronounced during pregnancy (Ringholm et al. 2012a). Uncon- sciousness, seizures and hospitalisation commonly accompany severe episodes during pregnancy (Kimmerle et al. 1992; Evers et al. 2002; Nielsen et al. 2008). Although road traffic accidents (Kimmerle et al. 1992) and maternal death (Leinonen et al. 2001) caused by hypoglycaemia in pregnancy are rare, they emphasise the potential consequences of strict glycaemic control.

HYPOGLYCAEMIA AND PREGNANCY OUTCOME

Animal studies have reported a teratogenic effect of maternal hypoglycaemia leading to congenital malformations (Buchanan et al. 1986) and growth retardation (Smoak and Sadler 1990). The potential risk of maternal severe hypoglycaemia on the fetus has not been evaluated prospectively in clinical studies. However, no positive association has been found between maternal severe hypoglycaemia during pregnancy in women with type 1 diabetes and fetal malformations (Kimmerle et al. 1992; Rosenn et al. 1995; The Diabetes Control and Complications Trial Research Group 1996). Similarly, no human studies have identified an association between maternal severe hypoglycaemia and intrauterine growth retardation. With a 3 to 5-fold increase in the incidence of severe hypoglycaemia in early pregnancy compared with the year preceding pregnancy (Evers et al. 2002; Nielsen et al. 2008), it is reasonable to consider whether blood glucose targets, identified in these recent studies are too strict. However, the frequency of pre-eclampsia and preterm delivery in recent reports is lower (Nielsen et al. 2008, 2009b) compared with previous reports (Evers et al. 2004; Ekbom et al. 2008) and no major congenital malformations have been reported in those women achieving the tightest glycaemic control (Nielsen et al. 2008, 2009b). Reports on the long-term consequences of maternal hypoglycaemia in the offspring are scarce. Rizzo and colleagues reported that children whose mothers experienced severe hypoglycaemia during the second and third trimesters had normal intelligence scores at 2 and 5 years of age (Rizzo et al. 1991). Maternal hypoglycaemia in the first or third trimester has not been shown to predict cognitive function in adult offspring of women with type 1 diabetes (Clausen et al. 2011).

HYPOGLYCAEMIA DURING BREASTFEEDING

Immediately after delivery the mother’s need for insulin declines to approximately 60% of the pre-pregnancy dose owing to lack of placental hormonal influence. There is a concomi- tant increase in risk of severe maternal hypoglycaemia because of fluctuating glucose levels during breastfeeding and excessive insulin dose. Insulin requirements gradually increase over the following weeks (Ringholm et al. 2012b). In women who are breastfeeding, the insulin dose requirement is still around 10% lower than before pregnancy, with wide indi- vidual variations (Stage et al. 2006; Riviello et al. 2009). 224 Hypoglycaemia in Clinical Diabetes

PREVENTION AND CLINICAL MANAGEMENT

It appears that during pregnancy, hyperglycaemia is more harmful to the fetus than hypogly- caemia. Therefore, it may from a fetal perspective be unwise to aim for less strict metabolic control in early pregnancy purely to prevent severe hypoglycaemia, and therefore additional ways of avoiding hypoglycaemia must be sought. Early identification of women at increased risk is important and should be basedon estimation of the number of severe hypoglycaemic events in the year preceding pregnancy and women’s self-estimated hypoglycaemia awareness. This knowledge can help to esti- mate the individual risk of severe episodes (Nielsen et al. 2008) and be used to tailor special education and individual modification of glucose monitoring, diet and insulin therapy for each woman (Box 11.2). Insulin requirements often decline late in the first trimester (Jovanovic et al. 2001; Garcia-Patterson et al. 2010), and overinsulinisation may contribute to hypoglycaemic risk (Atkin et al. 1996; Jovanovic et al. 2001; Evers et al. 2002). The insulin dose often needs to be reduced by around 10% at 10–16 weeks when the risk of severe hypoglycaemia is highest (Rosenn et al. 1995; Evers et al. 2002; Nielsen et al. 2008; Robertson et al. 2009; Ringholm et al. 2012a). The second and third trimesters are characterised by increasing insulin resistance and an increasing insulin requirement in women with type 1 diabetes (Garcia-Patterson et al. 2010). In the third trimester, despite large insulin doses, the inci- dence of severe hypoglycaemia is lower than the incidence outside of pregnancy and episodes mainly occur in women with repeated severe hypoglycaemia before, and during early pregnancy (Nielsen et al. 2008; Ringholm Nielsen et al. 2008). This may permit strict metabolic control to be applied safely in the second part of pregnancy in many women. This applies particularly to those without a history of severe hypoglycaemia before 20 weeks, as their risk is low (Nielsen et al. 2008).

Pre-conception counselling is associated with earlier referral, lower HbA1c at the first visit, and lower occurrence of very preterm deliveries, but a beneficial effect on the rate of severe hypoglycaemia during pregnancy has not generally been demonstrated (Temple et al. 2006; Robertson et al. 2009). Focusing on hypoglycaemia awareness training, carbo- hydrate counting and appropriate corrective insulin seems sensible, but should be formally evaluated.

Box 11.2 Preventive measures to reduce the risk of severe hypoglycaemia during pregnancy • Identify high-risk patients early, particularly women with a history of severe hypogly- caemia the year preceding pregnancy and/or self-estimated impaired hypoglycaemia awareness. • Reduce insulin dose by around 10% at 10–16 weeks. • Use supplementary insulin cautiously in early pregnancy. • Carry oral glucose. • Use rapid- and long-acting insulin analogues. • Avoid pre-bedtime blood glucose values below 6.0 mmol/l. • Perform frequent blood glucose monitoring, including between 02:00–04:00 h. • Make glucagon available for administration at home by partner/family members. • Use continuous subcutaneous insulin infusion (with insulin pump) therapy and con- tinuous glucose monitoring. Chapter 11: Hypoglycaemia During Pregnancy in Women with Pregestational Diabetes 225

In non-pregnant individuals with type 1 diabetes, rapid- and long-acting insulin ana- logues are associated with a trend towards fewer severe hypoglycaemic events (Heller et al. 2004; Monami et al. 2009). Likewise, during pregnancy, studies of rapid-acting insulin analogues (Mathiesen et al. 2007; Heller et al. 2010; Garcia-Dominguez et al. 2011) and of long-acting insulin analogues (Poyhonen-Alho et al. 2007; Gallen et al. 2008; Imbergamo et al. 2008; Lapolla et al. 2009) have shown a tendency towards less frequent hypoglycaemia compared with human insulin. In a randomised study comparing the effect of insulin detemir and human NPH (isophane) insulin in combination with insulin aspart, fasting blood glucose values were modestly lower without an increase in severe hypogly- caemia (Mathiesen et al. 2012), while HbA1c levels did not differ. Selected individuals with severe hypoglycaemia during pregnancy, particularly those with a history of severe hypoglycaemia in the year preceding pregnancy and impaired hypoglycaemia awareness, may benefit from continuous subcutaneous insulin infusion (insulin pump) therapy (Coustan et al. 1986; Chico et al. 2010). This has been combined with continuous glucose monitoring to record periods of hypoglycaemia and glycaemic fluctuations (Kerssen et al. 2006; Worm et al. 2006; Murphy et al. 2007, 2008). Anecdotal reports have indicated the potential benefits of insulin pump therapy combined with real- time continuous glucose monitoring with alarms for low glucose values in preventing severe hypoglycaemia in pregnant women with impaired hypoglycaemia awareness among selected individuals (Worm et al. 2006). A randomised study of 123 unselected women with type 1 diabetes did not show a protective effect of real-time continuous glucose moni- toring on severe hypoglycaemia during pregnancy (Secher et al. 2013).

TREATMENT

Women and their family members should receive oral and written information about changes in insulin dose that are anticipated during pregnancy, including information about the high risk and treatment of severe hypoglycaemia during pregnancy, especially in the first trimester (Figure 11.4). A glucagon kit should be prescribed to all pregnant diabetic

None 1% Friend/colleague 4%

Family 5% Partner 75%

Ambulance/hospital staff 15%

Figure 11.4 Persons giving treatment for severe hypoglycaemia in pregnant women with type 1 diab­etes. 226 Hypoglycaemia in Clinical Diabetes women taking rapid-acting insulin. Mild hypoglycaemia should be treated with oral glucose; based on clinical experience 20 g administered as glucose tablets, a glucose drink or fruit juice is sufficient to restore normal glucose values without subsequent hyperglycaemia occurring through overcorrection. Severe hypoglycaemia may be treated with oral carbohydrate. If this is not sufficient to obtain recovery, 1 mg of glucagon can be administered by intramuscular injection at home by the partner or other family members. Recovery usually occurs within 15 minutes. If required, intravenous glucose is usually given as 20–50 ml of a 20% solution and produce recovery within minutes.

REFERENCES

Akram, K., Pedersen-Bjergaard, U., Carstensen, B., Borch-Johnsen, K. & Thorsteinsson, B. (2009) Pro- spective and retrospective recording of severe hypoglycaemia, and assessment of hypoglycaemia aware- ness in insulin-treated Type 2 diabetes. Diabetic Medicine, 26, 1306–1308 (letter). American Diabetes Association Workgroup on Hypoglycemia (2005) Defining and reporting hypoglycemia in diabetes: a report from the American Diabetes Association Workgroup on Hypoglycemia. Diabetes Care, 28, 1245–1249. Atkin, S.L., Lindow, S.W., Walton, C. & Masson, E.A. (1996) Recurrent hypoglycaemia associated with first and second trimester insulin sensitivity in a patient with insulin-dependent diabetes.Diabetic Medic­ ine, 13, 589–591. Bjorklund, A., Adamson, U., Andreasson, K. et al. (1998) Hormonal counterregulation and subjective symptoms during induced hypoglycemia in insulin-dependent diabetes mellitus patients during and after pregnancy. Acta Obstetricia et Gynecologica Scandinavica, 77, 625–634. Bougneres, P.F., Artavia-Loria, E., Ferre, P., Chaussain, J.L. & Job, J.C. (1985) Effects of hypopituitarism and growth hormone replacement therapy on the production and utilization of glucose in childhood. Journal of Clinical Endocrinology and Metabolism, 61, 1152–1157. ter Braak, E.W., Evers, I.M., Willem, E.D. & Visser, G.H. (2002) Maternal hypoglycemia during pregnancy in type 1 diabetes: maternal and fetal consequences. Diabetes/Metabolism Research and Reviews, 18, 96–105. Buchanan, T.A., Schemmer, J.K. & Freinkel, N. (1986) Embryotoxic effects of brief maternal insulin-hypoglycemia during organogenesis in the rat. The Journal of Clinical Investigation, 78, 643–649. Caufriez, A., Frankenne, F., Englert, Y. et al. (1990) Placental growth hormone as a potential regulator of maternal IGF-I during human pregnancy. The American Journal of Physiology, 258, E1014–E1019. Chico, A., Saigi, I., Garcia-Patterson, A. et al. (2010) Glycemic control and perinatal outcomes of pregnan- cies complicated by type 1 diabetes: influence of continuous subcutaneous insulin infusion and lispro insulin. Diabetes Technology and Therapeutics, 12, 937–945. Clausen, T.D., Mathiesen, E., Ekbom, P., Hellmuth, E., Mandrup-Poulsen, T. & Damm, P. (2005) Poor pregnancy outcome in women with type 2 diabetes. Diabetes Care, 28, 323–328. Clausen, T.D., Mortensen, E.L., Schmidt, L. et al. (2011) Cognitive function in adult offspring of women with Type 1 diabetes. Diabetic Medicine, 28, 838–844. Coustan, D.R., Reece, E.A., Sherwin, R.S. et al. (1986) A randomized clinical trial of the insulin pump vs intensive conventional therapy in diabetic pregnancies. Journal of the American Medical Association, 255, 631–636. Diamond, M.P., Reece, E.A., Caprio, S. et al. (1992) Impairment of counterregulatory hormone responses to hypoglycemia in pregnant women with insulin-dependent diabetes mellitus. American Journal of Obstetrics and Gynecology, 166, 70–77. Ekbom, P., Damm, P., Feldt-Rasmussen, B., Feldt-Rasmussen, U., Jensen, D.M. & Mathiesen, E.R. (2008) Elevated third-trimester haemoglobin A(1c) predicts preterm delivery in type 1 diabetes. Journal of Diabetes and its Complications, 22, 297–302. Evers, I.M., ter Braak, E.W.M.T., de Valk, H.W., van der Schoot, B., Janssen, N. & Visser, G.H.A. (2002) Risk indicators predictive for severe hypoglycemia during the first trimester of type 1 diabetic pregnancy. Diabetes Care, 25, 554–559. Chapter 11: Hypoglycaemia During Pregnancy in Women with Pregestational Diabetes 227

Evers, I.M., de Valk, H.W. & Visser, G.H.A. (2004) Risk of complications of pregnancy in women with type 1 diabetes: nationwide prospective study in the Netherlands. British Medical Journal, 328, 915. Gallen, I.W., Jaap, A., Roland, J.M. & Chirayath, H.H. (2008) Survey of glargine use in 115 pregnant women with Type 1 diabetes. Diabetic Medicine, 25, 165–169. Garcia-Dominguez, M., Herranz, L., Hillman, N. et al. (2011) Use of insulin lispro during pregnancy in women with pregestational diabetes mellitus. Medicina Clinica (Barcelona), 137, 581–586. Garcia-Patterson, A., Gich, I., Amini, S.B., Catalano, P.M., de Leiva, A. & Corcoy, R. (2010) Insulin requirements throughout pregnancy in women with type 1 diabetes mellitus: three changes of direction. Diabetologia, 53, 446–451. Heller, S.R., Colagiuri, S., Vaaler, S. et al. (2004) Hypoglycaemia with insulin aspart: a double-blind, randomised, crossover trial in subjects with Type 1 diabetes. Diabetic Medicine, 21, 769–775. Heller, S., Damm, P., Mersebach, H. et al. (2010) Hypoglycemia in type 1 diabetic pregnancy: role of preconception insulin aspart treatment in a randomized study. Diabetes Care, 33, 473–477. Hellmuth, E., Damm, P., Molsted-Pedersen, L. & Bendtson, I. (2000) Prevalence of nocturnal hypoglyc- emia in first trimester of pregnancy in patients with insulin treated diabetes mellitus. Acta Obstetricia et Gynecologica Scandinavica, 79, 958–962. Ilic, S., Jovanovic, L. & Wollitzer, A.O. (2000) Is the paradoxical first trimester drop in insulin requirement due to an increase in C-peptide concentration in pregnant Type I diabetic women? Diabetologia, 43, 1329–1330. Imbergamo, M.P., Amato, M.C., Sciortino, G. et al. (2008) Use of glargine in pregnant women with type 1 diabetes mellitus: a case-control study. Clinical Therapeutics, 30, 1476–1484. Jorgensen, H.V., Pedersen-Bjergaard, U., Rasmussen, A.K. & Borch-Johnsen, K. (2003) The impact of severe hypoglycemia and impaired awareness of hypoglycemia on relatives of patients with type 1 diabetes. Diabetes Care, 26, 1106–1109. Jovanovic, L., Knopp, R.H., Brown, Z. et al. (2001) Declining insulin requirement in the late first trimester of diabetic pregnancy. Diabetes Care, 24, 1130–1136. Kerssen, A., de Valk, H.W. & Visser, G.H.A. (2006) Do HbA1c levels and the self-monitoring of blood glucose levels adequately reflect glycaemic control during pregnancy in women with type 1 diabetes mellitus? Diabetologia, 49, 25–28. Kimmerle, R., Heinemann, L., Delecki, A. & Berger, M. (1992) Severe hypoglycemia incidence and pre- disposing factors in 85 pregnancies of type I diabetic women. Diabetes Care, 15, 1034–1037. Lapolla, A., Dalfra, M.G., Spezia, R. et al. (2008) Outcome of pregnancy in type 1 diabetic patients treated with insulin lispro or regular insulin: an Italian experience. Acta Diabetologica, 45, 61–66. Lapolla, A., Di Cianni, G., Bruttomesso, D. et al. (2009) Use of insulin detemir in pregnancy: a report on 10 Type 1 diabetic women. Diabetic Medicine, 26, 1181–1182. Lauenborg, J., Mathiesen, E., Ovesen, P. et al. (2003) Audit on stillbirths in women with pregestational type 1 diabetes. Diabetes Care, 26, 1385–1389. Leinonen, P.J., Hiilesmaa, V.K., Kaaja, R.J. & Teramo, K.A. (2001) Maternal mortality in type 1 diabetes. Diabetes Care, 24, 1501–1502. Madsbad, S., Krarup, T., Reguer, L., Faber, O.K. & Binder, C. (1981) Effect of strict blood glucose control on residual B-cell function in insulin-dependent diabetics. Diabetologia, 20, 530–534. Madsbad, S., Hilsted, J., Krarup, T. et al. (1982) Hormonal, metabolic and cardiovascular responses to hypoglycaemia in Type 1 (insulin-dependent) diabetes with and without residual B cell function. Dia­ betologia, 23, 499–503. Mathiesen, E.R., Kinsley, B., Amiel, S.A. et al. (2007) Maternal glycemic control and hypoglycemia in type 1 diabetic pregnancy: a randomized trial of insulin aspart versus human insulin in 322 pregnant women. Diabetes Care, 30, 771–776. Mathiesen, E.R., Hod, M., Ivanisevic, M. et al. (2012) Maternal efficacy and safety outcomes in a rand- omized, controlled trial comparing insulin detemir with NPH insulin in 310 pregnant women with type 1 diabetes. Diabetes Care, 35, 2012–2017. Monami, M., Marchionni, N. & Mannucci, E. (2009) Long-acting insulin analogues vs. NPH human insulin in type 1 diabetes. A meta-analysis. Diabetes, Obesity and Metabolism, 11, 372–378. Murphy, H.R., Rayman, G., Duffield, K. et al. (2007) Changes in the glycemic profiles of women with type 1 and type 2 diabetes during pregnancy. Diabetes Care, 30, 2785–2791. Murphy, H.R., Rayman, G., Lewis, K. et al. (2008) Effectiveness of continuous glucose monitoring in pregnant women with diabetes: randomised clinical trial. British Medical Journal, 337, a1680. 228 Hypoglycaemia in Clinical Diabetes

Nielsen, L.R., Pedersen-Bjergaard, U., Thorsteinsson, B., Johansen, M., Damm, P. & Mathiesen, E.R. (2008) Hypoglycemia in pregnant women with type 1 diabetes: predictors and role of metabolic control. Diabetes Care, 31, 9–14. Nielsen, L.R., Rehfeld, J.F., Pedersen-Bjergaard, U., Damm, P. & Mathiesen, E.R. (2009a) Pregnancy- induced rise in serum C-peptide concentrations in women with type 1 diabetes. Diabetes Care, 32, 1052–1057. Nielsen, L.R., Damm, P. & Mathiesen, E.R. (2009b) Improved pregnancy outcome in type 1 diabetic women with microalbuminuria or diabetic nephropathy: effect of intensified antihypertensive therapy? Diabetes Care, 32, 38–44. Nisell, H., Persson, B., Hanson, U. et al. (1994) Hormonal, metabolic, and circulatory responses to insulin- induced hypoglycemia in pregnant and nonpregnant women with insulin-dependent diabetes. American Journal of Perinatology, 11, 231–236. Novak, E.J., Ortqvist, E., Nord, E. et al. (2005) Stability of disease-associated antibody titers in pregnant women with type 1 diabetes with or without residual beta-cell function. Diabetes Care, 23, 1019–1021. Pedersen-Bjergaard, U., Agerholm-Larsen, B., Pramming, S., Hougaard, P. & Thorsteinsson, B. (2001) Activity of angiotensin-converting enzyme and risk of severe hypoglycaemia in type 1 diabetes mellitus. Lancet, 357, 1248–1253. Pedersen-Bjergaard, U., Pramming, S. & Thorsteinsson, B. (2003) Recall of severe hypoglycaemia and self-estimated state of awareness in type 1 diabetes. Diabetes/Metabolism Research and Reviews, 19, 232–240. Pedersen-Bjergaard, U., Pramming, S., Heller, S.R. et al. (2004) Severe hypoglycaemia in 1076 adult patients with type 1 diabetes: influence of risk markers and selection. Diabetes/Metabolism Research and Reviews, 20, 479–486. Penney, G.C., Mair, G. & Pearson, D.W.M. (2003) Outcomes of pregnancies in women with type 1 diabetes in Scotland: a national population-based study. British Journal of Obstetrics and Gynaecology, 110, 315–318. Pirttiaho, H.I., Hartikainen-Sorri, A.L., Kaapa, P., Kaila, J.M. & Puukka, R. (1987) Maternal residual beta- cell function and the outcome of diabetic pregnancy. Journal of Perinatal Medicine, 15, 83–89. Poyhonen-Alho, M., Ronnemaa, T., Saltevo, J., Ekblad, U. & Kaaja, R.J. (2007) Use of insulin glargine during pregnancy. Acta Obstetricia et Gynecologica Scandinavica, 86, 1171–1174. Pramming, S., Thorsteinsson, B., Bendtson, I. & Binder, C. (1991) Symptomatic hypoglycaemia in 411 type 1 diabetic patients. Diabetic Medicine, 8, 217–222. Ringholm, L., Pedersen-Bjergaard, U., Thorsteinsson, B., Damm, P. & Mathiesen, E.R. (2012a) Hypogly- caemia during pregnancy in women with type 1 diabetes. Diabetic Medicine, 29, 558–566. Ringholm, L., Mathiesen, E.R., Kelstrup, L. & Damm, P. (2012b) Glycemic control and antihypertensive therapy in women with type 1 diabetes: from pregnancy planning to the breastfeeding period. Nature Reviews Endocrinology, 8, 659–667. Ringholm Nielsen, L., Juul, A., Pedersen-Bjergaard, U., Thorsteinsson, B., Damm, P. & Mathiesen, E.R. (2008) Lower levels of circulating IGF-I in Type 1 diabetic women with frequent severe hypoglycaemia during pregnancy. Diabetic Medicine, 25, 826–833. Riviello, C., Mello, G. & Jovanovic, L.G. (2009) Breastfeeding and the basal insulin requirement in type 1 diabetic women. Endocrine Practice, 15, 187–193. Rizzo, T., Metzger, B.E., Burns, W.J. & Burns, K. (1991) Correlations between antepartum maternal metabolism and intelligence of offspring. New England Journal of Medicine, 325, 911–916. Robertson, H., Pearson, D.W.M. & Gold, A.E. (2009) Severe hypoglycaemia during pregnancy in women with Type 1 diabetes is common and planning pregnancy does not decrease the risk. Diabetic Medicine, 26, 824–826. Rosenn, B.M., Miodovnik, M., Holcberg, G., Khoury, J.C. & Siddiqi, T.A. (1995) Hypoglycemia: the price of intensive insulin therapy for pregnant women with insulin-dependent diabetes mellitus. Obstetrics and Gynecology, 85, 417–422. Rosenn, B.M., Miodovnik, M., Khoury, J.C. & Siddiqi, T.A. (1996) Counterregulatory hormonal responses to hypoglycemia during pregnancy. Obstetrics and Gynecology, 87, 568–574. Saint Vincent Declaration (1990) Diabetes care and research in Europe: the Saint Vincent declaration. Diabetic Medicine, 7, 360. Secher, A.L., Ringholm, L., Andersen, H.U., Damm, P. & Mathiesen, E.R. (2013) The effect of real-time continuous glucose monitoring in pregnant women with diabetes: a randomized controlled trial. Diabetes Care, 36, 1877–1883. Chapter 11: Hypoglycaemia During Pregnancy in Women with Pregestational Diabetes 229

Smoak, I.W. & Sadler, T.W. (1990) Embryopathic effects of short-term exposure to hypoglycemia in mouse embryos in vitro. American Journal of Obstetrics and Gynecology, 163, 619–624. Stage, E., Norgard, H., Damm, P. & Mathiesen, E. (2006) Long-term breast-feeding in women with type 1 diabetes. Diabetes Care, 29, 771–774. Temple, R.C., Aldridge, V.J. & Murphy, H.R. (2006) Prepregnancy care and pregnancy outcomes in women with type 1 diabetes. Diabetes Care, 29, 1744–1749. The Diabetes Control and Complications Trial Research Group (1991) Epidemiology of severe hypoglyc- emia in the Diabetes Control and Complications Trial. The American Journal of Medicine, 90, 450–459. The Diabetes Control and Complications Trial Research Group (1996) Pregnancy outcomes in the Diabetes Control and Complications Trial. American Journal of Obstetrics and Gynecology, 174, 1343–1353. The Diabetes Control and Complications Trial Research Group (1997) Hypoglycemia in the Diabetes Control and Complications Trial. Diabetes, 46, 271–286. Worm, D., Nielsen, L.R., Mathiesen, E.R. & Norgaard, K. (2006) Continuous glucose monitoring system with an alarm: a tool to reduce hypoglycemic episodes in pregnancy with diabetes. Diabetes Care, 29, 2759–2760 (letter). 12 Hypoglycaemia in Type 2 Diabetes and in Elderly People Nicola N. Zammitt1 and Brian M. Frier2 1 Royal Infirmary of Edinburgh, Edinburgh, UK 2 University of Edinburgh, Edinburgh, UK

INTRODUCTION

Hypoglycaemia is thought to be related almost exclusively to the management of type 1 diabetes and has received limited attention in the context of type 2 diabetes. However, as the prevalence of type 2 diabetes escalates and life expectancy increases, it is inevitable that the number of older people with insulin-treated diabetes is steadily increasing. Fur- thermore, the current emphasis on adherence to stricter glycaemic targets is influencing the prevalence and severity of hypoglycaemia in this group. Because hypoglycaemia is increas- ing as a clinical problem in people with type 2 diabetes, many of whom are elderly, it is pertinent to consider how age and type 2 diabetes per se affect the pathophysiology and frequency of hypoglycaemia.

PATHOPHYSIOLOGY OF HYPOGLYCAEMIA

Most studies that have examined counterregulatory responses to hypoglycaemia have been conducted in young adults with, and without, type 1 diabetes. The pathophysiological differences that exist between type 1 and type 2 diabetes have the potential to modify counterregulatory mechanisms. The increasing age of the population with diabetes will also influence counterregulation. The responses to hypoglycaemia can be considered in terms of the generation of symptom responses and of counterregulatory hormonal changes. The two are interdepend- ent, although some hormones such as epinephrine (adrenaline) can contribute to symptom generation; both responses are mediated centrally through glucose sensors in the brain. It is convenient to examine symptoms and counterregulation separately.

The effects of ageing on the responses to hypoglycaemia Symptoms The symptoms of hypoglycaemia (Chapter 2) have been classified as autonomic, neuro- glycopenic and non-specific (Dearyet al. 1993). In older and elderly people (aged 70 years

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 231

Table 12.1 Common symptoms of hypoglycaemia in the elderly as derived by Principal Components Analysis. (Source: Adapted from Jaap AJ 1998, with permission from John Wiley & Sons, Ltd).

Neuroglycopenic Autonomic Neurological

Poor concentration Sweating Unsteadiness Lightheadedness Shaking Poor coordination Weakness Anxiety Blurred vision Confusion Pounding heart Double vision Drowsiness Slurred speech Dizziness

or older) who have diabetes, an additional group of neurological symptoms affecting vision and coordination has been identified (Jaap et al. 1998; McAulay et al. 2001) (Table 12.1) and non-specific symptoms of general malaise are less prominent than in young adults. These neurological symptoms that are generated as manifestations of hypoglycaemia in elderly people may be misinterpreted as features of cerebrovascular disease, such as tran- sient ischaemic attacks or vaso-vagal episodes, or as features of neurological or psychiatric disorders. Many elderly people with type 2 diabetes have limited knowledge of the symptoms and treatment of hypoglycaemia (Pegg et al. 1991; Thomson et al. 1991) and their relatives and carers are equally uninformed (Mutch and Dingwall-Fordyce 1985). This heightens the risk that early hypoglycaemia will neither be identified nor treated, and will therefore progress to the point where external help is required. Structured education can achieve improved glycaemic control without increasing the frequency of hypoglycaemia. For example, in the INITIATEplus study, 4875 insulin-naive patients with type 2 diabetes were commenced on twice daily biphasic insulin and randomised to receive varying levels of telephone counselling from a dietitian: either no counselling, one session or three sessions.

Similar improvements in HbA1c were achieved in all groups with the intensive counselling group experiencing less hypoglycaemia (4 vs 9 episodes of severe hypoglycaemia per 100 patient years in the group receiving intensive counselling versus those with no counselling) (Oyer et al. 2009). However, even in young adults, knowledge of diabetes and its treatment declines with time (Lawrence and Cheely 1980), and in elderly people the need for regular educational reinforcement is greater, although seldom undertaken. The increasing prevalence of cognitive impairment with age may also affect the identi- fication of hypoglycaemia and its treatment. In a study of adults aged 75 and above, more than one-fifth of those with diabetes had cognitive impairment (Mini Mental State Exami- nation [MMSE] score ≤26) and, when questioned on the use of carbohydrate in the treat- ment of hypoglycaemia, 30% of these made mistakes compared to 7% of those with normal cognitive function. Furthermore, 51% of those with cognitive impairment made errors, such as attempting to increase their treatment after hypoglycaemia, compared to 22% of those with normal MMSE scores (Hewitt et al. 2011). In the ADVANCE study the participants with MMSE scores of <24 (‘severe’ dysfunction, n = 212) had a hazard ratio of developing severe hypoglycaemia of 2.1 (95% CI 1.14–3.87, P = 0.018 compared to those with MMSE scores of 28–30 (n = 8689) (de Galan et al. 2009). Recognition of hypoglycaemia in the elderly is influenced by factors other than cogni- tive impairment and level of education. Elderly people report a profile of hypoglycaemic 232 Hypoglycaemia in Clinical Diabetes symptoms that is different from that of young adults, and the intensity of their symptoms is less (Brierley et al. 1995; Matyka et al. 1997), with autonomic and neuroglycopenic symptoms being affected equally (Brierley et al. 1995). Two Canadian studies have impli- cated diminished autonomic activation as the reason for the lower magnitude of sympto- matic response in the elderly and have suggested that the attenuation in symptom intensity is a feature of increasing age, independent of any effects of diabetes (Meneilly et al. 1994a, 1994b). The interaction between age and awareness of hypoglycaemia was examined in a clamp study that compared 13 adults with type 2 diabetes aged between 39 and 64 years with 13 patients aged 65 or older. The hypoglycaemic symptoms were blunted in the older group, only one of whom correctly identified when their blood glucose was <3.3 mmol/l (Bremer et al. 2009). The glycaemic thresholds at which symptomatic responses to hypoglycaemia are gener- ated (Chapter 6) alter with age. In young adults, symptoms are generated at a blood glucose level that is approximately 1.0 mmol/l higher than the level at which cognitive dysfunction becomes apparent (Schwartz et al. 1987). However, in older subjects these thresholds are much closer together and occur almost simultaneously (around 3.0 ± 0.2 mmol/l) (Matyka et al. 1997) (Figure 12.1), which limits the time available for self-treatment before the development of potentially incapacitating neuroglycopenia. Thus, in older people the symptom profile of hypoglycaemia is different’ with neurologi- cal symptoms having greater prominence. At the same time, symptoms appear to be of a lesser intensity and are experienced at lower glucose levels. This altered symptomatic response to hypoglycaemia, combined with the presence of co-morbidities, such as mild cognitive impairment, increases the risk of developing hypoglycaemia of greater severity.

Arterialised blood glucose (mmol/l)

4.0

Symptoms Younger men 3.5 (3.6 mmol/l)

Symptoms 3.0 Older men (3.0 mmol/l) Reaction time

2.5 Four choice Younger men reaction time (2.6 mmol/l)

2.0

Figure 12.1 Glycaemic thresholds for subjective symptomatic awareness of hypoglycaemia and for the onset of cognitive dysfunction in young and elderly non-diabetic males. (Source: Data derived from Matyka K 1997. With permission from John Wiley & Son’s, Ltd). Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 233

Counterregulation Early studies examining counterregulatory responses in elderly non-diabetic adults yielded conflicting results, partly because of the confounding effect of co-morbidities inmany participants. In addition, the age cut-offs used to define an elderly population differ con- siderably between studies. However, subsequent studies have suggested that the nature and intensity of counterregulatory responses may change with age. When hypoglycaemia was induced with an intravenous infusion of insulin to compare counterregulatory responses in older non-diabetic adults (65.1 ± 0.9 years old) and young adults (23.8 ± 0.6 years old), the secretion of growth hormone and cortisol diminished with advancing age (Marker et al. 1992). A modest attenuation of blood glucose recovery was observed in older adults, in whom the rate of rise in plasma epinephrine was slower than in the younger subjects (Marker et al. 1992). The secretion of glucagon and the rate of insulin clearance were also lower. A reduced rate of clearance of insulin has been noted in several studies (Minaker et al. 1982; Reaven et al. 1982; Fink et al. 1985). The changes observed by Marker and colleagues were unaffected by preceding physical training, suggesting that they are not simply a consequence of a more sedentary lifestyle associated with ageing (Marker et al. 1992). The glycaemic thresholds for the secretion of glucagon and epinephrine in response to hypoglycaemia also occur at a lower blood glucose level than in younger subjects. In young non-diabetic adults (aged <30 years), these hormones are released at a blood glucose level of 3.3 mmol/l, compared to approximately 2.8 mmol/l in older adults (aged >65 years) (Meneilly et al. 1994a). With increasing age, the blood glucose nadir appears to influence the magnitude of the counterregulatory response. In clamp studies comparing elderly and young non-diabetic subjects, the magnitude of the glucagon and epinephrine responses was lower in the elderly group during mild hypoglycaemia (blood glucose 3.3 mmol/l). However, both age groups achieved a similar magnitude of response at a blood glucose of 2.8 mmol/l, indicating that in the elderly, counterregulatory responses are preserved during more profound hypogly- caemia (Ortiz-Alonso et al. 1994). Two other similar studies in non-diabetic elderly subjects have not demonstrated any significant age-related impairment of the counterregulatory hormonal responses to hypogly- caemia (Brierley et al. 1995; Matyka et al. 1997). Furthermore, although the symptomatic and counterregulatory responses to hypoglycaemia may be modified by advancing age, it is not known at what age these changes become apparent, whether this varies between individuals and whether these effects are influenced by gender or the menopause in women. When designing studies and analysing results, few investigations of hypoglycaemia in type 2 diabetes have taken age or gender into account and few studies have included people aged over 70.

The effects of type 2 diabetes on the responses to hypoglycaemia Symptoms Type 2 diabetes is a heterogeneous disorder, which can be treated with several different classes of drug. This raises the possibility that the agent used to treat the condition could influence the symptoms of hypoglycaemia. An early study in healthy volunteers compared the symptomatic responses to hypoglycaemia induced with insulin and a sulfonylurea (tolbutamide) and found no differences in the nature and intensity of symptoms (Peacey 234 Hypoglycaemia in Clinical Diabetes et al. 1996). However, a study using a stepped glucose clamp has suggested that when patients with type 2 diabetes are treated with sulfonylureas, they experience more intense hypoglycaemic symptoms, which occur at higher blood glucose concentrations than in people with insulin-treated type 2 diabetes, even when the two groups are matched for glycaemic control and duration of diabetes (Choudhary et al. 2009). While this may offer some protection from severe hypoglycaemia it may also discourage strict glycaemic control. In a retrospective population survey, people with insulin-treated type 2 diabetes reported a similar symptom profile to people with type 1 diabetes, who were matched for duration of insulin therapy, but not for age or duration of diabetes (Hepburn et al. 1993). Further- more, glucose clamp studies have shown that after controlling for age, no significant dif- ferences in the symptoms of hypoglycaemia exist between people with type 1 and type 2 diabetes (Levy et al. 1998). The symptoms of hypoglycaemia in type 2 diabetes appear to be unaffected by the nature of the disorder but may be affected by the choice of treatment modality.

Counterregulation Early studies of the counterregulatory responses to hypoglycaemia in type 2 diabetes were limited by technical factors (de Galan and Hoekstra 2001). Three later studies that exam- ined counterregulatory responses to hypoglycaemia in type 2 diabetes demonstrated that the secretion of counterregulatory hormones occurs at higher blood glucose levels in indi- viduals with type 2 diabetes than in non-diabetic subjects (Korzon-Burakowska et al. 1998; Spyer et al. 2000) and in people with type 1 diabetes (Levy et al. 1998). In one study, a stepped glucose clamp was used to examine how glycaemic control influences the counter- regulatory response to hypoglycaemia in 11 subjects with type 2 diabetes who were either diet-treated or were taking sulfonylureas, and they were compared with 10 subjects with type 1 diabetes (Levy et al. 1998). The group with type 2 diabetes was older than the group with type 1 diabetes, so two non-diabetic control groups were matched for age and body weight to the diabetic subjects. In the subjects with type 2 diabetes, counterregulatory hormones were released in response to stepped hypoglycaemia at higher blood glucose levels than in those with type 1 diabetes. However, males were over-represented in the group with type 2 diabetes, and six of the subjects with type 2 diabetes required high insulin infusion rates, so introducing potentially confounding variables and limiting interpretation of the results of this study. There is evidence that hyperinsulinaemia per se can suppress the release of glucagon in response to hypoglycaemia (Diamond et al. 1991; Liu et al. 1991a; Liu et al. 1991b; Fanelli et al. 1994) while increasing catecholamine and cortisol release (Davis et al. 1993) in non-diabetic and type 1 diabetic subjects. Similarly, in both non-diabetic and type 1 diabetic subjects, women have attenuated counterregulatory responses to hypoglycaemia compared to men (Amiel et al. 1993; Davis et al. 2000; San- doval et al. 2003). The effects of gender and hyperinsulinaemia on counterregulation have not yet been determined in type 2 diabetes. People with type 2 diabetes may have greater protection against hypoglycaemia because the counterregulatory responses commence at higher blood glucose levels than in non- diabetic people or in those with type 1 diabetes (Korzon-Burakowska et al. 1998; Levy et al. 1998; Spyer et al. 2000). However, when HbA1c is lowered with intensive therapy in type 1 diabetes, the thresholds for the counterregulatory responses shift to a lower glycae- mic level (Cryer 2002; Cryer et al. 2003) and the same phenomenon occurs in type 2 diabetes (Korzon-Burakowska et al. 1998; Levy et al. 1998). Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 235

Counterregulatory deficiencies are associated with increasing duration of type 1 diabe- tes; it has been suggested that similar counterregulatory deficiencies may develop in type 2 diabetes. The secretory response of glucagon to hypoglycaemia is lost early in the natural history of type 1 diabetes (Chapter 3) and, although the catecholamine response compen- sates for several years, it too declines with time (Cryer 2002). There is disagreement over whether the glucagon response in type 2 diabetes is modestly diminished (Bolli et al. 1984; Meneilly et al. 1994b; Shamoon et al. 1994) or preserved (Boden et al. 1983; Heller et al. 1987; Korzon-Burakowska et al. 1998; Levy et al. 1998). It is probable that the ability of an individual to secrete glucagon in response to hypoglycaemia is related to their residual insulin secretory capacity and the ability to suppress endogenous insulin secretion when blood glucose falls (Israelian et al. 2005). Most investigators who have reported an intact glucagon response have studied people with type 2 diabetes who were treated effectively with oral antidiabetes agents and are therefore unlikely to have been insulin deficient (Boden et al. 1983; Heller et al. 1987; Levy et al. 1998; Spyer et al. 2000). Furthermore, with the exception of one study (Meneilly et al. 1994b), all of these studies have examined counterregulatory responses in middle-aged subjects in their fifth or sixth decade, despite the fact that most people with type 2 diabetes are older than this. The effects of ageing on counterregulation have not been addressed in these studies. To assess the role of insulin deficiency in promoting counterregulatory hormonal defi- ciencies, the counterregulatory responses to hypoglycaemia were examined in 15 non- diabetic controls and in 13 people with type 2 diabetes, six of whom were treated with insulin, and who were demonstrated to be insulin-deficient (having a low plasma C-peptide), while the remaining seven were being treated with oral antidiabetes agents (Segel et al. 2002). The glucagon response to hypoglycaemia was preserved in the patients on oral agents and in the non-diabetic controls, but was virtually absent in the insulin- deficient patients, demonstrating an association of acquired counterregulatory­ abnor malities and insulin deficiency in type 2 diabetes. Deficient counterregulatory responses were also observed in a group of patients with type 2 diabetes who had moderate beta cell failure, in whom the expected reduction in endogenous insulin secretion during hypoglycaemia was delayed and reduced, and responses of glucagon and growth hormone were impaired (Israelian et al. 2006) (Figure 12.2). Thus, multiple defects in the counter- regulatory responses to hypoglycaemia emerge as beta cell function fails in type 2 diabetes. In type 1 diabetes, individuals who experience frequent hypoglycaemia can develop a condition that has been termed Hypoglycaemia-Associated Autonomic Failure (HAAF) (Cryer 1992, 2013), where recurrent hypoglycaemia leads to failure of the centrally medi- ated counterregulatory response to hypoglycaemia, resulting in impaired awareness of hypoglycaemia (Chapter 6). In the study by Segel and colleagues (Segel et al. 2002), sub- jects with type 2 diabetes underwent a hypoglycaemic clamp on the first day of the study followed by a subsequent period of hypoglycaemia later that same day. They were then exposed to a second hypoglycaemic clamp during the following day, at which time the plasma glucose levels required to activate the hormonal and symptomatic responses were observed to be significantly lower than during the first hypoglycaemic clamp. These find- ings implicate a role for antecedent hypoglycaemia underlying the modification of the glycaemic thresholds at which responses to hypoglycaemia are triggered in type 2 diabetes. Although impaired hypoglycaemia awareness has been associated primarily with type 1 diabetes, this study suggests that individuals with insulin-treated type 2 diabetes may also develop HAAF. 236 Hypoglycaemia in Clinical Diabetes

150 9.0

125 6.0

100 3.0 Glucagon (pg/ml)

75 hormone ( µ g/l) Growth 0

) = non-diabetic subjects 3 3.0 = T2DM subjects 750 MEAN ± SEM 2.0 500

1.0 250 Cortisol (nM)

Epinephrine (pM × 10 0.0 0 –30 0 30 60 90 120 –30 0 30 60 90 120 Time (min) Time (min)

Figure 12.2 Plasma concentrations of glucagon, epinephrine, cortisol and growth hormone in non- diabetic subjects and subjects with type 2 diabetes mellitus (T2DM) during hypoglycaemia. Significantly lower responses of glucagon (45% reduction, P < 0.05) and growth hormone (50% reduction, P < 0.04) occur during hypoglycaemia, but no significant difference occurred in the responses of epinephrine and growth hormone. (Source: Israelian Z 2006. Reproduced with permission from Elsevier).

Changes typical of HAAF are not exclusive to insulin therapy. Fifteen adults with type 2 diabetes were treated intensively with metformin, glipizide and acarbose, with an associ- ated fall in mean HbA1c from 10.2 ± 0.5 to 6.7 ± 0.3% (88 to 50 mmol/mol). Hypoglycaemic clamps that were performed before and after the improvement in HbA1c demonstrated that counterregulatory responses to hypoglycaemia were blunted following intensive therapy (Davis et al. 2009). The preservation of the epinephrine response to hypoglycaemia, in conjunction with possible insulin resistance in some individuals, may provide further defensive mechanisms to protect people with type 2 diabetes against hypoglycaemia. The lipolytic effects of epinephrine can overcome the anti-lipolytic action of insulin on insulin-resistant adipose tissue, resulting in a rise in plasma free fatty acids in response to hypoglycaemia in type 2 diabetes (Bolli et al. 1984; Heller et al. 1987; Shamoon et al. 1994), that is not seen in more insulin-sensitive individuals with type 1 diabetes (Maggs et al. 1997). Epinephrine secretion during hypoglycaemia may therefore have a greater protective effect in insulin- resistant patients. Epinephrine also stimulates the mobilisation­ of glucose stored in the kidney, and this may compensate in part for the impaired production of hepatic glucose observed in individuals who have a deficient glucagon response to hypoglycaemia (Woerle et al. 2003).

Moderators of hypoglycaemia in type 2 diabetes Various factors can influence the risk of hypoglycaemia in type 2 diabetes (Table 12.2). Prolonged treatment with insulin (greater than 10 years) is a reliable predictor of increased risk of severe hypoglycaemia in type 2 diabetes (Donnelly et al. 2005), and when people Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 237

Table 12.2 Moderators of hypoglycaemia in type 2 diabetes

Moderator Reference

ESTABLISHED MODERATORS Duration of insulin treatment Donnelly et al. (2005) Hepburn et al. (1993) Impaired awareness of hypoglycaemia Henderson et al. (2003) Schopman et al. (2010) Hypoglycaemia-Associated Autonomic Segel et al. (2002) Failure Davis et al. (2009) Intensification of glycaemic control Chan et al. (2009) UKPDS (1998c) The Action to Control Cardiovascular Risk in Diabetes Study Group (2008) de Galan et al. (2009); Zoungas et al. (2010) Comorbidities Gitt et al. (2012) Exercise Minuk et al. (1981) Trovati et al. (1984) Devlin et al. (1987) Larsen et al. (1997) POSSIBLE MODERATORS ACE genotype Evidence against: Freathy et al. (2006) Evidence for: Davis et al. (2011) Glucose variability Murata et al. (2004) Monnier et al. (2011) MODERATORS IN TYPE 1 DIABETES NOT INVESTIGATED IN TYPE 2 DIABETES Absent C-peptide levels Sleep Pregnancy

with type 2 diabetes become insulin-deficient, their frequency of severe hypoglycaemia approaches that of individuals with type 1 diabetes who are matched for duration of treat- ment with insulin (Hepburn et al. 1993). Impaired awareness of hypoglycaemia is a major risk factor for severe hypoglycaemia in people with type 1 diabetes (Gold et al. 1994), but is less common in type 2 diabetes (Hepburn et al. 1993; Henderson et al. 2003; Schopman et al. 2010), with only 8% having impaired awareness in a retrospective survey of 215 individuals with insulin-treated type 2 diabetes (Henderson et al. 2003). However, the incidence of hypoglycaemia was nine times greater in this subgroup than in those with normal awareness. In a subsequent retro- spective survey from the same centre, 122 people with insulin-treated type 2 diabetes demonstrated a prevalence of impaired awareness of hypoglycaemia of 9.8% while the incidence of severe hypoglycaemia in the preceding year was 17 times higher in this sub- group than in the group with normal awareness (Schopman et al. 2010). The development of HAAF in type 2 diabetes and its relationship to impaired awareness has been discussed earlier. Continuous glucose monitoring (CGM) can detect asymptomatic hypoglycaemia. In a prospective study of 30 individuals with type 2 diabetes and 40 patients with type 1 dia- betes, asymptomatic hypoglycaemia was detected in 47% of the group with type 2 diabetes and in 63% of the group with type 1 diabetes (Chico et al. 2003). An Australian study used 238 Hypoglycaemia in Clinical Diabetes

CGM to examine the frequency of hypoglycaemia over 6 days in 25 patients treated with sulfonylureas. In 56% of these subjects, asymptomatic readings of <2.2 mmol/l lasting for at least 15 minutes were recorded (Hay et al. 2003). Thus, CGM data suggest that asymp- tomatic hypoglycaemia in type 2 diabetes is more common than has been appreciated previously. When interpreting CGM data it is important to appreciate that the relationship between interstitial glucose and plasma glucose differs between younger and older subjects. In type 1 diabetes and young adults without diabetes, interstitial glucose is approximately 10–15% lower than plasma glucose during hypoglycaemia, whereas in older adults, both with and without type 2 diabetes, interstitial glucose is significantly higher than blood glucose at all levels of hypoglycaemia, with the difference increasing by 0.32 mmol/l for every 1.0 mmol/l fall in blood glucose (P < 0.001) (Choudhary et al. 2011). The mechanism for this differ- ence is unclear but these data raise the possibility that these studies may underestimate the incidence of hypoglycaemia in older people.

Low HbA1c levels are associated with an increased risk of hypoglycaemia. In the UKPDS (United Kingdom Prospective Diabetes Study Group 1998c), a higher frequency of hypogly- caemia was observed in association with intensive as compared to conventional treatment. Several randomised controlled trials of treatment intensification with combinations of oral agents and insulin have also confirmed the association between lower HbA1c and a higher frequency of hypoglycaemia (The Action to Control Cardiovascular Risk in Diabetes Study Group 2008; de Galan et al. 2009; Zoungas et al. 2010). In a post hoc analysis of three studies comparing different insulin analogues combined with metformin, each 1% decre- ment in HbA1c was associated with an increase in the rate of all hypoglycaemia of 1.4 episodes per patient per year (Chan et al. 2009). However, a high HbA1c value may not protect against hypoglycaemia, as 55% of patients with insulin-treated type 2 diabetes with moderate glycaemic control (mean HbA1c 8.0 ± 1.5% [64 mmol/mol], mean age 80.2 ± 6 years) reported having experienced hypoglycaemia over the preceding 6 months (Munshi et al. 2009). Using CGM over a 3-day period in adults with diabetes over the age of 75 it was shown that 65% were recording interstitial glucose levels <3.9 mmol/l despite their mean HbA1c being 9.3% (78 mmol/mol) (Munshi et al. 2011). In the Fremantle Study, the patients who had the greatest risk of recurrent severe hypoglycaemia were those with a high

HbA1c and low fasting plasma glucose (Davis et al. 2010). These data suggest that blood glucose variability is associated with an increased risk of hypoglycaemia. In a prospective observational trial, a predominantly male cohort of people with insulin-treated type 2 diabetes recorded blood glucose profiles for 8 weeks and reported all episodes of hypoglycaemia over one year. The probability of all hypoglycaemia was highest in those who had a low mean blood glucose with a high standard deviation, implicating variability of blood glucose values as a predictor of risk of hypoglycaemia in this group (Murata et al. 2004). An observational study of insulin detemir in patients with type 1 diabetes and insulin-treated type 2 diabetes suggested an association between vari- ability of fasting glucose and the frequency of symptomatic hypoglycaemia (Luddeke et al. 2007). A CGM study in adults with type 2 diabetes treated either with insulin sensi- tisers, sulfonylureas or insulin has also suggested that the frequency of hypoglycaemia may be associated with increased glycaemic variability, as reflected by the standard deviation around the mean glucose value (Monnier et al. 2011). Other co-morbidities can affect the risk of hypoglycaemia. A German registry study of adults with type 2 diabetes compared the frequency of severe hypoglycaemia in patients with and without vascular disease. Despite comparable glycaemic control in both groups, Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 239 the group with vascular disease had an odds ratio of 3.73 (95% CI 1.31–10.65) for devel- oping severe hypoglycaemia (Gitt et al. 2012). Exercise can influence the risk of hypoglycaemia. The rate of skeletal muscle glucose uptake during physical exertion is normal in people with type 2 diabetes, but hepatic glucose output is impaired compared to non-diabetic individuals (Chipkin et al. 2001); this can result in exercise-related hypoglycaemia during physical exertion. Exercise reduces postprandial plasma glucose concentrations (Minuk et al. 1981; Larsen et al. 1997) and improves insulin sensitivity for up to 16 hours after physical activity (Trovati et al. 1984; Devlin et al. 1987), exposing the individual to an increased risk of hypoglycaemia for a prolonged period. However, a combination of moderate exercise and ingestion of alcohol did not provoke acute hypoglycaemia in C-peptide-positive, middle-aged individuals with type 2 diabetes, regardless of whether they had eaten or fasted before exercise (Rasmussen et al. 1999). Although alcohol impairs the counterregulatory responses to hypoglycaemia in type 1 diabetes (Avogaro et al. 1993), it does not appear to affect recovery from hypogly- caemia in type 2 diabetes (Rasmussen et al. 2001). Some genetic variations may also modify the risk of hypoglycaemia. For example, in type 1 diabetes, the Arg16Gly polymorphism of the beta-2 adrenergic receptor is linked with receptor desensitisation. This is associated in turn with a 3.4-fold increase in the frequency of impaired awareness of hypoglycaemia (Schouwenberg et al. 2008). Serum angiotensin- converting enzyme (ACE) activity is another possible marker for risk of hypoglycaemia. Variation in serum ACE levels is mediated by gene polymorphisms, via I (insertion) and D (deletion) alleles. The DD genotype is associated with higher serum ACE activity and, in type 1 diabetes, has been linked with increased risk of severe hypoglycaemia in some studies (Pedersen-Bjergaard et al. 2001, 2003a) although not in others (Bulsara et al. 2007; Zammitt et al. 2007). Similarly, in type 2 diabetes, the influence of ACE is uncertain. In a study of 308 people with type 2 diabetes, no evidence was found for a relationship between carriage of the ACE D allele and an increased risk of severe hypoglycaemia (Freathy et al. 2006). However, when 602 adults with type 2 diabetes were reviewed in the Fremantle study, the DD genotype was associated with an increased risk of developing a first episode of severe hypoglycaemia (hazard ratio 2.34; CI 1.29–4.26) (Davis et al. 2011). Studies have also examined the effects of CYP2C9, a cytochrome P450 enzyme which is mainly responsible for hepatic metabolism of sulfonylureas. One study genotyped 283 non-diabetic adults and 176 adults with type 2 diabetes treated with sulfonylureas, 92 of whom had experienced hypoglycaemia at some time previously. The CPY2C9*3 allele was associated with an increased frequency of hypoglycaemia (odds ratio 1.69, P = 0.011), which may be an effect of altered sulfonylurea metabolism (Ragia et al. 2009). However, a later study genotyped 203 patients with type 2 diabetes, all of whom were treated with sulfonylureas but only half of whom had ever experienced severe hypoglycaemia, and did not find any over-representation of slow metaboliser genotypes in the severe hypoglycae- mia group. The roles of CYP2C9 and ACE genotype in influencing hypoglycaemia risk remain inconclusive (Holstein et al. 2011).

FREQUENCY OF HYPOGLYCAEMIA IN TYPE 2 DIABETES

Mild hypoglycaemia is defined clinically by the ability to self-treat, while the needfor external assistance is used to denote a severe episode. The frequency of hypoglycaemia in people with type 1 diabetes is described in Chapter 4. Mild hypoglycaemia occurs on 240 Hypoglycaemia in Clinical Diabetes average around twice weekly (Pramming et al. 1991; Pedersen-Bjergaard et al. 2004) and the estimated incidence of severe hypoglycaemia ranges from 1.0 to 3.2 episodes/patient/ year (MacLeod et al. 1993; ter Braak et al. 2000; Pedersen-Bjergaard et al. 2004; UK Hypoglycaemia Study Group 2007), with an annual prevalence of 30–40% (Diabetes Control and Complications Trial Research Group 1993; MacLeod et al. 1993; Stephenson and Fuller 1994; ter Braak et al. 2000). The frequency of hypoglycaemia in type 2 diabetes cannot be summarised as succinctly because of the heterogeneity of this disorder and the range of treatment modalities available, resulting in a wide range of figures as reported in different reviews (Zammitt and Frier 2005; Akram et al. 2006; Amiel et al. 2008). Furthermore, many people with type 2 diabetes are elderly and the frequency of hypoglycaemia is generally underestimated in this group (McAulay and Frier 2001) and has seldom been measured. When hypoglycaemia rates were examined in over 64 000 adults with type 1 diabetes in a multicentre German database, the risk of hypoglycaemia was almost double in the over 60-year-old group when compared to the under 60 year olds (40.1 vs 24 per 100 patient years) (Schutt et al. 2012). Most studies have been conducted retrospectively and both the definitions of hypoglycaemia and the treatment modalities that have been examined, differ between studies. Individuals with type 1 diabetes can reliably remember episodes of severe hypoglycaemia after an interval of one year, but recall of mild hypoglycaemia becomes unreliable within a week (Pramming et al. 1991; Pedersen-Bjergaard et al. 2003b). In people with insulin-treated type 2 diabetes, recall of severe hypoglycaemia is similarly robust over a period of one year (Akram et al. 2009), but the reliability of recall of mild hypoglycaemia has not been examined and is unlikely to be preserved. Finally, the treatment regimens and patient inclusion criteria used in clinical trials are seldom representative of clinical practice within an unselected diabetic outpatient group and limit the extrapolation of such data to all people with type 2 diabetes. Table 12.3 summarises relevant epidemiological data on hypoglycaemia in type 2 diabetes.

Hypoglycaemia and oral antidiabetes agents Hypoglycaemia caused by oral antidiabetes agents is primarily associated with the insulin secretagogues. It is seldom a side effect of treatment with thiazolidinediones and has been reported only infrequently in association with metformin, usually when food intake is restricted (United Kingdom Prospective Diabetes Study Group 1998a, 1998b). In a meta- analysis of parallel-design randomised controlled trials comparing various non-insulin antidiabetes agents given along with metformin, all drug classes provided similar reduc- tions in HbA1c (0.64–0.97%; 6.5–10.6 mmol/mol), but the sulfonylureas and glinides were associated with an increased risk of hypoglycaemia compared to placebo (relative risk 4.57–7.50) (Table 12.4) (Phung et al. 2010). The frequency of hypoglycaemia secondary to sulfonylurea therapy is considerably lower than that attributed to insulin (United Kingdom Prospective Diabetes Study Group 1998c; Miller et al. 2001; Leese et al. 2003) but is mostly underestimated (Hartling et al. 1987). In a retrospective 6-month study in England of 219 people with type 2 diabetes treated with sulfonylureas and metformin, 20% of those taking sulfonylureas (either alone or in combination with metformin) had experienced hypoglycaemic symptoms (Jennings et al. 1989). Sulfonylureas can also cause severe hypoglycaemia. Over a 7-year period, the Swedish Adverse Drug Reactions Advisory Committee reported 19 cases of glipizide- associated severe hypoglycaemia presenting with reduced consciousness, and two cases were fatal (Asplund et al. 1991). Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 241

The frequency of hypoglycaemia relates to the individual pharmacokinetic properties of each sulfonylurea (Table 12.5) with the long-acting agents, such as glibenclamide and the modified release form of glipizide, being associated with the greatest risk (Stahl and Berger 1999; Del Prato et al. 2002; Rendell 2004). Glibenclamide is associated with an increased risk of severe hypoglycaemia (Tessier et al. 1994) because active metabolites prolong its effects for 24 hours (Jonsson et al. 2001; Rendell 2004). Glibenclamide also attenuates the glucagon response to hypoglycaemia in non-diabetic volunteers (ter Braak et al. 2002) and in people with type 2 diabetes (Landstedt-Hallin et al. 1999; Banarer et al. 2002). A meta- analysis examining hypoglycaemic events caused by different insulin secretagogues con- cluded that glibenclamide is associated with the greatest risk of severe hypoglycaemia (relative risk 1.83; 95% CI 1.35–2.49) (Gangji et al. 2007). The American Diabetes Asso- ciation and the European Association for the Study of Diabetes now recommend avoiding the use of glibenclamide (Nathan et al. 2009). Several drugs may potentiate the hypoglycaemic effects of sulfonylureas (Table 12.6). Risk factors for severe hypoglycaemia associated with sulfonylurea therapy include age, a past history of cardiovascular disease or stroke, renal failure, reduced food intake, alcohol ingestion and interactions with other drugs (Hartling et al. 1987; Seltzer 1989; Asplund et al. 1991; Campbell et al. 1994; Shorr et al. 1997; Ben-Ami et al. 1999; Burge et al. 1999; Harrigan et al. 2001; Holstein et al. 2010). In a retrospective study of hospital admis- sions resulting from sulfonylurea-associated hypoglycaemia (155 admissions over 8 years), most patients were elderly, a third had recently had their medications changed, 43% had a concurrent infection, 44% had a plasma creatinine >120 μmol/l and 31% admitted to poor intake of food prior to hospital admission (Schejter et al. 2012). Some of the newer sulfonylureas such as glimepiride are associated with a lower risk of hypoglycaemia. In a population-based study in Germany, the incidence of hypoglycae- mia was examined in people with type 2 diabetes who had attended a hospital emergency department over a 4-year period (Holstein et al. 2001). A total of 45 episodes were recorded in individuals who had been taking a sulfonylurea, of which 38 were associated with glib- enclamide. Glimepiride was implicated in only six episodes, despite being prescribed more frequently. When glimepiride was compared with a modified release form of gliclazide in a multicentre, European trial (Schernthaner et al. 2004a), glycated haemoglobin values improved by around 1.0% in both groups, but modified release gliclazide was implicated in fewer cases of hypoglycaemia than glimepiride (3.7% vs 8.9%). The group that is particularly vulnerable to sulfonylurea-induced hypoglycaemia is elderly people with type 2 diabetes living in residential homes. In Hong Kong, where glibenclamide comprises 50–80% of all sulfonylureas prescribed, a retrospective study of Chinese patients admitted to hospital with drug-induced hypoglycaemia revealed that more than half (54.4%) were being treated with sulfonylureas (So et al. 2002). Two-thirds of those requiring recurrent admission for hypoglycaemia were living in residential homes for old people; their clinical presentation was generally more severe and they had a poorer prognosis. Similarly, among 1149 cases of severe hypoglycaemia treated with i.v. glucose at one hospital in Germany from 2000 to 2009, 141 (10%) had taken sulfonylureas, one- third of whom were living in nursing homes or being cared for by a home nursing service (Holstein et al. 2010). The oral prandial glucose regulators (or meglitinides), repaglinide and nateglinide, induce less frequent hypoglycaemia than the sulfonylureas because of their rapid onset of action and their selective stimulation of insulin secretion in the presence of carbohydrate but not in the fasting state (Strange et al. 1999; Nattrass and Lauritzen 2000). In a randomised 242 Hypoglycaemia in Clinical Diabetes

Table 12.3 Epidemiological data on hypoglycaemia (hypo) in type 2 diabetes. Figures are mean unless otherwise stated. (Source: Adapted from Zammitt NN 2005. Reproduced with permission from the American Diabetes Association).

United Kingdom UK Prospective Hypoglycaemia Diabetes Study Hepburn Abraira Henderson Leese et al. Donnelly Akram et al. Study Group Study Jennings et al. (1989) Group (1998c) et al. (1993)* et al. (1995) et al. (2003) (2003)* et al. (2005)* (2006) (2007)*

Design Retrospective, structured Prospective Retrospective, Prospective Retrospective Population-based Prospective Retrospective Prospective interview multicentre RCT questionnaire multicentre questionnaire dataset analysis questionnaire multi-centre RCT observational Subjects T2D, OHA only T2D. OHA & Insulin-treated All T2D Insulin-treated 69 T1D; T2D SU Insulin-treated Insulin-treated T1D, T2D on SU insulin T1D & T2D T2D (22), insulin T1D & T2D T2DM and T2D on (66) insulin Number 219 (203: SU, 16: MF) 3935 104 T1D, 104 153 215 160 94 T1D 173 401 108 SU, 166 T2D T2D insulin Duration 6 months 10 years 1 year 18–35 months 1 year 1 year One month 1 year 9–12 months Prevalence OHA: all MF 0% Glibenclamide NA NA NA NA NA NA SU: 39% hypos SUs: 20.2% Glibenclamide 17.0% 31.3% Gliclazide 13.1% OHA: SH Glibenclamide NA NA NA 0.8% NA NA SU: 7% 0.6% Insulin: all NA 36.5% 82.7% 56% 64% NA 45% NA T2 < 2 yr 51% hypos conventional T2 > 5 yr 64% 93% intensive Insulin: SH NA 2.3% 10% NA 15% 7.3% 3% 16.5% T2 < 2 yr 7% T2 > 5 yr 25% Incidence Median (range) OHA: all NA NA NA NA NA NA NA NA 0 (0–25) hypos OHA: SH NA NA NA NA NA SU: 0.009, MF: NA NA 0 (0–7.2) 0.0005 Insulin: all NA NA NA 1.5 NA NA 16.37 NA T2 < 2 yr 1 hypos (standard) (0–44) 16.5 T2 > 5 yr 2.7 (intensive) (0–144) Insulin: SH NA NA NA 0.02 0.28 0.12 (T1 & T2D) 0.35 0.44 T2 < 2 yr 0 (0–10.5) T2 > 5 yr 0 (0–10)

hypo, hypoglycaemia; LOC, loss of consciousness; MF, metformin; NA, not applicable; OHA, oral hypoglycaemic agent; RCT, randomised clinical trial; SH, severe hypoglycaemia; SU, sulfonylurea; T1D, type 1 diabetes; T2D, type 2 diabetes; * only figures for type 2 DM given. Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 243

Table 12.3 Epidemiological data on hypoglycaemia (hypo) in type 2 diabetes. Figures are mean unless otherwise stated. (Source: Adapted from Zammitt NN 2005. Reproduced with permission from the American Diabetes Association).

United Kingdom UK Prospective Hypoglycaemia Diabetes Study Hepburn Abraira Henderson Leese et al. Donnelly Akram et al. Study Group Study Jennings et al. (1989) Group (1998c) et al. (1993)* et al. (1995) et al. (2003) (2003)* et al. (2005)* (2006) (2007)*

Design Retrospective, structured Prospective Retrospective, Prospective Retrospective Population-based Prospective Retrospective Prospective interview multicentre RCT questionnaire multicentre questionnaire dataset analysis questionnaire multi-centre RCT observational Subjects T2D, OHA only T2D. OHA & Insulin-treated All T2D Insulin-treated 69 T1D; T2D SU Insulin-treated Insulin-treated T1D, T2D on SU insulin T1D & T2D T2D (22), insulin T1D & T2D T2DM and T2D on (66) insulin Number 219 (203: SU, 16: MF) 3935 104 T1D, 104 153 215 160 94 T1D 173 401 108 SU, 166 T2D T2D insulin Duration 6 months 10 years 1 year 18–35 months 1 year 1 year One month 1 year 9–12 months Prevalence OHA: all MF 0% Glibenclamide NA NA NA NA NA NA SU: 39% hypos SUs: 20.2% Glibenclamide 17.0% 31.3% Gliclazide 13.1% OHA: SH Glibenclamide NA NA NA 0.8% NA NA SU: 7% 0.6% Insulin: all NA 36.5% 82.7% 56% 64% NA 45% NA T2 < 2 yr 51% hypos conventional T2 > 5 yr 64% 93% intensive Insulin: SH NA 2.3% 10% NA 15% 7.3% 3% 16.5% T2 < 2 yr 7% T2 > 5 yr 25% Incidence Median (range) OHA: all NA NA NA NA NA NA NA NA 0 (0–25) hypos OHA: SH NA NA NA NA NA SU: 0.009, MF: NA NA 0 (0–7.2) 0.0005 Insulin: all NA NA NA 1.5 NA NA 16.37 NA T2 < 2 yr 1 hypos (standard) (0–44) 16.5 T2 > 5 yr 2.7 (intensive) (0–144) Insulin: SH NA NA NA 0.02 0.28 0.12 (T1 & T2D) 0.35 0.44 T2 < 2 yr 0 (0–10.5) T2 > 5 yr 0 (0–10) hypo, hypoglycaemia; LOC, loss of consciousness; MF, metformin; NA, not applicable; OHA, oral hypoglycaemic agent; RCT, randomised clinical trial; SH, severe hypoglycaemia; SU, sulfonylurea; T1D, type 1 diabetes; T2D, type 2 diabetes; * only figures for type 2 DM given. 244 Hypoglycaemia in Clinical Diabetes

Table 12.4 Results of mixed treatment comparison meta-analysis comparing different classes of antidiabetes drugs added to metformin (compared to metformin plus placebo). (Source: Data compiled using data from Phung et al. 2010).

Relative risk of overall 95% credible Drug class hypoglycaemia interval

Sulfonylureas 4.57 2.11–11.45 Meglitinides 7.50 2.12–41.52 Thiazolidinediones 0.56 0.19–1.69 DPP-4 inhibitors 0.63 0.26–1.71 GLP-1 analogues 0.89 0.22–3.96

Table 12.5 Pharmacokinetics of the sulfonylureas. (Data sourced from Kobayashi et al. 1984; Campbell et al. 1994; DeFronzo 1999; Davis et al. 2000; Harrower 2000; Harrigan et al. 2001; Schernthaner 2003).

Duration

tmax t½ of action Renal excretion of Generation Name (hours) (hours) (hours) active metabolite

First Chlorpropamide 2–7 36 60 Yes Tolbutamide 3–4 3–28 6–12 Insignificant Second Glipizide 1–3 7 12–24 No Glipizide GITS 6–12 7 24 No (Glucotrol XL) Glibenclamide 2–6 10 12–24 Yes Gliclazide 2–3 12–14 Minimal (5%) Gliclazide MR 4–6 24 Minimal Third Glimepiride 2–3 5–9 16–24 Yes (?)

tmax, time to peak; t½, half-life.

Table 12.6 Drugs that potentiate the effects of sulfonylurea drugs used to treat type 2 diabetes. (Data sourced from Campbell et al. 1994; Harrigan et al. 2001).

Displacement Decreased Decreased of SU from Increases hepatic renal albumin plasma Inhibition of Mechanism metabolism excretion binding sites concentration gluconeogenesis

Drugs Chloramphenicol Allopurinol Fibrates Fluconazole Alcohol H2 blockers Probenecid Trimethoprim Miconazole Ciprofloxacin Monoamine- oxidase inhibitors Warfarin Warfarin Aspirin Aspirin Sulfonamides Sulfonamides

SU, sulfonylurea. Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 245

multicentre trial comparing repaglinide with nateglinide, slightly lower HbA1c values were achieved after 16 weeks on repaglinide at the expense of mild hypoglycaemic episodes in 7% of patients compared to no episodes in the nateglinide group (Rosenstock et al. 2004).

Studies comparing hypoglycaemia secondary to insulin or oral antidiabetes agents It is often difficult to compare the frequency of hypoglycaemia ascribed to insulin and oral antidiabetes agents, because in many studies the participants were being treated with both modalities. The UK Prospective Diabetes Study (United Kingdom Prospective diabetes Study Group 1998c) reported the prevalence of hypoglycaemia in people with type 2 dia- betes using different therapies. A higher frequency of hypoglycaemia was observed in association with intensive, compared to conventional, treatment whether using sulfonylu- reas or insulin. The prevalence of hypoglycaemia across the different treatment groups is summarised in Table 12.7. Of the oral agents used in this study, metformin had the lowest recorded rate of hypoglycaemia while glibenclamide had the highest. With intensive treat- ment, hypoglycaemia occurred most frequently in the insulin-treated patients, and the prevalence of hypoglycaemia was lower in the first decade of the study than in later years. The prevalence of hypoglycaemia was lower when the groups were analysed on an ‘inten- tion to treat’ basis because several patients in the conventional-treatment groups required an increasing number of different therapies with escalating doses as their glycaemic control deteriorated. Although the patients were questioned about the occurrence of hypoglycaemia at every 4-monthly review, only the most severe episode was recorded each time. This study cannot therefore provide an accurate estimate of the incidence of hypoglycaemia in type 2 diabetes, and the overall prevalence of severe hypoglycaemia of 2% is deceptive as it does not indicate the rise in prevalence with increasing duration of type 2 diabetes (United Kingdom Prospective Diabetes Study Group 1998c).

Table 12.7 Hypoglycaemia episodes per year by intention to treat analysis and actual therapy for intensive and conventional treatment in the UKPDS. (Table compiled from data in papers by United Kingdom Prospective Diabetes Study 1998a, 1998c).

Mean proportion of patients per year with hypoglycaemia (%)

Conventional Chlorpropamide Glibenclamide Insulin Metformin

≥1 episode MH in Normal BMI 1.2 11.0 17.7 36.5 – first 10 years Overweight 0.9 12.1 17.5 34.0 4.2 participants ≥1 episode SH in Normal BMI 0.1 0.4 0.6 2.3 – first 10 years Overweight 0.7 0.6 2.5 0.3 0 participants ≥1 episode MH in Normal BMI 10 16 21 28 – first 10 years Overweight 7.9 15.2 20.5 25.5 8.3 (intention to participants treat analysis) ≥1 episode SH in Normal BMI 0.7 1.0 1.4 1.8 – first 10 years Overweight 0.7 1.2 1.0 2.0 0.6 (intention to participants treat analysis)

BMI, body mass index; MH, mild hypoglycaemia; SH, severe hypoglycaemia. 246 Hypoglycaemia in Clinical Diabetes

A systematic review of randomised controlled trials comparing insulin monotherapy with combination therapy with insulin and oral antidiabetes agents confirmed the relative safety of the latter treatment regimen and 13 out of 14 studies did not show any significant difference in hypoglycaemia rates between the two therapeutic strategies (Goudswaard et al. 2004). In an observational study of 41 people with type 2 diabetes treated with oral antidiabetes drugs and bedtime isophane (NPH) insulin, 49% had experienced mild hypoglycaemia since commencing insulin with an incidence of 4 episodes/patient/year and no episodes of severe hypoglycaemia (Allen et al. 2004). Several studies have compared the frequency of hypoglycaemia secondary to oral anti­ diabetes agents to that associated with insulin treatment. A retrospective population-based study in Tennessee examined the frequency of ‘serious’ hypoglycaemia (an event resulting in hospitalisation or death) over a 4-year period in 19 932 Medicaid patients with type 2 diabetes who were aged 65 or older (Shorr et al. 1997). The frequency of severe hypogly- caemia was underestimated by employing such a restrictive definition. The reported inci- dence of ‘serious’ hypoglycaemia with sulfonylureas was 1.23 episodes per 100 person years and 2.76 episodes per 100 person years with insulin treatment. In the DISTANCE study, which was a cross-sectional survey of adults with type 2 diabetes, the prevalence of severe hypoglycaemia was 59% in insulin-treated patients compared to a prevalence of 23% associated with combination oral therapies, 13% for insulin secretagogue mono- therapy and 5% for metformin monotherapy (Sarkar et al. 2010). A 12-month prospective multicentre British study tested the hypothesis that the risk of hypoglycaemia in individuals with insulin-treated type 2 diabetes of short duration is com- parable to those taking sulfonylureas and is lower than in patients with newly diagnosed type 1 diabetes (UK Hypoglycaemia Study Group 2007). Monthly questionnaires were used to record hypoglycaemia and a 72-hour period of CGM was applied at the beginning and the end of the study period to detect asymptomatic hypoglycaemia. No difference was observed in the prevalence of severe hypoglycaemia between those with type 2 diabetes treated with sulfonylureas and individuals with type 2 diabetes treated with insulin for less than 2 years (7% in both groups, Figure 12.3a) or of mild symptomatic hypoglycaemia (39% vs 51%, Figure 12.3b) or of low interstitial glucose identified with CGM (22% vs 20%).

Hypoglycaemia and insulin In the USA, the Veterans Affairs Cooperative Study in type 2 Diabetes Mellitus (VA CSDM) examined glycaemic control and complications and compared a ‘standard’ insulin regimen (administered once daily) with an intensive (‘stepped’) regimen (Abraira et al. 1995). The participants in this trial had diabetes of relatively short duration (mean ± SD 7.8 ± 4 years), they were all insulin-treated males, and the study lasted for only 18–35 months. The overall incidence of severe hypoglycaemia was 0.02 episodes/patient/year; no significant difference was observed between the standard and stepped treatment groups. The frequency of mild hypoglycaemia was significantly higher in the stepped group (16.5 vs 1.5 episodes/patient/year). However, blood glucose was monitored less frequently in the standard treatment group, which may have resulted in asymptomatic hypoglycaemia being under-reported. A retrospective survey of 215 people with insulin-treated type 2 diabetes observed that the frequency of severe hypoglycaemia increased with the duration of insulin therapy

(Henderson et al. 2003) (Figure 12.4), and was inversely proportional to HbA1c concentra- Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 247

0.8 Severe hypoglycaemia

0.7

0.6

0.5

0.4

0.3

0.2 one hypoglycaemic episode one hypoglycaemic Proportion reporting at least Proportion 0.1

0.0 SUs T2DM T2DM T1DM T1DM (a) < 2 yrs > 5 yrs > 2 yrs > 15 yrs

1.0 Mild hypoglycaemia

0.8

0.6

0.4 one hypoglycaemic episode one hypoglycaemic Proportion reporting at least Proportion 0.2

0.0 SUs T2DM T2DM T1DM T1DM (b) < 2 yrs > 5 yrs > 2 yrs > 15 yrs

Figure 12.3 (a) Proportion of each group experiencing at least one severe episode of self-reported hypoglycaemia during 9–12 months of follow-up. (b) Proportion of each group experiencing at least one mild (self-treated) episode of self-reported hypoglycaemia during 9–12 months of follow-up. SU: sulfo- nylurea ; T1DM, type 1 diabetes mellitus; T2DM, type 2 diabetes mellitus. (Source: UK Hypoglycaemia Study Group 2007. Reproduced with kind permission from Springer Science+Business Media). tion. This relationship between duration of insulin treatment and prevalence of severe hypoglycaemia was replicated in the prospective multicentre British survey (UK Hypogly- caemia Study Group 2007) (Figure 12.5). The annual prevalence of severe hypoglycaemia was 15% with an overall incidence of 0.28 episodes/patient/year. In this study, symptomatic hypoglycaemia in patients with type 2 diabetes recently commenced on insulin therapy (<2 years) was considerably lower than in those with type 1 diabetes of less than 5-years dura- tion (median rate 1 vs 22 episodes/subject/ year, P < 0.001; Figure 12.3), but was higher in patients with type 2 diabetes >5 years, showing a significantly higher prevalence in those with a longer duration of insulin therapy. It should be noted that in type 1 diabetes, the frequency of mild hypoglycaemia did not change with duration of treatment, whereas in 248 Hypoglycaemia in Clinical Diabetes

40

35 n = 29

30

25

20 n = 39 15

Prevalence (%) Prevalence n = 147 10

5

0 1–5 6–10 > 10 Duration of insulin therapy (years)

Figure 12.4 Prevalence of severe hypoglycaemia in relation to duration of diabetes. (Source: Hend- erson JN 2003. Reproduced with permission from John Wiley & Sons, Ltd).

n = 54

50

40 Annual 30 prevalence of n = 75 n = 46 severe hypoglycaemia 20 (%) 10 n = 85 0

T2DM T2DM T1DM T1DM < 2 yrs > 5 yrs < 5 yrs > 15 yrs

Figure 12.5 Prevalence of severe hypoglycaemia in relation to duration of treatment with insulin. (Source: Based on data from UK Hypoglycaemia Study Group 2007). insulin-treated type 2 diabetes, the prevalence of mild hypoglycaemia rises with increasing duration of treatment, just as it does for severe episodes (Figure 12.3b, UK Hypoglycaemia Study Group 2007). A similar incidence was reported in a retrospective Danish study of 401 patients with insulin-treated type 2 diabetes, 66 (16.5%) of whom had experienced at least one episode of severe hypoglycaemia in the preceding year, giving an overall incidence of 0.44 episodes/ patient/year, but no relationship was observed to HbA1c (Akram et al. 2006), in contrast with the findings of Chan et al. (2009). A retrospective study of 600 unselected insulin- Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 249 treated patients performed a decade earlier in Edinburgh had documented an incidence of severe hypoglycaemia of 0.73 episodes/patient/year in the 56 people with type 2 diabetes compared with 1.7 episodes/patient/year in the 544 with type 1 diabetes (MacLeod et al. 1993). Another survey in the same centre compared the frequency of severe hypoglycaemia in 86 people with insulin-treated type 2 diabetes with 86 people with type 1 diabetes, matched for duration of insulin treatment and insulin dose (Hepburn et al. 1993). The frequency of severe hypoglycaemia was similar in the two groups and a significant correla- tion was found between the frequency of severe hypoglycaemia and increasing duration of treatment with insulin (r = 0.39, P < 0.001). The use of insulin analogues has been claimed to lower the risk of hypoglycaemia. In some studies, the risk of hypoglycaemia has been reported to be lower with long-acting insulin glargine (Yki-Järvinen et al. 2000; Rosenstock et al. 2001; Rosenstock 2003; Riddle et al. 2003) and with insulin detemir (Hermansen et al. 2006) compared with isophane (NPH) insulin. Insulin glargine was also associated with a lower frequency of hypoglycae- mia than premixed insulins (Janka et al. 2005; Raskin et al. 2005). Rapid-acting insulin analogues, such as lispro and glulisine, also appeared to limit the frequency of hypogly- caemia in people with type 2 diabetes when compared to short-acting soluble (regular) insulins (Anderson et al. 1997; Bastyr et al. 2000; McAulay and Frier 2003; Dailey et al. 2004). However, other studies have not observed the incidence of hypoglycaemia to be significantly lower when insulin analogues were compared with conventional insulins (Ross et al. 2001; Raslova et al. 2004; Schernthaner et al. 2004b; Haak et al. 2005). An online survey of people with type 2 diabetes using insulin analogues reported that self- treated hypoglycaemia was common in approximately one-third of patients (Brod et al. 2012). Basal insulin analogues are not associated with lower rates of severe hypoglycaemia (Scottish Intercollegiate Guidelines Network (SIGN) 2010; Waugh et al. 2010). Therefore, both NICE and SIGN guidelines as well as a recent Cochrane review all recommend the use of isophane (NPH) insulin rather than a basal analogue insulin in type 2 diabetes unless the patient is having significant problems with hypoglycaemia (Horvath et al. 2007; National Institute for Health and Clinical Excellence 2009; Scottish Intercollegiate Guide- lines Network (SIGN) 2010). Continuous subcutaneous insulin infusion (CSII) using an insulin pump is not widely used as a method of insulin delivery in people with type 2 diabetes. Although it is often claimed that CSII is less likely to cause severe hypoglycaemia than a basal-bolus insulin regimen, a 12-month prospective randomised study of 107 adults with insulin-treated type 2 diabetes showed no significant difference between CSII and multiple insulin injections in the rates of mild or severe hypoglycaemia (Herman et al. 2005). A meta-analysis of studies, mainly conducted in patients with type 1 diabetes, was unable to demonstrate a significant benefit (Fatourechi et al. 2009). Therefore, while it is not possible to give a precise estimate of the frequency of hypogly- caemia in people with insulin-treated type 2 diabetes, some general conclusions can be drawn. Firstly, the frequency of hypoglycaemia rises with increasing duration of insulin treatment (Hepburn et al. 1993; Henderson et al. 2003; UK Hypoglycaemia Study Group 2007), which probably reflects the degree of pancreatic failure and insulin deficiency. Secondly, although the frequency of hypoglycaemia is generally greater with insulin therapy than with oral agents, the prevalence of hypoglycaemia in the first 2 years of insulin treatment is comparable to that of patients treated with sulfonylureas (UK Hypoglycaemia Study Group 2007). This should be considered when evaluating a patient’s individual risk of hypoglycaemia. 250 Hypoglycaemia in Clinical Diabetes

Hypoglycaemia and incretin mimetics Glucagon-like peptide 1 (GLP-1) is an incretin hormone that promotes glucose-dependent insulin secretion and inhibition of glucagon production. GLP-1 is rapidly degraded in vivo by the ubiquitous enzyme dipeptidyl peptidase 4 (DPP-4). GLP-1 analogues are associated with improvements in glycaemic control (Zander et al. 2002; Fineman et al. 2003; Kolter- man et al. 2003; Degn et al. 2004a) without causing hypoglycaemia in people with type 2 diabetes (Knop et al. 2003; Madsbad et al. 2004). Exenatide, a synthetic GLP-1 analogue, stimulates insulin release only in the presence of glucose (Degn et al. 2004b) and, although it suppresses glucagon production, the effect on glucagon suppression dissipates during hypoglycaemia and the counterregulatory response to insulin-induced hypoglycaemia is preserved (Nauck et al. 2002; Degn et al. 2004b). Although incretin mimetics may cause reactive hypoglycaemia in non-diabetic individuals (Meier and Nauck 2005), they do not appear to cause hypoglycaemia in people with type 2 diabetes (Vilsbøll et al. 2001; Knop et al. 2003). Generally, DPP-4 inhibitors are considered to cause hypoglycaemia only in the context of combination therapy (Chia and Egan 2008; Drucker et al. 2010). The following three studies compared different DPP-4 inhibitors to sulfonylureas as add-on therapy to met- formin. In a one-year trial comparing vildagliptin and glimepiride, only 1.7% of the vild- agliptin group reported any hypoglycaemia, all of which was self-treated, while 16% of patients treated with glimepiride experienced at least one episode of hypoglycaemia, including 10 episodes requiring external assistance (Ferrannini et al. 2009). Similarly, in a study comparing saxagliptin and glipizide, 3% of the saxagliptin group experienced hypoglycaemia with no severe episodes during the one-year study period, compared to 36% who reported any hypoglycaemia and 1.6% who reported severe hypoglycaemia in the glipizide group (Goke et al. 2010). A 2-year randomised study comparing sitagliptin and glipizide also demonstrated a lower rate of hypoglycaemia in the sitagliptin group (5% of 588 patients experienced at least one episode of hypoglycaemia, compared to 34% of 584 patients in the glipizide group) (Seck et al. 2010). Of note, during the 2-year study period there were two cases of severe hypoglycaemia in the sitagliptin group, but this was lower than the 15 episodes of severe hypoglycaemia reported in the glipizide group.

MORBIDITY OF HYPOGLYCAEMIA AND NEED FOR EMERGENCY TREATMENT

People with type 1 diabetes view severe hypoglycaemia with the same degree of trepidation as developing blindness from retinopathy or renal failure from nephropathy (Pramming et al. 1991). The impact of hypoglycaemia on quality of life in type 2 diabetes has been reviewed (Barendse et al. 2012), and it is clear that fear of hypoglycaemia is also com- monplace in type 2 diabetes, as assessed by measures such as the worry subscale of the Hypoglycaemia Fear Survey-II (HFS-II). Mean HFS-II scores are higher in sulfonylurea- treated patients who have experienced hypoglycaemia compared to those with no history of hypoglycaemia (Vexiau et al. 2008; Stargardt et al. 2009; Marrett et al. 2011); the sever- ity of hypoglycaemic symptoms was significantly associated with higher worry scores (Vexiau et al. 2008). Quality-of-life scales also demonstrate reductions in health status associated with hypoglycaemia in type 2 diabetes (Jermendy et al. 2008; Vexiau et al. 2008; Marrett et al. 2011). A Canadian survey assessing fear of hypoglycaemia in adults with Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 251 type 1 and insulin-treated type 2 diabetes found that while 63.6% of individuals with type 1 diabetes and recent severe hypoglycaemia admitted to having increased fear of future hypoglycaemia, this was even greater in people with type 2 diabetes, affecting 84.9% (Leiter et al. 2005). Fear of hypoglycaemia was also commonly reported in a survey of people with insulin-treated type 2 diabetes (Brod et al. 2012). Among people with type 2 diabetes treated with oral agents, symptoms of hypoglycaemia tend to be associated with a significantly lower quality of life and higher levels of healthcare resource utilisation (Williams et al. 2011). Blood glucose awareness training and cognitive behavioural therapy have been shown to reduce levels of fear and improve diabetes self-management (Currie et al. 2006). Hypoglycaemia is not only extremely unpleasant for the individual concerned, it can be associated with severe morbidity and may precipitate major vascular events such as stroke, myocardial infarction, acute cardiac failure and ventricular arrhythmias (Landstedt- Hallin et al. 1999; McAulay and Frier 2001; Desouza et al. 2003; Wright and Frier 2008) (see Chapter 13). Healthcare professionals may not always recognise the causative role of hypoglycaemia when treating these secondary events. The recent concern that strict gly- caemic control may result in an increase in cardiovascular mortality (The Action to Control Cardiovascular Risk in Diabetes Study Group 2008) is discussed in more detail in Chapter 13. The ADVANCE study has suggested that severe hypoglycaemia is associated with a range of adverse outcomes, despite the fact that this study did not report any excess mor- tality in its intensive treatment arm (Zoungas et al. 2010). These include more frequent major macrovascular events (HR 2.88; CI 2.01–4.12), cardiovascular deaths (HR 2.68; CI 1.72–4.19) and deaths from any cause (HR 2.69; CI 1.97–3.67) (P < 0.01 for all compari- sons). However, these associations do not prove causality and severe hypoglycaemia may simply be a marker of greater vulnerability rather than a specific trigger for the observed outcomes. In a 2-year prospective survey of hospital admissions of patients with type 2 diabetes aged 80 or over, of 124 people (mean HbA1c 5.1%) who were admitted to hospital, 31 (25%) admissions were precipitated by severe hypoglycaemia (Greco and Angileri 2004). In a German population-based observational study, the incidence of severe sulfonylurea- induced hypoglycaemia was 7 cases per 100 000 inhabitants per year (Holstein et al. 2010). These findings are worrying because the elderly are particularly at risk of hypoglycaemia- related injury and bone fractures as a consequence of their frailty and the presence of co-morbidities, such as osteoporosis (McAulay and Frier 2001). In a 7-year review of 102 cases of hypoglycaemic coma secondary to either insulin or glibenclamide, 92 patients had type 2 diabetes; seven sustained physical injury, five died, two suffered myocardial ischae- mia and one patient had a stroke (Ben-Ami et al. 1999). Increasing interest is being focused on whether severe hypoglycaemia may be associated with an increased risk of dementia in elderly patients with type 2 diabetes. In a longitudinal cohort study of a Kaiser Permanente database of patients with type 2 diabetes (mean age 65 years), episodes of severe hypoglycaemia requiring treatment in hospital between 1980 and 2002 were reviewed, and cohort members with no past history of dementia were examined after an interval of 4 years (Whitmer et al. 2009). Cox proportional hazards regression models demonstrated a graded risk of developing dementia; one episode of severe hypoglycaemia was associated with a hazard ratio of developing dementia of 1.26, while two episodes had a hazard ratio of 1.8 and those with three or more episodes had a hazard ratio of 1.94. The conclusions of this study should be interpreted with caution, as observational studies cannot prove causation, confounders such as cerebrovascular disease 252 Hypoglycaemia in Clinical Diabetes can contribute to both hypoglycaemia and dementia and only episodes resulting in hospital admission were included in this study (Graveling and Frier 2009). This issue has also been explored as part of the Edinburgh Type 2 Diabetes Study, which is a cross-sectional population based study of 1066 adults with type 2 diabetes aged between 60 and 75 years old. After adjustment for confounders such as duration of diabetes and age, general cognitive ability scores were lower in participants who reported at least one previous episode of severe hypoglycaemia, with scores on verbal fluency tests, digit symbol tests, letter-number sequencing and trail-making tests being particularly affected (Aung et al. 2012). The fact that this association persisted after adjustment for prior cognitive ability suggests that hypoglycaemia may contribute to age-related cognitive decline. A recently reported prospective population-based study in the USA of 783 older people with diabetes has suggested the existence of a bidirectional association between hypoglycaemia and dementia (Yaffe et al. 2013). In type 1 diabetes, relatives often treat severe hypoglycaemia at home, but while people with insulin-treated type 2 diabetes experience severe hypoglycaemia less frequently, they appear to require the assistance of the emergency services with equal frequency. This might suggest that people with insulin-treated type 2 diabetes are at greater risk of morbidity and disability during hypoglycaemia, and they and their relatives may be less able to cope than younger people with type 1 diabetes, particularly if they live alone. A population survey in Tayside (Scotland) indicated that the annual rate of severe hypoglycaemia requiring emergency medical intervention was similar in these groups (Leese et al. 2003). All epi- sodes of severe hypoglycaemia requiring input from the emergency medical services in one year were identified. A total of 160 people with diabetes required treatment for 244 episodes of severe hypoglycaemia. Emergency treatment was required for 7.1% of those with type 1 diabetes, 7.3% of those with insulin-treated type 2 diabetes and 0.8% of people taking oral antidiabetes agents (principally sulfonylureas). In a different prospective survey in Tayside, the occurrence of hypoglycaemia was monitored over one month in a cohort of 267 people with insulin-treated diabetes of both types (Donnelly et al. 2005). The prevalence of all hypoglycaemia in the group with insulin- treated type 2 diabetes was 45%, with an incidence of 16.4 episodes/patient/year, compared to an incidence of 42.9 episodes/patient/year in type 1 diabetes. The incidence of severe hypoglycaemia was 0.35 episodes/patient/year in the group with type 2 diabetes and 1.15 episodes/patient/year in those with type 1 diabetes. The figures for the incidences were extrapolated from prospective data collected over one month, but these calculated rates for people with type 1 diabetes are consistent with those recorded in other European studies (Pramming et al. 1991; MacLeod et al. 1993; ter Braak et al. 2000; Pedersen-Bjergaard et al. 2004). In the Tayside study, only 10% of the group with type 1 diabetes experiencing severe hypoglycaemia required emergency service treatment compared to one in three of the group with type 2 diabetes. Thus the frequency of severe hypoglycaemia recorded in people with type 2 diabetes was higher than anticipated and their need to enlist the help of the emergency services was greater than in those with type 1 diabetes. Studies investigating the need for emergency services in the context of hypoglycaemia draw attention to the economic burden of hypoglycaemia. One study has compared the costs of severe hypoglycaemia associated with insulin-treated diabetes in Germany, Spain and the UK. Costs were lowest in the UK, but average costs were higher for patients with type 2 diabetes than patients with type 1 diabetes across all three countries (€537 per patient with type 2 diabetes, compared to €236 per patient with type 1 diabetes in the UK). This may relate to the greater likelihood of needing hospital input for treatment of severe hypoglycaemia in type 2 diabetes (53% vs 22% in type 1 diabetes) (Hammer et al. 2009). Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 253

Mild hypoglycaemic events in people with type 2 diabetes, particularly when they have occurred at night, have been shown to be associated with reduced work productivity, which has significant economic consequences, with effects on self-management of diabetes (Brod et al. 2011). In a systematic review, an assessment of the indirect costs of hypoglycaemia confirmed an association between lower work productivity and a greater likelihood of being unemployed (Zhang et al. 2010).

CONCLUSIONS • Ageing modifies the counterregulatory and symptomatic responses to hypoglycaemia. • In older people, self-treatment of hypoglycaemia may be compromised by the fact that the glycaemic thresholds for the onset of symptoms and cognitive dysfunction occur almost simultaneously. • In type 2 diabetes, counterregulatory responses to hypoglycaemia commence at higher blood glucose levels than those observed in non-diabetic adults or in people with type 1 diabetes, which may have a protective effect. When insulin therapy is introduced and

HbA1c is reduced, these thresholds shift to lower blood glucose levels. • With progressive insulin deficiency, people with type 2 diabetes develop counterregula- tory hormonal deficiencies and impaired symptomatic awareness, similar to the acquired hypoglycaemia syndromes of type 1 diabetes. • Hypoglycaemia in type 2 diabetes occurs most frequently with insulin therapy, but sulfonylurea-induced hypoglycaemia is also a significant and underestimated problem. • Although less common than in type 1 diabetes, the frequency of hypoglycaemia in insulin-treated type 2 diabetes rises progressively with increasing duration of insulin treatment. • People with insulin-treated type 2 diabetes are more likely to require the assistance of emergency medical services to treat severe hypoglycaemia than those with type 1 diabetes. • The paucity of data in elderly people with diabetes is a cause for concern, as the mor- bidity of hypoglycaemia is greater in this age group, their presenting features of hypoglycaemia may be misinterpreted and prompt treatment may not be provided.

REFERENCES

Abraira, C., Colwell, J.A., Nuttall, F.Q. et al. (1995) Veterans Affairs Cooperative Study on glycemic control and complications in Type II Diabetes (VA CSDM). Diabetes Care, 18, 1113–1123. Akram, K., Pedersen-Bjergaard, U., Carstensen, B., Borch-Johnsen, K. & Thorsteinsson, B. (2006) Fre- quency and risk factors of severe hypoglycaemia in insulin-treated Type 2 diabetes: a cross-sectional survey. Diabetic Medicine, 23, 750–756. Akram, K., Pedersen-Bjergaard, U., Carstensen, B., Borch-Johnsen, K. & Thorsteinsson, B. (2009) Pro- spective and retrospective recording of severe hypoglycaemia and assessment of hypoglycaemia aware- ness in insulin-treated Type 2 diabetes. Diabetic Medicine, 26, 1306–1308 (letter). Allen, K.V., McAulay, V., Sommerfield, A.J. & Frier, B.M. (2004) Hypoglycaemia is uncommon with a combination of antidiabetic drugs and bedtime NPH insulin for type 2 diabetes. Practical Diabetes International, 21, 179–182. Amiel, S.A., Maran, A., Powrie, J.K., Umpleby, A.M. & Macdonald, I.A. (1993) Gender differences in counterregulation to hypoglycaemia. Diabetologia, 36, 460–464. Amiel, S.A., Dixon, T., Mann, R. & Jameson, K. (2008) Hypoglycaemia in type 2 diabetes. Diabetic Medicine, 25, 245–254. 254 Hypoglycaemia in Clinical Diabetes

Anderson, J.H., Jr., Brunelle, R.L., Keohane, P. et al. (1997) Mealtime treatment with insulin analog improves postprandial hyperglycemia and hypoglycemia in patients with non-insulin-dependent diabetes mellitus. Multicenter Insulin Lispro Study Group. Archives of Internal Medicine, 157, 1249–1255. Asplund, K., Wilholm, B.-E. & Lundman, B. (1991) Severe hypoglycaemia during treatment with glipizide. Diabetic Medicine, 8, 726–731. Aung, P.P., Strachan, M.W., Frier, B.M. et al. (2012) Severe hypoglycaemia and late-life cognitive ability in older people with type 2 diabetes: the Edinburgh Type 2 Diabetes Study. Diabetic Medicine, 29, 328–336. Avogaro, A., Beltramello, P., Gnudi, L. et al. (1993) Alcohol intake impairs glucose counterregulation during acute insulin-induced hypoglycemia in IDDM patients. Evidence for a critical role of free fatty acids. Diabetes, 42, 1626–1634. Banarer, S., McGregor, V. & Cryer, P.E. (2002) Intraislet hyperinsulinemia prevents the glucagon response to hypoglycemia despite an intact autonomic response. Diabetes, 51, 958–965. Barendse, S., Singh, H., Frier, B.M. & Speight, J. (2012) The impact of hypoglycaemia on quality of life and related patient-reported outcomes in type 2 diabetes: a narrative review. Diabetic Medicine, 29, 293–302. Bastyr, E.J., Huang, Y., Brunelle, R.L., Vignati, L., Cox, D.J. & Kotsamos, J.G. (2000) Factors associated with nocturnal hypoglycemia among patients with Type 2 diabetes new to insulin therapy. Diabetes, Obesity and Metabolism, 2, 39–46. Ben-Ami, H., Nagachandran, P., Mendelson, A. & Edoute, Y. (1999) Drug-induced hypoglycemic coma in 102 diabetic patients. Archives of Internal Medicine, 159, 281–284. Boden, G., Soriano, M., Hoeldtke, R.D. & Owen, O.E. (1983) Counterregulatory hormone release and glucose recovery after hypoglycemia in non-insulin-dependent patients. Diabetes, 32, 231–237. Bolli, G., Tsalikian, E., Haymond, M., Cryer, P. & Gerich, J.E. (1984) Defective glucose counterregulation after subcutaneous insulin in noninsulin dependent diabetes mellitus: paradoxical lack of compensatory increase in glucose production, roles of insulin resistance, abnormal neuroendocrine responses and islet paracrine interactions. The Journal of Clinical Investigation, 73, 1532–1541. Bremer, J.P., Jauch-Chara, K., Hallschmid, M., Schmid, S. & Schultes, B. (2009) Hypoglycemia unaware- ness in older compared with middle-aged patients with type 2 diabetes. Diabetes Care, 32, 1513–1517. Brierley, E.J., Broughton, D.L., James, O.F.W. & Alberti, K.G.M.M. (1995) Reduced awareness of hypo­ glycaemia in the elderly despite an intact counterregulatory response. The Quarterly Journal of Medicine, 88, 439–445. Brod, M., Christensen, T., Thomsen, T.L. & Bushnell, D.M. (2011) The impact of non-severe hypoglycemic events on work productivity and diabetes management. Value in Health, 14, 665–671. Brod, M., Rana, A. & Barnett, A.H. (2012) Impact of self-treated hypoglycaemia in type 2 diabetes: a multinational survey in patients and physicians. Current Medical Research and Opinion, 28, 1947–1958. Bulsara, M.K., Holman, C.D.J., van Bockxmeer, F.M. et al. (2007) The relationship between ACE genotype and risk of severe hypoglycaemia in a large population-based cohort of children and adolescents with type 1 diabetes. Diabetologia, 50, 965–971. Burge, M.R., Zeise, T.-M., Sobhy, T.A., Rassam, A.G. & Schade, D.S. (1999) Low-dose ethanol predisposes elderly fasted patients with Type 2 diabetes to sulfonylurea-induced low blood glucose. Diabetes Care, 22, 2037–2043. Campbell, I.W., Chalmers, J. & Herlihy, O.M. (1994) Sulphonylurea-induced hypoglycaemia in elderly people with diabetes. Practical Diabetes, 11, 102–103. Chan, J.C., Leyk, M., Frier, B.M. & Tan, M.H. (2009) Relationship between HbA1c and hypoglycaemia in patients with type 2 diabetes treated with different insulin regimens in combination with metformin. Diabetes/Metabolism Research and Reviews, 25, 224–231. Chia, C.W. & Egan, J.M. (2008) Incretin-based therapies in type 2 diabetes mellitus. Journal of Clinical Endocrinology and Metabolism, 93, 3703–3716. Chico, A., Vidal-Rios, P., Subira, M. & Novials, A. (2003) The continuous glucose monitoring system is useful for detecting unrecognised hypoglycemias in patients with type 1 and type 2 diabetes but is not better than frequent capillary glucose measurements for improving metabolic control. Diabetes Care, 26, 1153–1157. Chipkin, S.R., Klugh, S.A. & Chasan-Taber, L. (2001) Exercise and diabetes. Cardiology Clinics, 19, 489–505. Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 255

Choudhary, P., Lonnen, K., Emery, C.J. et al. (2009) Comparing hormonal and symptomatic responses to experimental hypoglycaemia in insulin- and sulphonylurea-treated type 2 diabetes. Diabetic Medicine, 26, 665–672. Choudhary, P., Lonnen, K., Emery, C.J., Freeman, J.V., MacLeod, K.M. & Heller, S.R. (2011) Relationship between interstitial and blood glucose during hypoglycemia in subjects with type 2 diabetes. Diabetes Technology and Therapeutics, 13, 1121–1127. Cryer, P.E. (1992) Iatrogenic hypoglycemia as a cause of hypoglycemia-associated autonomic failure in IDDM. A vicious cycle. Diabetes, 41, 255–260. Cryer, P.E. (2002) Hypoglycaemia: the limiting factor in the glycaemic management of Type I and Type II diabetes. Diabetologia, 45, 937–948. Cryer, P.E. (2013) Mechanisms of hypoglycemia-associated autonomic failure in diabetes. The New England Journal of Medicine, 369, 369–372. Cryer, P.E., Davis, S.N. & Shamoon, H. (2003) Hypoglycemia in diabetes. Diabetes Care, 26, 1902–1912. Currie, C.J., Morgan, C.L., Poole, C.D., Sharplin, P., Lammert, M. & McEwan, P. (2006) A critical review of the literature on fear of hypoglycemia in diabetes: implications for diabetes management and patient education. Current Medical Research and Opinion, 22, 1523–1534. Dailey, G., Rosenstock, J., Moses, R.G. & Ways, K. (2004) Insulin glulisine provides improved glycemic control in patients with type 2 diabetes. Diabetes Care, 27, 2363–2368. Davis, S.N., Goldstein, R.E., Jacobs, J., Price, L., Wolfe, R. & Cherrington, A.D. (1993) The effects of differing insulin levels on the hormonal and metabolic response to equivalent hypoglycemia in normal humans. Diabetes, 42, 263–272. Davis, S.N., Fowler, S. & Costa, F. (2000) Hypoglycemic counterregulatory responses differ between men and women with type 1 diabetes. Diabetes, 49, 56–72. Davis, S.N., Mann, S., Briscoe, V.J., Ertl, A.C. & Tate, D.B. (2009) Effects of intensive insulin therapy and antecedent hypoglycemia on counterregulatory responses to hypoglycemia in type 2 diabetes. Dia­ betes, 58, 701–709. Davis, T.M.E., Brown, S.G.A., Jacobs, I.G., Bulsara, M., Bruce, D.G. & Davis, W.A. (2010) Determinants of severe hypoglycemia complicating type 2 diabetes: the Fremantle diabetes study. Journal of Clinical Endocrinology and Metabolism, 95, 2240–2247. Davis, W.A., Brown, S.G., Jacobs, I.G., Bulsara, M., Bruce, D.G. & Davis, T.M. (2011) Angiotensin- converting enzyme insertion/deletion polymorphism and severe hypoglycemia complicating type 2 diabetes: the Fremantle diabetes study. Journal of Clinical Endocrinology and Metabolism, 96, E696–E700. Deary, I.J., Hepburn, D.A., MacLeod, K.M. & Frier, B.M. (1993) Partitioning the symptoms of hypogly- caemia using multi-sample confirmatory factor analysis. Diabetologia, 36, 771–777. DeFronzo, R.A. (1999) Pharmacologic therapy for type 2 diabetes mellitus. Annals of Internal Medicine, 131, 281–303. Degn, K.B., Juhl, C.B., Sturis, J. et al. (2004a) One week’s treatment with the long-acting glucagon-like peptide 1 derivative liraglutide (NN2211) markedly improves 24h glycemia and α- and β-cell function and reduces endogenous glucose release in patients with type 2 diabetes. Diabetes, 53, 1187–1194. Degn, K.B., Brock, B., Juhl, C.B. et al. (2004b) Effect of intravenous infusion of exenatide (synthetic Exendin-4) on glucose-dependent insulin secretion and counterregulation during hypoglycemia. Diab­ etes, 53, 2397–2403. Del Prato, S., Aragona, M. & Coppelli, A. (2002) Sulfonylureas and hypoglycaemia. Diabetes, Nutrition and Metabolism, 15, 444–451. Desouza, C., Salazar, H., Cheong, B., Murgo, J. & Fonseca, V. (2003) Association of hypoglycemia and cardiac ischemia. Diabetes Care, 26, 1485–1489. Devlin, J.T., Hirshman, M., Horton, E.D. & Horton, E.S. (1987) Enhanced peripheral and splanchnic insulin sensitivity in NIDDM men after single bout of exercise. Diabetes, 36, 434–439. Diamond, M.P., Hallerman, L., Starick-Zych, K. et al. (1991) Suppression of counterregulatory hormone response to hypoglycemia by insulin per se. Journal of Clinical Endocrinology and Metabolism, 72, 1388–1390. Donnelly, L.A., Morris, A.D., Frier, B.M. et al. (2005) Frequency and predictors of hypoglycaemia in Type 1 and insulin-treated Type 2 diabetes: a population-based study. Diabetic Medicine, 22, 749–755. Drucker, D., Sherman, S.I., Gorelick, F.S., Bergenstal, R.M., Sherwin, R.S. & Buse, J.B. (2010) Incretin- based therapies for the treatment of type 2 diabetes: evaluation of the risks and benefits. Diabetes Care, 33, 428–433. 256 Hypoglycaemia in Clinical Diabetes

Fanelli, C., Pampanelli, S., Epifano, L. et al. (1994) Relative roles of insulin and hypoglycaemia on induc- tion of neuroendocrine responses to, symptoms of, and deterioration of cognitive function in hypogly- caemia in male and female humans. Diabetologia, 37, 797–807. Fatourechi, M.M., Kudva, Y.C., Murad, M.H., Elamin, M.B., Tabini, C.C. & Montori, V.M. (2009) Hypoglycemia with intensive insulin therapy: a systematic review and meta-analysis of randomized trials of continuous subcutaneous insulin infusion versus multiple insulin injections. Journal of Clinical Endocrinology and Metabolism, 94, 729–740. Ferrannini, E., Fonseca, V., Zinman, B. et al. (2009) Fifty-two-week efficacy and safety of vildagliptin vs.glimepiride in patients with type 2 diabetes mellitus inadequately controlled on metformin mono- therapy. Diabetes, Obesity and Metabolism, 11, 157–166. Fineman, M.S., Bicsak, T.A., Shen, L.Z. et al. (2003) Effect on glycemic control of exenatide (synthetic exendin-4) additive to existing metformin and/or sulfonylurea treatment in patients with type 2 diabetes. Diabetes Care, 26, 2370–2377. Fink, R.I., Revers, R.R., Kolterman, O.G. & Olefsky, J.M. (1985) The metabolic clearance of insulin and the feedback inhibition of insulin secretion are altered with ageing. Diabetes, 34, 275–280. Freathy, R.M., Lonnen, K.F., Steele, A.M. et al. (2006) The impact of the Angiotensin-converting enzyme insertion/deletion polymorphism on severe hypoglycemia in type 2 diabetes. The Review of Diabetic Studies, 3, 76–81. de Galan, B.E. & Hoekstra, J.B.L. (2001) Glucose counterregulation in Type 2 diabetes mellitus. Diabetic Medicine, 18, 519–527. de Galan, B.E., Zoungas, S., Chalmers, J. et al. (2009) Cognitive function and risks of cardiovascular diease and hypoglycaemia in patients with type 2 diabetes: the Action in Diabetes and Vascular Disease: Pre- terax and Diamicron Modified Release Controlled Evaluation (ADVANCE) trial. Diabetologia, 52, 2328–2336. Gangji, A.S., Cukierman, T., Gerstein, H.C., Goldsmith, C.H. & Clase, C.M. (2007) A systematic review and meta-analysis of hypoglycemia and cardiovascular events: a comparison of glyburide with other insulin secretagogues and with insulin. Diabetes Care, 30, 389–394. Gitt, A.K., Bramlage, P., Binz, C. et al. (2012) Hypoglycaemia is more frequent in type 2 diabetic patients with co-morbid vascular disease: an analysis of the DiaRegis registry. European Journal of Preventive Cardiology, 19, 765–772. Goke, B., Gallwitz, B., Eriksson, J., Hellqvist, A., Gause-Nilsson, I. & for the D1680C00001 Investigators (2010) Saxagliptin is non-inferior to glipizide in patients with type 2 diabetes mellitus inadequately controlled on metformin alone: a 52-week randomised controlled trial. International Journal of Clinical Practice, 64, 1619–1631. Gold, A.E., MacLeod, K.M. & Frier, B.M. (1994) Frequency of severe hypoglycemia in patients with type 1 diabetes with impaired awareness of hypoglycemia. Diabetes Care, 17, 697–703. Goudswaard, A., Furlong, N.J., Rutten, G., Stolk, R. & Valk, G. (2004) Insulin monotherapy vs. combina- tions of insulin with oral hypoglycaemic agents in patients with type 2 diabetes mellitus (Cochrane review). Cochrane Database of Systematic Reviews, (4), CD003418. Graveling, A.J. & Frier, B.M. (2009) Dementia and hypoglycemic episodes in patients with type 2 diabetes mellitus. Journal of the American Medical Association, 302, 843. (letter). Greco, D. & Angileri, G. (2004) Drug-induced severe hypoglycaemia in type 2 diabetic patients aged 80 years or older. Diabetes, Nutrition and Metabolism, 17, 23–26. Haak, T., Tiengo, A., Draeger, E., Suntum, M. & Waldhausl, W. (2005) Lower within-subject variability of fasting blood glucose and reduced weight gain with insulin detemir compared to NPH insulin in patients with type 2 diabetes. Diabetes, Obesity and Metabolism, 7, 56–64. Hammer, M., Lammert, M., Monereo Mejias, S., Kern, W. & Frier, B.M. (2009) Costs of managing severe hypoglycaemia in three European countries. Journal of Medical Economics, 12, 281–290. Harrigan, R.A., Nathan, M.S. & Beattie, P. (2001) Oral agents for the treatment of type 2 diabetes mellitus: pharmacology, toxicity and treatment. Annals of Emergency Medicine, 38, 68–78. Harrower, A. (2000) Gliclazide modified release: from once-daily administration to 24-hour blood glucose control. Metabolism: Clinical and Experimental, 49 (Suppl. 2), 7–11. Hartling, S.G., Faber, O.K., Wegmann, M.-L., Wahlin-Boll, E. & Melander, A. (1987) Interaction of ethanol and glipizide in humans. Diabetes Care, 10, 683–686. Hay, L.C., Wilmhurst, E.G. & Fulcher, G. (2003) Unrecognized hypo- and hyperglycemia in well-controlled patients with type 2 diabetes mellitus: the results of continuous glucose monitoring. Diabetes Technology and Therapeutics, 5, 19–26. Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 257

Heller, S.R., Macdonald, I.A. & Tattersall, R.B. (1987) Counterregulation in Type II (non-insulin-depend- ent) diabetes mellitus. Normal endocrine and glycaemic responses up to 10 years after diagnosis. Dia­ betologia, 30, 924–929. Henderson, J.N., Allen, K.V., Deary, I.J. & Frier, B.M. (2003) Hypoglycaemia in insulin-treated Type 2 diabetes: frequency, symptoms and impaired awareness. Diabetic Medicine, 20, 1016–1021. Hepburn, D.A., MacLeod, K.M., Pell, A.C.H., Scougal, I.J. & Frier, B.M. (1993) Frequency and symptoms of hypoglycaemia experienced by patients with type 2 diabetes treated with insulin. Diabetic Medicine, 10, 231–237. Herman, W.H., Ilag, L.L., Johnson, S.L. et al. (2005) A clinical trial of continuous subcutaneous insulin infusion versus multiple daily injections in older adults with type 2 diabetes. Diabetes Care, 28, 1568–1573. Hermansen, K., Davies, M., Derezinski, T. et al. (2006) A 26-week, randomised, parallel, treat-to-target trial comparing insulin detemir with NPH insulin as add-on therapy to oral glucose-lowering drugs in insulin-naïve people with type 2 diabetes. Diabetes Care, 29, 1269–1274. Hewitt, J., Smeeth, L., Chaturvedi, N., Bulpitt, C.J. & Fletcher, A.E. (2011) Self-management and patient understanding of diabetes in the older person. Diabetic Medicine, 28, 117–122. Holstein, A., Plaschke, A. & Egberts, E.-H. (2001) Lower incidence of severe hypoglycemia in patients with type 2 diabetes treated with glimepiride versus glibenclamide. Diabetes/Metabolism Research and Reviews, 17, 467–473. Holstein, A., Hammer, C., Hahn, M., Kulmadavil, N.S. & Kovacs, P. (2010) Severe sulfonylurea-induced hypoglycemia: a problem of uncritical prescription and deficiencies of diabetes care in geriatric patients. Expert Opinion in Drug Safety, 9, 675–681. Holstein, A., Hahn, M., Patzer, O., Seeringer, A., Kovacs, P. & Stingl, J. (2011) Impact of clinical factors and CYP2C9 variants for the risk of severe sulfonylurea-induced hypoglycemia. European Journal of Clinical Pharmacology, 67, 471–476. Horvath, K., Jeitler, K., Berghold, A. et al. (2007) Long-acting insulin analogues versus NPH insulin (human isophane insulin) for the treatment of type 2 diabetes mellitus. Cochrane Library, (2), CD005613. Israelian, Z., Gosmanov, N.R., Szoke, E. et al. (2005) Increasing the decrement in insulin secretion improves glucagon responses to hypoglycemia in advanced type 2 diabetes. Diabetes Care, 28, 2691–2696. Israelian, Z., Szoke, E., Woerle, J. et al. (2006) Multiple defects in counterregulation of hypoglycemia in modestly advanced type 2 diabetes mellitus. Metabolism: Clinical and Experimental, 55, 593–598. Jaap, A.J., Jones, G.C., McCrimmon, R.J., Deary, I.J. & Frier, B.M. (1998) Perceived symptoms of hypogly- caemia in elderly Type 2 diabetic patients treated with insulin. Diabetic Medicine, 15, 398–401. Janka, H.U., Plewe, G., Riddle, M.C., Klieb-Frisch, C., Schweitzer, M.A. & Yki-Järvinen, H. (2005) Comparison of basal insulin added to oral agents versus twice-daily premixed insulin as initial therapy for type 2 diabetes. Diabetes Care, 28, 254–259. Jennings, A.M., Wilson, R.M. & Ward, J.D. (1989) Symptomatic hypoglycemia in NIDDM patients treated with oral hypoglycemic agents. Diabetes Care, 12, 203–207. Jermendy, G., Hungarian RECAP Group, Erdesz, D. et al. (2008) Outcomes of adding second hypoglyc- emic drug after metformin monotherapy failure among type 2 diabetic patients in Hungary. Health Quality of Life Outcomes, 6, 88. Jonsson, A., Chan, J.C., Rydberg, T. et al. (2001) Pharmacodynamics and pharmacokinetics of intravenous glibenclamide in Caucasian and Chinese patients with type-2 diabetes. European Journal of Clinical Pharmacology, 55, 721–727. Knop, F.K., Vilsbøll, T., Larsen, S., Madsbad, S., Holst, J. & Krarup, T. (2003) No reactive hypoglycemia after subcutaneous administration of glucagon-like peptide-1 in lean type 2 diabetic patients and in patients with diabetes secondary to chronic pancreatitis. Diabetes Care, 26, 2581–2587. Kobayashi, K., Kimura, M., Sakuguchi, T., Hase, A. & Matsuoka, A. (1984) Pharmacokinetics of gliclazide in healthy and diabetic subjects. Journal of Pharmacological Sciences, 73, 1684–1687. Kolterman, O.G., Buse, J.B., Fineman, M.S. et al. (2003) Synthetic Exendin-4 (exenatide) significantly reduces postprandial and fasting plasma glucose in subjects with type 2 diabetes. Journal of Clinical Endocrinology and Metabolism, 88, 3082–3088. Korzon-Burakowska, A., Hopkins, D., Matyka, K. et al. (1998) Effects of glycemic control on protective responses against hypoglycemia in type 2 diabetes. Diabetes Care, 21, 283–290. Landstedt-Hallin, L., Adamson, U. & Lins, P.-E. (1999) Oral glibenclamide suppresses glucagon secretion during insulin-induced hypoglycemia in patients with type 2 diabetes. Journal of Clinical Endocrinology and Metabolism, 84, 3140–3145. 258 Hypoglycaemia in Clinical Diabetes

Larsen, J.J., Dela, F., Kjaer, M. & Galbo, H. (1997) The effect of moderate exercise on postprandial glucose homeostasis in NIDDM patients. Diabetologia, 40, 447–453. Lawrence, P.A. & Cheely, J. (1980) Deterioration of diabetic patients’ knowledge and management skills as determined during out-patient visits. Diabetes Care, 3, 214–218. Leese, G.P., Wang, J., Broomhall, J. et al. (2003) Frequency of severe hypoglycemia requiring emergency treatment in Type 1 and Type 2 diabetes: a population-based study of health service resource use. Dia­ betes Care, 26, 1176–1180. Leiter, L.A., Yale, J.-F., Chiasson, J.-L., Harris, S.B., Kleinstiver, P. & Sauriol, L. (2005) Assessment of the impact of fear of hypoglycemic episodes on glycemic and hypoglycemia management. Canadian Journal of Diabetes, 29, 186–192. Levy, C.J., Kinsley, B.T., Bajaj, M. & Simonson, D.C. (1998) Effect of glycemic control on glucose coun- terregulation during hypoglycemia in NIDDM. Diabetes Care, 21, 1330–1338. Liu, D., Moberg, E., Kollind, M., Lins, P.E. & Adamson, U. (1991a) A high concentration of circulating insulin suppresses the glucagon response to hypoglycemia in normal man. Journal of Clinical Endo­ crinology and Metabolism, 73, 1123–1128. Liu, D.T., Adamson, U.C., Lins, P.E., Kollind, M.E., Moberg, E.A. & Andreasson, K. (1991b) Inhibitory effect of circulating insulin on glucagon secretion during hypoglycemia in type I diabetic patients. Diabetes Care, 15, 59–65. Luddeke, H.-J., Sreenan, S., Aczel, S. et al. (2007) PREDICTIVE - a global, prospective observational study to evaluate insulin detemir treatment in types 1 and 2 diabetes: baseline characteristics and predic- tors of hypoglycaemia from the European cohort. Diabetes, Obesity and Metabolism, 9, 428–434. MacLeod, K.M., Hepburn, D.A. & Frier, B.M. (1993) Frequency and morbidity of severe hypoglycaemia in insulin-treated diabetic patients. Diabetic Medicine, 10, 238–245. Madsbad, S., Schmitz, O., Ranstam, J., Jakobsen, G., Matthews, D.R. & on behalf of the NN2211-1310 international study group (2004) Improved glycemic control with no weight increase in patients with type 2 diabetes after once-daily treatment with the long-acting Glucagon-Like Peptide 1 analog Liraglutide (NN2211): a 12-week, double-blind, randomized, controlled trial. Diabetes Care, 27, 1335–1342. Maggs, D.G., Jacob, R., Rife, F., Caprio, S., Tamborlane, W.V. & Sherwin, R.S. (1997) Counterregulation in peripheral tissues: effect of systemic hypoglycemia on levels of substrates and catecholamines in human skeletal muscle and adipose tissue. Diabetes, 46, 70–76. Marker, J.C., Cryer, P.E. & Clutter, W.E. (1992) Attenuated glucose recovery from hypoglycemia in the elderly. Diabetes, 41, 671–678. Marrett, E., Radican, L., Davies, M.J. & Zhang, Q. (2011) Assessment of severity and frequency of self- reported hypoglycemia on quality of life in patients with type 2 diabetes treated with oral antihypergly- cemic agents: a survey study. BMC Research Notes, 4, 251–258. Matyka, K., Evans, M., Lomas, J., Cranston, I., Macdonald, I. & Amiel, S.A. (1997) Altered hierarchy of protective responses against severe hypoglycemia in normal aging in healthy men. Diabetes Care, 20, 135–141. McAulay, V. & Frier, B.M. (2001) Hypoglycaemia. In: A.J. Sinclair & P. Finucane (eds), Diabetes in Old Age, 2nd edn, pp. 133–152. John Wiley and Sons, Chichester. McAulay, V. & Frier, B.M. (2003) Insulin analogues and other developments in insulin therapy for diabetes. Expert Opinion in Pharmacotherapy, 4, 1141–1156. McAulay, V., Deary, I.J. & Frier, B.M. (2001) Symptoms of hypoglycaemia in people with diabetes. Dia­ betic Medicine, 18, 690–705. Meier, J.J. & Nauck, M.A. (2005) Glucagon-like peptide 1 (GLP-1) in biology and pathology. Diabetes/ Metabolism Research and Reviews, 21, 91–117. Meneilly, G.S., Cheung, E. & Tuokko, H. (1994a) Altered responses to hypoglycemia of healthy elderly people. Journal of Clinical Endocrinology and Metabolism, 78, 1341–1348. Meneilly, G.S., Cheung, E. & Tuokko, H. (1994b) Counterregulatory hormone responses to hypoglycemia in the elderly patient with diabetes. Diabetes, 43, 403–410. Miller, C.D., Philips, L.S., Ziemer, D.C., Gallina, D.L., Cook, C.B. & El-Kebbi, I.M. (2001) Hypoglycemia in patients with Type 2 diabetes mellitus. Archives of Internal Medicine, 161, 1653–1659. Minaker, K.L., Rowe, J.W., Torino, R. & Pallotta, J.A. (1982) Influence of age on clearance of insulin in man. Diabetes, 31, 851–855. Minuk, H.L., Vranic, M., Marliss, E.B., Hanna, A.K., Albisser, A.M. & Zinman, B. (1981) Glucoregulatory and metabolic response to exercise in obese noninsulin-dependent diabetes. The American Journal of Physiology, 240, E458–E464. Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 259

Monnier, L., Wojtusciszyn, A., Colette, C. & Owens, D. (2011) The contribution of glucose variability to asymptomatic hypoglycemia in persons with type 2 diabetes. Diabetes Technology and Therapeutics, 13, 813–818. Munshi, M., Hayes, M., Sternthal, A. & Ayres, A. (2009) Use of serum C-peptide level to simplify diabetes treatment regimens in older adults. The American Journal of Medicine, 122, 395–397. Munshi, M., Segal, A.R., Suhl, E. et al. (2011) Frequent hypoglycemia among elderly patients with poor glycemic control. Archives of Internal Medicine, 171, 362–364. Murata, G.H., Hoffman, R.M., Shah, J.H., Wendel, C.S. & Duckworth, W.C. (2004) A probabilistic model for predicting hypoglycemia in type 2 diabetes mellitus. Archives of Internal Medicine, 164, 1445–1450. Mutch, W.J. & Dingwall-Fordyce, I. (1985) Is it a hypo? Knowledge of the symptoms of hypoglycaemia in elderly diabetic patients. Diabetic Medicine, 2, 54–56. Nathan, D.M., Buse, J.B., Davidson, M.B. et al. (2009) Medical management of hyperglycemia in type 2 diabetes: a consensus algorithm for the initiation and adjustment of therapy: a consensus statement of the American Diabetes Association and the European Association for the Study of Diabetes. Diabetes Care, 32, 193–203. National Institute for Health and Clinical Excellence (2009) Type 2 diabetes: the managment of type 2 diabetes. National Institute for Health and Clinical Excellence. URL http://www.nice.org.uk/CG87 [accessed on 16 June 2013]. Nattrass, M. & Lauritzen, T. (2000) Review of prandial glucose regulation with repaglinide: a solution to the problem of hypoglycemia in the treatment of Type 2 diabetes? International Journal of Obesity and Related Metabolic Diseases, 24 (Suppl. 3), S21–S31. Nauck, M.A., Heimesaat, M.M., Behle, K. et al. (2002) Effects of glucagon-like peptide 1 on counterregu- latory hormone responses, cognitive functions and insulin secretion during hyperinsulinemic stepped hypoglycemic clamp experiments on healthy volunteers. Journal of Clinical Endocrinology and Metabo­ lism, 87, 1239–1246. Ortiz-Alonso, F.J., Galecki, A., Herman, W.H., Smith, M.J., Jacquez, J.A. & Halter, J.B. (1994) Hypogly- cemia counterregulation in elderly humans: relationship to glucose levels. The American Journal of Physiology, 267, E497–E506. Oyer, D.S., Shepherd, M.D., Coulter, F.C. et al. (2009) A1c control in a primary care setting: self-titrating an insulin analoge pre-mix (INITIATEplus Trial). The American Journal of Medicine, 122, 1043– 1049. Peacey, S.R., George, E., Rostami-Hodjegan, A. et al. (1996) Similar physiological and symptomatic responses to sulphonylurea and insulin induced hypoglycaemia in normal subjects. Diabetic Medicine, 13, 634–641. Pedersen-Bjergaard, U., Agerholm-Larsen, B., Pramming, S., Hougaard, P. & Thorsteinsson, B. (2001) Activity of angiotensin-converting enzyme and risk of severe hypoglycaemia in type 1 diabetes mellitus. Lancet, 357, 1248–1253. Pedersen-Bjergaard, U., Agerholm-Larsen, B., Pramming, S., Hougaard, P. & Thorsteinsson, B. (2003a) Prediction of severe hypoglycaemia by angiotensin-converting enzyme activity and genotype in type 1 diabetes mellitus. Diabetologia, 46, 89–96. Pedersen-Bjergaard, U., Pramming, S. & Thorsteinsson, B. (2003b) Recall of severe hypoglycaemia and self-estimated state of awareness in Type 1 diabetes. Diabetes/Metabolism Research and Reviews, 19, 232–240. Pedersen-Bjergaard, U., Pramming, S., Heller, S.R. et al. (2004) Severe hypoglycaemia in 1076 adult patients with Type 1 diabetes: influence of risk markers and selection. Diabetes/Metabolism Research and Reviews, 20, 479–486. Pegg, A., Fitzgerald, D., Wise, D., Singh, B.M. & Wise, P.H. (1991) A community-based study of diabetes- related skills and knowledge in elderly people with insulin-requiring diabetes. Diabetic Medicine, 8, 778–781. Phung, O.J., Scholle, J.M. & Talwar, M. (2010) Effect of noninsulin antidiabetic drugs added to metformin therapy on glycemic control, weight gain, and hypoglycemia in type 2 diabetes. Journal of the American Medical Association, 303, 1410–1418. Pramming, S., Thorsteinsson, B., Bendtson, I. & Binder, C. (1991) Symptomatic hypoglycaemia in 411 Type 1 diabetic patients. Diabetic Medicine, 8, 217–222. Ragia, G., Petridis, I., Tavridou, A., Christakidis, D. & Manolopoulos, V.G. (2009) Presence of CYP2C9*3 allele increases risk for hypoglycemia in type 2 diabetic patients treated with sulphonylureas. Pharma­ cogenomics, 10, 1781–1787. 260 Hypoglycaemia in Clinical Diabetes

Raskin, P., Allen, E., Hollander, P. et al. (2005) Initiating insulin therapy in type 2 diabetes: a comparison of biphasic and basal insulin analogs. Diabetes Care, 26, 260–265. Raslova, K., Bogoev, M., Raz, I., Leth, G., Gall, M.A. & Hancu, N. (2004) Insulin detemir and insulin aspart: a promising basal-bolus regimen for type 2 diabetes. Diabetes Research and Clinical Practice, 66, 193–201. Rasmussen, B.M., Christiansen, C., Rasmussen, O.W., Hansen, C. & Hermansen, K. (1999) Alcohol and postexercise metabolic responses in type 2 diabetes. Metabolism: Clinical and Experimental, 48, 597–602. Rasmussen, B.M., Orskov, L., Schmitz, O. & Hermansen, K. (2001) Alcohol and glucose counterregulation during acute insulin-induced hypoglycemia in type 2 diabetic subjects. Metabolism: Clinical and Experi­ mental, 50, 451–457. Reaven, G.M., Greenfield, M.S., Mondon, C.E., Rosenthal, M., Wright, D. & Reaven, E.P. (1982) Does insulin removal rate from plasma decline with age? Diabetes, 31, 670–673. Rendell, M. (2004) The role of sulphonylureas in the management of type 2 diabetes mellitus. Drugs, 64, 1339–1358. Riddle, M.C., Rosenstock, J., Gerich, J. & on behalf of the insulin glargine 4002 study investigators (2003) Randomized addition of glargine or human NPH insulin to oral therapy of type 2 diabetic patients. Diabetes Care, 26, 3080–3086. Rosenstock, J. & on behalf of the HOE 901/204 Study Investigators Group (2003) Safety and efficacy of insulin glargine (HOE 901) versus NPH insulin in combination with oral treatment in type 2 diabetic patients. Diabetic Medicine, 20, 545–551. Rosenstock, J., Schwartz, S.L., Clark, C.M., Park, G.D., Donley, D.W. & Edwards, M.B. (2001) Basal insulin therapy in Type 2 diabetes. Diabetes Care, 24, 631–636. Rosenstock, J., Hassman, D.R., Madder, R.D. et al. (2004) Repaglinide versus nateglinide monotherapy. Diabetes Care, 27, 1265–1270. Ross, S.A., Zinman, B., Campos, R.V., Strack, T. & Canadian Lispro Study Group (2001) A comparative study of insulin lispro and human regular insulin in patients with type 2 diabetes mellitus and secondary failure of oral hypoglycemic agents. Clinical Investigative Medicine, 24, 292–298. Sandoval, D.A., Ertl, A.C., Richardson, M.A., Tate, D.B. & Davis, S.N. (2003) Estrogen blunts neuroen- docrine and metabolic responses to hypoglycemia. Diabetes, 52, 1749–1755. Sarkar, U., Karter, A.J., Liu, J.Y., Moffet, H.H., Adler, N.E. & Schillinger, D. (2010) Hypoglycemia is more common among type 2 diabetes patients with limited health literacy: the diabetes study of Northern California (DISTANCE). Journal of General Internal Medicine, 25, 962–968. Schejter, Y.D., Turvall, E. & Ackerman, Z. (2012) Characteristics of patients with sulfonylurea-induced hypoglycemia. Journal of the American Medical Directors Association, 13, 234–238. Schernthaner, G. (2003) Gliclazide Modified Release: a critical review of pharmacodynamic, metabolic and vasoprotective effects. Metabolism: Clinical and Experimental, 52 (Suppl. 1), 29–34. Schernthaner, G., Grimaldi, A., Di Mario, U. et al. (2004a) GUIDE study: double blind comparison of once-daily gliclazide MR and glimepiride in type 2 diabetic patients. European Journal of Clinical Investigation, 34, 535–542. Schernthaner, G., Kopp, H.P., Ristic, S., Muzyka, B., Peter, L. & Mitteregger, G. (2004b) Metabolic control in patients with type 2 diabetes using Humalog Mix50 injected three times daily: crossover comparison with human insulin 30/70. Hormone Metabolism Research, 36, 188–193. Schopman, J.E., Geddes, J. & Frier, B.M. (2010) Prevalence of impaired awareness of hypoglycaemia and frequency of hypoglycaemia in insulin-treated type 2 diabetes. Diabetes Research and Clinical Practice, 87, 64–68. Schouwenberg, B.J., Veldman, B.A., Spiering, W. et al. (2008) The Arg16Gly variant of the beta2- adrenergic receptor predisposes to hypoglycemia unawareness in type 1 diabetes. Pharmacogenetic Genomics, 18, 369–372. Schutt, M., Fach, E.-M., Seufert, J. et al. (2012) Multiple complications and frequent severe hypoglycaemia in “elderly” and “old” patients with Type 1 diabetes. Diabetic Medicine, 29, e176–e179. Schwartz, N.S., Clutter, W.E., Shah, S.D. & Cryer, P.E. (1987) Glycemic thresholds for activation of glucose counterregulatory systems are higher than the thresholds for symptoms. The Journal of Clinical Inves­ tigation, 79, 777–781. Scottish Intercollegiate Guidelines Network (SIGN) (2010) Management of diabetes. SIGN publication number 116 [WWW document]. URL http://www.sign.ac.uk/guidelines/fulltext/116/index.html [accessed on 31 July 2013]. Chapter 12: Hypoglycaemia in Type 2 Diabetes and in Elderly People 261

Seck, T., Nauck, M., Sheng, D. et al. (2010) Safety and efficacy of treatment with sitagliptin or glipizide in patients with type 2 diabetes inadequately controlled on metformin: a 2-year study. International Journal of Clinical Practice, 64, 562–576. Segel, S.A., Paramore, D.A. & Cryer, P.E. (2002) Hypoglycemia-Associated Autonomic Failure in advanced Type 2 diabetes. Diabetes, 51, 724–733. Seltzer, H.S. (1989) Drug-induced hypoglycemia. Endocrinology and Metabolism Clinics of North America, 18, 163–183. Shamoon, H., Friedman, A., Canton, C., Zacharowicz, L., Hu, M. & Rossetti, L. (1994) Increased epine- phrine and skeletal muscle responses to hypoglycemia in non-insulin-dependent diabetes mellitus. The Journal of Clinical Investigation, 93, 2562–2571. Shorr, R.I., Ray, W.A., Daugherty, J.R. & Griffin, M.R. (1997) Incidence and risk factors for serious hypoglycemia in older persons using insulin or sulfonylureas. Archives of Internal Medicine, 157, 1681–1686. So, W.Y., Chan, J.C.N., Yeung, V.T.F. et al. (2002) Sulphonylurea-induced hypoglycaemia in institutional- ized elderly in Hong Kong. Diabetic Medicine, 19, 966–971. Spyer, G., Hattersley, A., Macdonald, I.A., Amiel, S. & MacLeod, K.M. (2000) Hypoglycemic counter- regulation at normal blood glucose concentrations in patients with well controlled Type 2 diabetes. Lancet, 356, 1970–1974. Stahl, M. & Berger, W. (1999) Higher incidence of severe hypoglycaemia leading to hospital admission in Type 2 diabetic patients treated with long-acting versus short-acting sulphonylureas. Diabetic Medicine, 16, 586–590. Stargardt, T., Gonder-Frederick, L., Krobot, K.J. & Alexander, C.M. (2009) Fear of hypoglycaemia: defin- ing a minimum clinically important difference in patients with type 2 diabetes. Health Quality of Life Outcomes, 7, 91–98. Stephenson, J., Fuller, J.H. on behalf of the EURODIAB IDDM Complications Study Group (1994) Micro- vascular and acute complications in IDDM patients: the EURODIAB IDDM complications study. Diabetologia, 37, 278–285. Strange, P., Schwartz, S.L., Graf, R.J. et al. (1999) Pharmacokinetics, pharmacodynamics, and dose- response relationship of repaglinide in type 2 diabetes. Diabetes Technology and Therapeutics, 1, 247–256. ter Braak, E.W.M.T., Appelman, A.M.M.F., van de Laak, M.F., Stolk, R.P., van Haeften, T.W. & Erkelens, D.W. (2000) Clinical characteristics of Type 1 diabetic patients with and without severe hypoglycemia. Diabetes Care, 23, 1467–1471. ter Braak, E.W.M.T., Appelman, A.M.M.F., van Der Tweel, I., Erkelens, D.W. & van Haeften, T.W. (2002) The sulfonylurea glibenclamide induces impairment of glucagon and growth hormone responses during mild insulin-induced hypoglycemia. Diabetes Care, 25, 107–112. Tessier, D., Dawson, K., Tetrault, J.P., Bravo, G. & Meneilly, G.S. (1994) Glibenclamide versus gliclazide in Type 2 diabetes of the elderly. Diabetic Medicine, 11, 974–980. The Action to Control Cardiovascular Risk in Diabetes Study Group (2008) Effects of intensive glucose lowering in type 2 diabetes. New England Journal of Medicine, 358, 2545–2559. The Diabetes Control and Complications Trial Research Group (1993) The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. New England Journal of Medicine, 329, 977–986. Thomson, F.J., Masson, E.A., Leeming, J.T. & Boulton, A.J.M. (1991) Lack of knowledge of symptoms of hypoglycaemia by elderly diabetic patients. Age and Ageing, 20, 404–406. Trovati, M., Carta, Q., Cavalot, F. et al. (1984) Influence of physical training on blood glucose control, glucose tolerance, insulin secretion and insulin action in non-insulin-dependent diabetic patients. Diab­ etes Care, 7, 416–420. UK Hypoglycaemia Study Group (2007) Risks of hypoglycaemia in types 1 and 2 diabetes: effects of treatment modalities and their duration. Diabetologia, 50, 1140–1147. United Kingdom Prospective Diabetes Study Group (1998a) Effect of intensive blood-glucose control with metformin on complications in overweight patients with Type 2 diabetes (UKPDS 34). Lancet, 352, 854–865. United Kingdom Prospective Diabetes Study Group (1998b) A 6 year randomised controlled trial comparing sulfonylurea, insulin and metformin therapy in patients with newly diagnosed Type 2 diabetes that could not be controlled with diet therapy (UKPDS 24). Annals of Internal Medicine, 128, 165–175. 262 Hypoglycaemia in Clinical Diabetes

United Kingdom Prospective Diabetes Study Group (1998c) Intensive blood-glucose control with sulpho- nylureas or insulin compared with conventional treatment and risk of complications in patients with Type 2 diabetes (UKPDS 33). Lancet, 352, 837–852. Vexiau, P., Mavros, P., Krishnarajah, G., Lyu, R. & Yin, D. (2008) Hypoglycaemia in patients with type 2 diabetes treated with a combination of metformin and sulphonylurea therapy in France. Diabetes, Obesity and Metabolism, 10 (Suppl. 1), 16–24. Vilsbøll, T., Krarup, T., Madsbad, S. & Holst, J.J. (2001) No reactive hypoglycemia in type 2 diabetic patients after subcutaneous administration of GLP-1 and intravenous glucose. Diabetes Care, 18, 144–149. Waugh, N., Cummins, E., Royle, P. et al. (2010) Newer agents for blood glucose control in type 2 diabetes. Health Technology Assessments, 14 (36). Whitmer, R.A., Karter, A.J., Yaffe, K., Quesenberry, C.P., Jr. & Selby, J. (2009) Hypoglycemic episodes and risk of dementia in older patients with type 2 diabetes. Journal of the American Medical Association, 301, 1565–1572. Williams, S.A., Pollack, M.F. & Dibonaventura, M. (2011) Effects of hypoglycemia on health-related quality of life, treatment satisfaction and healthcare resource utilization in patients with type 2 diabetes mellitus. Diabetes Research and Clinical Practice, 91, 363–370. Woerle, H.J., Meyer, C., Popa, E.M., Cryer, P.E. & Gerich, J.E. (2003) Renal compensation for impaired hepatic glucose release during hypoglycemia in type 2 diabetes. Further evidence for hepatorenal reci- procity. Diabetes, 52, 1386–1392. Wright, R.J. & Frier, B.M. (2008) Vascular disease and diabetes: is hypoglycaemia an aggravating factor? Diabetes/Metabolism Research and Reviews, 24, 353–363. Yaffe, K., Falvey, C.M., Hamilton, N., et al. (2013) Association between hypoglycemia and dementia in a biracial cohort of older adults with diabetes mellitus. Journal of the American Medical Association, Internal Medicine, published on-line doi:10.1001/jamainternmed.2013.6176 Yki-Järvinen, H., Dressler, A., Ziemen, M. & the HOE 901/3002 study group (2000) Less nocturnal hypoglycemia and better post-dinner glucose control with bedtime insulin glargine compared with bedtime NPH insulin during insulin combination therapy in Type 2 diabetes. Diabetes Care, 23, 1130–1136. Zammitt, N.N. & Frier, B.M. (2005) Hypoglycemia in type 2 diabetes: pathophysiology, frequency and effects of different treatment modalities. Diabetes Care, 28, 2948–2961. Zammitt, N.N., Geddes, J., Warren, R.E., Marioni, R., Ashby, J.P. & Frier, B.M. (2007) Serum angiotensin- converting enzyme and frequency of severe hypoglycaemia in Type 1 diabetes: does a relationship exist? Diabetic Medicine, 24, 1449–1454. Zander, M., Madsbad, S., Madsen, J.L. & Holst, J.J. (2002) Effect of 6-week course of glucagon-like peptide 1 on glycaemic control, insulin sensitivity and beta-cell function in type 2 diabetes: a parallel group study. Lancet, 359, 824–830. Zhang, Y., Wieffer, H., Modha, R., Balar, B., Pollack, M. & Krishnarajah, G. (2010) The burden of hypogly- caemia in type 2 diabetes: a systematic review of patient and economic perspectives. Journal of Clinical Outcomes Management, 17, 547–557. Zoungas, S., Patel, A., Chalmers, J. et al. (2010) Severe hypoglycemia and risks of vascular events and death. The New England Journal of Medicine, 363, 1410–1418. 13 Mortality, Cardiovascular Morbidity and Possible Effects of Hypoglycaemia on Diabetic Complications Elaine Chow1, Miles Fisher2 and Simon R. Heller1 1 University of Sheffield, Sheffield, UK 2 University of Glasgow, Glasgow, UK

INTRODUCTION

The risk of hypoglycaemia restrains many patients from pursuing intensive insulin therapy and trying to achieve the levels of strict glycaemic control that are necessary to prevent the development of microvascular diabetic complications. Many patients share the worries expressed by some diabetes healthcare professionals about the possible long-term effects of recurrent hypoglycaemia on the brain (see Chapter 14). An additional factor that may dissuade people from improving their glycaemic control is the fear of dying during an episode of hypoglycaemia, especially when low blood glucose occurs during sleep. Recent trials of intensive glycaemic therapy in people with type 2 diabetes have suggested a pos- sible association between hypoglycaemia and an increased risk of death. In this chapter, mortality that can be directly attributed to hypoglycaemia is reviewed. The phenomenon of the ‘dead-in-bed’ syndrome, a particular pattern of sudden nocturnal death that has been associated with hypoglycaemia in individuals with type 1 diabetes, is discussed. In addition, the association is explored between hypoglycaemia and the increase in mortality that has been observed in trials of intensive glycaemic therapy in patients with type 2 diabetes and other contexts. Putative mechanisms by which hypoglycaemia can aggravate microvascular and macrovascular complications are also described.

DEATHS DIRECTLY ATTRIBUTED TO HYPOGLYCAEMIA

Crude estimates of mortality secondary to hypoglycaemia Many episodes of hypoglycaemia, including nocturnal hypoglycaemia, are not recognised by patients, their relatives or carers. If we add the problems of death certification and confirming hypoglycaemia post-mortem, it is not surprising that considerable uncertainty and variation surround the estimated number of deaths attributed to hypoglycaemia in people with diabetes. The cause of death is often recorded inaccurately on the death certificate, even to the extent of omitting ‘diabetes’ altogether (Muhlhauser et al. 2002; Waernbaum et al. 2006). Subsequent problems with analysis of death certificates can be

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 264 Hypoglycaemia in Clinical Diabetes compounded in the UK by the fact that ‘hypoglycaemia’ is not coded under one single heading. Furthermore, the exact coding of the cause of death is often left to the discretion of individual coding clerks who seldom have any knowledge of the clinical details (Tat- tersall and Gale 1993). Carbohydrate metabolism continues after death, and post-mortem changes in blood glucose can subvert confirmation of a suspected hypoglycaemic death. The continuing breakdown of glycogen (glycogenolysis) increases the blood glucose con- centration in the inferior vena cava, so that the presence of a normal or high blood glucose concentration on the right side of the heart does not exclude ante-mortem hypoglycaemia (a false negative result for a diagnosis of hypoglycaemia). Glucose continues to be utilised in the peripheral circulation and vitreous humour. Indeed low blood glucose is often found after death in those without diabetes (a false positive result for a diagnosis of hypo­ glycaemia). In addition to the biochemical problems in diagnosing hypoglycaemia after death, errors may be introduced by attribution bias of the pathologist performing the post-mortem. Estimates of mortality from hypoglycaemia vary widely, from some reports indicating no deaths, to between 20 and 25% in some Scandinavian centres. Studies suggest the pro- portion of deaths caused by hypoglycaemia in young adults is between 7 and 10% (Skri- varhaug et al. 2006; Feltbower et al. 2008), slightly lower than that associated with ketoacidosis. If deaths caused by renal failure or coronary heart disease in people with diabetes continue to decline as diabetes care improves, then the relative proportion of deaths caused by hypoglycaemia may increase. This is particularly likely as intensive insulin therapy is adopted more widely (Nathan et al. 2005; Holman et al. 2008). However, since hypoglycaemia is so very common, it can be concluded that the risk of dying during an individual episode is extremely low.

Risk factors for death from hypoglycaemia The risk factors that are commonly cited as increasing the risk of death from hypoglycaemia have been based on individual cases and have owed more to the prejudices of individual clinicians than to scientific evidence. Some proposed are detailed in Box 13.1 and include alcohol abuse and/or inebriation (Arky et al. 1968; Kalimo and Olsson 1980; Critchley et al. 1984; MacCuish 1993), psychiatric illness or personality disorder (Shenfield et al. 1980; Tunbridge 1981), self-neglect (Tunbridge 1981), resistance to education (Shenfield et al. 1980), hypopituitarism following pituitary ablation therapy for proliferative retinopa- thy (Nabarro et al. 1979; Shenfield et al. 1980), and patients who have diabetes secondary to pancreatic disease (MacCuish 1993).

Box 13.1 Possible risk factors for death from hypoglycaemia • Alcoholism and/or inebriation. • Psychiatric illness or personality disorder. • Self-neglect; inanition. • Fecklessness/resistance to education. • Diabetes secondary to pancreatic disease. • Hypopituitarism following pituitary ablation. Chapter 13: Mortality, Cardiovascular Morbidity and Possible Effects of Hypoglycaemia 265

Box 13.2 Syndromes of sudden death in non-diabetic young people • Hypertrophic obstructive cardiomyopathy (HOCM). • Coronary heart disease (severe coronary artery occlusion or myocardial infarction). • Other cardiac anatomical abnormalities (congenital anomalies of the coronary arter- ies, right ventricular dysplasia). • Syndromes of QT prolongation. • Epilepsy. • Phaeochromocytoma. • Sudden death in water. • Toxic substance abuse.

HYPOGLYCAEMIA AND SUDDEN DEATH IN TYPE 1 DIABETES

Although sudden deaths do occur, albeit rarely, in young people who do not have diabetes (Box 13.2), in those with type 1 diabetes the risks appear to be substantially higher. Since estimates in the non-diabetic population are based on larger numbers, they are probably more accurate than those reporting risks in individuals with type 1 diabetes. In young people without diabetes the rates are around 1.3 to 8.5 deaths per 100 000 person-years. In people with diabetes, the risk appears to be 3-fold higher, though it is difficult to be exact.

DEAD-IN-BED SYNDROME

At a British inquest in the early 1990s, it was suggested that the frequency of sudden death in people with type 1 diabetes was increasing. The British Diabetic Association (now Dia- betes UK) requested notification of sudden, unexpected deaths of young people with type 1 diabetes. A detailed analysis was published by Tattersall and Gill (1991). A total of 53 cases were referred, but the analysis was confined to the 50 cases who were under 50 years of age. The largest group comprised 22 sudden deaths which shared similar characteristics. All individuals had apparently been well the preceding day, were found dead in an undis- turbed bed suggesting no seizure activity, and no cause of death was revealed at post- mortem. The accompanying editorial coined the term ‘dead-in-bed syndrome’ to describe this particular pattern of sudden death (Campbell 1991) (Figure 13.1). Analysis of the 22 ‘dead in bed’ individuals showed that they were aged between 12 and 43 years, with duration of diabetes from 3 to 27 years. Although all had been taking human insulin at the time of death, most had been transferred from animal insulin between 6 months and 2 years earlier, and the authors concluded that no temporal relationship between the change in insulin species and the fatal event could be demonstrated. Informa- tion on diabetic complications was not available for all of the cases, but 13 had none and only four had severe complications. Of note, 14 of the 22 patients had a history of severe nocturnal hypoglycaemia. Neuropathological evidence of hypoglycaemia was rare, however, suggesting that protracted neuroglycopenia had not occurred in most patients. However, given the high proportion of sudden death victims with a history of nocturnal hypoglycae- mia, the authors proposed that ‘circumstantial evidence implicates nocturnal hypoglycae- mia in many cases’. 266 Hypoglycaemia in Clinical Diabetes

2

5 “Dead in bed” Unexplained 22 Diabetic ketoacidosis 11 Suicide

Definite cause of death

6 Hypoglycaemic brain damage 4

Figure 13.1 Sudden deaths of 50 young type 1 diabetic patients. (Source: Data derived from Tattersall and Gill 1991).

50 100

40 80

30 60

20 40

Number of sudden deaths 10 20 Human insulin in % of total sale Human insulin in % of total

0 0 1982 1983 1984 1985 1986 1987 1988

Figure 13.2 Number of sudden deaths (bars) and human insulin as percentage of total sales of insulin (line) in Denmark from 1982 to 1988. (Source: Borch-Jensen K 1993. Reproduced with permission from John Wiley & Sons, Ltd).

The phenomenon of the ‘dead-in-bed’ syndrome has subsequently been described in other countries. In a study in Sweden of over 4000 patients with type 1 diabetes diagnosed before the age of 14 and followed up for 23 years, 22% (17 out of 78) of deaths were unexplained, with subjects being found dead in bed without any cause of death identified at autopsy (Dahlquist and Kallen 2005). Another study from Norway reported sudden deaths that fulfilled the criteria of ‘dead-in-bed’ syndrome, which represented 6.7% of all deaths occurring in people with type 1 diabetes under the age of 40 (Thordarson and Sovik 1995). The number of subjects who were found dead in bed remained remarkably constant over a 7-year period, and no association was observed with an increasing usage of human insulin in Denmark during that time (Figures 13.2 and 13.3). Sudden deaths consistent with Chapter 13: Mortality, Cardiovascular Morbidity and Possible Effects of Hypoglycaemia 267

20

15

10 Number of deaths 5

0 1982 1983 1984 1985 1986 1987 1988

Figure 13.3 Number of deaths due to definite (shaded bars) and possible (unshaded bars) hypogly- caemia from 1982 to 1988. (Source: Borch-Johnsen K 1993. Reproduced with permission from John Wiley & Sons, Ltd).

Box 13.3 Possible risk factors for ‘dead-in-bed’ syndrome • Previous nocturnal hypoglycaemia • Living/sleeping alone • Intensive therapy • Multiple injections of insulin • Alcohol ingestion • Male gender • Genetic factors a description of ‘dead-in-bed’ syndrome have continued to afflict a small but significant number of patients, based on childhood-onset diabetes registries in Norway and the USA (Skrivarhaug et al. 2006; Secrest et al. 2011). Combined estimates from UK and Scandi- navian countries suggest the ‘dead-in-bed’ syndrome accounts for 6% of all deaths in people with type 1 diabetes below the age of 40 or an incidence of 20–60 per 100 000 patient years (Sovik and Thordarson 1999).

RISK FACTORS FOR DEAD-IN-BED SYNDROME

The risk factors for dead-in-bed syndrome are difficult to determine given the rarity of these events and incomplete data on disease from deceased patients. However, a number of factors appear more common in ‘dead-in-bed’ patients based on epidemiological obser- vations (Box 13.3). Most studies report a male preponderance (Thordarson and Sovik 1995; Tu et al. 2010; Secrest et al. 2011). One study showed an association with chronic alcohol abuse or acute alcohol intoxication (Borch-Johnsen and Helweg-Larsen 1993). All of the investigators who have reported sudden death in patients with type 1 diabetes have impli- cated hypoglycaemia as a contributing factor for the ‘dead-in-bed’ syndrome, and hypogly- caemia is perhaps the strongest risk factor. 268 Hypoglycaemia in Clinical Diabetes

Hypoglycaemia In most reports, death had occurred during the night – a time when hypoglycaemia is a common problem – and some had strict glycaemic control and had experienced nocturnal hypoglycaemia previously. Around two-thirds of patients had a history of nocturnal hypoglycaemia (Tattersall and Gill 1991; Thordarson and Sovik 1995). Further support for the role of hypoglycaemia comes from a case report of a 23-year-old man with type 1 diabetes with a history of recurrent nocturnal hypoglycaemia. He was fitted with a continu- ous glucose monitor and was subsequently found dead in an undisturbed bed (Tanenberg et al. 2010). The interstitial glucose value at the time of death was recorded as 1.7 mmol/l with minimal evidence that counterregulation had occurred. Intensive insulin therapy increases the risk of hypoglycaemia and thus may contribute to the ‘dead-in-bed’ syndrome. In a 10-year study in Norway, three-quarters of sudden deaths occurred in the last 3 years (1988–1990), coinciding with greater use of intensive insulin therapy (Thordarson and Sovik 1995). Few epidemiological studies have detailed information on insulin treatment, but some studies suggest a high proportion were on multiple daily injections (Thordarson and Sovik 1995).

Autonomic neuropathy Support for a role of autonomic neuropathy in the ‘dead-in-bed’ syndrome is derived from observations in diabetic (Ewing et al. 1980) and non-diabetic (La Rovere et al. 2001) populations. Autonomic neuropathy increases the risk of sudden death in people with dia- betes in some, but not all studies (Sampson et al. 1990). In the Rochester diabetic neuropa- thy study, although autonomic neuropathy was associated with increased risk of sudden death, this was generally correlated with atherosclerosis and coronary heart disease (Suarez et al. 2005). In ‘dead-in-bed’ victims, autonomic function has seldom been formally tested, although few had symptomatic disease. Some subjects had advanced diabetic complications and would probably have had some degree of autonomic neuropathy, but a significant proportion of those who died had a relatively short duration of diabetes with no evidence of microvascular disease. Nevertheless, subclinical autonomic changes in heart rate vari- ability and baroreceptor sensitivity, undetected by standard cardiovascular reflex tests (Weston et al. 1996), can be present as early as 2 years after diagnosis (Vinik et al. 2003). However, it seems unlikely that autonomic neuropathy alone accounts for the greater risk of sudden death in young people with type 1 diabetes. The exact cause of death in autonomic neuropathy remains uncertain, although most authors have suggested that a cardiac arrhythmia is responsible. Impaired cardiac auto- nomic function reduces protective vagal reflexes leading to unopposed sympathetic activity, increasing arrhythmic risk. Some groups have reported lengthened QT intervals in patients with autonomic neuropathy (Ewing and Neilson 1990), highlighting the association between prolonged QT intervals and the risk of sudden death in other conditions such as the con- genital long QT syndrome (Ewing et al. 1991).

MECHANISMS OF SUDDEN DEATH

Examination of recognised causes in non-diabetic patients (Box 13.2) may give some insight into possible mechanisms of sudden death in those with type 1 diabetes (Box 13.4). In the general population, the most frequent cause of sudden death is a cardiac arrhythmia, Chapter 13: Mortality, Cardiovascular Morbidity and Possible Effects of Hypoglycaemia 269

Box 13.4 Possible mechanisms contributing to sudden death • Ventricular arrhythmias/fibrillation (HOCM, coronary artery occlusion). • Increased epinephrine (sport, phaeochromocytoma, sudden death in water). • Decreased potassium (sport, phaeochromocytoma). • Autonomic stimulation/bradycardia (cold water immersion). • Severe hyperkalaemia and red cell lysis (fresh water drowning). • Respiratory arrest (autonomic neuropathy). • Asystole (epilepsy).

mostly related to coronary heart disease; it is likely that the same problem occurs in people with diabetes. Nevertheless, if sudden death is occurring more frequently in patients, then additional factors are probably responsible. Some of this increase may be a consequence of the more advanced or premature ischaemic heart disease which is associated with dia- betes. In some people the development of hypoglycaemia-induced seizures may impose an additional insult, as sudden death is relatively common in people with epilepsy and epi- leptiform seizures can provoke cardiac arrhythmias. However, the fact that most subjects were found with their bedclothes undisturbed argues against the pre-terminal development of a tonic-clonic seizure. This scenario does not exclude other forms of seizure activity and a striking similarity exists between the syndromes of sudden death in epilepsy and diabetes (Brown et al. 1990; Nashef and Brown 1996; Neligan et al. 2011). A study using an implantable electrocardiogram (ECG) recorder in 20 patients with epilepsy identified three who developed a potentially life-threatening arrhythmia (ventricular asystole), requiring insertion of a permanent pacemaker (Rugg-Gunn et al. 2004). Given the strong association between ‘dead-in-bed’ syndrome and hypoglycaemia, the crucial question is whether any mechanism exists through which hypoglycaemia could cause sudden death? Hypoglycaemia can cause irreversible brain damage but those who suffer this complication require prolonged exposure to a low blood glucose and are unlikely to die suddenly. Previous authors have emphasised the likelihood of an arrhythmic death and have pointed out that the demonstration of a plausible mechanism by which hypogly- caemia caused cardiac arrhythmias would strongly implicate hypoglycaemia (Tattersall and Gale 1993).

Cardiac arrhythmia The evidence concerning the effects of hypoglycaemia on the ECG has been obtained from clinical episodes of arrhythmia at the time of spontaneous hypoglycaemia, terminated by glucose treatment and experimental studies of insulin-induced hypoglycaemia in the labora- tory. Hypoglycaemia causes an increase in autonomic neural activity and plasma epine- phrine. Ischaemic changes have been observed during hypoglycaemia (Wright and Frier 2008). In the presence of ischaemic heart disease it is not difficult to see how some of these changes might lead to malignant tachyarrhythmias and sudden death (see below). It is less easy to explain how even these profound, but brief, physiological changes could precipitate a fatal cardiac event in those whose heart is otherwise healthy. Lengthening of the QT interval is associated with sudden death in other conditions. The congenital long QT syndrome is an inherited disorder in which mutations in cardiac membrane ion channels cause profound lengthening of the QT interval (Jervell and 270 Hypoglycaemia in Clinical Diabetes

QTc = 456 ms QTc = 610 ms HR = 66 bpm HR = 61 bpm

(a) Euglycaemia (b) Hypoglycaemia

Figure 13.4 Typical QT measurement with a screen cursor placement from one subject (a) during euglycaemia showing a clearly defined T wave, and (b) during hypoglycaemia showing prolonged repolarisation and prominent U wave. Horizontal: 700 ms epoch, vertical 1.33 mV full scale. (Source: Marques JLB 1997. Reproduced with permission from John Wiley & Sons, Ltd).

Lange-Neilsen 1957; Jackman et al. 1988; Curran et al. 1995). Among the mutations that have been reported, one (within the SCN5A gene causing LQT3) leads to impaired sodium channel inactivation, whereas others (LQT1-2, LQT5-6) produce abnormalities within potassium channels leading to decreased outward potassium current (Roden et al. 2002). These can lead to a form of fatal polymorphic ventricular tachychardia (Torsade de Pointe), often triggered by activities that promote an acute autonomic stimulus (e.g. exercise, emo- tional stress, sudden noise, sleep). Certain therapeutic agents can also cause an acquired long QT syndrome and sudden death by blocking movement of K+ ions through one of the cardiac ion channels (IKr coded by the HERG gene) (Roden et al. 2002). In a post-mortem study of 22 ‘dead-in-bed’ cases in Australia, no differences were found in the SCN5A gene compared with a control population (Tu et al. 2010). However, such studies are relatively underpowered and in any event do not eliminate the contribution of mutations in other genes such as the HERG channel. It has been demonstrated that the QT interval can lengthen during experimental hypogly- caemia both in diabetic and non-diabetic subjects (Marques et al. 1997) (Figure 13.4). Some subjects had quite pronounced increments in QT interval and a strong relationship was observed between the increase in QT interval and the rise in plasma epinephrine. Beta- blockade and potassium infusion both prevented QT lengthening during hypoglycaemia (Robinson et al. 2003). Epinephrine infusion in healthy non-diabetic subjects also causes QT lengthening that can be prevented in part by simultaneous potassium infusion (Lee et al. 2003). Similar changes in QTc have subsequently been described during clinical episodes of nocturnal hypoglycaemia in adults and children with type 1 diabetes (Murphy et al. 2004; Robinson et al. 2004). A major problem in proving or refuting the overall hypothesis is that it is very difficult to test directly because sudden death occurs so rarely. Isolated case reports of transient cardiac dysrhythmias have not provided much additional useful information. Another predisposing factor for an arrhythmia during hypoglycaemia might be an imbal- ance of autonomic activation. As mentioned earlier, hypoglycaemia leads to activation of Chapter 13: Mortality, Cardiovascular Morbidity and Possible Effects of Hypoglycaemia 271 the sympathetic and parasympathetic nervous system. Bradyarrhythmias, suggesting excess parasympathetic tone, have been observed in response to clinical episodes of nocturnal hypoglycaemia (Gill et al. 2009). It is difficult to examine cardiac autonomic activity during human studies. This has been inferred from studies of heart rate variability during insulin- induced hypoglycaemia, but these do not report consistent effects on cardiac sympathetic and parasympathetic tone (Laitinen et al. 2003; Schächinger et al. 2004; Koivikko et al. 2005). Lee et al. (2004) appeared to refute the hypothesis that autonomic neuropathy con- tributes to hypoglycaemia-induced QT lengthening by demonstrating that those with auto- nomic neuropathy had the smallest increments in QT interval induced by experimental hypoglycaemia when compared to other diabetic groups. However, since these individuals also had the longest duration of diabetes, unsurprisingly they also experienced the smallest rise in sympathoadrenal activation. It seems likely that interactions between genetic factors and environmental effects (including age, gender, state of sympathoadrenal activation and K+ concentration) deter- mine whether a cardiac arrhythmia is triggered. Fatal events may be infrequent because the combination of factors that together lead to a fatal cardiac arrhythmia occur only rarely. Indeed severe hypoglycaemia, which causes a profound sympathoadrenal discharge without alerting the patient or partner, is itself relatively unusual. Thus, the ‘substrate’ for a lethal cardiac arrhythmia might be the combination of a severe hypoglycaemic attack at night (which fails to wake people unless symptoms are intense) and a cardiac conduction system affected by subclinical autonomic neuropathy in an individual who has inherited polymor- phisms in the LQT genes causing exaggerated QT lengthening during sympathoadrenal activation. In summary, the majority of sudden deaths in young patients with diabetes remain unexplained, and it has been hypothesised that these deaths were caused by ventricular arrhythmia. These might have resulted from the increase in plasma epinephrine and fall in plasma potassium that accompany hypoglycaemia, so producing prolongation of the QT interval. It is possible that this occurs on a background of autonomic instability caused by early autonomic neuropathy.

What do we say to patients? It is clear that the frequency of deaths resulting from hypoglycaemia is underestimated, and some deaths related to hypoglycaemia may be attributed to other causes at the time of certification. Nevertheless, when we consider that nocturnal hypoglycaemia is very common, affecting as many as 60% of patients with type 1 diabetes every night, patients can be reassured that sudden death as a consequence of acute hypoglycaemia is very rare. However, the available evidence prevents a categorical statement that there is no risk of death from hypoglycaemia during sleep, or at other vulnerable times when treatment is not rendered promptly.

HYPOGLYCAEMIA AS A RISK FACTOR FOR INCREASED ALL CAUSE AND CARDIOVASCULAR MORTALITY

Deaths at the time of hypoglycaemia are relatively rare events. However, there are recent concerns that the risk of death from hypoglycaemia is not limited to the time of the acute episode. Patients with a history of severe hypoglycaemia are at increased risk of death and 272 Hypoglycaemia in Clinical Diabetes

Table 13.1 Summary of main features and outcomes of trials evaluating effect of intensive versus conventional glycaemic control in type 2 diabetes patients with cardiovascular risk

ACCORD ADVANCE VADT

n 10 251 11 140 1791 Mean age (years) 62 66 60 Duration of diabetes 10 8 11.5 (years) History of CV disease 35 32 40 (%)

Target HbA1c (%)* <6.0 vs 7.0–7.9 <6.5 vs based on <6.0 vs planned local guidelines separation of 1.5

Achieved HbA1c (%)* 6.4 vs 7.5 6.3 vs 7.0 6.9 vs 8.5 Median follow up 3.5 5 5.6 (years) Severe hypoglycaemia* 16.2 vs 5.1 2.7 vs 1.5 21.2 vs 9.9 (%) HR for total mortality 1.22 (1.01–1.46) 0.93 (0.83–1.06) 1.07 (0.81–1.42) (95% CI) HR for CV mortality 1.35 (1.04–1.76) 0.88 (0.74–1.04) 1.32 (0.81–2.14) (95% CI)

* Intensive versus conventional glycaemic control groups; CV, cardiovascular; HR, hazard ratio. of fatal cardiovascular events in subsequent months compared with patients with no history of severe hypoglycaemia. These concerns were raised following trials (reported in 2008 and 2009) that employed intensive therapy to obtain very strict glycaemic control in patients with type 2 diabetes who had a significant cardiovascular risk. The Action to Control Cardiovascular Risk in Diabetes (ACCORD), Action in Diabetes and Vascular Disease: Preterax and Diamicron MR Controlled Evaluation (ADVANCE) and Veterans Affairs Diabetes Trial (VADT) randomised 24 000 patients with high cardiovascular risk to very strict glycaemic control or to standard control (Gerstein et al. 2008; Patel et al. 2008; Duckworth et al. 2009). The characteristics of these trials are summarised in Table 13.1. In all three trials, no significant benefit in overall mortality or cardiovascular events was observed in the intensively treated groups. In 2008, the diabetes community was alarmed by the premature termination of the ACCORD trial because an excess of deaths had occurred in the intensive treatment arm (hazard ratio [HR] 1.22, 95% CI 1.01–1.46, P = 0.04). A number of explanations have been proposed, including greater weight gain, medication effects or just chance. However, a strong candidate was the effect of hypogly- caemia, since this was 3-fold higher in frequency in the intensive arm (Gerstein et al. 2008). The increased likelihood of hypoglycaemia during strict glycaemic control is not unex- pected (Turnbull et al. 2009), particularly during a strategy whose overall aim was to lower the HbA1c to below 6.5% as rapidly as possible. In the ACCORD trial, patients who had one or more severe hypoglycaemic episodes had higher mortality rates than those with no hypoglycaemia across both study arms (HR 1.41, 95% CI (1.03–1.93) (Bonds et al. 2010). One-third of all deaths were attributed to cardiovascular disease, and hypoglycaemia was associated with higher cardiovascular mortality. In VADT, a recent severe hypoglycaemic event was the strongest independent predictor of death at 90 days (Duckworth et al. 2009). A similar pattern has also been shown in an analysis of the ADVANCE trial (Zoungas et al. 2010). Patients with a history of Chapter 13: Mortality, Cardiovascular Morbidity and Possible Effects of Hypoglycaemia 273 hypoglycaemia had higher rates of death and macrovascular events. This relationship was attenuated after adjusting for multiple covariates, but remained significant. However, similar associations were reported with a number of seemingly unrelated non-vascular outcomes, such as respiratory and skin diseases. The crucial question is whether a true causal relationship exists between hypoglycaemia and mortality, and this is difficult to determine from post hoc analyses of trial data. The possibility cannot be excluded that the association between hypoglycaemia and death could, at least in part, be a marker for vulnerability. The relationship between hypoglycaemia, mortality and intensity of therapy is more complex. In ACCORD, although hypoglycaemia was associated with increased risk of death in both study arms, the relationship was stronger in the standard therapy group. This counterintuitive observation seems to reduce the likelihood of a causal relationship between hypoglycaemia and mortality. However, such an observation might be consistent with the known pathophysiology associated with repeated episodes of hypoglycaemia. If the adverse effects of hypoglycaemia are mediated by sympathoadrenal activation, then those receiving intensive treatment, might have greater protection as repeated exposure to hypoglycaemia would attenuate counterregulatory responses. Overall, the burden of hypoglycaemia would still be higher in the intensive group. Intensive insulin therapy was claimed to reduce mortality in surgical critical care patients by reducing the effects of stress-related hyperglycaemia, inflammation and depressed fibri- nolysis (Van den Berghe et al. 2001). These benefits have not been consistently replicated in other medical or mixed intensive care settings (Van den Berghe et al. 2006; Brunkhorst et al. 2008; Finfer et al. 2009). It is possible that the adverse effects of hypoglycaemia, increased by up to 6-fold in some reports (Griesdale et al. 2009), offset the benefit derived from intensive insulin therapy. As in studies of type 2 diabetes, hypoglycaemia was observed to be an independent risk factor for death in several intensive care trials, follow- ing adjustment for disease severity (Van den Berghe et al. 2006; Brunkhorst et al. 2008). Insulin-glucose infusions have also been used in patients with diabetes during treatment of acute myocardial infarction (AMI) to reduce hyperglycaemia and improve myocardial energy balance. These may be combined with an intensive strategy of subcutaneous multi- dose insulin to control blood glucose tightly in the subsequent post-infarction period. Despite early promise, these regimens have not consistently reduced mortality rates (Malm- berg et al. 2005; Goyal et al. 2009). The more recent trials had several problems with failure to achieve the intended glycaemic targets and inadequate power (Malmberg et al. 2005), but hypoglycaemia remained a potential explanation for the lack of reduction in mortality as it was more prevalent in the groups treated with insulin-glucose infusions. However, in post hoc analyses, no significant relationship was found between hypoglycae- mia with this treatment and increased death or cardiovascular mortality after adjustment for confounding factors (Goyal et al. 2009; Mellbin et al. 2009). In summary, evidence from trials comparing intensive versus conventional glycaemic treatment strategies in a number of settings do not confirm or refute the possibility that hypoglycaemia may be linked to increased mortality. Hypoglycaemia may have been dif- ferentially recorded in the two study arms as glucose was measured more frequently in intensively-treated patient groups. Further, patients who experience frequent hypoglycae- mia will exhibit attenuated symptomatic responses, so that studies that are based on record- ing symptomatic hypoglycaemia may provide an unreliable measure of the true relationship with mortality. Hypoglycaemia may be a marker of vulnerability rather than be causally related to adverse cardiovascular outcomes and death. A number of shared factors increase 274 Hypoglycaemia in Clinical Diabetes both the risk of hypoglycaemia and of death, such as renal impairment, liver disease, cogni- tive impairment, and weight loss secondary to other chronic diseases. Despite these poten- tial confounders, a significant association remains after most of these factors have been accounted for. Although a causal relationship between hypoglycaemia and adverse mortal- ity and cardiovascular outcomes remains hypothetical, it is possible that adverse effects of hypoglycaemia can offset the benefit of strict glycaemic control. This may explain the lack of difference in mortality in ADVANCE and VADT and the apparent harm of intensive therapy in ACCORD. In the face of this controversy, should aggressive therapy to obtain strict glycaemic control be recommended in all people with diabetes? The Diabetes Control and Complica- tions Trial (DCCT) clearly demonstrated that intensive glycaemic control reduces the risk of cardiovascular events in people with type 1 diabetes (Nathan et al. 2005). In those with type 2 diabetes, the benefit of intensive glucose lowering appears to be limited to those whose cardiovascular disease burden is low. In subgroup analyses of ACCORD and VADT, a benefit of strict glycaemic control was observed only in patients with low baseline car- diovascular risk (Gerstein et al. 2008; Reaven et al. 2009). In the UK Prospective Diabetes Study, people with type 2 diabetes randomised to intensive therapy at diagnosis had sub- stantially lower mortality and cardiovascular event rates, an effect that was sustained beyond the intervention period (Holman et al. 2008). This contrasts with trials of patients at high cardiovascular risk (ACCORD/ADVANCE/VADT) who either had no benefit or experienced worse outcomes with strict glycaemic control. Based on current evidence, intensive treatment to obtain strict glycaemic control continues to be recommended for the majority of people with type 1 and type 2 diabetes. However, a less stringent glycaemic target is recommended for patients with type 2 diabetes with advanced macrovascular complications, limited life expectancy, long disease duration or those who are prone to hypoglycaemia (Frier et al. 2011).

POTENTIAL MECHANISMS BY WHICH HYPOGLYCAEMIA CAN INCREASE VASCULAR RISK

Although a definitive causal link between hypoglycaemia and cardiovascular events or death is not proven, a number of experimental studies have demonstrated the potentially adverse effects of hypoglycaemia on the vasculature. Hypoglycaemia promotes activation of the autonomic nervous system, catecholamine release and haemodynamic changes. These would be expected to affect both the microcirculation and the heart. Potential mecha- nisms by which hypoglycaemia could aggravate microvascular (Box 13.5) and macrovas- cular (Box 13.6) complications are discussed below.

Box 13.5 Postulated effects of hypoglycaemia on the microcirculation • Changes in capillary blood flow • Increased coagulation factors • Platelet activation • Neutrophil activation • Increased free-radical activity Chapter 13: Mortality, Cardiovascular Morbidity and Possible Effects of Hypoglycaemia 275

Box 13.6 Potential mechanisms by which hypoglycaemia can aggravate macrovascular disease • Inflammation and endothelial dysfunction • Reduced fibrinolysis • Cardiac ischaemia • Cardiac autonomic dysfunction • Pro-arrhythmic effects

HYPOGLYCAEMIA AND MICROVASCULAR COMPLICATIONS

It has been suggested that acute hypoglycaemia, by releasing vasoactive hormones and provoking changes in regional and capillary blood flow, might worsen established micro- vascular complications of diabetes (Frier and Hilsted 1985). Although microvascular com- plications are recognised to be the consequence of chronic hyperglycaemia and can be prevented or delayed by strict glycaemic control, the effect of recurrent hypoglycaemia on an already compromised microvasculature may be deleterious and cause further damage (Box 13.5). Exposure to recurrent hypoglycaemia may precipitate capillary closure inducing local- ised tissue ischaemia and producing deterioration in retinopathy. A sudden fall in intraocu- lar pressure occurs during hypoglycaemia, and could precipitate vitreous haemorrhage in patients with proliferative retinopathy, as friable new vessels are vulnerable to sudden changes in perfusion pressure or mechanical stresses. This could explain the occasional anecdotal reports of vitreous haemorrhage described by individual patients, often following nocturnal hypoglycaemia. Acute hypoglycaemia reduces renal plasma flow and glomerular filtration in normal subjects and in people with diabetes who have no significant complications. In those who have established nephropathy with glomerular sclerosis and arteriolar narrowing, the reduc- tion in renal plasma flow may precipitate further closure of arterioles and progression in renal impairment.

HYPOGLYCAEMIA AND MACROVASCULAR COMPLICATIONS

Precipitation of acute vascular events (such as myocardial infarction or stroke) as a con- sequence of the effects of hypoglycaemia on established macrovascular disease appears to occur relatively infrequently, considering how commonly episodes of hypoglycaemia occur in everyday life. A number of experimental studies have shown that hypoglycaemia can promote inflammation and may thereby accelerate atherogenesis and precipitate acute thrombosis. Hypoglycaemia has the potential to initiate cardiac ischaemia, autonomic dysfunction or even fatal arrhythmias predisposing to cardiac death (Box 13.6). Recurrent hypoglycaemic episodes may also be sufficient to precipitate a macrovascular event in those already at high risk. Whether hypoglycaemia can produce more sustained changes in the cardiovascular system that can increase the risk of death in the weeks and months after an episode is currently unknown. 276 Hypoglycaemia in Clinical Diabetes

Inflammation and atherogenesis The role of subclinical inflammation in atherogenesis and development of the vulnerable plaque is increasingly recognised. Acute hypoglycaemia has been reported to aggravate inflammatory processes in experimental studies in non-diabetic individuals and type1 diabetes patients (Gogitidze Joy et al. 2010; Wright et al. 2010). Hypoglycaemia can impair endothelial dysfunction as shown by changes in von Willebrand factor (vWF) and endothe- lin levels (Wright et al. 2007, 2010). Hypoglycaemia can also induce expression of athero- genic cell adhesion molecules (Gogitidze Joy et al. 2010) and increased leukocyte trafficking and atherogenesis (Wright et al. 2010). The effects on other downstream proinflammatory cytokines such as interleukin-6 (IL-6) and acute phase proteins such as C-reactive protein are less consistent (Galloway et al. 2000; Dotson et al. 2008; Gogitidze Joy et al. 2010; Wright et al. 2010).

Thrombosis Platelet activation is increased in response to hypoglycaemia in non-diabetic (Wright et al. 2010) and type 1 diabetic volunteers (Gogitidze Joy et al. 2010). Platelet aggregability increases during hypoglycaemia (Trovati et al. 1986). The mechanisms underlying platelet aggregation appear to be adrenaline-mediated (Hutton et al. 1979; Lande et al. 1985) and inhibited by alpha 2 adrenergic blockade (Kishikawa et al. 1987). Our current understand- ing of acute coronary syndromes has highlighted the importance of development of the vulnerable plaque and acute thrombosis as leading from stable to acute coronary heart disease. The fate of a ruptured plaque and whether it becomes a sustained occlusive throm- bus is dictated by fibrinolytic balance. In previous studies where hypoglycaemia was induced by an insulin bolus, there is activation of the fibrinolytic system as evidenced by increase in tissue plasminogen activator and fall in tissue plasminogen activator inhibitor (PAI) (Fisher et al. 1991; Wieczorek et al. 1993). However, it is unclear whether these resulted from the effects of insulin per se, which has fibrinolytic properties. Other studies have observed increased clot formation and reduced fibrinolysis during more sustained insulin-induced hypoglycaemia in comparison with euglycaemia (Ajjan et al. 2010; Gog- itidze Joy et al. 2010).

Ischaemia Acute hypoglycaemia can provoke an intense haemodynamic response secondary to activa- tion of the autonomic nervous system with the secretion of epinephrine, particularly in those individuals whose counterregulatory defences are intact (DeRosa and Cryer 2004). The heart rate increases over a period of 15 to 20 minutes, but rarely rises above 100 beats/ minute. A modest but significant increase in systolic blood pressure is accompanied by a slight but significant fall in diastolic blood pressure (Fisheret al. 1987; Russell et al. 2001). The pulse pressure widens, with a substantial increase in cardiac output and a fall in total peripheral vascular resistance (Figure 13.5). These haemodynamic changes are relatively short lived, and appear to exert no significant after effects on the 24-hour heart rate or blood pressure (Avogaro et al. 1994). Central blood pressure falls during hypoglycaemia, with a transient increase in elasticity of arteries (Sommerfield et al. 2007), which is diminished in people with type 1 diabetes of longer duration (>15 years), and premature arterial stiff- ening is common in diabetes of both types. This may accelerate the return of the pressure Chapter 13: Mortality, Cardiovascular Morbidity and Possible Effects of Hypoglycaemia 277

100 * 80 *

60 Heartrate (beats/min) 40 (a)

140 *

Systolic 100

Mean (mmhg) 60 Diastolic Blood pressure *

20 (b)

80 * * * * * ** 60 * * (%)

40 Left ventricular ejection fraction ventricular Left 0 R R+30 R+60 R+90 (c) Time (min)

Figure 13.5 Mean response of (a) heart rate, (b) systolic blood pressure and mean arterial blood pressure, and (c) left ventricular ejection fraction following intravenous injection of insulin at time 0. (R, autonomic reaction). (Source: Fisher M 1987. Reproduced with kind permission from Springer Science and Business Media). wave (generated during systole) to the heart and so affect coronary filling, which takes place in diastole. Myocardial blood flow reserve has been shown to decline during hypogly- caemia (Rana et al. 2011). In a person with a normal heart these haemodynamic changes are probably of no clinical significance, but in a patient who has underlying coronary heart disease the increase in cardiac workload may provoke a cardiac arrhythmia, myocardial ischaemia and even myocardial infarction (Box 13.5). The provocation of angina and myocardial ischaemia by exercise is well documented in clinical practice. By contrast, acute hypoglycaemia, which can provoke a more intense haemodynamic response and in particular a greater increase in plasma epinephrine, has seldom been documented as provoking anginal chest pain, either in the experimental 278 Hypoglycaemia in Clinical Diabetes situation or in clinical practice, other than occasional anecdotal case reports. This may reflect the fact that to avoid hypoglycaemia, clinicians generally accept higher ambient blood glucose concentrations in people with diabetes who have known coronary heart disease. It is also possible that the haemodynamic changes of hypoglycaemia are so profound that they are frequently fatal in patients with coronary heart disease, and hypoglycaemia is then overlooked as a provoking cause when determining the cause of death. It is now well estab- lished that many episodes of ST segment depression on the ECG are not associated with angina, and constitute ‘silent ischaemia’. One case has been described during 24-hour ECG monitoring of hypoglycaemia that provoked silent ischaemia in a diabetic patient with suspected coronary heart disease (Pladziewicz and Nesto 1989). More recently, 72-hour continuous glucose monitoring with simultaneous cardiac Holter monitoring was performed in 21 patients who had coronary heart disease and insulin-treated type 2 diabetes (Desouza et al. 2003). Hypoglycaemia was frequently associated with ECG abnormalities, some of which were associated with chest pain, in this group of well-controlled patients. Myocardial infarction has rarely been documented as a consequence of hypoglycaemia (Fisher and Frier 1993). In a series of non-diabetic patients with schizophrenia who were treated with hypoglycaemic shock therapy, 6 of 44 deaths were ascribed to cardiac causes, with most deaths caused by cerebral damage (Maclay 1953). It should be emphasised that this long-abandoned form of treatment of psychiatric disease necessitated prolonged and profound hypoglycaemia. Only a few cases have been published of myocardial infarction and hypoglycaemia in diabetic patients (Purucker et al. 2000; Chang et al. 2007). Thus a possible association is very difficult to establish. In addition, the release of counterregula- tory hormones such as glucagon, cortisol and epinephrine, which are released during stress, may raise blood glucose subsequently and make the contribution of preceding hypoglycae- mia difficult to confirm.

Autonomic function It is well established that hypoglycaemia is accompanied by neurally-mediated activation of the sympathetic nervous system and epinephrine release. Hypoglycaemia may shift the balance in favour of sympathetic activation and may be pro-arrhythmic in susceptible patients. In a small ambulatory study of people with type 2 diabetes with coronary heart disease, parasympathetic activity was reduced during low glucose levels, but overall sym- pathovagal balance was unchanged (Infusino et al. 2010). Recent hypoglycaemia impairs the subsequent hypoglycaemic counterregulatory auto- nomic response. The effect of hypoglycaemia on autonomic function may be more gener- alised and not limited to hypoglycaemic stimuli. The effect of antecedent hypoglycaemia on cardiac autonomic function has been examined in healthy non-diabetic subjects (Adler et al. 2009). Autonomic function tests were performed the day before and the day after the hypoglycaemic clamp. The study demonstrated attenuated muscle nerve sympathetic activity following antecedent hypoglycaemia, decreased baroreceptor reflex sensi­tivity as well as an impaired response to nitroprusside-induced hypotension. In studies of rats, antecedent hypoglycaemia also reduced subsequent epinephrine responses to non- hypoglycaemic stimuli such as hypotension by 2 to 3-fold (Herlein et al. 2006). This could mean that in people with intensively-treated diabetes who are experiencing recurrent hypoglycaemia, their response to hypotensive stress may be attenuated – for example in cardiogenic, septic or hypovolaemic shock, increasing the risk of death from circulatory Chapter 13: Mortality, Cardiovascular Morbidity and Possible Effects of Hypoglycaemia 279 failure. This is a potential mechanism whereby hypoglycaemia might increase the risk of death in the days to weeks after an episode.

Cardiac arrhythmias Experimental studies have suggested a variety of mechanisms by which hypoglycaemia can be pro-arrhythmic. Direct glucose effects, catecholamine release and electrolyte imbal- ance can all contribute to electrical instability of the myocardium (Robinson et al. 2003; Zhang et al. 2003). In patients with existing coronary heart disease, structural abnormalities and ischaemia further act as substrates for arrhythmia. It is not ethical to study patients with known coronary heart disease under experimental hypoglycaemia, for reasons of safety. However, sporadic reports of cardiac arrhythmias have been associated with clinical episodes of hypoglycaemia. Sinus bradycardia has been reported in a very small number of cases (Pollock et al. 1996; Navarro-Gutierrez et al. 2003). Atrial fibrillation has been described in some patients, and in addition there are several case reports of atrial fibrillation following hypoglycaemia in insulin-treated patients who had no overt evidence of heart disease (Collier et al. 1987; Baxter et al. 1990; Odeh et al. 1990; Navarro-Gutierrez et al. 2003). There is a single report of a transient ventricular tachycardia occurring during experimental hypoglycaemia in a non-diabetic patient with coronary heart disease, and ventricular tachycardia was documented in an elderly non-diabetic man who developed hypoglycaemia during emergency surgery (Chelliah 2000). There have been case reports, with ECG evidence, of ventricular ectopics, sustained ventricular tachycardia, ventricular fibrillation and asystole during hypoglycaemia in diabetic patients (Shimada et al. 1984; Burke and Kearney 1999). Obviously this does not exclude the possibility of these arrhyth- mias occurring more frequently in clinical practice, as these arrhythmias will be fatal if uncorrected. In most instances it is unlikely that any precipitating cause of the arrhythmia would be sought; hypoglycaemia may not have been recognised and the difficulties of establishing a putative diagnosis of hypoglycaemia at post-mortem have been alluded to previously.

CONCLUSIONS • Biochemical and mild symptomatic hypoglycaemia occur commonly in the treatment of people with type 1 diabetes. Even severe episodes are not infrequent, but sudden and unexpected deaths from hypoglycaemia are rare. • There appears to be an increased risk of sudden death in people with diabetes compared to those without, which may have become more numerous as greater attempts are made to control blood glucose much more strictly. These deaths, referred to as the ‘dead-in- bed’ syndrome, are possibly related to hypoglycaemia through hypoglycaemia-induced arrhythmias. • Experimental hypoglycaemia can provoke abnormalities of the ECG, which are associ- ated with sudden death in other conditions, and this observation offers a possible mechanism to explain the phenomenon. • Trials of intensive glycaemic control did not reduce overall mortality over 5 years or cardiovascular death in people with type 2 diabetes with existing cardiovascular disease. Intensive glycaemic control was associated with higher rates of hypoglycaemia, but a 280 Hypoglycaemia in Clinical Diabetes

causal relationship with adverse cardiovascular outcomes and death has not been proven. • The intense sympathoadrenal response provoked by severe hypoglycaemia may also provoke changes that could worsen established microvascular complications, although to date this is primarily hypothetical. • Under experimental conditions, hypoglycaemia has been shown to provoke subclinical inflammation, endothelial dysfunction, thrombosis, autonomic dysfunction and pro- arrhythmic effects. The clinical relevance of these mechanisms in precipitating cardio- vascular events remains to be established.

REFERENCES

Adler, G., Bonyhay, I., Failing, H., Waring, E., Dotson, S. & Freeman, R. (2009) Antecedent hypoglycemia impairs autonomic cardiovascular function: implications for rigorous glycemic control. Diabetes, 58, 360–366. Ajjan, R., Robinson, E., Alzahrani, S., Grant, P.J. & Heller, S. (2010) The effects of hypoglycemia on fibrin clot structure and fibrinolysis in individuals with type 1 diabetes. Diabetes, 59 (Suppl. 1), 402P (abstract). Arky, R.A., Veverbrants, E. & Abramson, E.A. (1968) Irreversible hypoglycemia. A complication of alcohol and insulin. Journal of the American Medical Association, 206, 575–578. Avogaro, A., Crepaldi, C., Miola, M. et al. (1994) Sequelae of acute hypoglycaemia on 24 hour blood pressure and metabolic parameters in normal and type 1 (insulin-dependent) diabetic individuals. Diab­ etic Medicine, 11, 573–577. Baxter, M.A., Garewal, C., Jordan, R., Wright, A.D. & Nattrass, M. (1990) Hypoglycaemia and atrial fibrillation. Postgraduate Medical Journal, 66, 981. Bonds, D., Miller, M., Bergenstal, R. et al. (2010) The association between symptomatic, severe hypogly- caemia and mortality in type 2 diabetes: retrospective epidemiological analysis of the ACCORD study. British Medical Journal, 340, b4909. Borch-Johnsen, K. & Helweg-Larsen, K. (1993) Sudden death and human insulin: is there a link? Diabetic Medicine, 10, 255–259. Brown, S.W., Mawer, G.E., Lawler, W. et al. (1990) Sudden death and epilepsy. Lancet, 355, 606–607. Brunkhorst, F.M., Engel, C., Bloos, F. et al. (2008) Intensive insulin therapy and pentastarch resuscitation in severe sepsis. New England Journal of Medicine, 358, 125–139. Burke, B.J. & Kearney, T.K. (1999) Hypoglycaemia and cardiac arrest. Practical Diabetes International, 16, 189–190. Campbell, I. (1991) Dead in bed syndrome: a new manifestation of nocturnal hypoglycaemia? Diabetic Medicine, 8, 3–4. Chang, J.-H., Tseng, C.-F. & Wang, J.-Y. (2007) Hypoglycemia-induced myocardial infarction: an unusual adverse effect of sulfonylureas. International Journal of Cardiology, 115, 414–416. Chelliah, Y.R. (2000) Ventricular arrhythmias associated with hypoglycaemia. Anaesthesia and Intensive Care, 28, 698–700. Collier, A., Matthews, D.M., Young, R.J. & Clarke, B.F. (1987) Transient atrial fibrillation precipitated by hypoglycaemia: two case reports. Postgraduate Medical Journal, 63, 895–897. Critchley, J.A.J.H., Proudfoot, A.T., Boyd, S.G., Campbell, I.W., Brown, N.S. & Gordon, A. (1984) Deaths and paradoxes after intentional insulin overdosage. British Medical Journal, 289, 225. Curran, M.E., Splawski, I., Timothy, K.W., Vincent, G.M., Green, E.D. & Keating, M.T. (1995) A molecular basis for cardiac arrhythmias: HERG mutations cause long QT syndrome. Cell, 80, 795–803. Dahlquist, G. & Kallen, B. (2005) Mortality in childhood-onset type 1 diabetes: a population-based study. Diabetes Care, 28, 2384–2387. DeRosa, M.A. & Cryer, P.E. (2004) Hypoglycemia and the sympathoadrenal system: neurogenic symptoms are largely the result of sympathetic neural, rather than adrenomedullary, activation. American Journal of Endocrinology, 287, E32–E41. Desouza, C., Salazar, H., Cheong, B., Murgo, J. & Fonseca, V. (2003) Association of hypoglycemia and cardiac ischemia: a study based on continuous monitoring. Diabetes Care, 26, 1485–1489. Chapter 13: Mortality, Cardiovascular Morbidity and Possible Effects of Hypoglycaemia 281

Dotson, S., Freeman, R., Failing, H. & Adler, G. (2008) Hypoglycemia increases serum interleukin-6 levels in healthy men and women. Diabetes Care, 31, 1222–1223. Duckworth, W., Abraira, C., Moritz, T. et al. (2009) Glucose control and vascular complications in veterans with type 2 diabetes. New England Journal of Medicine, 360, 129–139. Ewing, D.J. & Neilson, J.M. (1990) QT interval length and diabetic autonomic neuropathy. Diabetic Medi­ cine, 7, 23–26. Ewing, D.J., Campbell, I.W. & Clarke, B.F. (1980) The natural history of diabetic autonomic neuropathy. The Quarterly Journal of Medicine, 49, 95–108. Ewing, D.J., Boland, O., Neilson, J.M., Cho, C.G. & Clarke, B.F. (1991) Autonomic neuropathy, QT interval lengthening, and unexpected deaths in male diabetic patients. Diabetologia, 34, 182–185. Feltbower, R.G., Bodansky, H.J., Patterson, C.C. et al. (2008) Acute complications and drug misuse are important causes of death for children and young adults with type 1 diabetes: results from the Yorkshire Register of diabetes in children and young adults. Diabetes Care, 31, 922–926. Finfer, S., Chittock, D.R., Su, S.Y. et al. (2009) Intensive versus conventional glucose control in critically ill patients. New England Journal of Medicine, 360, 1283–1297. Fisher, B.M., Quin, J.D., Rumley, A. et al. (1991) Effects of acute insulin-induced hypoglycaemia on haemo­stasis, fibrinolysisand haemorheology in insulin-dependent diabetic patients and control subjects. Clinical Science, 80, 525–531. Fisher, B.M. & Frier, B.M. (1993) Effect on vascular disease. In: B.M. Frier & B.M. Fisher (eds), Hypogly­ caemia and Diabetes: Clinical and Physiological Aspects, pp. 355–361. Edward Arnold, London. Fisher, B.M., Gillen, G., Dargie, H.J., Inglis, G.C. & Frier, B.M. (1987) The effects of insulin-induced hypoglycaemia on cardiovascular function in normal man: studies using radionuclide ventriculography. Diabetologia, 30, 841–845. Frier, B.M. & Hilsted, J. (1985) Does hypoglycaemia aggravate the complications of diabetes? Lancet, 326, 1175–1177. Frier, B.M., Schernthaner, G. & Heller, S.R. (2011) Hypoglycemia and cardiovascular risks. Diabetes Care, 34 (Suppl. 2), S132–S137. Galloway, P., Thomson, G., Fisher, B. & Semple, C. (2000) Insulin-induced hypoglycemia induces a rise in C-reactive protein. Diabetes Care, 23, 861–862. Gerstein, H.C., Miller, M.E., Byington, R.P. et al. (2008) Effects of intensive glucose lowering in type 2 diabetes. New England Journal of Medicine, 358, 2545–2559. Gill, G., Woodward, A., Casson, I.F. & Weston, P.J. (2009) Cardiac arrhythmia and nocturnal hypoglycae- mia in type 1 diabetes–the ‘dead in bed’ syndrome revisited. Diabetologia, 52, 42–45. Gogitidze Joy, N., Hedrington, M.S., Briscoe, V.J., Tate, D.B., Ertl, A.C. & Davis, S.N. (2010) Effects of acute hypoglycemia on inflammatory and pro-atherothrombotic biomarkers in individuals with type 1 diabetes and healthy individuals. Diabetes Care, 33, 1529–1535. Goyal, A., Mehta, S.R., Diaz, R. et al. (2009) Differential clinical outcomes associated with hypoglycemia and hyperglycemia in acute myocardial infarction. Circulation, 120, 2429–2437. Griesdale, D.E., De Souza, R.J., Van Dam, R.M. et al. (2009) Intensive insulin therapy and mortality among critically ill patients: a meta-analysis including NICE-SUGAR study data. Canadian Medical Associa­ tion Journal, 180, 821–827. Herlein, J., Morgan, D., Phillips, B., Haynes, W. & Sivitz, W. (2006) Antecedent hypoglycemia, catecho- lamine depletion, and subsequent sympathetic neural responses. Endocrinology, 147, 2781–2788. Holman, R., Paul, S., Bethel, M., Matthews, D. & Neil, H. (2008) 10-year follow-up of intensive glucose control in type 2 diabetes. New England Journal of Medicine, 359, 1577–1589. Hutton, R., Mikhailidis, D., Dormandy, K. & Ginsburg, J. (1979) Platelet aggregation studies during tran- sient hypoglycaemia: a potential method for evaluating platelet function. Journal of Clinical Pathology, 32, 434–438. Infusino, F., Pitocco, D., Zaccardi, E. et al. (2010) Low glucose blood levels are associated with abnormal cardiac sympatho-vagal balance in type 2 diabetic patients with coronary artery disease. European Review of Medical and Pharmacological Sciences, 14, 203–207. Jackman, W.M., Friday, K.J., Anderson, J.L., Aliot, E.M., Clark, M. & Lazzara, R. (1988) The long QT syndromes: a critical review, new clinical observations and a unifying hypothesis. Progress in Cardio­ vascular Diseases, 31, 115–172. Jervell, A. & Lange-Neilsen, F. (1957) Congenital deaf-mutism, functional heart disease with prolongation of the QT interval and sudden death. American Heart Journal, 54, 59–68. Kalimo, H. & Olsson, Y. (1980) Effects of severe hypoglycemia on the human brain. Neuropathological case reports. Acta Neurologica Scandinavica, 62, 345–356. 282 Hypoglycaemia in Clinical Diabetes

Kishikawa, H., Takeda, H., Kiyota, S. et al. (1987) Role of alpha 2-adrenergic receptor in platelet activation during insulin-induced hypoglycemia in normal subjects. Diabetes, 36, 407–412. Koivikko, M.L., Salmela, P.I., Juhani Airaksinen, K.E. et al. (2005) Effects of sustained insulin-induced hypoglycemia on cardiovascular autonomic regulation in type 1 diabetes. Diabetes, 54, 744–750. La Rovere, M., Pinna, G., Hohnloser, S. et al. (2001) Baroreflex sensitivity and heart rate variability in the identification of patients at risk for life-threatening arrhythmias: implications for clinical trials. Circula­ tion, 103, 2072–2077. Laitinen, T., Huopio, H., Vauhkonen, I. et al. (2003) Effects of euglycaemia and hypoglycaemic hyperin- sulinaemia on sympathetic and parasympathetic regulation of haemodynamics in healthy subjects. Clini­ cal Science, 105, 315–322. Lande, K., Gjesdal, K., Fønstelien, E., Kjeldsen, S. & Eide, I. (1985) Effects of adrenaline infusion on platelet number, volume and release reaction. Thrombosis and Haemostasis, 54, 450–453. Lee, S., Harris, N.D., Robinson, R.T., Yeoh, L., Macdonald, I.A. & Heller, S.R. (2003) Effects of adrenaline and potassium on QTc interval and QT dispersion in man. European Journal of Clinical Investigation, 33, 93–98. Lee, S.P., Yeoh, L., Harris, N.D. et al. (2004) Influence of autonomic neuropathy on QTc interval lengthen- ing during hypoglycemia in type 1 diabetes. Diabetes, 53, 1535–1542. MacCuish, A.C. (1993) Treatment of hypoglycaemia. In: B.M. Frier & B.M. Fisher (eds), Hypoglycaemia and Diabetes: Clinical and Physiological Aspects, pp. 212–221. Edward Arnold, London. Maclay, W.S. (1953) Death due to treatment. Proceedings of the Royal Society of Medicine, 46, 13–20. Malmberg, K., Ryden, L., Wedel, H. et al. (2005) Intense metabolic control by means of insulin in patients with diabetes mellitus and acute myocardial infarction (DIGAMI 2): effects on mortality and morbidity. European Heart Journal, 26, 650–661. Marques, J.L.B., George, E., Peacey, S.R. et al. (1997) Altered ventricular repolarization during hypogly- caemia in patients with diabetes. Diabetic Medicine, 14, 648–654. Mellbin, L., Malmberg, K., Waldenström, A., Wedel, H. & Rydén, L. (2009) Prognostic implications of hypoglycaemic episodes during hospitalisation for myocardial infarction in patients with type 2 diabetes: a report from the DIGAMI 2 trial. Heart, 95, 721–727. Muhlhauser, I., Sawicki, P.T., Blank, M., Overmann, H., Richter, B. & Berger, M. (2002) Reliability of causes of death in persons with type I diabetes. Diabetologia, 45, 1490–1497. Murphy, N.P., Ford-Adams, M.E., Ong, K.K. et al. (2004) Prolonged cardiac repolarisation during spon- taneous nocturnal hypoglycaemia in children and adolescents with type 1 diabetes. Diabetologia, 47, 1940–1947. Nabarro, J.D.N., Mustaffa, B.E., Morris, D.V., Walport, M.J. & Kurtz, A.B. (1979) Insulin deficient dia- betes. Contrasts with other endocrine deficiencies. Diabetologia, 16, 5–12. Nashef, L. & Brown, S. (1996) Epilepsy and sudden death. Lancet, 348, 1324–1325. Nathan, D., Cleary, P., Backlund, J. et al. (2005) Intensive diabetes treatment and cardiovascular disease in patients with type 1 diabetes. New England Journal of Medicine, 353, 2643–2653. Navarro-Gutierrez, S., Gonzalez-Martinez, F., Fernandez-Perez, M.T., Garcia-Moreno, M.T., Ballester- Vidal, M.R. & Pulido-Morillo, F.J. (2003) Bradycardia related to hypoglycaemia. European Journal of Emergency Medicine, 10, 331–333. Neligan, A., Bell, G.S. & Sander, J.W. (2011) Sudden death in epilepsy. British Medical Journal, 343, d7303. Odeh, M., Oliven, A. & Bassan, H. (1990) Transient atrial fibrillation precipitated by hypoglycemia.Annals of Emergency Medicine, 19, 565–567. Patel, A., Macmahon, S., Chalmers, J. et al. (2008) Intensive blood glucose control and vascular outcomes in patients with type 2 diabetes. New England Journal of Medicine, 358, 2560–2572. Pladziewicz, D.S. & Nesto, R.W. (1989) Hypoglycemia-induced silent myocardial ischemia. The American Journal of Cardiology, 63, 1531–1532. Pollock, G., Brady, W.J., Jr., Hargarten, S., DeSilvey, D. & Carner, C.T. (1996) Hypoglycemia manifested by sinus bradycardia: a report of three cases. Academic Emergency Medicine, 3, 700–707. Purucker, E., Nguyen, H.-N., Lammert, F., Koch, A. & Matern, S. (2000) Central pontine myelinosis and myocardial infarction following severe hypoglycemia. Intensive Care Medicine, 26, 1406–1407. Rana, O., Byrne, C.D., Kerr, D. et al. (2011) Acute hypoglycemia decreases myocardial blood flow reserve in patients with type 1 diabetes mellitus and in healthy humans. Circulation, 124, 1548–1556. Reaven, P.D., Moritz, T.E., Schwenke, D.C. et al. (2009) Intensive glucose-lowering therapy reduces cardiovascular disease events in veterans affairs diabetes trial participants with lower calcified coronary atherosclerosis. Diabetes, 58, 2642–2648. Chapter 13: Mortality, Cardiovascular Morbidity and Possible Effects of Hypoglycaemia 283

Robinson, R.T.C.E., Harris, N.D., Ireland, R.H., Lee, S., Newman, C. & Heller, S.R. (2003) Mechanisms of abnormal repolarization during insulin-induced hypoglycemia. Diabetes, 52, 1469–1474. Robinson, R.T.C.E., Harris, N.D., Ireland, R.H., Macdonald, I.A. & Heller, S.R. (2004) Changes in cardiac repolarization during clinical episodes of nocturnal hypoglycaemia in adults with Type 1 diabetes. Dia­ betologia, 47, 312–315. Roden, D.M., Balser, J.R., George, A.L., Jr. & Anderson, M.E. (2002) Cardiac ion channels. Annual Review of Physiology, 64, 431–475. Rugg-Gunn, F.J., Simister, R.J., Squirrell, M., Holdright, D.R. & Duncan, J.S. (2004) Cardiac arrhythmias in focal epilepsy: a prospective long-term study. Lancet, 364, 2212–2219. Russell, R.R., Chyun, D., Song, S. et al. (2001) Cardiac responses to insulin-induced hypoglycemia in non-diabetic and intensively treated type 1 diabetic patients. American Journal of Endocrinology, 281, E1029–E1036. Sampson, M.J., Wilson, S., Karagiannis, P., Edmonds, M. & Watkins, P.J. (1990) Progression of diabetic autonomic neuropathy over a decade in insulin-dependent diabetics. The Quarterly Journal of Medicine, 75, 635–646. Schächinger, H., Port, J., Brody, S. et al. (2004) Increased high-frequency heart rate variability during insulin-induced hypoglycaemia in healthy humans. Clinical Science, 106, 583–588. Secrest, A.M., Becker, D.J., Kelsey, S.F., LaPorte, R.E. & Orchard, T.J. (2011) Characterizing sudden death and dead-in-bed syndrome in Type 1 diabetes: analysis from two childhood-onset Type 1 diabetes reg- istries. Diabetic Medicine, 28, 293–300. Shenfield, G.M., Bhalla, I.P., Elton, R.A. & Duncan, L.J.P. (1980) Fatal coma in diabetes. Diabete et Metabolisme (Paris), 6, 151–155. Shimada, R., Nakashima, T., Nunoi, K., Kohno, Y., Takeshita, A. & Omae, T. (1984) Arrhythmia during insulin-induced hypoglycemia in a diabetic patient. Archives of Internal Medicine, 144, 1068–1069. Skrivarhaug, T., Bangstad, H.-J., Stene, L.C., Sandvik, L., Hanssen, K.F. & Joner, G. (2006) Long-term mortality in a nationwide cohort of childhood-onset type 1 diabetic patients in Norway. Diabetologia, 49, 298–305. Sommerfield, A.J., Wilkinson, I.B., Webb, D.J. & Frier, B.M. (2007) Vessel wall stiffness in Type 1 diabetes and the central hemodynamic effects of acute hypoglycemia. American Journal of Physiology. Endo­ crinology and Metabolism, 293, E1274–E1279. Sovik, O. & Thordarson, H. (1999) Dead-in-bed syndrome in young diabetic patients. Diabetes Care, 22 (Suppl. 2), B40–B42. Suarez, G., Clark, V., Norell, J. et al. (2005) Sudden cardiac death in diabetes mellitus: risk factors in the Rochester diabetic neuropathy study. Journal of Neurology, Neurosurgery, and Psychiatry, 76, 240–245. Tanenberg, R.J., Newton, C.A. & Drake, A.J. (2010) Confirmation of hypoglycemia in the ‘dead-in-bed’ syndrome, as captured by a retrospective continuous glucose monitoring system. Endocrine Practice, 16, 244–248. Tattersall, R.B. & Gale, E.A.M. (1993) Mortality. In: B.M. Frier & B.M. Fisher (eds), Hypoglycaemia and Diabetes: Clinical and Physiological Aspects, pp. 191–198. Edward Arnold, London. Tattersall, R.B. & Gill, G.V. (1991) Unexplained deaths of type 1 diabetic patients. Diabetic Medicine, 8, 49–58. Thordarson, H. & Sovik, O. (1995) Dead in bed syndrome in young diabetic patients in Norway. Diabetic Medicine, 12, 782–787. Trovati, M., Anfossi, G., Cavalot, F. et al. (1986) Studies on mechanisms involved in hypoglycemia- induced platelet activation. Diabetes, 35, 818–825. Tu, E., Bagnall, R.D., Duflou, J., Lynch, M., Twigg, S.M. & Semsarian, C. (2010) Post-mortem pathologic and genetic studies in ‘dead in bed syndrome’ cases in type 1 diabetes mellitus. Human Pathology, 41, 392–400. Tunbridge, W.M.G. (1981) Factors contributing to deaths of diabetics under fifty years of age. Lancet ii, 569–572. Turnbull, F., Abraira, C., Anderson, R. et al. (2009) Intensive glucose control and macrovascular outcomes in type 2 diabetes. Diabetologia, 52, 2288–2298. Van den Berghe, G., Wouters, P., Weekers, F. et al. (2001) Intensive insulin therapy in the critically ill patients. New England Journal of Medicine, 345, 1359–1367. Van den Berghe, G., Wilmer, A., Hermans, G. et al. (2006) Intensive insulin therapy in the medical ICU. New England Journal of Medicine, 354, 449–461. 284 Hypoglycaemia in Clinical Diabetes

Vinik, A., Maser, R., Mitchell, B. & Freeman, R. (2003) Diabetic autonomic neuropathy. Diabetes Care, 26, 1553–1579. Waernbaum, I., Blohme, G., Ostman, J. et al. (2006) Excess mortality in incident cases of diabetes mellitus aged 15 to 34 years at diagnosis: a population-based study (DISS) in Sweden. Diabetologia, 49, 653– 659. Weston, P.J., Panerai, R.B., McCullough, A. et al. (1996) Assessment of baroceptor-cardiac reflex sensitiv- ity using time domain analysis in patients with IDDM and the relation to left ventricular mass index. Diabetologia, 39, 1385–1391. Wieczorek, I., Pell, A.C.H., McIver, B., Macgregor, I.R., Ludlam, C.A. & Frier, B.M. (1993) Coagulation and fibrinolytic systems in type I diabetes: effects of venous occlusion and insulin-induced hypoglycae- mia. Clinical Science (London), 84, 79–86. Wright, R.J., Macleod, K.M., Perros, P., Johnston, N., Webb, D.J. & Frier, B.M. (2007) Plasma endothelin response to acute hypoglycaemia in adults with Type 1 diabetes. Diabetic Medicine, 24, 1039–1042. Wright, R.J. & Frier, B.M. (2008) Vascular disease and diabetes: is hypoglycaemia an aggravating factor? Diabetes/Metabolism Research and Reviews, 24, 353–363. Wright, R.J., Newby, D.E., Stirling, D., Ludlam, C.A., Macdonald, I.A. & Frier, B.M. (2010) Effects of acute insulin-induced hypoglycemia on indices of inflammation: putative mechanism for aggravating vascular disease in diabetes. Diabetes Care, 33, 1591–1597. Zhang, Y., Han, H., Wang, J., Wang, H., Yang, B. & Wang, Z. (2003) Impairment of human ether-à-go-go- related gene (HERG) K+ channel function by hypoglycemia and hyperglycemia. Similar phenotypes but different mechanisms. The Journal of Biological Chemistry, 278, 10417–10426. Zoungas, S., Patel, A., Chalmers, J. et al. (2010) Severe hypoglycemia and risks of vascular events and death. New England Journal of Medicine, 363, 1410–1418. 14 Long-Term Effects of Hypoglycaemia on Cognitive Function in Diabetes Christopher M. Ryan University of Pittsburgh School of Medicine, Pittsburgh, PA, USA

INTRODUCTION

Cognitive functioning is exquisitely sensitive to acute changes in blood glucose. As blood glucose begins to fall, there is a corresponding decline in one’s ability to pay attention, learn new information, and perform a variety of complex information-processing tasks quickly. Depending on one’s age, and whether or not one has type 1 or type 2 diabetes, these cogni- tive sequelae may begin to appear at blood glucose values as high as 3.3–3.5 mmol/l in diabetic children (Ryan et al. 1990). As glucose values continue to fall, there is a correspond- ing increase in the severity of cognitive dysfunction, whereas re-establishment of euglycae- mia leads to what appears to be complete cognitive recovery – albeit after a delay that may exceed an hour or more (Zammitt et al. 2008). These observations have obvious implications for the activities of daily living for any person with diabetes in so far as even mild acute hypoglycaemia may transiently compromise performance in school or in the workplace, and may interfere with highly skilled, cognitively demanding activities like driving (Cox et al. 2000). Less well established is whether recurrent episodes of moderate to severe hypoglycae- mia can lead to long-lasting changes in brain structure that eventuate in chronic cognitive complications. Early work suggested that this was indeed the case, but more recent studies have either failed to find any relationship between recurrent hypoglycaemia and cognition, or have found only limited effects. To further complicate the interpretation of this literature, the strength of such a relationship may vary appreciably, depending on the patient popula- tion studied – children with type 1 diabetes appear to show a greater sensitivity to recurrent hypoglycaemia than adults with either type 1 or type 2 diabetes. The focus of this chapter is on identifying the neurocognitive phenotypes associated with severe hypoglycaemia and delineating the biomedical and metabolic factors that may interact with hypoglycaemia to alter the integrity of the brain.

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 286 Hypoglycaemia in Clinical Diabetes

ADULTS WITH TYPE 1 DIABETES

Profound hypoglycaemia and brain dysfunction Clinical case reports have indicated that even a single episode of profound hypoglycaemia (i.e. blood glucose levels ≤∼1.5 mmol/l that are associated with an extended loss of con- sciousness) may lead to extensive structural damage to the brain, as well as to clinically significant cognitive impairment. Neuroimaging suggests that structural damage evolves over a period of 7–14 days (Fujioka et al. 1997) and typically affects frontal and/or temporal cortical regions bilaterally and symmetrically, as well as the basal ganglia, hippocampus and brainstem (Auer et al. 1989; Jung et al. 2005). The few cases that have been autopsied also show neuronal necrosis widely distributed throughout the cerebral hemispheres – par- ticularly in superficial layers of the cortex (layers I-III), with extensive neuronal loss in the hippocampus, as well as marked myelin pallor – with astrocytosis, in multiple white matter tracts (Mori et al. 2006). A key brain region that is only now beginning to receive attention is the basal ganglia, which is almost invariably affected following profound hypoglycaemia, and its integrity may be useful in predicting long-term prognosis (Finelli 2001; Ma et al. 2009). Two well-established biomarkers of neuronal injury – serum levels of neurone- specific enolase and protein S-100 – are also elevated in patients who experienced very profound hypoglycaemia. In contrast, those adults who experienced only a moderately severe episode (blood glucose <2.8 mmol/l, without extended loss of consciousness) had normal levels of both markers (Strachan et al. 1999). This work suggests that profound hypoglycaemia may affect brain metabolism in a way that is qualitatively different from what occurs during a less severe hypoglycaemic episode. Few case reports have included a neurocognitive assessment, but in those that have, learning and memory skills were most often compromised, as one might expect given the involvement of the hippocampus and mesial temporal cortex (Boeve et al. 1995). In one case, a highly-educated young adult with type 1 diabetes who experienced a hypoglycaemic coma for approximately 6–8 hours before being treated showed evidence of a classic Minutengedächtnis, characterised by an inability to hold information in memory for more than a few minutes (Chalmers et al. 1991). Whether these lesions – and the associated neurocognitive deficits – are permanent or whether they may ameliorate over time remains controversial. Several writers have sug- gested that under some very limited circumstances, lesions may be reversible. In one case, the initial lesions were in the internal capsule, corona radiata, and frontoparietal cortex, and regressed when the MRI was repeated 10 days later, and the subject showed complete neurological recovery (Aoki et al. 2004). Interestingly, both the hippocampus and basal ganglia had been spared. Similar lesion regression – with corresponding improvement in neurological status – has also been reported by others performing serial MRI studies (Lo et al. 2006; Terakawa et al. 2007). These case reports demonstrate that a prolonged period of profound hypoglycaemia has the potential to damage the brain. However, these cases are quite atypical in a number of ways. Not only are exceedingly low blood glucose values very rare events, but in the vast majority of published cases, the patients tended to be over the age of 50, had experi- enced a coma lasting for several hours, and often had a poorly managed form of ‘brittle’ diabetes characterised by wildly fluctuating blood glucose levels, and/or had some other co-morbid medical condition (e.g. chronic alcohol abuse; HIV) known to affect the integrity of the CNS. Chapter 14: Long-Term Effects of Hypoglycaemia on Cognitive Function in Diabetes 287

100 Subiculum Caudate

80

Dentate gyrus 60

CA1 hippocampus 40 Neuronal necrosis (%)

20

0 0 10 20 30 40 50 60 Duration of Isoelectric EEG (minutes)

Figure 14.1 Degree of neuronal necrosis varies as a consequence of duration of profound hypogly- caemia (defined as duration of isoelectric EEG) and specific brain region. Neurones in the subiculum (inferior component of hippocampus) and caudate (basal ganglia) are more vulnerable than those in the dentate gyrus and CA1 region of the hippocampus. (Source: Adapted using data from Auer et al., 1984, and Suh et al., 2007).

Experimental animal studies have demonstrated that the primary determinant of cerebral damage is both the depth and duration of hypoglycaemia – regardless of other pre-existing brain insults or metabolic disorders like diabetes. Classical studies conducted by Auer and his colleagues have shown that neuronal necrosis develops only after blood glucose has fallen below 1.5 mmol/l and those values have been maintained for a period of time suf- ficient to reach the threshold for energy failure, as indexed by a flat, or isoelectric, electro- encephalogram (EEG) (Auer 2004). At that point, a cascade of neurochemical changes is initiated that includes a rapid reduction in adenosine triphosphate, the disruption of ion homeostasis and a corresponding cellular influx of calcium, and the release of the excitatory amino acids – aspartate and glutamate, which bind to dendrites and initiate the process of neuronal necrosis (Auer and Siesjö 1993). As the duration of the EEG isoelectric period increases from 10 to 60 minutes, there is a corresponding increase in the proportion of dead neurones, with some brain regions being more sensitive than others (Figure 14.1). Although earlier writers considered nerve cell death to be a direct consequence of energy failure secondary to neuroglycopenia, it is now considered to be the result of an excitotoxic process that results from the sustained activation of glutamate receptors (Suh et al. 2007).

Recurrent episodes of moderate to severe hypoglycaemia Neuropsychological manifestations If there is a linear relationship between the severity of hypoglycaemia and the extent of brain dysfunction, one might expect that individuals who experienced moderately severe hypoglycaemia would also manifest neurocognitive impairment, but to a lesser degree. 288 Hypoglycaemia in Clinical Diabetes

Several early, relatively small studies reached exactly that conclusion, but a large number of more recent studies, using better ascertainment of hypoglycaemic events, have failed to support that assertion. The premise underlying the use of neuropsychological testing is that abnormalities in brain structure induce changes in behaviour that can be detected by administering a battery of tests that are sensitive to specific cognitive abilities. By assessing multiple cognitive domains (e.g. language, visuospatial skills, attention, learning, memory, problem solving, mental flexibility, psychomotor coordination, and speed of information processing), the neuropsychologist can identify a pattern of strengths and weaknesses and draw inferences about the integrity of discrete brain systems known to underlie those skills. Further, by comparing patients with various metabolic/biomedical characteristics, researchers can iden- tify risk factors associated with the development of cognitive dysfunction. Early studies recruited groups of subjects with diabetes who had a history of moderate or severe hypoglycaemia and compared them to demographically similar people with diabetes with no such history. Although there is no universally accepted definition of severe hypoglycaemia, most researchers operationalised this construct in one of two ways. When the Diabetes Control and Complications Trial (DCCT) began in 1982, it defined moderate/severe hypoglycaemia as any event that required the assistance of another person and in which the blood glucose level was less than 2.8 mmol/l, or the symptoms were subsequently reversed by oral carbohydrate, injected glucagon, or intravenous glucose (The Diabetes Control and Complications Trial Research Group 1993). Most studies, particularly those conducted in North America, routinely used this definition. During the subsequent DCCT follow-up study (Epidemiology of Diabetes Interventions and Complications – EDIC), a more conservative definition of severe hypoglycaemia was adopted that required evidence of a seizure and/or a coma that was accompanied by blood glucose values less than 2.8 mmol/l (The Diabetes Control and Complications Trial/ Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Study Research Group 2007). In a classical early study, Deary and his colleagues evaluated 100 insulin-treated adults who developed diabetes after adolescence, and assigned them to subgroups based on prior episodes of severe hypoglycaemia as defined by the earlier DCCT criteria (Langan et al. 1991). Their subjects who had repeated episodes of hypoglycaemia earned lower Per­ formance IQ scores (indicative of impaired abstract reasoning and problem-solving skills) and performed more slowly on tasks requiring rapid decision-making. In contrast, no meaningful differences were observed between those with and without a history of severe hypoglycaemia on measures of verbal intelligence or learning and memory. Additional analyses revealed statistically reliable correlations between the number of episodes of severe hypoglycaemia and degree of cognitive decrement on these same measures (Deary et al. 1992, 1993). More recent research has generally failed to find any relationship between severe hypoglycaemia and cognition, particularly after taking into account the presence of diabetes- related microvascular changes. Results from the first published meta-analysis of cognitive outcomes in adults with type 1 diabetes included data from 33 published studies and showed that people with diabetes performed more poorly than their non-diabetic counterparts on measures of intelligence, psychomotor efficiency, attention, mental flexibility and visuo- perception (Brands et al. 2005). However, when adults with type 1 diabetes with and without a history of severe hypoglycaemia were compared, no meaningful between-group difference was evident in any cognitive domain (Table 14.1). These authors concluded that Chapter 14: Long-Term Effects of Hypoglycaemia on Cognitive Function in Diabetes 289

Table 14.1 Effects of recurrent episodes of hypoglycaemia on cognition in adults with type 1 diabetes, based on a meta-analysis of published papers. Standardised effect sizes (Cohen’s d) for each cognitive domain reflect differences between those subjects with and without a history of hypoglycaemia. An effect size (d) of 0.2 to 0.49 is considered ‘small’; values of 0.5 to 0.79 are considered ‘medium’. (Source: Adapted from Brands AMA 2005) (Figure 14.2).

Domain Effect Size Significance Total N

Overall Cognition <0.10 NS 487 Intelligence 0.20 NS 118 Attention <0.10 NS 391 Learning and Memory • Working Memory <0.10 NS 369 • Learning and Immediate Recall 0 NS 402 • Delayed Memory 0 NS 378 Psychomotor Activity and Speed of Information Processing • Psychomotor Efficiency <0.10 NS 420 • Speed of Information Processing 0.22 NS 50 • Motor Speed <0.10 NS 377 Cognitive Flexibility 0 NS 419 Visual Perception 0 NS 363

while diabetes itself may adversely affect cognition – either via the metabolic disruptions associated with chronically elevated blood glucose values and/or the micro- and macrov- ascular changes that accompany diabetes of long duration – there is no evidence that recur- rent hypoglycaemia compromises cognition in a meaningful way. Because patients who experience severe hypoglycaemia often have diabetes for a longer period of time and also have poorer levels of metabolic control or higher rates of micro- vascular complications (Langan et al. 1991), it is critically important to evaluate the potential role of chronically high blood glucose values when assessing the effects of severe hypoglycaemia on cognition. Most subsequent studies that have done this have tended to find that variables like duration of diabetes, degree of poor metabolic control (higher glycated haemoglobin [HbA1c] values), and microvascular complications are the best pre- dictors of cognitive dysfunction (Ferguson et al. 2003; Ryan et al. 2003; Brands et al. 2006). The most definitive evidence demonstrating that episodes of moderately severe hypoglycaemia have little or no impact on cognition comes from the DCCT/EDIC study group. Because of concerns that intensive insulin therapy could increase the risk of severe hypoglycaemia and thereby adversely affect brain function, a comprehensive battery of cognitive tests was administered at study entry to adolescents and young adults with type 1 diabetes, with repeated assessments over a follow-up period that averaged 18 years (The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Study Research Group 2007). Data from 1144 adults showed that although rates of prospectively ascertained severe hypoglycaemia (defined as seizure or coma) were high, with a total of 1355 episodes in 453 subjects, there was no statistically reliable relationship between the total number of hypoglycaemic episodes and changes in performance in any of the eight cognitive domains that were assessed (Figure 14.2). 290 Hypoglycaemia in Clinical Diabetes

No hypo 0.4 Hypo: 1-5 Hypo: 5+ 0.3 Delayed Learning Attention Motor speed 0.2 recall

0.1

0.0

–0.1

–0.2 Z score change from baseline

–0.3 Problem Immediate Spatial –0.4 solving memory processing Psychomotor

Figure 14.2 Change in performance within each cognitive domain between baseline cognitive testing and follow-up testing (mean: 18-year interval) expressed as z scores, and stratified on the basis of number of episodes of severe hypoglycaemia during the study (none; 1–5 episodes; >5 episodes). Within each cognitive domain there were no statistically reliable differences. (Source: Adapted from The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/ EDIC) Study Research Group 2007).

The very long follow-up period, the extensive cognitive assessment, the very high rates of severe hypoglycaemia and the large number of subjects who were assessed repeatedly, makes a very compelling case that if moderate hypoglycaemia disrupts brain function in a chronic manner, its effects are not readily detected. One criticism of this work is that these findings may not be widely applicable because the patients who were recruited into the DCCT may not be entirely representative of the ‘typical’ diabetic patient (Frier 2011). Not only were participants young (13–39 years old at baseline), generally healthy and highly motivated, but potential subjects were excluded if they had a past history of recurrent severe hypoglycaemia prior to enrolment, and if they were unable to perform a series of demand- ing blood glucose monitoring tests. Furthermore, regardless of study treatment arm, they received a level of monitoring and healthcare that is consistently better than what is afforded to most diabetes patients – at least in North America. All of these factors could have reduced the risk of developing cognitive impairment – for any reason – in this par- ticular group of patients. Nevertheless, subsequent published analyses of the DCCT/EDIC data have demonstrated that this possibility is unlikely, in so far as cognitive decline was detected over the 18 year follow-up, primarily on measures of psychomotor and mental speed. However, these cognitive changes were associated most strongly not with hypogly- caemia but with several markers of chronic hyperglycaemia, including elevated glycated haemoglobin levels as well as the development of two clinically significant microvascular complications – serious diabetic retinopathy and diabetic renal disease. Those data suggest that chronic hyperglycaemia is the primary driver of cognitive change. Chapter 14: Long-Term Effects of Hypoglycaemia on Cognitive Function in Diabetes 291

Electrophysiological changes The integrity of the CNS has been investigated with a variety of other measures, but few studies have found strong evidence that hypoglycaemia has a clinically significant impact on brain function or structure. One small study recorded EEGs to assess resting brain activity and found that a recent history of recurrent severe hypoglycaemia was associ- ated with a reduction in spectral power in the beta frequency band (13–35 Hz) and reduced vigilance – defined in terms of the alpha slow wave index (Howorka et al. 2000). By contrast, a larger study failed to find a relationship between severe hypogly- caemia and any EEG parameter, although as a group, subjects with diabetes – regardless of whether or not they had a past history of severe hypoglycaemia – showed evidence of slowing at multiple brain regions, including a loss of fast oscillations (alpha, beta and gamma activity) in temporal regions, and a reduction in the alpha peak frequency (Brismar et al. 2002). Adults with diabetes also manifest significant neural slowing when event-related poten- tial (ERP) latencies are recorded. This experimental paradigm differs from the standard EEG study in so far as cortical changes are measured not during rest, but when the brain is activated during the performance of a demanding test of attention (e.g. ‘count the number of occurrences of an infrequently presented high-pitched tone and ignore the more common low pitched tones’). Approximately 300 msec after the occurrence of a rare tone, a distinctive, positively deflected wave appears (P300). Latencies longer than 300 msec signal the presence of neural abnormalities. In what remains the largest study to use this paradigm, researchers found significantly longer P300 latencies in their well-controlled group of adults with diabetes, as compared to a non-diabetic reference sample – suggest- ing that diabetes per se disrupts normal neural functioning. Contrary to expectations, however, these effects were unrelated to a past history of severe hypoglycaemia (Kramer et al. 1998).

Brain structure anomalies Results from both the neuropsychological and electrophysiological literature suggest that at least in adults, hypoglycaemia does not appear to have a clinically significant impact on brain function. The possibility remains, however, that brain structure has been com- promised but that other, still intact, brain regions are able to somehow compensate and take over or otherwise normalise specific neurocognitive functions. Two recent neuroim- aging studies have used sophisticated MRI techniques to explicitly test the hypothesis that recurrent moderately severe hypoglycaemia induces structural abnormalities in the brain. Both studies failed to find compelling evidence to support their hypothesis, but both did find reductions in grey matter density in multiple brain regions (Musen et al. 2006), as well as white matter hyperintensities in the basal ganglia (Ferguson et al. 2003), which were associated with the development of retinopathy secondary to chronic hyperglycaemia. Brain structure has also been examined in several other very small MRI studies, but in most instances, the effects of hypoglycaemia were either not reported, or no meaningful relationships were observed (Lunetta et al. 1994). One small clinical study did find an elevated rate of cortical atrophy in adults with diabetes who had experienced severe hypoglycaemia compared to those without such a history (rate of atrophy: 45% vs 0%) (Perros et al. 1997). There is no good explanation as to why these data differ from more 292 Hypoglycaemia in Clinical Diabetes recent larger studies, with one exception: these subjects were considerably older (mean age = 45 years) than subjects in studies that found no structural anomalies (mean age = 28 years (Ferguson et al. 2003); mean age = 32 years (Musen et al. 2006)), and it is possible that the higher rates in older subjects with diabetes may reflect some form of an accelerated ageing process.

CHILDREN AND ADOLESCENTS WITH TYPE 1 DIABETES

Profound hypoglycaemia and brain dysfunction The neurocognitive correlates and consequences of neonatal hypoglycaemia have been well characterised, but there are essentially no reports of the neurocognitive effects of profound hypoglycaemia in children or adolescents with diabetes. One case series of neonates without diabetes who experienced profound hypoglycaemia showed that extensive brain abnormalities are common, although in a minority of cases – particularly those who had somewhat higher blood glucose values – the abnormalities tended to reverse over time (Alkalay et al. 2005). Studies that have followed infants over time have shown that the severity of white matter injury shortly after birth is associated with marked neurodevelop- mental delays at 18 months (Burns et al. 2008). These and other studies provide unequivo- cal evidence that in individuals without diabetes, profound hypoglycaemia very early in life can cause very significant damage to the CNS – particularly in the posterior cortex, basal ganglia, and white matter tracts (Inder 2008). The extent to which such episodes disrupt brain development when they occur after the neonatal period remains to be deter- mined. More research is also required to assess the consequences of less severe hypogly- caemia (e.g. blood glucose values >2.7 mmol/l) early in life.

Recurrent episodes of moderate to severe hypoglycaemia Neuropsychological manifestations The possibility that children and adolescents with diabetes may develop neurocognitive complications has intrigued clinicians and researchers for more than 75 years and has resulted in a very extensive literature indicating that diabetes per se not only increases the risk of mild cognitive dysfunction, but it is also associated with neural slowing, neurometa- bolic abnormalities, and brain structure anomalies (Ryan 2009). The strongest predictor of poorer neurocognitive outcomes is age at onset of diabetes. All other things being equal, children who develop diabetes within the first 5–7 years of life show the greatest degree of functional and structural change to the CNS. This well-established observation has led researchers to speculate on underlying mechanisms, with neuroglycopenia frequently being identified as a likely agent. Not only are episodes of hypoglycaemia more common – and more difficult to detect – in very young children (Wagner et al. 2005), but the devel- oping nervous system is exquisitely sensitive to a wide range of insults (Taylor and Alden 1997), including neuroglycopenia (Caksen et al. 2011). For those reasons, many studies have explicitly tested the hypothesis that one or more episodes of moderately severe hypoglycaemia may affect brain integrity – particularly if they occur early in life. As described in more detail below, several very well-conducted studies have reported some support for that position, but there are multiple contradictory findings, and as of yet, no consensus. Chapter 14: Long-Term Effects of Hypoglycaemia on Cognitive Function in Diabetes 293

Compared to their non-diabetic peers, children and adolescents with diabetes – regard- less of history of hypoglycaemia – are more likely to attain somewhat lower grades in school, fail or repeat a grade, experience more school absences, perform more poorly on tests of academic achievement, and when followed over an extended period of time, have lower secondary school graduation rates (68% vs 85%) and earn lower IQ scores (101.3 vs 105.1) (Northam et al. 2009; Parent et al. 2009). Psychosocial factors may make some contribution to the poorer classroom performance seen in some children with diabetes, but it is more likely that these classroom-related problems are actually a reflection of CNS dysfunction secondary to diabetes and its treatment. Support for that possibility comes from neuropsychological research demonstrating that youngsters with type 1 diabetes consistently earn somewhat lower scores on tests of atten- tion, visuospatial skills, verbal intelligence, executive function and psychomotor speed (Gaudieri et al. 2008; Patiño-Fernández et al. 2010). Examination of specific test results suggests that speed of information processing is particularly compromised, and is most evident when the task has several different components that need to be quickly integrated and executed. Regardless of the type of task, the magnitude of diabetes-associated cognitive dysfunction is especially pronounced in those children who developed diabetes within the first few years of life. According to one early study, children diagnosed before the age of six had a greatly elevated rate of clinically significant impairment (24%), as compared to those diagnosed after that age (6%) or to non-diabetic comparison subjects (6%) (Ryan et al. 1985). Multiple research teams have explicitly examined the effects of severe hypoglycaemia on cognition and have typically found either no, or very weak, relationships (Schoenle et al. 2002; Wysocki et al. 2003; Hannonen et al. 2012). This may be best illustrated by research conducted solely on children who developed diabetes before the age of six (Strud- wick et al. 2005). When assessed at 10 years of age, those subjects who experienced one or more hypoglycaemic seizures or comas performed no worse on any cognitive test than the 43 without a history of severe hypoglycaemia. Furthermore, the two groups earned identical scores on all memory tests despite the fact that the structures involved in memory processing (e.g. hippocampus) are also known to be damaged during an episode of severe hypoglycaemia (Auer et al. 1984). When these subjects were subsequently re-evaluated as young adults, similar null effects were found – with one exception. A small subgroup (n = 6) who experienced severe hypoglycaemia before 6 years of age performed signifi- cantly more poorly on a single measure of abstract reasoning and concept formation, as compared to those who experienced severe hypoglycaemia after that age (n = 14) or those who had no history of severe hypoglycaemia (n = 13) (Ly et al. 2011). These data must be interpreted cautiously, given the highly circumscribed nature of this deficit (i.e. perform- ance on only a single test was affected) and the very small number of subjects studied. Other paediatric studies have also reported relationships between recurrent hypoglycae- mia and cognition, although in virtually all cases, those subjects who experienced an episode of severe hypoglycaemia had also developed diabetes very early in life. The most compelling data have come from Bjørgass and her colleagues, who followed diabetic sub- jects with and without early severe hypoglycaemia and compared them to healthy non- diabetic subjects (Åsvold et al. 2010). Within the severe hypoglycaemia group, four subjects experienced one or more episodes during the first 5 years of life; five experienced episodes after that period. At study entry, those children with type 1 diabetes who had been exposed to an early episode of severe hypoglycaemia performed far worse (on average, 0.7 standard deviation units) than those without a history of early severe hypoglycaemia. When assessed 294 Hypoglycaemia in Clinical Diabetes

Severe hypoglycaemia Overall No hypoglycaemia Verbal functions Spatial functions Problem solving Attention Psychomotor

Motor speed Memory

–2.0 –1.5 –1.0 –0.5 0.0

Changes from childhood to adulthood over 16 years In standard deviation units (z scores)

Figure 14.3 Differences between young adults with, and without, a history of severe hypoglycaemia in childhood, who were followed over a 16-year period. Z-scores are based on the performance of non-diabetic comparison subjects who were assessed at the same time points (baseline: mean age = 12 yrs; follow-up: mean age = 28 yrs). Differences between the two groups are significant (P < 0.001) only for Problem Solving. (Source: Adapted from Åsvold BO (2010).

as adults, 16 years later, performance worsened further, with the hypoglycaemia subgroup achieving scores that were, when averaged across all domains, an additional 0.3 standard deviation units poorer. Differences between those with, and without, severe hypoglycaemia were most pronounced on measures of problem solving (∼2 z score units) and attention (∼1.5 z score units), but because of the small sample size, only the problem-solving findings were statistically significant. On the other hand, the performance of those without a history of severe hypoglycaemia was comparable to healthy non-diabetic comparison subjects at both points in time (Figure 14.3). Because both the diagnosis of diabetes and the severe hypoglycaemic events occurred during the first 5–6 years of life – a critical period of brain development – it is difficult to determine whether the resulting brain insult was primarily a consequence of neuroglyco- penia, or whether the metabolic changes associated with the onset of diabetes (e.g. ketoaci- dosis; increased brain glucose; altered insulin availability) changed brain development to such an extent that it became more vulnerable to any subsequent insult (e.g. from an episode of severe hypoglycaemia). As reviewed elsewhere (Ryan 2006), there is now growing support for the latter, so-called ‘diathesis’ or vulnerability model, but at this time the pos- sibility cannot entirely be ruled out that neuroglycopenia has a direct effect on brain matu- ration by somehow disrupting the normal process of neuronal proliferation and pruning, as well as axonal myelination and the formation of synaptic connections (Caviness et al. 1997). Relationships between severe hypoglycaemic events and poorer cognitive test perform- ance in children have also been reported by other investigators, although the magnitude of these effects has tended to be quite modest and, once again, is most prominent in the sub- jects who developed type 1 diabetes early in life (Perantie et al. 2008; Lin et al. 2010). Although a very recent meta-analysis has similarly concluded that recurrent hypoglycaemia Chapter 14: Long-Term Effects of Hypoglycaemia on Cognitive Function in Diabetes 295 adversely affects cognition in diabetic children and adolescents (Blasetti et al. 2011), it appears to this reviewer that those authors may have overinterpreted their findings. From their analysis of the results of 12 published studies, Blasetti and her colleagues concluded that recurrent severe hypoglycaemia selectively affects certain cognitive domains, particu- larly intelligence, learning, memory and verbal fluency. However, examination of these data suggests that those conclusions may be driven by a very small number of ‘influential outliers’. For example, of the 11 measures of learning ability that were considered, only two (18%) were significantly different from 0. Similarly, of the 31 tests of memory evalu- ated, only seven (23%) were significantly different from 0. The very high level of hetero- geneity seen across these studies may reflect small sample sizes, differences in defining hypoglycaemia, and a failure to take into account age at onset of diabetes. More systematic research on this topic is necessary to thoroughly disentangle the effects of hypoglycaemia from other putative risk factors for brain dysfunction.

Electrophysiological changes Speed of central neural transmission is typically altered in children and adolescents with type 1 diabetes. One large EEG study found increments in slow wave activity (delta and theta) and reductions in alpha peak frequency, particularly in frontal brain regions (Hyllienmark et al. 2005). Past episodes of severe hypoglycaemia as well as a long history of poor metabolic control were both associated with these neurophysiological abnormalities, but because both of these factors tended to occur together in their subjects and the regional distributions of the EEG parameters were similar for each, the investiga- tors were unable to determine the unique contribution made by severe hypoglycaemia. When statistical algorithms were used to identify clinically significant pathology, between 10 and 23% of subjects met criteria for an abnormal recording, with the highest rate of abnormal findings in the frontal region. Similar rates of pathology (26% of adolescents with diabetes versus 7% of non-diabetic comparison subjects) have been reported in older studies that graded EEG recordings visually, using clinical rather than statistical classi- fication criteria (Soltész and Acsádi 1989). The strongest predictors of a clinically sig- nificant EEG abnormality were an early age at onset, and multiple episodes ofsevere hypoglycaemia. Longitudinal data from Bjørgass’s group suggests that EEG abnormalities that are present in childhood may revert to normal as the brain continues to mature (Åsvold et al. 2011). For example, when quantitative topographical EEG techniques were used to evaluate 13-year-old subjects with and without severe hypoglycaemia, increased theta activity bilat- erally was found in the frontocentral region with reductions in relative alpha amplitude in those who had a history of severe hypoglycaemia (Bjørgaas et al. 1996). When these same subjects were re-evaluated 16 years later, no relationship was found between episodes of severe hypoglycaemia and any EEG amplitude variable in any brain region – a finding that contrasts with their observation that severe hypoglycaemia continues to affect cognition adversely (Åsvold et al. 2010). Because of the small number of subjects studied, this finding needs to be replicated, primarily because the authors provide no plausible mechanism that could account for this electrophysiological ‘normalisation’. It does suggest, however, that whatever impact hypoglycaemia may have on the developing nervous system, its magni- tude is remarkably small. 296 Hypoglycaemia in Clinical Diabetes

Brain structure anomalies Multiple studies have identified small structural changes in the CNS of children and ado- lescents with diabetes. Data from the Melbourne Longitudinal Study showed that 12 years after study enrolment, diabetic subjects (now aged 20.5 years, on average) manifested a 6–10% decrease in grey matter volume in several regions, as compared to healthy controls (Northam et al. 2009). Reductions were most prominent in the thalamus, para-hippocampal gyrus, insula and both frontal and parietal cortical regions. White matter volume was also reduced, particularly in left mesial temporal and middle frontal regions. The pathophysi- ological basis for these effects remains unknown, although both severe hypoglycaemia and poor metabolic control were associated with changes in one structure – the thalamus – but not in other areas. Brain changes are also seen, but to a lesser extent, in younger children (mean age = 12 years) with a much briefer duration of diabetes (5.7 years, on average) (Perantie et al. 2007). Although white and grey matter volumes were comparable for diabetic and non- diabetic children, within the diabetic sample multiple relationships were found between metabolic variables and brain structure. Severe hypoglycaemia was associated with smaller grey matter volumes in a single area – the left superior temporal region – whereas a history of poorer metabolic control was associated with reductions in grey and white matter volumes in multiple posterior regions, as well as increments in circumscribed prefrontal regions. A somewhat different pattern of results has been reported in a study of still younger children (mean age = 10 years), all of whom had developed type 1 diabetes before 6 years of age (Ho et al. 2008). Within this early-onset cohort, 16% manifested mesial temporal sclerosis (MTS) – a very rare structural abnormality that ordinarily occurs in less than 1% of the paediatric population. Somewhat unexpectedly, the occurrence of hypoglycaemic seizures was not a risk factor for the development of MTS; if anything, there was a trend for a somewhat higher prevalence of MTS in the ‘no seizure’ group (20% vs 12.5%). Further, the subgroup of 16 subjects that had seizures before 6 years of age, were no more likely to manifest MTS than the 16 subjects who experienced their first seizure after that age. Although completely unexpected, these findings are consistent with the view that having diabetes early in life – regardless of whether or not one experiences hypoglycaemic seizures – may adversely affect normal brain development and that, in turn, may influence cognition (Ryan 2006). Because brain maturation is most dramatic within the first 5–7 years of life, one might expect to find, in very young children, structural abnormalities that are associated with diabetes and/or severe hypoglycaemia. The first study to use MRI to measure brain volumes in 7-year-old diabetic children reported data consistent with that hypothesis (Aye et al. 2011). Although as a group, children with diabetes had grey and white matter brain volumes that were similar to non-diabetic comparison subjects, there was a significant age by group interaction, such that those children with diabetes did not show the expected age-associated increase in white matter volume that was seen in the healthy comparison subjects. Further- more, those subjects with diabetes who had a history of severe hypoglycaemia (n = 8) had smaller grey and white matter volumes when compared to those (n = 19) without seizures. Neuropsychological testing showed a similar pattern: those with a history of severe hypoglycaemia performed significantly more poorly on measures of working memory, perceptual reasoning, and processing speed. This is the first study to examine brain integrity in very young children, and it will be of great interest to follow these children over time Chapter 14: Long-Term Effects of Hypoglycaemia on Cognitive Function in Diabetes 297 to determine whether these effects worsen over time, or whether normalisation occurs, as has been reported in EEG studies (Åsvold et al. 2011).

ANIMAL MODELS OF RECURRENT HYPOGLYCAEMIA

Animal research has demonstrated repeatedly that the structure and function of the nervous system is disrupted shortly after the experimental induction of diabetes in adult rats (Bies- sels and Gispen 2005). Reductions in neurometabolities occur after only 2 weeks of dia- betes, neural slowing appears after 3–4 months of diabetes, and learning and memory skills are compromised when measured after 3–4 months of untreated diabetes, but are normal when insulin treatment is provided (Biessels et al. 1996). Only recently have investigators begun to examine the extent to which diabetes and moderately severe hypoglycaemia affect brain structure and function in younger animals. Not surprisingly, given the generally nega- tive findings from the clinical literature, this animal work has generally failed to support the hypothesis that bouts of hypoglycaemia disrupt the CNS. The best, albeit imperfect, research on this topic has been conducted by Malone and his associates who studied three groups of young rats over an 8-week period, beginning at 4 weeks of age. Animals received either streptozotocin (STZ) to induce diabetes, ex­­ perienced 3 hours of insulin-induced moderate hypoglycaemia (blood glucose values <3.3 mmol/l), three times a week for 8 weeks, or were untreated (Malone et al. 2008). Behavioural testing showed that while the diabetic rats learned as well as a non-diabetic control group, their delayed recall of the maze was poor; in contrast, the rats that experi- enced multiple bouts of hypoglycaemia showed no evidence of impairment on either learn- ing or memory tests. At autopsy, the diabetic rats – but not the hypoglycaemic rats – showed marked abnormalities in neuronal morphology, including a decline in dendritic branching, fewer dendritic spines per branch, and alterations in the actual shape of the dendritic spines (Figure 14.4). How diabetes induces those neuronal changes has not yet been established, but the pos- sibility exists that early in the course of diabetes, excessive levels of glucose enter the brain via a transient breach in the blood–brain barrier (BBB), which in turn is triggered by the very high levels of circulating glucose that are present around the time of diagnosis. The neurotoxic effects of glucose then initiate – at least for a limited period of time – some degree of brain remodelling. Consistent with that view are data from animal studies dem- onstrating that the BBB became hyper-permeable to [14C] sucrose 2 weeks after the experi- mental induction of diabetes (Hawkins et al. 2007). Not only were these changes concurrent with a decreased production of tight junction proteins at the BBB – which presumably enables the breach – but the magnitude of the effect was reduced following treatment with insulin. From this reviewer’s perspective, more animal research is clearly needed, particularly studies that systematically manipulate the degree and duration of hypoglycaemia. Malone’s group studied ‘moderate’ hypoglycaemia (blood glucose <3.3 mmol/l), which in children is sufficient to induce acute changes in brain function, and is typical of the degreeof hypoglycaemia experienced by many children (Levine et al. 2001). Nevertheless, before concluding from these animal studies that hypoglycaemia has no, or only minimal, effects on the developing brain, it is imperative that researchers evaluate the effects of recurrent bouts of somewhat greater degrees of ‘moderately severe’ hypoglycaemia (e.g. 2.8 or 2.2 mmol/l) and study these effects in very young animals that have been made diabetic. 298 Hypoglycaemia in Clinical Diabetes

(a) (b) (c)

Figure 14.4 Camera lucida drawings of the basilar tree of layer II-III pyramids from (a) control rat brain, (b) hypoglycaemic rat brain, and (c) hyperglycaemic rat brain. Repeated bouts of hypoglycaemia (b) had no significant impact on neuronal morphology, whereas STZ-induced diabetes (c) reduced the density and morphology of dendritic spines. (Source: Malone JI 2008. Reproduced with permission from John Wiley & Sons, Ltd).

This is an important consideration because of other data indicating that animals with STZ diabetes that subsequently experience an experimentally-induced profound hypoglycaemia (blood glucose ∼0.6–0.8 mmol/l) for an hour manifest a 2.3-fold increase in neuronal necrosis in cortical brain regions, as compared to animals without diabetes that were exposed to the same degree of hypoglycaemia (Bree et al. 2009). Whether similar effects would be seen in studies comparing diabetic and non-diabetic animals that are exposed to less severe hypoglycaemia remains to be determined.

RECURRENT HYPOGLYCAEMIA IN OLDER ADULTS WITH TYPE 2 DIABETES

Neurocognitive complications are common in older adults with type 2 diabetes, and are phenotypically quite similar to what has been reported in younger adults with type 1 dia- betes. Compared to their non-diabetic peers, adults with type 2 diabetes are more likely to show modest decrements on cognitive tasks that require rapid information processing, as well as those that rely on the ability to pay attention, learn new information, and remember previously learned material (Brands et al. 2007; Reijmer et al. 2010). Neural slowing is also evident (Mooradian et al. 1988; Kurita et al. 1996) as are reductions in brain volume. Less grey matter volume, greater subcortical atrophy, and larger white matter lesions have been reported in patients with type 2 diabetes, and these tend to be correlated with the degree of chronic hyperglycaemia – indexed by duration of diabetes, glycated haemoglobin values, and presence of microvascular complications (Jongen et al. 2007). As one might Chapter 14: Long-Term Effects of Hypoglycaemia on Cognitive Function in Diabetes 299 predict, greater cortical atrophy is associated with more severe performance decrements on cognitive tests – particularly in those subjects who also had significant comorbid vascular disease (Tiehuis et al. 2009). Most researchers have ignored the possibility that these neurocognitive complications could be related to recurrent hypoglycaemia, for two very salient reasons. First, until recently, it was widely believed that rates of hypoglycaemia were quite low in patients with type 2 diabetes. (Shorr et al. 1997). Although more recent studies have found somewhat higher rates, there is much variability across studies, largely as a consequence of the meth- odological challenges in accurately estimating rates of mild and severe hypoglycaemia in this older patient population who often have several co-morbidities (Zammitt and Frier 2005). The few studies that have explicitly tested the hypothesis that recurrent, severe hypoglycaemia is associated with neurocognitive dysfunction in people with type 2 diabe- tes have yielded contradictory findings. One largely negative study examined 302 surviving participants from the Fremantle Diabetes Study who were 70 years or older, and used cognitive assessment to identify subjects who met criteria for clinically significant dementia (9.3% of sample) or for cognitive impairment without dementia (19.9%) (Bruce et al. 2009). While rates of severe hypoglycaemia (defined as requiring hospital treatment) were significantly higher in the two cognitively impaired groups (14 to 16% rate of hypoglycae- mia) as compared to cognitively normal subjects (3.8% rate), multiple logistic regression analyses showed that hypoglycaemia was unrelated to cognitive impairment (with or without dementia), whereas older age, longer duration of diabetes, presence of peripheral vascular disease and lower body mass index independently predicted poorer cognitive outcomes. By contrast, data from a longitudinal cohort study found a small but statistically reli- able relationship between number of episodes of severe hypoglycaemia and the risk of hospital-diagnosed dementia. Reviewing medical records of 16 667 older adults with type 2 diabetes, Whitmer and her associates (Whitmer et al. 2009) reported that those who had experienced at least one episode of hospital-diagnosed hypoglycaemia had an elevated increase in the risk of diagnosed dementia, and that risk rose linearly as the number of episodes of hypoglycaemia increased. This paper has engendered some controversy, with commentators expressing disbelief that a single episode of severe hypoglycaemia could increase the risk of dementia by 50% (Graveling and Frier 2009), particularly in the absence of other compelling data – from either human or animal studies – that would support such a strong positive relationship between hypoglycaemia and dementia (McNay 2005). Moreover, Whitmer’s analyses failed to include measures of subclinical cerebrov- ascular disease, as well as ascertainment of other variables (e.g. history of depression; alcohol/drug use; head injury; lower premorbid IQ scores) known to increase the risk of dementia. Taking a different approach, the Edinburgh Type 2 Diabetes Study Group assessed cogni- tive functioning in 1066 adults with type 2 diabetes, aged 60 to 75 years, and compared the performance of those 113 who had experienced one or more episodes of severe hypoglycae- mia (requiring the assistance of a third party) with those who had no such history (Aung et al. 2012). After adjustments for age and gender, a preceding history of severe hypoglycae- mia was associated with a general cognitive ability score that was 0.39 z score units lower. Analyses that stratified subjects based on the number of previous episodes of severe hypoglycaemia (1–2; 2–4; 5+) revealed a linear relationship, although there was much vari- ability (Figure 14.5), perhaps because the sample size for some groups was quite modest, and subjects experienced other age and/or diabetes-related conditions (e.g. cardiovascular 300 Hypoglycaemia in Clinical Diabetes

0.20

0.00 0 1–2 3–4 5 –0.20

–0.40

–0.60

–0.80

–0.10

–1.20

–1.40

Figure 14.5 Age- and gender-adjusted mean (95% CI) late-life general cognitive ability, by frequency of severe hypoglycaemia in the year preceding study entry. (Source: Aung PP 2012. Reproduced with permission from John Wiley & Sons, Ltd). disease) that could also affect cognitive test performance. This study has now followed subjects prospectively for 4 years and will be in a position to further validate these findings over time. Despite these intriguing findings, it is clear that more research is needed before one can conclude unequivocally that recurrent hypoglycaemia has a major impact on the development of neurocognitive dysfunction in older adults with type 2 diabetes.

CONCLUSIONS • At any age, individuals who experience profound hypoglycaemia, defined as blood glucose values less than 1.5 mmol/l, have a greatly increased risk of manifesting signifi- cant neurocognitive impairments that reflect damage to the brain. While the severity of these deficits may reduce over time, most people continue to show lasting impairments. • Adults with type 1 diabetes rarely show neurocognitive sequelae that can be attributed to one or more episodes of severe hypoglycaemia. When deficits are found, they are more likely to be associated with a long history of poor metabolic control and with the development of microvascular complications. • Children with type 1 diabetes often show cognitive deficits, which are particularly pronounced in those who were diagnosed with diabetes early in life. The extent to which these deficits may be triggered by recurrent, moderately severe hypoglycaemia remains unresolved. Although most studies find only weak support for that possibility, the very recent brain imaging results reported by Aye and colleagues (Aye et al. 2011) are very disconcerting and indicates that additional research is needed, with a particular focus on very young children with diabetes. • Older adults with type 2 diabetes often manifest cognitive impairments that are most strongly associated with the development of diabetes-related complications, including vascular disease. Few studies have formally assessed the role played by recurrent hypoglycaemia and so at this time, it is impossible to attribute neurocognitive outcomes to those hypoglycaemic events. Chapter 14: Long-Term Effects of Hypoglycaemia on Cognitive Function in Diabetes 301

REFERENCES

Alkalay, A.L., Flores-Sarnat, L., Sarnat, H.B., Moser, F.G. & Simmons, C.F. (2005) Brain imaging findings in neonatal hypoglycemia: case report and review of 23 cases. Clinical Pediatrics, 44, 783–790. Aoki, T., Sato, T., Hasegawa, K., Ishizaki, R. & Saiki, M. (2004) Reversible hyperintensity lesion on diffusion-weighted MRI in hypoglycemic coma. Neurology, 63, 392–393. Åsvold, B.O., Sand, T., Hestad, K. & Bjørgaas, M.R. (2010) Cognitive function in type 1 diabetic adults with early exposure to severe hypoglycemia: a 16-year follow-up study. Diabetes Care, 33, 1945– 1947. Åsvold, B.O., Sand, T., Hestad, K.A. & Bjorgaas, M.R. (2011) Quantitative EEG in type 1 diabetic adults with childhood exposure to severe hypoglycaemia: a 16 year follow-up study. Diabetologia, 54, 2404–2408. Auer, R.N. (2004) Hypoglycemic brain damage. Metabolic Brain Disease, 19, 169–175. Auer, R.N. & Siesjö, B.K. (1993) Hypoglycaemia: brain neurochemistry and neuropathology. Baillière’s Clinical Endocrinology and Metabolism, 7, 611–625. Auer, R.N., Wieloch, T., Olsson, Y. & Siesjo, B.K. (1984) The distribution of hypoglycemic brain damage. Acta Neuropathologica, 64, 177–191. Auer, R.N., Hugh, J., Cosgrove, E. & Curry, B. (1989) Neuropathologic findings in three cases of profound hypoglycemia. Clinical Neuropathology, 8, 63–68. Aung, P.P., Strachan, M.W., Frier, B.M., Butcher, I., Deary, I.J. & Price, J.F. (2012) Severe hypoglycaemia and late-life cognitive ability in older people with Type 2 diabetes: the Edinburgh Type 2 Diabetes Study. Diabetic Medicine, 29, 328–336. Aye, T., Reiss, A.L., Kesler, S. et al. (2011) The feasibility of detecting neuropsychologic and neuroana- tomic effects of type 1 diabetes in young children. Diabetes Care, 34, 1458–1462. Biessels, G.-J., Kamel, A., Ramakers, G.M. et al. (1996) Place learning and hippocampal synaptic plasticity in streptozotocin-induced diabetic rats. Diabetes, 45, 1259–1266. Biessels, G.J. & Gispen, W.H. (2005) The impact of diabetes on cognition: what can be learned from rodent models? Neurobiology of Aging, 26s, S36–S41. Bjørgaas, M., Sand, T. & Gimse, R. (1996) Quantitative EEG in type 1 diabetic children with and without episodes of severe hypoglycemia: a controlled, blind study. Acta Neurologica Scandinavica, 93, 398–402. Blasetti, A., Chiuri, R.M., Tocco, A.M. et al. (2011) The effect of recurrent severe hypoglycemia on cogni- tive performance in children with type 1 diabetes: a meta-analysis. Journal of Child Neurology. Boeve, B.F., Bell, D.G. & Noseworthy, J.H. (1995) Bilateral temporal lobe MRI changes in uncomplicated hypoglycemic coma. The Canadian Journal of Neurological Sciences, 22, 56–58. Brands, A.M.A., Biessels, G.J., de Haan, E.H.F., Kappelle, L.J. & Kessels, R.P.C. (2005) The effects of type 1 diabetes on cognitive performance: a meta-analysis. Diabetes Care, 28, 726–735. Brands, A.M.A., Kessels, R.P.C., Biessels, G.J. et al. (2006) Cognitive performance, psychological well- being, and brain magnetic resonance imaging in older patients with type 1 diabetes. Diabetes, 55, 1800–1806. Brands, A.M.A., Biessels, G.J., Kappelle, L.J. et al. (2007) Cognitive functioning and brain MRI in patients with type 1 and type 2 diabetes mellitus: a comparative study. Dementia and Geriatric Cognitive Dis- orders, 23, 343–350. Bree, A.J., Puente, E.C., Daphna-Iken, D. & Fisher, S.J. (2009) Diabetes increases brain damage caused by severe hypoglycemia. American Journal of Physiology. Endocrinology and Metabolism, 297, E194–E201. Brismar, T., Hyllienmark, L., Ekberg, K. & Johansson, B.-L. (2002) Loss of temporal lobe beta power in young adults with type 1 diabetes mellitus. Neuroreport, 13, 2469–2473. Bruce, D.G., Davis, W.A., Casey, G.P. et al. (2009) Severe hypoglycaemia and cognitive impairment in older patients with diabetes: the Fremantle Diabetes Study . Diabetologia, 52, 1808–1815. Burns, C.M., Rutherford, M.A., Boardman, J.P. & Cowan, F.M. (2008) Patterns of cerebral injury and neurodevelopmental outcomes after symptomatic neonatal hypoglycemia. Pediatrics, 122, 65–74. Caksen, H., Güven, A.S., Yilmaz, C. et al. (2011) Clinical outcome and magnetic resonance imaging find- ings in infants with hypoglycemia. Journal of Child Neurology, 26, 25–30. Caviness, V.S., Kennedy, D.N., Bates, J.F. & Makris, N. (1997) The developing human brain: a morpho- metric profile. In: R.W. Thatcher, G.R. Lyon, J. Rumsey & N. Krasnegor (eds),Developmental Neuroim- aging: Mapping the Development of Brain and Behavior, pp. 3–14. Academic Press, San Diego. 302 Hypoglycaemia in Clinical Diabetes

Chalmers, J., Risk, M.T.A., Kean, D.M., Grant, R., Ashworth, B. & Campbell, I.W. (1991) Severe amnesia after hypoglycemia: clinical, psychometric, and magnetic resonance correlations. Diabetes Care, 14, 922–925. Cox, D.J., Gonder-Frederick, L.A., Kovatchev, B.P., Julian, D.M. & Clarke, W.L. (2000) Progressive hypoglycemia’s impact on driving simulation performance. Diabetes Care, 23, 163–170. Deary, I.J., Langan, S.J., Graham, K.S., Hepburn, D.A. & Frier, B.M. (1992) Recurrent severe hypoglyc- emia, intelligence, and speed of information processing. Intelligence, 16, 337–359. Deary, I.J., Crawford, J.R., Hepburn, D.A., Langan, S.J., Blackmore, L.M. & Frier, B.M. (1993) Severe hypoglycemia and intelligence in adult patients with insulin-treated diabetes. Diabetes, 42, 341–344. Ferguson, S.C., Blane, A., Perros, P. et al. (2003) Cognitive ability and brain structure in type 1 diabetes: relation to microangiopathy and preceding severe hypoglycemia. Diabetes, 52, 149–156. Finelli, P.F. (2001) Diffusion-weighted MR in hypoglycemic coma. Neurology, 57, 933–935. Frier, B.M. (2011) Cognitive functioning in type 1 diabetes: the Diabetes Control and Complications Trial (DCCT) revisted. Diabetologia, 54, 233–236. Fujioka, M., Okuchi, K., Hiramatsu, K., Sakaki, T., Sakaguchi, S. & Ishii, Y. (1997) Specific changes in human brain after hypoglycemic injury. Stroke, 28, 584–587. Gaudieri, P.A., Chen, R., Greer, T.F. & Holmes, C.S. (2008) Cognitive function in children with type 1 diabetes: a meta-analysis. Diabetes Care, 31, 1892–1897. Graveling, A.J. & Frier, B.M. (2009) Dementia and hypoglycemic episodes in patients with type 2 diabetes. Journal of the American Medical Association, 302, 843–844 (letter). Hannonen, R., Komulainen, J., Riikonen, R. et al. (2012) Academic skills in children with early-onset type 1 diabetes: the effects of diabetes-related risk factors. Developmental Medicine & Child Neurology, 54, 457–463. Hawkins, B.T., Lundeen, T.F., Norwood, K.M., Brooks, H.L. & Egleton, R.D. (2007) Increased blood-brain barrier permeability and altered tight junctions in experimental diabetes in the rat: contribution of hyper- glycaemia and matrix metalloproteinases. Diabetologia, 50, 202–211. Ho, M.S., Weller, N.J., Ives, F.J. et al. (2008) High prevalence of structural CNS abnormalities in early onset type 1 diabetes. The Journal of Pediatrics, 153, 385–390. Howorka, K., Pumprla, J., Saletu, B., Anderer, P., Krieger, M. & Schabmann, A. (2000) Decrease of vigi- lance assessed by EEG-mapping in type 1 diabetic patients with history of recurrent severe hypoglycae- mia. Psychoneuroendocrinology, 25, 85–105. Hyllienmark, L., Maltez, J., Dandenell, A., Ludviggson, J. & Brismar, T. (2005) EEG abnormalities with and without relation to severe hypoglycaemia in adolescents with type 1 diabetes. Diabetologia, 48, 412–419. Inder, T. (2008) How low can I go? The impact of hypoglycemia on the immature brain. Pediatrics, 122, 440–441. Jongen, C., Van Der Grond, J., Kappelle, L.J. et al. (2007) Automated measurement of brain and white matter lesion volume in type 2 diabetes mellitus. Diabetologia, 50, 1509–1516. Jung, S.L., Kim, B.S., Lee, K.S., Yoon, K.H. & Byun, J.Y. (2005) Magnetic resonance imaging and diffusion- weighted imaging changes after hypoglycemic coma. Journal of Neuroimaging, 15, 193–196. Kramer, L., Fasching, P., Madl, C. et al. (1998) Previous episodes of hypoglycemic coma are not associated with permanent cognitive brain dysfunction in IDDM patients on intensive insulin treatment. Diabetes, 47, 1909–1914. Kurita, A., Katayama, K. & Mochio, S. (1996) Neurophysiological evidence for altered higher brain func- tions in NIDDM. Diabetes Care, 19, 361–364. Langan, S.J., Deary, I.J., Hepburn, D.A. & Frier, B.M. (1991) Cumulative cognitive impairment following recurrent severe hypoglycaemia in adult patients with insulin-treated diabetes mellitus. Diabetologia, 34, 337–344. Levine, B.-S., Anderson, B.J., Butler, D.A., Antisdel, J.E., Brackett, J. & Laffel, L.M.B. (2001) Predictors of glycemic control and short-term adverse outcomes in youth with type 1 diabetes. The Journal of Pediatrics, 139, 197–203. Lin, A., Northam, E.A., Rankins, D., Werther, G.A. & Cameron, F.J. (2010) Neuropsychological profiles of young people with type 1 diabetes 12 years after disease onset. Pediatric Diabetes, 11, 235–243. Lo, L., Tan, C.H.A., Umapathi, T. & Lim, C.C. (2006) Diffusion-weighted MR imaging in early diagnosis and prognosis of hypoglycemia. American Journal of Neuroradiology, 27, 1222–1224. Lunetta, M., Damanti, A.R., Fabbri, G., Lombardo, M., Di Mauro, M. & Mughini, L. (1994) Evidence by magnetic resonance imaging of cerebral alterations of atrophy type in young insulin-dependent diabetic patients. Journal of Endocrinological Investigation, 17, 241–245. Chapter 14: Long-Term Effects of Hypoglycaemia on Cognitive Function in Diabetes 303

Ly, T.T., Anderson, M., McNamara, K.A., Davis, E.A. & Jones, T.W. (2011) Neurocognitive outcomes in young adults with early-onset type 1 diabetes: a prospective follow-up study. Diabetes Care, 34, 2192–2197. Ma, J.-H., Kim, Y.-J., Yoo, W.-J. et al. (2009) MR imaging of hypoglycemic encephalopathy: lesion dis- tribution and prognosis prediction by diffusion-weighted imaging. Neuroradiology, 51, 641–649. Malone, J.I., Hanna, S., Saporta, S. et al. (2008) Hyperglycemia not hypoglycemia alters neuronal dendrites and impairs spatial memory. Pediatric Diabetes, 9, 531–539. McNay, E.C. (2005) The impact of recurrent hypoglycemia on cognitive function in aging. Neurobiology of Aging, 26s, S76–S79. Mooradian, A.D., Perryman, K., Fitten, J., Kavonian, G.D. & Morley, J.E. (1988) Cortical function in elderly non-insulin dependent diabetic patients. Behavioral and electrophysiologic studies. Archives of Internal Medicine, 148, 2369–2372. Mori, F., Nishie, M., Houszen, H., Yamaguchi, J. & Wakabayashi, K. (2006) Hypoglycemic encephalopathy with extensive lesions in the cerebral white matter. Neuropathology, 26, 147–152. Musen, G., Lyoo, I.K., Sparks, C.R. et al. (2006) Effects of type 1 diabetes on gray matter density as measured by voxel-based morphometry. Diabetes, 55, 326–333. Northam, E.A., Rankins, D., Lin, A. et al. (2009) Central nervous system function in youth with type 1 diabetes 12 years after disease onset. Diabetes Care, 32, 445–450. Parent, K.B., Wodrich, D.L. & Hasan, K.S. (2009) Type 1 diabetes mellitus and school: a comparison of patients and healthy siblings. Pediatric Diabetes, 10, 554–562. Patiño-Fernández, A.M., Delamater, A.M., Applegate, E.B. et al. (2010) Neurocognitive functioning in preschool-age children with type 1 diabetes mellitus. Pediatric Diabetes, 11, 424–430. Perantie, D.C., Wu, J., Koller, J.M. et al. (2007) Regional brain volume differences associated with hyper- glycemia and severe hypoglycemia in youth with type 1 diabetes. Diabetes Care, 30, 2331–2337. Perantie, D.C., Lim, A., Wu, J. et al. (2008) Effects of prior hypoglycemia and hyperglycemia on cognition in children with type 1 diabetes mellitus. Pediatric Diabetes, 9, 87–95. Perros, P., Deary, I.J., Sellar, R.J., Best, J.J.K. & Frier, B.M. (1997) Brain abnormalities demonstrated by magnetic resonance imaging in adult IDDM patients with and without a history of recurrent severe hypoglycemia. Diabetes Care, 20, 1013–1018. Reijmer, Y.D., van den Berg, E., Ruis, C., Jaap Kappelle, L. & Biessels, G.J. (2010) Cognitive dysfunction in patients with type 2 diabetes. Diabetes/Metabolism Research and Reviews, 26, 507–519. Ryan, C.M. (2006) Why is cognitive dysfunction associated with the development of diabetes early in life? The diathesis hypothesis. Pediatric Diabetes, 7, 289–297. Ryan, C.M. (2009) Cognition in children and adolescents with type 1 diabetes. In: G.J. Biessels & J.A. Luchsinger (eds), Diabetes and the Brain, pp. 251–275. Humana Press (Springer), New York. Ryan, C., Vega, A. & Drash, A. (1985) Cognitive deficits in adolescents who developed diabetes early in life. Pediatrics, 75, 921–927. Ryan, C.M., Atchison, J., Puczynski, S., Puczynski, M., Arslanian, S. & Becker, D. (1990) Mild hypogly- cemia associated with deterioration of mental efficiency in children with insulin-dependent diabetes mellitus. The Journal of Pediatrics, 117, 32–38. Ryan, C.M., Geckle, M.O. & Orchard, T.J. (2003) Cognitive efficiency declines over time in adults with type 1 diabetes: effects of micro- and macrovascular complications. Diabetologia, 46, 940–948. Schoenle, E.J., Schoenle, D., Molinari, L. & Largo, R.H. (2002) Impaired intellectual development in

children with Type 1 diabetes: association with HbA1c, age at diagnosis, and sex. Diabetologia, 45, 108–114. Shorr, R.I., Ray, W.A., Daugherty, J.R. & Griffin, M.R. (1997) Incidence and risk factors for serious hypoglycemia in older persons using insulin or sulfonylureas. Archives of Internal Medicine, 157, 1681–1686. Soltész, G. & Acsádi, G. (1989) Association between diabetes, severe hypoglycemia, and electroencepha- lographic abnormalities. Archives of Disease in Childhood, 64, 992–996. Strachan, M.W.J., Abraha, H.D., Sherwood, R.A. et al. (1999) Evaluation of serum markers of neuronal damage following severe hypoglycaemia in adults with insulin-treated diabetes mellitus. Diabetes/ Metabolism Research and Reviews, 15, 5–12. Strudwick, S.K., Carne, C., Gardiner, J., Foster, J.K., Davis, E.A. & Jones, T.W. (2005) Cognitive function- ing in children with early onset type 1 diabetes and severe hypoglycemia. The Journal of Pediatrics, 147, 680–685. Suh, S.W., Hamby, A.M. & Swanson, R.A. (2007) Hypoglycemia, brain energetics, and hypoglycemic neuronal death. Glia, 55, 1280–1286. 304 Hypoglycaemia in Clinical Diabetes

Taylor, H.G. & Alden, J. (1997) Age-related differences in outcomes following childhood brain insults: an introduction and overview. Journal of the International Neuropsychological Society, 3, 555–567. Terakawa, Y., Tsuyugughi, N., Nunomura, K. et al. (2007) Reversible diffusion-weighted imaging changes in the splenium of the corpus callosum and internal capsule associated with hypoglycemia. Neurologia Medico-Chirurgica (Tokyo), 47, 486–488. The Diabetes Control and Complications Trial Research Group (1993) The effect of intensive treatment on the development and progression of long-term complications in insulin-dependent diabetes mellitus. New England Journal of Medicine, 329, 977–986. The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Study Research Group (2007) Long-term effect of diabetes and its treatment on cognitive function. New England Journal of Medicine, 356, 1842–1852. Tiehuis, A.M., Mali, W.P.T.M., Van Raamt, A.F. et al. (2009) Cognitive dysfunction and its clinical and radiological determinants in patients with symptomatic arterial disease and diabetes. Journal of the Neurological Sciences, 283, 170–174. Wagner, V.M., Grabert, M. & Holl, R.W. (2005) Severe hypoglycaemia, metabolic control and diabetes management in children with type 1 diabetes in the decade after the Diabetes Control and Complications Trial – a large scale multicentre study. European Journal of Pediatrics, 164, 73–79. Whitmer, R.A., Karter, A.J., Yaffe, K., Quesenberry, C.P. & Selby, J.V. (2009) Hypoglycemic episodes and risk of dementia in older patients with type 2 diabetes mellitus. Journal of the American Medical Asso- ciation, 301, 1565–1572. Wysocki, T., Harris, M.A., Mauras, N. et al. (2003) Absence of adverse effects of severe hypoglycemia on cognitive function in school-aged children with diabetes over 18 months. Diabetes Care, 26, 1100– 1105. Zammitt, N.N. & Frier, B.M. (2005) Hypoglycemia in type 2 diabetes: pathophysiology, frequency, and effects of different treatment modalities. Diabetes Care, 28, 2948–2961. Zammitt, N.N., Warren, R.E., Deary, I.J. & Frier, B.M. (2008) Delayed recovery of cognitive function following hypoglycemia in adults with type 1 diabetes: effect of impaired awareness of hypoglycemia. Diabetes, 57, 732–736. 15 Neurological Sequelae of Hypoglycaemia Petros Perros1 and Marit R. Bjørgaas2 1 Royal Victoria Infirmary, Newcastle upon Tyne, UK 2 St. Olavs Hospital, Trondheim, Norway

INTRODUCTION

People with type 1 diabetes experience an average of two mild episodes of hypoglycaemia per week, and one-third experience one or more episodes of severe hypoglycaemia annu- ally, requiring the help of another person to recover (Pedersen-Bjergaard et al. 2004). Thus, hypoglycaemia is common despite substantial advances in insulin treatment during the last three decades. Strict glycaemic control, intensified insulin treatment, long diabetes duration and impaired awareness of hypoglycaemia are established risk factors for severe hypogly- caemia (see Chapter 4). The brain depends upon a constant supply of glucose from the bloodstream, and neu- roglycopenic symptoms are generated because of neuronal glucose deprivation when blood glucose falls below a critical level. Symptoms include difficulty concentrating, drowsiness and difficulties in speech and coordination. Prolonged severe hypoglycaemia may result in neuronal death (Auer et al. 1984a; Tkacs et al. 2005; Ennis et al. 2008) and long-term brain damage, which can be fatal (Patrick and Campbell 1990; Finelli 2001; Yoneda and Yamamoto 2005), although structural damage is relatively rare. The vast majority of people recover completely from a severe episode. However, it is possible that repeated exposure to severe hypoglycaemia may progressively damage brain structure and function in people with type 1 diabetes. The concept of ‘diabetic encephalopathy’, a disorder of the brain associated with diabetes, has been proposed. (Fazeli 2009; Sima 2010). Its causes are complex and may be related to several factors, only one of which is hypoglycaemia (Dej- gaard et al. 1991; McCall 1992; Biessels et al. 1994; Makimattila et al. 2004; Bauduceau et al. 2010; Sima 2010). In this chapter the neurological sequelae following severe hypogly- caemia are reviewed.

FUNCTIONAL EFFECTS OF HYPOGLYCAEMIA

Hypoglycaemia-induced neurological syndromes Hypoglycaemia causes a wide range of neurological symptoms and clinical signs, which can be mild or severe, reversible or permanent. These effects depend on several factors:

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 306 Hypoglycaemia in Clinical Diabetes

Box 15.1 Transient neuropsychological manifestations of severe hypoglycaemia Neurological • Focal or generalised convulsions • Hemiplegia • Focal neurological syndromes

Psycho-social • Mental slowness • Inappropriate behaviour • Confusional state • Aggressive behaviour

• the glucose nadir; • the duration of hypoglycaemia; • the frequency of episodes; • the age of the patient; • the presence of previous brain insults (e.g. head injury, chronic alcohol abuse).

Reversible effects of hypoglycaemia on the brain An acute fall in blood glucose causes cognitive impairment, which, if untreated, can proceed to loss of consciousness. Recovery is usually rapid (within 30–45 minutes) after the glucose levels return to normal, but some individuals complain of headache, malaise and memory problems for several hours. Although many aspects of intellectual perform- ance recover fully within a day, altered mood may take longer to recover (Strachan et al. 2000). In some persons, hypoglycaemia triggers transient focal neurological deficits (Box 15.1), including visual disturbances, hemiplegia or dysphasia (Foster and Hart 1987; Lahat et al. 1995; Pocecco and Ronfani 1998). Hypoglycaemia may also cause aggressive behav- iour, and in some medico-legal cases, automatism (a legal term) associated with hypogly- caemia has been cited as a defence against assault or even murder (Beaumont 2007). Diagnosis may be difficult if the presentation is atypical, in a post-ictal state, or if the blood glucose concentration has increased to normal or elevated levels following a counter- regulatory response.

Convulsions and associated morbidity The risk of focal or generalised convulsions precipitated by hypoglycaemia has been esti- mated to be 0.02 convulsions per patient per year (MacLeod et al. 1993). There is an obvious risk of injury during convulsions (Hepburn et al. 1989) from fracture-dislocation of joints, vertebral compression fractures (Figure 15.1) or a head injury. Seizures in people with diabetes are often assumed to be hypoglycaemic without a confirmatory blood glucose measurement, making it difficult to distinguish idiopathic epilepsy from hypoglycaemia-induced convulsions. The EEG is often unhelpful as epi­ leptiform activity may occur in acute hypoglycaemia without the development of convul- sions (Bjørgaas et al. 1998) (Figure 15.2), and EEG-abnormalities can persist for several days. EEG should therefore be deferred for at least a week and undertaken at a time when Chapter 15: Neurological Sequelae of Hypoglycaemia 307

Figure 15.1 Lateral X-ray of thoracic spine demonstrating a vertebral compression fracture sustained during a hypoglycaemia-induced convulsion in a young man with type 1 diabetes (Courtesy of Professor B. M. Frier). blood glucose is at, or just above, within the normal range. Treatment with anticonvulsant drugs is generally not recommended if the diagnosis of epilepsy is uncertain, but medica- tion may be beneficial for selected individuals with type 1 diabetes who have recurrent hypoglycaemic seizures and a lowered seizure threshold (Eeg-Olofsson 1977). A 4-fold increased risk of type 1 diabetes in young adults with idiopathic epilepsy has been reported (McCorry et al. 2006), although an increased risk of epilepsy in young people with type 1 diabetes has not been found (O’Connell et al. 2008). Cerebral oedema, a feared complication of severe insulin-induced hypoglycaemia, should be suspected if further deterioration or false localising signs ensue following treat- ment (MacCuish 1993). Hypoglycaemia-associated cerebral oedema is often resistant to treatment and frequently fatal. Urgent imaging of the brain is important in excluding other potentially remediable causes of neurological abnormalities or coma.

Permanent neurological effects of hypoglycaemia on the brain In rare cases, severe and protracted hypoglycaemia causes permanent brain damage, hypoglycaemic encephalopathy, resulting in death or a persistent vegetative state (Finelli 2001; Yoneda and Yamamoto 2005; Ma et al. 2009; Kang et al. 2010). This may be associ- ated with excessive consumption of alcohol and occasionally follows attempted suicide or 308 Hypoglycaemia in Clinical Diabetes

f3–c3

f4–c4

p3–o1

p4–o2

t3–t5

t4–t6

fpz–fz

t1–t2 200 µV 1 sec Basal Target four Target three Target two Post hypoglycaemia Blood glucose (mmol/l): 4.9 4.0 2.9 2.1 7.3 Subject 10 (MM)

Figure 15.2 EEG changes during hypoglycaemic clamp in a 13-year old child with type 1 diabetes displaying progressive slow activity as blood glucose declines. Epileptiform spikes occurred at blood glucose of 2.1 mmol/l but there were no convulsions. (Source: Bjørgaas M 1998. Reproduced with per- mission from John Wiley & Sons, Ltd). unintentional insulin overdose. Some individuals recover partially with focal neurological deficits such as hemiparesis, ataxia or severe memory loss (Malouf and Brust 1985; Lins and Adamson 1993; Sontineni et al. 2008) (Box 15.2). Some of the patients who require hospital admission for treatment of severe hypoglycaemia have a poor prognosis with an increased risk of dying within a few months of discharge (Hart and Frier 1998; Hsu et al. 2013).

FUNCTIONAL AND STRUCTURAL CHANGES IN THE CENTRAL NERVOUS SYSTEM

EEG and hypoglycaemia A loss of fast EEG-oscillations, notably the dominant alpha rhythm, is an established finding in dementia, and the occurrence of oscillations with faster frequencies (beta and gamma) is related to cognitive function (Brismar et al. 2002). Acute insulin-induced hypoglycaemia is accompanied by EEG-deceleration (Pramming et al. 1988; Tribl et al. 1996; Bjørgaas et al. 1998), reflecting cortical neuronal dysfunction. During controlled, moderate hypoglycaemia, the EEGs of children with diabetes were more disturbed than those of non-diabetic children (Bjørgaas et al. 1998), indicating increased susceptibility to metabolic disturbance in diabetic compared to non-diabetic subjects. Chapter 15: Neurological Sequelae of Hypoglycaemia 309

Box 15.2 Long-term neuropsychological manifestations of severe insulin-induced hypoglycaemia Neurological • Persistent vegetative state • Hemiparesis • Focal abnormalities (motor, sensory) • Brainstem syndrome • Ataxia; choreoathetosis • Epilepsy

Psychological • Cognitive impairment • Behavioural abnormalities • Chronic anxiety state; psychosis • Psycho-social problems

Early studies described pathological changes with an increased amount of slow EEG activity in about 50% of children with diabetes. After episodes of severe hypoglycaemia, almost three-quarters of children had EEG-slowing, and in some, epileptiform activity was described (Haumont et al. 1979; Halonen et al. 1983; Soltesz and Acsadi 1989). EEGs were generally interpreted visually and the investigator was seldom blinded as to the patients’ clinical status, which may have overestimated the degree of EEG pathology. More recent studies (employing quantitative EEG interpretation and a blinded investigator) have reported EEG-decelerations in association with severe hypoglycaemia but no persistence of epilep- tiform activity (Bjørgaas et al. 1996; Howorka et al. 2000; Hyllienmark et al. 2005). Patients with impaired awareness of hypoglycaemia and recurrent severe hypoglycaemia may have more pronounced EEG-deceleration during hypoglycaemia than those with normal awareness (Howorka et al. 1996). During euglycaemia, subjects with impaired awareness had persistent EEG-slowing, and reduced vigilance, resembling the type of changes observed in pathological ageing (Howorka et al. 2000). Reduced vigilance may reflect impaired hypoglycaemia perception leading to a vicious spiral of recurrent severe episodes and increasing loss of awareness. EEG-deceleration has also been demonstrated in association with poor glycaemic control (Hauser et al. 1995; Hyllienmark et al. 2005). Tupola and co-workers found that EEG abnormalities at the time of diagnosis of diabetes predicted an increased risk of convulsions associated with subsequent hypoglycaemia (Tupola et al. 1998) and interpreted these results as an increased cerebral vulnerability in some individuals with diabetes, perhaps related to genetic or perinatal factors. A recent follow-up of hypoglycaemia-associated EEG-slowing in childhood did not reveal major EEG pathology 16 years later (Åsvold et al. 2011), suggesting that EEG abnormalities normalise as the brain matures.

Evoked potentials Evoked potentials describe signals generated in the central nervous system in response to external stimuli. The event-related potential P300 is considered to reflect neuropsychologi- cal changes that are not detectable by psychometric tests. 310 Hypoglycaemia in Clinical Diabetes

Some people with type 1 diabetes have delayed P300 wave latencies (Pozzessere et al. 1991; Parisi and Uccioli 2001; Elia et al. 2005; Shehata and Eltayeb 2010), possibly reflect- ing impaired cognitive function. In most studies, the association between abnormal evoked potentials and hypoglycaemia has not been explored. However, one recent study in 97 children with diabetes duration of around 8 years (Wysocka-Mincewicz et al. 2007) dem- onstrated abnormal auditory wave latencies and peripheral nerve conduction in children with recurrent hypoglycaemia. The authors concluded that hypoglycaemic episodes could damage the peripheral as well as the central nervous system.

Cerebral blood flow Hypoglycaemia leads to a redistribution of regional cerebral blood flow (Tallroth et al. 1992; MacLeod et al. 1994; Kennan et al. 2005; Bie-Olsen et al. 2009; Teh et al. 2010), which could cause localised neuronal ischaemia, particularly if the cerebral macro- or microcirculation is already compromised. During acute hypoglycaemia in non-diabetic subjects, a relative increase in the blood flow to the frontal lobes, thalamus, globus pallidus, prefrontal and motor cortex has been demonstrated (Tallroth et al. 1992; Teves et al. 2004; Kennan et al. 2005) together with a relative decrease in blood flow in the hippocampus (Teves et al. 2004). This might reflect increased vulnerability of the hippocampus to injury. In individuals with a history of previous severe hypoglycaemia (MacLeod et al. 1994) and impaired hypoglycaemia awareness (MacLeod et al. 1996) this altered pattern in regional cerebral blood flow appears to be permanent (Figures 15.3 and 15.4). A permanent increase in regional cerebral blood flow to the frontal lobes may be an adaptive response to protect an area of the brain that is most vulnerable to acute hypoglycaemia. Such a susceptibility of the frontal areas has been suggested by other techniques, including EEG (Pramming et al. 1988) and tests of cognitive function (see Chapter 14).

Frontal

Calcarine

Figure 15.3 Schematic representation of regions of interest in the brain in a neuroanatomical template used for transaxial (horizontal) slices at the level of the basal ganglia, examining cerebral blood flow. By using single photon emission tomography (SPET) during acute hypoglycaemia, increased uptake of isotope in the frontal area indicated increased blood flow while in the calcarine area it was reduced compared to euglycaemia, thus demonstrating redistribution of regional blood flow (Source: MacLeod KM et al. 1994. Reproduced with kind permission from Springer Science+Business Media). Chapter 15: Neurological Sequelae of Hypoglycaemia 311

Cerebral cortex hippocampus

Basal ganglia anterior thalamus

Brain stem spinal cord

Figure 15.4 Diagram indicating the sensitivity of regions of the brain to acute neuroglycopenia. The cortex and hippocampus are most vulnerable and the brainstem and spinal cord are most resistant.

Magnetic resonance imaging (MRI) Acute effects of profound hypoglycaemia Recent MRI studies in individuals with hypoglycaemic encephalopathy have reported bilateral and often symmetrical lesions in grey or white matter, or both (Kim and Koh 2007; Ma et al. 2009; Kang et al. 2010; Johkura et al. 2012). The frontal, temporal and parietal regions of the cerebral cortex are often affected, as well as the hippocampus and basal ganglia. White matter lesions are frequently found in the internal capsule and periventricul­ar areas. Absent or focal lesions in the internal capsule alone, at the time of presentation with hypoglycaemic coma, predicted a better clinical outcome than cases with hemispheric white matter lesions (Johkura et al. 2012). Although these studies suggest that white matter is particularly sensitive to profound hypoglycaemia, the findings reflect previous neuro­ pathological observations indicating susceptibility to neuroglycopenia in a rostro-caudal direction. Thus the cerebral cortex, basal ganglia and hippocampus are most sensitive and the brainstem and spinal cord most resistant to hypoglycaemia (Auer et al. 1984a) (Figure 15.5).

Long-term effects after severe hypoglycaemia White matter A general reduction in white matter has been reported in young children with type 1 dia- betes who had early exposure to hypoglycaemia-induced seizures (Aye et al. 2011). This involves myelinated axons which relay messages between neuronal cells; lesions in white matter are correlated with poorer performance in selected neurocognitive tests (Kodl et al. 312 Hypoglycaemia in Clinical Diabetes

(a) (b)

(c)

Figure 15.5 Common asymptomatic neurological abnormalities observed with MRI in patients with type 1 diabetes. (a) Cortical atrophy; (b) ventricular dilatation; (c) leukoaraiosis.

2008; Aye et al. 2011). In adults and adolescents with type 1 diabetes, pathological changes have not been found in white matter after episodes of severe hypoglycaemia (Ferguson et al. 2003; Perantie et al. 2007; Kodl et al. 2008; Weinger et al. 2008).

Grey matter Children and young adults with type 1 diabetes who have a history of severe hypoglycaemia exhibit reduced grey matter in the cerebellum, temporo-occipital cortex and thalamus (Musen et al. 2006; Perantie et al. 2007; Northam et al. 2009). Some of these areas support memory and verbal abilities; a decline in these functions has also been reported after severe hypoglycaemia (Perantie et al. 2008; Åsvold et al. 2010). Children with early onset of hypoglycaemic seizures appear to have a general reduction in grey matter (Ho et al. 2008; Chapter 15: Neurological Sequelae of Hypoglycaemia 313

Aye et al. 2011), supporting the notion of increased vulnerability of the immature brain to hypoglycaemia (Ennis et al. 2008). The hippocampus appears to be particularly vulnerable (Auer et al. 1984a). MRI studies have identified hippocampal damage in children with diabetes who have a history of recur- rent severe hypoglycaemia (Hershey et al. 2010), and in those with early-onset diabetes (Ho et al. 2008). The hippocampus is important for spatial navigation and memory (Igloi et al. 2010); reduced spatial memory after recurrent hypoglycaemia in childhood (Hershey et al. 2005) may indicate hippocampal damage.

STRUCTURAL AND FUNCTIONAL CHANGES IN THE CENTRAL NERVOUS SYSTEM

Structural changes of the brain in diabetes Cerebrovascular disease is more extensive and occurs at an earlier age in diabetic patients compared with the non-diabetic population (McCall 1992; Mankovsky and Ziegler 2004). Microvascular disease may also affect the brain (Strachan 2011). This is relevant to the hypothesis that haemodynamic and haemorrheological changes induced by hypoglycaemia may precipitate ischaemia in tissues with established macro- and microvascular disease (see Chapter 13). Voxel-based morphometry now allows the density of white and grey matter to be studied in different brain regions. Young adults with diabetes have lower grey matter density in various brain regions. Other imaging techniques of the brain have yielded complementary information (Figure 15.6). Studies using CT and MRI scanning report a

100 Diabetic

80

Non-diabetic 60

40 % with brain atrophy

20

0 0–9 10–19 20–29 30–39 40–49 50–59 60–69 70–79 80–89 Age (years)

Figure 15.6 The prevalence of brain atrophy with increasing age as demonstrated by MRI. This is more common in people with diabetes (type 1 and type 2) at an earlier age. (Source: Araki Y 1994. Reproduced with kind permission from Springer Science+Business Media). 314 Hypoglycaemia in Clinical Diabetes

Box 15.3 Structural abnormalities of the brain associated with diabetes

Gross pathology Histological abnormalities*

Severe cerebral atheroma Meningeal fibrosis Pseudocalcinosis Cerebral infarction Lacunar strokes Diffuse degeneration of grey and white matter Increased endothelial basement membrane thickness Microaneurysms Abnormal imaging (Figure 15.5) Cortical atrophy Ventricular dilatation Leukoaraiosis

* Some changes were observed after death in patients who had coexisting uraemia and hypertension – the changes may not therefore be specific to diabetes.

Box 15.4 Structural abnormalities of the brain associated with profound hypoglycaemia Lethal hypoglycaemia Cortical necrosis Hippocampal necrosis

Survivors of severe hypoglycaemia with gross neurological deficit Cortical atrophy Hippocampal atrophy Ventricular dilatation Lesions in white matter (leukoencephalopathy)

Patients with severe recurrent hypoglycaemia and no neurological signs Cortical atrophy high prevalence of cerebral atrophy in people with diabetes (36–53% compared to 12% in age-matched non-diabetic controls), which occurs earlier in life than in non-diabetic indi- viduals (Figure 15.6) (Araki et al. 1994; Chen et al. 2011). Ventricular enlargement is also more frequent in patients with diabetes (Lunetta et al. 1994) (Box 15.3).

Structural changes associated with hypoglycaemia (Box 15.4) After fatal severe hypoglycaemia areas of cortical necrosis can be demonstrated in the brain at autopsy, particularly in the frontal lobes and hippocampus, with relative sparing of the hindbrain (Yoneda and Yamamoto, 2005; Finelli, 2001). Cortical and hippocampal atrophy and ventricular enlargement have been described in long-term survivors of severe hypoglycaemia (McCall 1992). These neurohistological features are non-specific and resemble those of anoxic brain damage. Human studies are further confounded by second- ary changes as a result of cardiorespiratory collapse (Patrick and Campbell 1990). In hypoglycaemic brain damage there is selective neuronal acidophilia with shrinkage of the cells which have a bright red cytoplasm (Figure 15.7). These cannot be differentiated from Chapter 15: Neurological Sequelae of Hypoglycaemia 315

Figure 15.7 Histopathological appearance of neurones in layer 2 of the parietal cortex destroyed by exposure to severe hypoglycaemia in a fatal case of a patient with type 1 diabetes, showing pronounced shrinkage of neurones which appeared acidophilic and were stained bright red (not demonstrable in black and white print). (Courtesy of Dr G. A. Lammie). ischaemic neurones, but the pattern of neuronal injury characterises hypoglycaemic damage with cells in specific layers of the cortex being destroyed. Some case reports have described abnormalities of brain structure detected by CT scan- ning or MRI, associated with focal neurological deficit following one or more episodes of severe hypoglycaemia. Marked global cerebral atrophy has been described in a young patient with type 1 diabetes within a few months of a severe episode of hypoglycaemia, which was associated with severe neurological deficit and cortical blindness (Gold and Marshall 1996). Following severe hypoglycaemia, lesions have been identified in the hippocampus in diabetic patients with severe amnesia (Chalmers et al. 1991; Boeve et al. 1995), and in the pons in patients with persistent ataxia and hemiparesis (Figure 15.8) (Perros et al. 1994; Sontineni et al. 2008). Using fluid-attenuated inversion recovery (FLAIR) sequences on MRI or diffusion weighted images (DWI), earlier structural abnor- malities could be seen in patients who had suffered severe hypoglycaemia (Finelli 2001). Some of these changes had disappeared after 14 days, coincident with an improvement in the patient’s clinical condition (Maekawa et al. 2006). The neuropathology of mild cognitive impairment (in the absence of abnormal neuro- logical signs) associated with recurrent severe hypoglycaemia is unknown, but may either be a milder form of structural neuronal damage similar to that described in lethal cases, or a functional (metabolic) defect. A study using brain MRI has reported a group of 11 diabetic patients with a history of severe recurrent hypoglycaemia who had a high prevalence of cortical atrophy (45%) compared to none in a matched diabetic control group (Perros et al. 1997). A subsequent larger study (Ferguson et al. 2003) found a strong association between leukoaraiosis and retinopathy, but not with hypoglycaemia, suggesting that leu- koaraiosis represents a microvascular complication of hyperglycaemia. This premise is supported by the results of a functional MRI study of people with type 1 diabetes with proliferative diabetic retinopathy (Wessels et al. 2006). 316 Hypoglycaemia in Clinical Diabetes

Figure 15.8 MRI scan showing an irregular area of high signal intensity in the left pons (arrow) in a patient with type 1 diabetes who suffered permanent ataxia and hemiparesis following a single episode of severe hypoglycaemia. (Source: Perros P 1994. Reproduced with permission of the American Diabetes Association).

Mechanisms of hypoglycaemia-induced brain Injury The principal mechanism by which hypoglycaemia leads to its acute neuropsychological manifestations is thought to be the direct effect of lack of glucose on neurones, causing energy failure. Cerebral glycogen stores (which are limited in amount) may be able to curtail some of the effects of hypoglycaemia, but the importance of this glucose source to protect the human brain is unknown (Gruetter et al. 2003; McCall 2004). It has been suggested that the harmful effect of severe hypoglycaemia depends on the stage of develop- ment of the neurones affected (Ennis et al. 2008). If hypoperfusion complicates hypogly- caemia, there is a focal loss in autoregulation, with grossly decreased blood blow to the frontal and parietal lobes and almost no decline in blood flow to the cerebellum and brain stem (Tallroth et al. 1992; Jarjour et al. 1995), and this may account for the increased vulnerability of the frontal and parietal cortical areas. The alterations in the cerebral circula- tion induced by hypoglycaemia may cause localised ischaemia, provoking focal neurologi- cal abnormalities such as hemiparesis (Sontineni et al. 2008). In animal models, activation of postsynaptic neurocytotoxin receptors by neurotransmit- ters (glutamate and N-acetyl aspartate) released from presynaptic neurones as a result of hypoglycaemia, appear to be an important cause of neuronal death (Cotman and Iversen 1987; Choi 1990; Auer et al. 1984b). Increased influx of calcium, which may be linked to stimulation of neurocytotoxin receptors, is also toxic and can cause cell death (Siesjö Chapter 15: Neurological Sequelae of Hypoglycaemia 317 et al. 1989). These mechanisms may explain the selective nature of hypoglycaemia-induced neuronal damage, which spares glial and vascular tissue in the brain.

Evidence for diabetic encephalopathy Considerable evidence indicates an association between neuropsychological dysfunction and diabetes. The nature of this association is unclear, but four main contributing factors have been identified:

• poor glycaemic control; • cerebrovascular disease; • hypoglycaemia; • the psychosocial impact of diabetes per se.

Hypoglycaemia is of particular importance because it is potentially avoidable, and the subtle cumulative effects on cognitive function may not be noticed until its severity com- promises the social and psychological functioning of those affected. The misplaced enthu- siasm by some health professionals (and patients) to pursue and implement strict glycaemic control when this is neither prudent nor appropriate (such as in individuals with impaired awareness of hypoglycaemia), may increase the risk of diabetic encephalopathy. In a clini- cal context, severe hypoglycaemia is encountered in three broad categories:

• individuals with type 1 diabetes who have strict glycaemic control with no or minimal microvascular complications; • people with long duration of type 1 diabetes, moderate or poor glycaemic control (often related to inadequate diabetes self-management, erratic lifestyle, inappropriate insulin dose or regimen, and co-existing social and psychological problems), associated with advanced microvascular complications; • individuals who experience a single major episode of hypoglycaemia because of a deliberate or accidental overdose of insulin or a sulfonylurea.

The accumulated evidence indicates that younger patients in the first category (resem- bling the highly selected population who participated in the DCCT) may not be suscep- tible to cumulative cognitive deterioration (Reichard and Pihl 1994; The Diabetes Control and Complications Trial Research Group 1996, 1997; Musen et al. 2008). In clinical practice a sizeable proportion of patients belongs to the second category. They have elevated glycated haemoglobin concentrations and established microvascular complica- tions. Hypoglycaemia may aggravate established micro- and macrovascular disease (Wright and Frier 2008) and thus potentiate the risk of damage to the brain. Evidence from retrospective studies suggests that chronic cognitive impairment may be a real risk should the DCCT approach be applied indiscriminately to these individuals (Deary and Frier 1996).

CONCLUSION • All patients with type 1 diabetes are potentially at risk of suffering long-term cerebral dysfunction as a consequence of hypoglycaemia. 318 Hypoglycaemia in Clinical Diabetes • The severity, duration and frequency of hypoglycaemia on the one hand, and the age and pre-existing macro- and microvascular disease on the other, determine outcome. • Recognition of hypoglycaemia as the underlying cause of subtle or unusual neuropsy- chiatric symptoms is important. • Therapeutic targets in patients treated with insulin must be tailored to the individual and should include an assessment of vulnerability of that person to hypoglycaemia.

REFERENCES

Araki, Y., Nomura, M., Tanaka, H., Yamamoto, T., Tsukaguchi, I. & Nakamura, H. (1994) MRI of the brain in diabetes mellitus. Neuroradiology, 36, 101–103. Åsvold, B.O., Sand, T., Hestad, K. & Bjørgaas, M.R. (2010) Cognitive function in type 1 diabetic adults with early exposure to severe hypoglycemia: a 16-year follow-up study. Diabetes Care, 33, 1945–1947. Åsvold, B.O., Sand, T., Hestad, K.A. & Bjorgaas, M.R. (2011) Quantitative EEG in type 1 diabetic adults with childhood exposure to severe hypoglycaemia: a 16 year follow-up study. Diabetologia, 54, 2404–2408. Auer, R.N., Olsson, Y. & Siesjo, B.K. (1984a) Hypoglycemic brain injury in the rat. Correlation of density of brain damage with the EEG isoelectric time: a quantitative study. Diabetes, 33, 1090–1098. Auer, R.N., Wieloch, T., Olsson, Y. & Siesjö, B.K. (1984b) The distribution of hypoglycemic brain damage. Acta Neuropathol., 64, 177–191. Aye, T., Reiss, A.L., Kesler, S. et al. (2011) The feasibility of detecting neuropsychologic and neuroana- tomic effects of type 1 diabetes in young children. Diabetes Care, 34, 1458–1462. Bauduceau, B., Doucet, J., Bordier, L., Garcia, C., Dupuy, O. & Mayaudon, H. (2010) Hypoglycaemia and dementia in diabetic patients. Diabetes and Metabolism, 36 (Suppl. 3), S106–S111. Beaumont, G. (2007) Automatism and hypoglycaemia. Journal of Forensic and Legal Medicine, 14, 103–107. Bie-Olsen, L.G., Kjaer, T.W., Pedersen-Bjergaard, U. et al. (2009) Changes of cognition and regional 15 cerebral activity during acute hypoglycemia in normal subjects: a H2 O positron emission tomographic study. Journal of Neuroscience Research, 87, 1922–1928. Biessels, G.J., Kappelle, A.C., Bravenboer, B., Erkelens, D.W. & Gispen, W.H. (1994) Cerebral function in diabetes mellitus. Diabetologia, 37, 643–650. Bjørgaas, M., Sand, T. & Gimse, R. (1996) Quantitative EEG in type 1 diabetic children with and without episodes of severe hypoglycemia: a controlled, blind study. Acta Neurologica Scandinavica, 93, 398–402. Bjørgaas, M., Sand, T., Vik, T. & Jorde, R. (1998) Quantitative EEG during controlled hypoglycaemia in diabetic and non-diabetic children. Diabetic Medicine, 15, 30–37. Boeve, B.F., Bell, D.G. & Noseworthy, J.H. (1995) Bilateral temporal lobe MRI changes in uncomplicated hypoglycemic coma. Can J Neurol Sci., 22, 56–58. Brismar, T., Hyllienmark, L., Ekberg, K. & Johansson, B.L. (2002) Loss of temporal lobe beta power in young adults with type 1 diabetes mellitus. Neuroreport, 13, 2469–2473. Chalmers, J., Risk, M.T., Kean, D.M., Grant, R., Ashworth, B. & Campbell, I.W. (1991) Severe amnesia after hypoglycemia. Clinical, psychometric, and magnetic resonance imaging correlations. Diabetes Care, 14, 922–925. Chen, Z., Li, L., Sun, J. & Ma, L. (2011) Mapping the brain in type II diabetes: voxel-based morphometry using DARTEL. European Journal of Radiology, 81, 1870–1876. Choi, D.W. (1990) Methods for antagonizing glutamate neurotoxicity. Cerbrovascular and Brain Metabo- lism Reviews, 2, 105–147. Cotman, C.W. & Iversen, L.L. (1987) Excitatory amino acids in the brain – focus on NMDA receptors. Trends in Neurosciences, 10, 263–265. Dejgaard, A., Gade, A., Larsson, H., Balle, V., Parving, A. & Parving, H.H. (1991) Evidence for diabetic encephalopathy. Diabetic Medicine, 8, 162–167. Deary, I.J. & Frier, B.M. (1996) Severe hypoglycaemia and cognitive impairment in diabetes. British Medical Journal, 313, 767–768. Eeg-Olofsson, O. (1977) Hypoglycemia and neurological disturbances in children with diabetes mellitus. Acta Paediatrica Scandinavica. Supplement, 270, 91–96. Chapter 15: Neurological Sequelae of Hypoglycaemia 319

Elia, Y.T., Daneman, D., Rovet, J. et al. (2005) Color visual evoked potentials in children with type 1 diabetes: relationship to metabolic control. Investigative Ophthalmology and Visual Science, 46, 4107–4113. Ennis, K., Tran, P.V., Seaquist, E.R. & Rao, R. (2008) Postnatal age influences hypoglycemia-induced neuronal injury in the rat brain. Brain Research, 1224, 119–126. Fazeli, S.A. (2009) Neuroprotection in diabetic encephalopathy. Neurodegenerative Diseases, 6, 213–218. Ferguson, S.C., Blane, A., Perros, P. et al. (2003) Cognitive ability and brain structure in type 1 diabetes: relation to microangiopathy and preceding severe hypoglycemia. Diabetes, 52, 149–156. Finelli, P.F. (2001) Diffusion-weighted MR in hypoglycemic coma. Neurology, 57, 933. Foster, J.W. & Hart, R.G. (1987) Hypoglycemic hemiplegia: two cases and a clinical review. Stroke, 18, 944–946. Gold, A.E. & Marshall, S.M. (1996) Cortical blindness and cerebral infarction associated with severe hypoglycemia. Diabetes Care, 19, 1001–1003. Gruetter, R., Adriany, G., Choi, I.Y., Henry, P.G., Lei, H. & Oz, G. (2003) Localized in vivo 13C NMR spectroscopy of the brain. Nuclear Magnetic Resonance Biomedicine, 16, 313–338. Halonen, H., Hiekkala, H., Huupponen, T. & Hakkinen, V.K. (1983) A follow-up EEG study in diabetic children. Annals of Clinical Research, 15, 167–172. Hart, S.P. & Frier, B.M. (1998) Causes, management and morbidity of acute hypoglycaemia in adults requiring hospital admission. Quarterly Journal of Medicine, 91, 505–510. Haumont, D., Dorchy, H. & Pelc, S. (1979) EEG abnormalities in diabetic children: influence of hypogly- cemia and vascular complications. Clinical Pediatrics, 18, 750–753. Hauser, E., Strohmayer, C., Seidl, R., Birnbacher, R., Lischka, A. & Schober, E. (1995) Quantitative EEG in young diabetics. Journal of Child Neurology, 10, 330–334. Hepburn, D.A., Steel, J.M. & Frier, B.M. (1989) Hypoglycemic convulsions cause serious musculoskeletal injuries in patients with IDDM. Diabetes Care, 12, 32–34. Hershey, T., Perantie, D.C., Warren, S.L., Zimmerman, E.C., Sadler, M. & White, N.H. (2005) Frequency and timing of severe hypoglycemia affects spatial memory in children with type 1 diabetes. Diabetes Care, 28, 2372–2377. Hershey, T., Perantie, D.C., Wu, J., Weaver, P.M., Black, K.J. & White, N.H. (2010) Hippocampal volumes in youth with type 1 diabetes. Diabetes, 59, 236–241. Ho, M.S., Weller, N.J., Ives, F.J. et al. (2008) Prevalence of structural central nervous system abnormalities in early-onset type 1 diabetes mellitus. Journal of Pediatrics, 153, 385–390. Howorka, K., Heger, G., Schabmann, A., Anderer, P., Tribl, G. & Zeitlhofer, J. (1996) Severe hypoglycae- mia unawareness is associated with an early decrease in vigilance during hypoglycaemia. Psychoneu- roendocrinology, 21, 295–312. Howorka, K., Pumprla, J., Saletu, B., Anderer, P., Krieger, M. & Schabmann, A. (2000) Decrease of vigi- lance assessed by EEG-mapping in type I diabetic patients with history of recurrent severe hypoglycae- mia. Psychoneuroendocrinology, 25, 85–105. Hsu, P.-F., Sung, S.-H., Cheng, H.-M. et al. (2013) Association of clinical symptomatic hypoglycemia with cardiovascular events and total mortality in type 2 diabetes. Diabetes Care, 36, 894–400. Hyllienmark, L., Maltez, J., Dandenell, A., Ludvigsson, J. & Brismar, T. (2005) EEG abnormalities with and without relation to severe hypoglycaemia in adolescents with type 1 diabetes. Diabetologia, 48, 412–419. Igloi, K., Doeller, C.F., Berthoz, A., Rondi-Reig, L. & Burgess, N. (2010) Lateralized human hippocampal activity predicts navigation based on sequence or place memory. Proceedings of the National Academy of Sciences of the United States of America, 107, 14466–14471. Jarjour, I.T., Ryan, C.M. & Becker, D.J. (1995) Regional cerebral blood flow during hypoglycaemia in children with IDDM. Diabetologia, 38, 1090–1095. Johkura, K., Nakae, Y., Kudo, Y., Yoshida, T.N. & Kuroiwa, Y. (2012) Early diffusion MR imaging findings and short-term outcome in comatose patients with hypoglycemia. American Journal of Neuroradiology, 33, 904–909. Kang, E.G., Jeon, S.J., Choi, S.S., Song, C.J. & Yu, I.K. (2010) Diffusion MR imaging of hypoglycemic encephalopathy. American Journal of Neuroradiology, 31, 559–564. Kennan, R.P., Takahashi, K., Pan, C., Shamoon, H. & Pan, J.W. (2005) Human cerebral blood flow and metabolism in acute insulin-induced hypoglycemia. Journal of Cerebral Blood Flow and Metabolism, 25, 527–534. Kim, J.H. & Koh, S.B. (2007) Extensive white matter injury in hypoglycemic coma. Neurology, 68, 1074. 320 Hypoglycaemia in Clinical Diabetes

Kodl, C.T., Franc, D.T., Rao, J.P. et al. (2008) Diffusion tensor imaging identifies deficits in white matter microstructure in subjects with type 1 diabetes that correlate with reduced neurocognitive function. Diabetes, 57, 3083–3089. Lahat, E., Barr, J. & Bistritzer, T. (1995) Focal epileptic episodes associated with hypoglycemia in children with diabetes. Clinical Neurology and Neurosurgery, 97, 314–316. Lins, P.E. & Adamson, U. (1993) Neurological manifestations of hypoglycaemia. In: B.M. Frier & M. Fisher (eds), Hypoglycaemia and Diabetes, Clinical and Physiological Aspects, pp. 347–354. Edward Arnold, London. Lunetta, M., Damanti, A.R., Fabbri, G., Lombardo, M., Di Mauro, M., & Mughini, L. (1994) Evidence by magnetic resonance imaging of cerebral alterations of atrophy type in young insulin-dependent diabetic patients. J Endocrinol Invest, 17, 241–245. Ma, J.H., Kim, Y.J., Yoo, W.J. et al. (2009) MR imaging of hypoglycemic encephalopathy: lesion distribu- tion and prognosis prediction by diffusion-weighted imaging. Neuroradiology, 51, 641–649. MacCuish, A.C. (1993) Treatment of hypoglycaemia. In: B.M. Frier & B.M. Fisher (eds), Hypoglycaemia and Diabetes: Clinical and Physiological Aspects, pp. 212–221. Edward Arnold, London. MacLeod, K.M., Hepburn, D.A. & Frier, B.M. (1993) Frequency and morbidity of severe hypoglycaemia in insulin-treated diabetic patients. Diabetic Medicine, 10, 238–245. MacLeod, K.M., Hepburn, D.A., Deary, I.J. et al. (1994) Regional cerebral blood flow in IDDM patients: effects of diabetes and of recurrent severe hypoglycaemia. Diabetologia, 37, 257–263. MacLeod, K.M., Gold, A.E., Ebmeier, K.P. et al. (1996) The effects of acute hypoglycemia on relative cerebral blood flow distribution in patients with type I (insulin-dependent) diabetes and impaired hypoglycemia awareness. Metabolism: Clinical and Experimental, 45, 974–980. Maekawa, S., Aibiki, M., Kikuchi, K., Kikuchi, S., Umakoshi, K. et al. (2006) Time related changes in reversible MRI findings after prolonged hypoglycemia. Clinical Neurology and Neurosurgery, 108, 511–513. Makimattila, S., Malmberg-Ceder, K., Hakkinen, A.M. et al. (2004) Brain metabolic alterations in patients with type 1 diabetes-hyperglycemia-induced injury. Journal of Cerebral Blood Flow and Metabolism, 24, 1393–1399. Malouf, R. & Brust, J.C. (1985) Hypoglycemia: causes, neurological manifestations, and outcome. Annals of Neurology, 17, 421–430. Mankovsky, B.N. & Ziegler, D. (2004) Stroke in patients with diabetes mellitus. Diabetes/Metabolism Research and Reviews, 20, 268–287. McCall, A.L. (1992) The impact of diabetes on the CNS. Diabetes, 41, 557–570. McCall, A.L. (2004) Cerebral glucose metabolism in diabetes mellitus. European Journal of Pharmacol- ogy, 490, 147–158. McCorry, D., Nicolson, A., Smith, D., Marson, A., Feltbower, R.G. & Chadwick, D.W. (2006) An asso­ ciation between type 1 diabetes and idiopathic generalized epilepsy. Annals of Neurology, 59, 204– 206. Musen, G., Lyoo, I.K., Sparks, C.R. et al. (2006) Effects of type 1 diabetes on gray matter density as measured by voxel-based morphometry. Diabetes, 55, 326–333. Musen, G., Jacobson, A.M., Ryan, C.M. et al. (2008) Impact of diabetes and its treatment on cognitive function among adolescents who participated in the Diabetes Control and Complications Trial. Diabetes Care, 31, 1933–1938. Northam, E.A., Rankins, D., Lin, A. et al. (2009) Central nervous system function in youth with type 1 diabetes 12 years after disease onset. Diabetes Care, 32, 445–450. O’Connell, M.A., Harvey, A.S., Mackay, M.T., Cameron, F.J. (2008) Does epilepsy occur more frequently in children with Type 1 diabetes? Journal of Paediatric Child Health, 44, 586–589. Parisi, V. & Uccioli, L. (2001) Visual electrophysiological responses in persons with type 1 diabetes. Diabetes/Metabolism Research and Reviews, 17, 12–18. Patrick, A.W. & Campbell, I.W. (1990) Fatal hypoglycaemia in insulin-treated diabetes mellitus: clinical features and neuropathological changes. Diabetic Medicine, 7, 349–354. Pedersen-Bjergaard, U., Pramming, S., Heller, S.R. et al. (2004) Severe hypoglycaemia in 1076 adult patients with type 1 diabetes: influence of risk markers and selection. Diabetes/Metabolism Research and Reviews, 20, 479–486. Perantie, D.C., Wu, J., Koller, J.M. et al. (2007) Regional brain volume differences associated with hyper- glycemia and severe hypoglycemia in youth with type 1 diabetes. Diabetes Care, 30, 2331–2337. Chapter 15: Neurological Sequelae of Hypoglycaemia 321

Perantie, D.C., Lim, A., Wu, J. et al. (2008) Effects of prior hypoglycemia and hyperglycemia on cognition in children with type 1 diabetes mellitus. Pediatric Diabetes, 9, 87–95. Perros, P., Sellar, R.J. & Frier, B.M. (1994) Chronic pontine dysfunction following insulin-induced hypoglycemia in an IDDM patient. Diabetes Care, 17, 725–727. Perros, P., Deary, I.J., Sellar, R.J., Best, J.J., Frier, B.M. (1997) Brain abnormalities demonstrated by magnetic resonance imaging in adult IDDM patients with and without a history of recurrent severe hypoglycemia. Diabetes Care, 20, 1013–1018. Pocecco, M. & Ronfani, L. (1998) Transient focal neurologic deficits associated with hypoglycaemia in children with insulin-dependent diabetes mellitus. Italian Collaborative Paediatric Diabetologic Group. Acta Paediatrica, 87, 542–544. Pozzessere, G., Valle, E., de Crignis, S. et al. (1991) Abnormalities of cognitive functions in IDDM revealed by P300 event-related potential analysis. Comparison with short-latency evoked potentials and psycho- metric tests. Diabetes, 40, 952–958. Pramming, S., Thorsteinsson, B., Stigsby, B. & Binder, C. (1988) Glycaemic threshold for changes in electroencephalograms during hypoglycaemia in patients with insulin dependent diabetes. British Medical Journal, 296, 665–667. Reichard, P., Pihl, M. (1994) Mortality and treatment side-effects during long-term intensified conventional insulin treatment in the Stockholm Diabetes Intervention Study. Diabetes, 43, 313–317. Shehata, G. & Eltayeb, A. (2010) Cognitive function and event-related potentials in children with type 1 diabetes mellitus. Journal of Child Neurology, 25, 469–474. Siesjö, B.K. & Bengtsson, F. (1989) Calcium fluxes, calcium antagonists, and calcium-related pathology in brain ischemia, hypoglycemia, and spreading depression: a unifying hypothesis. Journal of Cerebral Blood Flow and Metabolism, 9, 127–140. Sima, A.A. (2010) Encephalopathies: the emerging diabetic complications. Acta Diabetologica, 47, 279–293. Soltesz, G. & Acsadi, G. (1989) Association between diabetes, severe hypoglycaemia, and electroencepha- lographic abnormalities. Archives of Disease in Childhood, 64, 992–996. Sontineni, S.P., Lee, J.M. & Porter, J. (2008) Hypoglycemia-induced pontine infarction in a diabetic male with basilar artery stenosis: insight into the mechanisms of hypoglycemic stroke. Cerebrovascular Dis- eases, 25, 281–282. Strachan, M.W.J. (2011) The brain as a target organ in Type 2 diabetes: exploring the links with cognitive impairment and dementia. Diabetic Medicine, 28, 141–147. Strachan, M.W.J., Deary, I.J., Ewing, F.M. & Frier, B.M. (2000) Recovery of cognitive function and mood after severe hypoglycemia in adults with insulin-treated diabetes. Diabetes Care, 23, 305–312. Tallroth, G., Ryding, E. & Agardh, C.D. (1992) Regional cerebral blood flow in normal man during insulin- induced hypoglycemia and in the recovery period following glucose infusion. Metabolism: Clinical and Experimental, 41, 717–721. The Diabetes Control and Complications Trial Research Group (1996) Effects of intensive diabetes therapy on neuropsychological function in adults in the Diabetes Control and Complications Trial. Ann Intern Med, 124, 379–388. The Diabetes Control and Complications Trial Research Group (1997) Hypoglycemia in the Diabetes Control and Complications Trial. Diabetes, 46, 271–286. Teh, M.M., Dunn, J.T., Choudhary, P. et al. (2010) Evolution and resolution of human brain perfusion responses to the stress of induced hypoglycemia. Neuroimage, 53, 584–592. Teves, D., Videen, T.O., Cryer, P.E. & Powers, W.J. (2004) Activation of human medial prefrontal cortex during autonomic responses to hypoglycemia. Proceedings of the National Academy of Sciences of the United States of America, 101, 6217–6221. Tkacs, N.C., Pan, Y., Raghupathi, R., Dunn-Meynell, A.A. & Levin, B.E. (2005) Cortical Fluoro-Jade staining and blunted adrenomedullary response to hypoglycemia after noncoma hypoglycemia in rats. Journal of Cerebral Blood Flow and Metabolism, 25, 1645–1655. Tribl, G., Howorka, K., Heger, G., Anderer, P., Thoma, H. & Zeitlhofer, J. (1996) EEG topography during insulin-induced hypoglycemia in patients with insulin-dependent diabetes mellitus. European Neurology, 36, 303–309. Tupola, S., Saar, P. & Rajantie, J. (1998) Abnormal electroencephalogram at diagnosis of insulin-dependent diabetes mellitus may predict severe symptoms of hypoglycemia in children. Journal of Pediatrics, 133, 792–794. 322 Hypoglycaemia in Clinical Diabetes

Weinger, K., Jacobson, A.M., Musen, G. et al. (2008) The effects of type 1 diabetes on cerebral white matter. Diabetologia, 51, 417–425. Wessels, A.M., Rombouts, S.A., Simsek, S., Kuijer, J.P., Kostense, P.J., Barkhof, F., Scheltens, P., Snoek, F.J., Heine, R.J. (2006) Microvascular disease in type 1 diabetes alters brain activation: a functional magnetic resonance imaging study. Diabetes, 55, 334–340. Wright, R.J. & Frier, B.M. (2008) Vascular disease and diabetes: is hypoglycaemia an aggravating factor? Diabetes/Metabolism Research and Reviews, 24, 353–363. Wysocka-Mincewicz, M., Trippenbach-Dulska, H., Emeryk-Szajewska, B., Zakrzewska-Pniewska, B., Kochanek, K. & Pankowska, E. (2007) Impact of hypoglycemic episodes on nerve conduction and audi- tory and visual evoked potentials in children with type 1 diabetes. Pediatric Endocrinology Diabetes and Metabolism, 13, 17–22. Yoneda, Y. & Yamamoto, S. (2005) Cerebral cortical laminar necrosis on diffusion-weighted MRI in hypoglycaemic encephalopathy. Diabetic Medicine, 22, 1098–1100. 16 Psychological Effects of Hypoglycaemia Frank J. Snoek, Tibor R.S. Hajos and Stefanie M.P.A. Rondags VU University Medical Centre, Amsterdam, The Netherlands

INTRODUCTION

Diabetes is a psychologically demanding chronic disorder, having a profound impact on daily living and quality of life (Rubin and Peyrot 2001). In addition to the necessity of adapting to living with a chronic medical condition, which includes coping with physical discomfort, managing negative emotions and preserving social functioning, people with diabetes are faced with stresses that are unique to diabetes and its treatment demands (Snoek 2000). The stresses and strains of coping with diabetes revolve around the need to continuously self-regulate blood glucose levels, trying to preserve a balance between hyperglycaemia and hypoglycaemia. Despite the psychological demands of living with diabetes, most adults with type 1 and type 2 diabetes report having a satisfactory sense of well-being. However, approximately one-third suffer from depressive symptoms or experi- ence psychological distress (Snoek et al. 2011). Psychological distress can be viewed both as an antecedent of, and also a consequence of, the inherent difficulty of regulating blood glucose. For example, emotional distress may result in erratic blood glucose profiles that in turn can induce more stress and lead to a variety of coping responses, such as overeating or drinking too much alcohol. This is not to infer that poor glycaemic control is always the result of psychological factors, but that psychological and social factors should always be considered, irrespective of the cause of unsatisfactory glycaemic control. This is certainly true for hypoglycaemia, one of the recognised major barriers to achieving and maintaining good control of diabetes (Cryer 1999).

EXPERIENCE OF HYPOGLYCAEMIA

Hypoglycaemia can have a profound effect on the psychological well-being and social functioning of a person with diabetes, particularly when faced with recurrent episodes of hypoglycaemia that are perceived as ‘unpredictable’, that is, beyond personal control (Ritholz and Jacobson 1998). Over time, exposure to recurrent hypoglycaemic events has a negative impact on an individual’s self-confidence, fuelling anxiety and generating pessimistic beliefs about diabetes management (‘I will never get it right’; ‘my diabetes is

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 324 Hypoglycaemia in Clinical Diabetes beyond my control’) and ultimately a sense of hopelessness. This is true not only for the person with diabetes but often affects the partner or spouse and other family members who live with them. Hypoglycaemia is one of the risk factors for the development of ‘diabetes burnout’ (Polonsky 1996). Hypoglycaemia can exert physiological effects that directly disrupt a person’s cognitive, emotional and behavioural performance. Even mild hypoglycaemia can affect a person’s normal functioning or at least distract him or her from their work and other activities. Although the adverse effects of hypoglycaemia are acute, full recovery may take hours, with those affected reporting headaches, nausea and feelings of exhaustion for long after the restoration of blood glucose. A clear association exists between the actual blood glucose level and the magnitude of cognitive and emotional derangement: the lower the blood glucose the more disturbed the person’s performance. However, hypoglycaemia is to a considerable extent a subjective experience, whereby at identical levels of low blood glucose, people with diabetes experi- ence varying degrees of cognitive, emotional and behavioural dysfunction. It is important to appreciate that these inter-individual differences occur and that they are largely explained by three factors:

• symptom perception; • emotional appraisal; • social context.

Symptom perception and awareness Symptom perception reflects a person’s awareness and interpretation of their bodily and mental sensations, which is described as an interoception (Pennebaker 1982). Two impor- tant factors appear to play a role in the process of information processing: ‘selective search’ and ‘competition of cues’ that have relevance for determining awareness of hypoglycaemia (Cox et al. 1993) (see Chapter 6). Selective search refers to the phenomenon that what a person observes and recognises is driven to a large extent by their emotional state and what they (unconsciously) expect to observe, given the circumstances and time of day (Gonder-Frederick et al. 1985). If hypoglycaemia is not anticipated to occur, symptoms indicative of a falling blood glucose are less likely to be perceived and interpreted correctly. Pohl et al. (1997) demonstrated the effects of expectancy in a study in which healthy non-diabetic volunteers were injected with either saline or insulin, with half of them being misinformed as to which they were receiving. Persons who were expecting to develop hypoglycaemia reported more intense neuroglycopenic symptoms, such as weakness, drowsiness and headache. Competition of cues refers to the phenomenon that the human brain is able to process only a limited amount of incoming information at a given time, as a result of which physi- cal symptoms may escape a person’s attention because the individual is concentrating on another activity such as reading a book or watching a film. Once the symptoms are suffi- ciently severe to interfere with the act of reading or viewing, they become hard to ignore, but this ‘awareness’ may be delayed – in some instances developing too late to avoid the development of confusion and eventual loss of consciousness. Once a person experiences symptoms, the correct interpretation of these symptoms is the next step necessary to prompt an appropriate behavioural response (Kovatchev et al. 1998). While some people recognise that they are experiencing symptoms that suggest the onset of hypoglycaemia, they may Chapter 16: Psychological Effects of Hypoglycaemia 325 still misinterpret them as features of stress, anxiety or of a high blood glucose. Conversely, symptoms of high blood glucose may be falsely interpreted as indicating hypoglycaemia. Misinterpreting the cues of hypoglycaemia or delaying the decision to take action can be hazardous during activities such as operating machinery or driving a vehicle.

Mood changes Hypoglycaemia can induce rapid changes in mood (Chapter 2), characterised by a feeling of tense-tiredness, agitation, fatigue and reduction in a positive mood state (Merbis et al. 1996; Gold et al. 1997a; Hermanns et al. 2003). Low blood glucose values may present with transient emotional instability and disproportional emotional reactions. Exaggerated emotional responses may alert relatives or friends that the individual is experiencing hypoglycaemia. The mood changes that occur can influence the cooperation and willing- ness of the affected individual to accept help. It can be challenging to get them to accept a snack or glucose drink, as this may make them more agitated and reject help. Conflicts over hypoglycaemia and acceptance of help can seriously burden the spouses of patients and damage the marital relationship (Gonder-Frederick et al. 1997). The experience of hypoglycaemia can indirectly influence a person’s mood state and behaviour through the process of cognitive and emotional appraisal that determines an individual’s coping response (Leventhal et al. 1992). The appraisal of hypoglycaemia is highly idiosyncratic and is dependent on social context. For example, the impact of hypoglycaemia may be experienced as less aversive when it occurs at home than in the workplace. Likewise, nocturnal hypoglycaemia is likely to be a less stressful experience when living with a supportive partner than when living alone. Knowledge of a person’s social history, occupation and usual environment should help the clinician to better under- stand a patient’s perceptions of, and ability to cope with, hypoglycaemia.

Fear of hypoglycaemia The effects that fear of hypoglycaemia can have on family life have received limited atten- tion in scientific research. In one qualitative study, 20 people with type 1 diabetes described a wide range of difficulties in their family relationships as a consequence of hypoglycaemia (Barendse et al. 2012). These included perceived lack of understanding by others, conflicts about accepting help during a hypoglycaemic episode, difficulty in addressing concerns about hypoglycaemia and making sense of the behavioural impairments associated with hypoglycaemia: fear of dependency, guilt, loss of control and difficulties in talking about hypoglycaemia both with their family and their medical attendants. Increasing interest is being taken in the negative effect that fear of hypoglycaemia has on a person’s quality of life, loss of productivity and the associated economic cost, both in type 1 and type 2 diabetes (Jonsson et al. 2006; Brod et al. 2011; Fidler et al. 2011; Barendse et al. 2012).

SPECIFIC WORRIES ABOUT HYPOGLYCAEMIA

Worrying about hypoglycaemia is one of the most frequent problems to cause distress in people with insulin-treated diabetes, along with ‘worrying about complications and the future’ and ‘feeling guilty or anxious about getting off track with diabetes’ (Snoek et al. 326 Hypoglycaemia in Clinical Diabetes

Box 16.1 Psychological consequences of hypoglycaemia

Short-term Long-term

Anxiety Stress Transient cognitive dysfunction Chronic anxiety state Aversion Avoidance behaviour Depersonalisation Obsessive self-monitoring Loss of control Relationship conflicts Guilt, frustration Guilt, frustration Embarrassment Work/school problems Dependence on others Social isolation

2000). Some of the potential psychological consequences of hypoglycaemia are shown in Box 16.1. After experiencing a hypoglycaemic episode in the presence of others, the affected person may feel ashamed and humiliated and become anxious about the negative social reactions that this may have promoted, particularly if consciousness had been lost. In many instances, patients have no recollection of what happened during an episode of severe hypoglycaemia, and may worry whether they had behaved oddly or inappropriately and fear subsequent criticism. In the workplace, hypoglycaemia can be a major concern as it may cause poor performance and endanger others. Avoiding even mild hypoglycaemia is an important objective for many people during working hours, which is most often achieved by underdosing insulin (Brod et al. 2011, 2012) or maintaining suboptimal gly- caemic control (Leckie et al. 2005). When a person has to frequently correct his/her blood glucose to treat or avoid hypogly- caemia, weight gain may be a specific worry. Defensive eating to avoid hypoglycaemia may force patients to consume additional calories that promote weight gain, which is par- ticularly frustrating for those trying to lose weight or prevent weight gain. Moreover, when treating hypoglycaemia it can be difficult to consume the amount of carbohydrate that is sufficient to normalise blood glucose, while avoiding hyperglycaemia. Too often, overcom- pensation causes the blood glucose to rise excessively with the need to correct by taking more insulin. Severe hypoglycaemia is particularly feared because of the loss of control or deperson- alisation and dependency on others for assistance that it promotes (Cox et al. 1987; Polon- sky et al. 1992; Wild et al. 2007). Some of the reasons why people fear hypoglycaemia are listed in Box 16.2. It is commonly believed that hypoglycaemic coma can seriously damage the brain and can be fatal. Loss of control because of hypoglycaemia may also be associated with feelings of shame. Patients may behave oddly or inappropriately and may lose consciousness. Following severe hypoglycaemia, many patients have no memory of the event or only a vague recollection, and are dependent on others for reconstructing and recounting what actually happened. Concerns that hypoglycaemia will lead to long-term cognitive impairment can lead to a pervasive fear of hypoglycaemia. During pregnancy (Chapter 11), the concerns of women (and their partners) about severe or even mild hypoglycaemia are mainly centred on the potential damage that this may cause to the fetus. In about two-thirds of diabetic pregnancies at least one episode of severe hypoglycaemia occurs during the first 20 weeks (Ter Braak et al. 2002). The demands of striving to maintain strict glycaemic control while attempting to prevent hypoglycaemia Chapter 16: Psychological Effects of Hypoglycaemia 327

Box 16.2 Reasons why hypoglycaemia is feared Loss of control Personal embarrassment Unpleasant symptoms and effects on mood Risk of losing consciousness Risk of injury to self (and others) Potential dependence on others for help Risk of occurrence during sleep (without wakening) Warning symptoms may be inadequate or absent Potential impact on daily activities

can cause significant psychological distress in pregnant women with diabetes (Wouters and Snoek 2005).

RISK FACTORS FOR FEAR OF HYPOGLYCAEMIA

A past history of psychological symptoms has been linked to heightened fear of hypogly- caemia (Irvine et al. 1992), as has a greater self-reported severity of diabetic problems (Gold et al. 1997a). This suggests that a pre-existing tendency to worry contributes to the development of fear of hypoglycaemia. Anxiety or panic attacks were more prevalent in people with type 1 diabetes who reported experiencing more frequent hypoglycaemia (Costea et al. 1993). Trait anxiety (disposition to worry, nervousness) and a history of severe hypoglycaemia are the two most important determinants of fear of hypoglycaemia in adults with type 1 diabetes (Hepburn et al. 1994; Gold et al. 1997b; Wild et al. 2007). Less research has been performed in type 2 diabetes, but patients taking sulfonylureas report greater levels of fear of hypoglycaemia compared to those on metformin (Marrett et al. 2011), a difference likely to be explained by the higher hypoglycaemia risk associated with sulfonylureas (Zammitt and Frier 2005). Likewise, people with type 2 diabetes who have commenced treatment with insulin, or have been on insulin for some time, reported greatest fear of hypoglycaemia, although this is somewhat lower than that observed in people with type 1 diabetes (Beléndez and Hernández-Mijares 2009; Swinnen et al. 2010; Hajos et al. 2011, 2012; Pettersson et al. 2011). Insulin is a recognised source of worry among insulin-naive patients with type 2 diabetes, many of whom develop anticipatory anxiety about the perceived burden of daily injections and the side effects of insulin, par- ticularly weight gain and hypoglycaemia (Snoek et al. 2007). As many as 30% of people with type 2 diabetes are reluctant to commence insulin when this is recommended by their physician, a phenomenon referred to as ‘psychological insulin resistance’ (Peyrot et al. 2005; Kunt and Snoek 2009; Polonsky et al. 2011).

Fear of hypoglycaemia: adaptive or maladaptive? Worry about hypoglycaemia, which is common among people with insulin-treated diabetes, in itself constitutes a psychological burden. The key question is whether the patient’s fear of hypoglycaemia and related avoidance behaviours are appropriate, given the risks they have encountered. From a treatment perspective, patients always need to be aware of the 328 Hypoglycaemia in Clinical Diabetes

(real) risks of lowering blood glucose in order to cope effectively with hypoglycaemia, while avoiding excessive worrying and panic attacks (Polonsky et al. 1992; Brod et al. 2011, 2012). In people with insulin-treated diabetes, hypoglycaemia can trigger a cascade of emo- tional and behavioural responses that may appear to be exaggerated or ‘phobic’ to an outsider. Some individuals are inclined to respond to even a single episode of mild hypogly- caemia by adjusting their insulin dose (Leiter et al. 2005; Brod et al. 2011, 2012). A patient’s behavioural response is best understood as a function of the level of fear: the greater the anxiety, the more likely the patient is to exhibit avoidance behaviour. Some patients, however, show maladaptive responses to the risk of hypoglycaemia. For people with high trait anxiety and/or a history of unexpected severe hypoglycaemia, the mere thought of a possible fall in blood glucose can be sufficient for them to become apprehensive and compel them to check their blood glucose. Unless a patient is constantly being informed about their actual blood glucose level, by means of a real-time continuous glucose sensor, they have to rely on self-monitoring and maintaining alertness to symptoms (‘internal’ cues) that signal that blood glucose may be drifting dangerously high or low and correct this if necessary. However, symptoms can be misleading, particularly in the pres- ence of emotional stress. Patients who are stressed about developing hypoglycaemia often report difficulty in distinguishing between anxiety and the initial symptoms of hypogly- caemia (Polonsky et al. 1992), the negative effect of which can promote fear of hypogly- caemia. Indeed, emotional arousal can result in physical symptoms such as sweating and pounding heart that mimic hypoglycaemia, further reinforcing the patient’s anxiety and urge to eat or drink to raise their blood glucose even though this may be normal or high. This is also true for partners who may be overanxious and overprotective, and urge the patient to test or make an immediate correction by consuming carbohydrate. In contrast, some patients (and partners) show (too) little concern about lowering blood glucose or may underestimate the risk of hypoglycaemia and delay a corrective response, thereby increas- ing the risk of progression to severe hypoglycaemia. Whether worrying about hypoglycae- mia is an adaptive or maladaptive response is a question that can be answered only by assessing the actual risk for a particular individual. If the person has never actually expe- rienced severe hypoglycaemia but is extremely worried about doing so, the anxiety may qualify as a phobia. Conversely, if the patient reports having experienced multiple episodes of severe hypoglycaemia, but has not the slightest worry about hypoglycaemia, this may suggest ‘denial’. With respect to adaptation, the lower the frequency of severe hypoglycaemia, the lower the level of anxiety (Figure 16.1). It is important to note, however, that patients who actu- ally have a low risk (as defined by a low frequency of severe hypoglycaemia in preceding months) may attribute the absence of hypoglycaemia to their own cautious avoidance behaviour (e.g. frequent self-monitoring, defensive snacking, lowering the insulin dose, and avoiding intensive exercise). The challenge is to help the patient reduce anticipatory fear and the associated avoidance behaviour, by gradually building up confidence in the prevention and timely recognition of falling blood glucose values. Low fear of hypoglycaemia combined with high risk is obviously problematic and almost always a serious burden to partners and relatives who frequently have to resuscitate the patient from hypoglycaemia or seek emergency medical assistance. Little is known about the subgroup of ‘deniers’. Anecdotal evidence suggests a high prevalence of underly- ing psychiatric disorders among this group of predominantly male patients, including obsessive compulsive disorder, alcohol abuse and an ‘addiction’ towards hypoglycaemia. Chapter 16: Psychological Effects of Hypoglycaemia 329

High ‘Phobia’

Fear of hypo

‘Denial’

Low High Actual risk of hypolycaemia

Figure 16.1 The relationship between actual risk and fear of hypoglycaemia. (Source: Adapted from Irvine A 1994. Reproduced with permission from Taylor & Francis).

Both extremes of high and absent fear of hypoglycaemia are often maladaptive and warrant professional attention (Rogers et al. 2012).

PSYCHOLOGICAL ASSESSMENT AND MANAGEMENT

Recognition of the psychological burden of hypoglycaemia to a patient is important and should also acknowledge their relatives who have to struggle with the stress of coping with intermittent neuroglycopenia, impaired awareness of hypoglycaemia and unpredictable fluctuations in blood glucose affecting the patient. Clinicians should avoid criticism of patients for poor self-management, but instead encourage them to express their worries and beliefs within a non-judgemental atmosphere. To assess the patient’s level of fear of hypoglycaemia, clinicians can administer the Hypoglycaemia Fear Survey (HFS), a vali- dated self-report questionnaire that was originally developed for people with type 1 diabetes (Gonder-Frederick et al. 2011a). The HFS comprises two subscales, a Worry and a Behav- iour subscale, that in combination provide insight into a person’s emotional and behavioural responses to (the risk of) hypoglycaemia. The HFS has been translated into several lan- guages, and is freely available for clinical use. A paediatric version of the HFS has been developed (HFS-C) and validated as well as a parent-version (HFS-P) (Gonder-Frederick et al. 2011b). Extreme item scores can serve as ‘red flags’ along with high total HFS scores for fear and avoidance behaviours. No cut-off score has yet been determined for extremely high anxiety as measured by the HFS, but in practice a score of one standard deviation above the mean can be considered to be clinically relevant. The HFS may also be helpful to review changes in a patient’s hypoglycaemia fear and avoidance behaviours over time, for example after changing their insulin regimen or an educational intervention (Stargardt et al. 2009). To facilitate discussion with the patient and learn about his or her specific concerns, beliefs and behaviours, it may be helpful to examine their fear score alongside the number of episodes of severe hypoglycaemia (Figure 16.1). In this context, self-monitored blood glucose levels (or glucose values monitored by a continuous sensor) could be examined to identify how high and low risk situations relate to self-management behaviours. Where a patient has evidence of a phobia (unrealistic concerns and expressions of extreme fear of 330 Hypoglycaemia in Clinical Diabetes hypoglycaemia), referral to a psychologist or psychotherapist is indicated (Green et al. 2000). Cognitive behaviour therapy (CBT) can help patients to identify unrealistic worries, and learn to gradually reduce their fears, with subsequent lessening of passive and active avoidance behaviours (Clark and Beck 2012). Referral to a mental health specialist is not always well received by patients (‘I am not mad’). Self-management educational programmes are recommended for all patients with dia- betes, and may be particularly helpful in preventing severe hypoglycaemia and associated anxieties. In people with type 1 diabetes, the UK-based group programme, Dose Adjust- ment For Normal Eating (DAFNE), has been shown to be effective in improving glycaemic control and psychological outcomes, without increasing the risk of severe hypoglycaemia (Hopkins et al. 2012).

Blood Glucose Awareness Training (BGAT) Blood glucose awareness training (BGAT) is a well-researched psycho-educational group programme, based on symptom perception and behavioural therapy. The programme is aimed at preventing severe hypoglycaemia and reducing fear of hypoglycaemia by helping participants improve their symptom recognition and ability to anticipate low blood glucose. BGAT does help patients to prevent extreme blood glucose levels and avoid severe hypogly- caemia (Cox et al. 2006). The programme comprises 8-weekly interactive sessions deliv- ered over a 2-month period, and homework assignments. BGAT has been translated into different languages and adapted for use in various countries, including the Netherlands, Germany, Switzerland and Japan. The programme is delivered by a team comprising a diabetologist, (usually) a nurse specialist, and a psychologist. To extend availability and lower cost, a web-based version of BGAT (BGAT home) has been developed and pilot tested in the USA, with encouraging results (Cox et al. 2008).

SUMMARY

Clinicians need to be aware of the psychological burden of hypoglycaemia and its emo- tional and behavioural sequelae, not only in people who have type 1 and type 2 diabetes but also in their families and partners. In addition to modern glucose sensor technology and diabetes education, individual and group psycho-education and therapy can help people with diabetes to effectively reduce their fear of hypoglycaemia and improve their self- management behaviour. Diabetes care teams should make an effort to educate and assist people with insulin-treated diabetes to cope effectively with the threat of hypoglycaemia.

REFERENCES

Barendse, S., Singh, H., Frier, B.M. & Speight, J. (2012) The impact of hypoglycaemia on quality of life and related patient-reported outcomes in Type 2 diabetes: a narrative review. Diabetic Medicine, 29, 293–302. Beléndez, M. & Hernández-Mijares, A. (2009) Beliefs about insulin as a predictor of fear of hypoglycaemia. Chronic Illness, 5, 250–256. Brod, M., Christensen, T., Thomsen, T.L. & Bushnell, D.M. (2011) The impact of non-severe hypoglycemic events on work productivity and diabetes management. Value in Health, 14, 665–671. Chapter 16: Psychological Effects of Hypoglycaemia 331

Brod, M., Rana, A. & Barnett, A.H. (2012) Impact of self-treated hypoglycaemia in type 2 diabetes: a multinational survey in patients and physicians. Current Medical Research and Opinion, 28, 1–12. Clark, D.A.N.D. & Beck, A.T. (2012) The Anxiety and Worry Workbook: The Cognitive Behavioral Solu- tion. Guilford, New York. Costea, M., Ionescu-Tirgoviste, C., Cheta, D. & Mincu, I. (1993) Fear of hypoglycemia in type 1 (insulin- dependent) diabetic patients. Romanian Journal of Internal Medicine, 31, 291–295. Cox, D.J., Irvine, A., Gonder-Frederick, L., Nowacek, G. & Butterfield, J. (1987) Fear of hypoglycemia: quantification, validation, and utilization. Diabetes Care, 10, 617–621. Cox, D.J., Gonder-Frederick, L.A., Antoun, B., Cryer, P.E. & Clarke, W.L. (1993) Perceived symptoms in the recognition of hypoglycemia. Diabetes Care, 16, 519–527. Cox, D.J., Gonder-Frederick, L., Ritterband, L. et al. (2006) Blood Glucose Awareness Training: what is it, where is it, and where is it going? Diabetes Spectrum, 19, 43–49. Cox, D.J., Ritterband, L., Magee, J., Clarke, W. & Gonder-Frederick, L. (2008) Blood Glucose Awareness Training delivered over the internet. Diabetes Care, 31, 1527–1528. Cryer, P.E. (1999) Hypoglycemia is the limiting factor in the management of diabetes. Diabetes/Metabolism Research and Reviews, 15, 42–46. Fidler, C., Elmelund, C.T. & Gillard, S. (2011) Hypoglycemia: an overview of fear of hypoglycemia, quality-of-life , and impact on costs. Journal of Medical Economics, 14, 646–655. Gold, A.E., Deary, I.J. & Frier, B.M. (1997a) Hypoglycaemia and non-cognitive aspects of psychological function in insulin-dependent (type 1) diabetes mellitus (IDDM). Diabetic Medicine, 14, 111–118. Gold, A.E., Frier, B.M., MacLeod, K.M. & Deary, I.J. (1997b) A structural equation model for predictors of severe hypoglycaemia in patients with insulin-dependent diabetes mellitus. Diabetic Medicine, 14, 309–315. Gonder-Frederick, L.A., Cox, D.J., Bobbitt, S.A. & Pennebaker, J.W. (1985) Mood changes associated with blood glucose fluctuations in insulin-dependent diabetes mellitus. Health Psychology, 8, 45–59. Gonder-Frederick, L.A., Clarke, W.L., Cox, D.J. (1997) The emotional, social, and behavioral implications of insulin-induced hypoglycemia. Seminars in Clinical Neuropsychiatry, 2, 57–65. Gonder-Frederick, L.A., Nyer, M., Shepard, J.A., Vajda, K. & Clarke, W. (2011a) Assessing fear of hypoglycemia in children with Type 1 diabetes and their parents. Diabetes Management (London, England), 1, 627–639. Gonder-Frederick, L.A., Schmidt, K.M., Vajda, K.A. et al. (2011b) Psychometric properties of the hypogly- cemia fear survey-ii for adults with type 1 diabetes. Diabetes Care, 34, 801–806. Green, L., Feher, M. & Catalan, J. (2000) Fears and phobias in people with diabetes. Diabetes/Metabolism Research and Reviews, 16, 287–293. Hajos, T.R., Pouwer, F., de Grooth, R. et al. (2011) Initiation of insulin glargine in patients with Type 2 diabetes in suboptimal glycaemic control positively impacts health-related quality of life. A prospective cohort study in primary care. Diabetic Medicine, 28, 1096–1102. Hajos, T.R.S., Pouwer, F., de Grooth, R. et al. (2012) The longitudinal association between glycaemic control and health-related quality of life following insulin therapy optimisation in type 2 diabetes patients. A prospective observational study in secondary care. Quality of Life Research, 21, 1359– 1365. Hepburn, D.A., Deary, I.J., MacLeod, K.M. & Frier, B.M. (1994) Structural equation modeling of symp- toms, awareness and fear of hypoglycemia, and personality in patients with insulin-treated diabetes. Diabetes Care, 17, 1273–1280. Hermanns, N., Kubiak, T., Kulzer, B. & Haak, T (2003) Emotional changes during experimentally induced hypoglycaemia in type 1 diabetes. Biological Psychology, 63, 15–44. Hopkins, D., Lawrence, I., Mansell, P. et al. (2012) Improved biomedical and psychological outcomes 1 year after structured education in flexible insulin therapy for people with type 1 diabetes: the UK DAFNE experience. Diabetes Care, 35, 1638–1642. Irvine, A., Cox, D. & Gonder-Frederick, L. (1994) The fear of hypoglycemia scale. In: C. Bradley (ed.), Handbook of Psychology and Diabetes. A Guide to Psychological Measurement in Diabetes Research and Practice, pp. 133–155. Harwood Academic, Chur. Irvine, A.A., Cox, D. & Gonder-Frederick, L. (1992) Fear of hypoglycemia: relationship to physical and psychological symptoms in patients with insulin-dependent diabetes mellitus. Health Psychology, 11, 135–138. Jonsson, L., Bolinder, B. & Lundkvist, J. (2006) Cost of hypoglycemia in patients with Type 2 diabetes in Sweden. Value in Health, 9, 193–198. 332 Hypoglycaemia in Clinical Diabetes

Kovatchev, B., Cox, D., Schlundt, D. & Gonder-Frederick, L. (1998) Sochastic model of self-regulation decision making exemplified by decisions concerning hypopglycemia. Health Psychology, 17, 277–284. Kunt, T. & Snoek, F.J. (2009) Barriers to insulin initiation and intensification and how to overcome them. International Journal of Clinical Practice, 63 (Suppl. 164), 6–10. Leckie, A.M., Graham, M.K., Grant, J.B., Ritchie, P.J. & Frier, B.M. (2005) Frequency, severity, and morbidity of hypoglycemia occurring in the workplace in people with insulin-treated diabetes. Diabetes Care, 28, 1333–1338. Leiter, L.A., Yale, J.-F., Chiasson, J.-L., Harris, S.B., Kleinstiver, P. & Sauriol, L. (2005) Assessment of the impact of fear of hypoglycemic episodes on glycemic and hypoglycemia management. Canadian Journal of Diabetes, 29, 186–192. Leventhal, H., Diefenbach, M. & Leventhal, E.A. (1992) Illness cognition: using common sense to under- stand treatment adherence and affect cognition interactions. Cognitive Therapy and Research, 162, 143–163. Marrett, E., Radican, L., Davies, M.J. & Zhang, Q. (2011) Assessment of severity and frequency of self- reported hypoglycemia on quality of life in patients with type 2 diabetes treated with oral antihypergly- cemic agents: a survey study. BMC Research Notes, 4, 251. Merbis, M.A.E., Snoek, F.J., Kanc, K. & Heine, R.J. (1996) Hypoglycaemia induces emotional disruption. Patient Education and Counseling, 29, 117–122. Pennebaker, J.W. (1982) The Psychology of Physical Symptoms. Springer Verlag, New York. Pettersson, B., Rosenqvist, U., Deleskog, A., Journath, G. & Wändell, P. (2011) Self-reported experience of hypoglycemia among adults with type 2 diabetes mellitus. Diabetes Research and Clinical Practice, 92, 19–25. Peyrot, M., Rubin, R.R., Lauritzen, T. et al. (2005) Resistance to insulin therapy among patients and pro- viders: results of the cross-national Diabetes Attitudes, Wishes, and Needs (DAWN) study. Diabetes Care, 28, 2673–2679. Pohl, J., Frohnau, G., Kerner, W. & Fehm-Wolfsdorf, G. (1997) Symptom awareness is affected by the subjects′ expectations during insulin-induced hypoglycemia. Diabetes Care, 20, 796–802. Polonsky, W.H. (1996) Understanding and treating patients with diabetes burnout. In: B.J. Anderson & R.R. Rubin (eds), Practical Psychology for Diabetes Clinicians, pp. 183–192. American Diabetes Asso- ciation, Alexandria. Polonsky, W.H., Davis, C.L., Jacobson, A.M. & Anderson, B.J. (1992) Correlates of hypoglycemic fear in type I and type II diabetes mellitus. Health Psychology, 11, 199–202. Polonsky, W.H., Hajos, T.R., Dain, M.P. & Snoek, F.J. (2011) Are patients with type 2 diabetes reluctant to start insulin therapy? An examination of the scope and underpinnings of psychological insulin resistance in a large, international population. Current Medical Research and Opinion, 27, 1169–1174. Ritholz, M.D. & Jacobson, A.M. (1998) Living with hypoglycemia. Journal of General Internal Medicine, 13, 799–804. Rogers, H.A., de Zoysa, N. & Amiel, S.A. (2012) Patient experience of hypoglycaemia unawareness in Type 1 diabetes: are patients appropriately concerned? Diabetic Medicine, 29, 321–327. Rubin, R.R. & Peyrot, M. (2001) Psychological issues and treatments for people with diabetes. Journal of Clinical Psychology, 57, 457–478. Snoek, F.J. (2000) Barriers to good glycaemic control: the patient’s perspective. International Journal of Obesity and Related Metabolic Disorders, 24 (Suppl. 3), S12–S20. Snoek, F.J., Pouwer, F., Welch, G.W. & Polonsky, W.H. (2000) Diabetes-related emotional distress in Dutch and U.S. diabetic patients: cross-cultural validity of the problem areas in diabetes scale. Diabetes Care, 23, 1305–1309. Snoek, F.J., Skovlund, S.E. & Pouwer, F. (2007) Development and validation of the insulin treatment appraisal scale (ITAS) in patients with type 2 diabetes. Health and Quality of Life Outcomes, 5, 69. Snoek, F.J., Kersch, N.Y.A., Eldrup, E. et al. (2011) Monitoring of individual needs in Diabetes (MIND): baseline data from the cross-national Diabetes Attitudes Wishes and Needs (DAWN) MIND study. Diabetes Care, 34, 601–603. Stargardt, T., Gonder-Frederick, L., Krobot, K.J. & Alexander, C.M. (2009) Fear of hypoglycaemia: defin- ing a minimum clinically important difference in patients with type 2 diabetes. Health and Quality of Life Outcomes, 7, 91. doi: 10.1186/1477-7525-7-91 Chapter 16: Psychological Effects of Hypoglycaemia 333

Swinnen, S.G., Dain, M.P., Aronson, R. et al. (2010) A 24-week, randomized, treat-to-target trial comparing initiation of insulin glargine once-daily with insulin detemir twice-daily in patients with type 2 diabetes inadequately controlled on oral glucose-lowering drugs. Diabetes Care, 33, 1176–1178. ter Braak, E.W., Evers, I.M., Erkelens, D.W. & Visser, G.H. (2002) Maternal hypoglycemia during preg- nancy in type 1 diabetes: maternal and fetal consequences. Diabetes/Metabolism Research and Reviews, 18, 96–105. Wild, D., von Maltzhahn, R., Brohan, E., Christensen, T., Clauson, P. & Gonder-Frederick, L. (2007) A critical review of the literature on fear of hypoglycemia in diabetes: implications for diabetes manage- ment and patient education. Patient Education and Counselling, 68, 10–15. Wouters, M.G.A.J. & Snoek, F.J. (2005) Psychological issues in the management of diabetes and pregnancy. In: F.J. Snoek & T.C. Skinner (eds), Psychology in Diabetes Care, pp. 53–59. Wiley, Chichester. Zammitt, N.N. & Frier, B.M. (2005) Hypoglycemia in type 2 diabetes: pathophysiology, frequency, and effects of different treatment modalities. Diabetes Care, 28, 2948–2961. 17 Exercise Management and Hypoglycaemia in Type 1 Diabetes Ian W. Gallen1 and Alistair Lumb2 1 Royal Berkshire NHS Foundation Trust, Reading, UK 2 Buckinghamshire Healthcare Trust, Buckinghamshire, UK

INTRODUCTION

Many people with diabetes enjoy participation in sporting activities or simply like to take regular exercise. However, people who are treated with insulin are likely to experience marked fluctuations in blood glucose and experience frequent hypoglycaemia in relation to exercise. The occurrence of hypoglycaemia may appear to be both unpredictable and inexplicable to the person with diabetes, which may encourage them to respond by ingest- ing an excessive amount of carbohydrate before and following exercise, with resultant hyperglycaemia, so adding to the burden of maintaining glycaemic control. Perhaps of more concern to people with diabetes is the risk of hypoglycaemia occurring during exercise, and an increase in the risk of nocturnal hypoglycaemia. This concern may restrict the desire of people with diabetes (and for children, their career advisers and, sports coaches) to take part in exercise and discourage involvement or progression in any chosen sporting activity (Brazeau et al. 2008). In the last decade significant advances have been made in understanding the metabolism of exercise in diabetes, why hypoglycaemia can develop following exercise, and how diabetes should be managed in the context of exercise. In this chapter the mechanisms underlying the propensity for hypoglycaemia to occur with exercise are reviewed, and the effect that antecedent exercise has on the physiological responses to hypoglycaemia. Strategies are outlined to assist the clinician when advising a patient about safety in relation to different forms of exercise.

EFFECT OF EXERCISE ON GLYCAEMIC CONTROL

Exercise increases cardiovascular and respiratory activity to support the increased oxygen and fuel requirements of muscles. These cardiovascular and respiratory responses are the same in young adults with type 1 diabetes as in people without diabetes (Nugent et al. 1997). There is no reason why young people with insulin-treated diabetes cannot perform in sports to the same level as their non-diabetic peers. In the initial stages of exercise, muscles utilise glucose derived from intramuscular glycogen stores as their primary fuel

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. Chapter 17: Exercise Management and Hypoglycaemia in Type 1 Diabetes 335

(Koivisto et al. 1996), and the supply of glucose is then provided by hepatic glycogenolysis (Petersen et al. 2004). During prolonged exercise, other fuels become important sources of energy. The metabolic response to exercise is under endocrine control, and thus has the potential to be altered in diabetes. With prolonged exercise a marked reduction in insulin secretion occurs (Robertson et al. 1976; Broadstone et al. 1987), with increased secretion of glu­ cagon and catecholamines (Wasserman et al. 1920; Hirsch et al. 1991; Howlett et al. 1999; Kreisman et al. 2000a, 2000b, 2001, 2003; Marliss et al. 2000). These hormones cause lipolysis, permitting the early utilisation of free fatty acids as the source of fuel, and sub- sequent activation of gluconeogenesis provides glucose as the substrate for energy (Stich et al. 2000; Kreisman et al. 2001). Muscle glucose uptake continues to increase despite a reduction in insulin secretion, because exercise stimulates the translocation of GLUT-4 receptors to the cell surface in a similar manner to insulin-stimulated GLUT-4 translocation (Thorell et al. 1999; Fujii et al. 2000). To some extent, the endurance athlete has almost a similar endocrine state as the person with untreated type 1 diabetes but does not accumulate glucose and ketones because these are all oxidised. This suggests that the optimum man- agement of people with type 1 diabetes during exercise, to maximise performance and reduce the risk of hypoglycaemia, may be to replicate this physiological hypoinsulinaemic state during exercise. Some physical activities produce a variation in these physiological responses. Short duration and high intensity exercise, such as sprinting or mixed sports that involve short bursts of exercise (e.g. rugby or football), may produce a greater catecholamine response, which increases hepatic glucose production, but there may not be time to increase the production of ketones or free fatty acids (Kreisman et al. 2000a, 2000b). Endogenous insulin secretion is higher during, and following, these types of exercise than in endurance exercise, with this effect being more marked in women (Marliss et al. 2000). With all forms of exercise, once physical activity has ceased, the circulating levels of catecholamines rapidly decline and plasma insulin rapidly rises so that euglycaemia is maintained (Marliss et al. 2000). Exogenous insulin treatment for type 1 diabetes therefore has significant effects on the endocrine response to exercise, and elevated plasma insulin levels are inappropriate either during, or following, exercise.

Effects of type 1 diabetes and insulin therapy on exercise physiology In type 1 diabetes, injected insulin resides in subcutaneous depots, from where absorption continues or may actually increase during exercise (Koivisto and Felig 1978). The ambient level of plasma insulin during prolonged exercise is usually inappropriately high (derived from both the basal and bolus components) (Riddell et al. 2000; Petersen et al. 2004). Plasma glucagon concentrations do not change, with the result that the insulin : glucagon ratio presented to the liver is inappropriately high, so that hepatic glucose production may not be increased sufficiently during exercise to cope with the increased fuel requirement (Petersen et al. 2004). This inappropriately high plasma insulin will promote glucose uptake by non-exercising muscles and other tissues (Goodyear and Kahn 1998). The catecho- lamine and growth hormone responses to exercise are also deficient in type 1 diabetes (Ahlborg and Lundberg 1996; Jenni et al. 2010). During prolonged submaximal exercise, the relatively high ambient plasma insulin levels combined with a deficient counterregul­ atory hormonal response will provoke a marked fall in blood glucose. In the same way that 336 Hypoglycaemia in Clinical Diabetes electricity is not stored in electrical wires, glucose is not stored in circulating blood (or in extra-cellular fluid), these media being simply the means of transit. The free glucose that is available, even during hyperglycaemic states, is limited. For example, a lean person weighing 70 kg, who has a blood glucose of 15 mmol/l will have only approximately 8 G of immediately available free glucose. Given this limited availability along with the impaired hepatic glucose production, it is not unusual for blood glucose to fall from values in the mid-teens to the hypoglycaemic range within a period of 30 minutes. It is essential to consider the flux of glucose rather than the concentration of glucose in the blood. Con- versely, as intense exercise may stimulate a marked catecholamine response, which is not balanced by ambient insulin levels, it may allow blood glucose to rise rapidly and produce marked hyperglycaemia.

Not all exercise is the same: the effect of exercise intensity on counterregulatory hormonal responses in diabetes Intense exercise produces a greater counterregulatory hormone response than submaximal endurance exercise (Marliss et al. 1991, 2000). This effect is preserved but diminished in type 1 diabetes, and has important implications for glycaemic control (Marliss et al. 1991; Bussau et al. 2006, 2007; Guelfi et al. 2007). Intermittent high intensity exercise is associ- ated with a lesser fall in glucose than moderate exercise of a similar workload. When people with type 1 diabetes were exercised for 30 minutes by continuous submaximal exercise or with 4-second sprints performed every 2 minutes, the decline in blood glucose was less with the intense exercise. Glucose levels remained stable during recovery, but continued to fall after submaximal exercise alone. The mechanism for this stabilisation of blood glucose levels appears to be mediated through increased secretion of catecho- lamines and growth hormone during the early recovery period after exercise (Guelfi et al. 2005). Short periods of intense exercise performed before, or following, endurance exercise can also attenuate the decline in blood glucose that occurs following exercise. When a maximal 10-second cycling sprint was performed immediately following exercise of mod- erate intensity, the fall in blood glucose was less in the 2 hours following exercise compared to that observed following moderate intensity exercise without the brief sprint. A similar protective effect is observed if the 10-second sprint is performed before continuous exercise (Bussau et al. 2006, 2007). This protective effect of high intensity exercise seems to be lost in the late post-exercise period. Nocturnal hypoglycaemia is more frequent following intermittent high intensity exercise than after moderate exercise. When glucose profiles were examined using con- tinuous glucose monitoring (CGM) during, and in the 20 hours after, a 30-minute session of either intermittent high-intensity exercise or moderate-intensity exercise, glucose levels declined during both types of exercise. However, between midnight and 06:00 h the glucose levels were significantly lower after the period of intermittent exercise, and the number of hypoglycaemic episodes was greater than after lower intensity exercise (Maran et al. 2010). While differences exist between glycaemic responses in people with type 1 diabetes during, and following, different forms of exercise, differences in responses are also observed when aerobic and resistance exercise are performed in the same session but in different exercise order (Figure 17.1) (Yardley et al. 2012). When aerobic exercise preceded resist- ance exercise, blood glucose fell during aerobic exercise but not during resistance exercise. Chapter 17: Exercise Management and Hypoglycaemia in Type 1 Diabetes 337

Rest Exercise 1 Exercise 2 Recovery 12

11 ‡ ‡ ‡ ‡ 10 ‡ ‡ 9 *

8 * * *† 7 *†*† * **

Plasma glucose (mmol/L) 6 *† * *† = RA = AR 5 –15 0 60453015 9075 105 120 135 150 Time (minutes)

Figure 17.1 Plasma glucose (mean ± SE) during exercise and recovery for aerobic exercise performed before resistance exercise (AR, dashed line with ○) and resistance exercise performed before aerobic exercise (RA, solid line with ●) (n = 11). * Difference from baseline during exercise where P < 0.05. † Difference between conditions where P < 0.05. ‡ Change throughout recovery from end-exercise level where P < 0.05. (Source: Yardley JEP 2012. Reproduced with permission of the American Diabetes Association).

When resistance exercise first took place, blood glucose initially rose, but then fell during subsequent aerobic exercise, although following exercise it was similar to baseline values. CGM demonstrated a trend to less nocturnal hypoglycaemia when resistance training had taken place first. These data suggest that performing resistance exercise before aerobic exercise improves glycaemic stability throughout exercise and reduces the duration and severity of post-exercise hypoglycaemia for individuals with type 1 diabetes.

The interaction between hypoglycaemia and exercise Hypoglycaemia can be difficult to detect during exercise as many of the symptomatic manifestations of hypoglycaemia (such as sweating and pounding heart) are similar to those experienced by an exercising athlete. A series of elegant studies has further demonstrated the complexity of hormonal and metabolic interactions between exercise and antecedent hypoglycaemia. People with type 1 diabetes were studied during 90 minutes of euglycaemic exercise in the form of cycling on the day after exposure to two 2-hour periods of either euglycaemia or stepped hypoglycaemia of 3.9, 3.3 or 2.8 mmol/l. Progressive blunting of endogenous glucose production and of the responses of glucagon, catecholamines, cortisol and lipolysis to exercise was observed when the exercise had been preceded by antecedent hypoglycaemia. This was paralleled by a graduated increase in the amount of exogenous glucose that was required to maintain euglycaemia during exercise. This development of acute counterregulatory failure during prolonged, moderate-intensity exercise may be induced in a dose-dependent fashion by differing depths of antecedent hypoglycaemia, commencing at blood glucose of 3.9 mmol/l (Galassetti et al. 2006). A person with type 1 diabetes who has experienced an episode of hypoglycaemia on the day before the period of exercise will therefore need to ingest more glucose to avoid developing hypoglycaemia during subsequent exercise. Moderate intensity exercise also impairs the counterregulatory response to hypoglycae- mia when this is induced on either the same day or the following day (Sandoval et al. 2004, 338 Hypoglycaemia in Clinical Diabetes

2006). Epinephrine, pancreatic polypeptide, muscle sympathetic neural activity and endog- enous glucose production were all significantly lower during hypoglycaemia induced on the subsequent day, after 90 minutes of moderate submaximal exercise, in comparison with hypoglycaemia induced in control subjects with type 1 diabetes who had rested. Total hypoglycaemia symptom scores were also significantly lower following exercise (Sandoval et al. 2004). These changes in response to hypoglycaemia are also observed within a few hours of exercise in people with type 1 diabetes. Exercise again significantly blunts epine- phrine and muscle sympathetic neural activity responses; endogenous glucose production is lower following subsequent hypoglycaemia induced 2 hours after exercise when com- pared with rested subjects with type 1 diabetes. These data suggest that exercise-induced counterregulatory failure can occur very rapidly in type 1 diabetes, increasing the risk of hypoglycaemia developing within a few hours (Sandoval et al. 2006). A gender difference has been observed as these interactions are less prominent in women, in whom the counter- regulatory response to exercise following hypoglycaemia is relatively preserved (Galassetti et al. 2004). Hypoglycaemia is also more likely to occur and less likely to be recognised during sleep following a preceding period of exercise. A prospective case-finding study suggested that late-onset, post-exercise hypoglycaemia occurred in 48 of 300 type 1 diabetes patients who had developed diabetes before the age of 20 and who were monitored for up to 2 years (MacDonald 1987). Typically, hypoglycaemia occurred 6–15 h after the completion of unusually strenuous exercise and was nocturnal. In more than half the cases, the hypogly- caemia resulted in loss of consciousness or seizures and necessitated treatment with sub- cutaneous glucagon or intravenous glucose and/or emergency medical assistance. The hypoglycaemia was not limited to people with good or strict glycaemic control and often occurred after a single bout of exercise in individuals who were unaccustomed to physical activity or in athletic patients who were making the transition from an untrained to a trained state (MacDonald 1987). The mechanism for this nocturnal hypoglycaemia is multifacto- rial. As discussed earlier, the counterregulatory response to hypoglycaemia is attenuated by antecedent exercise, reducing endogenous glucose production, which is protective. While glucose uptake increases during exercise to provide fuel for exercising muscle, a second peak of glucose uptake occurs several hours after exercise, which probably results from replenishment of the body’s glycogen stores. Glucose infusion rates to maintain stable blood glucose levels are elevated during and shortly after exercise, and again from 7 to 11 hours after exercise (McMahon et al. 2007). This biphasic increment in glucose require- ment to maintain euglycaemia after exercise contributes to the risk of developing delayed nocturnal hypoglycaemia. The combination of increased nocturnal glucose disposal into skeletal muscle, along with an impaired counterregulatory response to hypoglycaemia following exercise, increases the risk of developing nocturnal hypoglycaemia. This risk is greater following alcohol ingestion (Cheyne et al. 2004; Richardson et al. 2005), and hypoglycaemia is less likely to be observed and treated in those who sleep alone. The physician advising a person with type 1 diabetes should explain the increased risk of the occurrence of hypoglycaemia during and following exercise, particularly in males, so that they can try to minimise or avoid late-onset hypoglycaemia. In the author’s own clinical practice, particular care is taken to advise young men who are going to college or university of this potential risk, and to recommend that they live in shared accommodation so that assistance is available should severe hypoglycaemia occur some time after exercise. Chapter 17: Exercise Management and Hypoglycaemia in Type 1 Diabetes 339

STRATEGIES TO MAINTAIN EUGLYCAEMIA DURING AND AFTER EXERCISE

Several strategies can be adopted to maintain euglycaemia during and following exercise (Table 17.1). The mainstay of management remains ingestion of carbohydrate before and during exercise. It is important not to over-replace, as this results in hyperglycaemia. Complex carbohydrates with a low glycaemic index may be successful at reducing hypogly- caemia during exercise without causing late post-exercise hyperglycaemia, but these carbohydrates are not currently widely available (West et al. 2011a). Simpler forms of carbohydrate, such as glucose in an accessible form, can be ingested regularly during exercise (Tamis-Jortberg et al. 1996; Ramires et al. 1997; Riddell et al. 2000; Mauvais- Jarvis et al. 2003; Francescato et al. 2004, 2008; West et al. 2011a). The ability to absorb glucose during exercise appears to be limited to as little as 60 G/h (Jeukendrup and Jentjens 2000). This implies that the provision of carbohydrate in excess of this amount will not help to avoid hypoglycaemia, but merely exacerbate post-exercise hyperglycaemia. As a consequence, ingestion of up to a dose of approximately 1 G/kg/h of exercise cannot be recommended during endurance exercise, such that a 70 kg adult may typically be ingesting 20 G of glucose every 20 minutes during exercise (Gallen et al. 2010). Where exercise is for less than one hour, simple carbohydrates should be consumed. When exercise is more prolonged, more palatable complex foods can be ingested.

Table 17.1 Strategies to prevent hypoglycaemia associated with exercise

Strategy Advantages Disadvantages

Reducing pre-exercise bolus Reduced hypoglycaemia Requires planning; not helpful for (preferably when exercise during exercise; reduces spontaneous or late postprandial is within 90–120 minutes CHO requirement; exercise; may result in of a bolus) beneficial for weight commencing exercise with management elevated blood glucose Adjusting pre-exercise and As above Requires planning as basal rate during exercise basal rate adjustments may need to be made (relevant to CSII) at least 60 minutes before the start of exercise Carbohydrate feeding Useful for unplanned or Counterproductive when purpose of during exercise prolonged exercise exercise is for weight control; not practical with all sports; may cause gastrointestinal discomfort Pre-exercise or post-exercise Reduces immediate Effect limited to shorter and less sprint post-exercise intense exercise, no effect on hypoglycaemia hypoglycaemia during exercise Reducing basal insulin Reduces nocturnal May cause raised fasting blood post-exercise (possible hypoglycaemia glucose with CSII and basal-bolus regimens) Taking caffeine before Less hypoglycaemia during Impairments or alterations of fine exercise and after exercise; motor control and technique, and reduced carbohydrate over-arousal (interfering with requirement recovery and sleep patterns)

CHO, carbohydrate; CSII, Continuous subcutaneous insulin infusion 340 Hypoglycaemia in Clinical Diabetes

People taking multiple daily insulin injections can adjust their pre-exercise bolus insulin dose and take extra carbohydrate during exercise to reduce the risk of hypoglycaemia (Hernandez et al. 2000; Rabasa-Lhoret et al. 2001; Mauvais-Jarvis et al. 2003). In a series of studies, people with type 1 diabetes performed graded exercise on a cycle ergometer following different doses of pre-exercise bolus of insulin. Male subjects with type 1 dia- betes, who were taking a basal-bolus insulin regimen, with ultralente insulin as the basal insulin and insulin lispro as the pre-meal insulin, were tested in a randomised, crossover fashion during postprandial exercise at 25% VO2max for 60 min, 50% VO2max for 30 and 60 min, and 75% VO2max for 30 min, commencing 90 min after a standardised mixed break- fast following either 100%, 50% or 25% of the usual insulin dose. The full (100%) pre-meal insulin dose was associated with an increased risk of hypoglycaemia. Lower pre-meal insulin doses were associated with a lower incidence of exercise-induced hypoglycaemia (Rabasa-Lhoret et al. 2001). An expected and predictable increase in the post-prandial glucose level occurred before exercise, so this approach is not beneficial to maintain gly- caemic control when exercise is taken in the late post-prandial period. These observations were confirmed in relation to running for 45 minutes (West et al. 2010). Furthermore, this pre-exercise reduction in insulin dose requires premeditated planning, and is not relevant to managing unplanned exercise, or may be ineffective where exercise is either more or less intense than expected (e.g. if a team game requires more or less exercise than usual).

Continuous subcutaneous insulin infusion Treatment with continuous subcutaneous insulin infusion (CSII) offers the capacity to modify the basal infusion rate by small increments and to obtain an effect relatively quickly. The subcutaneous reservoir of insulin is very small, and CSII provides an ability to give multiple additional bolus doses of insulin without needing to give additional injections. With an insulin pump it is possible to reduce or suspend insulin infusion during exercise. Young people with type 1 diabetes who suspend basal insulin infusion during exercise experience significantly less hypoglycaemia compared to those who continue with their usual basal rate of insulin infusion during exercise (Diabetes Research in Children Network (DirecNet) Study Group et al. 2006). However, this does result in more post-exercise hyperglycaemia. This is important as it increases the risk of provoking a deterioration in overall glycaemic control and, more importantly, of inducing ketosis. These results were confirmed in adults in a study when pre-meal insulin infusion rates were reduced by 50%, and suspended during 60 minutes of exercise (Sonnenberg et al. 1990). Very little data are currently available to indicate by how much the insulin infusion rates may have to be modified for different types of exercise, and the most appropriate time at whichthese changes in rate should be made. It is likely that either a large reduction in the rate of insulin infusion or complete suspension of CSII will be required at least 30 minutes before com- mencing prolonged endurance exercise. By contrast, short periods of high intensity exercise may not require much change in the insulin infusion rate. Following intense exercise, hyperglycaemia may be problematic. This can be controlled by giving a corrective bolus dose of fast-acting insulin, and with experience this necessary modification can be antici- pated. While CSII seems to offer a potential solution to many of the problems of insulin adjustment and exercise, full understanding of how to optimise its use remains to be determined. An alternative strategy is to combine reduction of the pre-exercise insulin dose with ingestion of carbohydrate that has a low glycaemic index. In one study, people with type Chapter 17: Exercise Management and Hypoglycaemia in Type 1 Diabetes 341

1 diabetes ingested 75 g isomaltulose before a period of running. They rested for 30, 60,

90 or 120 minutes before completing 45 minutes of running at 70% VO2max. No hypogly- caemia occurred when exercise took place after resting for 30 minutes, but hypoglycaemia was increasingly frequent when exercise was progressively delayed and was most marked after a delay of 120 minutes. Therefore, the combination of low glycaemic index carbohy- drate and reduced insulin dose, administered 30 minutes before running, has the potential to abolish the risk of hypoglycaemia associated with intense and prolonged exercise. This method lowers the risk of hypoglycaemia with exercise, and is associated with lesser excur- sions of pre-exercise glucose, but although helpful, it is subject to the same limitations as a pre-prandial reduction of insulin dose (West et al. 2009, 2011b). One concern is that reducing the insulin dose may incur a greater risk of ketosis. This was addressed in a study that examined the effects of up to a 75% reduction in the dose of rapid-acting insulin before exercise on the concentrations of blood beta-hydroxybutyrate following prolonged running in individuals with type 1 diabetes. Ketogenesis following running was not influenced by reducing the dose of pre-exercise rapid-acting insulin, sug- gesting that this important preparatory strategy helps to maintain blood glucose but does not pose a greater risk of inducing ketone body formation during exercise (Bracken et al. 2011). These observations also suggest that the excessive ketone production associated with a reduction in insulin dose is controlled by an appropriate physiological increase in ketone oxidation by skeletal muscles, which may improve performance and endurance in type 1 diabetes. The occurrence of delayed hypoglycaemia in the hours following exercise can be mini- mised by lowering the dose of post-exercise basal insulin (Hernandez et al. 2000; Gallen 2006; Taplin et al. 2010). For people treated with CSII with an insulin pump, post-exercise adjustment of the basal rate of insulin infusion is feasible. The tendency to develop post- exercise hypoglycaemia during the night can also be managed more effectively with CSII. From clinical experience, in people who take exercise less frequently than alternate days, nocturnal hypoglycaemia can be limited without compromising overall glycaemic control (including morning fasting glucose readings) by reducing the basal CSII rate during the night after the exercise. Following 60 minutes of exercise, the most effective rate adjust- ment appears to be a reduction in the basal insulin infusion rate of 25% during the post- exercise period. A greater reduction of 50% of the basal insulin infusion rate during this post-exercise period was associated with elevated blood glucose levels (Sonnenberg et al. 1990). For all forms of exercise, a reduction in infusion rate of 25% is likely to be required in the late post-exercise period and also overnight.

Continuous glucose monitoring The advent of commercially available CGM devices may seem to offer a significant advance in the management of exercise and type 1 diabetes. These devices measure sub- cutaneous interstitial glucose values following calibration with capillary glucose measure- ments (see Chapter 9). Blood glucose values change very rapidly during exercise, and unfortunately the correlation between the CGM measurements and capillary blood glucose is not reliable, so that CGM is not yet a useful tool for identifying significant changes in glucose, and particularly hypoglycaemia, in real time (Davison et al. 2010; Gallen et al. 2010). However, CGM may have a role in alerting the user as to the trajectory and speed of fall of the glucose response to exercise, and to identifying later nocturnal hypoglycaemia, and may be useful to an individual to identify the patterns of glucose change that occur 342 Hypoglycaemia in Clinical Diabetes during different forms of exercise (Cauza et al. 2005a, 2005b; Iscoe et al. 2006, 2008; Kapitza et al. 2010; Maran et al. 2010). In contrast to the difficulties experienced with the accuracy of CGM values during exercise, a further technological development offers another means of reducing the risk of hypoglycaemia following exercise. In one study, a sensor-augmented insulin pump system was equipped with an automatic suspension of insulin delivery with exercise-induced hypoglycaemia. The low glucose suspend (LGS) feature stops insulin delivery for 2 hours following a sensor glucose value of 3.9 mmol/l or less (≤70 mg/dl). People with type 1 diabetes exercised until their plasma glucose value reached ≤4.7 mmol/l (85 mg/dl). When exercise sessions were done with the LGS feature turned on, the mean nadir and end glucose values were higher than when the feature was switched off. Automatic suspension of insulin delivery significantly reduced the duration and severity of exercise-induced hypoglycaemia without causing rebound hyperglycaemia (Garg et al. 2012).

Pharmacological adjuncts Ingestion of sympathomimetic agents including caffeine (Debrah et al. 1996; Richardson et al. 2005) or a β2 adrenoreceptor agonist may potentially offer an additional treatment option for people who wish to exercise in the late post-prandial period (see Chapter 8), and who would like to reduce or avoid taking additional carbohydrate. Acute ingestion of caf- feine 30 minutes before prolonged endurance exercise reduced the duration of hypoglycae- mia following exercise in people with type 1 diabetes (Taplin et al. 2010; Gallen et al. 2011). Terbutaline may also protect against nocturnal hypoglycaemia developing following exercise in children (Taplin et al. 2010).

CONCLUSIONS • Managing diabetes in the context of exercise and sporting activities can be challenging but has become an increasingly important and necessary skill for clinicians who look after people with insulin-treated diabetes. • Careful elucidation of the type, timing, duration and intensity of exercise is required with a review of the intake of food and timing of insulin injections. • Particular care is essential to provide advice about the risk and avoidance of hypogly- caemia both before and after exercise. • The judicious use of CSII with an insulin pump to mimic physiological levels of plasma insulin through adjustments to insulin dose and delivery may provide a relatively rapid response to exercise. • With the support of a skilled clinician, there is no reason why people with type 1 diab­ etes should not be able to achieve high levels of athletic performance while maintaining good glycaemic control and avoiding hypoglycaemia (Gallen et al. 2003).

REFERENCES

Ahlborg, G. & Lundberg, J.M. (1996) Exercise-induced changes in neuropeptide Y, noradrenaline and endothelin-1 levels in young people with type I diabetes. Clinical Physiology, 16, 645–655. Chapter 17: Exercise Management and Hypoglycaemia in Type 1 Diabetes 343

Bracken, R.M., West, D.J., Stephens, J.W., Kilduff, L.P., Luzio, S. & Bain, S.C. (2011) Impact of pre- exercise rapid-acting insulin reductions on ketogenesis following running in Type 1 diabetes. Diabetic Medicine, 28, 218–222. Brazeau, A.-S., Rabasa-Lhoret, R., Strychar, I. & Mircescu, H. (2008) Barriers to physical activity among patients with type 1 diabetes. Diabetes Care, 31, 2108–2109. Broadstone, V.L., Pfiefer, M.A., Bajaj, V., Stagner, J.I. & Samols, E. (1987) Alpha-adrenergic blockade improves glucose-potentiated insulin secretion in non-insulin-dependent diabetes mellitus. Diabetes, 36, 932–937. Bussau, V.A., Ferreira, L.D., Jones, T.W. & Fournier, P.A. (2006) The 10-s maximal sprint: a novel approach to counter an exercise-mediated fall in glycemia in individuals with type 1 diabetes. Diabetes Care, 29, 601–606. Bussau, V.A., Ferreira, L.D., Jones, T.W. & Fournier, P.A. (2007) A 10-s sprint performed prior to moderate- intensity exercise prevents early post-exercise fall in glycaemia in individuals with type 1 diabetes. Diabetologia, 50, 1815–1818. Cauza, E., Hanusch-Enserer, U., Strasser, B., Kostner, K., Dunky, A. & Haber, P. (2005a) Strength and endurance training lead to different post exercise glucose profiles in diabetic participants using a con­ tinuous subcutaneous glucose monitoring system. European Journal of Clinical Investigation, 35, 745–751. Cauza, E., Hanusch-Enserer, U., Strasser, B. et al. (2005b) Continuous glucose monitoring in diabetic long distance runners. International Journal of Sports Medicine, 26, 774–780. Cheyne, E.H., Sherwin, R.S., Lunt, M.J., Cavan, D.A., Thomas, P.W. & Kerr, D. (2004) Influence of alcohol on cognitive performance during mild hypoglycaemia; implications for Type 1 diabetes. Diabetic Medi- cine, 21, 230–237. Davison, R., Aitken, G., Charlton, J., McKnight, J. & Kilbride, L. (2010) Comparison of patient blood glucose monitoring with continuous blood glucose monitoring during exercise. Diabetes Care, 59, 184–P [Abstract]. Debrah, K., Sherwin, R.S., Murphy, J. & Kerr, D. (1996) Effect of caffeine on recognition of and physi- ological responses to hypoglycaemia in insulin-dependent diabetes. Lancet, 347 (8993), 19–24. Francescato, M.P., Geat, M., Fusi, S., Stupar, G., Noacco, C. & Cattin, L. (2004) Carbohydrate requirement and insulin concentration during moderate exercise in type 1 diabetic patients. Metabolism: Clinical & Experimental, 53, 1126–1130. Francescato, M.P., Zanier, M. & Gaggioli, F. (2008) Prediction of glucose oxidation rate during exercise. International Journal of Sports Medicine, 29, 706–712. Fujii, N., Hayashi, T., Hirshman, M.F. et al. (2000) Exercise induces isoform-specific increase in ′5 AMP- activated protein kinase activity in human skeletal muscle. Biochemical and Biophysical Research Communications, 273, 1150–1155. Galassetti, P., Tate, D., Neill, R.A., Morrey, S., Wasserman, D.H. & Davis, S.N. (2004) Effect of sex on counterregulatory responses to exercise after antecedent hypoglycemia in type 1 diabetes. American Journal of Physiology. Endocrinology & Metabolism, 287, E16–E24. Galassetti, P., Tate, D., Neill, R.A., Richardson, A., Leu, S.-Y. & Davis, S.N. (2006) Effect of differing antecedent hypoglycemia on counterregulatory responses to exercise in type 1 diabetes. American Journal of Physiology. Endocrinology and Metabolism, 290, E1109–E1117. Gallen, I. (2006) Diabetes and sport: managing the complex interactions. British Journal of Hospital Medicine, 67, 512–515. Gallen, I.W., Redgrave, A. & Redgrave, S. (2003) Olympic diabetes. Clinical Medicine, 3, 333–337. Gallen, I.W., Hume, C. & Lumb, A. (2010) Fuelling the athlete with type 1 diabetes. Diabetes, Obesity and Metabolism, 13, 130–136. Gallen, I.W., Ballav, C., Lumb, A. & Carr, J. (2011) Caffeine supplementation reduces exercise induced decline in blood glucose and subsequent hypoglycaemia in adults with Type 1 Diabetes (T1DM) treated with Multiple Daily Insulin Injection (MDI). Diabetes, 59 (S1), A318 (abstract). Garg, S., Brazg, R.L., Bailey, T.S. et al. (2012) Reduction in duration of hypoglycemia by automatic suspen- sion of insulin delivery: the in-clinic ASPIRE study. Diabetes Technology & Therapeutics, 14, 205–209. Goodyear, L.J. & Kahn, B.B. (1998) Exercise, glucose transport, and insulin sensitivity. Annual Review of Medicine, 49, 235–261. Guelfi, K.J.B., Jones, T.W. & Fournier, P.A. (2005) The decline in blood glucose levels is less with inter- mittent high-intensity compared with moderate exercise in individuals with Type 1 diabetes. Diabetes Care, 28, 1289–1294. 344 Hypoglycaemia in Clinical Diabetes

Guelfi, K.J., Ratnam, N., Smythe, G.A., Jones, T.W. & Fournier, P.A. (2007) Effect of intermittent high- intensity compared with continuous moderate exercise on glucose production and utilization in individu- als with type 1 diabetes. American Journal of Physiology. Endocrinology & Metabolism, 292, E865–E870. Hernandez, J.M., Moccia, T., Fluckey, J.D., Ulbrecht, J.S. & Farrell, P.A. (2000) Fluid snacks to help persons with type 1 diabetes avoid late onset postexercise hypoglycemia. Medicine & Science in Sports & Exercise, 32, 904–910. Hirsch, I.B., Marker, J.C., Smith, L.J. et al. (1991) Insulin and glucagon in prevention of hypoglycemia during exercise in humans. American Journal of Physiology. Endocrinology and Metabolism, 260, E695–E704. Howlett, K., Febbraio, M. & Hargreaves, M. (1999) Glucose production during strenuous exercise in humans: role of epinephrine. American Journal of Physiology. Endocrinology and Metabolism, 276, E1130–E1135. Iscoe, K.E., Campbell, J.E., Jamnik, V., Perkins, B.A. & Riddell, M.C. (2006) Efficacy of continuous real-time blood glucose monitoring during and after prolonged high-intensity cycling exercise: spinning with a continuous glucose monitoring system. Diabetes Technology & Therapeutics, 8, 627–635. Iscoe, K.E., Corcoran, M. & Riddell, M.C. (2008) High rates of nocturnal hypoglycemia in a unique sports camp for athletes with type 1 diabetes: lessons learned from continuous glucose monitoring systems. Canadian Journal of Diabetes, 32, 182–189. Jenni, S., Christ, E.R. & Stettler, C. (2010) Exercise-induced growth hormone response in euglycaemia and hyperglycaemia in patients with Type 1 diabetes mellitus. Diabetic Medicine, 27, 230–233. Jeukendrup, A.E. & Jentjens, R. (2000) Oxidation of carbohydrate feedings during prolonged exercise: current thoughts, guidelines and directions for future research. Sports Medicine, 29, 407–424. Kapitza, C., Hovelmann, U., Nosek, L., Kurth, H.J., Essenpreis, M. & Heinemann, L. (2010) Continuous glucose monitoring during exercise in patients with type 1 diabetes on continuous subcutaneous insulin infusion. Journal of Diabetes Science & Technology, 4, 123–131. Koivisto, V.A. & Felig, P. (1978) Effects of leg exercise on insulin absorption in diabetic patients. New England Journal of Medicine, 298, 79–83. Koivisto, V.A., Stevens, I.K., Mattock, M. et al. (1996) Cardiovascular disease and its risk factors in IDDM in Europe. Diabetes Care, 19, 689–697. Kreisman, S.H., Ah, M.N., Arsenault, M. et al. (2000a) Epinephrine infusion during moderate intensity exercise increases glucose production and uptake. American Journal of Physiology. Endocrinology and Metabolism, 278, E949–E957. Kreisman, S.H., Manzon, A., Nessim, S.J. et al. (2000b) Glucoregulatory responses to intense exercise performed in the postprandial state. American Journal of Physiology. Endocrinology and Metabolism, 278, E786–E793. Kreisman, S.H., Ah, M.N., Halter, J.B., Vranic, M. & Marliss, E.B. (2001) Norepinephrine infusion during moderate-intensity exercise increases glucose production and uptake. Journal of Clinical Endocrinology and Metabolism, 86, 2118–2124. Kreisman, S.H., Halter, J.B., Vranic, M. & Marliss, E.B. (2003) Combined infusion of epinephrine and norepinephrine during moderate exercise reproduces the glucoregulatory response of intense exercise. Diabetes, 52, 1347–1354. MacDonald, M.J. (1987) Postexercise late-onset hypoglycemia in insulin-dependent diabetic patients. Diabetes Care, 10, 584–588. Maran, A., Pavan, P., Bonsembiante, B. et al. (2010) Continuous glucose monitoring reveals delayed nocturnal hypoglycemia after intermittent high-intensity exercise in nontrained patients with type 1 diabetes. Diabetes Technology & Therapeutics, 12, 763–768. Marliss, E.B., Simantirakis, E., Miles, P.D.G. et al. (1991) Glucoregulatory and hormonal responses to repeated bouts of intense exercise in normal male subjects. Journal of Applied Physiology, 71, 924–933. Marliss, E.B., Kreisman, S.H., Manzon, A., Halter, J.B., Vranic, M. & Nessim, S.J. (2000) Gender differences in glucoregulatory responses to intense exercise. Journal of Applied Physiology, 88, 457–466. Chapter 17: Exercise Management and Hypoglycaemia in Type 1 Diabetes 345

Mauvais-Jarvis, F., Sobngwi, E., Porcher, R. et al. (2003) Glucose response to intense aerobic exercise in type 1 diabetes: maintenance of near euglycemia despite a drastic decrease in insulin dose. Diabetes Care, 26, 1316–1317 [letter]. McMahon, S.K., Ferreira, L.D., Ratnam, N. et al. (2007) Glucose requirements to maintain euglycemia after moderate-intensity afternoon exercise in adolescents with type 1 diabetes are increased in a biphasic manner. Journal of Clinical Endocrinology & Metabolism, 92, 963–968. Nugent, A-M., Steele, I.C., Al-Modaris, F. et al. (1997) Exercise responses in patients with IDDM. Diabetes Care, 20, 1814–1821. Petersen, K.F., Price, T.B. & Bergeron, R. (2004) Regulation of net hepatic glycogenolysis and gluconeo- genesis during exercise: impact of type 1 diabetes. Journal of Clinical Endocrinology & Metabolism, 89, 4656–4664. Rabasa-Lhoret, R., Bourque, J., Ducros, F. & Chiasson, J-L. (2001) Guidelines for premeal insulin dose reduction for postprandial exercise of different intensities and durations in type 1 diabetic subjects treated intensively with a basal-bolus insulin regimen (ultralente-lispro). Diabetes Care, 24, 625–630. Ramires, P.R., Forjaz, C.L., Strunz, C.M. et al. (1997) Oral glucose ingestion increases endurance capacity in normal and diabetic (type I) humans. Journal of Applied Physiology, 83, 608–614. Richardson, T., Weiss, M., Thomas, P. & Kerr, D. (2005) Day after the night before: influence of evening alcohol on risk of hypoglycemia in patients with type 1 diabetes. Diabetes Care, 28, 1801–1802. Riddell, M.C., Bar-Or, O., Hollidge-Horvat, M., Schwarcz, H.P. & Heigenhauser, G.J. (2000) Glucose ingestion and substrate utilization during exercise in boys with IDDM. Journal of Applied Physiology, 88, 1239–1246. Robertson, R.P., Halter, J.B. & Porte, D., Jr. (1976) A role for alpha-adrenergic receptors in abnormal insulin secretion in diabetes mellitus. Journal of Clinical Investigation, 57, 791–795. Sandoval, D.A., Guy, D.L., Richardson, M.A., Ertl, A.C. & Davis, S.N. (2004) Effects of low and moderate antecedent exercise on counterregulatory responses to subsequent hypoglycemia in type 1 diabetes. Diabetes, 53, 1798–1806. Sandoval, D.A., Guy, D.L., Richardson, M.A., Ertl, A.C. & Davis, S.N. (2006) Acute, same-day effects of antecedent exercise on counterregulatory responses to subsequent hypoglycemia in type 1 diabetes mel- litus. American Journal of Physiology. Endocrinology & Metabolism, 290, E1331–E1338. Sonnenberg, G.E., Kemmer, F.W. & Berger, M. (1990) Exercise in type 1 (insulin-dependent) diabetic patients treated with continuous subcutaneous insulin infusion. Prevention of exercise induced hypogly- caemia. Diabetologia, 33, 696–703. Stich, V., de Glisezinski, I., Berlan, M. et al. (2000) Adipose tissue lipolysis is increased during a repeated bout of aerobic exercise. Journal of Applied Physiology, 88, 1277–1283. Tamis-Jortberg, B., Downs, D.A., Jr. & Colten, M.E. (1996) Effects of a glucose polymer sports drink on blood glucose, insulin, and performance in subjects with diabetes. Diabetes Educator, 22, 471–487. Taplin, C.E., Cobry, E., Messer, L., McFann, K., Chase, H.P. & Fiallo-Scharer, R. (2010) Preventing post-exercise nocturnal hypoglycemia in children with type 1 diabetes. Journal of Pediatrics, 157, 784–788. The Diabetes Research in Children Network (DirecNet) Study Group, et al. (2006) Prevention of hypogly- cemia during exercise in children with type 1 diabetes by suspending basal insulin. Diabetes Care, 29, 2200–2204. Thorell, A., Hirshman, M.F., Nygren, J. et al. (1999) Exercise and insulin cause GLUT-4 translocation in human skeletal muscle. American Journal of Physiology. Endocrinology and Metabolism, 277, E733–E741. Wasserman, D.H., Spalding, J.A., Lacy, D.B., Colburn, C.A., Goldstein, R.E. & Cherrington, A.D. (1989) Glucagon is a primary controller of hepatic glycogenolysis and gluconeogenesis during muscular work. American Journal of Physiology. Endocrinology and Metabolism, 257, E108–E117. West, D., Morton, R., Stephens, J.W., Bain, S.C. & Bracken, R.M. (2009) Effects of rapid-acting insulin reductions on post-exercise glycaemia and carbohydrate intake in people with Type 1 diabetes. Diabetic Medicine, 26, 60 [Abstract]. West, D.J., Morton, R.D., Bain, S.C., Stephens, J.W. & Bracken, R.M. (2010) Blood glucose responses to reductions in pre-exercise rapid-acting insulin for 24 h after running in individuals with type 1 diabetes. Journal of Sports Sciences, 28, 781–788. 346 Hypoglycaemia in Clinical Diabetes

West, D.J., Morton, R.D., Stephens, J.W. et al. (2011a) Isomaltulose improves postexercise glycemia by reducing CHO oxidation in T1DM. Medicine & Science in Sports & Exercise, 43, 204–210. West, D.J., Stephens, J.W., Bain, S.C. et al. (2011b) A combined insulin reduction and carbohydrate feeding strategy 30 min before running best preserves blood glucose concentration after exercise through improved fuel oxidation in type 1 diabetes mellitus. Journal of Sports Sciences, 29, 279– 289. Yardley, J.E., Kenny, G.P., Perkins, B.A. et al. (2012) Effects of performing resistance exercise before versus after aerobic exercise on glycemia in Type 1 diabetes. Diabetes Care, 35, 669–675. 18 Living with Hypoglycaemia Brian M. Frier University of Edinburgh, Edinburgh, UK

INTRODUCTION

Hypoglycaemia is the single most important factor that limits the achievement and main- tenance of good glycaemic control in people with insulin-treated diabetes. Because hypoglycaemia can occur at any time of day or night, is often unpredictable, affects intel- lectual and physical performance and disrupts the life of the affected individual and others, its effects can impinge on every aspect of everyday living. Irrespective of the causes and risk factors for hypoglycaemia, the impact on those affected is generally unpleasant, fright- ening and can have wide ramifications that include psychological sequelae. Adverse experi- ences of severe hypoglycaemia can influence the subsequent behaviour of an individual as he or she attempts to avoid further events, and the effect that this has on a patient’s self- care of diabetes may encourage poor glycaemic control.

PSYCHO-SOCIAL EFFECTS

Fear of hypoglycaemia The psychological effects of hypoglycaemia are discussed in detail in Chapter 16, but the potential impact on everyday life merits brief mention here. In addition to the subjective experience of symptoms and physical changes induced by acute hypoglycaemia, the effects on cognitive and non-cognitive functions can be disabling, leading to loss of control of events and reliance on others for assistance. The emotional consequences of living with the ever-present risk of hypoglycaemia may affect the personal lives of not only the affected individual but also of partners and other family members. It is not surprising that most individuals with recurrent exposure to severe hypoglycaemia will seek to avoid it. Many rate their fear of severe hypoglycaemia as equivalent to their concern about developing serious long-term complications of diabetes (Pramming et al. 1991) (Figure 18.1). In a group of 60 people with type 1 diabetes, 11 (18%) described severe hypoglycaemia as the most frightening event in their lives (Sanders et al. 1975), and many associated this with feelings of insecurity, tension and depression.

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 348 Hypoglycaemia in Clinical Diabetes

‘Mid’ hypoglycaemia Not worried Very worried

‘Severe’ hypoglycaemia Not worried Very worried

Diabetes thoughts Not worried Very worried

Blindness Not worried Very worried

Kidney complications Not worried Very worried

Figure 18.1 Attributes towards different aspects of diabetes indicated by 411 patients with type 1 diabetes using a visual analogue scale. (Source: Pramming S 1991. Reproduced with permission from John Wiley & Sons, Ltd).

People with insulin-treated diabetes who have experienced frequent severe hypoglycae- mia suffer greater levels of psychological distress, including increased anxiety, depression and fear of future hypoglycaemia (Wredling et al. 1992; Gold et al. 1994a). Fear of hypoglycaemia is also a common source of anxiety for relatives, and may strain marital and family relationships. Spouses have a greater fear of hypoglycaemia, and report expe- riencing sleep disturbance through worrying about nocturnal hypoglycaemia when com- pared with the spouses of those without experience of severe episodes (Gonder-Frederick et al. 1997; Jorgensen et al. 2003). The negative consequences of hypoglycaemia not only affect partners and spouses, but also the parents of children with type 1 diabetes (Clarke et al. 1998), the children of diabetic parents and other family members. Two-thirds of a group of 60 spouses of people with type 1 diabetes said that the risk of severe hypogly- caemia was a major source of concern to them, and when their partner is late, one in five considered hypoglycaemia to be the principal cause (Stahl et al. 1998). About 10% felt that severe hypoglycaemia was ‘always’ a burden. Similar worries may afflict a child who has previously discovered a comatose, hypoglycaemic diabetic parent. Anecdotal evidence suggests that the stress of dealing with episodic severe hypoglycaemia can sometimes lead to marital breakdown. Cox et al. (1987) devised a simple questionnaire, the Hypoglycaemia Fear Survey, to assess an individual’s fear of hypoglycaemia. In addition to measuring fear, it is possible Chapter 18: Living with Hypoglycaemia 349 to evaluate the way that people worry about hypoglycaemia and the behavioural responses that they take to avoid it. In some patients, hypoglycaemia may profoundly influence the impact of diabetes on their daily life and also their approach to self-management. The consequences of evasion and subsequent fear of hypoglycaemia may promote a phobia and encourage behavioural changes that maintain a high blood glucose to avoid future hypogly- caemia (see Chapter 16). Although the psychological consequences rarely produce frank psychiatric illness, the long-term effect of hypoglycaemia on subsequent behaviour may be much greater in many patients than is recognised by clinicians. This may partly explain the resistance shown by some individuals to therapeutic recommendations that would improve their glycaemic control. Fear of hypoglycaemia was cited as the main reason why many young patients with type 1 diabetes were discouraged from attempting strict glycae- mic control despite the outcome of the Diabetes Control and Complications Trial (DCCT) (Thompson et al. 1996). The non-cognitive effects of acute hypoglycaemia on mood and behaviour are extensive (Gold et al. 1997) and are discussed in Chapters 2 and 16. Hypogly- caemia also exerts a negative effect on quality of life in many people with insulin-treated type 2 diabetes (Barendse et al. 2012). The disruption to domestic life caused by an episode of severe hypoglycaemia, the feeling of helplessness in preventing further episodes, and the increased tension and anxiety that is generated, both in the person with diabetes and their relatives, is not conducive to a relaxed home environment. Unfortunately the psycho-social implications and conse- quences of hypoglycaemia on the family and on home life are rarely appreciated by many health professionals, who do not empathise with the domestic problems presented by hypoglycaemia.

EXERCISE

Regular exercise has long-term health benefits for people with insulin-treated diabetes (Wasserman and Zinman 1994). It increases insulin sensitivity, helps to control weight and improves several metabolic parameters, including lipids and cardiovascular risk factors (Lehmann et al. 1997). However, exercise may not improve overall glycaemic control unless it is frequent and intense (as pursued by many athletes), and can increase the risk of hypoglycaemia during and after the activity in people with insulin-treated diabetes. The relationship between exercise and the risk of hypoglycaemia is discussed in detail in Chapter 17. The temporal relationship between the time of exercise, the time of administration of insulin, and its time-action profile, are major determinants of the blood glucose curve (Riddell and Perkins 2006). Other relevant factors include the time of the ingestion of food and its nature, the intensity, type and duration of the exercise, and the site of insulin injec- tion. Exercise of a limb into which insulin has recently been injected will increase the rate of absorption by muscle action (Box 18.1). The risk of hypoglycaemia is also increased if intramuscular injection is made inadvertently (Frid et al. 1990).

Prevention of hypoglycaemia following exercise The potential risk of associated hypoglycaemia requires differing strategies to prevent an undesirable fall in blood glucose. The individual can either ingest additional short-acting carbohydrate, in liquid or solid form, or reduce the dose of insulin prior to physical activity. 350 Hypoglycaemia in Clinical Diabetes

Box 18.1 Factors influencing blood glucose response to exercise in people with insulin-treated diabetes • Time of previous insulin administration • Type of insulin used; insulin regimen • Site of insulin injection • Time of previous meal or snack • Nature and quality of food consumed before exercise • Duration, intensity, nature (aerobic, anaerobic or resistance) of exercise • Time of day of exercise

Both measures may be necessary, determined both by the duration and the intensity of the exercise intended. Strenuous anaerobic exercise as a short burst of physical activity, such as a sprint or a game of squash, can cause blood glucose to rise and does not therefore require prophylactic consumption of short-acting carbohydrate in advance. However, more prolonged aerobic exercise (such as running for 30 minutes) will cause a decline in blood glucose so necessitating prophylactic carbohydrate. A fall in blood glucose can be pre- vented if the aerobic exercise is preceded by resistance exercise (lifting weights), which also reduces the risk of post-exercise hypoglycaemia (Yardley et al. 2012). Protracted physical activity (aerobic exercise) lasting for several hours, such as hill walking or mara- thon running, requires a substantial reduction in total insulin dose, in addition to an increase in consumption of carbohydrate. It is difficult to advocate general measures that apply to everyone and all types of exercise, because the response can be idiosyncratic and also depends upon the overall quality of glycaemic control. The risk of inducing hypoglycaemia with exercise is obviously heightened in individuals who have strict glycaemic control and whose blood glucose is lower at the start of exercise (particularly 5.0 mmol/l or less), compared to those with moderate hyperglycaemia, while in those with poor glycaemic control with high prevailing blood glucose levels, there is a risk of strenuous exercise inducing ketosis. Trial and error may be necessary to assess the effect of specific activities on blood glucose in individuals. A fall in blood glucose may not occur during, or immediately following, physical exer- tion but may be delayed for several hours, sometimes occurring as much as 15 hours later (MacDonald 1987). Following exercise in the late afternoon or early evening, hypoglycae- mia may occur during the night or even the following day. The response depends on factors such as the efficiency of mobilisation of glucose from glycogen stores, how effectively these are replenished after exercise, the magnitude of the coexisting hormonal response and sympathoadrenal activation, and the prevailing plasma insulin concentration during and after exercise. For many people one of the safer times of day to exercise is in the fasting state (before breakfast) and before the administration of rapid-acting insulin, because insulin resistance is greatest at this time of day and plasma insulin levels are usually low (Ruegemer et al. 1990). However, this may vary depending on the type of basal insulin and its time of administration. In the fasting state, the fall in blood glucose during moderate exercise is small or absent in subjects with insulin-treated diabetes; prophylactic ingestion of carbo- hydrate may cause an unwanted rise in blood glucose and may need to be avoided (Soo et al. 1996). The use of continuous subcutaneous insulin infusion (CSII) does not remove the risk of exercise-induced hypoglycaemia, and interrupting the basal insulin infusion may Chapter 18: Living with Hypoglycaemia 351

Box 18.2 Measures to prevent hypoglycaemia induced by exercise • Take extra carbohydrate (20–30 g short-acting) before, and possibly during, exercise (especially if prolonged). • Reduce insulin dose before exercise. • Monitor blood glucose frequently. • Avoid peak absorption and time of action of insulin (for strenuous exercise). • Use anterior abdominal wall for injection of insulin (avoid active limbs). • Avoid exercise if blood glucose is high (>15.0 mmol/l) especially if ketosis is present. • Learn the glycaemic response to different types of exercise. • Carry identification re insulin therapy.

not be sufficient to prevent a fall in blood glucose after post-prandial exercise if hyperin- sulinaemia is present (Edelmann et al. 1986). In well-controlled patients using CSII the insulin dose should be reduced before, during and after exercise to minimise the risk of acute and late hypoglycaemia (Sonnenberg et al. 1990). This is discussed in detail in Chapter 17. Measures to limit hypoglycaemia occurring in relation to exercise are shown in Box 18.2. Undertaking protracted strenuous exercise (such as long-distance running) within 24 hours of an episode of severe hypoglycaemia is unwise and should be strongly discouraged (Graveling and Frier 2010). Strenuous exercise may sometimes be unpremeditated, as in an emergency situation. In addition, it is very easy for individuals to become distracted during activities like home decorating or gardening, such that they may work harder or longer than originally intended. It is therefore essential that all people with insulin-treated diabetes carry glucose tablets or an alternative source of quick-acting refined carbohydrate at all times to counter a sudden decline in blood glucose but without over-treating or inducing hyperglycaemia.

Sport Some sports are inherently dangerous and may be inadvisable for individuals at high risk of developing hypoglycaemia. Dangerous activities are those involving height, water, extremes of climate and exposure to inhospitable terrain. Potentially hazardous activities such as hang-gliding and parachuting, sub-aqua diving or unaccompanied rock climbing are inadvisable for people treated with insulin because of the risk of hypoglycaemia, and people taking insulin should not swim alone. Boxing is inadvisable because a decline of blood glucose into the hypoglycaemic range will impair performance and increase vulnerability to sustain a serious head injury, and it may be difficult for the hypoglycaemic boxer to distinguish the early symptoms of hypoglycaemia from those generated by the physical exertion. Apart from the risk to the individual, the safety of others must be considered, and an individual experiencing acute hypoglycaemia should not put them at risk. However, with adequate precautions and careful preparation, people with insulin-treated diabetes can tackle most sports safely, and many athletes or profes- sional sportsmen and women who have diabetes have achieved distinction at the highest levels of sporting prowess. For the average individual, for whom sport is principally a form of recreation and a means of obtaining regular exercise, measures to avoid hypogly- caemia are relatively straightforward, as described earlier. Gallen has described how to 352 Hypoglycaemia in Clinical Diabetes manage insulin-treated diabetes in relation to elite sporting activities (Gallen 2004), and information and advice about many different sports is available on a valuable website at http://www.runsweet.com. Most team sports, such as soccer and hockey, and competitive games such as squash or tennis have a relatively predictable duration, but other activities such as swimming, cycling or running are more variable. Endurance sports and protracted and demanding physical activities require more elaborate planning. Sustained exercise such as hill walking requires a premeditated reduction in total insulin dose of at least 20–30%. Teenagers with diabetes attending an outdoor activity holiday in Scotland reported that hill walking, canoeing and mountain biking, all activities that involved intense exercise of long duration, were most commonly associated with frequent and severe hypoglycaemia (Thompson et al. 1999). Long-distance running requires a considerable reduction in insulin dose. In 13 runners with insulin-treated diabetes who participated in the New York City Marathon, the total insulin dose was reduced by a mean of 38% (Grimm and Muchnick 1993). Frequent ingestion of beverages and snacks that are rich in carbohydrate is also necessary. A personal account of a marathon run (Kjeldby 1997) emphasised the difficulty in determining how much intermediate-acting insulin to inject in the evening following the run, and the need for frequent measurements of blood glucose over the next 24 to 48 hours to avoid delayed hypoglycaemia. Outward Bound mountain courses and holidays for young people with type 1 diabetes, which include rock climbing, canoeing, horse riding, caving and mountain expeditions, have been described by Hillson (1984, 1987) who has outlined the measures necessary to avoid and treat hypoglycaemia (Box 18.3). Anticipation of potential hazards for people at risk of hypoglycaemia must be considered for all activities, with consideration given to the timing of meals and administration of insulin, travelling time and how much energy is likely to be expended. It may be difficult to distinguish between physical exhaustion and hypoglycaemia, both of which may coexist (Hillson 1984), and hypothermia can be induced by hypoglycaemia as well as by cold and wet conditions. Cold-induced hypoglycaemia has been associated particularly with water sports such as canoeing and windsurfing, especially in adverse weather (Thompson et al. 1999). It may be necessary to reduce insulin dosage

Box 18.3 Measures to prevent hypoglycaemia in outdoor activities and holidays (derived from Hillson 1984, 1987) • Reduce total insulin dose (by 10–15%). • Ensure a good intake of high-fibre carbohydrate with plentiful quick-acting carbohydrate. • Increase carbohydrate at main meals and double the amount taken at snacks, or the number of snacks between main meals. • Consume glucose tablets or drinks immediately before climbing up or down anything high, or during water activities. • Carry glucose at all times and keep by the bed at night. • Monitor blood glucose four times daily and respond appropriately to the results. Aim for a blood glucose of ∼10.0 mmol/l. Blood glucose may be difficult to measure in cold or wet weather. • Take an hourly snack during prolonged exercise such as cycling or mountain walking. • Take a large pre-bedtime snack to avoid delayed hypoglycaemia. Chapter 18: Living with Hypoglycaemia 353 and increase carbohydrate intake during holidays and at summer camps for children with type 1 diabetes (Braatvedt et al. 1997); the sudden burst of energetic activities after children arrive at camp is associated with a considerably heightened hypoglycaemic risk. In chil- dren, this policy should be applied to holidays in general (see Chapter 10). The management of sporting activities involving children with type 1 diabetes has been reviewed in a valu- able Canadian article (Riddell and Iscoe 2006).

Recreation Strenuous and protracted exercise is not confined to sport and may occur during recreational activities, such as prolonged and vigorous dancing. These social events may also involve the consumption of alcohol, another potential cause of promoting and protracting hypo­ glycaemia. Some ‘recreational’ drugs such as amphetamines have been associated with promoting frenetic behaviour and increased metabolic rate, which may then induce hypogly- caemia in people treated with insulin (Jenks and Watkinson 1998). Young people with type 1 diabetes who attend clubs or parties often avoid the potential risk and embarrassment of hypoglycaemia by not taking their insulin before the event. Although this may seem to be a pragmatic approach, such an exercise may worsen the pre-existing hyperglycaemia, and even promote development of ketosis. A modest reduction of insulin dose, combined with appropriate high carbohydrate snacks and the judicious consumption of alcohol, should avoid hypoglycaemia, although this requires forward planning and may not be conducive to the spontaneity of social activity desired by many young adults. Similarly, unpremedi- tated and energetic sexual intercourse can precipitate unexpected hypoglycaemia, depend- ing on the related metabolic circumstances, and care should be taken to avoid this hazard if possible. It is easy to see why meticulous self-care of diabetes could inhibit the social activities of teenagers and young adults and interferes with late nights and parties. It is also understandable why many people do not strive for good glycaemic control in this situation. An episode of severe hypoglycaemia will ruin a social outing, but chronic hyperglycaemia is equally undesirable.

DRIVING

For people with diabetes who are treated with insulin, the potential risks of hypoglycaemia are always present and have therefore influenced the ways in which modern society regu- lates and restricts their activities. This principally affects driving licences and some forms of employment. Although most restrictions are reasonable and important for public safety, much lay, and even medical, ignorance exists about hypoglycaemia and its effects, so that discriminatory practices still occur, particularly with regard to employment.

Effect of hypoglycaemia on driving Driving is a common and everyday activity that demands complex psychomotor skills, including good visuo-spatial functions, rapid information processing, vigilance and satis- factory judgement. Because hypoglycaemia rapidly interferes with cognitive functions, even modest degrees of neuroglycopenia may affect driving skills, without necessarily provoking symptomatic awareness of hypoglycaemia. In seminal studies using a so­­ phisticated driving simulator, the driving abilities of drivers with type 1 diabetes were 354 Hypoglycaemia in Clinical Diabetes examined at different blood glucose concentrations, maintained by a glucose clamp (Cox et al. 1993, 2000). Driving performance started to deteriorate when blood glucose declined below 3.8 mmol/l, and typical driving deficiencies included speeding and inappropriate braking, driving off the road, crossing the centre line, ignoring ‘STOP’ signs and causing an increased number of ‘crashes’. Allowing for the artificial conditions of a driving simula- tor, it is evident that hypoglycaemia has an adverse effect on driving performance. A particularly disconcerting observation was that none of the drivers took action to treat hypoglycaemia until their blood glucose had declined to <2.8 mmol/l, and even then only 30% of the participants responded (Cox et al. 2000). Many did not experience any warning symptoms of hypoglycaemia, and fewer than 25% said they did not feel competent to drive when their blood glucose was low (Cox et al. 1993, 2000). The driving simulator studies also demonstrated that driving has a substantial metabolic demand that can lower blood glucose (Cox et al. 2001, 2002), leading the authors to recommend that a prophylactic snack should be consumed before driving if blood glucose is 5.0 mmol/l or less. Various studies have shown that many drivers with insulin-treated diabetes believe that it is safe to drive when their blood glucose is low (Clarke et al. 1999; Weinger et al. 1999; Graveling et al. 2004); this misperception may be influenced by progressive neuroglycopenia. Hypoglycaemia can impair cognitive function and judgement without necessarily pro- voking warning symptoms or altering consciousness. Driving can therefore continue while apparently not being under conscious control, a condition that has been given the strictly legal term of ‘automatism’ (but it should be noted that there are no medical or scientific publications about ‘automatism’). Irrational and compulsive behaviour during hypoglycae- mia has certainly been described by insulin-treated diabetic drivers (Frier et al. 1980). The police have on occasion arrested diabetic drivers who are hypoglycaemic, suspecting alcohol and inebriation to be the cause of a driver’s altered behaviour and symptoms.

Risk of accidents and restriction of driving licences Hypoglycaemia can adversely affect the ability to drive, and in individual cases hypogly- caemia has been implicated as a precipitating cause of road traffic accidents, causing the occasional fatality. In a study of insulin-treated drivers in Northern Ireland, the number of hypoglycaemic episodes that occurred while driving in the preceding year was shown to be associated with the total number of accidents during the previous 5 years (Stevens et al. 1989), consistent with Scottish studies that showed a greater rate of accidents among diabetic drivers who experienced hypoglycaemia while driving (Frier et al. 1980; Eading- ton and Frier 1989; MacLeod et al. 1993). It is difficult to quantitate how often hypogly- caemia occurs during driving and how often this precipitates a road traffic accident, particularly as hypoglycaemia-induced incidents in which the diabetic driver is killed are seldom identified after the event. In the UK, around a third of insulin-treated diabetic drivers have reported hypoglycaemia while driving (Frier et al. 1980; Eadington and Frier 1989; Stevens et al. 1989; Graveling et al. 2004). In the USA, a prospective multicentre study of 452 drivers with type 1 diabetes, all of whom were reviewed monthly for one year, documented 503 episodes of moderate hypoglycaemia (where self-treatment was possible but the individual could no longer drive safely) in 185 (41%) of the incidents while 23 drivers (5%) experienced 31 episodes of severe hypoglycaemia (inability to self-treat) that occurred while driving (Cox et al. 2009). There is no doubt that this is a significant problem among drivers treated with insulin. Chapter 18: Living with Hypoglycaemia 355

The rate of hypoglycaemia-induced motor vehicle accidents is difficult to evaluate and, of necessity, is often anecdotal. Most road traffic accidents involve several contributory factors, and it may be difficult to isolate hypoglycaemia as the principal cause. Studies in the UK have suggested that the accident rate of diabetic drivers is similar to non-diabetic drivers (Eadington and Frier 1989; Stevens et al. 1989; Lonnen et al. 2008), and this premise is supported by studies from Germany (Chantelau et al. 1990) and the USA (Songer et al. 1988). In an assessment of medical factors causing road traffic accidents, a study in Iceland showed that disorders such as diabetes were not over-represented (Gisla- son et al. 1997). One American study that observed a ‘slight increase’ in the risk of motor vehicle accidents in diabetic drivers (Hansotia and Broste 1991) concluded this increase to be insufficient to ‘warrant further restrictions on driving privileges’. These studies can be criticised for being retrospective, for excluding fatal accidents and being influenced by the removal of drivers with diabetes who had ceased driving either by their own volition or through the efforts of the regulatory authorities. They may no longer be representative of current driving conditions and accident risk. However, in the UK, notifications by the police to the Driver and Vehicle Licensing Agency (DVLA) of serious accidents associ- ated with hypoglycaemia have risen in recent years. This probably represents improved identification and increased reporting, rather than an escalating risk of hypoglycaemia- related road traffic accidents, but 3–5 fatalities associated with hypoglycaemia are reported annually in the UK (Dr Simon Rees, DVLA, personal communication). Many national licensing authorities in westernised countries restrict driving licences for people with insulin-treated diabetes, usually by restricting duration (known as a ‘period-restricted’ licence) with a necessity for regular medical review, but in other parts of the world no such regulation is in place. Other than visual impairment, the principal factors that com- monly lead to an ordinary driving licence being revoked are related to hypoglycaemia. Impaired awareness of hypoglycaemia with its increased risk of severe hypoglycaemia (see Chapter 6), and a history of recurrent severe episodes, can affect medical fitness to drive and present hazards to safe driving; they are common reasons for driving licences to be revoked.

Changes in European Union regulations for driving licences The European Union (EU) has altered the regulations for licensing drivers with insulin- treated diabetes, following the publication of their 3rd Directive on Driving in 2006. These changes affect people holding ordinary (Group 1) car licences and were implemented in 2010. In the UK these were implemented (but not initiated) by the DVLA. Medical fitness to drive must now be reviewed at no more than an interval of 5 years (for many years this has been performed at 3-yearly intervals in the UK), and a driver must not experience more than one episode of severe hypoglycaemia (defined as requiring help for recovery) within a period of 12 months, irrespective of when this occurs during the day, so nocturnal events during sleep are included. This was based on evidence that a history of severe hypogly- caemia was predictive of a greater risk of motor vehicle accidents (Cox et al. 2009). Impaired awareness of hypoglycaemia (IAH) has also been included as a reason to debar a driver with diabetes, although the evidence for this as a significant risk factor is debatable (American Diabetes Association 2012), and the syndrome has not been defined in the EU regulations, allowing some latitude in interpretation. In the UK, IAH has been defined as ‘total loss of warning symptoms of hypoglycaemia’, a situation that is uncommon in 356 Hypoglycaemia in Clinical Diabetes clinical practice and therefore allows many Group 1 drivers with IAH to continue driving, although frequent blood glucose monitoring and prophylactic snacks are strongly advised. Taxi driving is controlled by local authorities, and for drivers with diabetes who work for emergency services (such as the police, fire and ambulance services), driving restric- tions are determined by the employer, usually with professional advice from occupational health physicians.

Group 2 (Vocational) driving licences The policies of international governments towards vocational licensing for drivers with diabetes vary greatly around the world (DiaMond Project Group on Social Issues 1993) and even differ between states in the USA (Gower et al. 1992). In many parts of the world they are non-existent. In Europe a more stringent approach has been taken with Group 2 (vocational) driving licensing since 1991 (2nd Directive on Driving) for professional drivers of large goods vehicles (LGV) such as lorries, and public carriage vehicles (PCV) such as buses and coaches. Insulin-treated drivers were debarred from holding a Group 2 driving licence other than ‘in exceptional circumstances’, although some EU countries had taken a lax approach to licensing, which affected other states when drivers of these large vehicles crossed national boundaries. People with insulin-treated diabetes were allowed to hold a licence to drive smaller commercial vehicles that were classified as C1 vehicles, which weighed between 3.5 and 7.5 tonnes (3500 to 7500 kg), such as vans or lorries, and D1 vehicles in the form of minibuses carrying up to 16 passengers (not allowed in the UK). Licensing was subject to much higher medical standards, and in the UK required annual review of medical fitness to drive for renewal of the licence. The 3rd Directive on Driving has relaxed the regulations for Group 2 driving licences for people treated with insulin, who can now apply, but again the standards required are much stricter than for a Group 1 driving licence. Applicants must not have a history of severe hypoglycaemia, should have normal awareness of hypoglycaemia, and must be undertaking regular blood glucose monitoring at times related to driving, with the results being recorded on a glucose meter with an inbuilt memory function. In the UK, the assess- ment of medical fitness to drive for Group 2 licences now involves a two-stage process, with an independent assessment being made by a consultant diabetologist. The main medical concern is the risk of hypoglycaemia affecting the driver of one of these large vehicles. The need to safeguard public safety has to be balanced against the privilege of the individual with diabetes to hold a driving licence, and this issue can often arouse an emotive response. The implications of the changes in EU regulations on driving in the UK have been reviewed (Gallen et al 2012; Inkster and Frier 2013). The Federal Highways Administration in North America previously explored a waiver scheme for insulin-treated diabetic drivers of commercial trucks who drive between states, provided they could meet strict medical criteria and were free of severe hypoglycaemia, but this has not been pursued and regulations still differ between states. The American Diabetes Association (2012) has issued a detailed position paper about driving and diabetes, which discusses all the key issues.

Advice for diabetic drivers The education of drivers with diabetes by healthcare professionals regarding safe driving practices has been shown to be deficient (Cox et al. 2003; Watson et al. 2004). Although Chapter 18: Living with Hypoglycaemia 357

Box 18.4 Drivers with diabetes – reasons to cease driving Hypoglycaemia • People with newly diagnosed type 1 diabetes or any patient commencing treatment with insulin should cease driving temporarily until glycaemic control and vision are stable. • Recurrent hypoglycaemia (especially if severe). • Impaired awareness of hypoglycaemia (if disabling).

Other • Reduced (corrected) visual acuity for distance (worse than 6/12 on Snellen chart) in both eyes. Care required after use of mydriatic for eye examination. • Sensori-motor peripheral neuropathy with loss of proprioception. • Severe peripheral vascular disease; lower limb amputation (hand controls and auto- matic transmission may be feasible).

Box 18.5 Advice for drivers with diabetes regarding hypoglycaemia • If hypoglycaemia occurs while driving, stop the vehicle in a suitable location; leave the driver’s seat. • Always keep an emergency supply of readily accessible fast-acting carbohydrate (e.g. glucose tablets or sweets) in the vehicle. • Check blood glucose before driving (even on short journeys) and estimate at regular intervals on long journeys. • Take regular meals and snacks, and rest periods on long journeys; avoid alcohol. • If hypoglycaemia is experienced, do not drive until 45 minutes after blood glucose is restored to normal (delayed recovery of cognitive function). • Carry personal identification indicating ‘diabetes’ in case of injury in a road traffic accident.

this chapter is primarily concerned with hypoglycaemia, there are various reasons why an individual driver who is taking insulin may be advised to cease driving, albeit temporarily (Box 18.4). Cox et al. (1994) have claimed that blood glucose awareness training in a small number of people with impaired awareness of hypoglycaemia led to fewer road traffic accidents in subsequent years, suggesting an indirect benefit of this approach to improving the recognition of blood glucose fluctuations (see Chapter 6). Prevention of hypoglycaemia while driving is essential (Box 18.5), and it is important for the driver to plan each journey (no matter how short) in advance. Blood glucose testing is important before driving, and at arbitrary intervals of about 2 hours during long journeys, and rest periods for snacks and meals should be taken. Unfortunately, this practice is not common. A survey in Edinburgh showed that 50% of 202 insulin-treated diabetic drivers never tested blood glucose in rela- tion to driving, and only 14% did this regularly, most of these individuals having impaired awareness of hypoglycaemia (Graveling et al. 2004). Around half of those questioned admitted to a variety of unsafe practices with respect to driving. A supply of both quick-acting and more substantial complex carbohydrate should be kept permanently in the vehicle in case of unexpected delays or emergencies (traffic jams, 358 Hypoglycaemia in Clinical Diabetes breakdowns) or unpremeditated exercise such as changing a wheel, and replaced immedi- ately after use. If hypoglycaemia occurs while driving, the driver should stop the vehicle, switch off the engine and leave the driver’s seat, as in British law a charge of driving under the influence of a drug (insulin) can be made even if the car is stationary. It is also important that driving is not recommenced immediately after blood glucose is restored to normal. In this situation, blood glucose does not accurately reflect the glucose concentration in the brain, with the rise in intra-cerebral glucose lagging behind that in the peripheral blood. The complete recovery of cognitive function following hypoglycaemia can take up to 45 minutes after blood glucose has returned to normal (Chapter 2), and an interval of this duration should be allowed before driving is recommenced.

Medico-legal aspects of driving Physicians who specialise in diabetes are often required to provide medical reports in rela- tion to road traffic accidents involving drivers with insulin-treated diabetes, in whom hypoglycaemia has been implicated as a possible cause. A detailed history of the circum- stances should be taken from the driver to identify whether hypoglycaemia was likely at the time of the accident, as contemporaneous blood glucose measurements are seldom available. Occasionally, blood glucose has been measured at the scene of the accident by paramedical ambulance staff or on subsequent admission to hospital. However, any signifi- cant delay before the blood glucose is estimated may obscure the glycaemic status at the time of the accident, through the effect of counterregulatory hormones released by the stress of the accident and/or hypoglycaemia per se, or as a result of treatment. The presentation of a convincing history of hypoglycaemia preceding the accident has to be accompanied by a careful description of the potential effects of hypoglycaemia on cognitive function and behaviour, comprehensible to a lay person. Although this mitigat- ing factor may not allow legal charges to be dismissed, in the author’s experience the penalty may be reduced if disabling hypoglycaemia is accepted to be the principal problem that has affected the individual’s driving ability and had precipitated the acci- dent. However, this must not be considered to be a foregone conclusion and may depend on the outcome of the accident, because the legal view of hypoglycaemia occurring in a person treated with insulin (or an oral hypoglycaemic drug such as a sulfonylurea) is that this represents ‘careless’ or ‘reckless’ behaviour on their part and is therefore the ‘fault’ of the individual, even though in clinical practice no specific cause can be determined for many episodes of hypoglycaemia. Although the difficulty of always being able to test blood glucose before driving is recognised in clinical practice, when this has a serious outcome, the judiciary may take a much stricter view. In a court case in Edinburgh in 2000, which involved a fatal car accident caused by a hypoglycaemic diabetic driver who had not measured his blood glucose before driving, the Sheriff commented: ‘There is a risk associated with diabetes and driving, and as a consequence, there is a need to monitor your blood sugar level. There is some culpability on your part’. The driver was found guilty of dangerous driving. By contrast, spontaneous hypoglycaemia is an accepted defence, and one patient with insulin-treated diabetes under the author’s care had charges of dangerous driving dismissed when it was shown that at the time of the offence he had developed undiagnosed and untreated Addison’s disease – a relatively rare but recognised cause of increased and unpredictable severe hypoglycaemia in type 1 diabetes. Medico- legal aspects of hypoglycaemia and diabetes have been reviewed elsewhere (Frier and Maher 1988). Chapter 18: Living with Hypoglycaemia 359

Rationing of blood glucose strips on economic grounds for people with type 2 diabetes, which is a common policy in primary care, cannot be justified for drivers who are receiv- ing treatment with insulin, and a robust case can be made for routine blood glucose testing in relation to driving in people with type 2 diabetes who are receiving insulin secretago- gues (sulfonylureas and meglitinides). As a result, general practitioners should consider carefully when refusing or limiting prescriptions for blood glucose strips to such patients, as their involvement in a subsequent hypoglycaemia-related motor vehicle accident that led to litigation and claims for damages and injury could imply culpability on the part of the doctor.

TRAVEL

Many of the measures recommended for longer car journeys (Box 18.5) are appropriate to long-distance travel, irrespective of the mode of transport used. Forward planning is essen- tial to avoid hypoglycaemia, with emphasis on adjustment of insulin dose (or regimen) if necessary, carrying equipment for blood glucose monitoring and ensuring an adequate supply both of quick-acting carbohydrate and of non-perishable emergency rations in case suitable food is not available during travel. Standard airline meals are often low in unrefined carbohydrate and may be unpalatable. Advice for travel and holidays is available from various sources, but with respect to avoiding (and treating) hypoglycaemia, some practical points can be made:

• For long-distance air travel, crossing several time zones, frequent administration of short-acting (soluble or analogue) insulin is much simpler to use than attempting to modify the times of administration and dosage of intermediate-acting (isophane) insu- lins. Disposable insulin pens are also very useful for this purpose. Rapid-acting insulin analogues have the advantage that their administration can be delayed until the food on offer is available and its palatability assessed, or can be taken after the meal, provid- ing greater flexibility for treatment. Adjustment of the long-acting (basal) insulin may not be necessary. • Some blood glucose meters are inaccurate in the hypoglycaemic range and many do not give accurate readings at high altitude or at extremes of temperature. Visually-read strips for blood glucose estimation may therefore be necessary in some situations. It is advisable to carry a spare blood glucose monitor in case of equipment failure. Insulin pumps may also malfunction during commercial air travel and induce hypoglycaemia by delivering more insulin than intended (King et al. 2011). Atmospheric pressure reduction causes insulin delivery from the pump by bubble formation and expansion of existing bubbles. • A supply of quick-acting carbohydrate is essential, but should be stored appropriately. Dextrose tablets may disintegrate or become very hard in hot and humid climates unless wrapped in silver foil or stored in a suitable container, and chocolate will melt if the temperature is high. At very cold temperatures, the wrapper may become welded to the chocolate. Cartons of orange juice cannot be re-used once opened, and so a plastic bottle or container with a screw top is preferable. Sealed packets of powdered glucose may be more suitable to carry in hot humid climates. • Travelling companions should carry a supply of fast-acting carbohydrate (and gluca- gon) for emergency use. 360 Hypoglycaemia in Clinical Diabetes

The nature of the travel undertaken, the amount of energy expended, the quality and nature of food and the risk of intermittent illness (such as travel sickness or gastroenteritis) are all factors that can influence blood glucose and potentially induce hypoglycaemia. Although many situations are predictable, the most important measure is frequent monitor- ing of blood glucose so that sensible adjustments in insulin therapy and ingestion of food can be made.

EMPLOYMENT

The risk of developing acute hypoglycaemia and its consequences (mainly in people with insulin-treated diabetes) provide reasons why some forms of employment are not available to individuals who require insulin therapy. Employment prospects are often restricted where the threat of hypoglycaemia poses a risk to the diabetic worker or to his or her colleagues and to the general public. With some occupations, particularly those involving public transport, any risk of hypoglycaemia is considered to be unacceptable. In other areas, the potential risks of hypoglycaemia are less well defined, and restrictions to employment have been established by individual industries or firms, rather than by legisla- tion. Although medical advice has usually been sought, this has not always been well informed, and may not have involved physicians with expertise in diabetes or occupational health. Some restrictions have been challenged successfully, with one example in the UK being the reinstatement of several active firefighters, based on individual medical assess- ment. The regulations on commercial pilots are being modified in Europe and have been relaxed for individuals who meet strict medical criteria and are considered to have a rela- tively low risk of developing hypoglycaemia; they will have to fly alongside a co-pilot who does not have diabetes. Air Canada have successfully adopted this policy for pilots treated with insulin for a number of years. In the USA, an air traffic controller appealed successfully against dismissal on grounds of discrimination, but the nature of this work could leave an individual very vulnerable to the effects of neuroglycopenia with potentially catastrophic consequences. People with insulin-treated diabetes are not usually permitted to work alone in isolated or dangerous areas or at unprotected heights. They are also debarred from serving in the armed forces. This is based on the grounds that all service personnel (including non- combatants) could be involved in a conflict at short notice, and maintaining provision of insulin and appropriate dietary requirements could present difficulties in a wartime situa- tion. Employment is not usually permitted in emergency teams, civil aviation, work in the offshore oil industry and in many forms of commercial driving (Waclawski 1989). A list of jobs in which the employment of people with insulin-treated diabetes (both types 1 and 2) is generally restricted is shown in Table 18.1.

Hypoglycaemia at work Although anecdotal accounts describe severe hypoglycaemia affecting employees with insulin-treated diabetes while at work, this does not appear to be a widespread problem in people with type 1 diabetes. Although isolated episodes of severe episodes occurring in the workplace are inevitable, it appears that most hypoglycaemia is mild, quickly self-treated and does not cause disruption. The times of day at which hypoglycaemia is most common Chapter 18: Living with Hypoglycaemia 361

Table 18.1 Employment restrictions placed on workers with insulin-treated diabetes in the UK. Absolute restrictions have been removed for categories marked * (Source: Adapted from Waclawski ER 1989. With permission from John Wiley & Sons, Ltd)

Vocational driving Large goods vehicles (LGV licences) * Public carriage vehicles (PCV licences) * Locomotives and underground trains Professional drivers (chauffeurs) Taxi drivers (variable: depends on local authority) Civil aviation Commercial pilots; flight engineers * Aircrew * Air traffic controllers National and emergency Armed forces (Army, Navy, Air Force) services Police force Fire brigade or Rescue services Prison and Security services Dangerous areas for work Offshore oil-rig work Moving machinery Incinerator loading Hot-metal areas Work on railway tracks Coal mining Work at heights Overhead linesmen Crane driving Scaffolding/high ladders or platforms

were recorded in a prospective study of 60 patients with type 1 diabetes, half of whom had impaired awareness of hypoglycaemia (Gold et al. 1994b). Most severe episodes occurred during the evening or night, or in the early morning before work (Figure 18.2). The higher frequency of severe hypoglycaemia in the evening or during the night was not a conse- quence of the insulin regimens being used, and may be related to relaxed vigilance or different behaviour when at home. A study of 243 people with insulin-treated diabetes in fulltime employment in Edinburgh, conducted prospectively for one year, indicated that the frequency of severe hypoglycaemia at work was low (15% of all episodes, affecting 11% of workers) (Leckie et al. 2005). Although severe hypoglycaemia occurred at home and at other times, the hypoglycaemia experienced in the workplace seldom caused disrup- tion or serious morbidity. People treated with insulin may be more vigilant while they are at work and actively try to avoid a low blood glucose, working activities may be more regular than at home or at weekends, or they self-treat low blood glucose promptly, so avoiding significant neuroglycopenia. However, in the Edinburgh study, very few of the people in employment had impaired awareness (a major risk factor for severe hypoglycae- mia), and overall glycaemic control was not strict, suggesting that many employees with insulin-treated diabetes may deliberately avoid having a low HbA1c to limit their hypogly- caemia risk (Leckie et al. 2005). People with IAH are just as active economically as those with normal awareness (Ogundipe et al. 2011). By contrast, many people with insulin-treated type 2 diabetes have reported problems with work in relation to non-severe episodes, ranging from loss of productivity to having to miss work or arrive late because of hypoglycaemia occurring outside working hours (particularly nocturnal), or not being able to perform satisfactorily while at work (Brod et al. 2011). 362 Hypoglycaemia in Clinical Diabetes

100

80

p = 0.03 60

NS 40 p = 0.05

20 NS

% of episodes severe hypoglycaemia 0 8 am–1 pm 1 pm–6 pm 6 pm–12 MN 00–8 am

Figure 18.2 Percentages of total number of episodes of severe hypoglycaemia occurring at different times of day in patients with type 1 diabetes with normal (solid bars) and impaired awareness of hypogly- caemia (speckled bars). NS, not significant; MN, midnight. (Source: Gold AE 1994b. Reproduced with permission of the American Diabetes Association).

Shift work is generally not a contraindication to the employment of people with type 1 diabetes. However, occasionally a frequent change of shift rota can cause difficulties with glycaemic control, although this usually causes deterioration in control rather than hypogly- caemia (Poole et al. 1992). Measures to avoid and/or treat hypoglycaemia at work are no different from at any other time or circumstance, although in some jobs, blood glucose monitoring may not be feasible while at work, and break times for snacks and meals may be variable. It is essential that workmates or colleagues are familiar with the emergency treatment of hypoglycaemia and that a supply of quick-acting carbohydrate is available in the workplace. Ideally, the individual’s employer is made aware that he or she has insulin- treated diabetes, although some people conceal this, fearing dismissal or discrimination. Legislation to avoid this is in place in many Western countries, but widespread ignorance remains about the nature of diabetes, its treatment and the possible side effects, and this has to be confronted by those with specialist expertise. Similar arrangements for emergency treatment should be in place in schools (see Chapter 10), colleges and other venues of tertiary education, including university accommodation. In separate cases in Edinburgh (Strachan et al. 2000), two young male students with type 1 diabetes died tragically after developing severe hypoglycaemia while living in university halls of residence. One of the students was left lying unconscious for 2 days on the floor of his room, despite being observed by a member of the domestic staff who thought he was asleep or drunk. Although the warden had known that the student had diabetes, other members of staff did not. At the subsequent fatal accident enquiry, the Sheriff discussed the dichotomy between maintaining the privacy and confidentiality of the individual, and the duty of pastoral care owed to young people living in university accommodation. As a result of his report and recommendations, the facilities and practical support for students with type 1 diabetes in residencies were modified in Scottish universities. However, these cases highlight the importance of disclosure of diabetes, and its potential metabolic prob- lems (especially hypoglycaemia), to the appropriate authorities and those who may be required to render emergency assistance. Chapter 18: Living with Hypoglycaemia 363

Specialist medical reports Physicians who specialise in diabetes are often required to provide medical reports for employers, either related to the suitability of specific types of work for a person with insulin-treated diabetes or to their capability of performing the job. It is often necessary to advocate on the behalf of patients when a problem at work is specifically related to some aspect of their diabetic management, such as hypoglycaemia. The introduction of insulin therapy sometimes has to be delayed because of the impact that this will have on the individual’s employment. By strict dietary measures, an oil tanker driver (with a LGV licence) attending the author’s clinic, who had longstanding type 2 diabetes, managed to maintain adequate glycaemic control on the maximum dose of combined oral antidiabetes therapy for nearly 2 years, delaying conversion to insulin. He was anxious to protect his pension rights until his retirement, at which point insulin was commenced. However, procrastination with starting insulin therapy is usually inadvisable in type 2 diabetes, and is not possible in those presenting with newly-diagnosed type 1 diabetes. Alternative treatment such as a GLP-1 agonist may be feasible but should not be taken with a sul- fonylurea. The relaxation of the EU regulations for Group 2 driving licences may per- suade more professional drivers approaching insulin therapy not to procrastinate, but they will have to stop driving Group 2 vehicles for at least 3 months while collecting blood glucose data during treatment with insulin, and this could result in loss of employment. Sometimes, sympathetic medical counselling may be necessary to recommend an alterna- tive occupation.

School and academic examinations It is recognised that school pupils and students with type 1 diabetes may be disadvantaged during school or college examinations by unpredictable fluctuations in blood glucose con- centrations, with pronounced cognitive impairment and mood changes occurring when blood glucose falls below 3.5 mmol/l. Fluctuations in blood glucose may be exacerbated by the stress associated with an exam, and sometimes children are unable to consume adequate carbohydrate in advance of an academic test because of pre-exam anxiety and temporary loss of appetite. The consumption of glucose by the brain is very substantial during a period of intense intellectual activity and concentration, and this may cause the blood glucose to fall considerably and provoke hypoglycaemia. This effect may be greater in the child than in an adult because of the larger size of the brain relative to the rest of the body, and its high energy requirements. The idiosyncratic nature of this effect and the variability in different circumstances and at different times of day makes accurate deter- mination of insulin requirement difficult in advance of an exam, but generally a lower dose of insulin should be given. However, excessive reductions in dose to avoid hypoglycaemia may lead to undesirable hyperglycaemia, causing osmotic symptoms and mood change. Although a pupil or student with type 1 diabetes can (and should) measure blood glucose immediately before an exam, it is generally impractical to do repeated monitoring during the exam, where the use of a blood glucose meter or the consumption of food or glucose drinks may be prohibited. It may be necessary to request special dispensation in advance. During an exam, because of multiple distractions and the need to focus on the exam itself, the affected individual may not identify the usual warning symptoms of hypoglycae- mia. In addition, because many of the typical symptoms of hypoglycaemia (such as tremor, sweating, pounding heart, feelings of anxiety) are caused by stimulation of the autonomic 364 Hypoglycaemia in Clinical Diabetes nervous system (see Chapter 2), even when they are generated in this situation, they may be incorrectly attributed by the student (or by an observer) to exam-induced anxiety; the falling blood glucose is therefore not identified and corrective treatment is not given. Neuroglycopenic symptoms such as difficulty concentrating, feeling lightheaded and drow- siness, may also be ignored in an exam situation, and cognitive dysfunction may contribute to the difficulty in initiating self-treatment. Hypoglycaemia affects most cognitive functions, with the severity depending on the depth and duration of the low blood glucose, and the complexity of the tasks being under- taken (Chapter 2). All of the major cognitive domains are affected, including memory, attention, concentration, reasoning, problem-solving ability, abstract thought, executive function, rapid decision-making and judgement, and skills like hand-eye coordination are impaired. In addition, hypoglycaemia provokes major mood changes, including tense- tiredness, pessimism, irritability and sometimes anger, which are not conducive to enhancing exam performance. Hyperglycaemia can also affect cognitive function but in more subtle ways and with less dramatic impact than hypoglycaemia, and is also associated with nega- tive mood changes.

PRISON AND POLICE CUSTODY

Police custody and hypoglycaemia Many features of acute hypoglycaemia simulate those of alcoholic inebriation, so that people with insulin-treated diabetes have occasionally been arrested by the police under the mistaken impression that they are drunk or under the influence of drugs. The danger of this situation is often compounded by their detention at a police station, with confine- ment in a cell, instead of receiving treatment or urgent transfer to hospital. Unfortunately, the consumption of alcohol can promote hypoglycaemia in people treated with insulin, and may be a contributory factor to inducing the low blood glucose, so causing further difficulty with identification of the underlying metabolic problem. This emphasises the importance of an individual with insulin-treated diabetes carrying some form of identification to indi- cate that they are taking insulin and may be at risk of developing hypoglycaemia-induced coma. In addition to the risk of being arrested during an episode of hypoglycaemia because of aggressive or abnormal behaviour, low blood glucose may develop while in custody. The police have limited comprehension of the needs of a person with diabetes and the risks of hypoglycaemia. The young male patient with undiagnosed Addison’s disease, described earlier, who was arrested on a driving charge, was profoundly neuroglycopenic when taken into custody. He was detained without treatment for 2 hours. When his father arrived at the police station, he recognised immediately that his son was severely hypoglycaemic and needed emergency treatment with dextrose. This type of situation is clearly alarming, and potentially could have a fatal outcome. An initiative in Edinburgh led to successful liaison with the local police force to improve the way in which people with insulin-treated diabetes are handled while in custody (Wright et al. 2008), and this model has been adopted in other centres.

Management of diabetes in prison The general problems of managing diabetes in prison were examined in two British studies (Gill and MacFarlane 1989; MacFarlane et al. 1992), and recommendations have been Chapter 18: Living with Hypoglycaemia 365 made to improve the care of diabetes in prison, by the American Diabetes Association (Eichold 1989) and by Diabetes UK (Gill et al. 1992). Imprisonment causes particular problems for the management of diabetes, which are conducive to the development of hypoglycaemia. These include the following:

• an inadequate or inappropriate prison diet; • long ‘lock-up’ periods necessitated by prison routine; • solitary confinement for individual prisoners; • restrictions in the time and place of insulin administration; • the use of some insulin regimens (e.g. basal-bolus and/or injection of bedtime isophane insulin) are precluded by prison routine; • a long time interval between the evening meal and breakfast the following morning (sometimes over 12 hours); • no blood glucose monitoring facilities being allowed in cells; • lack of medical knowledge among most prison officers with few personnel having any medical training.

Many of these problems predispose to a risk of nocturnal hypoglycaemia, and actively discourage any attempt at achieving strict glycaemic control. Various measures can be suggested to try and prevent hypoglycaemia in prisoners with insulin-treated diabetes:

• availability of dextrose tablets (or an alternative source of carbohydrate, e.g. biscuits) in cells; • provision of a late evening snack; • avoidance of solitary confinement if possible; • sharing cells with prisoner(s) who can recognise hypoglycaemia; • arranging access to specialist advice on management of diabetes.

CONCLUSIONS • Fear of hypoglycaemia is common and may influence self-management of blood glucose by individual patients. Worries about hypoglycaemia extend to relatives, spouses and partners of the person with diabetes, and recurrent hypoglycaemia can disrupt family life. • The risk of hypoglycaemia occurring during exercise depends on the prevailing plasma concentrations of insulin and glucose and the duration, nature and intensity of the physi- cal activity. Strategies to avoid a fall in blood glucose include the ingestion of additional carbohydrate and a reduction in insulin dose. The contributory effect of alcohol may be important. Sport, recreational activities and travel all require forward planning and preventative measures. Some dangerous recreational activities should be avoided. • Although the risk of hypoglycaemia-related driving accidents is difficult to quantitate, hypoglycaemia is a potential hazard when driving, and impaired awareness of hypogly- caemia may cause revocation of the driving licence. The insulin-treated driver must carry a supply of glucose in the vehicle and take suitable precautions to ensure safe driving on all journeys, including blood glucose monitoring. • Hypoglycaemia at work is uncommon but the potential risk debars people with insulin- treated diabetes from certain occupations. Discrimination by employers against people with insulin-treated diabetes still occurs. 366 Hypoglycaemia in Clinical Diabetes • The problems of managing diabetes in prison include inadequate facilities for prevent- ing hypoglycaemia, especially overnight. Retention of individuals with diabetes in police custody may cause difficulties by failure of the custodians to recognise hypogly- caemia and its similarity to the features of inebriation.

REFERENCES

American Diabetes Association (2012) Diabetes and driving. Diabetes Care, 35 (Suppl. 1), S81–S86. Barendse, S., Singh, H., Frier, B.M. & Speight, J. (2012) The impact of hypoglycaemia on quality of life and related patient-reported outcomes in Type 2 diabetes: a narrative review. Diabetic Medicine, 29, 293–302. Braatvedt, G.D., Mildenhall, L., Patten, C. & Harris, G. (1997) Insulin requirements and metabolic control in children with diabetes mellitus attending a summer camp. Diabetic Medicine, 14, 258–261. Brod, M., Christensen, T., Thomsen, T.L. & Bushnell, D.M. (2011) The impact of non-severe hypoglycemic events on work productivity and diabetes management. Value in Health, 14, 665–671. Chantelau, E., Dannbeck, S., Kimmerle, R. & Ross, D. (1990) Zur verkehrstüchtigkeit insulin-behandelter diabetiker. Münchener Medizinische Wochenschrift, 132, 468–471. Clarke, W.L., Gonder-Frederick, L.A., Miller, S., Richardson, T. & Snyder, A. (1998) Maternal fear of hypoglycemia in their children with insulin-dependent diabetes mellitus. Journal of Pediatric Endo- crinology and Metabolism, 4, 189–194. Clarke, W.L., Cox, D.J., Gonder-Frederick, L.A. & Kovatchev, B. (1999) Hypoglycemia and the decision to drive a motor vehicle by persons with diabetes. Journal of the American Medical Association, 282, 750–754. Cox, D.J., Irvine, A., Gonder-Frederick, L., Nowacek, G. & Butterfield, J. (1987) Fear of hypoglycemia: quantification, validation and utilization. Diabetes Care, 10, 617–621. Cox, D.J., Gonder-Frederick, L. & Clarke, W. (1993) Driving decrements in type 1 diabetes during moder- ate hypoglycemia. Diabetes, 42, 239–243. Cox, D.J., Gonder-Frederick, L., Julian, D.M. & Clarke, W. (1994) Long-term follow-up evaluation of blood glucose awareness training. Diabetes Care, 17, 1–5. Cox, D.J., Gonder-Frederick, L.A., Kovatchev, B.P., Julian, D.M. & Clarke, W.L. (2000) Progressive hypoglycemia’s impact on driving simulation performance. Occurrence, awareness and correction. Dia- betes Care, 23, 163–170. Cox, D.J., Gonder-Frederick, L.A., Kovatchev, B.P. & Clarke, W.L. (2001) Self-treatment of hypoglycemia while driving. Diabetes Research and Clinical Practice, 54, 17–26. Cox, D.J., Gonder-Frederick, L.A., Kovatchev, B.P. & Clarke, W.L. (2002) The metabolic demands of driving for drivers with type 1 diabetes mellitus. Diabetes/Metabolism Research and Reviews, 18, 381–385. Cox, D.J., Penberthy, J.K., Zrebiec, J. et al. (2003) Diabetes and driving mishaps: frequency and correla- tions from a multinational survey. Diabetes Care, 26, 2329–2334. Cox, D.J., Ford, D., Gonder-Frederick, L.G. et al. (2009) Driving mishaps among individuals with type 1 diabetes: a prospective study. Diabetes Care, 32, 2177–2180. DiaMond Project Group on Social Issues (1993) Global regulations on diabetics treated with insulin and their operation of commercial motor vehicles. British Medical Journal, 307, 250–253. Eadington, D.W. & Frier, B.M. (1989) Type 1 diabetes and driving experience: an eight-year cohort study. Diabetic Medicine, 6, 137–141. Edelmann, E., Staudner, V., Bachmann, W., Walter, H., Haas, W. & Mehnert, H. (1986) Exercise-induced hypoglycaemia and subcutaneous insulin infusion. Diabetic Medicine, 3, 526–531. Eichold, S. (1989) Diabetes in correctional institutions. Diabetes Care, 12, 509–510. Frid, A., Östman, J. & Linde, B. (1990) Hypoglycemia risk during exercise after intramuscular injection of insulin in thigh in IDDM. Diabetes Care, 13, 473–477. Frier, B.M. & Maher, G. (1988) Diabetes and hypoglycaemia: medico-legal aspects of criminal responsibil- ity. Diabetic Medicine, 5, 521–526. Frier, B.M., Matthews, D.M., Steel, J.M. & Duncan, L.J.P. (1980) Driving and insulin-dependent diabetes. Lancet, 315, 1232–1234. Chapter 18: Living with Hypoglycaemia 367

Gallen, I.W. (2004) Helping the athlete who has type 1 diabetes. British Journal of Diabetes and Vascular Disease, 4, 87–92. Gallen, I., Amiel, S., Robinson, T. & McKnight, J. on behalf of the ABCD Committee (2012) Driving and hypoglycaemia: questions and answers. Practical Diabetes, 29, 13–14. Gill, G.V. & MacFarlane, I.A. (1989) Problems of diabetics in prison. British Medical Journal, 298, 221–223. Gill, G.V., MacFarlane, I.A. & Tucker, N.H. (1992) Diabetes care in British prisons: existing problems and potential solutions. Diabetic Medicine, 9, 109–113. Gislason, T., Tomasson, K., Reynisdottir, H., Björnsson, J.K. & Kristbjarnarson, H. (1997) Medical risk factors amongst drivers in single-car accidents. Journal of International Medicine, 241, 213–219. Gold, A.E., Deary, I.J., Jones, R.W., O’Hare, J.P., Reckless, J.P.D. & Frier, B.M. (1994a) Severe deteriora- tion in cognitive function and personality in five patients with long-standing diabetes: a complication of diabetes or a consequence of treatment? Diabetic Medicine, 11, 499–505. Gold, A.E., MacLeod, K.M. & Frier, B.M. (1994b) Frequency of severe hypoglycemia in patients with type 1 diabetes with impaired awareness of hypoglycemia. Diabetes Care, 17, 697–703. Gold, A.E., Deary, I.J. & Frier, B.M. (1997) Hypoglycaemia and non-cognitive aspects of psychological function in insulin-dependent (Type 1) diabetes mellitus (IDDM). Diabetic Medicine, 14, 111–118. Gonder-Frederick, L., Cox, D., Kovatchev, B., Julian, D. & Clarke, W. (1997) The psychosocial impact of severe hypoglycemic episodes on spouses of patients with IDDM. Diabetes Care, 20, 1543–1546. Gower, I.F., Songer, T.J., Hylton, H. et al. (1992) Epidemiology of insulin-using commercial motor vehicle drivers. Diabetes Care, 15, 1464–1467. Graveling, A.J. & Frier, B.M. (2010) Risks of marathon running and hypoglycaemia in type 1 diabetes. Diabetic Medicine, 27, 585–588. Graveling, A.J., Warren, R.E. & Frier, B.M. (2004) Hypoglycaemia and driving in people with insulin-treated diabetes: adherence to recommendations for avoidance. Diabetic Medicine, 21, 1014–1019. Grimm, J.-J. & Muchnick, S. (1993) Type 1 diabetes and marathon running. Diabetes Care, 16, 1624 (letter). Hansotia, P. & Broste, S.K. (1991) The effect of epilepsy or diabetes mellitus on the risk of automobile accidents. New England Journal of Medicine, 324, 22–26. Hillson, R.M. (1984) Diabetes Outward Bound mountain course, Eskdale, Cumbria. Diabetic Medicine, 1, 59–63. Hillson, R.M. (1987) British Diabetic Association activities for young people – safety while adventuring. Practical Diabetes, 4, 233–234. Inkster, B. & Frier, B.M. (2013) Diabetes and driving. Diabetes, Obesity and Metabolism, 15, 775–783. Jenks, J. & Watkinson, M. (1998) Minimising the risks of amphetamine use for young adults with diabetes. Journal of Diabetes Nursing, 2, 179–182. Jorgensen, H.V., Pedersen-Bjergaard, U., Rasmussen, A.K. & Borch-Johnsen, K. (2003) The impact of severe hypoglycemia and impaired awareness of hypoglycemia on relatives of patients with type 1 diabetes. Diabetes Care, 26, 1106–1109. King, B.R., Goss, P.W., Paterson, M.A., Crock, P.A. & Anderson, D.G. (2011) Changes in altitude cause unintended insulin delivery from insulin pumps. Diabetes Care, 34, 1932–1933. Kjeldby, B. (1997) Running the New York marathon with diabetes. British Medical Journal, 314, 1053–1054. Leckie, A.M., Graham, M.K., Grant, J.B., Ritchie, P.J. & Frier, B.M. (2005) Frequency, severity, and morbidity of hypoglycemia occurring in the workplace in people with insulin-treated diabetes. Diabetes Care, 28, 1333–1338. Lehmann, R., Kaplan, V., Bingisser, R., Bloch, K.E. & Spinas, G.A. (1997) Impact of physical activity on cardiovascular risk factors in IDDM. Diabetes Care, 20, 1603–1611. Lonnen, K.F., Powell, R.J., Taylor, D., Shore, A.C. & MacLeod, K.M. (2008) Road traffic accidents and diabetes: insulin use does not determine risk. Diabetic Medicine, 25, 578–584. MacDonald, M.J. (1987) Postexercise late-onset hypoglycemia in insulin-dependent diabetic patients. Diabetes Care, 10, 584–588. MacFarlane, I.A., Gill, G.V., Masson, E. & Tucker, N.H. (1992) Diabetics in prison: can good diabetic care be achieved? British Medical Journal, 304, 152–155. MacLeod, K.M., Hepburn, D.A. & Frier, B.M. (1993) Frequency and morbidity of severe hypoglycaemia in insulin-treated diabetic patients. Diabetic Medicine, 10, 238–245. 368 Hypoglycaemia in Clinical Diabetes

Ogundipe, O.O., Geddes, J., Leckie, A.M. & Frier, B.M. (2011) Impaired hypoglycaemia awareness and employment in people with type 1 diabetes. Occupational Medicine, 61, 241–246. Poole, C., Wright, A. & Nattrass, M. (1992) Control of diabetes mellitus in shift workers. British Journal of Industrial Medicine, 49, 513–515. Pramming, S., Thorsteinsson, B., Bendtson, I. & Binder, C. (1991) Symptomatic hypoglycaemia in 411 type 1 diabetic patients. Diabetic Medicine, 8, 217–2. Riddell, M.C. & Iscoe, K.E. (2006) Physical activity, sport and pediatric diabetes. Pediatric Diabetes, 7, 60–70. Riddell, M.C. & Perkins, B.A. (2006) Type 1 diabetes and vigorous exercise: applications of exercise physiology to patient management. Canadian Journal of Diabetes, 30, 63–71. Ruegemer, J.J., Squires, R.W., March, H.M. et al. (1990) Differences between pre-breakfast and late after- noon glycemic responses to exercise in IDDM patients. Diabetes Care, 13, 104–110. Sanders, K., Mills, J., Martin, F.I.R. & Horne, D.J.D. (1975) Emotional attitudes in adult insulin-dependent diabetics. Journal of Psychosomatic Research, 19, 241–246. Songer, T.J., LaPorte, R.E., Dorman, J.S. et al. (1988) Motor vehicle accidents and IDDM. Diabetes Care, 11, 701–707. Sonnenberg, G.E., Kemmer, F.W. & Berger, M. (1990) Exercise in type 1 (insulin-dependent) diabetic patients treated with continuous subcutaneous insulin infusion. Prevention of exercise induced hypogly- caemia. Diabetologia, 33, 696–703. Soo, K., Furler, S.M., Samaras, K., Jenkins, A.B., Campbell, L.V. & Chisholm, D.J. (1996) Glycemic responses to exercise in IDDM after simple and complex carbohydrate supplementation. Diabetes Care, 19, 575–579. Stahl, M., Berger, W., Schaechinger, H. & Cox, D.J. (1998) Spouse’s worries concerning diabetic partner’s possible hypoglycaemia. Diabetic Medicine, 15, 619–620 (letter). Stevens, A.B., Roberts, M., McKane, R., Atkinson, A.B., Bell, P.M. & Hayes, J.R. (1989) Motor vehicle driving among diabetics taking insulin and non-diabetics. British Medical Journal, 299, 591–595. Strachan, M.W.J., MacCuish, A.C. & Frier, B.M. (2000) The care of students with insulin-treated diabetes mellitus living in university accommodation: scope for improvement? Diabetic Medicine, 17, 70–73. Thompson, C.J., Cummings, J.F.R., Chalmers, J., Gould, C. & Newton, R.W. (1996) How have patients reacted to the implications of the DCCT? Diabetes Care, 19, 876–878. Thompson, C., Connacher, A., Rewt, D. & Newton, R. (1999) Exercise in children and adolescents. In: B. Burr & D. Nagi (eds), Exercise and Sport in Diabetes, pp. 83–96. John Wiley & Sons Ltd, Chichester. Waclawski, E.R. (1989) Employment and diabetes: a survey of the prevalence of diabetic workers known by occupation physicians, and the restrictions placed on diabetic workers in employment. Diabetic Medicine, 6, 16–19. Wasserman, D.H. & Zinman, B. (1994) Exercise in individuals with IDDM. Diabetes Care, 17, 924– 927. Watson, W.A., Currie, T., Lemon, J.S. & Gold, A.E. (2004) Driving and insulin-treated diabetes: who knows the rules and recommendations? Practical Diabetes International, 24, 201–206. Weinger, K., Kinsley, B.T., Levy, C.J. et al. (1999) The perception of safe driving ability during hypogly- cemia in patients with type 1 diabetes mellitus. American Journal of Medicine, 107, 246–253. Wredling, R.A.M., Theorell, P.G.T., Roll, H.M., Lins, P.E.S. & Adamson, U.K.C. (1992) Psychosocial state of patients with IDDM prone to recurrent episodes of severe hypoglycemia. Diabetes Care, 15, 518–521. Wright, R.J., Barclay, J.I.P., Plastow, B. & Frier, B.M. (2008) Management of diabetes in police custody: a liaison initiative between a diabetes specialist service and the police force. Practical Diabetes Inter- national, 25, 72–75. Yardley, J.E., Kenny, G.P., Perkins, B.A. et al. (2012) Effects of performing resistance exercise before versus after aerobic exercise on glycemia in Type 1 diabetes. Diabetes Care, 35, 669–675. Index

academic examinations 363–364 Diamicron MR Controlled Evaluation) acarbose 106 251, 272, 274 ACCORD trial (Action to Control aetiology see causes and risk factors Cardiovascular Risk in Diabetes) 272, age 273, 274 CGM interpretation 238 ACE (angiotensin-converting enzyme) 84–86, counterregulation 50–51 239 hypoglycaemia risk 89, 205–206 acquired hypoglycaemia syndromes see symptom variations 31 counterregulation, defective; impaired see also children; elderly people awareness of hypoglycaemia aggressive behaviour 30, 174, 306, 325 ACTH (adrenocorticotrophic hormone) 11 air travel 359–360 acute myocardial infarction 273, 278 alcohol 76, 167–168, 239, 353 Addison’s disease 54, 89, 175, 358, 364 American Diabetes Association (ADA), adipose tissue 2, 3 definition of hypoglycaemia 64, see also lipolysis 147–148 adolescents amino acids 177 cognitive function 292–293 amygdala 134 glycaemic control 203–204 angina 277–278 social and sporting activities 338, 352, 353 angiotensin-converting enzyme (ACE) 84–86, adrenaline 239 in diabetes 53–54, 55, 83 antecedent hypoglycaemia 32, 82–83, 130 children 50, 206 cardiac autonomic function 278–279 IAH 126–127 cognitive impairment 130 nocturnal hypoglycaemia 100 counterregulatory failure 53–54, 55–58, pregnancy 220 130 type 2 diabetes 236 exercise following 77, 337–338 normal physiology 8, 13–14, 50 and IAH 130–134 in the elderly 233 in type 2 diabetes 235 exercise and 335, 336 anticonvulsants 307 QT interval 270 anxiety 28, 40, 117 adrenocorticotrophic hormone (ACTH) 11 see also fear of hypoglycaemia ADVANCE trial (Action in Diabetes and appetite, effect of insulin 153 Vascular Disease: Preterax and arrest and imprisonment 364–365

Hypoglycaemia in Clinical Diabetes, Third Edition. Edited by Brian M. Frier, Simon R. Heller, and Rory J. McCrimmon. © 2014 John Wiley & Sons, Ltd. Published 2014 by John Wiley & Sons, Ltd. 370 Index arterialised blood glucose 64 blood glucose levels artificial pancreas (closed loop systems) 108, bedtime, predictive of nocturnal 176, 193, 211, 342 hypoglycaemia 103, 219 asymptomatic hypoglycaemia defining hypoglycaemia 63–64, 147–148, associated with IAH 120 171, 172 definition 65 in children 199 frequency 69–71, 74 during exercise 336–337, 350 nocturnal 70, 71, 101, 103 and IAH 51, 120, 123–126, 137–138 atherogenesis 276 monitoring atrial fibrillation 279 capillary blood 169, 185 attentional control 40 continuous see continuous glucose auditory processing 39–40 monitoring automatism 306, 354 before driving 357, 358–359 autonomic nervous system 10 in IAH 137–138, 191–192 dead-in-bed syndrome 268, 270–271 investigation of asymptomatic HAAF 54, 83–84, 127, 235–236 hypoglycaemia 69 IAH 124, 126–127 during travel 359 neuropathy 88, 126–127, 268 in type 2 diabetes 158 response to hypoglycaemia 9–11, 14–18 postmortem 264 cardiac function 14, 276–277, 278–279 relationship to symptoms of hypoglycaemia during sleep 100–101 26–27, 38, 64, 115, 147, 232 symptoms associated with hypoglycaemia altered glycaemic thresholds 51, 16–18, 28–29, 115–116, 202 123–126, 232, 234 in children 29–30, 202 blood pressure 14, 276 in the elderly 231, 232 blood–brain barrier 297 raised glycaemic threshold 124 bovine insulin 135 in type 2 diabetes 231, 232, 235–236 brain awareness of hypoglycaemia 24, 32–33, acute effects 102, 207, 311 116–118, 324–325 blood flow 14–15, 133, 137, 310, 316 see also impaired awareness of in children hypoglycaemia cognitive impairment 102, 206–208, 292–295 basal ganglia 286, 311 glucose metabolism 197–198 bedtime snacks 105–106, 179, 211 structural changes 208, 296–297, 311–313 behavioural responses to hypoglycaemia risk counterregulatory role 8–9, 49 41, 88, 327–329 mechanisms of defective counterregulation in IAH 138–139, 151 56–57 behavioural symptoms of hypoglycaemia 306 electrophysiology 291, 295, 306, 308–310 in children 29–30, 41 encephalopathy 307–308, 311, 317 β2-adrenergic receptors IAH pathogenesis 56, 101, 127–134, 151 agonists used to prevent hypoglycaemia long-term effects 307–308 106, 177, 212, 342 in adults 286–292, 298–300 genetics 86, 239 animal models 297–298 in IAH 126–127 in children 102, 207–208, 292–297 BGAT (Blood Glucose Awareness Training) profound hypoglycaemia 286–287, 292 41, 138, 330 severe hypoglycaemia 206–207, biochemical hypoglycaemia see asymptomatic 287–292, 298–300, 311–315 hypoglycaemia mechanisms of injury 5, 316–317 blood flow 14–16, 276–278 metabolism 5, 48 cerebral 14–15, 133, 137, 310, 316 in children 197–198 Blood Glucose Awareness Training (BGAT) effects of hypoglycaemia 5–6, 57, 127 41, 138, 330 PET studies in IAH 134 Index 371

oedema 174, 307 causes and risk factors 31, 75–89, 90 seizures 269, 306–307 acquired hypoglycaemia syndromes 82–84, structural changes 99–101, 121, 151, 237 in children 208, 296–297, 311–313 alcohol use 76, 167–168, 239, 353 in diabetes 313–314 carbohydrate absorption variability 88, 166 after profound hypoglycaemia 286, 311 in children 203–206 after severe hypoglycaemia 291–292, education programmes and 80, 81, 311–313, 314–315 150–152 see also cognitive function/impairment; endocrine disorders 89 neuroglycopenic symptoms endogenous insulin absent 86, 149 brainstem 311 exercise 76–77, 205, 239 breastfeeding 223 genetic 84–86, 239 insulin therapy 165–166 C-peptide 149 absorption variability 76, 166 in pregnancy 221 clearance variability 87–88 caffeine 32, 137, 179, 342 formulation 76, 86–87, 97–99, 135 capillary blood glucose monitoring 169, intensive 46, 77–82, 148–153, 165–166 185 new technologies 80–81, 152, 186 carbohydrate in type 2 diabetes 152–153, 236–237 consumption in acute hypoglycaemia for nocturnal hypoglycaemia 87, 97–101, 172–174, 226 204, 206 consumption during exercise 339, 340–341, in pregnancy 221–222 350 psychological 88, 151 emergency supplies 351, 357, 359, 362 renal disease 87–88 low glycaemic index 106, 178, 211, in type 2 diabetes 152–153, 236–239 340–341 central nervous system see brain variability in absorption from the GI tract cerebellum 5, 15 18, 88, 106–107, 166 cerebral blood flow 14–15, 133, 137, 310, cardiac arrhythmias 279 316 and dead-in-bed syndrome 102, 268, cerebral cortical loss 291, 311, 314 269–271 in children 208, 296, 312–313 cardiac autonomic function cerebral oedema 174, 307 effect of antecedent hypoglycaemia CGM see continuous glucose monitoring 278–279 children 197–212 response to hypoglycaemia 14, 27, 28, cognitive impairment 206–207 276–277 long-term effects 207–208, 292–295 cardiovascular disease short-term effects 102, 207 hypoglycaemia increases risk 251, counterregulation 50, 206 271–274, 275–279 diet 205, 211 strict glycaemic control reduces risk electrophysiology 295 145–146, 157 exercise 205, 211–212, 353 catecholamines fear of hypoglycaemia 208, 329 in diabetes 53–54, 55, 83 glucose metabolism 197–198 children 50, 206 hemiplegia 208 IAH 126–127 IAH 121, 202–203 nocturnal hypoglycaemia 100 intensive insulin therapy 157, 204, 210, 211 pregnancy 220 management of hypoglycaemia 209–212 type 2 diabetes 236 nocturnal hypoglycaemia 198 normal physiology 8, 13–14, 50 counterregulatory defects 206 in the elderly 233 diet 105, 211 exercise and 335, 336 effect on cognitive performance 102, 207 QT interval and 270 insulin regimens 204, 210, 211 372 Index children (contd) continuous subcutaneous insulin infusion prevalence 97, 199 (CSII) unawareness of 97, 202 air travel 359 prevalence of hypoglycaemia 199 in children 204, 210 risk factors for hypoglycaemia 203–206 compared with multiple daily injection schooling 207, 293, 362, 363–364 80–81, 184–185 structural changes in the brain 208, costs 192 296–297, 311–313 DKA and 154 symptoms of hypoglycaemia 29–30, 41, exercise and 340–341, 350–351 199–202 hypoglycaemia risk 152, 176, 185 chlorpropamide 245 nocturnal hypoglycaemia 107– closed loop delivery systems 108, 176, 193, 109 211, 342 in IAH 137 cognitive function/impairment in pregnancy 225 in academic exams 363–364 in type 2 diabetes 249 in acute hypoglycaemia 34–40, 207, 285 control, fear of loss of 326 and antecedent hypoglycaemia 130 convulsions 269, 296, 306–307 in children 206–207 core body temperature 18 long-term effects 207–208, 292–295 cornstarch, uncooked 106, 179, 211 short-term effects 102, 207 corticotrophin releasing hormone (CRH) 11, driving ability 38–39, 353–355 56 in the elderly 231, 251–252, 298–300 cortisol 8, 11, 50 IAH and 124–125, 134–135 counterregulatory defects 54, 55 long-term changes 285–300 cost issues in older adults with type 2 diabetes CSII and CGM 192 251–252, 298–300 type 2 diabetes 252–253 after profound hypoglycaemia 286, 292 counselling see patient education after recurrent severe hypoglycaemia in counterregulation type 1 diabetes 207–208, 287–290, defective 51 292–295 alcohol and 167, 239 nocturnal hypoglycaemia and 102, 207 antecedent hypoglycaemia and 53–54, in the offspring of mothers with 55–58, 130 hypoglycaemia 223 in children 50, 206 college education 338, 362, 363–364 hormonal changes 52–55, 83, 100, 167, coma 174, 286–287, 292, 311 220–221, 235, 337, 338 competition of cues 32–33, 117, 324 IAH 124 concentration, loss of, as a symptom 25, 27, intensive insulin therapy 148 28, 35 nocturnal 99–101, 206 continuous glucose monitoring (CGM) 148 in pregnancy 220–221 in children 211 in type 2 diabetes 234–236 costs 192 exercise and 8, 77, 130, 335–338 description of devices 169–171, 185 normal 6–9, 48–51 in the elderly 238 in the elderly 50–51, 233 during exercise 341–342 CRH (corticotrophin releasing hormone) 11, hypoglycaemia risk 80–81, 152, 175–176 56 in IAH 137–138, 191–192 CSII see continuous subcutaneous insulin mild hypoglycaemia 190 infusion nocturnal hypoglycaemia 71, 97, 107–109, cytochrome P450 (CYP2C9) 239 189, 190–191 in pregnancy 225 DAFNE (Dose Adjustment for Normal problems and limitations 70–71, 97, Eating) 80, 175, 330 170–171, 186–189 dangerous activities 351, 360 in type 2 diabetes 237–238 see also driving Index 373

DCCT see Diabetes Control and Complications early symptoms 25 Trial economic issues death 263–264 CSII and CGM 192 cardiovascular-related 251, 271–274, type 2 diabetes 252–253 275–279 Edinburgh Hypoglycaemia Scale 29, 34 dead-in-bed syndrome 102, 265–271 education definitions academic 207, 293, 338, 362, 363–364 hypoglycaemia concerning diabetes/hypoglycaemia biochemical 63–64, 147–148, 171, 172, CGM use 186 198–199 pre-conception counselling 224 clinical 64–65, 147 structured programmes 80, 81, 150–152, IAH 118–119 175–176, 210, 231, 330 dementia 231, 251–252, 299 warning symptoms 25–27, 28, 31, 138 denial 88, 328 EEG studies 291, 295, 306, 308–309 detemir elderly people lower weight gain 153 CGM 238 in pregnancy 225 cognitive impairment in type 2 diabetes prevention of nocturnal hypoglycaemia 231, 251–252, 298–300 106–107, 210 frequency of hypoglycaemia 240, 241 in type 2 diabetes 249 intensive insulin therapy 157 developmental delay 292 normal counterregulation 50–51, 233 Diabetes Control and Complications Trial risk of hypoglycaemia 89 (DCCT) 77–79 symptoms of hypoglycaemia 30–31, 33, cognitive function assessed in follow-up trial 230–232 (DCCT/EDIC) 289–290 see also type 2 diabetes intensive insulin therapy increased risk of electrophysiology, cerebral 291, 295, 306, hypoglycaemia 46, 79, 81, 165–166, 308–310 203 emergency hospital admission 72–74, 172, patient selection 66, 79, 290 174, 251, 252 vascular complications 145–146, 156 emotional changes 30, 40–41, 325 Diabetes UK recommendations 64 employment 326, 360–362, 363 diabetic ketoacidosis (DKA) 154 driving 356, 363 diet encephalopathy 307–308, 311, 317 bedtime snacks 105–107, 179, 211 endorphins 11 carbohydrate ingestion during exercise 339, epidemiology 65–90 340–341, 350 asymptomatic hypoglycaemia 69–71, 74 in children 205, 211 in children 199 management of acute hypoglycaemia death due to hypoglycaemia 263–264, 172–174, 226 265–267 in pregnancy 222 in the elderly 240, 241 variability in food absorption 18, 88, IAH 83, 120–123 106–107, 166 mild hypoglycaemia weight control 153, 326 type 1 diabetes 66–69, 74, 220 dipeptidyl peptidase 4 (DPP-4) inhibitors 250 type 2 diabetes 246, 248 distraction 32–33, 117, 324 nocturnal hypoglycaemia 87, 96–97, 190, dizziness 25, 27 199, 219 DKA (diabetic ketoacidosis) 154 in pregnancy 218–220 Dose Adjustment for Normal Eating (DAFNE) severe hypoglycaemia 80, 175, 330 type 1 diabetes 71–74, 121, 199, 218–219 driving 38–39, 353–359, 363 type 2 diabetes 240–241, 245, 246–247, drowsiness 28, 202 248–249 Düsseldorf education programme 80, in type 1 diabetes 65–75, 199, 218–220, 175–176 239–240 374 Index epidemiology (contd) fibrinolysis 276 in type 2 diabetes 220, 240–250, 252 fluoxetine 178 see also causes and risk factors frequency of hypoglycaemia 90 epilepsy 307 asymptomatic 69–71, 74 epinephrine see adrenaline in children 199 ethanol 76, 167–168, 239, 353 in the elderly 240, 241 EURODIAB IDDM Complications Study of IAH 120–121 81–82, 88 mild European Medicines Agency 148 type 1 diabetes 66–69, 74, 220 event-related (evoked) potentials 291, 309–310 type 2 diabetes 246, 248 everyday living 347–366 nocturnal 87, 96–97, 190, 199 driving ability 38–39, 353–359, 363 in pregnancy 218–220 education 207, 293, 338, 362, 363–364 severe employment 326, 356, 360–362, 363 type 1 diabetes 71–74, 121, 199, 218–219 exercising 334–335, 339, 349–353 type 2 diabetes 240–241, 245, 246–247, fear and embarrassment 41, 250–251, 248–249 325–327, 347–349, 353 in type 1 diabetes 65–75, 199, 218–220, imprisonment 364–365 239–240 long-distance travel 359–360 in type 2 diabetes 220, 240–250, 252 examinations (academic) 363–364 fructose 178 exenatide (exendin 4) 5, 250 exercise 334–342, 349–353 gamma(γ)-aminobutyric acid (GABA) 177 as cause of hypoglycaemia 76–77, 205, 239 gastric emptying 18, 88, 166 in children 205, 211–212, 353 gastrointestinal blood flow 15–16 glucose metabolism 8, 77, 130, 334–338 gastroparesis 88 nocturnal hypoglycaemia following 336, gender 337, 338, 341 counterregulation 51, 338 prevention of hypoglycaemia 339–342, symptoms 26, 32, 37 349–351 genetics 84–86, 239 sport 334–335, 339, 350, 351–353 long QT syndrome 270 in type 2 diabetes 5, 239 GH (Growth Hormone) 13, 50, 221 counterregulation 7–8, 54 family GHRH (Growth Hormone Releasing Hormone) fear of hypoglycaemia 328, 348 13 recognising symptoms 29–30, 203, 231 glargine (insulin) treatment of hypoglycaemia episodes 173, prevention of nocturnal hypoglycaemia 252, 325 106–107, 210 fast-acting insulin analogues see rapid-acting in type 2 diabetes 249 insulin analogues glibenclamide 241, 245 fasting state gliclazide 241 exercise during 350 glimepiride 241 hyperglycaemia after nocturnal glipizide 241 hypoglycaemia 103–105, 109, 204 glucagon normal glucose homeostasis 2–3, 197–198 alcohol and 167 fat metabolism 3, 8, 335 amino acids stimulate release 177 alcohol inhibits 167 counterregulation 7–8, 49–50 insulin inhibits 52 defective 52–53, 83, 206 fear of hypoglycaemia 41, 88, 250–251, in the elderly 233 325–330, 347–349 normal glucose homeostasis 1–4, 7–8, 49–50 in children 208, 329 pancreatic secretion 13 in pregnancy 223, 326–327 treatment of acute hypoglycaemia 173, 226 fed state 3–4 in type 2 diabetes 235 fetus, effects of maternal glucose levels 223, glucagon-like peptide 1 (GLP-1) 4, 5, 250 224 glucokinase 57, 178 Index 375 gluconeogenesis 1, 3, 8, 52, 335 hypoglycaemia increases risk 251, glucose administration in acute hypoglycaemia 271–274, 275–279 intravenous 174 strict glycaemic control reduces risk oral 172–173 145–146, 157 glucose homeostasis hemiplegia/hemiparesis 208 alcohol and 167 hepatic metabolism see liver in children 197–198 hippocampus 5, 208, 287, 310, 311, 313 in diabetic patients 46–48, 51–59 holidays 352–353, 359–360 exercise and 8, 77, 130, 334–338 hospitalisation 72–74, 172, 174, 251, 252 normal 1–9, 48–51, 75, 335 human insulin see also counterregulation dead-in-bed syndrome and 265, 266 glucose sensors in the body 9, 114–115 IAH and 86–87, 135–136 glucose transporters (GLUTs) 6, 129, 335 see also long-acting insulin analogues; and antecedent hypoglycaemia 56–57 rapid-acting insulin analogues glycaemic control hunger in children and adolescents 157, 203–204 suppressed by insulin 153 improved using CGM 186 as a symptom 25, 27, 28, 202 poor, and appearance of symptoms 64, 124 hyperglycaemia strict see strict glycaemic control effect on cognition 289, 290 treatment of IAH 136–137 fasting, after nocturnal hypoglycaemia in type 2 diabetes 238 103–105, 109, 204 glycated haemoglobin see HbA1c post-exercise 339, 340 glycogen 1, 3 hyperinsulinaemia brain glycogen supercompensation effect on glucose metabolism 52, 58 hypothesis 57 intensive insulin therapy 154 and hepatic alcohol metabolism 167 hyperinsulinaemic clamps 8, 35, 49 glycogenolysis 1, 3, 8 hypoglycaemia associated autonomic failure glycolysis 1 (HAAF) 54, 83–84, 127 grey matter loss 291, 311, 314 in type 2 diabetes 235–236 in children 208, 296, 312–313 Hypoglycaemia Fear Survey 41, 88, 329 Growth Hormone (GH) 13, 50, 221 hypothalamus 5 counterregulation 7–8, 54 blood flow 15 Growth Hormone Releasing Hormone counterregulatory response 9, 49, 57 (GHRH) 13 neuroendocrine response 11, 13 VMH as glucose sensor 9, 11, 57, 177 HAAF (hypoglycaemia associated autonomic hypothermia 18, 352 failure) 54, 83–84, 127 in type 2 diabetes 235–236 IAH see impaired awareness of hypoglycaemia haemodynamic changes 14–16, 276–278 IGF-1 (insulin-like growth factor-1) 221 cerebral 14–15, 133, 137, 310, 316 imaging see neuroimaging

HbA1c (glycated haemoglobin) impaired awareness of hypoglycaemia (IAH) hypoglycaemia risk 46, 81–82, 238 118–139 vascular complication risk 156 alcohol and 167 hearing 39–40 in children 121, 202–203 heart cognitive impairment 124–125, 134 arrhythmias 279 definition 118–119 and dead-in-bed syndrome 102, 268, diagnosis 120 269–271 driving restrictions 355–356 autonomic function EEG changes 309 effect of antecedent hypoglycaemia epidemiology 83, 120–121 278–279 human insulin 86–87, 135–136 response to hypoglycaemia 14, 27, 28, management 136–139, 151 276–277 CGM 138, 191–192 cardiovascular disease pathogenesis 139 376 Index impaired awareness of hypoglycaemia (IAH) microvascular disease 146, 155–156, 158 (contd) weight gain 153, 157–158 altered glycaemic threshold for symptom in type 2 diabetes 152–153, 157–158, 234, initiation 38, 51, 123–126, 234 236 autonomic nervous system 126–127 intermediate-acting insulin (isophane insulin) central nervous system 56, 101, 76, 151, 249 127–134, 151 internal capsule 311 progression 119 interstitial (tissue) glucose levels in CGM and risk of severe hypoglycaemia 83, 121, 70–71, 168, 169–170, 188 237 age and 238 in type 2 diabetes 121, 237 intravenous glucose 174 inflammation 276 isophane insulin 76, 151, 249 insulin counterregulatory defects in diabetes 52, 58 ketone bodies/ketosis 3 endogenous production is protective 86, 149 DKA 154 in pregnancy 221 following exercise 341 normal glucose homeostasis 1–4, 49, 75 kidney exercise 335 blood flow 16, 275 insulin pumps see closed loop systems; diabetic nephropathy 87–88, 156, 275 continuous subcutaneous insulin infusion insulin therapy lateral orbitofrontal cortex 134 absorption variability 76, 166 legal issues administration errors 76 arrest 364 in children and adolescents 157, 203–204, automatism 306, 354 210, 211 driving 354–356, 358–359, 363 exercise and 335–336, 340–342, 350–351 employment 360, 362 formulation/delivery methods 76, 86–87, leukoaraiosis 312, 315 97–99, 135–136 see also white matter changes human vs animal 135–136 lipohypertrophy/lipoatrophy 147 hypoglycaemic risk 38, 46, 77–82, 86–87, lipolysis 3, 8, 335 87–88, 148–153, 165–166, 236–237 alcohol inhibits 167 IAH 120–121, 124, 135–136, 137 insulin inhibits 52 intensive see intensive insulin therapy lispro insulin 107, 210, 249 multiple daily injections compared with CSII liver 80–81, 184–185 alcohol metabolism 167 and nocturnal hypoglycaemia 97–99, glucose homeostasis 2, 3 106–108 hepatic autoregulation 8, 9 in pregnancy and postpartum 218, 223, sulfonylurea metabolism 239 224–225 long-acting insulin analogues 176 during travel 359 lower weight gain 153 in type 2 diabetes 152–153, 157–158, in pregnancy 225 236–237, 245, 246–249, 327, 363 prevention of nocturnal hypoglycaemia insulin-like growth factor-I (IGF-I) 221 107–108, 210 intensive insulin therapy 145–159 in type 2 diabetes 249 benefits 145–146, 154–155 long QT syndrome 269–270, 271 in children and adolescents 157, 203–204, low glycaemic index foods 106, 179, 211, 210, 211 340–341 patient suitability 156–157, 158 risks 146–147 macrovascular disease see cardiovascular disease death 268, 272–274 malaise 28, 231 DKA 154 management increased hypoglycaemia 46, 77–82, acute hypoglycaemia 172–175, 226 148–153, 165–166 in children 209–212 Index 377

driving 356–358 nervousness 25, 27, 28 drugs and dietary supplements 105–107, neuroendocrine responses to hypoglycaemia 137, 177–180, 211, 342 11–14, 49–50 fear of hypoglycaemia 329–330 neuroglycopenic symptoms 28–30, 33, 34–41 IAH 136–139, 151, 191–192 in children 29–30, 202 nocturnal hypoglycaemia 105–109, 179, in the elderly 231–232 210, 211 in IAH 124–125, 134–135 in pregnancy 224–226 neuroimaging prevention of hypoglycaemia during/after children 208, 296–297, 311–313 exercise 339–342, 349–351 in diabetes 296–297, 313–314 recurrent hypoglycaemia 175–177 functional studies in IAH 134 marathon running 352 after profound hypoglycaemia 286, 311 meglitinides 241–245 after severe hypoglycaemia 291–292, memory 35, 40, 208, 286, 289, 313 311–313, 314–315 see also cognitive function/impairment neurological effects of hypoglycaemia 30, mesial temporal sclerosis 208, 296 305–318 metformin 240, 245 electrophysiology 291, 295, 306, 308–310 microvascular disease encephalopathy 307–308, 311, 317 effect of hypoglycaemia on 275, 313 hemiplegia 208 effect of intensive insulin therapy on 146, seizures 306–307 155–157, 158 structural changes in the brain 208, 286, and hyperglycaemia 145, 315 291–292, 296–297, 311–316 increases hypoglycaemia risk 87–88 see also cognitive function/impairment mild hypoglycaemia neuropathy acquired hypoglycaemia syndromes 83 autonomic 88, 126–127, 268 CGM devices 190 insulin neuritis 156 definition 65, 147 neurotransmitters 5, 11, 13, 316 modified for children 199 in management of hypoglycaemia 178 driving ability 38, 353–354 newborn infants 292 frequency NMDA (N-methyl-D-aspartate) receptors 5 type 1 diabetes 66–69, 74, 220 nocturnal hypoglycaemia 87, 96–109 type 2 diabetes 246, 248 asymptomatic 70, 71, 101, 103 in pregnancy 220, 226 causes and risk factors 87, 97–101, 204, treatment 172, 226 206 modafinil 137, 177 CGM 71, 97, 108, 189, 190–191 moderate hypoglycaemia, use of term 65, in children 198 147, 199 counterregulatory defects 206 mood changes 30, 40–41, 325 diet 105, 211 morning sickness 222 effect on cognitive performance 102, 207 mortality see death insulin regimens 204, 210, 211 motor coordination 35, 39 prevalence 97, 199 muscle unawareness of 97, 202 blood flow 16 consequences 101–102, 207 exercise 335 sudden death 102, 265–271 glucose homeostasis 2, 3 exercise preceding 336, 337, 338, 341 sympathetic activity 10–11 frequency 96–97, 190, 199, 219 myocardial infarction 273, 278 hyperglycaemia following 103–105, 108, 204 nateglinide 241–245 prediction 103, 219 nausea, in pregnancy 222 in pregnancy 219 neonates 292 prevention 105–108, 177, 179, 190–191, nephropathy 156, 275 210, 211 and risk of hypoglycaemia 87–88 in type 2 diabetes 107 378 Index

NPH (isophane) insulin 76, 151, 249 denial 88, 328 nutrition see diet fear of hypoglycaemia 41, 88, 250–251, 325–330, 347–349 older people see elderly people in children 208, 329 oral antidiabetes agents in pregnancy 223, 326–327 combined with insulin 153, 246 mood changes 30, 40–41, 325 frequency of hypoglycaemia 239, 240–246 response to IAH 138–139, 151 pharmacology 239, 244 see also cognitive function/impairment symptoms of hypoglycaemia 233–234 puberty 203–204 orbitofrontal cortex 134 oxytocin 13 QT interval, prolonged 269–270, 271 quality of life (QOL) 155, 250–251, 323–324 paediatrics see children see also everyday living pallor 16, 29 pancreas rapid eye movement (REM) sleep 100–101 neuroendocrine response 13, 49–50 rapid-acting insulin analogues 87, 184 residual beta cell function 86, 149, 221 in children 210 transplantation 152, 177 exercise and 340, 341 pancreatic polypeptide (PP) 11, 13 in IAH 139 parasympathetic nervous system 10, 11 in pregnancy 225 parents prevention of nocturnal hypoglycaemia diabetic 348 106–107, 210 of diabetic children, recognising during travel 359 hypoglycaemia 29–30, 203 in type 2 diabetes 249 patient education reaction times 35, 39 CGM use 186 rebound hyperglycaemia (Somogyi pre-conception counselling 224 phenomenon) 103–105, 109, 204 structured education programmes 80, 81, recombinant insulin 150–152, 175, 210, 231, 330 dead-in-bed syndrome and 265, 266 warning symptoms 25–27, 28, 31, 138 IAH and 86–87, 135–136 patient error 76 see also long-acting insulin analogues; pessimism 40 rapid-acting insulin analogues phobia 41, 328, 329–330 recreational activities 353 see also fear of hypoglycaemia sport 334–335, 339, 350, 351–353 physical activity see exercise reimbursement issues 192 pituitary function 11–13, 54, 89 renal blood flow 16, 275 platelets 276 renal disease, diabetic 87–88, 156, 275 police detention 364 renin–angiotensin–aldosterone system 13–14 porcine insulin 135–136 repaglinide 241–245 pre-conception counselling 224 retinopathy 146, 155, 156–157, 158, 275, 315 pre-conditioning hypothesis 57–58 risk factors for death 264 pregnancy 218–226, 326–327 cardiovascular-related 251, 271–274, 275–279 prevalence see frequency dead-in-bed syndrome 267–268 prison 364–365 risk factors for hypoglycaemia 31, 75–89, 90 profound hypoglycaemia (coma) 174–175, acquired hypoglycaemia syndromes 82–84, 286–287, 292, 311 99–101, 121, 151, 237 prolactin 13 alcohol use 76, 167–168, 239, 353 psychological effects 323–330 carbohydrate absorption variability 88, 166 attitude to intensive insulin therapy 154– in children 203–206 155, 158 education programmes and 80, 81, 150–152 awareness of symptoms of hypoglycaemia endocrine disorders 89 33, 117–118, 324–325 endogenous insulin absent 86, 149 behavioural symptoms 29–30, 41, 306 exercise 76–77, 205, 239 Index 379

genetic 84–86, 239 driving ability 38–39, 353–359, 363 insulin therapy 165–166 education 207, 293, 338, 362, 363–364 absorption variability 76, 166 employment 326, 356, 360–362, 363 clearance variability 87–88 exercising 334–335, 349–353 formulation 76, 86–87, 97–99, 135–136 fear and embarrassment 326, 347–349, 353 intensive 38, 46, 77–82, 148–153, imprisonment 364–365 165–166 long-distance travel 359–360 new technologies 80–81, 152, 186 somatostatin 13 in type 2 diabetes 152–153, 236–237 Somogyi phenomenon 103–105, 109, 204 for nocturnal hypoglycaemia 87, 97–101, speech difficulties 28, 35 204, 206 splanchnic blood flow 15–16 in pregnancy 221–222 sport 334–335, 339, 350, 351–353 psychological 88, 151 Stockholm Diabetes Intervention Study (SDIS) renal disease 87–88 79 in type 2 diabetes 152–153, 236–239 stomach, emptying 18, 88, 166 road traffic accidents 354–355, 358 strict glycaemic control benefits 145–146, 154–155 safety issues cardiovascular disease and 146, 157, driving 38–39, 353–358, 363 272–274 sport 351 in children and adolescents 157, 203–204, workplace 326, 360 210, 211 saxagliptin 250 IAH and 38, 124, 128, 136 schools 207, 293, 362, 363–364 patient suitability 156–157, 158 scoring systems for symptoms 29, 34 risks 146–147 seizures 269, 296, 306–307 DKA 154 selective search 324 increased hypoglycaemia 46, 77–82, selective serotonin reuptake inhibitors 178 148–153, 165–166, 238 severe hypoglycaemia microvascular disease 146, 155–156, 158 definition 65, 147, 199 weight gain 153, 157–158 frequency 71–74 in type 2 diabetes 157–158, 238, 272–274 in children 199 structured education programmes 80, 81, and IAH 83, 121, 237 150–152, 175–176, 330 in pregnancy 218–219 for childhood diabetes 210 in type 2 diabetes 240–241, 245, for the elderly 231 246–247, 248–249 sudden death 102, 265–271 treatment 173–174, 226 sulfonylureas sex (gender) frequency of hypoglycaemia 239, 240–241, counterregulation 51, 338 246 symptoms 26, 32, 37 pharmacology 239, 244 sexual intercourse 353 symptoms of hypoglycaemia 233–234 shaking (tremor) 16–18, 25, 27, 28, 29, 202 supine posture 100 short-acting insulins 107, 139 sweating 11, 16, 25, 28, 29 see also rapid-acting insulin analogues sympathetic nervous system 9–11, 16 sitagliptin 250 symptoms skin autonomic 16–18, 28–29, 115–116, 124, reactions to hypoglycaemia 11, 16 202, 231, 232 reactions to insulin therapy 147 blood glucose levels and 26–27, 38, 64, sleep 115, 147, 232 effect of nocturnal hypoglycaemia on 207 in children 29–30, 41, 199–202 as a risk factor for nocturnal hypoglycaemia clinical definitions of hypoglycaemia 87, 100–101, 206 64–65, 147 smoking 89 diminished awareness see impaired social aspects 347–366 awareness of hypoglycaemia 380 Index symptoms (contd) risk factors for hypoglycaemia 152–153, in the elderly 30–31, 33, 230–232 236–239 neuroglycopenic 28–30, 33, 34–41, 115, symptoms 124–125, 202, 231–232 compared with type 1 diabetes 233–234 neurological 231–232, 306 in the elderly 30–31, 33, 230–232 non-specific 28, 116–117, 202, 231 in type 2 diabetes compared with type 1 UK Hypoglycaemia Study Group 71, 171, diabetes 233–234 243, 246, 247 warning of hypoglycaemia 23–34, UK Prospective Diabetes Study (UKPDS) 116–118, 231–232, 324–325 Group 46, 146, 242, 245 altered glycaemic thresholds 51, unawareness of hypoglycaemia see impaired 123–126, 232, 234 awareness of hypoglycaemia systemic mediator theory 55–56 university education 338, 362, 363–364 urocortins 1–3 (UCN1–3) 56 tachycardia 14, 279 teenagers vacations 352–353, 359–360 cognitive function 292–293 VADT (Veterans Affairs Diabetes Trial) 272, glycaemic control 203–204 274 social and sporting activities 338, 352, 353 vascular disease teratogenicity of maternal hypoglycaemia 223 effect of hypoglycaemia on 251, 271–279, terbutaline 106, 178, 212, 342 313 thalamus 5, 134, 208, 296 effect of intensive insulin therapy on theophylline 137, 178 145–146, 155–157, 158 thrombosis 276 and hyperglycaemia 145, 315 tiredness 28, 202 increases hypoglycaemia risk 87–88, tolerance of hypoglycaemia 57–58 238–239 transplantation, pancreatic 152, 177 vasopressin 13 travel 359–360 venous blood glucose 64 driving 38–39, 353–359, 363 ventromedial nucleus of the hypothalamus treatment see management (VMH) 9, 11, 57, 177 tremor 16–18, 25, 27, 28, 29, 202 verbal skills 28, 35 triglyceride metabolism 3, 8 vertigo (dizziness) 25, 27 type 2 diabetes 230–253 Veterans Affairs Diabetes Trial (VADT) 272, cardiovascular disease 251, 272–274 274 cognitive impairment 231, 251–252, vildagliptin 250 298–300 visual acuity 39 counterregulation 234–236 VMH (ventromedial nucleus of the driving 359, 363 hypothalamus) 9, 11, 57, 177 economic costs 252–253 vomiting, in pregnancy 222 emergency admissions 251, 252 employment 361 warning symptoms 23–34, 116–118 exercise and 5, 239 altered glycaemic thresholds 51, 123–126, fear of hypoglycaemia 250–251, 327 232, 234 frequency of hypoglycaemia 220, 240–250, in the elderly 231–232 252 weakness 27, 28, 40 IAH 121, 237 weight gain 326 insulin therapy 152–153, 157–158, intensive insulin therapy 153, 157–158 236–237, 245, 246–249, 327, 363 white matter changes 291, 311–312, 315 cardiovascular-related mortality 272–274 in children 208, 292, 296 nocturnal hypoglycaemia 107–108 workplace issues 326, 360–362, 363 in pregnancy 220 driving 356, 363