Prevention of Type 2 Diabetes

Derek LeRoith Editor

From Science to Therapy Editor Derek LeRoith, MD, PhD Department of Medicine Mount Sinai School of Medicine New York, NY, USA

ISBN 978-1-4614-3313-2 ISBN 978-1-4614-3314-9 (eBook) DOI 10.1007/978-1-4614-3314-9 Springer New York Heidelberg Dordrecht London

Library of Congress Control Number: 2012936664

© Springer Science+Business Media New York 2012 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifi cally the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfi lms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. Exempted from this legal reservation are brief excerpts in connection with reviews or scholarly analysis or material supplied specifi cally for the purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the work. Duplication of this publication or parts thereof is permitted only under the provisions of the Copyright Law of the Publisher’s location, in its current version, and permission for use must always be obtained from Springer. Permissions for use may be obtained through RightsLink at the Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specifi c statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. While the advice and information in this book are believed to be true and accurate at the date of publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein.

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com) Preface

Epidemiological data demonstrate very convincingly that there is a world-wide epidemic of obesity. This epidemic has driven a similar increase in cases of metabolic syndrome and eventually Type 2 diabetes. Furthermore, a very worrisome increase in obesity and Type 2 diabetes is occurring in the adolescence. The health risks associated with these epidemics are well known and indeed affect most organs and tissues of the human body. A critical point of discussion has been how to deal with the obesity epidemic and therefore prevent the increase in Type 2 diabetes with all its rami ﬁ cations. This book has been complied to discuss the various aspects of both the understanding of the pathophysiology and ways to implement prevention. It deals with the epidemic in adults as well as children, it discusses local, national, and international programs set up to prevent diabetes and also the recent trials using changes in lifestyle and certain oral medications that were shown to either prevent, delay, or suppress the onset of Type 2 diabetes. The chapters are all brought to the reader by world experts and we believe that the reader will ﬁ nd valuable information in the book.

New York, NY, USA Derek LeRoith

Acknowledgement

My thanks to the authors for their contributions.

Derek Leroith

vii

Contents

1 Prevention of Type 2 Diabetes; from Science to Therapies ...... 1 Emily Jane Gallagher and Derek LeRoith 2 Pathophysiology: Loss of b-Cell Function ...... 11 Ele Ferrannini and Andrea Mari 3 Pathophysiology of Insulin Resistance: Implications for Prevention ...... 31 Shamsa Ali and Vivian A. Fonseca 4 Epidemiology Including Youth Through Adulthood and Socioeconomic Impact ...... 41 Helen Looker 5 Prediabetes Genes in Pima and Amish ...... 61 Leslie J. Baier 6 Predicting Diabetes ...... 81 Rachel Dankner and Jesse Roth 7 Screening for Prediabetes and Diabetes ...... 103 Amir Tirosh 8 Neuropathy in Prediabetes and the Metabolic Syndrome ...... 117 Aaron I. Vinik and Marie-Laure Nevoret 9 Interventional Trials to Prevent Diabetes: Diabetes Prevention Program ...... 143 Vanita R. Aroda and Robert E. Ratner 10 Decreasing Postprandial Plasma Glucose Using an a-Glucosidase Inhibitor in Subjects with IGT for the Prevention of Type 2 Diabetes Mellitus: The STOP-NIDDM Trial ...... 167 Jean-Louis Chiasson, Markku Laakso, and Markolf Hanefeld

ix x Contents

11 Da Qing, Finnish DPP, Tripod, and Dream: Lifestyle and Thiazolidinediones in the Prevention of Diabetes ...... 189 Mariela Glandt and Zachary Bloomgarden 12 Community Approaches to Diabetes Prevention ...... 203 Ann Albright and David Williamson 13 Think Locally, Act Locally, Extend Globally: Diabetes Prevention Through Partnerships with Local Communities ...... 221 Carol R. Horowitz and Brett Ives 14 Global Challenge in Diabetes Prevention from Practice to Public Health ...... 239 Peter E.H. Schwarz

Index ...... 251 Contributors

Ann Albright , PhD, RD Division of Diabetes Translation, Centers for Disease Control and Prevention , Atlanta , GA , USA Shamsa Ali , MD Department of Medicine, Section of Endocrinology, Tulane University Health Science Center , New Orleans , LA , USA Vanita R. Aroda , MD MedStar Health Research Institute , Hyattsville , MD , USA Georgetown University School of Medicine , Washington , DC , USA Leslie J. Baier , PhD Diabetes Molecular Genetics Section, Phoenix Epidemiology and Clinical Research Center, National Institute of Diabetes, Digestive and Kidney Diseases , National Institutes of Health , Phoenix , AZ , USA Zachary Bloomgarden , MD Department of Medicine , Mount Sinai School of Medicine , New York , NY , USA Jean-Louis Chiasson , MD Department of Medicin, Centre de recherche du Centre hospitalier de l’Université de Montréal (CRCHUM), Université de Montréal , Montréal , QC , Canada Rachel Dankner , MD, MPH Unit for Cardiovascular Epidemiology, The Gertner Institute for Epidemiology and Health Policy Research, Sheba Medical Center , Tel Hashomer , Israel Department of Epidemiology and Preventive Medicine, Sackler Faculty of Medicine, School of Public Health , Tel Aviv University, Ramat Aviv, Tel Aviv , Israel Patient Oriented Research, The Feinstein Institute for Medical Research, Manhasset, North Shore, New York Ele Ferrannini , MD Department of Internal Medicine , University of Pisa School of Medicine , Pisa , Italy Vivian A. Fonseca , MD, FRCP Department of Medicine, Section of Endocrinology , Tulane University Health Science Center , New Orleans , LA , USA

xi xii Contributors

Emily Jane Gallagher Division of Endocrinology Diabetes and Bone Disease, Mount Sinai Medical Center, New York, NY, USA Mariela Glandt , MD Medical Director, Diabetes Medical Center, Tel Aviv, Israel Markolf Hanefeld , MD, PhD Centre for Clinical Studies , Dresden , Germany Carol R. Horowitz , MD, MPH Departments of Health Evidence and Policy and Medicine , Mount Sinai School of Medicine , New York , NY , USA Brett Ives , MSN, NP, CDE Division of Endocrinology, Diabetes, and Bone Diseases, Department of Medicine, Mount Sinai School of Medicine, New York, NY , USA Markku Laakso , MD, PhD Department of Medicine , University of Eastern Finland , Kuopio , Finland Derek LeRoith, MD, PhD Division of Endocrinology Diabetes and Bone Disease, Mount Sinai Medical Center , New York , NY , USA Helen Looker , MD Division of Endocrinology Diabetes and Bone Disease , Mount Sinai Medical Center , New York , NY, USA Medical Research Institute, Wellcome Trust Centre for Molecular Medicine, Clinical Research Centre, Level 7, University of Dundee, Ninewells Hospital & Medical School , Dundee , Scotland, UK Andrea Mari , PhD Institute of Biomedical Engineering, National Research Council , Padova , Italy Marie-Laure Nevoret , MD Strelitz Diabetes Center for Endocrine and Metabolic Disorders and Neuroendocrine Unit , Eastern Virginia Medical School , Norfolk , VA , USA Robert E. Ratner , MD MedStar Health Research Institute, Georgetown University School of Medicine , Washington , DC , USA Jesse Roth , MD Laboratory of Diabetes & Metabolic Disorders, Elmezzi Graduate School of Molecular Medicine, The Feinstein Institute for Medical Research, Hofstra North Shore-LIJ School of Medicine , Manhasset , New York Endocrinology Division, Department of Medicine, Albert Einstein College of Medicine, Yeshiva University, Bronx, NY, USA Peter E. H. Schwarz , MD Abteilung Prävention und Versorgung des Diabetes, Medizinische Klinik III, Universitätsklinikum Carl Gustav Carus, Technischen Universität Dresden , Dresdon , Germany Amir Tirosh , MD, PhD Division of Endocrinology, Diabetes and Hypertension, Brigham and Women’s Hospital and Harvard School of Public Health , Boston , MA , USA Contributors xiii

Aaron I. Vinik , MD, PhD, FCP, MACP Strelitz Diabetes Center for Endocrine and Metabolic Disorders and Neuroendocrine Unit , Eastern Virginia Medical School , Norfolk , VA , USA David Williamson , PhD Hubert Department of Global Health , Rollins School of Public Health, Emory University , Atlanta , GA , USA

Chapter 1 Prevention of Type 2 Diabetes; from Science to Therapies

Emily Jane Gallagher and Derek LeRoith

Introduction

At the beginning of the twentieth century, a newly diagnosed diabetic patient had a life expectancy of 44 years. Little could be offered to treat diabetes except for dietary restriction and starvation, and death from hyperglycemic coma was common. After the discovery of insulin in 1921, the average life expectancy in patients with diabetes rapidly increased to 61 years, death rates from coma declined, while death from cardiovascular disease, gangrene, and renal complications began to increase [ 1 ] . Diabetes research, for many years to follow, focused on understanding the pathophysiology behind diabetes and its complications, and the development of new treatments and technologies to improve the care of diabetic patients. Improvements in sanitation in the early 1900s resulted in an increase in life expectancy for the entire population, and by 1933, it was already recognized that the risk of developing diabetes increased with advancing age. In addition to advancing age, it was also noted that genetics and obesity contributed to one’s risk of developing diabetes. The social changes at the time allowed easier access to food, and the transition of labor from manual to mechanical in both urban and rural societies led greater numbers of people becoming overweight. Therefore, as the population was living longer and becoming more overweight, it was predicted that a rise in the incidence of diabetes was inevitable [ 2 ] . By the end of the twentieth century, diabetes was being described as a global epidemic. Life expectancy continued to rise; the social changes leading to overconsumption of food and an increasingly sedentary lifestyle caused the rates of obesity to escalate worldwide (Fig. 1.1 ). As diabetes and its complications could be effectively managed with modern medicine, the cost of diabetes care was growing annually. Therefore, in the twenty-ﬁ rst century, the focus of diabetes care is

E. J. Gallagher • D. LeRoith, MD, PhD (*) Division of Endocrinology Diabetes and Bone Disease , Mount Sinai Medical Center , 1 Gustave L. Levy Place , New York , NY 10029 , USA e-mail: [email protected]

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 1 DOI 10.1007/978-1-4614-3314-9_1, © Springer Science+Business Media New York 2012 2 E.J. Gallagher and D. LeRoith

Fig. 1.1 Factors contributing to the increased prevalence of type 2 diabetes mellitus shifting toward diabetes prevention. In 1933, Joslin commented that diabetes is not a contagious disease and is a disease for the doctor to treat, rather than the state, the city, or the boards of health. However, given the extent of the current diabetes epidemic, professional societies, health boards, and government are becoming increasingly involved in diabetes prevention and there have been suggestions that the epidemic should be addressed in a similar manner to the outbreak of an infectious disease [3 ] .

The Scope of the Diabetes Epidemic

In 1998, the World Health Organization (WHO) predicted that the global diabetes epidemic would affect 154 million people by the year 2000 [ 4 ] . The actual rise in diabetes, however, surpassed this projection, and the most recent fi gures from the WHO suggest that, by the year 2030, more than 366 million people, or 4.4% of all adults worldwide, will have diabetes. India is expected to have the greatest number of individuals with diabetes (79.4 million), followed by China, the USA, Indonesia, and Pakistan [ 5] . The WHO estimates are more conservative than those of the International Diabetes Federation (IDF) that anticipates a rise in the number of people with diabetes to 438 million by 2030 [6 ] . It is conceivable, however, that the actual increase in diabetes may surpass even these projections, due to the aging population, the growing obesity epidemic, the increasing rates of T2DM in children and adolescents, progressive urbanization, and the rise in prevalence of T2DM in developing countries. In parallel with the increase in diabetes prevalence, the age of onset of T2DM has been decreasing. T2DM has been traditionally a disease of adulthood, but in recent 1 Prevention of Type 2 Diabetes; from Science to Therapies 3 years there has been an increase in the prevalence of T2DM in children and adolescents. While in general T1DM is still more common than T2DM in this age group, certain ethnic groups, including the Pima Indians in Arizona, have very high rates of T2DM in adolescents, and in Japanese school children T2DM is now more common than T1DM [7, 8 ] . A greater prevalence of T2DM in younger adults will result in an increased rate of T2DM in women of childbearing age, increasing the risk of congenital anomalies and neonatal complications. In Pima Indians, the offspring of women with diabetes during pregnancy were more obese than the offspring of those without T2DM and over 70% of the offspring of diabetic mothers had T2DM at 25–34 years of age [ 9] . Therefore, T2DM begets T2DM in this vicious cycle of metabolic derangement. Along with the increasing prevalence of diabetes, mortality from diabetes-related conditions has also been increasing. From 2000 to 2010, it increased from 5.2 to 6.8% and this fi gure is expected to climb further over the next 10 years [10, 11 ] . In addition to increasing mortality, diabetes is also associated with signifi cant morbidity. Diabetes is the leading cause of blindness among adults aged 20–74 years in the United States and is also the leading cause of chronic kidney disease. Neuropathy affects 60–70% of people with diabetes and more than 60% of nontraumatic amputations occur in those with diabetes. Severe periodontal disease affects more than a third of people with diabetes. Furthermore, people with diabetes have a risk of heart disease and stroke that is 2–4 times greater than those without diabetes [ 7 ] . These complications can cost an individual in life expectancy and quality of life, and cost a nation in healthcare expenses. Globally, diabetes-related healthcare costs accounted for approximately 11.1% of the total healthcare expenditure in 2010. This economic cost is expected to rise from USD 376 billion to USD 490 billion between the year 2010 and 2030 [12 ] . Medical expenses for those with diabetes are 2.3-fold higher than for those without diabetes and 2–5.5 times higher in diabetic patients with complications, compared to those with diabetes with no complications [13 ] . Of the total economic cost of diabetes in the United States, approximately a third is indirect cost, due to disability, loss of work and premature mortality [ 7 ] . With the growing prevalence of diabetes in children and adolescents, there is enormous concern that there will be great morbidity and early mortality in the young population from the development of diabetes-related complications at a younger age [ 14 ] . Therefore, preventing or delaying the onset of T2DM has become the subject of many recent studies as well as the focus of healthcare policy and national and international agencies.

Diabetes and Prediabetes

Most studies have focused on the prevention of T2DM in those at highest risk of developing the condition, speci ﬁ cally those with impaired fasting glucose (IFG) or impaired glucose tolerance (IGT). Six to ten percent of individuals with IGT alone progress to T2DM every year, while the 6-year cumulative incidence of T2DM for those with both IFG and IGT is up to 65% [ 15 ] . In the United States, 79 million 4 E.J. Gallagher and D. LeRoith

Table 1.1 Current de ﬁ nitions of diabetes ADA 2011 WHO 2011 IDF 2006 AACE/ACE 2010 Fasting plasma ³ 126 ³ 126 ³ 126 ³ 126 glucose (mg/dL) Two-hour plasma ³ 200 ³ 200 ³ 200 ³ 200 glucose during a 75 g OGTT (mg/dL) A1c ³ 6.5% ³ 6.5% Not Not recommended recommended as primary diagnostic test Additional criteria Random plasma glucose of ³ 200 mg/dL classic symptoms of hyperglycemia or hyperglycemic crisis ADA American Diabetes Association; WHO World Health Organization; IDF International Diabetes Federation; AACE American Association of Clinical Endocrinologist; ACE American College of Endocrinology

people or 35% of the population aged 20 years and older have “pre-diabetes” (IFG or IGT or both), while 50% of those over the age of 65 years have prediabetes (NIH Statistics 2011). In addition to being associated with an increased risk of developing T2DM, prediabetes confers an increased risk of cardiovascular mortality [16– 18 ] . The current diagnostic criteria for the diagnosis of diabetes from the American Diabetes Association (ADA), WHO, and the American Association of Clinical Endocrinologists/American College of Endocrinology (AACE/ACE) are outlined in Table 1.1 . All three committees endorse the use of fasting plasma glucose (FPG) and the 2 h 75 g oral glucose tolerance test (OGTT) for the diagnosis of diabetes, with repeat testing to confi rm the diagnosis in the absence of unequivocal hyperglycemia [19 ] . The ADA additionally adopted the use of A1c to diagnose diabetes in nonpregnant adults in 2010. They recommend that the A1c test should be performed using a method that is certifi ed by the National Glycohemoglobin Standardization Program (NGSP) and an assay that is standardized to the Diabetes Control and Complications Trial (DCCT) reference assay. An A1c ³ 6.5% is diagnostic of diabetes according to the current ADA criteria [ 19] . In 2011, the WHO also recommended that an A1c of ³ 6.5% be used to diagnose diabetes, provided that the tests and assays are standardized to the international reference values, and that no condition is present that impairs its accurate measurement (e.g., hemoglobinopathies, certain drugs, increased red cell turnover) [ 20 ] . The AACE/ACE position statement from 2010 does not support the use of A1c as a primary test to diagnose diabetes, but states that it may be considered an additional optional diagnostic criterion [ 21 ] . Prediabetes is de fi ned by the ADA as IFG, with a FPG level of 100–125 mg/dL (5.6–6.9 mmol/L), or IGT with a 2 h plasma glucose value of 140–199 mg/dL 1 Prevention of Type 2 Diabetes; from Science to Therapies 5

Table 1.2 Current de ﬁ nitions of prediabetes ADA WHO/IDF AACE/ACE Impaired fasting glucose (mg/dL) 100–125 110–125 100–125 Impaired glucose tolerance (plasma 140–199 140–199 140–199 glucose 2 h after 75 g oral glucose tolerance test) (mg/dL) A1c 5.7–6.4% Not recommended Not recommended ADA American Diabetes Association; WHO World Health Organization; IDF International Diabetes Federation; AACE American Association of Clinical Endocrinologist; ACE American College of Endocrinology

(7.8–11 mmol/L) after a 75 g OGTT, or an A1c level of 5.7–6.4% (Table 1.2 ). The current cutoff value for IFG was modifi ed by the ADA in 2003, reducing it from a FPG level of 110 mg/dL (6.1 mmol/L) to 100 mg/dL (5.6 mmol/L). The AACE and ACE use the same IFG (100–125 mg/dL) and IGT (140–199 mg/dL) levels as the ADA to defi ne prediabetes. The WHO and IDF defi ne IFG as 110–125 mg/dL (6.1–6.9 mmol/L) rather than 100 mg/dL (5.6 mmol/L) as defi ned by the ADA, AACE, and ACE (Table 1.1 ). The WHO uses the term “intermediate hyperglycemia” instead of prediabetes, to describe glucose levels between normoglycemia and diabetes [22 ] . The use of A1c to diagnose prediabetes was added to the criteria used to defi ne prediabetes, as individuals with A1c levels between 5.5 and 6% have a 5-year incidence of diabetes of 9–25%, and an A1c of 6–6.5% have a 5-year risk of 25–50% [ 19 ] . However, its use as a diagnostic tool for diabetes and prediabetes remains controversial. Recent studies have demonstrated that there are signifi cant discrepancies in the number of individuals classi fi ed as having prediabetes, depending on which criteria are used to defi ne the condition [ 23, 24] . Both of these studies report a low sensitivity for A1c in the diagnosis of prediabetes. Based on the 2005–2008 NHANES data, of those individuals with IGT, only 58.2% had a fasting glucose level of 100– 125 mg/dL, 23.4% had a fasting glucose of 110–125 mg/dL, and 32.3% had an A1c of 5.7–6.4% [ 24] . However, using OGTT as the “gold standard” for diagnosing diabetes is also controversial, as A1c is highly correlated with the risk of retinopathy, and therefore, although A1c may detect a different population, it is not necessarily inferior [ 25 ] . The arguments for and against using A1c for the diagnosis of diabetes and prediabetes are outlined in Table 1.3 . A1c is a more convenient than fasting glucose or an OGTT, as the patient does not have to fast or suffer the incon- venience of the 2 h OGTT. There is less intraindividual variation with the A1c related to stress and illness, when compared to plasma glucose. It has less variation over a short period of time (coef fi cient of variation, CV = 3.6%), compared to FPG (CV 5.7%) or 2 h postprandial glucose (CV = 16.6%). A1c is also a better predictor of complications than plasma glucose and is a better re fl ection of overall glycemia than a plasma glucose, which re fl ects glucose only at a particular point in time. Additionally, A1c has less preanalytic variability and the NGSP assay is regulated and standardized to the DCCT, while plasma glucose may decrease prior to analysis and the assay is not internationally standardized. However, A1c diagnoses less people 6 E.J. Gallagher and D. LeRoith

Table 1.3 Advantages and disadvantages of A1c for diagnosing diabetes and prediabetes Advantages of using A1c Disadvantages of using A1c • More convenient • Missing certain cases will lead to • Less intraindividual variability lost opportunities at prevention • Easier to follow up baseline and subsequent • More expensive readings • Limited availability of standardized • Better predictor of complications than fasting tests in many countries plasma glucose or postprandial glucose • Variation in correlation with mean • More preanalytical stability glucose in some ethnic groups • Re ﬂ ects long-term control, not a point in time • Altered A1c with certain medical • Greater testing due to more convenience, conditions, medications therefore more overall cases diagnosed • Need more research to validate range for prediabetes

with prediabetes and therefore will lead to missed opportunities to impact their future risk of developing diabetes. A1c is a more expensive assay than plasma glucose and although most developed countries have a standardized assay, labs in many developing countries do not, and point of care testing is not recommended by the ADA due to lack of standardization; therefore, doctors and other healthcare professionals need to be certain that the lab processing their blood samples has a standardized assay before using A1c as a screening test. Many factors are emerging apart from glucose levels that inﬂ uence A1c, including ethnicity, pregnancy, blood loss, hemoglobinopathies, anemias, certain medications, and chronic diseases. A1c is not recommended to diagnose diabetes in pregnant women. It is important to be aware of these limitations before using A1c to diagnose diabetes or prediabetes and so the diagnostic test of choice should be individualized for the patient [26, 27] . In addition, if A1c measurements become widely used to de ﬁ ne prediabetes, an apparent decline in the prevalence of prediabetes may result.

Obesity, Insulin Resistance, and Diabetes

One of the major driving forces behind the diabetes epidemic is the escalating obesity epidemic. Obesity rates are ever increasing due to sedentary lifestyle and caloric excess. In the United States, over a third of the population are obese, while over two thirds are overweight or obese [ 28 ] . Increasing body mass index (BMI) is associated with increased prevalence of diabetes in all ethnic groups. However, results of the NHANES III and the 1999–2004 NHANES show that the Mexican American and Non-Hispanic Black population were more likely than the Non-Hispanic White population to develop diabetes in the normal and overweight BMI categories [ 29 ] . Similarly, Asian Americans are more likely to develop T2DM than White Americans, despite having a lower BMI. In addition to a possible genetic predisposition of 1 Prevention of Type 2 Diabetes; from Science to Therapies 7 certain ethnicities to developing T2DM, Asian Americans are more likely to have greater visceral adiposity than their White counterparts for any given BMI [ 30 ] . Increased visceral adiposity is associated with insulin resistance, and insulin resistance is known to be an important factor underlying the development of T2DM [ 31, 32] . Therefore, the global spread of obesity will not affect all populations equally; certain populations such as South Asians will be at higher risk of developing T2DM, with lesser degrees of obesity [30, 33 ] .

Preventing Type 2 Diabetes

Many intervention studies have demonstrated that lifestyle modi fi cation in the setting of a clinical trial is at least as effective as pharmacological therapy for reducing the progression from prediabetes or the metabolic syndrome to T2DM. In a meta-analysis of studies examining the effect of lifestyle, diabetic medication, and antiobesity medication on the cumulative incidence of diabetes over 5 years, the number needed to treat (NNT) to prevent or delay one case of diabetes was 6.4 for lifestyle, 10.8 for antidiabetic medication, and 5.4 for orlistat [ 34 ] . Bariatric surgery has also been reported to decrease the prevalence of prediabetes and T2DM [35 ] . Therefore, treating individuals at risk for T2DM with lifestyle intervention, pharmacotherapy, or surgery can potentially delay or prevent the onset of T2DM. The increasing morbidity and mortality associated with diabetes, in addition to the rising diabetes-related healthcare expenditure, has led to the recognition of T2DM as a major public health concern. The IDF, WHO, ADA, the National Cholesterol Education Program—Third Adult Treatment Panel (NCEP-ATP III), along with representatives from every con- tinent convened in Lisbon, Portugal in 2006 to create a consensus statement on T2DM prevention. In this statement, they proposed that T2DM prevention strategies should not only be targeted toward those individuals at high risk of developing T2DM, but also the general population. Those at higher risk of developing T2DM should be identifi ed and lifestyle modifi cation strategies should be advised, with the possible addition of pharmacological agents if lifestyle modi fi cation fails to achieve the desired results. Targeting the general population should go beyond the scope of the healthcare sector; governments should be involved to establish health policy initiatives related to transportation and urban planning to promote physical activity; food pricing and advertising should promote healthy eating. Education programs need to target children and adults to raise awareness of the risk of diabetes to the whole population and to help people understand the importance of healthy eating, maintaining a healthy weight, and exercising regularly. This consensus statement implores governments to change policies in order to empower people to improve their physical activity and develop healthy eating habits [ 36 ] . Others have suggested that the epidemic of obesity and T2DM is not only related to lack of exercise and poor nutrition but also to a host of medical and environmental factors in a predisposed person that contribute to disease development. They therefore suggest that a wider approach needs to be taken to resolve the multiple causes of obesity and 8 E.J. Gallagher and D. LeRoith diabetes [ 37 ] . Overall, diabetes prevention is becoming a priority for healthcare professionals and governments. The best method of identifying those at highest risk and the best method of prevention of the disease and its complications are the subject of intense research and heated debate.

References

1. Joslin EP. Heredity and prevention. Bull N Y Acad Med. 1933;9(9):532–7. 2. Mosenthal HO. Diabetes mellitus—problems of present day treatment. Bull N Y Acad Med. 1933;9(9):505–22. 3. Matthews DR, Matthews PC. Type 2 diabetes as an ‘infectious’ disease: is this the Black Death of the 21st century? Diabet Med. 2011;28:2–9. 4. King H, Aubert RE, Herman WH. Global burden of diabetes, 1995–2025. Diabetes Care. 1998;21:1414–31. 5. Wild D, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes. Diabetes Care. 2004;27:1047–53. 6. Shaw JE, Sicree RA, Zimmet PZ. Global estimates of the prevalence of diabetes for 2010 and 2030. Diabetes Res Clin Pract. 2010;87(1):4–14. 7. Centers for Disease Control and Prevention. National diabetes fact sheet: national estimates and general information on diabetes and prediabetes in the United States, 2011. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention; 2011. 8. Kitagawa T, Owada M, Urakami T, Yamanchi K. Increased incidence of non-insulin dependent diabetes mellitus among Japanese school children correlates with an increased intake of animal protein and fat. Clin Pediatr. 1998;37:111–6. 9. Dabelea D, Knowler WC, Pettitt DJ. Effect of diabetes in pregnancy on offspring: follow-up research in Pima Indians. J Matern Fetal Med. 2000;9(1):83–8. 10. Roglic G, Unwin N, Bennett PH, Mathers C, Tuomilehto J, Nag S, et al. The burden of mortality attributable to diabetes: realistic estimates for the year 2000. Diabetes Care. 2005;28:2130–5. 11. Roglic G, Unwin N. Mortality attributable to diabetes: estimates for the year 2010. Diabetes Res Clin Pract. 2010;87(1):15–9. 12. Zhang P, Zhang X, Brown J, Vistisen D, Sicree J, Nichols G. Global healthcare expenditure on diabetes for 2010 and 2030. Diabetes Res Clin Pract. 2010;87(3):293–301. 13. IDF Diabetes Atlas. The global burden, 4th ed. Brussels: International Diabetes Federation; 2009. p. 21–7. 14. Steinberger J, Daniels SR. Obesity, insulin resistance, diabetes and cardiovascular risk in children. Circulation. 2003;107:1448–53. 15. De Vegt F, Dekker JM, Jager A, Hienkens E, Kostense PJ, Stehouwer CD, et al. Relation of impaired fasting and postload glucose with incident type 2 diabetes in a Dutch population: the Hoorn Study. JAMA. 2001;285:2109–13. 16. DECODE Study Group, European Diabetes Epidemiology Group. Is the current de ﬁ nition for diabetes relevant to mortality risk from all causes and cardiovascular and noncardiovascular diseases? Diabetes Care. 2003;26:688–96. 17. Qiao Q, Jousilahti P, Eriksson J, Tuomilehto J. Predictive properties of impaired glucose tolerance for cardiovascular risk are not explained by the development of overt diabetes during follow-up. Diabetes Care. 2003;26:2910–4. 18. Rodriguez BL, Lau N, Burchﬁ el CM, Abbott RD, Sharp DS, Yano K, et al. Glucose intolerance and 23-year risk of coronary heart disease and total mortality: the Honolulu Heart Program. Diabetes Care. 1999;22:1262–5. 19. American Diabetes Association. Standards of medical care in diabetes-2011. Diabetes Care. 2011;34 Suppl 1:S11–61. 1 Prevention of Type 2 Diabetes; from Science to Therapies 9

20. WHO Consultation. Use of glycated haemoglobin (HbA1C) in the diagnosis of diabetes mellitus. 2011. http://www.who.int/cardiovascular_diseases/report-hba1c_2011_edited.pdf . Accessed May 2011. 21. American Association of Clinical Endocrinologists Board of Directors; American College of Endocrinologists Board of Trustees. American Association of Clinical Endocrinologists/ American College of Endocrinology Statement on the use of hemoglobin A1c for the diagnosis of diabetes. Endocr Pract. 2010;16(2):155–6. 22. WHO. Defi nition and diagnosis of diabetes mellitus and intermediate hyperglycemia: report of a WHO/IDF consultation. Geneva, Switzerland: WHO Press; 2006. 23. Mann DM, Carson AP, Shimbo D, Fonseca V, Fox CS, Muntner P. Impace of HbA1c screening criteria on the diagnosis of pre-diabetes among US adults. Diabetes Care. 2010;33(10):2190–5. 24. James C, Bullard KM, Rolka DB, Geiss LS, Williams DE, Cowie CC, et al. Implications of alternative defi nitions of prediabetes for prevalence in US adults. Diabetes Care. 2011;24(2):387–91. 25. Buse JB. Screening for diabetes and prediabetes with proposed A1c-based diagnostic criteria. Diabetes Care. 2010;33(12):e174. 26. Bonora E, Tuomilehto J. The Pros and Cons of diagnosing diabetes with A1c. Diabetes Care. 2011;34 Suppl 2:S184–90. 27. Malkani S, Mordes JP. Implications of using hemoglobin A1c for diagnosing diabetes mellitus. Am J Med. 2011;124:395–401. 28. Flegal KM, Carroll M, Ogden CL, Curtin LR. Prevalence and trends in obesity among US adults, 1999–2008. JAMA. 2010;303(3):235–41. 29. Zhang Q, Wang Y, Huang ES. Changes in racial/ethnic disparities in the prevalence of type 2 diabetes by obesity level among US adults. Ethn Health. 2009;14(5):439–57. 30. Lee JWR, Brancati FL, Yeh HC. Trends in the prevalence of type 2 diabetes in Asians versus whites. Diabetes Care. 2011;34:353–7. 31. Hutley L, Prins JB. Fat as an endocrine organ: relationship to the metabolic syndrome. Am J Med Sci. 2005;330(6):280–9. 32. Reaven GM. Banting Lecture 1988. Role of insulin resistance in human disease. Diabetes. 1988;37(12):1595–607. 33. Lear SA, Humphries SK, Birmingham CL. The use of BMI and waist circumference as surrogates of body fat differs by ethnicity. Obesity. 2007;15:2817–24. 34. Gilles CL, Abrams KR, Lambert PC, Cooper NJ, Sutton AJ, Hsu RT, et al. Pharmacological and lifestyle interventions to prevent or delay type 2 diabetes in people with impaired glucose tolerance: systematic review and meta-analysis. BMJ. 2007;344(7588):299–307. 35. De la Cruz-Munoz N, Messiah SE, Arheart KL, Lopez-Mitnik G, Lipshultz SE, Livingstone A. Bariatric surgery signi fi cantly decreases the prevalence of type 2 diabetes mellitus and prediabetes among morbidly obese multiethnic adults: long term results. J Am Coll Surg. 2011;212:505–13. 36. Alberti KGMM, Zimmet P, Shaw J. International diabetes federation: a consensus on type 2 diabetes prevention. Diabet Med. 2007;24:451–63. 37. Egger J, Dixon J. Non-nutrient causes of low-grade, systemic infl ammation: support for ‘a canary in the mineshaft’ view of obesity in chronic disease. Obes Rev. 2011;12:339–45. Chapter 2 Pathophysiology: Loss of b -Cell Function

Ele Ferrannini and Andrea Mari

Introduction

The pathophysiology of prediabetes is a direct extension of the physiology of glucose control. In fact, all evidence indicates that progression from normoglycemia to dysglycemia to frank hyperglycemia occurs along a continuum not just of plasma glucose concentrations but also of underlying mechanisms. Therefore, the pathophysiology of prediabetes can be described equally well as shifts in glucose tolerance category and in terms of continuous changes in glucose parameters [1 ] . The glucose system is highly homeostatic, swings in plasma glucose concentrations rarely exceeding 3 mmol/L (54 mg/dL) in normal subjects. At any given time, the plasma glucose concentration represents the balance between entry of glucose into and exit from the circulation via cellular metabolism or excretion: excessive release or defective removal (or combinations of the two) will result in rising glucose levels. Entry and exit of glucose are subject to multiple regulatory mechanisms, with insulin and glucagon principally controlling entry and insulin governing exit. The role of the endocrine pancreas in the pathophysiology of prediabetes can therefore be reduced to the following questions: Are there changes in b -cell or a -cell function (or sensitivity to these hormones)? What consequences do these changes have for glucose homeostasis? A preliminary consideration is the unique organization of the insulin/glucagon system. For many protein and nonprotein hormones, action is modulated by at least one, often two, hierarchical hormonal feedback pathways (e.g., CRH and ACTH hormone for cortisol, GnRH and gonadotrophins for sex steroids). In these cases,

E. Ferrannini, MD (*) Department of Internal Medicine , University of Pisa School of Medicine , Via Roma, 67 , 56100 Pisa , Italy e-mail: [email protected] A. Mari, PhD Institute of Biomedical Engineering, National Research Council , Padova , Italy

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 11 DOI 10.1007/978-1-4614-3314-9_2, © Springer Science+Business Media New York 2012 12 E. Ferrannini and A. Mari sensitivity is provided by the circulating hormone concentrations acting upon speci ﬁ c hormone receptors located on target tissues as well as on the master gland of the feedback loop (e.g., the pituitary). In the case of insulin and glucagon, there is no pituitary or hypothalamic relay; target tissues control secretion directly. Thus, the circulating concentrations of substrates (mostly glucose, but also amino acids, free fatty acids, and ketone bodies), which result from insulin action on intermediary metabolism in different tissues, feed signals back to the b -cell and the a -cell. Sensitivity gating is provided by insulin and glucagon receptors on target tissues. An additional level of regulation is autocrine/paracrine in nature, i.e., insulin receptors on the b-cell and the a -cell, respectively. b -Cell function

A normal b -cell integrates multiple hormonal and substrate inputs to mount a secretory response precisely geared at limiting plasma glucose excursions [2 ] . The chain of events leading from stimulation of biosynthesis, processing, packaging, and release of the hormone is highly complex and tightly regulated at multiple steps. Therefore, it is not surprising that the repertoire of in vivo b-cell responses is ample; corre- spondingly, no single clinical test of insulin secretion captures the overall ability of b -cells to govern glucose homeostasis. However, modes of insulin secretory response can be categorized into two main groups, static and dynamic, simply on the basis of the time course of stimulation. By this criterion, static properties are those that represent adaptation to chronic or prolonged stimuli, such as fasting hyperglycemia, obesity, and insulin resistance; dynamic properties are those that determine the response to acute stimulation. Fasting insulin concentration and fasting secretion rate are the primary static parameters, refl ecting the b -cell secretory setpoint. In addition, in nondiabetic subjects the total amount of insulin released over a specifi ed period of time is directly related to fasting insulin secretion, presumably also refl ecting the level of the setpoint: Fig. 2.1 shows this covariation in a large group of individuals in whom the deconvolution technique was used to reconstruct secretion rates from plasma C-peptide concentrations [ 3 ] . Dynamic properties can be investigated in response to a variety of acute stimuli, the most popular being an intravenous glucose bolus (IVGTT), a hyperglycemic clamp, and oral glucose (OGTT) or mixed meal administration. Analysis of these clinical tests has been performed in a large number of variants, employing diverse doses and timing of the stimulus, sampling schedules, and data analysis. Time- honored empirical indices are the acute insulin response to an IVGTT or to a hyperglycemic plateau (AIR, as the sum of the incremental plasma insulin or C-peptide concentrations over the fi rst 8 min following the glucose challenge) [ 4 ] , and the insulinogenic index on the OGTT (as the ratio of plasma insulin increments at 30 min to the corresponding plasma glucose increments). It is now recognized that the use of intravenous or oral glucose as stimulus may yield different, even contrasting, information on b-cell function [ 5] . More insight into in vivo dynamics of insulin secretion 2 Pathophysiology: Loss of b-Cell Function 13

Fig. 2.1 Relationship between fasting insulin secretion rate and total insulin output (over the 2 h following the ingestion of 75 g of glucose) in nondiabetic subjects (n = 1,318). Unpublished data from the RISC study [43 ] can be gained with the use of mathematical models. We have developed and validated one such model [6 ] , which incorporates the main features of the in vitro response of isolated rodent or human islets to glucose stimulation: glucose sensitivity, rate sensitivity, and potentiation. Glucose sensitivity is the dose–response of insulin release to glucose concentrations; variably modeled (e.g., as a Michaelis–Menten function), the slope of this relationship measures the ability of b -cells to sense glucose, synchro- nize, and rev up their insulin secretion accordingly. Rate sensitivity is the ability to sense the speed of change of glucose concentrations and to further augment insulin release proportionally. Potentiation is the fact that pre-exposure to glucose enhances the response to glucose itself; this phenomenon can be also thought of as a priming effect or a glucose memory. In the isolated perfused pancreas, the timing of successive glucose applications is crucial for glucose potentiation: if the time interval between two stimulating glucose levels is too long, potentiation vanishes, if too short potentiation regresses to inhibition [ 7 ] . Each of these three dynamic properties of b -cell function can be demonstrated to be present in vivo in man. We shall illustrate their emergence and their role in prediabetes by using data from the RISC study [3 ] . This cohort of women and men of European descent, ranging in age between 30 and 60 years, were carefully phenotyped using the hyperinsulinemic euglycemic clamp technique, an IVGTT, and an OGTT. Based on the latter, the majority of the participants had normal glucose tolerance (fasting glucose <6.1 mmol/L and 2-h glucose <7.8 mmol/L), roughly 10% of them had impaired glucose tolerance (IGT, fasting glucose <7.0 mmol/L and a 2-h glucose of 7.8–11.1 mmol/L), and a small group had isolated impaired fasting glucose (IFG, fasting glucose between 6.1 and 7.0 mmol/L and 2-h glucose <11.1 mmol/L). IFG and IGT are categories of prediabetes, which sometimes are lumped together as IGR (impaired glucose regulation). 14 E. Ferrannini and A. Mari

Fig. 2.2 Plasma glucose and insulin concentrations during a standard OGTT in subjects with normal glucose tolerance (NGT), impaired fasting glycemia (IFG), or impaired glucose tolerance (IGT). Plots are mean ± SD. Unpublished data from the RISC study [43 ]

The plasma glucose and insulin excursion in response to a 75-g oral glucose load, depicted in Fig. 2.2 , are typical: an upward shift in glucose and insulin for IFG, an exaggerated and delayed response of both glucose and insulin in IGT. If the plasma insulin concentrations are plotted against the corresponding plasma glucose concentrations at each sampling time-point, one obtains loops of different shape: steep and short in the NGT group, ﬂ atter and elongated for IFG and IGT (Fig. 2.3 ). At each plasma glucose concentration, insulin secretion rate is higher in NGT than IFG or IGT; conversely, a given insulin secretion rate intercepts the IFG/IGT loops at higher plasma glucose levels than the NGT loop. This simply means that the ability of b -cells to induce a rise in plasma hormone concentrations in response to any given increment in glucose concentrations during the test is impaired in IFG/IGT as compared to NGT individuals; this is equivalent to saying that in IFG/IGT subjects glucose sensitivity (or sensing) is compromised. Also evident from these plots is that the plasma insulin levels associated with a given glucose concentration are higher on the descending than the ascending part of the loop, i.e., at later than earlier times during the test: this systematic difference is glucose potentiation. This simple method of interpreting insulin response in the context of the corresponding glucose stimulus can be used with any format of stimulation. The mathematical model uses the C-peptide concentrations to convert plasma insulin concentrations into insulin secretion rates (by deconvolution [ 6 ] ) and then calculates the slope of the relationship between insulin secretion and glycemia (Fig. 2.4 ): the mean slope over the observed plasma glucose span of each individual is termed b -cell glucose sensitivity. 2 Pathophysiology: Loss of b-Cell Function 15

Fig. 2.3 The data in Fig. 2.2 are here presented as plots of the plasma insulin against plasma glucose concentrations at each time-point of the OGTT. The loops connect the points of each group in time sequence

Fig. 2.4 The data in Fig. 2.3 are converted (by mathematical modeling) into insulin secretion rates as a function of the concomitant plasma glucose concentrations during the OGTT

Potentiation is calculated as a time-dependent modulation of the dose–response curve, but is constrained to average unity throughout the test time; the ratio of its value at 2 h to the baseline value is called potentiation factor (Fig. 2.5). Rate sensitivity is computed from the ﬁ rst derivative of plasma glucose concentrations for successive time intervals. 16 E. Ferrannini and A. Mari

Fig. 2.5 Time course of potentiation in NGT, IFG, and IGT subjects. Note that, for each series the average potentiation is constrained to average unity. The potentiation factor is therefore taken to be the ratio of the value at 2 h to that at baseline. This ratio is clearly diminished in IGT as compared to NGT or IFG

An important issue is the relationship between b -cell function and insulin sensitivity. It has been argued that insulin secretory parameters must be viewed in relation to the prevailing degree of insulin sensitivity [ 8 ] . Thus, data from frequently sampled IVGTTs have been used to show an inverse relationship between AIR and

S I (i.e., the minimal-model derived parameter of insulin sensitivity). In several publications (reviewed in [9 ] ), this reciprocal relationship has been reported to ﬁ t an equilateral hyperbola, and the product of AIR and SI has been termed disposition index: this index has been proposed to represent the inherent ability of the b -cell to cope with the extant insulin resistance. This conceptual approach has merit and but also limitations. In fact, while the mathematical relationship between the two quan- tities depends on how they are measured (and rarely is an equilateral hyperbola [ 10 ] ), the main limitation is that some parameters of b -cell function are strongly related to insulin sensitivity while others are not. For example, fasting insulin secretion and total insulin output both are inversely related to insulin sensitivity in a curvilinear fashion, with IFG/IGT subjects falling to the right of the curve of the NGT individuals (Fig. 2.6 ). This pattern is explained by the dependency of absolute insulin secretion rates on the b-cell setpoint, which, as previously noted, is an adaptive response to chronic pressure by obesity and insulin resistance; the relative position of the IFG/ IGT subjects is explained by their slightly, but signiﬁ cantly, higher plasma glucose concentrations. In contrast, dynamic parameters such as glucose sensitivity, rate sensitivity, and potentiation are largely independent of insulin sensitivity [11 ] . Therefore, a full portrait of b-cell function should include analysis of both static and dynamic parameters; separate consideration of insulin sensitivity helps assessing the 2 Pathophysiology: Loss of b-Cell Function 17

Fig. 2.6 Relationship between total insulin output over the 2 h following glucose ingestion and insulin sensitivity (as the M/I). The lines are the separate power function ﬁ t for the NGT subjects and the IFG/IGT group. The intercept of the two lines are signiﬁ cantly different (p < 0.0001). Data from the RISC Study [3 ]

relative role of b -cell dysfunction and insulin resistance in glucose intolerance, but creating composite indices of secretion and action—such as the disposition index— may be misleading depending on which b -cell function parameter is used [10 ] . b -Cell Function in Prediabetes

Modeling of the OGTT data in Figs. 2.2 and 2.3 demonstrates that b -cell glucose sensitivity is signi ﬁ cantly reduced in both IFG and IGT subjects, possibly to a slightly greater extent in the latter than in the former (Fig. 2.7 ). In contrast, fasting insulin secretion rate and total insulin output (over 2 h following glucose ingestion) are both increased in the prediabetic groups as compared to NGT individuals (Fig. 2.8). Rate sensitivity is unaltered in both dysglycemic groups, whereas potentiation is impaired in IGT but not IFG (Fig. 2.9 ). All in all, b -cell dysfunction in prediabetes is basically a problem of glucose sensing: b-cells do not adequately “read” the degree of glucose rise. It is important to recognize that the hyperglycemia that develops as a result of defective secretory dynamics eventually acts upon the b -cell, inducing augmented insulin release. As a consequence, prediabetes is a state of impaired b -cell function and 18 E. Ferrannini and A. Mari

Fig. 2.7 Box plots of b -cell glucose sensitivity in the NGT, IFG, and IGT group. p Values refer to the comparison with NGT

Fig. 2.8 Box plots of fasting insulin secretion rate and total insulin output in the NGT, IFG, and IGT group. p Values refer to the comparison with NGT 2 Pathophysiology: Loss of b-Cell Function 19

Fig. 2.9 Box plots of rate sensitivity and potentiation factor in the NGT, IFG, and IGT group. p Values refer to the comparison with NGT hyperinsulinemia/hypersecretion [ 12 ] . Ignoring the dynamic aspects of b -cell function leads to the erroneous view (also known as the Starling curve of the pancreas [ 13 ] ) that the endocrine pancreas “copes” adequately with the prevailing insulin resistance in the early stages of dysglycemia, and that once this compensation becomes insuf ﬁ cient—and absolute insulin secretion starts to fall—then hyperglycemia ensues. On the contrary, not only is a degree of b -cell incompetence present in prediabetes, but it extends to NGT [ 14 ] . As depicted in Fig. 2.10, the relationship between b-cell glucose sensitivity and glucose tolerance (as the average glucose concentration during the OGTT) describes a continuum through NGT into IFG/ IGT. In fact, b -cell glucose sensitivity varies manifold between individuals, while mean glucose levels vary three- to fourfold, thereby attesting to the vast operative range of the b-cell as a controller of glucose homeostasis. An important point is that, in prediabetic individuals the defect in glucose sensing is more severe than can be accounted for by age. As shown in Fig. 2.11 , age is consistently and independently associated with a decline in b -cell glucose sensitivity; however, IFG and IGT subject—and even more so patients with overt type 2 diabetes—fall way below the prediction of age-related glucose insensitivity. The empirical indices of b -cell function derived from either the OGTT or the IVGTT in the same population generally support the model-derived picture. 20 E. Ferrannini and A. Mari

NGT 742 IFG/IGT ) -1 273 .mM -2

.m 100 -1

37 (pmol.min ß-cell glucose sensitivity 14

5 3 3.5 4 4.7 5.5 6.4 7.4 8.6 10 11.6 13.4 Plasma glucose (mmol/L)

Fig. 2.10 Reciprocal association between b -cell glucose sensitivity and mean plasma glucose level during an OGTT. Note the log–log scale. Line of best ﬁ t and its 95% conﬁ dence intervals are shown

200

175

) 150 -1

.mM 125

-2 IFG .m

-1 100

75 IGT (pmol.min 50 T2D ß-cell glucose sensitivity 25

0 25 30 35 40 45 50 55 60 65 Age (years)

Fig. 2.11 Sex- and BMI-adjusted dependence of b -cell glucose sensitivity on age in the NGT segment of the RISC cohort (line of best ﬁ t and 95% conﬁ dence intervals). The mean ± SD values for the IFG and IGT subjects, as well as those for a group of 133 patients with overt type 2 diabetes (T2D), are largely lower than predicted by age

The OGTT-based insulinogenic index, for example, reproduces the impaired insulin response to “early” rises in glycemia of both IFG and IGT subjects. Of interest is that even fasting serum proinsulin concentrations carry a robust predictive weight. The IVGTT-derived indices, on the other hand, perform rather poorly both in accuracy and precision (Table 2.1 ). First, the glucose excursions in response to the intravenous 2 Pathophysiology: Loss of b-Cell Function 21

Table 2.1 Empirical indices of b -cell function in subjects with normal glucose tolerance (NGT), impaired fasting glycemia (IFG), or impaired glucose tolerance (IGT)a NGT (n = 1,154) IFG ( n = 32) IGT ( n = 121) p Empirical indices Insulinogenic index (OGTT) 79 [72] 53 [54] 58 [53] <0.0001 (pmol/mmol) ∂ a a AUCG (IVGTT) (mmol/L) 6.80 ± 0.07 7.56 ± 0.41 7.56 ± 0.21 <0.001 ∂ AUCI (IVGTT) (pmol/L) 90 [183] 41 [246] 89 [154] ns ∂ AUCC-pep (IVGTT) (pmol/L) 714 [606] 524 [649] 682 [538] ns ∂ ∂ AUCC-pep / AUC G (IVGTT) 106 [83] 83 [85] 90 [65] <0.02 (pmol/mmol) ∂ ∂ AUCI / AUCG (IVGTT) 14 [26] 6 [35] 11 [19] ns (pmol/mmol) ∂ ∂ a AUCsecr / AUCG (IVGTT) 472 [353] 346 [317] 392 [277] <0.03 (pmol/min m 2 mM) Proinsulin (pmol/L) 5.0 [7.0] 14.5 [12.0]a 7.0 [7.0] a ** <0.0001 a ∂AUC = incremental area-under-curve, expressed as mean incremental concentration between 0 and 8 min following the glucose bolus; ∂AUCsecr = incremental area-under-curve expressed as mean incremental insulin secretion rate (reconstructed by C-peptide deconvolution); p = p value by Kruskal–Wallis test; a p < 0.0001 vs. NGT and **p = 0.003 vs. IFG glucose bolus are signiﬁ cantly higher in the prediabetic groups, which introduces the need to measure them and adjust the insulin (or C-peptide) changes for the concomitant changes in glycemia. Second, the interindividual scatter is large and negative numbers are not unusual. Finally, and consequently, the discriminating power across groups is reduced. Fasting and postprandial plasma glucagon levels have been reported to be abnormal in IFG/IGT, either in absolute value or in the face of the prevailing plasma glucose/ insulin concentrations [15, 16 ] . Whether this re ﬂ ects an intrinsic a -cell dysfunction or is related to the b -cell dysfunction via a paracrine effect is not clear. Recent pathology data show that the higher proportion of a -cells to b -cells in the islets of some type 2 diabetic subjects is due to a decrease in b-cell number rather than an increase in a -cell number [17 ] . In the RISC cohort, fasting hyperglucagonemia is a feature of insulin resistance independently of insulin levels and glucose tolerance [ 18 ] . Likewise, release of GLP-1 and other gastrointestinal hormones is impaired in IFG [ 19 ] and IGT [20 ] , again with little information as to whether this secretory incretin defect is primary or secondary to the b -cell dysfunction. b -Cell Mass

Recent autopsy studies in relatively large groups of subjects have reexamined the question, whether and to what extent there is loss of b -cells in patients with type 2 diabetes [21, 22 ] . While both reports concluded that in long-standing diabetes there is an average loss of b -cell volume [21 ] or mass [ 22 ] of 40–50%, one study also 22 E. Ferrannini and A. Mari found a 50% reduction of b -cell volume in patients labeled as IFG [ 23 ] , whereas the other study concluded that b -cell mass is essentially preserved in patients with recent-onset type 2 diabetes. With all the technical limitations of postmortem examinations and the uncertainties about the clinical phenotype and cause of death of the study subjects, the question remains wide open. On the other hand, it is relevant to recall that, in humans undergoing subtotal pancreatectomy (~70%), mild degrees of glucose intolerance are usually the only clinical correlate postsurgery [24 ] . The clinical relevance of loss of function viz. loss of mass is best appreciated from longitudinal observations and intervention studies (see below).

Insulin Resistance

As shown in many studies, prediabetes is an insulin resistant state. When assessed by the euglycemic clamp technique (and expressed as the total amount of glucose utilized normalized by fat-free mass as well as steady-state clamp insulin concentrations), insulin sensitivity is found to be progressively impaired from NGT to IFG to IGT (Fig. 2.12 ). To emphasize the continuous nature of the relationship between insulin sensitivity and glucose tolerance, Fig. 2.13 shows the regression of insulin sensitivity on the OGTT 2-h plasma glucose concentration adjusted for gender, age, and BMI: all else being equal, insulin sensitivity decreases by 11 m mol/min kgffm nM (or ~10% of the central value in NGT subjects) per each mmol/L increase in 2-h plasma glucose concentrations. Thus, peripheral insulin resistance is an inherent metabolic feature of prediabetes independent of factors—such as sex, age, and obesity— which themselves affect insulin action. Even within the domain of NGT, subjects with higher glucose increments during a standard dynamic test such as the OGTT are more insulin resistant than individuals whose glucose excursions are lower.

Fig. 2.12 Box plots of insulin sensitivity in the NGT, IFG, and IGT group. p Values refer to the comparison with NGT 2 Pathophysiology: Loss of b-Cell Function 23

Fig. 2.13 Reciprocal association between insulin sensitivity (as the M/I) and 2-h plasma glucose concentration on a standard OGTT. The relation shown by the solid line and its 95% con ﬁ dence intervals is adjusted for center, sex, age, and BMI. Data from the RISC study [3 ]

Importantly, ethnicity may contribute to insulin resistance independent of glucose tolerance and other determinants. In a study employing the insulin clamp technique, Mexican-Americans were shown to be more insulin resistant than non-Hispanic whites regardless of whether they were NGT, IGT, or diabetic [25 ] . The contribution of insulin sensitivity and the dynamic parameters of b -cell function to the plasma glucose concentrations seen at different during a 2-h OGTT can be examined by plotting the correlation coef fi cients linking each control mechanism (insulin sensitivity, glucose and rate sensitivity, and potentiation) to the glucose concentrations measured at each time-point [11 ] . As is evident from the trajectories in Fig. 2.14 , rate sensitivity is prominent early after glucose ingestion, when glucose rises rather rapidly, then wanes; the impact of glucose sensitivity peaks midway but is signifi cant throughout, insulin sensitivity lags behind glucose sensitivity, and potentiation is only signifi cant toward the end of the test. These associations are statistical features of the data, but they do shed light on the interplay of insulin secretion and action in shaping the glycemic curve that we interpret as glucose tolerance.

Cause and Evolution of b -Cell Dysfunction in Prediabetes

The cellular mechanisms underlying the glucose “blindness” described above are still uncertain but certainly very complex, and need not be the same in every dysglycemic subject. Membrane glucose transport, initial glucose processing by glucokinase, mitochondrial ATP production, generation of intracellular signals such as calcium and c-AMP and associated electrical transduction, and insulin signaling itself each and all are potential sites of inherited or acquired abnormalities [2 ] . 24 E. Ferrannini and A. Mari

Fig. 2.14 Partial correlation coeffi cients (with inverted sign, adjusted for sex, age, and BMI) between plasma glucose concentrations during the OGTT and metabolic parameters. Lines are spline functions connecting the coef fi cients of the four metabolic parameters at the fi ve time-points during the OGTT. The gray area includes nonsigni fi cant values for the partial correlation coef fi cients. Adapted from Mari et al. [11 ]

Factors predisposing to the development of prediabetes or to progression from prediabetes to overt type 2 diabetes have been identi fi ed in epidemiological studies. Male sex, a low birthweight, obesity and weight gain, smoking, sedentariness, and a low-quality diet are risk factors in prediabetes [15 ] . Interestingly, most if not all of these factors have been shown to affect insulin sensitivity rather than primarily b -cell function [15, 26] . In contrast, most of the gene variants that have been associated with diabetes—as a phenotype—or glycemia—as a continuous trait—are involved in some aspect of b -cell function [27– 32 ] (reviewed in [33 ] ). For example, variants in the gene encoding for transcription factor-7-like 2 (TCF7L2), which strongly predicted future diabetes in two independent cohorts, are associated with impaired insulin secretion, incretin effects, and enhanced rate of hepatic glucose production. Furthermore, overexpression of TCF7L2 in human islets reduced glucose-stimulated insulin secretion [ 34 ] . Another striking example is the glucokinase mutation identi fi ed in a young girl with severe neonatal hypoglycemia, which was associated with abnormally large islets and a fi vefold decrease in the threshold for glucose-stimulated insulin secretion [ 35 ] . At present, it seems conceivable that a large number of mutations in multiple genes involved in the regulation of b -cell function may constitute the substrate for the predisposition to prediabetes and diabetes in the general population. From the pathophysiological standpoint, the two main defects responsible for loss of glucose tolerance—i.e., insulin resistance and b -cell glucose insensitivity— occur together in prediabetes as they do in overt diabetes [ 12 ] , and co-predict incident dysglycemia in NGT subjects [ 36, 37] . Thus, despite the fact that defective insulin action can occur in the presence of perfectly preserved b -cell function—as 2 Pathophysiology: Loss of b-Cell Function 25 is the case of the obese individual with NGT—and, conversely, failing b -cells can cause hyperglycemia in subjects with normal insulin sensitivity—as typi fi ed by well-controlled type 1 diabetes—in human prediabetes, it has not been possible to identify a stage when only one abnormality can be detected. Surrogate measures of b -cell dysfunction and insulin resistance coexist even in children and adolescents at enhanced risk of developing type 2 diabetes [ 38 ] . Furthermore, when IGT or IFG regress to NGT, both insulin resistance and b -cell function improve just as they deteriorate in parallel when IFG/IGT progress to overt diabetes [ 37 ] . The reason(s) for this covariance of physiological functions are imperfectly understood. There may be genetic variants or epigenetic modifi cations that impact both insulin action and aspects of b -cell function. Another possibility is insulin resistance of b -cells: like classical target tissues, b-cells are richly endowed with insulin receptors. When b-cell insulin receptors are selectively knocked out, some of the transgenic mice develop hyperglycemia with defective b -cell glucose sensing [ 39, 40 ] . Conversely, in healthy volunteers pre-exposure to high physiological hyperinsulinemia potentiates insulin release in response to intravenous glucose [ 41 ] . These fi ndings have led to the hypothesis that insulin resistance in the b -cell synergizes with insulin resistance in the periphery to produce the two classic defects of diabetes/prediabetes [42 ] . More work is needed to produce a complete proof of this predicament.

Conclusions

There appears to be little doubt that prediabetes is characterized by a loss of b -cell function qualitatively similar to that of overt type 2 diabetes but of a lesser severity. It has been argued that b -cell dysfunction is more severe in IFG while insulin resistance is more marked in IGT [43, 44 ] ; the opposite has also been claimed [45 ] . The arbitrary nature of diagnostic thresholds (e.g., IFG deﬁ ned as a fasting glucose between 6.1–6.9 and 5.7–6.9 mmol/L) viz. the continuous nature of physiological functions (cf, Figs. 2.10 and 2.13) may well explain the divergent results obtained in small-size clinical studies or in datasets of surrogate measures. Given the tight dependency of fasting glucose levels on endogenous glucose output [46 ] , it is obvious that picking subjects with a higher fasting glucose will also select for more severe hepatic insulin resistance, possibly linked with a low pre-hepatic insulin-to-glucagon ratio. By the same token, extracting subjects with low fasting but high post-OGTT glucose levels will bias the results toward insulin resistance. In fact, any mix of b-cell dysfunction and insulin resistance can be predicted by variably combining the glucose levels in Fig. 2.14 . Along the same line of reasoning, the literature is replete with analyses of predictivity of individual plasma glucose concentrations, fasting and post-OGTT [47 ] . In reality, both values carry independent predictive power [1 ] as a result of the partially different underlying physiological mechanisms. What is clinically more relevant to recall is that the glucose abnormalities are part of a constellation of subclinical abnormalities that consistently occur together in prediabetic individuals. In comparison with NGT controls, IFG/IGT subjects 26 E. Ferrannini and A. Mari

Fig. 2.15 Schematic representation of the natural history of b -cell function and insulin sensitivity. See text for further explanation

have a higher family history of diabetes, are slightly more often men than women, somewhat older, de ﬁ nitely heavier, and with a more central distribution of body fat; values of heart rate, systolic and diastolic blood pressure are higher as are serum lipid levels (LDL-cholesterol, triglycerides, and FFA) while HDL cholesterol concentrations are lower [48 ] . The clustering also suggests that insulin resistance/ hyperinsulinemia may link diabetes with clinical hypertension and dyslipidemia either mechanistically or by genetic linkage (or both). Of further note is that recent work has emphasized that the prediabetic phenotype is closely predictive of nonalcoholic fatty liver disease [49 ] . In summary (Fig. 2.15 ), prediabetes encompasses conventional diagnostic categories of IFG and IGT, but really is a bandwidth of glucose concentrations and a temporal phase over a continuum extending from conventional NGT to overt type 2 diabetes. Insulin resistance and defective glucose sensing at the b-cell are the central pathophysiologic determinants, with insulin hypersecretion acting as a compensatory mechanism. While genetic in ﬂ uences impact on b -cell function, becoming overweight is the main acquired challenge to insulin action. However, there may be inherent genetic components that bring together insulin resistance and b -cell dysfunction in the same individual, thereby signaling his/her predisposition to progression (asterisks in Fig. 2.15 ). Treatment of high glucose levels by whatever means (lifestyle intervention, hypoglycemic agents, bariatric surgery) is accompanied by recovery of b -cell function, typically rapid and occasionally complete. This phenomenon constitutes incontrovertible evidence that b -cells in dysglycemic states are stunned but mostly alive [50 ] . This forms the rationale for early and vigorous treatment of any degree of dysglycemia if the b -cell demise characterizing long- standing diabetes is to be prevented. 2 Pathophysiology: Loss of b-Cell Function 27

References

1. Ferrannini E, Massari M, Nannipieri M, et al. Plasma glucose levels as predictors of diabetes: the Mexico City diabetes study. Diabetologia. 2009;52:818–24. 2. Boitard C, Cerasi E, Efendic S, Henquin JC, Steiner DF, Ferrannini E. Novel factors in the regulation of beta-cell function. Diabetes. 2004;53 Suppl 1:S1–3. 3. Ferrannini E, Balkau B, Coppack SW, et al.; RISC Investigators. Insulin resistance, insulin response, and obesity as indicators of metabolic risk. J Clin Endocrinol Metab. 2007;92:2885–92. 4. Brunzell JD, Robertson RP, Lerner RL, et al. Relationships between fasting plasma glucose levels and insulin secretion during intravenous glucose tolerance tests. J Clin Endocrinol Metab. 1976;42:222–9. 5. Mari A, Tura A, Pacini G, Kautzky-Willer A, Ferrannini E. Relationships between insulin secretion after intravenous and oral glucose administration in subjects with glucose tolerance ranging from normal to overt diabetes. Diabet Med. 2008;25:671–7. 6. Mari A, Schmitz O, Gastaldelli A, Oestergaard T, Nyholm B, Ferrannini E. Meal and oral glucose tests for assessment of beta-cell function: modeling analysis in normal subjects. Am J Physiol Endocrinol Metab. 2002;283:E1159–66. 7. Nesher R, Cerasi E. Modeling phasic insulin release: immediate and time-dependent effects of glucose. Diabetes. 2002;51 Suppl 1:S53–9. 8. Kahn SE, Prigeon RL, McCulloch DK, et al. Quantiﬁ cation of the relationship between insulin sensitivity and beta-cell function in human subjects: evidence for a hyperbolic function. Diabetes. 1993;42:1663–72. 9. Kahn SE, Zraika S, Utzschneider KM, et al. The beta cell lesion in type 2 diabetes: there has to be a primary functional abnormality. Diabetologia. 2009;52:1003–12. 10. Ferrannini E, Mari A. Beta cell function and its relation to insulin action in humans: a critical appraisal. Diabetologia. 2004;47:943–56. 11. Mari A, Tura A, Natali A, et al.; RISC Investigators. Impaired beta cell glucose sensitivity rather than inadequate compensation for insulin resistance is the dominant defect in glucose intolerance. Diabetologia. 2010;53:749–56. 12. Ferrannini E, Gastaldelli A, Miyazaki Y, et al. Predominant role of reduced beta-cell sensitivity to glucose over insulin resistance in impaired glucose tolerance. Diabetologia. 2003;46:1211–9. 13. Stumvoll M, Häring H, Fritsche A. For debate: Starling’s curve of the pancreas—overuse of a concept? Horm Metab Res. 2003;35(7):391–5. 14. Ferrannini E, Gastaldelli A, Miyazaki Y, et al. Beta-cell function in subjects spanning the range from normal glucose tolerance to overt diabetes: a new analysis. J Clin Endocrinol Metab. 2005;90:493–500. 15. Faerch K, Borch-Johnsen K, Holst JJ, Vaag A. Pathophysiology and aetiology of impaired fasting glycemia and impaired glucose tolerance: does it matter for prevention and treatment of type 2 diabetes? Diabetologia. 2009;52:1714–23. 16. Faerch K, Vaag A, Holst JJ, et al. Impaired fasting glycaemia vs impaired glucose tolerance: similar impairment of pancreatic alpha and beta cell function but differential roles of incretin hormones and insulin action. Diabetologia. 2008;51:853–61. 17. Henquin JC, Rahier J. Pancreatic alpha cell mass in European subjects with type 2 diabetes. Diabetologia. 2011;54(7):1720–5. 18. Ferrannini E, Muscelli E, Natali A, et al. Relationship between Insulin Sensitivity and Cardiovascular Disease Risk (RISC) Project Investigators. Association of fasting glucagon and proinsulin concentrations with insulin resistance. Diabetologia. 2007;50:2342–7. 19. Færch K, Vaag A, Holst J, et al. Impaired fasting glycaemia vs impaired glucose tolerance: similar impairment of pancreatic alpha and beta cell function but differential roles of incretin hormones and insulin action. Diabetologia. 2008;51:853–61. 20. Muscelli E, Mari A, Natali A, et al. Impact of incretin hormones on beta-cell function in subjects with normal or impaired glucose tolerance. Am J Physiol Endocrinol Metab. 2006;291:E1144–50. 28 E. Ferrannini and A. Mari

21. Butler AE, Janson J, Bonner-Weir S, et al. Beta-cell defi cit and increased beta-cell apoptosis in humans with type 2 diabetes. Diabetes. 2003;52:102–10. 22. Rahier J, Guiot Y, Goebbels RM, et al. Pancreatic beta-cell mass in European subjects with type 2 diabetes. Diabetes Obes Metab. 2008;10 Suppl 4:32–42. 23. Ritzel RA, Butler AE, Rizza RA, et al. Relationship between beta-cell mass and fasting blood glucose concentration in humans. Diabetes Care. 2006;29:717–8. 24. Clark A, Jones LC, de Koning E, et al. Decreased insulin secretion in type 2 diabetes: a problem of cellular mass or function? Diabetes. 2001;50 Suppl 1:S169–71. 25. Ferrannini E, Gastaldelli A, Matsuda M, et al. In fl uence of ethnicity and familial diabetes on glucose tolerance and insulin action: a physiological analysis. J Clin Endocrinol Metab. 2003;88:3251–7. 26. Ferrannini E. Insulin resistance versus insulin de fi ciency in non-insulin-dependent diabetes mellitus: problems and prospects. Endocr Rev. 1998;19:477–90. 27. Pascoe L, Tura A, Patel SK, et al; RISC Consortium; U.K. Type 2 Diabetes Genetics Consortium. Common variants of the novel type 2 diabetes genes CDKAL1 and HHEX/IDE are associated with decreased pancreatic beta-cell function. Diabetes. 2007;56:3101–4. 28. Pascoe L, Frayling TM, Weedon MN, et al; RISC Consortium. Beta cell glucose sensitivity is decreased by 39% in non-diabetic individuals carrying multiple diabetes-risk alleles compared with those with no risk alleles. Diabetologia. 2008;51:1989–92. 29. Soranzo N, Sanna S, Wheeler E, et al. Common variants at 10 genomic loci in fl uence hemo-

globin A1 (C) levels via glycemic and nonglycemic pathways. Diabetes. 2010;59:3229–39. 30. Ingelsson E, Langenberg C, Hivert MF, et al. Detailed physiologic characterization reveals diverse mechanisms for novel genetic loci regulating glucose and insulin metabolism in humans. Diabetes. 2010;59:1266–75. 31. Dupuis J, Langenberg C, Prokopenko I, et al. New genetic loci implicated in fasting glucose homeostasis and their impact on type 2 diabetes risk. Nat Genet. 2010;42:105–16. 32. Saxena R, Hivert MF, Langenberg C, et al. Genetic variation in GIPR in ﬂ uences the glucose and insulin responses to an oral glucose challenge. Nat Genet. 2010;42:142–8. 33. Groop L, Lyssenko V. Genetic basis of beta-cell dysfunction in man. Diabetes Obes Metab. 2009;11 Suppl 4:149–58. 34. Lyssenko V, Lupi R, Marchetti P, et al. Mechanisms by which common variants in the TCF7L2 gene increase risk of type 2 diabetes. J Clin Invest. 2007;117:2155–63. 35. Kassem S, Bhandari S, Rodríguez-Bada P, et al. Large islets, beta-cell proliferation, and a glucokinase mutation. N Engl J Med. 2010;362:1348–50. 36. Walker M, Mari A, Jayapaul MK, et al. Impaired beta cell glucose sensitivity and whole-body insulin sensitivity as predictors of hyperglycaemia in non-diabetic subjects. Diabetologia. 2005;48:2470–6. 37. Ferrannini E, Natali A, Muscelli E, et al; RISC Investigators. Natural history and physiological determinants of changes in glucose tolerance in a non-diabetic population: the RISC study. Diabetologia. 2011;54:1507–16. 38. Berenson GS, Agirbasli M, Nguyen QM, Chen W, Srinivasan SR. Glycemic status, metabolic syndrome, and cardiovascular risk in children. Med Clin North Am. 2011;95:409–17. 39. Kulkarni RN, Brüning JC, Winnay JN, et al. Tissue-speciﬁ c knockout of the insulin receptor in pancreatic beta cells creates an insulin secretory defect similar to that in type 2 diabetes. Cell. 1999;96:329–39. 40. Otani K, Kulkarni RN, Baldwin AC, et al. Reduced beta-cell mass and altered glucose sensing impair insulin-secretory function in betaIRKO mice. Am J Physiol Endocrinol Metab. 2004;286:E41–9. 41. Bouche C, Lopez X, Fleischman A, et al. Insulin enhances glucose-stimulated insulin secretion in healthy humans. Proc Natl Acad Sci USA. 2010;107:4770–5. 42. Kahn CR, Brüning JC, Michael MD, Kulkarni RN. Knockout mice challenge our concepts of glucose homeostasis and the pathogenesis of diabetes mellitus. J Pediatr Endocrinol Metab. 2000;13 Suppl 6:1377–84. 2 Pathophysiology: Loss of b-Cell Function 29

43. Abdul-Ghani MA, Tripathy D, DeFronzo RA. Contributions of beta-cell dysfunction and insulin resistance to the pathogenesis of impaired glucose tolerance and impaired fasting glucose. Diabetes Care. 2006;29:1130–9. 44. Faerch K, Vaag A, Holst JJ, et al. Natural history of insulin sensitivity and insulin secretion in the progression from normal glucose tolerance to impaired fasting glycemia and impaired glucose tolerance: the Inter99 study. Diabetes Care. 2009;32:439–44. 45. Hanefeld M, Koehler C, Fuecker K, et al. Insulin secretion and insulin sensitivity pattern is different in isolated impaired glucose tolerance and impaired fasting glucose: the risk factor in Impaired Glucose Tolerance for Atherosclerosis and Diabetes study. Diabetes Care. 2003; 26:868–74. 46. Ferrannini E, Groop LC. Hepatic glucose production in insulin-resistant states. Diabetes Metab Rev. 1989;5:711–26. 47. Abdul-Ghani MA, Lyssenko V, Tuomi T, Defronzo RA, Groop L. The shape of plasma glucose concentration curve during OGTT predicts future risk of type 2 diabetes. Diabetes Metab Res Rev. 2010;26:280–6. 48. Ferrannini E, Gastaldelli A, Iozzo P. Pathophysiology of prediabetes. Med Clin North Am. 2011;95:327–39. 49. Gastaldelli A, Kozakova M, Højlund K, et al; RISC Investigators. Fatty liver is associated with insulin resistance, risk of coronary heart disease, and early atherosclerosis in a large European population. Hepatology. 2009;49:1537–44. 50. Ferrannini E. The stunned beta cell: a brief history. Cell Metab. 2010;11:349–52. Chapter 3 Pathophysiology of Insulin Resistance: Implications for Prevention

Shamsa Ali and Vivian A. Fonseca

Introduction

Insulin resistance is a condition in which there is decreased ability of insulin to stimulate glucose disposal by muscle, adipose tissue, and liver. Insulin resistance (IR) is characterized by decreasing sensitivity of target tissues to the action of insulin, elevated blood glucose concentration, and increased hepatic production of atherogenic lipids. IR is associated with declining insulin production by the pancreas, the emergence of T2DM, and increasing risk of cardiovascular disease (CVD). Insulin resistance has a major contribution to the pathogenesis of T2DM. Insulin resistance was present in 83.9% individuals with T2DM and 65.9% of those with IGT [1 ] . Impaired fasting glucose (IFG) and impaired glucose tolerance (IGT) are intermediate states between individuals with normal glucose tolerance and those who have T2DM [2 ] . There is a difference between the site of IR in patients with IFG and IGT. The main site of insulin resistance in subjects with IFG is liver whereas there is moderate to severe IR in muscle in individuals with IGT [2 ] . Clinical trials have shown that lifestyle changes to promote weight loss and medical therapy with insulin-sensitizing agents can reduce the likelihood of progression from early stages of IR to T2DM. Pharmacologic therapies should focus not only on promoting weight loss but also on improving the cardiometabolic risk in patients with and without diabetes. The social and economic changes within the last century led to an increase of sedentary lifestyle associated with positive energy balance and increased health risks. The parallel escalation in the epidemics of type 2 diabetes (T2DM) and obesity, both of which are characterized by defects of insulin action, led researches to investigate the potential role of insulin resistance (IR) in the pathogenesis of common disorders in association with the resulting health defects.

S. Ali , MD • V. A. Fonseca , MD, FRCP (*) Department of Medicine, Section of Endocrinology , Tulane University Health Science Center , 1430 Tulane Avenue, SL-53 , New Orleans , LA 70112 , USA e-mail: [email protected]

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 31 DOI 10.1007/978-1-4614-3314-9_3, © Springer Science+Business Media New York 2012 32 S. Ali and V.A. Fonseca

It is estimated that 10–25% of the general population presents with some degree of IR [ 2] . The ability to estimate the severity of IR is important to understand its pathogenesis, to examine the epidemiology, and to assess the effects of any intervention. The scope of this chapter is to cast light on the pathophysiology, detrimental effects of IR on metabolism and body systems as well as to highlight the current and future therapeutic approaches related to prevention of IR.

Mechanism of Insulin Resistance

Circulating insulin binds to the alpha subunit of the insulin receptor which is a trans-membrane tyrosine kinase, that leads to autophosphorylation of the beta subunit of the receptor [3 ] . Binding of insulin to its receptor leads to activation of tyrosine kinase activity of the beta subunit of the insulin receptor. It causes activation of insulin signaling pathways. There are two major pathways involved in insulin receptor signaling namely, MAP (mitogen-activated protein) kinase and the PI3-K (phosphatidylinositol 3-kinase), which results in most of the biologic actions of insulin, e.g., protein and glycogen synthesis, glucose transport, anti-lipolysis and anti-apoptosis [3 ] . The PI3-K pathway involves insulin receptor substrates IRS-1 and IRS-2 in transmission of signals from receptor to downstream proteins [ 4 ] . Insulin induced tyrosine phosphorylation of the insulin receptor, and IRS-1 is reduced in obese nondiabetic individuals and in patients with T2DM [4 ] . In individuals with insulin resistance the target tissue fails to respond to circulating insulin levels. Insulin resistance can result from defects in binding to the insulin receptor [5 ] or defect in insulin receptor kinase activity [6 ] . The glucose transport system is a possible site for a post-receptor defect. Skeletal muscle, adipocytes, and cardiac muscle express Glut 4, which in the basal state is primarily in an intracellular vesicular location. Insulin stimulates glucose transport in these tissues by causing the recruitment of Glut 4 proteins from the intracellular pool to the plasma membrane. In the vast majority of type 2 diabetic patients, Glut 4 gene coding sequence and muscle Glut 4 protein levels are normal but insulin-stimulated translocation of Glut 4 to the plasma membrane is impaired [7, 8 ] .

Methods

The “gold standard” for assessing insulin sensitivity in vivo is hyperinsulinemic euglycemic glucose clamp technique. However, the glucose clamp is time consuming, complex, and requires intravenous infusion of insulin and frequent blood samples. The technique entails a constant intravenous infusion of insulin for several hours, while blood glucose is kept at a predetermined level by a feedback-controlled infusion of glucose. As the insulin acts to stimulate tissue glucose uptake and suppress hepatic 3 Pathophysiology of Insulin Resistance: Implications for Prevention 33 glucose output (HGO), the amount of glucose needed to maintain the target glucose level increases progressively until a steady state is reached at which time the rate of whole body glucose disposal is the sum of the glucose infusion rate and the rate of any residual output of glucose from the liver. The latter can be quanti ﬁ ed if a radioactive or stable isotope of glucose is infused during the study. The more sensitive the subject the higher the glucose disposal rate at any given glucose and insulin level. An alternative method is the minimal model analysis of a frequently sampled intravenous glucose tolerance test (FSIVGTT). It is easier to perform but still requires timed blood samples for measurement of plasma glucose and insulin for about 3 h following an intravenous glucose bolus. The glucose and insulin values are entered into a computer model to generate an index of insulin sensitivity [9 ] . Homeostasis Model Assessment of Insulin Resistance (HOMA IR) method of assessing insulin sensitivity calculates insulin sensitivity index as a product of fasting plasma insulin and blood glucose values divided by a constant [9 ] . Quantitative insulin sensitivity check index (QUICKI) is another method to assess insulin resistance. This method is simple and fast, providing a gross estimate of insulin resistance and can be used in the ofﬁ ce setting. It uses fasting glucose and insulin levels to assess insulin sensitivity [10 ] . The glucose clamp have been used to plot dose–response curve for insulin- stimulated whole body glucose uptake. Type 2 diabetic patients exhibit both a rightward shift (diminished sensitivity) and a marked decrease in the maximal rate (decreased responsiveness). The changes tend to be more pronounced in obese diabetic patients. Obesity is associated with a variable degree of insulin resistance: some subjects are mildly resistant and display only a rightward shift in their dose– response curve while others are more resistant and exhibit both a rightward shift and a decreased maximal response [11 ] . When glucose is infused intravenously or given orally 80–85% of overall insulin- stimulated glucose uptake is accounted for by skeletal muscle. Insulin has a major role in disposal of glucose at different target tissues especially skeletal muscle. Some glucose is oxidized but most is stored as muscle glycogen. In type 2 diabetes defects in both oxidative and nonoxidative glucose reductions are found, although the defect in the latter is greater [ 12 ] . A decreased rate of muscle glycogen synthesis in type 2 diabetes has been directly shown, and the magnitude of this defect correlates well with the impairment of whole body glucose uptake [ 12 ] . Studies using nuclear magnetic resonance spectroscopy to examine muscle metabolism strongly suggest that the lower rates of muscle glucose uptake and glycogen synthesis in type 2 diabetic patients are due primarily to a defect in glucose transport [ 13 ] . This does not preclude additional defects that may affect glucose metabolism. An impairment of glucose phosphorylation contributes to insulin resistance. Activation of glycogen synthase (rate-limiting enzyme for glycogen synthesis from glucose-6- phosphate) and pyruvate dehydrogenase (rate limiting for oxidation of pyruvate produced by glycolysis) is also impaired in diabetes [ 14, 15 ] Glucose transport activity in these tissues correlates well with whole body insulin sensitivity in both obese and type 2 diabetic subjects. In addition, intrahepatic and intrahepatocellular lipid accumulation is associated with obesity and may exacerbate insulin resistance. 34 S. Ali and V.A. Fonseca

Clinical Implications of Insulin Resistance

Insulin resistance is associated with number of cardiovascular risk factors including endothelial dysfunction. [ 16 ] . Endothelial dysfunction has shown to promote proco- agulant state in patients with insulin resistance including endothelial cell activation, thrombin generation, platelet aggregation, and suppression of endogenous ﬁ brinolytic substances [ 17] . Individuals from northern Manhattan study who had insulin resistance were shown to have increased risk of incident stroke among nondiabetic patients [18 ] . Cross-sectional data was obtained from the Relationship between Insulin Sensitivity and Cardiovascular Risk (RISC) study which was performed in individuals who during glucose tolerance test exhibit 1-h excursion of plasma glucose as high as individuals with impaired glucose tolerance. Individuals who have higher 1-h glucose had larger waist circumference, higher BMI, lower insulin sensitivity, higher fasting glucose, and higher insulin secretion than individuals with normal glucose tolerance [19 ] . Severity of glucose intolerance and insulin resistance in nondiabetic patients correlates with not only functional and clinical severity of heart failure but are also independent predictors of outcome. Metabolism shift from glucose to fatty acid due to insulin resistance contributes to the pathophysiological development of heart failure [20 ] . A cross-sectional study has shown that insulin resistance is independently associated with left ventricular diastolic dysfunction in patients without overt diabetes [21 ] . Distribution of body fat is important in identifying individuals with insulin resistance [22 ] . Waist circumference was found to be a more accurate method to identify individuals with insulin resistance than any other components of the metabolic syndrome [ 23] . However, there is a well-established relationship between elevated plasma FFA and development of insulin resistance. Elevated plasma FFA levels correlate with decreased insulin-stimulated glucose disposal in skeletal muscles [24, 25] . The ability of insulin to inhibit lipolysis and decrease plasma FFA concentrations is reduced in patients with insulin resistance which leads to increased rates of lipolysis and elevated plasma FFA level [ 25 ] . There is a well-established relationship between chronic elevations of FFA levels and the development of insulin resistance that has been referred to as “lipotoxicity” [26 ] . The GOCADAN study showed association between saturated content in red blood cells and insulin resistance and glucose intolerance in Alaska Inuits [27 ] . Not all the individuals with insulin resistance develop hyperglycemia. In fact the majority of obese individuals do not develop T2DM. Obese individuals with normal glucose tolerance have marked insulin resistance. There is increased intramuscular triglyceride content shown by muscle biopsy studies in obese individuals who do not have diabetes [28, 29 ] . Studies using magnetic resonance spectroscopy have shown an increase in intramyocellular fat accumulation in the skeletal muscle of obese nondiabetic individuals and it strongly correlates to muscle insulin resistance. Intramyocellular fat in skeletal muscle plays a key role in development of insulin resistance [29, 30 ] . 3 Pathophysiology of Insulin Resistance: Implications for Prevention 35

Prevention and Treatment of Insulin Resistance

Interventions implemented at an early stage to decrease insulin resistance, and improved insulin sensitivity postpones onset of diabetes and other associated comor- bidities. Simple life style modifi cations such as being more active have been shown to have an affect on insulin resistance. Physical inactivity was found to be a greater risk for the development of insulin resistance than obesity alone. A study done to evaluate the role of physical inactivity in obese individual showed that relative fat mass and cardiovascular fi tness are better predictors of insulin resistance [31 ] . Exercise not only improves insulin resistance but moderate to vigorous exercise training improves beta cell function as well [ 32] . Studies have been performed to evaluate which type of exercise has most benefi cial effect in improving insulin resistance. A randomized controlled trial was performed to observe the effects of resistance vs. aerobic exercise on insulin resistance. Hundred and thirty-six sedentary, obese older men and women were assigned in resistance, aerobic, combined resistance and aerobic or no exercise groups. Improvement in insulin resistance, measured by hyperinsulinemic-euglycemic clamp, was observed in the aerobic exercise and the combined exercise groups but not in the resistance exercise or control group [ 33 ] . Another study showed water-based exercise utilizing a combination of aerobic and resistance exercises reduces fasting insulin levels by 44% over a 12-week period [34 ] . Hence aerobic exercise seems to be more effective in improving insulin resistance. Aerobic activity should be probably combined with other type of exercise to achieve maximum benefi t in terms of reducing insulin resistance, although the optimal mix of types of exercise is unknown and therefore best left to patient preference. Medications known to improve insulin sensitivity such as metformin and thiazolidinediones (TZDs), peroxisome proliferator-activated receptor gamma (PPARgamma) agonists, can be used to reduce insulin resistance. Metformin is a useful insulin sensitizer which not only improves insulin resistance but also has favorable effects on lipids and BMI. Metformin improves glycemia by improving peripheral glucose disposal and decreasing endogenous production of glucose at the level of liver. The rate of peripheral glucose disposal under hyperinsulinemic-clamp conditions was shown to be increased with metformin and troglitazone by 13 and 54%, respectively [35 ] . TZDs improves insulin sensitivity in liver, muscle, and adipose tissue. The target of action for TZDs are the PPAR gamma receptors which are present in various tissues but most abundantly in adipose tissue. Their action results in altered adipokine release and promotes fatty acid uptake and storage. In euglycemic clamp studies TZDs have been shown to improve peripheral uptake of glucose and insulin, resulting in improved insulin sensitivity; however, there was minimal effect on hepatic glucose production [35 ] . Pioglitazone reduces insulin resistance by increasing aerobic capacity of skeletal muscle by improving high energy phosphate metabolism, fatty acid oxidation, and decreasing intramyocellular lipid content [36 ] . Some of the mechanisms by which TZDs improve insulin resistance also involves a decrease in the elevated free fatty acid levels present in insulin-resistant patients and also changes in the body distribution of adipose tissue [37 ] . By keeping lipids in adipose tissue the lipotoxicity at the level of liver and skeletal muscle is avoided. 36 S. Ali and V.A. Fonseca

The fi rst TZD, troglitazone, was taken off of market due to reports of severe liver failure. The second TZD, rosiglitazone, is still available but there is some hesitation for its use as in a meta-analysis it was shown to increase cardiovascular events. The third TZD, pioglitazone, has been shown to improve not only glycemia but also cardiovascular out comes in high-risk patients with T2DM in a large randomized controlled trial [38 ] . Another randomized controlled trial, ACTNOW, was performed to examine the effect of pioglitazone in preventing/delaying the development of T2DM. It showed improvement in insulin resistance and beta cell function as fundamental mechanisms by which it reduces conversion of IGT to T2DM [36 ] . In the BARI 2D trial effect of intensive medical therapy in patients with T2DM with stable CAD was compared with percutaneous coronary intervention (PCI) or coronary-artery bypass grafting (CABG). There was no advantage of revasculariza- tion on mortality or major cardiovascular event over intensive medical therapy. Among the medical therapies insulin-sensitizing drugs (metformin and TZDs) were superior to the insulin provision group (insulin, sulfonylureas) in frequency of hypoglycemia, weight gain, and HDL levels [39 ] . Hence insulin resistance should not only be addressed in early stages of disease but is an excellent fi rst-line strategy in patients with severe disease. Adiponectin and leptin are other potential insulin sensitizers discovered in mid-1990s. Adiponectin and leptin are two of many adipokines secreted by adipose tissue, an active endocrine tissue. Adiponectin levels are positively correlated with insulin resistance whereas leptin levels are negatively correlated [ 40 ] . Adiponectin effects multiple tissues but its receptors are present most abundantly in liver, adipose tissue, and skeletal muscle [ 41 ] . It increases fatty acid oxidation and decreases intracellular lipid and triglyceride content [42 ] . Leptin is positively correlated with the amount of body fat. An open-label study performed in patients with lipodystrophy and leptin de fi ciency showed improvement in hepatic and peripheral glucose metabolism and decreased intracellular triglyceride content in muscle and liver [ 43 ] . Understanding of the actions and pathophysiologic mechanisms of adipokines is still not well understood. However, with continuing research and better understanding it does hold promise for new pharmacotherapeutic approaches to treat insulin resistance in the near future. Weight loss surgery is well known to improve insulin resistance and is an effective treatment for T2DM. Any kind of weight loss surgery resulting in decreased fat mass affects adipokine levels leading to a favorable impact on insulin resistance [ 44 ] . In a recent study duodenojejunal bypass liner (DJBL) was shown to improve insulin resistance as well as reducing cardiovascular risk among morbidly obese patients with T2DM. Triglyceride/high-density lipoprotein (HDL) cholesterol ratio of 3.5 was used to de fi ne patients with insulin resistance. After a 6-month period, there was statistically signifi cant improvement in insulin resistance shown by signi fi cant reductions of the TG/HDL ratio from 5.75 to 4.36 (p < 0.001) and 42.6% of the patients had TG/HDL ratio lower than 3.5 [45 ] . Salicylates are known to improve glycemia by reducing insulin resistance. Salicylates reduce insulin clearance which leads to higher insulin levels [ 46 ] . 3 Pathophysiology of Insulin Resistance: Implications for Prevention 37

Aspirin treatment has been shown to reduce hepatic glucose production and improvement in peripheral glucose uptake during euglycemic-hyperglycemic clamp [46 ] . Although salicylates are cheap, easily available, and data from studies are promising there is a need for larger randomized studies of longer duration. Longer studies will also answer the question of safe and tolerable doses to achieve maximum bene ﬁ ts with no or minimum side effects.

References

1. Bonora E, Kiechl S, Willeit J, Oberhollenzer F, Egger G, Targher G, et al. Prevalence of insulin resistance in metabolic disorders: the Bruneck study. Diabetes. 1998;47(10):1643–9. 2. Abdul-Ghani MA, Tripathy D, DeFronzo RA. Contributions of beta-cell dysfunction and insulin resistance to the pathogenesis of impaired glucose tolerance and impaired fasting glucose. Diabetes Care. 2006;29(5):1130–9. 3. Le Roith D, Zick Y. Recent advances in our understanding of insulin action and insulin resistance. Diabetes Care. 2001;24(3):588–97. 4. Thies RS, Molina JM, Ciaraldi TP, Freidenberg GR, Olefsky JM. Insulin-receptor autophosphorylation and endogenous substrate phosphorylation in human adipocytes from control, obese, and NIDDM subjects. Diabetes. 1990;39(2):250–9. 5. Roach P, Zick Y, Formisano P, Accili D, Taylor SI, Gorden P. A novel human insulin receptor gene mutation uniquely inhibits insulin binding without impairing posttranslational processing. Diabetes. 1994;43(9):1096–102. 6. Grigorescu F, Flier JS, Kahn CR. Characterization of binding and phosphorylation defects of erythrocyte insulin receptors in the type A syndrome of insulin resistance. Diabetes. 1986;35(2):127–38. 7. Kruszynska YT, Olefsky JM. Cellular and molecular mechanisms of non-insulin dependent diabetes mellitus. J Investig Med. 1996;44(8):413–28. 8. Kelley DE, Mintun MA, Watkins SC, Simoneau JA, Jadali F, Fredrickson A, et al. The effect of non-insulin-dependent diabetes mellitus and obesity on glucose transport and phosphorylation in skeletal muscle. J Clin Invest. 1996;97(12):2705–13. 9. Matthews DR, Hosker JP, Rudenski AS, Naylor BA, Treacher DF, Turner RC. Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia. 1985;28(7):412–9. 10. Katz A, Nambi SS, Mather K, Baron AD, Follmann DA, Sullivan G, et al. Quantitative insulin sensitivity check index: a simple, accurate method for assessing insulin sensitivity in humans. J Clin Endocrinol Metab. 2000;85(7):2402–10. 11. Kolterman OG, Gray RS, Grif ﬁ n J, Burstein P, Insel J. Receptor and postreceptor defects contribute to the insulin resistance in noninsulin-dependent diabetes mellitus. J Clin Invest. 1981;68(4):957–69. 12. Ferrannini E, Mari A. How to measure insulin sensitivity. J Hypertens. 1998;16(7):895–906. 13. Cline GW, Petersen KF, Krssak M, Shen J, Hundal RS, Trajanoski Z, et al. Impaired glucose transport as a cause of decreased insulin-stimulated muscle glycogen synthesis in type 2 diabetes. N Engl J Med. 1999;341(4):240–6. 14. Kruszynska YT, Mulford MI, Baloga J, Yu JG, Olefsky JM. Regulation of skeletal muscle hexokinase II by insulin in nondiabetic and NIDDM subjects. Diabetes. 1998;47(7):1107–13. 15. Thorburn AW, Gumbiner B, Bulacan F, Brechtel G, Henry RR. Multiple defects in muscle glycogen synthase activity contribute to reduced glycogen synthesis in non-insulin dependent diabetes mellitus. J Clin Invest. 1991;87(2):489–95. 16. Petersen JL, McGuire DK. Impaired glucose tolerance and impaired fasting glucose—a review of diagnosis, clinical implications and management. Diab Vasc Dis Res. 2005;2(1):9–15. 38 S. Ali and V.A. Fonseca

17. Grundy SM. Hypertriglyceridemia, insulin resistance, and the metabolic syndrome. Am J Cardiol. 1999;83(9B):25F–9. 18. Rundek T, Gardener H, Xu Q, Goldberg RB, Wright CB, Boden-Albala B, et al. Insulin resistance and risk of ischemic stroke among nondiabetic individuals from the northern Manhattan study. Arch Neurol. 2010;67(10):1195–200. 19. Manco M, Panunzi S, Macfarlane DP, Golay A, Melander O, Konrad T, et al. One-hour plasma glucose identi fi es insulin resistance and beta-cell dysfunction in individuals with normal glucose tolerance: cross-sectional data from the Relationship between Insulin Sensitivity and Cardiovascular Risk (RISC) study. Diabetes Care. 2010;33(9):2090–7. 20. Mamas MA, Deaton C, Rutter MK, Yuille M, Williams SG, Ray SG, et al. Impaired glucose tolerance and insulin resistance in heart failure: underrecognized and undertreated? J Card Fail. 2010;16(9):761–8. 21. Dinh W, Lankisch M, Nickl W, Scheyer D, Scheffold T, Kramer F, et al. Insulin resistance and glycemic abnormalities are associated with deterioration of left ventricular diastolic function: a cross-sectional study. Cardiovasc Diabetol. 2010;9:63. 22. Abrahamson MJ. A 74-year-old woman with diabetes. JAMA. 2007;297(2):196–204. 23. Sierra-Johnson J, Johnson BD, Allison TG, Bailey KR, Schwartz GL, Turner ST. Correspondence between the adult treatment panel III criteria for metabolic syndrome and insulin resistance. Diabetes Care. 2006;29(3):668–72. 24. Groop LC, Barzilai N, Ratheiser K, Luzi L, Wahlin-Boll E, Melander A, et al. Dose-dependent effects of glyburide on insulin secretion and glucose uptake in humans. Diabetes Care. 1991;14(8):724–7. 25. Belfort R, Mandarino L, Kashyap S, Wirfel K, Pratipanawatr T, Berria R, et al. Dose-response effect of elevated plasma free fatty acid on insulin signaling. Diabetes. 2005;54(6):1640–8. 26. Unger RH. Lipotoxicity in the pathogenesis of obesity-dependent NIDDM. Genetic and clinical implications. Diabetes. 1995;44(8):863–70. 27. Ebbesson SO, Tejero ME, Lopez-Alvarenga JC, Harris WS, Ebbesson LO, Devereux RB, et al. Individual saturated fatty acids are associated with different components of insulin resistance and glucose metabolism: the GOCADAN study. Int J Circumpolar Health. 2010;69(4):344–51. 28. Perseghin G, Scifo P, De Cobelli F, Pagliato E, Battezzati A, Arcelloni C, et al. Intramyocellular triglyceride content is a determinant of in vivo insulin resistance in humans: a 1H-13C nuclear magnetic resonance spectroscopy assessment in offspring of type 2 diabetic parents. Diabetes. 1999;48(8):1600–6. 29. Phillips DI, Caddy S, Ilic V, Fielding BA, Frayn KN, Borthwick AC, et al. Intramuscular triglyceride and muscle insulin sensitivity: evidence for a relationship in nondiabetic subjects. Metabolism. 1996;45(8):947–50. 30. Pan DA, Lillioja S, Kriketos AD, Milner MR, Baur LA, Bogardus C, et al. Skeletal muscle triglyceride levels are inversely related to insulin action. Diabetes. 1997;46(6):983–8. 31. Kasa-Vubu JZ, Lee CC, Rosenthal A, Singer K, Halter JB. Cardiovascular fi tness and exercise as determinants of insulin resistance in postpubertal adolescent females. J Clin Endocrinol Metab. 2005;90(2):849–54. 32. Slentz CA, Tanner CJ, Bateman LA, Durheim MT, Huffman KM, Houmard JA, et al. Effects of exercise training intensity on pancreatic beta-cell function. Diabetes Care. 2009;32(10):1807–11. 33. Davidson LE, Hudson R, Kilpatrick K, Kuk JL, McMillan K, Janiszewski PM, et al. Effects of exercise modality on insulin resistance and functional limitation in older adults: a randomized controlled trial. Arch Intern Med. 2009;169(2):122–31. 34. Jones LM, Meredith-Jones K, Legge M. The effect of water-based exercise on glucose and insulin response in overweight women: a pilot study. J Womens Health (Larchmt). 2009;18(10):1653–9. 35. Inzucchi SE, Maggs DG, Spollett GR, Page SL, Rife FS, Walton V, et al. Effi cacy and metabolic effects of metformin and troglitazone in type II diabetes mellitus. N Engl J Med. 1998;338(13):867–72. 36. Yokota T, Kinugawa S, Hirabayashi K, Suga T, Takada S, Takahashi M, et al. Pioglitazone increases the aerobic capacity with improved skeletal muscle energetics in patients with insulin resistance. Diabetologia. 2010;53:S276. 3 Pathophysiology of Insulin Resistance: Implications for Prevention 39

37. Lebovitz HE, Banerji MA. Insulin resistance and its treatment by thiazolidinediones. Recent Prog Horm Res. 2001;56:265–94. 38. Dormandy JA, Charbonnel B, Eckland DJ, Erdmann E, Massi-Benedetti M, Moules IK, et al. Secondary prevention of macrovascular events in patients with type 2 diabetes in the PROactive study (PROspective pioglitAzone clinical trial in macroVascular events): a randomised controlled trial. Lancet. 2005;366(9493):1279–89. 39. BARI 2D Study Group, Frye RL, August P, Brooks MM, Hardison RM, Kelsey SF, et al. A randomized trial of therapies for type 2 diabetes and coronary artery disease. N Engl J Med. 2009;360(24):2503–15. 40. Weyer C, Funahashi T, Tanaka S, Hotta K, Matsuzawa Y, Pratley RE, et al. Hypoadiponectinemia in obesity and type 2 diabetes: close association with insulin resistance and hyperinsulinemia. J Clin Endocrinol Metab. 2001;86(5):1930–5. 41. Saha AK, Schwarsin AJ, Roduit R, Masse F, Kaushik V, Tornheim K, et al. Activation of malonyl-CoA decarboxylase in rat skeletal muscle by contraction and the AMP-activated protein kinase activator 5-aminoimidazole-4-carboxamide-1-beta-D-ribofuranoside. J Biol Chem. 2000;275(32):24279–83. 42. Chandran M, Phillips SA, Ciaraldi T, Henry RR. Adiponectin: more than just another fat cell hormone? Diabetes Care. 2003;26(8):2442–50. 43. Petersen KF, Oral EA, Dufour S, Befroy D, Ariyan C, Yu C, et al. Leptin reverses insulin resistance and hepatic steatosis in patients with severe lipodystrophy. J Clin Invest. 2002;109(10):1345–50. 44. Gumbs AA, Modlin IM, Ballantyne GH. Changes in insulin resistance following bariatric surgery: role of caloric restriction and weight loss. Obes Surg. 2005;15(4):462–73. 45. de Moura EG, Orso IR, Martins BD, Lopes GS, de Oliveira SL, Galvao-Neto MD, et al. Improvement of insulin resistance and reduction of cardiovascular risk among obese patients with type 2 diabetes with the duodenojejunal bypass liner. Obes Surg. 2011;21(7):941–7. 46. Gold ﬁ ne AB, Silver R, Aldhahi W, Cai D, Tatro E, Lee J, et al. Use of salsalate to target in ﬂ ammation in the treatment of insulin resistance and type 2 diabetes. Clin Transl Sci. 2008;1(1):36–43. Chapter 4 Epidemiology Including Youth Through Adulthood and Socioeconomic Impact

Helen Looker

Introduction

Type 2 diabetes (T2D) is a major public health issue affecting populations around the world. Projections based on demographics and expected shifts in population density suggest that this is going to be a long-term problem with major increases in the number of individuals with T2D over the course of the next few decades. Understanding the extent of the existing health problem caused by T2D puts into perspective the urgent need for programs to prevent T2D and its complications. In this chapter I will address recent estimates for the prevalence of diabetes along with changes in the patterns of prevalence, especially the emergence of T2D in youth. The evidence for diabetes risk factors, categorized as either ﬁ xed, related to early life, or modiﬁ able, will be assessed including underlying issues with the means of measurement. Again special mention will be given to the evidence for measures of risk in childhood. Finally, the data for the current socioeconomic impact of diabetes will be considered along with the evidence for the cost-effectiveness of T2D prevention.

Statistics

The quality of data available on the current prevalence of T2D varies from country to country [ 1 ] . While the USA has major health surveys and data from healthcare providers to assist it in estimating the number of people with diabetes nationally this

H. Looker, MD (*) Division of Endocrinology Diabetes and Bone Disease , Mount Sinai Medical Center , New York , NY , USA Medical Research Institute , Wellcome Trust Centre for Molecular Medicine, Clinical Research Centre, Level 7 , University of Dundee, Mail Box 11, Ninewells Hospital & Medical School, Dundee DD1 9SY , Scotland , UK e-mail: [email protected]

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 41 DOI 10.1007/978-1-4614-3314-9_4, © Springer Science+Business Media New York 2012 42 H. Looker

Fig. 4.1 Estimated glycemia breakdown for the US population 2010. Data for fi gure taken from National Institute of Diabetes and Digestive and Kidney Diseases website http://diabetes.niddk. nih.gov/dm/pubs/statistics/#ddY20 data is not available in every region. The World Health Organization (WHO) uses data collected from population studies around the world to create its estimates for the current global impact of diabetes as well as for building models to predict future global prevalence. Using data from studies carried out in 39 countries the most recent estimate from the WHO is that there were 171 million adults globally with diabetes in 2000. Using expected demographic changes and the population shifts in developing countries to greater urban populations the WHO estimate that by 2030 those numbers will have more than doubled to 366 million. These numbers corre- spond to an overall prevalence of 2.8% in 2000 rising to 4.4% in 2030 [1 ] . While these estimates are not broken down by diabetes type, T2D accounts for 90–95% of all cases of diabetes in adults [ 2 ] . The United States ranks third overall in number of adults with diabetes with only India and China having more cases in both 2000 and projected by 2030. The National Institutes of Health estimated that there are 18.8 million adults in the USA with diagnosed diabetes and another 7.0 million with undiagnosed diabetes in 2010 (8.3% of the whole population). If the number of individuals in the USA with prediabetes is included that means two in fi ve of the US population having some degree of glucose abnormality (Fig. 4.1 ) [ 1 ] . Along with the increasing number of people with diabetes, there is also the worrying trend that the mean age of onset for T2D is falling. According to data from the National Health and Nutrition Examination Survey (NHANES), the mean age of diagnosis of T2D in the USA has fallen from 52 years in 1988 to 46 years in 2000 [3 ] . Earlier onset of T2D has major implications for the health of the population as the complications of diabetes are strongly related to diabetes duration. Among the Pima Indians where youth onset T2D is less rare than in many other populations, younger onset of T2D is associated with increased risk of developing end-stage renal disease by middle age as well as being associated with an increased risk of premature mortality [4 ] . In keeping with this, diabetes is also becoming more common among children and this is due to increases in both type 1 and T2D. While type 1 diabetes accounts for most cases of diabetes among children aged under 10 years, T2D is becoming 4 Epidemiology Including Youth Through Adulthood and Socioeconomic Impact 43 increasingly common in older children especially among particular ethnic groups. The National Institutes of Health estimate that there are 215,000 cases of diabetes in the USA among the under 20s. The SEARCH for Diabetes in Youth Study is a multicenter study that collected data on both prevalent and incident cases of diabetes among children in the USA along with detailed collection of data for determin- ing type of diabetes. They found that while type 1 diabetes remained the commonest form of diabetes among children that T2D is no longer the rarity that it was in past decades. Among children under the age of 10 years, the incidence of T2D was 0.4 cases per 100,000 children but among children aged 10–19 years the incidence was 8.5 cases per 100,000. For Asian/Paci fi c Islander Americans and Native American children aged 10–19 years, the incidence of T2D was higher than that for type 1 diabetes while for Hispanic and non-Hispanic Blacks the rates for T2D and type 1 diabetes were similar in this age group [5 ] . A study of mortality among Pima Indians with T2D found that the introduction of more effective treatment for renal disease in diabetes had led to a marked increase in mortality due to cardiovascular disease. So as deaths due to diabetic nephropathy declined, the death rate overall was maintained due to the rise in deaths due to cardiovascular disease [ 6 ] . While treatment options for diabetes continue to expand and management of diabetic complications improves, the increase in prevalence is predicted to push the global mortality rates, due to diabetes, upwards. Whereas diabetes was the 11th most common cause of death globally in 2002, it is projected to have increased to the 7th leading cause by 2030 [7 ] .

Risk Factors for T2D

When considering the risk factors for T2D, one can divide them into those that are modifi able and those that are unmodifi able or fi xed. This is to some degree arbitrary but generally the fi xed risk factor is one that an at-risk individual has no power to change. The fi xed risk factors may not be anything that we can intervene to change, but they are important for understanding an individual’s baseline risk and as they interact with modifi able risk factors. However, when considering diabetes prevention the interventions are primarily focused on the modifi able or behavioral risk factors. It is becoming apparent that early life elements may have an impact on later disease and these have been classi fi ed here as early life risk factors as they could be open to modi fi cation if we aim to target early life but in most studies are no longer modifi able.

Fixed Risk Factors

Genetic predisposition: The role genetics play in T2D has long been accepted with the observations of the risks associated with ethnicity and family history and that these factors remain strong even after adjustment for other individual risk factors such as obesity. 44 H. Looker

1 . Ethnicity: As outlined above prevalence of T2D is higher among certain ethnic groups than others and these differences are not all explained by adjustment for lifestyle-related risk factors [8 ] . The ethnic groups that have the highest risks for T2D include Native Americans, Australian Aborigines, South Asians, Hispanics, and African Americans. The Pima Indians of Arizona have the highest prevalence of T2D of any population [ 9] . However, it should be noted that the Pima Indians who settled in the mountains of Mexico have far lower prevalence despite being genetically identical showing that the relationship between risk and ethnicity is also dependent on other risk factors such as lifestyle and obesity [ 10 ] . South Asians also have a higher prevalence of T2D and of note risk for T2D becomes apparent at lower levels of body mass index (BMI) than for other populations [11 ] . This may be a refl ection on different patterns of body fat distribution. 2 . Familial : A family history of T2D is a strong risk factor for T2D especially when considering T2D in fi rst degree relatives [12 ] . Concordance for T2D among monozygotic twins is high and ranges from 0.29 to 1.0 compared to concordance rates for dizygotic twins of 0.1–0.43 [13– 18 ] . Concordance is in fl uenced by time of follow-up for assessment of the twins and concordance is highest with the longest follow-up time [ 14] . Heritability (a measure of the proportion of a phenotype explained by genetics) is calculated as being 26% for T2D [13 ] and there are also strong heritability estimates for T2D-related traits such as insulin resistance [19 ] . Of note the risk associated with having a parent with T2D is higher if in addition the child is exposed to diabetes in utero [20 ] (see below). 3 . Specifi c genes: There have been major efforts over the course of the last 20 years to determine the genes underlying T2D and with the advent of large-scale genome- wide association studies the last decade has seen a rapid increase in the number of risk genes identifi ed. At the time of writing over 40 genes have been identifi ed that are statistically signifi cantly associated at the genome-wide level with an increased risk of diabetes [21 ] . However, despite these advances the identifi ed genes account for less than 10% of the observed heritability of T2D [ 22 ] . Effect sizes for the genes identi fi ed to this point have tended to be in the 1.1–1.2 range so that the presence of any one risk allele only raises the individuals risk for T2D by only a small degree. The gene with the strongest effect so far described is TCF7L2 where the risk allele is associated with a 1.4-fold increased risk of T2D [ 19 ] . At this stage models that include information on genetic variants do predict risk for diabetes but these models are currently not statistically signifi cantly better than those that instead include traditional risk factors such as age, ethnicity, and obesity [23 ] . There is a linear association between number of risk alleles and presence of T2D indicating that the genes identi fi ed thus far have an additive effect in conferring risk for T2D [21 ] . While there is no denying that more genes with modest effects on diabetes are likely to be identi fi ed, the size of the studies that have been undertaken suggests it is unlikely we will fi nd another gene with an effect size greater than that of TCF7L2 which has led to a reexamination of the idea that instead of common genetic variants explaining the majority of diabetes there may be a number of rare variants with large effects within families. To this end there has recently been a move to genetic sequencing and the search for rare variants [22 ] . 4 Epidemiology Including Youth Through Adulthood and Socioeconomic Impact 45

Age: T2D has traditionally been a condition of middle and old age but recent trends in the USA have shown that the mean age of onset of diabetes is reducing. This does not mean that the prevalence of T2D does not remain strongly correlated with age. Again based on NHANES data for the period 2005–2008 the prevalence of diabetes by age in the USA was 3.7% for 20–44-year olds, 13.7% for 45–64-year olds, and 26.9% for people aged 65 years or older [2 ] .

Early Life Risk Factors

Uterine environment : Poor in utero growth has been identi fi ed as being associated with several adult conditions including cardiovascular disease and T2D and these associations were the foundations of Hales and Barkers “thrifty gene” hypothesis for the pathogenesis of T2D [24 ] . According to this theory T2D in later life is a result of a fetal adaptation to maternal malnutrition during pregnancy. Events during World War II in Europe allowed for two studies—the Dutch Hunger Study [25 ] and the Leningrad Siege study [ 26 ] —to try and address this issue. Both centered on places where severe malnutrition was suffered by a population for a defi ned period of time and where the children born during this period were available for assessment in adulthood. The Dutch Hunger Study found an association with poorer glucose tolerance among adults whose mothers had been exposed to malnutrition during the latter stages of gestation with higher 2-h glucose and insulin concentrations [ 25 ] . In contrast the Leningrad Siege Study found no association between experience of maternal malnutrition and glycemia in the adult offspring [ 26 ] . The differences in fi ndings between the two studies may refl ect the importance not only of maternal nutrition in pregnancy but mother and child nutrition after delivery. Exposure to T2D itself in utero is associated with an increased risk of developing both obesity and T2D in later life. There is evidence that this is not simply a threshold effect and that as glycemia in the mother increases (even at levels below the diagnostic cut point for gestational diabetes) the risk of subsequent T2D in the offspring increases [27 ] . This is traditionally viewed as an unmodifi able risk factor as it is something an individual cannot infl uence but maternal glycemia remains a potential target for intervention when considering possible long-term prevention steps. As exposure to greater degrees of hyperglycemia in utero also increases risk for obesity later in life [ 28] this may well be an important phase for intervention, though complex to study due to the time of follow-up required. However, as yet there is no evidence that aggressive management of glycemia during pregnancy affects the long-term risks for T2D in the offspring though there is now good evidence that it has bene fi ts in the short term [29 ] . Birth weight : Birth weight can be considered a combination of genetic factors and intrauterine environment. It is also one of the few pieces of data reliably recorded on a population level and as a result birth weight has been a useful measure to assess as a risk factor for numerous diseases which are believed to have their roots in fetal development. A recent meta-analysis of studies found primarily a relationship 46 H. Looker between T2D and low birth weight with an overall odds ratio adjusted for age and sex of 0.75 (95% confi dence interval 0.70–0.91) per kilogram of birth weight [ 30 ] . However, some studies did report either a positive association or a U-shaped relationship [ 31, 32] . These reports came from studies of Native American populations where there was a high prevalence of maternal diabetes. Thus, it is suggested that the association between high birth weight and T2D is due to the effects of exposure to diabetes in utero and resultant macrosomia. As diabetes becomes more common among women of reproductive age, then it is possible that the U-shaped relationship will become the dominant one in the general population. Breast feeding : Prevention of T2D is one of the many potential positive effects of breast feeding. A review of published studies found an overall odds ratio of 0.61 (95% confi dence limits 0.44–0.85) for T2D in those who had been breast fed as infants [ 33 ] . Early growth: In addition to birth weight certain patterns of early growth have been associated with a risk for diabetes. The primary pattern that has been studied is termed “catch-up” growth and is essentially the observation that small for gestational age babies who then experience excess growth in the early years of life are at higher risk for metabolic dysregulation and obesity [ 34 ] . A study in the Pima Indians showed that this pattern was not seen among offspring of mothers with diabetes and there a pattern of “catch-down” seemed to be the dominant one [35 ] .

Modi ﬁ able Risk Factors

Body fat : Traditionally, the gold standard for measurement of body fat is underwater weighing though imaging techniques including dual-energy X-ray absorptiometry (DEXA), CT, and MRI are also highly accurate means of assessment of body fat [ 36] . However, these methods are not practical for large-scale epidemiological studies nor are they measures that will be available to physicians in their offi ce when assessing an individual’s risk. Body weight can be used as a surrogate measure for an individual’s body fat mass. BMI is a measure of weight adjusted for height and is the commonest measure of body weight used in epidemiological studies as it is easy to measure accurately and correlates reasonably well with body fat measures [ 37, 38] but has also been proven to predict obesity-related conditions such as T2D [39 ] . Through studies clear cut points for BMI have been defi ned, which are associated with risk for diabetes and other conditions. In adults a BMI between 25 and 29.9 kg/m 2 indicates being overweight while a BMI above 30 kg/m2 indicates being obese [ 40 ] . Of note these cut points may not be applicable to all ethnic groups. In particular there is evidence that Asians with a BMI of 22 kg/m 2 are at as great a risk for T2D as a Caucasian or African American with a BMI of 25 kg/m2 [ 41 ] . BMI will also misclassify some individuals [42 ] . Despite these limitations BMI remains the commonest available assessment of body fat used in epidemiological studies. Other measures of obesity include waist circumference. There is good evidence that waist circumference is independently associated with diabetes risk and may at 4 Epidemiology Including Youth Through Adulthood and Socioeconomic Impact 47 times be a stronger risk factor than BMI [ 43 ] though this is not always the case [ 39 ] . However, while it is an easily made measure it has not as yet become commonly used in clinical practice. In part this may be due to confusion as to the best place to make the measurement—waist circumference can be measured at the widest part of the abdomen, at a fi xed point de fi ned by distance from bony landmarks or at the level of the umbilicus. There is no evidence that measurement at any speci fi c site gives a better measure for risk assessment over another but obviously when wanting to compare measures in studies or to track an individual’s measurements a uniform means of measurement is needed. Waist/Hip ratio is also on occasion used but does not seem to be superior to waist circumference alone [44 ] . Skinfold thickness can be used as a means of assessment of body fat and fat distribution can be accounted for by use of different sites [ 45 ] . While some studies have found it a useful measure, it has not been adopted clinically as a means of risk assessment, again possibly due to the lack of standardized protocols. Diet : At its most simplistic level body weight is driven by energy balance—excessive calorie intake for a given level of physical activity will result in a positive energy balance and weight gain. Measurement of dietary intake is fraught with methodological issues [ 46 ] . While it is possible to closely monitor food intake in a controlled setting (such as an in-patient facility), how this relates to actual behaviors in free living conditions is unknown. The primary means for assessing diet in epidemiological studies are the use of food diaries, food recall interviews, and food frequency questionnaires. They all have their strengths and weaknesses—generally the more accurate tools are the ones that require more time to complete. Questionnaires have been validated and several standard instruments are in general use [47, 48 ] . It is important for these questionnaires to be validated in new populations to ensure that the questions do relate to the general diet of the population being studied. The relationship between calorie intake and risk of diabetes is established but great interest remains in trying to identify specifi c elements of a given diet that are either protective or particularly high risk. At this point the evidence for or against particular diets or dietary elements is mixed as epidemiological studies fi nd it hard to exclude the possibility that any specifi c measured differences in diet are not related to other, unmeasured aspects of lifestyle. Possibly one of the most commonly reported elements is the association between sugar-sweetened beverages (which may be better termed high-fructose corn syrup sweetened beverages in the USA) and risk of T2D. A recent meta-analysis of eight studies of the relationship between sugar-sweetened beverages and T2D found a 26% increased risk for diabetes among individuals in the top quartile for intake (drinking 1–2 drinks a day) compared to the lowest quartile, where intake was at most once a month [ 49] . However, it is still not clear whether this is a speci fi c risk associated with those drinks or that their use correlates with an overall less healthy lifestyle especially as no mechanistic explanation can explain why intake of sugar-sweetened beverages would be specifi cally related to risk over the calories ingested. Other studies have examined the role of whole grains, fruits and vegetables and diabetes risk. A study based on pooled data from 48 H. Looker six cohort studies found that an increase of two servings per day of whole grain was associated with a 21% decrease in risk for T2D and this was statistically signi fi cant even after adjustment for confounders such as BMI [ 50 ] . A meta-analysis of six studies of fruit and vegetable intake found no association with risk for T2D for any measure other than the intake of fresh leafy greens, where the people with the highest intake had a 14% reduction in risk compared to those with the lowest intake [51 ] . Among Pima Indians no speci fi c element of diet related to risk for T2D. However, individuals who considered their diet “traditional” as opposed to “anglo” (a more Western style diet) were at lower risk in keeping with the argument that lifestyle as a whole is the key factor rather than individual elements [52 ] . Physical activity: Many of the same issues that make diet hard to assess apply to physical activity. The gold standard currently is the use of doubly labeled water which can be used to calculate energy expenditure [53 ] . This is not a suitable method for large-scale studies, for these studies objective measures of physical activity rely on simpler methods such as the use of pedometers, accelerometers, heart rate monitors, and combination devices [ 53 ] . At present pedometers are the commonest device used to directly measure physical activity in epidemiological studies. They are used because they are cheap, easy to use, and as walking is a form of physical activity which is often the focus of physical activity interventions. For large epidemiological studies it is more common for physical activity to be measured by questionnaire and as with diet a number of validated instruments exist. The National Cancer Institute has collected data on physical activity questionnaires that include data for walking and/or cycling and have a list of over 100 separate questionnaires including 72 which have been validated [ 54 ] . As with diet questionnaires it is important for the speci fi c instruments used to be considered in regards of the population to ensure the activities data are collected on are appropriate to the population. Questionnaires usually differentiate between work-related and leisure physical activity. Leisure physical activity tends to be reported more than occupational physical activity though there is evidence that occupational physical activity is associated with risk for T2D in that it correlates with abdominal obesity [55 ] . A study of men who attended the University of Pennsylvania found that higher levels of leisure time physical activity recorded at the time of entry to college in the USA was inversely associated with incident T2D. With every 500-kcal increment in energy expenditure, the age-adjusted risk of T2D was reduced by 6%. This association remained after adjustment for covariates including BMI [ 56 ] . The Nurses Health Study found that risk of T2D was associated with lower physical activity levels and that this association was seen even if restricting the analysis to the nonobese women where the risk associated with low physical activity was twice that of the more physically active women [ 57] . Despite these studies adjusting for measured covariates it is hard to fully account for all lifestyle elements that may correlate with physical activity. Another metric that has been studied is the time spent undertaking low energy activities—most frequently television watching. The Nurses Health Study reported increased risk of T2D with increased time spent watching television independent of overall level of reported physical activity [58 ] . Similarly, 4 Epidemiology Including Youth Through Adulthood and Socioeconomic Impact 49 analysis of data from the European Prospective Investigation into Cancer and Nutrition–Potsdam study found that the adjusted hazard ratio for diabetes in people who watched at least 4 h of television a day was 1.63 (95% confi dence interval 1.17–2.27) compared to those who watched less than an hour a day [ 59 ] . In interpreting these data it is important to consider whether one measure is measured more accurately than the other. In this instance it may be easier for individuals to estimate their daily television watching time than it is for them to estimate the time spent undertaking a variety of different levels of physical activity. If that is the case it is not surprising that the former might remain signi fi cant even after adjustment from the latter with the mechanism still relating to overall levels of physical activity. Another approach has been to measure fi tness, assessed by treadmill tests, instead of physical activity. There have been confl icting reports from these studies with one study garnering much attention for fi nding that the risk for T2D was equal between the non-obese non- fi t and the obese fi t [60 ] though these fi ndings have not been reported by other studies [61 ] . Hyperglycemia: Diabetes is defi ned by specifi c cut points of blood glucose—measured either while fasting or after a glucose load—and thus it is no surprise that current glucose is a risk factor for future T2D. The presence of either impaired fasting glucose (a blood glucose of 100–125 mg/dL) or impaired glucose tolerance (a 2-h blood glucose of 140–199 mg/dL) has now been termed prediabetes which indicates how strongly they are associated with risk for diabetes. Of note the cut point for impaired fasting glucose was adjusted in 2004 where it shifted from 110 to 125 mg/dL to its current level [62 ] . As a result studies that took place prior to the adjustment cannot be directly compared to those taking place more recently. Current estimates of prevalence for prediabetes in the USA indicate that it was present in ~30% of adults aged over 20 years in 2005 [ 63 ] . The move to identify people with prediabetes is founded on the knowledge that this group has a very high risk of T2D [ 64 ] . In the Diabetes Prevention Program (DPP) 11% of individuals in the placebo group developed T2D each year [65 ] . Metabolic syndrome : That metabolic syndrome predicts incidence of diabetes is unsurprising as it is de fi ned by the presence of a combination of factors that include such strong risk factors for diabetes as obesity (measured by BMI or waist circumference) and abnormal glycemia (such as prediabetes) [ 21, 66, 67 ] . Despite the variations in the de fi nition meta-analyses suggest that in all its forms metabolic syndrome is predictive of T2D risk. In addition to considering the condition as a single entity there was a positive association with how many individual components of the syndrome were present and risk of T2D [68 ] . Gestational diabetes : The de fi nition of gestational diabetes is the onset of hyperglycemia fi rst noted in pregnancy. For the purposes of gestational diabetes hyperglycemia is defi ned as the presence of prediabetes or diabetic levels of glucose. Studies of glucose regulation in pregnancy show that insulin resistance does rise in pregnancy and while the majority of women can adjust their insulin production to maintain normal glycemia a small proportion develop gestational diabetes [ 69 ] . 50 H. Looker

When gestational diabetes was initially described, it was as a condition that identifi ed women at high risk for later onset of T2D [ 70 ] . The cumulative incidences reported vary with time of follow-up but the greatest rise in cumulative incidence is seen in the fi rst 5 years [71 ] . Socioeconomic status : Socioeconomic status can be measured using a variety of metrics with education, occupation, class, and income all reported. Incident diabetes was related to wealth for both men and women though this association was attenuated after adjustment for lifestyle and BMI with obesity in women strongly associated with lower wealth [21, 72 ] . Occupation has also been reported as a risk factor, with T2D incidence associated with lower employment grades [ 72, 73 ] . Education level is inversely related to risk of T2D [ 72, 74] . A recent study of risk factors for T2D among African American women found that measures of socioeconomic status for the area of residence were associated with risk of individuals developing diabetes—with women living in areas of low socioeconomic status having the highest incidence of T2D. This fi nding was signi fi cant even after adjustment for individual socioeconomic status [75 ] . Whether one socioeconomic risk factor is more strongly associated than another must always be viewed in light of accuracy of measure. Educational attainment tends to be less infl uenced by possible effects of reverse causation than occupation or income, at least for individuals who develop diabetes as adults. Smoking : Cigarette smoking is associated with risk for T2D [ 76 ] but a recent study of people within the Atherosclerosis Risk in Communities Study (ARIC) found that, at least in the short term, smoking cessation was associated with an increased risk of T2D with the people who stopped smoking during the study having a higher risk of T2D than those who continued to smoke [ 77 ] . This may well be associated with weight gain during this period though steps were taken to adjust for that as much as possible. Coffee: There has been confl icting evidence as to the role of coffee intake and T2D. As with most diet issues accurately collecting data on coffee intake is complicated not only by need to assess number and volume of drinks but also the variation in strength of coffee and these methodological issues may account for the disparate fi ndings; however, a recent meta-analysis concluded that there was a 7% reduction in risk for T2D with each additional cup of coffee drunk [78 ] . Vitamin D: Vitamin D has attracted interest as a potential risk factor for T2D. The handful of longitudinal studies that have looked at vitamin D intake have been suggestive of a protective effect due to vitamin D intake but as yet there are no conclu- sive studies in this area [ 79 ] . It should also be remembered that studies of nutrients often taken as supplements are particularly liable to confounding by lifestyle (cf. vitamin E). Liver disease: Recently studies have focused on a potential role for liver disease as a risk factor for T2D with associations being found for speci fi c liver enzymes as well as for the condition Non-Alcoholic Fatty Liver Disease (NAFLD) [80– 82 ] . 4 Epidemiology Including Youth Through Adulthood and Socioeconomic Impact 51

Risk Factor Assessment in Childhood

As diabetes becomes more prevalent in childhood there is increasing interest in assessment of risk during childhood. The risk factors identi fi ed are as listed for adults but there are speci fi c issues as to assessment of some risk factors—primarily body fat and glycemia. BMI is a less useful measure in childhood than it is for adults as it varies with age [83 ] . As a result growth in children is monitored with reference to age-specifi c BMI. The Center for Disease Control has produced BMI charts indicating the percentiles for BMI and has suggested defi ning childhood obesity as a BMI over the 95th percentile (or a BMI over the 85th percentile in the presence of complications of obesity) and a BMI above the 85th percentile as being overweight [84 ] . It should be noted that these data are only truly applicable to the US population. The International Obesity Taskforce have tried to create international guidelines that are not country specifi c and suggested calculating the BMI at any given age that would track to the cut points of the adult BMI categories and thus created age- and sex-specifi c cut points for children [ 85 ] . At present none of these de fi nitions is fully validated on the basis of being shown to be associated with a step up in risk for future obesity or complications of obesity. As the Pima Indian population has such a high prevalence of T2D, it was possible to examine data collected at different stages in childhood and assess the predictive value for incident diabetes. In children aged 5–9 years waist circumference was the strongest predictor of later diabetes and additional information on risk factors (including glycemia, blood pressure, and BMI) did not greatly add to the power of the model including only waist circumference. In contrast among 10–14-year olds and 15–19-year olds 2 h glucose, HbA1c and BMI were the strongest risk factors, with waist circumference being an informative addition to the model for the older age group. As expected exposure to diabetes in utero was by far the strongest single risk factor measured; however, even after stratifi cation for this the models based on other risk factors remained positively predictive of diabetes. Figure 4.2 shows the hazard ratios for all children (aged 5–19 years) for tertiles of each risk factor considered strati fi ed by exposure to diabetes in utero [86 ] .

Overview of Risk Factors for T2D

Above we have detailed some of the evidence for the leading risk factors identi fi ed for T2D. In reality the interplay between these factors is still to be fully understood. The schematic in Fig. 4.3 outlines a possible way of conceptualizing the interplay between risk factors. The fi xed risk factors exert infl uence not only on the early life risk factors but also on the lifestyle risk factors. Similarly, early life risk factors have an in fl uence on lifestyle risk factors. All three groups then have an impact on the main risk phenotypes for T2D (obesity and hyperglycemia) with reciprocal infl uence back from these phenotypes onto the behavioral risk factors.

Fig. 4.2 Risk factors for diabetes measured in Pima Indian children and youth aged 5–19 years. Taken from Franks et al. [86 ]

Fig. 4.3 Schematic of interplay of risk factors and T2D. Obesity and hyperglycemia are seen as intermediate steps with risk factors in ﬂ uencing risk of diabetes through these factors as well as directly on the risk of T2D 4 Epidemiology Including Youth Through Adulthood and Socioeconomic Impact 53

One element to consider is whether we should be desperately searching for the individual components of lifestyles that put people at risk for diabetes and instead accept that the real power may come from the combined elements of the lifestyle. We should also consider that all factors are also open to being in fl uenced by the environment; as access to areas to exercise or shops with affordable healthy food may be highly in fl uential on behavioral risk factors. Equally, it should be recalled that the individual may also in fl uence the environment so even this interaction is dynamic.

Socioeconomic Impact of Diabetes

The impact of diabetes on society as a whole is hard to calculate. Any estimate must include the costs required for health care to people with diabetes—including both costs of hospitalizations and medicines—but also the lost income due to loss of work days for adults, and premature death. As T2D is a chronic disease, the direct healthcare costs are high as there are long-term treatment requirements. In addition to treatment to control glycemia there is the recognition that other associated conditions also need active management and, as in the case of blood pressure, more aggressive management than for people without T2D. Thus, in addition to hypoglycemic agents many people with diabetes will also be treated with antihypertensive agents and lipid lowering drugs. There are additional laboratory investigations undertaken for people with T2D and screening for diabetes complications. The economic costs associated with the treatment of the complications of diabetes are also important. Diabetic nephropathy can lead to the need for renal replacement therapy while diabetic retinopathy may require repeated laser treatments and vascular disease, be it coronary, cerebral, or peripheral, may require surgical intervention. The complications will also contribute to the indirect costs of diabetes due to loss of working time following the onset of renal failure, amputation or cardiovascular disease. The costs to the individual with diabetes extend beyond those for health care alone. There is loss of earnings due to lost workdays because of diabetes or its complications. There are the costs incurred by requiring changes to homes and requirements for assistance to deal with disabilities due to diabetes. Figures from the USA in 2007 suggested that overall absenteeism is 0.8% higher in people with diabetes than people without diabetes, and overall diabetes accounts for 107 million work days lost due to unemployment and disability secondary to diabetes. The costs of the absenteeism and reduced work productivity were estimated at 22.6 billion with lost productivity due to early death calculated as $26.9 billion [87 ] . The International Diabetes Federation estimated that the global cost of health care for diabetes treatment and diabetes prevention would total $465 billion in 2011 increasing to over $595 billion by 2030 [88 ] . Estimates for the direct costs of health care by percentage of total annual healthcare budget are 2.5–15% depending on the prevalence in the population and the treatment options available. The American 54 H. Looker

Diabetes Association estimates that diabetes cost the US $174 billion in 2007 of which $116 billion were due to direct medical costs and the remaining $58 billion due to the indirect costs due to diabetes-related disability, work loss, and premature mortality. These estimates do not consider the costs related to morbidity and mortality among the people with undiagnosed diabetes which was estimated at an additional $18 billion. An individual with diabetes has 2.3 times more healthcare costs in a year than they would if they did not have diabetes [87 ] . In some parts of the world the cost of diabetes care is beyond the budget available. The International Diabetes Fund estimates that in 2007 in Burundi the healthcare cost spent per person with diabetes was $6 while in Haiti it was only $48.

Cost-Effectiveness of Diabetes Prevention

From the above it is clear that at the level of the individual and society diabetes has a terrible cost. This would suggest that diabetes prevention is a key means to reduce both the direct and indirect costs of T2D. The interventions used in the trials that demonstrated that T2D could be prevented can be costed leading to attempts to predict the cost-effectiveness of diabetes prevention. The results of these studies re fl ect in part the variation that rests on the underlying assumptions in such models. When trying to determine whether a treatment is “cost-effective,” costs per QALY have been used as a way to combine costs with effectiveness. As a rule a treatment is deemed to be cost-effective if the cost per QALY is less than $50,000. The DPP proved that T2D could be delayed most effectively by a lifestyle intervention, though the use of metformin was associated with a lesser degree of prevention [ 65 ] . However, economic modeling showed that both lifestyle and metformin for prevention were more costly than placebo alone. Modeling the fi ndings of the DPP suggested that lifestyle intervention was associated with a delay of 11 years for onset of diabetes, while metformin was associated with a delay of 3 years. A cost-effectiveness analysis based on the DPP interventions found that the cost for the lifestyle intervention was ~$1,100 per QALY. This was lower than for metformin where the cost per QALY was ~$31,300 [ 89 ] . Thus, both the lifestyle intervention and metformin would be considered to be cost-effective. Of note there was an age interaction with metformin not being cost-effective for use among adults over the age of 65 years. To put these estimates in context, statin use in a patient with T2D, no coronary disease, and a total cholesterol above 200 mg/dL has a cost of $52,000 per QALY while intensive glycemic control costs $41,000 per QALY over the course of a lifetime [ 90 ] . In contrast a model again using DPP data but this time using the Archimedes model found there was only a 0.1% chance of the cost per QALY of the lifestyle intervention being below $50,000 [91 ] . As more data becomes available as to the long-term bene fi ts of the various diabetes prevention studies, the models assessing cost-bene fi t will hope- fully be improved. However, work is still on going to fi nd low cost means of diabetes prevention. 4 Epidemiology Including Youth Through Adulthood and Socioeconomic Impact 55

Conclusions

T2D is already a major chronic disease affecting people across the world, with predictions based on demographics suggesting that this is only going to get worse in coming decades. In addition to the increasing prevalence is the evidence that diabetes is affecting younger people now than in the past and the consequences this may have for the onset of diabetic complications at an earlier age. The main modi ﬁ able risk factors for T2D are body fat, diet, and exercise with most of successful diabetes prevention studies addressing all three elements to one degree or another. The most common means of identifying high-risk individuals are obesity, family history of diabetes, and presence of prediabetes. Traditionally, prevention for T2D has only been considered in adults but with the emergence of more T2D in youth, and the strong tracking of obesity interventions aimed at young people is emerging. However, it may also be necessary not just to address high-risk individuals but society as a whole to look for ways to increase the level of public health. The costs due to diabetes, both economic and noneconomic, are huge already and expected to increase further. Thus, we have never more needed a successful and cost-effective means for prevention of T2D. The existing studies suggest that it is within reach but translation of study ﬁ ndings into real-world health initiatives is still in its early stages with little in the way of evidence that diabetes has been prevented on a wider level.

References

1. Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care. 2004;27:1047–53. 2. National Diabetes Statistics. 2011. http://diabetes.niddk.nih.gov/dm/pubs/statistics/ . Accessed 5 June 2011. 3. Koopman RJ, Mainous III AG, Diaz VA, Geesey ME. Changes in age at diagnosis of type 2 diabetes mellitus in the United States, 1988 to 2000. Ann Fam Med. 2005;3:60–3. 4. Pavkov ME, Bennett PH, Knowler WC, Krakoff J, Sievers ML, Nelson RG. Effect of youth- onset type 2 diabetes mellitus on incidence of end-stage renal disease and mortality in young and middle-aged Pima Indians. JAMA. 2006;296:421–6. 5. Dabelea D, Bell RA, D’Agostino Jr RB, et al. Incidence of diabetes in youth in the United States. JAMA. 2007;297:2716–24. 6. Pavkov ME, Bennett PH, Sievers ML, et al. Predominant effect of kidney disease on mortality in Pima Indians with or without type 2 diabetes. Kidney Int. 2005;68:1267–74. 7. Mathers CD, Loncar D. Projections of global mortality and burden of disease from 2002 to 2030. PLoS Med. 2006;3:e442. 8. Shai I, Jiang R, Manson JE, et al. Ethnicity, obesity, and risk of type 2 diabetes in women: a 20-year follow-up study. Diabetes Care. 2006;29:1585–90. 9. Knowler WC, Pettitt DJ, Savage PJ, Bennett PH. Diabetes incidence in Pima Indians: contributions of obesity and parental diabetes. Am J Epidemiol. 1981;113:144–56. 10. Schulz LO, Bennett PH, Ravussin E, et al. Effects of traditional and western environments on prevalence of type 2 diabetes in Pima Indians in Mexico and the U.S. Diabetes Care. 2006;29:1866–71. 56 H. Looker

11. Misra A, Khurana L. The metabolic syndrome in South Asians: epidemiology, determinants, and prevention. Metab Syndr Relat Disord. 2009;7:497–514. 12. Harrison TA, Hindorff LA, Kim H, et al. Family history of diabetes as a potential public health tool. Am J Prev Med. 2003;24:152–9. 13. Poulsen P, Kyvik KO, Vaag A, Beck-Nielsen H. Heritability of type II (non-insulin-dependent) diabetes mellitus and abnormal glucose tolerance—a population-based twin study. Diabetologia. 1999;42:139–45. 14. Medici F, Hawa M, Ianari A, Pyke DA, Leslie RD. Concordance rate for type II diabetes mellitus in monozygotic twins: actuarial analysis. Diabetologia. 1999;42:146–50. 15. Kaprio J, Tuomilehto J, Koskenvuo M, et al. Concordance for type 1 (insulin-dependent) and type 2 (non-insulin-dependent) diabetes mellitus in a population-based cohort of twins in Finland. Diabetologia. 1992;35:1060–7. 16. Gottlieb MS, Root HF. Diabetes mellitus in twins. Diabetes. 1968;17:693–704. 17. Barnett AH, Eff C, Leslie RD, Pyke DA. Diabetes in identical twins. A study of 200 pairs. Diabetologia. 1981;20:87–93. 18. Newman B, Selby JV, King MC, Slemenda C, Fabsitz R, Friedman GD. Concordance for type 2 (non-insulin-dependent) diabetes mellitus in male twins. Diabetologia. 1987;30:763–8. 19. Freeman MS, Mans ﬁ eld MW, Barrett JH, Grant PJ. Heritability of features of the insulin resistance syndrome in a community-based study of healthy families. Diabet Med. 2002;19:994–9. 20. Pettitt DJ, Aleck KA, Baird HR, Carraher MJ, Bennett PH, Knowler WC. Congenital susceptibility to NIDDM. Role of intrauterine environment. Diabetes. 1988;37:622–8. 21. Dupuis J, Langenberg C, Prokopenko I, et al. New genetic loci implicated in fasting glucose homeostasis and their impact on type 2 diabetes risk. Nat Genet. 2010;42:105–16. 22. Manolio TA, Collins FS, Cox NJ, et al. Finding the missing heritability of complex diseases. Nature. 2009;461:747–53. 23. Meigs JB, Shrader P, Sullivan LM, et al. Genotype score in addition to common risk factors for prediction of type 2 diabetes. N Engl J Med. 2008;359:2208–19. 24. Hales CN, Barker DJ. The thrifty phenotype hypothesis. Br Med Bull. 2001;60:5–20. 25. Roseboom T, de Rooij S, Painter R. The Dutch famine and its long-term consequences for adult health. Early Hum Dev. 2006;82:485–91. 26. Stanner SA, Bulmer K, Andres C, et al. Does malnutrition in utero determine diabetes and coronary heart disease in adulthood? Results from the Leningrad siege study, a cross sectional study. BMJ. 1997;315:1342–8. 27. Franks PW, Looker HC, Kobes S, et al. Gestational glucose tolerance and risk of type 2 diabetes in young Pima Indian offspring. Diabetes. 2006;55:460–5. 28. Pettitt DJ, Baird HR, Aleck KA, Bennett PH, Knowler WC. Excessive obesity in offspring of Pima Indian women with diabetes during pregnancy. N Engl J Med. 1983;308:242–5. 29. Crowther CA, Hiller JE, Moss JR, McPhee AJ, Jeffries WS, Robinson JS. Effect of treatment of gestational diabetes mellitus on pregnancy outcomes. N Engl J Med. 2005;352:2477–86. 30. Whincup PH, Kaye SJ, Owen CG, et al. Birth weight and risk of type 2 diabetes: a systematic review. JAMA. 2008;300:2886–97. 31. McCance DR, Pettitt DJ, Hanson RL, Jacobsson LT, Knowler WC, Bennett PH. Birth weight and non-insulin dependent diabetes: thrifty genotype, thrifty phenotype, or surviving small baby genotype? BMJ. 1994;308:942–5. 32. Dyck RF, Klomp H, Tan L. From “thrifty genotype” to “hefty fetal phenotype”: the relationship between high birthweight and diabetes in Saskatchewan Registered Indians. Can J Public Health. 2001;92:340–4. 33. Owen CG, Martin RM, Whincup PH, Smith GD, Cook DG. Does breastfeeding in ﬂ uence risk of type 2 diabetes in later life? A quantitative analysis of published evidence. Am J Clin Nutr. 2006;84:1043–54. 34. Ong KK, Dunger DB. Birth weight, infant growth and insulin resistance. Eur J Endocrinol. 2004;151 Suppl 3:U131–9. 35. Touger L, Looker HC, Krakoff J, Lindsay RS, Cook V, Knowler WC. Early growth in offspring of diabetic mothers. Diabetes Care. 2005;28:585–9. 4 Epidemiology Including Youth Through Adulthood and Socioeconomic Impact 57

36. Silver HJ, Welch EB, Avison MJ, Niswender KD. Imaging body composition in obesity and weight loss: challenges and opportunities. Diabetes Metab Syndr Obes. 2010;3:337–47. 37. Gallagher D, Visser M, Sepulveda D, Pierson RN, Harris T, Heyms ﬁ eld SB. How useful is body mass index for comparison of body fatness across age, sex, and ethnic groups? Am J Epidemiol. 1996;143:228–39. 38. Roubenoff R, Dallal GE, Wilson PW. Predicting body fatness: the body mass index vs estimation by bioelectrical impedance. Am J Public Health. 1995;85:726–8. 39. Tulloch-Reid MK, Williams DE, Looker HC, Hanson RL, Knowler WC. Do measures of body fat distribution provide information on the risk of type 2 diabetes in addition to measures of general obesity? Comparison of anthropometric predictors of type 2 diabetes in Pima Indians. Diabetes Care. 2003;26:2556–61. 40. WHO. Obesity: preventing and managing the global epidemic. Report on a WHO consultation on obesity. Geneva: WHO; 1998. 41. WHO Expert Consultation. Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet. 2004;363:157–63. 42. Rothman KJ. BMI-related errors in the measurement of obesity. Int J Obes (Lond). 2008;32 Suppl 3:S56–9. 43. Wei M, Gaskill SP, Haffner SM, Stern MP. Waist circumference as the best predictor of noninsulin dependent diabetes mellitus (NIDDM) compared to body mass index, waist/hip ratio and other anthropometric measurements in Mexican Americans—a 7-year prospective study. Obes Res. 1997;5:16–23. 44. Wannamethee SG, Papacosta O, Whincup PH, et al. Assessing prediction of diabetes in older adults using different adiposity measures: a 7 year prospective study in 6,923 older men and women. Diabetologia. 2010;53:890–8. 45. Taylor AE, Ebrahim S, Ben-Shlomo Y, et al. Comparison of the associations of body mass index and measures of central adiposity and fat mass with coronary heart disease, diabetes, and all-cause mortality: a study using data from 4 UK cohorts. Am J Clin Nutr. 2010;91:547–56. 46. Beaton GH, Burema J, Ritenbaugh C. Errors in the interpretation of dietary assessments. Am J Clin Nutr. 1997;65:1100S–7. 47. NHANES Food Frequency Questionnaire. National Cancer Institute. 2011. http://riskfactor. cancer.gov/diet/usualintakes/ffq.html . Accessed 6 June 2011. 48. McKeown NM, Day NE, Welch AA, et al. Use of biological markers to validate self-reported dietary intake in a random sample of the European Prospective Investigation into Cancer United Kingdom Norfolk cohort. Am J Clin Nutr. 2001;74:188–96. 49. Malik VS, Popkin BM, Bray GA, Despres JP, Willett WC, Hu FB. Sugar-sweetened beverages and risk of metabolic syndrome and type 2 diabetes: a meta-analysis. Diabetes Care. 2010;33:2477–83. 50. de Munter JS, Hu FB, Spiegelman D, Franz M, van Dam RM. Whole grain, bran, and germ intake and risk of type 2 diabetes: a prospective cohort study and systematic review. PLoS Med. 2007;4:e261. 51. Carter P, Gray LJ, Troughton J, Khunti K, Davies MJ. Fruit and vegetable intake and incidence of type 2 diabetes mellitus: systematic review and meta-analysis. BMJ. 2010;341:c4229. 52. Williams DE, Knowler WC, Smith CJ, et al. The effect of Indian or Anglo dietary preference on the incidence of diabetes in Pima Indians. Diabetes Care. 2001;24:811–6. 53. Westerterp KR. Assessment of physical activity: a critical appraisal. Eur J Appl Physiol. 2009;105:823–8. 54. Standardized questionnaires of walking and bicycling database. 2011. http://appliedresearch. cancer.gov/tools/paq/ . Accessed 6 June 2011. 55. Steeves JA, Bassett Jr DR, Thompson DL, Fitzhugh EC. Relationships of occupational and non-occupational physical activity to abdominal obesity. Int J Obes (Lond). 2012;36(1):100–6. 56. Helmrich SP, Ragland DR, Leung RW, Paffenbarger Jr RS. Physical activity and reduced occurrence of non-insulin-dependent diabetes mellitus. N Engl J Med. 1991;325:147–52. 57. Rana JS, Li TY, Manson JE, Hu FB. Adiposity compared with physical inactivity and risk of type 2 diabetes in women. Diabetes Care. 2007;30:53–8. 58 H. Looker

58. Hu FB, Li TY, Colditz GA, Willett WC, Manson JE. Television watching and other sedentary behaviors in relation to risk of obesity and type 2 diabetes mellitus in women. JAMA. 2003;289:1785–91. 59. Ford ES, Schulze MB, Kroger J, Pischon T, Bergmann MM, Boeing H. Television watching and incident diabetes: fi ndings from the European Prospective Investigation into Cancer and Nutrition-Potsdam Study. J Diabetes. 2010;2:23–7. 60. Wei M, Kampert JB, Barlow CE, et al. Relationship between low cardiorespiratory fi tness and mortality in normal-weight, overweight, and obese men. JAMA. 1999;282:1547–53. 61. Fogelholm M. Physical activity, fi tness and fatness: relations to mortality, morbidity and disease risk factors. A systematic review. Obes Rev. 2010;11:202–21. 62. Genuth S, Alberti KG, Bennett P, et al. Follow-up report on the diagnosis of diabetes mellitus. Diabetes Care. 2003;26:3160–7. 63. Cowie CC, Rust KF, Ford ES, et al. Full accounting of diabetes and pre-diabetes in the U.S. population in 1988–1994 and 2005–2006. Diabetes Care. 2009;32:287–94. 64. Edelstein SL, Knowler WC, Bain RP, et al. Predictors of progression from impaired glucose tolerance to NIDDM: an analysis of six prospective studies. Diabetes. 1997;46:701–10. 65. Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346:393–403. 66. Lorenzo C, Williams K, Hunt KJ, Haffner SM. The National Cholesterol Education Program— Adult Treatment Panel III, International Diabetes Federation, and World Health Organization defi nitions of the metabolic syndrome as predictors of incident cardiovascular disease and diabetes. Diabetes Care. 2007;30:8–13. 67. Boyko EJ, de Courten M, Zimmet PZ, Chitson P, Tuomilehto J, Alberti KG. Features of the metabolic syndrome predict higher risk of diabetes and impaired glucose tolerance: a prospective study in Mauritius. Diabetes Care. 2000;23:1242–8. 68. Ford ES, Li C, Sattar N. Metabolic syndrome and incident diabetes: current state of the evidence. Diabetes Care. 2008;31:1898–904. 69. Kuhl C. Insulin secretion and insulin resistance in pregnancy and GDM. Implications for diagnosis and management. Diabetes. 1991;40 Suppl 2:18–24. 70. O’Sullivan JB. Diabetes mellitus after GDM. Diabetes. 1991;40 Suppl 2:131–5. 71. Kim C, Newton KM, Knopp RH. Gestational diabetes and the incidence of type 2 diabetes: a systematic review. Diabetes Care. 2002;25:1862–8. 72. Robbins JM, Vaccarino V, Zhang H, Kasl SV. Socioeconomic status and diagnosed diabetes incidence. Diabetes Res Clin Pract. 2005;68:230–6. 73. Kumari M, Head J, Marmot M. Prospective study of social and other risk factors for incidence of type 2 diabetes in the Whitehall II study. Arch Intern Med. 2004;164:1873–80. 74. Williams ED, Tapp RJ, Magliano DJ, Shaw JE, Zimmet PZ, Oldenburg BF. Health behaviours, socioeconomic status and diabetes incidence: the Australian Diabetes Obesity and Lifestyle Study (AusDiab). Diabetologia. 2010;53:2538–45. 75. Krishnan S, Cozier YC, Rosenberg L, Palmer JR. Socioeconomic status and incidence of type 2 diabetes: results from the Black Women’s Health Study. Am J Epidemiol. 2010;171:564–70. 76. Wannamethee SG, Shaper AG, Perry IJ. Smoking as a modifi able risk factor for type 2 diabetes in middle-aged men. Diabetes Care. 2001;24:1590–5. 77. Yeh HC, Duncan BB, Schmidt MI, Wang NY, Brancati FL. Smoking, smoking cessation, and risk for type 2 diabetes mellitus: a cohort study. Ann Intern Med. 2010;152:10–7. 78. Huxley R, Lee CM, Barzi F, et al. Coffee, decaffeinated coffee, and tea consumption in relation to incident type 2 diabetes mellitus: a systematic review with meta-analysis. Arch Intern Med. 2009;169:2053–63. 79. Pittas AG, Dawson-Hughes B. Vitamin D and diabetes. J Steroid Biochem Mol Biol. 2010;121:425–9. 80. Vozarova B, Stefan N, Lindsay RS, et al. High alanine aminotransferase is associated with decreased hepatic insulin sensitivity and predicts the development of type 2 diabetes. Diabetes. 2002;51:1889–95. 4 Epidemiology Including Youth Through Adulthood and Socioeconomic Impact 59

81. Lee DH, Silventoinen K, Jacobs Jr DR, Jousilahti P, Tuomileto J. Gamma-glutamyltransferase, obesity, and the risk of type 2 diabetes: observational cohort study among 20,158 middle-aged men and women. J Clin Endocrinol Metab. 2004;89:5410–4. 82. Shibata M, Kihara Y, Taguchi M, Tashiro M, Otsuki M. Nonalcoholic fatty liver disease is a risk factor for type 2 diabetes in middle-aged Japanese men. Diabetes Care. 2007;30:2940–4. 83. Rolland-Cachera MF, Sempe M, Guilloud-Bataille M, Patois E, Pequignot-Guggenbuhl F, Fautrad V. Adiposity indices in children. Am J Clin Nutr. 1982;36:178–84. 84. Barlow SE, Dietz WH. Obesity evaluation and treatment: expert committee recommendations. The Maternal and Child Health Bureau, Health Resources and Services Administration and the Department of Health and Human Services. Pediatrics. 1998;102:E29. 85. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard defi nition for child overweight and obesity worldwide: international survey. BMJ. 2000;320:1240–3. 86. Franks PW, Hanson RL, Knowler WC, et al. Childhood predictors of young-onset type 2 diabetes. Diabetes. 2007;56:2964–72. 87. American Diabetes Association. Economic costs of diabetes in the U.S. In 2007. Diabetes Care. 2008;31:596–615. 88. The economic impact of diabetes. International Diabetes Federation. http://www.idf.org/ diabetessatlas/5e/healthcare-expenditures . 89. Herman WH, Hoerger TJ, Brandle M, et al. The cost-effectiveness of lifestyle modifi cation or metformin in preventing type 2 diabetes in adults with impaired glucose tolerance. Ann Intern Med. 2005;142:323–32. 90. The CDC Diabetes Cost Effectiveness Group. Cost-effectiveness of intensive glycemic control, intensi fi ed hypertension control, and serum cholesterol level reduction for type 2 diabetes. JAMA. 2002;287:2542–51. 91. Eddy DM, Schlessinger L, Kahn R. Clinical outcomes and cost-effectiveness of strategies for managing people at high risk for diabetes. Ann Intern Med. 2005;143:251–64. Chapter 5 Prediabetes Genes in Pima and Amish

Leslie J. Baier

Type 2 Diabetes Is an Environmentally In ﬂ uenced, Polygenic Disease

Type 2 diabetes mellitus has a global distribution, yet its prevalence varies from country to country with the highest rates being reported in developed and developing countries [1– 3 ] . Studies comparing rural vs. urban dwelling, as well as migration studies, indicate that change towards a “Westernized” lifestyle is associated with a dramatic increase in the prevalence rates for this disease [ 4 ] . This environmental impact has been documented in urbanized Paci fi c Island populations and migrant Asian Indians [5– 9 ] . However, among individuals living in a similar environment, genetics has a clear infl uence on prevalence rates of type 2 diabetes. Different ethnic groups living within the same geographic region often have different prevalence rates of this disease [ 10] . For example, Latinos are the largest minority population in the United States and have a two to fourfold higher prevalence of diagnosed diabetes as compared to Caucasians [ 11, 12] . Rates of diabetes also differ between families, where siblings of affected individuals have an increased risk, which again suggests a familial component. Some of the strongest evidence for a genetic basis for type 2 diabetes comes from studies in twin pairs. Both monozygotic and dizygotic twin pairs share equally in a family environment, yet there is a higher concordance rate for type 2 diabetes among monozygotic twins who share 100% of their DNA as compared to dizygotic twins who share on average 50% of their DNA [ 13, 14 ] . In recent years, there has been an alarming increase in the prevalence of type 2 diabetes which has led to signi fi cant efforts to understand the etiology of this disease. Progress has been made in identifying several lifestyle factors that contribute to type

L. J. Baier, PhD (*) Diabetes Molecular Genetics Section, Phoenix Epidemiology and Clinical Research Center , National Institute of Diabetes, Digestive and Kidney Diseases, National Institutes of Health , 445 North 5th Street , Phoenix , AZ 85004 , USA e-mail: [email protected]

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 61 DOI 10.1007/978-1-4614-3314-9_5, © Springer Science+Business Media New York 2012 62 L.J. Baier

2 diabetes [ 15, 16] , but identifying the genetic basis for this disease has been far more diffi cult than anticipated. A portion of the complexity arises from the fact that obesity, which is a major risk factor for type 2 diabetes, is also infl uenced by both genetics and lifestyle [17 ] . Studies of body weight in twins, and particularly in twins reared apart, have produced high heritability estimates for body mass index (BMI, weight in kg/height in m 2, an estimate of degree of obesity) [ 18, 19 ] . However, eating behavior and activity levels, independent of genetics, can in fl uence body weight. It is generally accepted that the high rates of prevalence and incidence of type 2 diabetes now being seen in adolescents and children is a direct consequence of the recent increase in rates of obesity in these age groups [ 20 ] . However, the observed effect of changes in diet and activity may not always be independent of genetics. The interaction of genes and environment can infl uence body weight and risk for type 2 diabetes [ 21] , and it has been proposed that certain populations such as American Indians have a genetic susceptibility to type 2 diabetes (“diabetes genotype”) where the disease is unmasked by changes in environment or lifestyle [22 ] . The relative contribution of genetics vs. environment may be variable, where in certain populations the genetic role may be more important than the environment or vice versa. Consequently, it becomes increasingly complicated to identify genes that predispose to a disease when the genetic composition of an individual alone does not determine the disease phenotype.

Genomic Studies for Complex Diseases Require Huge Sample Sizes

Before the advent of genome-wide technology (prior to the year 2,000), genetic studies for complex diseases were typically designed as candidate gene analyses. Genes were selected based on their known function being consistent with the known physiology of type 2 diabetes. These studies were successful at identifying genes for rare, extreme forms of type 2 diabetes that segregate as monogenic Mendelian disorders. These single-gene disorders include maturity onset diabetes of the young (MODY) [23– 28 ] , neonatal diabetes [29 ] , and mitochondrial diabetes with deafness [ 30] . Subtypes of MODY have been clinically assigned based on the speciﬁ c gene whose variation gave rise to the disease [31 ] . However, candidate gene studies were not highly successful in identifying genes that have a role in common forms of type 2 diabetes, where the mode of inheritance is unknown. Many nominal genetic associations with type 2 diabetes can be found in the literature, but few could be independently replicated. Although it is arguable what constitutes a “validated” gene for type 2 diabetes, it is generally accepted that nonsynonymous variants in two genes identiﬁ ed via candidate gene analysis affect susceptibility for common type 2 diabetes. A proline to alanine substitution (Pro12Ala) in the peroxisome proliferator-activated receptor-gamma 2 is protective of type 2 diabetes [ 32, 33 ] , whereas a glutamic acid to lysine substitution (E23K) in the inward rectifying potassium channel (KCNJ11) increases risk for type 2 diabetes [34 ] . 5 Prediabetes Genes in Pima and Amish 63

The ability to interrogate the entire genome became possible as a result of the Human Genome Project [35, 36 ] and the International HapMap project which has identi fi ed nearly three million single nucleotide polymorphisms (SNPs) [ 37, 38 ] . Knowledge of the chromosomal location of each of these SNPs allowed them to be used as markers to interrogate the entire human genome for association studies. In 2007, fi ve reports emerged which utilized genome-wide association studies (GWASs) to identify SNPs that were reproducibly associated with type 2 diabetes [ 39– 43] . In contrast to candidate gene studies, GWASs are hypothesis-free, and these initial studies as well as subsequent GWAS publications have reported a large number of genes where their association with a phenotype is reproducible across many studies. The success of these studies was largely due to technical advance- ments that allowed genotyping of thousands of DNA samples using a microchip technology. In addition, since studies were being done globally using similar technology, data sharing of an unprecedented number of samples also became feasible. The ability to dramatically increase statistical power through international collaborative data sharing has allowed detection of dozens of susceptibility genes with nominal affect size, but whose associations replicate across many studies. The type 2 diabetes risk gene with the largest population effect size is TCF7L2 [ 44, 45 ] . SNPs in this gene have been consistently associated with type 2 diabetes in studies of individuals of various ethnicities, one exception being American Indians [ 46 ] . A meta-analysis of 28 individual studies gave a pooled odds ratio (OR) of 1.46 with a con fi dence interval of 1.42–1.51 and highly signi fi cant p value of 5.4 × 10 −140 for variants in this gene [ 47] . In 2011, the number of independent loci showing genome- wide signi fi cant associations with type 2 diabetes was 44 [48 ] . Combined, these loci still only account for ~10% of the observed familial clustering in Europeans, leaving much of the heritability unexplained [48 ] .

Genetic Isolates Reduce Complexity of Heterogeneous Populations

Genome-wide studies in Europeans have shown that common forms of type 2 diabetes are highly polygenic [ 49 ] and “successful identiﬁ cation of genes” (i.e., identifying SNPs whose effect can be replicated between studies) has typically required analysis of tens of thousands of samples [ 47 ] . However, huge sample sizes may not always be required to detect genes underlying common complex diseases. Studies in genetically isolated populations, with a limited number of founders, have proven successful in uncovering rare recessive disease genes, where these disease alleles are enriched, thus resulting in homozygote individuals affected by the disease [ 50 ] . Population studies to identify genes for a complex disease such as type 2 diabetes might also be more amenable in genetically isolated populations. In addition to having reduced allelic and locus heterogeneity and extended linkage disequilibrium across chromosomal regions, genetic isolates may also have more environmental 64 L.J. Baier and phenotypic homogeneity. Isolated populations also offer easier access to family members spanning multiple generations that allow for analysis of extended pedi- grees. Studies of type 2 diabetes are currently being pursued in Pima Indians living in Arizona [ 51 ] , as well as other genetic isolates, such as individuals living in a western region of Finland [52 ] . The Ashkenazi Jews [53 ] and the Amish living in Lancaster County are similarly being studied as societal isolates [54 ] . As a consequence of founder effects and genetic drift, each isolate may have a unique set of rare disease alleles, although it might be expected that older variants that are more common will be shared among different isolates. Some rare disease-associated alleles that are readily detected in one population isolate may go undetected in others, necessitating the use of multiple isolates to get a picture of the full spectrum of variants that affect risk for disease. Identiﬁ cation of rare high-impact alleles may be of critical importance for our understanding of metabolic pathways that lead to common diseases or traits. Knowledge of these pathways could be utilized in drug targeting or personalized medicine.

Prediabetic Traits Reduce Complexity of Phenotype

In addition to utilizing genetic isolates to reduce genotypic and environmental variability in the sample, the complexity of the disease phenotype can be reduced. Clinically, type 2 diabetes mellitus is a heterogenous disorder deﬁ ned by high blood sugar. Hyperglycemia occurs through the combination and interaction of two mechanisms: (1) insulin resistance in skeletal muscle, liver, and adipose tissue; and (2) abnormal insulin secretion due to pancreatic beta-cell defects (reviewed in [55 ] ). Reduced insulin action (insulin resistance) can be the consequence of obesity or could be caused by insulin resistance independent of obesity. Reduced insulin secretion can be the consequence of either beta-cell dysfunction or a reduced beta- cell mass.

Obesity

Typically, both younger and older individuals who develop type 2 diabetes are overweight or obese. Among ten populations representing different areas of the world, strong correlations were observed between the mean percent of the standard weight of a population and its prevalence of type 2 diabetes [ 56 ] . In addition to being more obese, on average, than individuals with normal glucose tolerance, individuals with type 2 diabetes tend to have a more central distribution of body fat. Hartz at al sur- veyed more than 30,000 American women and reported a higher prevalence of diabetes in individuals with a greater proportion of body fat at the waist as compared to the hips (waist to hip ratio) even at comparable measures of BMI [57 ] . Independent of the severity of obesity, the duration of obesity is an important risk factor for type 5 Prediabetes Genes in Pima and Amish 65

2 diabetes [58 ] , suggesting that a cumulative exposure to excess body weight may hasten the onset of type 2 diabetes among individuals. In Pima Indians, the risk for diabetes is twice as high in individuals who have been obese for 10 or more years as compared to individuals who have been obese for less than 5 years [ 59 ] . BMI and percentage of body fat are heritable traits [ 60 ] , suggesting that genetic variation contributes to these prediabetic traits.

Insulin Resistance

Defects in Insulin Action

Individuals with type 2 diabetes have impaired insulin action due to insulin resistance [55 ] . In the postprandial state, adipose tissue accounts for less than 5% of whole-body glucose uptake, whereas skeletal muscle accounts for more than 80%. The hyperinsulinemic, euglycemic clamp technique has made it possible to quantify skeletal muscle and hepatic insulin sensitivity [ 61 ] . Studies have demonstrated reduced insulin-mediated glucose uptake in muscle of individuals with type 2 diabetes [62 ] ; however, nondiabetic individuals who are at high risk for developing type 2 diabetes also have decreased insulin-stimulated glucose disposal rates. For example, offspring and ﬁ rst-degree relatives of patients with type 2 diabetes, who themselves are normal glucose-tolerant, exhibit moderate-to-severe insulin resistance [63 ] . Among nondiabetic Pima Indians who were studied over a period of 12 years, insulin action as assessed using the hyperinsulinemic, euglycemic clamp technique was highly predictive of future onset of type 2 diabetes [64 ] . Insulin action is a heritable trait [60 ] suggesting that genetic variation contributes to insulin resistance.

Insulin Secretion

The relationship between glycemia and insulin secretory function has been studied using the intravenous glucose tolerance test. An intravenous glucose bolus induces a biphasic insulin response and individuals with type 2 diabetes have a marked deﬁ cit in the ﬁ rst-phase response, or the acute insulin response [ 65 ] . Among Pima Indians with normal glucose tolerance who were studied over a period of 12 years, a reduction in acute insulin response was predictive of future onset of type 2 diabetes [ 64 ] . Figure 5.1 shows Pima Indians who were clinically characterized when they were normal glucose-tolerant. Those who were both insulin-resistant and had an AIR below the median at baseline had the highest cumulative incidence rate for type 2 diabetes (48%) at a 12-year follow-up and those who were both insulin-sensitive and had an AIR above the median at baseline had the lowest cumulative incidence for type 2 diabetes (11%) at follow-up [ 64 ] . Early in the disease progress leading to 66 L.J. Baier

Fig. 5.1 Cumulative incidence rates by 12 years for subjects who at baseline had either: (1) insulin sensitivity and a high AIR (above the population median); (2) insulin sensitivity and a low AIR (below the population median); (3) insulin resistance and a high AIR; (4) insulin resistance and a low AIR. Adapted from reference [64 ]

type2 diabetes, hyperinsulinemia is already present as a compensatory mechanism for insulin resistance. As the insulin resistance worsens with progression from NGT to an impaired fasting glucose (IFG)/IGT state, there is further decline in beta-cell function and this eventually leads to overt diabetes [ 66 ] . AIR is a highly familial trait, even after controlling for percentage of body fatness and insulin action [ 60 ] . Genetic variation that could affect either the insulin secretory pathway or islet mass could lead to an abnormal insulin response.

Identifying Genes for These Prediabetic Traits

Type 2 diabetes is very polygenic [49 ] and is independently predicted by obesity, abnormal insulin secretion, and insulin resistance which are all heritable traits themselves [55, 60 ] . Figure 5.2 depicts a scenario where the genes that underlie a single prediabetic trait may be fewer in number and therefore easier to detect than trying to identify all genes that underlie the complex heterogenic disease of type 2 diabetes. Therefore, several genetic studies have been designed to identify genes for prediabetic traits. These studies must be done in groups of individuals for whom these prediabetic clinical measures are available. Two such populations are the Pima Indians of Arizona and the Amish in Lancaster County. 5 Prediabetes Genes in Pima and Amish 67

Fig. 5.2 Reducing the complexity of type 2 diabetes. Identifying genes for individual traits that predict type 2 diabetes may be easier than identifying genes for the disease itself

Population Isolates for the Study of Prediabetic Traits

The Pima Indians of Arizona

The Pima Indians are believed to have lived in Arizona for more than 2,000 years. They are Paleoindians, whose ancestors migrated from Asia to North America in the ﬁ rst of three migrations across the Bering land bridge. They initially settled in Mexico. The modern-day Pima Indian likely descended from the Hohokam who moved from Mexico into the Gila River Valley of what is now Arizona in approximately 300 bc [22 ] . Modern-day Pima Indians living in Arizona have minimal European admixture [67 ] . The Pima were historically an agricultural-based society and dug irrigation canals to transport water from the Gila River to their crops in the desert (reviewed in [ 68] ). Descriptions of Pima Indians from the early 1900s suggest that diabetes was either rare or not diagnosed at that time [ 69, 70] . In subsequent years, the increasing settlement of the west by people of European ancestry led to diversion of the water supply and disruption of the Pima agriculture. The loss of water resulted in curtailment of subsistence farming and led to fundamental changes in the Gila River Pima lifestyle where most of the tribe became dependent upon government- issued foods. In the late 1930s, Joslin reviewed medical records from hospitals serving the Pima population and identi ﬁ ed 21 individuals with diabetes [71 ] . From these records, he concluded that the prevalence of diabetes was similar to that in the U.S. population. By the 1950s, many more Pima Indians were known to have diabetes and the trend in the incidence rate of type 2 diabetes began to be documented. From 1965 to 2008, systematic testing for diabetes was performed in Pima Indians 68 L.J. Baier

Fig. 5.3 Diabetes prevalence in the Gila River Indian Community in Arizona

living in the Gila River Indian community as part of a longitudinal study of type 2 diabetes [ 72 ] . Individuals (> 5 years of age) who resided in the Gila River Indian community participated in a research examination approximately every 2 years regardless of health [72 ] . These individuals were predominately Pima or the closely related Tohono O’odham (Papago) Indians. These biennial examinations included a 75 g oral glucose tolerance test (OGTT) where diabetes status was assessed according to the criteria of the World Health Organization [ 73 ] . In addition, height and weight were measured, information on pregnancy and health of children was gathered, and diabetes complications were assessed. Genealogic information was also documented. This longitudinal study has shown that the Pima Indians of Arizona have the highest rates of prevalence and incidence of type 2 diabetes of any population in the world [ 74 ] . The prevalence of type 2 diabetes in individuals examined between the years 1995 and 2003 is shown in Fig. 5.3 . Although the reasons for this high prevalence and incidence of diabetes among the Pimas living in Arizona are not known with certainty, genetic factors and an increasing prevalence of obesity due to lifestyle changes are likely [ 75 ] . As shown in Fig. 5.4 , adult (at least 20 years of age) Pima Indians are, on average, more obese than adult non-Hispanic Whites, non- Hispanic Blacks, or Mexican Americans living in the United States [ 76 ] . Studies of Pima Indians from Maycoba, Mexico, who share considerable genetic similarity with Pima Indians living in the United States but are much leaner and have remarkably low prevalence rates of type 2 diabetes, support the notion of an epidemic of diabetes in the Gila River Community coinciding with increased contact with European 5 Prediabetes Genes in Pima and Amish 69

Fig. 5.4 Prevalence of obesity (BMI > 30 kg/m2 ) in adults (>20 years) living in the United States. Data for non-Hispanic White, non-Hispanic Black, and Mexican Americans adapted from reference [76 ]

Americans and the ensuing change in lifestyle [ 22, 75 ] . This rapid rise in incidence of diabetes in Pimas living in Arizona has been followed by a relatively stable incidence rate, but with a shift to younger age at onset of diabetes [77 ] . During the past 43 years, the incidence of diabetes among Pima Indians less than 15 years of age has increased nearly sixfold [78 ] . This shift to younger ages of diabetes onset is likely a consequence of increasing rates of obesity in children and young adults [ 77, 78 ] . The declining incidence of diabetes among Pimas aged 25–34 years may re fl ect, in part, a shift to a younger age at onset in those at greatest risk. Despite diabetes being diagnosed in Pima children as young as 3 years of age, diabetes in this American Indian tribe is exclusively type 2. Diabetes among Pima children is characterized by the lack of insulin dependence, absent or low levels of islet cell and glutamic acid decarboxylase antibodies, and absence of strong linkage or association with known genes for MODY [ 79– 81] . The degree of heritability of diabetes is greater among Pima Indians who develop the disease at a younger age (onset age <45 years) as compared to subjects who develop diabetes at older ages [82 ] . It is assumed that diabetes with an onset at older ages may be more environmentally infl uenced. The absence of type 1 diabetes and the minimal European admixture in this population may indicate limited genetic and environmental variability in the etiology of type 2 diabetes in the Pima Indians (i.e., the disease is more “homogeneous”). In addition, the relatively young age of onset in Pima Indians allows for a better estimate of “affected” vs. “unaffected” status for a given individual. Beginning in 1982, a subset of the Pima Indian population living in Arizona was asked to volunteer yearly for a ~10-day inpatient stay in a Clinical Research Center 70 L.J. Baier of the National Institutes of Health [83 ] . The purpose of these examinations was to assess metabolic traits in individuals who were at risk for developing type 2 diabetes but were nondiabetic at the time of examination. Physiologic tests included a 75 g OGTT where venous blood was collected for measures of insulin and glucose levels at fasting, 30, 60, and 120 min. Body composition, such as percentage of body fatness, was measured by underwater weighing and later by dual X-ray absorptometry. The fi rst-phase, acute insulin release, was measured in response to a 25 g intravenous glucose tolerance. Insulin resistance was assessed using the euglycemic hyperinsulinemic clamp with glucose tracers to measure insulin action in vivo as well as rates of endogenous glucose production [ 83 ] . DNA for genetic analysis of these prediabetic traits is available on approximately 600 of these nondiabetic subjects.

The Amish of Lancaster County

In 1995, the Amish Family Diabetes Study was initiated in Lancaster PA to identify genetic determinants for type 2 diabetes and diabetes-related traits in this closed founder population [ 84 ] . The Amish, named after their leader Jacob Ammann, immigrated to the United States from Western Europe (primarily Switzerland) to escape religious persecution. The earliest immigrants arrived in 1727 and settled in Pennsylvania, whereas later groups settled in other midwestern states. Approximately 200 families settled in Lancaster County Pennsylvania and are considered to be the founders of the current Lancaster Amish community [ 85 ] . In the year 2000, the Amish population near Lancaster exceeded 30,000 [86 ] . The Amish are a conservative Christian sect whose rural lives are guided by the Ordnung, which promotes religious devotion, family, and community cohesion. Their livelihood is predominately based on farming, although modern technologies such as electric power and self-powered farm equipment are banned. The Amish community remains fairly isolated from their surrounding culture and are known for their antiquated style of dress and transportation by horse and buggy since personal automobiles are also banned. They do not allow outsiders to marry into their sect and rarely relocate and thus represent a genetic isolate [ 87 ] . First cousin marriages are not allowed, but on average Old Order Amish married couples are more closely related than second cousins once removed but less related than second cousins [ 88 ] . Families are large (average number of children per family is 6–7) and family gene- alogies are well documented dating back to the early 1700s (12–14 generations) [84 ] . The accurate and extensive genealogic records of the Amish are a tremendous asset for genetic studies. The Amish are very physically active due to their abstinence from modern con- veniences, yet Snitker et al. found that Amish adults are just as obese as the general US non-Hispanic population (mean BMI = 27.9 kg/m2 in Amish as compared to 27.0 kg/m 2 in non-Hispanic Whites) [ 89 ] . As shown in Fig. 5.5 , Snitker et al. also determined that the Amish are nearly as likely to have an abnormal OGTT as non- Hispanic Whites, yet have half the rate of type 2 diabetes as compared to the 5 Prediabetes Genes in Pima and Amish 71

Fig. 5.5 Prevalence of impaired glucose tolerance (IGT) or type 2 diabetes (a ) and of type 2 diabetes (b ) in non-Hispanic Whites and Amish (from reference [89 ] )

non-Hispanic White general US population (6.7% in Amish as compared to 14.3% in non-Hispanic Whites) [89 ] . These cross-sectional data suggest that the Amish have a low rate of conversion from IGT to diabetes. The lower prevalence of type 2 diabetes in the Amish compared with other Caucasians, despite the fact that both groups are, on average, equally obese, suggests that physical activity protects against type 2 diabetes independent of weight loss. It is possible that the high levels of physical activity among the Amish is prohibiting or prolonging the transition from IGT to full-blown diabetes [89 ] . As part of the Amish Family Diabetes Study, 953 adults (53% female) from 45 multigenerational families were studied over a 3-year period [84 ] . The mean sibship size was 4.5 (range 1–16). DNA and phenotypic data collected as part of this study have been utilized in candidate gene and hypothesis-free genome-wide studies. For example, several genes whose association with type 2 diabetes has been well replicated in other studies of Caucasians, such as TCF7L2 and HNFA, are also associated with type 2 diabetes in the Amish [90, 91 ] . However, other genes such as GRB10 which was the strongest signal in an Amish GWAS but was not among the top signals in other GWAS studies may have a larger effect in the Amish as compared with other Caucasian populations [ 92 ] . GWAS data from the Amish study have also been included in a large meta-analysis to identify genetic determinants of fasting glucose homeostasis [ 93 ] . 72 L.J. Baier

Genes Associated with Prediabetic Traits in Pima Indians and Amish

Table 5.1 summarizes genes that have been identifi ed by either candidate gene or genome-wide studies that contain variant(s) associated with a prediabetic trait in Pima Indians. Some of these variants were also nominally associated with type 2 diabetes, while others likely had an effect size on diabetes that was too small to be detected. For example, the Pro12Ala variant in PPARgamma is associated with increased insulin sensitivity and is protective for type 2 diabetes in Pima Indians [94 ] . In contrast, two genes (CDKAL1 and HHEX) that are reproducibly associated with type 2 diabetes in many studies of Caucasians are associated with insulin secretion among Pima Indians with normal glucose tolerance, but a signi fi cant association with type 2 diabetes cannot be detected in these American Indians [95 ] . Some of these genes associated with prediabetic traits in Pima Indians are also associated with either type 2 diabetes or prediabetes traits in the Amish. For example, MBL2, which encodes the mannose-binding lectin protein, was studied as a candidate gene for diabetes [ 96] . A variant in the promoter region of the MBL2 has been reported to in fl uence expression levels of this gene and this variant was associated with a decreased acute insulin response to an intravenous glucose bolus infusion in Pima Indians with normal glucose tolerance. This variant is also reproducibly associated with type 2 diabetes in two groups of American Indians as well as the Amish [ 96 ] . The frequency of the risk allele for prediabetes and overt diabetes was more common in Pima Indians as compared to the Amish. Despite the fact that the Amish are of European descent, this variant was not signi fi cantly associated with type 2 diabetes in a larger study of Europeans, which again suggests heterogeneity of the etiological factors of this disease [47 ] . Although insulin resistance is a predominant clinical feature of type 2 diabetes, most of the susceptibility genes identifi ed for type 2 diabetes affect insulin secretion, but not insulin sensitivity [112 ] . One exception is the apoptosis signal-regulating kinase 1 gene (ASK1; also known as MAP3K5 ). A noncoding variant in the sequence of ASK1 increases risk for both insulin resistance and type 2 diabetes in Pima Indians [ 99 ] . This variant appears to function by reducing ASK1 expression levels in skeletal muscle, and ASK1 expression levels in muscle biopsies from nondiabetic Pima Indians are correlated with the individual’s level of insulin resistance [ 98 ] . A surprising prediabetes gene among the Pima Indians was HLA-DRB1. A SNP was identi fi ed that was highly correlated with HLA-DR2, with the molecular allele DRB1 * 1602 [ 99] . Individuals with the nucleotide variant (marker for DR2+ ) had a lower prevalence of type 2 diabetes, and among Pima Indians with normal glucose tolerance, individuals who were DR2+ had signifi cantly higher plasma insulin levels in response to an intravenous glucose challenge even after adjusting for body fatness and insulin sensitivity. This was confi rmed by measures from an OGTT where the mean 30 min plasma insulin concentration was also higher in those who were DR2+ even after adjusting for their 30-min plasma glucose correlation which was lower in those who were DR2+ .These data suggest that individuals who have the HLA 5 Prediabetes Genes in Pima and Amish 73

Table 5.1 Genes associated with prediabetic traits in nondiabetic Pima Indians or Old Order Amish Metabolic traits affected Insulin Insulin % Body Gene symbol Gene name secretion action fat References GRB10 Growth factor receptor-bound + [92 ] protein 10 PPARg Peroxisome proliferator- + [ 94 ] activated receptor gamma CDKAL1 CDK5 regulatory subunit- + [ 95 ] associated protein 1 HHEX Hematopoietically expressed + [ 95 ] homeobox MBL2 Mannose-binding lectin + [ 96 ] (protein C) 2 ASK1 Apoptosis signal-regulating + [ 97 ] kinase 1 (ASK1) HLA-DRB1 Major histocompatibility + + [ 98 ] complex, class II, DR beta 1 ACAD10 Acyl-CoA dehydrogenase + [ 99 ] family, member 10 PCLO Presynaptic cytomatrix protein + [100 ] CACNA1E Calcium channel, voltage- + [ 101 ] dependent, R type, alpha 1E subunit PTP1B Protein tyrosine phosphatase, + [ 102 ] nonreceptor type 1 ARHGEF11 Rho guanine nucleotide + [ 103, 104 ] exchange factor (GEF) 11 CHRM3 Cholinergic receptor, + [ 105 ] muscarinic 3 ARHGEF12 Rho guanine nucleotide + [ 106 ] exchange factor (GEF) 12 IRS-1 Insulin receptor substrate 1 + [107 ] A2BP1 Ataxin-2 binding protein 1 + + [108 ] FTO Fat mass and obesity-associated + [95, 109 ] TCF7L2 Transcription factor 7-like 2 + + [110 ] Sim1 Single-minded homolog 1 + [ 111 ]

DRB1* 02 allele have a lower risk of type 2 diabetes due to protection from an autoimmune-mediated loss of insulin secretion, resulting from loss of beta-cell mass or beta-cell function. This is remarkable in that for many years the DR2 allele has been known to be predictive of type 1 diabetes, but was thought to have no role in other types of diabetes. It appears that the autoimmune-mediated loss of insulin secretion in people with type 1 diabetes is just an extreme case of a more mild, and much more common, biologic process apparent even in people with normal glucose tolerance who are not predisposed to type 1 diabetes. 74 L.J. Baier

The Future of Genetics of Complex Diseases

The genetics of diabetes and prediabetes is still in its early phases where most of the studies have focused on genome-wide variation at the SNP level. Although GWASs have uncovered many SNPs with small effects on type 2 diabetes and prediabetic traits, they have explained relatively little of the heritability of these traits. This situation is not unique to type 2 diabetes and has led researchers to question the “missing heritability” of complex diseases [113, 114 ] . Until recently, it was generally assumed that common diseases, such as type 2 diabetes, were caused by common variants [113 ] . GWASs seemed to be a valid technique to uncover these variants because they provided genotypic information on common variants (minor allele frequency >0.05). However, it is now believed that susceptibility to common complex diseases involves the contribution of both common variants and rare mutations, and the relative impact of each in a particular trait or disease may vary among populations. Rare variants may have a substantial effect size on a given phenotype. Although rare variants are, by de fi nition, rare by themselves, a particular population could harbor a large number of unique rare variants which, in combination, explains a considerable portion of the variance in a phenotype. Consequently, rare variants may explain a portion of the “missing heritability” and there has been a major shift towards whole-genome sequencing to identify rare variants that may have important roles in susceptibility for complex diseases [ 115 ] . Another explanation for the “missing heritability” from GWASs is that these studies provide information on variation due to single nucleotide changes, yet the human genome has much diversity beyond SNPs [116 ] . Other sources of variation such as large deletions, duplications, and inversions have yet to be thoroughly analyzed in studies of complex disease, yet it is estimated that 8% of individuals have a large (>500 kb) deletion or duplication that occurs at an allele frequency of <0.05% [ 114, 117 ] . Copy-number variants (CNVs) are important in that they have been subjected to sudden, rapid, and often adaptive, evolution in human populations. CNV analysis is in its early stages [ 118 ] , and although of great biological and evo- lutionary interest, the highly repetitive nature of these genes make them dif fi cult to genotype or sequence with a high degree of certainty with current technology. The parent that transmitted a particular SNP to the offspring (“parent of origin”) may also determine whether or not a variant has functional consequences [ 114 ] . Susceptibility variants for cancer and type 2 diabetes have been identi fi ed which confer risk only when inherited from a specifi c parent, and a variant was discovered that can either increase or reduce risk for type 2 diabetes depending on the parent of origin [ 119] . Few studies have DNA on both parents and offspring, and therefore cannot assess parent of origin effects. Therefore, these variants may contribute to missing heritability in that they are more diffi cult to discover and, even if discovered, their contribution to heritability would be underestimated when evaluated under models that do not take parental origin into account [114 ] . Epigenetic modi fi cations are heritable changes in gene function that can be passed from one generation to the next, yet are not caused by changes in nucleotide 5 Prediabetes Genes in Pima and Amish 75 sequence. Epigenetic factors include DNA methylation, histone modifi cation, RNA processing, and microRNA expression, and their effects on phenotype may additionally be in fl uenced by the environment. Transgenerational epigenetic effects can persist across multiple generations, and their effect size can be as strong as more conventional Mendelian inheritance [114 ] . Because transgenerational effects loosen conventional genotype–phenotype associations, genomic sequencing of offspring may fail to uncover susceptibility factors derived from a prior generation. Studies on the relationship between epigenetic factors and risk for type 2 diabetes are ongoing and may explain another portion of the “missing heritability.” As our understanding of these genetic mechanisms increases, it is likely that we will have better tools to predict, prevent, and treat type 2 diabetes in people of all ethnicities.

References

1. Wild S, Roglic G, Green A, et al. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care. 2004;27:1047–53. 2. King H, Aubert RE, Herman WH. Global burden of diabetes, 1995–2025: prevalence, numeri- cal estimates, and projections. Diabetes Care. 1998;21:1414–31. 3. Zimmet P, Alberti KG, Shaw J. Global and societal implications of the diabetes epidemic. Nature. 2001;414:782–7. 4. Zimmet P. Type 2 (non-insulin-dependent) diabetes—an epidemiological overview. Diabetologia. 1982;22(6):399–411. 5. Papoz L, Ben Khalifa F, Eschwege E, et al. Diabetes mellitus in Tunisia: description in urban and rural populations. Int J Epidemiol. 1988;17(2):419–22. 6. Zimmet P. Epidemiology of diabetes and its macrovascular complications in Paciﬁ c populations: the medical effects of social progress. Diabetes Care. 1979;2:144–53. 7. Zimmet P, Taylor R, Parshu R, et al. Prevalence of diabetes and impaired glucose tolerance in the biracial (melanesian and indian) population of Fiji: a rural-urban comparison. Am J Epidemiol. 1983;118:673–88. 8. Chow CK, Raju PK, Raju R, et al. The prevalence and management of diabetes in rural India. Diabetes Care. 2006;29:1717–8. 9. Garduño-Diaz SD, Khokhar S. Prevalence, risk factors and complications associated with type 2 diabetes in migrant South Asians. Diabetes Metab Res Rev. 2012;28:6–24. 10. McBean AM, Li S, Gilbertson DT, et al. Differences in diabetes prevalence, incidence, and mortality among the elderly of four racial/ethnic groups: Whites, Blacks, Hispanics, and Asians. Diabetes Care. 2004;27(10):2317–24. 11. Stern MP, Gaskill SP, Hazuda HP, et al. Does obesity explain excess prevalence of diabetes among Mexican Americans? Results of the San Antonio Heart Study. Diabetologia. 1983;24(4):272–7. 12. Baxter J, Hamman RF, Lopez TK, et al. Excess incidence of known non-insulin-dependent diabetes mellitus (NIDDM) in Hispanics compared with non-Hispanic whites in the San Luis Valley, Colorado. Ethn Dis. 1993;3(1):11–21. 13. Newman B, Selby JV, King MC, et al. Concordance for type II (non-insulin-dependent) diabetes mellitus in male twins. Diabetologia. 1987;30:763–8. 14. Kaprio J, Tuomilehto J, Koskenvuo M, et al. Concordance for type 1 (insulin-dependent) and type 2 (non-insulin-dependent) diabetes mellitus in a population-based cohort of twins in Finland. Diabetologia. 1992;35:1060–7. 76 L.J. Baier

15. Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346:393–403. 16. Hu FB. Globalization of diabetes: the role of diet, lifestyle and genes. Diabetes Care. 2011;34:1249–57. 17. Wardle J, Carnell S, Haworth CMA, et al. Evidence for a strong genetic in ﬂ uence on childhood adiposity despite the force of the obesogenic environment. Am J Clin Nutr. 2008;87:398–404. 18. Stunkard AJ, Sorensen TI, Hanis C, et al. An adoption study of human obesity. N Engl J Med. 1986;314(4):193–8. 19. Segal NL, Allison DB. Twins and virtual twins: bases of relative body weight revisited. Int J Obes Relat Metab Disord. 2002;26(4):437–41. 20. Amed S, Daneman D, Mahmud FH, et al. Type 2 diabetes in children and adolescents. Expert Rev Cardiovasc Ther. 2010;8:393–406. 21. Wareham NJ, Franks PW, Harding AH. Establishing the role of gene-environment interac- tions in the etiology of type 2 diabetes. Endocrinol Metab Clin North Am. 2002;31:553–66. 22. Ravussin E, Valencia ME, Esparza J, et al. Effects of a traditional lifestyle on obesity in Pima Indians. Diabetes Care. 1994;17(9):1067–74. 23. Yamagata K, Furuta H, Oda N, et al. Mutations in the hepatocyte nuclear factor-4alpha gene in maturity-onset diabetes of the young (MODY1). Nature. 1996;384:458–60. 24. Vionnet N, Stoffel M, Takeda J, et al. Nonsense mutation in the glucokinase gene causes early-onset non-insulin-dependent diabetes mellitus. Nature. 1992;23:721–2. 25. Malecki MT, Jhala US, Antonellis A, et al. Mutations in NEUROD1 are associated with the development of type 2 diabetes mellitus. Nat Genet. 1999;23:323–8. 26. Stoffers DA, Ferrer J, Clarke WL, et al. Early-onset type-II diabetes mellitus (MODY4) linked to IPF1. Nat Genet. 1997;17:138–9. 27. Horikawa Y, Iwasaki N, Hara M, et al. Mutation in hepatocyte nuclear factor-1 beta gene (TCF2) associated with MODY. Nat Genet. 1997;17:384–5. 28. Yamagata K, Oda N, Kaisaki PJ, et al. Mutations in the hepatocyte nuclear factor-1alpha gene in maturity-onset diabetes of the young (MODY3). Nature. 1996;384:455–8. 29. Flanagan SE, Clauin S, Bellanne-Chantelot C, et al. Update of mutations in the genes encoding the pancreatic beta-cell K(ATP) channel subunits Kir6.2 (KCNJ11) and sulfonylurea receptor 1 (ABCC8) in diabetes mellitus and hyperinsulinism. Hum Mutat. 2009;30(2):170–80. 30. van den Ouweland JMW, Lemkes HHPJ, Ruitenbeek W, et al. Mutation in mitochondrial tRNA Leu(UUR) gene in a large pedigree with maternally transmitted type II diabetes mellitus and deafness. Nat Genet. 1992;1:368–71. 31. Murphy R, Ellard S, Hattersley AT. Clinical implications of a molecular genetic classiﬁ cation of monogenic beta-cell diabetes. Nat Clin Pract Endocrinol Metab. 2008;4(4):200–13. 32. Beamer BA, Yen C-J, Andersen RE, et al. Association of the Pro12Ala variant in the peroxisome proliferator-activated receptor-g 2 gene with obesity in two Caucasian populations. Diabetes. 1998;47:1806–8. 33. Altshuler D, Hirschhorn JN, Klannemark M, et al. The common PPAR gamma Pro12Ala polymorphism is associated with decreased risk of type 2 diabetes. Nat Genet. 2000;26(1):76–80. 34. Gloyn AL, Weeden MN, Owen KR, et al. Large-scale association studies of variants in genes

encoding the pancreatic b -cell KATP channel subunits Kir6.2 (KCNJ11 ) and SUR1 (ABCC8 ) con ﬁ rm that the KCNJ11 E23K variant is associated with type 2 diabetes. Diabetes. 2003;52:568–72. 35. Lander ES, et al. Initial sequencing and analysis of the human genome. Nature. 2001; 409:860–921. 36. Venter JC, et al. The sequence of the human genome. Science. 2001;291:1304–51. 37. The International HapMap Consortium. The international HapMap project. Nature. 2003; 426:789–96. 38. Frazer KA, et al. A second generation human haplotype map of over 3.1 million SNPs. Nature. 2007;449:851–61. 5 Prediabetes Genes in Pima and Amish 77

39. Sladek R, Rocheleau G, Rung J, et al. A genome-wide association study identi fi es novel risk loci for type 2 diabetes. Nature. 2007;445:881–5. 40. Steinthorsdottir V, Thorleifsson G, Reynisdottir I, et al. A variant in CDKAL1 infl uences insulin response and risk of type 2 diabetes. Nat Genet. 2007;39:770–5. 41. Saxena R, Voight BF, Lyssenko V, Burtt NP, et al. Genome-wide association analysis identifi es loci for type 2 diabetes and triglyceride levels. Science. 2007;316:1331–6. 42. Scott LJ, Mohlke KL, Bonnycastle LL, et al. A genome-wide association study of type 2 diabetes in Finns detects multiple susceptibility variants. Science. 2007;316:1341–5. 43. Zeggini E, Weedon MN, Lindgren CM, et al. Replication of genome-wide association signals in UK samples reveals risk loci for type 2 diabetes. Science. 2007;316:1336–41. 44. Grant SF, Thorleifsson G, Reynisdottir I, et al. Variant of transcription factor 7-like 2 (TCF7L2) gene confers risk of type 2 diabetes. Nat Genet. 2006;38(3):320–3. 45. Cauchi S, El Achhab Y, Choquet H, Dina C, Krempler F, Weitgasser R, et al. TCF7L2 is reproducibly associated with type 2 diabetes in various ethnic groups: a global meta-analysis. J Mol Med. 2007;85(7):777–82. 46. Guo T, Traurig M, Muller YL, et al. TCF7L2 is not a susceptibility gene for type 2 diabetes in Pima Indians. Diabetes. 2007;56:3082–8. 47. Zeggini E, Scott LJ, Saxena R, et al. Meta-analysis of genome-wide association data and large-scale replication identifi es additional susceptibility loci for type 2 diabetes. Nat Genet. 2008;40(5):638–45. 48. Wheeler E, Barroso I. Genome-wide association studies and type 2 diabetes. Brief Funct Genomics. 2011;10:52–60. 49. McCarthy MI. Genomics, type 2 diabetes, and obesity. N Eng J Med. 2010;363:2339–50. 50. Kristiansson K, Naukkarinen J, Peltonen L. Isolated populations and complex gene identi fi cation. Genome Biol. 2008;9:109. 51. Baier LJ, Hanson RL. Genetic studies of the etiology of type 2 diabetes mellitus in Pima Indians: hunting for pieces to a complicated puzzle. Diabetes. 2004;53:1181–6. 52. Groop L, Forsblom C, Lehtovirta M, et al. Metabolic consequences of a family history of NIDDM (the Botnia Study): evidence for sex-specifi c parental effects. Diabetes. 1996;45: 1585–93. 53. Love-Gregory LD, Wasson J, Ma J, et al. A common polymorphism in the upstream promoter region of the hepatocyte nuclear factor-4 alpha gene on chromosome 20q is associated with type 2 diabetes and appears to contribute to the evidence for linkage in an Ashkenazi Jewish population. Diabetes. 2004;53(4):1134–40. 54. Hsueh WC, Mitchell BD, Aburomia R, et al. Diabetes in the Old Order Amish: characterization and heritability analysis of the Amish Family Diabetes Study. Diabetes Care. 2000;23(5):595–601. 55. Tataranni PA, Bogardus C. Metabolic abnormalities in the development of type 2 diabetes mellitus. In: LeRoith D, Taylor SI, Olefsky JM, editors. Diabetes mellitus. A fundamental and clinical text. Philadelphia: Lippincott Williams & Wilkins; 2004. p. 797–807. 56. West KM, Kalbfl eisch JM. Infl uence of nutritional factors on prevalence of diabetes. Diabetes. 1971;20:99. 57. Hartz AJ, Rupley DC, Rimm AA. The association of girth measurements with disease in 32,856 women. Am J Epidemiol. 1984;119:71. 58. Wannamethee SG, Shaper AG. Weight change and duration of overweight and obesity in the incidence of type 2 diabetes. Diabetes Care. 1999;22:1266–72. 59. Everhart JE, Pettitt DJ, Bennett PH, et al. Duration of obesity increases the incidence of NIDDM. Diabetes. 1992;41(2):235–40. 60. Sakul H, Pratley R, Cardon L, et al. Familiality of physical and metabolic characteristics that predict the development of non-insulin-dependent diabetes mellitus in Pima Indians. Hum Genet. 1997;60(3):651–6. 61. DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol. 1979;237:E214–23. 62. DeFronzo RA. Lilly Lecture 1987. The triumvirate: B-cell, muscle, liver: a collusion responsible for type 2 diabetes. Diabetes. 1988;637:667. 78 L.J. Baier

63. Warram JH, Martin BC, Krolewski AS, et al. Slow glucose removal and hyperinsulinemia precede the development of type 2 diabetes in the offspring of diabetic parents. Ann Intern Med. 1990;113:909. 64. Bunt JC, Krakoff J, Ortega E, et al. Acute insulin response is an independent predictor of type 2 diabetes mellitus in individuals with both normal fasting and 2-h plasma glucose concentrations. Diabetes Metab Res Rev. 2007;23(4):304–10. 65. DeFronzo RA, Ferrannini E. The pathogenesis of non-insulin—dependent diabetes: an update. Medicine (Baltimore). 1982;62:125. 66. Weyer C, Bogardus C, Mott DM, et al. The natural history of insulin secretory dysfunction and insulin resistance in the pathogenesis of type 2 diabetes mellitus. Clin Invest. 1999;104(6):787–94. 67. Williams RC, Steinberg AG, Knowler WC, et al. Gm 3;5,13,14 and stated-admixture: independent estimates of admixture in American Indians. Am J Hum Genet. 1986;39(3):409–13. 68. Knowler WC, Pettitt DJ, Saad MF, et al. Diabetes mellitus in the Pima Indians: incidence, risk factors and pathogenesis. Diabetes Metab Rev. 2009;6:1–27. 69. Hrdlicka A. The Bureau of American Ethnology: physiological and medical observations among the Indians of Southwestern United States and Northern Mexico. Washington, DC: U.S. Government Printing Of ﬁ ce; 1908 (Bulletin 34:1–347). 70. Russell F. The Pima Indians. In Twenty-Sixth Annual Report of the Bureau of American Ethnology to Secretary of Smithsonian Institution. Washington, DC: U.S. Government Printing Of ﬁ ce; 1908. p. 3–389. 71. Joslin EP. The universality of diabetes. JAMA. 1940;115:2033–8. 72. Knowler WC, Bennett PH, Hamman RF, Miller M. Diabetes incidence and prevalence in Pima Indians: a 19-fold greater incidence than in Rochester, Minnesota. Am J Epidemiol. 1978;108:497–505. 73. World Health Organization. Report of a WHO Study Group: Diabetes Mellitus. Geneva, World Health Org; 1985 (Tech. Rep. Ser., no. 727). 74. Knowler WC, Bennett PH, Hamman RF, Miller M. Diabetes incidence and prevalence in Pima Indians: a 19-fold greater incidence than in Rochester, Minnesota. Am J Epidemiol. 1978;108(6):497–505. 75. Schulz LO, Bennett PH, Ravussin E, Kidd JR, Kidd KK, Esparza J, et al. Effects of traditional and western environments on prevalence of type 2 diabetes in Pima Indians in Mexico and the U.S. Diabetes Care. 2006;29:1866–971. 76. Ogden CL, Carroll MD, Curtin LR, et al. Prevalence of overweight and obesity in the United States, 1999–2004. JAMA. 2006;295:1549–55. 77. Pavkov ME, Hanson RL, Knowler WC, et al. Changing patterns of type 2 diabetes incidence among Pima Indians. Diabetes Care. 2007;30:1758–63. 78. Dabelea D, Hanson RL, Bennett PH, et al. Increasing prevalence of type II diabetes in American Indian children. Diabetologia. 1998;41:904–10. 79. Savage PJ, Bennett PH, Senter RG, et al. High prevalence of diabetes in young Pima Indians. Diabetes. 1979;28:937–42. 80. Baier LJ, Permana PA, Traurig M, et al. Mutations in the genes for hepatocyte nuclear factor (HNF)-1alpha, -4alpha, -1beta, and -3beta; the dimerization cofactor of HNF-1; and insulin promoter factor 1 are not common causes of early-onset type 2 diabetes in Pima Indians. Diabetes Care. 2000;23(3):302–4. 81. Dabalea D, Palmer JP, Bennett PH, et al. Absence of glutamic acid decarboxylase antibodies in Pima Indian children with diabetes mellitus. Diabetologia. 1999;2:1265–6. 82. Hanson RL, Elston RC, Pettitt DJ, et al. Segregation analysis of non-insulin-dependent diabetes mellitus in Pima Indians: evidence for a major-gene effect. Am J Hum Genet. 1995;57(1):160–70. 83. Lillioja S, Mott DM, Spraul M, et al. Insulin resistance and insulin secretory dysfunction as precursors of non-insulin-dependent diabetes mellitus. Prospective studies of Pima Indians. N Engl J Med. 1993;329(27):1988–92. 84. Hsueh W-C, Mitchell BD, Aburomia R, et al. Diabetes in the old Order Amish: characterization and heritability of the Amish Family Diabetes Study. Diabetes Care. 2000;23(5):595–601. 85. Cross HE. Population studies and the old Order Amish. Nature. 1976;262:17–20. 5 Prediabetes Genes in Pima and Amish 79

86. Church Directory of the Lancaster County Amish. In association with Katie Beiler. Gordonsville, PA: Pequea publishers. 1996;1:320pp, 2:322pp. 87. McKusick VA. Medical genetic studies of the Amish. Baltimore, MD: Johns Hopkins University; 1978. 88. Khoury MJ, Cohen BH, Diamond EL, et al. Inbreeding and prereproductive mortality in the Old Order Amish. I. Genealogic epidemiology of inbreeding. Am J Epidemiol. 1987;125:453–61. 89. Snitker S, Mitchell BD, Shuldiner AR. Physical activity and prevention of type 2 diabetes. Lancet. 2003;361:87–8. 90. Damcott CM, Pollin T, Reinhart LJ, et al. Polymorphisms in the transcription factor 7-like 2 (TCF7L2) gene are associated with type 2 diabetes in the Amish: replication and evidence for a role in both insulin secretion and insulin resistance. Diabetes. 2006;55:2654–9. 91. Damcott CM, Hoppman N, Ott SH, et al. Polymorphisms in both promoters of hepatocyte nuclear factor 4-a are associated with type 2 diabetes in the Amish. Diabetes. 2004;53: 3337–41. 92. Rampersaud E, Damcott CM, Fu M, et al. Identi ﬁ cation of novel candidate genes for type 2 diabetes from a genome-wide association scan in the Old Order Amish. Diabetes. 2007; 56:3053–62. 93. Dupuis J, Langenberg C, Prokopenko I, et al. New genetic loci implicated in fasting glucose homeostasis and their impact on type 2 diabetes risk. Nat Genet. 2010;42:105–16. 94. Muller Y-L, Bogardus C, Beamer BA, et al. A functional variant in the PPARã2 promoter is associated with predictors of obesity and type II diabetes. Diabetes. 2003;52:1864–71. 95. Rong R, Hanson RL, Ortiz D, et al. Association analysis of FTO, CDKAL1, SLC30A8, HHEX, EXT2, IGF2BP2, OC387761 and CDKN2B with type 2 diabetes and pre-diabetic traits in Pima Indians. Diabetes. 2009;58:478–88. 96. Muller YL, Hanson RL, Bian L, et al. Functional variants in MBL2 are associated with type 2 diabetes and pre-diabetic traits in Pima Indians and the Old Order Amish. Diabetes. 2010;59:2080–5. 97. Bian L, Hanson RL, Ossowski V, et al. Variants in ASK1 are associated with skeletal muscle mRNA expression, in vivo insulin resistance, and type 2 diabetes in Pima Indians. Diabetes. 2010;59:1276–82. 98. Williams RC, Muller YL, Knowler WC, et al. HLA-DRB1 reduces the risk of type 2 diabetes mellitus by increased insulin secretion. Diabetologia. 2011;54:1684–92. 99. Bian L, Muller YL, Ma L, et al. Variants in the acyl-coenzyme A dehydrogenase 10 (ACAD10) gene are associated with type 2 diabetes, insulin resistance and decreased lipid oxidation in Pima Indians. Diabetologia. 2010;53:1349–53. 100. Ma L, Hanson RL, Que LN, et al. PCLO variants are nominally associated with early onset type 2 diabetes and insulin resistance in Pima Indians. Diabetes. 2008;57:3156–60.

101. Muller YL, Hanson RL, Zimmerman C, et al. Variants in the Cav 2.3 ( a1E) subunit of voltage- activated Ca2+ channels are associated with insulin resistance and type 2 diabetes in Pima Indians. Diabetes. 2007;56:3089–94. 102. Traurig M, Hanson RL, Kobes S, et al. Protein tyrosine phosphatase 1B gene is not a major susceptibility gene for type 2 diabetes mellitus or obesity among Pima Indians. Diabetologia. 2007;50:985–9. 103. Ma L, Hanson RL, Que LN, et al. Variants in ARHGEF11, a candidate gene for the linkage to type 2 diabetes mellitus on chromosome 1q, are nominally associated with insulin resistance and type 2 diabetes mellitus in Pima Indians. Diabetes. 2007;56:1454–9. 104. Fu M, Sabra MM, Damcott C, et al. Evidence that Rho guanine nucleotide exchange factor 11 (ARHGEF11) on 1q21 is a type 2 diabetes susceptibility gene in the Old Order Amish. Diabetes. 2007;56:1363–8. 105. Guo Y, Traurig M, Ma L, et al. A CHRM3 gene variation is associated with decreased acute insulin secretion and increased risk for early onset type 2 diabetes mellitus in Pima Indians. Diabetes. 2006;55:3625–9. 106. Kovacs P, Stumvoll M, Bogardus C, et al. A functional Tyr1306Cys variant in LARG is associated with increased insulin action in vivo. Diabetes. 2006;55:1497–505. 80 L.J. Baier

107. Kovacs P, Hanson R, Lee Y-H, et al. The role of insulin receptor substrate-1 gene (IRS1) in type 2 diabetes mellitus in Pima Indians. Diabetes. 2003;52:3005–9. 108. Ma L, Traurig MT, Hanson RL, et al. Evaluation of A2BP1 as an obesity gene. Diabetes. 2010;59:2837–45. 109. Rampersaud E, Mitchell BD, Pollin TI, et al. Physical activity and the association of common FTO gene variants with body mass index and obesity. Arch Intern Med. 2008;168(16): 1791–7. 110. Damcott CM, Pollin TI, Reinhart LJ, et al. Polymorphisms in the transcription factor 7-like 2 (TCF7L2) gene are associated with type 2 diabetes in the Amish: replication and evidence for a role in both insulin secretion and insulin resistance. Diabetes. 2006;55(9):2654–9. 111. Traurig M, Mack J, Hanson RL, et al. Common variation in SIM1 is reproducibly associated with body mass index in Pima Indians. Diabetes. 2009;58:1682–9. 112. Florez JC. Newly identiﬁ ed loci highlight beta cell dysfunction as a key cause of type 2 diabetes: where are the insulin resistance genes? Diabetologia. 2008;51(7):1100–10. 113. Manolio TA, Collins FS, Cox NJ, et al. Finding the missing heritability of complex diseases. Nature. 2009;461(7265):747–53. 114. Evan E, Eichler EE, Flint J, et al. Missing heritability and strategies for ﬁ nding the underlying causes of complex disease. Nat Rev Genet. 2010;11:446–50. 115. Park KS. The search for genetic risk factors of type 2 diabetes mellitus. Diabetes Metab J. 2011;35(1):12–22. 116. Scherer SW, et al. Challenges and standards in integrating surveys of structural variation. Nat Genet. 2007;39(Suppl):S7–15. 117. Itsara A, et al. Population analysis of large copy number variants and hotspots of human genetic disease. Am J Hum Genet. 2009;84:148–61. 118. Wellcome Trust Case Control Consortium, Craddock N, Hurles ME, et al. Genome-wide association study of CNVs in 16,000 cases of eight common diseases and 3,000 shared controls. Nature. 2010;464(7289):713–20. 119. Kong A, et al. Parental origin of sequence variants associated with complex diseases. Nature. 2009;462:868–74. Chapter 6 Predicting Diabetes

Rachel Dankner and Jesse Roth

Introduction

The astute reader, seeing the length of this chapter, will immediately surmise that the ability of the biomedical community to predict diabetes is quite limited, despite a larger number of studies. We encourage our readers to learn with us from past studies and to prepare for the more promising approaches presented near the close of the chapter.

Predicting Diabetes

These two commonly used words, “predicting” and “diabetes” are more ambiguous than they may seem. Ideally, to predict is to know about an event before it happens. A true predictor of diabetes would be a phenomenon that presents in advance and foretells the disease. However, a marker, such as obesity, is generally considered a

R. Dankner , MD, MPH (*) Unit for Cardiovascular Epidemiology, The Gertner Institute for Epidemiology and Health Policy Research, Sheba Medical Center , Tel Hashomer 52621, Israel Department of Epidemiology and Preventive Medicine, Sackler Faculty of Medicine , School of Public Health , Tel Aviv University, Ramat Aviv, Tel Aviv 69978 , Israel Patient Oriented Research, The Feinstein Institute for Medical Research, Manhasset, North Shore 11030, New York e-mail: [email protected]; [email protected] J. Roth , MD Laboratory of Diabetes & Metabolic Disorders, Elmezzi Graduate School of Molecular Medicine, The Feinstein Institute for Medical Research , Hofstra North Shore-LIJ School of Medicine , 350 Community Drive , Manhasset 11030 , New York Endocrinology Division, Department of Medicine, Albert Einstein College of Medicine, Yeshiva University, Bronx, NY 10461, USA e-mail: [email protected]

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 81 DOI 10.1007/978-1-4614-3314-9_6, © Springer Science+Business Media New York 2012 82 R. Dankner and J. Roth predictor of diabetes because it associates with a higher probability of disease occurrence than does its absence. Predicting diabetes with confi dence can now be done in only a small fraction of patients, such as those with inborn errors in the insulin receptor or insulin gene. Other conditions associated with the development of diabetes include cystic fi brosis after many years of disease, pancreatic atresia, and surgical extirpation of more than 50% of the pancreas. However, the rare occurrence of such disorders limits their use as predictors. Thus, for the vast majority of patients with type 2 diabetes, the accuracy of predictions is poor. Like others who perform badly, we, the authors, shall begin by blaming our tools, our predecessors, Mother Nature, and the diabetes establishment. We will survey major efforts in the fi eld to cull the most promising ideas and fi nally try to construct new pathways that may give more accurate predictions. During the past 30 years, a number of factors have been associated with the risk of diabetes, some due to empirical relationships without any known causal mechanisms, and others that have helped elucidate the pathophysiology of diabetes. This chapter traces the search for factors predicting diabetes by focusing on key longitudinal research studies. Some of these studies are considered pivotal, and others have received less attention, despite the contribution they have made to our accumulating knowledge. The strengths and limitations of population-based studies in identifying the risk of diabetes in individuals will be emphasized.

The Essence of Diabetes

Before continuing, we need to defi ne what is being predicted, i.e., what is diabetes. The authors in Chap. 2 reviewed the changing defi nitions of diabetes. Cut-off thresholds of fasting and postload glucose that have been used to diagnose diabetes have been lowered, as more modest elevations in blood glucose have been linked to macro- and microvascular complications. However, the question arises as to whether elevated blood glucose and the ensuing complications actually defi ne the disease. Remarkably, no specifi c biological marker has been identifi ed that presents in all individuals with diabetes and is absent in all those without. Neither does the threshold level of any biological marker unambiguously distinguish between those with and those without diabetes, or between those who will develop complications from diabetes, and those who will not. Retinopathy seemed to approach being such a biological marker. However, as we describe later in this chapter, current diagnostic tools enable detection of retinopathy at glucose levels well within the normal range of glucose, suggesting that this complication does not present at a threshold level but rather along a continuum of plasma glucose values. We face the philosophical question as to whether the ambiguity regarding the diagnosis of diabetes refl ects an ambiguity regarding its essence; if the essence of diabetes is an increased risk for other diseases, should the real goal be prediction of risk for the complications of diabetes? While rapid, user-friendly, and accurate methods of measuring glucose in blood have emerged, the concept of type 2 diabetes as a disease process has become more complicated. It has become increasingly clear that diabetes affects virtually all systems 6 Predicting Diabetes 83 in the body (not just eyes, peripheral nerves, kidney glomeruli, and arteries), and that blood glucose levels in the upper limits of the normal range can produce reversible and irreversible changes [1, 2 ] . Adding to this complexity are aging and obesity, very frequent partners of type 2 diabetes. Decreases over the lifespan in muscle mass, fatty acid oxidation, and mitochondrial function in skeletal muscle [ 3 ] , together with increased oxidative stress upon fat accretion add to and synergize with changes in type 2 diabetes [ 4– 6 ] . The metabolic syndrome (de fi ned by disruptions in insulin secretion and insulin action, dyslipidemias, and hypertension) occurs widely among obese, elderly, and those with diabetes, contributing as (1) mediators of pathology, (2) potential diagnostic aids, and (3) obfuscators of the borders between diseases. For instance, risks for cardiovascular disease differ according to the de fi nitions of the metabolic syndrome proposed by the International Diabetes Federation (IDF) and the Adult Treatment Panel (ATP) III [7 ] . For an individual considered at risk for developing diabetes, genetic, environmental, clinical, and lifestyle factors apparently determine whether (1) overt disease will actually manifest, (2) a high-risk state will persist, or (3) reversion to a healthy status will be achieved. Such factors may themselves be considered predictors. Ideally, the power of predictors should be compared through a prospective trial in which individuals with a certain characteristic, such as elevated fasting glucose, are exposed to controlled conditions, i.e., particular nutritional and physical regimens. Since such controlled trials are not possible in humans, we are left with observational studies. In fact, we do not really know why one individual develops type 2 diabetes and another does not. Nor do we know what primary defect causes the chain of events that leads to overt diabetes, though we do know which organs participate in this cascade, either by contributing to it or by being affected by it. Studies in recent years have complicated matters, raising questions as to the starting point of the pathology. Dozens of genetic variations have been identi fi ed, each of which may be contributing from the very beginning of life. In utero and perinatal epigenetic infl uences have been considered important; it is widely believed that both undernutrition and overnutrition early in life leave their marks into adulthood, possibly inducing lifelong effects. At the point at which current knowledge enables intervention, we cannot distinguish between causes and results. Nevertheless, prediction is important in light of growing evidence that diabetes can be prevented, or at least delayed. The goal seems to be identi fi cation of the earliest triggers or markers of disease.

Plasma Insulin and Glucose Levels as Predictors of Diabetes

In type 2 diabetes, insulin and glucose share the diagnostic spotlight. Both are major agents of pathology. In an individual destined to become diabetic (frankly hyperglycemic) in the future, the ﬁ rst change will be an elevation in the basal insulin concentration. Elevated basal insulin is typically associated with insulin resistance, the degree of which varies widely among tissues and among insulin-sensitive pathways within a single cell [ 8 ] . Instead of insulin resistance, dysinsulinism may be a better description. 84 R. Dankner and J. Roth

Basal Hyperinsulinemia Precedes Diabetes

Prospective studies initiated during the 1970s and published during the 1980s provided early evidence of the relationship between basal (fasting) and poststimulus hyperinsulinemia, and insulin resistance. In 1987, Sicree et al. reported an association between changes in insulin response and the progression to type 2 diabetes [ 9 ] . They found that the insulin level at 2 h after an oral glucose load was the best predictor of diabetes among Nauruans over 6 years. Interestingly, they found conversion to diabetes to associate with a high insulin response among those initially with normal glucose tolerance (2-h glucose less than 7.8 mM) and with a low insulin response among those initially with impaired glucose tolerance (IGT) (2-h glucose greater than 7.8 mM). These ﬁ ndings suggested that diabetes onset is preceded by a period of hyperinsulinemia followed by a period of hypoinsulinemia. Similarly, in their study of offspring of couples in which both parents had diabetes, Warram et al. [10 ] found hyperinsulinemia, basal and poststimulus, and not hypoinsulinemia, to accompany reduced glucose clearance antedating by 1–2 decades the diagnosis of diabetes. Irrespective of whether hyperinsulinemia is a consequence or cause of insulin resistance, it is insulin resistance, and not impaired pancreatic beta-cell function (insulin secretory dysfunction), that appears earlier in the course of conversion to diabetes. A 6-year prospective study of Pima Indians supported the above. Insulin resistance, assessed with the euglycemic-hyperinsulinemic clamp (EHC), was the strongest single predictor of type 2 diabetes [ 11 ] . A low acute insulin response (AIR) to glucose, as assessed by a glucose tolerance test, and indicative of insulin secretory dysfunction, was found to be an additional, though weaker risk factor. Further analysis of the Pima Indian cohort identi ﬁ ed an independent role for basal hyperinsulinemia in the development of diabetes, with greater impact than fasting plasma glucose (FPG), and distinct from that of insulin resistance, assessed as insulin- stimulated glucose disposal (using EHC) [ 12 ] . Distinguishing between hyperinsulinemia and insulin resistance challenges the common use of basal hyperinsulinemia as a surrogate for insulin resistance. Weyer et al. showed that before the onset of diabetes, individuals with elevated fasting plasma insulin concentration, relative to their degree of adiposity and insulin resistance, are at increased risk for decreased early-phase insulin secretion, but not for increased insulin resistance. The Israeli Glucose Intolerance Obesity and Hypertension (GOH) study, initiated by Modan et al. in 1980 [ 13] , was the basis for a recent report showing that basal hyperinsulinemia is a strong predictor of the cumulative incidence of type 2 diabetes over a 24-year period [ 14 ] . Even in normoglycemic subjects with normal body mass index (BMI), hyperinsulinemia in the basal state was the strongest predictor for conversion to diabetes.

Fasting Glucose as a Surrogate for Basal Hyperinsulinemia

While the euglycemic-hyperinsulinemic clamp is considered the gold standard for assessing insulin resistance, it is not practical for clinical evaluation or large research studies. Consequently, fasting glucose has become a common and easily assessed, but weak surrogate for insulin resistance. 6 Predicting Diabetes 85

Diagnostic threshold levels for fasting glucose delineate the normal range as all values under a cut-off level and above a certain minimum level. By de ﬁ nition, values within a normal range are all associated with the same level of risk, and thus essentially the same. The investigation of elevated glucose levels within the normal range as a possible predictor of diabetes actually challenges the validity of a normal glucose range. Already in 1990, Haffner et al. found fasting glucose measurements in the upper reaches of the normal range, as well as high fasting insulin, to associate with the development of diabetes in Mexican-Americans 8 years later [15 ] . A number of other studies have since found elevated levels of fasting glucose within the normal range to be associated with diabetes years later. In a prospective study of 1947 normoglycemic men (fasting glucose levels less than 110 mg/dL), those in the highest quartile of fasting blood glucose, as well as those in the lowest quartile of glucose disappearance rate, as assessed from an intravenous glucose tolerance test, were more likely to have diabetes 22.5 years later than men without these characteristics. Interestingly, no correlation was found between these two risk factors [16 ] . Another study (restricted to men) reported similar results, even after de ﬁ ning the normal FPG level as less than 100 mg/dL [ 17] . Here, an elevated risk of diabetes was observed even in men with FPG levels as low as 87 mg/dL, compared with those with levels less than 81 mg/dL. An increase in the level of fasting glucose was associated with an increased risk of diabetes. In addition, elevated serum triglyceride levels greater than 150 mg/dL and a body mass index (BMI) greater than 30 kg/m2 further increased the risk for diabetes. Importantly, a recent study reported FPG levels in the upper 50th percentile of the normal adult range (86 mg/dL) during childhood to be associated with increased risk for type 2 diabetes in adulthood, on average 21 years later [ 18 ] . In light of decreases in the level of fasting glucose that is designated as the cut- off point for diagnosis of diabetes, the question arises as to whether an elevated glucose level within the normal range is actually an early manifestation of the disease rather than a predictor of its future incidence.

Insulin Resistance as Predictor of Diabetes

A number of sophisticated indices of insulin sensitivity (the reciprocal of insulin resistance) have been used in research. Gutt et al. [ 19 ] developed an index derived from body weight, and from glucose and insulin levels, both fasting and during an OGTT. In a pooled analysis of prospective data from the San Antonio Heart Study, the Mexico City Diabetes Study (MCDS), and the Insulin Resistance Atherosclerosis Study, Gutt et al.’s insulin sensitivity index at 0 and 120 min, abbreviated ISI (0,120), demonstrated better predictive power than other measures, as shown by the largest area under the receiver operator characteristic (AROC) curve for predicting incident diabetes (78.5%) [20 ] . In contrast to glucose levels, insulin levels are generally compared within a population, and not to universal standards. The 75th percentile of individuals without diabetes within a study population has been commonly used as a threshold level for assessment of insulin resistance. However, three surrogate measures of insulin resistance: fasting insulin, the homeostasis model assessment of insulin resistance 86 R. Dankner and J. Roth

(HOMA-IR), and the reciprocal of the Gutt insulin sensitivity index, did not reveal threshold effects at this level, or at any other level, among 2,720 Framingham Offspring Study subjects over a 7- to 11-year follow-up [21 ] . These ﬁ ndings challenge the notion that conversion to diabetes occurs according to population-based thresholds.

Decreased Insulin Secretory Function as a Predictor of Diabetes

While hyperinsulinemia apparently precedes impaired insulin secretory function during the conversion from normoglycemia to diabetes, beta-cell failure is an inevitable condition for diabetes onset. Thus, measures of beta-cell function might be expected to be a key predictor for type 2 diabetes, albeit occurring at a later stage than hyperinsulinemia. The hyperglycemic clamp is considered to be the gold standard method for the measurement of both ﬁ rst- and second-phase insulin secretion. However, like the euglycemic-hyperinsulinemic clamp, it is impractical for both clinical and large- scale research purposes. The OGTT has thus become a commonly used surrogate for beta-cell function. Low ratios of change in insulin to change in glucose levels in the

30 min following glucose ingestion (delta I30 /delta G 30) were found to predict the development of diabetes in Mexican-Americans [ 22 ] . While the 2-h postload OGTT is most commonly used, a 20-year prospective study showed noninferiority of the 1-h postload OGTT [ 23 ] . The 1-h postload OGTT saves time in both the clinic and in the research environment. Nevertheless, since the OGTT is not routinely performed, most investigations of beta-cell function are from relatively small prospective studies.

Glycated Hemoglobin (HbA1c) as a Predictor of Diabetes

Glucose in a time-dependent fashion forms covalent bonds—initially reversible, later irreversible—with all proteins including hemoglobin. These glucose moieties, which then remain for the lifetime of the molecule, serve as a measure of hyperglycemia over the preceding weeks. The percent of hemoglobin that is glycated (hemoglobin A1c or HbA1c) reﬂ ects average plasma glucose levels over the preceding 3-month period. HbA1c is the major fraction, created during the nonenzymatic glycation of hemoglobin. Recent interest in HbA1c values as a predictor of diabetes parallels interest in its use as a diagnostic tool. In a prospective study of Japanese men, FPG and HbA1c were independently associated with the risk of type 2 diabetes over 4 years. The combination of these two factors increased the predictive power, as demonstrated by a greater area under the receiver operating characteristic (AUROC) curve [ 24 ] . In a retrospective analysis of 14 years follow-up of the prospective Atherosclerosis Risk in Communities (ARIC) cohort, the 2010 American Diabetes Association HbA1c cut-off of 6.5% was at least as effective as fasting glucose in identifying individuals with retinopathy and at future risk for kidney disease. Though individuals with diabetes were found to be at increased risk for microvascular disease, there was no evidence of a natural “glycemic threshold” [ 25 ] . 6 Predicting Diabetes 87

The Pattern of Changes in Glucose and Insulin Levels in the Conversion to Diabetes

In prospective studies, repeated testing of glucose levels in healthy individuals has revealed a pattern of changes in the conversion to diabetes. A survey of 35- to 64-year- old men and women in Mexico City demonstrated rapid conversion from normoglycemia to hyperglycemia [26 ] . A sharp increase in FPG was coincident with a decrease in 2-h postglucose insulin, the latter indicating impaired beta-cell function. The Botnia Study, which prospectively followed 2,115 individuals without diabetes in western Finland over the course of 6 years, also found a marked deterioration in beta- cell function, as assessed by repeated OGTTs, to precede the onset of type 2 diabetes [27 ] . In addition, a longitudinal study based on annual measurements with Pima Indians revealed a multistage pattern in the conversion to diabetes, characterized by a gradual, linear increase in 2-h plasma glucose, followed by an exponential rise, the latter beginning within 4.5 years of diabetes onset [28 ] . These studies are noteworthy in their tracing glucose and insulin levels in individuals over time and in their emphasis on the changes occurring in the period preceding diagnosis. A pattern emerges of gradual, followed by relatively abrupt changes in the conversion to diabetes. Thirteen-year trajectories of metabolic measures from the British Whitehall II study con ﬁ rm and expand on these patterns [29 ] (see Fig. 6.1 ). For those who converted to diabetes, fasting glucose and 2-h postload glucose increased, ﬁ rst linearly and then more steeply, from about 3 years prior to the diagnosis of diabetes. Insulin sensitivity was calculated based on homeostasis model assessments (HOMA), which account for both fasting glucose and insulin levels. A steep decrease in HOMA insulin sensitivity was observed already 5 years before diagnosis. HOMA beta-cell function increased during the 3–4 period prior to diagnosis, apparently compensating for decreased insulin sensitivity; and subsequently decreased until diagnosis. In contrast, for the group of individuals who did not convert to diabetes, trajectories increased mildly, gradually, and linearly, for all metabolic measures except insulin secretion, which did not change during follow-up. In the above studies, conversion to diabetes was determined by the accepted diagnostic threshold level of FPG. Nevertheless, the observation that a steep increase in glycemia precedes diagnosis, and persists after it, challenges the relevance of the threshold level. The transition from linear to exponential progression in plasma glucose and insulin levels may in fact be the more relevant indication of pathology. Since such transition does not seem to start at a threshold level, we endorse the idea that regular testing of individuals at yearly intervals may be the best way to predict prospectively the point at which an individual starts to convert to diabetes.

Anthropometric Measures as Predictors of Diabetes

While obesity (excess body fat) is an accepted risk factor for type 2 diabetes, the best means of its assessment is less clear. Anthropometric measures assess physical dimensions and can thereby serve as noninvasive proxies for body fat. BMI, 88 R. Dankner and J. Roth

Fig. 6.1 These graphs compare trajectories for fasting and 2-h postload glucose ((a , b ) respectively) in 505 individuals in the 13 years before their conversion to diabetes, vs. 6,033 nondiabetic controls. As can be seen, in those who converted to diabetes, both fasting glucose and 2 h postload glucose increased linearly with a shallow slope, and then more steeply during the 3 years prior to the diagnosis of diabetes. For the number of measurements for each year, see original fi gure in Tabak et al. [ 29 ] (adapted from the British Whitehall II study [29 ] ) calculated as body weight in kilograms divided by the square of height in meters, is commonly used as a measure of obesity. The association that has been demonstrated between abdominal fat and cardio-metabolic risk [30 ] suggests that measures of central fat distribution such as waist circumference and waist-hip ratio may predict type 2 diabetes better than measures of general obesity such as BMI. Anthropometric measures raise particular interest as predictors of diabetes due to their noninvasiveness. In a 6-year prospective study of Pima Indians, the 6 Predicting Diabetes 89 easy-to-measure anthropometric measures of waist circumference, waist-to-thigh ratio, weight, and BMI were as useful for identifying risk for diabetes as more complicated measures as percentage body fat [ 31 ] . Differences in predictive power depended on types of analysis used, such as receiver operating characteristic (ROC) curves or stepwise regression analysis, with no single measure demonstrating an overall advantage. A large study of African-American and White participants supports these fi ndings [ 32 ] . From a sample of 12,814 total, 1,515 new cases of diabetes were diagnosed over a 9-year follow-up. BMI, waist circumference, and waist-to- hip ratio were similar in their predictive capability (0.66–0.73 for single measures in ROC curves), with differences observed based on gender and ethnicity. In a prospective study of a Swedish population, BMI in the highest quartile was as good a predictor of diabetes 6–10 years later as was FPG or 2-h plasma glucose in the highest quartiles [ 33] . A recent prospective study of adults aged 60–79 years found waist circumference to be a superior predictor of diabetes than BMI in women, with no difference between these measures in men [34 ] . Recently, neck circumference was found to associate with cardiovascular risk factors including diabetes and impaired fasting glucose [35 ] . As surrogates for measuring adiposity, anthropometric measures are limited in their capability to re fl ect differences either between individuals or between population groups. BMI, for example, is not sensitive to differences in percent body fat between individuals or to characteristic differences in body shape between ethnic groups. A dramatic demonstration of the effects of ethnicity was presented in a 2011 publication from Ontario, Canada [36 ] . With mean follow-up of 6 years, the study, comprising nearly 60,000 residents, showed South Asian, Chinese, and Black subjects to develop diabetes at a higher rate, at an earlier age, and at a lower range of BMI than their White counterparts (Fig. 6.2). Differences between the ethnic groups in central adiposity at similar BMI levels were suggested as a possible modest contributor to the ethnic differences. The fi ndings highlight the limitation of universal anthropometric cut-off levels and support the idea that changes in an individual at regular intervals of time, considering a variety of factors, including body shape, ethnicity, and environment, may be a more sensitive indicator of metabolic change than is coincidence with de fi ned normal ranges.

Triglycerides as Predictors of Diabetes

Triglycerides are the chemical form of most fat, both in food and in the body. Elevated triglyceride levels, which commonly present in individuals with type 2 diabetes or those determined to be at a risk for such, have been associated with atherosclerosis, heart disease, and stroke [37 ] . In a random population sample of 1,351 Swedish women, aged 39–65 years, even serum triglycerides as low as 1.0–1.4 mmol/L were associated with an increased risk of diabetes, compared with a level <1.0 mmol/L, independent of age, BMI, blood pressure, and physical activity [ 38 ] . In a large study of healthy Israeli men, aged 26–45 [ 39 ] , baseline triglycerides were an independent risk factor for the development of type 2 diabetes. Importantly, 90 R. Dankner and J. Roth

Fig. 6.2 Ethnic inﬂ uences on diabetes incidence as a function of body mass index (BMI)—the incidence of diabetes per 1,000 patient years is plotted as a function of BMI. After appropriate adjustments, the risk of diabetes was almost twice as high among South Asians than Chinese and Blacks, who were substantially more affected than Whites. The median ages for diabetes diagnosis were 49, 55, 57, and 58, respectively, in these four ethnic groups. Diabetes incidence rates were equivalent at BMI levels of 24, 25, 26, and 30, respectively (adapted from Chiu et al. [36 ] )

the risk for diabetes was shown to be responsive to changes in triglycerides over a 5-year period, as well as to their baseline levels. For example, an increase in triglycerides carried a greater risk than did a persistently high level. Moreover, a decrease in triglycerides from the high to the low tertile, without lipid-lowering medication, was associated with such lifestyle changes as decreased BMI, increased physical activity, and routine consumption of breakfast. These ﬁ ndings support the role of triglycerides as a sensitive lifestyle biomarker. The observation that plasma triglycerides remained a signi ﬁ cant determinant of diabetes risk even after adjustment for lifestyle factors suggests an independent effect of triglycerides on diabetes risk.

Hepatic Markers as Predictors of Diabetes

Iron Storage

As in haemochromatosis, iron overload is recognized as a risk factor for diabetes [40 ] . A prospective nested case–control study from Finland found elevated stores of iron (below those in haemochromatosis) to be a risk factor for diabetes [ 41 ] . During 4 years of follow-up, men in Finland with high stores of iron i.e. those in the lowest 6 Predicting Diabetes 91 quarter of the ratio of transferrin receptors to ferritin, (<9.4 nanogram/ml) were 2.5 times more likely to develop diabetes than men with lower stores of iron, after adjustment for other variables including serum triglycerides. Analysis of the much larger Nurses’ Health Study cohort supported these ﬁ ndings in women [42 ] . Importantly, the association between serum ferritin and the development of diabetes was shown to be independent of in ﬂ ammatory factors [43 ] . The basis of the association between elevated ferritin levels and diabetes is not clear. Possibly, high storage of iron in the liver may promote insulin resistance and beta-cell dysfunction. Alternatively, elevated ferritin may be a manifestation of the metabolic abnormalities that result in diabetes. Irrespective of whether elevated iron stores are a causal factor or a marker of diabetes, low ratios of transferrin receptors to ferritin in healthy individuals may represent an elevated risk for diabetes.

Hepatic Enzymes

The role of the liver in the pathogenesis of diabetes, namely the contribution of hepatic glucose output to hyperglycemia, has led to the investigation of hepatic enzymes as possible markers for diabetes risk. Alanine aminotransferase (ALT), aspartate aminotransferase (AST), and gamma-glutamyltranspeptidase (GGT) are three enzymes found in the liver and commonly measured in plasma for diagnostic testing. Since ALT is found predominantly in the liver, it serves as a more speci fi c indication of liver infl ammation than AST and GGT, which are also found in other tissues. In Pima Indians, diabetes incidence over an average follow-up of 7 years was increased in those with ALT levels in the upper reaches of the normal range (but not with AST or GGT) [ 44] . In another prospective study, ALT was higher at 18 months before diagnosis in those who converted to diabetes than in those who did not [45 ] . Further, during the subsequent 18 months, ALT levels continued to rise in converters, simultaneous with a lesser rise in triglycerides and a rapid rise in fasting glucose levels. In a sample of Black and Hispanic individuals, aged 40–69, baseline ALT and AST were associated with the risk of type 2 diabetes 5.2 years later, after adjustment for BMI, insulin sensitivity, AIR, and C-reactive protein (CRP—a marker of subclinical infl ammation discussed below) [ 46 ] . Since baseline elevations of AST and ALT may re fl ect nonalcoholic fatty liver disease (NAFLD), this study suggests that pathologies of the liver may predict type 2 diabetes. The question as to whether hepatic markers are mediated by triglycerides or are associated directly with the development of diabetes remains to be answered.

Adipokines as Predictors of Diabetes

Adipokines are proteins secreted from adipose tissue. Due to the known association between diabetes and obesity, two major adipokines—leptin and adiponectin—have been investigated for possible association with diabetes. 92 R. Dankner and J. Roth

Human leptin is a protein of 167 amino acids that is produced primarily in the adipocytes of white adipose tissue. The level of circulating leptin is directly proportional to the total amount of fat in the body. Leptin acts on receptors in the brain, speci fi cally in the hypothalamus, where it diminishes appetite. Elevated levels of leptin correlate positively with fasting insulin [ 47, 48] . Weight reduction is associated with decreased levels of leptin (and of fasting insulin) [ 49 ] .The link between plasma leptin levels and total adipose tissue, a recognized risk factor for diabetes, has led to the investigation of leptin as a risk factor. Baseline leptin levels were reported higher in Japanese American men who converted to diabetes during 5–6 years follow-up, but not in women, after controlling for total fat, and for insulin and glucose levels [ 50] . Findings of the large ARIC Study (570 incident diabetes cases and 530 normoglycemics) also showed that elevated leptin levels associate with diabetes incidence [ 51 ] . However, when BMI, waist-to-hip ratio (WHR), fasting insulin, in fl ammatory factors, hypertension, and triglycerides were accounted for, the association between leptin and diabetes risk reversed and even appeared to be protective. Those with the highest adjusted leptin levels had an incidence of diabetes that was only 40% of those with the lowest levels. Adiponectin is a protein hormone secreted exclusively from adipose tissue that modulates a number of metabolic processes, including glucose regulation and fatty acid catabolism. It also has anti-in fl ammatory properties. Adiponectin levels are inversely correlated with body fat measures in adults. Weight reduction has been found to associate with increased serum adiponectin levels [ 52, 53 ] . A number of case–control studies have reported higher levels of adiponectin to associate with lower risk for diabetes [ 54– 56] . The Hoorn study found elevated adiponectin to associate prospectively with a lower risk of diabetes, particularly in women, after adjustments for waist-hip ratio, leptin, and glucose levels [57 ] .

In ﬂ ammatory Markers as Predictors of Diabetes

In fl ammatory markers are molecules, mostly proteins, that are released during infl ammation and are readily detected in plasma. The ability of cytokines like TNF and IL-6 to interfere with the action of insulin, causing insulin resistance, as well as the association of infl ammatory biomarkers with cardiovascular diseases [ 58, 59 ] , has raised the possibility that infl ammation may also be involved in the pathogenesis of diabetes. Associations have been proposed between type 2 diabetes and a number of infl ammatory markers, including CRP, fi brinogen, interleukin 6 (IL-6), and plasminogen activator inhibitor-1 (PAI-1). Using a nested case–control design, investigators in the Women’s Health Study found higher levels of interleukin 6 (IL-6) and CRP in women who converted to diabetes during a 4-year follow-up than in women who remained normoglycemic [ 60 ] . The West of Scotland Coronary Prevention Study found a similar association of CRP with diabetes in middle-age men followed for 5 years [61 ] . In the Insulin Resistance Atherosclerosis Study, individuals who converted to diabetes within 5 years had higher baseline levels of fi brinogen, CRP, and PAI-1 [62 ] . However, only the association with PAI-1 remained strong after adjusting for insulin 6 Predicting Diabetes 93 resistance and body fat (BMI and waist circumference). This study implies that PAI-1 is a more robust risk factor for diabetes than is CRP and also supports the notion that chronic infl ammation precedes the development of diabetes. Further analysis of the same study showed changes in PAI-1 over time, as well as baseline levels, are associated with incident diabetes [63 ] .

Endogenous Sex Hormones as Predictors of Diabetes

Endogenous sex hormones are synthesized de novo by the ovaries, testes and adrenal cortex, or by conversion from other steroids in the liver and fat. Differences between these hormones by gender (and age) have led to investigations of their potential effect on the differential rates of diabetes incidence between men and women. A systematic analysis of 43 prospective and cross-sectional studies showed that higher testosterone levels are associated with a decreased risk for diabetes in men and an increased risk in women [64 ]. A low level of circulating sex hormone– binding globulin was reported to be a strong predictor of risk for type 2 diabetes, in women and in men, beyond that of traditional risk factors [ 64, 65 ] .

The Metabolic Syndrome and Other Combinations of Risk Factors as Predictors of Diabetes

Many combinations of risk factors have been proposed and tested in certain populations as predictors of diabetes. In recent years, the metabolic syndrome has been one of the most common of them [65 ] . In the 1980s, Modan et al. [13 ] proposed that hyperinsulinemia (de fi ned as the upper quartile of the combined measure of basal and postload insulin) is linked to hypertension, obesity, dyslipidemia, and glucose intolerance, a cluster initially termed Syndrome X [65 ] and later “metabolic syndrome.” Insulin resistance, as assessed by hyperinsulinemia, is currently considered the dominant component of the metabolic syndrome. Hypertension and dyslipidemias—elevated triglycerides, elevated LDL, and depressed HDL—are considered major cocontributors to pathology. In addition to this core of abnormalities, some investigators have included obesity and in fl ammatory markers as well as disturbances in glucose metabolism and insulin secretion. A number of investigators have tried to discern whether the metabolic syndrome is a better predictor of diabetes than any of its individual components. In the San Antonio Heart Study, with 7–8 years of follow-up, the metabolic syndrome was considered an independent predictor of diabetes, though with lower positive predictive power than 2-h postload glucose levels [ 66 ] . However, with further analysis of the same cohort, together with that of the Mexico City Diabetes Study (MCDS), the Diabetes Predicting Model was a better predictor of diabetes than was the metabolic syndrome [67, 68] . The Diabetes Predicting Model characterizes the risk factors of age, gender, ethnicity, fasting glucose, systolic blood pressure, HDL cholesterol, 94 R. Dankner and J. Roth and family history of diabetes as continuous variables. This contrasts with the dichotomization of variables in de fi nitions of the metabolic syndrome. Dichoto- mization of continuous variables results in a loss of discrimination. Further, a selection of cut-off points may be population-specifi c. Accordingly, a quantitative analysis of 16 cohorts failed to show the metabolic syndrome to have an advantage over fasting glucose as a predictor of diabetes [68 ] . Clinical and statistical investigation of a large number of biomarkers involved in a variety of pathways led to the PreDx Diabetes Risk Score (DRS). The seven biomarkers: adiponectin, CRP, ferritin, fasting glucose, HbA1c, insulin, and interleukin-2 receptor alpha, were selected as the most informative predictors for the incidence of diabetes within 5 years [ 69 ] . The PreDx DRS was more effective than any single measure tested, including fasting glucose, in predicting diabetes in the Inter99 cohort, a population-based primary intervention of 30–60-year-olds in Denmark [ 69 ] . We now turn to studies investigating four scales that are free of laboratory testing and also noninvasive. During 16 years of follow-up in the Nurses’ Health Study, a combination of lifestyle factors: BMI less than 25 kg/m2 , routine exercise, abstinence from smoking, moderate consumption of alcohol, and a diet high in cereal fi ber and polyunsaturated fats and low in saturated and trans fats, as well as in glycemic load, was associated with a lower incidence of type 2 diabetes; 91% of the incidence of diabetes (95%CI 83–95) was found among women who did not maintain such lifestyle. While overweight was the single most important predictor of diabetes, other lifestyle factors were shown to signi fi cantly affect the risk for diabetes in this large cohort of middle-age women [70 ] . A score combining age, BMI, waist circumference, history of anti-hypertensive drug treatment and high blood glucose, physical inactivity, and low consumption of fruits, berries, and vegetables was associated with the risk of diabetes in 35- to 64-year-old men and women in Finland over a 10-year follow-up [ 71 ] . This simple noninvasive tool for identifying diabetes risk, since known as the FINDRISC, has been applied to studies in other cultures. Developed in the European Prospective Investigation into Cancer and Nutrition (EPIC)-Potsdam study, the German DRS combines age, height, waist circumference, history of hypertension, physical activity, smoking, and consumption of red meat, whole-grain bread, coffee, and alcohol [72 ] . This scale includes more lifestyle factors than does the FINDRISC and uses continuous variables, without attributing a risk classi fi cation. The results of the FINDRISC study raise the question as to whether the German DRS provides enough additional predictive power to justify the additional data collection and calculations that it requires [73 ] . The QDScore is a noninvasive diabetes risk algorithm that was developed to predict the 10-year risk of conversion to diabetes in diverse populations [ 74 ] . It includes self-assigned ethnicity and a factor re fl ecting social deprivation, in addition to age, sex, BMI, smoking status, family history of diabetes, treated hypertension, cardiovascular disease, and current use of corticosteroids. The importance of ethnicity in the assessment of diabetes is highlighted by the four to fi vefold variation in the risk of diabetes between ethnic groups revealed by the QDScore. This tool is easy to use in clinical practice and available to the public through a simple web calculator ( www.qdscore.org ). 6 Predicting Diabetes 95

Genetic Factors as Predictors of Diabetes

The observation that type 2 diabetes seems more common in some families than in others has led to the investigation of genetic factors that may associate with the disease. No direct relationship has been identi fi ed between any one genetic factor and type 2 diabetes. Single nucleotide polymorphisms (SNPs) are DNA sequence variations of a single nucleotide—A, T, C, or G—in the genome. The frequency of SNP alleles differs between populations. Certain SNPs have been identifi ed with increased risk for particular pathologies. In two separate investigations, the Framingham Offspring cohort and the EPIC-Potsdam cohort, SNPs that had been identifi ed as associating with diabetes did not substantially enhance the prediction of diabetes compared to common risk factors [ 75, 76] . An 18-year follow-up analysis of the Whitehall II cohort found associations between common genetic variants and fasting glucose to be constant over time in nondiabetic individuals, i.e., to remain stable over age trajectories. In contrast, genetic effects on 2-h glucose were found to be dependent on age, i.e., differences in 2-h glucose per additional risk allele were found to increase with advancing age, as depicted by steep rises along age trajectories [77 ] . This study suggests fundamental differences in the predictive power of fasting and 2-h glucose and highlights the importance of age and related environmental factors in the consideration of risks of genetic variations. The large multicenter prospective design exempli fi es the research conditions necessary for the investigation of genetic factors. Chapter 4 discusses in more depth the genetic aspects of diabetes.

Predictors of Diabetes: Threshold Levels or a Continuum

Recurring questions in the search for early markers of diabetes are whether threshold levels contribute to the prediction of diabetes, and whether they should be used as diagnostic criteria instead of increases along a continuum. Regarding the importance of threshold levels as diabetes diagnostic criteria, the World Health Organization and the American Diabetes Association set 7.0 mmol/L (126 mg/dL) as the FPG threshold for diabetes diagnosis following three pivotal epidemiological studies [ 78– 80 ] that reported the rare appearance of signs of retinopathy below this threshold, and much greater prevalence above it. It has since been claimed that retinopathy was incompletely and imprecisely assessed in these studies [2 ] . Subsequent analysis of the association of retinopathy to FPG in four other cross-sectional adult populations, in Australia, North America, and China, using sophisticated assessments of retinopathy, did not support the 7.0 mmol/L level, or any other FPG value, as a threshold for signs of retinopathy [2, 81 ] . As shown above, neither was there evidence for a cut-off threshold for glycated hemoglobin [ 25 ] . Rather, accumulating evidence suggests a continuous relationship between glycemic measures and retinopathy. These fi ndings challenge the FPG cut-off of 7.0 mmol/L, as well as the 2010 American Diabetes Association HbA1c threshold of 6.5%, as threshold levels for the 96 R. Dankner and J. Roth identifi cation of retinopathy. Since diagnostic criteria for diabetes are based on the association between retinopathy and FPG, these fi ndings also question the relevance of threshold levels to the diagnosis of diabetes. Interestingly, the relationship between glucose and macrovascular complications such as cardiovascular disease seems continuous, with no threshold [ 1 ] . This continuous relationship is analogous to that between end-organ damage and other cardiovascular risk factors such as blood pressure and serum cholesterol levels. The possibility that threshold levels of glucose for diagnosing diabetes may become obsolete challenges the value of threshold levels to the prediction of diabetes. In light of the evidence presented above of continuous rather than threshold-based relationships between risk factors and diabetes incidence, prediabetes is now considered a continuous abnormality, which, similar to hypercholesterolemia, hypertension, and metabolic bone disease, can be treated based on a defi nition of continuous risk [ 82 ] . Perhaps, like hypertension and hyperlipidemia, prediabetes should be risk- stratifi ed regarding the intensity and advent of intervention. This would entail setting clinical goals according to underlying risk factors. As such, individuals diagnosed with prediabetes who have additional risk factors, such as elevated BMI, hyperlipidemia, hypertension, and family history, would receive earlier and more aggressive intervention than those presenting only with prediabetes. Supporting such an approach, lifestyle modi fi cation in individuals with isolated impaired fasting glucose was shown to be only minimally effective in thwarting progression to diabetes compared to similar intervention in individuals with both impaired fasting glucose and IGT [83 ] .

Predicting Diabetes in the Future

We have shown that the presently available tools are inadequate for predicting diabetes for individuals. Rather than being seers with crystal balls, we are like actuaries in an insurance company—gauging risk for a group but not predicting individual outcomes. Imagining ourselves as actuaries, we are aware of the weakness of population- based normal ranges for glucose and other key metabolic variables. We expect that personalized measures will be much more valuable in predicting the risk for diabetes over the next 3–10 years. While we started with glucose and hemoglobin A1c, we believe that levels of insulin, C-peptide, and possibly triglycerides will give earlier warnings. With time, more markers will be tested and validated, e.g., the branched chain amino acids or metabolites that change with the onset of the metabolic syndrome. Again, assessments based on an individual’s own past will be much more sensitive than those based on population norms. Future studies will likely deepen our knowledge of how obesity and diabetes are both causing reversible and irreversible damage to tissues throughout the body well before glucose reach levels diagnosed as diabetes. The brain, with evidence for functional and anatomical changes early on, seems particularly vulnerable. The inescapable conclusion is that earlier treatment is vital. 6 Predicting Diabetes 97

Advice to the Practicing Physician

Today, it seems unwise to refrain from therapy until patients cross the line demar- cating hyperglycemia according to current criteria, given (1) our recognition that pathological processes start early, (2) reliable warning signs are evident, and (3) the most effective therapies (calorie restriction, modiﬁ ed food choices, moderate weight loss, and regular exercise) are affordable, safe, and with very minimal side effects. When an individual’s blood glucose levels repeatedly reach values higher than their own previous ones, a state of “dysglycemia” should be declared, and the introduction of nonpharmacologic therapies strongly considered. Hypertension should probably be dealt with similarly; blood pressure readings within the normal range on a population basis but reproducibly above a patient’s own norm should prompt lifestyle modi ﬁ cation, i.e., nutritional therapy and an exercise program.

Personalized Risk Pro ﬁ les of Diabetes

The future risk for type 2 diabetes has been shown to differ among individuals in any glucose tolerance category [ 84 ] . Even along a continuum of glucose levels, the same fasting or postload value may signify different degrees of risk to different people. Age, gender, ethnicity, and genes are some of the factors that have been shown to affect the risk for developing diabetes. Environmental and genetic effects on the development of diabetes have received great attention in recent years. We expect that with the introduction of new sophisticated tools for genetics research, the effect of genetic variants may take center stage. However, very large-scale collaboration is required for the investigation of such factors. Individual characteristics may re fi ne disease risk even more than do population characteristics. Individual risk levels have become acceptable for a number of health conditions. For example, BMI of 26 kg/m2 generally prompts treatment for an individual who is identifi ed with risk factors for disease, yet may be ignored otherwise. Moreover, low-density-cholesterol under 100 mg/dL is recommended for those with diabetes, compared to a target of 130 for the general population. Risk factors and risk scales are often identi fi ed, developed, and tested within particular populations. The population may be of a similar ethnicity, geographical area, age range, or gender. Accounting for the combined effects of these interactive factors leads to a proposition that diabetes risk may best be determined by personalized profi les rather than by universal or population norms [ 85 ] . Such risk profi les will incorporate combinations of risk factors. Individual changes, within so-called population norms, for a wide range of anthropometric and biochemical characteristics, may indicate increased risk for diabetes. We note that personalized standards are in wide use for interpreting the clinical data on linear growth, heart rate, and body temperature. As noted earlier, for a number of risk factors, the rate and magnitude of change, as well as baseline values, apparently in fl uence the incidence of diabetes. Conversion to diabetes seems to include a phase of steep change in certain factors, in addition to con- 98 R. Dankner and J. Roth tinuous progression of others. Yet, the earliest triggers or markers of type 2 diabetes, and the levels at which they become risk factors, seem to differ between individuals. The burden of diabetes, and the growing evidence that its onset can be delayed, motivates the elucidation of risk pro fi les. Identi fi cation of the earliest signs, and of the characteristics of conversion, contributes to our understanding of its pathophysiology. The real goal may be identifi cation of personalized risk profi les for particular symptoms and complications of diabetes, rather than for the disease as it is currently de fi ned.

Glossary

AIR Acute insulin response ALT Alanine aminotransferase AST Aspartate aminotransferase BMI Body mass index DRS Diabetes risk score EHC Euglycemic-hyperinsulinemic clamp FPG Fasting plasma glucose GGT Gamma-glutamyltranspeptidase GL Glycemic load HbA1c Glycated hemoglobin HDL High-density lipoprotein IFG Impaired fasting glucose IGT Impaired glucose tolerance IL-6 Interleukin 6 LDL Low-density lipoprotein MCDS Mexico City Diabetes Study NAFLD Nonalcoholic fatty liver disease PAI-1 Plasminogen activator inhibitor-1 SHBG Sex hormone binding globulin SNPs Single nucleotide polymorphisms WHR Waist-to-hip ratio

Acknowledgments We are indebted to Michael Bergman for his insightful comments and to Ian Whitford and Sana Qureshi for excellent editorial assistance.

References

1. Jesudason DR, Dunstan K, Leong D, Wittert GA. Macrovascular risk and diagnostic criteria for type 2 diabetes: implications for the use of FPG and HbA(1c) for cost-effective screening. Diabetes Care. 2003;26(2):485–90. 2. Wong TY, Liew G, Tapp RJ, et al. Relation between fasting glucose and retinopathy for diagnosis of diabetes: three population-based cross-sectional studies. Lancet. 2008;371:736–43. 6 Predicting Diabetes 99

3. Toth MJ, Tchernof A. Lipid metabolism in the elderly. Eur J Clin Nutr. 2000;54 Suppl 3:S121–5. 4. Nakao C, Ookawara T, Sato Y, et al. Extracellular superoxide dismutase in tissues from obese (ob⁄ob) mice. Free Radic Res. 2000;33:229–41. 5. Furukawa S, Fujita T, Shimabukuro M, et al. Increased oxidative stress in obesity and its impact on metabolic syndrome. J Clin Invest. 2004;114:1752–61. 6. Park J, Chung JJ, Kim JB. New evaluations of redox regulating system in adipose tissue of obesity. Diabetes Res Clin Pract. 2007;77 Suppl 1:S11–6. 7. Brown TM, Voeks JH, Bittner V, Safford MM. Variations in prevalent cardiovascular disease and future risk by metabolic syndrome classi ﬁ cation in the REasons for Geographic and Racial Differences in Stroke (REGARDS) study. Am Heart J. 2010;159(3):385–91. 8. Rabinowitz D, Zierler KL. Forearm metabolism in obesity and its response to intra-arterial insulin: characterization of insulin resistance and evidence for adaptive hyperinsulinism. J Clin Invest. 1962;41:2173–81. 9. Sicree RA, Zimmet PZ, King HO, et al. Plasma insulin response among Nauruans. Prediction of deterioration in glucose tolerance over 6 yr. Diabetes. 1987;36(2):179–86. 10. Warram JH, Martin BC, Krolewski AS, et al. Slow glucose removal rate and hyperinsulinemia precede the development of type II diabetes in the offspring of diabetic parents. Ann Intern Med. 1990;113(12):909–15. 11. Lillioja S, Mott DM, Spraul M, et al. Insulin resistance and insulin secretory dysfunction as precursors of non-insulin-dependent diabetes mellitus. Prospective studies of Pima Indians. N Engl J Med. 1993;329(27):1988–92. 12. Weyer C, Hanson RL, Tataranni PA, et al. A high fasting plasma insulin concentration predicts type 2 diabetes independent of insulin resistance: evidence for a pathogenic role of relative hyperinsulinemia. Diabetes. 2000;49(12):2094–101. 13. Modan M, Halkin H, Almog S, et al. Hyperinsulinemia: a link between hypertension, obesity and glucose intolerance. J Clin Invest. 1985;75:809–17. 14. Dankner R, Chetrit A, Shanik MH, et al. Basal-state hyperinsulinemia in healthy normoglycemic adults is predictive of type 2 diabetes over a 24-year follow-up: a preliminary report. Diabetes Care. 2009;32(8):1464–6. 15. Haffner SM, Stern MP, Mitchell BD, et al. Incidence of type II diabetes in Mexican Americans predicted by fasting insulin and glucose levels, obesity, and body-fat distribution. Diabetes. 1990;39(3):283–8. 16. Bjørnholt JV, Erikssen G, Liestøl K K, et al. Prediction of type 2 diabetes in healthy middle- aged men with special emphasis on glucose homeostasis. Results from 22.5 years’ follow-up. Diabet Med. 2001;18(4):261–7. 17. Tirosh A, Shai I, Tekes-Manova D, et al., Israeli Diabetes Research Group. Normal fasting plasma glucose levels and type 2 diabetes in young men. N Engl J Med. 2005;353(14):1454–62. 18. Nguyen QM, Srinivasan SR, Xu JH, et al. Fasting plasma glucose levels within the normoglycemic range in childhood as a predictor of prediabetes and type 2 diabetes in adulthood. The Bogalusa Heart Study. Arch Pediatr Adolesc Med. 2010;164(2):124–8. 19. Gutt M, Davis CL, Spitzer SB, et al. Validation of the insulin sensitivity index (ISI(0,120)): comparison with other measures. Diabetes Res Clin Pract. 2000;47(3):177–84. 20. Hanley AJ, Williams K, Gonzalez C, et al., San Antonio Heart Study; Mexico City Diabetes Study; Insulin Resistance Atherosclerosis Study. Prediction of type 2 diabetes using simple measures of insulin resistance: combined results from the San Antonio Heart Study, the Mexico City Diabetes Study, and the Insulin Resistance Atherosclerosis Study. Diabetes. 2003;52(2): 463–9. 21. Rutter MK, Wilson PWF, Sullivan LM, et al. Use of alternative thresholds de ﬁ ning insulin resistance to predict incident type 2 diabetes mellitus and cardiovascular disease. Circulation. 2008;117:1003–9. 22. Haffner SM, Miettinen H, Gaskill SP, et al. Decreased insulin secretion and increased insulin resistance are independently related to the 7-year risk of NIDDM in Mexican-Americans. Diabetes. 1995;44(12):1386–91. 100 R. Dankner and J. Roth

23. Dankner R, Abdul-Ghani MA, Gerber Y, et al. Predicting the 20-year diabetes incidence rate. Diabetes Metab Res Rev. 2007;23(7):551–8. 24. Sato KK, Hayashi T, Harita N, et al. Combined measurement of fasting plasma glucose and A1C is effective for the prediction of type 2 diabetes. Diabetes Care. 2009;32:644–6. 25. Selvin E, Ning Y, Steffes MW, et al. Glycated hemoglobin and the risk of kidney disease and retinopathy in adults with and without diabetes. Diabetes. 2011;60:298–305. 26. Ferrannini E, Nannipieri M, Williams K, et al. Mode of onset of type 2 diabetes from normal or impaired glucose tolerance. Diabetes. 2004;53(1):160–5. 27. Lyssenko V, Almgren P, Anevski D, et al., Botnia study group. Predictors of and longitudinal changes in insulin sensitivity and secretion preceding onset of type 2 diabetes. Diabetes. 2005;54(1):166–74. 28. Mason CC, Hanson RL, Knowler WC. Progression to type 2 diabetes characterized by moderate then rapid glucose increases. Diabetes. 2007;56:2054–61. 29. Tabák AG, Jokela M, Akbaraly TN, et al. Trajectories of glycemia, insulin sensitivity and insulin secretion preceding the diagnosis of type 2 diabetes: the Whitehall II study. Lancet. 2009;373:2215–21. 30. Klein S, Allison DB, Heyms fi eld SB, Association for Weight Management and Obesity Prevention; NAASO; Obesity Society; American Society for Nutrition; American Diabetes Association; Association for Weight Management and Obesity Prevention; NAASO, the Obesity Society; the American Society for Nutrition; and the American Diabetes Association et al. Waist circumference and cardiometabolic risk: a consensus statement from shaping America’s health. Diabetes Care. 2007;30:1647–52. 31. Warne DK, Charles MA, Hanson RL, et al. Comparison of body size measurements as predictors of NIDDM in Pima Indians. Diabetes Care. 1995;18(4):435–9. 32. Stevens J, Couper D, Pankow J, et al. Sensitivity and specifi city of anthropometrics for the prediction of diabetes in a biracial cohort. Obes Res. 2001;9(11):696–705. 33. Rolandsson O, Hägg E, Nilsson M, et al. Prediction of diabetes with body mass index, oral glucose tolerance test and islet cell autoantibodies in a regional population. J Intern Med. 2001;249(4):279–88. 34. Wannamethee SG, Papacosta O, Whincup PH, et al. Assessing prediction of diabetes in older adults using different adiposity measures: a 7 year prospective study in 6,923 older men and women. Diabetologia. 2010;53(5):890–8. 35. Preis SR, Pencina MJ, D’Agnostino RB. Abstract 1469: neck circumference and development of cardiovascular disease risk factors in the Framingham heart study. Circulation. 2009;120:S509. 36. Chiu M, Austin PC, Manuel DG, Shah BR, Tu JV. Deriving ethnic-specifi c BMI cutoff points for assessing diabetes risk. Diabetes Care. 2011;34(8):1741–8. 37. Ginsberg HN, Zhang YL, Hernandez-Ono A. Regulation of plasma triglycerides in insulin resistance and diabetes. Arch Med Res. 2005;36:232–40. 38. Dotevall A, Johansson S, Wilhelmsen L, et al. Increased levels of triglycerides, BMI and blood pressure and low physical activity increase the risk of diabetes in Swedish women. A prospective 18-year follow-up of the BEDA study. Diabet Med. 2004;21(6):615–22. 39. Tirosh A, Shai I, Bitzur R, et al. Changes in triglyceride levels over time and risk of type 2 diabetes in young men. Diabetes Care. 2008;31(10):2032–7. 40. Yaouanq JM. Diabetes and haemochromatosis: current concepts, management and prevention. Diabete Metab. 1995;21(5):319–29. 41. Salonen JT, Tuomainen TP, Nyyssönen K, et al. Relation between iron stores and non-insulin dependent diabetes in men: case-control study. BMJ. 1998;317:727. 42. Jiang R, Manson JE, Meigs JB, et al. Body iron stores in relation to risk of type 2 diabetes in apparently healthy women. JAMA. 2004;291(6):711–7. 43. Forouhi NG, Harding AH, Allison M, et al. Elevated serum ferritin levels predict new-onset type 2 diabetes: results from the EPIC-Norfolk prospective study. Diabetologia. 2007;50(5):949–56. 44. Vozarova B, Stefan N, Lindsay RS, et al. High alanine aminotransferase is associated with decreased hepatic insulin sensitivity and predicts the development of type 2 diabetes. Diabetes. 2002;51(6):1889–95. 6 Predicting Diabetes 101

45. Sattar N, McConnachie A, Ford I, et al. Serial metabolic measurements and conversion to type 2 diabetes in the west of Scotland coronary prevention study: speci fi c elevations in alanine aminotransferase and triglycerides suggest hepatic fat accumulation as a potential contributing factor. Diabetes. 2007;56(4):984–91. 46. Hanley AJG, Williams K, Festa A, et al. Elevations in markers of liver injury and risk of type 2 diabetes. The Insulin Resistance Atherosclerosis Study. Diabetes. 2004;53:2623–32. 47. Malmstrom R, Taskinen MR, Karonen SL, et al. Insulin increases plasma leptin concentrations in normal subjects and patients with NIDDM. Diabetologia. 1996;39:993–6. 48. Zimmet P, Hodge A, Nicolson M, et al. Serum leptin concentration, obesity, and insulin resistance in Western Samoans: cross sectional study. BMJ. 1996;313:965–9. 49. Shih LY, Liou TH, Chao JC, et al. Leptin, superoxide dismutase, and weight loss: initial leptin predicts weight loss. Obesity (Silver Spring). 2006;14(12):2184–92. 50. McNeely MJ, Boyko EJ, Weigle DS, et al. Association between baseline plasma leptin levels and subsequent development of diabetes in Japanese Americans. Diabetes Care. 1999;22(1):65–70. 51. Schmidt MI, Duncan BB, Vigo A, et al., ARIC Investigators. Leptin and incident type 2 diabetes: risk or protection? Diabetologia. 2006;49(9):2086–96. 52. Hotta K, Funahashi T, Arita Y, et al. Plasma concentrations of a novel, adiposespeci fi c protein, adiponectin, in type 2 diabetic patients. Arterioscler Thromb Vasc Biol. 2000;20(6):1595–9. 53. Yang WS, Lee WJ, Funahashi T, et al. Weight reduction increases plasma levels of an adipose- derived anti-infl ammatory protein, adiponectin. J Clin Endocrinol Metab. 2001;86(8):3815–9. 54. Lindsay RS, Funahashi T, Hanson RL, et al. Adiponectin and development of type 2 diabetes in the Pima Indian population. Lancet. 2002;360:57–8. 55. Daimon M, Oizumi T, Saitoh T, et al. Decreased serum levels of adiponectin are a risk factor for the progression to type 2 diabetes in the Japanese Population: the Funagata study. Diabetes Care. 2003;26(7):2015–20. 56. Duncan BB, Schmidt MI, Pankow JS, et al. Adiponectin and the development of type 2 diabetes: the atherosclerosis risk in communities study. Diabetes. 2004;53(9):2473–8. 57. Snijder MB, Heine RJ, Seidell JC, et al. Associations of adiponectin levels with incident impaired glucose metabolism and type 2 diabetes in older men and women: the Hoorn study. Diabetes Care. 2006;29(11):2498–503. 58. Ridker PM, Cushman M, Stampfer MJ, et al. In fl ammation, aspirin, and the risk of cardiovascular disease in apparently healthy men. N Engl J Med. 1997;336:973–9. 59. Ridker PM, Hennekens CH, Buring JE, et al. C-reactive protein and other markers of infl ammation in the prediction of cardiovascular disease in women. N Engl J Med. 2000;342: 836–43. 60. Pradhan AD, Manson JE, Rifai N, et al. C-reactive protein, interleukin 6, and risk of developing type 2 diabetes mellitus. JAMA. 2001;286(3):327–34. 61. Freeman DJ, Norrie J, Caslake MJ, et al., West of Scotland Coronary Prevention Study. C-reactive protein is an independent predictor of risk for the development of diabetes in the West of Scotland Coronary Prevention Study. Diabetes. 2002;51(4):1131–7. 62. Festa A, D’Agostino R Jr, Tracy RP, et al., Insulin Resistance Atherosclerosis Study. Elevated levels of acute-phase proteins and plasminogen activator inhibitor-1 predict the development of type 2 diabetes: the insulin resistance atherosclerosis study. Diabetes. 2002;51(4):1131–7. 63. Festa A, Williams K, Tracy RP, et al. Progression of plasminogen activator inhibitor-1 and fi brinogen levels in relation to incident type 2 diabetes. Circulation. 2006;113(14):1753–9. 64. Ding EL, Song Y, Malik VS, et al. Sex differences of endogenous sex hormones and risk of type 2 diabetes: a systematic review and meta-analysis. JAMA. 2006;295(11):1288–99. 65. Ding EL, Song Y, Manson JE, Hunter DJ, Lee CC, Rifai N, Buring JE, Gaziano JM, Liu S. Sex hormone-binding globulin and risk of type 2 diabetes in women and men. N Engl J Med. 2009 Sep 17;361(12):1152–63. Epub 2009 Aug 5. 66. Reaven GM. Banting lecture: role of insulin resistance in human disease. Diabetes. 1988;37: 1595–607. 67. Lorenzo C, Okoloise M, Williams K, et al., San Antonio Heart Study. The metabolic syndrome as predictor of type 2 diabetes: the San Antonio heart study. Diabetes Care. 2003;26(11):3153–9. 102 R. Dankner and J. Roth

68. Stern MP, Williams K, González-Villalpando C, et al. Does the metabolic syndrome improve identi fi cation of individuals at risk of type 2 diabetes and/or cardiovascular disease? Diabetes Care. 2004;27(11):2676–81. 69. Ford ES, Li C, Sattar N. Metabolic syndrome and incident diabetes: current state of the evidence. Diabetes Care. 2008;31:1898–904. 70. Urdea M, Kolberg J, Wilber J, et al. Validation of a multimarker model for assessing risk of type 2 diabetes from a fi ve-year prospective study of 6784 Danish people (Inter99). J Diabetes Sci Technol. 2009;3(4):748–55. 71. Hu FB, Manson JE, Stampfer MJ, et al. Diet, lifestyle, and the risk of type 2 diabetes mellitus in women. N Engl J Med. 2001;345(11):790–7. 72. Lindström J, Tuomilehto J. The diabetes risk score: a practical tool to predict type 2 diabetes risk. Diabetes Care. 2003;26(3):725–31. 73. Schulze MB, Hoffmann K, Boeing H, et al. An accurate risk score based on anthropometric, dietary, and lifestyle factors to predict the development of type 2 diabetes. Diabetes Care. 2007;30:510–5. 74. Bergmann A, Li J, Wang L, et al. A simplifi ed Finnish diabetes risk score to predict type 2 diabetes risk and disease evolution in a German population. Horm Metab Res. 2007;39(9):677–82. 75. Hippisley-Cox J, Coupland C, Robson J, et al. Predicting risk of type 2 diabetes in England and Wales: prospective derivation and validation of QDScore. BMJ. 2009;338:b880. 76. Meigs JB, Shrader P, Sullivan LM, et al. Genotype score in addition to common risk factors for prediction of type 2 diabetes. N Engl J Med. 2008;359(21):2208–19. 77. Schulze MB, Weikert C, Pischon T. Use of multiple metabolic and genetic markers to improve the prediction of type 2 diabetes: the EPIC-Potsdam study. Diabetes Care. 2009;32:2116–9. 78. Jensen AC, Barder A, Kumari M, et al. Associations of common genetic variants with age- related changes in fasting and postload glucose. Evidence from 18 years of follow-up of the Whitehall II Cohort. Diabetes. 2011;60(5):1617–23. 79. McCance DR, Hanson RL, Charles MA, et al. Comparison of tests for glycated haemoglobin and fasting and two hour plasma glucose concentrations as diagnostic methods for diabetes. BMJ. 1994;308:1323–8. 80. Engelgau MM, Thompson TJ, Herman WH, et al. Comparison of fasting and 2-hour glucose and HbA1c levels for diagnosing diabetes. Diagnostic criteria and performance revisited. Diabetes Care. 1997;20(5):785–91. 81. Expert Committee on the Diagnosis and Classi fi cation of Diabetes Mellitus. Report of the expert committee on the diagnosis and classifi cation of diabetes mellitus. Diabetes Care. 1997;20(7):1183–97. 82. Jonas JB, Xu L, Xie XW, et al. Relationship between fasting glucose and retinopathy for diagnosis of diabetes: results from a population-based study in urban and rural China. Retina. 2010;30(8):1223–7. 83. Bergman M. Inadequacies of absolute threshold levels for diagnosing prediabetes. Diabetes Metab Res Rev. 2010;26:3–6. 84. Toshikazu S, Makoto W, Junko N, et al., Zensharen Study for Prevention of Lifestyle Diseases Group. Lifestyle modi fi cation and prevention of type 2 diabetes in overweight Japanese with impaired fasting glucose levels. A Randomized Controlled Trial. Arch Intern Med. 2011; 171(15):1352–60. 85. Abdul-Ghani MA, Defronzo RA. Plasma glucose concentration and prediction of future risk of type 2 diabetes. Diabetes Care. 2009;32 Suppl 2:S194–8. 86. Dankner R, Danoff A, Roth J. Can ‘personalized diagnostics’ promote earlier intervention for dysglycaemia? Hypothesis ready for testing. Diabetes Metab Res Rev. 2010;26:7–9. Chapter 7 Screening for Prediabetes and Diabetes

Amir Tirosh

Screening populations at risk for a disease, applying interventions to halt or reverse the progression to overt clinical syndrome, and preventing future morbidity and/or mortality in a cost-effective manner have been a subject for extensive research for several decades for many common health problems. In 1968, the World Health Organization commissioned a report on screening by Wilson and Jungner entitled Principles and practice of screening for disease which has since become a classic and widely accepted doctrine in the science of public health [ 1 ] . Wilson and Jungner attempted to defi ne screening criteria to guide the selection of conditions that would be suitable for screening, based on the following, among other, principles: (1) The condition sought should be an important health problem. (2) There should be an accepted treatment for patients with recognized disease. (3) There should be a rec- ognizable latent or early symptomatic stage. (4) There should be a suitable test or examination. (5) The test should be acceptable to the population. (6) The natural history of the condition, including development from latent to declared disease, should be adequately understood. (7) There should be an agreed policy on whom to treat as patients. (8) The cost of case- fi nding (including diagnosis and treatment of patients diagnosed) should be economically balanced in relation to possible expenditure on medical care as a whole. Based on Wilson and Jungner principles, type 2 diabetes seems to be the ideal disease for an ongoing screening effort in order to implement treatments early in the course of the disease for otherwise undiagnosed, asymptomatic patients as well as to detect individuals at risk for the disease (in the prediabetic state) for whom primary prevention approaches may be bene fi cial. In the last decade, there has been a bulk of epidemiological data enhancing our understanding of the natural history of the disease, the screening tools of either hemoglobin A1c (HbA1c) or glucose levels

A. Tirosh , MD, PhD (*) Division of Endocrinology, Diabetes and Hypertension , Brigham and Women’s Hospital and Harvard School of Public Health , 221 Longwood Avenue , Boston , MA 02115 , USA e-mail: [email protected]

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 103 DOI 10.1007/978-1-4614-3314-9_7, © Springer Science+Business Media New York 2012 104 A. Tirosh

(fasting or postloading) have become widely available and validated, and strong evidence exists about the effi cacy of glucose monitoring early in the course of the disease in preventing late diabetes complications. In addition, lessons from the Diabetes Prevention Program (DPP) as well as from many other recently published prevention trials have clearly shown that diabetes can be prevented or at least delayed by either lifestyle modifi cations or pharmacological agents [ 2 ] . Nevertheless, while screening for detection and treatment of undiagnosed patients with diabetes and for applying primary prevention programs for at-risk groups seems obvious, a debate continues over the bene fi ts and harms of screening adults for diabetes or prediabetes.

Screening for Prediabetes and Type 2 Diabetes: What Is the Evidence?

Based on the above-mentioned principles for disease, screening the data which argues in favor of screening for this condition consists of the following; 1 . Diabetes as important public health issue : Type 2 diabetes is a major public health problem with rates that are rapidly increasing in both women and men, among all age groups and across all ethnic groups and geographic locations (though at different incident rates). It has been recently estimated that the disease affects approximately 8% of the population in the United States and with many millions being undiagnosed with the condition [3 ] . Worldwide, the prevalence of type 2 diabetes continues to rise, and while the estimated number of patients with diabetes at the year 2030 was 333 million, a more recent updated prediction model is anticipating 438 million to have diabetes by 2030 with a much higher proportion having prediabetes [3 ] . 2 . There is an effective treatment for diabetes: Well-established treatments for type 2 diabetes exist with proven bene fi ts in controlling the disease and attenuating the risk for diabetes-related complications. Clearly, lessons from the United Kingdom Prospective Diabetes Study (UKPDS) have shown that intensive treatment of hyperglycemia early in the course of the disease can substantially decrease the risk of microvascular complications [4 ] . Not before a decade after the UKPDS study had been concluded that we realize that early treatment of patients with diabetes, early in the course of the disease with intensive glucose lowering regimens, can also decrease the incidence rate of myocardial infarction and mortality rates [5 ] , an effect which could not be observed when intensive treatment has been initiated late in the course of the disease [6 ] . For prediabetes, using either nonpharmacologic or pharmacologic approaches could be ef fi cient in delaying the progression to type 2 diabetes [ 2] . It is important to note, however, that while the above-mentioned studies provide compelling evidence to believe early intervention among patients with diabetes would be benefi cial, to date there are no clinical trials that directly explored this issue by comparing the 7 Screening for Prediabetes and Diabetes 105

long-term effects of initiating treatment in screening-detected vs. clinically detected patients with diabetes. Regarding initiating treatment in the prediabetic state, several clinical trials have studied the effects of introducing lifestyle modi fi cations and various medications in patients with prediabetes following screening. In the DPP after 3 years of intervention with either lifestyle modifi cations or metformin, there were no differences in cardiovascular events between the various groups. However, the study was not powered to assess cardiovascular outcomes among the treatment arms. Encouraging was a signi fi cant improvement in the overall cardiovascular risk following intervention which may be translated in the future to decreased cardiovascular morbidity and mortality [7 ] . Intensive lifestyle intervention signi fi cantly increased the HDL-cholesterol level, reduced triglyceride level and the proatherogenic LDL phenotype B, and resulted in less people requiring antihypertensive medications suggesting that longer intervention may be proved bene fi cial in reducing cardiovascular event rate. In the Study to Prevent Non- Insulin Dependent Diabetes Mellitus (STOP-NIDDM) trial, despite a disappointing attrition rate, decreasing postprandial hyperglycemia with acarbose among participants with impaired glucose tolerance (IGT) was associated with a 49% relative risk reduction in the development of cardiovascular events (hazard ratio, 0.51; 95% con fi dence interval; 0.28–0.95; P = 0.03) and a 2.5% absolute risk reduction [8, 9 ] . Acarbose was also associated with a 34% relative risk reduction in the incidence of hypertension. In the Diabetes Reduction Assessment with Ramipril and Rosiglitazone Medication (DREAM) trial, in addition to prevention of diabetes, rosiglitazone was found to reduce the incidence of renal disease [ 9 ] , although the incidence rate of the primary composite outcome of renal and cardiovascular events was similar between the treatment groups. 3 . Asymptomatic phase in type 2 diabetes: A prolonged asymptomatic period exists for many patients with type 2 diabetes, to the point that at the time diabetes becomes clinically apparent, diabetes-related complications have been reported to already exist at rates as high as 48% for neuropathy, 10–27% for microalbuminuria, and up to 20% for retinopathy [10– 12] . Early detection with adequate treatment to reduce the risk for microvascular complications and to halt the progression of complications among those who are already affected may be proved benefi cial, though this has not been addressed yet in an intervention trial. 4 . Suitable tests for prediabetes / diabetes screening : There are several approaches for diabetes and prediabetes screening, which consist of either screening using the gold standard methods for diagnosis of diabetes or using risk assessment tools to prescreen the population to detect high-risk groups before using invasive and more expensive tests. • Fasting plasma glucose ( FPG ), glucose tolerance test ( GTT ), and Hemoglobin A1c : As opposed to coronary heart disease or various cancers, the gold standard for diagnosis of prediabetes and/or diabetes are indices of glycemia, which are simple and minimally invasive procedures therefore argue in favor of using the same tools also for screening. Using this approach, either with FPG, 2 h post a 75 g of glucose load (GTT) or hemoglobin A1c (HbA1c), one 106 A. Tirosh

can both screen for undiagnosed patients with preexisting diabetes as well as to detect patients at high risk for the disease, such as those with impaired fasting glucose (IFG), IGT, and HbA1c of 5.7–6.4%. All three tests are reasonable to be used for screening, though they detect somewhat different populations, with different sensitivities for diabetes prediction as well as for prediction of cardiovascular outcomes and microvascular complications. Most of the professional association guidelines do not suggest screening the entire population directly with tests for glycemia and recommend risk assessment as summarized in Table 7.1 . Nevertheless, based on those recommendations, usually the presence of one risk factor is suf fi cient for screening and as suggested by the American Diabetes Association, even in the absence of risk factors testing should be started for the entire population aged 45 years or above. Clearly, with signi fi cant improvement in instrumentation and standardization, the HbA1c assay has become accurate and precise. In addition, samples can be obtained at any time and do not require 8 h fasting as for FPG measurement or for the patient to wait for 2 h in the clinic as for GTT. It has been also shown that HbA1c has substantially less biologic variability and is relatively unaffected by acute illness or stress. As a measure of long-term glycemic exposure, HbA1c has been shown to be better and more consistently correlated with retinopathy in diabetic patients, which have established widely accepted HbA1c treatment goals for diabetes. A limitation of HbA1c for diagnosis is that a specifi c intermediate threshold at which increased risk for diabetes clearly begins cannot be easily defi ned, though this seems to be the case also with direct glucose measurements [ 13] . In a recent study using the US National Health and Nutrition Examination Survey (NHANES ) database, it has been shown that screening population using HbA1c criteria results in substantially lower prevalence of undiagnosed and total diabetes, and of prediabetes, than the prevalences estimated from FPG or GTT [14 ] (Fig. 7.1 ). Standardized prevalence of undiagnosed diabetes using HbA1c criteria was found to be one third that using either FPG or GTT criteria, resulting in a prevalence of total diabetes using HbA1c criteria one quarter less than that of total diabetes using either glucose criteria. However, in practice, a large portion of the population with type 2 diabetes remains unaware of their condition. Thus, it is conceivable that the lower sensitivity of A1C at the designated cut point will be offset by the test’s greater practicality, and that wider application of a more convenient test (HbA1c) may actually increase the number of diagnoses made [15 ] . Given the likelihood of HbA1c becoming a more widely used test for diabetes as well as prediabetes screening, it is important to note that at this point, most of the data available of the bene fi ts of lifestyle modi fi cations or medications in diabetes preventions are obtained from randomized controlled trials using FPG and GTT for detection of high-risk groups. It should also be noted that HbA1c measurement is relatively inac- curate in estimating glycemia status under certain circumstances such as in the presence of hemoglobinopathies or disorders affecting red cell turnover. In these settings, screening using FPG or GTT would be more appropriate. 7 Screening for Prediabetes and Diabetes 107

Table 7.1 Guidelines and consensus statements for diabetes screening The American Diabetes Association Testing should be considered in all adults who are (ADA), 2011 overweight (BMI ³ 25 kg/m2 , or lower in some ethnic groups) with at least one additional risk factor 1. Physical inactivity 2. First-degree relative with diabetes 3. High-risk race/ethnicity (e.g., African American, Latino, Native American, Asian American, Paci ﬁ c Islander) 4. Women who delivered a baby weighing ³ 9 lb or were diagnosed with GDM 5. Hypertension 6. HDL-cholesterol level £ 35 mg/dL and/or a triglyceride level ³ 250 mg/dL 7. Women with polycystic ovarian syndrome (PCOS) 8. A1C ³ 5.7%, IGT, or IFG on previous testing 9. Other clinical conditions associated with insulin resistance (e.g., severe obesity, acanthosis nigricans) 10. History of CVD In the absence of the above criteria, testing for diabetes should begin at age 45 years If results are normal, testing should be repeated at least at 3-year intervals, with consideration of more frequent testing depending on initial results and risk status American Association of Clinical Diabetes risk factors that justify screening: Endocrinologists (AACE), 2011 Family history of diabetes mellitus Cardiovascular disease Being overweight or obese Sedentary lifestyle Nonwhite ancestry Previously identi ﬁ ed impaired glucose tolerance, impaired fasting glucose, and/or metabolic syndrome Hypertension Increased levels of triglycerides, low concentrations of high-density lipoprotein cholesterol, or both History of gestational diabetes mellitus Delivery of a baby weighing more than 4 kg (9 lb) Polycystic ovary syndrome Antipsychotic therapy for schizophrenia and/or severe bipolar disease The ADA and the European Screening for prediabetes and type 2 diabetes in Association for the Study asymptomatic people should be considered in adults of Diabetes (EASD), 2009 who are overweight or obese (BMI ³ 25 kg/m2 ) with at least one more additional risk factor. Otherwise, testing should begin at age 45 years, and if results are normal, testing should be repeated at least at 3-year intervals (continued) 108 A. Tirosh

Table 7.1 (continued) US Preventive Services Task Force All adults with a sustained blood pressure of greater than (USPSTF), 2008 135/80 mmHg should be screened for diabetes Current evidence is insuf fi cient to assess balance of bene fi ts and harms of routine screening for type 2 diabetes in asymptomatic, normotensive patients The European Society of Cardiology Primary screening for the potential type 2 diabetes can be (ESC) and the EASD, 2007 done most ef fi ciently using a noninvasive risk score, combined with a diagnostic oral glucose tolerance testing in people with high score values

Fig. 7.1 Prevalence of diabetes and high risk for diabetes using A1C criteria in the US Population in 1988–2006. Undiagnosed diabetes in the US population aged ³ 20 years by three diagnostic criteria—NHANES 2005–2006. The thresholds of diagnostic criteria for diabetes were A1C ³ 6.5%, FPG ³ 7.0 mmol/l, and 2-h glucose ³11.1 mmol/1. Point estimates (%) and 95% CIs for the categories are: A1C alone = 0.3 (0.0–0.7); FPG alone = 0.2 (0.0–0.5); 2-h glucose alone = 2.5 (1.9– 3.2); A1C and FPG not 2-h glucose = 0.0; A1C and 2-h glucose not FPG = 0.1 (0.0–0.3); FPG and 2-h glucose not A1C = 1.0 (0.3–1.8); A1C, FPG, and 2-h glucose = 1.2 (0.5–2.0); total A1C = 1.6 (0.7–2.5); total FPG = 2.5 (1.2–3.8); total 2-h glucose = 4.9 (3.4–6.4); diagnosed diabetes = 7.8 (6.7–8.8); nodiabetes = 86.9 (84.6–89.1). Adapted from Cowie et al. [14 ]

• Diabetes risk scores : As mentioned above, while some risk assessment is recommended even when using glycemia indices for screening (Table 7.1 ), a large proportion of the adult population, especially in Western societies, would be qualiﬁ ed for screening using HbA1c or glucose testing. In an attempt to better identify subgroups at higher risk for the disease, several risk assessment tools have been developed. Using these tools, which comprise simple questionnaires about important diabetes risk factors (such as age, weight, family history of diabetes, personal history of hypertension, physical activity) and 7 Screening for Prediabetes and Diabetes 109

basic anthropometric measurements, followed by invasive glucose or HbA1c assessments of the moderate-to-high risk groups only, is suggested to provide a more cost-effective approach for diabetes screening than using invasive tests with a minimal or no prescreen. The most widely studied is the Finnish Type 2 Diabetes Risk Score (FINDRISC). The FINDRISC uses a composite of eight risk parameters in the adult population to calculate the overall estimated incidence rate of diabetes in 10 years in the Finnish population (Fig. 7.2). However, the risk score has been validated in various populations with a good sensitivity and speci fi city [16– 18 ] . A moderate risk score which corresponds to approximately 15–20% incidence rate could be predicted with a sensitivity and specifi city of 60–80%. Comparable results were reported with a German diabetes risk score which includes more lifestyle parameters such as physical activity and nutrition habits [ 19] . Other risk scores have been developed and validated in diverse populations, although in contrast to the FINDRISC, few have been validated prospectively [20, 21 ] . Despite the potential clinical utility of diabetes risk scores, there is a lack of data from randomized controlled trials investigating the long-term benefi ts from interventions based on these tools. Additional data is also required about the utility of the various risk scores in predicting diabetes in different ethnic groups. • Combination of risk scores and laboratory testing : Other more complex models using risk factor assessment combined with laboratory testing have been devised to predict the likelihood of developing type 2 diabetes [ 22 ] . While this approach may be less practical for community-based screening of mass populations, it does help to better predict the risk for diabetes in specifi c populations while taking into account a combination of risk factors, including both family history, lifestyle parameters, anthropometric measurements, and blood biomarkers. In this regard, several models for diabetes prediction were compared in the Framingham Offspring Study [ 23 ] . The model included information about age, parental history of diabetes, BMI, blood pressure, HDL-cholesterol, triglycerides, and IFG. Addition of GTT and measurements of insulin sensitivity as well as the presence of diabetes risk alleles did not improve the power of the model [24 ] . It is important to note that a strong interaction between risk factors exists for estimation of diabetes incidence in different population. For example, among young normal-weight men, normoglycemia (FPG < 100 mg/dL) is not associated with an increased risk for diabetes, whereas FPG > 90 mg/dL is already associated with approximately fi vefold increase in diabetes risk in overweight men (BMI of 25–29.9 kg/m2 ) and nearly eightfold increase in the risk was observed with FPG > 87 mg/dL in obese men (BMI ³ 30 kg/m 2 ) [13 ] . Thus, the concept of strict “cut-off” values may be misleading for assessing diabetes risk [ 25 ] . This concept has recently been incorporated to an Israeli Diabetes Risk Score matrix (Fig. 7.3 ) derived from the Metabolic Life-style and Nutrition Assessment in Young Adult (MELANY) cohort, a model which has not yet been validated in other populations [ 13, 25, 26] . These studies reveal that at least in young adults, the risk for diabetes increases almost as a continuum across the range of various 110 A. Tirosh

Fig. 7.2 Finnish type 2 diabetes risk score. FINnish Diabetes Risk SCore (FINDRISC) to assess the 10-year risk of type 2 diabetes in adults. Available at: http://www.diabetes. fi /english . ( http://www.diabetes. fi / fi les/502/eRiskitestilomake.pdf ) 7 Screening for Prediabetes and Diabetes 111

Fig. 7.3 Israeli diabetes risk score for young adults

parameters such as BMI, FPG, and triglyceride levels. Thus, the overall risk for diabetes even with FPG in the “pre-diabetes” range may not be signifi cantly elevated in the setting of normal BMI and low level of triglycerides, while high incidence rate for the disease could be observed among normoglycemic men but with positive family history of diabetes, high triglyceride level, and obesity. This may be especially important for enrollment participants to DPPs and in calculating their effi cacy as using only a single parameter (such as FPG) may result in misclassifi cation of patients to “high risk” group or excluding participants who may be at high risk otherwise. • Using genetic markers for diabetes prediction : In recent years, using genome- wide association studies, several genotypes have been identifi ed to alter the 112 A. Tirosh

susceptibility to type 2 diabetes [27 ] . It was therefore anticipated that using these fi ndings on their own or in combination with existing phenotype-based risk algorithms will improve risk models. Unfortunately, the results at this point are disappointing. The addition of genetic data to clinical models had a minimal effect on prediction of type 2 diabetes [28– 30 ] . In one such study, the phenotype-based risk models provided greater discrimination for type 2 diabetes than did models based on 20 common independently inherited diabetes risk alleles. The addition of genotypes to phenotype-based risk models produced only minimal improvement in accuracy of risk estimation [30 ] . Thus, at the current time, there is insuf fi cient evidence to support genotyping for risk assessment in clinical practice. 5 . Cost- effectiveness—Assessment of both the effectiveness and the cost of diabetes screening and prevention is a matter of ongoing debate. The wide variety of screening tools, the availability of various approaches for prevention and/or treatment for prediabetes and diabetes, and the lack of data from long-term clinical trials on the benefi ts of targeted screening (such as delay or prevention of late complications) make a reliable cost-bene fi t analysis almost impossible. In addition, while the ef fi cacy of lifestyle modi fi cations in a “real life” setting, the most effective approach for diabetes prevention in clinical trials, may be disappointing [ 31, 32 ] , the long-term benefi ts from successful implementation of lifestyle modifi cations are expected to exceed diabetes prevention per se and to result also in reduced incidence rates of hypertension, dyslipidemia, cardiovascular events, and potentially even cancer. Given the lack of data from clinical trials, a comparison in cost-effectiveness of eight screening strategies was conducted using a computer- based model of simulated US population of 325,000 people aged 30 years without diabetes [33 ] . The benefi ts of early detection for all screening strategies included a reduced incidence of myocardial infarction, microvascular complications, and an increase in quality-adjusted life years (QALYs) over 50 years of age. The most cost-effective strategies were those that started between the ages of 30 and 45 years, with screening repeated every 3–5 years. In addition, two studies suggested that targeted screening for type 2 diabetes among persons with hypertension may be relatively cost-effective in terms of macrovascular outcomes [34, 35 ] . Based on these models, older persons benefi tted more than younger persons, and limited screening of obese persons was more cost-effective than mass screening [ 34 ] . In addition, screening for prediabetes and undiagnosed type 2 diabetes, followed by intervention (lifestyle or pharmacological), was more cost-effective than no screening [ 36 ] . A more direct evidence of the cost-effectiveness of diabetes screening awaits the completion of the ADDITION-Cambridge trial aimed at evaluating the cost and effectiveness of a stepwise screening strategy for type 2 diabetes and intensive treatment for people with screen-detected diabetes in a primary care setting. Primary endpoints are the overall cardiovascular risk assessed at 1 year and cardiovascular mortality and morbidity at 5 years after diagnosis of diabetes. Secondary endpoints include all-cause mortality, development of renal and visual impairment, peripheral neuropathy, health service costs, self-reported quality of life, functional status, and health utility [37 ] . 7 Screening for Prediabetes and Diabetes 113

Guidelines Recommendations

In recent years, with detection of better screening tools, new treatment options for diabetes and prediabetes and with more evidence obtained from clinical trials, the screening recommendations by expert groups have been frequently modi fi ed. The two main approaches to screening are either to screen the entire population above a certain age or to limit the screening to “high-risk” groups. The American Diabetes Association in a joint statement with the European Association for the Study of Diabetes (EASD) in 2009 and again in its updated guidelines in 2011 [ 38 ] recommends testing for diabetes or prediabetes in all adults with BMI ³ 25 kg/m2 and one or more additional risk factors for diabetes (Table 7.1 ). Regardless of additional risk factors, the entire population should be tested starting at the age of 45 years. Either A1C, FPG, or 2-h GTT is appropriate for testing. If the tests are normal, the patient should be retested in 3 years. The American Association of Clinical Endocrinologists (AACE) has also adopted similar approach in its 2011 guidelines with the addition of high triglycerides, low HDL-cholesterol, and antipsychotic therapy for schizophrenia and/or severe bipolar disease as recommended conditions in which diabetes screening should be considered [ 39 ] . Another important difference is that the AACE, as opposed to the ADA, recommends using HbA1c only for screening, while diabetes diagnosis should continue to require glucose testing. The US Preventive Services Task Force (USPSTF) recommends limiting screening to adults with a sustained blood pressure of greater than 135/80 mmHg [40 ] . The Canadian Task Force on Preventive Health Care recommended in 2005 screening all patients with hypertension or hyperlipidemia [ 41 ] , though an update process has been initiated during 2011. In a retrospective analysis, compared with the ADA recommendations, the new USPSTF guidelines result in a lower number of patients eligible for screening (26% vs. 66%) and in a decrease case- fi nding rate [42 ] . In contrary to the above approaches, and given the good experience in using the FINDRISC in European cohorts, the European Society of Cardiology has recommended that screening for type 2 diabetes can be done most ef fi ciently using a noninvasive risk score, combined with a diagnostic oral glucose tolerance testing in people with high score values [43 ] .

References

1. Wilson JMG, Jungner G. Principles and practice of screening for disease. Geneva: WHO; 1968. 2. Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346:393–403. 3. Federation ID. Diabetes atlas. 4th ed. Brussels: International Diabetes Federation; 2009. 4. Intensive blood-glucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS 33). UK Prospective Diabetes Study (UKPDS) Group. Lancet. 1998;352:837–53. 5. Holman RR, Paul SK, Bethel MA, Matthews DR, Neil HA. 10-year follow-up of intensive glucose control in type 2 diabetes. N Engl J Med. 2008;359:1577–89. 114 A. Tirosh

6. Gerstein HC, Miller ME, Byington RP, et al. Effects of intensive glucose lowering in type 2 diabetes. N Engl J Med. 2008;358:2545–59. 7. Ratner R, Goldberg R, Haffner S, et al. Impact of intensive lifestyle and metformin therapy on cardiovascular disease risk factors in the diabetes prevention program. Diabetes Care. 2005;28:888–94. 8. Chiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A, Laakso M. Acarbose for prevention of type 2 diabetes mellitus: the STOP-NIDDM randomised trial. Lancet. 2002;359:2072–7. 9. Chiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A, Laakso M. Acarbose treatment and the risk of cardiovascular disease and hypertension in patients with impaired glucose tolerance: the STOP-NIDDM trial. JAMA. 2003;290:486–94. 10. Harris MI, Klein R, Welborn TA, Knuiman MW. Onset of NIDDM occurs at least 4–7 yr before clinical diagnosis. Diabetes Care. 1992;15:815–9. 11. Klein R, Klein BE, Moss S, DeMets DL. Proteinuria in diabetes. Arch Intern Med. 1988;148:181–6. 12. Spijkerman AM, Dekker JM, Nijpels G, et al. Microvascular complications at time of diagnosis of type 2 diabetes are similar among diabetic patients detected by targeted screening and patients newly diagnosed in general practice: the hoorn screening study. Diabetes Care. 2003;26:2604–8. 13. Tirosh A, Shai I, Tekes-Manova D, et al. Normal fasting plasma glucose levels and type 2 diabetes in young men. N Engl J Med. 2005;353:1454–62. 14. Cowie CC, Rust KF, Byrd-Holt DD, et al. Prevalence of diabetes and high risk for diabetes using A1C criteria in the U.S. population in 1988–2006. Diabetes Care. 2010;33:562–8. 15. American Diabetes Association. Diagnosis and classi fi cation of diabetes mellitus. Diabetes Care. 2011;34 Suppl 1:S62–9. 16. Makrilakis K, Liatis S, Grammatikou S, et al. Validation of the Finnish diabetes risk score (FINDRISC) questionnaire for screening for undiagnosed type 2 diabetes, dysglycaemia and the metabolic syndrome in Greece. Diabetes Metab. 2011;37(2):144–51. 17. Bergmann A, Li J, Wang L, Schulze J, Bornstein SR, Schwarz PE. A simpli fi ed Finnish diabetes risk score to predict type 2 diabetes risk and disease evolution in a German population. Horm Metab Res. 2007;39:677–82. 18. Saaristo T, Moilanen L, Korpi-Hyovalti E, et al. Lifestyle intervention for prevention of type 2 diabetes in primary health care: one-year follow-up of the Finnish National Diabetes Prevention Program (FIN-D2D). Diabetes Care. 2010;33:2146–51. 19. Schulze MB, Hoffmann K, Boeing H, et al. An accurate risk score based on anthropometric, dietary, and lifestyle factors to predict the development of type 2 diabetes. Diabetes Care. 2007;30:510–5. 20. Spijkerman AM, Yuyun MF, Grif fi n SJ, Dekker JM, Nijpels G, Wareham NJ. The performance of a risk score as a screening test for undiagnosed hyperglycemia in ethnic minority groups: data from the 1999 health survey for England. Diabetes Care. 2004;27:116–22. 21. Mohan V, Deepa R, Deepa M, Somannavar S, Datta M. A simpli fi ed Indian Diabetes Risk Score for screening for undiagnosed diabetic subjects. J Assoc Physicians India. 2005;53: 759–63. 22. Norberg M, Eriksson JW, Lindahl B, et al. A combination of HbA1c, fasting glucose and BMI is effective in screening for individuals at risk of future type 2 diabetes: OGTT is not needed. J Intern Med. 2006;260:263–71. 23. Wilson PW, Meigs JB, Sullivan L, Fox CS, Nathan DM, D’Agostino Sr RB. Prediction of incident diabetes mellitus in middle-aged adults: the Framingham Offspring Study. Arch Intern Med. 2007;167:1068–74. 24. Meigs JB, Shrader P, Sullivan LM, et al. Genotype score in addition to common risk factors for prediction of type 2 diabetes. N Engl J Med. 2008;359:2208–19. 25. Tirosh A, Shai I, Afek A, et al. Adolescent BMI trajectory and risk of diabetes versus coronary disease. N Engl J Med. 2011;364:1315–25. 26. Tirosh A, Shai I, Bitzur R, et al. Changes in triglyceride levels over time and risk of type 2 diabetes in young men. Diabetes Care. 2008;31:2032–7. 7 Screening for Prediabetes and Diabetes 115

27. McCarthy MI. Genomics, type 2 diabetes, and obesity. N Engl J Med. 2010;363:2339–50. 28. Lyssenko V, Jonsson A, Almgren P, et al. Clinical risk factors, DNA variants, and the development of type 2 diabetes. N Engl J Med. 2008;359:2220–32. 29. Cornelis MC, Qi L, Zhang C, et al. Joint effects of common genetic variants on the risk for type 2 diabetes in U.S. men and women of European ancestry. Ann Intern Med. 2009;150:541–50. 30. Talmud PJ, Hingorani AD, Cooper JA, et al. Utility of genetic and non-genetic risk factors in prediction of type 2 diabetes: Whitehall II prospective cohort study. BMJ. 2010;340:b4838. 31. King DE, Mainous 3rd AG, Carnemolla M, Everett CJ. Adherence to healthy lifestyle habits in US adults, 1988–2006. Am J Med. 2009;122:528–34. 32. Serour M, Alqhenaei H, Al-Saqabi S, Mustafa AR, Ben-Nakhi A. Cultural factors and patients’ adherence to lifestyle measures. Br J Gen Pract. 2007;57:291–5. 33. Kahn R, Alperin P, Eddy D, et al. Age at initiation and frequency of screening to detect type 2 diabetes: a cost-effectiveness analysis. Lancet. 2010;375:1365–74. 34. Waugh N, Scotland G, McNamee P, et al. Screening for type 2 diabetes: literature review and economic modelling. Health Technol Assess. 2007;11:iii–iv; ix–xi; 1–125. 35. Hoerger TJ, Harris R, Hicks KA, Donahue K, Sorensen S, Engelgau M. Screening for type 2 diabetes mellitus: a cost-effectiveness analysis. Ann Intern Med. 2004;140:689–99. 36. Gillies CL, Lambert PC, Abrams KR, et al. Different strategies for screening and prevention of type 2 diabetes in adults: cost effectiveness analysis. BMJ. 2008;336:1180–5. 37. Echouffo-Tcheugui JB, Simmons RK, Williams KM, et al. The ADDITION-Cambridge trial protocol: a cluster—randomised controlled trial of screening for type 2 diabetes and intensive treatment for screen-detected patients. BMC Public Health. 2009;9:136. 38. American Diabetes Association. Standards of medical care in diabetes—2011. Diabetes Care. 2011;34 Suppl 1:S11–61. 39. Handelsman Y, Mechanick JI, Blonde L, et al. American Association of Clinical Endocrinologists Medical Guidelines for Clinical Practice for developing a diabetes mellitus comprehensive care plan. Endocr Pract. 2011;17 Suppl 2:1–53. 40. U.S. Preventive Services Task Force. Screening for type 2 diabetes mellitus in adults: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2008;148:846–54. 41. Feig DS, Palda VA, Lipscombe L. Screening for type 2 diabetes mellitus to prevent vascular complications: updated recommendations from the Canadian Task Force on Preventive Health Care. CMAJ. 2005;172:177–80. 42. Sheehy AM, Flood GE, Tuan WJ, Liou JI, Coursin DB, Smith MA. Analysis of guidelines for screening diabetes mellitus in an ambulatory population. Mayo Clin Proc. 2010;85:27–35. 43. Ryden L, Standl E, Bartnik M, et al. Guidelines on diabetes, pre-diabetes, and cardiovascular diseases: executive summary. The Task Force on Diabetes and Cardiovascular Diseases of the European Society of Cardiology (ESC) and of the European Association for the Study of Diabetes (EASD). Eur Heart J. 2007;28:88–136. Chapter 8 Neuropathy in Prediabetes and the Metabolic Syndrome

Aaron I. Vinik and Marie-Laure Nevoret

Introduction

Distal symmetric polyneuropathy (DSPN) and cardiac autonomic neuropathy (CAN) are major chronic complications of diabetes and are associated with loss of quality-of-life (QOL) [1 ] and increased morbidity and mortality [ 2, 3 ] . Diabetes is known to be a major cause of peripheral neuropathy [ 4 ] . Neuropathy was considered a chronic complication only occurring after many years of diabetes, but there is now evidence that symptoms, signs, and objective evidence of neuropathy are found as early as from the time of diagnosis of diabetes [ 5 ] . While there have been suggestions of neuropathy preceding the onset of diabetes [ 6 ] , there has been some debate as to the validity of these claims [ 7] . Yet despite these doubts, a careful and detailed review of the relationship between glycemic control and evidence for neuropathy suggests that the clock starts ticking before the advent of diabetes by current de fi nition [8 ] . Perhaps what led to the controversy is not whether or not there was a neuropathy in prediabetes, but what the defi nition of neuropathy should be. Indeed, there is now widespread recognition that neuropathy may be small or large fi ber, proximal or distal, acute or chronic [ 9] , and that it may occur with symptoms or signs demonstrable only on nerve conduction study (NCS), quantitative sensory testing (QST), quantitative autonomic function test (QAFT), or using modern techniques such as skin biopsy with quantifi cation of intraepidermal nerve fi ber density (IENF) [10 ] . With this revision in the thinking of what neuropathy constitutes, prediabetes is without doubt increasingly being recognized as an important contributor to neuropathy and has moved neurologists to reexamine patients with idiopathic or

A. I. Vinik , MD, PhD, FCP, MACP (*) • M.-L. Nevoret , MD Strelitz Diabetes Center for Endocrine and Metabolic Disorders and Neuroendocrine Unit , Eastern Virginia Medical School , Norfolk , VA 23510 , USA e-mail: [email protected]

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 117 DOI 10.1007/978-1-4614-3314-9_8, © Springer Science+Business Media New York 2012 118 A.I. Vinik and M.-L. Nevoret cryptogenic neuropathy by carrying out glucose tolerance tests to reveal the true nature of their disorder [ 4 ] . Thus, it is now conceded that the diagnostic workup for idiopathic neuropathy must include evaluation of glucose homeostasis. Here I will review the current understanding of what constitutes neuropathy and how this relates to the role of prediabetes and the metabolic syndrome.

De ﬁ nition of Neuropathy

Diabetic peripheral neuropathy (DPN) is a common late complication of diabetes. It results in a variety of syndromes for which there is no universally accepted unique classi fi cation. They are generally subdivided into focal/multifocal neuropathies, including diabetic amyotrophy, and symmetric polyneuropathies, including sensorimotor polyneuropathy (DSPN). The latter is the most common type, affecting about 30% of diabetic patients in hospital care and 25% of those in the community [9, 11] . DSPN has been recently defi ned as a symmetrical, length-dependent sensorimotor polyneuropathy attributable to metabolic and microvascular alterations as a result of chronic hyperglycemia exposure (diabetes) and cardiovascular risk covariates [12 ] . Its onset is generally insidious, and without treatment, its course is chronic and progressive. The loss of small fi ber-mediated sensation results in the loss of thermal and pain perception, whereas large fi ber impairment results in loss of touch and vibration perception. Sensory fi ber involvement may also result in “positive” symptoms, such as paresthesias and pain, although up to 50% of neuropathic patients are asymptomatic. DSPN can be associated with the involvement of the autonomic nervous system, i.e., diabetic autonomic neuropathy which rarely causes severe symptoms [13, 14 ] , but in its cardiovascular form is de fi nitely associated with at least a threefold increased risk for mortality [3, 15, 16 ] . Pain is the reason for 40% of patient visits in a primary care setting and about 20% of these patients have had pain for greater than 6 months [17 ] . Chronic pain may be nociceptive which occurs as a result of disease or damage to tissue wherein there is no abnormality in the nervous system or there may be no somatic abnormality. By contrast, Experts in the Neurology and Pain community defi ne neuropathic pain as “pain arising as a direct consequence of a lesion or disease affecting the somatosensory system” [18 ] . Persistent neuropathic pain interferes considerably with QOL, impairing sleep and recreation; signifi cantly impacts emotional well- being; and is associated with, if not the cause of, depression, anxiety, and noncom- pliance with treatment [19 ] . Diabetic neuropathy pain (DNP) is a dif fi cult-to-manage clinical problem. It is often associated with mood and sleep disturbances and patients with DNP are more apt to seek medical attention than those with other types of diabetic neuropathy. Two population-based studies showed that neuropathic pain is associated with a greater psychological burden than nociceptive pain [ 20 ] and is considered to be more severe than other pain types. Early recognition of psychological problems is critical to the management of pain and physicians need to go beyond the management of pain per se if they are to achieve success. Patients may 8 Neuropathy in Prediabetes and the Metabolic Syndrome 119 also complain of decreased physical activity and mobility, increased fatigue, and negative effects on their social lives. Providing signifi cant pain relief markedly improves quality-of-life measures, including sleep and vitality [1, 21 ] .

Epidemiology of Neuropathic Pain

Neuropathic pain is not uncommon. A population-based survey of 6,000 patients treated in family practice in the UK reported a 6% prevalence of pain predominantly of neuropathic origin [ 22 ] . Similarly, a large population-based study in France showed that 6.9% of the population had neuropathic pain [20 ] . Interestingly, in a Dutch population survey of >362,000 people, younger people with pain tended to be mostly women, but with advancing age the gender differences disappeared. Perhaps a little recognized fact is that mononeuritis and entrapments were three times as common as DPN, and fully one third of the diabetic population has some form of entrapment [ 23 ] , which when recognized is readily amenable to intervention [24 ] . Even more salutary is the mounting evidence that even with impaired glucose tolerance (IGT), patients may experience pain [25– 27 ] . In the general population (region of Augsburg, Southern Germany), the prevalence of painful PN was 13.3% in the diabetic subjects, 8.7% in those with IGT, 4.2% in those with impaired fasting glucose (IFG), and 1.2% in those with normal glucose tolerance (NGT) [ 28 ] . Among survivors of myocardial infarction (MI) from the Augsburg MI Registry, the prevalence of neuropathic pain was 21.0% in the diabetic subjects, 14.8% in those with IGT, 5.7% in those with IFG, and 3.7% in those with NGT [27 ] . Thus, subjects with macrovascular disease appear to be prone to neuropathic pain. The most important risk factors of DSPN and neuropathic pain in these surveys were age, obesity, and low physical activity, while the predominant comorbidity was peripheral arterial disease, highlighting the paramount role of cardiovascular risk factors and diseases prevalent in DSPN. As a corollary, patients presenting with painful neuropathy have IFG or IGT, and about 50% of the time are overweight and have autonomic dysfunction [ 25 ] . Even in the absence of elevated fasting blood glucose (<100 mg/dL), pain may be the presenting feature of the metabolic syndrome and cosegregates with elevated triglycerides and a low HDL-C [ 6] . Indeed, a risk factor for neuropathic pain in diabetic and nondiabetic populations is an impairment of peripheral vascular function [ 27, 29 ] .

Diagnostic Aspects of Diabetic Neuropathy

DPN is a common late complication of diabetes. It results in a variety of syndromes for which there is no universally accepted unique classiﬁ cation. They are generally subdivided into focal/multifocal neuropathies, including diabetic amyotrophy, and symmetric polyneuropathies, including sensorimotor polyneuropathy (DSPN). 120 A.I. Vinik and M.-L. Nevoret

The latter is the most common type, affecting about 30% of diabetic patients in hospital care and 25% of those in the community [ 12 ] . Because of the lack of agreement on the de fi nition and diagnostic assessment of neuropathy, several consensus conferences were convened to overcome the current problems, the most recent of which has rede fi ned the minimal criteria for the diagnosis of typical DSPN: 1 . Possible DSPN: The presence of symptoms or signs of DSPN including any of the following: symptoms—decreased sensation; positive neuropathic sensory symptoms (e.g., “asleep numbness,” prickling or stabbing, burning or aching pain) predominantly in the toes, feet, or legs; or signs—symmetric decrease of distal sensation or unequivocally decreased or absent ankle re flexes. 2 . Probable DSPN : The presence of a combination of symptoms and signs of neuropathy including ³ 2 of any of the following: neuropathic symptoms, decreased distal sensation, or unequivocally decreased or absent ankle re fl exes. 3 . Con fi rmed DSPN : The presence of an abnormality of nerve conduction and a symptom or symptoms or a sign or signs of neuropathy confi rms DSPN. If nerve conduction is normal, a validated measure of small fi ber neuropathy (SFN) (with class 1 evidence) may be used. To assess for the severity of DSPN, several approaches can be recommended: the graded approach outlined above; various continuous measures of sum scores of neurologic signs, symptoms, or nerve test scores; scores of function of activities of daily living (ADLs); or scores of predetermined tasks or of disability. 4 . Subclinical DSPN : The presence of no signs or symptoms of neuropathy is con fi rmed with abnormal nerve conduction or a validated measure of SFN (with class 1 evidence). De fi nitions 1, 2, or 3 can be used for clinical practice, and de fi nitions 3 or 4 can be used for research studies. 5 . Small fi ber neuropathy ( SFN): SFN should be graded as follows: (1) possible: the presence of length-dependent symptoms and/or clinical signs of small fi ber damage; (2) probable: the presence of length-dependent symptoms, clinical signs of small fi ber damage, and normal sural nerve conduction; and (3) de fi nite: the presence of length-dependent symptoms, clinical signs of small fi ber damage, normal sural nerve conduction, and altered IENFD at the ankle and/or abnormal thermal thresholds at the foot [12 ] . Figure 8.1 shows that measurements of small fi ber function have greater diagnostic sensitivity than large fi ber sensory or motor function.

Objective Devices for Diagnosis of Neuropathy

Minimum criteria for the clinical diagnosis of neuropathy according to the American Academy of Neurology are: a positive nerve symptom score and evidence of neurologic impairment demonstrated by a positive score using the nerve impairment 8 Neuropathy in Prediabetes and the Metabolic Syndrome 121

Fig. 8.1 Small fi ber measures have greater diagnostic sensitivity. Reproduced from Shun et al. [143 ] scoring system (NISS). However, this means that moderate signs without symptoms or mild signs with moderate symptoms would not qualify as neuropathy. This also certainly means that exclusive presence of neuropathic symptoms without objective evidence is insuffi cient to diagnose DPN. Therefore, early stages of DPN or a painful SFN with or without minimal defi cits are insuffi cient evidence without the cor- roborative verifi cation using more sophisticated tests such as the quantifi cation of thermal perception or skin biopsy.

Skin Biopsy and Quantitation of Intraepidermal Nerve Fiber Density

Skin biopsy has become a widely used tool to investigate small caliber sensory nerves including somatic unmyelinated intraepidermal nerve fi bers (IENF), dermal myelinated nerve fi bers, and autonomic nerve fi bers in peripheral neuropathies and other conditions [ 6, 30, 31] . Different techniques for tissue processing and nerve fi ber evaluation have been used. For diagnostic purposes in peripheral neuropathies, a recent guideline has recommended a 3-mm punch skin biopsy at the distal leg and quantifying the linear density of IENF in at least three 50-m m thick sections per biopsy, fi xed in 2% PLP or Zamboni’s solution, by bright-fi eld immunohistochem- istry or immuno fl uorescence with anti-protein gene product (PGP) 9.5 antibodies. Quanti fi cation of IENF density appeared more sensitive than sensory NCS and sural nerve biopsy in diagnosing SFN. 122 A.I. Vinik and M.-L. Nevoret

Corneal Confocal Microscopy

Corneal confocal microscopy (CCM) is a noninvasive technique used to detect small nerve ﬁ ber loss in the cornea which correlates with both increasing neuropathic severity and reduced IENFD in diabetic patients [ 32, 33 ] . A novel technique of real-time mapping permits an area of 3.2 mm 2 to be mapped with a total of 64 theoretically nonoverlapping single 400 m m 2 images [34 ] .

Contact Heat-Evoked Potentials

Contact heat-evoked potentials have now been studied in healthy controls, newly diagnosed diabetics, established diabetics, and patients with the metabolic syndrome. It does appear that CHEPS is capable of detecting SFN in the absence of other indices and that there is a correlation with quantitative sensory perception and objective tests of small ﬁ ber function such as the cooling detection threshold and cold pain [35 ] .

Impaired Fasting Glucose or Impaired Glucose Tolerance?

The American Diabetes Association defi nes prediabetes as two conditions, IFG ranging between 100 and 125 mg/dL and IGT, with plasma glucose ranging between 140 and 199 mg/dL at 2 h postglucose administration during an oral glucose tolerance test (OGTT) [36 ] . IGT is generally considered a better predictor of cardiovascular morbidity and mortality than IFG [ 37– 40 ] . However, the situation is less clear for microvascular disease. In DETECT 2, thresholds for diabetic retinopathy from receiver-operating curve analyses indicated a narrow threshold range for fasting plasm glucose (FPG), but not for 2-h post prandial glucose (PG) [ 41 ] . The major reason for this discrepancy is that the FPG may have been inappropriately high for the 2 h PG levels. Thus, the authors suggested that the current diagnostic level for diabetes by FPG could be lowered [ 41 ] . Sosenko et al. showed that only subjects with FPG ³ 126 mg/dL exhibited higher sensory thresholds for vibration and warm/ cold perception [42 ] . By contrast, the 2-h glucose levels (as in OGTT) could identify a substantially greater number of individuals earlier in the natural history of T2DM, before the development of complications [42 ] . Differences in the metabolic mechanisms between IFG and IGT appear to exist and account for the apparent differences in contribution to the complications of diabetes, suggesting that it would not be advisable to lump IFG and IGT under one bracket as impaired glucose regulation (IGR) [43, 44 ] . The vast majority of studies examining the association between pre diabetes and neuropathy have focused on IGT rather than IFG [44– 48 ] . More importantly, there is evidence from both 8 Neuropathy in Prediabetes and the Metabolic Syndrome 123 clinic-based [48– 50 ] and population-based studies [ 26– 28 ] that peripheral neuropathy and neuropathic pain are more frequent in IGT than in IFG [ 48– 50 ] . Carrying out an OGTT yields a higher diagnostic rate (62%) than FPG (39%) in the identifi cation of patients with IGR [50 ] , supporting the notion that the setpoint for FPG may still be too high. The MONICA/KORA surveys [26– 28 ] report a prevalence of neuropathy of 28% among diabetic subjects, 13% in IGT, 11.3% in IFG, and 7.4% in the general population with normal glucose tolerance (NGT) [26 ] . Prevalence of neuropathic pain was 13.3% in diabetes, 8.7% in IGT, 4.2% on IFG, and 1.2% among those with NGT [28 ] . Neuropathic pain was two to threefold more frequent among subjects with diabetes or IGT than among those with IFG or NGT [ 28 ] . Neuropathic pain was also more frequent among survivors of myocardial infarction, with a comparable gradient: 21% in diabetes, 14.8% in IGT, 5.7% in IFG, and 3.7% in NGT ( p < 0.001) [ 27 ] . This data provides the strongest evidence that in the general population neuropathic pain is approximately two to fourfold more frequent among subjects with diabetes or IGT than IFG or NGT [27 ] . While most studies suggest a stronger relationship of neuropathy with IGT than with IFG, diagnosis of IFG is not adequately speci fi ed. It appears that the majority of researchers have used the earlier cut-off value of glucose ³ 110 mg/dL [ 36 ] and not glucose ³ 100 mg/dL [ 51] . Thus, the question of an increased predisposition to neuropathy in IFG remains open. The notion that neuropathy may be associated with IGT launched the Impaired Glucose Tolerance Neuropathy (IGTN) Study [ 25, 52] and the Rochester Diabetic Neuropathy Study of patients with Impaired Glucose Metabolism (RDNS-IGM) [ 43] . The IGTN Study followed 71 patients with IGT and neuropathy for up to 3 years to ascertain early neuropathic changes and the potential benefi cial effect of lifestyle intervention [ 25, 52] . The RDNS-IGM Study aims to recruit about 600 subjects from the Olmsted County population (300 with IGT and 300 matched subjects with NGT), who will be followed for at least 10 years, in order to monitor the development of neuropathy in each group [ 43 ] . A preliminary report at the Peripheral Neuropathy Society meetings may be premature to conclude that there is no increased predisposition to neuropathy with IGM, and the criteria for neuropathy really need to be reevaluated.

Prediabetes in Patients with Neuropathy

In studies examining patients with unexplained neuropathy, diabetes has been sought as a cause [ 44, 46, 48– 50, 53– 55] . Neuropathy was either described in general as peripheral [53 ] , or speci ﬁ ed as chronic distal symmetric predominantly sensory polyneuropathy (DSPN) [ 44, 55 ] , painful [ 46, 49 ] , or axonal (chronic idiopathic axonal neuropathy [CIAP], in which electrophysiology demonstrated mainly axonal loss, without demyelination) [ 48, 50, 54] . Prediabetes was more frequent among neuropathic patients than in historical controls [44, 46, 48– 50, 53 ] . 124 A.I. Vinik and M.-L. Nevoret

However, this is not universal. Hughes et al. [54 ] compared 50 consecutive CIAP patients with 50 control subjects from the same region and found a 28.6% (14/49) frequency of prediabetes among patients vs. 12.2% (6/49) among controls. Frequency of prediabetes was 45.5% (10/22) in the subgroup of patients with pain vs. 14.8% (4/27) in those without pain [54 ] . After adjustment for age and sex, these differences were not signi fi cant [54 ] , but the direction was consistent with other observations. It appears that prediabetes is common in patients with neuropathy. The major fl aw in the argument is [ 44, 46, 48– 50, 53 ] the reliance on literature-based controls [ 56, 57 ] . These controls came from population-based studies that had been conducted some years ago and included subjects with extremely wide age ranges (20– 74 years) [ 56, 57 ] and, as already identi fi ed [ 58, 59] , this is not entirely valid. As pointed out above, Hughes et al. [ 54 ] recruited matched controls and, after controlling for age and gender, concluded that there was no increased predisposition for prediabetes. His report, however, included very small numbers of patients and may have been underpowered, especially in the subgroup analysis of painful vs. painless subjects which was suggestive of a relationship between painful neuropathy and prediabetes [ 54 ] . This again reinforces the notion that small fi ber neuropathies and those with pain in which objective evidence of neuropathy may be lacking are the principal form of prediabetic neuropathy. A further concern is the small number of subjects, both in patients with and without neuropathy, and the very limited data of detailed glucose metabolic studies using the OGTT. Of course clinic-based studies are always hampered by referral bias which could distort the observations [44, 46, 48– 50, 53– 55 ] .

Peripheral Neuropathy in Subjects with Prediabetes

Population studies have examined the presence of neuropathy in prediabetes [26– 28, 45, 60– 68] . Subjects with IGT have been reported to have neuropathy [ 26, 45 ] , neuropathic pain [ 27, 28 ] , impaired nerve conduction [62 ] , reduced sweat secretion [ 61 ] , and diminished sympathetic skin response [62 ] . The population-based MONICA/KORA surveys evaluated the relationship between IGT and neuropathy in the general population [26– 28 ] . A cut point >2 of the validated Michigan Neuropathy Screening Instrument (MNSI) was used for diagnosis of DSPN [69 ] . Prevalence of neuropathy was 28% among diabetic subjects, 13% among subjects with IGT, 11.3% among those with IFG, and 7.4% among those with NGT [26 ] . Prevalence of neuropathic pain in the lower limbs was 13.3% in diabetics, 8.7% among subjects with IGT, 4.2% among those with IFG, and 1.2% among those with NGT [26 ] . Neuropathic pain was two to threefold more frequent among subjects with diabetes or IGT than among those with IFG or NGT [ 28 ] . Neuropathic pain was also more frequent among survivors of myocardial infarction and the same gradient was observed: 21% in diabetes, 14.8% in IGT, 5.7% in IFG, and 3.7% in NGT (p < 0.001) [ 27 ] . Overall, neuropathic pain was two to fourfold more frequent among subjects with diabetes or IGT than IFG or NGT [27 ] . The strengths of the 8 Neuropathy in Prediabetes and the Metabolic Syndrome 125

KORA surveys [ 26– 28 ] are their population-based setting, the relatively large number of subjects enrolled, the separate evaluation of neuropathy and neuropathic pain, and the use of a validated diagnostic tool for neuropathy [ 70 ] . The results of the MONICA/KORA study suggest a continuum of neuropathy prevalence, particularly painful DSPN, in the general population from NGT through prediabetes to manifest diabetes [26– 28] . These observations are in agreement with the fi ndings from the Hoorn study [60 ] . Among 267 subjects with NGT, 167 with IGT, 90 with newly diagnosed diabetes mellitus, and 73 with previously known diabetes, prevalence of large fi ber dysfunction was highest in known diabetes, and large fi ber dysfunction showed a progressive increase in prevalence within the glucose dysregulation NGT- IGT-DM [60 ] . Objective evidence of this relationship was also found for thermal discrimination threshold (a measure of small fi ber function), which also increased with increasing fasting and postload insulin levels (p < 0.05) [ 60 ] . Conversely, some studies have not found an increased prevalence of DSPN in prediabetes [63– 68 ] . However, important limitations of these studies include the very low frequency of neuropathy in the population examined [ 64 ] ; the reliance on vibration perception threshold, which may not be an adequate measure of overall neuropathy, especially small fi ber function [64 ] ; the inclusion of subjects from very speci fi c populations, which may be different from the Caucasian subjects included in most studies [63, 64 ] ; the absence of a speci fi c aim to examine neuropathy in prediabetes [ 64] or of a control population for comparison; and the evaluation of daily chronic pain (which in diabetes is very different from neuropathic pain [ 71 ] with at least 3-month duration) rather than specifi cally addressing neuropathic pain and neuropathy [ 67, 68 ] . Very few natural history studies have been done. Eriksson et al. [ 66 ] followed age-matched men with IGT and diabetes mellitus prospectively for 12–15 years to examine the impact of long-term glucose intolerance on the development and progression of neuropathy. After 12–15 years of IGT, peripheral nerve function did not differ between subjects with IGT and NGT [66 ] . Conversely, diabetic patients exhibited lower nerve conduction velocity than IGT and control subjects [66 ] . The authors concluded that diabetes but not IGT is linked to peripheral nerve dysfunction. However, this study addressed the impact of long-term IGT on peripheral nerve function using measures that may not have been able to detect neuropathy; it did not address the question of whether neuropathy is frequent among IGT patients [66 ] .

Cardiac Autonomic Neuropathy in Subjects with Prediabetes

CAN detected by reduced heart rate variability (HRV) is associated with increased mortality in diabetes [72– 74 ] and cardiovascular disease [75 ] . In the general population [ 76 ] an imbalance in the autonomic nervous system between sympathetic and parasympathetic function may be the strongest predictor of poor cardiovascular outcome and the risk of sudden death [ 15 ] . This may be relevant to the observation that two clinic-based studies showed signi ﬁ cantly lower HRV and adrenergic innervation defects [ 77 ] in subjects with IGT. 126 A.I. Vinik and M.-L. Nevoret

Population-based surveys [62, 63, 78– 84 ] have demonstrated reduced HRV [ 78, 84] , impaired systolic pressure response to handgrip test [ 62 ] , and decreased parasympathetic activity (manifested by HF power and 30/15 ratio, resulting in a shift toward augmented sympathetic tone, expressed as an increased LF/HF ratio) [ 81] in IGT. In IFG, one study showed reduced HRV [ 79 ] , another showed reduction in some but not all HRV parameters [ 80] , while a third study was negative [ 81 ] . According to two de ﬁ nitions of IFG, HRV was more impaired in IFG (6.1– 6.9 mmol/L) and diabetes than in normal fasting glucose (4.5–5.5 mmol/L) or IFG (5.6–6.0 mmol/L) [ 83 ] . Diminished HRV has also been associated with overweight and obesity [82 ] , and with the metabolic syndrome, independently of fasting glucose levels [83 ] . Taken together, the population studies suggest that IGT rather than IFG is associated with cardiac autonomic dysfunction. In the lifestyle modi ﬁ cation arm of the Diabetes Prevention Program, parameters of autonomic function improved (i.e., heart rate decreased and HRV increased) in subjects with prediabetes. Improvements in these parameters were inversely associated with the development of diabetes, independently of weight change [85 ] . Furthermore, in the Steno memorial studies multiple risk factor reduction controlling for blood pressure, lipids, and glucose showed a 64% reduction in the development of cardiac autonomic dysfunction. It would seem fair, then, to reconsider the impact of glycemic control on the autonomic nervous system and label the abnormality as a loss of sympathetic/parasympathetic balance that is made worse as the continuum from IGT to diabetes progresses and is potentially reversible, as the pathogenesis clearly embraces more than just hyperglycemia per se [86 ] .

Pathogenesis of Neuropathy in Prediabetes

The pathogenesis of neuropathy in prediabetes is clearly multifactorial. There is a genetic predisposition, a long history of a prodrome of in fl ammation and oxidative/ nitrosative stress, and the subsequent intrusion of hyperglycemia, microvascular abnormalities, dyslipidemia, epigenetic phenomena, and miscellaneous factors associated with the metabolic syndrome [9, 86 ] . Hyperglycemia is directly neurotoxic by leading to increased oxidative stress, accumulation of advanced glycation endproducts, activation of protein kinase C and the polyol pathway, as well as other mechanisms [ 87, 88 ] . In prediabetes, hyperglycemia may be transient, predominantly in the postprandial phase. Experimental evidence in rats rendered acutely hyperglycemic with glucose infusion suggests that hyperglycemia promotes apoptosis in dorsal root ganglion neurons and Schwann cells [ 89] . In genetically diabetic mice, short-term hyperglycemic exposure has been shown to exert an unfavorable effect on mitochondrial function, which in the longer term results in mitochondrial fi ssion and shortened neuronal survival [90 ] . Similarly, in the diabetic BB/Wistar rat, transient hyperglycemia generates spontaneous impulses in nociceptive primary afferent neurons, and this activation leads to 8 Neuropathy in Prediabetes and the Metabolic Syndrome 127 neuropathic pain [ 91 ] . In humans, incubation of endothelial cells in high glucose induces production of reactive nitrogen and oxygen species, which, in turn, leads to single-strand DNA breakage, activation of poly (ADP-ribose) polymerase (PARP), and both metabolic and functional impairment [ 92 ] . Acute glycemic excursions in prediabetes trigger oxidative stress and related endothelial dysfunction, both established neurotoxic factors [93, 94 ] . Microvascular pathology has also been documented [ 95 ] . Increased capillary density was associated with current or future diabetes, and decreased capillary luminal area with future deterioration in glucose tolerance. Interestingly, total capillary basement membrane area was increased in patients with peripheral neuropathy, in comparison to those without (114.6 vs. 75.3 m m 2 , p < 0.0084) [ 95 ] . These fi ndings lend support to the hypothesis that endoneurial capillary microangiopathy presages impairment of glucose regulation (IGR) and is an early feature in the mechanism underlying DPN [ 95 ] . Using laser Doppler perfusion imaging (LDI) to measure vasodilation in the forearm skin in response to iontophoresis of acetylcholine, reduced endothelium- dependent vasodilation was demonstrated in subjects with prediabetes [ 96 ] . More recently, the axon re fl ex-elicited fl are areas on (LDI) were shown to be signi fi cantly reduced in subjects with prediabetes [97 ] . Such endothelial dysfunction is largely related to glycemic excursions. Oxidative stress induced by the latter results in both neurotoxicity and endothelial damage [93, 94 ] . The pivotal role of oxidative stress has been recently reaf fi rmed by experimental evidence [98, 99 ] . Furthermore, experimental data have shown that hyperlipidemia exerts a direct neurotoxic effect [100 ] . There is accumulating evidence that hyperlipidemia plays an important role in the pathogenesis of neuropathy in both type 1 [29 ] and type 2 diabetes [ 101 ] and should be addressed as a major therapeutic target [ 101 ] . Hughes et al. have reported that hypertriglyceridemia rather than hyperglycemia is a signi fi cant risk factor for CIAP [54 ] .

Neuropathy and the Metabolic Syndrome

There appears to be a role for miscellaneous factors associated with the metabolic syndrome (MS) in the development of neuropathy. In the initial reports of the loss of IENF in metabolic syndrome, the strongest association was with increasing levels of triglycerides and a low HDL-C. In the Australian natural history studies, the lowest prevalence of neuropathy occurred in people on lifetime use of a fi bric acid derivative which lowered triglycerides and raised HDL-C [102 ] . Other factors include waist circumference [ 26 ] , body weight [28 ] , and abdominal obesity [27 ] . The role of abdominal obesity may relate to the overproduction of in fl ammatory cytokines such as TNF-a and IL6 [ 16 ] . The KORA surveys [ 26– 28] have provided evidence for the association between obesity and neuropathy in the general population. In addition, small fi ber dysfunction (assessed by pain and vasodilatation refl ex responses) has been shown in subjects with morbid obesity, further strengthening 128 A.I. Vinik and M.-L. Nevoret this association [ 103 ] . There may be a genetic as well as an ethnic predisposition. Studies using Laser Doppler blood fl ow measurements have shown that family members of people with diabetes and African Americans have a greater predisposition to amputations and have signifi cantly impaired microvascular perfusion compared to Caucasians and healthy controls [ 104 ] . As mentioned above, patients with metabolic syndrome have signi fi cantly reduced IENF compared to healthy controls [10 ] . In support of the role of metabolic factors other than glycemia, 87 patients with idiopathic neuropathy had on average two additional features of MS [ 4 ] . Similarly, among 219 patients with idiopathic peripheral neuropathy, normoglycemic patients with idiopathic neuropathy had signifi cantly higher total and LDL-cholesterol and triglycerides, and lower HDL-cholesterol, than diabetic subjects without neuropathy [ 105] . Normoglycemic neuropathy subjects had signifi cantly more features of MS (other than hyperglycemia) than diabetic patients, suggesting an association between neuropathy and metabolic syndrome features other than hyperglycemia, in particular lipid abnormalities [105 ] . The association with MS is mainly attributable to obesity [ 4, 29, 105– 107] . Obesity may lead to increased levels of tumor necrosis factor alpha (TNF-a ) and circulating lipids (triglycerides and free fatty acids), which, in turn, may aggravate hyperglycemia by promoting hepatic gluconeogene- sis but are also likely to act independently on nerve function [100 ] . Eventually, both TNF- a and serum lipids increase oxidative stress and endothelial dysfunction, resulting in neurotoxicity [4, 106 ] . Growing evidence suggests that enhanced oxidative/nitrosative stress, in particular increased production of the potent oxidant peroxynitrite (a product of superoxide anion radicals with nitric oxide), is a characteristic feature of both experimental and clinical diabetes mellitus [ 108 ] . Peroxynitrite causes damage to a variety of tissues by diverse effects which include: nitration and nitrosylation of protein; damage to DNA; altered gene expression and changes in transcriptional regulation and signal transduction; altered mitochondrial function; and induction of microvascular endothelium necrosis in a variety of tissues [ 108– 110 ] including peripheral nerve, spinal cord, dorsal root ganglion neurons, and vasa nervorum of several different models of both type 1 and type 2 diabetes [ 111 ] . Several markers of oxidative stress in plasma including superoxide and peroxynitrite are elevated in diabetic patients with CAN [112 ] . These fi ndings suggest the presence of peroxynitrite cytotoxicity at both early and advanced stages of T1DM and T2DM, and furthermore, at the prediabetic stage. Enhanced nitrosative stress has also been documented in the circulation [113 ] and cutaneous microvasculature of human subjects with diabetes mellitus [ 114 ] . Monocyte nitrosylated protein expression is a new biomarker of metabolic control and in fl ammation in diabetic individuals with macroangiopathy and correlates with measures of in fl ammation such as CRP [115 ] . Moreover, COX-2 activation appears to play an important role in the development of CAN since COX-2 gene inactivation is protective against indices of CAN, oxidative stress, and infl ammation, and pre- vents LV dysfunction and myocardial fi brosis in experimental diabetes [116 ] . A more detailed assessment of diabetic microvascular complications and autonomic function is needed to determine if these variables can: (1) be employed as biomarkers 8 Neuropathy in Prediabetes and the Metabolic Syndrome 129 of the presence, severity, and progression of diabetic autonomic neuropathy and (2) account in part for the relatively low risk conferred by hyperglycemia alone (DCCT and EDIC) and the greater risk reduction for autonomic neuropathy with multiple risk factor reduction [117 ] . More recently, a causative role for chronic pain and depression has been suggested [118 ] . In patients with CIAP and IGT, the number of features of MS has been found to correlate with pain and depression scores, giving rise to the hypothesis that the latter might be common etiologic factors for both CIAP and MS [118 ] . In addition, while neuropathy and reduced ADLs are associated with depression, painful neuropathy is associated with anxiety and autonomic imbalance.

Conditions of Autonomic Imbalance Associated with Cardiovascular Risk

Metabolic Syndrome

In a study of healthy individuals, prolonged mild hyperinsulinemia was shown to disrupt the circadian rhythm of cardiac autonomic activity. Thus, the authors suggested that early changes in the neural control of cardiac activity may provide a potential mechanism mediating a pathophysiological link between IGT and cardio vascular disease (CVD) [119 ] . Individuals with the MS have alterations in the function of the autonomic nervous system (ANS), as increased activity of the sympathetic nervous system (SNS) is associated with several of the speciﬁ c MS components (e.g., obesity, hypertension, insulin resistance) [ 120, 121 ] . Unresolved, however, is whether the aberrations of the ANS contribute to the development of the MS or are a consequence of the MS. A recent study of 1,298 individuals with different numbers of metabolic abnormalities showed that altered cardiac autonomic function existed in individuals with one or two metabolic abnormalities. Furthermore, due to evidence of cardiac autonomic dysfunction but not insulin resistance in persons with one metabolic abnormality, the authors suggested that altered cardiac autonomic function precedes the presence of insulin resistance in the MS [ 122 ] . Prospective studies are needed, however, to answer this question with regard to the natural history. One prospective study of 433 nonobese, normotensive men followed for 5 years showed that autonomic dysfunction contributed to the development of obesity as sympathetic overactivity (i.e., plasma norepinephrine concentrations) and serum uric acid levels predicted future weight gain and elevation of blood pressure [ 123 ] . In individuals with a history of diabetes, it is well known that dysfunction of the ANS is a potential complication. Impaired autonomic function may, however, be a mechanism associated with early glucose dysmetabolism and thus autonomic dysfunction may be involved in the pathogenic pathway leading to the development of diabetes [124 ] . 130 A.I. Vinik and M.-L. Nevoret

Neuropathy in Prediabetes: Clinical Manifestations

Generally, subjects with prediabetes have less severe neuropathy than those with manifest diabetes. Sumner et al. [ 53 ] have reported less severe neuropathy in prediabetes as compared to diabetes. This was documented by differences in sural nerve amplitudes and nerve conduction velocities as well as distal leg intraepidermal nerve fi ber density (IENFD) [ 44 ] . Similarly, Singleton et al. [ 49 ] have shown that axonal injury based on nerve conduction studies was less pronounced in prediabetes. In terms of clinical manifestations, sensory modalities are more commonly affected than motor modalities [ 25, 49] . In terms of nerve fi ber type, small nerve fi bers are prominently affected and may be the earliest detectable sign. In the study by Sumner et al. [ 53 ] , patients with prediabetes had predominantly SFN, compared to diabetic patients, who had more involvement of large nerve fi bers. In the study by Singleton et al. [ 49] , again small fi ber involvement was predominant. A more recent study has confi rmed that prediabetic neuropathy mainly affects small fi bers, as manifested by impaired warm/ cold perception [125 ] . Small fi ber involvement is also documented by skin biopsy [ 53, 126 ] , and pain is a major symptom [ 25, 27, 28] . All subjects reported some degree of pain and mean pain visual analogue scale (VAS range = 0–100) was 36.4 ± 19.4 in one study [ 25 ] .

Distinguishing CAN from Autonomic Imbalance

The traditional view of CAN is that there is an early phase of loss of parasympathetic function with increased resting heart rate and abnormalities in the expiration:inspiration ratio of HRV. There may, however, be no parasympathetic denervation as such, but simply early augmentation of sympathetic tone. Early in the natural history of diabetes, there is impairment of parasympathetic function with a relative increase of sympathetic function causing an imbalance of the sympathetic/ parasympathetic tone. Later, sympathetic denervation follows beginning at the apex of the ventricles and progressing towards the base of the heart [ 3 ] , leading to yet another imbalance with an increased propensity for dysrhythmias. Analysis of HRV coupled with analysis of respiratory activity provides a noninvasive and objective method for assessing CAN and may be derived from ECG recordings [ 3 ] . Incorporating respiratory signal analysis enables one to independently measure each branch of the ANS. Spectral analysis of HRV is an important tool to evaluate CAN [3 ] . It decomposes a series of sequential R–R intervals into a set of sinusoidal waves. The power spectrum is displayed with the magnitude of variability as a function of frequency. The main frequency components are: the very low-frequency components (<0.04 Hz) related to ﬂ uctuations in vasomotor tone associated with thermoregulation; the low-frequency components (0.04–0.15 Hz) 8 Neuropathy in Prediabetes and the Metabolic Syndrome 131 associated with the baroreceptor reﬂ ex; and the high-frequency components (0.15–0.4 Hz) related to respiratory activity [3 ] . The cardiogram (from the EKG) only provides one number (HRV) for a two part system (i.e., parasympathetic and sympathetic). If one number (HRV) changes, one cannot tell which part (parasympathetic or sympathetic) changed. Respiratory analysis adds the second number, identifying the parasympathetic activity that generates respiratory sinus arrhyth- mia (RSA), thereby enabling RSA to be analyzed separately to identify parasympathetic activity (i.e., Rfa (respiration frequency area)). HRV, which is mixed parasympathetic and sympathetic, can now be separated (i.e., Rfa and Lfa (low- frequency area)).

Diagnosis of Neuropathy in Prediabetes

Diagnosis of neuropathy should follow standard guidelines for the assessment of diabetic neuropathy [12 ] . Clinical examination may be supplemented by nerve conduction studies if required [ 44, 49, 53 ] . At the same time, emphasis should be placed on the evaluation of small fi bers, for example by measurement of cold and warm perception thresholds [126 ] . Interestingly, Novella et al. [44 ] reported normal NCV in 12% of subjects, who were, however, found to exhibit SFN [ 40 ] . Skin biopsy has proven the most sensitive modality for the assessment of SFN [ 25, 31, 127 ] , but this cannot be advocated for routine use. Quantitative sudomotor axon refl ex test (QSART) has also been used, but its reproducibility was relatively poor in individuals with prediabetes [128 ] . Finally, CCM may prove useful. This is a novel, noninvasive method of examining human corneal nerve fi bers in vivo, enabling repeat evaluation over time [32 ] . It has high accuracy for the diagnosis of small fi ber impairment, as con fi rmed by skin biopsy [32 ] . CCM has also been shown to detect underlying pathology in 25 subjects with idiopathic SFN, including 8 subjects with IGT [129 ] . Nerve damage did not relate to metabolic abnormalities, but the number of IGT patients was very small [ 129 ] . Contact heat-evoked potentials have the capacity to detect changes recorded in the CNS upon stimulation using nociceptive heat in the periphery. CHEPS can potentially identify defective C fi ber and A delta fi ber function, with the added capacity of de fi ning nerve conduction, interpeak amplitudes, and the level of the lesion (from the peripheral nerve, the dorsal root ganglia, the spinal cord, or the CNS). Recent studies have shown its capacity to detect abnormalities in nerve function in metabolic syndrome and newly diagnosed diabetes before the advent of abnormalities in other measures of nerve function [130 ] . The Sudoscan is a newly developed means of quantifi cation of sweat generation using microcurrents. Early reports suggest that it is capable of detecting early peripheral autonomic dysfunction and may be able to provide an estimate of risk for development of diabetes and the metabolic syndrome [131 ] . 132 A.I. Vinik and M.-L. Nevoret

Towards Treatment of Neuropathy in Prediabetes

Little is known about effi cacious treatment of neuropathy in prediabetes. Given the aforementioned role of hyperglycemia, it is reasonable to achieve glycemic control as a medical priority. In the IGTN Study [25, 52] , 72 subjects with IGT and neuropathy were followed for up to 3 years. Neuropathy was examined at baseline and yearly with clinical exam, pain scales, NCS, QST, QSART, skin biopsy for IENFD, and LDI. Subjects received quarterly diet and exercise counseling to lower weight by 7% and increase weekly exercise to 150 min. OGTT, HbA 1c, lipid panels, exercise min- utes, weight, and BP were followed. Diet and exercise counseling signifi cantly reduced weight (p < 0.02) and LDL, increased weekly exercise after 1 year, and improved 2 h OGTT after 2 years (p < 0.02) [ 25, 52 ] . These were accompanied by a signifi cant ( p < 0.004) improvement in proximal IENFD of 1.3 ± 2.2 fi bers/mm and by a very modest improvement in (p < 0.12) distal IENFD of 0.3 ± 1.1 fi bers/mm [25 ] . Moreover, there was a very modest improvement in pain as measured by VAS from 36.4 ± 19.4 to 32.8 ± 26.3 (p < 0.4), a signifi cant (p < 0.05) increase in foot sweat volume by QSART (mL) of 0.3 ± 0.8 and a signi fi cant improvement in peroneal motor conduction velocity (p < 0.004) [ 25] . Metabolic improvements were sustained, but only lipid improvements remained signifi cant at 3 years. Neuropathy exam scales showed nonsignifi cant improvement, and measures of small fi ber nerve function (IENFD at ankle and thigh, QSART, LDI) improved signifi cantly over the fi rst year. Nonetheless, by 2 years neuropathy measures were not signifi cantly improved, and by 3 years there was a trend toward worsening compared to baseline. The Utah Early Neuropathy Scale (UENS), a sensitive examination measure of sensory nerves, was signi fi cantly worse by the year 3 [ 52 ] . The IGTN study suggests that improved metabolic abnormalities and weight loss achieved in IGTN subjects with diet and exercise result in transient improvement of neuropathy. However, counseling alone did not sustain signifi cant metabolic improvement, with consequent neuropathy progression over 3 years. The China Da Qing Diabetes Prevention Outcome Study examined the effect of a 6-year lifestyle intervention in 577 adults with IGT on the development of retinopathy, nephropathy, and neuropathy after 20 years. Follow-up examination was carried out in 542 (94%) of the 577 original participants [132 ] . In survivors, neuropathy was assessed by loss of sensation at one or more sites in the mono fi lament test [132 ] . Loss of sensation did not differ between the intervention and control groups (8.6% vs. 9.1%, respectively, p = 0.89) [ 132 ] . However, it must not escape our notice that there are important limitations. First, the low prevalence of neuropathy limited the statistical power to detect differences, as the authors themselves acknowledged [132 ] . Secondly, only follow-up but not baseline examination for neuropathy was performed, and so information was completely missing in the 25% who died during the study, leading to survival bias [132 ] . Finally, neuropathy was only evaluated by the monofi lament, which is a very crude test revealing severe neuropathy and high risk of foot ulceration, and can be less well accepted as a sensitive overall measure of neuropathy. Unfortunately, changes in NCV and any improvement in small fi ber function by QST could not be assessed in this study [132 ] . 8 Neuropathy in Prediabetes and the Metabolic Syndrome 133

It is also likely that treatment of hypertension and lipids, as well as smoking cessation, could contribute to improvement of nerve function in a multifactorial intervention, as in diabetes [133 ] , but this remains to be shown. Lifetime use of ﬁ bric acid derivatives has been associated with a lower incidence of neuropathy [ 102 ] . Interestingly, the use of an antiepileptic drug has been shown to relieve pain, improve features of the metabolic syndrome, and induce an increase in IENF [134, 135 ] .

Lifestyle Changes to Restore Balance

In the Diabetes Prevention Program, therapeutic lifestyle changes which included appropriate diet and exercise induced a 25% reduction in risk of autonomic dysfunction [85 ] . Endurance training improves HRV in patients with minimal abnormalities. However, one cannot use the normal calculations for heart rate of 220-age to calculate the maximum intensity and to derive a target for intensity of exercise because of the resting tachycardia in patients with autonomic dysfunction. Therefore, individuals must rely on use of perceived exertion to prescribe exercise intensity in diabetic autonomic neuropathy [136 ] . In T2DM patients chronic exercise is associated with enhanced cutaneous blood ﬂ ow [ 137 ] , restoration of baroreceptor sensitivity [138 ] , and improved vagal activity and exercise capacity after 12 weeks of endurance training in early CAN but not severe CAN [ 139 ] . These results emphasize the need for early aggressive intervention at the stage of physiological deﬁ cits in nerve function. Perhaps the most enlightening study was that of Motooka et al. [ 140 ] who showed that older people walking their pet animal improved sympathetic/parasympathetic balance; whereas the same exercise without the animal was quite stressful, increasing the prevalence of low-frequency amplitudes indicating loss of balance and sympathetic activation.

Medications to Restore Sympathovagal Balance

Disturbances in autonomic balance can arise from an increase or decrease in either sympathetic or parasympathetic tone. Each can be restored to normal using appropriate medications (see Vinik et al. [16 ] ). The use of alpha lipoic acid [ 141 ] , a powerful antioxidant, has been shown to improve HRV. The role for reactive oxygen species (ROS) in diabetes-associated nerve blood fl ow and conduction de fi cits was demonstrated in studies with the “universal” antioxidant dl - a-lipoic acid which combines free radical and metal-chelating properties with an ability (after conversion to dehydrolipoic acid) to regenerate levels of other antioxidants. Pharmacological tools targeting peroxynitrite formation or promoting its decomposition have recently been examined for their effects on autonomic neuropathy. Several peroxynitrite decomposition catalysts have 134 A.I. Vinik and M.-L. Nevoret been employed for preclinical studies of early peripheral diabetic neuropathy and autonomic neuropathy. The peroxynitrite decomposition catalysts, Fe(III)tetrakis-2- ( N -triethylene glycol monomethyl ether)-pyridyl porphyrin (FP15) and Fe(III) tetra-mesitylporphyrin octasulfonate (FeTMPS), improve nerve function in STZ- diabetic rats and STZ-diabetic and ob / ob mice. The bene fi cial effects of the protein nitration inhibitor epicatechin have also been reported. Multiple risk factor reduction has been shown to lower the hazard ratio for autonomic neuropathy by 63% [ 117 ] . Which elements were important in the Steno Type 2 study have yet to be determined. Even the maldistribution of cardiac sympathetic innervation can be restored with excellent diabetes control [3, 142 ] . An increase in HRV has also been described with the use of aldose reductase inhibitors (ARIs), C-peptide, angiotensin converting enzyme (ACE)-inhibitors (qui- napril, ramipril, perindopril), angiotensin II reception blocker (ARB), b Blocker (meto- prolol), digoxin, and verapamil [3 ] . Autonomic imbalance and high postinfarction morbidity and mortality are frequently observed in diabetic patients. Because the reduction in both recurrent myocardial infarction and mortality in postinfarction patients treated with b blockers without intrinsic sympathetic activity (ISA) was higher in diabetic than nondiabetic subjects, it has been suggested that this high-risk group could particularly bene fi t from these agents. To date, no results are available for advanced glycation end product inhibitors (AGEIs), statins, carnitine, peroxisome proliferator-activated receptor activators (PPAR), protein kinase C-b inhibitors, and anti-in fl ammatory agents, and their effect on HRV. These agents are therefore some of the suggested targets for future attention. A paradigm of pharmaceutical approaches to reducing infl ammation and oxidative/nitrosative stress is illustrated in Fig. 8.2 modi fi ed from that previously published [ 114 ] .

Conclusions

There is now, despite some negative reports, emerging evidence for an association between neuropathy and IGT, rather than IFG. This relationship is bidirectional: a substantial proportion of patients with idiopathic neuropathy (24.6–62%) have prediabetes (mainly IGT) [ 44, 46, 48– 50, 53] , and a substantial proportion of subjects with prediabetes (mainly IGT) exhibit peripheral neuropathy (11.2–24.3%) or neuropathic pain (12.9–20.5%) [ 26– 28, 45] . Nonetheless, important caveats must be borne in mind. Indeed, clinic-based studies included selected patients referred to specialized centers, resulting in selection bias [ 44, 46, 48– 50, 53– 56 ] . Moreover, these studies included small numbers of patients, and OGTT was usually performed only in a subset of these [ 44, 46, 48– 50, 53– 56] . Of greater importance, most studies [ 44, 46, 48, 49, 53 ] relied on the comparison between observed frequency of prediabetes and that reported in the general population, as based on historical control studies [ 56, 57 ] . The absence of control groups in these studies seriously undermines their level of evidence, as already suggested [ 58, 59] . In this respect, population-based studies provide 8 Neuropathy in Prediabetes and the Metabolic Syndrome 135

Fig. 8.2 This fi gure suggests pharmaceutical approaches to the treatment or prevention of autonomic dysfunction in diabetes. Central to this evolving concept is the role of adipocytokines and in fl ammation. Adapted with permission from Vinik et al. [114 ] stronger evidence [26– 28, 60– 62 ] . The MONICA/KORA surveys [26– 28 ] showed that a substantial portion of subjects with prediabetes exhibited painful neuropathy [ 26– 28 ] . Other population studies [60– 62 ] also provide evidence of incipient neuropathy in prediabetes. There are negative studies [ 63– 65] , but these have their limitations as well. Indeed, some of them have included very speci fi c populations which may be different from the Caucasian subjects included in most studies [ 63, 64 ] . One study [64 ] assessed only VPT, which may not be an adequate measure of overall neuropathy, especially small fi ber function. The same study [64 ] did not aim to examine neuropathy in prediabetes. In one study, no NGT population was included [65 ] . The most sensitive test to evaluate glucose homeostasis is the OGTT [ 26– 28, 50 ] . Many underlying mechanisms appear to be implicated in the pathogenesis of neuropathy in prediabetes: hyperglycemia [93, 94] , microvascular abnormalities [ 96, 97] , dyslipidemia [ 100, 101] , and miscellaneous factors associated with the metabolic syndrome [ 26– 28, 105, 107 ] . Postchallenge hyperglycemia could represent a major mechanism [ 93, 94] . In general, subjects with prediabetes have less severe neuropathy than those with manifest diabetes [ 49, 53 ] . Sensory modalities are more commonly affected than motor modalities [ 25, 44] , while small nerve fi bers may be prominently affected [49, 53, 125 ] . Pain is frequently a major symptom, but the neuropathic process may also remain asymptomatic [25, 27, 28 ] . Diagnosis should rely 136 A.I. Vinik and M.-L. Nevoret on careful clinical examination, with emphasis on the evaluation of small fi bers [125 ] . Treatment should mainly aim at optimizing glycemic control [25 ] . This may contribute to a transient improvement in nerve function [25, 52 ] , but its long-term effi cacy is questionable [ 52, 132] and needs further study. Multiple risk factor reduction including weight management, exercise, cessation of smoking, and control of blood pressure and lipids may be salutary.

References

1. Vinik EJ, Hayes RP, Oglesby A, Bastyr E, Barlow P, Ford-Molvik SL, et al. The development and validation of the Norfolk QOL-DN, a new measure of patients’ perception of the effects of diabetes and diabetic neuropathy. Diabetes Technol Ther. 2005;7:497–508. 2. Boulton A, Vinik A, Arezzo J, Bril V, Feldman E, Freeman R, et al. Position statement: diabetic neuropathies. Diabetes Care. 2005;28(4):956–62. 3. Vinik AI, Ziegler D. Diabetic cardiovascular autonomic neuropathy. Circulation. 2007;115:387–97. 4. Smith AG, Singleton JR. Idiopathic neuropathy, prediabetes and the metabolic syndrome (Abstract). J Neurol Sci. 2006;242:9–14. 5. Rota E, Quadri R, Fanti E, Isoardo G, Poglio F, Tavella A, et al. Electrophysiological fi ndings of peripheral neuropathy in newly diagnosed type II diabetes mellitus. J Peripher Nerv Syst. 2005;10:348–53. 6. Pittenger G, Mehrabyan A, Simmons K, Rice A, Dublin C, Barlow P, et al. Small fi ber neuropathy is associated with the metabolic syndrome. Metab Syndr Relat Disord. 2005;3:113–21. 7. Dyck P, Clark V, Overland C, et al. Does impaired glycemia cause polyneuropathy and other diabetic complications? (Abstract). Peripher Nerve Soc. 2011;16(Suppl):S30. 8. Papanas N, Vinik AI, Ziegler D. Neuropathy in prediabettes: does the clock start ticking early? Nat Rev Endocrinol. 2011;7:682–90. 9. Vinik A, Ullal J, Parson HK, Casellini CM. Diabetic neuropathies: clinical manifestations and current treatment options. Nat Clin Pract Endocrinol Metab. 2006;2:269–81. 10. Pittenger GL, Ray M, Burcus NI, McNulty P, Basta B, Vinik AI. Intraepidermal nerve fi bers are indicators of small-fi ber neuropathy in both diabetic and nondiabetic patients. Diabetes Care. 2004;27:1974–9. 11. Sadosky A, McDermott AM, Brandenburg NA, Strauss M. A review of the epidemiology of painful diabetic peripheral neuropathy, postherpetic neuralgia, and less commonly studied neuropathic pain conditions. Pain Pract. 2008;8:45–56. 12. Tesfaye S, Boulton AJ, Dyck PJ, Freeman R, Horowitz M, Kempler P, et al. Diabetic neuropathies: update on defi nitions, diagnostic criteria, estimation of severity, and treatments. Diabetes Care. 2010;33:2285–93. 13. Ziegler D. Painful diabetic neuropathy: treatment and future aspects. Diabetes Metab Res Rev. 2008;24 Suppl 1:S52–7. 14. Boulton AJ, Malik RA, Arezzo JC, Sosenko JM. Diabetic somatic neuropathies. Diabetes Care. 2004;27:1458–86. 15. Vinik AI, Maser RE, Ziegler D. Neuropathy: the crystal ball for cardiovascular disease? Diabetes Care. 2010;33:1688–90. 16. Vinik A, Maser R, Ziegler D. Autonomic imbalance: prophet of doom or scope for hope? Diabet Med. 2011;28:643–51. 17. Mantyselka P, Ahonen R, Kumpusalo E, Takala J. Variability in prescribing for musculoskeletal pain in Finnish primary health care. Pharm World Sci. 2001;23:232–6. 18. Treede RD, Jensen TS, Campbell JN, Cruccu G, Dostrovsky JO, Grif fi n JW, et al. Neuropathic pain: rede fi nition and a grading system for clinical and research purposes. Neurology. 2008;70:1630–5. 8 Neuropathy in Prediabetes and the Metabolic Syndrome 137

19. Jensen MP, Chodroff MJ, Dworkin RH. The impact of neuropathic pain on health-related quality of life: review and implications. Neurology. 2007;68:1178–82. 20. Bouhassira D, Lanteri-Minet M, Attal N, Laurent B, Touboul C. Prevalence of chronic pain with neuropathic characteristics in the general population. Pain. 2008;136:380–7. 21. Vinik E, Paulson J, Ford-Molvik S, Vinik A. German-Translated Norfolk Quality of Life (QOL-DN) identifi es the same factors as the English version of the tool and discriminates different levels of neuropathy severity. J Diabetes Sci Technol. 2008;2:1075–86. 22. Torrance N, Smith BH, Bennett MI, Lee AJ. The epidemiology of chronic pain of predominantly neuropathic origin. Results from a general population survey. J Pain. 2006;7:281–9. 23. Dieleman JP, Kerklaan J, Huygen FJ, Bouma PA, Sturkenboom MC. Incidence rates and treatment of neuropathic pain conditions in the general population. Pain. 2008;137:681–8. 24. Cornblath DR, Vinik A, Feldman E, Freeman R, Boulton AJ. Surgical decompression for diabetic sensorimotor polyneuropathy. Diabetes Care. 2007;30:421–2. 25. Smith AG, Russell J, Feldman EL, Goldstein J, Peltier A, Smith S, et al. Lifestyle intervention for pre-diabetic neuropathy. Diabetes Care. 2006;29:1294–9. 26. Ziegler D, Rathmann W, Dickhaus T, Meisinger C, Mielck A. Prevalence of polyneuropathy in pre-diabetes and diabetes is associated with abdominal obesity and macroangiopathy: the MONICA/KORA Augsburg Surveys S2 and S3. Diabetes Care. 2008;31:464–9. 27. Ziegler D, Rathmann W, Meisinger C, Dickhaus T, Mielck A. Prevalence and risk factors of neuropathic pain in survivors of myocardial infarction with pre-diabetes and diabetes. The KORA Myocardial Infarction Registry. Eur J Pain. 2009;13:582–7. 28. Ziegler D, Rathmann W, Dickhaus T, Meisinger C, Mielck A. Neuropathic pain in diabetes, prediabetes and normal glucose tolerance: the MONICA/KORA Augsburg Surveys S2 and S3. Pain Med. 2009;10:393–400. 29. Tesfaye S, Chaturvedi N, Eaton SE, Ward JD, Manes C, Ionescu-Tirgoviste C, et al. Vascular risk factors and diabetic neuropathy. N Engl J Med. 2005;352:341–50. 30. Lauria G, Hsieh S, Johansson O, Kennedy W, Leger J, Mellgren S, et al. European Federation of Neurological Societies/Peripheral Nerve Society Guideline on the Use of Skin Biopsy in the Diagnosis of Small Fiber Neuropathy. Report of a joint task force of the European Federation of Neurological Societies and the Peripheral Nerve Society. Eur J Neurol. 2010;17:903–12. 31. Lauria G, Bakkers M, Schmitz C, Lombardi R, Penza P, Devigili G, et al. Intraepidermal nerve fi ber density at the distal leg: a worldwide normative reference study. J Peripher Nerv Syst. 2010;15:202–7. 32. Quattrini C, Tavakoli M, Jeziorska M, Kallinikos P, Tesfaye S, Finnigan J, et al. Surrogate markers of small fi ber damage in human diabetic neuropathy. Diabetes. 2007;56:2148–54. 33. Tavakoli M, Quattrini C, Abbott C, Kallinikos P, Marshall A, Finnigan J, et al. Corneal confocal microscopy: a novel noninvasive test to diagnose and stratify the severity of human diabetic neuropathy. Diabetes Care. 2010;33:1792–7. 34. Zhivov A, Blum M, Guthoff R, Stachs O. Real-time mapping of the subepithelial nerve plexus by in vivo confocal laser scanning microscopy. Br J Ophthalmol. 2010;94:1133–5. 35. Parson HK, Nguyen VT, Boyd AL, Vinik A. CHEPS detects neuropathic changes earlier than traditional clinical measures. Diabetes. 2009;58(Suppl):829P. 36. Expert Committee on the Diagnosis and Classi fi cation of Diabete Mellitus. Report of the expert committee on the diagnosis and classi fi cation of diabetes mellitus (Abstract). Diabetes Care. 2003;26:S5–20. 37. DECODE Study Group, European Diabetes Epidemiology Group. Glucose tolerance and cardiovascular mortality: comparison of fasting and 2-hour diagnostic criteria. Arch Intern Med. 2001;161:397–405. 38. Qiao Q, Pyorala K, Pyorala M, Nissinen A, Lindstrom J, Tilvis R, et al. Two-hour glucose is a better risk predictor for incident coronary heart disease and cardiovascular mortality than fasting glucose. Eur Heart J. 2002;23:1267–75. 39. Hyvarinen M, Qiao Q, Tuomilehto J, Laatikainen T, Heine RJ, Stehouwer CD, et al. Hyperglycemia and stroke mortality: comparison between fasting and 2-h glucose criteria. Diabetes Care. 2009;32:348–54. 138 A.I. Vinik and M.-L. Nevoret

40. Ning F, Tuomilehto J, Pyorala K, Onat A, Soderberg S, Qiao Q. Cardiovascular disease mortality in Europeans in relation to fasting and 2-h plasma glucose levels within a normoglycemic range. Diabetes Care. 2010;33:2211–6. 41. Colagiuri S, Lee CM, Wong TY, Balkau B, Shaw JE, Borch-Johnsen K. Glycemic thresholds for diabetes-specifi c retinopathy: implications for diagnostic criteria for diabetes. Diabetes Care. 2011;34:145–50. 42. Sosenko JM, Kato M, Goldberg RB. Sensory function and albumin excretion according to diagnostic criteria for diabetes. Diabetes Care. 2004;27:1716–20. 43. Dyck PJ, Dyck PJ, Klein CJ, Weigand SD. Does impaired glucose metabolism cause polyneuropathy? Review of previous studies and design of a prospective controlled population- based study. Muscle Nerve. 2007;36:536–41. 44. Novella SP, Inzucchi SE, Goldstein JM. The frequency of undiagnosed diabetes and impaired glucose tolerance in patients with idiopathic sensory neuropathy. Muscle Nerve. 2001;24: 1229–31. 45. Franklin GM, Kahn LB, Baxter J, Marshall JA, Hamman RF. Sensory neuropathy in non- insulin-dependent diabetes mellitus. The San Luis Valley Diabetes Study. Am J Epidemiol. 1990;131:633–43. 46. Singleton JR, Smith AG, Bromberg MB. Painful sensory polyneuropathy associated with impaired glucose tolerance. Muscle Nerve. 2001;24:1225–8. 47. Sumner CJ, Sheth S, Griffi n JW, Cornblath DR, Polydefkis M. The spectrum of neuropathy in diabetes and impaired glucose tolerance. Neurology. 2003;60:108–11. 48. Smith AG, Singleton JR. The diagnostic yield of a standardized approach to idiopathic sensory- predominant neuropathy. Arch Intern Med. 2004;164:1021–5. 49. Singleton JR, Smith AG, Bromberg MB. Increased prevalence of impaired glucose tolerance in patients with painful sensory neuropathy. Diabetes Care. 2001;24:1448–53. 50. Hoffman-Snyder C, Smith BE, Ross MA, Hernandez J, Bosch EP. Value of the oral glucose tolerance test in the evaluation of chronic idiopathic axonal polyneuropathy. Arch Neurol. 2006;63:1075–9. 51. Expert Committee on the Diagnosis and Classifi cation of Diabetes Mellitus. Report of the expert committee on the diagnosis and classi fi cation of diabetes mellitus (Abstract). Diabetes Care. 1997;20:1183–97. 52. Singleton JR et al. Diet and exercise counseling alone does not prevent long term neuropathy progression in IGTN (Abstract). American Academy of Neurology 59th Annual Meeting. May 4, 2007, Boston. 53. Sumner C, Sheth S, Griffi n J, Cornblath D, Polydefkis M. The spectrum of neuropathy in diabetes and impaired glucose tolerance. Neurology. 2003;60:108–11. 54. Hughes RA, Umapathi T, Gray IA, Gregson NA, Noori M, Pannala AS, et al. A controlled investigation of the cause of chronic idiopathic axonal polyneuropathy. Brain. 2004;127: 1723–30. 55. Nebuchennykh M, Loseth S, Jorde R, Mellgren SI. Idiopathic polyneuropathy and impaired glucose metabolism in a Norwegian patient series. Eur J Neurol. 2008;15:810–6. 56. Harris MI, Flegal KM, Cowie CC, Eberhardt MS, Goldstein DE, Little RR, et al. Prevalence of diabetes, impaired fasting glucose, and impaired glucose tolerance in U.S. adults. The Third National Health and Nutrition Examination Survey, 1988–1994. Diabetes Care. 1998;21:518–24. 57. Centers for Disease Control and Prevention (CDC). Prevalence of diabetes and impaired fasting glucose in adults, United States, 1999––2000. MMWR Morb Mortal Wkly Rep. 2003;52: 833–7. 58. Russell JW, Feldman EL. Impaired glucose tolerance-does it cause neuropathy? Muscle Nerve. 2001;24(9):1109–12. 59. Kissel JT. Peripheral neuropathy with impaired glucose tolerance: a sweet smell of success? Arch Neurol. 2006;63:1055–6. 60. de Neeling JN, Beks PJ, Bertelsmann FW, Heine RJ, Bouter LM. Peripheral somatic nerve function in relation to glucose tolerance in an elderly Caucasian population: the Hoorn study. Diabet Med. 1996;13:960–6. 8 Neuropathy in Prediabetes and the Metabolic Syndrome 139

61. Grandinetti A, Chow DC, Sletten DM, Oyama JK, Theriault AG, Schatz IJ, et al. Impaired glucose tolerance is associated with postganglionic sudomotor impairment. Clin Auton Res. 2007;17:231–3. 62. Isak B, Oﬂ azoglu B, Tanridag T, Yitmen I, Us O. Evaluation of peripheral and autonomic neuropathy among patients with newly diagnosed impaired glucose tolerance. Diabetes Metab Res Rev. 2008;24:563–9. 63. Fujimoto WY, Leonetti DL, Kinyoun JL, Shuman WP, Stolov WC, Wahl PW. Prevalence of complications among second-generation Japanese-American men with diabetes, impaired glucose tolerance, or normal glucose tolerance. Diabetes. 1987;36:730–9. 64. Shaw JE, Hodge AM, de Courten M, Dowse GK, Gareeboo H, Tuomilehto J, et al. Diabetic neuropathy in Mauritius: prevalence and risk factors. Diabetes Res Clin Pract. 1998;42:131–9. 65. Barr EL, Wong TY, Tapp RJ, Harper CA, Zimmet PZ, Atkins R, et al. Is peripheral neuropathy associated with retinopathy and albuminuria in individuals with impaired glucose metabolism? The 1999–2000 AusDiab. Diabetes Care. 2006;29:1114–6. 66. Eriksson KF, Nilsson H, Lindgarde F, Osterlin S, Dahlin LB, Lilja B, et al. Diabetes mellitus but not impaired glucose tolerance is associated with dysfunction in peripheral nerves. Diabet Med. 1994;11:279–85. 67. Mantyselka P, Miettola J, Niskanen L, Kumpusalo E. Glucose regulation and chronic pain at multiple sites. Rheumatology (Oxford). 2008;47:1235–8. 68. Mantyselka P, Miettola J, Niskanen L, Kumpusalo E. Chronic pain, impaired glucose tolerance and diabetes: a community-based study. Pain. 2008;137:34–40. 69. Feldman EL, Stevens MJ, Thomas PK, Brown MB, Canal N, Greene DA. A practical two- step quantitative clinical and electrophysiological assessment for the diagnosis and staging of diabetic neuropathy. Diabetes Care. 1994;17:1281–9. 70. Moghtaderi A, Bakhshipour A, Rashidi H. Validation of Michigan neuropathy screening instrument for diabetic peripheral neuropathy. Clin Neurol Neurosurg. 2006;108:477–81. 71. Vinik AI, Ullal J, Parson HK, Barlow PM, Casellini CM. Pioglitazone treatment improves nitrosative stress in type 2 diabetes. Diabetes Care. 2006;29:869–76. 72. Vinik AI, Maser RE, Mitchell BD, Freeman R. Diabetic autonomic neuropathy. Diabetes Care. 2003;26:1553–79. 73. Maser RE, Mitchell BD, Vinik AI, Freeman R. The association between cardiovascular autonomic neuropathy and mortality in individuals with diabetes: a meta-analysis. Diabetes Care. 2003;26:1895–901. 74. Ziegler D, Zentai CP, Perz S, Rathmann W, Haastert B, Doring A, et al. Prediction of mortality using measures of cardiac autonomic dysfunction in the diabetic and nondiabetic population: the MONICA/KORA Augsburg Cohort Study. Diabetes Care. 2008;31:556–61. 75. Task Force of the European Society of Cardiology, North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation and clinical use. Circulation. 1996;93:1043–65. 76. Beijers HJ, Ferreira I, Bravenboer B, Dekker JM, Nijpels G, Heine RJ, et al. Microalbuminuria and cardiovascular autonomic dysfunction are independently associated with cardiovascular mortality: evidence for distinct pathways: the Hoorn Study. Diabetes Care. 2009;32:1698–703. 77. Diakakis GF, Parthenakis FI, Patrianakos AP, Koukouraki SI, Stathaki MI, Karkavitsas NS, et al. Myocardial sympathetic innervation in patients with impaired glucose tolerance: relationship to subclinical in ﬂ ammation. Cardiovasc Pathol. 2008;17:172–7. 78. Gerritsen J, Dekker JM, TenVoorde BJ, Bertelsmann FW, Kostense PJ, Stehouwer CD, et al. Glucose tolerance and other determinants of cardiovascular autonomic function: the Hoorn Study. Diabetologia. 2000;43:561–70. 79. Singh JP, Larson MG, O’Donnell CJ, Wilson PF, Tsuji H, Lloyd-Jones DM, et al. Association of hyperglycemia with reduced heart rate variability (The Framingham Heart Study). Am J Cardiol. 2000;86:309–12. 80. Schroeder EB, Chambless LE, Liao D, Prineas RJ, Evans GW, Rosamond WD, et al. Diabetes, glucose, insulin, and heart rate variability: the atherosclerosis risk in communities (ARIC) study. Diabetes Care. 2005;28:668–74. 140 A.I. Vinik and M.-L. Nevoret

81. Wu JS, Yang YC, Lin TS, Huang YH, Chen JJ, Lu FH, et al. Epidemiological evidence of altered cardiac autonomic function in subjects with impaired glucose tolerance but not isolated impaired fasting glucose. J Clin Endocrinol Metab. 2007;92:3885–9. 82. Wu JS, et al. Epidemiological evidence of altered cardiac autonomic function in overweight but not underweight subjects. Int J Obes (Lond). 2008;32:788–95. 83. Stein PK, Barzilay JI, Domitrovich PP, Chaves PM, Gottdiener JS, Heckbert SR, et al. The relationship of heart rate and heart rate variability to non-diabetic fasting glucose levels and the metabolic syndrome: the Cardiovascular Health Study. Diabet Med. 2007;24:855–63. 84. Perciaccante A, Fiorentini A, Paris A, Serra P, Tubani L. Circadian rhythm of the autonomic nervous system in insulin resistant subjects with normoglycemia, impaired fasting glycemia, impaired glucose tolerance, type 2 diabetes mellitus. BMC Cardiovasc Disord. 2006;6:19. 85. Carnethon MR, Prineas RJ, Temprosa M, Zhang ZM, Uwaifo G, Molitch ME. The association among autonomic nervous system function, incident diabetes, and intervention arm in the Diabetes Prevention Program. Diabetes Care. 2006;29:914–9. 86. Vinik AI, Strotmeyer ES, Nakave AA, Patel CV. Diabetic neuropathy in older adults. Clin Geriatr Med. 2008;24:407–35; v. 87. Sugimoto K, Murakawa Y, Sima AA. Diabetic neuropathy—a continuing enigma. Diabetes Metab Res Rev. 2000;16:408–33. 88. Dobretsov M, Romanovsky D, Stimers JR. Early diabetic neuropathy: triggers and mechanisms. World J Gastroenterol. 2007;13:175–91. 89. Russell JW, Sullivan KA, Windebank AJ, Herrmann DN, Feldman EL. Neurons undergo apoptosis in animal and cell culture models of diabetes. Neurobiol Dis. 1999;6:347–63. 90. Edwards JL, et al. Diabetes regulates mitochondrial biogenesis and fi ssion in mouse neurons. Diabetologia. 2010;53:160–9. 91. Burchiel KJ, Russell LC, Lee RP, Sima AA. Spontaneous activity of primary afferent neurons in diabetic BB/Wistar rats. A possible mechanism of chronic diabetic neuropathic pain. Diabetes. 1985;34:1210–3. 92. Garcia SF. Diabetic endothelial dysfunction: the role of poly(ADP-ribose) polymerase activation. Nat Med. 2001;7:108–13. 93. Heine RJ, Balkau B, Ceriello A, Del PS, Horton ES, Taskinen MR. What does postprandial hyperglycaemia mean? Diabet Med. 2004;21:208–13. 94. Su Y, Liu XM, Sun YM, Jin HB, Fu R, Wang YY, et al. The relationship between endothelial dysfunction and oxidative stress in diabetes and prediabetes. Int J Clin Pract. 2008;62:877–82. 95. Thrainsdottir S, Malik RA, Dahlin LB, Wiksell P, Eriksson KF, Rosen I, et al. Endoneurial capillary abnormalities presage deterioration of glucose tolerance and accompany peripheral neuropathy in man. Diabetes. 2003;52:2615–22. 96. Caballero AE, Arora S, Saouaf R, Lim SC, Smakowski P, Park JY, et al. Microvascular and macrovascular reactivity is reduced in subjects at risk for type 2 diabetes. Diabetes. 1999;48:1856–62. 97. Green AQ, Krishnan S, Finucane FM, Rayman G. Altered C- fi ber function as an indicator of early peripheral neuropathy in individuals with impaired glucose tolerance. Diabetes Care. 2010;33:174–6. 98. Watcho P, Stavniichuk R, Ribnicky DM, Raskin I, Obrosova IG. High-fat diet-induced neuropathy of prediabetes and obesity: effect of PMI-5011, an ethanolic extract of Artemisia dracunculus L. Mediators In fl amm. 2010;2010:268547. 99. Stavniichuk R, Drel VR, Shevalye H, Vareniuk I, Stevens MJ, Nadler JL, et al. Role of 12/15-lipoxygenase in nitrosative stress and peripheral prediabetic and diabetic neuropathies. Free Radic Biol Med. 2010;49:1036–45. 100. Vincent AM, Hayes JM, McLean LL, Vivekanandan-Giri A, Pennathur S, Feldman EL. Dyslipidemia-induced neuropathy in mice: the role of oxLDL/LOX-1. Diabetes. 2009;58: 2376–85. 101. Vincent AM, Hinder LM, Pop-Busui R, Feldman EL. Hyperlipidemia: a new therapeutic target for diabetic neuropathy. J Peripher Nerv Syst. 2009;14:257–67. 102. Davis TM, Yeap BB, Davis WA, Bruce DG. Lipid-lowering therapy and peripheral sensory neuropathy in type 2 diabetes: the Fremantle Diabetes Study. Diabetologia. 2008;51:562–6. 8 Neuropathy in Prediabetes and the Metabolic Syndrome 141

103. Herman RM, Brower JB, Stoddard DG, Casano AR, Targovnik JH, Herman JH, et al. Prevalence of somatic small ﬁ ber neuropathy in obesity. Int J Obes (Lond). 2007;31:226–35. 104. Parson H, Bridge J, Dublin C, Ullal J, Vinik A. African-Americans exhibit differences in neurovascular and endothelial dysfunction when compared to Caucasians. Diabetes. 2005;54:A220. 105. Smith AG, Rose K, Singleton JR. Idiopathic neuropathy patients are at high risk for metabolic syndrome. J Neurol Sci. 2008;273:25–8. 106. Smith AG, Singleton JR. Impaired glucose tolerance and neuropathy. Neurologist. 2008;14: 23–9. 107. Singleton JR, Smith AG, Russell JW, Feldman EL. Microvascular complications of impaired glucose tolerance. Diabetes. 2003;52:2867–73. 108. Pacher P, Beckman JS, Liaudet L. Nitric oxide and peroxynitrite in health and disease. Physiol Rev. 2007;87:315–424. 109. Szabo C, Mabley JG, Moeller SM, Shimanovich R, Pacher P, Virag L, et al. Part I: pathoge- netic role of peroxynitrite in the development of diabetes and diabetic vascular complications: studies with FP15, a novel potent peroxynitrite decomposition catalyst. Mol Med. 2002;8:571–80. 110. Virag L, Szabo E, Gergely P, Szabo C. Peroxynitrite-induced cytotoxicity: mechanism and opportunities for intervention. Toxicol Lett. 2003;140–141:113–24. 111. Obrosova IG, Drel VR, Oltman CL, Mashtalir N, Tibrewala J, Groves JT, et al. Role of nitrosative stress in early neuropathy and vascular dysfunction in streptozotocin-diabetic rats. Am J Physiol Endocrinol Metab. 2007;293:E1645–55. 112. Ziegler D, Sohr CG, Nourooz-Zadeh J. Oxidative stress and antioxidant defense in relation to the severity of diabetic polyneuropathy and cardiovascular autonomic neuropathy. Diabetes Care. 2004;27:2178–83. 113. Ceriello A, Assaloni R, Da RR, Maier A, Quagliaro L, Piconi L, et al. Effect of irbesartan on nitrotyrosine generation in non-hypertensive diabetic patients. Diabetologia. 2004;47:1535–40. 114. Vinik A, Parson H, Ullal J. The role of PPARs in the microvascular dysfunction in diabetes. Vascul Pharmacol. 2006;45:54–64. 115. Julius U, Drel VR, Grassler J, Obrosova IG. Nitrosylated proteins in monocytes as a new marker of oxidative-nitrosative stress in diabetic subjects with macroangiopathy. Exp Clin Endocrinol Diabetes. 2009;117:72–7. 116. Kellogg AP, Converso K, Wiggin T, Stevens M, Pop-Busui R. Effects of cyclooxygenase-2 gene inactivation on cardiac autonomic and left ventricular function in experimental diabetes. Am J Physiol Heart Circ Physiol. 2009;296:H453–61. 117. Gaede P, Vedel P, Larsen N, Jensen GV, Parving HH, Pedersen O. Multifactorial intervention and cardiovascular disease in patients with type 2 diabetes. N Engl J Med. 2003;348:383–93. 118. Rezania K, Soliven B, Rezai KA, Roos RP. Impaired glucose tolerance and metabolic syndrome in idiopathic polyneuropathy: the role of pain and depression. Med Hypotheses. 2011;76:538–42. 119. Petrova M, Townsend R, Teff KL. Prolonged (48-hour) modest hyperinsulinemia decreases nocturnal heart rate variability and attenuates the nocturnal decrease in blood pressure in lean, normotensive humans. J Clin Endocrinol Metab. 2006;91:851–9. 120. Ziegler D, Zentai C, Perz S, Rathmann W, Haastert B, Meisinger C, et al. Selective contribution of diabetes and other cardiovascular risk factors to cardiac autonomic dysfunction in the general population. Exp Clin Endocrinol Diabetes. 2006;114:153–9. 121. Licht CM, Vreeburg SA, van Reedt Dortland AK, Giltay EJ, Hoogendijk WJ, Derijk RH, et al. Increased sympathetic and decreased parasympathetic activity rather than changes in hypothalamic-pituitary-adrenal axis activity is associated with metabolic abnormalities. J Clin Endocrinol Metab. 2010;95:2458–66. 122. Chang CJ, Yang YC, Lu FH, Lin TS, Chen JJ, Yeh TL, et al. Altered cardiac autonomic function may precede insulin resistance in metabolic syndrome. Am J Med. 2010;123:432–8. 123. Masuo K, Kawaguchi H, Mikami H, Ogihara T, Tuck ML. Serum uric acid and plasma norepinephrine concentrations predict subsequent weight gain and blood pressure elevation. Hypertension. 2003;42:474–80. 142 A.I. Vinik and M.-L. Nevoret

124. Carnethon MR, Jacobs Jr DR, Sidney S, Liu K. Infl uence of autonomic nervous system dysfunction on the development of type 2 diabetes: the CARDIA study. Diabetes Care. 2003;26:3035–41. 125. Putz Z, Tabak AG, Toth N, Istenes I, Nemeth N, Gandhi RA, et al. Noninvasive evaluation of neural impairment in subjects with impaired glucose tolerance. Diabetes Care. 2009;32: 181–3. 126. Smith AG, Ramachandran P, Tripp S, Singleton JR. Epidermal nerve innervation in impaired glucose tolerance and diabetes-associated neuropathy. Neurology. 2001;57:1701–4. 127. Hays AP. Utility of skin biopsy to evaluate peripheral neuropathy. Curr Neurol Neurosci Rep. 2010;10:101–7. 128. Peltier A, Smith AG, Russell JW, Sheikh K, Bixby B, Howard J, et al. Reliability of quantitative sudomotor axon re fl ex testing and quantitative sensory testing in neuropathy of impaired glucose regulation. Muscle Nerve. 2009;39:529–35. 129. Tavakoli M, Marshall A, Pitceathly R, Fadavi H, Gow D, Roberts ME, et al. Corneal confocal microscopy: a novel means to detect nerve fi bre damage in idiopathic small fi bre neuropathy. Exp Neurol. 2010;223:245–50. 130. Parson H, Orciga M, Huertas H, Vinik A. Evidence of early involvement of A delta and C-fi bers in proximal sites in type 2 diabetes Abstract No. 14-LB. American Diabetes Association 71st Scienti fi c Session, June 24–28, 2011, San Diego, California. 131. Selvarajah D, Rao G, Tesfaye S. Diagnosing diabetic peripheral neuropathy using Sudoscan, a new, rapid method of assessing Sudometer function Abstract No. 2119-PO. American Diabetes Association 71st Scienti fi c Session, June 24–28, 2011, San Diego, California. 132. Gong Q, et al. Long-term effects of a randomised trial of a 6-year lifestyle intervention in impaired glucose tolerance on diabetes-related microvascular complications: the China Da Qing Diabetes Prevention Outcome Study. Diabetologia. 2011;54:300–7. 133. Gaede P, Vedel P, Larsen N, Jensen G, Parving H, Pedersen O. Multifactorial intervention and cardiovascular disease in patients with type 2 diabetes. N Engl J Med. 2003;348:383–93. 134. Boyd A, Casellini C, Vinik E, Vinik A. Quality of life and objective measures of diabetic neuropathy in a prospective placebo controlled trial of ruboxistaurin and topiramate. J Diabetes Sci Technol. 2011;5(3):714–22. 135. Boyd AL, Barlow P, Pittenger G, Simmons K, Vinik A. Topiramate improves neurovascular function, epidermal nerve fi ber morphology, and metabolism in patients with type 2 diabetes mellitus. Diabetes Metab Syndr Obes. 2010;3:431–7. 136. Colberg S, Swain D, Vinik A. Use of heart rate reserve and rating of perceived exertion to prescribe exercise intensity in diabetic autonomic neuropathy. Diabetes Care. 2003;26:986–90. 137. Colberg SR, Stansberry KB, McNitt PM, Vinik AI. Chronic exercise is associated with enhanced cutaneous blood fl ow in type 2 diabetes. J Diabetes Complications. 2002;16:139–45. 138. Michalsen A, Knoblauch NT, Lehmann N, Grossman P, Kerkhoff G, Wilhelm FH, et al. Effects of lifestyle modi fi cation on the progression of coronary atherosclerosis, autonomic function, and angina—the role of GNB3 C825T polymorphism. Am Heart J. 2006;151:870–7. 139. Howorka K, Pumprla J, Haber P, Koller-Strametz J, Mondrzyk J, Schabmann A. Effects of physical training on heart rate variability in diabetic patients with various degrees of cardiovascular autonomic neuropathy. Cardiovasc Res. 1997;34:206–14. 140. Motooka M, Koike H, Yokoyama T, Kennedy NL. Effect of dog-walking on autonomic nervous activity in senior citizens. Med J Aust. 2006;184:60–3. 141. Ziegler D, Schatz H, Conrad F, Gries FA, Ulrich H, Reichel G. Effects of treatment with the antioxidant alpha-lipoic acid on cardiac autonomic neuropathy in NIDDM patients. A 4-month randomized controlled multicenter trial (DEKAN Study). Deutsche Kardiale Autonome Neuropathie. Diabetes Care. 1997;20:369–73. 142. Ziegler D, Weise F, Langen KJ, Piolot R, Boy C, Hubinger A, et al. Effect of glycaemic control on myocardial sympathetic innervation assessed by metaiodobenzylguanidine scintigraphy: a 4-year prospective study in IDDM patients. Diabetologia. 1998;41:443–51. 143. Shun C, et al. Brain. Skin denervation in type 2 diabetes: correlations with diabetic duration and functional impairments. 2004;127:1593–605. Chapter 9 Interventional Trials to Prevent Diabetes: Diabetes Prevention Program

Vanita R. Aroda and Robert E. Ratner

Setting the Stage: Backdrop to the Diabetes Prevention Program

The largest randomized controlled clinical trial in diabetes prevention is the Diabetes Prevention Program (DPP), funded by the US National Institutes of Health. By design, the DPP is the most diverse prospective study to date in diabetes prevention. As part of the planning of the DPP, a combined analysis of six prospective studies in diverse populations was performed to identify predictors of progression from impaired glucose tolerance (IGT) to type 2 diabetes (T2DM) [ 1 ] . The six studies included the Baltimore Longitudinal Study of Aging and ﬁ ve population-based studies: the Rancho Bernardo Study, the San Antonio Heart Study, a 6-year study of the natural history of IGT in the Micronesian population of Nauru, the San Luis Valley Diabetes Study, and the Pima Indian Study in the Gila River Indian Community in Arizona. These studies followed men and women 2–27 years after the recognition of IGT. In this combined analysis, incidence of T2DM was signiﬁ cantly higher in the top quartile of fasting plasma glucose levels and increased linearly with increasing 2-h postchallenge glucose quartiles. Incidence rates of diabetes were higher in the Hispanic, Mexican-American, Pima, and Nauruan populations

V. R. Aroda , MD (*) MedStar Health Research Institute , 6525 Belcrest Road, Suite #700 , Hyattsville , MD 20782 , USA Georgetown University School of Medicine , Washington , DC , USA e-mail: [email protected] R. E. Ratner , MD MedStar Health Research Institute, Georgetown University School of Medicine , Washington , DC , USA

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 143 DOI 10.1007/978-1-4614-3314-9_9, © Springer Science+Business Media New York 2012 144 V.R. Aroda and R.E. Ratner compared to the Caucasian populations. Effect of age was not consistent across the studies. Measures of obesity, including body mass index, waist-to-hip ratio, and waist circumference, were positively associated with T2DM progression in all studies, whereas sex and family history of diabetes were not. This baseline analysis guided key design elements of the DPP, insuring enrollment of a diverse population of individuals at high risk of progression to type 2 diabetes [1 ] . The interventions for the DPP were also carefully evaluated based on existing data [ 2] . Prior to the DPP, two studies demonstrated the potential for behavioral interventions to prevent type 2 diabetes: the Malmo Study [ 3] and the Da Qing study [ 4 ] . In Malmo, Sweden, men aged 47–49 with IGT underwent a 5-year protocol of dietary treatment and/or increase in physical activity. Glucose tolerance normalized in >50% of the individuals with IGT, and improvements in glucose tolerance correlated with weight loss and increased fi tness. Blood pressure, lipids, and hyperinsulinemia improved [ 3 ] . Although the treatment groups were not randomized and differed in medical conditions at baseline, this study demonstrated the feasibility of carrying out a long-term lifestyle intervention program. The Da Qing study in China evaluated the effects of diet and exercise intervention in the prevention of diabetes in individuals with IGT. In this 6-year study, over 500 Chinese participants with IGT were cluster-randomized by clinic to a control group or to one of three active treatment groups: diet only, exercise only, or a combination of diet plus exercise. Compared to control, diet intervention reduced the risk of diabetes by 31%, exercise alone by 46%, and diet plus exercise by 42%. There were no signifi cant differences between the three intervention groups [ 4 ] . Here, too, the Da Qing study established the feasibility of long-term behavioral interventions to prevent progression to diabetes. Candidates for pharmacotherapy intervention were considered based on a record of lowering glycemia in individuals similar to the proposed DPP cohort, acceptable safety profi le, and favorable adherence/retention. Based on their mechanisms of actions of improved insulin sensitivity and improved glycemia and available safety pro fi le, both metformin and troglitazone were initially chosen as medications for intervention. Sulfonylureas were considered but not chosen due to risk of hypoglycemia [2 ] .

DPP: Clinical Trial Design and Methods

Eligibility criteria : Based on the initial analysis of predictors of progression from IGT to type 2 diabetes in six different populations [ 1 ] , the main inclusion criteria were the presence of IGT and elevated FPG (95–125 mg/dL). There was no FPG requirement for Native Americans due to high level of progression at even lower levels of FPG in this population. Recruitment was directed at overweight/obese individuals with BMI ³ 24 kg/m 2 , or ³ 22 kg/m 2 for Asian Americans due to high incidence of diabetes in this BMI range. Age criteria were set at ³ 25 years to allow inclusion of groups at high risk of diabetes in early adulthood, such as Native 9 Interventional Trials to Prevent Diabetes: Diabetes Prevention Program 145

Fig. 9.1 Locations of the 27 DPP investigative sites in the US Courtesy of the Diabetes Prevention Program ( http://www.bsc.gwu.edu/dpp/slides.htmlvdoc )

Americans and women with a history of gestational diabetes [5 ] . Eligible persons were excluded if they were taking medicines known to alter glucose tolerance (e.g., thiazides, beta blockers), if they had illnesses that could seriously reduce their life expectancy or their ability to participate in the trial, or if they experienced a CVD event £6 months before randomization [ 5 ] . To ensure a representative diverse population, recruitment aimed for at least half of the study group to be women, approximately 20% to be ³ 65 years of age, and half to consist of African-American, Hispanic, Native American, Asian American, and Paci fi c Islander minorities [6 ] . Overall study design: The initial study design of the DPP was a 4-group randomized trial across 27 clinical sites in the US (Fig. 9.1 ), with standardized protocols and procedures at all sites [ 7 ] . Eligible participants were randomly assigned to placebo, metformin 850 mg twice daily, troglitazone 400 mg every day, or an intensive program of lifestyle modifi cation. All groups received standard lifestyle recommendations in addition to the study intervention. The troglitazone arm was terminated during recruitment in June 1998 due to liver toxicity, resulting in the fi nal design of the DPP as a 3-group randomized trial (Fig. 9.2 ) [ 7 ] . The primary goal of the DPP was to prevent or delay the development of type 2 diabetes in individuals with IGT, and thus the primary outcome was the development of diabetes according to the American Diabetes Association criteria, con fi rmed by a repeat test. OGTT was performed annually, FPG performed every 6 months, and symptoms of hyperglycemia assessed to monitor for development of diabetes. Secondary outcomes included cardiovascular risk profi le and disease, changes in glycemia, insulin secretion and sensitivity, obesity, physical activity and nutrient intake, quality of life, and occurrence of adverse events [5, 7 ] . 146 V.R. Aroda and R.E. Ratner

Fig. 9.2 Study design of the diabetes prevention program. Courtesy of the Diabetes Prevention Program ( http://www.bsc.gwu.edu/dpp/slides.htmlvdoc )

Fig. 9.3 Timeline of the diabetes prevention program and diabetes prevention program outcomes study. Courtesy of Maranella Temprosa, Diabetes Prevention Program

Recruitment : Participant recruitment began in July 1996, and completed in May 1999 (Fig. 9.3 ). A four-step staged screening process determined participant eligibility (Fig. 9.4 ) [6 ] . Recruitment strategies included direct mailing, media advertisements, and work site and community screenings with mass direct mailing and print media appearing the most effective for overall study enrollment. The combination of strategies was successful in enrolling a diverse study population. While direct mail was cited as the primary source of information by Caucasian, African American, 9 Interventional Trials to Prevent Diabetes: Diabetes Prevention Program 147

Fig. 9.4 Screening and recruitment in the diabetes prevention program. Courtesy of the Diabetes Prevention Program ( http://www.bsc.gwu.edu/dpp/slides.htmlvdoc ) and Hispanic-American participants, screenings and community events, print media, and direct mail were equally cited by Asian Americans and Paci fi c Islanders. Native Americans were more responsive to community screenings and events or medical referral and Hispanic Americans and Asian Americans were responsive to phone calls and information from family and friends [6 ] . The eligibility criteria set forth permitted identi fi cation of individuals at high risk of diabetes across the different ethnic groups. Age, BMI, and elevated fasting glucose were key elements in predicting eligibility in the DPP. In all ethnic groups, older age increased the yield of high-risk individuals and in most ethnic groups with the exception of Asian Americans, BMI predicted high risk of diabetes [8 ] . Elevated fasting glucose, even when measured on capillary blood, predicted high yield of DPP eligibility in all ethnic groups, independent of age and BMI [8 ] . As shown in Fig. 9.4, 158,177 individuals were screened, of which 2.5% yielded the fi nal randomized cohort [6 ] . Adaptive randomization was employed to ensure balance among the treatment groups.

The Interventions

Standard lifestyle recommendation: Lifestyle interventions, standard or intensive, were targeted at reducing obesity and increasing physical activity. All participants received standard lifestyle recommendations following randomization and annually. Both written information and individual sessions were provided, addressing the 148 V.R. Aroda and R.E. Ratner importance of a healthy lifestyle for the prevention of T2DM. Participants were advised to lose 5–10% of their initial weight through diet and exercise and were counseled on the Food Pyramid guidelines, the National Cholesterol Education Program step 1 diet, and on gradually increasing physical activity such as walking to 30 min a day 5 days a week. Smoking cessation and avoiding excess alcohol were also reviewed [2, 5 ] . Intensive lifestyle intervention: The core goals for intensive lifestyle intervention were to achieve and maintain a weight loss of at least 7% of initial body weight through healthy eating and physical activity and to achieve and maintain a level of physical activity of at least 150 min/week through moderate intensity activity. In principle, the goals of the intensive lifestyle intervention were similar to standard lifestyle goals, though the approaches and resources to support implementation were extensive. The initial intervention phase was 24 weeks, followed by long-term efforts to sustain these goals. During this initial intervention phase, participants went through a 16-session core curriculum on diet, exercise, and behavior strategies. Dietary counseling focused initially on reduction of dietary fat intake, followed by setting of calorie goals if needed [ 2, 5 ] . Case managers, or lifestyle coaches, trained in nutrition, exercise, or behavior modifi cation met with individual participants for these sessions in the fi rst 24 weeks and had contact with the participants at least monthly thereafter. Supervised group exercise sessions were conducted twice a week. A toolbox approach was also available to allow the use of individualized speci fi c strategies to help each participant reach his/her goals [9 ] . Quarterly postcore classes were also offered and included the delivery of national campaign programs. Examples of such campaigns included the “10,000 Steps” campaign using a pedometer to advocate the goal of 10,000 steps per day, holiday campaigns, and lifestyle survival skills to prevent relapse [9 ] . To meet the needs of a diverse cohort, culturally appropriate materials were included into the curriculum. For example, at the Native American sites, oral translation of the curriculum was available in the local language. Lifestyle coaches gained familiarity with the participants’ cultural values to develop cultural compe- tency. Cultural beliefs and foods were incorporated into the curriculum. Study teams worked closely with the tribal health systems and Indian Health Service health providers to reinforce individual goals [10 ] . Metformin: Metformin or matching placebo were provided at a starting dose of 850 mg once daily, then increased to 850 mg twice daily. Adjustment of dose was allowed due to gastrointestinal symptoms. Medication case managers worked with participants to enhance adherence [2, 5 ] . Concomitant conditions : Adherence to standard treatment guidelines for hypertension, lipid management, and smoking cessation were encouraged. The use of medications that could worsen glucose intolerance was discouraged (e.g., thiazides, beta blockers, nicotinic acid). Women of child-bearing potential were asked to practice reliable contraception. Any new episodes of cardiovascular disease were adjudicated by an adjudication committee that was blinded to treatment assign- ment [ 2, 5 ] . 9 Interventional Trials to Prevent Diabetes: Diabetes Prevention Program 149

Sample size determinations : Data for the six population-based cohort studies were used to determine conversion rates from IGT to diabetes [ 1 ] . Based on the entry criterion of an elevated FPG and presence of IGT, the conversion rate from IGT to diabetes was estimated at 7.7 per 100 person-years of follow-up. To allow for a margin of error, the sample size was estimated based on an expected conversion of 6.5 per 100 person-years in participants assigned to the standard lifestyle-placebo group. Assuming intensive lifestyle or metformin would reduce the diabetes development hazard rate by ³33% and accounting for a 10% rate of loss to follow-up per year, the randomized goal of the DPP was 3,000 participants, or 1,000 per treatment group. It was estimated that >150,000 individuals would need to be screened to reach randomization goals [2, 5 ] .

DPP: Results

Randomized cohort : Over 158,000 individuals were screened to meet randomization goals [ 6 ] . True to the design of the study, the DPP randomized a diverse cohort of 3,234 participants at high risk of type 2 diabetes (IGT and elevated FPG) to placebo, intensive lifestyle intervention (ILS), or metformin [ 5 ] . An additional 585 participants were initially randomized to troglitazone and the results of this arm were reported separate from the primary publication [ 11 ] . Figures 9.5 and 9.6 illustrate the breakdown of the randomized cohort: 55% were Caucasian, 20% were African-American, 16% Hispanic, 5% Native American, and 4% Asian American. Average age at entry was 51 ± 10.7 years (mean ± SD) with 20% ³ 60 years of age.

Fig. 9.5 Ethnic distribution of DPP cohort. Courtesy of the Diabetes Prevention Program ( http:// www.bsc.gwu.edu/dpp/slides.htmlvdoc ) 150 V.R. Aroda and R.E. Ratner

Fig. 9.6 Sex and age distribution of DPP cohort. Courtesy of the Diabetes Prevention Program ( http://www.bsc.gwu.edu/dpp/slides.htmlvdoc )

67.7% of DPP participants were women. Overall, BMI was 34.0 ± 6.7 kg/m2 . Fasting glucose was 106.5 ± 8.3 mg/dL, and baseline glycosylated hemoglobin was 5.91 ± 0.50%. Gender, ethnic distribution, and risk factors for diabetes were similar among the three treatment groups [5, 12 ] . Of note, in June 1997, the American Diabetes Association rede ﬁ ned the threshold for diabetes at ³ 126 mg/dL from the previous cutoff of ³ 140 mg/dL. Fifty-four participants (1.67%) who were randomized under the former criteria (fasting glucose 100–139 or ±139 mg/dL in Native Americans) would have been considered diabetic by the new criteria. These individuals were followed as per the DPP protocol [ 2, 5, 12 ] .

Effects of Intensive Lifestyle Intervention and Metformin on Diabetes Prevention

Primary results: The blinded treatment phase was terminated 1 year early based on the advice of the data monitoring board (Fig. 9.3 ) [ 5 ] . Ef ﬁ cacy was clearly demonstrated on 65% of the planned person-years of observation. Compared to placebo, intensive lifestyle intervention reduced the incidence of diabetes by 58%, and metformin by 31% at an average follow-up of 2.8 years (Fig. 9.7 ). The average incidence of diabetes was 11.0, 7.8, and 4.8 cases per 100 person-years in the placebo, metformin, and lifestyle groups, respectively. Impressively, 6.9 persons would need to be treated in intensive lifestyle intervention to prevent one case of diabetes and 13.9 would need to be treated with metformin to prevent one case of diabetes [5 ] . 9 Interventional Trials to Prevent Diabetes: Diabetes Prevention Program 151

Fig. 9.7 Incidence of diabetes in the placebo, metformin, and lifestyle intervention groups of the diabetes prevention program. Courtesy of the Diabetes Prevention Program ( http://www.bsc.gwu. edu/dpp/slides.htmlvdoc )

Effects in subgroups: By its clinical design and recruitment strategy, the DPP was able to evaluate the effectiveness of the interventions in diverse populations. There were no signiﬁ cant differences in treatment effects by gender, race, or ethnicity. While lifestyle intervention was highly effective in all groups, it was particularly effective with increasing age (incident rates of 6.3, 4.9, and 3.3 cases per 100 person- years, in the 25–44, 45–59, and 60–85 year age groups, respectively) [ 13 ] . Likewise, participants aged 60–85 years had the greatest weight loss, greatest reduction in waist circumference, and metabolic equivalent hours of physical activity [13 ] . ILS was also more effective in individuals with lower 2-h postchallenge glucose levels than in those with higher 2-h postchallenge glucose values. Metformin had greater effect in diabetes reduction in individuals with higher BMI, reducing progression to diabetes by 53% in individuals with a BMI > 35 kg/m 2 , compared to only 3% in individuals with a BMI 22 to <30 kg/m 2 . Metformin was also more effective in the setting of higher fasting glucoses, consistent with its mechanism of action of suppressing endogenous glucose production [5 ] . Another subgroup analysis of interest was in women with a history of gestational diabetes (GDM). In the DPP cohort, 350 women reported a past history of GDM, and these were compared to the 1,416 women with a previous live birth but no history of GDM. As in previous studies, GDM conferred a very high risk of diabetes, with women with GDM randomized to placebo having a 71% higher incidence rate of diabetes than women without GDM. ILS reduced the incidence of diabetes by 50% in this population, similar to the 49% in the subgroup without GDM. Results from metformin treatment were striking, with a 50.4% reduction in diabetes in the 152 V.R. Aroda and R.E. Ratner

Fig. 9.8 Adverse events in the diabetes prevention program. Courtesy of the Diabetes Prevention Program ( http://www.bsc.gwu.edu/dpp/slides.htmlvdoc )

GDM subgroup compared to a 14.4% reduction in the non-GDM subgroup (interaction p = 0.06). This may in part be related to younger age of the GDM group, as metformin was as effective as ILS women aged 25–44 in the DPP. Only fi ve to six women with a history of GDM would need to be treated with either ILS or metformin to prevent one case of diabetes over a 3-year period [14 ] . Adherence and tolerability : With all of the tools and resources available to the participants, 50% were able to achieve the initial goal of weight loss of 7% or more by the end of the 24-week core curriculum and 38% had this level of weight loss by their last visit. Seventy-four percent met the activity goal of at least 150 min of moderate intensity physical activity per week by the end of the 24-week curriculum and 58% demonstrated this long-term by the last visit [5 ] . Seventy-two percent of participants in the metformin group took at least 80% of the study medication compared to 77% in the placebo group (p < 0.001). Eighty- four percent of those taking metformin were able to take the fully prescribed dose of 850 mg twice a day [ 5 ] . Within the metformin group, older age groups were more adherent in taking metformin than the youngest group, and women reported more adverse effects of metformin. Of note, the effect of metformin on risk of diabetes was more marked in those adherent to the metformin, with a 38.2% risk reduction for developing diabetes in those adherent to metformin compared to those adherent to placebo [15 ] . Both interventions were well tolerated. Gastrointestinal symptoms were more common in the metformin arm and musculoskeletal symptoms were more common in the intensive lifestyle arm (Fig. 9.8 ) [ 5 ] . Predictors of success: Although the DPP was not designed to fi gure out the separate contributions of dietary changes, increased physical activity, and weight loss on the reduction on risk of diabetes, it was able to examine factors related to achieving 9 Interventional Trials to Prevent Diabetes: Diabetes Prevention Program 153 weight loss and physical activity goals, and diabetes reduction. Caucasians were more likely to meet the initial weight loss goals, and men and those with lower BMI were more likely to meet initial activity goals [ 16 ] . Male gender, lower BMI, greater readiness for change for physical activity level as determined by validated questionnaires, higher exercise self-effi cacy, and lower perceived stress, anxiety, and depression scores all correlated with higher levels of physical activity levels at baseline, 1 year, and end of study [ 17 ] . Success of achieving long-term weight loss goals and both initial and long-term activity goals increased with age. Dietary self-monitoring and meeting activity goals were positively relating to achieving and sustaining weight loss. Participants who met the initial goals were more likely to meet these goals long-term [16 ] , highlighting the importance of trying to achieve these intervention goals by the end of the initial intervention phase. Other psychosocial variables such as employment, marital status, income, and depressive symptoms were unrelated to achieving weight loss or activity goals [ 16 ] . Depression scores at baseline or during the study were also not related to diabetes risk, although baseline or continuous antidepressant medication use were in the ILS and placebo groups [ 18] . Modest alcohol intake was associated with a reduced risk of incident diabetes in the ILS and metformin groups and was associated with lower insulin secretion [ 19 ] . Evaluating the success in the ILS arm alone, weight loss was the primary predictor of reduced diabetes incidence, with every kilogram of weight loss conferring a 16% reduction in risk of progression to diabetes, adjusted for changes in diet and activity [20 ] . Reductions in central body fat distribution, as determined by computed tomography, BMI, and waist circumference, were all associated with decreased diabetes risk in the ILS group [21 ] . In turn, lower percent of calorie intake from fat and increased physical activity predicted weight loss [20 ] . Baseline waist circumference was a strong predictor of diabetes development in the placebo and ILS groups [ 22] . Even in individuals who did not meet weight loss goals, achieving the physical activity goal alone lowered the risk of diabetes by 44% [20 ] . In the metformin group, weight loss alone explained 64% of its benefi cial effect compared to placebo. Improvements in insulin sensitivity, proinsulin, and reductions in fasting glucose also contributed to metformin’s effects on diabetes reduction. While metformin signi fi cantly decreased fasting glucose, it did not impact the 120-min glucose, suggesting its primary action in this population via suppressing endogenous hepatic glucose production rather than by enhancing peripheral glucose uptake [23 ] . Biomarkers as predictors of diabetes in the DPP : Adiponectin, a biomarker of metabolic and cardiovascular disease risk, was also evaluated in the DPP. Baseline adiponectin was strongly inversely related to progression to diabetes in all three treatment groups. Increases in adiponectin were associated with weight loss and changes in insulin sensitivity, but not changes in beta-cell function. In the placebo and ILS groups, increases in adiponectin were associated with decrease in diabetes risk, though less so than baseline adiponectin levels [24 ] . Urine albumin to creatinine ratio (ACR), also linked to cardiovascular risk factors, vascular disease, and insulin resistance, was also evaluated in the DPP. 154 V.R. Aroda and R.E. Ratner

Elevated ACR was present in 6.2% at entry into the DPP. In this population, however, at levels below microalbuminuria level, ACR did not independently predict diabetes [25 ] . Effects on weight and nutrient intake: Weight loss was greatest in the intensive lifestyle group (5.6 kg), compared to placebo (0.1 kg) and metformin (2.1 kg) [ 5 ] . A substudy in 758 participants at baseline and 1 year showed that ILS reduced both visceral adiposity and subcutaneous adiposity in both men and women, whereas metformin reduced subcutaneous fat in men only [21 ] . The ILS group had the greatest reduction in daily energy intake and average fat intake and the greatest increase in leisure physical activity. In the intensive lifestyle group, total caloric intake decreased 452 kcal/day by 1 year and percent energy from fat reduced 6.6%, whereas in the metformin and placebo groups, total energy intake decreased 294 and 250 kcal/day, respectively, and percent energy intake from fat decreased 0.8% in both groups after 1 year [26 ] . Effects on glucose variables: Fasting glucose, insulin, and proinsulin concentrations decreased during the ﬁ rst year in the ILS and metformin groups. In the subsequent years, levels of each increased in all, though remained signi ﬁ cantly lower than the placebo group [27 ] . Relative to the other groups, ILS was more effective in restor- ing normal postload glucose values [ 5, 27 ] . At 1 year, ILS produced to the greatest improvement in insulin sensitivity, as calculated by the insulin sensitivity index and 1/fasting insulin, and enhanced beta-cell function, with metformin effects intermediate to those of ILS of placebo. Of note, baseline FPG, 2-h postchallenge glucose, HbA1c, proinsulin, insulin sensitivity, and insulin secretion were all independently predictive of the development of diabetes [27 ] .

Effects of Troglitazone

The initial design of the DPP included a fourth randomized arm with troglitazone, a potent insulin-sensitizing agent. The troglitazone arm was discontinued in June 1998 due to concern regarding liver toxicity, and participants continued follow-up (Fig. 9.3 ). During the mean 0.9 years of treatment with troglitazone (n = 585), diabetes was reduced 75% compared to placebo, signi fi cantly lower than placebo and metformin, though not signi fi cantly lower than intensive lifestyle intervention. Troglitazone signifi cantly improved insulin sensitivity, as determined by the insulin sensitivity index, more than placebo or metformin though again, not more than intensive lifestyle intervention. Marked elevations in liver enzymes to at least ten times the upper limit of normal occurred in more participants in the troglitazone arm (1.2%) compared to metformin (0.0%) or placebo (0.2%) (p < 0.01). There was one death in the troglitazone group due to liver disease, contributing to the decision to terminate this arm in 1998. The rate of development of diabetes reached that of placebo following discontinuation of troglitazone, although the cumulative incidence of diabetes from the date of randomization remained overall lower in the troglitazone group [ 11 ] . 9 Interventional Trials to Prevent Diabetes: Diabetes Prevention Program 155

Additional Effects of Diabetes Prevention

Effects on cardiovascular risk factors : Persons with IGT, metabolic syndrome, and type 2 diabetes are at increased risk of cardiovascular disease, and thus impacts on cardiovascular risk factors and CVD during the DPP and Diabetes Prevention Program Outcomes Study (DPPOS) are of considerable interest. At baseline, prevalence of hypertension was 30% across the study participants and was higher in African Americans (36.4%) and Asian Americans (37.7%) and lowest in Hispanics (21.5%) and Native Americans (13.8%). Prevalence of hypertension was also strongly associated with age, male gender, higher levels of fasting glucose, fasting insulin, proinsulin, adiposity, and urinary albumin/creatinine ratio [ 28 ] . In the DPP, the prevalence of hypertension increased in the placebo and metformin groups and did not change in the intensive lifestyle group. Compared to placebo, both ILS and metformin decreased systolic and diastolic blood pressure, though the impact was greatest with ILS (−3.27 ± 0.5 SBP, −3.82 ± 0.3 DBP) [29 ] . At baseline, among the 3,819 patients randomized to the initial four treatment arms, over 40% had elevated triglyceride, LDL cholesterol, C-reactive protein (CRP) levels, and reduced HDL levels. There was signifi cant heterogeneity in baseline levels by gender, age, and ethnicity, with men having higher triglyceride and lower HDL levels with smaller LDL particle size and women having higher CRP and fi brinogen levels. Increasing age was associated with increases in total, LDL, and HDL cholesterol. African Americans had lower triglycerides and higher HDL and higher fi brinogen levels while Asian Americans had lower CRP and fi brinogen levels than Caucasians and Hispanics. The degree of insulin resistance, as measured by homeostasis model assessment of insulin resistance, was a signifi cant determinant of triglyceride, HDL cholesterol, LDL particle size, and tPA, while BMI in fl uenced CRP and fi brinogen levels [30 ] . Compared to placebo and metformin, intensive lifestyle intervention demonstrated favorable effects on lipid parameters, causing a signi fi cantly greater decrease in triglyerides (−11.9 mg/dL placebo, −7.4 mg/dL metformin, −25.4 mg/dL ILS) and increase in HDL cholesterol (−0.1 mg/dL placebo, +0.3 mg/dL metformin, +1.0 mg/dL ILS) compared to placebo (−11.9 mg/dL) or metformin (−7.4 mg/dL). Furthermore, ILS reduced the prevalence of proatherogenic LDL phenotype B representing the smaller, more dense, more atherogenic LDL [29 ] . Consistent with these improvements, intensive lifestyle participants required less pharmacotherapy for risk factor modifi cation. Intensive lifestyle signifi cantly decreased the point prevalence of antihypertensive therapy, and fewer ILS participants required drug therapy for elevated triglyceride or LDL cholesterol levels compared to metformin or placebo participants [29 ] . Improvements in the nontraditional cardiovascular risk factors of CRP and fi brinogen were also seen with ILS and to a lesser extent metformin. ILS reduced CRP levels by 33% in men and by 29% in women at 1 year. Metformin reduced CRP by −7% in men and −14% in women. ILS, but not metformin, also decreased fi brinogen levels at 1 year. Changes in CRP and fi brinogen in the ILS group were explained in part by weight loss rather than by changes in physical activity [31 ] . 156 V.R. Aroda and R.E. Ratner

Approaching cardiovascular risk as a cluster of metabolic risk factors, the prevalence of metabolic syndrome was also evaluated in the DPP. Metabolic syndrome was defi ned by the National Cholesterol Education Program Adult Treatment Panel III criteria and was present in 53% of the participants at baseline. There were no differences in the prevalence of metabolic syndrome or its components among treatment groups, with the exception of a greater prevalence of low HDL levels in the placebo group. Over a mean 3.2 years, ILS signifi cantly reduced the prevalence of metabolic syndrome by 41% and metformin decreased the prevalence by 17% compared to placebo [32 ] . Did these improvements in cardiovascular risk factors translate to reduction in CVD? During the DPP, there were a total of 89 confi rmed cardiovascular disease events with no difference in CVD or CVD-related deaths among the treatment groups [29 ] . The observation period of the DPP was likely too short to detect a difference in CVD, particularly with enforcement of standardized treatment goals for cardiovascular risk factors. Prolonged observation in the Diabetes Prevent Program Outcomes Study (DPPOS) will determine whether these initial improvements in cardiovascular risk factors ultimately translate to reduction in CVD. Effects on microvascular disease: The follow-up in the DPP was too short to determine de fi nitive effects of ILS or metformin on development of microvascular disease, but did provide insight and suggestion of potential bene fi t. Effects on albumin excretion : Despite improvements in hypertension and cardiovascular risk factors, ILS and metformin had no signifi cant effect on ACR within the duration of the DPP [ 33 ] . Effects on retinopathy: Fundus photography was performed at a mean 5.6 years after study entry in a subset of 898 nondiabetic and diabetic participants. Diabetic retinopathy was detected in 12.6% of the diabetic participants, within approximately 3 years of diagnosis of diabetes and in 7.9% of the nondiabetic participants, suggesting that retinopathy may begin in the prediabetic state and increase signi fi cantly shortly after the development of diabetes. Numbers in the treatment groups were too small for comparison, but may be more revealing in the long-term follow-up of the DPP participants [ 34 ] . Effects on autonomic nervous system : Autonomic nervous system dysfunction is associated with obesity, insulin resistance, and diabetes. As such, measures of fi tness and autonomic nervous system function, including heart rate, heart rate variability, and EKG QT duration, were evaluated at baseline and annually in 2,980 DPP participants. Increased heart rate at baseline was signi fi cantly associated with the development of diabetes, even after adjustment for weight change and physical activity. Here again, ILS showed benefi cial effects on autonomic nervous system function, decreasing heart rate and QT indices and increasing heart rate variability. These changes were also associated with lower risk of diabetes overall, independent of weight change. These results suggest improved fi tness and improvement in autonomic function as one contributory mechanism of lowering risk of diabetes in the lifestyle arm [35 ] . Effects on urinary incontinence in women : Because diabetes is associated with increased risk of urinary incontinence, data were collected on incontinence symptoms 9 Interventional Trials to Prevent Diabetes: Diabetes Prevention Program 157 by frequency and type at the end-of-trial visit. Although this information was not collected at baseline, at the end of study, women randomized to ILS had signifi cantly lower prevalence of weekly incontinence (38.3%), particularly stress incontinence compared to metformin (48.1%) or placebo (45.7%). This bene fi t persisted after adjusting for baseline hormone therapy use, general health status, and 2-h postchallenge glucose categories. Urge incontinence did not differ among the treatment groups. There was no signi fi cant difference by age, ethnicity, or BMI [36 ] .

Genetic Insights from the DPP

3,548 (92.9%) of the randomized participants in the DPP provided consent for genetic investigation, affording the opportunity to evaluate genetic susceptibility to the progression to type 2 diabetes in a diverse population [ 37 ] . Highlighted are just a few selected completed explorations, with further investigations underway. Previous studies in Caucasian patients by Grant and colleagues identi fi ed a common allele in the transcription factor 7-like 2 gene (TCF7L2) that increased the risk of type 2 diabetes. Specifi cally, they identifi ed the T alleles at two single-nucleotide polymorphisms (rs12255372 and rs7903146) as risk variants. They speculated that genetic variation in TCF7L2 would impair the expression of glucagon-like peptide 1 in enteroendocrine cells [ 38 ] . These variants were also evaluated in the diverse cohort of the DPP. DPP participants who were homozygous for the T allele at rs7903146 or at rs12255372 also had increased risk of progression compared to those homozygous for the C allele or G allele, respectively, with the greatest effect in the placebo group. The frequency of the minor T alleles was similar in Caucasians and African Americans, but lower in Hispanics, Asians, and Native Americans. At baseline, the TT genotype was associated with decreased insulin secretion, but not increased insulin resistance, which could be consistent with the earlier model that variation in TCF7L2 impairs GLP-1 expression [37 ] . Several other polymorphisms have also been evaluated in the DPP. A proline (P) allele at codon 12 of the PPARG gene (P12A variant) is thought to increase risk of type 2 diabetes and may affect therapeutic response to thiazolidinediones. Previous reports were contradictory, possibly due to sample size, different ascertainment, or analytical methods. In the DPP, P/P homozygotes at PPARG p12A appeared more likely to develop diabetes than alanine carriers (HR 1.24, 95% CI 0.99–1.57, p = 0.07). The apparent protective effect of alanine was less in more obese individuals. Moreover, there was no signifi cant effect of genotype at PPARG P12A in response to troglitazone treatment [39 ] . In a separate analysis of the KCNJ11 gene, which encodes the islet ATP-sensitive potassium channel Kir6.2, lysine change at position 23 (E23K) appeared protective in the DPP participants, contrary to earlier reports. However, carriers of the lysine variant did have impaired insulin secretion and responded less well to metformin treatment at 1 year than E/E homozygotes. These fi ndings suggested that the lysine allele may have deleterious effects at earlier stages of diabetes rather than in a population that already has IGT [40 ] . 158 V.R. Aroda and R.E. Ratner

Pharmacogenomics analyses in the DPP suggest that there may be differential responses to treatment based on one’s genotype. In one exploration, carriers of the protective genotype CDKN2A/B showed differential improvement in beta-cell function after 1 year of troglitazone treatment (p = 0.01) and possibly lifestyle modi fi cation ( p = 0.05) [41 ] . In another analysis, the ENPP1 K121Q polymorphism, involved in insulin signaling, conferred an increased risk of diabetes incidence, which was eliminated with lifestyle intervention [42 ] . Moreover, potential genetic determinants of response to metformin therapy were identifi ed in the DPP population, which may ultimately have therapeutic implications if con fi rmed [43 ] . A number of other investigations beyond the scope of this chapter have been conducted or are underway to evaluate genetic in fl uences on diabetes risk and response to treatment.

Are the Effects of Intensive Lifestyle and Metformin on Diabetes Prevention Sustainable?

To evaluate whether the effect of metformin on diabetes prevention was a pharmacological effect or a sustained effect, a short washout was performed at the end of the DPP in participants who had not yet developed diabetes. During this washout, study drug was discontinued for 1–2 weeks and OGTT was repeated. Following the washout, those in the metformin arm had a reduction in incidence of diabetes by 25%, reduced from the original fi ndings of 31% reduction in diabetes with metformin [44 ] . These fi ndings suggest that metformin may exert a combined pharmacological and physiologic effect on the prevention or delay of diabetes. A longer washout period would have been better able to delineate this. Another way to assess sustainability of diabetes prevention is by evaluating per- sistence of these effects years after the initial intervention. Just as was shown in the long-term follow-up of the Da Qing participants [ 45 ] and Finnish Diabetes Prevention Study [ 46 ] , 10-year follow-up since randomization into the DPP shows ongoing diabetes reduction years past the initial intervention. Eighty-eight percent of the active DPP participants enrolled into the DPPOS for ongoing follow-up. In the DPPOS, all groups were offered group-based lifestyle intervention based on the fi ndings of the DPP and the metformin group continued unmasked metformin treatment. New incidence rates in the follow-up study alone were similar across the treatment groups and may represent the loss of a true placebo group with the introduction of lifestyle intervention for all groups. However, during the median 10-year follow-up, cumulative diabetes incidence in the ILS group was reduced by 34% and in the metformin group was reduced by 18% compared to placebo, suggesting a persistent effect of ILS and metformin on diabetes prevention or delay. Moreover, the modest weight loss in the metformin group was maintained while the ILS group partly regained weight [47 ] . 9 Interventional Trials to Prevent Diabetes: Diabetes Prevention Program 159

Is It Cost-Effective to Prevent or Delay Type 2 Diabetes?

When the DPP primary results were fi rst released, some of the fi rst questions raised were of feasibility and cost-effectiveness [48 ] . Herman et al. estimated progression of disease, costs, and quality of life to determine cost-effectiveness of the DPP. They estimated that over a lifetime, lifestyle intervention would reduce the absolute incidence of diabetes by 20% and delay its onset by 11.1 years, increase life expectancy by 0.5 years, reduce the incidence of blindness by 39%, end-stage renal disease by 38%, amputation by 35%, stroke by 9%, and coronary heart disease by 8%. Compared to placebo, lifestyle intervention would cost $635 more over a lifetime and produce a gain of 0.57 quality-adjusted life years (QALY), or a cost per QALY gained of ~$1,100. Intensive lifestyle intervention was not only cost- effective in all age groups, but cost-saving in those under the age of 45 [49 ] . Analysis from the DPP also suggest cost-effectiveness of metformin in younger age groups, but a signi fi cant cost of more than $100,000 per QALY in participant 65 years of age or older. Over a lifetime, metformin is expected to prevent diabetes in 8% and delay the onset of diabetes by 3.4 years, increase life expectancy by 0.2 year, and reduce the cumulative incidence of blindness by 16%, end-stage renal disease by 17%, amputation by 16%, stroke by 3%, and coronary heart disease by 2%. Assuming the use of generic metformin, the estimated cost per QALY gained for metformin intervention extended over the lifetime compared to placebo is $1,755 [49 ] . Assuming real-world application with group lifestyle intervention and use of generic metformin, a sensitivity analysis estimated the cost of preventing a single case of diabetes during a 3-year period to $4,301 with lifestyle intervention and $11,141 with metformin [ 50, 51 ] . The full impact of the initial interventions on lifetime cost and benefi t remains to be seen. Interim health economic analysis from the DPP and DPPOS presented at the American Diabetes Association Scienti fi c Sessions, 2011, suggests ongoing cost-effectiveness of intervention. Specifi cally, metformin therapy for diabetes prevention appears cost - saving compared to placebo and intensive lifestyle is cost- effective in the range of treating diastolic hypertension >105 mmHg [52 ] .

Impact of the Diabetes Prevention Program on Guidelines and Policy

Diabetes prevention guidelines: The DPP and its results have had far-reaching effects and will continue to do so in the coming decades. By demonstrating de fi nitive effi cacy of interventions to prevent or delay the progression to type 2 diabetes in large population diverse in age, gender, and ethnicity, it has impacted the recommended approach to diabetes screening and prevention. This impact is demonstrated in the recommended approach to diabetes screening and prevention by multiple medical societies, as outlined in Table 9.1 [ 53– 61 ] . All of the guidelines recommend Table 9.1 Recommendations of selected organizations on the screening and prevention of type 2 diabetes Population to screen Method of screening Recommended treatment(s) Follow-up American Diabetes In asymptomatic adults, A1c Ongoing support program targeting weight In those with prediabetes, Association, 2011 screen all individuals FPG loss of 7% of body weight, increased monitor for the develop- [ 53, 54 ] who are overweight 2-hPG physical activity to at least 150 min/week ment of diabetes by BMI with any of moderate activity every year additional risk factor Consider metformin in those at highest risk, If screening is normal and no for diabetes such as those with multiple risk factors, risk factors, repeat every progression of hyperglycemia (e.g., A1c 3 years ³ 6%) despite lifestyle interventions Identify and treat other cardiovascular risk factors American College Screen individuals FPG, 2-hPG, or screening Lifestyle modi fi cation, weight loss 5–10%, Annual measurement of FPG of Endocrinology with risk factors for metabolic regular moderate intensity physical activity and A1c, optional OGTT and the American for diabetes syndrome Consider additional pharmacotherapy if suspected progression Association in individuals at high risk of hyperglycemia of Clinical Cardiovascular risk factor treatment, including Annual assessment of blood Endocrinologists, low-dose aspirin in absence of contraindica- pressure, fasting lipid, 2008 [55 ] tion, blood pressure, and lipid management and microalbuminuria More frequent monitoring in patients at high risk of progression Endocrine Society, Individuals with Screen for evidence of Weight reduction (goal weight loss 5–10%) Screen individuals with IFG 2008 [56 ] “metabolic risk” metabolic risk, e.g., if overweight or obese, weight maintenance or IGT for overt type 2 presence of multiple if not diabetes and metabolic risk components of Regular moderateintensity physical activity factors at 1- to 2-year metabolic syndrome Diet low in total and saturated fat, low in trans intervals fatty acids, with adequate fi ber Priority to lifestyle modi fi cation rather than drug therapies Canadian Diabetes Screen individuals with FPG, 2hPG in individuals Lifestyle modi fi cation, including moderate If glucose is normal, rescreen Association, risk factors for diabetes with FPG 110–125 mg/ weight loss and regular physical activity every 3 years, more often 2008 [57 ] In the absence of risk dL (6.1–6.9 mmol/L), Consider pharmacotherapy in individuals with if high risk or evidence factors, begin screening or FPG 5.6–6.0 mmol/L IGT of prediabetes at the age of 40 and ³ 1 risk factors Indian Health Annual testing of FPG, 2-hPG as resources Lifestyle modi fi cation Monitor glucose values every Service, individuals at risk for permit Consider metformin on an individualized basis 6 months 2008 [58 ] developing diabetes United States Asymptomatic adults with Does not recommend Recommends intense interventions for obese Optimal screening interval for Preventive sustained blood speci fi c screening for persons who desire to lose weight and diabetes not known Services Task pressure, treated or IFG or IGT, but population-based approaches to increase Force, 2008 [59 ] untreated, greater than supports screening for physical activity and reduce obesity 135/80 mmHg diabetes Australian Diabetes Screen individuals at risk FPG fi rst; if FPG between Lifestyle modi fi cation for a minimum of 6 75 g OGTT, initially performed Society and for diabetes 5.5 and 6.9 mmol/L, months before consideration of annually, then individual- Australian Diabetes then proceed with 75 g pharmacotherapy ized retesting every 1–3 Educators OGTT years Association, 2007 [60 ] International Diabetes Identify high-risk FPG, then OGTT if FPG is Lifestyle modi fi cation Not speci fi ed Federation, 2007 individuals by risk ³ 110–125 mg/dL Consider pharmacotherapy if desired weight [61 ] factor assessment (6.1–6.9 mmol/L) loss and/or improved glucose tolerance (central obesity, family Assess other risk factors goals not achieved history, age, history of (waist circumference, Cardiovascular risk factor treatment elevated blood pressure blood pressure, family and/or heart disease, history of diabetes, history of gestational triglycerides, evidence diabetes, use of drugs of preexisting that predispose to type cardiovascular disease) 2 diabetes) Presented by date, beginning with most recent 162 V.R. Aroda and R.E. Ratner screening individuals at high risk of type 2 diabetes, recognizing the merit of early detection of a preventable disease. Nearly all guidelines base their recommendations for treatment based on results of the DPP and similar studies, recommending intensive lifestyle intervention targeting a weight loss of 5–10% and increased moderate intensity physical activity to 150 min/week. Recommendations for pharmacotherapy also factor in fi ndings from the DPP, recognizing the greater effect of metformin in younger individuals, individuals with higher BMI, or other risk factors such as gestational diabetes. Policy: Recognizing the merit of disease prevention, efforts are underway to translate diabetes prevention efforts into policy. US Senators Al Franken and Richard Lugar introduced the “Diabetes Prevention Amendment” in 2009 to facilitate translation of evidence from the DPP and similar studies. Congress passed the Patient Protection and Affordable Care Act in March, 2010, which allowed for the creation of the CDC-led National DPP [ 62] . Based in principle on the DPP, this program includes grant programs for community-based diabetes prevention, recognition and determination of entities able to deliver community-based diabetes prevention services, training and outreach programs for lifestyle intervention instructors, an evaluation and monitoring process, and potential for further research [63 ] . The Y, also known as the YMCA of the USA, and United Health Group are two instrumental partners of the National DPP. Early pilot studies utilizing the YMCA demonstrated the feasibility in disseminating lifestyle interventions adapted from the DPP on a widespread scale in the community. In the DEPLOY pilot study, group-based DPP lifestyle intervention delivery by the YMCA decreased body weight 6.0% and decreased total cholesterol 21.6 mg/dL after 6 months, signi fi cantly more so than brief counseling alone [64 ] . In partnership with United Health Group and the CDC, lifestyle intervention programs adapted from the DPP are now being offered at numerous YMCA facilities throughout the country, with continued expansion in the number of YMCA facilities and other organizations getting involved in diabetes prevention. In summary, the DPP is the largest prospective study to date demonstrating the wide-reaching impact and feasibility of diabetes prevention efforts in a diverse population. Current efforts are now focused on translating this evidence into practice. In fact, the Strategic Planning Report of the Diabetes Mellitus Interagency Coordinating Committee, posted by the NIH/NIDDK in February 2011, has identifi ed several key areas for future research in diabetes prevention that build upon the DPP. These are: (1) How can the outcomes of the DPP be translated in diverse settings and populations to prevent type 2 diabetes in youth and adults? (2) What are the key behavioral and environmental factors that need to be assessed along with genetic markers to better tailor type 2 diabetes prevention approaches? (3) How can the structures and policies of communities, worksites, and other systems infl uence behavioral change in individuals to prevent type 2 diabetes? (4) How can interventions to prevent type 2 diabetes be cost-effective at the societal level and fi nancially feasible from the perspective of individual payers and healthcare organizations [65 ] ? 9 Interventional Trials to Prevent Diabetes: Diabetes Prevention Program 163

The DPP has indeed established the necessary foundation to answer these questions and more, and its ongoing follow-up, the DPPOS, will continue to provide valuable insights to the lifetime effects of diabetes prevention.

References

1. Edelstein SL, Knowler WC, Bain RP, Andres R, Barrett-Connor EL, Dowse GK, et al. Predictors of progression from impaired glucose tolerance to NIDDM: an analysis of six prospective studies. Diabetes. 1997;46:701–10. 2. Group The Diabetes Prevention Program Research. The Diabetes Prevention Program: design and methods for a clinical trial in the prevention of type 2 diabetes. Diabetes Care. 1999;22:623–34. 3. Eriksson KF, Lindgarde F. Prevention of type II (non-insulin-dependent) diabetes mellitus by diet and physical exercise: the 6-year Malmo Feasibility Study. Diabetologia. 1991;34: 891–8. 4. Pan XR, Li GW, Hu YH, Wang JX, Yang WY, An ZX, et al. Effects of diet and exercise in preventing NIDDM in people with impaired glucose tolerance. The Da Qing IGT and Diabetes Study. Diabetes Care. 1997;20:537–44. 5. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, et al.; Diabetes Prevention Program Research Group. Reduction in the incidence of type 2 diabetes with lifestyle or metformin. N Engl J Med. 2002;346:393–403. 6. Rubin RR, Fujimoto WY, Marrero DG, Brenneman T, Charleston JB, Edelstein SL, et al.; DPP Research Group. The Diabetes Prevention Program: recruitment methods and results. Control Clin Trials. 2002;23:157–71. 7. Fujimoto WY. Background and recruitment data for the U.S. Diabetes Prevention Program. Diabetes Care. 2000;23 Suppl 2:B11–3. 8. Diabetes Prevention Program Research Group. Strategies to identify adults at high risk for type 2 diabetes: the Diabetes Prevention Program. Diabetes Care. 2005;28:138–44. 9. Wylie-Rosett J, Delahanty L. An integral role of the dietitian: implications of the Diabetes Prevention Program. J Am Diet Assoc. 2002;102:1065–8. 10. Hoskin M, Begay SH, Bolin P, Hermes J, Ingraham LE, Killean T, et al. Providing lifestyle interventions in Native American communities. Obesity Management. 2005;1:251–5. 11. Knowler WC, Hamman RF, Edelstein SL, Barrett-Connor E, Ehrmann DA, Walker EA, et al.; Diabetes Prevention Program Research Group. Prevention of type 2 diabetes with troglitazone in the Diabetes Prevention Program. Diabetes. 2005;54:1150–6. 12. Group The Diabetes Prevention Program Research. The Diabetes Prevention Program: baseline characteristics of the randomized cohort. Diabetes Care. 2000;23:1619–29. 13. Diabetes Prevention Program Research Group, Crandall J, Schade D, Ma Y, Fujimoto WY, Barrett-Connor E, et al. The in ﬂ uence of age on the effects of lifestyle modi ﬁ cation and metformin in prevention of diabetes. J Gerontol A Biol Sci Med Sci. 2006;61:1075–81. 14. Ratner RE, Christophi CA, Metzger BE, Dabelea D, Bennett PH, Pi-Sunyer X, et al.; Diabetes Prevention Program Research Group. Prevention of diabetes in women with a history of gestational diabetes: effects of metformin and lifestyle interventions. J Clin Endocrinol Metab. 2008;93:4774–9. 15. Walker EA, Molitch M, Kramer MK, Kahn S, Ma Y, Edelstein S, et al. Adherence to preventive medications: predictors and outcomes in the Diabetes Prevention Program. Diabetes Care. 2006;29:1997–2002. 16. Wing RR, Hamman RF, Bray GA, Delahanty L, Edelstein SL, Hill JO, et al.; Diabetes Prevention Program Research Group. Achieving weight and activity goals among diabetes prevention program lifestyle participants. Obes Res. 2004;12:1426–34. 164 V.R. Aroda and R.E. Ratner

17. Delahanty LM, Conroy MB, Nathan DM; Diabetes Prevention Program Research Group. Psychological predictors of physical activity in the diabetes prevention program. J Am Diet Assoc. 2006;106:698–705. 18. Rubin RR, Ma Y, Marrero DG, Peyrot M, Barrett-Connor EL, Kahn SE, et al.; Diabetes Prevention Program Research Group. Elevated depression symptoms, antidepressant medicine use, and risk of developing diabetes during the diabetes prevention program. Diabetes Care. 2008;31:420–6. 19. Crandall JP, Polsky S, Howard AA, Perreault L, Bray GA, Barrett-Connor E, et al.; Diabetes Prevention Program Research Group. Alcohol consumption and diabetes risk in the Diabetes Prevention Program. Am J Clin Nutr. 2009;90:595–601. 20. Hamman RF, Wing RR, Edelstein SL, Lachin JM, Bray GA, Delahanty L, et al. Effect of weight loss with lifestyle intervention on risk of diabetes. Diabetes Care. 2006;29:2102–7. 21. Fujimoto WY, Jablonski KA, Bray GA, Kriska A, Barrett-Connor E, Haffner S, et al.; Diabetes Prevention Program Research Group. Body size and shape changes and the risk of diabetes in the diabetes prevention program. Diabetes. 2007;56:1680–5. 22. Group The Diabetes Prevention Program Research. Relationship of body size and shape to the development of diabetes in the diabetes prevention program. Obesity. 2006;14:2107–17. 23. Lachin JM, Christophi CA, Edelstein SL, Ehrmann DA, Hamman RF, Kahn SE, et al.; DDK Research Group. Factors associated with diabetes onset during metformin versus placebo therapy in the diabetes prevention program. Diabetes. 2007;56:1153–9. 24. Mather KJ, Funahashi T, Matsuzawa Y, Edelstein S, Bray GA, Kahn SE, et al.; Diabetes Prevention Program. Adiponectin, change in adiponectin, and progression to diabetes in the Diabetes Prevention Program. Diabetes. 2008;57:980–6. 25. Friedman AN, Marrero D, Ma Y, Ackermann R, Narayan KM, Barrett-Connor E, et al.; Diabetes Prevention Program Research Group. Value of urinary albumin-to-creatinine ratio as a predictor of type 2 diabetes in pre-diabetic individuals. Diabetes Care. 2008;31:2344–8. 26. Mayer-Davis EJ, Sparks KC, Hirst K, Costacou T, Lovejoy JC, Regensteiner JG, et al.; Diabetes Prevention Program Research Group. Dietary intake in the diabetes prevention program cohort: baseline and 1-year post randomization. Ann Epidemiol. 2004;14:763–72. 27. Kitabchi AE, Temprosa M, Knowler WC, Kahn SE, Fowler SE, Haffner SM, et al.; Diabetes Prevention Program Research Group. Role of insulin secretion and sensitivity in the evolution of type 2 diabetes in the diabetes prevention program: effects of lifestyle intervention and metformin. Diabetes. 2005;54:2404–14. 28. Diabetes Prevention Program Research Group. Hypertension, insulin, and proinsulin in participants with impaired glucose tolerance. Hypertension. 2002;40:679–86. 29. Ratner R, Goldberg R, Haffner S, Marcovina S, Orchard T, Fowler S, et al.; Diabetes Prevention Program Research Group. Impact of intensive lifestyle and metformin therapy on cardiovascular disease risk factors in the diabetes prevention program. Diabetes Care. 2005;28:888–94. 30. Diabetes Prevention Program Research Group. Lipid, lipoproteins, C-reactive protein, and hemo- static factors at baseline in the diabetes prevention program. Diabetes Care. 2005;28:2472–9. 31. Haffner S, Temprosa M, Crandall J, Fowler S, Goldberg R, Horton E, et al.; Diabetes Prevention Program Research Group. Intensive lifestyle intervention or metformin on in ﬂ ammation and coagulation in participants with impaired glucose tolerance. Diabetes. 2005;54:1566–72. 32. Orchard TJ, Temprosa M, Goldberg R, Haffner S, Ratner R, Marcovina S, et al.; Diabetes Prevention Program Research Group. The effect of metformin and intensive lifestyle intervention on the metabolic syndrome: the Diabetes Prevention Program randomized trial. Ann Intern Med. 2005;142:611–9. 33. Diabetes Prevention Program Research Group. Changes in albumin excretion in the diabetes prevention program. Diabetes Care. 2009;32:720–5. 34. Diabetes Prevention Program Research Group. The prevalence of retinopathy in impaired glucose tolerance and recent-onset diabetes in the Diabetes Prevention Program. Diabet Med. 2007;24:137–44. 35. Carnethon MR, Prineas RJ, Temprosa M, Zhang ZM, Uwaifo G, Molitch ME; Diabetes Prevention Program Research Group. The association among autonomic nervous system 9 Interventional Trials to Prevent Diabetes: Diabetes Prevention Program 165

function, incident diabetes, and intervention arm in the Diabetes Prevention Program. Diabetes Care. 2006;29:914–9. 36. Brown JS, Wing R, Barrett-Connor E, Nyberg LM, Kusek JW, Orchard TJ, et al.; Diabetes Prevention Program Research Group. Lifestyle intervention is associated with lower prevalence of urinary incontinence: the Diabetes Prevention Program. Diabetes Care. 2006;29:385–90. 37. Florez JC, Jablonski KA, Bayley N, Pollin TI, de Bakker PI, Shuldiner AR, et al.; Diabetes Prevention Program Research Group. TCF7L2 polymorphisms and progression to diabetes in the Diabetes Prevention Program. N Engl J Med. 2006;355:241–50. 38. Grant SF, Thorleifsson G, Reynisdottir I, Benediktsson R, Manolescu A, Sainz J, et al. Variant of transcription factor 7-like 2 (TCF7L2) gene confers risk of type 2 diabetes. Nat Genet. 2006;38:320–3. 39. Florez JC, Jablonski KA, Sun MW, Bayley N, Kahn SE, Shamoon H, et al.; Diabetes Prevention Program Research Group. Effects of the type 2 diabetes-associated PPARG P12A polymorphism on progression to diabetes and response to troglitazone. J Clin Endocrinol Metab. 2007;92:1502–9. 40. Florez JC, Jablonski KA, Kahn SE, Franks PW, Dabelea D, Hamman RF, et al. Type 2 diabetes-associated missense polymorphisms KCNJ11 E23K and ABCC8 A1369S in fl uence progression to diabetes and response to interventions in the Diabetes Prevention Program. Diabetes. 2007;56:531–6. 41. Moore AF, Jablonski KA, McAteer JB, Saxena R, Pollin TI, Franks PW, et al.; Diabetes Prevention Program Research Group. Extension of type 2 diabetes genome-wide association scan results in the diabetes prevention program. Diabetes. 2008;5:2503–10. 42. Moore AF, Jablonski KA, Mason CC, McAteer JB, Arakaki RF, Goldstein BJ, et al.; Diabetes Prevention Program Research Group. The association of ENPP1 K121Q with diabetes incidence is abolished by lifestyle modi fi cation in the diabetes prevention program. J Clin Endocrinol Metab. 2009;94:449–55. 43. Jablonski KA, McAteer JB, de Bakker PI, Franks PW, Pollin TI, Hanson RL, et al.; Diabetes Prevention Program Research Group. Common variants in 40 genes assessed for diabetes incidence and response to metformin and lifestyle intervention in the diabetes prevention program. Diabetes. 2010;59:2672–81. 44. Diabetes Prevention Program Research Group. Effects of withdrawal from metformin on the development of diabetes in the diabetes prevention program. Diabetes Care. 2003;26:977–80. 45. Li G, Zhang P, Wang J, Gregg EW, Yang W, Gong Q, et al. The long-term effect of lifestyle interventions to prevent diabetes in the China Da Qing Diabetes Prevention Study: a 20-year follow-up study. Lancet. 2008;371:1783–9. 46. Lindström J, Ilanne-Parikka P, Peltonen M, Aunola S, Eriksson JG, Hemiö K, et al.; Finnish Diabetes Prevention Study Group. Sustained reduction in the incidence of type 2 diabetes by lifestyle intervention: follow-up of the Finnish Diabetes Prevention Study. Lancet. 2006;368:1673–9. 47. Diabetes Prevention Program Research Group. 10-Year follow-up of diabetes incidence and weight loss in the Diabetes Prevention Program Outcomes Study. Lancet. 2009;374:1677–86. 48. Benjamin SM, Valdez R, Vinicor F. Diabetes prevention. Author reply. N Engl J Med. 2002;346:1829–30. 49. Herman WH, Hoerger TJ, Brandle M, Hicks K, Sorensen S, Zhang P, et al.; Diabetes Prevention Program Research Group. The cost-effectiveness of lifestyle modi fi cation or metformin in preventing type 2 diabetes in adults with impaired glucose tolerance. Ann Intern Med. 2005;142:323–32. 50. Ratner RE; Diabetes Prevention Program Research Group. An update on the Diabetes Prevention Prgoram. Endocr Pract. 2006;12 Suppl 1:20–4. 51. Diabetes Prevention Program Research Group. Within-trial cost-effectiveness of lifestyle intervention or metformin for the primary prevention of type 2 diabetes. Diabetes Care. 2003;26:2518–23. 52. Herman WH, Edelstein SL, Ratner RE, Montez MG, Ackermann RT, Orchard TJ, et al.; The Diabetes Prevention Program Research Group. The 10-year cost-effectiveness of lifestyle 166 V.R. Aroda and R.E. Ratner

intervention or metformin for the primary prevention of type 2 diabetes mellitus: an intent-to-treat analysis of diabetes prevention. American Diabetes Association 71st Scienti ﬁ c Sessions; 2011. 53. American Diabetes Association. Standards of medical care in diabetes—2011. Diabetes Care. 2011;4:S11–61. 54. Nathan DM, Davidson MB, DeFronzo RA, Heine RJ, Henry RR, Pratley R, et al.; American Diabetes Association. Impaired fasting glucose and impaired glucose tolerance: implications for care. Diabetes Care. 2007;30:753–9. 55. Garber AJ, Handelsman Y, Einhorn D, Bergman DA, Bloomgarden ZT, Fonseca V, et al. Diagnosis and management of prediabetes in the continuum of hyperglycemia: when do the risks of diabetes begin? A consensus statement from the American College of Endocrinology and the American Association of Clinical Endocrinologists. Endocr Pract. 2008;14:933–46. 56. Rosenzweig JL, Ferrannini E, Grundy SM, Haffner SM, Heine RJ, Horton ES, et al. Primary prevention of cardiovascular disease and type 2 diabetes in patients at metabolic risk: an endocrine society clinical practice guideline. J Clin Endocrinol Metab. 2008;93:3671–89. 57. Canadian Diabetes Association Clinical Practice Guidelines Expert Committee. Canadian Diabetes Association for the prevention and management of diabetes in Canada. Can J Diabetes. 2008;32 Suppl 1:S17–9. 58. Indian Health Service. Guidelines for care of adults with pre-diabetes and/or the metabolic syndrome in clinical settings. 2008. http://www.ihs.gov/MedicalPrograms/Diabetes/HomeDocs/ Tools/ClinicalGuidelines/PreDiabetes_Guidelines_0209.pdf. Accessed 19 April 2011. 59. U.S. Preventive Services Task Force. Screening for type 2 diabetes mellitus in adults: U.S. Preventive Services Task Force recommendation statement. Ann Intern Med. 2008;148:846–54. 60. Twigg SM, Kamp MC, Davis TM, Neylon EK, Flack JR; Australian Diabetes Society; Australian Diabetes Educators Association. Prediabetes: a position statement from the Australian Diabetes Society and Australian Diabetes Educators Association. Med J Aust. 2007;186:461–5. 61. Alberti KG, Zimmet P, Shaw J. International diabetes federation: a consensus on Type 2 diabetes prevention. Diabet Med. 2007;24:451–63. 62. Patient Protection and Affordable Care Act, Pub. L. No. 111-148, §10409, 124 Stat. 119, 978; 2010. 63. Centers for Disease Control and Prevention. National Diabetes Prevention Program. 2011. http://www.cdc.gov/diabetes/projects/prevention_program.htm . Accessed 16 Feb 2011. 64. Ackermann RT, Finch EA, Brizendine E, Zhou H, Marrero DG. Translating the Diabetes Prevention Program into the community. The DEPLOY Pilot Study. Am J Prev Med. 2008; 35:357–63. 65. National Institutes of Health/National Institute of Diabetes and Digestive and Kidney Diseases. Advances and emerging opportunities in diabetes research: a strategic planning report of the DMICC, Final Plan. 2011. http://conferences.thehillgroup.com/DRSP/PDF_508/DRAFT_ DSP_Full02.22.10.pdf . Accessed 16 Feb 2011. Chapter 10 Decreasing Postprandial Plasma Glucose Using an a -Glucosidase Inhibitor in Subjects with IGT for the Prevention of Type 2 Diabetes Mellitus: The STOP-NIDDM Trial

Jean-Louis Chiasson , Markku Laakso , and Markolf Hanefeld

Introduction

We are currently witnessing a worldwide explosion in the prevalence of type 2 diabetes mellitus [1 ] . Because of the high morbidity and the excess mortality associated with diabetes, it has major societal implications [ 2, 3] . Diabetes still remains the most common cause of blindness, end-stage renal disease and non-traumatic amputation and a major cause of cardiovascular disease (CVD) [ 4, 5] . Consequently, it has a strong impact on healthcare cost [ 6 ] . Given the magnitude of the problem, type 2 diabetes is one of the major challenges of the twenty- ﬁ rst century. The only way that we can curtail this ever-growing problem is by developing and implementing prevention strategies. The concept for the prevention of type 2 diabetes has grown from a better understanding of the pathophysiology of the disease. Although a number of susceptibility genes are involved in the development of type 2 diabetes [ 7, 8] , it is usually precipi- tated by a number of environmental determinants such as sedentary lifestyle, nutritional over-indulgence and obesity [9– 11 ] . All these factors contribute to the development of insulin resistance, one of the major metabolic impairments leading to diabetes [ 12, 13] . Normal glucose tolerance will be maintained as long as the b cells can compensate for insulin resistance; however, glucose intolerance will

J.-L. Chiasson, MD (*) Department of Medicine , Centre de recherche du Centre hospitalier de l’Université de Montréal (CRCHUM), Université de Montréal , 3850 St. Urbain Street, Room 8-202 , Montréal , QC , Canada H2W 1T8 e-mail: [email protected] M. Laakso, MD, PhD Department of Medicine , University of Eastern Finland , Kuopio , Finland M. Hanefeld, MD, PhD Centre for Clinical Studies , Dresden , Germany

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 167 DOI 10.1007/978-1-4614-3314-9_10, © Springer Science+Business Media New York 2012 168 J.-L. Chiasson et al. appear when the b cells fail to fully compensate for insulin resistance [13, 14 ] . It will generally manifest itself at ﬁ rst as impaired glucose tolerance (IGT) characterized by postprandial hyperglycaemia and/or impaired fasting glucose (IFG), which are now recognized as prediabetic states [ 15 ] . Although prediabetes is generally not associated with diabetes-speci ﬁ c complications, the DPP showed in a random sample of their study population that retinopathy was detected in 12.6% of those who converted to diabetes and in 7.9% of those who remained IGT [ 16, 17 ] . However, it is generally recognized that prediabetes, particularly IGT, is associated with an increased risk of CVD [18– 20 ] .

Postprandial Hyperglycaemia as a Risk Factor for Diabetes and CVD

The relatively mild postprandial hyperglycaemia characterizing IGT is believed to be suffi cient to induce glucotoxicity and further decrease insulin action and insulin response to a glucose challenge, thus accelerating the progression of IGT to diabetes [21 ] . We do recognize that the concept of glucotoxicity is based more on animal than on human data. However, the observation that tight metabolic control in human diabetics, independent of the method by which it is achieved, leads to improvement in both insulin secretion and insulin sensitivity supports this hypothesis [ 22– 26 ] . Most of the animal data is derived from the 90% pancreatectomy rat model. The advantages of this model are that the b cells in the remaining 10% of the pancreas are known to be normal and that the resulting hyperglycaemia is relatively mild [ 27] . Using this model, Rossetti et al. [ 28 ] have shown that partial pancreatectomy resulted in moderate glucose intolerance that was associated with a signifi cant reduction in insulin sensitivity. Furthermore, correction of hyperglycaemia with phlorizin normalized tissue sensitivity to insulin. Using the same model, a number of studies have shown that chronic hyperglycaemia resulted in impaired insulin response to a hyperglycaemic clamp, and normalization of plasma glucose completely corrected the insulin secretion defect [27, 29, 30 ] . The mechanism(s) through which postprandial hyperglycaemia can exert its toxic effects is still controversial. However, much data have accumulated suggesting that it may be related to the production of reactive oxygen species (ROS) induced by hyperglycaemia, particularly the rise in plasma glucose following a meal containing carbohydrates [31 ] . It has been well documented that markers of oxidative stress were increased in correlation with plasma glucose in diabetic subjects, and that these markers improved under intensive glycaemic treatment [ 32, 33 ] . Furthermore, a hyperglycaemic response to an oral glucose tolerance test (OGTT) was shown to be associated with a reduction in anti-oxidant capacity in both healthy and type 2 diabetic subjects [34 ] . Also in healthy subjects, a 2-h hyperglycaemic clamp was associated with a 2.5-fold increase over baseline in nitrotyrosine, a marker of oxidative stress [ 35 ] . Ceriello et al. [36 ] have shown 10 Decreasing Postprandial Plasma Glucose Using an a-Glucosidase Inhibitor... 169 that both a high carbohydrate load (75 g glucose) and a high fat load induced a rise in nitrotyrosine, and that a mixed meal containing both carbohydrate and fat resulted in a cumulative effect on the postprandial rise in the production of oxidative stress markers. It is known that in sedentary obese non-diabetic subjects, the associated elevated FFA levels contribute to the impairment in insulin secretion [ 24, 37– 39 ] and insulin action [ 40, 41 ] , major defects leading to the development of type 2 diabetes. It can therefore be postulated that in genetically predisposed subjects, the oxidative stress induced by hyperlipidemia could lead to decreased insulin secretion and action, resulting in the development of IGT characterized by postprandial hyperglycaemia. This moderate postprandial hyperglycaemia will then exacerbate the resulting ROS production and further contribute to the deterioration of insulin secretion and action, thus accelerating the progression to diabetes [42– 46 ] . There is also evidence that postprandial hyperglycaemia and hyperlipidemia are involved in the development of atherosclerosis in both diabetic and non-diabetic subjects [ 47 ] . A number of prospective observational studies and meta-analyses have confi rmed the relationship between postprandial hyperglycaemia and cardiovascular events and mortality [48– 51 ] . LDL cholesterol oxidation increases after meals and has been shown to be directly related to the degree of postprandial hyperglycaemia [52 ] . The postprandial state is also associated with subclinical in fl ammation, augmented coagulation activity, increased expression of adhesion molecules and endothelial dysfunction [53– 56 ] . Endothelial dysfunction observed in subjects with IGT in response to a 75 g OGTT is an early step in atherosclerosis. This is further supported by the observation that the 2-h plasma glucose during an OGTT is an independent predictor of the intima media thickness (IMT) in nondiabetic subjects, an accepted surrogate marker of atherosclerosis [ 57 ] .

Acarbose for the Treatment of Postprandial Hyperglycaemia

Acarbose is a competitive inhibitor of the a -glucosidases of the brush border of the small intestine, enzymes that are necessary for the hydrolysis of the disaccharides, oligosaccharides and polysaccharides to monosaccharides for absorption [58 ] . Acarbose is a pseudo-tetrasaccharide of microbial origin. Its chemical structure closely resembles that of an oligosaccharide obtained by digestion of starch. The active part of the molecule is an acarvosine unit linked to a maltose unit. This nitrogen link of the acarvosine unit confers to the compound its high af fi nity for the glucosyl site of the different a -glucosidases, which is more than 10,000– 100,000 times that of the common oligosaccharides from carbohydrates ingestion. Because of this nitrogen link, acarbose cannot be hydrolyzed by the enzyme. However, its high affi nity binding is reversible and has the kinetics of a competitive inhibitor [ 58 ] . Carbohydrates are usually absorbed rapidly in the proximal portion of the small intestine. The use of acarbose with meal will therefore slow the digestion and the 170 J.-L. Chiasson et al. absorption of carbohydrates such that they will be hydrolyzed and absorbed throughout the whole length of the small intestine. This will attenuate the rise in postprandial plasma glucose in a dose-dependent manner, and consequently moderate the rise in postprandial plasma insulin [59 ] . Because of its specifi city for a-glucosidases, acarbose has no effect on the b -glucosidases such as lactase. The digestion and absorption of milk (lactose) is therefore not affected. More importantly, the intestinal absorption of monosaccharides such as glucose is not affected. The acarbose molecule is virtually non- absorbed (less than 1–2%) and can therefore exert its inhibiting effect throughout the small intestine up to the ileum. However, bacterial enzymes in the colon cleave it into a number of metabolites, of which 35% are found in the urine [58 ] . Acarbose has to be present at the site of enzymatic activity at the same time as the carbohydrates to exert its competitive inhibitory activity on the disaccharides and oligosaccharides in the small intestine. Consequently, the drug has to be taken after the fi rst bite of the meal and not later than 15 min after the beginning of the meal [60 ] . By delaying the digestion and the absorption of carbohydrates in the small intestine, acarbose can increase the amount of carbohydrates reaching the colon where they can be fermented and induce gastro-intestinal symptoms such as fl atulence and occasionally diarrhoea. These side effects can be prevented or minimized by starting at low dose and titrating very slowly to a maximal dose of 100 mg with each meal [61 ] . Numerous studies including a Cochrane meta-analysis have shown that acarbose was effective in treating postprandial hyperglycaemia and in reducing HbA 1c in subjects with type 2 diabetes, whether it was used as monotherapy or in combination with other antidiabetic medications [62– 66 ] . In fact, even when other oral hypogly- caemic agents have failed, acarbose is still effective in improving glycaemic control. Therefore, acarbose is an effective drug in reducing postprandial hyperglycaemia and postprandial hyperinsulinemia, resulting in improved glycaemic control [ 58 ] . By its mechanism of action, acarbose decreases the glucose stress on b cells. By reducing postprandial hyperglycaemia, it decreases glucose toxicity and improves insulin sensitivity in subjects with IGT and diabetes [67, 68 ] . Furthermore, acarbose has been shown to reduce postprandial rise in markers of oxidative stress, of in fl ammation, of activated coagulation and to reduce postprandial endothelial dysfunction in animal models as well as in human subjects with IGT or diabetes [ 69– 73 ] . Some studies have reported that acarbose given with a carbohydrate-rich meal was associated with an increase in GLP-1 [ 74, 75] . Acarbose has also been shown to increase adiponectin in subjects with type 2 diabetes and to reduce the size of myocardial infarctions in animal models [ 72, 76, 77] . In some studies with type 2 diabetes, acarbose treatment has been shown to be associated with a reduction in postprandial hypertriglyceridemia, systolic blood pressure and body weight [ 78– 82 ] . Thus, it has bene fi cial effects on major components of metabolic syndrome. All these observations support a potential role for acarbose in the prevention of type 2 diabetes mellitus and CVD. 10 Decreasing Postprandial Plasma Glucose Using an a-Glucosidase Inhibitor... 171

The STOP-NIDDM Trial: The Study Design

The STOP-NIDDM Trial was an international study including Austria, Canada, Denmark, Finland, Germany, Israel, Norway, Spain and Sweden. Subjects with IGT were recruited from a high-risk population based on fasting plasma glucose between 5.6 and 7.8 mmol/L and a 2-h plasma glucose post-75 g glucose between 7.8 and 11.0 mmol/L inclusively. During the study, the fasting plasma glucose criterion for the diagnosis of diabetes was lowered to 7.0 mmol/L, such that 10% of our population would have been considered diabetic based on that single measurement. Overall, 1,429 subjects with IGT were randomized in a double-blind fashion to either placebo or acarbose starting at 50 mg once a day and titrated gradually to 100 mg three times a day with meals. At entry, all subjects were advised on a weight maintaining or reducing diet and were encouraged to exercise regularly. The primary objective was to determine the frequency of the development of diabetes based on a single 75 g OGTT performed yearly; however, at the 3-month visits, fasting plasma glucose was performed and if the level was ³7.0 mmol/L, the subject was scheduled for an OGTT. Secondary objectives included conversion of IGT to normal glucose tolerance, and quantify new cases of hypertension and cardiovascular events as well as changes in anthropometric measurements, blood pressure, lipid pro fi le and HbA 1c . Blood pressure and dyslipidemia were treated according to local guidelines. The required sample size was calculated using a two-tailed a of 0.05 and a 1-b of 90% assuming an annual conversion rate of 7%, a 36% reduction in the acarbose- treated group and a 10% dropout rate. It was calculated that 600 subjects needed to be randomized per treatment group. Analysis on the intent-to-treat population was done using the Cox proportional hazards model including covariates such as treatment and any baseline variable that could in fl uence outcome. Interim analysis was done every 6 months after the fi rst year of follow-up by an independent Data Safety and Quality Review Committee. ECG readings and evaluation of cardiovascular events were done by independent cardiologists blinded to treatment. The subjects were seen every 3 months by the coordinating nurse and every 6 months by the investigator for a median follow-up of 3.3 years. At the end of the treatment period, all subjects who had not converted to diabetes were put on placebo in a single-blind fashion for a 3-month washout period at the end of which the outcome measures were repeated.

The STOP-NIDDM Trial: The Prevention of Diabetes

Overall, 1,429 subjects were randomized to acarbose (n = 714) or placebo (n = 715) [ 83 ] . We excluded 17 subjects (8 on acarbose and 9 on placebo) because they did not meet the criteria for IGT and another 44 because they had no valid post-randomization data, leaving 1,368 subjects (682 on acarbose and 686 on placebo) for analysis (Fig. 10.1). Altogether, 24.6% discontinued treatment prematurely, but these were maintained in the intent to treat population. 172 J.-L. Chiasson et al.

Fig. 10.1 Trial pro ﬁ le of the STOP-NIDDM trial (from Chiasson et al. [83 ] , with permission)

The baseline characteristics of the intent to treat population are listed in Table 10.1 . Men and women were equally represented (49 and 51% respectively) with a mean age of 54.5 years, a mean BMI of 31 kg/m 2 and a mean waist circumference of 102.2 cm. The mean fasting plasma glucose was 6.24 mmol/L and the mean 2-h plasma glucose 9.26 mmol/L. One hundred and thirty- ﬁ ve subjects (9.7%) had a fasting plasma glucose ³ 7.0 mmol/L but <7.8 mmol/L. Forty-six per cent had hypertension, 58% dyslipidemia and 61% had metabolic syndrome according to the NCEP-ATP III de ﬁ nition [84 ] . These patients were thus at very high risk of developing diabetes and CVD. Based on a single OGTT, the cumulative incidence of diabetes over 3.3 years was 221 (32.4%) in the acarbose-treated group compared to 285 (41.2%) in the placebo group. Figure 10.2 illustrates the effect of acarbose vs. placebo on the probability of remaining free of diabetes over time. Decreasing postprandial glucose with acarbose therefore resulted in a relative reduction of 25% and an absolute reduction of 8.8% in the risk of progressing to diabetes (p = 0.0015). These results suggest that 11 patients with IGT would have to be treated (NNT) for 3.3 years to prevent one case of diabetes. Two years after the beginning of the study, the diagnostic criteria 10 Decreasing Postprandial Plasma Glucose Using an a-Glucosidase Inhibitor... 173

Table 10.1 Baseline characteristics of intent-to-treat population Acarbose ( n = 682) Placebo ( n = 686) Gender Male 329 (48%) 344 (50%) Female 353 (52%) 342 (50%) Age (years) 54.3 (7.9) 54.6 (7.9) White 664 (97%) 670 (98%) Weight (kg) 87.6 (15.3) 87.1 (14.1) Body mass index (kg/m 2 ) 31.0 (4.3) 30.9 (4.2) Waist circumference (cm) 102.1 (11.7) 102.2 (11.2) Plasma glucose (mmol/L) Fasting 6.23 (0.50) 6.24 (0.53) 2-h 9.26 (1.06) 9.25 (1.01) Plasma insulin (pmol/L) Fasting 99.34 (57.64) 98.13 (56.78) 2-h 606.37 (437.46) 597.99 (414.38) Serum lipids (mmol/L) Total cholesterol 5.76 (1.04) 5.61 (0.99) HDL-cholesterol 1.19 (0.32) 1.17 (0.33) LDL-cholesterol 3.66 (0.91) 3.54 (0.90) Triglycerides 2.07 (1.10) 2.07 (1.17) Blood pressure (mmHg) Systolic 131.4 (16.3) 130.9 (16.2) Diastolic 82.8 (9.4) 82.0 (9.3) Smoking (%) 79 (12%) 99 (14%) Data are mean (SD) or number (%)

Fig. 10.2 Effect of acarbose and placebo on the cumulative probability of remaining free of diabetes over time in subjects with impaired glucose tolerance (IGT) (from Chiasson et al. [ 83 ] , with permission) 174 J.-L. Chiasson et al.

Fig. 10.3 Effect of acarbose on the development of diabetes in subjects with IGT according to age, gender and BMI (from Chiasson et al. [83 ] , with permission) for the diagnosis of diabetes changed: the fasting plasma glucose was lowered to 7.0 mmol/L, and whichever criterion was used had to be confi rmed on a separate day. Accordingly, if we use a fasting plasma glucose ³ 7.0 mmol/L on two consecutive visits as the criterion, 117 (17%) patients developed diabetes in the acarbose group compared to 178 (26%) in the placebo group resulting in an absolute reduction of 8.7% and a relative reduction of 32.4% ( p = 0.001). Similarly, if we used two positive OGTTs, 105 (15%) patients converted to diabetes in the acarbose group and 165 (24%) in the placebo group for an absolute reduction of 8.7% and a relative reduction of 36.4% (p = 0.0003). No matter which criteria were used for the diagnosis, the absolute risk reduction was essentially the same, 8.7%. Figure 10.3 illustrates the analysis using the Cox proportional-hazards model adjusted for age, gender and BMI. Interestingly, acarbose was still effective in reducing the risk of diabetes whether the subjects were older or younger, male or female, or had a higher or lower BMI. However, it was particularly effective in the elderly, in moderately overweight subjects and in women. Furthermore, acarbose treatment was associated with an increase in the conversion of IGT to normal glucose tolerance [hazard ratio 1.42 (95% CI: 1.24–1.62); p = 0.0001]. At the end of the treatment period, all subjects were put on placebo in a single- blind fashion for another 3-month period after which all outcome variables were measured. During this short period, the benefi t of acarbose on the conversion of IGT to diabetes totally disappeared. It is obvious that any intervention for the prevention of diabetes has to be maintained inde fi nitely to remain effective [83, 85, 86 ] . The most common side effects to acarbose treatment were gastrointestinal symptoms, mainly fl atulence and diarrhoea. But these were judged to be mild to moderate in severity and mostly disappeared over time. In a secondary analysis, we evaluated the impact of single traits and overall metabolic syndrome on the conversion of IGT to diabetes [87 ] . The prevalence of metabolic syndrome in the STOP-NIDDM population was 61% based on the NCEP-ATP III de fi nition [84 ] . Multivariate analysis revealed that the treatment group, 2-h plasma glucose, triglycerides and leukocyte counts were all independent predictors 10 Decreasing Postprandial Plasma Glucose Using an a-Glucosidase Inhibitor... 175

Fig. 10.4 The effects of acarbose on the probability of remaining free of diabetes in subjects with IGT and its modulation by the presence of the metabolic syndrome (from Hanefeld et al. [ 87 ] , with permission)

of diabetes. In the placebo group, the annual incidence of diabetes was 18.7% vs. 11.2% in subjects with and without metabolic syndrome respectively. In the acarbose-treated group, the corresponding incidence rates were 13.5% vs. 9.4% respectively. Subjects with metabolic syndrome treated with acarbose had the same risk for diabetes as those on placebo without metabolic syndrome (Fig. 10.4 ). Interestingly, the number needed to treat to prevent one new case of diabetes was 5.8 in patients with metabolic syndrome compared to 16.5 in those without the metabolic syndrome [87 ] . The absolute ef ﬁ cacy of acarbose in patients with metabolic syndrome was similar to that found in prevention trials using lifestyle intervention [ 88, 89 ] . We also looked at the effects of a number of candidate gene polymorphisms on the incidence of diabetes in a sub-group ( n = 770) of the STOP-NIDDM population (Table 10.2 ) [90– 94 ] . With the exception of the 3 ¢ UTR polymorphism of the leptin receptor gene and the Pro12Ala of the PPARg gene, all other single nucleotide polymorphisms (SNPs) studied increased the risk for diabetes. Three of the polymorphisms were associated with gender differences; women carrying the combination of the G-allele of SNP +45 and the T-allele of SNP +276 of the adiponectin gene had an especially high risk of developing diabetes (odds ratio 22.2%, 95% CI 2.7– 183.3) [ 90] . Interestingly, acarbose treatment completely neutralized the effect of 176 J.-L. Chiasson et al.

Table 10.2 Impact of candidate gene polymorphisms on the conversion of IGT to diabetes and to normal glucose tolerance, on weight loss and on the effect of acarbose on the incidence of diabetes IGT Acarbose Weight Gene Polymorphisms to diabetes effect lost IGT to NGT PPAR g Pro12A1a – ↑♀ – – PGC-1 a Gly482Ser ↑ ×1.6 ↑ – – Hepatic lipase Ala250Ala ↑ ×2.74 – – ↓ Adiponectin SNP + 45G ↑ ×1.8 – – – SNP +276T ↑ ×4.5 – – – + 45G +276T ↑ ×22.2 ↑ ↑ – PPAR d Crs6902123 ↑ ×2.7 – – – + Gly482Ser ↑ ×2.5 ↑ – – + Pro12Pro ↑ ×3.9♀ – – – HNF4 a Rs4810424 ↑ ×1.7♀ – – – Leptin receptor gene 3¢ UTR – – ↑ – ♀Female risk genotypes of the adiponectin gene on the risk of type 2 diabetes. Three other genotypes were associated with an increase in the effect of acarbose on the prevention of diabetes: SNPs of the PPAR d , Gly482Ser of the PGC-1 a and Pro12Ala of the PPARg gene [92 ] . Two of the SNPs, 3¢ UTR of the leptin receptor gene and the combination SNP +45 and SNP +276 of the adiponectin gene, were associated with increased weight loss [90, 94 ] . Finally, the Gly250Ala polymorphism of the hepatic lipase gene was associated with an increased risk of diabetes and a reduction in the conversion of IGT to normal glucose tolerance [ 91 ] . Many polymorphisms are associated with an increased risk of conversion of IGT to diabetes and some modulate the effect of acarbose on the prevention of diabetes.

The STOP-NIDDM Trial: The Prevention of Cardiovascular Disease

CVD remains the leading cause of death in type 2 diabetes mellitus, accounting for 40–50% of all deaths [95 ] . In these patients, there is a two- to tenfold increase in mortality risk from coronary heart disease, cerebrovascular disease and peripheral vascular disease [ 96– 98] . Though type 2 diabetes is generally associated with other cardiovascular risk factors such as hypertension and dyslipidemia [ 99, 100 ] , it is believed that hyperglycaemia per se, particularly postprandial hyperglycaemia, is an independent risk factor for CVD in both diabetic and non-diabetic subjects [ 48, 51, 100, 101] . It has been acknowledged that macrovascular disease starts years before the development of diabetes [ 102 ] . Many studies have now conﬁ rmed that prediabetes, particularly IGT, is associated with increased risk of CVD even after adjusting for other classical risk factors [ 19, 103– 106] . Even a moderate increase in postprandial plasma glucose was a strong predictor of atherosclerosis [57, 107– 111 ] . 10 Decreasing Postprandial Plasma Glucose Using an a-Glucosidase Inhibitor... 177

Fig. 10.5 Effect of acarbose on the development of cardiovascular events in subjects with IGT (from Chiasson et al. [112 ] , with permission)

Evidence supports the hypothesis that postprandial hyperglycaemia may be linked to CVD through the generation of oxidative stress [31 ] . In the STOP-NIDDM Trial, 47 subjects had at least one cardiovascular event; 15 were in the acarbose-treated group compared to 32 in the placebo group (Fig. 10.5 ) [ 112] . All the cardiovascular events were evaluated and confi rmed by an independent committee of cardiologists blinded to treatment. Acarbose treatment was associated with a 49% reduction in any cardiovascular event with a hazard ratio of 0.51 (95% CI 0.28–0.95). Even myocardial infarction, despite the small number of events, was signifi cantly reduced by acarbose with a hazard ratio of 0.09 (95% CI 0.01–0.72). Though all other cardiovascular events were not reduced signi fi cantly, they were all favourably affected by acarbose. Figure 10.6 shows the probability of developing CVD over time, with the two groups starting to separate after 1½ years in favour of acarbose (p = 0.04 by log rank test and 0.03 by Cox proportional- hazards model). All subjects had an ECG done at randomization and at the end of the study. The reading of the ECGs by independent cardiologists blinded to treatment revealed 8 silent myocardial infarctions which had not been identi fi ed clinically, 1 was in the acarbose-treated group and 7 in the placebo group for a total of 2 myocardial infarctions under acarbose treatment and 19 under placebo (p = 0.0002 by Chi square analysis) [113 ] . Furthermore, in a sub-group of patients from the STOP-NIDDM trial ( n = 132), the IMT of the carotids, an accepted surrogate of atherosclerosis, was measured by B-mode ultrasound before randomization and at the end of the study [ 114 ] . The mean annual IMT increase was 0.02 (SD 0.07) mm in the acarbose group vs. 0.05 (0.06) mm in the placebo group ( p = 0.027). The annual increase in IMT was therefore reduced by 50% with acarbose treatment compared to placebo. The STOP-NIDDM trial also showed that treating postprandial hyperglycaemia with 178 J.-L. Chiasson et al.

Fig. 10.6 Effect of acarbose on the probability of developing cardiovascular disease (CVD) in subjects with IGT (from Chiasson et al. [112 ] , with permission)

acarbose over 3 years was associated with a reduction of other cardiovascular risk factors that are part of metabolic syndrome such as excess body weight, dyslipidemia and hypertension. Anthropometric measurements were all favourably affected by acarbose treatment. Over 3 years, body weight decreased by 1.2 kg under acarbose compared to an increase of 0.3 under placebo; the 1.4 kg difference was signi fi cant at p < 0.001 [112 ] . Acarbose was associated with a reduction in body mass index of 0.60 kg/m2 compared to 0.12 kg/m2 under placebo (p < 0.001). Finally, waist circumference decreased by 0.6 cm under acarbose treatment vs. a reduction of 0.2 under placebo ( p = 0.001). The lipid profi le was also affected favourably by acarbose treatment. Triglycerides decreased by 0.18 mg/dL in the acarbose treatment group compared to 0.04 mg/dL in the placebo group (p = 0.01). Acarbose signi fi cantly reduced the mean systolic (−0.92 mmHg) and diastolic (−1.4 mmHg) blood pressure compared to placebo ( p < 0.001). But more importantly, acarbose signi fi cantly reduced the incidence of new cases of hypertension. In subjects normotensive at baseline ( n = 666), 96 developed hypertension based on the most recent criteria ( ³140/90 mmHg). Acarbose treatment in subjects with IGT resulted in a relative reduction in the incidence of hypertension of 41% [ 115 ] (Fig. 10.7). The STOP- NIDDM trial is the fi rst prospective intervention study in subjects with IGT showing that an a-glucosidase inhibitor was associated with a signifi cant reduction in cardiovascular events and in the incidence of hypertension. All these observations lend further support to the hypothesis that postprandial hyperglycaemia is an independent risk factor for CVD. However, its cardiovascular effects still have to be 10 Decreasing Postprandial Plasma Glucose Using an a-Glucosidase Inhibitor... 179

Fig. 10.7 Effect of acarbose on the probability of remaining free of hypertension and its modulation by the presence of the metabolic syndrome (unpublished data) con ﬁ rmed in a well-designed and well-powered prospective study where the primary outcome will be CVD. The ongoing Acarbose Cardiovascular Evaluation (ACE) trial is such a study.

The STOP-NIDDM Trial: Cost-Effectiveness of Acarbose Treatment for the Prevention of Diabetes and CVD in Subjects with IGT

The high prevalence of prediabetes, IGT and IFG constitutes a huge population at high risk for the development of diabetes and CVD with a major impact on healthcare costs. Consequently, there is an emerging body of cost-effectiveness literature in the management of prediabetes. For acarbose, economic analyses have been done for Spain, Germany, Sweden and Canada within the STOP-NIDDM trial [116– 120 ] . Each of these within-trial analyses compared only the direct cost to the healthcare system in each country for acarbose vs. placebo, using the risk reduction in new cases of diabetes and cardiovascular events over the mean 3.3-year follow-up in the STOP-NIDDM Trial. Healthcare resource utilization data and unit costs were used to calculate the cost outcomes estimated by modelling techniques. The main economic outcome for all these analyses 180 J.-L. Chiasson et al. was the incremental cost per cases of diabetes averted and, for the German and Swedish analyses, the incremental cost per subject free of cardiovascular events [118 , 119 ] . The incremental cost ratios were obtained for the total STOP-NIDDM population and in groups at high risk for diabetes, CVD or both. The high-risk groups were deﬁ ned as the upper quartiles of validated risk scores for type 2 diabetes and CVD applied to individual patient data from the STOP-NIDDM Trial [121, 122 ] . All these within-trial analyses in each country demonstrated favourable cost- effectiveness results for acarbose treatment in the prevention of diabetes. In Germany, acarbose treatment was cost-effective for the total population at 772€ (2004) per cases of diabetes averted. However, acarbose was the preferred strategy compared to placebo for all high-risk groups [119 ] . In Spain, where the analysis was done only for the total population, acarbose treatment was estimated to be both cost- savings and improved outcomes [ 120 ] . In Sweden, acarbose treatment was estimated to be the dominant strategy vs. placebo in subjects at high risk for CVD and combined high risk for diabetes and CVD. For the total population and for those at high risk for diabetes, the incremental cost-effectiveness ratios were approximately 3,000 and 825€ (2005) respectively [118 ] . In Canada, the analysis was done using the Markov model over a 10-year follow-up [117 ] . Acarbose treatment was superior to placebo in regard to cost per life year gained. All these within-trial cost-effectiveness analyses of the STOP-NIDDM trial population suggest that acarbose treatment for the prevention of diabetes would be cost-effective. Therefore, investing in the prevention of diabetes would be a good investment for the future. The decision to invest in the prevention of diabetes implies that we have to invest in screening strategies. Screening everybody would not be cost-effective. However, there is general support for the screening of high-risk populations [ 123, 124 ] . A number of screening strategies based on opportunistic screening have been proposed based on fasting and/or OGTT [125, 126 ] . Strategies based on fasting plasma glucose only would miss 30–60% of subjects with IGT depending on the ethnic group, gender and age of the population [ 127– 129] . Although OGTT is the gold standard, it lacks reproducibility, is cumbersome, time-consuming and not without cost. More recently, a number of risk-score models have been developed and validated in different populations. Two simple risk scores have been developed and validated from the STOP-NIDDM trial data to predict the development of diabetes and cardiovascular events in individuals with IGT, and these scores can also be used to estimate the risk reduction with acarbose [ 130 ] . Such a screening procedure is recommendable. It provides a simple, inexpensive and sensitive test for the screening of subjects at high risk for diabetes.

Conclusion

Type 2 diabetes mellitus remains one of the major challenges of the twenty- ﬁ rst century because of its high and growing prevalence, its high morbidity, its excess mortality and its impact on healthcare costs. Our only hope to curtail this 10 Decreasing Postprandial Plasma Glucose Using an a-Glucosidase Inhibitor... 181 ever-growing problem is by developing and implementing prevention strategies. The STOP-NIDDM trial has addressed and validated the concept of postprandial plasma glucose as a risk factor for the development of diabetes and CVD. This international randomized controlled trial evaluated the effect of an a -glucosidase inhibitor, acarbose, on the incidence of diabetes and CVD in a high-risk population with IGT and a fasting plasma glucose ³ 5.6 mmol/L over a period of 3.3 years. Acarbose treatment resulted in a relative risk reduction of 25% when the diagnosis was based on a single OGTT and an absolute reduction of 8.8% (p = 0.0015). The number needed to treat to prevent one case of diabetes was 11. When the diagnosis was based on two OGTTs as now required, the relative risk reduction was 36.4% with an absolute reduction of 8.7% (p = 0.0003). Acarbose was even more effective in IGT subjects with metabolic syndrome where the NNT was 5.8 subjects to prevent the development of one case of diabetes. Acarbose also reduced most of the risk factors associated with metabolic syndrome: BMI, triglycerides and hypertension. Furthermore, acarbose treatment was associated with a 49% relative risk reduction of any cardiovascular events and 41% reduction of newly diagnosed hypertension. The major effect was on the reduction in myocardial infarctions, both those diagnosed clinically as well as the silent myocardial infarctions diagnosed on ECG reading. It was also shown in a sub-group of the STOP-NIDDM population that acarbose resulted in a 50% reduction in the progression of the IMT of the carotids. These observations are now being tested in a well-powered prospective study, the ACE trial. The use of acarbose for the prevention of diabetes and CVD has been shown to be cost-saving or cost-effective in the perspective of the healthcare system of most developed countries. Opportunistic screening for high-risk individuals using a simple, inexpensive and validated tool such as the STOP-NIDDM risk-score would make it even more cost-effective. The high and ever-growing prevalence of type 2 diabetes mellitus worldwide is exerting an enormous stress on the health of the populations and on the healthcare systems. We now know that type 2 diabetes can be prevented, or at least delayed. It is imperative that we implement prevention strategies to curtail this ever-growing problem. The STOP-NIDDM trial has shown that the use of acarbose was an effective and cost-saving strategy that should be considered for the prevention of diabetes.

References

1. Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care. 2004;27:1047–53. 2. Zimmet P, Alberti KG, Shaw J. Global and societal implications of the diabetes epidemic. Nature. 2001;414:782–7. 3. Amos AF, McCarty DJ, Zimmet P. The rising global burden of diabetes and its complications: estimates and projections to the year 2010. Diabet Med. 1997;14 Suppl 5:S1–85. 4. Klein R, Klein BEK, Moss SE. Relation of glycemic control to diabetic microvascular complications in diabetes mellitus. Ann Intern Med. 1996;124(1 Pt 2):90–6. 182 J.-L. Chiasson et al.

5. Fox CS, Pencina MJ, Wilson PW, Paynter NP, Vasan RS, D’Agostino Sr RB. Lifetime risk of cardiovascular disease among individuals with and without diabetes stratifi ed by obesity status in the Framingham heart study. Diabetes Care. 2008;31:1582–4. 6. American Diabetes Association. Economic costs of diabetes in the U.S. in 2007. Diabetes Care. 2008;31:596–615. 7. Saxena R, Voight BF, Lyssenko V, Burtt NP, de Bakker PI, Chen H, et al. Genome-wide association analysis identi fi es loci for type 2 diabetes and triglyceride levels. Science. 2007;316:1331–6. 8. Sladek R, Rocheleau G, Rung J, Dina C, Shen L, Serre D, et al. A genome-wide association study identi fi es novel risk loci for type 2 diabetes. Nature. 2007;445:881–5. 9. Hossain P, Kawar B, El Nahas M. Obesity and diabetes in the developing world—a growing challenge. N Engl J Med. 2007;356:213–5. 10. Dwyer T, Magnussen CG, Schmidt MD, Ukoumunne OC, Ponsonby AL, Raitakari OT, et al. Decline in physical fi tness from childhood to adulthood associated with increased obesity and insulin resistance in adults. Diabetes Care. 2009;32:683–7. 11. Arnlov J, Sundstrom J, Ingelsson E, Lind L. Impact of BMI and the metabolic syndrome on the risk of diabetes in middle-aged men. Diabetes Care. 2011;34:61–5. 12. Lillioja S, Bogardus C. Obesity and insulin resistance: lessons learned from the Pima Indians. Diabetes Metab Rev. 1988;4:517–40. 13. Schinner S, Scherbaum WA, Bornstein SR, Barthel A. Molecular mechanisms of insulin resistance. Diabet Med. 2005;22:674–82. 14. Kahn SE. The relative contributions of insulin resistance and beta-cell dysfunction to the pathophysiology of type 2 diabetes. Diabetologia. 2003;46:3–19. 15. Unwin N, Shaw J, Zimmet P, Alberti KG. Impaired glucose tolerance and impaired fasting glycaemia: the current status on de fi nition and intervention. Diabet Med. 2002;19:708–23. 16. Gavin III JR, Alberti KGMM, Davidson MB, DeFronzo RA, Drash AL, Gabbe SG, et al. Report of the expert committee on the diagnosis and classi fi cation of diabetes mellitus. Diabetes Care. 1997;20:1183–97. 17. Diabetes Prevention Program Research Group. The prevalence of retinopathy in impaired glucose tolerance and recent-onset diabetes in the diabetes prevention program. Diabet Med. 2007;24:137–44. 18. Eschwege E, Richard JL, Thibult N, Ducimetière P, Warnet JM, Claude JR, et al. Coronary heart disease mortality in relation with diabetes, blood glucose and plasma insulin levels. Horm Metab Res. 1985;15(Suppl Series):41–6. 19. Tominaga M, Eguchi H, Manaka H, Igarashi K, Kato T, Sekikawa A. Impaired glucose tolerance is a risk factor for cardiovascular disease, but not impaired fasting glucose. The Funagata Diabetes Study. Diabetes Care. 1999;22:920–4. 20. Bonora E. Postprandial peaks as a risk factor for cardiovascular disease: epidemiological perspectives. Int J Clin Pract Suppl. 2002;129:5–11. 21. Rossetti L, Giaccari A, DeFronzo RA. Glucose toxicity. Diabetes Care. 1990;13:610–30. 22. Garvey WT, Olefsky JM, Griffi n J, Hamman RF, Kolterman OG. The effect of insulin treatment on insulin secretion and insulin action in type II diabetes mellitus. Diabetes. 1985;34:222–34. 23. Kosaka K, Kuzuya T, Akanuma Y, Hagura R. Increase in insulin response after treatment of overt maturity onset diabetes mellitus is independent of the mode of treatment. Diabetologia. 1980;18:23–8. 24. Paolisso G, Tagliamonte MR, Rizzo MR, Gualdiero P, Saccomanno F, Gambardella A, et al. Lowering fatty acids potentiates acute insulin response in fi rst degree relatives of people with type II diabetes. Diabetologia. 1998;41:1127–32. 25. Kolterman OG, Gray RS, Shapiro G, Scarlett JA, Grif fi n J, Olefsky JM. The acute and chronic effects of sulfonylurea therapy in type II diabetic subjects. Diabetes. 1984;33:346–54. 26. Beck-Nielsen H, Richelsen B, Hasling C, Nielsen OH, Heding L, Sorensen NS. Improved in vivo insulin effect during continuous subcutaneous insulin infusion in patients with IDDM. Diabetes. 1984;33:832–7. 10 Decreasing Postprandial Plasma Glucose Using an a-Glucosidase Inhibitor... 183

27. Bonner-Weir S, Trent DF, Weir GC. Partial pancreatectomy in the rat and subsequent defect in glucose-induced insulin release. J Clin Invest. 1983;71:1544–53. 28. Rossetti L, Smith D, Shulman GI, Papachristou D, DeFronzo RA. Correction of hyperglycemia with phlorizin normalizes tissue sensitivity to insulin in diabetic rats. J Clin Invest. 1987;79:1510–5. 29. Zawalich WS, Zawalich KC, Shulman GI, Rossetti L. Chronic in vivo hyperglycemia impairs phosphoinositide hydrolysis and insulin release in isolated perfused rat islets. Endocrinology. 1990;126:253–60. 30. Rossetti L, Shulman GI, Zawalich W, DeFronzo RA. Effect of chronic hyperglycemia on in vivo insulin secretion in partially pancreatectomized rats. J Clin Invest. 1987;80:1037–44. 31. Ceriello A, Davidson J, Hanefeld M, Leiter L, Monnier L, Owens D, et al. Postprandial hyperglycaemia and cardiovascular complications of diabetes: an update. Nutr Metab Cardiovasc Dis. 2006;16:453–6. 32. Ceriello A, Giugliano D, Quatraro A, Dello RP, Lefebvre PJ. Metabolic control may infl uence the increased superoxide generation in diabetic serum. Diabet Med. 1991;8:540–2. 33. Berg TJ, Nourooz-Zadeh J, Wolff SP, Tritschler HJ, Bangstad H-J, Hanssen KF. Hydroperoxides in plasma are reduced by intensi fi ed insulin treatment. A randomized controlled study of IDDM patients with microalbuminuria. Diabetes Care. 1998;21:1295–300. 34. Ceriello A, Bortolotti N, Crescentini A, Motz E, Lizzio S, Russo A, et al. Antioxidant defences are reduced during the oral glucose tolerance test in normal and non-insulin-dependent diabetic subjects. Eur J Clin Invest. 1998;28:329–33. 35. Marfella R, Quagliaro L, Nappo F, Ceriello A, Giugliano D. Acute hyperglycemia induces an oxidative stress in healthy subjects. J Clin Invest. 2001;108:635–6. 36. Ceriello A, Quagliaro L, Piconi L, Assaloni R, Da Ros R, Maier A, et al. Effect of postprandial hypertriglyceridemia and hyperglycemia on circulating adhesion molecules and oxidative stress generation and the possible role of simvastatin treatment. Diabetes. 2004;53:701–10. 37. Lam TK, Carpentier A, Lewis GF, van de Werve G, Fantus IG, Giacca A. Mechanisms of the free fatty acid-induced increase in hepatic glucose production. Am J Physiol Endocrinol Metab. 2003;284:E863–73. 38. Carpentier A, Zinman B, Leung N, Giacca A, Hanley AJ, Harris SB, et al. Free fatty acid- mediated impairment of glucose-stimulated insulin secretion in nondiabetic Oji-Cree individuals from the Sandy Lake community of Ontario, Canada: a population at very high risk for developing type 2 diabetes. Diabetes. 2003;52:1485–95. 39. Carpentier A, Mittelman SD, Bergman RN, Giacca A, Lewis GF. Prolonged elevation of plasma free fatty acids impairs pancreatic beta-cell function in obese nondiabetic humans but not in individuals with type 2 diabetes. Diabetes. 2000;49:399–408. 40. Boden G. Role of fatty acids in the pathogenesis of insulin resistance and NIDDM. Diabetes. 1997;46:3–10. 41. McGarry JD. Banting lecture 2001: dysregulation of fatty acid metabolism in the etiology of type 2 diabetes. Diabetes. 2002;51:7–18. 42. Henriksen EJ. Exercise training and the antioxidant alpha-lipoic acid in the treatment of insulin resistance and type 2 diabetes. Free Radic Biol Med. 2006;40:3–12. 43. Ogihara T, Asano T, Katagiri H, Sakoda H, Anai M, Shojima N, et al. Oxidative stress induces insulin resistance by activating the nuclear factor-kappa B pathway and disrupting normal subcellular distribution of phosphatidylinositol 3-kinase. Diabetologia. 2004;47:794–805. 44. Evans JL, Gold fi ne ID, Maddux BA, Grodsky GM. Are oxidative stress-activated signaling pathways mediators of insulin resistance and beta-cell dysfunction? Diabetes. 2003;52:1–8. 45. Robertson RP, Harmon J, Tran PO, Poitout V. Beta-cell glucose toxicity, lipotoxicity, and chronic oxidative stress in type 2 diabetes. Diabetes. 2004;53 Suppl 1:S119–24. 46. Zraika S, Hull RL, Udayasankar J, Aston-Mourney K, Subramanian SL, Kisilevsky R, et al. Oxidative stress is induced by islet amyloid formation and time-dependently mediates amyloid-induced beta cell apoptosis. Diabetologia. 2009;52:626–35. 47. Ceriello A. The post-prandial state and cardiovascular disease: relevance to diabetes mellitus. Diabetes Metab Res Rev. 2000;16:125–32. 184 J.-L. Chiasson et al.

48. Coutinho M, Gerstein HC, Wang Y, Yusuf S. The relationship between glucose and incident cardiovascular events. A metaregression analysis of published data from 20 studies of 95,783 individuals followed for 12.4 years. Diabetes Care. 1999;22:233–40. 49. Qiao Q, Nakagami T, Tuomilehto J, Borch-Johnsen K, Balkau B, Iwamoto Y, et al. Comparison of the fasting and the 2-h glucose criteria for diabetes in different Asian cohorts. Diabetologia. 2000;43:1470–5. 50. Qiao Q, Tuomilehto J, Borch-Johnsen K. Post-challenge hyperglycaemia is associated with premature death and macrovascular complications. Diabetologia. 2003;46 Suppl 1:M17–21. 51. The DECODE Study Group. Glucose tolerance and cardiovascular mortality: comparison of fasting and 2-hour diagnostic criteria. Arch Intern Med. 2001;161:397–405. 52. Ceriello A, Bortolotti N, Motz E, Pieri C, Marra M, Tonutti L, et al. Meal-induced oxidative stress and low-density lipoprotein oxidation in diabetes: the possible role of hyperglycemia. Metabolism. 1999;48:1503–8. 53. Ceriello A, Assaloni R, Da Ros R, Maier A, Piconi L, Quagliaro L, et al. Effect of atorvastatin and irbesartan, alone and in combination, on postprandial endothelial dysfunction, oxidative stress, and in fl ammation in type 2 diabetic patients. Circulation. 2005;111:2518–24. 54. Ceriello A. Fibrinogen and diabetes mellitus: is it time for intervention trials? Diabetologia. 1997;40:731–4. 55. Ceriello A, Falleti E, Bortolotti N, Motz E, Cavarape A, Russo A, et al. Increased circulating intercellular adhesion molecule-1 levels in type II diabetic patients: the possible role of metabolic control and oxidative stress. Metabolism. 1996;45:498–501. 56. Kawano H, Motoyama T, Hirashima O, Hirai N, Miyao Y, Sakamoto T, et al. Hyperglycemia rapidly suppresses fl ow-mediated endothelium-dependent vasodilation of brachial artery. J Am Coll Cardiol. 1999;34:146–54. 57. Hanefeld M, Koehler C, Schaper F, Fuecker K, Henkel E, Temelkova-Kurktschiev T. Postprandial plasma glucose is an independent risk factor for increased carotid intima-media thickness in non-diabetic individuals. Atherosclerosis. 1999;144:229–35. 58. Bischoff H. The mechanism of a -glucosidase inhibition in the management of diabetes. Clin Invest Med. 1995;18:303–11. 59. Rabasa-Lhoret R, Chiasson J-L. Potential of a -glucosidase inhibitors in elderly patients with diabetes mellitus and impaired glucose tolerance. Drugs Aging. 1998;13:131–43. 60. Rosak C, Nitzsche G, König P, Hofmann U. The effect of the timing and the administration of acarbose on postprandial hyperglycaemia. Diabet Med. 1995;12:979–84. 61. Lebovitz HE. a -Glucosidase inhibitors as agents in the treatment of diabetes. Diabetes Rev. 1998;6:132–45. 62. Chiasson JL. a -Glucosidase inhibitors. In: Fonseca VA, editor. Excerpt from clinical diabetes translating research into practice. Toronto: Elsevier Canada; 2007. p. 321–30. 63. Chiasson J-L, Josse RG, Hunt JA, Palmason C, Rodger NW, Ross SA, et al. The effi cacy of acarbose in the treatment of patients with non-insulin-dependent diabetes mellitus. A multicenter controlled clinical trial. Ann Intern Med. 1994;121:928–35. 64. Hoffmann J, Spengler M. Effi cacy of 24-week monotherapy with acarbose, glibenclamide, or placebo in NIDDM patients. The Essen Study. Diabetes Care. 1994;17:561–6. 65. Hoffmann J, Spengler M. Ef fi cacy of 24-week monotherapy with acarbose, metformin, or placebo in dietary-treated NIDDM patients: the Essen-II Study. Am J Med. 1997;103:483–90. 66. van de Laar FA, Lucassen PL, Akkermans RP, van de Lisdonk EH, Rutten GE, van Weel C. Alpha-glucosidase inhibitors for patients with type 2 diabetes: results from a Cochrane systematic review and meta-analysis. Diabetes Care. 2005;28:154–63. 67. Chiasson J-L, Josse RG, Leiter LA, Mihic M, Nathan DM, Palmason C, et al. The effect of acarbose on insulin sensitivity in subjects with impaired glucose tolerance. Diabetes Care. 1996;19:1190–3. 68. Meneilly GS, Ryan EA, Radziuk J, Lau DCW, Yale J-F, Morais J, et al. Effect of acarbose on insulin sensitivity in elderly patients with diabetes. Diabetes Care. 2000;23:1162–7. 69. Inoue I, Shinoda Y, Nakano T, Sassa M, Goto S, Awata T, et al. Acarbose ameliorates athero- genecity of low-density lipoprotein in patients with impaired glucose tolerance. Metabolism. 2006;55:946–52. 10 Decreasing Postprandial Plasma Glucose Using an a-Glucosidase Inhibitor... 185

70. Lu JM, et al. The effect of acarbose on postprandial CRP levels in subjects with impaired glucose tolerance. Chin J Endocrinol Metab. 2003;19:254–6. 71. Ceriello A, Taboga C, Tonutti L, Giacomello R, Stel L, Motz E, et al. Post-meal coagulation activation in diabetes mellitus: the effect of acarbose. Diabetologia. 1996;39:469–73. 72. Ochiai H, Ooka H, Shida C, Ishikawa T, Inoue D, Okazaki R. Acarbose treatment increases serum total adiponectin levels in patients with type 2 diabetes. Endocr J. 2008;55:549–56. 73. Kato T, Inoue T, Node K. Postprandial endothelial dysfunction in subjects with new-onset type 2 diabetes: an acarbose and nateglinide comparative study. Cardiovasc Diabetol. 2010;9:12. 74. Seifarth C, Bergmann J, Holst JJ, Ritzel R, Schmiegel W, Nauck MA. Prolonged and enhanced secretion of glucagon-like peptide 1 (7-36 amide) after oral sucrose due to alpha-glucosidase inhibition (acarbose) in type 2 diabetic patients. Diabet Med. 1998;15:485–91. 75. Qualmann C, Nauck MA, Holst JJ, Orskov C, Creutzfeldt W. Glucagon-like peptide 1 (7-36 amide) secretion in response to luminal sucrose from the upper and lower gut. A study using alpha-glucosidase inhibition (acarbose). Scand J Gastroenterol. 1995;30:892–6. 76. Miyamura M, Schnell O, Yamashita C, Yoshioka T, Matsumoto C, Mori T, et al. Effects of acarbose on the acceleration of postprandial hyperglycemia-induced pathological changes induced by intermittent hypoxia in lean mice. J Pharmacol Sci. 2010;114:32–40. 77. Minatoguchi S, Zhang Z, Bao N, Kobayashi H, Yasuda S, Iwasa M, et al. Acarbose reduces myocardial infarct size by preventing postprandial hyperglycemia and hydroxyl radical production and opening mitochondrial KATP channels in rabbits. J Cardiovasc Pharmacol. 2009;54:25–30. 78. Hanefeld M, Fischer S, Schulze J, Spengler M, Wargenau M, Schollberg K, et al. Therapeutic potentials of acarbose as fi rst-line drug in NIDDM insuf fi ciently treated with diet alone. Diabetes Care. 1991;14:732–7. 79. Kado S, Murakami T, Aoki A, Nagasse T, Katsura Y, Noritake M, et al. Effect of acarbose on postprandial lipid metabolism in type 2 diabetes mellitus. Diabetes Res Clin Pract. 1998;41:49–55. 80. Hanefeld M, Cagatay M, Petrowitsch T, Neuser D, Petzinna D, Rupp M. Acarbose reduces the risk for myocardial infarction in type 2 diabetic patients: meta-analysis of seven long- term studies. Eur Heart J. 2004;25:10–6. 81. Wolever TMS, Chiasson J-L, Josse RG, Hunt JA, Palmason C, Rodger NW, et al. Small weight loss on long-term acarbose therapy with no change in dietary pattern or nutrient intake of individuals with non-insulin-dependent diabetes. Int J Obes. 1997;21:756–63. 82. Mertes G. Safety and effi cacy of acarbose in the treatment of type 2 diabetes: data from a 5-year surveillance study. Diabetes Res Clin Pract. 2001;52:193–204. 83. Chiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A, Laakso M. Acarbose for prevention of type 2 diabetes mellitus: the STOP-NIDDM randomised trial. Lancet. 2002;359:2072–7. 84. Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults. Executive summary of the third report of the National Cholesterol Education Program (NCEP) expert panel on detection, evaluation, and treatment of high blood cholesterol in adults (Adult Treatment Panel III). JAMA. 2001;285:2486–97. 85. The Diabetes Prevention Program Research Group. Effects of withdrawal from metformin on the development of diabetes in the diabetes prevention program. Diabetes Care. 2003;26:977–80. 86. Knowler WC, Hamman RF, Edelstein SL, Barrett-Connor E, Ehrmann DA, Walker EA, et al. Prevention of type 2 diabetes with troglitazone in the diabetes prevention program. Diabetes. 2005;54:1150–6. 87. Hanefeld M, Karasik A, Koehler C, Westermeier T, Chiasson JL. Metabolic syndrome and its single traits as risk factors for diabetes in people with impaired glucose tolerance: the STOP- NIDDM trial. Diab Vasc Dis Res. 2009;6:32–7. 88. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346:393–403. 89. Tuomilehto J, Lindstrom J, Eriksson JG, Valle TT, Hamalainen H, Ilanne-Parikka P, et al. Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med. 2001;344:1343–50. 186 J.-L. Chiasson et al.

90. Zacharova J, Chiasson JL, Laakso M. The common polymorphisms (single nucleotide polymorphism [SNP] +45 and SNP +276) of the adiponectin gene predict the conversion from impaired glucose tolerance to type 2 diabetes: the STOP-NIDDM trial. Diabetes. 2005;54:893–9. 91. Zacharova J, Todorova BR, Chiasson JL, Laakso M. The G-250A substitution in the promoter region of the hepatic lipase gene is associated with the conversion from impaired glucose tolerance to type 2 diabetes: the STOP-NIDDM trial. J Intern Med. 2005;257:185–93. 92. Andrulionyte L, Peltola P, Chiasson JL, Laakso M. Single nucleotide polymorphisms of PPARD in combination with the Gly482Ser substitution of PGC-1A and the Pro12Ala substitution of PPARG2 predict the conversion from impaired glucose tolerance to type 2 diabetes: the STOP-NIDDM trial. Diabetes. 2006;55:2148–52. 93. Andrulionyte L, Laukkanen O, Chiasson JL, Laakso M. Single nucleotide polymorphisms of the HNF4alpha gene are associated with the conversion to type 2 diabetes mellitus: the STOP-NIDDM trial. J Mol Med. 2006;84:701–8. 94. Zacharova J, Chiasson JL, Laakso M. Leptin receptor gene variation predicts weight change in subjects with impaired glucose tolerance. Obes Res. 2005;13:501–6. 95. de Marco R, Locatelli F, Zoppini G, Verlato G, Bonora E, Muggeo M. Cause-speci ﬁ c mortality in type 2 diabetes. The Verona Diabetes Study. Diabetes Care. 1999;22:756–61. 96. Stamler J, Vaccaro O, Neaton JD, Wentworth D. Diabetes, other risk factors, and 12-yr cardiovascular mortality for men screened in the Multiple Risk Factor Intervention Trial. Diabetes Care. 1993;16:434–44. 97. Manson JE, Colditz GA, Stampfer MJ, Willett WC, Krolewski AS, Rosner B, et al. A prospective study of maturity-onset diabetes mellitus and risk of coronary heart disease and stroke in women. Arch Intern Med. 1991;151:1141–7. 98. Laakso M. Hyperglycemia and cardiovascular disease in type 2 diabetes. Diabetes. 1999;48:937–42. 99. U.K.Prospective Diabetes Study Group. U.K. Prospective Diabetes Study 27. Plasma lipids and lipoproteins at diagnosis of NIDDM by age and sex. Diabetes Care. 1997;20:1683–7. 100. Hypertension in Diabetes Study (HDS): I. Prevalence of hypertension in newly presenting type 2 diabetic patients and the association with risk factors for cardiovascular and diabetic complications. J Hypertens. 1993;11:309–17. 101. Hanefeld M, Fischer S, Julius U, Schulze J, Schwanebeck U, Schmechel H, et al.; The DIS Group. Risk factors for myocardial infarction and death in newly detected NIDDM: the Diabetes Intervention Study, 11-year follow-up. Diabetologia. 1996;39:1577–83. 102. Haffner SM, Stern MP, Hazuda HP, Mitchell BD, Patterson JK. Cardiovascular risk factors in con ﬁ rmed prediabetic individuals. JAMA. 1990;263:2893–998. 103. Fuller JH, Shipley MJ, Rose G, Jarrett RJ, Keen H. Coronary-heart-disease risk and impaired glucose tolerance. The Whitehall study. Lancet. 1980;1:1373–6. 104. Fontbonne A, Eschwège E, Cambien F, Richard J-L, Ducimetière P, Thibult N, et al. Hypertriglyceridaemia as a risk factor of coronary heart disease mortality in subjects with impaired glucose tolerance or diabetes. Diabetologia. 1989;32:300–4. 105. Pyorala K. Relationship of glucose tolerance and plasma insulin to the incidence of coronary heart disease: results from two population studies in Finland. Diabetes Care. 1979;2:131–41. 106. Barzilay JI, Spiekerman CF, Wahl PW, Kuller LH, Cushman M, Furberg CD, et al. Cardiovascular disease in older adults with glucose disorders: comparison of American Diabetes Association criteria for diabetes mellitus with WHO criteria. Lancet. 1999;354:622–5. 107. Temelkova-Kurktschiev TS, Koehler C, Henkel E, Leonhardt W, Fuecker K, Hanefeld M. Postchallenge plasma glucose and glycemic spikes are more strongly associated with atherosclerosis than fasting glucose or HbA1c level. Diabetes Care. 2000;23:1830–4. 108. Bonora E, Kiechl S, Oberhollenzer F, Egger G, Bonadonna RC, Muggeo M, et al. Impaired glucose tolerance, type II diabetes mellitus and carotid atherosclerosis: prospective results from the Bruneck Study. Diabetologia. 2000;43:156–64. 109. O’Leary DH, Polak JF, Kronmal RA, Kittner SJ, Bond MG, Wolfson Jr SK, et al. Distribution and correlates of sonographically detected carotid artery disease in the Cardiovascular Health Study. The CHS Collaborative Research Group. Stroke. 1992;23:1752–60. 10 Decreasing Postprandial Plasma Glucose Using an a-Glucosidase Inhibitor... 187

110. Yamasaki Y, Kawamori R, Matsushima H, Nishizawa H, Kodama M, Kubota M, et al. Asymptomatic hyperglycaemia is associated with increased intimal plus medial thickness of the carotid artery. Diabetologia. 1995;38:585–91. 111. Beks PHJ, Mackaay AJC, De Vries H, De Neeling JND, Bouter LM, Heine RJ. Carotid artery stenosis is related to blood glucose level in an elderly Caucasian population: the Hoorn Study. Diabetologia. 1997;40:290–8. 112. Chiasson JL, Josse RG, Gomis R, Hanefeld M, Karasik A, Laakso M. Acarbose treatment and the risk of cardiovascular disease and hypertension in patients with impaired glucose tolerance. The STOP-NIDDM Trial. JAMA. 2003;290:486–94. 113. Zeymer U, Schwarzmaier-D’assie A, Petzinna D, Chiasson JL. Effect of acarbose treatment on the risk of silent myocardial infarctions in patients with impaired glucose tolerance: results of the randomised STOP-NIDDM trial electrocardiography substudy. Eur J Cardiovasc Prev Rehabil. 2004;11:412–5. 114. Hanefeld M, Chiasson JL, Koehler C, Henkel E, Schaper F, Temelkova-Kurktschiev T. Acarbose slows progression of intima-media thickness of the carotid arteries in subjects with impaired glucose tolerance. Stroke. 2004;35:1073–8. 115. Hanefeld M, Koehler C, Westermeier T, Saunders G, Chiasson JL. Predictors of newly diagnosed hypertension in people with IGT: impact of acarbose treatment. A scondary analysis of the STOP-NIDDM trial. Diabetes. 2010;59:Abstract 325-OR. 116. Josse RG, McGuire AJ, Saal GB. A review of the economic evidence for acarbose in the prevention of diabetes and cardiovascular events in individuals with impaired glucose tolerance. Int J Clin Pract. 2006;60:847–55. 117. Caro JJ, Getsios D, Caro I, Klittich WS, O’Brien JA. Economic evaluation of therapeutic interventions to prevent type 2 diabetes in Canada. Diabet Med. 2004;21:1229–36. 118. Quilici S, Chancellor J, Maclaine G, McGuire A, Andersson D, Chiasson JL. Cost- effectiveness of acarbose for the management of impaired glucose tolerance in Sweden. Int J Clin Pract. 2005;59:1143–52. 119. Evers T, Luddeke HJ, Liebl A, et al. Economic value of acarbose treatment in persons with impaired glucose tolerance (IGT) for the German health care system. Value Health. 2004;7:741. 120. Sabes R. Cost-effectiveness analysis of acarbose in the treatment of patients with impaired glucose tolerance. Gac Sanit. 2004;18:431–9. 121. Stern MP, Williams K, Haffner SM. Identifi cation of persons at high risk for type 2 diabetes mellitus: do we need the oral glucose tolerance test? Ann Intern Med. 2002;136:575–81. 122. Anderson KM, Odell PM, Wilson PW, Kannel WB. Cardiovascular disease risk pro fi les. Am Heart J. 1991;121:293–8. 123. Ur E. De fi nition, classi fi cation and diagnosis of diabetes of other dysglycemic categories. Can J Diabetes. 2008;32 Suppl 1:S5–9. 124. Ur E, Chiasson JL, Ransom T, Rowe R. Prevention of diabetes. Can J Diabetes. 2008;32 Suppl 1:S17–9. 125. Colagiuri S, Hussain Z, Zimmet P, Cameron A, Shaw J. Screening for type 2 diabetes and impaired glucose metabolism: the Australian experience. Diabetes Care. 2004;27:367–71. 126. Zhang P, Engelgau MM, Valdez R, Cadwell B, Benjamin SM, Narayan KM. Effi cient cutoff points for three screening tests for detecting undiagnosed diabetes and pre-diabetes: an economic analysis. Diabetes Care. 2005;28:1321–5. 127. Kousta E, Lawrence NJ, Penny A, Millauer BA, Robinson S, Dornhorst A, et al. Implications of new diagnostic criteria for abnormal glucose homeostasis in women with previous gestational diabetes. Diabetes Care. 1999;22:933–7. 128. Larsson H, Lindgarde F, Berglund G, Ahren B. Prediction of diabetes using ADA or WHO criteria in post-menopausal women: a 10-year follow-up study. Diabetologia. 2000;43:1224–8. 129. Simmons D, Thompson CF, Engelgau MM. Controlling the diabetes epidemic: how should we screen for undiagnosed diabetes and dysglycaemia? Diabet Med. 2005;22:207–12. 130. Tuomilehto J, Lindstrom J, Hellmich M, Lehmacher W, Westermeier T, Evers T, et al. Development and validation of a risk-score model for subjects with impaired glucose tolerance for the assessment of the risk of type 2 diabetes mellitus-The STOP-NIDDM risk-score. Diabetes Res Clin Pract. 2010;87:267–74. Chapter 11 Da Qing, Finnish DPP, Tripod, and Dream: Lifestyle and Thiazolidinediones in the Prevention of Diabetes

Mariela Glandt and Zachary Bloomgarden

Introduction

Nearly one in seven adults in the United States has diabetes—but more than twice that number are at risk of diabetes, based on impaired glucose tolerance (IGT, 2-h glucose >140 and <200 mg/dL) or impaired fasting glucose (IFG, fasting glucose ³100 and £126 mg/dL) based on an oral glucose tolerance test (OGTT) [ 1 ] . In the world, the prevalence of these diabetes risk states exceeds 8% of the adult population [ 2] . Given the high complication rates of diabetes, a great deal of interest has focused on measures to reduce the progression of this risk state. We review the results of two studies of lifestyle modiﬁ cation and two studies of thiazolidinedione administration in the prevention of diabetes.

Da Qing

110,660 residents age 25–74 of Da Qing, Hei Long Jiang Province, China were screened in 1986 by measurement of plasma glucose concentrations 2 h after breakfast containing ³ 100 g steamed bread, with approximately 80 g carbohydrate. 4,209 persons with levels >6.67 mmol/L (120 mg/dL) were referred for a 75 g OGTT, with 3,956 having the examination [ 3 ] , ﬁ nding 190 (0.17%) with previously known

M. Glandt, MD Medical Director, Diabetes Medical Center , Boney ha Ir 9 , Tel Aviv 61480 , Israel e-mail: [email protected] Z. Bloomgarden, MD (*) Department of Medicine , Mount Sinai School of Medicine , 35 East 85th Street , New York , NY 10028 , USA e-mail: [email protected]

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 189 DOI 10.1007/978-1-4614-3314-9_11, © Springer Science+Business Media New York 2012 190 M. Glandt and Z. Bloomgarden

diabetes, 630 (0.6%) with newly diagnosed diabetes, and 577 (0.55%) with IGT, using the 1985 WHO criteria [ 4] . Retinopathy was found in 30.5% of newly diagnosed diabetic persons, and their prevalence of hypertension, dyslipidemia, and obesity was approximately twice that in nondiabetic persons [ 1 ] . IGT was a risk factor for CHD after adjustment for age, sex, cigarette smoking, plasma cholesterol, BP, and obesity, and was associated with a doubling of the frequency of hypertension, obesity, and abnormal albumin excretion. Approximately 60% of those with IGT were overweight or obese; using BMI >23 rather than >25 kg/m2 was felt to be the appropriate cutoff for increased risk. BMI tracked with hypertension, elevated fi brinogen, and hypertriglyceridemia and interacted with fasting glucose in predicting the 6-year likelihood of development of diabetes. The highest risk of diabetes was at BMI >27 kg/m2 [ 5 ] . Similarly, waist circumferences cutoffs of 80 and 90 cm, respectively, were indicative of moderate and marked increase in risk, both in men and in women. The persons with IGT identi fi ed in the survey formed the cohort for the long-term follow-up and intervention Da Qing study. Of the 577 persons with IGT who were randomized in the study, 530 completed the intervention. Their baseline age was 45, BMI 25.8 kg/m2 , and fasting and 2-h blood glucose 5.6 and 9.0 mmol/L. Treatment assignments were made on a clinic- wide basis (rather than differently assigning individuals within a given clinic) to a control group and to groups with diet alone, to exercise alone (primarily walking), and to both diet and exercise interventions; over 6 years, the incidences of 2-h glucose >200 were 15.7, 10, 8.3, and 9.6, and incidences of fasting glucose >140 mg/ dL were 9.6, 3.7, 5.3, and 5.5 per 100-patient years in the four groups, respectively. Cumulatively, diabetes developed in 66, 47, 44, and 45% of the members of the respective groups [6 ] . Based on the 2-h glucose, and on both fasting and 2-h glucose, the interventions led to 36–47% and to 29–33% reductions in diabetes, respectively. The exercise intervention was particularly effective in persons with BMI < 25, with diabetes developing in 60% of controls, 38% of those with diet, 26% with exercise, and 35% with both interventions, while overweight participants developed diabetes among 72% of controls, 48% with diet alone, 51% with exercise alone, and 53% with diet plus exercise. Baseline fasting and 2-h insulin measurements were performed in 284 of the 577 persons in the trial [ 7] . Those with higher insulin levels had lesser responses to all the interventions. The diet plus exercise intervention trended to be particularly useful for this group. Fascinating studies reported 20-year follow-up results of all but 26 of the 577 persons randomized in the DaQing interventions. The reduction in diabetes development seen in the intervention vs. control groups at the end of 6 years was maintained through the subsequent 14 years, with very high cumulative diabetes development rates of 80% vs. 93% [8 ] . The authors note that there was little difference in weight-change between the intervention and control groups, suggesting a different mechanism of prevention from that operative in the “Finnish Diabetes Prevention Study ” and in the US Diabetes Prevention Program [ 9 ] , where weight loss appeared to be the principal mechanism of bene fi t. There was a trend to reduction in cumulative CV mortality at 12.5% vs. 17.4%, with two thirds of the deaths being attributed to stroke and the remainder to heart disease. First CVD events 11 Da Qing, Finnish DPP, Tripod, and Dream… 191 occurred in 39% of those in the lifestyle groups vs. 42% of controls. A1c was 7.3% vs. 7.8%, and severe retinopathy occurred in 9.2% vs. 16.2%, respectively, but there were no differences in neuropathy and nephropathy [10 ] .

Conclusions

Diet, exercise, and the combination of both lifestyle interventions in Chinese persons with prediabetes reduced the absolute risk of diabetes by approximately 20%, without a major effect on weight, and, interestingly, appearing particularly to reduce risk in insulin-deﬁ cient rather than insulin-resistant persons. In 20-year follow-up, the absolute diabetes development rate was 13% lower among those persons originally randomized to the lifestyle interventions, and their level of glycemic control appeared to be better than among the original control population. The population had high likelihood of diabetic complications, particularly retinopathy and CVD, with reduction in rates of severe retinopathy among those undergoing lifestyle interventions and a trend to reduction in CV mortality.

Finnish Diabetes Prevention Study

Another study of lifestyle intervention to prevent type 2 diabetes in high-risk individuals was the Finnish Diabetes Prevention Study. In addition to the main goal of assessing the effi cacy of an intensive diet-exercise program in preventing or delaying type 2 diabetes mellitus in subjects with IGT, it also aimed to evaluate the effects of the intervention program on cardiovascular risk factors and to assess the determinants for the progression to diabetes in persons with IGT. Recruitment began after a pilot study in 1993 and was completed in May 1998. The study subjects were recruited in fi ve different centers in Finland through population screenings with special emphasis on high-risk groups such as those with obesity or fi rst-degree relatives of patients with type 2 diabetes. In the study, 523 overweight subjects with IGT ascertained by two OGTTs were randomized to either a control or intervention group. IGT was de fi ned as a plasma glucose concentration of 140–198 mg/dL (7.8–11.0 mmol/L) 2 h after the oral administration of 75 g of glucose in subjects whose plasma glucose concentration after an overnight fast was less than 140 mg/dL, based on criteria adopted by the WHO in 1985. Participants also had to be between 40 and 64 years of age at randomization and had to be overweight with BMI > 25 kg/m2 to be eligible for the study. Subjects who already had a diagnosis of diabetes (except for gestational diabetes) or were already involved in a vigorous exercise program and those with other diseases were excluded. The subjects in the control group received general information at the start of the trial about the lifestyle changes necessary to prevent diabetes and about annual follow-up visits, but no speci fi c individualized programs were offered to them. A 3-day food record was fi lled out once a year. The subjects in the intervention group had 192 M. Glandt and Z. Bloomgarden seven sessions with a nutritionist during the fi rst year and a visit every 3 months thereafter. The dietary goals of the intervention were (1) reduction in weight of 5% or more, (2) reduction in total intake of fat to less than 30% of energy consumed, (3) reduction in intake of saturated fat to less than 10% of energy consumed, and (4) increase in fi ber intake to at least 15 g/1,000 kcal. The patient fi lled out a 3-day food record before the fi rst appointment, and every 3 months thereafter. After 6 months, the use of a very-low-calorie diet (VLCD) for 2–5 weeks or as a substitute for one to two meals per day was considered to boost weight loss. Subjects were also guided individually to increase their physical activity. The physical activity goal was to achieve moderate exercise for at least 30 min/day. Endurance exercise (such as walking, jogging, swimming, aerobic ball games, or skiing) was recommended as a way to increase aerobic capacity and improve cardiorespiratory fi tness. Supervised, progressive, individually tailored, circuit-type resistance-training sessions were also offered with the aim of improving large muscle group functional capacity and strength. Participants were instructed to perform a moderate to high number of repetitions and to take a break of 15–60 s between the stations on the circuit. The study subjects completed the validated Kuopio Ischaemic Heart Disease Risk Factor Study 12-month Leisure Time Physical Activity (LTPA) questionnaire at baseline and at every annual visit. All participants had an annual OGTT, fasting lipid panel, medical history, and physical examination with measurements of height, weight, waist circumference, and systolic and diastolic blood pressure. Of the 522 subjects in the study, 265 were randomized to the intervention group with intensive diet-exercise counseling and 257 were randomized to the control group. The two groups were similar at baseline with mean age 55 and BMI 31.1 kg/m2 . The mean baseline fasting plasma glucose was 109 mg/dL and mean plasma glucose 2 h after the 75 g oral glucose load was 160.3 mg/dL without signi fi cant difference between the groups. Of the participants, 10% in the intervention group and 8% in the control group were lost to follow-up, with median follow-up of 4 years. The primary outcome was development of diabetes, as measured by OGTT, repeated for confi rmation if diagnostic of diabetes. Diabetes was diagnosed in a total of 86 subjects—27 in the intervention group and 59 in the control group, with incidences of 32 and 78 cases per 1,000 person-years in the intervention and control groups, respectively, a 58% reduction. The intervention was effective for both sexes, with diabetes decreasing 63 and 54% in men and in women, respectively. The study included a number of secondary outcomes. Mean weight reduction was 4.5 kg in the intervention group and 1.0 kg in the control group at 1 year. Some regain of weight appeared during the following 2 years. Excluding intervention group participants on VLCD, whose weight reduction was 6.2–7.0 kg at year 1 and 4.8–7.2 kg at year 3, mean weight reduction was 4.1–4.3 kg at year 1 and 3.2–4.5 kg at year 3. During the fi rst year, the prevalences of abdominal obesity, IFG, elevated blood pressure, and low HDL cholesterol decreased signifi cantly in the intervention group, although that of hypertriglyceridemia was unchanged. In the control group, only the prevalence of hypertension decreased. From baseline to study-end, a signifi cant decrease in the prevalence of abdominal obesity, elevated blood pressure, low HDL 11 Da Qing, Finnish DPP, Tripod, and Dream… 193 cholesterol, and elevated triglycerides was observed in the intervention group, but only low HDL cholesterol improved in controls. The lifestyle intervention reduced abdominal obesity, adjusted for age, sex, and baseline value [11 ] . The prevalence of metabolic syndrome decreased during the fi rst year from 74.0 to 58.0% vs. from 74.0 to 67.7% in the intervention and control groups, respectively. At the end of the study, 62.6% of subjects in the intervention group and 71.2% of subjects in the control group had metabolic syndrome, which corresponds to an age- and sex-adjusted odds ratio of 0.62 [12 ] . A subset of patients were studied in detail to determine the impact of the 4-year lifestyle intervention on insulin sensitivity and insulin secretion, endeavoring to determine which factor was more important in reducing the incidence of diabetes. All of these patients were studied in the Kupio clinic, where in addition to the OGTT, 87 participants at baseline and 52 participants at the end of the study underwent a frequently sampled intravenous glucose tolerance test (FSIGT). In this subset, at the end of the study, the intervention group lost on average 3.5 kg more than the control group and had a reduction in waist circumference 2.4 cm greater than that of the control group. There were strong correlations between the 4-year changes in insulin sensitivity and in weight. In the entire group, insulin sensitivity improved by 64% among those in the highest tertile of weight loss, but deteriorated by 24% in those who gained weight (lowest tertile). The acute insulin response declined signifi cantly in the control group [ 13 ] . Thus, both improvement in insulin sensitivity and protection against the reduction in insulin secretion appeared to play roles in the bene fi t of the intervention. Cardiovascular mortality and morbidity were monitored through computerized register linkage to two nationwide health registers: the Hospital Discharge Register and the Causes of Death Register, using the national personal identi fi cation number. After a median follow-up time of 10.2 years, there was no statistically signi fi cant difference in cardiovascular mortality between the two groups, with 57 new cardiovascular events in the intervention group and 54 in the control group. Men and women had a similar CVD incidence, whether the groups were analyzed separately or combined. There were no signi fi cant differences between the intervention and control groups in coronary artery angioplasty, in by-pass surgery, or in treatment for dyslipidemia and blood pressure [14 ] . All participants in the study who had not developed diabetes were invited to take part in the postintervention follow-up, with yearly nurse visits during which the same procedures were carried out as during the intervention period. No specifi c diet or exercise counseling was provided during this 3-year follow-up. Those who originally participated in the intervention vs. control groups had diabetes incidence rates of 4.6 vs. 7.2 per 100 person-years, a 36% reduction in relative risk [15 ] . Post hoc analyses were carried out to assess the determinants for the progression to diabetes. Even after adjustment for other risk factors, dietary fat and fi ber intake were signifi cant predictors of sustained weight reduction and of progression to type 2 diabetes [16 ] . Greater levels of moderate-to-vigorous LTPA were associated with decreased likelihood of developing metabolic syndrome and an increased likelihood of its resolution, after adjustment for changes in dietary intakes of total and 194 M. Glandt and Z. Bloomgarden

saturated fat, fi ber, and energy, and change in BMI. The increase in moderate- to-vigorous LTPA was most strongly associated with improvement of glycemia [17 ] . Another post hoc analysis found long sleep duration to be associated with increased risk of type 2 diabetes. In the control group, participants who slept more than 9 h had greater rates of developing diabetes, while patients in the intervention group implementing lifestyle changes did not show an effect of sleep duration on the incidence of diabetes. The lifestyle intervention resulted in similar improvement in body weight, insulin sensitivity, and immune mediator levels regardless of sleep duration [18 ] . Another study examined which individual components of the comprehensive lifestyle intervention were most likely to reduce subclinical in fl ammation, which confers increased risks of type 2 diabetes, cardiovascular disease, neurodegenerative disorders, and other age-related chronic diseases. C-reactive protein and interleukin-6 levels, thought to represent the best characterized pro-infl ammatory risk factors for type 2 diabetes, were compared at baseline and 1 year after follow-up in a subsam- ple of 406 of the participants, both infl ammatory markers decreasing with the lifestyle intervention. The decrease was predicted by increases in fi ber intake and in moderate-to-vigorous LTPA, but not by increase in total LTPA or by changes in carbohydrate or fat intake [19 ] . A post hoc analysis showed that the higher the baseline risk for diabetes, the greater the risk reduction achieved during the intervention, so that those who were most at risk bene fi tted the most. Risk was calculated using the FINDRISC questionnaire, a validated screening tool for undiagnosed type 2 diabetes, dysglycemia, and the metabolic syndrome [ 20 ] . In participants with low baseline risk, the risk of developing diabetes was low whether they were in the interventional or control group. Participants at high risk lowered the risk if they were in the intervention group, actively dieting and exercising, while those at high risk who were randomized to the control group had a very high incidence rate of diabetes. The study also showed that the intervention was most effective among the oldest (age 61 years) individuals, with a relative risk reduction of 64% compared with that in the control group. This was similar to the fi ndings in the US Diabetes Prevention Program, where the participants in the oldest age-group (60–85 years at baseline) achieved the largest risk reduction [21 ] . A recent study investigated whether a family history of diabetes or genetic variants of type 2 diabetes modulated the decreased incidence of diabetes achieved with lifestyle changes [22 ] . As of today, 30 genetic variants have been found to be associated with an increased risk of type 2 diabetes [23– 25 ] . Generally, each variant has a limited effect, since the risk may increase by 10–15% per copy of each risk allele, with the exception of TCF7L2, which has a more pronounced in fl uence, increasing the risk 1.4-fold [ 26 ] . These known genetic variants account for only 10% of the genetic basis of type 2 diabetes and seem to have limited ability in predicting the development of type 2 diabetes[ 27, 28] . The study found that, at 4-year follow-up, those participants with a family history of diabetes seemed to have a lower incidence of diabetes than those with no family history. However, this difference between the two groups disappeared when the entire 7-year follow-up was 11 Da Qing, Finnish DPP, Tripod, and Dream… 195

analyzed, showing that the effect of diet and exercise was signi fi cant both among those with and without a family history of diabetes. The study did not fi nd aggregation of known genetic risk variants in persons with a family history of diabetes, and genetic risk variants did not modify the incidence of diabetes. Two substudies provided supportive evidence for the involvement of genetic variation as a modifi er in the effect of lifestyle changes. One showed that genetic variation in adiponectin, found in the ADIPOQ locus, contributed to variation in body size and serum adiponectin concentrations and may also modify the risk of developing type 2 diabetes [ 29 ] . Another study showed that patients who were homozygous for a polymorphism of the hepatic lipase gene were less likely to benefi t from the lifestyle intervention, and therefore were more likely to develop diabetes (13% vs. 1% in subjects who had at least one normal allele), suggesting that polymorphism of the hepatic lipase gene is a risk factor for type 2 diabetes [30 ] .

Conclusions

The Finnish Diabetes Prevention Study showed that a moderate degree of weight loss achieved through diet and exercise led to a very signiﬁ cant change in the risk of developing diabetes. Twenty-two patients with IGT need to be treated in order to prevent one case of diabetes. This very feature-rich study offered a plethora of insights and consistently highlights the importance of diet and exercise for prevention. The intervention was particularly effective in those at the highest risk. It is possible that many of the results of the study are less robust than they would be in a nonstudy, real-world situation, as the control arm also received advice and received consistent medical follow-up. Although there was no difference in cardiovascular mortality, the study was not powered to examine this effect. More time might be required to see changes in this endpoint. Furthermore, the study showed that the mortality in both arms of the study was low compared to mortality in individuals with IGT in the general population, a not uncommon ﬁ nding among participants in randomized controlled clinical trials. It is also notable that the study had a very low dropout rate, compared to most weight loss studies, suggesting that patients at risk for diabetes are willing to participate in an intensive prevention program when given the opportunity.

TRIPOD/PIPOD

The TRoglitazone In Prevention Of Diabetes (TRIPOD) study investigated the development of diabetes in 266 high-risk Hispanic women with previous gestational diabetes treated with either troglitazone or placebo [ 31 ] . The women were identi fi ed through chart reviews and patient interviews from Los Angeles County Women’s 196 M. Glandt and Z. Bloomgarden and Children’s Hospital from August 1995 to May 1998. Women met the inclusion criteria if they were older than 18 years of age, had gestational diabetes mellitus within the previous 4 years, and were willing to use effective contraception. Exclusion criteria included evidence of chronic disease, serum alanine aminotransferase concentration >1.5 times the laboratory upper normal, or known diabetes. Women also were excluded if they had a sum of fi ve OGTTs plasma glucose concentrations >625 mg/dL predicting a >70% 5-year risk of diabetes. Enrolled subjects received dietary advice and were advised to walk for 30 min 3 days each week. Patients underwent a FSIGT within 4 weeks of the screening OGTT to assess baseline insulin sensitivity and pancreatic b -cell function, and then were randomized to receive troglitazone 400 mg/day or placebo in a double-blind fashion. Fasting glucose was measured at 3-month intervals and OGTTs were performed annually to detect diabetes. Measurements of height, weight, sitting blood pressure, fasting serum lipids, and carotid intima-media thickness (cIMT) were performed at the times of OGTTs. The trial was scheduled to continue until August 2000, but was terminated in March 2000, when troglitazone was withdrawn from the market after reports of hepatotoxicity. At that time, 79% of 105 subjects active in TRIPOD had not reached their annual OGTT visit for the year 2000. They were noti fi ed of their treatment status, asked to discontinue study medications, and scheduled for an end-of-trial OGTT. Development of diabetes was the primary study endpoint. During the median follow-up of 30 months, average annual diabetes incidence rates in women who returned for follow-up were 12.1 and 5.4% in the placebo and troglitazone groups, respectively, a 55% reduction. Treatment with troglitazone signifi cantly improved insulin sensitivity, while acute insulin response remained the same, resulting in a signi fi cant improvement in disposition index from baseline and vs. placebo, suggesting improved b -cell compensation for insulin resistance [32 ] . The most responsive group to intervention were those who early in the study, while on troglitazone, showed the greatest reduction in insulin resistance and fall in insulin secretion to a glucose challenge. The Pioglitazone In Prevention Of Diabetes (PIPOD) study was conducted as a single arm, open-label follow-up study for those participants who had completed the TRIPOD study. OGTTs were performed annually on pioglitazone and at the end of the 6-month postdrug washout. Intravenous glucose tolerance tests (IVGTTs) for assessment of insulin sensitivity and b -cell function were conducted at baseline, after 1 year on pioglitazone, and at the end of the postdrug washout [33 ] . Incidence rates of diabetes were calculated from 86 women (42 from the active treatment arm of the TRIPOD study) who had at least one follow-up visit after enrollment. Overall, 11 of them had diabetes at one or more OGTTs during a median of 35.9 months of pioglitazone treatment. Average annual incidence rates of diabetes were 5.2% during pioglitazone treatment and 4.6% during the entire observation period, including the post drug washout. Similar to the TRIPOD study, the risk was lowest in women with the largest reduction in total IVGTT insulin area after 1 year of treatment. The study showed that pioglitazone stopped the decline in b -cell 11 Da Qing, Finnish DPP, Tripod, and Dream… 197

function that occurred during placebo treatment in the TRIPOD study. Pioglitazone also maintained the stability of b -cell function that occurred during troglitazone treatment in the TRIPOD study. A follow-up study reported results on the progression of subclinical atherosclerosis, measured by cIMT in the women in PIPOD who did not develop diabetes. The study analyzed the 61 women who completed at least one annual follow-up cIMT measurement. Pioglitazone treatment was associated with a signi ﬁ cant reduction in the rate of progression of cIMT compared to rates that had been observed in the same individuals during placebo treatment in the TRIPOD study. Pioglitazone also maintained a persistently low cIMT progression rate in women who had been on troglitazone treatment in TRIPOD and pioglitazone in PIPOD. The impact of pioglitazone on cIMT was not explained by alterations in body weight, blood pressure, or circulating concentrations of glucose, insulin, or standard lipids [34 ] .

Conclusion

Both troglitazone and pioglitazone treatment given to Hispanic women with prior gestational diabetes were associated with stable pancreatic b -cell function and a relatively low rate of diabetes. The lowest rate of diabetes occurred in association with the greatest reduction in insulin secretory demands during the ﬁ rst year of treatment. Pioglitazone was also shown to slow progression of subclinical atherosclerosis as measured by cIMT. The ﬁ ndings from these two trials suggest that thiazolidinedione drugs may modify the natural history of progression to type 2 diabetes in high-risk Hispanic patients.

DREAM (Diabetes REduction Assessment with Ramipril and Rosiglitazone Medication)

In the DREAM trial, 5,269 persons with prediabetes were randomized to receive ramipril 15 mg/day, rosiglitazone 8 mg/day, both, or neither. Of these, 35% had isolated IGT, 14% had isolated IFG, and 51% had both. At semiannual OGTTs, diabetes was diagnosed if two consecutive plasma glucose levels within a 3-month period exceeded 125 mg/dL fasting or 199 mg/dL at 2 h [ 35 ] . In total, 24,872 individuals in 21 countries were screened with OGTT over 2 years; 14,661 were women, with glycemic abnormality 3% more likely per child born, 14% more likely in those with a history of eclampsia or preeclampsia, 21% more likely in those with history of irregular menses, and 53% more likely in those with history of gestational diabetes. There was 5% greater likelihood per year of age, and 9% greater likelihood among those of non-European ancestry [36 ] . 198 M. Glandt and Z. Bloomgarden

Patients were followed in the study for a median of 3 years. The likelihood of diabetes was not signifi cantly reduced by ramipril, but the drug increased the likelihood of regression to normoglycemia by 16%, with 2-h glucose 135 vs. 141 mg/dL in those receiving vs. not receiving the agent [ 37 ] . Among those receiving vs. not receiving rosiglitazone, diabetes developed in 11% vs. 25%. There was regression to normoglycemia in 51% vs. 30%, no signifi cant change in atherosclerotic events, and 0.5% vs. 0.1% likelihood of congestive heart failure [ 38 ] . The glycemic bene fi t of rosiglitazone was greater among those of Latino ethnicity and lesser among South Asians, although all ethnic groups showed signifi cant reduction in diabetes development [39 ] . Interestingly, ramipril was neither associated with improvement nor even with lesser degrees of worsening of albuminuria or estimated glomerular fi ltration rate (eGFR), while rosiglitazone was associated with a trend to reduced likelihood of deterioration in eGFR, and signifi cant 18% reduction in albuminuria progression and 20% improvement in the composite renal endpoint of fi rst occurrence of progression of albuminuria, decreased eGFR by ³30%, or dialysis or transplanta- tion [40 ] . The single nucleotide polymorphism rs6123045, in the nuclear factor of activated T-cells cytoplasmic calcineurin-dependent 2 (NFATC2) gene, was signifi cantly associated with development of edema in the rosiglitazone-treated group, although neither with cardiovascular endpoints nor with congestive heart failure [41 ] . The endothelium-derived vasoconstrictive factor endothelin-1 induces calcineurin activity, and rosiglitazone suppresses nuclear translocation of NFATc4 and enhances the association of PPAR g with NFATc4/calcineurin in isolated cardiomyocytes. This genetic fi nding may improve our understanding of thiazolidinedione-induced edema. A subset of 1,425 participants in the trial had serial measurement of cIMT. There was a trend to slower progression in patients receiving rosiglitazone, with signifi cant reduction in the average of two IMT measurements in the common carotid far wall; ramipril failed to change either measure [ 42 ] . It should be noted that there were low levels of carotid atherosclerosis at baseline, and that demonstrating regression of such subclinical disease would not be expected to occur during the trial. Another interesting subset analysis showed marked improvement in b -cell function with rosiglitazone, either as measured with the fasting proinsulin-to-C-peptide ratio, or with the change from baseline to 30 min in insulin divided by that in glucose, in turn divided by the homeostasis model assessment of insulin resistance; this was particularly the case in patients with elevation in the baseline 2-h glucose, and lesser improvement was seen in those with isolated IFG; ramipril did not improve any of these measures [43 ] . A 1–2-year posttreatment follow-up was performed on a representative subset of 47% of participants in the study. Those previously receiving rosiglitazone remained heavier, with BMI 31 vs. 30 kg/m2 , and 12% vs. 7% had edema; those previously receiving ramipril had 5/2 mmHg lower blood pressure. Among those randomized to rosiglitazone vs. placebo, 12% vs. 26% had developed the primary endpoint of diabetes or death while on treatment, 19% vs. 31% after an initial washout period, and 27% vs. 39% at the end of the 1–2-year follow-up period, suggesting that a subset of patients had durable diabetes prevention with and following use of the agent [44 ] . Ramipril did not show a signi fi cant long-term effect on glycemia. 11 Da Qing, Finnish DPP, Tripod, and Dream… 199

Conclusions

Thiazolidinedione treatment had major effect in preventing diabetes in this trial as in TRIPOD, with evidence of durable posttrial bene fi t. The drugs may both reduce insulin resistance and improve b -cell function. The expected side effect of edema was seen, and an explanatory gene polymorphism was found, without evidence of increase in atherosclerotic endpoints and with modest suggestion of antiatherosclerotic bene fi t. There was also evidence in this relatively low-risk population of renal benefi t of the treatment. Surprisingly, no benefi t was seen with a high dose of ramipril either in preventing diabetes or in reducing renal or cardiovascular complications. Given the cost-effectiveness of lifestyle intervention and the evidence of fl uid retention as a side effect with rosiglitazone, its practical use in diabetes prevention was ques- tioned at the time of the study [ 45 ] , although thiazolidinedione use in patients at the higher ranges of prediabetes, perhaps based on A1c values, was suggested to potentially be of bene fi t [ 46 ] .

Summary

In the DaQing study, lifestyle intervention not only reduced the development of diabetes, but, at 20-year follow-up, improved glycemic control in participants, reduced diabetic retinopathy, and showed a trend to reduction in CVD. The Finnish Diabetes Prevention Study showed similar bene ﬁ t in reducing development of diabetes, suggesting this to be a feasible and effective approach. Both troglitazone and pioglitazone, given in TRIPOD to women who had had gestational diabetes, were associated with diabetes prevention, with a suggestion of reduction in otherwise progressive b -cell failure. Rosiglitazone was similarly effective in DREAM, with evidence of durable posttrial bene ﬁ t, and with a suggestion of reduction in intermediate renal and atherosclerosis endpoints. Taken together, these trials suggest that measures to reduce development of diabetes are possible and should be strongly considered by healthcare planners to prevent the burden of disease which is otherwise likely to ensue.

References

1. Cowie CC, Rust KF, Ford ES, Eberhardt MS, Byrd-Holt DD, Li C, et al. Full accounting of diabetes and pre-diabetes in the U.S. population in 1988–1994 and 2005–2006. Diabetes Care. 2009;32:287–94. 2. International Diabetes Federation. Fact sheet: impaired glucose tolerance (IGT); 2011. http:// www.idf.org/fact-sheets/impaired-glucose-tolerance . Accessed 31 March 2011. 3. Hu YH, Pan XR, Liu PA, Li GW, Howard BV, Bennett PH. Coronary heart disease and diabetic retinopathy in newly diagnosed diabetes in Da Qing, China: the Da Qing IGT and Diabetes Study. Acta Diabetol. 1991;28:169–73. 200 M. Glandt and Z. Bloomgarden

4. Pan XR, Hu YH, Li GW, Liu PA, Bennett PH, Howard BV. Impaired glucose tolerance and its relationship to ECG-indicated coronary heart disease and risk factors among Chinese. Da Qing IGT and diabetes study. Diabetes Care. 1993;16:150–6. 5. Li G, Chen X, Jang Y, Wang J, Xing X, Yang W, et al. Obesity, coronary heart diseaase risk factors and diabetes in Chinese: an approach to the criteria of obesity in the Chinese population. Obes Rev. 2002;3:167–72. 6. Pan XR, Li GW, Hu YH, Wang XJ, Yang WY, An ZX, et al. Effects of diet and exercise in preventing NIDDM in people with impaired glucose tolerance. Diabetes Care. 1997;20: 537–44. 7. Li GW, Hu YH, Yang WY, Jiang YY, Wang JP, Xiao JZ, et al. Effects of insulin resistance and insulin secretion on the ef fi cacy of interventions to retard development of type 2 diabetes mellitus: the DA Qing IGT and Diabetes Study. Diabetes Res Clin Pract. 2002;58:193–200. 8. Li GW, Zhang P, Wang JP, Gregg EW, Yang WY, Gong QH, et al. The long-term effect of lifestyle interventions to prevent diabetes in the China Da Qing Diabetes Prevention Study: a 20-year follow-up study. Lancet. 2008;371:1783–9. 9. Knowler WC, Barrett-Connor E, Fowler SE, Hamman RF, Lachin JM, Walker EA, et al.; Diabetes Prevention Program Research Group. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346:393–403. 10. Gong Q, Gregg EW, Wang J, An Y, Zhang P, Yang W, et al. Long-term effects of a randomised trial of a 6-year lifestyle intervention in impaired glucose tolerance on diabetes-related microvascular complications: the China Da Qing Diabetes Prevention Outcome Study. Diabetologia. 2011;54:300–7. 11. Tuomilehto J, Lindström J, Eriksson JG, Valle TT, Hämäläinen H, Ilanne-Parikka P, et al.; Finnish Diabetes. Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med. 2001;344:1343–50. 12. Ilanne Parikka P, Eriksson JG, Lindstrom J, et al. Effect of lifestyle intervention on the occurrence of metabolic syndrome and its components in the Finnish Diabetes Prevention Study. Diabetes Care. 2008;31:805–7. 13. Uusitupa M, Lindi V, Louheranta A, Salopuro T, Lindstro J, Tuomilehto J; Finnish Diabetes Prevention Study Group. Changing lifestyles of people with impaired glucose tolerance 4-year results from the Finnish Diabetes Prevention Study. Diabetes. 2003;52:2532–8. 14. Uusitupa M, Peltonen M, Lindström J, Aunola S, Ilanne-Parikka P, Keinänen-Kiukaanniemi S, et al.; Finnish Diabetes Prevention Study Group. Ten-year mortality and cardiovascular morbidity in the Finnish Diabetes Prevention Study—secondary analysis of the randomized trial. PLoS One. 2009;4:e5656. 15. Lindström J, Ilanne Parikka P, Peltonen M, Aunola S, Eriksson JG, Hemiö K, et al.; Finnish Diabetes Prevention Study Group. Sustained reduction in the incidence of type 2 diabetes by lifestyle intervention: follow-up of the Finnish Diabetes Prevention Study. Lancet. 2006;368:1673–9. 16. Lindström J, Peltonen M, Eriksson JG, Louheranta A, Fogelholm M, Uusitupa M, et al. High- fi bre, low-fat diet predicts long-term weight loss and decreased type 2 diabetes risk: the Finnish Diabetes Prevention Study. Diabetologia. 2006;49:912–20. 17. Ilanne-Parikka P, Laaksonen DE, Eriksson JG, Lakka TA, Lindstr J, Peltonen M, et al.; Finnish Diabetes Prevention Study Group. Leisure-time physical activity and the metabolic syndrome in the Finnish diabetes prevention study. Diabetes Care. 2010;33:1610–7. 18. Tuomilehto H, Peltonen M, Partinen M, Lavigne G, Eriksson JG, Herder C, et al.; Finnish Diabetes Prevention Study Group. Sleep duration, lifestyle intervention, and incidence of type 2 diabetes in impaired glucose tolerance: the Finnish Diabetes Prevention Study. Diabetes Care. 2009;32:1965–71. 19. Herder C, Peltonen M, Koenig W, Sütfels K, Lindström J, Martin S, et al.; Finnish Diabetes Prevention Study Group. Anti-in fl ammatory effect of lifestyle changes in the Finnish Diabetes Prevention Study. Diabetologia. 2009;52:433–42. 11 Da Qing, Finnish DPP, Tripod, and Dream… 201

20. Makrilakis K, Liatis S, Grammatikou S, Perrea D, Stathi C, Tsiligrosa P, Katsilambros N. Validation of the Finnish diabetes risk score (FINDRISC) questionnaire for screening for undiagnosed type 2 diabetes, dysglycaemia and the metabolic syndrome in Greece. Diabetes Metab. 2011;37:144–51. 21. Schulze MB. Determinants for the effectiveness of lifestyle intervention in the Finnish Diabetes Prevention Study: response to Lindstrom et al. Diabetes Care. 2008;31:857–62. 22. Uusitupa MI, Stančáková A, Peltonen M, Eriksson JG, Lindström J, Aunola S, et al. Impact of positive family history and genetic risk variants on the incidence of diabetes the Finnish Diabetes Prevention Study. Diabetes Care. 2011;34:418–23. 23. McCarthy MI, Zeggini E. Genome-wide association studies in type 2 diabetes. Curr Diab Rep. 2009;9:164–71. 24. Grant RW, Hivert M, Pandiscio JC, Florez JC, Nathan DM, Meigs JB. The clinical application of genetic testing in type 2 diabetes: a patient and physician survey. Diabetologia. 2009;52:2299–305. 25. Voight BF, Scott LJ, Steinthorsdottir V, et al.; MAGIC Investigators; GIANT Consortium. Twelve type 2 diabetes susceptibility loci identiﬁ ed through large-scale association analysis. Nat Genet. 2010;42:579–89. 26. Grant RW, Hivert M, Pandiscio JC, Florez JC, Nathan DM, Meigs JB. The clinical application of genetic testing in type 2 diabetes: a patient and physician survey. Diabetologia. 2009;52:2299–305. 27. Lyssenko V, Jonsson A, Almgren P, et al. Clinical risk factors, DNA variants, and the development of type 2 diabetes. N Engl J Med. 2008;359:2220–32. 28. Cornelis MC, Qi L, Zhang C, et al. Joint effects of common genetic variants on the risk for type 2 diabetes in U.S. men and women of European ancestry. Ann Intern Med. 2009;150:541–50. 29. Todorova B, Kubaszek A, Pihlajamäki J, Lindström J, Eriksson J, Valle TT, et al. The G-250A promoter polymorphism of the hepatic lipase gene predicts the conversion from impaired glucose tolerance to type 2 diabetes mellitus: the Finnish Diabetes Prevention Study. J Clin Endocrinol Metab. 2004;89:2019–23. 30. Siitonen N, Pulkkinen L, Lindström J, Kolehmainen M, Eriksson JG, Venojärvi M, et al. Association of ADIPOQ gene variants with body weight, type 2 diabetes and serum adiponectin concentrations: the Finnish Diabetes Prevention Study. BMC Med Genet. 2011;12:5. 31. Buchanan TA, Xiang AH, Peters RK, Kjos SL, Marroquin A, Goico J, et al. Preservation of pancreatic beta-cell function and prevention of type 2 diabetes by pharmacological treatment of insulin resistance in high-risk Hispanic women. Diabetes. 2002;51:2796–803. 32. Buchanan TA, Xiang AH, Peters RK, Kjos SL, Berkowitz K, Marroquin A, et al. Response of pancreatic beta-cells to improved insulin sensitivity in women at high risk for type 2 diabetes. Diabetes. 2000;49:782–8. 33. Xiang AH, Peters RK, Kjos SL, Marroquin A, Goico J, Ochoa C, et al. Effect of pioglitazone on pancreatic beta-cell function and diabetes risk in Hispanic women with prior gestational diabetes. Diabetes. 2006;55:517–22. 34. Xiang AH, Hodis HN, Kawakubo M, Peters RK, Kjos SL, Marroquin A, et al. Effect of pioglitazone on progression of subclinical atherosclerosis in non-diabetic premenopausal Hispanic women with prior gestational diabetes. Atherosclerosis. 2008;199:207–14. 35. Gerstein HC, Yusuf S, Holman R, Bosch J, Pogue J; DREAM Trial Investigators. Rationale, design and recruitment characteristics of a large, simple international trial of diabetes prevention: the DREAM trial. Diabetologia. 2004;47:1519–27. 36. McDonald SD, Yusuf S, Sheridan P, Anand SS, Gerstein HC; Diabetes Reduction Assessment with Ramipril and Rosiglitazone Medication Trial Investigators. Dysglycemia and a history of reproductive risk factors. Diabetes Care. 2008;31:1635–8. 37. DREAM Trial Investigators, Bosch J, Yusuf S, Gerstein HC, Pogue J, Sheridan P, et al. Effect of ramipril on the incidence of diabetes. N Engl J Med. 2006;355:1551–62. 38. DREAM (Diabetes REduction Assessment with ramipril and rosiglitazone Medication) Trial Investigators, Gerstein HC, Yusuf S, Bosch J, Pogue J, Sheridan P, et al. Effect of rosiglitazone on the frequency of diabetes in patients with impaired glucose tolerance or impaired fasting glucose: a randomised controlled trial. Lancet. 2006;368:1096–105. 202 M. Glandt and Z. Bloomgarden

39. Boyko EJ, Gerstein HC, Mohan V, Yusuf S, Sheridan P, Anand S, et al.; DREAM Trial Investigators. Effects of ethnicity on diabetes incidence and prevention: results of the Diabetes REduction Assessment with ramipril and rosiglitazone Medication (DREAM) trial. Diabet Med. 2010;27:1226–32. 40. DREAM Trial Investigators, Dagenais GR, Gerstein HC, Holman R, Budaj A, Escalante A, et al. Effects of ramipril and rosiglitazone on cardiovascular and renal outcomes in people with impaired glucose tolerance or impaired fasting glucose: results of the Diabetes REduction Assessment with ramipril and rosiglitazone Medication (DREAM) trial. Diabetes Care. 2008;31:1007–14. 41. Bailey SD, Xie C, Do R, Montpetit A, Diaz R, Mohan V, et al.; DREAM Investigators. Variation at the NFATC2 locus increases the risk of thiazolidinedione-induced edema in the Diabetes REduction Assessment with ramipril and rosiglitazone Medication (DREAM) study. Diabetes Care. 2010;33:2250–3. 42. Lonn EM, Gerstein HC, Sheridan P, Smith S, Diaz R, Mohan V, et al.; DREAM (Diabetes REduction Assessment with ramipril and rosiglitazone Medication) and STARR Investigators. Effect of ramipril and of rosiglitazone on carotid intima-media thickness in people with impaired glucose tolerance or impaired fasting glucose: STARR (STudy of Atherosclerosis with Ramipril and Rosiglitazone). J Am Coll Cardiol. 2009;53:2028–35. 43. Hanley AJ, Zinman B, Sheridan P, Yusuf S, Gerstein HC; Diabetes Reduction Assessment with Ramipril and Rosiglitazone Medication (DREAM) Investigators. Effect of Rosiglitazone and Ramipril on {beta}-cell function in people with impaired glucose tolerance or impaired fasting glucose: the DREAM trial. Diabetes Care. 2010;33:608–13. 44. DREAM On (Diabetes Reduction Assessment with Ramipril and Rosiglitazone Medication Ongoing Follow-up) Investigators, Gerstein HC, Mohan V, Avezum A, Bergenstal RM, Chiasson JL, et al. Long-term effect of rosiglitazone and/or ramipril on the incidence of diabetes. Diabetologia. 2011;54:487–95. 45. Tuomilehto J, Wareham N. Glucose lowering and diabetes prevention: are they the same? Lancet. 2006;368:1218–9. 46. Davidson MB. Clinical implications of the DREAM Study. Diabetes Care. 2007;30:418–20. Chapter 12 Community Approaches to Diabetes Prevention

Ann Albright and David Williamson

Introduction

Preventing type 2 diabetes is a public health challenge that cannot be met by the clinical care sector acting alone. It requires complimentary and shared public health and clinical approaches that together achieve more than each can accomplish unaided (Fig. 12.1 ). The clinical sector must be involved in assessing patients’ risk for type 2 diabetes, discussing risk status with patients and their support network, referring (or encouraging) high-risk patients to participate in proven, community- based structured lifestyle programs, and, where necessary, prescribing medications for those at risk for and treating those who go on to develop diabetes. Glasgow et al. has de fi ned a public health approach to diabetes as “a broad, multidisciplinary perspective that is concerned with improving outcomes in all people who have (or are at risk for) diabetes, with attention to equity and the most effi cient use of resources in ways that enhance patient and community quality of life” [1 ] . The role of the public health sector is described by the Ten Essential Public Health Services ( [2 ] ; Fig. 12.2 ). When applied specifi cally to diabetes prevention, public health services include such actions as monitoring diabetes risk; informing, educating, and empowering people about prediabetes; mobilizing partnerships to reduce new cases of diabetes; linking people to proven diabetes prevention services; and

A. Albright , PhD, RD (*) Division of Diabetes Translation , Centers for Disease Control and Prevention , 2877 Brandywine Road, Williams Building/4770 Buford Highway, MS K-10 , Atlanta , GA 30341 , USA e-mail: [email protected] D. Williamson , PhD Hubert Department of Global Health , Rollins School of Public Health, Emory University , Atlanta , GA , USA

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 203 DOI 10.1007/978-1-4614-3314-9_12, © Springer Science+Business Media New York 2012 204 A. Albright and D. Williamson

Fig. 12.1 Prevention of type 2 diabetes. The community–clinic partnership model. Elements in the clinical component are adapted from the Chronic Care Model, MacColl Institute for Healthcare Innovation. The elements listed in this ﬁ gure are not intended to be all-inclusive, but to provide information on the kinds of elements contributed by each sector and shared across sectors (provided by the Centers for Disease Control and Prevention (CDC), Division of Diabetes Translation)

developing policies that support individual risk reduction and community change that makes it easier to practice healthy behaviors. Since diabetes risk progresses along a continuum from low risk to high risk, it is important that effective interventions exist along the continuum. Most of the evidence currently available for diabetes prevention involves those at high risk for type 2 diabetes. There is some evidence that provides insights into broader population risk reduction. Such a reduction may help those at high risk maintain healthy behaviors and prevent others from moving to high-risk status. This chapter examines evidence for community approaches to diabetes prevention and describes current national efforts in the United States to bring together the clinical and community (public health) sectors to prevent type 2 diabetes in high-risk persons. 12 Community Approaches to Diabetes Prevention 205

1. Monitor health status to identify community health problems. 2. Diagnose and investigate health problems and health hazards I the community. 3. Inform, educate, and empower people about health issues. 4. Mobilize community partnerships to identify and solve health problems. 5. Develop policies and plans that support individual and community health efforts. 6. Enforce laws and regulations that protect health and insure safety. 7. Link people to needed personal health services and ensure the provision of health care when otherwise unavailable. 8. Ensure a competent public health and personal health care workforce. 9. Evaluate effectiveness, accessibility, and quality of personal and population-based health services. 10. Research for new insights and innovative solutions to health.

Fig. 12.2 The Ten Essential Public Health Services, developed by the Core Public Health Functions Steering Committee, describes the public health activities that should be undertaken in all communities [2 ]

Lessons and Questions from U.S. Diabetes Prevention Translation Research

Several published research studies, carried out in real-world settings, have implemented modi fi ed versions of the lifestyle intervention developed by the US Diabetes Prevention Program (DPP) clinical research trial. These studies have very similar goals. For example, Aldana et al. ( [ 3 ] , p. 499) state, “The purpose of this study was to determine if the U.S. National Institutes of Health Diabetes Prevention Program (DPP) could be successfully implemented in a worksite setting.” Such studies are translation research studies, defi ned as “… applied research that strives to translate the available knowledge and render it operational in clinical and public health practice” ( [4 ] , p. 1794). Several of these studies were recently described [5 ] . It has been pointed out that “The availability of health-related interventions now in the marketplace exceeds by a considerable margin our societal ability to afford them” ( [ 6 ] , p. v). This observation suggests that economic costs are an important factor limiting wide-scale adoption of effective clinical and public health interventions. Therefore, a major focus of translation research for diabetes prevention is how best to use limited resources to deliver lifestyle intervention, while ensuring that weight loss is adequate to signi fi cantly decrease future incidence of—and health costs associated with—type 2 diabetes. To address this challenge, translation research studies use diverse approaches that affect key factors related to wide-scale implementation of diabetes prevention. These factors include participants’ diabetes risk; lifestyle curriculum; duration and intensity of intervention; attendance; body weight, diet, and physical activity monitoring; type of intervention staff; and weight loss achieved. 206 A. Albright and D. Williamson

A formal systematic review of all available diabetes prevention translation research studies is beyond the scope of this chapter. Instead, our purpose is to highlight some of the lessons learned, and still being learned and practical questions being raised, by examining eight US studies that represent varied approaches to translating the DPP clinical trial lifestyle intervention [3, 6– 13 ] . We have organized this narrative review along the key factors listed above. For comparison we list the original DPP trial in the ﬁ rst row of each table describing various aspects of the studies (Tables 12.1 – 12.3 ).

Basic Characteristics of the Translation Research Studies

The studies took place in diverse settings including community-based organizations such as Ys (also known as YMCAs) and churches, community-based outpatient clinics, health maintenance organizations, and worksites (Table 12.1). Sample sizes for the studies were generally modest, ranging from 8 participants to 295, with most studies having fewer than 100 participants. The studies’ “evaluation periods” (elapsed time between ﬁ rst and last outcome measurement) ranged from 3 months to 1 year. For some studies, the evaluation period was the same as the duration of the lifestyle intervention, while, for others, the evaluation period was substantially longer. Loss to follow-up of study participants during the evaluation period ranged from 0 to 43%, with most studies losing about 20% of their participants. The typical study participant was a moderately obese woman in her early-to-mid 50s. As in most weight studies, men were in the minority in nearly all studies. Since men and women are at similar risk of developing diabetes, more applied research is clearly needed on how to recruit and retain men for lifestyle intervention programs.

Factors Important for Broad and Effective Implementation

Participants’ Diabetes Risk

The proportion of study participants with diagnosed prediabetes ranged between 8 and 100%, with only three of the studies having 100% of participants with prediabetes (Table 12.1 ). Only one study [8 ] relied on a physician-reported diagnosis of prediabetes (based on either the fasting blood test or 2-h glucose tolerance test). The remaining seven studies performed their own diagnostic tests; three used capillary (“ ﬁ nger-stick”) blood tests (one was non-fasting [7 ] ) and three administered fasting plasma blood tests. Only one study [ 3 ] used the oral glucose tolerance test. All of the studies that performed their own diagnostic tests also used some form of riskfactor screening test to reduce the number of participants that required the more costly and less convenient diagnostic blood tests (not shown in table). 12 Community Approaches to Diabetes Prevention 207 Mean BMI at baseline vention vention 22 54.8 19 35.7 Loss to follow-up Loss to follow-up total over period evaluation (%) Age Men (%) of sample, 12 month for remainder 11 Month 1 Year 18 20 51.9 41 37.4 nr 30 35.7 2.8 Years 1 Year 6.5 1 Year 15 50.6 17 32 1 Year 6 Month 33.9 56.5 50 3 Month 0 53.6 20 0 32.0 3 Month for part 43 35.9 nr nr 34 54 nr 32.0 16 31.6 nr Duration of evaluation period glucose test glucose test and OGTT glucose test diagnosis of impaired glucose or fasting impaired glucose tolerance test glucose test glucose test glucose test

37 (8) 10 (100) plasma blood Fasting capillary blood Fasting 1,079 (100) plasma blood test Fasting 295 (52) Primary care provider 35 (89) 8 (100) Oral glucose tolerance 88 (42) capillary blood Fasting 93 (46) plasma blood Fasting plasma blood Fasting N (% prediabetes) Diagnosis of prediabetes Maintenance organizations church centers clinic, community health center, YMCA employer church hoods in poor urban community primary care practices Academic clinical Characteristics of some US research studies that have translated the Diabetes Prevention Program (DPP) trial lifestyle inter translated the Diabetes Prevention Characteristics of some US research studies that have Research Group Not reported In lifestyle arm of study In lifestyle arm of study McBride Local Health Davis-Smith African American nr Ackermann Ackermann Semi-urban YMCA Amundson 46 (100) Local outpatient Aldana Random capillary blood Boltri Worksite of local Seidel African American Kramer Distinct neighbor- Urban and rural Table Table 12.1 First author Setting a Knowler; Knowler; DPP Table 12.2 Characteristics of lifestyle interventions in some US translation research studies that have translated the DPP trial lifestyle intervention # Core Physical sessions Attendance Weight Diet activity Lifestyle interven- First author Curriculum Format (weeks) (%) Post-core monitoring monitoring monitoring tion staff Staff training Knowler; DPP Individual 16 (24) 15 (95) Monthly 1:1 Each Participants From Registered Not described DPP (1:1) or group session recorded partici- dietitians, Research sessions 7 days of pant logs masters level Group dietary exercise intake physiologists, each psychologists, week in health booklet educators Ackermann Closely Group 16 (20) 9 (57) Monthly Each nr nr YMCA staff with 2½-Day training modeled (8–12 group session associate or by former after DPP mem- sessions bachelor degree DPP trial bers) or equivalent staff training/ certi fi cation in exercise or health Amundson Closely Group 16 (16) 13 (83) Monthly Each From From Registered 2-Day training modeled (8–34 group session partici- partici- dietitians by former after DPP mem- sessions pant logs pant logs DPP trial bers) staff Aldana Closely Group and 16 (24) 11 (67) Monthly Measured From Logs and Registered nurses nr modeled 1:1 group by nurse partici- pedom- and certi fi ed after DPP sessions monthly pant logs eter health educator Boltri Closely Group (8 16 (16) 10 (62) None Frequency nr nr “Volunteer medical One training modeled mem- not personnel” session after DPP bers) reported Seidel Modi fi ed Group 12 (12) 6 (52) None Charted From Logs and A dietitian and an 2-Day training DPP (5–13 weekly partici- pedom- exercise by former mem- pant logs eter specialist DPP trial bers) staff Kramer Modi fi ed Group 12 8 (67) Monthly Self- From Monitoring Registered nurses, 2-Day training DPP (size (12– group moni- partici- books health educator, by staff that nr) 15) sessions tored in pant logs and registered originally for part charts pedom- dietitian, and developed of twice eter exercise DPP sample weekly specialist only McBride Modi fi ed Group 12 (12) nr Monthly Frequency Participant Weekly Registered nr DPP (size group not weekly exercise dietitian, nr) and 1:1 reported food logs clinical sessions records exercise physiologist Davis-Smith Modeled Group (10 6 (7) 5 (78) None Each From From “Healthcare One, 60-min after DPP mem- session partici- partici- professional training bers) pant logs pant logs (HCP)” session nr Not reported (this does not mean the activity did/did not occur, only that it was not reported in the publication) 210 A. Albright and D. Williamson

a 5% Loss (%) ³ PP trial lifestyle 7% Loss (%) ³ Time between ﬁ rst and last ﬁ between Time “Core” weight measure (weeks) loss (kg) Weight loss (%) Weight # Core sessions (weeks) Weight loss outcomes during the “core sessions” component of some US translation research studies that have translated the D loss outcomes during the “core sessions” component of some US translation research studies that have Weight Not reported Personal communication with Dr. Ronald Ackermann Ronald Ackermann Personal communication with Dr. intervention intervention Knowler; Knowler; DPP Research Group Ackermann 16 (24) 24 16 (20) 24 6.5 7 5.5 50 6 nr 36 59 Table Table 12.3 First author a Amundson Aldana Boltri Seidel Kramer McBride Davis-Smith nr 16 (16) 16 (24) 16 16 (16) 12 (12) 12 (14) 6 12 (7) (12) 24 16 12 12 12 24 6.7 2.9 3.4 nr 3.4 6.7 5.0 4.0 3.3 3.6 nr 3.5 45 4.6 3.8 nr nr 67 24 26 nr nr nr nr 52 46 nr nr 12 Community Approaches to Diabetes Prevention 211

Because the over-arching goal for public health translation of diabetes prevention is to ensure that biologically effective lifestyle programs are provided at a sustainable economic cost, participants’ risk status is of great importance. Persons with prediabetes have annual risks of developing diabetes that can be 10–15 times higher than people with normal glucose levels [14 ] . Because persons with normal glucose levels have substantially lower risk for future diabetes—and diabetes-related health costs—diabetes prevention programs that include participants with normal glucose levels are much less likely to save money [ 15] . It is unlikely that third party payers (private health insurers, employers, or federal, state, and local governments) will pay for new health interventions, such as diabetes prevention, unless there is conﬁ dence that the intervention will at least pay for itself by reducing future healthcare costs. The practical reality, however, is that most persons with prediabetes do not know they have it [16 ] . Diagnosis requires access to a health professional and medical laboratory. Therefore, an important challenge of community-based lifestyle intervention programs is how to effectively partner with the clinical sector to accurately and conveniently identify potential lifestyle intervention participants with prediabetes.

Lifestyle Curriculum and Intervention Format

All studies reported using versions of the original DPP lifestyle intervention curriculum that had been modiﬁ ed for use in group settings (Table 12.2). Some studies reported modifying the sequence of topics introduced during the 16-week “core” phase, such as introducing the topic of physical activity earlier [10, 11 ] , as well as introducing calorie-counting at the same time that fat-gram counting is discussed [ 10, 11 ] . Only one study [3 ] reported offering both individual and group sessions. Reported group sizes ranged from 5 [ 10 ] to 34 [8 ] , with most studies having groups between 8 and 12 participants. Two studies did not report group sizes [ 11, 12 ] . It is likely that differences among the curricula were much smaller than the simi- larities. Indeed, the DPP trial curriculum was developed for a very diverse study group (45% ethnic minorities) and had a very similar impact on reduction in diabetes incidence regardless of ethnic background [17 ] . To rapidly scale-up diabetes prevention in the United States, especially in nonacademic settings where research is not the main objective, access to a standardized, easily available lifestyle curriculum is essential.

Duration and Intensity of Intervention

Like the DPP trial, the intervention period in most studies had two phases, a “core” and a “post-core” phase (Table 12.2 ). The core phase in these studies consisted of intensive group sessions, often held weekly; the subsequent post-core phase consisted of less intensive group sessions held monthly. The purpose of the post-core 212 A. Albright and D. Williamson phase was intended to help participants improve and maintain weight loss, dietary, and physical activity behaviors learned during the core phase. The studies generally reported that group sessions lasted about 1 h (not shown). All of the studies included a core phase. Four of the studies had 16 core sessions, as did the DPP trial, and these were offered over a 16–24-week period; three studies included 12 core sessions offered over a 12–15-week period; and one study included 6 core sessions, offered over a 7-week period. Core sessions met no more frequently than weekly. Five of the eight studies included post-core phases with monthly sessions. Depending on the core phase’s duration, the post-core sessions occurred over the remaining 6–9 months. The impact on weight loss and economic costs of offering post-core sessions more or less frequently than monthly has not, to our knowledge, been studied in translation research. In addition, the utility and economic implications of offering post-core sessions for additional periods beyond the initial year of lifestyle intervention has not been studied in translation research. Only one of the four studies that used fewer than 16 core sessions explicitly reported a reason, and this was to reduce the cost of the intervention [ 11 ] . An additional reason may be for the convenience of the participants, which might increase session attendance and ultimately weight loss.

Attendance

The absolute number of core sessions attended was highest in those studies with 16 sessions, ranging from 9 to 13 sessions attended. Mean attendance in studies offering 12 sessions ranged between 6 and 8, and, in the study with 6 core sessions, average attendance was 5 sessions. These limited data suggest that offering more core sessions will mean more attendance. Attendance has important implications for weight loss, which we discuss later in this chapter.

Body Weight, Diet, and Physical Activity Monitoring

A key component of successful behavioral weight loss programs is self-monitoring of body weight, diet, and physical activity, which involves daily weighing and daily logs ﬁ lled out by participants [ 18] . Further monitoring occurs during program sessions when participants are weighed and their diet and physical activity logs are reviewed by lifestyle intervention staff. Self-monitoring was one of the most emphasized components of the original DPP trial lifestyle intervention [19 ] . All but two of the studies reported that self-monitoring of weight, diet, and physical activity were included in the lifestyle intervention; three of the studies reported that pedometers were used, in addition to paper logs, for physical activity monitoring. 12 Community Approaches to Diabetes Prevention 213

Intervention Staff

Healthcare professionals (HCPs), including physicians, certifi ed diabetes educators, registered dietitians, nurses, and exercise specialists, can all play key roles in diabetes prevention in high-risk persons. HCPs can assess risk status through screening and diagnostic tests commonly performed in the clinical setting. HCPs can help patients who are identifi ed with prediabetes interpret test results and understand future implications of their risk status. Importantly, HCPs can ensure that these high-risk patients are actively and effectively referred to community-based lifestyle intervention programs in the local community. On follow-up visits, HCPs can review progress of the high-risk patients who participate in the community-based lifestyle intervention program and reinforce improvements in body weight and dietary behaviors that have been achieved. For those high-risk patients in whom lifestyle intervention is, for whatever reason, not effective, timely decision-making about starting pharmacotherapy is also under the purview of the HCP. Substantial improvements in the cost-effectiveness of the DPP lifestyle intervention are made by offering the intervention in group, rather than individual, format [20 ] . Further reductions in the cost of the lifestyle intervention can occur when less expensive staff deliver the intervention. The original DPP trial employed staff that had Master’s degrees in clinical disciplines (Table 12.2 ). In contrast, one of the translation studies used regular employees of a local YMCA who had Associate or Bachelor degrees in health-related areas [7 ] . The other studies commonly employed dietitians, nurses, health educators, or exercise specialists. Some studies reported using “volunteer medical personnel” or “healthcare professionals,” not otherwise described. The study that used Y staff to deliver the intervention reported that “… the hourly wage of Y group instructors was approximately half that of behavioral experts in the DPP” [7 ] . In a separate publication, the same study reported the total 1-year cost for the lifestyle intervention including supplies, personnel time, and program administration was $275–325/participant [ 21] . The study by Kramer et al. [ 11 ] , which used HCPs, including registered dietitians, nurses, health educators, and exercise specialists, to deliver the intervention, estimated the cost/participant of a 1-year program with 12 core sessions and 9 post-core sessions was about $300. None of the other studies reported program costs. It is noteworthy that the DPP trial lifestyle intervention reported that the fi rst year cost was $1,399/participant [20 ] . Six of the eight studies reported training of intervention staff. Training ranged in duration from a single 60-min session to 2½ days. Four of the studies used expert staff from the original DPP study to conduct training. Academic or clinical certi fi cation may not be necessary to be an effective “lifestyle coach.” Rather, empa- thy and group leadership skills may be more important attributes for effective lifestyle intervention [22 ] . If so, further reductions in program cost might be achieved, without reducing effectiveness, by using intervention staff without clinical quali fi cations or college degrees. To effectively scale-up diabetes prevention in the United States, an accessible and cost-effective national system must be developed to train the large numbers of lifestyle intervention staff. This effort requires that a signi fi cant cadre of “master trainers” be developed to train intervention staff. 214 A. Albright and D. Williamson

Weight Loss Achieved

The DPP trial found that weight loss was the single most important factor in reducing the incidence of diabetes in high-risk persons, with those losing 5 kg or more experiencing a 58% reduction in 3-year diabetes incidence [23 ] . In addition, after statistically adjusting for changes in diet and physical activity, the DPP trial found that for every 1 kg of weight loss there was a 16% reduction in risk of developing diabetes, but the impact of diet and physical activity was not statistically signi fi cant after adjusting for weight loss. This occurred because weight loss is the key variable that mediates reduction in diabetes incidence brought about by improvements in diet and physical activity. In the absence of signifi cant weight loss, however, physical activity still had an impact on diabetes risk, albeit somewhat less than losing 5 kg or more. Among the 30% of DPP trial lifestyle participants who lost <3.5 kg and reported meeting the 150 min/week physical activity goal, there was a 44% reduction in 3-year diabetes incidence [23 ] . In the DPP trial, brisk walking was the recommended form of physical activity. This underscores the importance of lifestyle intervention programs emphasizing physical activity goals for all participants. In the DPP trial, a 5 kg weight loss was equivalent to a 5% weight loss. Therefore, translation studies reporting average weight losses of 5 kg/5% or more are likely to be successful in reducing diabetes incidence. In these eight translation research studies, the time periods over which weight loss was measured varied from 3 months to 1 year. Thus, for purposes of comparison, we identifi ed the weight losses achieved during each study’s core intervention phase, which ranged from 3 to 6 months duration (Table 12.3 ). The core phase is critical in weight loss because most persons reach their maximum loss 5–6 months after starting a lifestyle program [24 ] , and an initial weight loss of 5% or more is predictive of longer-term weight maintenance [25 ] . Four of the eight translation research studies reported weight losses of 5 kg/5% or greater during the core phase, and three of these four studies had 16 core sessions. Studies with 16 core sessions reported the highest incidence of both 7 and 5% weight loss. However, the eight studies varied in mean BMI, proportion of male participants, losses to follow-up, and proportion of participants that had prediabetes, as well as other potentially important characteristics that could confound the association between number of core sessions and achieved weight loss.

Questions

This review has identifi ed several practical questions about implementing biologically effective and economically sustainable lifestyle intervention programs for diabetes prevention in high-risk persons that would be helpful for further re fi nement of what is known about preventing type 2 diabetes. These questions include: How best to identify participants with diagnosed prediabetes? Which of the modifi ed DPP curricula are most effective and how to make them widely available? What should be the duration and frequency of core and post-core phases of the lifestyle intervention? 12 Community Approaches to Diabetes Prevention 215

What level of education and training is necessary to be an effective lifestyle interventionist? How can a workforce of lifestyle program staff be rapidly trained to meet the needs of the large US population of persons with prediabetes? Answering these questions will clearly involve translation research, but it is not clear that these questions can be answered by small, independent studies whose design and participants vary in characteristics that are likely to confound any infer- ences that can be drawn. For example, to convincingly answer the question about the optimal number of core sessions, lifestyle participants could be randomized to 8-, 12-, and 16-core sessions. In addition, participants could be further randomized— within each category of core sessions—to different types of lifestyle staff such as lay persons or clinical specialists (e.g., dietitians, exercise physiologists, or nurses). Such a study would be extremely valuable if data on economic costs were collected. The study would of course require a large number of participants to have adequate statistical power. A multicenter study could maximize the expertise of investigators, improve recruitment of participants, and achieve standardized implementation of study design.

U.S. National Diabetes Prevention Program

While questions remain about how to maximize the prevention of type 2 diabetes, the urgency of the growing incidence of diabetes demands that action is taken to implement what is currently known to be effective. In 2009, the U.S. Centers for Disease Control and Prevention (CDC) began to plan how to take the proven intervention from the DPP trial and fi ndings from translation studies to scale to achieve a population impact in diabetes prevention. The goal is to, in collaboration with community-based organizations that have necessary infrastructure, health payers, health professionals, public health, academia, and others, systematically scale the translated model of the DPP for high-risk persons, to reduce the incidence of type 2 diabetes in the United States. The strategic approach CDC has developed to help accomplish this goal includes four components: • Training—helping train the workforce that can implement the lifestyle intervention cost effectively (Essential Public Health Service #8) • Program Recognition—setting standards that will help ensure program quality and consistency (Essential Public Health Service #9) • Intervention Sites—implementing sites that will deliver the intervention (Essential Public Health Services #4, 5, 7) • Health marketing—raising awareness among both healthcare providers and high- risk populations to increase referral and use of the intervention (Essential Public Health Services #3, 4, 5) ( http://www.cdc.gov/diabetes/consumer/prevent.htm ) During this same year, Indiana University, CDC, and the National Institutes of Health convened a group that included third-party payers, state public health offi cials, community-based organizations, and other Federal health offi cials to 216 A. Albright and D. Williamson examine how to overcome barriers to implementing diabetes prevention on a national scale. This meeting provided insights which forged important partnerships for increasing access to cost-effective, community-based diabetes prevention services. In March 2010, Congress passed legislation that specifi cally addresses diabetes prevention through H.R. 3590—the Patient Protection and Affordable Care Act, SEC. 3999V-3 National Diabetes Prevention Program. The legislation authorizes CDC to manage the National Diabetes Prevention Program and establish a network of evidence-based lifestyle intervention programs for those at high risk of developing type 2 diabetes. Speci fi cally, the legislation states that the Program will include (1) A grant program for community-based diabetes prevention model sites; (2) A program within the CDC to determine eligibility of entities to deliver community- based diabetes prevention services; (3) A training and outreach program for lifestyle instructors; and (4) Evaluation, monitoring, and technical assistance, and applied research carried out by the CDC [26 ] . Over the last year, signi fi cant progress has been achieved in each of the four components of the National Diabetes Prevention Program: (1) Training—CDC has established the Diabetes Training and Technical Assistance Center (DTTAC) at Emory University to provide assistance and coordination for training the workforce, particularly “master trainers.” DTTAC has worked with leaders in delivering the DPP curriculum to prepare a master trainer curriculum for training intervention staff, along with a standardized curriculum for program participants that will be easily accessible on the CDC and DTTAC websites; (2) Program Recognition— CDC has established the CDC Diabetes Prevention Recognition Program (DPRP). The DPRP’s purpose is assurance of quality and standardized reporting on performance of recognized programs, so that decisions about participant referral and participation, and program funding are based on accurate, reliable, and trustworthy information. In addition to assuring program quality, the DPRP will allow CDC to provide technical assistance and maintain a registry of organizations that are recognized for their ability to deliver effective diabetes prevention programs. CDC worked with a diverse group of stakeholders to develop the standards for diabetes prevention lifestyle programs and began accepting applications in early 2012; (3) Intervention Sites—the delivery of community-based diabetes prevention programs requires organizations with an infrastructure that supports consistent, effective program delivery and also requires consistent funding. The inaugural organizations in the National Diabetes Prevention Program that have stepped up to deliver the program and provide reimbursement are the Y and UnitedHealth Group (UHG), respectively. UHG is offering the program through the Y to fully insured members and self-funded employers covered by UHG health plans. UHG also has an agreement with Medica, a Minnesota-based health insurer, to offer the program to employer- sponsored plans in the state. As the DPRP is implemented, more organizations will become involved in delivering the program; (4) Health Marketing—effective programs are really only effective if people who would benefi t participate in them. This means that people who are at risk know their risk status and receive the support and guidance needed to help them reduce their risk. HCPs are a critical source for encouraging program participation. CDC is currently conducting focus tests and 12 Community Approaches to Diabetes Prevention 217 interviews with various groups that are at risk for diabetes and their family members, as well as health professionals, to help develop messages and tools that will facilitate program referral and participation. In addition, other health marketing efforts are underway by various groups to increase awareness of prediabetes and the National Diabetes Prevention Program.

Community/Environmental Policies That Support Diabetes Prevention Efforts

As demonstrated in the DPP trial, the main drivers of the lifestyle intervention’s impact on reducing diabetes incidence were modest weight loss and increased physical activity. In reality the major effort and challenge to achieve weight control and increase physical activity necessarily take place outside the lifestyle program sessions and in the real-world environment where participants live. Environmental policies that have impact at the community level should, in theory, help high-risk persons with prediabetes maintain behaviors necessary to control body weight and be physically active. At the same time, such policies could help prevent persons with normal glucose levels from developing prediabetes. Faith et al. have carefully reviewed the effectiveness of macro-level environmental approaches to reducing population levels of obesity [ 27 ] . Based on fi ve well- conducted experimental studies, they found strong evidence that price subsidies for healthier foods in fl uence food purchases, but not necessarily total caloric consumption or body weight. In both cafeteria and vending machine settings, a 50% reduction in prices of fruit, salad, and other low-fat foods led to as much as a threefold increase in consumption. A signifi cant impact of reduced food prices on the purchase of low fat items was also seen in a study carried out in the setting of a delicatessen- style restaurant. In a “review of reviews” that included over 400 observational and experimental studies it was concluded that “…availability and accessibility of healthy and less healthy foods are important for nutrition behaviours in youth and in adulthood; schools and worksites offer good opportunities to improve availability of healthful foods.” ( [ 28 ] , p. 53). The CDC Guide to Community Preventive Services includes thorough, evidence- based reviews of obesity and physical activity interventions in community settings [ 29 – 30] . For obesity prevention and control the Community Guide recommends the following interventions: behavioral interventions to reduce time spent in television and computer monitor viewing, computer or web applications of multi-component coaching or counseling interventions to reduce weight and to maintain weight loss, and weigh control programs in worksite settings. For increasing population-wide levels of physical activity, the Community Guide recommends the following interventions: community-wide campaigns, individually adapted health behavior change programs, social support interventions in community settings, and enhanced school- based physical education. 218 A. Albright and D. Williamson

Conclusion

Diabetes prevention is a goal that is too large and is impacted by many factors outside of health care to expect it to be accomplished by the clinical sector alone. Public health (community) approaches are necessary for the prevention of type 2 diabetes. These public health approaches include delivery of proven structured lifestyle interventions in community settings by various personnel and population-wide efforts to create healthier environments. The most widely effective evidence currently available for prevention of type 2 diabetes supports structured lifestyle interventions for those with a clinical diagnosis of prediabetes to achieve and maintain modest weight loss and increase physical activity. Consequently, this provides a critical opportunity to achieve a more effective partnership between these sectors to signi fi cantly reduce new cases of type 2 diabetes. The alternative is an unsustain- able increase in new cases of type 2 diabetes that bring with it excessive health and fi nancial burdens. The fi ndings and conclusions in this report are those of the authors and do not necessarily represent the of fi cial position of the Centers for Disease Control and Prevention (CDC).

References

1. Glasgow RE, Wagner EH, Kaplan RM, Vinicor F, Smith L, Norman J. If diabetes is a public health problem, why not treat it as one? A population-based approach to chronic illness. Ann Behav Med. 1999;21:159–70. 2. Centers for Disease Control and Prevention. National Public Health Performance Standards Program (NPHPSP). http://www.cdc.gov/nphpsp/essentialServices.html . Accessed 28 March 2011. 3. Aldana SG, Barlow M, Smith R, et al. The diabetes prevention program: a worksite experience. AAOHN J. 2005;53:499–505. 4. Narayan KMV, Gregg EW, Engelgau MM, et al. Translation research for chronic disease—the case of diabetes. Diabetes Care. 2000;23:1794–8. 5. Williamson DF, Marrero DG. Scaling up type 2 diabetes prevention programs for high risk persons: progress and challenges in the United States. In: Schwarz P, Reddy P, Greaves C, Dunbar J, Schwarz J, editors. Diabetes prevention in practice, World congress on prevention of diabetes. Dresden: Tumani Institute for Prevention Management; 2010. p. 69–81. ISBN 978-3-00-03070765-2. 6. Gold MR, Siegel JE, Russell LB, Weinstein MC, editors. Cost-effectiveness in health and medicine. New York: Oxford University Press; 1996. 425 pp. 7. Ackermann RT, Finch EA, Brizendine E, et al. Translating the diabetes prevention program into the community: the DEPLOY pilot study. Am J Prev Med. 2008;35:357–63. 8. Amundson HA, Butcher MK, Gohdes D, et al. Translating the diabetes prevention program into practice in the general community: ﬁ ndings from the Montana cardiovascular disease and diabetes prevention programs. Diabetes Educ. 2009;35:209. doi: 10.1177/0145721709333269 . Accessed 4 Aug 2011. 9. Boltri JM, Davis-Smith MY, Seale JP, et al. Diabetes prevention in a faith-based setting: results of translational research. J Public Health Manag Pract. 2008;14:29–32. 12 Community Approaches to Diabetes Prevention 219

10. Seidel MC, Powell RO, Zgibor JC, et al. Translating the diabetes prevention program into an urban medically underserved community. Diabetes Care. 2008;31:684–9. 11. Kramer MK, Kriska AM, Venditti EM, et al. Translating the diabetes prevention program: a comprehensive model for prevention training and program delivery. Am J Prev Med. 2009;37:505–11. 12. McBride PE, Einerson JA, Grant H, et al. Putting the diabetes prevention program into practice: a program for weight loss and cardiovascular risk reduction for patients with metabolic syndrome or type 2 diabetes mellitus. J Nutr Health Aging. 2008;12:745s–9. 13. Davis-Smith M. Implementing a diabetes prevention program in a rural African American Church. J Natl Med Assoc. 2007;99:440–6. 14. Gerstein HC, Santaguida P, Raina P, et al. Annual incidence and relative risk of diabetes in people with various categories of dysglycemia: a systematic overview and meta-analysis of prospective studies. Diabetes Res Clin Pract. 2007;78:305–12. 15. Narayan KMV, Williamson DF. Prevention of type 2 diabetes: risk status, clinic, and community. J Gen Intern Med. 2009. doi: 10.1007/s11606-009-1148-9. Published online 5 Nov 2009. Accessed 4 Aug 2011. 16. Geiss LS, James C, Gregg EW, et al. Diabetes risk reduction behaviors among U.S. adults with prediabetes. Am J Prev Med. 2010. doi: 0.1016/j.amepre.2009.12.029. Published online 2 March 2010. Accessed 4 Aug 2011. 17. Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346:393–403. 18. Burke LE, Wang J, Sevick MA. Self-monitoring in weight loss: a systematic review of the literature. J Am Diet Assoc. 2011;111:92–102. 19. The Diabetes Prevention Program (DPP) Research Group. The diabetes prevention program (DPP): description of lifestyle intervention. Diabetes Care. 2002;25:2165–71. 20. The Diabetes Prevention Program (DPP) Research Group. Costs associated with the primary prevention of type 2 diabetes mellitus in the diabetes prevention program. Diabetes Care. 2003;26:36–47. 21. Ackermann RT, Marrero DG. Adapting the diabetes prevention program lifestyle intervention for delivery in the community: the YMCA model. Diabetes Educ. 2007;33:1–6. 22. Katula J, Vitolins MZ, Rosenberger EL, et al. Healthy living partnerships to prevent diabetes (HELP PD): design and methods. Contemp Clin Trials. 2009. doi: 10.1016/j.cct.2009.09.002 . Accessed 4 Aug 2011. 23. Hamman RF, Wing RR, Edelstein SL, et al. Effect of weight loss with lifestyle intervention on risk of diabetes. Diabetes Care. 2006;29:2102–7. 24. Jeffrey RW, Wing RR, Mayer RR. Are smaller weight losses or more achievable weight loss goals better in the long term for obese patients? J Consult Clin Psychol. 1998;66:641–5. 25. Anderson JW, Konz EC, Frederich RC, et al. Long-term weight-loss maintenance: a meta- analysis of US studies. Am J Clin Nutr. 2001;74:579–84. 26. Affordable Care Act. 2010. http://www.gpo.gov/fdsys/pkg/PLAW-111publ148/pdf/ PLAW-111publ148.pdf . Accessed 25 March 2011. 27. Faith MS, Fontaine KR, Baskin ML, Allison DB. Toward the reduction of population obesity: macrolevel environmental approaches to the problems of food, eating, and obesity. Psychol Bull. 2007;133:205–26. 28. Brug J. Determinants of healthy eating: motivation, abilities and environmental opportunities. Fam Pract. 2008. doi: 10.1093/fampra/cmn063 . Published online 30 Sept 2008. 29. Guide to Community Preventive Services. Obesity prevention and control interventions in community settings. www.thecommunityguide.org/obesity/communitysettings.html . Accessed 4 Aug 2011. 30. Guide to Community Preventive Services. Promoting physical activity environmental and policy approaches. www.thecommunityguide.org/pa/environmental-policy/index.html . Accessed 4 Aug 2011. Chapter 13 Think Locally, Act Locally, Extend Globally: Diabetes Prevention Through Partnerships with Local Communities

Carol R. Horowitz and Brett Ives

Introduction

Preventing diabetes requires lifestyle modifi cations to change the most basic habits— how we eat and how active we are in our free time. From a patient perspective, these changes are recommended to address the results of a blood test showing an elevated sugar, with no associated symptoms, to reduce the risk of a disease they do not offi cially have. Many individuals and families confront added barriers to diabetes prevention, including inadequate access to healthy foods, physical activity opportunities, and healthcare services. In fact, among the root causes of prediabetes are “obesogenic” factors beyond their control, such as public policies and marketing strategies encouraging unhealthy eating and sedentary behavior, food pricing favoring calorie dense, nutrient poor foods, limited access to healthy lifestyle education and counseling, schools with no physical education, and unsafe neighborhoods thwarting inexpensive forms of exercise [ 1, 2] . Without fi nancial, environmental, family, and culturally appropriate support, diabetes prevention becomes even more of an uphill battle. The best efforts of well-meaning clinicians may also prove futile, without an understanding of the factors contributing to the lifestyles they aim to address (in the limited time they have), to prevent diabetes. And while there is a burgeoning body of literature on the clinical, research, and policy efforts to thwart diabetes, its overall prevalence and racial and ethnic disparities in prevalence continue to grow

C. R. Horowitz , MD, MPH (*) Departments of Health Evidence and Policy and Medicine , Mount Sinai School of Medicine , 1479 Madison Avenue , Box 1087 , New York , NY 10029 , USA e-mail: [email protected] B. Ives , MSN, NP, CDE Division of Endocrinology, Diabetes, and Bone Diseases, Department of Medicine , Mount Sinai School of Medicine , One Gustave L. Levy Place Box 1055/New York , NY 10029 , USA e-mail: [email protected]

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 221 DOI 10.1007/978-1-4614-3314-9_13, © Springer Science+Business Media New York 2012 222 C.R. Horowitz and B. Ives dramatically. This failure begs new approaches. The poorest and most marginalized communities bear the greatest burden of diabetes and have the fewest resources for diabetes prevention. One promising approach may be to work with and in these communities, to uncover new, practical ideas that resonate with and impact those living in diabetes epicenters and the clinicians who work with them. This chapter will discuss how partnerships with local communities can yield the most relevant and effective diabetes prevention programs and initiatives. We will describe ways to employ community expertise and partnered approaches to build community capacity and improve clinical systems to prevent diabetes. By harness- ing the expertise of the community of front-line clinicians, partnered approaches can also redesign clinical care, enabling clinicians to effectively and efﬁ ciently recognize and address root causes of high rates of prediabetes in their efforts to prevent diabetes among their patients.

Disparities in Diabetes Prevalence, Morbidity, and Mortality

Certain groups are disproportionately impacted by prediabetes, diabetes, and their sequelae. In the United States, this includes seniors, and urban, overweight/obese, and people of non-White race and ethnicity. The burden of diabetes and its associated costs falls disproportionately on Blacks, Latinos, and Native Americans, who are more obese and have nearly twice the diabetes prevalence and mortality rates than Whites [3, 4 ] . Half of Latinos and nearly half of Blacks born in the year 2000, and nearly one third of Whites will develop diabetes if adequate preventive measures are not taken [5 ] . It is also crucial to look beyond the US borders to future epicenters of diabetes. In fact, the United Nations labeled diabetes as the fi rst noninfectious global epidemic [ 6 ] . The prevalence of diabetes will increase by two- to fourfold in many developing countries, far outpacing the United States, and leading to a doubling of world-wide diabetes cases by 2030 [7 ] . Given that the number of adults with prediabetes is estimated to be triple the diabetes prevalence [ 4, 8] , there may well be over one billion adults with prediabetes within the next 2 decades. To develop robust diabetes prevention approaches locally, regionally, nationally, and internationally, it is necessary to recognize the groups at greatest risk, and uncover and address the root causes of this increased prevalence. Inequalities in diabetes prevalence and outcomes are, in large part, due to social determinants of health: the social, economic, political, and environmental conditions in which people live [9 ] . These conditions include housing, transportation, food access, social support, employment, and education. More specifi cally, socioeconomic status (income, education, and occupation) largely determines health behaviors and access to health care, which in turn affects health status [ 10 ] and health outcomes [ 11 ] . Social determinants of health greatly in fl uence diabetes prevalence, morbidity, and mortality on an international scale as well, with poorer countries and the poorer, more marginalized people within these countries starting to demonstrate epidemic levels of diabetes [7, 12 ] . 13 Think Locally, Act Locally, Extend Globally… 223

There is no doubt, therefore, that challenges abound to prevent diabetes among those disproportionately impacted and least prepared to bear the brunt of the disease. What can be done? Surely global efforts to improve living conditions, reduce poverty, and improve education would help stem the diabetes tide. Short of such huge-scale transformative projects, medical therapies and clinical support will be of benefi t, particularly among those with access to the highest quality services, but may not be suffi cient to reverse predicted trends. Many would argue that communities are the optimal places to begin the multifactorial interventions needed to trans- form places from epidemics for disease to models for wellness [13Ð 15 ] . To begin to plan such approaches, it may be useful to compare communities with large disparities in both health and socioeconomic status. One such example is found in the adjacent neighborhoods of East Harlem and the Upper East Side in northeastern Manhattan (Fig. 13.1 ). As demonstrated in Table 13.1 , East Harlem, which is a predominantly non-White, has the lowest median household incomes, the poorest self-rated health, and highest obesity prevalence and diabetes-related mortality in New York City (Table 13.1 ) [ 16Ð 19] . In stark comparison, the immediately adjacent Upper East Side is predominantly White, and is the wealthiest neighborhood in NYC, with the best self-rated health, and lowest rates of the above conditions in NYC [ 19, 20 ] . Health-promoting behaviors critical for the prevention of diabetes require access to healthy foods and physical activity. Compared to the Upper East Side, East Harlem residents have a lower percentage of stores selling healthy foods [ 21 ] . Being physically active is a challenge in East Harlem, with violent crime and impassable sidewalks making exercise less safe [ 22 ] . Cultural attitudes towards exercise and general fatigue among East Harlem residents also contribute to physical inactivity and unhealthy eating [23, 24 ] . Social, economic, political, and environmental conditions of the Upper East Side promote good health and stave off chronic disease, while these same factors led East Harlem to suffer from obesity, diabetes, and its complications. With numerous barriers to healthy eating and physical activity, diabetes prevention in neighborhoods such as East Harlem is very dif fi cult on an individual, family, and community level. Weight loss of 5Ð10% and increased physical activity for up to 150 min/week prove to delay or prevent diabetes [ 25Ð 28] . Through initiatives such as the YMCA Diabetes Prevention Program and East Harlem’s Project HEED (Help Educate to Eliminate Diabetes), lifestyle modi fi cation programs to support weight loss and physical activity are just beginning to appear in community settings [ 24, 29 ] . However, it remains to be seen whether traditional lifestyle modi fi cation programs can be implemented and sustained in underserved, low-income urban neighborhoods such as East Harlem or Tepito, Mexico City, and rural areas like Pineville, WV or Bihar, India. Even in modifi ed forms, structured lifestyle interventions require funding and infrastructure, precluding some communities’ ownership of or access to such a program. In this respect, the communities most affected by diabetes may not have access to the proven effective diabetes prevention models [30 ] . East Harlem’s clinicians are also challenged with managing high volumes of poorly insured patients, who differ tremendously in terms of culture and social class. 224 C.R. Horowitz and B. Ives

Fig. 13.1 Map of East Harlem and Upper East Side neighborhoods, New York City 13 Think Locally, Act Locally, Extend Globally… 225

Table 13.1 Population characteristics of East Harlem and the Upper East Side, New York City, 2006 East Harlem Upper East Side (n = 108,100) [ 18 ] (n = 218,200) [ 20 ] Race/ethnicity [18, 20 ] White, non-Hispanic 7 83 Black 33 3 Hispanic 55 6 Other 5 8 Persons living in poverty, % [ 18, 20 ] 38 7 Less than HS education, % [18, 20 ] 46 5 Rate health fair/poor, % [18, 20 ] 30 6 No physical activity, % [18, 20 ] 48 16 Adults with obesity, % [18, 20 ] 31 7 Adults with diabetes, % [18, 20 ] 15 3 Diabetes hospitalization rate/100,000 [19 ] 820 138 Diabetes death rate/100,000 [19 ] 47 10

This can make even the most rudimentary discussions about risk prevention, eating, and physical activity (things few clinicians receive adequate training in) even more dif ﬁ cult and less effective. East Harlem has the highest rates of diabetes hospitalizations considered preventable by adequate outpatient care, and the highest preventable death rate in NYC [19 ] . Upper East Side clinicians, whose patients often share their backgrounds, education levels, and incomes, also have the lowest rates of preventable diabetes hospitalizations and preventable deaths [ 19 ] . However, community-based work in underserved, multicultural neighborhoods can yield many promising approaches for diabetes prevention. Through partnership with community experts, diabetes prevention programs can be more accessible and relevant to its local community members [13, 14 ] . The community-partnered work can also enable clinicians of East Harlem to better understand and then bridge the gap between themselves and their patients, to provide more relevant and effective counseling and care.

Diabetes Prevention in a Real-Life, Community Context: Yvette’s Story

Diabetes prevention efforts are extremely dif fi cult; they require signi fi cant lifestyle and behavioral changes in people who generally feel no symptoms from being overweight or having mildly increased glucoses, for risk reduction proven at a population level, but that may not prove to benefi t an individual, no matter how great his or her effort. Most people need a great deal of family, community, and perhaps clinical 226 C.R. Horowitz and B. Ives support to make fundamental changes in how they eat and how active they are. These changes are even more daunting if a person is living in poverty, is underem- ployed, has insecure housing or food supply, lacks social support, or is a recent immigrant. In these settings, where diabetes and prediabetes are often more norms rather than exceptions, thoughts, feelings, and attitudes towards diabetes prevention are also very complex. A case study can illustrate these ideas. Yvette is a 35-year-old Black woman, born in East Harlem, who lives with her unemployed husband and three children. She works part-time as a cashier and cleans at two schools 40 min away by subway. Her annual household income is $18,720, which is above the median income for the neighborhood [16 ] . Her rent is $900/month for a one-bedroom apartment, leaving her $140/week for all other expenses, including food. Yvette’s “entire family,” including her husband, has diabetes. “I try to be there for my husband, help him with it, but I also want to avoid going down that road myself.” Yvette’s BMI is 32, and her HbA1c is 5.9. She has Medicaid, and regularly attends a local health center where her primary care physician encourages her to lose weight through diet (emphasizing fresh produce, whole grains and fi sh, rather than other starches and meats) and exercise (just walk more!). To fi nd fresh vegetables and whole grains, Yvette must walk 20 blocks or take three busses to reach the one supermarket in her neighborhood. Her local corner stores (bodegas) do not carry low-fat dairy, whole grain products, or fresh vegetables. In her bimonthly trip to the supermarket, she aims to stretch her weekly food budget of $50 for her family of fi ve to include healthier products. These, however, are far more costly than rice and beans, foods her family has eaten for generations, are easy for her to prepare, and which her children are happy to eat. She argues with her overweight son every night when he comes home from school with junk food. Yvette, who is skeptical of big corporations and media, says “how can I compete against the food commercials directed towards my children? … they have millions of dollars backing them up. It’s like David and Goliath. But I’m not sure this David is going to win.” Yvette is also trying to increase her physical activity, but at the end of the day, “I’m so fatigued, I hardly have enough time or energy to do the laundry and put the kids to bed.” The doctor referred her to a nutritionist, but in describing that visit, Yvette said, “she told me not to eat bagels, but I’ve never eaten a bagel in my life. When I asked her how big a portion size of plantain I should have, she did not know. When I said her suggestions sounded expensive, she seemed to act as if I just didn’t care enough to be healthy.” Patients like Yvette needs to be considered beyond her individual characteristics such as attitude, self-ef fi cacy, and adherence to diet and physical activity. Whether she develops diabetes is also related to a larger network of factors that many clinicians and policy-makers may not have been exposed to or readily understand, such as food insuffi ciency contributing to obesity [ 31] . Newer diabetes prevention strategies are incorporating nuanced, local knowledge that comes from those most directly impacted by diabetes and its consequences, from communities like East Harlem. These strategies can empower both Yvette’s community and the community of clinicians. 13 Think Locally, Act Locally, Extend Globally… 227

By paying attention to possible root causes of obesity and diabetes, and openly acknowledging how little we still understand, community people and clinicians can start to communicate and act as a uniﬁ ed force.

Community-Partnered (Hybrid Approaches) to Diabetes Prevention

Community is defi ned as a group of people, linked by social ties, that shares common perspectives or interests [ 32 ] . The group may or may not share a geographic location. Two main communities interact in preventing diabetes; local geographic communities and the community of healthcare clinicians. In recent years, responses to health disparities and disease prevention including diabetes often follow one of two paths; they target neighborhoods or cities and the social and cultural determinants of health, or they target clinicians and health centers, seeking to improve quality of care [ 15 ] . These disparate paths have failed to signifi cantly impact rates of diabetes, suggesting that a hybrid clinicalÐcommunity approach is needed. The overlapping space in this hybrid model is ripe for innovative partnerships between community and clinical spheres (see Fig. 13.2 ). Using a hybrid approach, researchers, clinicians, policy-makers (henceforth called “academics”), and community members can impact diabetes prevention efforts. Implicit in such work is shared ownership, decision-making, and benefi t among community members and academics [33 ] . The community “inside experts” benefi t from the partnership process through expanded local capacity, resources, and ability to advocate for local needs [34 ] . Such partnerships allow academics to gain the “insider perspective,” the voices and viewpoints of people in a local community, who directly experience diabetes and who may hold the key that unlocks effective prevention strategies [35 ] .

Fig. 13.2 Two paths to approach diabetes prevention (adapted from [15 ] ) 228 C.R. Horowitz and B. Ives

CommunityÐacademic hybrid approaches may also guide policymakers through the myriad interventions to improve healthy and limit unhealthy food consumption and improve physical activity. This is particularly important because some recent high-proﬁ le policy approaches have either had disappointing impact to date (such as calorie labeling in fast-food venues) [ 36 ] or have met with insurmountable resistance (such as sugary drink taxes) [37, 38] from both businesses and those concerned about public health. Policies such as the proposed sugar tax hold an assumption of what is best for a society, taking the element of choice away from the individual [ 38] . These approaches must be handled with caution in low-income, minority communities in which people already lack sufﬁ cient autonomy and opportunities. Otherwise, they may cause individuals to further mistrust and disengage from efforts to help them positively change their lifestyles. Through community partnerships, we can uncover factors that motivate people to engage in health prevention activities, and use this information to drive diabetes prevention initiatives [ 38 ] . Partnered efforts have become more widely adopted in the past 10 years, with a growing body of research on their effectiveness in understanding [ 23, 39, 40 ] and addressing diabetes disparities, from urban NYC, Chicago, Detroit and Los Angeles [ 41, 42 ] , to rural areas such as the Lower Mississippi Delta and Native American communities in Canada [43, 44 ] . Partnered approaches can be adopted in other countries suffering from the diabetes epidemic. Many developing nations, such as India and Brazil, have a strong tradition in community-based public health programs incorporating the lay health worker model, especially in areas with low access to clinical services and technology. CommunityÐclinical partnerships can fortify existing traditions of community-based health promotion, while placing a greater emphasis on clinical expertise in the process [45Ð 48 ] .

Community–Academic Partnerships to Promote Diabetes Prevention in Clinical Settings

Clinician Knowledge and Behavior

Community expertise can improve cultural and socioeconomic relevance of clinical diabetes care and education. Community mentors can help academics, particularly clinicians and administrators, think about diabetes prevention in new ways. In the case of Yvette, her healthcare providers have a sense for what her barriers to diabetes prevention efforts are, but lack most of the details to come up with relevant self-management plans. If there are mechanisms for clinicians to receive feedback, dieticians like Yvette’s can recognize how to better work with patients (i.e., discussing portion sizes of plantains, not bagels with Caribbean Latino patients). Similarly, if Yvette’s health providers understand the barriers to physical activity in the local community, they can better tailor ideas and support efforts for diabetes prevention. Thus, we will demonstrate how community partnerships can inform diabetes 13 Think Locally, Act Locally, Extend Globally… 229

prevention, with a focus on physical activity and nutrition, the cornerstones of diabetes prevention. Increased physical activity is critical to diabetes prevention, yet it is one of the most dif fi cult tasks for people to carry out. Many researchers and clinicians believe that lack of safe public space is a large barrier to physical activity for people in underserved urban neighborhoods. In East Harlem, community and academic partners chose to investigate common academic statements that people do not exercise because of safety concerns and cost of gym memberships (factors that did not resonate with many community partners). Two surveys of over 300 local residents each found that these were barriers reported by less than 20% of respondents, but that general fatigue, lack of motivation, and lack of interest were the most common barriers [23, 24 ] . This local knowledge may help guide strategies. In this case partners chose to develop new and extremely popular exercise classes that would motivate attendance (such as Zumba) and to build educational curricula focused on overcoming exercise inertia, rather than channeling limited resources to subsidizing gym memberships. Availability of healthy foods in underserved neighborhoods has been a fertile area of exploration using community partnerships in the past 10 years. It may be very dif fi cult to lose weight if fresh, healthy foods are inaccessible or cost prohibitive. Just as physicians would not prescribe a patient a medicine that is unavailable in a local pharmacy or that the patient cannot afford, clinicians must be aware that recommendations for healthy eating may not be adopted by patients—not because patients do not care—but because, like Yvette, they are untenable. While this observation seems quite obvious, the idea of a “food desert” was not always common knowledge. The East Harlem Diabetes Center of Excellence, a communityÐacademic coalition, learned that the affl uent and predominantly White Upper East Side’s corner stores, or bodegas, were over fi ve times more likely than East Harlem bodegas to carry recommended foods for diabetes prevention and control [ 21] . This study also found that East Harlem residents did have stores in close proximity to their homes that carry healthy food items, but that the neighborhood was overwhelmed with stores that have no healthy items, making it challenging for residents to fi nd the “good” stores. Nationally, partnerships have arisen to address healthy food access. In East Harlem, the partnership has informed the expansion of healthy “Greencart” mobile produce stands, an initiative to bring healthier foods to bodegas, and offering healthier choices at local food pantries. NYC’s Department of Health and Mental Hygiene is ensuring food stamps are accepted at many local farmers’ markets and offering coupons for vegetable and fruit purchases [49 ] . In East Detroit, the East Side Village Health Worker Partnership identi fi ed a lack of grocery stores selling fresh produce and created a monthly minimarket at locations throughout the neighborhood [ 50 ] . Clinicians can also use the fi ndings of community-partnered work on healthy food availability to better serve their patients. Recognizing that patients can only eat foods they can fi nd and afford (similarly to only taking medicines available and reasonably priced at their pharmacies), clinicians can point patients to the healthiest available options and suggest portion reduction or cost-neutral food substitutions (i.e., diet instead of regular soda). 230 C.R. Horowitz and B. Ives

Clinician–Patient Communication

Community partnerships can promote clinicianÐpatient communication and mutual understanding. Through building of long-term trusting relationships, in which clinicians own their cultural biases and seek to understand their patient’s beliefs and attitudes, clinicianÐpatient communication can improve drastically. With time, clinicians become active members in the community and in turn may deliver more effective care for diabetes prevention. Some approaches to promote clinicianÐcommunity dialogue include bringing community members in to teach medical students and residents about the local community [51 ] . Through clinician education on local cultures and economic realities, a doctor or nurse may better anticipate a patient’s needs, concerns, and values. A doctor may, for example, need to understand how Ramadan might affect their patient’s eating behaviors. By recognizing that the patient will be fasting during day- light hours and eating a large meal in the evening, they can suggest how to negotiate this in the setting of weight loss efforts [ 52 ] . Even if the doctor fails to anticipate Ramadan, if they have shown basic understanding of the community and cultural con- texts of the patient, the patient may feel more comfortable raising important issues. In building trusting clinicianÐpatient relationships, obesity stigma should also be at the front of clinicians’ minds when working with patients with prediabetes. A patient may be shamed into avoiding the doctor due to fear of being embarrassed by weight and personal appearance [53Ð 55 ] . With awareness of this potential concern, a clinician could focus on increasing exercise and healthy eating, without compounding the shame the patient feels regarding their weight.

Clinical Systems

Through community-partnered work, clinical systems can be developed and revised to improve its fi nancial, structural, and cultural-linguistic elements [ 56 ] . By making these aspects more community- and patient-centered and friendly, the clinical community better refl ects the larger community it belongs to. When possible, fl exible payment plans should be available for patients. Staff should be knowledgeable in assistance programs for medicines and other services. The clinical staff should be culturally and linguistically diverse, and the clinical system should promote this as a valued strength of the system. Staff should undergo relevant training in interpersonal communication and cultural competence. Finally, leaders should elicit and respond to feedback from patients [ 56 ] . Through these deliberate decisions in clinical design and operations, community members will be more likely to positively infl uence clinical practice and patients will be more likely to adopt newer, more relevant prevention strategies.

Clinical Environment

Not only the clinical encounters but also the clinical space should ideally refl ect the local community. Community partners can provide insight on all elements of the 13 Think Locally, Act Locally, Extend Globally… 231 clinic’s programmatic content and environment: clinical delivery, educational healthcare messages, sound, color, décor, and much more. The Southcentral Foundation (SCF), the Alaska Native-owned nonprofi t healthcare system, places community members, called “community-owners,” at the center of decision-making of their “whole system design.” [ 14] Community members designed and built SCF physical spaces to promote comfort and pride among patients [ 14 ] . In their highly ef fi cient and successful model, integrated clinical teams include traditional healers and chiropractors, as a re fl ection of local Alaska Native preference for holistic approaches to health care.

Uncovering Novel Barriers to Diabetes Prevention

Not only can community partnerships inform content of clinical care and education, they can also uncover barriers to diabetes prevention that were not even considered previously by researchers. Community partners within East Harlem’s Communities IMPACT Diabetes Center recognized that while health literacy is important, people simply may not be able to see well enough to choose healthy foods, take medications, and exercise. They therefore assessed the prevalence of visual impairment and vision-related physical environment in their neighborhood. Of the 555 respondents to a visual function survey, over half reported that vision makes it difﬁ cult to perform tasks including reading food and medicine labels [ 22 ] . The environmental survey also found that only half of sidewalks were in good condition, exacerbating challenges of walking due to visual impairment. Furthermore, many of the vision problems found were correctible with provision of glasses [22 ] . The Communities IMPACT Diabetes Center will ﬁ nd creative ways to share these data with healthcare providers in East Harlem, so that they may screen and coordinate care for these problems with visual function.

Improving Clinical Practice: Educating Clinicians on Standards of Diabetes Care

Communities of clinicians, involved in diabetes prevention, can be deployed in effective ways to enhance diabetes prevention efforts. Primary care practice-based research networks work to improve clinical care for diabetes with an emphasis on translation of ﬁ ndings to clinical and the community. Federally quali ﬁ ed community health centers bring care of undeserved, underinsured, and uninsured individuals through comprehensive primary care services. These primary care networks touch a large segment of people with prediabetes. If a person with prediabetes engages with the healthcare system, it is most likely with a primary care provider. Presumably, most healthcare providers proactively work with patients with prediabetes. A community-partnered project in East Harlem indicated otherwise. Two board members from the East Harlem Partnership for Diabetes Prevention learned that their blood glucose fell in prediabetes range; however, their 232 C.R. Horowitz and B. Ives primary care doctors told them that the blood glucose levels were normal. The board then developed and administered a clinician survey to 229 clinicians at two local hospitals and a community health center. Only one in ten clinicians could correctly indicate prediabetes fasting and postprandial glucose levels [ 57 ] . The communityÐ academic board then educated all East Harlem clinicians about prediabetes. In Southeast Raleigh, community and academic partners in Project DIRECT enacted a clinical diabetes quality improvement initiative in which clinicians were educated and monitored on whether they met all standards of diabetes care. While clinicians improved processes of care, outcomes (i.e., glycemic control) did not improve [ 58] . Future projects will need to evaluate how to further improved such initiatives so as to improve end points, such as weight loss and diabetes prevention.

Community–Academic Partnerships to Bring Clinical Diabetes Prevention Information and Tools to Communities

Just as community knowledge can enrich community-based diabetes prevention programs, clinical expertise can be brought to the community to promote diabetes prevention. Community partnerships can channel clinical knowledge for community capacity-building and resource development. They can also provide knowledge and tools on research design and evaluation that can bolster community advocacy work and in ﬂ uence policy [45 ] . In a community partnership approach, clinicians are located as part of, rather than apart from, the local community. As members of a community, clinicians can lend expertise to community-driven programs by bringing information about prediabetes to communities through lectures and discussions as well as by helping with health screenings. For example, community and academic partners in East Harlem found that the six groups all conducting community-based diabetes screenings used different techniques, had different cutoffs for abnormal glucoses, and gave different advice about results. They then worked to improve the quality of diabetes screenings and incorporate detection messages about prediabetes into the materials [59 ] . Another approach that grounds clinicians in their role as community members as well as generates dialogue is a “clinician corner” series. Bringing forth social and community health topics, doctors or nurses set up on street corners to discuss social and community health topics with community members, resulting in greater trust and understanding. Care navigators (i.e., community health workers and patient navigators) can enter communities and with their knowledge of health and health systems, improve diabetes control [60, 61] . The community navigator system could prove effective for prediabetes as well; navigators can inform community members on topics such as diabetes risk and signs and symptoms of diabetes, as well as facilitate routine primary care to promote early detection of and clinical support for prediabetes. Community partnerships increase capacity of community-based diabetes prevention education through clinical knowledge and tools. The local community experts then take the information to create relevant programs, most of them peer-led. In the 13 Think Locally, Act Locally, Extend Globally… 233 case of the Project HEED, the East Harlem Partnership for Diabetes Prevention trained peer diabetes prevention educators based on clinical diabetes education guidelines by the Stanford Chronic Disease Self-Management Program [24, 62 ] , which resulted in proven-effective strategy to identify people with prediabetes in community settings and help them achieve signi ﬁ cant and sustained weight loss [ 24] . The Healthy Connections Program in Detroit has held over 100 “house par- ties” that screen for and inform women on diabetes and hypertension [ 63 ] . In addition to their community “diabetes resource centers,” the Chicago Southeast Diabetes Community Action Coalition increased capacity among local social service agencies, schools, labor unions and businesses, so that they too can educate on diabetes prevention and control, with a larger goal of “changing community norms” towards diabetes prevention [41 ] .

Community–Academic Partnerships to Shape Public Policy for Diabetes Prevention

ClinicalÐcommunity partnerships often evolve to include other stakeholders, including social agencies, businesses, insurance companies, and local and state governments, which in turn can lead to innovative public policy that promotes diabetes prevention.

Supermarket Development Incentive Programs

The Fresh Food Financing Initiative of Pennsylvania was the fi rst statewide fi nancing program to increase supermarkets in underserved areas. This policy was the product of a Pennsylvania-based communityÐacademic partnership, which leveraged their fi ndings on Phildelphia’s food insecurity with the Philadelphia city Council, Acme Markets, and United Way of Southeastern Pennsylvania [64 ] . The East Harlem Partnership’s work also informed the NYC Departments of Health and City Planning Food Retail Expansion to Support Health (FRESH) Program, which provides fi nancial incentives for grocery store renovation and development in underserved neighborhoods [ 21, 65 ] .

Merchant Incentive Programs

Recognizing the importance of city planning and development, the community partnership African Americans Building a Legacy of Health (AABLH) and nonprofi t organization Community Health Councils brought their fi ndings on food insecurity to the City of Los Angeles Planning Department and the Los Angeles Metropolitan Transit Authority and City Planning [66 ] . Two major policy innovations emerged: a merchant incentive program for healthier food options and a city ordinance 234 C.R. Horowitz and B. Ives banning fast-food restaurants in low-income neighborhoods for at least 2 years [ 66 ] . A San Francisco-based coalition working on food access and diabetes joined forces with a local nonprofi t youth environmental justice organization focused on reducing tobacco use among youth [ 67 ] . The partnership reduced tobacco subsidiary food products and tobacco advertisements and placed healthier foods in corner stores in a low-income neighborhood. With four city agency partners, coalition members are working with a California State assemblyman on legislation to replicate this program throughout California [67 ] .

Conclusion: Next Steps for Community Partnerships in Diabetes Prevention

There is a needed shift in collective consciousness to even slow the diabetes epidemic. Deeply ingrained policies, pervasive marketing, and poverty work against the individual or family in the huge efforts required to prevent diabetes. Due to the complexity of the diabetes crisis, no magic bullet will work for every person and place. However, any diabetes prevention work should begin with the community, the people living the realities of diabetes and prediabetes. Community-engaged models bring together community members and the clinician community to approach diabetes prevention from multiple angles. As society becomes ever more global, our efforts to address the diabetes epidemic need to start with both thinking and acting locally, then translating the approach to other societies around the world that are suffering from diabetes. The lessons learned from these partnerships can then inform large-scale, effective, sustainable diabetes prevention initiatives and policies.

References

1. Wallinga D. Agricultural policy and childhood obesity: a food systems and public health commentary. Health Aff (Millwood). 2010;29(3):405Ð10. doi: 10.1377/hlthaff.2010.0102 . 2. Harrington J, Friel S, Thunhurst C, Kirby A, McElroy B. Obesogenic island: the ﬁ nancial burden of private transport on low-income households. J Public Health (Oxf). 2008;30(1):38Ð44. doi:10.1093/pubmed/fdm084 . 3. Cowie CC, Rust KF, Byrd-Holt DD, et al. Prevalence of diabetes and impaired fasting glucose in adults in the U.S. population: National Health and Nutrition Examination Survey 1999- 2002. Diabetes Care. 2006;29(6):1263Ð8. doi: 10.2337/dc06-0062 . 4. Centers for Disease Control. CDC: National Diabetes Fact Sheet. 2011. http://www.cdc.gov/ diabetes/pubs/pdf/ndfs_2011.pdf . 6/3/2011. 5. Tuomilehto J, Lindstrom J, Eriksson JG, et al. Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med. 2001;344(18):1343Ð50. doi: 10.1056/NEJM200105033441801 . 6. United Nations Secretary General. Prevention and control of non-communicable disease. 2010. http://www.who.int/mediacentre/news/notes/2010/noncommunicable_diseases_20100514/en/ index.html . Retrieved on 4/9/11. 7. Wild S, Roglic G, Green A, Sicree R, King H. Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care. 2004;27(5):1047Ð53. 13 Think Locally, Act Locally, Extend Globally… 235

8. Cowie CC, Rust KF, Ford ES, et al. Full accounting of diabetes and pre-diabetes in the U.S. population in 1988-1994 and 2005-2006. Diabetes Care. 2009;32(2):287Ð94. doi: 10.2337/ dc08-1296 . 9. Marmot M, Wilkinson R, editors. Social determinants of health. New York: Oxford University Press; 1999. 10. Adler NE, Newman K. Socioeconomic disparities in health: pathways and policies. Health Aff (Millwood). 2002;21(2):60Ð76. 11. Pamuk E, Makuc D, Heck K, Reuben C, Lochner K. Socioeconomic status and health chart- book. Health, United States, 1998. Hyattsville: National Center for Health Statistics; 1998. 12. Marmot M. Social determinants of health inequalities. Lancet. 2005;365(9464):1099Ð104. doi: 10.1016/S0140-6736(05)71146-6 . 13. Bell J, Lee M. Why place and race matter. 2010. http://www.policylink.org/atf/cf/%7B97c6d565- bb43-406d-a6d5-eca3bbf35af0%7D/WPRM%20EXSUM%20(LORES).PDF . Retrieved on 4/9/11. 14. Gottlieb K, Sylvester I, Eby D. Transforming your practice: what matters most. Fam Pract Manag. 2008;15(1):32Ð8. 15. Horowitz CR, Lawlor EF. Community approaches to addressing health disparities. Report for the Institute of Medicine; 2008. 16. U.S. Census Bureau. Census 2000. Summary fi le 1: census of population and housing; 2000. 17. Thorpe LE, Mostahari F, Berger DK, Cobb LK, Helgerson SD, Frieden TR. Diabetes is epidemic: New York City. Vital Signs. 2003;2:1Ð4. 18. Olson EC, Van Wye G, Kerker B, Thorpe L, Frieden TR. Take care East Harlem. NYC Community Health Pro fi les. 2006;21(42):1Ð16. 19. Kim M, Berger D, Matte T. Diabetes in New York City: Public Health Burden and Disparities. New York: New York City Dept of Health and Mental Hygiene; 2006. 20. Olson EC, Van Wye G, Kerker B, Thorpe L, Frieden TR. Take care Upper East Side. NYC Community Health Pro fi les. 2006;23(2):1Ð16. 21. Horowitz CR, Colson KA, Hebert PL, Lancaster K. Barriers to buying healthy foods for people with diabetes: evidence of environmental disparities. Am J Public Health. 2004;94(9):1549Ð54. 22. Ramos M, Macmillan T, Horowitz C. Vision impairment among older adults in an urban, low-income neighborhood: implications for diabetes prevention and management. JGIM Suppl. (in press). 23. Horowitz CR, Williams L, Bickell NA. A community-centered approach to diabetes in East Harlem. J Gen Intern Med. 2003;18(7):542Ð8. 24. Parikh P, Simon EP, Fei K, Looker H, Goytia C, Horowitz CR. Results of a pilot diabetes prevention intervention in East Harlem, New York City: Project HEED. Am J Public Health. 2010;100 Suppl 1:S232Ð9. doi: 10.2105/AJPH.2009.170910 . 25. Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346(6):393Ð403. doi: 10.1056/ NEJMoa012512 . 26. American Diabetes Association. Standards of medical care in diabetes—2011. Diabetes Care. 2011;34 Suppl 1:S11Ð61. doi: 10.2337/dc11-S011 . 27. Biuso TJ, Butterworth S, Linden A. A conceptual framework for targeting prediabetes with lifestyle, clinical, and behavioral management interventions. Dis Manag. 2007;10(1):6Ð15. doi: 10.1089/dis.2006.628 . 28. Franz MJ. The evidence is in: lifestyle interventions can prevent diabetes. Am J Lifestyle Med. 2007;1(5):327Ð34. 29. Ackermann RT, Marrero DG. Adapting the diabetes prevention program lifestyle intervention for delivery in the community: the YMCA model. Diabetes Educ. 2007;33(1):69, 74Ð5, 77Ð8. doi: 10.1177/0145721706297743 . 30. Hernan WH, Brandle M, Zhang P, et al. Costs associated with the primary prevention of type 2 diabetes mellitus in the diabetes prevention program. Diabetes Care. 2003;26(1):36Ð47. 31. Drewnowski A, Specter SE. Poverty and obesity: the role of energy density and energy costs. Am J Clin Nutr. 2004;79(1):6Ð16. 32. Smith, MK. ‘Community’ in the encyclopedia of informal education, 2001. http://www.infed. org/community/community.htm . Retrieved on 6/10/11. 236 C.R. Horowitz and B. Ives

33. Wallerstein NB, Yen IH, Syme SL. Integration of social epidemiology and community-engaged interventions to improve health equity. Am J Public Health. 2011;101(5):822Ð30. doi: 10.2105/ AJPH.2008.140988 . 34. Israel BA, Schulz AJ, Parker EA, Becker AB. Review of community-based research: assessing partnership approaches to improve public health. Annu Rev Public Health. 1998;19:173Ð202. doi: 10.1146/annurev.publhealth.19.1.173 . 35. Minkler M. Ethical challenges for the “outside” researcher in community-based participatory research. Health Educ Behav. 2004;31(6):684Ð97. doi: 10.1177/1090198104269566 . 36. Elbel B, Gyamfi J, Kersh R. Child and adolescent fast-food choice and the infl uence of calorie labeling: a natural experiment. Int J Obes (Lond). 2011;35(4):493Ð500. doi: 10.1038/ ijo.2011.4 . 37. Brownell KD, Frieden TR. Ounces of prevention—the public policy case for taxes on sugared beverages. N Engl J Med. 2009;360(18):1805Ð8. doi: 10.1056/NEJMp0902392 . 38. Buchanan DR. Autonomy, paternalism, and justice: ethical priorities in public health. Am J Public Health. 2008;98(1):15Ð21. doi: 10.2105/AJPH.2007.110361 . 39. Wallerstein NB, Duran B. Using community-based participatory research to address health disparities. Health Promot Pract. 2006;7(3):312Ð23. doi: 10.1177/1524839906289376 . 40. Russell BE, Gurrola E, Ndumele CD, et al. Perspectives of non-Hispanic Black and Latino patients in Boston’s urban community health centers on their experiences with diabetes and hypertension. J Gen Intern Med. 2010;25(6):504Ð9. doi: 10.1007/s11606-010-1278-0 . 41. Giachello AL, Arrom JO, Davis M, et al. Reducing diabetes health disparities through community-based participatory action research: the Chicago Southeast Diabetes Community Action Coalition. Public Health Rep. 2003;118(4):309Ð23. 42. Gold fi nger JZ, Arniella G, Wylie-Rosett J, Horowitz CR. Project HEAL: peer education leads to weight loss in Harlem. J Health Care Poor Underserved. 2008;19(1):180Ð92. doi: 10.1353/ hpu.2008.0016 . 43. Downey LH, Castellanos DC, Yadrick K, et al. Capacity building for health through community-based participatory nutrition intervention research in rural communities. Fam Community Health. 2010;33(3):175Ð85. doi: 10.1097/FCH.0b013e3181e4bb58 . 44. Bisset S, Cargo M, Delormier T, Macaulay AC, Potvin L. Legitimizing diabetes as a community health issue: a case analysis of an Aboriginal community in Canada. Health Promot Int. 2004;19(3):317Ð26. doi: 10.1093/heapro/dah305 . 45. Horowitz CR, Robinson M, Seifer S. Community-based participatory research from the margin to the mainstream: are researchers prepared? Circulation. 2009;119(19):2633Ð42. doi: 10.1161/CIRCULATIONAHA.107.729863 . 46. World Health Organization. Action on the social determinants of health: a background paper prepared for the Commission on Social Determinants of Health; 2005. 47. Lehmann U, Sanders D. Community health workers: what do we know about them? The state of evidence on programmes, activities, costs and impact on health outcomes of using community health workers. Report for the WHO, Department of Human Resources for Health; 2007. 48. Djukanovic V, Mach EE. Alternative approaches to meeting basic health needs of populations in developing countries, a joint UNICEF/WHO Study; 1975. 49. Tester JM, Stevens SA, Yen IH, Laraia BL. An analysis of public health policy and legal issues relevant to mobile food vending. Am J Public Health. 2010;100(11):2038Ð2046 . 50. Schulz AJ, Zenk S, Odoms-Young A, et al. Healthy eating and exercising to reduce diabetes: exploring the potential of social determinants of health frameworks within the context of community-based participatory diabetes prevention. Am J Public Health. 2005;95(4):645Ð51. doi: 10.2105/AJPH.2004.048256 . 51. Ross PT, Wiley Cene C, Bussey-Jones J, et al. A strategy for improving health disparities education in medicine. J Gen Intern Med. 2010;25 Suppl 2:S160Ð3. doi: 10.1007/s11606-010- 1283-3 . 52. A fi fi ZE. Daily practices, study performance and health during the Ramadan fast. J R Soc Health. 1997;117(4):231Ð5. 13 Think Locally, Act Locally, Extend Globally… 237

53. Puhl RM, Heuer CA. Obesity stigma: important considerations for public health. Am J Public Health. 2010;100(6):1019Ð28. doi: 10.2105/AJPH.2009.159491 . 54. Puhl RM, Brownell KD. Psychosocial origins of obesity stigma: toward changing a powerful and pervasive bias. Obes Rev. 2003;4(4):213Ð27. 55. Puhl R, Brownell KD. Ways of coping with obesity stigma: review and conceptual analysis. Eat Behav. 2003;4(1):53Ð78. 56. Washington DL, Bowles J, Saha S, et al. Transforming clinical practice to eliminate racial- ethnic disparities in healthcare. J Gen Intern Med. 2008;23(5):685Ð91. doi: 10.1007/s11606- 007-0481-0 . 57. Sigworth S, Weiner A, Talavera S, Horowitz CR. Most clinicians do not recognize pre-diabetes: results of a survey. J Gen Intern Med. 2007;22 Suppl 1. 58. Din-Dzietham R, Porter ﬁ eld DS, Cohen SJ, Reaves J, Burrus B, Lamb BM. Quality care improvement program in a community-based participatory research project: example of Project DIRECT. J Natl Med Assoc. 2004;96(10):1310Ð21. 59. West B, Parikh P, Arniella G, Horowitz CR. Observations and recommendations for community- based diabetes screenings. Diabetes Educ. 2010;36(6):887Ð93. doi: 10.1177/0145721710386973 . 60. Babamoto KS, Sey KA, Camilleri AJ, Karlan VJ, Catalasan J, Morisky DE. Improving diabetes care and health measures among Hispanics using community health workers: results from a randomized controlled trial. Health Educ Behav. 2009;36(1):113Ð26. doi: 10.1177/1090198108325911 . 61. Castillo A, Giachello A, Bates R, et al. Community-based diabetes education for Latinos: the Diabetes Empowerment Education Program. Diabetes Educ. 2010;36(4):586Ð94. doi: 10.1177/0145721710371524 . 62. Lorig K, Ritter PL, Villa FJ, Armas J. Community-based peer-led diabetes self-management: a randomized trial. Diabetes Educ. 2009;35(4):641Ð51. doi: 10.1177/0145721709335006 . 63. Harvey I, Schulz A, Israel B, et al. The Healthy Connections project: a community-based participatory research project involving women at risk for diabetes and hypertension. Prog Community Health Partnersh. 2009;3(4):287Ð300. doi: 10.1353/cpr.0.0088 . 64. Giang T, Karpyn A, Laurison HB, Hillier A, Perry RD. Closing the grocery gap in underserved communities: the creation of the Pennsylvania Fresh Food Financing Initiative. J Public Health Manag Pract. 2008;14(3):272Ð9. doi: 10.1097/01.PHH.0000316486.57512.bf . 65. Department of City Planning, NYC Department of Health and Mental Hygiene, NYC Economic Development Corporation. Going to market: New York City’s neighborhood grocery store and supermarket shortage; 2009. 66. Lewis LB, Galloway-Gilliam L, Flynn G, Nomachi J, Keener LC, Sloane DC. Transforming the urban food desert from the grassroots up: a model for community change. Fam Community Health. 2011;34 Suppl 1:S92ÐS101. 67. Vasquez VB, Lanza D, Hennessey-Lavery S, Facente S, Halpin HA, Minkler M. Addressing food security through public policy action in a community-based participatory research partnership. Health Promot Pract. 2007;8(4):342Ð9. doi: 10.1177/1524839906298501 . Chapter 14 Global Challenge in Diabetes Prevention from Practice to Public Health

Peter E. H. Schwarz

Prevention of Type 2 Diabetes Mellitus

Prevention of type 2 diabetes (T2D) has become a major issue in health care during the last 10 years. In less than a decade we have gone from ascertaining that diabetes is preventable to implement diabetes prevention programmes at local, regional and national scale. There have been some notable landmarks. In 2001 the Finnish Diabetes Prevention Study and in 2002 the US Diabetes Prevention Program demonstrated that lifestyle modi fi cation interventions focused on losing weight, increasing physical activity and improving diet could reduce the risk of progression to diabetes by nearly 60% [1Ð 3 ] . The potential to prevent T2D in high-risk individuals by lifestyle intervention has been fi rmly established by these clinical trials. These studies have a strong focus on increased physical activity and dietary modi fi cation as well as weight reduction among overweight participants. The key issue seems to be a comprehensive approach to correct several risk factors simultaneously. Furthermore, long-term follow-up studies of lifestyle interventions, lasting for a limited time period, seem to have a long-lasting carry-over effect on risk factors and diabetes incidence. A major barrier to population-based screening for people at risk of diabetes has been the time-consuming oral glucose tolerance test. Using the results of risk factor studies conducted over a quarter of a century, Finnish researchers were able to create FINDRISC—an eight-item questionnaire, which has now been validated in several countries and allows simpli fi ed screening of large numbers of people [4, 5 ] . Another landmark was a profusion of published implementation trials including GOAL and the Saxon DPP in Europe [6 ] , the Greater Green Triangle DPP in Australia and programmes in Indianapolis, Pittsburgh and Montana in the United States [ 7 ] .

P. E. H. Schwarz, MD (*) Abteilung Prävention und Versorgung des Diabetes, Medizinische Klinik III , Universitätsklinikum Carl Gustav Carus, Technischen Universität Dresden , Fetscherstrasse 74 , 01307 Dresden , Germany e-mail: [email protected]

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 239 DOI 10.1007/978-1-4614-3314-9_14, © Springer Science+Business Media New York 2012 240 P.E.H. Schwa rz

A great challenge is scaling up from implementation trials to sizeable regional and national programmes. Finland has led the way with FIN-D2D, a large-scale implementation covering a quarter of the Finnish population [8 ] . Australia has the largest systematic programme in the world—Life! Taking action on diabetes in which a strictly deﬁ ned intervention is linked to a highly developed training programme and continuous quality improvement [ 9 ] .

The Global Need for Diabetes Prevention

Currently, we are experiencing a fast growing interest in the practice of diabetes prevention in several countries due to the fact that those countries are struggling with the current epidemic. The evidence for the prevention of diabetes is overwhelm- ingly good and a fast growing number of randomised controlled trials to investigate efﬁ cacy in diabetes prevention are currently followed by numerous implementation trials, which study the effectiveness in a real-world setting. This gains us a lot of knowledge about feasible prevention strategies but more and more also about barriers for implementation. This needs to be addressed by studying the ef ﬁ ciency to identify how to gain the best effect for most of the people on a population basis. Talking then about implementation, the availability of prevention programmes becomes a major fact. The chronic disease, which affects around 10% of the average population, needs to be addressed by preventive activities that are available to at least twice as many people. This will make it necessary to carefully plan the distribution of programmes and intervention concepts for diabetes prevention, which must be developed as easy as possible to reach this number of people. Only if all those strategic milestones are addressed, activities for the prevention of diabetes can have an effect on the disease prevalence on a long term. However, the way to address these milestones will be varying, depending on the economic situation of countries world wide, the healthcare structure, ethnicity and disease perception. All these aspects have to be taken into account if we plan to address the global challenge of diabetes prevention. Currently, there is an enormous chance, after the United Nations declaration “Unite for diabetes” in 2006, followed by the United Nations Summit in 2011, to address the need for prevention of non-communicable chronic diseases. Those activities can lead to the development of a global diabetes plan including a global diabetes prevention strategy. This action will be of high value to invite and motivate all relevant stakeholders for commitment to network together in order to address the needs for implementing diabetes prevention programmes.

From the Global Few to the Practice Strategy

In order to start the development of an intervention programme to prevent diabetes in clinical practice, we fi rst have to think about which milestones are necessary to reach the best effectiveness in a real-world setting. Those items were much depending on setting, 14 Global Challenge in Diabetes Prevention from Practice to Public Health 241 reimbursement structure and risk distribution in the environment, but certain basic items, which are directly related to effectiveness and ef fi ciency, include the following: 1. Evidence for diabetes prevention [10 ] (guideline) 2. Evidence for diabetes prevention practice [11 ] (implementation trial, practice guidelines) 3. Political support [12 ] (diabetes plan, prevention plan, educational activities, etc.) 4. Partners at different levels of care [ 12 ] (stakeholder involvement, multidisciplinary teams, etc.) 5. Adequate intervention concepts and material [4 ] (exchange with others, networking, etc.) 6. Training of the trainer (licence, reimbursement, work plan prevention) 7. Quality management in the process [13 ] (comparable QM, benchmarking) 8. Business planning in prevention including high-risk and public health approach [9 ] All the above practice aspects would be necessary in order to get into practice. To develop a prevention strategy, the programme must be well structured and easy to understand. As part of the programme, the managers have to fi nd the people where they are and go into the settings to identify the people with increased risk and focus on the individual empowerment of the participants in the intervention. Regular contact with the individuals seems to be a good indicator for success. On a structural level, the recruitment of an adequate number of lifestyle managers will be necessary as well as the use of screening tools, which are applicable in a population setting. The quality management will become relevant and important for the continuous evaluation, fi nally leading to tight prevention management.

What Is Needed Now?

What is needed now to successfully implement a diabetes prevention programme— there is a clear answer. Political support is needed and this requires the development of a national or international action plan for diabetes prevention, which needs involvement of a number of stakeholders on governmental and nongovernmental level as well as scienti fi c and practical input. Furthermore, practical guidance is needed and this includes the presentation of the evidence in the fi eld for diabetes prevention on the scientifi c and also practical level as well as the training of people to deliver preventive intervention. Two European funded projects DE-PLAN [14 ] and IMAGE [15 ] have been addressing the implementation process. Especially the IMAGE project was able to take a step ahead and to collate this information in a systematic manner. These include an evidence-based guideline on T2D prevention [ 10 ] , a toolkit for the prevention [ 16 ] and a paper on quality indicators in T2D prevention [17 ] . The information in these papers will represent a major further step in the work to make T2D prevention reality in Europe. The major objectives of the IMAGE project were the development of: ¥ European practice-oriented guidelines for the primary prevention of T2D ¥ A European curriculum for the training of prevention managers 242 P.E.H. Schwa rz

¥ European standards for the quality management in diabetes prevention ¥ A European e-health training portal for prevention managers As part of the IMAGE project a practical guideline “Toolkit” was developed for the prevention of T2D. This toolkit is meant for people who would like to implement a diabetes prevention programme, like educators and physicians, but also stakeholders and politicians. The toolkit includes all that is necessary to build up a diabetes prevention programme covering management, fi nancial, interventional and quality assurance aspects. Furthermore, IMAGE developed a curriculum for the training of prevention managers. This training includes a 7-day curriculum for educators to be qualifi ed and to learn necessary skills to deliver preventive intervention. Especially the toolkit is a landmark, because it combines international expertise on scientifi c and practical level and enables practical implementation with a scientifi c evidence basis.

Toolkit for the Prevention of Type 2 Diabetes

The major output of the IMAGE project—relevant for prevention practice—is the practical guideline called “Toolkit for the prevention of type 2 diabetes”. This toolkit is meant for all people involved with diabetes prevention: those working in primary and specialised healthcare services, physicians, physical activity experts, dieticians, nurses, teachers, but also stakeholders and politicians. In a condensed form the toolkit [ 18 ] includes the essence of what is necessary to build up a diabetes prevention programme covering management, fi nancial, intervention and quality assurance aspects and refers to the latest evidence in the science of diabetes prevention and allows translating this knowledge into practice. The toolkit addresses issues such as how to budget and fi nance a prevention programme and how to identify people at risk. The core of the toolkit describes elements of an effective lifestyle intervention programme. A process model for supporting lifestyle behaviour change is presented and described in its phases (motivation, action and maintenance). The toolkit gives the core goals of lifestyle (physical activity and diet) and gives practical instructions about how to address these with the client. Other behaviours to consider in diabetes prevention are, e.g. smoking, stress/depression and sleeping patterns. The toolkit fi nishes with an overview on how to evaluate intervention programmes and how to establish quality assurance. It provides several recommendations that may help with planning T2D prevention programmes. The toolkit aims to provide a good balance between clear, accurate information and practical guidance. It is not intended to be a comprehensive source of information. Speci fi cally, detailed instructions about how to achieve and maintain weight reduction, which evidently is one of the main issues in diabetes prevention, are not given because local and national guidelines as well as other information are abundantly available elsewhere. Furthermore, intervention delivery staff is assumed to have basic knowledge about e.g. diet and physical activity and their health effects and about supporting behaviour change. Finally, the toolkit is not designed to be used as 14 Global Challenge in Diabetes Prevention from Practice to Public Health 243 intervention material to be delivered directly to those participating in prevention interventions, although it does contain some examples of information sheets and materials which might be used with participants.

Content of the Toolkit

The toolkit starts with an executive summary including the rationale for diabetes prevention. It is followed by a chapter representing the background (T2D prevalence, risk factors, consequences, evidence of successful prevention) and giving instructions about the planning and development of prevention programmes and the identi ﬁ cation and recruitment of participants at high risk for T2DM. One of the core items of the toolkit is the description of what to do and how to do it. Behaviour change is a process which requires individual attention and effective communication to achieve motivation, self-monitoring, sustained support and other intervention to prevent and manage relapses. This section includes a model of intervention including empowerment and patient-centred messages. It is followed by key messages on behaviour (physical activity and diet) that are important in prevention of diabetes and practical advice for patient-centred counselling. The focus is on long-term, sustainable lifestyle changes. Finally, a brief guide for evaluation and quality assurance in reference to the “quality and outcome indicators” is included. This section is followed by a consideration of possible risks and adverse effects. The IMAGE Toolkit main text ends with a positive mission statement, emphasising what can be achieved if we work together. The appendices give the reader a set of easy-to-use tools including a checklist for prevention programme development, templates for goal-setting and for food and physical activity diaries, an example of a risk screening questionnaire (the FINDRISC questionnaire) and a template for evaluation and quality assurance data collection.

European Curriculum for the Training of Prevention Managers

The IMAGE Study Group collated information in a systematic manner and delivered additionally a European curriculum for the training of prevention managers. Work package 5 of the IMAGE project was assigned to this development and the aim was the elaboration of a European curriculum for the training of prevention managers, supported by an online e-health portal (Work package 7) to provide learning materials and tools and to support the curriculum delivery as well as the interaction between teachers and course participants [19 ] . The development followed a standardised strategy with speci ﬁ c steps as mentioned below: ¥ Review of existing training curricula of professionals working in the ﬁ eld. ¥ Decision upon the tasks of the prevention manager. ¥ Development of the curriculum. 244 P.E.H. Schwa rz

A review of existing prevention programmes showed heterogeneous activities in respect to lifestyle change interventions in various European countries (e.g. unstruc- tured programmes in Bulgaria, France, Latvia, Portugal, Serbia, UK; or structured programmes in Finland, France, Germany) including a variety of teaching materials for the target group (people at risk), but without any existing written curriculum for the training of prevention managers, which could be taken as a basis. Discussions with the majority of the work package partners were carried out about the tasks of a prevention manager, the structure and duration of a training course and suitable entrance quali ﬁ cations.

Tasks of a Prevention Manager

The prevention manager, qualifi ed by the training, defi ned in the IMAGE curriculum, should be responsible at his place of work/institution to implement a lifestyle change intervention programme for people at risk in order to prevent the manifestation of T2D mellitus. This task includes management as well as counselling and training aspects: con- sisting of Ð The organisation of the programme (including number and duration of teaching units, contents of these units, teaching material, costs for the participants, time line/dates, places, co-workers, reimbursement, etc.) Ð The motivation and recruitment of participants (target group: persons at risk) Ð The required inter-organisational or intra-organisational networking Furthermore, the prevention manager has a responsibility to teach, counsel and train participants in speci fi c aspects of nutrition in diabetes prevention and speci fi c aspects of physical activity in diabetes prevention, based on recommended methods of behaviour change and motivation. Depending on his or her basic profession/quali fi cation and depending on the national (health care) context, the prevention manager may take over all the above de fi ned tasks (management as well as counselling/training of the persons at risk) by him/herself; as an alternative he or she may form a prevention team, e.g. collaborating with a physical activity expert who takes over the physical activity training units.

Structure and Duration of the Training Course

The PM-Training is designed to be offered and carried out by experienced national institutions/universities and it contains ¥ 7 Face-to-face training units (Modules 1Ð6 and 8) covering 1 training day/unit (contact hours: 55Ð60) 14 Global Challenge in Diabetes Prevention from Practice to Public Health 245

¥ Associated pre- and post-course assignments, further learning materials (PowerPoint sheets, time schedules, protocols, guidelines, drafts) and online communication forums , all supported by the IMAGE e-learning platform (see results of IMAGE Work package 7) ¥ A longitudinal project report (Module 7) which has to be written by every participant, describing the status of the organisation and implementation activities at his/her place of intervention Regional or national alumni networks for subsequent quality assurance and sustainability of the intended lifestyle change interventions are recommended.

Core Contents of the Eight Modules of the Curriculum

Module 1: Problems, evidence and tasks Module 2: Course organisation, recruitment, networking and evaluation management Modules 3 and 5: Behaviour change and motivation Module 4: Speci fi c aspects of physical activity in diabetes prevention Module 6: Speci fi c aspects of nutrition in diabetes prevention Module 7: Elaboration of an individual longitudinal project report Module 8: Presentation and discussion of the project reports The overall time span of the PM-training should cover about 6 months (e.g. one training unit every 2 or 3 weeks); the participants start their activities to plan and organise a local prevention programme immediately after the fi rst training unit, so that they could interchange their experience and encountered problems with their colleagues and teachers in the following training units. The individual longitudinal project report (Module 7) should reveal the respective implementation activities describing Ð The steps of implementation of the prevention programme Ð The intervention team and facilities Ð The methods and contents of the prevention programme Ð The concepts of recruitment and networking Ð The evaluation instruments and the collected evaluation data if already available The overall work load of the training course for prevention managers (including contact hours in the seven face-to-face-modules, working hours for pre- and post- course assignments, for organisation and implementation activities during the overall time span and for the elaboration of the longitudinal project report) can be estimated at approximately 150 h. The written individual project report including its presentation and discussion/disputation by each participant during the last module/ day of the training course is recognised as the course exam. If the report is approved, the participant receives her/his certifi cation as a “Prevention Manager T2Dm ” issued by the national institution responsible for the organisation of the training course. 246 P.E.H. Schwa rz

The Relevance of National Policy Involvement

Political support for establishing the framework for diabetes prevention programmes is extremely important and mandatory for national implementation [ 12 ] . The driving force for a positive outcome includes the structure, transparency, daily life relevance of the intervention and practicality of the programme. The global experience of diabetes prevention programmes demonstrated that lack of political support or endorsement by stakeholders predicted little chance of achieving public health relevance. For successful national public health implementation, it is abso- lutely necessary to involve relevant stakeholders in the political arena, health insurance companies supporting nongovernmental organisations (NGO), the scienti ﬁ c community and the target population. Within the European Union, only 5 of 27 countries have a national diabetes plan and only one has a national diabetes prevention programme [ 12 ] . In Asia, the situation is similar with a progressive increase in the number of countries including diabetes prevention in their national policies. The United States are at the forefront of governmental initiatives for developing a diabetes prevention programme with the Centers for Disease Control and Prevention being the driving force for coordinating the national effort. The existence of a national policy for supporting diabetes prevention does not equate with a positive outcome, but it is a mandatory ﬁ rst step for successful public health implementation.

Public Health Model for Implementation of Diabetes Prevention Programmes

Diabetes prevention schemes in clinical and public health practice have provided a variety of experiences over the past years identifying essential milestones and road- blocks, which ensure successful programme implementation (Fig. 14.1 ) [ 12 ] . It is also understood that we require certain levels in the public health environment including state and government sectors, which are necessary to create awareness and a suitable political environment in order to implement national programmes. The community level, which is responsible for implementation, requires educational elements and guidelines for screening. Diabetes prevention programmes are only achievable in the context of a comprehensive prevention organisation, which is why structures, which constitute the next level, are crucial for developing such programmes. It is also important to assure quality programme management, to involve physician education as well as to establish a connection to secondary prevention programmes. The last level in this model is the personal level, which addresses the attitude, behaviour and needs of potential risk persons (Fig. 14.1 ). 14 Global Challenge in Diabetes Prevention from Practice to Public Health 247

Fig. 14.1 Four-level public health model for the implementation of diabetes prevention programs (adapted from Schwarz PE. Public health implications: translation into diabetes prevention initiatives—four-level public health concept. Med Clin North Am. 2011;95:397Ð407, ix)

State Level

The state level, which is represented by governmental stakeholders is essential for policy development. It is the basis for primary diabetes prevention public health activities and assures the effectiveness of intervention programmes. A prevention strategy that is included in a national diabetes plan constitutes the public health framework for implementation. As a part of this strategy, national or local health insurance companies should be involved in order to assure adequate reimbursement. Structured plans for screening, instructing healthy behaviour, exercise and physical activity are required. Furthermore, introducing healthy lifestyle education in schools, especially in primary schools, would lay the foundation for life-long sustainability and could decrease the burden of obesity and T2D.

Community Level

The community provides the underlying structural platform for implementing an intervention programme and therefore bears most of the responsibility. The community forms the framework for screening, network structures and intervention programmes. It gives people healthy food options and supplies facilities for physical activity. It would be very valuable for public health to establish a healthy work place 248 P.E.H. Schwa rz environment. This is an ambitious goal, especially for small businesses, but this would enormously reduce the risk for metabolic diseases at the work site.

Intervention Structures

Intervention depends on programme management structures. Therefore, evidence- based practice has its strongest impact at this level. After the community has set the conditions for intervention, the according structures have to be established. The programme can only be successful if arrangements for intervention management and networking in the community have been made. In order to precisely target high- risk individuals, there need to be clear guidelines for prevention practice and intervention programme development. To secure further support by the stakeholders at the state level we need to assure quality management in the intervention process, which involves routine reporting of programme performance and accomplishments. Furthermore, quality management allows direct comparison of performance. If the intervention is organised in a medical setting, it is necessary that physicians are involved. Furthermore, relevant education of physicians as well as correspondence with available secondary prevention programmes is necessary.

Personal Level

The personal level is critical as it concerns individuals at increased risk who are therefore eligible for intervention. If all other levels have been adequately addressed, this level is the overall indicator of quality and success of the programme. The individuals can be motivated to participate in the programme if they have realised their personal gain and if any possible drawbacks have been eliminated. This applies especially to minorities or ethnic groups in the community setting. Intervention material that is easily accessible and easy to understand is essential to ensure participation with solicitation of feedback to further motivate the subjects to stick to the programme and to change their lifestyle. The personal level will be beneﬁ ted as well by choosing healthier food alternatives. The four-level public health model requires milestones in order to develop an effective prevention programme. Failing to address one milestone can result in road- blocks for programme management and implementation. Milestones are usually highly interrelated and generally do not stand alone. Therefore, the initial challenge of the overall effort becomes easier as incremental progress is achieved by progressing from one milestone or one level to the next. This public health model can be applied to most westernised healthcare systems and is adaptable to other systems as well. It is most important to help high-risk persons to understand the added value an intervention could have on their lives. Only if these people are addressed on a personal level and if they understand that a healthy lifestyle is achievable in the short- term, implementation of diabetes prevention programmes can be successful. 14 Global Challenge in Diabetes Prevention from Practice to Public Health 249

Network “Who Are Active in Diabetes Prevention”

An important current new initiative is the start of an international network “Who are active in diabetes prevention”. The aim of this network is that people, who are interested in the prevention of diabetes and those, who want to start being active in the fi eld, meet in one professional network. The network itself encourages exchanging knowledge, recent intervention material as well as educational standards, but the most important focus of the network is the exchange of experiences in diabetes prevention practice. It is also thought as a platform to change scienti fi c information or up to date study information between research groups and people active in diabetes prevention. People from more than 140 countries are already part of the network and especially many participants come from low- and middle-income countries, who spread the information about diabetes prevention practice and have already initiated new prevention programmes and will increase the intervention quality. Everyone who is interested in the prevention of diabetes is invited to register for free and benefi t from this global network— http://www.activeindiabetesprevention.com .

Future Challenges and Opportunities

As estimated by the International Diabetes Federation, the number of T2D patients is likely to increase during the forthcoming years and maybe decades in Europe, but with e.g. the implementation of the recommendations of the IMAGE project the prevention of T2D can become reality in clinical care and the increase of the T2D epidemic may be eventually controlled and the burden of diabetes be gradually diminished. The fi rst prerequisite is political support and the development of a national action plan for diabetes prevention, which necessitates the involvement of a number of stakeholders on governmental and nongovernmental level as well as scientifi c and practical input. The International Diabetes Federation, together with other organisations representing chronic diseases, plans a non-communicable Disease Alliance and hold a United Nations Summit in September 2011. One part of it is to develop a Global United Nations Diabetes Plan which will call for political action and support. Furthermore, practical guidance is needed and this includes the presentation of the evidence in the fi eld for diabetes prevention on the scientifi c, but even more important on the practical level as well as the training of people to deliver preventive intervention. There will be a lot of work in implementing these recommendations in the future. Also, there is a need to continue systematic research into the aetiology, prevention and management of T2D. In particular, translational research regarding the implementation of existing knowledge into public health and clinical practice must be carried out. This is not possible without proper research funding that should become available through various national and international funding sources. It may not be possible to eradicate chronic non-communicable diseases, but their burden can be 250 P.E.H. Schwa rz reduced dramatically. Good examples are coronary heart disease, stroke, lung cancer, etc. From these actions to reduce the burden of non-communicable diseases we have learned that a successful prevention must be based on both population-based actions targeted to the entire community and the high-risk approach targeted to individuals at the highest risk of a particular disease simultaneously. In any case, the future of T2D prevention has never been as bright as it is today.

References

1. Knowler WC, Barrett-Connor E, Fowler SE, et al. Reduction in the incidence of type 2 diabetes with lifestyle intervention or metformin. N Engl J Med. 2002;346(6):393Ð403. 2. Tuomilehto J. Counterpoint: evidence-based prevention of type 2 diabetes: the power of lifestyle management. Diabetes Care. 2007;30(2):435Ð8. 3. Tuomilehto J, Lindstrom J, Eriksson JG, et al. Prevention of type 2 diabetes mellitus by changes in lifestyle among subjects with impaired glucose tolerance. N Engl J Med. 2001; 344(18):1343Ð50. 4. Schwarz PE, Li J, Lindstrom J, Tuomilehto J. Tools for predicting the risk of type 2 diabetes in daily practice. Horm Metab Res. 2009;41(2):86Ð97. 5. Schwarz PE, Li J, Reimann M, et al. The Finnish Diabetes Risk Score is associated with insulin resistance and progression towards type 2 diabetes. J Clin Endocrinol Metab. 2009;94(3):920Ð6. 6. Schwarz PE, Schwarz J, Schuppenies A, Bornstein SR, Schulze J. Development of a diabetes prevention management program for clinical practice. Public Health Rep. 2007;122(2):258Ð63. 7. Jackson L. Translating the Diabetes Prevention Program into practice: a review of community interventions. Diabetes Educ. 2009;35(2):309Ð20. 8. Saaristo T, Peltonen M, Keinanen-Kiukaanniemi S, et al. National type 2 diabetes prevention programme in Finland: FIN-D2D. Int J Circumpolar Health. 2007;66(2):101Ð12. 9. Schwarz PEH, Greaves C, Reddy P, Dunbar J, Schwarz J. Diabetes prevention in practice, vol. 1. Dresden: TUMAINI Institute for Prevention management; 2010. 10. Paulweber B, Valensi P, Lindström J, et al. A European evidence-based guideline for the prevention of type 2 diabetes. Horm Metab Res. 2010;42 Suppl 1:S3Ð36. 11. Lindstrom J, Neumann A, Sheppard KE, et al. Take action to prevent diabetes—the IMAGE toolkit for the prevention of type 2 diabetes in Europe. Horm Metab Res. 2010;42 Suppl 1:S37Ð55. 12. Schwarz PE, Muylle F, Valensi P, Hall M. The European perspective of diabetes prevention. Horm Metab Res. 2008;40(8):511Ð4. 13. Pajunen P, Landgraf R, Muylle F, et al. Quality indicators for the prevention of type 2 diabetes in Europe—IMAGE. Horm Metab Res. 2010;42 Suppl 1:S56Ð63. 14. Schwarz PE, Lindstrom J, Kissimova-Scarbeck K, et al. The European perspective of type 2 diabetes prevention: diabetes in Europe—prevention using lifestyle, physical activity and nutritional intervention (DE-PLAN) project. Exp Clin Endocrinol Diabetes. 2008;116(3):167Ð72. 15. Schwarz PE, Gruhl U, Bornstein SR, Landgraf R, Hall M, Tuomilehto J. The European perspective on diabetes prevention: development and implementation of a European guideline and training standards for diabetes prevention (IMAGE). Diab Vasc Dis Res. 2007;4(4):353Ð7. 16. Lindström J, Neumann A, Sheppard KE, et al. Take action to prevent diabetes—the IMAGE toolkit for the prevention of type 2 diabetes in Europe. Horm Metab Res. 2010;42 Suppl 1:37Ð55. 17. Pajunen P, Landgraf R, Muylle F, et al. Quality indicators for the prevention of type 2 diabetes in Europe—IMAGE. Horm Metab Res. 2010;42 Suppl 1:56Ð63. 18. Lindstrom J, Neumann A, Sheppard KE, et al. Take action to prevent diabetes—the IMAGE toolkit for the prevention of type 2 diabetes in Europe. Horm Metab Res. 2010;42 Suppl 1:S37Ð55. 19. Tolks D, Fischer M. E-Learning Portal Präventionsmanager. In: Deutsche-Diabetes-Stiftung, editor. Diabetes in Deutschland: Fakten—Zahlen: 20 Jahre nach St. Vincent, vol. 1. München: Edition Lipp; 2010. p. 227Ð36. Index

A Apoptosis signal-regulating kinase 1 (ASK1) AABLH. See African Americans Building gene , 72 a Legacy of Health (AABLH) Area under the receiver operating characteristic AACE/ACE. See American Association (AUROC) curve , 85, 87 of Clinical Endocrinologists/ ARIC. See Atherosclerosis Risk in American College of Endocrinology Communities (ARIC) (AACE/ACE) ASK1 gene. See Apoptosis signal-regulating A1c , 5–6 kinase 1 (ASK1) gene Acarbose , 169–170. See also Study to Prevent Aspartate aminotransferase (AST) , 91 Non-Insulin Dependent Diabetes Asymptomatic phase , 105 Mellitus (STOP-NIDDM) trial Atherosclerosis , 169 Acarbose Cardiovascular Evaluation (ACE) , 179 Atherosclerosis Risk in Communities ACR. See Albumin to creatinine ratio (ACR) (ARIC) , 50, 87 Acute insulin response (AIR) , 84 ATP. See Adult Treatment Panel (ATP) ADA. See American Diabetes AUROC curve. See Area under the receiver Association (ADA) operating characteristic (AUROC) Adipokines , 92 curve Adiponectin , 36, 92, 153 Autonomic imbalance , 129–131 Adult Treatment Panel (ATP) , 83 Autonomic nervous system (ANS) , 129, 156 African Americans Building a Legacy of Health (AABLH) , 233 Age-related glucose insensitivity , 19–20 B AIR. See Acute insulin response (AIR) Basal hyperinsulinemia , 84 Alanine aminotransferase (ALT) , 91 Biomarkers , 153–154 Albumin to creatinine ratio (ACR) , 153–154 Birth weight , 45–46 Alpha lipoic acid , 133 BMI. See Body mass index (BMI) A L T . See Alanine aminotransferase (ALT) Body fat , 46–47 American Association of Clinical Body mass index (BMI) , 6, 7, 46, 51, 89–90 Endocrinologists (AACE) , 113 Breast feeding , 46 American Association of Clinical Endocrinologists/American College of Endocrinology (AACE/ACE) , 4, C 113 CABG. See Coronary-artery bypass American Diabetes Association (ADA) , 4 grafting (CABG) Amish, prediabetes genes , 70–71 Cardiac autonomic neuropathy (CAN) , 117 ANS. See Autonomic nervous system (ANS) autonomic imbalance , 130–131 Anthropometric measures , 89–90 prediabetes and , 125–126

D. LeRoith (ed.), Prevention of Type 2 Diabetes: From Science to Therapy, 251 DOI 10.1007/978-1-4614-3314-9, © Springer Science+Business Media New York 2012 252 Index

Cardiovascular disease (CVD) , 31, 168–169. novel barriers uncovering , 231 See also Study to Prevent Non- systems and clinician-patient Insulin Dependent Diabetes communication , 230 Mellitus (STOP-NIDDM) trial information and tools , 232–233 Cardiovascular risk factors , 155–156 public policy shaping , 233–234 Carotid intima-media thickness Community approaches, diabetes prevention. (cIMT) , 196, 197 See Diabetes prevention, Catch-up growth , 46 community approaches CCM. See Corneal confocal microscopy Community-clinic partnership model , 203–204 (CCM) Community-owners , 231 CDC. See U.S. Centers for Disease Control Con fi rmed DSPN , 120 and Prevention (CDC) Contact heat-evoked potentials (CHEPS) , 122 b -Cell dysfunction , 23–25 Copy-number variants (CNVs) , 74 b -Cell function Corneal confocal microscopy (CCM) , 122 fasting insulin secretion rate vs. total Coronary-artery bypass grafting (CABG) , 36 insulin output , 12–13 COX-2 gene , 128 insulin/glucagon system , 11–12 Cox proportional hazards model , 171 insulin resistance , 22–24 C-reactive protein (CRP) , 91–93, 155 insulin sensitivity and , 16–17 CVD. See Cardiovascular disease (CVD) mass , 21–22 mathematical models , 13 natural history schematic D representation , 26 Da Qing study , 144, 189–191 plasma glucose and insulin excursion , Data Safety and Quality Review 14–15 Committee , 171 plasma glucose concentrations , 11 DCCT. See Diabetes Control and potentiation factor , 15–16 Complications Trial (DCCT) prediabetes Decreased insulin secretory function , 86 age-related glucose insensitivity , 19–20 DE-PLAN , 241 empirical indices , 20–21 DEXA. See Dual-energy X-ray glucose sensitivity , 17–18 absorptiometry (DEXA) rate sensitivity and potentiation Diabetes Control and Complications factor , 17, 19 Trial (DCCT) , 4 b -Cell glucose sensitivity , 14 Diabetes predicting model , 94 b -Cell mass , 21–22 Diabetes prediction CHEPS. See Contact heat-evoked potentials anthropometric measures , 89–90 (CHEPS) endogenous sex hormones , 93 China Da Qing Diabetes Prevention Outcome essence , 82–83 Study , 132 future prospects , 96–97 Chronic idiopathic axonal neuropathy genetic factors , 95 (CIAP) , 123, 124 hepatic markers cIMT. See Carotid intima-media adipokines , 92 thickness (cIMT) in fl ammatory markers, 92–93 CNVs. See Copy-number variants (CNVs) iron storage and hepatic enzymes , 91 Coffee , 50 markers , 81–82 Communities IMPACT Diabetes Center , 231 metabolic syndrome and other risk factors , Community-academic hybrid approaches , 93–95 227–228. See also Community- personalized risk pro fi les , 97–98 academic partnerships plasma insulin and glucose levels Community-academic partnerships basal hyperinsulinemia , 84 clinical settings decreased insulin secretory function , 86 clinical environment , 230–231 fasting glucose , 84–85 clinical practice improvement , 231–232 HbA1c , 86–87 clinician knowledge and insulin resistance , 85–86 behavior , 228–229 pattern of changes , 87–88 Index 253

practicing physician , 97 T2DM , 239–240 threshold levels , 95–96 toolkit , 242–243 triglycerides , 90 Diabetes prevention, local community Diabetes prevention partnerships Da Qing , 189–191 community-academic partnerships DREAM , 197–199 clinical settings (see Community- Finnish Diabetes Prevention Study academic partnerships) genetic variation , 194–195 information and tools , 232–233 insulin sensitivity and CVD public policy shaping , 233–234 incidence , 193 two paths , 227 post hoc analyses , 193–194 future prospects , 234 recruitment , 191 prevalence, morbidity, and mortality VLCD and LTPA , 192 disparities TRIPOD/PIPOD , 195–197 East Harlem and the Upper East Diabetes Prevention Amendment , 162 Side , 223–224 Diabetes prevention, community approaches population characteristics , 223, 225 community-clinic partnership model , social determinants , 222 203–204 real-life, community context , 225–227 Ten Essential Public Health root causes and efforts , 221–222 Services , 203, 205 Diabetes Prevention Program (DPP) , 104 U.S. Diabetes Prevention Translation combined analysis , 143–144 Research cost-effectiveness , 159 attendence, body weight, diet, and Da Qing study , 144 physical activity monitoring , 212 effects basic characteristics , 206–209 adherence and tolerability , 152 community/environmental policies , 217 biomarkers , 153–154 de ﬁ nition and key factors , 205 cardiovascular risk factors , 155–156 duration and intervention genetic insights , 157–158 intensity , 211–212 placebo, metformin, and lifestyle intervention staff , 213 intervention , 150–151 lifestyle curriculum and intervention randomized cohort , 149–150 format , 211 retinopathy and autonomic nervous participants’ diabetes risk , 206, 211 system , 156 questions , 214–215 subgroup effects , 151–152 U.S. National DPP , 215–217 success predictors , 152–153 weight loss outcomes , 210, 214 troglitazone , 154 Diabetes prevention, global challenges urinary incontinence , 156–157 future challenges and opportunities , weight and nutrient intake , 154 249–250 eligibility criteria , 144–145 national policy involvement , 246 guidelines , 159–162 need , 240 interventions network , 249 intensive lifestyle and practice strategy , 240–241 metformin , 148 prevention manager training sample size determinations , 149 eight modules , 245 standard lifestyle recommendation , European curriculum , 243–244 147–148 structure and duration , 244–245 locations , 145 tasks , 244 policy , 162–163 public health model implementation recruitment and screening , 146–147 community level , 247–248 study design , 145–146 intervention structures and personal sustainability , 158 level , 248 Diabetes Prevention Program Outcomes schematic representation , 246–247 Study (DPPOS) , 155, 156 state level , 247 Diabetes Prevention Recognition Program requirements , 241–242 (DPRP) , 216 254 Index

Diabetes REduction Assessment with Ramipril Ethnicity , 44 and Rosiglitazone Medication Euglycemic-hyperinsulinemic clamp (EHC) , 84 (DREAM) , 105, 197–199 European Association for the Study Diabetes risk scores (DRS) , 94, 108–110 of Diabetes (EASD) , 113 Diabetes screening. See Prediabetes and diabetes screening Diabetes Training and Technical Assistance F Center (DTTAC) , 216 Fasting plasma glucose (FPG) , 105–106, Diabetic nephropathy , 53 122–123 Diabetic neuropathy pain (DNP) , 118, 119 FIN-D2D , 240 Diabetic peripheral neuropathy (DPN) , 118 FINDRISC. See FINnish Diabetes Risk SCore Diet , 47–48 (FINDRISC) Disposition index , 16 Finnish Diabetes Prevention Study Distal symmetric polyneuropathy (DSPN) , genetic variation , 194–195 117, 118 insulin sensitivity and CVD incidence , 193 DJBL. See Duodenojejunal bypass post hoc analyses , 193–194 liner (DJBL) recruitment , 191 DNP. See Diabetic neuropathy pain (DNP) VLCD and LTPA , 192 DPN. See Diabetic peripheral FINnish Diabetes Risk SCore (FINDRISC) , neuropathy (DPN) 94, 109, 110, 239 DPP. See Diabetes Prevention Program (DPP) Fixed risk factors , 43–45 DPPOS. See Diabetes Prevention Program Food Retail Expansion to Support Health Outcomes Study (DPPOS) (FRESH) Program , 233 DPRP. See Diabetes Prevention Recognition FPG. See Fasting plasma glucose (FPG) Program (DPRP) Frequently sampled intravenous glucose DREAM. See Diabetes REduction Assessment tolerance test (FSIVGTT) , 33, 193 with Ramipril and Rosiglitazone FRESH Program. See Food Retail Expansion Medication (DREAM) to Support Health (FRESH) DRS. See Diabetes risk scores (DRS) Program DSPN. See Distal symmetric polyneuropathy FSIVGTT. See Frequently sampled (DSPN) intravenous glucose tolerance DTTAC. See Diabetes Training and Technical test (FSIVGTT) Assistance Center (DTTAC) Dual-energy X-ray absorptiometry (DEXA) , 46 G Duodenojejunal bypass liner (DJBL) , 36 Gamma-glutamyltranspeptidase (GGT) , 91 Dysglycemia , 97 GDM. See Gestational diabetes mellitus (GDM) Genetic insights , 157–158 E Genetic markers , 111–112 Early life risk factors , 45–46 Genetic predisposition , 43–45 EASD. See European Association for Genome-wide association studies (GWASs), 63 the Study of Diabetes (EASD) Gestational diabetes , 49–50 eGFR. See Estimated glomerular fi ltration Gestational diabetes mellitus (GDM) , 151–152 rate (eGFR) GGT. See Gamma-glutamyltranspeptidase EHC. See Euglycemic-hyperinsulinemic (GGT) clamp (EHC) Global United Nations Diabetes Plan , 249 Endogenous sex hormones , 93 Glucose Intolerance Obesity and Hypertension Epidemiology and socioeconomic impact, (GOH) study , 84 T2D. See Type 2 diabetes (T2D) Glucose sensitivity , 13 youth epidemiology Glucose tolerance test (GTT) , 105–106 Epigenetic modi fi cations , 74 a -Glucosidases , 169, 170 Estimated glomerular fi ltration rate Glut 4 gene , 32 (eGFR) , 198 Glycated hemoglobin (HbA1c) , 86–87 Index 255

Glycemic threshold , 87 Impairment of glucose regulation (IGR) , 127 GOH study. See Glucose Intolerance Obesity IMT. See Intima media thickness (IMT) and Hypertension (GOH) study In ﬂ ammatory markers , 92–93 GTT. See Glucose tolerance test (GTT) Insulin clamp technique , 23 GWASs. See Genome-wide association Insulin/glucagon system , 11–12 studies (GWASs) Insulin receptor substrates (IRS) , 32 Insulin resistance (IR) , 6–7, 22–24, 65, 85–86 H clinical implications , 34 HCPs. See Healthcare professionals (HCPs) mechanism , 32 HDL. See High-density lipoprotein (HDL) methods , 32–33 Healthcare professionals (HCPs) , 213 prevention and treatment , 35–37 Health disparities , 227 T2DM and , 31 Healthy Connections Program , 233 Insulin sensitivity , 32–34 Heart rate variability (HRV) , 125–126 Intensive lifestyle intervention , 148 Hemoglobin A1c (HbA1c) , 105–106 Interleukin 6 (IL-6) , 92–93 Hepatic enzymes , 91 Intermediate hyperglycemia , 5 Hepatic glucose output (HGO) , 32–33 International Diabetes Federation (IDF) , 2, 53, Hepatic lipase gene , 195 83, 249 Hepatic markers. See Diabetes prediction Interventional trials. See Diabetes Prevention Heritability , 44 Program (DPP) Heterogeneous populations complexity , 63–64 Intima media thickness (IMT) , 169 HGO. See Hepatic glucose output (HGO) Intraepidermal nerve ﬁ ber (IENF) density High-density lipoprotein (HDL) , 36 quantitation , 121 HLA-DRB1 , 72 Intravenous glucose tolerance tests Homeostasis model assessment of insulin (IVGTTs) , 196 resistance (HOMA-IR) , 33, 86, 87 Intrinsic sympathetic activity (ISA) , 134 H R V. See Heart rate variability (HRV) Iron storage , 91 Hyperglycemia , 49, 126 IRS. See Insulin receptor substrates (IRS) Hyperinsulinemia , 93 ISA. See Intrinsic sympathetic activity (ISA) Hyperinsulinemic euglycemic glucose clamp Israeli Diabetes Risk Score matrix , 109, 111 technique , 32 IVGTTs. See Intravenous glucose tolerance Hyperlipidemia , 127, 169 tests (IVGTTs)

I K IDF. See International Diabetes KCNJ11 gene , 157 Federation (IDF) IENF density quantitation. See Intraepidermal nerve ﬁ ber (IENF) density L quantitation Laser Doppler perfusion imaging (LDI) , 127 IFG. See Impaired fasting glucose (IFG) Leisure time physical activity (LTPA) , 192 IGR. See Impairment of glucose Leptin , 36, 92 regulation (IGR) Lifestyle interventions IGT. See Impaired glucose tolerance (IGT) Da Qing , 189–191 IGTN Study. See Impaired Glucose Tolerance Finnish Diabetes Prevention Study (see Neuropathy (IGTN) Study Finnish Diabetes Prevention Study) IL-6. See Interleukin 6 (IL-6) Lipotoxicity , 34 IMAGE , 241–242 Liver disease , 50 Impaired fasting glucose (IFG) , 3–4, 106, Local community partnerships, diabetes 107, 122–123 prevention. See Diabetes Impaired glucose tolerance (IGT) , 3–4, prevention, local community 105, 107, 122–123, 168, 173 partnerships Impaired Glucose Tolerance Neuropathy LTPA. See Leisure time physical (IGTN) Study , 123 activity (LTPA) 256 Index

M NCEP-ATP III. See National Cholesterol MAP kinase. See Mitogen-activated protein Education Program-Third Adult (MAP) kinase Treatment Panel (NCEP-ATP III) Maturity onset diabetes of the young NCS. See Nerve conduction study (NCS) (MODY) , 62 Nerve conduction study (NCS) , 117 MBL2 , 72 Nerve impairment scoring system (NISS) , MCDS. See Mexico City Diabetes 120–121 Study (MCDS) Neuropathy and prediabetes MELANY cohort. See Metabolic Life-style CAN , 125–126 and Nutrition Assessment in Young CCM and CHEPS , 122 Adult (MELANY) cohort clinical manifestations , 130–131 Merchant Incentive Programs , 233–234 de ﬁ nition , 118–119 Metabolic Life-style and Nutrition diagnosis , 131 Assessment in Young Adult diagnostic aspects , 119–121 (MELANY) cohort , 109 epidemiology , 119 Metabolic syndrome (MS) , 49, 83, IFG or IGT , 122–123 93–95, 127–129 metabolic syndrome and , 127–129 Metformin , 35, 148 objective devices , 120–121 Mexico City Diabetes Study (MCDS) , pathogenesis , 126–127 85, 94 peripheral neuropathy , 124–125 Michigan neuropathy screening instrument relationship , 123–124 (MNSI) , 124 skin biopsy and IENF density Microvascular disease , 156 quantitation , 121 Missing heritability , 74, 75 treatment Mitogen-activated protein (MAP) kinase , 32 diet and exercise , 132 MNSI. See Michigan neuropathy screening lifestyle changes , 133 instrument (MNSI) pharmaceutical approaches , 134, 135 Modi ﬁ able risk factors sympathovagal balance , 133–134 body fat , 46–47 NFATC2 gene. See Nuclear factor of activated diet , 47–48 T-cells cytoplasmic calcineurin- gestational diabetes , 49–50 dependent 2 (NFATC2) gene hyperglycemia and metabolic NGO. See Nongovernmental organisations syndrome , 49 (NGO) physical activity , 48–49 NGSP. See National Glycohemoglobin socioeconomic status and smoking , 50 Standardization Program (NGSP) MODY. See Maturity onset diabetes of the NGT. See Normal glucose tolerance (NGT) young (MODY) NHANES. See National Health and Nutrition MONICA/KORA surveys , 124–125 Examination Survey (NHANES) MS. See Metabolic syndrome (MS) NISS. See Nerve impairment scoring system (NISS) Non alcoholic fatty liver disease (NAFLD) , N 50, 91 NAFLD. See Nonalcoholic fatty liver Nongovernmental organisations (NGO) , 246 disease (NAFLD) Normal glucose tolerance (NGT) , 123 National Cancer Institute , 48 Nuclear factor of activated T-cells cytoplasmic National Cholesterol Education Program- calcineurin-dependent 2 (NFATC2) Third Adult Treatment Panel gene , 198 (NCEP-ATP III) , 7 Nurses Health Study , 48 National Glycohemoglobin Standardization Program (NGSP) , 4 National Health and Nutrition Examination O Survey (NHANES) , 42 Obesity , 6–7, 64–65 National Institutes of Health , 42, 43 Oral glucose tolerance test (OGTT) , 4, 5 Index 257

P Prediabetes and diabetes screening. PAI-1. See Plasminogen activator inhibitor-1 See also b -Cell function; Type 2 (PAI-1) diabetes mellitus (T2DM) PARP. See Poly (ADP-ribose) polymerase asymptomatic phase , 105 (PARP) cost-effectiveness , 112 PCI. See Percutaneous coronary criteria , 103 intervention (PCI) effective treatment , 104–105 Pedometers , 48 guidelines recommendations , 113 Percutaneous coronary intervention (PCI) , 36 public health issue , 104 Peripheral neuropathy , 124–125 suitable tests Peroxisome proliferator-activated A1C criteria , 106, 108 receptor gamma (PPARgamma) diabetes risk scores , 108–110 agonists , 35 FPG, GTT and HbA1c , 105–106 Peroxynitrite , 128, 133–134 genetic markers , 111–112 Personalized risk pro ﬁ les , 97–98 guidelines and consensus Phenotype complexity. See Prediabetes genes statements , 106–108 Phosphatidylinositol 3-kinase (PI3-K) , 32 risk scores and laboratory Physical activity , 48–49 testing , 109, 111 PI3-K. See Phosphatidylinositol 3-kinase Prediabetes and neuropathy. See Neuropathy (PI3-K) and prediabetes Pima Indians, prediabetes genes Prediabetes genes children , 69 future prospects , 74–75 description , 67 genomic studies and huge sample sizes , examinations , 70 62–63 obesity , 68–69 heterogeneous populations prevalence , 68 complexity , 63–64 Pioglitazone , 35, 36, 196–197 phenotype complexity Pioglitazone In Prevention Of Diabetes genes identi ﬁ cation , 66–67 (PIPOD) study , 196–197 insulin resistance and secretion , 65–66 Placebo , 150–151 obesity , 64–65 Plasma insulin and glucose levels. polygenic disease , 61–62 See Diabetes prediction population isolates Plasminogen activator inhibitor-1 Amish , 70–71 (PAI-1) , 92–93 genes associated , 72–73 PM training. See Prevention manager (PM) Pima Indians (see Pima Indians, training prediabetes genes) Poly (ADP-ribose) polymerase (PARP) , 127 PreDx DRS , 94 Population isolates, prediabetes genes. Prevention manager (PM) training See Prediabetes genes eight modules , 245 Possible DSPN , 120 European curriculum , 243–244 Postprandial hyperglycaemia structure and duration , 244–245 acarbose treatment , 169–170 tasks , 244 diabetes and CVD risk factor , 168–169 Prevention practice , 241, 242 insulin resistance , 167–168 Probable DSPN , 120 STOP-NIDDMN trial (see Study to Public health model implementation. See Prevent Non-Insulin Dependent Diabetes prevention, global Diabetes Mellitus (STOP-NIDDM) challenges trial) Potentiation , 13, 15 Potentiation factor , 15–16 Q PPARgamma agonists. See Peroxisome QAFT. See Quantitative autonomic function proliferator-activated receptor test (QAFT) gamma (PPARgamma) agonists QALYs. See Quality-adjusted life years PPARG gene , 157 (QALYs) 258 Index

QDScore , 95 Socioeconomic impact , 53–54 QOL. See Quality-of-life (QOL) Socioeconomic status , 50 QSART. See Quantitative sudomotor axon Southcentral Foundation (SCF) , 231 re fl ex test (QSART) Standard lifestyle recommendation , 147–148 QST. See Quantitative sensory testing (QST) Stanford Chronic Disease Self-Management Quality-adjusted life years (QALYs) , 112, 159 Program , 233 Quality-of-life (QOL) , 117 Study to Prevent Non-Insulin Dependent Quantitative autonomic function test Diabetes Mellitus (STOP-NIDDM) (QAFT) , 117 trial , 105 Quantitative insulin sensitivity check index acarbose treatment cost-effectiveness , (QUICKI) , 33 179–180 Quantitative sensory testing (QST) , 117 CVD prevention Quantitative sudomotor axon re fl ex test acarbose effects , 177–178 (QSART) , 131 hypertension and , 178–179 QUICKI. See Quantitative insulin sensitivity risk factors , 176 check index (QUICKI) diabetes prevention acarbose effects and diabetes , 174–175 R baseline characteristics , 172, 173 Ramipril , 197–199 candidate gene polymorphisms , Rate sensitivity , 13, 15 175–176 RDNS-IGM. See Rochester Diabetic pro fi le , 171–172 Neuropathy Study of patients study design , 171 with Impaired Glucose Metabolism Subclinical DSPN , 120 (RDNS-IGM) Sudoscan , 131 Reactive oxygen species (ROS) , 168 Sugar-sweetened beverages , 47 Relationship between Insulin Sensitivity and Supermarket Development Incentive Cardiovascular Risk (RISC) study , Programs , 233 34 Sympathovagal balance , 133–134 Retinopathy , 82, 156 RISC study. See Relationship between Insulin Sensitivity and Cardiovascular Risk T (RISC) study TCF7L2 , 44. See also Transcription Risk factors, T2D. See Type 2 diabetes (T2D) factor-7-like 2 (TCF7L2) gene youth epidemiology T2D. See Type 2 diabetes (T2D) youth Rochester Diabetic Neuropathy Study of epidemiology patients with Impaired Glucose T2DM. See Type 2 diabetes mellitus (T2DM) Metabolism (RDNS-IGM) , 123 Ten Essential Public Health Services , 203, 205 ROS. See Reactive oxygen species (ROS) TG. See Triglycerides (TG) Rosiglitazone , 36, 197–199 Thiazolidinediones (TZDs) , 35–36 and diabetes DREAM , 197–199 S TRIPOD/PIPOD , 195–197 Salicylates , 36–37 Threshold levels , 95–96 SCF. See Southcentral Foundation (SCF) Thrifty gene hypothesis , 45 SFN. See Small fi ber neuropathy (SFN) TNF- a . See Tumor necrosis factor Single nucleotide polymorphisms (SNPs) , alpha (TNF-a ) 63, 95, 175–176 Tolerability , 152 Skin biopsy , 121 Toolkit , 242–243 Skinfold thickness , 47 Transcription factor-7-like 2 (TCF7L2) Small fi ber neuropathy (SFN) , 120, 121 gene , 24, 157 Smoking , 50 Triglycerides (TG) , 36, 90 SNPs. See Single nucleotide TRIPOD. See TRoglitazone In Prevention polymorphisms (SNPs) Of Diabetes (TRIPOD) Index 259

Troglitazone , 36, 154, 195–197 UnitedHealth Group (UHG) , 216 TRoglitazone In Prevention Of Diabetes United Kingdom Prospective Diabetes Study (TRIPOD) , 195–197 (UKPDS) , 104 Tumor necrosis factor alpha (TNF-a ) , 128 Urinary incontinence , 156–157 Type 2 diabetes mellitus (T2DM) , 239–240 U.S. Centers for Disease Control and epidemic scope , 2–3 Prevention (CDC) , 215 factors contributing , 1–2 U.S. Diabetes Prevention Translation obesity and insulin resistance , 6–7 Research. See Diabetes prevention, and prediabetes community approaches A1c advantages and disadvantages , 5–6 U-shaped relationship , 46 current de fi nitions , 4–5 U.S. National Diabetes Prevention Program diagnostic criteria , 4 (DPP) , 215–217 IFG and IGT , 3–4 US Preventive Services Task Force prevention , 7–8 (USPSTF) , 113 Type 2 diabetes (T2D) youth epidemiology Uterine environment , 45 prevention cost-effectiveness , 54 risk factors assessment , 51, 52 V early life , 45–46 Very-low-calorie diet (VLCD) , 192 fi xed , 43–45 Vitamin D , 50 modi fi able (see Modi fi able risk factors) overview , 51–53 socioeconomic impact , 53–54 W statistics , 41–43 Waist circumference , 46–47 Tyrosine kinase , 32 Weight loss surgery , 36 TZDs. See Thiazolidinediones (TZDs) World Health Organization (WHO) , 2, 42

U Y UHG. See UnitedHealth Group (UHG) YMCA , 162 UKPDS. See United Kingdom Prospective Youth and T2D. See Type 2 diabetes (T2D) Diabetes Study (UKPDS) youth epidemiology