Manual on Vitamin a De Ciency Disorders (VADD)

Building bridges for better nutrition.

Sight and Life We care about the world’s most vulnerable populations and exist to help improve their nutritional status. Acting as their advocates, we guide original nutrition research, disseminate its ndings and facilitate dialog to bring about positive change.

For more information please visit www.sightandlife.org Donald S. McLaren, Klaus Kraemer Sight and Life Press, Third Edition 2012

Manual on Vitamin A De ciency Disorders ( VADD) ( Disorders ciency De A Vitamin on Manual Manual on Vitamin A De ciency Disorders (VADD) Donald S. McLaren ISBN 978-3-906412-58-0 Sight and Life | P.O. Box 2116 | 4002 Basel | Switzerland Klaus Kraemer

Sight and Life Press We care about the world’s most vulnerable pop- c/o Sight and Life ulations, and exist to help improve their nutri- DSM Nutritional Products Ltd tional status. Acting as their advocates, we PO Box 2116 guide original nutrition research, disseminate 4002 Basel its findings and facilitate dialog to bring about Switzerland positive change. We believe that the right mix of Phone +41 (0) 61 815 88 23 funding, knowledge, technology and policy will Fax +41 (0) 61 815 81 90 lead to better health. We work through advo- Email [email protected] cacy and leadership to address societal chang- Web www.sightandlife.org es affecting dietary quality. We act responsibly to disseminate information that reflects current Publications fromSight and Life can be obtained knowledge and evolving scientific evidence. We from the above address. Requests for permis- also aim to promote partnerships that are able sion to reproduce or translate Sight and Life to solve outstanding problems. publications should also be submitted to this address. The paper used in this book is acid-free and falls within the guidelines established to ensure per- This publication contains information derived manence and durability. from third party sources that do not necessarily represent the point of view of Sight and Cover design by S1 | Communication Design, Life and are the sole responsibility of the origi- Augsburg, Germany nal source. The mention of specific companies Proofreading by transparent – and trademarks does not imply that these are translation & text services, Berlin, Germany endorsed by Sight and Life. Although Sight and Typeset and print by Burger Druck, Life has used diligent care to ensure that the Waldkirch, Germany content of this publication is accurate and up to date, Sight and Life makes no representation ISBN 978-3-906412-58-0 or warranty regarding the accuracy, reliability or completeness of the content. This publication does not constitute or provide scientific or medical advice and is distributed without warranty of any kind, either express or implied. The reader shall be solely responsible for any interpretation or use of the material contained here- Carbon-neutral production in. In no event shall Sight and Life be liable for any damages arising from the reader‘s reliance upon, or use of, these materials. Donald S. McLaren, Klaus Kraemer Sight and Life Press, Third Edition 2012 Manual on Vitamin A Deficiency Disorders (VADD) II

Acknowledgements

The Sight and Life Manual on Vitamin A We would like to take this opportunity to Deficiency Disorders (VADD) is now appearing acknowledge the support we have experienced in its 3rd edition. Although our scope has broad- during the last couple of years in preparing the ened from “sight”, a single focus on vitamin A, to 3rd edition of this book. There are many people “life”, a multiple micronutrient focus, we believe to acknowledge who have been associated with that vitamin A remains sufficiently important and this project. We would therefore like to start with interesting for us to dedicate a whole book to this apologies to those who we may have omitted nutrient. to mention. Extraordinary thanks go to Svenia Sayer-Ruehmann and Anne-Catherine Frey from This book not only draws on the latest scient- the Sight and Life team for their skilful tech- ific research and expert insights, but also benefits nical support, without which this publication greatly from the work of previous generations would have never been completed. Further, our of scientists and policy-makers. As a result, it is Sight and Life interns Laurence Curty and Sarah almost impossible to acknowledge all those who Diouf both merit special mention. We continue have contributed their expertise. Key sources in- to appreciate Yvonne Bearne for her proofread- clude, among others, journal articles, policy state- ing, and the Burger Druck team for laying out the ments, reports, fact sheets, guidelines, reference book. We also wish to thank the World Health books, websites and – of course – the data and Organization for its kind permission to reprint the research to be found in Sight and Life Magazine vitamin A deficiency world maps. Last but not itself. least, Royal DSM deserves our special apprecia- tion for its continuous support of Sight and Life; in particular, we would also like to thank Stephan Tanda, Chairman of the Sight and Life Steering Committee, for writing his encouraging foreword to this volume. III3

Preface

This Third Edition of the Sight and Life demands of the time do not seem to permit the Manual on Vitamin A Deficiency Disorders luxury of extreme specialization. Henceforth (VADD) is likely to be the last. This is not be- source materials like this manual will also need cause there is nothing more to learn about the sub- to focus more on VMD in general. Sight and ject. It is because in recent years there has been Life sees this as a key role for itself in the future a major shift in the approach to the control of as we continue to evolve to ensure that we make a vitamin and mineral deficiencies (VMD) or multi- valuable contribution. micronutrient deficiencies (MMD) – including VADD. This is largely due to the fact that it is now The Third Edition has not surprisingly in- well recognized that VMD tend to occur together creased considerably in size and includes three rather than on their own – a tendency that is most new chapters. The first of these is specifically likely explained by the fact that most micronutri- titled ‘Foreground’ (as opposed to Background), ents are found in the same classes of foodstuffs, in order to emphasize, from the start, that the especially fruit, vegetables, fish and meat. Until focus of the Manual is on the control of VADD recently, it was general practice to identify and and that this objective should be kept in mind combat the most serious forms of deficiency, usu- throughout the book. Chapter 2 illustrates how ally of a single micronutrient. Prominent among essential and versatile the vitamin A molecules these have been vitamin A, iron, iodine, zinc and are – together with their close relatives found folate. Today, although there remains a great deal in nature, the carotenoids. Without them human to be learned about the individual roles and func- life, vision, growth and development would not tions of most micronutrients in human nutrition, be possible. The bioavailability of carotenoids interest is increasingly focused on their interrela- has become so important that it now requires its tionship, especially as micronutrient intervention own chapter (Chapter 5). In addition, the effect programs are on the increase and many are in the that the acute phase response (APR) has on serum process of being scaled-up. retinol levels is of such significance that, after an introduction to the issue in Chapter 7, it remains a The recognition of the interconnected relation- prominent feature in several subsequent chapters. ship of the micronutrients is further highlighted As so often in the field of VMD, the work on APR by the disbanding of the International Vitamin A has been pioneered in relation to VADD, but its Consultative Group (IVACG) and International effect in relation to other micronutrient deficien- Nutritional Anemia Consultative Group (INACG), cies has received little attention and much more which were set up more than thirty years ago to research is necessary. assist in the control of specific nutritional deficiencies. They have been replaced by the all en- This edition also includes in Chapter 12 the compassing Micronutrient Forum (MF), which latest available data on the global occurrence held its first scientific meeting in April 2007. It is of VAD from the World Health Organization no exaggeration to say that hundreds of nutrition- (WHO). Attention is drawn to the inability, over related scientists, whose knowledge and research the past several years, to determine accurate prev- efforts were virtually confined to a single micro- alence figures for the number of blind persons liv- nutrient, will now have to broaden their area of ing in any given country or region of the world or expertise and research. Rightly or wrongly, the within any vulnerable group etc., due to the deci- IV Preface

sion to no longer report non-blinding xerophthal- rophthalmia) had been brought under control to a mia and blinding xerophthalmia separately. As a large extent. However, at around the same time it consequence, while data on other causes of blind- was recognized that the problem had several other ness are becoming increasingly precise, vitamin A concerning elements: deficiency blindness is included under the heading of ‘other causes of childhood blindness’ – some- • Vitamin A deficiency at a subclinical and so thing which should be rectified in view of its im- often unnoticed level is extremely widespread portance and public health impact. in developing countries; • Even at this subclinical level, it has a significant adverse effect on morbidity and mortality In Memoriam rates; Martin Frigg, 1943–2010 • Other groups besides young children, such as pregnant and lactating women and school-age While this third edition of the Sight and children, are also highly susceptible to vitamin Life Manual on Vitamin A Deficiency Dis- A deficiency; orders was in preparation, we received news • Vitamin A and both pro- and non-pro-vitamin A of the sad death of Martin Frigg, Director- carotenoids are increasingly being recognized General of Sight and Life from 1994 to as playing an important role in diseases which 2005. Although Martin did not contribute to affect all age groups throughout the world, in- the present volume, it would not exist without cluding age-related macular degeneration, can- his invaluable editorial work on the previous cer, metabolic syndrome and cardiovascular two editions. disease; and finally, • New terminology is now generally accepted We gratefully acknowledge Martin’s past con- for the expanded and enhanced role that vita- tribution and deeply lament his passing. min A is now recognized to play – vitamin A deficiency disorders (VADD). The Editors Although many micronutrients are intimately involved in human health and the relationship Medicine is often considered to be both a between the micronutrients is increasingly impor- science and an art. In reality it is neither, although tant, the pioneer in vitamin deficiency research, it employs both science and art to meet its ends. vitamin A, remains an important and interesting It is better understood as a human endeavor dedi- research field. We trust that this Third Edition cated to the prevention and cure of disease and of the Vitamin A Manual will add to the body of the alleviation of suffering. Diseases, like the literature that contributes towards interventions civilizations that suffer from them, rise and fall, that ultimately ensure sustainable and significant occasionally without the cause(s) or the true ef- improvements in human nutrition, health and fects of the measures employed against them be- well-being. ing known. Amongst those diseases attributed to an inadequacy of nutrients in the diet, vitamin A deficiency was shown to hold a prominent place in the middle of the 20th century. At that point, it was the most common cause of blindness in young children worldwide. Within several decades, as a result of research that led to interven- Donald McLaren tions, severe blinding vitamin A deficiency (xe- Klaus Kraemer V5

Foreword

A Powerful Commitment able benefits to some of the world’s poorest and most disadvantaged people. Sight and Life Hidden hunger is the number one cause achieves this by invoking an array of approaches. of death in the world, killing more people than These range from publishing authoritative works AIDS, malaria and tuberculosis combined. The such as this volume through disseminating up- term describes the malnutrition that results from to-the-minute information and insights via Sight a lack of essential vitamins and minerals. Malnu- and Life Magazine to staging conferences and trition and hunger in combination currently cause workshops, as well, of course, as funding original the deaths of 3.5 million children under the age of research work and field programs on the ground. five every year. The achievement of no fewer than six out of the eight United Nations Millennium The role of Sight and Life is therefore Development Goals is contingent on the elimina- to analyze what is good and what is new in the tion of hidden hunger. micronutrient arena but also to be aware of what is tried and tested so that it can effectively advocate Among the vitamins essential for healthy good policy- and program-making. I know from growth and development, vitamin A plays a key my work with them that the Sight and Life- role. It was the recognition of the link between team are passionate about the work they do and vitamin A deficiency and blindness in malnour- deeply dedicated to ensuring a sustainable and ished populations that inspired the founding significant improvement in the health and well- of Sight and Life in 1986 and, indeed, gave the being of some of the world’s poorest and most humanitarian initiative its name. vulnerable people.

Since that time much has been learned about DSM fully supports Sight and Life. In our the characteristics and effects on the human partnership with the World Food Programme physiology of a wide range of vitamins and miner- (WFP), which has been running since 2007, als. Moreover, scientists have come to appreciate Sight and Life plays a key role. It is DSM’s as never before the complex interactions of these belief that together DSM and WFP can make a micronutrients. A great deal of original research difference to the lives of millions of people. Im- remains to be done, but the way forward is clear: proving nutrition means improving lives: it means the elimination of hidden hunger worldwide can breaking the vicious circle of poverty and hunger only be achieved if scientists, policymakers, non- and malnutrition which leads to suffering, disease governmental organizations and the private sector and underachievement. join forces to defeat this public health scourge. We believe that by improving people’s lives Sight and Life has an important role to in this way we can make a major contribution to play in this effort. It is the role of a committed unleashing the untapped human and economic advocate. Sight and Life maintains close and potential of many countries around the globe. productive relationships with policymakers, other As our CEO Feike Sijbesma has often said of non-governmental organizations, researchers and DSM, we cannot be successful, nor can we call private sector organizations to advocate targeted ourselves successful, in a society that fails. This policies and effective programs that bring measur- powerful corporate and personal commitment VI Foreword

was recognized in October 2010 when the United It is an honor of which all of us at DSM are Nations Association of New York awarded Feike deeply proud. And we will be proud to continue Sijbesma the prestigious 2010 Humanitarian of our efforts to improve the lives of millions of the Year award for his outstanding commitment people who so urgently need our help. to corporate social responsibility (CSR) and in particular for DSM’s partnership with the United Nations World Food Programme. Stephan Tanda Member of the DSM Managing Board with responsibility for DSM’s Nutrition cluster of businesses VII7

Table of Contents

1. FOREGROUND Setting of VADD 11 Nature of VADD 13 Significance ofVADD 14

2. VITAMIN A IN NATURE Light 17 Carotenoids and retinoids 17 Photosynthesis 20 Rhodopsin where least expected 21 Bacteriorhodopsin 22 A “typical” eye 22 Variations in chromophores 25 Photoentrainment 26 Vitamin A functions unrelated to light 26

3. METHODS OF ANALYSIS Analytical procedures 27 Bioassay procedures 27 Carotenoid analysis in foods 28

4. FOOD SOURCES Introduction 29 Units of vitamin A activity 30 Provitamin A carotenoid sources 31 Preformed vitamin A 34

5. BIOAVAILABILITY OF CAROTENOIDS Introduction 37 Recent advances in technology 38 Isotope dilution techniques 38 Factors affecting the bioavailability of carotenoids 39 Other techniques 42 Conversion factors 42 VIII Table of Contents

6. VITAMIN A IN HEALTH Digestion and absorption 43 Transport to the liver 44 Metabolism in the liver 45 Transport between cells 46 Cellular retinol-binding proteins (CRBP) 48 Gene function and retinoids 50 Functions of vitamin A 51 Activities of carotenoids 58 Human requirements 59

7. ASSESSMENT OF VITAMIN A STATUS Introduction 61 Indicators at the subclinical level 63 Indicators at the clinical level 71 Assessment of dietary intake of vitamin A 72

8. XEROPHTHALMIA Definitions 75 Night blindness (XN) 76 Conjunctival xerosis (X1A) 80 Bitot’s spot (X1B) 80 Corneal xerosis (X2) 82 Keratomalacia (X3A, X3B) 84 Corneal scar (XS) – vitamin A related 86 Xerophthalmic fundus (XF) 86

9. MORTALITY AND MORBIDITY ESPECIALLY IN RELATION TO INFECTIONS Introduction 89 Mortality associated with non-corneal xerophthalmia 89 Vitamin A intervention and mortality in young children 90 Cause-specific mortality 93 Contribution of VAD to child mortality 93 Measles mortality and vitamin A treatment 95 Vitamin A suplementation and maternal mortality 96 Vitamin A supplementation and infectious morbidity 97 Remarks 106 Table of Contents IX

10. OTHER EFFECTS OF VAD Anemia 107 Growth 109 Immune response 112 Normal and diseased skin 114 Thyroid function 114 Reproductive systems 114 Bone 114

11. INTERACTION OF VITAMIN A AND OTHER MICRONUTRIENTS Introduction 115 Vitamin A and iron 115 Vitamin A and zinc 116 Vitamin A, iron and zinc supplementation 117 Vitamin A and iodine 118 Vitamin A, riboflavin and other micronutrients 118 Other multi-micronutrient combinations 118

12. GLOBAL OCCURRENCE Prevalance of VADD 121 VAD and blindness 121 Vitamin A in global food supplies 133 Methodology 134 Trends 135 Country situations 137

13. EPIDEMIOLOGY Introduction 141 Age 141 Physiological status 143 Diet 143 Breastfeeding 144 Cultural factors 145 Infectious diseases 146 Protein-energy malnutrition (PEM) 148 Season 149 Socioeconomic status 150 Location 150 Sex 152 The VADD cycle 154 X Table of Contents

14. RETINOIDS AND CAROTENOIDS IN GENERAL MEDICINE Introduction 155 Secondary VAD 155 Hypervitaminosis A 156 Non-nutritional diseases responsive to vitamin A and/or carotenoids 159 Eye diseases 160 Dermatological pharmaceutical use of synthetic retinoids 162 Cancer chemoprevention and treatment 162 Carotenoids and chronic diseases 163 Cardiovascular disease 164 Other systems 164

15. CONTROL Introduction 167 General considerations 168 Treatment 169 Prophylactic supplementation 170 The special issues of 1) newborn VAS and 2) effects of vaccines 173 Prevention and management of infectious diseases 175 Food fortification 176 Dietary modification 178 General issues 180 Plant breeding and genetic modification 181 Disaster relief 182 11

1 Foreground

The story of how the complex of disorders providing an outline of their setting, nature, and of human health now known as Vitamin A Defi- significance. This outline is meant to introduce ciency Disorders (VADD) came to be recognized the subject in such a way that a degree of confi- over the course of time is a truly remarkable dence, involvement and concern is instilled in the one. Detailed accounts have been published reader from the outset. in recent years (McLaren 1999 and 2004a). Ta- ble 1.1 provides the dates of key events in the story. SETTING OF VADD

Table 1.1: Historical background. Vitamin A deficiency is just one among many 1816 Xerophthalmia in experimental vitamin deficiencies that occur in humans. The animals form that damages the eye and vision, known as xerophthalmia, has been documented since ancient times. Other vitamin deficiencies, similarly 1904 First large human epidemic in Japan long since recorded, include beriberi (thiamine deficiency), pellagra (niacin deficiency), scurvy 1913 “Fat soluble A” discovered in the (vitamin C deficiency), rickets and osteomalacia USA (vitamin D deficiency). Vitamins and certain minerals make up the 1930 Formulae of β-carotene and vitamin A micronutrients – substances required for health in discovered (Karrer, Switzerland) very small amounts, i.e. just a few milligrams or micrograms per day. There are also several lipids, 1947 Industrial vitamin A synthesis known as essential fatty acids, which are needed (Isler at Roche, Switzerland) in the diet in small amounts. Finally, there are macronutrients, the fats, proteins, and carbohydrates needed in relatively 1964 First global survey of VADD large amounts to help provide the body with energy and fulfil its structural requirements. Some 1980s Importance of vitamin A in child of the building blocks of proteins, amino acids, survival shown are “essential” nutrients. Varying combinations of deficiencies in all 1990s Maternal mortality reduced with these substances lead to what is probably the retinol or β-carotene supplements most widespread nutritional deficiency disease Discovery of retinoid receptor gene of all: protein-energy malnutrition (PEM). In its family most severe clinical forms this syndrome includes kwashiorkor and marasmus, but the majority of cases are characterized by failure to thrive and It is intended that detailed and up-to-date in- underweight. formation on every relevant aspect of VADD will The latest press release on global young child be systematically covered in the chapters that fol- annual mortality data is that of UNICEF (10 Sep- low, with attention being focused in particular on tember 2009). According to these data the abso- 12 1 Foreground

lute number of child deaths in 2008 declined to an response (see Chapter 6). Infections in turn often estimated 8.8 million, from 12.5 million in 1990. impair vitamin A status by decreasing appetite, The data show that global under-five mortality increasing demands on the body, and sometimes has decreased steadily over the past two decades. causing loss of the vitamin through the urine (see The average rate of decline from 2000 to 2008 is Chapter 14). 2.3%, compared to a 1.4% average decline from Recent estimates by WHO (Bryce, Boschi- 1990 to 2000. Pinto, Shibuya et al. 2005) suggest that approxi- Table 1.2 lists the five main nutritional de- mately 10.6 million children under the age of five ficiency diseases. When the WHO first named die annually, most in developing countries. In the four major deficiency diseases as being of global same report, pneumonia, surprisingly, turned out public health magnitude (Anon 1972), zinc defi- to be the leading killer of children, responsible for ciency in humans had been barely recognized (see about one third of all deaths under five (Figure Chapter 10). It is rather depressing to have to note 1.1). that after more than 30 years the original four de- Undernutrition was an underlying cause in ficiency diseases still remain major problems and 53% of all deaths presented here. Further details have been joined by a fifth. are not available, but it would be reasonable to assume, from work referred to below (see Bryce, Table 1.2: Major nutritional deficiency diseases. Black, Walker et al. 2005), that VAD and zinc 1. Protein-Energy Malnutrition (PEM) deficiency play an important role. However, it appears that the term “undernutrition” only ap- 2. Vitamin A Deficiency Disorders (VADD) plies to the presence of wasting and/or stunting. Growth retardation is associated with deficiencies of vitamin A or zinc, and possibly with other mi- 3. Nutritional Anemias (especially Iron) cronutrient deficiencies. In 2002 a groundbreak- ing study was published which showed the global 4. Iodine Deficiency Disorders (IDD) and regional burden of disease and the 26 selected major risk factors responsible (Ezzati, Lopez, 5. Zinc Deficiency Rodgers et al. 2002). The burden of disease was measured as death and disability, represented by a unit termed “disability-adjusted life year” More than a third of all children under the (DALYs). Child and maternal underweight were age of five in developing countries are affected the fourth highest cause of mortality, with high by PEM: by international criteria they are either blood pressure as the first. When deficiency of wasted, stunted, or underweight (de Onis, Mon- iron, zinc and vitamin A are added to under- teiro, Akre et al. 1993). It is probable that many weight, to form a category of malnutrition of of them also suffer from VADD, but statistics are their own, the outcome is then second overall. not available because in recent years the terms For DALYs, undernutrition is the single leading PEM, kwashiorkor and marasmus for the severe global cause of loss of health constituting 16% clinical forms of childhood malnutrition have of the total burden. For iron, vitamin A, and zinc been abandoned and there are no records of their deficiency, the DALYs are 2.4%, 1.8%, and 1.9%, prevalence, trends, or mortality worldwide. with iodine deficiency only 0.1%. Another important feature of the setting of In 2006 (Lopez, Mathers, Ezzati et al.) an up- VADD is the frequency of the infectious diseases date of these data showed some changes. Child that accompany them. The relationship is syner- underweight for age became the fourth attribut- gistic: VAD weakens the defenses of the body able cause of deaths, with zinc, vitamin A, and and has specific damaging effects on the immune iron deficiencies in 12th, 14th and 15th positions 1 Foreground 13

AIDS

Diarrheal diseases

Note that undernutrition is implicated in 53% of all deaths among children under five.

Pneumonia causes 19% of all under-5 deaths. This figure, however, does not include deaths caused by pneumonia during the neonatal period. It is estimated that 26% of neonatal deaths, or 10% of under-5 deaths, are caused by severe infection. A large proportion of these infections is caused by pneumonia/sepsis. If these deaths were added to the overall estimate, pneumonia would account for up to 3 million, or as many as one-third (29%) of under-5 deaths worldwide. Figure 1.1: Pneumonia is the leading killer of children worldwide. This figure shows global distribution of cause- specific mortality among children under five in 2004 (Bryce, Boschi-Pinto, Shibuya et al. 2005).

respectively. For the attributable disease burden vitamin A and zinc supplementation were among (% global DALYs), child underweight for age the 18 measures chosen. The total annual cost of was first, with zinc, iron, and vitamin A deficien- new resources required was estimated at US$ 5.1 cies 11th, 13th and 14th respectively. In the first billion. of a series of papers, known as The Lancet Series, devoted to the latest assessment of “Maternal and child malnutrition”, Black, Allen, Bhutta et al. NATURE OF VADD (2008) dealt with global and regional exposures and health consequences. It was estimated that in The term VADD has been defined as “a com- children under five years in developing countries prehensive term that covers all aspects of the de- stunting, wasting, and IUGR (intra-uterine growth ficiency state of vitamin A and includes adverse retardation) together account for 22% of deaths effects on health, survival and vision.” and 21% of DALYs. In the same age group VAD Until quite recently, vitamin A deficiency, in and zinc deficiency account for 0.6 million and the form of xerophthalmia, was considered to 0.4 million deaths respectively, and combined for constitute a major cause of preventable blind- 9% DALYs. ness in developing countries. Similarly, iron defi- The aim of a recent study (Bryce, Black, ciency was regarded as the most widespread form Walker et al. 2005) was to calculate the cost of of anemia, mainly in children and women of saving the lives of 6 million children each year child-bearing age. Iodine deficiency was recog- throughout all of the 42 countries in which nized as the primary cause of endemic goiter. 90% of child deaths occurred in 2000. Both Rickets and osteomalacia were acknowledged 14 1 Foreground

forms of bone disease, in young children and SIGNIFICANCE OF VADD elderly adults respectively, caused by vitamin D deficiency. Just over four decades ago xerophthalmia was It has become evident that even mild de- demonstrated to be of worldwide significance grees of these and all other forms of nutritional (Oomen, McLaren, Escapini 1964). Latest avail- deficiency have important adverse implications able health statistics show the impact of VADD for health. And as a consequence, these deficien- on the health of young children (Table 1.3) and cies are much more widespread than previously the impact of vitamin A intervention on child thought, and are thus of much greater public survival: health importance. “Improvement of vitamin A status in young In the specific case of VADD, much prog- child populations… leads to a reduction in all- ress has been made in the identification of the cause mortality rates of about 23%.” (Beaton, subclinical stages of deficiency (see Chapter 7). Martorell, Aronson et al. 1993). Following upon the great efforts that were made “Improved vitamin A nutriture would be ex- in recent decades to combat xerophthalmia (see pected to prevent approximately 1.3 – 2.5 million McLaren 1999), this severe form of the disease deaths annually among children aged under 5 has been greatly reduced. Even so, there remains years.” (Humphrey, West, Sommer 1992). much more work to be done (Table 1.3). In recent years attention has been focused on A further aspect of the nature of nutritional vitamin A deficiency in women. It transpires that diseases that is becoming increasingly appar- they are indeed a susceptible group (Table 1.4). ent is the fact that, at the subclinical level, most Recently estimates were made of the loss of of the underprivileged members of developing DALYs attributable to the present prevalence of countries, of all ages, are subject to a lifetime of VAD in 0–4 year old children worldwide (Table dietary inadequacy that affects many nutrients. 1.5) (Mason, Musgrave, Habicht 2003). Another In such cases multi-micronutrient (MMN) defi- group (Levine, Pollit, Galloway et al. 1993) ciency becomes the rule. This situation, which is calculated the return on investments from VAD complex, serious, and urgent, poses a new chal- control in US$ (Table 1.6). These data are indeed lenge to those concerned with human health and impressive, concerning both the potential benefit welfare (see Chapter 11). from the reduction in disease burden from elimi-

Table 1.3: Impact of vitamin A deficiency on health of preschool-age children (latest estimates of global numbers affected). 190 million with low vitamin A status indicated by serum retinol <0.70 μmol/L (WHO 2009 ) 5.7 million with night blindness (XN) (WHO 2009) 4.424 million with clinicial xerophthalmia (one or more of WHO 1982 xerophthalmia classification eye signs present) (West, Darnton-Hill 2008) 0.4 million blind due to VAD (no data being collected, but previous studies suggested about 10% of those with clinical xerophthalmia. Also responsible for about 10% of all blind children. See previous editions of this book). 1 Foreground 15

nating VADD in 0–4 year old children, and also cluded that for each US$ spent on vitamin A and the possible return on investments from VAD zinc supplementation a return on investment of control. US$ 17 can be expected (see also Chapter 15). This brief overview may be concluded on an In 2008 (Rosenbloom, Kaluski, Berry) a new optimistic note. As long ago as 1993, in its De- concept was introduced, called the global nutri- velopment Report, Investing in Health, the World tional index (GNI), modelled on the human de- Bank published a study in which the benefits and velopment index. For each country it is based on costs of 47 health interventions were evaluated. three indicators of nutritional status: deficits, ex- As Figure 1.2 shows, vitamin A supplementation cess and food security, and for the first time pro- was among the very best performers in both re- vides a single statistic for each country according gards. In 2008, the Copenhagen Consensus con- to its overall level of nutrition.

Table 1.4: Impact of VADD on health in pregnancy (from WHO 2009).* 19.1 million with serum retinol <0.70 μmol/L 9.75 million with night blindness (XN) *no data given for xerophthalmia or blindness

Table 1.5: Reduction in disease burden from eliminating VAD in 0–4 year old children through decrease in mortality risk (from Mason, Musgrave, Habicht 2003). Region Prevalence of clinical DALYs lost from % Decrease in VAD (%) VAD (millions) DALYs India 1.00 20.10 19.26 China 0.20 1.30 3.76 Other Asia 0.76 6.60 12.22 Sub-Saharan Africa 0.93 20.80 15.59 Latin America 0.25 1.20 5.00 Middle East Crescent 0.28 4.10 6.75

Table 1.6: Return on investments from VAD control in US$ (from Levine, Pollitt, Galloway et al. 1993).

Discounted value of Cost per DALY Intervention Cost of lives saved productivity gained gained/program Supplementation (children <5 years) 0.50 1.40

Fortification 154 47.50 4.20 16 1 Foreground

Costs per intervention year (US$)

Figure 1.2: Benefits and costs of 47 health interventions. Dots include ages under and over 15. (Reproduced with permission from World Bank). 17

2 Vitamin A in Nature

Substances closely related to vitamin A abound in both plant and animal life and take part in a most amazing variety of processes. In one way or another, many of these processes are associated with the action of light. Without light there would be no life as we know it. As far as is known, there is no other molecular group in nature that matches the versatility of vitamin A and its relatives.

LIGHT

Light rays are composed of small packets of energy called photons. In that very small part of the electromagnetic spectrum that can be detected by our photoreceptors, termed the visual spectrum, the energy level of photons specifies a col- or, ranging from red to violet. The longest wavelength (red) has the lowest intensity of brightness, Figure 2.1: Electromagnetic radiation spans an enor- or flux, and the shortest wavelength (violet) has mous range of frequencies or wavelengths. The spectrum is customarily designated by fields, waves and the highest intensity (Figure 2.1). Thus light particles in increasing magnitudes of frequencies, with can be conceived of in two ways; as quanta, or the visible spectrum occupying a very small fraction. photons of varying flux, and as waves of varying (Adapted from Electromagnetic radiation. Encyclo- wavelength. paedia Britannica Macropaedia 1999:195-211).

The chemical structure of carotenoids and CAROTENOIDS AND RETINOIDS retinoids is very similar (Figure 2.2). Carotenoids have a basic C40 skeleton which is made up from Each molecule addressed here belongs to one successive additions of C5 isoprene units. Reti- of two related classes of substances, carotenoids noids approximate to half of the carotenoid mol- or retinoids. Members of each are found in both ecule, consisting of four isoprenoid units joined plants and animals, but carotenoids predominate together in a head-to-tail manner and customar- in plants and retinoids in animals. The major- ily containing five conjugated double bonds. The ity of retinoids – perhaps all – are derived from term vitamin A is used as a generic descriptor for carotenoids, which appear to have originated in retinoids that qualitatively exhibit the biological archebacteria several billion years ago towards activity of all-trans retinol. the beginning of evolutionary time, long before Most readers are probably unaware that re- plant life proper. They were part of the process cently electrophysiology as the standard method that provided our previously anoxic earth with its for investigating neuronal signaling is being re- oxygen-rich atmosphere (Margulis, Sagan 1995). placed by various optical probes as a result of new 18 2 Vitamin A in Nature

forms of microscopy. Among these amazing new Carotenoids are yellow, orange or red pig- probes the type includes forms of rhodopsin (e.g. ments. Over 600 have been identified and the list channelrhodopsin and holorhodopsin) that will is constantly growing. Of the 600 over 50 have appear several times later (Scanziani, Häusser a structure that enables them to be converted into 2009). vitamin A, which includes at least one unsub- Examples of some non-provitamin A carot- stituted β-ionone ring and a polyene side chain. enoids appear below (Figure 2.3). The other end of the molecule may have a cyclic

Figure 2.2: (A) Vitamin A (retinol); (B) all-trans-β-carotene; (C) all-trans-retinoic acid; (D) all-trans-retinal; (E) 11-cis-retinal; (F) 3-dehydroretinal.

Figure 2.3: (A) lycopene; (B) lutein; (C) β-ionone. 2 Vitamin A in Nature 19

or an acyclic structure. It may be lengthened but isomer to increased thermodynamic instability. not shortened to less than an 11-carbon polyene The ability of carotenoids to absorb visual light is chain. Chain lengthening decreases activity. Ul- related to the presence of delocalized pi electrons. timately, all vitamin A in nature originates from In plants, energy can transfer from excited carot- one or other of these provitamin carotenoids. enoids, generating excited chlorophyll, which is All-trans β-carotene is by far the commonest but active in photosynthesis (see below). In leaves α-carotene, γ-carotene and β-cryptoxanthin con- carrying out photosynthesis, the physical struc- tribute to lesser extents (Table 2.1). ture of the chloroplast, the subcellular structure containing the carotenoid, facilitates this transfer of energy to chlorophyll (see Figure 2.4) (Britton Table 2.1: Relative provitamin A activity of various 1995). carotenoids (Simpson, Tsou 1986). Under conditions of high light intensity chlo- Carotenoid Activity (%) rophyll in the triplet state can accumulate and β-carotene 100 cause damage. Carotenoids can counteract this α-carotene 50–54 effect in two ways: by deactivating the triplet state of chlorophyll or by converting single-state γ-carotene 42-50 oxygen to its ground triplet state and again allow- 3,4-dehydro-β-carotene 75 ing release of the transferred energy as heat (Pas- β-carotene-5,6-epoxide 21 cal, Liu, Broess et al. 2005). With the advance α-carotene-5,6-epoxide 25 of global warming, this protective function is increasingly significant. 3-oxo-β-carotene 52 Because of their high degree of unsaturation, 3-hydroxy-β-carotene carotenoids can extract or donate electrons, re- (cryptoxanthin) 50–60 sulting in radical anions and cations which can 4-hydroxy-β-carotene 48 react with oxygen or other molecules; they show β-2’-apo-carotenal active both antioxidant and prooxidant properties under β-8’-apo-carotenal 72 various conditions (Britton 1995). These properties are being investigated in relation to the lycopene inactive prevention or treatment of various diseases (see lutein inactive Chapter 14). Of special interest has been the 3,3’-dihydroxy-β-carotene discovery that high concentrations of two (zeaxanthin) inactive carotenoids, zeaxanthin and lutein, are normally deposited in the human retina. Low concen- The most characteristic feature of the carot- trations appear to be associated with develop- enoid structure is the long system of alternating ment of age-related macular degeneration (see double and single bonds, usually 9–13 in number, Chapter 14). the polyene chain that forms the central part of The attractive red, orange, and yellow colors the molecule. It is this that gives carotenoids their of many autumn leaves are due to their xantho- distinctive molecular shape, chemical reactivity phyll content. Xanthophylls are non-provitamin and light-absorbing properties. Usually, the most A carotenoids that have one or more oxygen stable form of the polyene chain is a linear, ex- molecules on the ring or in the chain. It has been tended conformation, as in lycopene. suggested that these changes in color amount to a Each of the many carotenoid conjugated dou- warning to pests that the tree would fight any in- ble bonds can form vast numbers of geometric festation (Hamilton, Brown 2001). Abscisic acid, isomers; theoretically hundreds of thousands. The an anti-growth hormone and carotenoid deriva- presence of a cis-double bond predisposes the tive, controls the fall of leaves in autumn. Xantho- 20 2 Vitamin A in Nature

phylls also contribute to the striking plumage of PHOTOSYNTHESIS many birds. Both provitamin and non-provitamin carotenoids are responsible for the appealing red By far the best known function of carotenoids and yellow colors of fruits and vegetables, and is the part they play in photosynthesis, which in both classes are of great nutritional and health simple terms is the process whereby plants, in significance. Carotenoids also are found in algae, the presence of light, cause carbon dioxide (CO2) fungi, yeasts, moulds, mushrooms, bacteria and in from the atmosphere to combine with water (H2O) all classes of plants and animals. to form sugar or starch, with oxygen (O2) as a by- Carotenoids and retinoids are nonpolar and product. hydrophobic. The formation of complexes with The green pigment chlorophyll, in particu- proteins allows them access to aqueous media, lar chlorophyll-a, is responsible for most light such as plasma. The many activities of carot- absorption, but the red, orange or yellow carotenoids have been classified into functions, actions enoids, covering other parts of the visual spec- and associations (Olson 1996) (see Chapter 4). trum, may account for about one third of total photosynthesis (Figure 2.4).

Figure 2.4: Photosynthesis at different wavelengths. (A) The ability of light of different wavelengths to support photosynthesis; (B) The absorption spectra for three photosynthetic pigments; chlorophyll a, chlorophyll b, and β-carotene. (Adapted from Lodish, Baltimore, Berk et al. 1995). 2 Vitamin A in Nature 21

Figure 2.5: The structure of a chloroplast with simplified membrane structure. (Adapted from Hopkins, Huener 1995).

Photosynthesis takes place in an organelle the retina responsible for vision under conditions known as a chloroplast (Figure 2.5). Like the of restricted illumination, or night vision. They mitochondrion it has been shown to have its own might be surprised to learn that there are many DNA, different from that of the cell, and it is now hundreds of different forms of rhodopsin found accepted to be an example of eusymbiosis, where in nature, many in simple organisms of the plant early in evolution one organism has been incor- kingdom. In all instances, rhodopsin consists of porated into another for the mutual benefit of both two parts: 1) a membrane protein known as a G- (Margulis, Sagan 1995). protein which differs slightly in each case; and 2) a prosthetic group, or chromophore, which is a special form of vitamin A and responsible for the RHODOPSIN WHERE LEAST EXPECTED unique process of phototransduction (see below for details) (Figure 2.6). Most readers are familiar with rhodopsin, or The importance of the class of proteins known visual purple, as the visual pigment in the rods of as G-proteins has been discovered only in recent

Figure 2.6: Rhodopsin molecule. The dashed circle in the centre of the transmembrane region indicates the site and plane of 11-cis-retinal to the lysine residues at position 206 and 209. (Adapted from Forrester, Dick, McMenami et al. 1996). 22 2 Vitamin A in Nature

years. Many receptors linked to effector systems The essence of this process is the photoiso- by trimeric guanosine triphosphate (GTP)-bind- merization of the all-trans retinal chromophore ing proteins (G-proteins) have been described. to the 13-cis form, which causes a bend in the These proteins regulate the activity of a specific all-trans molecule that amounts to only 2 Å plasma membrane enzyme or ion channel (as in (1 angstrom = a hundred millionth of a centi- the case of rhodopsin). meter). At the same time, a proton (H+ atom) is pumped from the cytosol (inside of the cell) to the outside. BACTERIORHODOPSIN Something very similar happens in the visual cycle in animals (see below). Of the many plant rhodopsins, the first discovered and the best understood is bacteriorhodopsin, produced by the archebacterium Halo- A “TYPICAL” EYE bacterium halobium (Stryer 1995). In the presence of abundant oxygen, this organism oxidizes The entire visual process is highly complex fuel molecules to generate ATP, as do most aerobic and requires lengthy description. In the present organisms. However, when oxygen is scarce, it context, we limit our discussion to the relevant synthesizes large amounts of bacteriorhodopsin, involvement of vitamin A in the initial process which takes part in a light-driven proton pump of phototransduction (conversion of light energy (Figure 2.7). into nerve cell energy). Differences between in-

Coupled isomerization Photoisomerization

Figure 2.7: Proposed mechanism of light-driven proton pumping by bacteriorhodopsin. (Adapted from Mathies, Lin, Amed et al. 1991). 2 Vitamin A in Nature 23

vertebrates and vertebrates and certain special bright light (where colors can be distinguished). cases will be commented upon, but the human eye Cones and rods have differences in both structure is here presented as a generally “typical” case, and function (Figure 2.9). The chromophore is particularly with regard to phototransduction. The the same in both, 11-cis-retinal (not 13-cis-retinal emergence in evolution of the earliest true eye (as as in bacteriorhodopsin – see above), but the G- opposed to primitive luminescent “eye spots”) proteins differ. Rhodopsin is in the rods, and in is believed to have occurred in a trilobite and to man there are three slightly different versions of have triggered the Cambrian explosion about 544 a similar G-protein, iodopsin, in the cones. These million years ago (Parker 2003). cone photoreceptors absorb either blue, green or The retina is composed of ten layers of spe- red light, giving man, and all other mammals, cialized cells, and in mammals the photoreceptors trichromatic vision. It has been proposed that form the outermost of these underneath the reti- complex color vision evolved to facilitate the nal pigment epithelium (RPE) (Figure 2.8). This gathering of colored fruits and young leaves for strange arrangement, which means that before food (Sumner 2002). Some animals have fewer phototransduction can take place light has to pass types of cones, some have more (up to four), through all other nine layers, has never been fully while others, including birds, can even see ultra- explained. Even so, it does not seem to interfere violet light. with the clarity of images. The photoreceptors of the human retina are There are two types of photoreceptor: rods, so sensitive that one can be triggered by a single for vision in poor light; and cones, for vision in photon. In a rod this generates a current of ap-

Figure 2.8: Schematic diagram of the ten layers of the retina. (Adapted from Cormack 1987:691). 24 2 Vitamin A in Nature

Figure 2.9: Diagram of a generalized mammalian retinal cone (right) and rod (left). (Adapted from Cormack 1987:692).

proximately 1–3 picoamperes; in a cone it gen- base of retinal to the all-trans form. This erates only about 10 femtoamperes or about one markedly alters the geometry of retinal. The hundredth that of a rod. However, the response Schiff base linkage of retinal moves approximate- time of a cone is about four times faster than that ly 5 Å (angstrom) in relation to the ring portion of a rod. Cones are better suited for discerning of the chromophore (Stryer 1995) (Figure 2.10). rapidly changing events; rods for low-light visual Before light excitation, 11-cis-retinal locks acuity. the receptor protein (opsin) in its inactive form. The process of phototransduction begins in The energy in a photon has been converted into the the outer segment of the rods and cones and the energy of atomic motion to the extent of 5 Å. In the signal generated passes down a fine cilium to the case of bacteriorhodopsin (see above) the 13-cis- inner segment. In the human eye, there are about isomer is used and the energy shift is 2 Å and not 100 million rods and only 3 million cones. 5 Å. Much of the isomerization of retinal takes The primary event in visual excitation is the place within 200 femtoseconds, one of the fastest isomerization of the 11-cis-isomer of the Schiff reactions in nature, giving rise to a series of tran- 2 Vitamin A in Nature 25

Figure 2.10: Isomerization of the 11-cis-isomer of the Schiff base of retinal to the all-transform. (Adapted from Stryer 1995:335).

sient intermediates. One of these, metarhodopsin frog to 625 nm for the red-absorbing cone of the II, called photoexcited or activated rhodopsin, goldfish. The G-protein (opsin family) variation, triggers an enzymatic cascade that activates a and not the chromophore, is responsible for most G-protein (transducin) cascade leading to cyclic of these differences. GMP hydrolysis, which closes cation-specific The main chromophores of the visual pig- channels to eventually generate a nerve signal. ments are derived from one of the two forms In this process, the rhodopsin molecule al- of vitamin A; vitamin A1 (all-trans-retinol) or ters its light-absorbing properties, from magenta vitamin A2 (all-trans-3-dehydroretinol). The through orange to yellow and ultimately to white chromophores are respectively retinal (the great or “bleached.” At this stage it has become opsin majority) and 3,4-didehydroretinal. So far only and dissociated all-trans-retinal. The opsin under- three other retinal congeners have been identified goes regeneration by binding another molecule throughout the animal kingdom. All are versions of 11-cis-retinal. In addition to restoring light- of hydroxyretinal and occur mainly in insects. absorbing capacity, this is critical for shutting off These chromophores are thought to be derived the pigment’s catalytic activity and thus allowing from the cleavage of β-carotene to retinal fol- full dark adaptation to occur. lowed by hydroxylation within the retina (Seki, Another protein, RPE65 in the RPE, performs Isono, Ozkaki et al. 1998). a key role in vision. The palmitoylation of RPE65 The visual pigments of freshwater fish differ serves to switch off the visual cycle in darkness from those of marine fish and almost all other ani- and to switch it on in light (Wolf 2005). Recent mals. Freshwater fish and some amphibians have work shows that another function taking place 3-dehydroretinal as chromophore and porphyrop- in human RPE cells is cleavage of β-carotene sin for visual pigment, with gamma max near by carotenoid 15,15ʼ-monooxygenase (CMO1) 540 nm (compared with all-trans-retinal and (Chichili, Nohr, Schaeffer et al. 2005). rhodopsin, with gamma max near 500 nm). There is evidence that the retinas of freshwater fish and some amphibians can convert VARIATIONS IN CHROMOPHORES retinal into 3-dehydroretinal. This process may be regulated largely by properties of light in Throughout the animal kingdom hundreds of the natural environment and is sometimes under visual pigments have been characterized by their hormonal control. Age and diet also play a absorption maximum (gamma max). The gamma part. The difference in the gamma max of the max range from 432 nm for the green rod of the different chromophores appears to be an adap- 26 2 Vitamin A in Nature

tation that allows fish to make use of the longer Vitamin A itself (all-trans retinol) has no wavelengths of light found in fresh water as known biological activity. The different retinoid compared with sea water. forms present in the body are generated mainly Euryhaline fish (those that tolerate a wide through modifications to the terminal polar end range of salinity in water), have a mixture of both of the molecule. Retinol and retinyl esters (main- rhodopsin and porphyropsin. In the frog Rana ly palmitic, oleic, stearic, and linoleic) are the pipiens the chromophore in the tadpole is 11-cis- most abundant forms. Formation of retinyl esters 3-dehydroretinal. In the adult, this is replaced by makes the retinoid less toxic and permits its stor- 11-cis-retinal. age, mainly in the liver, within intracellular droplets. Aside from in the eye (see above), retinal has PHOTOENTRAINMENT no known function other than serving as an intermediate in the synthesis of retinoic acid (RA). The term photoentrainment applies to the phe- All-trans RA is formed through the irreversible nomenon of regulation and control of a wide vari- oxidation of all-trans-retinal. ety of metabolic and other processes in which light The all-trans- and 9-cis-isomers of RA are entering the eyes brings about daily or other regu- transcriptionally active and account for the gene lar changes also known as circadian rhythm or bio- regulatory properties of retinoids within most logical clock (Foster, David-Gray, Lucas 2001). cells and tissues. RA can bind three retinoid A In mammals this clock is located within a receptors (RARα, RARβ, and RARγ) and three paired nucleus in the brain above the crossing of retinoid X receptors (RXRα, RXRβ, and RXRγ). the optic nerves, known as the suprachiasmatic Well over 500 genes may be regulated by RA. nucleus or SCN. Until recently, it was assumed Because there are no known enzymes that can that photoentrainment was carried out by the pho- reduce RA to retinal, excess RA must be catabo- toreceptors, the rods and cones. Recently it has lized. This probably accounts for the presence of been shown that so-called intrinsically photosen- small amounts of many other forms of retinol and sitive retinal ganglion cells (ipRGCs), compos- RA. ing one of the ten layers of the retina, contain an This is only an outline of the subject and more opsin/11-cis-retinal-based photopigment called details of certain aspects are given in Chapters melanopsin (Dkhissi-Benyaha, Rieux, Hut et al. 4–6, which consider the roles vitamin A and its 2006). Consequently, the circadian processes that relatives play in health. melanopsin controls are sensitive to vitamin A deprivation. In recent years one of these processes, the contraction of the pupil, known as the pupillary threshold, has been exploited in assessment of vitamin A status (see Chapter 7, p. 72).

VITAMIN A FUNCTIONS UNRELATED TO LIGHT

The involvement of light has been absolutely essential for the widely differing functions of vitamin A and its relatives described above, but for most functions of vitamin A derivatives throughout the mammalian body, all-trans retinoic acid (RA), light plays no part at all. 27

3 Methods of Analysis

ANALYTICAL PROCEDURES infections. The need for this has only been recognized in recent years, and is discussed later in Vitamin A has several physical properties several sections (see especially Chapter 7 and the which in the past have been utilized in its ana- acute phase response). lysis. These include a characteristic ultraviolet In recent years, notable advances have been (UV) absorption with an absorption maximum made in the application of stable isotope tech- of 325 nm and greenish fluorescence at 470 nm niques for the assessment of vitamin A status when excited at 325 nm. Recently, a portable (Ribaya-Mercado, Solomons 2003). A principle fluorometer (iCheck™) for measuring vitamin A similar to that of Fick’s, known for decades in has become available using these properties. A experimental physiology, uses measurement by blue chromophore, which is unfortunately tran- mass spectrometry of deuterium-labeled vitamin A sient, forms on exposure to certain Lewis acids, (non-radioactive) which is swallowed and al- such as antimony trichloride (Carr-Price reaction) lowed to equilibrate throughout the body before and trifluoroacetic acid (Neeld-Pearson reac- being measured again. The technology involved tion). Carotenoids also have characteristic UV is highly complicated. When a subject’s vitamin A absorption spectra. Nowadays, vitamin A and status is low, less dilution of an oral dose of carotenoids are most generally measured by high labeled vitamin A by endogenous vitamin A oc- performance liquid chromatography (HPLC). curs. Under these circumstances, a relatively high Straight-phase HPLC is most suitable for sepa- ratio of labeled to nonlabeled retinol is found in rating cis- and trans-isomers, and reverse-phase the serum. Application of this technique in hu- HPLC best separates compounds of different man populations is discussed later (Chapter 7). polarity, such as carotenoids and xanthophylls. To Using isotopes to determine bioavailability or prevent oxidation and polymerization, samples bioconversion has been critically reviewed by should be immediately analyzed or stored frozen van Lieshout, West, and van Breeman (2003) and in the dark at -70°C (Furr, Barua, Olson 1992). Yeum and Russell (2002). Recently there has been increased interest in Double stable isotope tests are being used the use of plasma retinol binding protein (RBP) to estimate the bioavailability of β-carotene and as a surrogate for plasma retinol. It has been mea- other carotenoids (see Chapter 5). sured by radial immunodiffusion (Almekinder, Using the very sensitive technique of accel- Manda, Kumwenda et al. 1999) and in a rapid erator mass spectrometry (Burri, Clifford 2005) field test using dried blood spots by fluorometry it has been possible to feed very small doses of (Craft 1999), and also by HPLC (Craft, Haitema, C-14-labeled β-carotene to volunteers and to Brindle et al. 2000). study its metabolism. Interesting variations are The use of dried blood spots after collecting being found in healthy subjects, and the implica- small amounts of blood (about 15-20 μL whole tions of this are discussed later (see Chapter 5). blood) has recently been perfected (Erhardt 2005). A standard sandwich ELISA technique can be used for measuring retinol binding protein BIOASSAY PROCEDURES (RBP) for vitamin A status, and also at the same time C-reactive protein (CRP) and α-1 glyco- In the past, biological tests played an impor- protein (AGP) as indicators of acute and chronic tant part in assessing vitamin A activity. These 28 3 Methods of Analysis

include the classical growth response tests in Broadly speaking, there are two main methods vitamin A deficient rats, liver storage assays in for estimating carotenoid content of foodstuffs rats and chicks, and the vaginal smear technique. – open-column method and HPLC (Rodriguez- Cell culture systems have been introduced for Amaya 1999). The open-column method is sim- assessing the biological activity of retinoids and pler and much less costly, and quite satisfactory carotenoids (Bertram 2004; Bertram, Vine 2005). for the main four or five carotenoids in a sample. More recently, cell culture models, using One specific technique for determination of CaCo-2 cells, have been established for studying carotenoids in tissues is of particular interest. intestinal uptake and metabolism of carotenoids Raman spectroscopy has been applied to the (During, Hussain, Morel et al. 2002). measurement of carotenoids in the macula of the eye in relation to diagnosis and treatment of age- related macular degeneration (see Chapter 13) CAROTENOID ANALYSIS IN FOODS and also in the skin (Gellermann, Zidichouski, Smidt et al. 2005). The technology of carotenoid analysis is be- For retinoid research, cell reporter assays or coming increasingly complex. This is not the ap- transgenic reporter animals are increasingly used propriate place to enter into detail, but there are (Gundersen 2006). authoritative sources for those who need them (Yeum, Booth, Sadowski et al. 1996; Pfander 2002; Britton, Liaaen-Jensen, Pfander 2004). 29

4 Food Sources

INTRODUCTION There is no single standard ideal diet for the whole of mankind; all we know is that some are As far as it is known, no species has been un- more healthy than others, while some are frankly able to survive and reproduce due to their natural nutritionally inadequate. The surprising thing is habitat being so improverished of the required that early man had the abilities to find edible wild- nutrients. Changes in climate, overpopulation, growing vegetables and fruits that are not fatally predation, or any of several inimical circum- toxic and eventually to cultivate some of these stances have developed later to bring many and to domesticate and cross-breed some grasses flourishing species to an end. Inadequate dietary as cereals and some animals, with no understand- intake of nutrients, rather than of overall food, has ing whatsoever of the science of nutrition. rarely been the cause. Man has been no exception The origins of both non-human and human to this. nutritional deficiencies are buried deep in evo- All living things are required to consume in lutionary time. The fact remains that man is re- some way an adequate amount of the elements quired to consume adequate amounts of more and molecules of which their own bodies are nutrients, numbering about 40, than any other spe- made. That this process usually takes place in a cies. Almost all of these are required in only very balanced manner is nothing short of a marvel of small amounts. They are the micronutrients; the nature. We have no idea how organisms select vitamins and essential elements. Most classes of their diet, either quantitatively or qualitatively, foodstuffs (cereals, vegetables, fruits, lentils, nuts for health. and fish, meat and dairy products) contain some It was made clear in Chapter 2 that through micronutrients, but some are devoid of or have the processes of photosynthesis and the absorp- very low concentrations of them. In addition, the tion of water and nutrients from the soil and at- amount of a foodstuff consumed is important. For mosphere, all the members of the plant kingdom instance, for most people in the West, bread and from trees, bushes and flowers to moulds and potatoes, which have very low concentrations of bacteria, are maintained. Examples are known, thiamine and ascorbic acid respectively, are the for instance, of inadequate mineral content of major source of each vitamin because they are soils and of resultant plant nutritional deficiency consumed in large amounts. disease. Most plant and animal species seem to In the present context, the sources in food have evolved successfully by consuming a rather of provitamin A carotenoids and preformed vita- uniform diet. min A are of paramount importance. It should be However, in the case of humans, as we have recognized that the underlying single major rea- seen (Chapter 1), nutritional deficiencies of one son for the existence of the problem of VADD, kind and another have developed through subsis- and consequently the impetus behind the writing tence on the very varied dietaries that have been of this book, is the fundamental and widespread consumed by mankind over the span of recorded deficiency in the basic diet in sources of vita- time. In recent times, along with everything else, min A in certain parts of the world and under cer- food has become increasingly globalized. In hard tain circumstances. Vitamin A itself has been economic times this brings its dangers (see Pref- known for almost 100 years (see Chapter 1). ace). Knowledge of how to cure the deficiency almost 30 4 Food Sources

as long, yet VADD in various forms (Chapter 12) be of very great significance. Firstly, in Chapter persists, as do the defective diets that cause them. 14 the consumption of some of the non-provi- There are several fascinating questions that tamin A carotenoids will be shown to be related still remain unanswered regarding why humans, both to the prevention and/or treatment of some especially young children, have proved to be so of the common chronic or degenerative diseases susceptible to becoming vitamin A deficient. For in general medicine. As an example, analysis instance, the liver stores much of the vitamin A of the third NHANES data in the United States ingested, and the level in the blood usually rises has revealed interesting population determinants steadily with increasing age, whereas most other of serum lycopene concentrations (Ganji, Kafai vitamins are readily excreted. The normal con- 2005). All told there are more than 40 carotenoids centration of provitamin A carotenoids, espe- present in the human diet, of which 25 are known cially in different vegetables and fruits, varies to be absorbed, metabolized, or utilized by the enormously. Substituting one for another in the body. The formulae and activities of these me- diet might prove disastrous. In particular rice, the tabolites are under active investigation (Khachik staple food of a large proportion of the world’s 2005). Secondly, in Chapter 12 on Global oc- population, has no provitamin A carotenoid currence of VADD, in the section on Vitamin A activity in any of its hundreds of different varie- in global food supply, Figure 12.5 shows, by ties in nature. Only through genetic manipulation region of the world, the daily per capita supply has so-called ‘Golden Rice’ been produced (see of vitamin A and the proportion of the total from Chapter 15). provitamin A. There are important differences For decades the scientific community re- between regions which will be commented on in mained in over-optimistic ignorance of the true Chapter 12 and also in relation to VAS in new- difference in relative vitamin A activity between borns in Chapter 15. preformed vitamin A and provitamin A carotenoids (see later in this Chapter). Consideration of the impact of this scientific knowledge on human health is delayed until Chapter 12. Vita- UNITS OF VITAMIN A ACTIVITY min A activity in the regional and world food supplies (see Figure 12.1) is then dealt with as a Earlier analytical methods failed to distin- fitting background for the global occurrence of guish between individual carotenoids, and as a re- VADD. sult non-provitamin carotenoids were frequently Human communities rely on a very wide included along with those with vitamin A activity range of plant and animal foods to meet their in concentrations reported in food composition dietary requirements for vitamin A. Ovo-lacto tables for fruits and vegetables. For example, one vegetarians and those who eat no foods of ani- analysis showed that as little as 7% of the caro- mal origin (vegans) can usually obtain all their tene value estimated was actually β-carotene or nutritional needs from plant sources alone, with provitamin A. The introduction of HPLC paved the exception of vitamin B12. Animal products are the way for the solution to this problem, but it usually expensive and are rarely relied on almost has taken time for accurate analyses to be made exclusively to meet requirements. From what is and for the results to replace the older, inaccu- known of the carotenoid composition of common rate values in food composition tables. The 1998 fruits and vegetables and surveys on carotenoid USDA–NCC Carotenoid Database for US Foods intakes, the intake of non-provitamin carotenoids (see Appendix) and McCance and Widdowson is greater than that of provitamin carotenoids. (Paul, Southgate eds. 1978) can be used as reli- There are several parts of the rest of this book able sources for values of individual carotenoid where dietary carotenoid intake will be shown to contents of foods. 4 Food Sources 31

Another problem that has to be addressed are good sources. This is because the carotenoid with regard to the vitamin A activity of foodstuffs content in chloroplasts is roughly proportional to is the need to make allowances for the differences the concentration of photosynthesis-associated in activity between retinol itself and β-carotene, chlorophyll (see Chapter 2). Edible dark green and also the need to differentiate between the ac- leaves are readily available in most areas where tivity of β-carotene and other provitamin carot- VADD are a problem. The species vary consid- enoids. The reasons for these differences will be erably from place to place and Table 4.2 shows considered later (see Chapter 5). The generally just a few examples of typical provitamin A val- accepted position on the relationships between ues for some that are commonly consumed and different expressions of vitamin A activity at the have proved useful in intervention programs. present time is as shown in Table 4.1. More exhaustive lists of similar data on a national or regional basis are becoming increasingly avail- Table 4.1: Retinol activity equivalent (RAE). able. 1 μg RAE = 1 μg retinol Table 4.2: Examples of common vegetable/fruit 2 μg β-carotene in oil carotenoid sources. 12 μg β-carotene in mixed μg RAE/100g foods edible portion 24 μg other provitamin A Mango (golden) 307 carotenoids in mixed foods Papaya (solo) 124 Cucurbita (mature pulp) 862 The rationale for the introduction of the new Buriti palm (pulp) 3,000 unit, retinol activity equivalent (RAE), is dis- Red palm oil 30,000 cussed further in Chapter 5. In practice, it means Carrot 2,000 that provitamin A activities of mixed foods have Dark green leafy vegetables 685 been halved. Some researchers maintain that for dark green leafy vegetables the reduction should Tomato 100 be even greater (West, Eilander, van Lieshout Apricot 250 2002). Sweet potato, red and yellow 670 The values for provitamin A activity are approximations because the bioavailability and bioconversion of carotenoids may be influenced Vegetables by many different factors. The doubt that has recently been cast on the efficacy of dark green Current analyses tend to report the content of leaves and some other sources of provitamin A each carotenoid identified, irrespective of pres- has stimulated much research activity in this field ence or absence of provitamin activity. This has (see Chapter 5). the additional value of indicating the value of the foodstuff as a source of antioxidant activity. Table 4.3 shows the lutein and β-carotene con- PROVITAMIN A CAROTENOID SOURCES centrations in green vegetables. β-carotene and lutein are the major carote- Dark green leafy vegetables, yellow fruits, noids in leafy vegetables, and together account for orange roots – mainly carrots – and the oils of over 80% of all carotenoids. α- and γ-carotene, palms are the main sources of provitamin A. cryptoxanthin and lycopene occur in lower con- Among leaves, only those that are dark green centrations. 32 4 Food Sources

Table 4.3: Lutein and β-carotene concentrations in green vegetables (μg/100 g fresh weight) (Ong, Tee 1992). Vegetable Lutein β-Carotene Green, leafy* (4 types) – 330–5,030 Green, nonleafy (6 types) – 217–763 “Cruciferous” vegetables (5 types) 280–34,200 80–14,600 L0eafy vegetables (32 types) – 1,000–44,400 Tuberous vegetables and beans (16 types) – 40–1,700 Green, leafy* (7 types) 250–10,200 1,000–5,600 Other vegetables (19 types) trace–440 11–430 Green, leafy* (27 types) 73–29,900 97–13,600 Green, nonleafy (8 types) 142–460 74–569 * Values from different references

Table 4.4: α- and β-carotene concentrations in carrots (μg/100 g fresh weight) (Ong, Tee 1992). Carrot α-Carotene β-Carotene Raw 2,000–5,000 4,600–12,500 Canned 3,200–4,800 7,000–11,000 Frozen 8,400–8,800 26,000–28,100

Raw 3,790 7,600

Raw, A+ hybrid 10,650 18,350 Freshly cooked, A+ hybrid 15,000 25,650 Canned 2,800 4,760

Line B6273 Lyophilized 3,400 6,000 Raw 3,200 5,200 Frozen 3,100 5,100 Line B9692 Lyophilized 6,100 13,800 Raw 6,600 11,700 Frozen 6,600 11,600 HCM line Lyophilized 20,300 28,200 Raw 20,600 25,100 Frozen 20,400 25,500 Raw 19 cultivars 2,200–4,900 4,600–10,300 Raw 3,410 6,770

Carrots are increasingly being grown in parts teristic feature is the relatively high concentra- of the developing world and may vary consider- tion of α-carotene, usually about half of that of ably in carotenoid content (Table 4.4). A charac- β-carotene. 4 Food Sources 33

In tomatoes, lycopene usually far exceeds central Brazil is almost as rich in provitamin A the concentration of β-carotene, although some carotenoids as is red palm oil (see below). varieties rich in β-carotene have been developed. Currently, quite extensive analysis is under- Tomato and tomato products are by far the best way in search of antioxidant and phytonutrient source of lycopene, which is of considerable properties among fruits that are less well-known interest in relation to some prevalent diseases in the West, but commonly consumed for puta- worldwide (see Chapter 14). The main carot- tive health reasons by more traditional societies enoids in pumpkin are α-carotene and β-carotene. (Burke, Smidt, Vuong 2005).

Fruits Roots and tubers

The vitamin A activity of fruits is generally Fewer analyses have been carried out on lower than that of leafy vegetables and their ca- roots and tubers than on vegetables or fruits, and rotenoid content is more complex. Their greater most of the varieties examined have had low con- acceptability, especially to young children, is an tents (see Table 4.6). However, the higher values advantage as far as intervention programs are for some pigmented varieties suggest that plant concerned (see Table 4.5). The “sweet paste” of breeding for these might be of importance. the buriti (Mauritia vinifera) fruit in north and

Table 4.5: Carotenoid concentrations in fruits (µg/100 g fresh weight) (Ong, Tee 1992). Fruit Lutein Cryptoxanthin Lycopene α-Carotene β-Carotene Banana 20–40 0 0 60–160 40–100 Berries, grapes, black 20–200 0 0 0–60 6–150 currant Mango – – – – 63–615 Orange, mandarine 20–30 7–300 – 20 25–80 Papaya, watermelon 0 450–1,500 2,000–5,300 0 228–324 Starfruit 60 1,070 0 0 28

Table 4.6: Carotenoid concentrations in selected roots and tubers (µg/100 g fresh weight) (Ong, Tee 1992). Root or tuber Lutein Cryptoxanthin Lycopene β-Carotene Sweet potato Different varieties – 0 – 5–551 – – – 1–4 Yellow variety 25 0 42 19 Orange variety 7 27 147 1,140 Cassava White variety 2 3 1 20 Yellowish – – – 40–790 Potato 13–60 Trace Trace 3–40 Taro 3–31 1 1–3 2–16 34 4 Food Sources

Vegetable oils appear more attractive. These have included extracts from leaves, carrots, tomatoes, algae and Carotenoids are present in most oils that are red palm oil. β-carotene was the first synthetic consumed but usually in low concentration. Red carotenoid to be used as a food color, others are palm oil (Elaeis sp. see Table 4.7) contains the β-apo-8’-carotenal and canthaxanthin. The for- highest carotenoid concentration in the vegetable mer two have vitamin A activity and therefore kingdom and has been extensively studied, for its also contribute to nutrient intake. unsaturated fatty acid content as much as for its high β-carotene content. It is also a rich source of other dietary substances such as vitamin E, PREFORMED VITAMIN A ubiquinones and phytosterols. Different species have different concentrations, as do different oil Mention has already been made of fish liver extracts. It is the main cooking oil in most regions oils as highly concentrated sources of vitamin A. of west and central Africa, but the tree is also They are used as pharmaceuticals rather than as cultivated in parts of Asia and of South America. diet items. Fish liver is often discarded along with Red palm oil used in cooking imparts a distinc- other soft organs, but if utilized could form a valu- tive taste to food that may not be readily accepted able source of the vitamin in many parts of the by some people. Highly refined red palm oil is developing world. The storage form is vitamin A1 increasingly being made available and is much alcohol (retinol) in saltwater fish. In freshwater more readily acceptable than the crude form. fish it is vitamin A2 alcohol (3-dehydroretinol), Many authorities see a major role for this oil in which has about 40% of the activity of retinol. the control of VADD in the future (Kritchevsky Because of the teratogenic effects of large doses 2000; Nagendran, Unnithan, Choo et al. 2000; of vitamin A, the consumption of liver is contra- Rao 2000 and Radhika, Bhaskaram, Balakrishna indicated during pregnancy (Ministry of Agricul- et al. 2003; see also Chapter 15). ture, Fisheries and Food, UK 1995). Most mammalian livers consumed, such as those of calves, Table 4.7: Carotenoid contents of palm oil extracts ox, lambs or chicken, have concentrations that are (mg/kg) (Ong. Tee 1992). Crude palm oil 630–700 Table 4.8: Examples of common animal vitamin A Crude palm olein 680–760 sources (μg retinol/100 g edible portion). Crude palm stearin 380–540 Fatty fish liver oils Second pressed oil 1,800–2,400 Halibut 900,000 Residual oil from fiber 4,000–6,000 Cod 18,000 Shark 180,000 Herring and mackerel 50 Other sources Dairy produce Butter 830 Hens’ eggs are often considered to be a rich Margarine, vitaminized 900 source of provitamin A because of the rich color, Eggs 140 but the major pigments are lutein and zeaxanthin, Milk 40 and β-carotene makes up less than 7% of the total. Cheese, fatty type 320 Some fish flesh is brightly colored, but most of Meats the pigment is xanthophyll. Natural extracts containing carotenoids have Liver of sheep and ox 15,000 long been used for coloring foods to make them Beef, mutton, pork 0–4 4 Food Sources 35

comparable to those in cod liver oil. Milk, butter, cheese and eggs are all moderate sources. Table 4.8 gives a selection of preformed vitamin A concentrations of some common foods. Over the years vitamin A has been added successfully to many foods. What is termed food fortification (or occasionally nutrification) is one of the major long-term approaches to the control of the problem of VADD, and is discussed further in Chapter 15. 36 37

5 Bioavailability of Carotenoids

INTRODUCTION ciency could lead to a significant increase in the risk of mortality and morbidity in young children. Major progress has been made in recent years It gradually became clear that while the con- in the research of bioavailability of carotenoids. sumption of dark green leaves and yellow fruits Largely because of the important implications could protect against severe VAD leading to xe- the subject has for measures aimed at controlling rophthalmia, it might not be capable of preventing VADD (see Chapter 15) and the increasing atten- subclinical vitamin A deficiency in a substantial tion given to it by research groups, it is logical to proportion of the young child and maternal popu- devote a separate chapter to it. lations of developing countries. Although it is all too easy to pass judgement The group at the Department of Human Nu- with the benefit of hindsight, it does nevertheless trition at Wageningen, Netherlands, was at the seem quite astounding that decisions that were forefront of the investigations into the bioavail- taken on little or no evidence in the mid 1960s ability of dietary carotenoids, non-provitamin A were allowed to remain unchallenged and formed and provitamin A. They also led the way in the the basis for action for more than 30 years. The study of the many complex factors that determine Joint FAO/WHO Expert Group (1976) introduced bioavailability. the term Retinol Equivalent (RE), and on the very De Pee, West, Muhilal et al. (1995) in In- scanty evidence available at that time assigned donesia found that an additional daily portion β-carotene one sixth and other provitamin A of dark green leafy vegetables given to lactat- carotenoids one twelfth the value of preformed ing women did not improve vitamin A status (as vitamin A (Hume, Krebs 1949). The concept has measured by serum retinol, breast milk retinol, proved to be of immense help in assessing the and serum β-carotene). In those who received a equivalence of different sources of vitamin A similar amount of isolated β-carotene on a wafer activity. The actual values remained untested, there was significant improvement. A review of despite a strong call from the committee for re- previous studies (de Pee, West 1996) found that search, and led to a false sense of security until many studies with positive results had design challenged in the mid 1990s. faults and that more recent, better controlled stud- In fairness, it should be recalled that in the ies tended to have negative results. Later work 1960s and long thereafter attention in this field showed that in central Java vegetables played a was focused on xerophthalmia and its prevention. lesser part in maintaining vitamin A status than Intermittent large-dose supplementation was the had been thought (de Pee, Bloem, Gorstein et al. prevailing control measure for this emergency 1998), and orange fruits were more effective in situation. The growing and consumption of dark improving vitamin A status than dark green leafy green leafy vegetables (DGLV) was encouraged vegetables (de Pee, West, Permaesih et al. 1998). as a long-term solution, but it received little at- The calculated β-carotene to retinol equivalence tention at international meetings. Most sources in leafy vegetables and carrots was 26:1 (compare of food composition data at this time did not dis- next page). A study in Guatemala, where 50 g of tinguish between provitamin and non-provitamin cooked carrots were added to the diet of children carotenoids. aged 7–12 years, failed to find evidence of nutri- In the mid 1980s evidence began to accumu- tional benefit (Bulux, Quan de Serrano, Giuliano late that widespread subclinical vitamin A defi- et al. 1994). A study in Ghana (Takyi 1999) found 38 5 Bioavailability of Carotenoids

that consumption of DGLV with added fat signifi- labeled retinyl ester/β-carotene concentrations in cantly increased serum retinol levels in children, the absorptive phase. While there was less cleav- but even so just over 50% still had inadequate age of β-carotene during vitamin A supplemen- vitamin A status. tation, higher absorption resulted in larger molar vitamin A values. A new phenomenon has been revealed as RECENT ADVANCES IN TECHNOLOGY a result of the introduction of isotope dilution techniques. It has become apparent that healthy The methodology for the accurate determi- people have an extremely wide range of ability nation of provitamin A carotenoid bioactivity in to convert β-carotene into vitamin A. For exam- man is still in an early and rapidly growing stage ple, 15 healthy adult Chinese were studied for 55 of its development. There are two main approach- days after receiving on day one isotope-labeled es; techniques that rely on changes in one or more β-carotene and retinyl acetate (Wang, Yin, Zhao indicators of vitamin A status, and those that pro- et al. 2004). Plasma samples were collected at fre- vide a more direct estimate of absorption and/or quent intervals. It was found that four of the sub- conversion efficiency of single doses of provita- jects had conversion factors of >29:1; they were min A carotenoids (Parker 2000; Yeum, Russell labeled “poor converters.” Eleven were “normal 2002; van Lieshout, West, van Breemen 2003). It converters” with conversion factors of 2.0:1 to seems unlikely that precise values will be set, as 12.2:1 (mean 4.8). The post-intestinal absorption in the past, but it should be possible eventually to conversion was estimated to be about 30% of the provide reliable, practical information for appli- total converted retinol. cation in interventions for the control of VADD. Similar results were obtained by Hickenbot- There is evidently a great deal more painstaking tom, Follett, Lin et al. (2002). Out of 11 healthy research that needs to be done (Traber 2000). men only six had sufficient measurable plasma

concentrations of D6 β-carotene and D3 retinol. Burri (2001) has reviewed the subject and pointed ISOTOPE DILUTION TECHNIQUES out its important implications for the control of VADD in developing countries. The possible rea- Equimolar doses of labeled β-carotene and sons for these interesting results are not fully es- retinyl acetate were administered orally to a single tablished. Variable intestinal absorption, variable healthy subject (Hickenbottom, Lemke, Dueker conversion to vitamin A, or accelerated clear- 2002). The results suggested that 8.5 mol of ance from the blood are possibilities. Recently, β-carotene were equivalent to 1 mol of preformed it has been demonstrated that the β-carotene low vitamin A. On a mass basis 15.5 μg of β-carotene responder phenotype could be caused by ge- was equivalent to 1 mg retinol. β-carotene was netic polymorphisms in the β-carotene 15,15’- found to be 55% absorbed. This gave precise con- monooxygenase gene (Leung, Hessel, Meplan et firmation of the low bioefficacy of β-carotene. al. 2009). Two common non-synonymous SNPs Vitamin A itself increases the absorption of exist in the human carotenoid 15,15̛-monooxy- β-carotene (Lemke, Dueker, Follett et al. 2003). genase (CMO1) gene that occur at high frequen- One nmol of [14C]-β-carotene was given to two cies and that alter β-carotene metabolism. Fe- human volunteers. Vitamin A supplementation male volunteers carrying either the 379V or the (10,000 IU/day) was given from day 53 and a combined 267S + 379V variant alleles showed a second dose of [14C]-β-carotene was adminis- reduced ability to convert β-carotene to retinol. tered on day 74. There were three main effects: The results provide a putative explanation for the 1) increased apparent absorption; 2) 10-fold re- poor responder phenotype in β-carotene biocon- duction in urinary excretion; 3) a lower ratio of version. 5 Bioavailability of Carotenoids 39

It has recently been shown in a healthy man vided an estimated vitamin A equivalence of that excentral cleavage of β-carotene can occur in 27:1 on average (see above). It should be vivo (Ho, de Moura, Kim et al. 2007). The fate of noted that to achieve this effect more than an oral dose of all-trans [10,10̓, 11,11̓- 14C]-β- 250 g of vegetables was consumed each day carotene (1.01 nmol; 100 nCi) and of its metabo- by these children, as young as 3 years old. An- lites was followed. This process had previously other study of the bioavailability of β-carotene been shown to occur in vitro and in animal mod- from oriental vegetables found higher bio- els. availabilities, but still much less than from Tang, Gu, Hu et al. (1999) in China studied purified β-carotene beadlets (Huang, Tang, Chen the effect of a two vegetable diet on vitamin A et al. 2000). body stores of children 3–7 years old by a double stable isotope method (see below). Vitamin A labeled with two different kinds of deuterium FACTORS AFFECTING THE BIO- (D4 and D8) was administered before and after AVAILABILITY OF CAROTENOIDS the intervention to estimate the occurrence of changes in total body vitamin A stores. One group In very general terms it has been possible to received their daily customary intake of 56 g rank different vegetable and fruit sources of car- of green-yellow vegetables and 224 g of light- otenoids, allowing for the influence of the food colored vegetables, and the other received matrix and food processing (Boileau, Moore, Erd- 238 g of green-yellow vegetables and 34 g of man Jr 1999) (Figure 5.1). light colored vegetables. Only the latter diet The Wageningen group has introduced a mne- maintained adequate vitamin A stores. It was monic (SLAMENGHI) to assist in the discussion calculated from the isotope studies that proof this complicated subject. It should be pointed vitamin A carotenoids (mainly β-carotene) pro- out that the question of bioavailability applies not

Very high bioavailability

Examples Food matrix Formulated carotenoids in water-dispersible beadlets Formulated natural or synthetic carotenoids Carotenoids-oil form Isolated natural or synthetic Carotenoids in oil

Papaya, peach, melon Fruits

Squash, yam, sweet potato Tubers

Tomato juice Processed juice with fat-containing meal

Carrots, peppers Mildly cooked yellow/orange vegetables

Tomato Raw juice without fat

Carrots, peppers Raw yellow/orange vegetables

Spinach Raw green leafy vegetables

Very low bioavailability (<10%)

Figure 5.1: Effect of food matrix and processing on bioavailability of carotenoids (after Boileau, Moore, Erdman Jr, 1999). 40 5 Bioavailability of Carotenoids

only to provitamin A carotenoids, but also to non- ‘M’: Matrix in which the carotenoid provitamin A carotenoids, such as lycopene, lutein is incorporated and others whose possible role in prevention or treatment of a number of diseases is discussed in In green leaves carotenoids exist within chlo- Chapter 14 (Chung, Rasmussen, Johnson 2004; roplasts as pigment-protein complexes which Boileau, Boileau, Erdman Jr 2002). require disruption of the cells for the carotenoid Before clarifying the mnemonic SLAMENGHI, to be released. In other vegetables and fruits it is first necessary to mention another term. Bio- carotenoids are sometimes found in lipid droplets conversion is not the same as bioavailability: it from which they may be readily released. Cook- relates to the efficiency with which the carot- ing assists in the release, but excessive heat may enoid in question is converted to vitamin A in the lead to oxidative destruction of the carotenoid. body. There is little information on this subject. The carbon-carbon double bonds of carotenoids The 1976 FAO/WHO Expert Group assumed that are subject to oxidation by oxygen in the air, and β-carotene has twice the efficiency of other provi- heat may bring about structural changes, includ- tamin A carotenoids. ing isomerization of all-trans carotenoids to cis forms.

‘S’: Species of carotenoids ‘E’: Effectors of absorption and The naturally occurring configuration of carot- bioconversion enoids in plants is usually the all-trans isomer. All-trans β-carotene is more readily absorbed in Dietary components influence absorption. A man than the 9-cis form. A significant propor- minimum amount of fat is required for adequate tion of 9-cis β-carotene is converted to the all- micelle formation. This appears to be sufficient trans form before entering the bloodstream. Cis- for optimal absorption of α-carotene, β-carotene lycopene, however, is preferentially absorbed and vitamin E, but lutein esters require higher compared with trans-lycopene (Boileau, Boileau, amounts of fat (Roodenburg, Leenen, van het Hof Erdman Jr. 2002). et al. 2000). Using stable-isotope-dilution methodology, Ribaya-Mercado, Maramag, Tengco et al. (2007) fed provitamin carotenoid-rich meals ‘L’: Molecular linkage with differing amounts (7, 15, or 29 g fat/day) to groups of more than thirty school children in Esters of carotenoids are common in fruit and the Philippines. Minimal fat intake enhanced the vegetables, but there has been little study of their vitamin A body pool size as much as did the high- absorption. Lutein esters and the free form have er fat levels. been shown to be equally bioavailable (Chung, Bioavailability of carotenoids is higher if Rasmussen, Johnson 2004). ingested with food than without, or with a low amount of fat (Dimitrov, Meyer, Ullrey et al. 1988). Adequate protein and zinc intake assists ‘A’: Amount of carotenoids consumed in the maintenance of vitamin A status. Vitamin E, in a meal as an antioxidant, protects vitamin A from being This has been studied in many human experi- oxidized. Fiber, chlorophyll and non-provitamin ments, and the consensus is that multiple dosing carotenoids, which are commonly present in the in a day is the most effective way to raise plasma diet, tend to reduce bioavailability. There is evi- levels. dence that alcohol ingestion interferes with the conversion of β-carotene to vitamin A. Dietary 5 Bioavailability of Carotenoids 41

intake of vitamin A itself has been shown to in- ‘I’: Mathematical interactions fluence absorption and retinol equivalence of β-carotene in humans (Lemke, Dueker, Follet et The final letter of the mnemonic refers to the al. 2003). possible synergistic or even antagonistic effect of two or more factors when acting together. At present this is only a theoretical concept, as data are ‘N’: Nutrient status of the host lacking, but it will have to be borne in mind as knowledge progresses. Absorption of carotenoids is influenced by In recent years, further contributions have vitamin A status. If status is low, conversion been made to our understanding of the complexi- of carotenoids to vitamin A is likely to be in- ties involved in the bioavailability of carotenoids. creased. There is evidence that zinc deficiency In particular, the important effects the food matrix impairs the efficiency of β-carotene conversion has on carotenoid absorption have been detailed to vitamin A. by Schwartz (2004). In summary, the type of matrix, the physical state of carotenoids, effects of processing, co-consumption of fat, carotenoid ‘G’: Genetic factors interactions, and structural considerations of the molecules all have profound effects on absorp- As clinical genetics advances it is likely that tion. there will be further reports of cases of rare genet- To date, one of the most revealing studies ic defects as the cause of vitamin A deficiency. To in this regard is that of Tang, Qin, Dolnikowski date, three types of defect are known: enzymatic et al. (2005). By feeding intrinsically deuterated failure to cleave β-carotene in the small intestinal spinach or carrots to volunteers, it was shown that mucosa (McLaren, Zekian 1971), heterozygotic the average provitamin A carotenoid to vitamin A reduction of plasma RBP (Matsuo, Matsuo, conversion efficiency is 14.8:1 by weight from Shiraga et al. 1988), and mutations in the gene carrot β-carotene, and 20.9:1 from spinach. These for retinol-binding protein (Biesalski, Frank, values are much less than those used previously, Beck et al. 1999); (see also Chapter 14). although significant amounts of vitamin A can be provided from these sources. The great importance of the nature of the matrix was confirmed, ‘H’: Host-related factors and the difference between the values of spinach and carrots is probably explained by the differ- The serum response to β-carotene is higher ence in the food matrix. β-carotene is contained in women than in men. In several ways, men are in spinach leaves in the form of pigment proteins more susceptible to VAD than women. Age does located in chloroplasts. In carrots, it is in the form not appear to be a factor. Diseases that interfere of crystals in chromoplasts and more readily re- with intestinal absorption, especially of lipids, leased in the digestive tract. are likely to impair carotenoid bioconversion. At the XXII International Vitamin A Consul- In terms of public health, especially in develop- tative Group (IVACG) Meeting, Wongsiriroj, Paik ing countries, intestinal parasites such as Ascaris and Blaner (2005) presented evidence for a new lumbricoides and Giardia lamblia are of great hypothesis to explain how tissues regulate vita- importance (see Chapter 13). min A formation in response to need. They have found that the presence of apo-cellular retinol- binding protein 1 (apo-CRBP-I) strongly stimulates the conversion of β-carotene to vitamin A. The presence of apo-CRBP-I in tissues such as 42 5 Bioavailability of Carotenoids

liver, lung, testis, eye and kidney indicates that focused on processing and other methods that the tissues have insufficient vitamin A. This has may considerably improve bioefficacy, the term the effect of stimulating the formation of vita- that is tending to replace bioavailability. min A to meet normal cellular and tissue functions. Thus tissues can use CRBP-I as a sensor to Table 5.1: Vitamin A conversion factors. assess vitamin A status. 1 μg retinol = 0.00349 μmol retinol 1.15 μg retinyl acetate OTHER TECHNIQUES 1.83 μg retinyl palmitate 3.33 IU (1 IU = 0.3 μg) In-vitro techniques are also being intro- 1 Retinol Activity Equivalent duced. A relationship has been shown to exist (RAE) between the β-carotene content of a diet and the 1 Retinol Equivalent β-carotene content of exfoliated colonic epithelial cells obtained from a subject consuming that diet (Gireesh, Nair, Sudhakaran 2004). A similar technique using the Caco-2 cell line from hu- In animal experiments (Deming, Baker, Erd- man intestine is being extensively applied (Gar- man 2002) it has been shown that the presence of rett, Failla, Sarama 1999; Liu, Glahn, Liu 2004). cis-isomers of β-carotene in a diet may affect the Simple in-vitro, rapid, cost-effective screening effectiveness of β-carotene as a vitamin A precur- techniques like these may contribute important sor. information in a field where much remains to be discovered.

CONVERSION FACTORS

In response to new studies of bioavailability and bioconversion of carotenoids the US Institute of Medicine made new recommendations (2001) (see also Chapter 4). The factors for β-carotene (1:12) and other provitamin A carotenoids (1:24) are both half of previous values. They are both equivalent to 1 μg of all-trans retinol, and all are equivalent to 1 μg of the new unit that has replaced the retinol equivalent (RE), called Retinol Activ- ity Equivalent (RAE) (Table 5.1). West, Eilander, and van Lieshout (2002) were of the view that these estimates might be overoptimistic. From their work, the bioefficacy for dark green leafy vegetables was about 1:26 or 1:28. For fruit the bioefficacy was 1:12. This suggests that, with a mixture of vegetables and fruits in a ratio of 4:1, typical for both developing and developed countries, the bioefficacy of β-carotene from a mixed diet would be only 1:21. Attention is now being 43

6 Vitamin A in Health

In recent years knowledge of the physiology, A proportion of provitamin A carotenoids biochemistry and molecular biology of vitamin A in the diet passes into the intestinal mucosal has advanced enormously. More is probably cells. A fraction of these, along with non-pro- known about vitamin A than about any other vita- vitamin A carotenoids, passes through unchanged min used by man. into the lymph and blood. The remainder of the provitamin A carotenoids undergoes cleavage with the formation of vitamin A. The key enzyme DIGESTION AND ABSORPTION in this process, carotenoid 15,15ʼ-monooxygenase (CMO1) was cloned independently by two Dietary preformed vitamin A and carotenoids research groups (von Lintig, Vogt 2000; and are released from protein in the stomach by Wyss, Wirtz, Woggon et al. 2001). This pro- proteolysis. They then aggregate with lipids and cess can also take place within the liver and pass into the upper part of the small intestine. some other tissues (Goodman, Huang, Shiratori Dietary fat and protein and their hydrolytic 1966). Symmetrical cleavage of the β-carotene products stimulate, through the secretion of the molecule yields two molecules of retinal, which hormone cholecystokinin, the secretion of bile. is mostly reduced and esterified to retinyl ester. This emulsifies lipids and promotes the formation Some cleavage is asymmetrical and less retinal of micelles which have lipophilic groups on their is produced (von Lintig, Voigt 2004; see Figure inside and hydrophilic groups on their outside. 6.1). In practice, β-carotene and other provita- This facilitates the absorption of fat. Bile salts min A carotenoids have only a fraction of the stimulate pancreatic lipase and other esterases activity of retinol, as was pointed out earlier (see that hydrolyze retinyl esters in intestinal mucosal Chapter 3). cells (enterocytes). Retinol, the product of the There is evidence from its structural and bio- hydrolysis, is well absorbed (70–90%) by intest- chemical properties that CMO1 belongs to an inal mucosa cells. ancient family of carotenoid-modifying enzymes

Figure 6.1: Carotenoid cleavage enzymes in E. coli. (Adapted from von Lintig and Vogt 2004). 44 6 Vitamin A in Health

first described in plants (Giuliano, Al-Babili, von TRANSPORT TO THE LIVER Lintig 2003). Within the mucosal cells retinol is esterified Chylomicrons consist of aggregates of thou- before incorporation into chylomicrons (see Fig- sands of molecules of triglycerides and phospho- ure 6.2). In this process a specific cellular retinol- lipids with fat-soluble vitamins, including retinyl binding protein (CRBPII) carries the lipid-soluble esters and carotenoids, cholesterol esters and retinol through the aqueous media and delivers some apolipoproteins (see Figure 6.3). it to the enzyme lecithin:retinol acyltransferase (LRAT) (Ong, Kakkad, MacDonald 1987). This They pass into the lymph and then the gen- enzyme appears to be the main intestinal enzyme eral circulation, and are broken down to some that normally esterifies retinol and then delivers it extent to produce chylomicron remnants. Almost to the chylomicrons. In mice lacking LRAT, only all retinyl esters remain with the chylomicron trace amounts of retinyl ester are stored in liver, remnants, which are mainly cleared by the liver. lung and kidney tissues, but adipose stores are Recent work has demonstrated that they also de- greatly elevated (O’Byrne, Wongsiriroj, Libien et liver retinyl esters to lung and some other tissues, al. 2005). and also to some cancer cells (Blomhoff 1994). In terms of intake of vitamin A in the diet, In these sites they are converted to retinoic acid, about 10% is not absorbed, 20% appears in the fe- which can be used for regulation of gene expres- ces through the bile, 17% is excreted in the urine, sion (see later).

3% is released as CO2 and 50% is stored in the liver (Olson 1994).

Figure 6.2: Absorption and metabolism of carotenoids. 6 Vitamin A in Health 45

transferred to the endoplasmic reticulum. There it binds to retinol-binding protein (RBP) and after the complex enters the Golgi complex it is secreted from the cell. In the absence of retinol, RBP is retained and degraded by the liver and tends to accumulate (Rask, Valtersson, Arundi et al. 1983). This ill-understood phenomenon forms the basis of relative dose-response tests (RDR) of vitamin A status. Most of the retinyl ester taken up by hepatocytes from chylomicron remnants is transferred as retinol attached to RBP to another type of liver cell called stellate cells (Wake 1994). Storage of vitamin A as retinyl esters seems to be one of the main functions of this type of liver cell. They also produce collagen type III, which when excessive may contribute to a form of cirrhosis of the liver (see also Chapter 13). 50–80% of vitamin A in the body is in the liver. 90–95% of this is in the stellate cells and 98% of this is in the form of retinyl esters, mostly as palmitate. This store is normally sufficient to last for several months. Bound to RBP, retinol is released from both stellate and parenchymal cells. When liver reserves of vitamin A are low parenchymal cells are the major site of storage (Batres, Olson 1987). Some parenchymal cells contain much more vitamin A than others. These are respectively termed “heavy” or “light” according to their density (Batres, Olson 1987a). Cells identical to the liver stellate cells also occur in many other tissues of the body, including the lung, intestine and kidney (Nagy, Holven, Roos et al. 1997). It has been shown that adipose tissue is an important storage site for retinol, as well as for β-carotene (Wei, Lai, Patel et al. 1997). Figure 6.3: Schematic drawing of chylomicron and chylomicron remnant (Blomhoff 1994). About 15–20% of the rat’s store of retinol is in adipocytes. This comes from chylomicron-bound esters, and not from circulating RBP as is the case METABOLISM IN THE LIVER of that in the liver. Furthermore, free retinol is secreted rather than retinol bound to RBP. The In the liver (see Figure 6.4), most vitamin A secretion of vitamin A from the liver is a com- from chylomicron remnants in the form of retinyl plicated process and by tracer kinetics in animal esters is taken up by hepatic parenchymal cells studies several pools have been identified (Green, (hepatocytes). There the esters are hydrolyzed Green 1996). Green and Green (2005) have very and after processing in endosomes retinol is recently reviewed this topic in relation to the 46 6 Vitamin A in Health

Figure 6.4: Schematic diagram of the liver structure (Blomhoff 1994).

assessment of vitamin A status in humans. Vit- Its three-dimensional structure is such that it con- amin A storage and action in the liver is the tains a specialized hydrophobic pocket within subject of a recent review by Wongsiriroj and which the fat-soluble retinol fits (Sivaprasadarao, Blaner (2004). Findlay 1994). In well-nourished adults total RBP concentration in plasma is 1.9–2.4 mmol/L (40–50 mg/ TRANSPORT BETWEEN CELLS mL). The value for children is about 60% of that of adults. Protein-energy malnutrition, infections Retinol in plasma bound to RBP is also al- and parasitic infestations lower concentrations most entirely associated with another protein, of RBP in plasma and in these cases caution transthyretin (TTR) (Ingenbleek, Young 1994). must be used in assessing vitamin A status (see The formation of this complex reduces the loss Chapter 7). of retinol in glomerular filtrate in the kidney. RBP RBP is a member of the lipocalin super- has been fully characterized and is a single poly- family, which includes proteins that bind to peptide chain with a molecular weight of 21,230. other lipid-soluble molecules (Table 6.1). RBP 6 Vitamin A in Health 47

Table 6.1: Properties and functions of some selected lipocalins (after Sivaprasadarao, Findlay 1994). Protein Size* Tissue Fluid Ligand Property/Function β-Lacto- 2 x 162 Mammary Milk Retinol (?) Possible involvement in transport globulin gland of vitamin A via gut receptor Serum retinol- 183 Liver and Serum Retinol, Retinol transport, delivery of binding protein other retinal, retinol to tissues via a receptor (RBP) tissues retinoic acid Complement 182 Liver Serum Retinol, Component of complement C8γ retinoic acid Apolipoprotein 169 Adrenal, Serum, Lecithin, Lipid transport, enhances D (ALPD) kidney, gut cholesterol lecithin:cholesterol pancreas, secretion acyltransferase activity liver * Number of amino acids

Table 6.2: Serum retinol concentrations (µmol/L) as a function of age and sex in American residents, 1971–74* (Olson 1996).

Age (years) Total Males Females 3–5 yr 1.28 (0.73–2.0) 1.29 (0.70–2.1) 1.26 (0.77–2.0)

n 1414 725 689 6–11 yr 1.31 (0.84–1.9) 1.30 (0.84–1.9) 1.32 (0.84–1.9)

n 1857 930 927 12–17 yr 1.58 (1.0–2.3) 1.62 (1.1–2.3) 1.53 (1.0–2.2)

n 2035 1026 1009 18–44 yr 1.94 (1.2–2.9) 2.08 (1.4–3.0) 1.80 (1.1–2.8)

n 7035 2164 4871 45–74 yr 2.20 (1.3–3.3) 2.29 (1.4–3.5) 2.11 (1.3–3.2)

n 6111 2911 3200 * The 5th and 95th percentile values are given in parentheses. (Derived from Pilch, 1987)

is synthesized in hepatocytes, but probably also shown in Table 6.2. In most age ranges, levels are in many other tissues, the knowledge of which is higher in males. rapidly growing. Retinol is extensively recycled between Mean values and ranges of serum retinol in plasma, liver and other tissues. At present it is healthy subjects as a function of age and sex are thought that most of the irreversible utilization 48 6 Vitamin A in Health

Table 6.3: Extracellular transport molecules for retinoids and carotenoids (Blomhoff 1994). Approximate Main ligand Suggested function MW (kD) Chylomicrons and 20,000–200,000 Retinyl esters Lymph and plasma transport of their remnants and carotenoids newly absorbed vitamin A and carotenoids VLDL and LDL 2,000–20,000 Carotenoids Plasma transport of carotenoids RBP 21 Retinol Blood plasma and intestinal fluid transport IRBP 140 Retinol, retinal Intercellular transport in visual cycle EBP 20 Retinoic acid Intercellular transport in epididymis

of retinol is a kind of detoxification, and only a in men (Olmedilla, Granado, Blanco et al. 1994), small fraction is functional. Small amounts of in contrast to retinol. other retinoids can be detected in plasma, including all-trans retinoic acid. There are several molecules for the extra- CELLULAR RETINOL-BINDING cellular transport of retinoids and carotenoids PROTEINS (CRBP) (LI, NORRIS 1996) (Table 6.3). It should be noted that there are no specific carrier proteins for carotenoids. Some retinol may enter cells by diffusion but the evidence is that most is taken up by membrane Concentrations of carotenoids in plasma (see RBP receptors which at the present time are being Figure 6.5) are highly dependent on diet. In the characterized (see Figure 6.6 and Table 6.4). In case of provitamin A carotenoids, normal levels bovine pigment epithelial cells a multitransmem- in plasma are significantly higher in women than brane protein, STRA6, has been identified as a

Figure 6.5: Typical carotenoid pattern in plasma. 6 Vitamin A in Health 49

Table 6.4: Intracellular retinoid-binding proteins (Blomhoff 1994). Approximate Main ligand Suggested function MW (kD) CRBP(I) 16 Retinol Donor for LRAT reaction and oxidative enzymes; regulates retinyl ester hydrolysis CRBP(II) 16 Retinol Donor for LRAT reaction CRABP(I) 16 Retinoic acid Regulates free retinoic acid concentration; donor for catabolizing enzymes CRABP(II) 15 Retinoic acid Regulates free retinoic acid concentration; donor for catabolizing enzymes CRALBP 36 Retinal Enzymatic reactions in the visual cycle

CRBP, cellular retinol-binding protein; CRABP, cellular retinoic acid-binding protein; CRALBP, cellular retinal-binding protein; LRAT, lecithin:retinol acyl transferase.

Figure 6.6: Hypothesis for cellular RBP metabolism. RBP-retinol may be recognized by a cell surface receptor. Retinol may be transferred to CRBPs either at the cell surface or after internalization into endosomes. CRBP may also deliver normal retinol to newly synthesized RBP in the endoplasmic reticulum (Blomhoff 1994). specific membrane receptor for RBP (Kawaguchi, enzymes (see Table 6.4). These proteins, like Yu, Honda et al. 2007). Recently, malformations RBP, also belong to a family with other members. in humans have been reported in association with In recent years mounting evidence has established mutations in STRA6. Carotenoids are mainly key functions for CRBP in retinoid metabolism. It carried in plasma by low-density lipoproteins acts as a kind of chaperone in protecting retinol (LDLs) and enter cells bearing the LDL receptor. within the cell and interacts with certain retinoid- Within cells there are many proteins that spe- metabolizing enzymes (Napoli 2000). cifically bind retinoids and direct them to specific 50 6 Vitamin A in Health

GENE FUNCTION AND RETINOIDS tive listing of vitamin A responsive genes thought to be activated through the action of RARs and/ It is known that the functions of vitamin A in or RXRs. different tissues, with the exception of the visual These retinoid hormone receptors are now process (see later), are mediated by its acid de- recognized to be part of a hormone receptor rivatives (Pfahl, Chytil 1996). All-trans retinoic superfamily whose members interact. It has acid (RA) is noncovalently bound to three nuclear been called the “Steroid-Thyroid-Retinoid Super- RA receptors (RARα, RARβ, RARγ) in gene family.” Table 6.6 gives a partial listing of these transcription regulation. When activated, the nu- nuclear receptors and their respective activat- clear receptors for vitamin A are able to bind to ing substances. If these nuclear receptors are to “response elements” in specific genes to increase recognize response elements within responsive or decrease the level of expression of the gene. genes they must act in pairs (dimers). A member The response elements are nucleotide sequences of the RXR family must serve as a partner if the which in the DNA compose the genes. dimer is to function. It is therefore evident that Soon afterwards it was shown that 9-cis RA vitamin A, usually in the form of RA, and the binds and activates three other nuclear receptors: RXRs can be regarded as regulators of several RXRα, RXRβ, RXRγ. RA only binds RAR re- hormone response systems. ceptors, whereas 9-cis RA can bind all six nuclear Major sources of RA in the diet are its precur- receptors, but from recent work it seems that it is sors: retinol and retinyl esters. The diets of ani- not the endogenous ligand for RXR (Wolf 2006). mals and man contain little RA itself. However, it The expression of these six receptors varies has been shown that both rat and human intestinal between cells, but most if not all cells express cytosol can produce RA from β-carotene (Napoli at least one of these receptors. Several hundred 1994). This suggests a new source for the genera- genes have so far been shown to be induced or re- tion of RA within the body. The concentration of pressed by retinoids. Table 6.5 gives a representa- RA in the body is much less than that of retinol;

Table 6.5: Partial and representative listing of vitamin A responsive genes thought to be regulated through the action of RARs and/or RXRs (Blaner 1998). Gene Gene function/role Oxytocin Reproduction Growth hormone Growth Phosphoenol pyruvate carboxykinase Gluconeogenesis Class I alcohol Alcohol oxidation dehydrogenase Transglutaminase Cell growth and cell death

Laminin B1 Cellular interactions Matrix Gla Protein Bone growth and strength Keratin Skin Cellular retinol-binding protein type I Vitamin A metabolism RARβ Vitamin A action Hox 1.6 Fetal development

Dopamine D2 receptor Central nervous system 6 Vitamin A in Health 51

Table 6.6: Partial listing of members of the “Steroid-Thyroid-Retinoid Superfamily” of nuclear receptors and their activators (Blaner 1998). Nuclear receptor Activating substance Retinoid X receptor (RXR) 9-cis retinoic acid Retinoic acid receptor (RAR) all-trans retinoic acid Vitamin D receptor (VDR) 1,25 dihydroxy vitamin D Thyroid hormone receptor (TR) Triiodo-thyronine (T3) Estrogen receptor (ER) Estrogen Progesterone receptor (PR) Progesterone Glucocorticoid receptor (GR) Cortisol Mineralocorticoid receptor (MR) Aldosterone Androgen receptor (AR) Testosterone Lipid X receptor (LXR) Cholesterol metabolite(s) Peroxisome proliferator-activated receptor (PPAR) Putative fatty acid metabolite

between 100 and 1000 times smaller. Even so, retinoids in cells and target tissues. CYP26 may it is RA that is responsible for the vast majority play a role in embryonic development (see below of vitamin A actions in the body. Most of the RA and Chapter 14) and in cancer (see Chapter 14). has to come from retinol and its esters by way of The great advances made in the understanding of oxidation to retinal. It is proposed that enzymes in the hormonal control retinoids exert over gene two families are of key importance in this process; function have highly significant implications for the SDR (short-chain dehydrogenase/reductase) the management and treatment of many diseases family, and the ADH (medium-chain alcohol de- that occur throughout the world (see Chapter 14). hydrogenase) family. The second and final step of retinal oxidation to RA is carried out by several retinal dehydrogenases (RALDH). This subject FUNCTIONS OF VITAMIN A has recently been reviewed (O’Byrne, Blaner 2005). In recent years knowledge of the functions of Cytochrome P450 (CYP) enzymes constitute vitamin A has increased greatly. Table 6.7 indi- a large superfamily with hundreds of members. cates the major functions of vitamin A as far as They catalyze the oxidative metabolism of many man is concerned. In view of the fact that nuclear endogenous compounds as well as numerous for- receptors for retinoids have been identified in vir- eign chemicals. It has been proposed that P450s tually every type of cell it could be argued that in regulate steady-state levels of ligands for the some way vitamin A plays a part in every bodily nuclear hormone receptor superfamily which in- process. For our present purposes we concentrate cludes retinoids (Sonneveld, van der Saag 1998). upon those included in Table 6.7. RA is metabolized through oxidation to more polar metabolites. Recently, a novel cytochrome P450 enzyme (CYP26) with specific RA 4-hy- Vision droxylase activity has been cloned from man and some animal species. It fulfils all the requirements In recent years, many details of the story of of an enzyme which could control levels of active the role of vitamin A and its metabolites in photo- 52 6 Vitamin A in Health

Table 6.7: An outline of functions of vitamin A. Vision Photopic and color, scotopic Cellular Gene transcription differentiation morphogenesis Immune response Non-specific, cell metabolism Hemopoiesis Iron metabolism Growth Skeletal Fertility Male and female Thyroid metabolism Regulation of TSH secretion

reception have been filled in. The main aspects of The latest scheme of the visual cycle is illus- the functions of these compounds in different as- trated in Figure 6.7 (Pepperburg, Crouch 2001). pects of vision have been described and illustrated This shows in outline a new, recently elucidated, in Chapter 2 and need not be repeated here. visual cycle by which a protein found in the RPE It has recently been shown (Wolf 2004) that called retinal G-protein-coupled receptor (RGR) the visual cycle in cones differs significantly can supplement during intense illumination the from that in rods. The main difference is in the 11-cis retinal required to regenerate rhodopsin. In mechanism of isomerization of all-trans-retinol addition, it has been shown that the protein RPE65 to 11-cis-retinol, and the latter’s oxidation to 11- performs a key role in the trans-cis isomerization cis-retinal. When light strikes the retina a series of of retinol. The palmitoylation of RPE65 switches complex biochemical changes takes place, result- off the visual cycle in darkness and switches it on ing in the generation of a nerve impulse. The ini- in the light (Wolf 2005). tial event in visual excitation is the isomerization The space between the RPE and the layer of 11-cis retinal to the all-trans form (see Figure of photoreceptor cells is taken up mostly by a 2.10). large glycoprotein called the interphotorecep- Vitamin A, in the form of all-trans retinol, is tor retinoid-binding protein (IRBP), the function delivered by the blood to the retinal pigment epi- of which is to transport two different molecular thelium (RPE) where it is either esterified for stor- species of vitamin A to and from two differ- age, or isomerized to 11-cis retinol, which is then ent cell types, as outlined above. Chen and oxidized to 11-cis retinal. This is transported to Noy (1994) have proposed mechanisms to ac- the photoreceptor rod or cone cells where it com- count for this process. Another retinoid-binding bines with the protein opsin to form rhodopsin or protein is also unique to the eye. It is cellular reti- iodopsin, respectively, which are light sensitive. naldehyde (retinal)-binding protein (CRALBP). It On exposure to light, 11-cis retinal is isomerized binds closely to 11-cis retinal and to 11-cis retinol, back to all-trans retinal and the nerve impulse is and protects the former from photoisomerization. generated. On release from the protein all-trans It has recently been shown that β-carotene retinal is reduced to all-trans retinol and is trans- may be converted to retinol in the RPE, thus pro- ported back to the RPE to complete the cycle. viding an alternative pathway of supply of retinol As in the liver, lecithin:retinol acyltransferase (Reddy, Nohr, Schaeffer et al. 2005). (LRAT) is essential for accumulation of all-trans- In the developing embryo RA is highly con- retinyl esters in the RPE (Batten, Imanishi, centrated in certain tissues (Dräger, Wagner, Maeda et al. 2004). McCaffery 1998). These patterns are generated 6 Vitamin A in Health 53

Figure 6.7: A scheme showing the newly described light-dependent, RGR-mediated isomerization of all-trans to 11-cis-retinal in the retinal pigment epithelium (dotted line). RGR = retinal G protein-coupled receptor. (Adapted from Pepperberg and Crouch 2001). by the local expression of RA-synthesizing al- vertebrate, photoreceptors. It might be speculated dehyde dehydrogenase enzymes. The eye, espe- that this may have accelerated the rapid evolution cially vulnerable to vitamin A deficiency, is one of RA-mediated transcriptional regulation, at of the RA-richest regions in the embryo. In the the transition from invertebrates to vertebrates. mature eye the pattern of these enzymes is stable, It may also explain the prominent role of RA but the amount of RA synthesized is variable and in the eye. Figure 6.8 shows a scheme of vita- is dependent on ambient light levels. This re- min A usage illustrating just how close the links sults from changing levels of the RA precursor between its role in vision and in transcriptional retinal, which is released from illuminated rho- control are. dopsin. Thus this is a mechanism whereby light can directly influence gene expression. This phenomenon of light-induced release of retinal from rhodopsin occurs only in vertebrate, but not in in- 54 6 Vitamin A in Health

Figure 6.8: A scheme of vitamin A usage in the eye illustrating the link between its roles in vision and in transcriptional control. As the regeneration of 11-cis-retinal takes place in the retinal pigment epithelium, some of the all-trans-retinal, released from rhodopsin, is converted to the transcriptional activator retinoic acid. (Adapted from Dräger UC, Wagner E, McCaffery et al. 1998).

Cellular differentiation Glycoprotein and glycosaminoglycan synthesis Retinoic acid can combine with the nuclear retinoic acid receptor leading to growth inhibi- Glycoproteins are polypeptides with short tion as in some tumors. RA can also bind to the chains of carbohydrates. Glycosaminoglycans orphan nuclear receptor peroxisome proliferator- are related compounds that are cell surface mol- activated receptor β/δ (PPAR-β/δ). This results ecules. Retinoids have been shown to control in stimulation of cell growth and inhibition of the expression of enzymes involved in the syn- apoptosis. To bind to RAR, RA is carried into thesis of some of these compounds (Vahlquist the nucleus by the cytosolic RA-BP-II. To bind 1994). Impairment of this function by VAD to PPAR-β/δ, it is transported into the nucleus by may contribute to lack of mucin secretion and the cytosolic fatty-acid-binding protein 5 (Wolf liquefaction of the cornea seen in xerophthalmia 2008). (see Chapter 8). In vitamin A deficiency keratin-producing cells replace mucus-secreting cells in many epithelial tissues of the body. This is the basis of the Embryogenesis pathological process termed xerosis that leads to the drying of the conjunctiva and cornea of the Severe vitamin A deficiency on the one hand, eye (see Chapter 8). The process can be reversed and excessive dosing with vitamin A and also RA by vitamin A. It has become clear recently that on the other, result in malformations of the em- vitamin A, mainly in the form of RA, plays a key, bryo affecting most organs of the body in many hormone-like role in cell differentiation through- vertebrate species. The embryo is only sensitive out the tissues and organs of the body (Figure to teratogenic influences, that is to say those that 6.9). induce congenital malformations, during a rela- 6 Vitamin A in Health 55

Figure 6.9: Suggested roles of CRBPs and CRABPs. See text for details (Blomhoff 1994). tively short period after conception. After that Recent animal work (Quadro, Hamberger, period, known as the organogenetic period, has Gottesman et al. 2005) has shown that retinol- passed, usually the first three months of fetal life RBP is the primary contributor to fetal develop- in man, there is no longer any risk in this regard ment, whereas retinyl esters are mainly respon- to the fetus (Gerster 1997; Wiegand, Hartmann, sible for accumulation of fetal retinoid stores. Hummler 1998; Ward, MacGowan, Hornby et al. An indication of the fundamental importance 2000). To date it has not been conclusively dem- of vitamin A, especially in the form of all-trans- onstrated that human VAD causes congenital mal- retinoic acid via specific nuclear transcription fac- formations. A number of individual case reports tors, the retinoid receptors, may be appreciated by have been published but the association might the following sampling of some of the recent ex- well be coincidental and not causal. Although perimental contributions in this field. some of the synthetic retinoids are clearly tera- Zile (2004) and colleagues have concentrat- togenic in man, there is very much less evidence ed on cardiovascular development in the early for a similar effect of large doses of retinol (see avian embryo during neurulation, correspond- Chapter 14). Some other aspects of this topic are ing to the first 2–3 weeks of human pregnancy. considered later in relation to the safe use of vita- Several developmental genes, linked to those min A (see Chapters 14 and 15). Many of the ef- events, are regulated by vitamin A. They include fects of vitamin A are now known to be mediated homeobox genes HoxB-1 and Msx-1; the cardio- by retinoid-regulated expression of homeobox genic transcription factors GATA-4 and GATA-6 genes and other pattern-forming genes. RA recep- and the heart asymmetry genes nodal, snail tors are closely involved (Ross 1999). and Pitx-2. RARα 2, RARγ and RXRα have 56 6 Vitamin A in Health

been identified as the retinoid receptors essen- ously considered unrelated, such as immunoglob- tial for transducing the RA signal during the criti- ulin production, are now known to be affected by cal RA-requiring developmental window. One retinoids (Semba 1998). It has been suggested study (LaRue, Argraves, Zile et al. 2004) demon- (Buck, Ritter, Dannecker et al. 1991) that a third strated that vascular abnormalities observed under oxidized metabolite of retinol, 14-hydroxy-4,14- conditions of retinol deficiency are due to a reduc- retro retinol is the active molecule in the immune tion in the number of angioblasts. This leads to system, but how it acts is not known. Recently it an insufficiency in the number of endothelial cells has been shown that retinoic acid regulates factors required to build complex vascular networks. known to be required for Ig class switch recombi- The addition of retinol in vitro to bovine em- nation and modulates the population dynamics of bryo culture media was shown to have a signifi- ligation-stimulated B cells (Chen, Ross 2005). cant positive effect on early embryonic develop- Mucida, Park, Kim et al. (2007) have shown ment (Lima, Oliveira, Goncalves et al. 2004). that RA is a key regulator of TGF-β-dependent The key role played by retinoids in lung de- immune responses, capable of inhibiting the velopment and regeneration (Maden 2004) gives IL-6-driven induction of pro-inflammatory TH17 hope that they may provide the first treatment for cells and promoting anti-inflammatory Treg cell diseases such as emphysema and chronic obstruc- differentiation. Thus RA is capable of regulating tive pulmonary disease. the balance between pro-inflammatory and anti- Excessive dietary intake of retinoic acid in- inflammatory immunity (seeTable 10.2). duces hindlimb and eye deformities in the frog Much of the evidence for the functioning of (Alsop, Brown, Van Der Kraak 2004). vitamin A in the immune response comes from During embryogenesis the bilateral symme- studies in both experimental animals and human try and orderly formation of somites is signaled subjects who are vitamin A deficient. This subject by retinoic acid (Kawakami, Raya, Raya et al. is pursued further in Chapters 9 and 10. 2005). Many other tissues and organs are under active study along similar lines; all provide increas- Thyroid function ing evidence for the universal importance of the presence of an appropriate concentration of Thyroid-stimulating hormone (TSH) of the RA in precise areas of the body and at critical anterior pituitary gland is regulated by the bind- times in embryogenesis for normal development ing of the thyroid hormone activated thyroid to occur. receptor to the TSH gene. At the same time, the binding of the 9-cis-retinoic acid activated retinoid X receptor to the same gene occurs. Both Immune response interactions, separately or simultaneously, can suppress and thus regulate the expression of the Many of the epithelial tissues are important TSH gene (Wolf 2002) barriers to infection and VAD impairs this function in a nonspecific way. In addition, vitamin A is known to be involved in maintaining immuno- Reproduction competence in more specific ways. It helps to maintain the lymphocyte pool. Vitamin A also It used to be thought that vitamin A in the functions in T-cell-mediated responses. RXRα form of retinoic acid could not prevent signs of has an essential role in the development of TH2 vitamin A deficiency in the organs of the repro- responses (Du, Tabeta, Mann et al. 2005). ductive system, as is the case in the eye. This area Some aspects of the immune response previ- remains under-researched, but certain aspects 6 Vitamin A in Health 57

have become clearer in recent years (Ross 2005). Growth In males, failure of gametogenesis, reduced testosterone production, and loss of fertility occur Several aspects of this subject have received in the presence of vitamin A deficiency. All three further attention in recent years. It is probable that main types of testicular cells, Sertoli, germinal there are many factors responsible for growth in and Leydig cells, undergo morphologic changes early life; it is likely that many have not been dis- in vitamin A deficiency. A recent study in mouse covered yet. In addition, the long-term and large- testis suggests that spermiation requires RXRβ/ scale studies that are necessary to provide definite RARα heterodimers in Sertoli cells, whereas answers have not yet been undertaken. In most spermatogonia proliferation involves, indepen- instances the aspect of growth that has been cho- dently of RXR, two distinct RAR-mediated sen for study has been height, rather than weight signaling pathways in both Sertoli cells and sper- or some other parameter. Height, compared with matogonia (Vernet, Dennefeld, Rochette-Egly et weight, has the advantage that it is short-term less al. 2006). readily influenced by such factors as infectious In females, vitamin A is required for normal diseases. On the other hand, it is subject to much estrus cycle and implantation of the blastocyst less variation so alterations are much more dif- after conception. The role of retinoic acid in regu- ficult to recognize. lating gene expression and morphogenesis in the Deficiency of certain nutrients, including vi- developing embryo was covered earlier in this tamin A, are likely to be partly responsible for chapter. the widespread occurrence throughout develop- In the human, there is evidence (see Chapter 9; ing countries of height retardation or “stunting”. West, Katz, Khatry et al. 1999) that suggests that Another important factor seems to be the effect vitamin A or β-carotene supplementation dur- of seasonal variation. This probably results from ing pregnancy has a beneficial effect on certain a combination of low nutrient intake and occur- aspects of gestation. The mechanism(s) is not rence of outbreaks of some infectious diseases at understood at present. the particular seasons when growth retardation is maximal. Several recent studies illustrate the difficulty and magnitude of the problem. Hemopoiesis Retinoic acid is known to play its hormone- like function in the control of growth and VAD in man and in experimental animals is development of tissues in the musculo-skeletal consistently associated with an iron deficiency system, just as it does elsewhere. In communi- type of anemia. It has been repeatedly shown ties subject to widespread VAD, the occurrence that in these circumstances, in addition to iron, of many other adverse factors has made the vitamin A is required for a full hematologic res- demonstration of the retarding effect of VAD ponse. The mechanism remains unclear. VAD difficult (Sommer, West 1996:163–188; see might interfere with the absorption, transport or Chapter 10). One possible mechanism for the storage of iron. On the other hand, it might act influence on growth is the recent demon- directly on hemopoiesis, although that seems stration that both vitamin A and retinoic acid less likely (Sommer, West 1996:150–162). The produce rapid release of cyclic AMP (adenosine demonstration that RA is necessary for erythro- monophosphate) and human growth hormone se- cyte differentiation suggests that it may control cretion (Djakoure, Guibourdeuche, Porquet et al. the usage of iron (Pfahl, Chytil 1996; see also 1996) Chapter 11). 58 6 Vitamin A in Health

Other functions Signal transduction Evidence for a cytoplasmic role for retinoids as an Energy balance effector in signal transduction has been reviewed Heat production (thermogenesis) by mitochon- recently (Hoyos, Hammerling 2005). Oxygen is dria in small collections of brown adipose tissue among a group of inorganic gases, including also plays a part in the regulation of energy balance ROS (reactive oxygen species), with intracellu- and consequently possibly in the control of obe- lar messenger properties. The role of retinol can sity. In mitochondria, an enzyme not present in be seen as a device for the efficient and accurate white adipose tissue controls the local produc- channeling of ROS to the zinc-finger domains of tion of energy as heat. This enzyme is regulated serine/threonine kinases involved in signal trans- by the sympathetic nervous system, but is under duction. transcriptional regulation by RA (Alvarez, De Andres, Yubero et al. 1995). Mitochondria and apoptosis Recent in-vitro studies have shown that various Regulation of the dopaminergic system synthetic retinoids appear to target mitochondria RA plays a major role in the development of to trigger apoptosis in transformed cells. This the fetal central nervous system. All-trans and may be important for their anticancer potential 9-cis RAs that are bound to their receptors (Hail, Lotan 2005). in the brain and the pituitary gland regulate the expression of the dopamine receptor D2R, which is part of the dopaminergic system. Do- Activities of carotenoids pamine is a signaling molecule in the central nervous system. It controls coordination of move- Olson (1999) has classified the many activities ment and also the synthesis of pituitary hormones. of carotenoids into functions, actions, or associa- Parkinson’s disease results from a defect in the tions, but these divisions are not strict and rigid. dopaminergic system. D2R-knockout mice lack In Table 6.8 an attempt is made to bring them these RA receptors and develop neurological together in outline form. The underlying mecha- damage similar to that in Parkinson’s disease nisms of some of these activities of carotenoids (Wolf 1998). have been discussed in Chapter 5. Functions usually denote a more or less clearly Gap junctional communication defined purpose. The conversion of provitamin A Gap junctions are narrow, hydrophilic pores con- carotenoids to vitamin A in the animal body is the necting the cytosol of two adjacent cells. There is most evident one. Others, as indicated in Table evidence that gap junctions play a role in regu- 6.8, are less well understood in chemical terms. lation of morphogenesis, cell differentiation, se- Actions, both biological and chemical, are read- cretion of hormones, and growth. Effects on gap ily demonstrated but their consequences are not junctional communication might be involved so clear. Associations of carotenoids are mostly in carcinogenesis and teratogenesis (Hix, Vine, known from epidemiological studies and usually Lockwood et al. 2005). RA and its analogues ex- have their importance in relation to occurrence of ert their effect by acting as ligands of nuclear re- diseases (see Chapter 14). ceptors RAR or RXR (see earlier and Stahl, Sies Evidence suggests that carotenoid cleavage 1998). It has also been shown that carotenoids and products may increase the risk of cancer devel- other micronutrients, as well as their metabolites opment by impairing mitochondrial function. and cleavage products, are also involved (Stahl, Depletion of mitochondrial sulfhydryl groups and Sies 2005). impaired oxidative phosphorylation may induce oxidative damage in DNA molecules and nuclei 6 Vitamin A in Health 59

Table 6.8: Functions and actions of carotenoids. Functions  Some carotenoids are the only known precursors of vitamin A and its derivatives  Accessory pigments in energy transfer in photosynthesis  Phototropism in simple and higher plant forms  Photoprotective role in bacteria and also man  Plant growth regulation  Reproduction regulation in fungi Actions  Quenching of singlet oxygen  Color attractants in flowers for insect pollination  Coloration of food for mankind

(Siems, Wiswedel, Alija et al. 2005). On the other sex, of different ages and in various physiological hand it has been shown (Zhao, Aldini, Johnson states, such as pregnancy and lactation, are kept et al. 2006) that carotenoid supplementation by under regular review. As mentioned earlier (see mouth, with a combination that includes lutein, Chapters 4 and 5) it is now customary to express β-carotene and lycopene, can exert protection the vitamin A activity of a diet in terms of retinol against DNA damage. activity equivalents (RAE) to take account of the Laboratory studies are contributing to our different activities in the body of retinol and the understanding of the antioxidant activity of carot- provitamin carotenoids which have recently been enoids in terms of quenching of singlet oxygen, altered. Table 6.9 gives criteria and dietary refer- as mentioned in Table 6.8. Very recently it has ence intake values for vitamin A for children and been shown that carotenoids normally present in adults. human serum are capable of inhibiting cellular Tables like these are designed to indicate proliferation (McDevitt, Tchao, Harrison et al. how deficiency may be prevented and how a 2005). safe intake for the large majority of a population The biological activity of carotenoid metabo- (covering about ±2 SDs from the mean) may be lites has recently been reviewed (Krinsky 2005). provided. They were originally meant to ensure It has to be appreciated that nutrients, such that, if met, deficiency would not occur, although as carotenoids, are by no means the only dietary a considerable proportion of a population might constituents that are capable of preventing cellu- consume more than they needed to. lar oxidative damage. Many studies have shown However, evidence has accumulated in recent that flavonoid components of fruit and vegetables, years for a protective effect against some chronic such as anthocyanins and flavonols, also have this diseases of increased amounts of antioxidant nu- ability (Prior 2003). trients, such as the carotenoids (see above and Chapter 14). These possible beneficial effects have not previously been taken into consideration HUMAN REQUIREMENTS when recommended dietary allowances (RDAs) have been formulated. The usual advice being Along with those of other nutrients, the re- given is to increase the intake of fresh fruits and quirements for vitamin A for humans of either vegetables. 60 6 Vitamin A in Health

Table 6.9: Criteria and dietary reference intake values for vitamin A by life-stage group. EAR* RDA* (μg RAE/d) (μg RAE/d) Life-stage group Criterion male female male female 0–6 mo Average vitamin A intake from human milk 7–12 mo Extrapolation from 0–6 mo AI * * 1–3 y Extrapolation from adult EAR 210 210 300 300 4–8 y Extrapolation from adult EAR 275 275 400 400 9–13 y Extrapolation from adult EAR 445 420 600 600 14–18 y Extrapolation from adult EAR 630 485 900 700 >18 y Adequate liver vitamin A stores 625 500 900 700 Pregnancy 14–18 y Adolescent female EAR plus estimated 530 750 daily accumulation by fetus 19–50 y Adult female EAR plus estimated daily 550 770 accumulation by fetus Lactation 14–18 y Adolescent female EAR plus average 885 1200 amount of vitamin A secreted in human milk 19–50 y Adolescent female EAR plus average 900 1300 amount of vitamin A secreted in human milk * Estimated average requirement (EAR) and recommended daily allowance (RDA) according to the Food and Nutrition Board (FNB) ** Adequate intakes (AI) according to FNB 61

7 Assessment of Vitamin A Status

INTRODUCTION and indirect assessment of liver reserve or body store are most commonly used in the case of vi- Nutritional status, or nutriture, is an impor- tamin A. tant concept that often tends to be misunderstood. The use of somatic measurements, such as It is sometimes thought to be represented by di- weight and height, is the mainstay in the detec- etary intake, but this is not the case. A financial tion of protein-energy malnutrition under field analogy may be helpful. A person may have an conditions. Although there is evidence that vita- income, probably from a number of sources, and min A deficiency does have an effect on growth also an expenditure on a variety of things. If in- (see Chapter 10), there are many other nutritional come exceeds expenditure then that person’s fi- factors and infections which act similarly. There nancial balance or status is positive (in the black) is no aspect of growth retardation which can be and if expenditure exceeds income the status is linked specifically to vitamin A deficiency. In negative (in the red). In this analogy “income” fact, to describe a community as “underweight” may be replaced by “nutrients in the diet” and or “stunted” contributes no understanding of the “expenditure” by “utilization of nutrients by the nature of the underlying causes. body.” Then the resulting nutritional status (bal- In the field of assessment of vitamin A sta- ance) may be either deficient (negative) or normal tus the first steps were taken in 1974, when WHO (positive). No analogy is perfect and whilst very convened an expert group to report on the knowl- few indeed would object to having a very high edge at that time on the problem of VAD and financial status, excess nutrient intake, including xerophthalmia (WHO Expert Group, 1976). A that of vitamin A, may be harmful. classification of the eye lesions was agreed upon. Although, as pointed out above, assessment With the limited experience at that time some of of dietary intake on its own is not equivalent to these eye signs were chosen, together with serum assessment of nutrient status, it is clearly closely retinol, and criteria were proposed for the iden- related. Consequently, the methodology of the as- tification of a vitamin A deficiency problem of sessment of dietary intake of vitamin A is given public health magnitude. some attention at the end of this chapter. Several years later the first national point It is clear that deviations in nutritional status prevalence survey of vitamin A deficiency was from the usually accepted normal range (which carried out in Indonesia (Sommer 1982). A sec- is, to a considerable extent, arbitrary) may vary ond WHO Expert Group on the Control of Vi- in degree. The terms “mild” “moderate” and “se- tamin A Deficiency and Xerophthalmia (WHO vere” deficiency are often used. Another useful Expert Group 1982) was largely influenced by division of status on the deficiency side is the use this study to make several alterations in the clas- of the terms “subclinical” and “clinical” deficien- sification and in the public health criteria (Tables cy (see below). 7.1 and 7.2). Nutritional status always concerns one par- The criteria shown in Table 7.2 were widely ticular aspect of the state of the body and conse- adopted and used successfully over the past 30 quently there are different types of test possible years or so. However, as the more severe eye le- for the assessment of status. Clinical signs in the sions and very low serum retinol levels gradually eye, several aspects of eye function, histological became less common, changes were advocated changes in the conjunctiva, serum level of retinol, (Sommer and Davidson 2002). These are shown 62 7 Assessment of Vitamin A Status

Table 7.1: Xerophthalmia Classification by ocular Table 7.3: Recently proposed prevalence criteria indi- signs (WHO Expert Group, 1982). cating significant VAD within a defined population.

Night blindness (XN) 1Plasma retinol 0.7 μmol/L in >15% Conjunctival xerosis (X1A) (20 μg/dL) 1 Bitot’s spot (X1B) Night blindness (XN) >5% in women 2 Corneal xerosis (X2) Total xerophthalmia >1% (XN, X1B, X2, X3, XS) Corneal ulceration/keratomalacia 1 2 <1/3 of corneal surface (X3A) Sommer and Davidson 2002, ACC/SCN 1997 Corneal ulceration/keratomalacia ≥1/3 of corneal surface (X3B) In addition a simplified approach has been advocated. This is to sum all the eye signs and if the Corneal scar (XS) total exceeds 1% a xerophthalmia problem exists. Xerophthalmic fundus (XF) A regrettable consequence of this simplification is the inability to distinguish between blinding and in Table 7.3. Table 7.3 has several weaknesses non-blinding xerophthalmia. If this practice is and does not appear to have come into general generally applied it will no longer be possible to use. estimate the role of VAD as a cause of blindness (see also Chapter 12). At the present time there seems to be a tendency to abolish the division traditionally set between clinical and subclinical deficiency. Al- Table 7.2: Criteria for assessing the public health significance of xerophthalmia and vitamin A deficiency, though there is no clear-cut threshold, this con- based on the prevalence among children less than six cept seems to have worked especially well in the years old in the community (WHO Expert Group, case of VAD. “Subclinical” implies that there is 1982). no clinical evidence of disease. The subject has Criteria Minimum no complaints of ill health and the examiner is un- prevalence able to elicit any physical signs of disease. Some Clinical (primary) indicators of vitamin A status span this borderline between subclinical and clinical. This is the case Night blindness (XN) 1.0% both for impairment of retinal rod function and for abnormal bulbar conjunctival histology. Both Bitot’s spot (X1B) 0.5% of these will be discussed in detail later, but it is useful at this point to see how they span the Corneal xerosis 0.01% subclinical-clinical divide (see Tables 7.4 and and/or ulceration/ 7.5). These tables show how impairment of func- keratomalacia tion and abnormality of structure, respectively, (X2 + X3A + X3B) increase progressively with increasing degree of deficiency. Xerophthalmia related 0.05% Although the following chapters deal with corneal scars (XS) the ocular lesions of xerophthalmia, after looking at the subclinical stages of vitamin A deficiency Biochemical (supportive) and the tests that have been developed for them it Serum retinol (vitamin A) 5.0% is important to recognize that in historical terms less than 0.35 μmol/L the eye signs were studied first. This is because (10 μg/dL) xerophthalmia, attributable to severe deficiency, 7 Assessment of Vitamin A Status 63

Table 7.4: Increasing impairment of retinal rod func- nificant increase in young child mortality (see tion, illustrating the subclinical–clinical divide. Chapter 9), what might be termed a second wave Tests of retinal rod Abnormal response of concern resulted. This realization provided the function response impetus for research into reliable methods for the Subclinical assessment of vitamin A status at the subclinical level. Dark adaptometry Abnormal final rod threshold Vision restoration Delayed response INDICATORS AT THE SUBCLINICAL after bleaching LEVEL Pupillary contraction Failure of dilation in low illumination Nearly thirty years on there is a variety of Clinical techniques on offer for assessment of VAD at this Night blindness Subjective level. It will be necessary to explain the varying impairment of vision state of popularity that each enjoys at the present in low illumination time. In addition, recent research has shown that the whole area of early detection of VAD is influ- Table 7.5: Progressive changes in conjunctiva and enced by accompanying subclinical and clinical cornea, illustrating the subclinical–clinical divide. infections or trauma. Assessment of the status of some other micronutrients is also similarly af- Test Abnormal appearance fected. Subclinical Conjunctival In epithelial and Acute phase response (APR) impression cytology goblet cells In the two previous editions of this Manual the Clinical phenomenon known as the acute phase response Conjunctival xerosis Dryness (APR) was referred to and described in relation (X1A) to VAD. However, recent years have seen a great Bitot᾿s spot (X1B) Foamy, cheesy, deal of research in this area and because of the heaping up of implications of this research for the assessment keratinized of vitamin A status, it deserves to be given promi- epithelial cells in nence at this point. Cross references are provided, interpalpebral fissure especially in Chapters 9, 11, and 13. Corneal xerosis (X2) Hazy cornea The acute phase response is the term given to the phenomenon that occurs in response to in- Keratomalacia (X3) Liquefaction of part flammation, infection or trauma. Proteins in the or all of cornea circulation fall into two classes with regard to the APR. In direct response to APR some proteins are released by tissues and increase in concentration was the first to be recognized as a serious public in the blood. These include C-reactive protein, se- health problem. It was only with the recognition rum amyloid, α1-antichymotrypsin, and α1-acid- of xerophthalmia as the most common cause of glycoprotein. The release of pro-inflammatory blindness in young children around the world (see circulating cytokines may be responsible. Other Chapter 12), that all the efforts expended to con- proteins, including many transport proteins, such trol it from the 1960s onwards could be justified. as retinol-binding protein and consequently retin- When, in the 1980s, it became evident that ol, are reduced in concentration. The mechanisms subclinical deficiency was associated with a sig- responsible for this fall are not fully understood, 64 7 Assessment of Vitamin A Status

but leakage into extracellular space and the de- ment of APR. The most appropriate, AGP (α1- pression of synthesis in the liver are thought likely acid glycoprotein), has been in short supply and causes (Rosales, Ritter, Zolfaghari et al. 1996). is relatively expensive. CRP (C-reactive protein) The fall in concentration is usually greatest is often misinterpreted by setting the cut-off point in acute infections and levels return to normal at 10 mg/L (mean CRP is <1 mg/L in absence of during recovery (Figure 7.1). It is now clear that inflammation). The correction of retinol data has even subclinical infection can cause significant not been taken up widely. A recent field study in falls (Thurnham, McCabe, Northrop-Clewes Kenya confirmed the conclusions reached in 2003 2003). Consequently, the fall in serum retinol (Thurnham, Mburu, Mwaniki et al. 2008). levels, applying equally to RBP, the ratio of RBP to TTR and RDR and MRDR (see below), is Serum retinol widespread and needs to be taken into account at The level of retinol in serum is under homeostatic all times where appropriate. Thurnham, Mburu, control over a wide range of body stores and re- Mwaniki et al. (2005) have recently reviewed flects these stores only when they are very high the importance of the acute phase response for or very low. Thus an isolated serum retinol value the assessment of micronutrient status in general. is not an accurate indicator of vitamin A status. Figure 7.1 shows an explanatory scheme, based Serum retinol is best used when a frequency dis- on field observations, of the effect of various de- tribution can provide useful information about the grees and stages of APR on plasma retinol. status of a population and about response to an David Thurnham (2009) states that there have intervention program (Gregory, Collins, Davies et been problems with the use of antigens for assess- al. 1995) (Figure 7.2).

(76%)

Figure 7.1: Effects of different stages of subclinical infection on plasma retinol. Changes are unidirectional except for in convalescence, during which people might recover or relapse. (Adapted from Thurnham, McCabe, Northrop-Clewes 2003). 7 Assessment of Vitamin A Status 65

Plasma retinol (µmol/L)

Figure 7.2: Distribution of plasma retinol in children aged 1.5 to 4.5 years. (Adapted from Gregory, Collins, Davies et al. 1995).

In the past it has usually been considered that Sowell 2001). Data from 16,058 participants aged a serum retinol level <0.35 μmol/L was “defi- 4–90 years were analyzed. In children serum reti- cient” and a level of <0.70 μmol/L was seen as nol increased with age, BMI, serum lipids, and “low.” The Annecy Accords (Sommer, Davidson use of supplements. Serum retinol <1.05 μmol/L 2002) concluded that a serum retinol level of prevalence was 16.7–33.9% in children aged <0.70 μmol/L (20 μg/dL) at a prevalence rate of 4–8 years and 3.6–14.2% at 9–13 years. Signifi- >15% should be a new prevalence criterion indi- cantly lower values occurred in non-Hispanic cating significant VAD within a defined popula- black and Mexican American children and wom- tion. The thinking behind this change (from >5% en. In circumstances where low levels of serum at <0.35 μmol/L) took into account evidence that retinol are common, it is difficult not to believe very low values are difficult to detect reproduc- that minimal infections are influencing these data ibly and reliably in routine laboratory tests. Since through the operation of the acute phase response the concentration has been doubled, more chil- (Stephensen, Gildengorin 2000). Subsequent ana- dren will be found to have these higher levels, lysis of these same data by the same group and in consequence the prevalence rate has been (Gillespie, Ballew, Bowman et al. 2004) for intra- increased automatically from 5% to 15%. individual variation in serum retinol caused them Although the use of serum retinol may appear to conclude that “the actual population prevalence to be satisfactory in the absence of inflammation of inadequate vitamin A status may be 75% lower and infection (i.e. where the acute phase response than the estimates previously reported.” This is is absent, see above and Chapters 9, 11, and 13), as an enormous difference and poses a serious chal- may be assumed to be the case in most developed lenge to acceptance of the currently recognized societies, doubts have to be raised. For example, standards for serum retinol data. serum retinol distribution data from residents of Figure 7.3 shows additional mean retinol val- the United States taken from the National Health ues in UK children surveyed in 1990, 1995 and and Nutrition Examination Survey, 1988–1994, 2000. As in the US (see Ballew, Bowman, Sowell have recently been studied (Ballew, Bowman, 2001, above) there was a marked increase with 66 7 Assessment of Vitamin A Status

2.2 Males 2.0

1.8 Females 1.6

1.4

1.2

1.0

0.8

0.6

Figure 7.3: Mean retinol concentrations at different ages for males and females obtained from three UK national surveys (Gregory et al. 1990, 1995, 2000). (◆), Males; ( ), females; (.....), 0.7 μmol retinol/L; (----), 1.05 μmol retinol/L (the two retinol concentrations represent the cut-off values used to indicate risk of vitamin A deficiency). (Adapted from Thurnham, Mburu, Mwaniki et al. 2005).

increasing age. This creates another problem in mothers (1.7 μmol/L). Serum all-trans retinoic using serum retinol to assess vitamin A status. acid was also lower (3.4 nmol/L) in the newborns It seems appropriate to raise another related than in their mothers (5.8 nmol/L). Serum con- problem at this point. It has been proposed that centrations of 13-cis retinoic acid were signifi- lutein may serve as a useful surrogate for retinol cantly lower (2.0 nmol/L) in the newborns than in plasma (Adelekan-Delana, Northrop-Clewes, in their mothers (2.6 nmol/L). The serum concen- Owa et al. 2003). It has been shown that lutein trations of all-trans retinoic acid and retinol did and some other carotenoids behave similarly to not correlate in any group. The authors conclude retinol in regard to the acute phase response. The that retinol concentrations do not accurately problems with serum retinol mentioned above reflect the concentrations of the biologically active would seem to apply equally to lutein; as we derivative all-trans retinoic acid. know they do to RBP. Serum retinol-binding protein (RBP) Retinoids in serum in mothers and newborns Several recent studies (Blaner, Piantedosi, Gam- Soderlund, Fex, and Nilsson-Ehle (2005) under- ble 1999; Donnen, Dramaix, Zihindula et al. 1999; took the first measurements of the biologically Almekinder, Manda, Kumwenda et al. 1999) have active metabolite all-trans retinoic acid and its shown that measurement of serum RBP correlates isomer 13-cis retinoic acid in newborns and their very closely with serum retinol (see Chapter 3). mothers, and in women in the first trimester of The immunodiffusion technique commonly pregnancy. Newborns had significantly lower employed is much simpler and cheaper than (1.0 μmol/L) retinol concentrations than their HPLC which is usually used for retinol. A simple, 7 Assessment of Vitamin A Status 67

portable microtechnique for measuring holo- the concept of relying upon the estimation of vi- RBP using a fluorometer was introduced by Craft tamin A concentration in the livers of patients at (1999). postmortem as a means of characterizing the vita- Using the immunodiagnostic kits referred to min A status of a population (WHO Expert Group, in Chapter 3, Gamble, Ramakrishnan, Palafox et 1976). In practice this has proved difficult, except al. (2001) confirmed the value of RBP as a sur- in the context of research. Furthermore, newer in- rogate for serum retinol. Criticisms of the latter direct and direct assessment methods have come concerning the APR, however, apply equally to to occupy centre stage. RBP. In healthy people serum RBP was found to correlate well with serum retinol (r = 0.88). Indirect assessment of liver stores This remained so, even in the presence of HIV-1 In Chapter 5 it was mentioned that apo-RBP ac- infection, protein malnutrition, or APR (Baeten, cumulates in the liver when retinol is in short sup- Richardson, Bankson et al. 2004). ply. Once retinol becomes available, holo-RBP is released into the circulation. This phenomenon is RBP/TTR molar ratio the basis of the relative dose response tests. The This ratio has been introduced in an attempt to magnitude of the retinol released from the liver assess vitamin A status in the presence of inflam- when a test dose is given is proportional to the mation (Rosales, Ross 1998). In the APR, RBP in degree of prior depletion of the liver. plasma falls, as in VAD. On the other hand, transthyretin (TTR), to which retinol is also bound in Relative dose response (RDR) plasma, falls in the APR but is unaffected by VAD. (Underwood 1990) Studies in experimental animals and children with Very recently, Fujita, Brindle, Rocha et al (2009) serious infections have indicated that a low mo- have shown that replacing serum retinol with lar ratio of RBP/TTR can distinguish VAD in the serum retinol-binding protein potentially allows presence of infection. Recently, further work was more accurate assessment of body reserves of carried out in malaria (Rosales, Topping, Smith et vitamin A than the usual employment of serum al. 1999). In HIV-1 infection the ratio is distorted, retinol in the RDR test. even in the absence of the APR (Baeten, Wener, A fasting serum retinol is measured (A0). Bankson et al. 2006). Vitamin A as 450–1000 μg retinyl ester in oily Several more studies using this technique solution is given orally. Five hours after dosing, have been carried out recently. A cut-off value another serum retinol is measured (A5). of less than or equal to 0.36 to indicate marginal The RDR is calculated as follows: vitamin A deficiency is recommended for young RDR = (A5-A0) x 100/A5 children by the work of Rosales, Chau, Haskell If the result of the calculation is >20% then et al. (2002). For adults, Zago, Dupraz, Sarchi et the test is considered to be positive (i.e. stores are al. (2002) proposed 0.37, for its high sensitivity, deficient). It has been shown that a result of 20% specificity and predictive value. in the test is approximately equivalent to a liver Recently (Sankaranarayanan, Suarez, Taren reserve of 0.07 mmol/g. Two blood samples are et al. 2005), it was shown that in deficient Nepal- required. ese women in late pregnancy the RBP:TTR index was a useful proxy for free holo-RBP concentra- Modified relative dose response (MRDR) tion. Thirty percent were judged to be vitamin A (Tanumihardjo, Koellner, Olson 1990) deficient by an index of <0.36. This test employs a metabolite of vitamin A, 3,4- didehydroretinyl acetate (DR, also known as de- Assessment of body reserves hydroretinol, vitamin A2). DR binds to RBP and In the past there was considerable advocacy for appears in the serum after a test dose is given if 68 7 Assessment of Vitamin A Status

liver reserves of vitamin A are low. A single oral always at hand. The compound DR, used in the dose of DR is given and only one blood test is MRDR test, is costly and is not readily available required, 4–6 hours after the dose. (WHO, 1996). Serum concentrations of retinol (SR) and Several studies have made comparisons be- de-hydroretinol (SDR) are measured. The molar tween various methods of assessment of vita- ratio is calculated in the following way: MRDR min A status (Rice, Stoltzfus, de Francisco et = SDR/SR. An MRDR value is considered to be al. 2000; Apgar, Makdani, Sowell et al. 1996; abnormal when the above ratio is >0.06 Wahed, Alvarez, Khaled et al. 1995; de Pee, Yuniar, West et al. 1997). Serum 30 day dose response (+S30DR) (Flores, Campos, Aranjo et al. 1984) This test is similar to the RDR described above Deuterated-retinol-dilution technique but the second blood sample is taken 30–45 days after the first. It has been used at the population This technique using deuterium, the stable level and has been used for monitoring the effec- isotope of hydrogen, measures total body vitamin tiveness of intervention programs for improving stores indirectly but quantitatively and is being vitamin A status. used increasingly in research projects (Haskell, Mazumder, Peerson et al. 1999; Ribaya-Mercado, Advantages Mazariegos, Tang et al. 1999; Haskell, Handel- It is known from work in depletion studies on man, Peerson et al. 1997). A dose of vitamin A experimental animals that vitamin A levels in the labeled with the stable isotope deuterium is giv- liver fall steadily over a considerable period of en by mouth and about three weeks allowed for time before serum retinol levels begin to fall, and equilibration with the reserves of the body. Blood even longer before any functional or structural is then sampled and the extent of dilution of the changes begin to occur. These tests appear there- labeled tracer relates to the amount of endoge- fore to be effective ways of measuring subclinical nous reserves. Olson (1997) has reviewed the dif- VAD. ficulties inherent in this new method but describes isotope-dilution techniques such as this one as “a Disadvantages wave of the future in human nutrition.” Blood samples are required. In many populations Recent work has vindicated Olson’s predic- this may be unacceptable for cultural reasons tion (above, 1997) in regard to this technique. An as well as in view of the risk of transmitting oral dose of deuterated retinyl acetate was found HIV, hepatitis etc.. The single sample for MRDR to have a mean equilibrium time of 16–17 days, raises less of a problem than the two samples uninfluenced by the size of hepatic reserves (Was- required for the RDR. antwisut 2002). Total vitamin A body reserves can Retinol-binding protein is among a number be estimated according to principles of isotope of proteins that take part in the acute phase re- dilution, a set of assumptions regarding retention action in infection and inflammation (see above). of the labeled dose, the ratio of specific activities The release of RBP from the liver is repressed by of plasma to liver vitamin A, and the irreversible a number of factors that are likely to be widely loss of vitamin A over time. prevalent in the communities under study. These The technique has been used to assess the vi- include acute and chronic infections and infesta- tamin A status of school children participating in a tions and protein-energy malnutrition. This sub- national sugar fortification program (Ribaya-Mer- ject is considered in more detail later (see Chapter cado, Solomons, Medrano et al. 2004). It has also 10). Other than in research projects, the means for been used to monitor dietary vitamin A intakes of estimating serum retinol on a routine basis are not elderly subjects of low vitamin A status (Ribaya- 7 Assessment of Vitamin A Status 69

Mercado, Solon, Fermin et al. 2004). In 31 elderly vitamin A supplementation was the measurement men and 31 elderly women who were marginally of breast milk vitamin A per gram of fat in casual nourished it was shown that a sufficient dietary breast milk samples (Rice, Stoltzfus, de Francisco intake of vitamin A was associated with an ad- et al. 2000). equate liver vitamin A concentration of 6.45 μg Plasma and breast milk retinol were measured RAE/kg body weight. For a reference 76 kg man as indicators of vitamin A status during pregnan- and 61 kg woman these values are approximately cy. Acute phase proteins, where infections were 500 and 400 μg RAE/day respectively. present, were also measured. It was concluded Octadeuterated and tetradeuterated vitamin A that the measurement of retinol in breast milk was have been given before and after dietary interven- superior to the measurement of retinol in serum tions respectively: in this way the impact on total under these conditions (Semba, Kumwenda, Taha body vitamin A stores of different dietary sources et al. 2000). of provitamin A carotenoids have been studied. It was later shown that breast milk retinol con- Tang, Qin, Hao et al. (2002) studied whether centrations are not affected by APR (Duncheck, the usual sample collection time of 11–26 days Nussenblatt, Ricks et al. 2005). could be reduced to make field studies easier. They gave an oral dose of deuterium-labeled vitamin A to 58 Chinese children aged 11–12 years. Histological indicator [conjunctival It was found that results using 3 days’ isotope en- impression cytology (CIC)] richment correlated well with 21 days, whereas 6 hours’ did not. In Chapter 6 attention was drawn to the major role played by vitamin A in cellular differentiation. The abnormal changes that occur when vi- Breast milk vitamin A concentration tamin A is deficient have been best documented for the conjunctiva and to a lesser extent for the It has long been known that the concentration cornea. Basically there is a drying process, xero- of vitamin A in the breast milk of undernourished sis, which, in the conjunctiva, may ultimately lead mothers is low. The proposal to use this as an in- to the development of a Bitot’s spot (X1B) (see dicator of the vitamin A status of a community Chapter 8). The earlier change that is visible to is relatively new and has been tested under vary- the naked eye, called conjunctival xerosis (X1A), ing conditions. It has the advantages of being is much less distinctive (see Chapter 8). This in non-invasive, readily acceptable and the sample turn is preceded by lesser degrees of change at the is easy to collect. It is important to follow stan- subclinical level. dardized methods of collection of the sample, and With the introduction of the technique of CIC the method of expression of the concentration (Wittpenn, Tseng, Sommer 1986) and a modifi- of vitamin A should be agreed upon. Results to cation termed impression cytology with transfer date suggest that the borderline of deficiency in a (ICT) (Luzeau, Carlier, Ellrodt et al. 1988) it has population be set at <1.05 μmol/L. become possible to study these very early chang- In one study in Indonesia (Stoltzfus, Hakimi, es under a microscope. Details of the techniques Miller et al. 1993) there was a close correla- that have been advocated are given in a training tion between breast milk concentrations below manual (Wittpenn, West, Keenum et al. 1988), 1.05 μmol/L and the prevalence of positive RDR in several original papers and also in the WHO tests in infants at 6 months of age. More recently 1996 report (WHO, 1996). In essence, a normal in Bangladesh a comparison of serum retinol, impression of conjunctival cells when stained will MRDR and breast milk found that overall the show one or more sheets of small regular epithe- most responsive indicator to postpartum maternal lial cells and numerous mucin-secreting goblet 70 7 Assessment of Vitamin A Status

cells. In VAD the epithelial cells become flat- goblet cells are markedly reduced in number or tened, enlarged and are reduced in number. The absent (see Figures 7.4–7.9).

Figure 7.4: Normal conjunctival impression with ab- Figure 7.7: Higher power of abnormal, enlarged con- undant goblet cells, sheets of small epithelial cells, and junctival cells (PAS and Harris’s hematoxylin, x 400). mucin spots (periodic-acid Schiff (PAS) and Harris’s hematoxylin, x 160).

Figure 7.5: Higher magnification of normal conjuncti- Figure 7.8: PAS-positive mucin spots representing va, showing contrast between PAS-positive goblet cells “impressions” of goblet cells on conjunctival surface and epithelial cells (PAS and Harris’s hematoxylin, (PAS and Harris’s hematoxylin, x 400). x 400).

Figure 7.6: Abnormal conjunctival impression with Figure 7.9: Impression cytology from normal child complete loss of goblet cells and mucin spots, along showing transition from abundant normal epithelium with appearance of enlarged epithelial cells (x 100). (lower left) to abnormal epithelium (upper right). Specimen was graded as normal (x100).

Figures 7.4–7.9: Conjunctival impression cytology (Wittpenn, Tseng, Sommer, 1986). 7 Assessment of Vitamin A Status 71

Limitations Night blindness Difficulties have arisen over the significance This is the most advanced stage of deficiency that of the appearances of some of the samples. While can be related to rod dysfunction. In a subject there is no doubt about those described above for able to cooperate this is the subjective sensation clearly normal and clearly abnormal impressions, of inability to see adequately in poor illumina- the vast majority of impressions will not be so tion, such as at dusk. It is sometimes erroneously distinct. A standardization of existing interpre- termed hemeralopia (Greek: hemera = day) or tation schemes is required if the technique is to nyctalopia (Greek: nyct = night; alaos = obscure). come into widespread use (WHO, 1996). The test Dusk blindness would be a better term. is well accepted after the age of about 3 years, but has posed problems in younger children. The dark adaptometer is a standard piece of The presence of acute conjunctivitis or trachoma equipment in the ophthalmology department for (Lietman, Dhital, Dean 1998) interferes with CIC. examination of night vision. Some versions are available for use under field conditions. These More recent experience suggests that con- may be used for investigating the problem of junctival impression cytology with transfer night blindness in cooperative subjects such (CICT) identical with ICT above is the preferable as school children and pregnant and lactating technique (Chowdhury, Kumar, Ganguly et al. women, among whom a high prevalence has 1996). These investigators reported that healing been reported in some countries (Sommer, West after vitamin A dosing takes about 70 –110 days 1996:338). (Chowdhury, Kumar, Ganguly 1997). This may seem to be a long time for a minimal lesion such In young children, the major vulnerable group as early xerosis and keratinization of the conjunc- for VADD, direct observation and an interview tiva, especially in comparison to the healing of have usually been employed. A standard set of more severe clinical lesions (see Chapter 8). In interview questions for the child’s guardians has this regard it should be remembered that micro- been developed (Sommer, Hussaini, Muhilal scopic techniques are likely to be more precise et al. 1982) and use is made of the fact that in than clinical observations. CIC is increasingly endemic areas terms frequently occur in the lo- being used in clinical ophthalmology and new re- cal language, for example likening poor vision finements are constantly being introduced (Thiel, at dusk to “chicken eyes” (chickens have no rod Bossart, Bernauer 1997). cells in their retinas and are therefore night blind) or Bitot’s spots to “fish scales.” If the child is old enough to walk, it may be observed to stumble in INDICATORS AT THE CLINICAL LEVEL a room at dusk, or it may fail to grasp a proffered toy or sweet (see also Chapter 8). Impaired dark adaptation As mentioned, the rod cells of the retina contain The recent development of sensitive tests vitamin A in the form of 11-cis retinal bound to such as those outlined below (as described more opsin to form rhodopsin (see Chapter 6). These fully in Chapter 8) have led to the replacement, cells are photosensitive under conditions of low wherever possible, of this subjective and unreli- levels of illumination (night or scotopic vision). able method. Table 7.4 above shows the various stages of increasing deficiency at which different kinds of Vision restoration time (VRT) testing may be applied. The ability of the bleached eye to recognize a letter under low levels of illumination has been measured in school-age children in Thailand 72 7 Assessment of Vitamin A Status

(Udomkesmalee, Dhanamitta, Sirisinha et al. the test object can be identified by the subject 1992). From this study, it appeared that zinc has recently been introduced. The apparatus is was a more important determinant of VRT than portable and early results are promising (see vitamin A (see also Chapter 11). The test is Chapter 8). applicable only to older children with some degree of literacy. A vision restoration time of Xerophthalmic fundus >180 seconds has been suggested as a cut-off This abnormality in the appearance of the retinal for abnormal night vision, suggestive of VAD. fundus has usually been considered too rare to be used in the field for recording the presence of xe- Pupillary threshold rophthalmia. A recent reappraisal suggests it may This test is based on the fact that the weakest have some practical value (see Chapter 8) threshold of light visible in the dark-adapted state is approximately of the same intensity as that needed to cause pupillary contraction. The pupils ASSESSMENT OF DIETARY INTAKE OF of subjects who are night blind due to VAD fail VITAMIN A to react or constrict normally in low illumination. Children as young as one year old have As mentioned above, this is strictly speaking been studied successfully (Congdon, Sommer, not a measure of vitamin A status but can provide Severns et al. 1995). Further testing in children useful and complementary information otherwise with VAD in India (Sanchez, Congdon, Som- unavailable. It has the advantage of being non- mer et al. 1997) and of pregnant women in Nepal invasive, inexpensive and uncomplicated, so that (Congdon, Dreyfuss, Christian et al. 1999) has large numbers of subjects can be readily targeted been distinctly promising. The test has proved and a profile of a population obtained. Knowledge readily acceptable, is associated closely with of available and utilized sources of vitamin A that serum retinol, and responds to vitamin A supple- may prove to be of value in subsequent dietary in- mentation. It also has the advantages of being terventions can be collected. Two organizations, noninvasive, applicable to young children and the International Vitamin A Consultative Group sensitive at the subclinical stage of VAD. (Figure (IVACG) and Helen Keller International (HKI) 7.10). have been largely instrumental in promoting the development of assessment methods and field Night vision threshold test (NVTT) studies. This test of the lowest level of luminance at which The IVACG (1989) introduced a simplified dietary assessment (SDA) as a method of identifying and monitoring groups at risk of suboptimal Figure 7.10: Response of light threshold to vitamin A vitamin A intake. Intake information is obtained (Sommer, West, 1996). by past 24-hour dietary recall and by history of

) the usual pattern of consumption of vitamin A rich 2 foods during the previous month. Food quantities were estimated. Some experience with this method has found it to be difficult, time-consuming, or that it correlates poorly with measurement of vitamin A status. The 24-VASQ (twenty-four hours vitamin A semiquantitative) method is partly based on the IVACG method. It estimates vitamin A

Fall in log threshold (log – c dm intake using 24-hour recall data, uses individual ingredients instead of dishes, and cites four 7 Assessment of Vitamin A Status 73

sources – vegetables, fruits, animal foods and areas this can provide 5–18% of the RDA and fortified foods. It has been used successfully in should therefore be considered. Experience with Indonesia, Bangladesh and Nepal (de Pee, Bloem, five communities in the Philippines, five in Tan- Kiess et al. 1999; de Pee, Bloem, Halati et al. zania and five in Guatemala, using serum retinol 1999). 0.70 μmol/L as deficient, found correct classifica- The Helen Keller International Food Fre- tion as deficient in 73.3%, but a high false-posi- quency Method (HKI FFM) was introduced in tive rate of 42.9% (Sloan, Rosen, Paz et al. 1997) 1993 (Rosen, Haselow, Sloan 1993). This method Unfortunately, the uncertainty now prevailing studies the diet of children 1–6 years of age. A as to the bioavailability and bioconversion fac- score is assigned to each child based on the num- tors for provitamin A carotenoids (see Chapter 5) ber of rich sources (>100 RE/100 g vitamin A overshadows these and all methods of assessment content), both animal and plant, consumed during of dietary vitamin A intake. Until these matters the previous week, regardless of amount. are resolved, results of field work will have to be The original method ignored animal milk, interpreted with caution. but it has subsequently been shown that in some 74 75

8 Xerophthalmia

DEFINITIONS To a certain extent the ocular signs in Table 7.2 have been listed in an increasing order of se- The term xerophthalmia literally means “dry verity. This means that retinal function tends to eye.” However, dryness or xerosis, which also be impaired before xerosis affects first the con- affects other parts of the body, is only part of junctiva and then the cornea. Liquefaction of the the abnormal process undergone by the eye in cornea, of increasing severity, is normally a very vitamin A deficiency. Xerosis is confined to the late stage. Corneal scars are not part of the active epithelial structures of the eye, the conjunctiva deficiency process. They may be considered to be and the cornea. Only the conjunctiva covering stigmata of previous deficiency, providing there the globe, known as the bulbar conjunctiva, and is circumstantial evidence suggestive of vitamin A not that lining the eyelids, the palpebral con- deficiency as their cause. These characteristics junctiva, is affected by xerosis. In addition, the cor- are discussed later. The xerophthalmic fundus is a nea undergoes other changes, known as kerato- rarity and does not fit neatly into the classification malacia, as will be described later. After recovery (see later). from acute vitamin A deficiency that affects more There are two other aspects of the subject than just the most superficial layer of the cornea, of the definition of xerophthalmia that need to scars of varying extent and depth remain (XS). be made quite clear. Firstly, there are causes of In vitamin A deficiency the retina is also af- xerophthalmia other than dietary vitamin A defi- fected. The rhodopsin system in the rod cells of ciency, the epidemic disease that primarily affects the retina is much more sensitive to deficiency young children in developing countries and that than is the iodopsin system in the cone cells. As a is our main concern in this Manual. These cases result, rod function is impaired early on, resulting arise sporadically and are caused by various de- when sufficiently marked in impairment of night fects in the utilization of vitamin A within the vision. Impairment of cone function (i.e., day vi- body. Most of these are quite uncommon. They sion and color vision) is rarely seen clinically. are instances of secondary or endogenous VAD There have been a few reports of structural dam- and are considered later in Chapter 14. age to the rod cells in the retina. These have been The other aspect that needs to be pointed out called xerophthalmic fundus (XF). is that some of the eye changes may be caused by All of these eye changes are included in the diseases that are not related to vitamin A deficien- term xerophthalmia. In other words, xerophthal- cy. For example, night blindness is a symptom of mia is synonymous with all of the clinical signs several rare degenerative disorders of the retina. It and symptoms that affect the eye in vitamin A is also a main feature of a not uncommon blinding deficiency. The various ocular appearances in disease called retinitis pigmentosa. For the most xerophthalmia were classified by WHO in 1976 part these diseases do not respond to treatment (WHO Expert Group 1976) and modified in with vitamin A (see Chapter 14). 1982 (WHO Expert Group 1982). Criteria of Dryness of the conjunctiva and cornea is also a public health problem were also developed a feature of several diffuse connective tissue dis- and were recently revised again. All of these orders, one of which is Sjogren’s syndrome. The recommendations are shown in the previous xerosis results from atrophy of the secretory epi- chapter (Tables 7.1; 7.2; 7.3) and should be re- thelium of the lacrimal glands, and does not referred to. spond to vitamin A. 76 8 Xerophthalmia

There are several facts that shed light on the ing burden on the family and the community at way in which xerophthalmia has tended to be ne- large. glected over the years. For reasons that are not clear, xerophthalmia was usually not included among the classical vitamin deficiency diseases NIGHT BLINDNESS (XN) when these were first identified in the 19th and early 20th centuries. Beriberi, scurvy, pellagra It will be recalled that the subject of retinal and rickets were usually considered together and dysfunction was considered previously in con- xerophthalmia was rarely included (Funk 1912). nection with indicators of VAD (see Chapter 7). Even though fat-soluble A was the first vitamin As shown in Table 7.3, night blindness is the identified (McCollum, Davis 1915) and was most extreme form of retinal dysfunction; suffi- therefore given the first letter in the alphabet, xe- ciently severe to cause subjective impairment of rophthalmia was not studied in detail at the time. vision at night. As an indicator of VADD it has Night blindness and Bitot’s spots tended to be de- both advantages and disadvantages. On the posi- scribed in isolation from the blinding changes of tive side, tests are noninvasive, and some may be corneal xerophthalmia. It may be that scientists applied by investigators with no specialized oph- and general physicians did not feel competent to thalmological training. In this category are ques- examine the eye and deal with eye disease. It is of tionnaires and observation of the performance of interest that the only vitamin deficiency to have young children under conditions of low illumina- reached epidemic proportions and to persist at tion (Table 8.1; WHO 1996). these levels today is xerophthalmia. This historical aspect of VADD has been recently discussed Table 8.1: Scheme for the classification of night blind- (McLaren 2004). ness by interview (WHO, 1996). Corneal xerophthalmia is usually associated 1) Does your child have any problem seeing with a severe degree of vitamin A deficiency, in the daytime? which is often accompanied by severe general- 2) Does your child have any problem seeing ized malnutrition (PEM) and serious infections at nighttime? (McLaren, Shirajian, Tchalian et al. 1965). Death 3) If (2) = yes, is this problem different from is the most likely outcome at this late stage, other children in your community? even in adequately treated patients. In this way (Note: this question is particularly appro- death “removes” the problem of long-term care priate where VAD is not very prevalent.) of the blinded survivors. Follow-up over a period of one year after the diagnosis of corneal 4) Does your child have night blindness (use xerophthalmia has shown that only about 40% local term that describes the symptom)? survive (ten Doesschate 1968). Of these survivors about 25% remain totally blind and 50–60% are On the negative side, the objective testing of partially blind. It is, consequently, uncommon night vision requires sophisticated and expensive to find a high proportion of inmates of blind equipment, operated by skilled ophthalmological schools in developing countries having xe- staff. The subjects need to be of sufficient age and rophthalmia as the cause of their blindness. Even education to cooperate fully in the testing. This those that do recover from acute severe vitamin A applies to dark adaptometry, rod scotometry, and deficiency when young are likely to fare badly to some extent to electroretinography, which is in the community. In such circumstances there is less sensitive. little possibility of their receiving an education. One study in Bangladesh (Hussain, Kvale, Without special training they will be unable to do Odland 1995) compared over a hundred children any work and their survival constitutes a continu- aged 2–15 years and reported to be night blind by 8 Xerophthalmia 77

their parents with a similar number of matched Khan 1984; Katz, Khatry, West et al. 1995) are controls. Both groups received an eye examina- also vulnerable groups in which the measurement tion, test of scotopic vision by a luxometer – a of rod vision may be a useful means of assessing simple form of dark adaptometer – and measure- vitamin A status in a community. These groups ment of serum retinol level. Although there was can cooperate in dark adaptometry, thus this is the fairly close correlation between the two methods method of choice. of diagnosing night blindness, the parents’ report Clinical dark adaptometers, designed for use appeared to be less sensitive. in hospitals, are expensive and sensitive pieces The occurrence of names for night blindness of equipment. There are simpler and more robust in local languages suggests that this rather dis- forms of equipment that may be readily transport- tinctive phenomenon occurs with some regularity ed for use in field studies. Vitamin A deficiency in a community. The commonest terms used are results in delay in rod adaptation to conditions of “night eyes” and “chicken eyes”. illumination of low intensity, followed by reduc- Little night blindness is detected in very tion in threshold sensitivity. Figure 8.1 shows the young children because the effects of deficien- curve that is obtained on testing a normal subject. cy only become evident when the child tries to Figure 8.2 illustrates how test curves rose pro- move around at dusk. From about 2 years onward gressively in a volunteer subject receiving a vita- rates of night blindness tend to rise, because of min A deficient diet. increased activity. Evidence has accumulated in In practice, it is usually sufficient to allow recent years to show that school age children and both examiner and a group of subjects for testing pregnant and lactating women (Khan, Haque, to undergo dark adaptation for about 30 minutes

Figure 8.1: Normal curve of dark adaptation showing cone-rod transition time, cone threshold, and final rod threshold (Hume, Krebs 1949). Log millilamberts are units of intensity of illumination. 78 8 Xerophthalmia

Figure 8.2: Curves of dark adaptation for one deprived subject at different stages of depletion (Hume, Krebs 1949). Log millilamberts are units of intensity of illumination, numbers on right are dates of tests.

in a darkened room. At this point maximum pos- rates reported have been surprisingly high, caus- sible rod adaptation will have occurred and the ing health authorities to devote much more at- tests are carried out at the final rod threshold (see tention to this more recently recognized public Figure 8.1). health problem. Night blindness is frequently the most preva- Accompanying this research activity has been lent form of xerophthalmia, as might be expected. a growing interest in the methodology of test- Although serum retinol of > 0.70 μmol/L has ing for night blindness; the salient feature of xe- customarily been considered to be “normal,” it rophthalmia in this group. was found that about 20% of children with night A large study in Nepal of the effects of vita- blindness and 10% with night blindness and min A or β-carotene supplementation on maternal Bitot’s spot had higher levels of serum retinol (see mortality was carried out during 1994–1997 (see Table 8.2). As might be expected, impaired dark Chapter 8). As part of this research it was found adaptation could be detected with serum retinol that 877 women had night blindness (as judged by between 20 and 30 μg/dL. weekly night blindness histories elicited through- Recent years have seen a surge of interest, out pregnancy). 9545 were negative, giving the especially in some of the countries of Southeast high prevalence of 9.1% (Christian, West, Khatry Asia, in the occurrence of xerophthalmia in wom- et al. 2000). In the placebo group, mortality of en of child-bearing age. In particular, attention night-blind women was 360/100,000 pregnancies. has focused on those who are, or have recently In comparison, the relative risk in the night-blind been, pregnant and may be lactating. Prevalence was 0.32 and in the non-night-blind 0.18. When 8 Xerophthalmia 79

Table 8.2: Association between xerophthalmia status and serum retinol (µg/dL). (From Sommer, West 1996). Clinical status Deficient Low Adequate Mean Number (<10) (10–20) (>20) of cases XN (+), X1B (-) 27% 55% 18% 13.9 174 XN (-), X1B (+) 31% 57% 12% 13.4 51 XN (+), X1B (+) 38% 53% 9% 12.1 79 Neighborhood controls – – – 17.7 282 Random sample 8% 37% 55% 20.0 268

XN = Night blindness; X1B = Bitot̓s spot; Controls = Neighborhood peers of same age and sex; random sample = peers from throughout the six villages studied infection-related causes of death were examined, in night-blind women of all groups. The improve- it was shown that night-blind women were five ment was greater in the group that received liver times more likely to have died of infections than than in that receiving vitamin A fortified rice. the non-night-blind. Plasma retinol response was greater in the higher dose capsule and liver groups than in the vegetable groups. Only the group receiving vitamin A Pupillary threshold fortified rice failed to improve (Haskell, Pandey, In Nepal the dark-adapted pupillary response Graham et al. 2005). was used to identify 298 pregnant and lactating A study was recently carried out to validate women with night blindness (Congdon, Dreyfuss, different definitions of reported night blindness Christian et al. 2000). In a placebo-controlled trial (Wedner, Ross, Congdon et al. 2004). No local with vitamin A and β-carotene, women receiving words for night blindness were in use in this ru- vitamin A had better dark adaptation thresholds ral population of Tanzania. Participants were 461 than women receiving placebo or β-carotene. children aged 24–71 months, 562 primary school Dark adaptation threshold was associated with children, and 191 pregnant or lactating women. serum retinol concentration in women receiving In all, 152 subjects complained of night blind- placebo or β-carotene, but not in those receiving ness and were group matched with 321 controls. vitamin A. Night blindness reports were validated by serum The same group (Christian, West, Khatry et retinol and pupillary dark adaptation. All children al. 2001) carried out a stratified analysis by ma- and women who reported night blindness or had ternal night blindness status during pregnancy other signs of xerophthalmia were treated with for 10,000 women. Mortality of infants of non- vitamin A and were followed up 3–4 weeks later. night-blind women was 63/1000 live births. Half of the control group were also treated and Among those night-blind, infant mortality was followed up. The overall prevalence of night relatively greater by 63% with placebo, 50% with blindness was 12.5%. At baseline, mean pupil- β-carotene, and only 14% with vitamin A. lary threshold and median serum retinol were Recently, pregnant Nepalese women suffer- not significantly different in cases and controls, ing from night blindness were studied during their either overall or by groups. More restricted case response to small daily doses of various sources definitions significantly reduced the prevalence of vitamin A. These included amaranth leaves, of reported night blindness to 5.5% (P<0.001), carrots, goat liver, vitamin A fortified rice, or red but there were no significant differences between palm oil. Pupillary threshold was used to assess cases and controls. After treatment dark adapta- vitamin A status. This test improved significantly tion improved significantly in cases in young chil- 80 8 Xerophthalmia

dren and primary schoolchildren, but only when obtain reproducible and comparable results it is restricted case definitions were used. It was con- of great importance that WHO guidelines be fol- cluded that night blindness reports were a poor lowed in matters like this (WHO Expert Group, indicator of vitamin A deficiency in this popula- 1982). tion. Both experimental and clinical evidence suggest that the process of xerosis, which also affects Night vision threshold test epithelial tissues other than the conjunctiva and NVTT has also been advocated for use in the cornea, is primarily due to changes in the proteins field. A detailed description of the apparatus has of the tissue itself. Loss of tears also occurs, but been published (Taren, Duncan, Greivenkamp et this is a secondary phenomenon, one that tends al. 2001). Using this test 1401 urban Nepalese to make existing xerosis worse. Loss of goblet pregnant women were examined (Taren, Duncan, cells and lack of secretion of mucin is an integral, Shrestha et al. 2004). Sixteen percent failed the but secondary, part of the process of xerosis in test, but only 6.4% reported having night blind- the conjunctiva. Local eye infections are frequent, ness. NVTT test failure was accompanied by making the clinical condition worse, but are not lower serum retinol. Serum retinol was correlated etiologically involved. with the NVTT scores (P<0.001).

BITOT’S SPOT (X1B) CONJUNCTIVAL XEROSIS (X1A) As was mentioned earlier, Bitot’s spot is the The term conjunctival xerosis could apply final part of the process of xerosis affecting the to any stage of xerotic change in the conjunc- bulbar conjunctiva. The typical Bitot’s spot (see tiva (Figure 8.3). This would range from abnor- Figure 8.3 and Figure 8.4) occurs in the exposed mal impression cytology, through dryness of the part of the conjunctiva, the area between the lids conjunctiva, to keratinization and heaping up of normally exposed, called the interpalpebral fis- material as in Bitot’s spot (X1B, see below). CIC sure. is subclinical and was dealt with previously (see The temporal (lateral) aspect of the eye is usu- Chapter 7). In the Xerophthalmia Classification ally first affected and consequently most Bitot’s (see Table 7.1) X1A distinguishes dryness from spots are found there. The nasal (inner) aspect is Bitot’s spot. Before this scheme was agreed upon affected later and only in extensive involvement a great deal of attention was given to various con- do the inferior and then finally the superior quad- junctival appearances, including thickening, wrin- rant undergo change. A Bitot’s spot consists of a kling and pigmentation. It is now recognized that heaping up of desquamated, keratinized epithe- these changes and the degree to which they occur lial cells, which form a slightly raised area that are not related directly to vitamin A deficiency. If may be readily wiped away. This leaves an un- the classification (see Table 7.1) and the criteria even, eroded base in the superficial epithelium, on (Table 7.2) are compared it will be evident that which more abnormal cells may accumulate over X1A is not one of those appearances chosen for a few days. The transient nature of Bitot’s spots use in surveys. This is because it has repeatedly creates a problem over their use in surveys. A sub- been found that, while advanced instances of X1A ject may “remove” a spot by vigorously rubbing can be readily identified, minor changes are sub- their own eyelids. ject to great inter- and intra-observer variation. Bitot’s spots vary considerably in size and This makes X1A an unreliable indicator of VAD. shape. The areas of conjunctiva affected may be Even so, it is often – wrongly – used in surveys, multiple, but more usually there is a single spot to making interpretation of results very difficult. To an eye. Some single spots are ovoid, others lin- 8 Xerophthalmia 81

Figure 8.3: Bitot’s spot (X1B) on temporal aspect of bulbar conjunctiva in interpalpebral fissure. Bubbles of foam are clearly visible.

Figure 8.4: Bitot’s spot without foam and of a “cheesy” appearance. The nature of the material is not known to have any significance. 82 8 Xerophthalmia

ear, and occasionally they take a roughly trian- Table 8.3: Characteristics of vitamin A responsive and gular form with the base close to the limbus, the unresponsive Bitot’s spots (X1B). junction between conjunctiva and cornea. None Responsive of these characteristics has any special clinical • Subject usually a child of less than significance (McLaren 1962). 6 years of age Broadly speaking, the appearance of Bitot’s • With the maximum vitamin A dosage, spots has been likened either to 1) foam; or 2) response usually evident within one week cheese (see Figure 8.3 and 8.4). Gasforming bac- • Usually accompanied by generalized teria may be responsible for the first appearance. conjunctival xerosis and night blindness There is no known significance to this difference • Males more commonly affected than in appearance. females Undoubtedly, the most significant way in which Bitot’s spots may be classified is as to Unresponsive whether they are 1) responsive to vitamin A treatment; or 2) unresponsive (Djunaedi, Sommer, • Commonly occur in children over 6 years Pandji et al. 1988). There are several character- and in otherwise healthy adults istics that may be of assistance in any attempt to • Usually a small, single spot identify the nature of a Bitot’s spot (see Table • No accompanying evidence of VAD 8.3). • Largely responsible for the apparent It is usually not possible to determine the increase of X1B prevalence with increasing cause of an unresponsive Bitot’s spot. Some ap- age pear to be stigmata of earlier deficiency and may therefore in this sense resemble corneal scars (XS) that are considered to relate to earlier defi- This suggests that the process of xerosis tends to ciency (see later). For others there is no evidence spread from the conjunctiva to later involve the of either earlier or present deficiency and these cornea. Clinically evident corneal xerosis (X2) in may result from local trauma of some kind. This which the cornea has a distinct hazy appearance might be due to a combination of environmental (see Figure 8.6) tends to last for a matter of only factors, among which might be included ultra- a day or two before advancing to deformation of violet exposure at high altitude, smoke-filled the cornea known as keratomalacia (see below). huts, or chronic eye infections, especially tra- Up to the stage of corneal xerosis (X2) choma. This hypothesis is supported by the fact prompt treatment with large doses of vitamin A that Bitot’s spots form most readily on the most can result in full preservation of sight without exposed part of the conjunctiva. Further evidence any residual impairment. It is of paramount im- for a role of exposure in Bitot’s spot formation is portance that all stages of xerophthalmia should reports of their presence on unusual aspects of the receive the maximum treatment with vitamin A conjunctiva as a result of distortions of the eyelids (see Chapter 15). In contrast to the lengthy full (McLaren 1980) (see Figure 8.5). response to vitamin A therapy of abnormal conjunctival cytology mentioned earlier (see Chapter 7), clinical improvement has been reported to take CORNEAL XEROSIS (X2) place in 70% of corneal lesions (X2/X3) within four days, in 95% within one week, and the vast Many instances of conjunctival xerosis are ac- majority were entirely healed within two weeks companied by a superficial punctate keratopathy, (Sommer, West 1996:285). Histological and clini- which can be seen when examined with the slit- cal healing are both subjective, and clearly are not lamp microscope (Sommer, Emran, Tamba 1979). strictly comparable. 8 Xerophthalmia 83

Figure 8.5: A single Bitot’s spot associated with abnormal exposure of the conjunctiva due to ectropion associated with scar on cheek. 84 8 Xerophthalmia

Figure 8.6: Bilateral haziness of the cornea (X2) in a young child identified in a survey.

KERATOMALACIA (X3A, X3B) loricrin, and keratin 10 begin to be expressed on the abnormal epithelial cells of the cornea. In the 1982 WHO Xerophthalmia Classifica- Many years ago (Pillat 1929) cases of kera- tion, keratomalacia was divided into two stages tomalacia in adults in China were also described according to the extent of the involvement of the as having similar colliquative changes in the skin. cornea (see Table 7.1). Keratomalacia is charac- This was termed “dermomalacia.” Figure 8.8 terized by softening of the corneal substance in shows one eye and the surrounding skin of a five- addition to increasing xerosis of the epithelium month-old Palestinian refugee child who died (see Figure 8.7). shortly after admission to hospital in Amman, Corneal softening is due to a unique patho- Jordan. There is advanced keratomalacia with logical process termed colliquative necrosis. The almost the entire cornea affected and the central stroma becomes edematous. It is suspected that area about to prolapse. In addition, the skin has activation of collagenases and other enzymes may a markedly burnished appearance, which also af- be responsible, but the precise pathogenesis is not fected much of the skin of the head and neck. The known. Now that the clinical condition has be- skin change is very reminiscent of that described come a comparative rarity, little effort has been earlier as dermomalacia. As a result of this pro- made in recent years to try to understand the cess in the cornea there will always be some de- underlying mechanism. A Japanese research gree of residual damage and deformity. group (Toshino, Shiraishi, Zhang et al. 2005) In corneal ulceration there is usually only one has shown in the severely vitamin A deficient rat ulcer per eye. The typical ulcer is infero-nasal that as the cornea keratinizes there is an increase in position, about 0–2mm in diameter and in keratinocyte transglutaminase (TG1, TG(K)) 0.25–0.5mm in thickness. In about 20% of cases activity and the envelope proteins involucrin, both eyes are affected and the characteristics tend 8 Xerophthalmia 85

Figure 8.7: Colliquative necrosis (keratomalacia) affecting the greater part of the cornea (X3B). The relative sparing of the superior aspect is typical. Plasma vitamin A, 4 µg/dL.

Figure 8.8: Entire cornea is undergoing liquefaction (X3B). Skin surrounding the eye shows hyperkeratinization suggestive of “dermomalacia” (see text). 86 8 Xerophthalmia

to be similar. Hypopyon (collection of sterile pus Table 8.4: Distinguishing characteristics of vitamin A in the anterior chamber) is common and infection related corneal scars (XS). is frequent. Eye examination • Location: typically nasal and inferior on the cornea if only a small part involved CORNEAL SCAR (XS) – • Often bilateral, not necessarily equal in VITAMIN A RELATED extent

Scarring of the cornea may result from a wide History variety of diseases affecting the eye. Visual impairment is inevitable; its degree depends on the • Onset between about 2 months and 5 years location and the density of the scar. Damage that • Accompanied at that time by severe PEM, is confined to the cornea may be overcome by sur- measles, severe diarrhea, respiratory gery. This is not possible when internal structures infection are also involved, usually as a result of accom- • Absence of trauma, or prolonged purulent panying infection. Prevention is clearly infinitely discharge preferable. Careful and detailed history taking and general physical examination, in addition to the eye examination, are of particular importance in this instance (see Figure 8.9). If, as a result of these, there is any residual doubt about the likely etiology of a corneal scar, then it should not be attributed to vitamin A deficiency. Table 8.4 gives the information to be elicited in order to come to a diagnosis.

XEROPHTHALMIC FUNDUS (XF)

This rare condition has been described mainly in school age children or young adults in South- east Asia (Teng-Khoen-Hing 1959; Sommer, Sugana, Djunaedi et al. 1978). It appears to result from prolonged deficiency of vitamin A in which impairment of rod function is succeeded by structural damage to the retina (Figure 8.10). Attention was recently drawn (McLaren 2004) to the potential value of XF in eye field sur- Figure 8.9: Bilateral corneal scars (leucomata XS) veys, especially in school-age children and adults. in an anemic and generally malnourished infant. The Figure 8.11 suggests that the lesion may not be as inferior situation of the scars is typical. rare as has sometimes been suggested. In view of the questions that are now being raised about the reliability of self-reporting of tion for XF, which is easily detectable by ophthal- night blindness, especially in areas where the phe- moscopic examination, has not been carried out. nomenon is not so common as to have a name in It was pointed out recently (McLaren 2004a) that the local language, it is surprising that examina- this sign might have been used more frequently in 8 Xerophthalmia 87

Figure 8.10: a) Numerous small yellowish-white lesions affecting most of this region of the retina. 24-year-old patient presented with night blindness and constricted visual fields. b) Two months after vitamin A treatment the lesions have largely disappeared (Sommer 1982). the past when xerophthalmia was a much greater and abnormal CIC (Lewallen, Taylor, Molyneux public health problem than at present. It should et al. 1998). Retinal changes and signs of raised be employed currently when high prevalence intracranial pressure responsive to vitamin A were rates are being reported in pregnant and lactating reported in a case of secondary VAD in the United women and school-age children (see earlier) and States (Panozzo, Babighian, Bonora 1998). when at the same time these high rates are being In conclusion, there are several general points questioned by some. that should be noted. Signs of xerophthalmia are Recently, a report from Malawi associated usually present in both eyes, but not necessarily to the whitening of the retina described in children the same degree. Keratomalacia may proceed very with cerebral malaria with low vitamin A status rapidly, noticeably within a matter of some hours

Figure 8.11: Prevalence of xerophthalmic fundus (XF) (Sommer 1982). 88 8 Xerophthalmia

rather than days. This is especially true in very involved, accompanying infection is probably the young children. In this age group keratomalacia rule. Frequently this obscures the classical picture may be present without any evidence of xerosis in of changes due to xerophthalmia. Unless this is the conjunctiva or cornea. It follows that a diag- remembered, the result may be underdiagnosis of nosis of vitamin A deficiency may be made in the xerophthalmia. absence of xerotic changes. When the cornea is 89

9 Mortality and Morbidity, Especially in Relation to Infections

INTRODUCTION these findings were not pursued further at that time. In explanation it should be pointed out that It can be instructive, on occasion, to look back epidemiological and statistical methodologies had in time and speculate about the reasons why events not advanced sufficiently in those times and there in history took the path they did. In Chapter 8 it was still great reticence on the part of governmen- was pointed out that recognition of xerophthalmia tal and other agencies to recognize the magnitude as a vitamin deficiency disease lagged behind that and severity of VADD. of several other diseases of nutritional etiology. In the present context it is a cause for surprise that widespread acceptance of the significant impact MORTALITY ASSOCIATED WITH of vitamin A on mortality and morbidity has come NON-CORNEAL XEROPHTHALMIA only in recent years (Semba 1999). Not long after the discovery of fat-soluble A the vitamin The significance of vitamin A deficiency in was dubbed “the anti-infective vitamin” (Green, child survival in the community was first sug- Mellanby 1928). This was on the basis of the sus- gested by a study of Sommer, Hussaini, Tarwotjo ceptibility of experimental animals fed on diets et al. (1983) that followed a point-prevalence deficient in the vitamin to infections resulting in survey of vitamin A deficiency in Indonesia (see death and of the response of humans with infec- later). The results, presented in Table 9.1, show tions to vitamin A. a significantly increased relative risk of death in Sommer and West (1996:pp19–61) reviewed those groups with evidence of xerophthalmia. The hospital-based reports of children with xerophthal- greatest risk is among those with both X1B and mia and severe PEM in whom mortality was sig- XN. It should be noted that only those with the nificantly increased in the presence of eye signs milder, non-corneal signs of xerophthalmia were accompanying general malnutrition. In some of included. these studies there was a fourfold (McLaren, Shi- It is probable that the mechanism of the great- rajian, Tchalian et al. 1965) or even greater mor- er mortality in those children who had clinical tality in children severely vitamin A deficient as VADD was in some way related to impairment compared with those equally generally malnour- in their immune response to an accompanying ished but with normal eyes. The implications of infectious disease. Whilst it is relatively easy to

Table 9.1: Mortality in relation to xerophthalmia (vitamin A) status (Sommer, Hussaini, Tarwotjo et al. 1983). Ocular status Child inter- Deaths Mortality Relative risk vals* (per 1000) Normal 19,889 108 5.4 1.0 XN (+), X1B (-) 547 8 14.6 2.7 XN (-), X1B (+) 269 6 35.5 6.6 XN (+), X1B (+) 215 10 46.5 8.6

* Examined at six three-monthly intervals 90 9 Mortality and Morbidity, Especially in Relation to Infections

ensure that all deaths are recorded in a study like As an indication of the magnitude of trials this, it is much more difficult to obtain accurate of this nature it should be noted that more than information as to the cause or causes of death. In 150,000 preschool-age children took part in the addition, in the circumstances that prevail in field eight trials described in outline below. All or studies it is usually not possible to make a defini- some of them have formed the basis of several tive diagnosis of any accompanying infections. meta-analyses that have been made. The results These matters are further discussed later when the of all the meta-analyses are in general agreement relationship between vitamin A deficiency and that improvement in the vitamin A status of defi- morbidity is considered. cient populations reduces significantly the overall preschool-age child mortality. Figure 9.2 is a summary of the results of one of these meta- VITAMIN A INTERVENTION AND analyses of the eight major mortality intervention MORTALITY IN YOUNG CHILDREN trials (Beaton, Martorell, Aronson et al. 1993). The overall reduction of mortality from this was 23%. The first of these trials was carried out in These results are remarkably consistent in Aceh, Indonesia (Sommer, Tarwotjo, Djunaedi view of the fact that the communities studied et al. 1986). A large dose of vitamin A (200,000 were very heterogeneous with regard to many IU) was given every 6 months. The initial report factors such as culture, ecology and disease pat- found a 34% reduction in mortality in the treated terns, and details of the interventions carried out group. Alternative analyses suggested a reduction differed considerably. of at least 40% and possibly >50% (Tarwotjo, It should also be noted that there are several Sommer, West et al. 1987; West, Pokhrel, Katz et reasons why the reported reductions in mortality al. 1991). Figure 9.1 shows some of the results may have underestimated the actual reductions. from this study. Some deaths may have been included that oc-

females, program villages

Figure 9.1: Cumulative mortality, by sex, of preschool-age children in Aceh, Indonesia. From Sommer, Tarwotjo, Djunaedi et al. 1986. 9 Mortality and Morbidity, Especially in Relation to Infections 91

Figure 9.2: Meta-analysis of the eight major community mortality intervention trials. For details, including explanation of fixed effect and random effects, see reference (Beaton, Martorell, Aronson et al. 1993). curred before the vitamin A intervention had a of this kind those who fail to participate, termed chance to act. In most studies children with xe- “noncompliers,” are usually those most in need of rophthalmia were identified, treated and excluded the intervention. Their mortality has been found from the main study. In the absence of the inter- to be greater than that of the control group (see vention these children would be expected to be Figure 9.3). among the most vulnerable. Periodic large-dose Knowledge of the level of noncompliance in vitamin A supplementation may not be the opti- any study is essential for assessing impact. It is mal means of improving vitamin A status in the also a reminder that when any measure is adop- long term. Consistent with this may be the high- ted for routine application there will always est reductions in mortality being reported from be those who for various reasons fail to partici- Bogor, Indonesia, and Tamil Nadu, India (see pate. Furthermore, they are likely to be the most Table 9.2), where vitamin A was provided in a at risk. more physiological way. Finally, each of the eight In a reassessment of most of these studies studies reported what is termed an “intention-to- Sommer and West (1996:34–37) found that in treat” analysis. In this it is assumed that a full in- children aged 6 months and older the reduction tervention was received by each subject supposed in mortality tended to increase in magnitude with to do so. It has been found (Tarwotjo, Sommer, age. The majority of deaths occurred in children West et al. 1987) that in most health interventions <2 years of age. These authors also discuss the 92 9 Mortality and Morbidity, Especially in Relation to Infections

Table 9.2: Major community mortality prevention trials (Sommer, West (1996:28)). Study Country Vitamin A supplement Reported Primary mortality reference reduction Aceh Indonesia Large dose every 6 mo 34% Sommer et al. 1986 Bogor Indonesia Vitamin A 45% Muhilal et al. 1988 fortified MSG * NNIPS Nepal Large dose every 4 mo 30% West et al. 1991 Jumla Nepal One large-dose follow- 29% Daulaire et al. 1992 up at 5 mo Tamil Nadu India Weekly RDA** 54% Rahmathullah et al. 1990 Hyderabad India Large dose every 6 mo 6% (ns) Vijayaraghavan et al. 1990 Khartoum Sudan Large dose every 6 mo +6% (ns) Herrera et al. 1992 VAST Ghana Large dose every 4 mo 19% Ghana VAST Study Team 1993

* Monosodium glutamate ** A dose providing the RDA was given once a week (ns) Not statistically significant

Program: no dose Program: dose Control

Figure 9.3: Mortality in the Aceh trial among children in treatment villages who did and did not receive their intended vitamin A dose, compared with children in control villages. Mortality was higher among children who were assigned to receive vitamin A but did not get it than for any other group, including children in control villages (Tarwotjo, Sommer, West et al. 1987). 9 Mortality and Morbidity, Especially in Relation to Infections 93

evidence for the effect on mortality of vitamin A tends to make more of an impact on measles and dosing in children under the age of 6 months. diarrhea than it does on respiratory infections in Under the age of 3 months 100,000 IU appeared this regard. to have a detrimental effect, but 50,000 IU was It should be noted that this seems to be consis- beneficial. In infants over the age of 3 months the tent with the results of studies that have attempted larger dose was effective. A single maternal dose to investigate the impact of vitamin A intervention of 300,000 IU has been shown to raise maternal on several aspects of the evolution of various infec- stores, vitamin A levels in breast milk and the se- tious diseases (see section on Morbidity later). rum retinol of breastfed children, and to reduce mortality in infants by about 30% (de Francisco, Yasui, Chakraborty 1994). CONTRIBUTION OF VAD TO CHILD Recently an analysis of the effect of Nepal’s MORTALITY vitamin A supplementation program on child mortality at age 12–59 months was made (Shyam, While subclinical vitamin A deficiency is Kim, Retherford 2005). It was shown that 100% much less fatal than overt xerophthalmia, it is community-level vitamin A coverage since the very much more widespread (see Chapter 12). child’s birth compared to no coverage reduced the Beaton, Martorell, Aronson et al. 1993 (see Fig- odds of dying by slightly more than 50%. ure 9.4), were unable to find any association in the field trials that have been cited above between relative risk (RR) of mortality and prevalence of CAUSE-SPECIFIC MORTALITY xerophthalmia. This may have been due in part to differences This was studied in four of the eight inter- in the reporting of xerophthalmia in the different ventions subject to meta-analysis and the results studies (McLaren 1991), but it probably indicates in Table 9.3 indicate that vitamin A intervention that xerophthalmia does not make a major contri-

Table 9.3: Cause-specific mortality, vitamin A supplementation community prevention trials (Sommer, West 1996:41). Symptoms/Diseases

Measles Diarrhea Respiratory

StudyC Vitamin A Control Vitamin A Control Vitamin A Control Tamil Nadu Deaths (n) 7 12 16 33 2 3 RRA 0.58 0.48 0.67 NNIPS 3 12 39 62 36 27 Deaths (n) 0.24 0.61 1.29/1.00B Jumla 3 4 94 129 18 17 Deaths (n) 0.67 0.65 0.95 Ghana VAST 61 72 69 111 47 45 Deaths (n) 0.82 0.66 1.00

A RR (relative risk): Cause-specific mortality rate of vitamin A group divided by rate in control group B Originally published results RR=1.29; reanalysis as an associated cause that recognizes other underlying causes RR≠1.00 C For details see Table 9.2 94 9 Mortality and Morbidity, Especially in Relation to Infections

Figure 9.4: Relative risk of mortality and prevalence of xerophthalmia. No relationship could be established (Beaton, Martorell, Aronson et al. 1993). bution to mortality within a community. The larg- ciency may be responsible for as many as 1.3 to est contribution is most likely to be attributable to 2.5 million deaths annually. In addition, PEM is subclinical vitamin A deficiency. considered to be a contributory cause of infant In Nepal a sustained reduction in child mor- and child deaths in nearly 50% of cases (Pelletier, tality with vitamin A was shown for the first time Frongillo, Schroeder et al. 1995). Recently much (Pokhrel, Khatry, West et al. 1994). Where defi- attention has been given to the importance in in- ciency is endemic vitamin A supplementation can fections of other micronutrient deficiencies, espe- achieve a rapid reduction in early childhood mor- cially of zinc (see Chapter 11). tality and a lower level of mortality can be sustained A recent account from Mozambique (Aguayo, as long as capsule coverage is adequate (>85%). Kahn, Ismael et al. 2005) of planning for a new In Malawi 377 HIV-negative women had se- vitamin A supplementation program for children rum retinol measured in the second and third tri- under the age of five, as a short- to medium-term mester of pregnancy. Their infants were observed strategy to assist in the reduction of a high under-5 from delivery until 12 months of age. Mothers mortality rate, is of interest. Low coverage (46%) whose infants died had serum retinol levels lower soon emerged as a critical problem. than mothers of the survivors. Infants born to By way of contrast, examination of the results mothers whose serum retinols were in the low- of a long-standing vitamin A supplementation est quartile had a threefold greater mortality risk program in Nepal with excellent coverage showed than those of mothers with serum retinols in the that in this way inequalities in child mortality higher quartiles (Semba, Miotti, Chiphangwi et by social class could be removed (Fiedler 2000; al. 1998). Bishai, Kumar, Waters et al. 2005) (see also Chap- It has been estimated (Humphrey, West, ter 15). In Zimbabwe Malaba, Iliff, Nathoo et al. Sommer 1992) that worldwide vitamin A defi- (2005) studied a group of HIV-negative mothers 9 Mortality and Morbidity, Especially in Relation to Infections 95

and their young infants. The effect of postpartum Measles generally takes a more serious form maternal and neonatal vitamin A supplementation in the generally undernourished child, resulting on infant mortality up to 12 months was studied. in more frequent and more serious complications No effect was shown in these mothers who were and a much higher death rate than in the well- shown to have adequate vitamin A status. nourished. This was shown in the community pre- Benn, Fisker, Diness et al. (2006) have drawn vention trials referred to already (see Table 9.3). attention to the adverse effect on mortality in girls Table 9.4 shows that in hospital the majority of in some programs where vaccines and VAS are deaths occur in children under the age of 2 years. combined. This important and complex issue is Dramatic reduction in mortality rates in hos- dealt with in Chapter 15 in a separate section (see pitalized cases of measles has been repeatedly p. 170). shown in those where treatment included vitamin A supplementation (Figure 9.5). In 2002 the UN General Assembly Special MEASLES MORTALITY AND VITAMIN A Session on Children adopted a goal to reduce TREATMENT deaths due to measles by half by the end of 2005, compared with 1999 estimates. This goal was Measles occupies a unique position among achieved; mortality due to measles was reduced common childhood infectious diseases and vita- by 60% from an estimated 873,000 deaths in 1999 min A deficiency. Of these diseases, it is in mea- to 345,000 in 2005. Nearly 7.5 million deaths sles alone that the infective agent, in this case a were prevented, with supplemental immunization virus, invades the eye. In serious cases it causes a activities (including vitamin A supplementation) damaging measles keratoconjunctivitis. Even im- and improved routine immunization accounting munization against measles has been shown to re- for 2.3 million of these prevented deaths (Wolf- sult in symptomless, minimal invasion of the cor- son, Strebel, Gacic-Dobo et al. 2007). The latest nea which may take months to disappear. This eye data (BMJ 2009) show that measles deaths world- pathology may predispose to corneal liquefaction wide fell by more than 78% between 2000 and in the presence of vitamin A deficiency. 2008 but in India the drop was only 46%. In

Table 9.4: Measles mortality – Cape Town Trial (Hussey, Klein 1990). Vitamin A Controls

Age Children Deaths Mortality Children Deaths Mortality Relative Risk (months) (%) (%) Control: Vitamin A <6 4 0 – 3 1 33.3 33 : 0 6–12 53 1 1.9 64 7 10.9 5.7 : 1 13–23 19 1 5.3 18 1 5.6 1.1 : 1

0–23 76 2 2.6 85 9 10.6 4.1 : 1

>24 16 0 – 12 1 8.3 8.3 : 0

Total 92 2 2.2 97 10 10.3 4.7 : 1 96 9 Mortality and Morbidity, Especially in Relation to Infections

Program: control Program: vitamin A

Figure 9.5: Measles case-fatality rates among hospitalized patients randomized to receive high-dose vitamin A (cod liver oil in the London trial) compared with those of their controls. Vitamin A supplementation reduced mortality by 50% in all three trials (Sommer, West 1996:53).

Africa the figure was 92%. “Delayed implementa- and were observed over 3½ years. Approximately tion of large-scale vaccination campaigns in equal numbers in three matched groups received India, the country with the majority of measles vitamin A (7,000 μg RE on a weekly basis), deaths, is largely accountable for this lack of β-carotene (7,000 μg RE on a weekly basis), or progress,” a spokesperson for Measles Initiative a placebo. said. Experts fear resurgence in measles mortality Maternal deaths from any cause during preg- if current strategies are not maintained. nancy or within 12 weeks of delivery were the end point. The preventive effects, expressed as relative risks, were 0.60 (40% reduction) for vi- VITAMIN A SUPPLEMENTATION AND tamin A and 0.51 (49% reduction) for β-carotene MATERNAL MORTALITY (see Table 9.5). It is suggested that the antioxidant effect of The preliminary results of the first trial of this β-carotene might have had some relationship to kind caused a considerable stir when they were the higher reduction rate in comparison with vi- reported at the XVIII IVACG Meeting in Cairo tamin A. There was neither reduction in fetal or (West, Khatry, Katz et al. 1997). The definitive infant mortality through until 6 months of age nor paper was published nearly 18 months later any impact on neonatal birth weight. (West, Katz, Khatry et al. 1999) together with an A preliminary report at the XIX IVACG editorial (Olsen 1999). An area of Nepal was cho- Meeting 1999 (Hakimi, Dibley, Suryono et al. sen for the study where VADD is known to be 1999) on a trial to show the effects of vitamin A common in pregnant women as well as in young and zinc supplements on maternal postpartum in- children. Well over 40,000 women participated fections showed that there was significant benefit 9 Mortality and Morbidity, Especially in Relation to Infections 97

Table 9.5: Impact of supplementation on mortality related to pregnancy up to 12 weeks postpartum (West, Katz, Khatry et al. 1999).

Placebo Vitamin A β-carotene Vitamin A or β-carotene Number of pregnancies* 7,241 7,747 7,201 14,948 Number of deaths 51 33 26 59 Mortality 704 426 361 395 (per 100,00 pregnancies) Relative risk (95% CI) 1.0 0.60 0.51 0.56 (0.37–0.97) (0.30–0.86) (0.37–0.84) P value <0.04 <0.01 <0.005

* Includes 157 pregnancies that were lost to follow up (43, 70, and 44 in the placebo, vitamin A, and β-carotene groups, respectively). in groups receiving vitamin A but zinc gave no tention than the much more numerous non-fatal improvement. ones. This may be partly because the latter have The results of two other large trials of a simi- been considered to be of only secondary impor- lar nature to that of the trial in Nepal have been tance. It may also be because they are much more reported (Palmer and Stewart 2009). difficult to document. A fatality is a discrete, one- In Bangladesh, where maternal mortality is off event; a morbidity or disability is quite vague. lower (than in Nepal) and maternal vitamin A Morbidity may be acute or chronic, mild or se- status better, there was no evidence of an impact vere. (Klemm, Labrique, Christian et al. 2008). A third The relationship of morbidity due to VADD trial from Ghana (Kirkwood, Hurt, Amenga- and infection is complicated by several further Etego et al. 2010), where maternal mortality and points. There is an apparent synergistic relation- vitamin A status were similar to those in the ship between impaired vitamin A status and in- Bangladesh trial, found no significant impacts fections; each seems to increase the risk of the on pregnancy-related mortality and pregnancy- other (see also Chapter 13). Other nutritional associated morbidities. The combined results of deficiencies, especially PEM of mild degree, are these trials confirm that, while potentially impor- often also present and this makes interpretation tant in high-mortality, deficient settings, a global of results difficult. Episodes of infectious diseases policy for maternal VAS is not warranted. vary greatly from case to case and may also differ to a considerable extent from one location to another. For many reasons their recognition in VITAMIN A SUPPLEMENTATION AND different studies may be based upon very differ- INFECTIOUS MORBIDITY ent criteria. Finally, although the phenomenon has been noted for many years, it is only relatively Mortality and morbidity are related but clear- recently that research workers have given due rely quite distinct. All disease states carry some gard to the fact that the acute-phase reaction as- element of morbidity. In a medical context the sociated with infections also causes a depression word relates in an unspecific way to illness. Some of serum retinol. This casts doubt on the value of instances of morbidity may result in death, but serum retinol as an indicator of vitamin A status that aspect is normally dealt with separately. Fatal under such circumstances. This subject will be diseases have, until recently, been given more at- considered in more detail in Chapter 13. 98 9 Mortality and Morbidity, Especially in Relation to Infections

Vitamin A deficiency has been induced in munity trial of the effect of vitamin A supple- experimental animals for many decades. The uni- mentation on diarrhea in young children in Brazil form experience has been that this nutritional de- revealed much enhancement of beneficial effects ficiency, more than any other, is associated from (Andreozzi, Bailey, Nobre et al. 2006). a relatively early stage with the development of Zinc supplementation has been shown to have infections, especially those affecting epithelial a marked effect in diarrheal disease in under- tissues. Most prominent have been those of the nourished children (see Chapter 11). In a trial in upper and lower respiratory tracts, parts of the which vitamin A was also given it was shown gut, and the genito-urinary tract (Moore 1957). that groups receiving zinc had better outcomes in These infections often became apparent early on every way measured compared with those receiving in the deficiency, long before the changes of xe- vitamin A or those serving as controls. Parameters rophthalmia occurred (McLaren 1959). measured were stool output, cumulative stool weight, body weight gain, and rapidity of clinical Diarrheal diseases recovery (Khatun, Malek, Black et al. 2001). This is probably a less well defined group of diseases than the others to be considered here. Respiratory infections There is a variety of bacterial and viral organ- The term respiratory infection is usually applied isms that have been isolated from patients with to diseases affecting the lower respiratory tract; diarrhea. Isolation does not necessarily mean in general this means one form or another of that an organism was responsible for the diar- bronchitis or pneumonia. Fever, cough, rhonchi rhea. In addition, even with full laboratory facili- or rales (sounds detected by stethoscope) are cus- ties, in some instances no organisms have been tomarily required to be present for the diagnosis isolated. Furthermore, the difficulties of defining to be made. what constitutes a case are probably greater for Animal studies, clinical and autopsy studies diarrhea than for any other infection or group of in children, and observational studies in human infections. populations all show an association between Cross-sectional studies of impaired vitamin A vitamin A deficiency and respiratory infec- status and various infections have been carried tions. Respiratory disease prevalence increased out in many developing countries (Sommer, in a linear fashion with increasing severity of West 1996:62–98). The strongest association xerophthalmia in a large hospital study (see in most was with diarrhea, especially when it Figure 9.6). was persistent, chronic, or severe. The relation- However, meta-analysis from mortality and ship is not clear from these studies. Which came morbidity intervention trials showed in general first? a lack of any impact of vitamin A supplementa- Intervention trials cast some light on this last tion on acute lower respiratory tract infections point. Vitamin A supplementation does not appear (ALRIs). There was some indication that there to have an effect on mild diarrhea (usually defined might be reduction in the severity, if not the as 3 or 4 stools/day). As was noted earlier, it does duration, of respiratory symptoms. It has been reduce diarrhea-related deaths. As expected then, pointed out that increase in cough in the supple- it also reduces the incidence of severe diarrhea mented group might be a favorable response and the benefit is proportional to the number of (Herrera, Fawzi, Nestel 1996). Strength of the stools/day. This evidence supports the suggestion cough reflex could be beneficial in removing made long ago (see Chapter 13) that in many in- infective material. These conclusions have been stances diarrhea is an integral part of VADD. supported in general by subsequent studies from Recently the application of more appropriate other countries. Treatment with vitamin A of re- and powerful statistical methods to a large com- spiratory syncytial virus (RSV) infection had no 9 Mortality and Morbidity, Especially in Relation to Infections 99

Figure 9.6: The prevalence of respiratory disease among Indonesian children presenting to the Cicendo Eye Hospital increased with the severity of their xerophthalmia (P<0.01 for linear trend) (Sommer 1982). significant benefit (Dowell, Papic, Bresee et al. tions. Although this was not investigated it might 1996). have related to the acute phase response (APR) In a meta-analysis of all adequate studies that (see Chapter 7). were available at the time (Vitamin A and Pneu- A study was undertaken in Tanzania (Fawzi, monia Working Group 1995) no major reduction Mbise, Spiegelman et al. 2000) to determine the occurred in the mortality or morbidity associated effect of vitamin A supplementation on the risk with respiratory infections as a result of vitamin A of young children with pneumonia developing supplementation. A meta-analysis of randomized diarrhea and acute respiratory infection. Supple- controlled trials of vitamin A supplementation for mentation resulted in less risk of severe watery both diarrheal disease and respiratory infections diarrhea, but higher risk of cough and elevated re- concluded that high-dose vitamin A supplementa- spiratory rate. The latter was confined to children tion should not be recommended routinely for all who were seronegative for HIV. preschool-age children (Grotto, Mimouni, Gda- In southern India, Coles, Kanungo, Rah- levich et al. 2003). A more recent meta-analysis mathullah et al. (2001) studied pneumococcal (Brown, Roberts 2004) involved 5 studies with nasopharyngeal colonization in young infants. acceptable criteria and included a total of 1,067 In their experience, Streptococcus pneumoniae intervention and 1,110 control children. Main out- is the most frequent bacterial cause of morbidity come measurements were: time to normalization and mortality in this group. Among 464 newborns of fever; respiratory rate and oxygen dependence; studied, samples collected at ages 2, 4, and 6 time to discharge; and mortality. There were no months were positive in 54%, 64.1%, and 70.2% significant differences for any measurement. respectively. The presence of maternal night Another study (Rodriguez, Hamer, Rivera et al. blindness increased the risk of colonization three- 2005) failed to find improvement with vitamin A fold. It was considered that the high rate of pneu- supplementation, but better results were associ- mococcal carriage in early infancy might explain ated with higher initial serum retinol concentra- the high risk for pneumonia. 100 9 Mortality and Morbidity, Especially in Relation to Infections

An interesting hypothesis has been put for- Measles ward by Sempertegui (2001). Based on impair- The evidence for the association between ment in lung function as measured by FEV1 vitamin A deficiency and measles, in its various (forced expiratory volume in one second) and aspects (see earlier and Chapter 13), is stronger FVC (forced vital capacity) in children with evi- than for any other infectious disease. Diagnosis dence of vitamin A deficiency it is proposed that probably has a firmer basis, as a single organism this might predispose them to respiratory infec- is responsible and the clinical picture is usually tions. typical. Measles is the only specific disease for It will not be possible to make real progress in which there is firm evidence that vitamin A sta- this area until the precise diagnosis of respiratory tus influences morbidity and mortality (Sommer, diseases is clarified. What they call a “descrip- West 1996:62–98). tive necropsy study” has been reported by Chintu, Hospital-based studies in South Africa dem- Mudenda, Lucas et al. (2002) from Lusaka Uni- onstrated that both the complications (see Table versity Teaching Hospital, Zambia. Autopsies 9.6) and outcome (see Table 9.7) in measles are were performed on 137 boys (93 HIV-1 positive) significantly improved by vitamin A supplemen- and 127 girls (87 HIV-1 positive) aged 1 month tation. to 15 years. The four most common respiratory Outcome during the prolonged recovery after disease diagnoses were acute pyogenic pneumo- discharge home is also markedly improved by a nia (39.1%), Pneumocystis carinii pneumonia further dose of vitamin A (see Table 9.8). (27.5%), tuberculosis (18.8%), and cytomegalo- Both measles and vitamin A deficiency are virus infection (20.2%). HIV-1 positivity predis- known to impair immune response and this composed to the occurrence of P. carinii pneumonia bination helps to explain the serious nature of the and cytomegalovirus infections. Eighty-eight per- disease in most developing countries. With the cent of P. carinii infections were in cases younger widespread take-up of measles immunization as than 12 months. Detailed pathology is bound part of the WHO Expanded Program on Immuni- to vary widely from place to place and would zation (EPI) there is already evidence of a marked seem to render invalid blanket statements fall in corneal blindness associated with measles being made at present. The same may apply for (Foster, Yorston 1992). The combination of vita- diarrheal diseases, and the nutritional status, in- min A supplementation with measles immuniza- cluding but not confined to vitamin A, will be tion is discussed later (see Chapter 15). important. Some work suggests that a single dose of 200,000 IU (210 mmol) retinol does not enhance

Table 9.6: Measles complications – Cape Town vitamin A controlled treatment trial (Hussey, Klein 1990). Complication Vitamin A Controls Relative Risk N=92 N=97 Vitamin A : Controls Pneumonia >10 days 12 29 0.44 (0.24, 0.80) Diarrhea <10 days 8 21 0.40 (0.19, 0.86) Post-measles croup 13 27 0.51 (0.28, 0.92) Requiring airway intervention 3 9 0.35 (0.10, 1.26) Herpes stomatitis 2 9 0.23 (0.05, 1.06) Intensive care 4 11 0.38 (0.13, 1.16) Hospital days 10.54 15.24 P<0.004 9 Mortality and morbidity, especially in relation to infections 101

Table 9.7: Hospital outcome – Durban measles vitamin A controlled treatment trial (Coutsoudis, Broughton, Coovadia 1991; Coutsoudis, Kiepiela, Coovadia et al. 1992). Outcome Vitamin A Placebo (N = 29) (N = 31) Duration (days) – pneumonia 3.8 ± 0.4 5.7 ± 0.8 Duration (days) – diarrhea 3.2 ± 0.7 4.5 ± 0.4 Duration (days) – fever 3.6 ± 0.3 4.2 ± 0.5 Clinical recovery in < days 28 (96%) 20 (65%) IMS on day 8* 0.24 ± 0.15 1.37 ± 0.40 * Integrated morbidity score the immune response nor is it effective in reduc- (Donnen, Dramaix, Brasseur et al. 1998). In one ing measles complications (Rosales, Kjolhede, study in Indonesia (Humphrey, Agoestina, Wu Goodman 1996) and doubling the dose may be et al. 1996) neonatal vitamin A supplementation required. Where access to healthcare and im- reduced infant mortality and the prevalence of se- munization is good and where VADD is mild to vere respiratory infection among young infants. moderate, vitamin A supplementation may have The most serious nonfatal complication of no effect on general morbidity (Ramakrishnan, measles is blindness. A recent review (Semba, Latham, Abel et al. 1995). High-dose treatment Bloem 2004) calculates that measles affects about (one dose of 200,000 IU vitamin A) did not im- 30 million children/year, of whom about 1 mil- prove morbidity in hospitalized, malnourished lion die. It is suggested by these authors that mea- children in Congo, while daily low dosage (5,000 sles is now the single leading cause of blindness IU) reduced the incidence of diarrheal disease but among children in low income countries; about had no effect on acute lower respiratory infections 15,000 – 60,000/year.

Table 9.8: Post-hospital outcome – Durban measles vitamin A controlled treatment trial (Coutsoudis, Broughton, Coovadia 1991; Coutsoudis, Kiepiela, Coovadia et al. 1992). 6 Weeks 6 Months

Vitamin A Placebo Vitamin A Placebo (N = 24) (N = 24) (N = 20) (N = 16) Weight gain (kg) 1.29 ± 0.17 0.90 ± 0.14 2.89 ± 0.23 2.37 ± 0.20 Diarrhea episodes 6 12 3 6 Score/episode – diarrhea 2.17 ± 0.31 2.25 ± 0.25 1.67 ± 0.67 2.17 ± 0.31 URI* episodes 7 9 3 8 Score/episode – URI 1.71 ± 0.28 2.66 ± 0.17 2.00 ± 0.58 2.37 ± 0.18 Pneumonia episodes 5 6 0 3 Score/episode – pneumonia 4.40 ± 0.98 6.67 ± 0.67 – 6.67 ± 0.67 Chest x-ray score >3 2 6 0 3 IMS ** 2.21 ± 0.45 5.74 ± 1.17 0.60 ± 0.22 4.12 ± 1.13

* Upper respiratory infections ** Integrated morbidity score 102 9 Mortality and Morbidity, Especially in Relation to Infections

Although up-to-date data are not available in in measles, as compared with other infections suitable form to settle this important issue it seems (Enwonwu, Phillips 2004). The hypothesis sug- more likely that xerophthalmia is still the major gests that measles, more than any other child- single cause of childhood blindness, although hood infection, causes hyporetinemia and marked greatly reduced in recent years. In the 1990s it hypercortisolemia. The latter promotes a shift in was estimated that there were worldwide about lymphocyte profile from TH1 to TH2 and influ- 300,000 young children blind from xerophthalmia; ences hepatic acute phase protein response through ten times that number had clinical xerophthalmia enhancement of the activity of interleukin 6. It also (i.e. night blindness and Bitot’s spot) (see Chapter antagonizes effects of retinoids at the cellular and 1). It is not possible to know whether that number transcriptional levels. Glucocorticoids decrease ex- has changed since. The world’s population is still pression of receptors for retinoic acid. It is there- rapidly growing; data collection is of better qual- fore suggested that increased tissue levels of corti- ity and wider in coverage; on the other hand many sol could explain the well-documented benefits of national and part-national control programs have vitamin A in acute infections, especially measles. been started in this period. In infants under the age of 9 months with About two decades ago there was general measles the presence of pneumonia on admission agreement that about 5–10 million cases of xeroph- and low vitamin A status (by conjunctival impres- thalmia (all stages) were occurring annually. Of sion cytology) were independently associated these about 10% (0.5–1 million) led to blindness. with mortality (overall 16%) (Courtright, Fine, About half of those affected would not survive, Broadhead et al. 2002). leaving a minimum of about 250,000 cases of blindness – very much greater than the range suggested for measles (see also Chapters 12 and 13). HIV/AIDS A recent study in the slums of Kolkota (Cal- cutta), India gives an impression of the very un- Evidence is accumulating that vitamin A satisfactory situation facing many children in the status in HIV infection is of considerable impor- ever growing shanty towns throughout the de- tance. Serum retinol levels are depressed and in veloping world (Ray, Mallik, Munsi et al. 2004). proportion to the severity of infection (Beach, Twenty clusters, each of 250 children under 5 Mantero-Atienza, Shor-Posner et al. 1992). Mor- years old who had a history of measles in the past tality in AIDS patients is higher in those with year, were studied. Only 19.7% had been immu- lower serum retinol levels (Tang, Graham, nized. Only 16.9% had received vitamin A before Kirby et al. 1993). Vitamin A supplementation measles and 8.6% afterwards. of a group of HIV-positive injection drug users A closer look has been taken at the results had no significant effect on HIV load or CD4 of the use of vitamin A supplementation in the lymphocyte count (Semba, Caiaffa, Graham et al. treatment of measles (Souza, Souza 2002). In a 1995). meta-analysis there was no significant reduction It was at first reported that mothers with in mortality in the vitamin A group when all stud- HIV infection have an increased chance of pas- ies were pooled. There was a 64% reduction with sing on the infection to their offspring if their two doses of 200,000 IU each. Two doses of water- serum retinol is low (Semba, Miotti, Chiphangwi dispersible vitamin A were associated with an et al. 1994). However, this has not been subse- 81% reduction (only 48% with oil). A single dose quently supported by three controlled trials of vita- of 200,000 IU was not effective. These effects min A supplementation of HIV-positive pregnant were greater in children under the age of 2 years. women, and an extensive review more recently A new hypothesis has been put forward to (Wiysonge, Shey, Sterne et al. 2005). No signifi- explain the greater beneficial effect of vitamin A cant differences were found in mother-to-child- 9 Mortality and Morbidity, Especially in Relation to Infections 103

transmission between control and supplemented progression are concerned. An exhaustive meta- groups (Humphrey 2000; Blaner, Gamble, Burger analysis of vitamin A supplementation and moth- et al. 1997). One study was unable to show any er-child transmission of HIV infection has shown beneficial effects of vitamin A supplementation that earlier hopes were not confirmed (Shey, on pregnancy outcomes and T-cell counts in HIV- Brocklehurst, Sterne 2002). However, evidence infected women, although these occurred with from supplementation trials conducted among multivitamins (Fawzi, Msamanga, Spiegelman et HIV-infected children suggests some potentially al. 1998). Maternal vitamin A deficiency during beneficial uses of vitamin A during HIV-1 infec- HIV infection is reported to predispose to growth tion. For example, there was a reduction in the failure (Semba, Miotti, Chiphangwi et al. 1997) occurrence of diarrhea-related morbidity and and to increased infant mortality (Semba, Miotti, mortality. Chiphangwi et al. 1995). Vitamin A supplementa- A disturbing report from Tanzania (Fawzi, tion of HIV-infected infants significantly reduced Msamanga, Hunter et al. 2002) found that among morbidity (Coutsoudis, Bobat, Coovadia et al. 1,078 HIV-infected pregnant women vitamin A 1994) and reduced mortality among both HIV- alone versus supplementation with vitamins B, infected and non-infected malnourished children C and E increased the rate of HIV transmission. (Fawzi, Mbise, Hertzmark et al. 1999). Multivitamin supplementation (vitamins B, C and An interesting study was carried out in Mom- E) reduced child mortality and HIV-transmission. basa, Kenya (Mostad, Overbaugh, DeVange et Further research to clarify the potential clinical al. 1997), on factors that affect HIV-1 shedding application of vitamin A supplementation is still in cervical and vaginal secretions. Both hormon- warranted. al contraceptive use and vitamin A deficiency Three studies have investigated genital shed- were associated with increased risk of vagi- ding of viral cells with differing results. No renal shedding. After adjustment for CD4 count, lation was found between genital viral burden in severe VAD, moderate VAD, and low vitamin A women infected with HIV and vitamin A status status were associated with 12.9, 8.0, and 4.9-fold (French, Cohen, Gange et al. 2002). In women increased risk of vaginal shedding, respectively. infected with both HIV and HSV (herpes simplex The public health implications of these results in virus) vitamin A supplementation did not influ- the control of HIV/AIDS spread are evident. ence genital HSV shedding (Baeten, McClelland, Some earlier conclusions about an associa- Corey et al. 2004). In a trial with vitamin A and/or tion between low serum retinol levels and certain multivitamins B-complex, C and E, significantly aspects of HIV infection need to be reconsidered more women had detectable levels of HIV-1 in in the light of what was said earlier about the in- cervicovaginal lavage compared with the mul- fluence of the acute phase response (APR) (see tivitamin group. The results raise concern about Chapter 7). the use of vitamin A supplementation in these cir- Kennedy, Kuhn, and Stein (2000) undertook cumstances. an exhaustive review of vitamin A and HIV in- Mastitis, or increased mammary permeabil- fection and came to certain conclusions. In-vitro ity, in an HIV-infected mother appears to increase studies have provided evidence for the role that the risk of mother-to-child-transmission (MTCT) retinoic acid (RA) plays in suppressing HIV-1 of HIV through breast milk. Vitamin A reduces viral expression. However, supplementation tri- the prevalence of bacterial mastitis, and vitamin A als have not provided consistent evidence of deficiency in HIV-infected women is associated therapeutic benefits of vitamin A. Randomized with increased risk of MTCT. Vitamin A supple- placebo-controlled trials conducted on adults mentation, surprisingly, increased the risk of have not been shown to confer benefit as far as MTCT (Dorosko 2005). vertical transmission, immune status, and disease In Uganda, (Melikian, Mmiro, Ndugwa et 104 9 Mortality and Morbidity, Especially in Relation to Infections

al. 2001) it was found in infants infected with effect of periodic vitamin A supplementation on HIV that serum retinol and carotenoids were mortality and morbidity of HIV-infected chil- both predictive of growth retardation, but only dren in Uganda (Semba, Ndugwa, Perry et al. carotenoids were related to risk of death. It 2005). 181 HIV-infected children were enrolled would seem that the acute phase response was at 6 months and randomized to receive vitamin A operative in these sick children. Carotenoids supplementation, 60 mg retinol equivalent, or pla- and retinol might have been affected differently. cebo every 3 months from age 15 to 36 months. Carotenoids are much more indicative of recent Mortality and some morbidity features were sig- dietary intake, which would have been very low nificantly reduced in the vitamin A supplemented near death. group. As in other contexts attention has been turned Another long-awaited contribution to this in recent years to supplementation with multi- complex area of field research on vitamin A and micronutrients. In particular three studies have HIV infection is a massive trial in Zimbabwe with been carried out by members of the Department of a total of over 14,000 mother-infant pairs (Hum- Nutrition of the Harvard School of Public Health. phrey, Iliff, Marinda et al. 2006). It is accompa- Villamor, Msamanga, Spiegelman et al. (2002) nied by a critical editorial by Fawzi (2006). From studied HIV-infected women during pregnancy. these papers and earlier evidence cited above the Multivitamin supplementation, not including vit- conclusions section of the paper may be quoted as amin A, significantly improved weight gain dur- reflecting a consensus view at present. “Targeted ing the 3rd trimester. Multivitamins, including vitamin A supplementation of HIV-positive chil- vitamin A, improved weight gain during the second dren prolongs their survival. However, postpartum trimester. Fawzi, Msamanga, Wei et al. (2003) maternal and neonatal vitamin A supplementation gave women with HIV supplementation dur- may hasten progression to death in breastfed chil- ing pregnancy and lactation. Children of women dren who are PCR (polymerase chain reaction) who received multivitamins (not including vitamin negative at 6 weeks. These findings raise concern A) had lower risk of diarrhea and higher CD4+ about universal maternal or neonatal vitamin A count. Those receiving vitamin A had children supplementation in HIV-endemic areas.” with a lower risk of pneumonia, but diarrhea and CD4+ cell count were unaffected. Fawzi, Malaria Msamanga, Spiegelman et al. (2004) monitored In an early study no relation was found be- progression of disease and mortality. Multivita- tween vitamin A status and malarial infection min supplements (without vitamin A) delayed the (Binka, Ross, Morris et al. 1995). A later study progression of HIV and resulted in higher CD4+ in Papua New Guinea (Shankar, Genton, Semba and CD8+ counts and lowered viral loads. Adding et al. 1999) found a substantially lower incidence vitamin A to the multivitamin regimen reduced of clinical episodes of Plasmodium falciparum the benefit with regard to some of the end points. malaria (P=0.001), parasite density (P=0.09), and Vitamin A alone was no better than placebo. prevalence of spleen enlargement (P=0.01) in chil- Miller, Stoltzfus, Iliff et al. (2005) reported that it dren attending health facilities. No such effects appears that while postpartum vitamin A supple- were found in Plasmodium vivax infection. The mentation did not affect mortality it reduced se- role of iron and zinc supplementation in malaria vere morbidity. Neonatal and maternal vitamin A is considered in Chapter 11. supplementation had a beneficial effect on ane- Shankar (2000) in a comprehensive review mia in infants of both HIV positive and negative pointed out that malaria had been on the increase mothers and should be considered for inclusion in for the past three decades. Morbid episodes stand programs to prevent infantile anemia. at about 200 million and deaths at about 203 mil- Recently a report has been published on the lion/year. Malnutrition increases both mortality 9 Mortality and Morbidity, Especially in Relation to Infections 105

and morbidity. Supplementation with vitamin A Rojanawatsirivet et al. 2005) and desbutyl-ben- or zinc can substantially reduce clinical attacks. flumetol (Samal, Rojanawatsirivet, Wernsdorfer Although iron supplementation may aggravate et al. 2005). some aspects of malarial infection, the author argues that these low-cost measures should be Tuberculosis added to current methods of malaria prevention Pulmonary tuberculosis has long been known to and treatment. be a common and serious infection throughout the Dreyfuss and Stoltzfus (2000) demonstrated developing world. It is on the increase there and the interrelationships between hookworm infesta- in the developed world, especially in the danger- tion, malaria, and vitamin A deficiency, iron defi- ous form that is multi-drug resistant. ciency and anemia in pregnant Nepalese women. A considerable body of both animal experi- Hookworm infestation was the strongest predic- mental and clinical human research on the re- tor of iron status. Low serum retinol was most lationship between vitamin A deficiency and strongly associated with mild anemia. P. vivax the spread and progress of the disease had been malaria and hookworm infestation were stron- carried out throughout the latter part of the 20th gest predictors of moderate to severe anemia. The century. This has been well summarized in the need for planners to take account of all these ad- chapter on nutrition and tuberculosis by Whalen verse factors is stressed. and Semba (2001). Unfortunately this topic had In Mozambique, children hospitalized with escaped our attention when the first two editions severe malaria, in addition to routine treat- of the Manual were written. More recent research ment, received either vitamin A or a placebo and is summarized below. were followed for six weeks after discharge In childhood tuberculosis it has been shown (Varandas, Julien, Gomes et al. 2001). Deaths that sCD30 (a putative marker of type-2 cytokine- in the two groups were 7 and 13% respectively. producing cells) is greatly increased. Malnutrition Although the difference was not statistically sig- made the levels even higher and vitamin A thera- nificant, routine supplementation with vitamin A py caused them to modulate over time. would seem justified under these circumstances. Pulmonary tuberculosis and HIV-infection Some rather puzzling results have recently frequently occur together; especially in mal- been published of the effects of vitamin A and nourished patients. Perhaps not surprisingly, this anti-malaria treatment on erythropoietin produc- concurrence leads to even greater impairment of tion in severely anemic and vitamin A deficient vitamin A status. Vitamin A supplementation, as children in Zanzibar (Cusick, Tielsch, Ramsan judged by raised serum retinol levels, appears to et al. 2005). Vitamin A decreased erythropoi- be of benefit in both diseases (Mugusi, Rusizoka, etin production, inflammation was reduced rap- Habib et al. 2003). idly, and iron was mobilized from stores and new A double-blind, placebo-controlled trial of pa- erythrocytes were produced. tients with pulmonary tuberculosis was carried out Vitamin A supplementation may reduce pla- in Indonesia (Karyadi, West, Schultink et al. 1997). cental parasitemia, believed to play a part in the The study group of newly diagnosed patients adverse effects of malaria infection in pregnancy received both vitamin A 1,500 RE, as retinyl ac- (Cox, Staalsoe, Arthur et al. 2005). etate, and 15 mg zinc, as zinc sulfate, daily for Increasingly, resistance of P. falciparum to 6 months. In this group sputum conversion and antimalarial drugs is posing a serious public resolution of X-ray lesion area occurred earlier. health problem. Two recent papers from the Several studies have shown that patients with same research group in Austria suggest that vita- active tuberculosis have lowered serum retinol min A in vitro inhibits schizont maturation in the levels. In error the conclusion has frequently presence of artemisinin (Thriemer, Wernsdorfer, been drawn that this is indicave of subnormal 106 9 Mortality and Morbidity, Especially in Relation to Infections

vitamin A status. The explanation, relating to the study in Rwanda (Semba, Bulterys, Munyeshuli acute phase response (APR) to infection and et al. 1996) found a case fatality rate of 20%, inflammation, is given in Chapter 7. most patients had low serum retinol levels, and Interestingly, a recent study (Ramachandran, mean CD4 lymphocyte percentage was higher Santha, Garg et al. 2004) of serum retinol levels and mean CD8 lymphocyte percentage lower in in sputum-positive pulmonary tuberculosis pa- children with meningitis compared with reference tients in comparison with household contacts and populations. Vitamin A intervention appears not healthy ‘normals’ demonstrated this point. Before to have been carried out yet. treatment serum retinol levels were significantly lower in tuberculosis patients. After anti-tuberculosis treatment, to which vitamin A supplementa- REMARKS tion was not added, serum retinol levels returned to normal. Despite a great deal of research there is still In Indonesia a double-blind, placebo-con- much we do not yet know about the precise controlled study of vitamin A and zinc supplementa- tribution that VAD makes to morbidity in infection in persons with tuberculosis (Karyadi, West, tious diseases. This will probably remain so until Schultink et al. 1997) was carried out. After anti- we are in possession of more accurate methods tuberculosis treatment plasma zinc concentrations of defining the various manifestations of morbid- were not altered. Plasma retinol concentrations ity and until some existing discrepancies have were significantly higher in the micronutrient been resolved. These latter have been discussed group than in the placebo group after 6 months by Kirkwood (1996). Epidemiological studies, (P<0.05), as were sputum conversion (P<0.05) described here, have indicated that vitamin A sup- and resolution of X-ray lesion area (P<0.01). De- plementation reduces child mortality and severe ficiencies of micronutrients, including vitamin A, morbidity and that most of the reduction is due to in adult patients with pulmonary tuberculosis are the effect on diarrhea and measles. These studies associated with wasting and higher HIV load (van have not shown a similar effect on acute lower Lettow, Harris, Kumwenda et al. 2004). respiratory tract infections (ALRIs). No explanation exists at present for this difference, especially in view of the evidence that supplementation Other infections with zinc benefits both diarrheal and respiratory infections (see Chapter 11). Observational stud- Urinary tract ies and animal experiments tend to show a ben- Bacteriuria was more than fourfold greater in eficial effect of supplementation on lung disease. children with xerophthalmia than in those without Moreover, in clinical trials on measles much of and the prevalence increased with the severity of the impact of vitamin A on mortality and morbid- xerophthalmia (Brown, Gaffar, Alamgir 1979). ity has been related to reduction in incidence and severity of associated pneumonia. Notwithstand- Otitis media ing, all are agreed that undernourished children Children with abnormal CIC had a significantly treated for pulmonary infections should be given greater risk of middle ear infection in a study supplemental vitamin A to improve their status. in Truk (Lloyd-Puryear, Humphrey, West et al. 1989).

Meningococcal disease Meningococcal disease is a major cause of child morbidity and mortality in sub-Saharan Africa. A 107

10 Other Effects of VAD

In the first edition of the Manual it was sug- (Sommer, West 1996:150–62). The best results gested that much of the evidence for the impor- were obtained when daily ingestion of vitamin A tance of VAD in this chapter was based on work in amounts close to recommended levels of intake in experimental animals and lacked firm evidence was used. as far as human disease was concerned. In recent Fortification of sugar with vitamin A was used years, there has been a steady increase in our in Guatemala to study the effect on iron status knowledge of the role of vitamin A, especially over a prolonged period (Sommer, West in the form of retinoic acid (see Chapter 6) in 1996:150–62). Table 10.1 shows the mean many systems of the body. We can with ever changes sampled at 6-monthly intervals over a increasing confidence use the term vitamin A 2-year period. It was concluded that initially en- deficiency disorders (VADD) to cover all pos- hanced iron mobilization led to lower iron stores. sible consequences of the lack of dietary intake This probably triggered increased efficiency of of vitamin A. iron absorption, which would gradually restore iron stores. These in turn would make iron available for hemopoiesis, at a level greater than pres- ANEMIA ent before sugar fortification began. An entirely different mechanism of vitamin It has long been known that deficiencies of A-iron interaction is suggested by the work of iron and vitamin A tend to exist together in popu- Layrisse, Garcia-Casal, Solano et al. (2000). lation groups. This might be expected to occur They measured double-labeled iron absorption because deficiencies in natural dietaries are likely in nearly 200 healthy adults. The object was to affect more than a single nutrient. Controlled to determine whether there was any interaction trials have been carried out in several countries between vitamin A or β-carotene and inhibitors in which vitamin A supplementation has brought of iron absorption. The presence of vitamin A about a significant increase in hemoglobin level increased iron absorption threefold from rice,

Table 10.1: Change in serum retinol and iron status in preschool children following sugar fortification with vitamin A in Guatemala (Mejia, Arroyave 1982). Duration of study (months) 6 12 18 24 No. of subjects analyzed 77 75 46 51 Retinol (mg/dL) +5.1* +5.2* +3.6* +2.5* Serum iron (mg/dL) +4.5* -3.6 +13.1* +9.1* Total iron-binding capacity (mg/dL) +18.3* -8.8 -7.8 -13.6 Transferrin saturation (%) +0.6 -1.0 +3.8* +2.2* Serum ferritin (ng/ml) -3.3* +2.1 +5.5* +4.9* * P ≤ 0.05 108 10 Other Effects of VAD

2.4-fold from wheat, and 1.8-fold from corn. to counteract any adverse effects of iron. The β-carotene increased iron absorption almost authors suggest that the beneficial effects of vita- threefold from the three cereals tested. It is sug- min A on iron status might be related to reduced gested that vitamin A and β-carotene may form a levels of infection. complex with iron reducing the binding to phytate The prevalence of anemia associated with vi- or polyphenols. tamin A deficiency has not been estimated but is It has been suggested that vitamin A may likely to be quite high, as both deficiencies are improve anemia by increasing the synthesis of common. Especially vulnerable are young chil- erythropoietin, a hormone produced in the kidney dren and women in the reproductive age period. that stimulates red cell production. Semba, Kum- In the Pacific region, Lloyd-Puryear, Humphrey wenda, Taha et al. (2001) could not confirm this and West (1989) showed that the prevalence of in anemic pregnant women. However, recently subclinical vitamin A deficiency, as evidenced Zimmermann, Biebinger, Rohner et al. (2006) by abnormal conjunctival impression cytology found that vitamin A supplementation in children (CIC), was inversely proportional to the hemo- with poor vitamin A and iron status increased globin level (see Figure 10.1). erythropoietin levels as iron was mobilized from The mechanism of vitamin A related anemia existing stores. is unclear. It has consistently been shown that There is some evidence that administration vitamin A deficiency restricts the release of iron of iron to treat anemia may exacerbate coexisting from the depots, resulting in evidence of iron infection. Many pathogenic bacteria require iron overload and, in the absence of increased absorp- and might proliferate as they compete for it. In tion of iron, eventually in anemia. Although in- one study (Northrop-Clewes, Paracha, McLoone fection also causes iron overload, it can occur in et al. 1996), an increase in dietary intake of the absence of infection. vitamin A under these circumstances appeared Animal work showed that a fall in hemoglo-

Hemoglobin (g/L)

Figure 10.1: Relationship between hemoglobin level and prevalence of vitamin A deficiency assessed by conjunctival impression cytology (CIC) in children 3–6 years of age in Truk, Micronesia (Lloyd-Puryear, Mahoney, Humphrey et al. 1991). 10 Other Effects of VAD 109

bin due to feeding low-iron diets was accompa- specific restrictions be applied, as in the case of nied by a lowering of plasma retinol, but by an experimental animals. Furthermore, there are fre- increase of hepatic retinyl esters. Low iron status quently accompanying infections and these influ- seems to interfere with hepatic release of vitamin A ence nutrient status in several ways. Appetite is (Rosales, Jang, Pinero et al. 2000). In animals it usually impaired, pyrexia tends to increase the was also shown that marginal vitamin A intake demand for nutrients, and absorption may be im- during lactation caused iron levels in milk to fall paired by gut infestations and infections. by depressing the action of iron transporters in Another difficulty is the multiplicity of ways mammary glands (Kelleher, Lonnerdal 2005). in which growth is customarily assessed. Skeletal A recent large human study in Zimbabwe measurements, such as height and head circum- (Miller, Stoltzfus, Iliff et al. 2006) raises some ference, cannot decrease during undernutrition questions. Nearly 1,600 infants were allocated, but undergo a decrease in the velocity of their rate with their mothers, to one of four treatment of increase. There is no absolute loss. Height/age groups – vitamin A to mothers and infants; vita- is usually employed to indicate “stunting.” min A to mothers and placebo to infants; placebo On the other hand, measurements of soft tis- to mothers and vitamin A to infants; placebo to sues like muscle and fat, including body weight, both. Over 8–14 months no effect of vitamin A on skinfold thickness and muscle circumference hemoglobin or anemia was observed. It is possible measurements, may involve absolute loss of tis- that in this, as in a similar study (van den Broek, sue. Weight/height indicates “wasting,” a more White, Flowers et al. 2006), initial vitamin A acute state of growth retardation. Some of these levels were mostly within the normal range. measurements are subject to quite large observer Since 2000, with the discovery of a peptide errors. While body weight can be measured with hormone produced by the liver, later named hep- considerable accuracy, there has to be some doubt cidin, it has become clear that iron hemostasis about its interpretation, especially in children. is maintained through meticulous regulation of Body composition is known to vary considerably circulating hepcidin levels. Many factors may in malnutrition; fat and muscle decrease but water influence this process, of which vitamin A status increases proportionately. Consequently weight is one. Details of the mechanisms involved are may not accurately reflect healthy body mass. the subject of intense current research (Wrighting, The relationship between anthropometric status Andrews 2006). and mortality is mentioned in Chapter 9. Children with corneal xerophthalmia have consistently been observed to be stunted. This GROWTH has perhaps best been documented in the countrywide survey in Indonesia carried out in 1978–79 The first clearly evident response of a young (Sommer 1982) (see Figure 10.2). The degree animal to dietary restriction of any kind is a de- of vitamin A deficiency was proportional to the crease in the velocity of growth. Young animals, degree of retardation of growth. There is also evi- and children, have relatively higher nutritional re- dence of frequent wasting in xerophthalmic chil- quirements partly because they need nutrients for dren. In this more acutely deficient state, mortal- growth as well as maintenance. Vitamin A is no ity may make interpretation difficult. exception (see Chapter 6), and in fact the grow- It appears that periodic large-dose vitamin A ing organism appears to be especially sensitive to supplementation has a significant impact on VAD. growth in children with xerophthalmia, but not The situation in the young child is infinitely on those with evidence of subclinical deficiency. more complicated than that in experimental However, three studies in which preformed animals. It cannot be so clearly defined nor can vitamin A was consumed in adequate amounts 110 10 Other Effects of VAD

Figure 10.2: Children with corneal disease (X2/X3) were shorter than children who had Bitot’s spots (X1B) (P<0.01); children with X1B were shorter than their matched controls (P<0.01). There were no significant differences in height between randomly sampled children without xerophthalmia (normal random sample) and matched controls (Sommer 1982).

regularly showed slight improvement in linear retinol levels, and in the long term affected main- growth (Sommer, West 1996:163–88). ly height. Soft tissue growth improvement tended Several studies of the effect of vitamin A to occur earlier on. supplementation on growth have failed to dem- A possible mechanism for an effect of vitamin onstrate any effect (Ramakrishnan, Latham, Abel A on growth is provided by a study carried out et al. 1995a; Kirkwood, Ross, Arthur et al. 1996; in a group of short prepubertal children (Evain- Fawzi, Herrera, Willett et al. 1997). It may be Brion, Porquet, Thérond et al. 1994). Fasting pointed out that “absence of evidence is not plasma retinol correlated with nocturnal growth evidence of absence.” This general principle in hormone (GH) secretion but not with stimulated interpretation of results should be applied to GH secretion. Dietary intake of vitamin A was the negative results found in these studies. A lower in children with low nocturnal GH secre- group in India (Bahl, Bhandari, Taneja et al. tion, and supplementation with vitamin A for 3 1997) observed improved weight gain with months increased their nocturnal GH secretion. vitamin A supplementation only in the summer In Sudan Sedgh, Herrera, Nestel et al. (2000) season, when subclinical vitamin A deficiency studied more than 8000 initially stunted children peaks. Studies in Nepal (West, Le Clerq, Shrestha aged 6–72 months. Their intake of vitamin A over et al. 1997) and Indonesia (Hadi, Stoltzfus, a 24-hour period and their height and weight Dibley et al. 2000) both reported improved were measured. The observations were repeated growth with supplemental vitamin A. The effect at 6-month intervals for 18 months of follow up. was seen most in children with the lowest serum Children who were initially in the highest quin- 10 Other Effects of VAD 111

tile grew on average 13 mm more than those in (Hadi, Dibley, West 2004). 1,405 children aged the lowest quintile. Carotenoid intake was pro- 6–48 months were given high dose vitamin A portional to improvement in height. Dietary ef- every 4 months. A significant effect of vitamin A fects on growth appeared to be strongest among supplementation on linear growth occurred at all very young children and upon those who had seasons in children with a low burden of respira- been most malnourished. Other factors, includ- tory infections (<21 days with respiratory illness). ing breastfeeding and socioeconomic status, were In each season, the highest effect was found in also important. children with low burden of respiratory infection In recent years, a great deal of evidence has and low vitamin A intake (<400 RAE/d). There been obtained that suggests that impaired growth was no benefit for linear growth from vitamin A in utero and during the first year or two of life is supplementation in children with both high bur- strongly associated with the development in later den of respiratory infection and high vitamin A life of such chronic degenerative diseases as coro- intake, regardless of season. Under these condi- nary heart disease, hypertension, central obesity, tions, it is evident that vitamin supplementation and diabetes type II (Barker 1998). While the as- did play an important part in counteracting sea- sociation seems definite, the mechanisms are un- sonal growth retardation. clear. Work from Brazil (Rondo, Abbott, Tomkins Research in Gambia and India points to the 2001) found a significant direct relationship be- possible part that impairment of gut integrity may tween several growth measurements at birth and play (Thurnham, Northrop-Clewes, McCullough serum retinol in cord blood. Several other studies et al. 2000). These workers showed progressive have not confirmed such a relationship. deterioration of integrity of the gut mucosa after It has become clear that supplementation with introduction of weaning food, which often in- vitamin A alone will not bring about optimal im- troduces bacterial contamination. A dual sugar provement in growth. Other micronutrients must permeability test was used. Impairment was least also be taken into account, just as they are in rela- from April–June, when vitamin A intake was tionship to control of infections (see Chapter 11). maximum, due mainly to this being the mango A review from Mexico concluded that zinc season. Evidence to support this association was had the greatest effect on children’s growth, fol- obtained by the finding that measurements of the lowed to a lesser extent by vitamin A and iron. acute phase protein alpha-1-antichymotrypsin (ACT) These beneficial effects were confined to children were lowest between April and June. This co- with severe deficiency (Rivera, Hotz, Gonzalez- incided with the lowest prevalence of subclini- Cossio et al. 2003). cal infection, best gut integrity, and most rapid The same group (Rivera, Gonzalez-Cossio, growth. More such studies would be very valu- Flores et al. 2001) found that feeding a formula able. containing 1.5 RDA of 13 vitamins and 6 essen- It is appropriate to end this section with ref- tial elements to infants daily for an average period erence to the wider picture of the importance of of just over 12 months resulted in significantly general malnutrition in child health and survival. greater increase than in those receiving placebo. A recent study (Pelletier, Frongillo 2003) has The main lesson from this unusual study is per- quantified the effects of changes in general mal- haps that the majority of young children in de- nutrition, as measured by child weight-for-age, veloping countries are stunted because they are on changes in child survival in 59 developing undernourished with regard to many nutrients. countries. They concluded that “continued reduc- An important contribution to our understand- tion in mortality will require improved targeting ing of the complex interaction of infection, diet of selective interventions and general nutritional and seasonal growth response was made recently improvement to the most marginal populations.” in a study of Indonesian preschool-age children Vitamin A supplementation is an important part 112 10 Other Effects of VAD

of this program, but, as recognized above, only tem. Phagocytic cells also belong to this system. a part. There is one aspect of growth and develop- It has been proposed (Ross, 1996) that there ment that has received little attention: a possible are broadly speaking two hypotheses to explain effect of vitamin A deficiency on brain growth and the protective action of vitamin A against infec- intellectual development. In a trial of vitamin A tion (see Table 10.2). and folic acid during pregnancy in Indonesia the The epithelial cell linings of organs and tis- results were negative (Schmidt, Muslimatun, West sues are generally considered to have a defensive et al. 2004), but it surely merits more thorough function. They have been considered as provid- research. ing a first line of defense in resistance to infection taking place. Ross argues that this is a misconcep- tion and that they primarily have an offensive role. IMMUNE RESPONSE On the other hand the immunologic response is a defensive response against infection once it has Interest in the possibility of vitamin A being taken place. The beneficial effect vitamin A sup- involved in the function of the immune system plementation has on childhood diarrhea may be was first derived from the association of vita- determined by the regulatory effect supplementa- min A deficiency with infectious diseases. More tion has on the mucosal immune response in the recently, it has been shown experimentally that gut (Long, Santos, Garcia et al. 2006). retinoids can stimulate immune responses. The available evidence for each of these theo- There are two distinct responses to exposure ries from studies of human populations is strictly to antigens; humoral and cellular (cell-mediated) limited (see Chapter 6). It would be expected that immunity. Humoral immunity results from anti- offensive protection would reduce the incidence body production mediated by B-lymphocytes, of infections. A defensive mechanism would be which is often T-lymphocyte-dependent. Cellular likely to reduce the duration or the severity of an immunity is mediated by T-lymphocytes. There infection. As was shown in Chapter 6, the few are two main effector mechanisms: cytolytic T- studies that addressed these points tended to sug- lymphocyte (CTL) responses, and delayed-type gest that the impact of vitamin A supplementation hypersensitivity (DTH) responses. has been on duration and severity of infection. There are also natural killer (NK) cells, which This supports the immunologic response hypoth- are part of the innate or nonspecific immune sys- esis.

Table 10.2: Two contrasting hypotheses for the protective role of vitamin A against infection. (After Ross 1996). Epithelial barrier hypothesis Immunologic response hypothesis The basic reaction is OFFENSIVE The basic reaction is DEFENSIVE This protects against invasion This enhances the body’s defense by infection against pathogens Structural integrity is of Functional integrity is most important, greatest importance as is cell differentiation Consequently resistance to establishment Consequently resistance to proliferation of of infection will be reduced by VAD infection will be reduced by VAD Major effect of vitamin A intervention Major effect of vitamin A intervention will be will be decreased incidence of infection decreased duration and/or severity of infection 10 Other Effects of VAD 113

Most of the evidence from both animal and As in other aspects of vitamin A activity, at- human studies suggests that the various aspects tention is beginning to be focused on the effect of humoral immunity are little affected by vita- of deficiency of other micronutrients as well (see min A deficiency. Cell-mediated immunity, how- Chapter 11). Wieringa, Dijkhuizen, and van der ever, has been shown to be markedly impaired. Ven-Jongekrijg et al. (2003) measured cytokine The production and maturation of lymphocytes production in healthy Indonesian infants. A strong are reduced by lack of vitamin A. In a study in positive correlation was found between produc- Indonesia it was found that the ratio of T-cells tion of interferon-γ, interleukin-12, and inter- bearing CD4+ and CD8+ antigens was lower in leukin-6. In contrast, interleukin-18 was negative- the peripheral blood lymphocytes of xerophthal- ly correlated to interleukin-12 and interleukin-6. mic children compared with non-xerophthalmic Interferon-γ was significantly lower in vitamin A controls (Semba, Muhilal, Ward et al. 1993). deficient infants than in controls. Vitamin A After vitamin A supplementation the proportion deficient infants had significantly higher neopter- of CD4+ to CD8+ T-cells and the percentage of in levels, reflecting increased macrophage activ- naive CD4+ T-lymphocytes increased. ity. In zinc deficient infants there was a tendency The precise mechanisms whereby vitamin A towards a lower level of production of all cyto- plays a part in maintaining normal function of the kines. immune response are not yet fully understood. In intervention studies, six groups of infants The active form at the cellular level appears to be received either iron (10 mg/day), zinc (10 mg/ retinoic acid, but it is possible that other metabo- day), iron and zinc (10 mg/day each), β-carotene lites of retinol may also be active. (2.4 mg/day), zinc and β-carotene (10 and 2.4 mg/ day respectively), or placebo. Iron supplementa- Table 10.3: Main immune response elements that may tion enhanced type 1 immune responses. Zinc and be regulated by retinoids. (After Semba 1998). β-carotene had immunomodulatory effects that are opposite to those of iron.  Keratinization Early studies with provitamin A carotenoids  Mucin production showed results similar to vitamin A, but more  Hematopoiesis recent work using non-provitamin carotenoids, (influences many types of cells) such as lycopene and lutein, showed these to be  Apoptosis (programmed cell death) very active in both cell-mediated and humoral im-  Function of neutrophil, natural killer, munity (Chew, Park 2004). monocyte/macrophage, various lympho- In mice, Stephensen, Jiang, and Freytag cyte and other cells (2004) recently demonstrated that vitamin A de-  Immunoglobulin production ficiency impairs both T-helper type 1 (Th1) and  Production of numerous interleukins type 2 (Th2)-mediated immune responses. At the time of initial antigen exposure, VAD enhances the development of IL-10-producing Th2 or T Semba (1998) has reviewed the role of vita- regulatory cells, and diminishes the development min A and related retinoids in immune function. of Th1 memory cells. Table 10.3 is a simplified version of the table in Finally, in the last few years several contro- his paper. A study of the effect of vitamin A sup- versial papers from the research group based in plementation on cytokine responses in pregnancy Denmark that has been investigating immuniza- and lactation showed suppression of proinflam- tion programs in Guinea-Bissau in West Africa matory type 1 immune responses in the presence for many years have appeared (Benn, Lisse, Balé of vitamin A deficiency (Cox, Arthur, Kirkwood et al. 2000; Kristensen, Aaby, Jensen et al. 2000). et al. 2006). These have been criticized and commented upon 114 10 Other Effects of VAD

by others (e.g., Fine 2000; Folb 2001). The first THYROID FUNCTION issue relates to the claim they make that some vaccines, particularly measles and BCG, have a Until recently, this area of possible impair- more beneficial effect on child survival than do ment of function resulting from vitamin A defi- others, such as diphtheria, tetanus, pertussis, or ciency has not been explored adequately. It has polio. At present agreement has not been reached been shown that in goitrous children the severity and immunization programs should certainly con- of vitamin A deficiency is proportional to thyroid tinue. volume and concentrations of TSH and thyro- A second issue directly involves the use of globulin. Vitamin A deficiency severity was also vitamin A supplementation together with immu- a strong predictor of higher concentrations of to- nization (Benn, Balé, George et al. 2002; Benn, tal T4. Supplementation with vitamin A reversed Balé, Sommerfeldt et al. 2003, with comments by these abnormalities (Zimmermann, Wegmuller, Stephensen 2003; McLaren 2004). The hypoth- Zeder et al. 2004). esis put forward here is, in essence, that the effect of VAS depends on the routine vaccines with which it is distributed. Vitamin A, for many years, REPRODUCTIVE SYSTEMS has been known to exert such adjuvant activity (Dresser 1963). This important issue is pursued Studies in experimental animals have shown further in Chapter 15 when the prophylaxis and that vitamin A is necessary for the proper func- management of infectious diseases in the control tioning of both male and female reproductive of VADD is considered. organs. The subject has been reviewed by Eskild A critical review of the known effects of vi- and Hansson (1994). In the male, vitamin A de- tamin A supplementation on immune responses ficiency impairs spermatogenesis and interferes and correlation with clinical outcomes (Villamor, with the functions of the Sertoli and Leydig cells. Fawzi 2005) shows how many uncertainties re- There is growing evidence that vitamin A, in main in this field. some form, is required for every stage in the reproductive process in the female. This is in addition to the implication of both deficiency and ex- NORMAL AND DISEASED SKIN cess of vitamin A during the organogenetic period in the production of malformations of the fetus The subject of skin has been extensively re- discussed elsewhere (see Chapters 6 and 14). viewed by Vahlquist (1994). Retinol and several other retinoids have been detected in the skin, mainly in the epidermis. Some years ago, when BONE clinical signs and symptoms were relied upon for the diagnosis of nutritional deficiency disease, Vitamin A deficiency in cattle has a marked hyperkeratosis of the skin around hair follicles effect on the modeling of bone. Most notably was frequently attributed to vitamin A deficiency. this has led to compression of the optic nerve in This change did not correlate with xerophthalmia the optic foramen, leading to blindness. Nothing or serum retinol. At present, the general view is comparable has been reported in man. However, that the skin does not undergo any characteristic there is increasing evidence that in man, over a clinical change in vitamin A deficiency. period of many years, the ingestion of a higher In recent years, numerous retinoids have been than average consumption of vitamin A may be synthesized for use in the treatment of a variety of associated with bone abnormalities in the elderly dermatological disorders (see Chapter 14). (see Chapter 14). 115

11 Interaction of Vitamin A and Other Micronutrients

INTRODUCTION field, there is a great deal that we do not know (McLaren 2008). Both during and for many years after what was called The Vitamin Era (McLaren 2004a), it was customary to think of each vitamin as hav- VITAMIN A AND IRON ing one particular function, at least so far as human health and nutrition was concerned. Thus Despite the need for a more detailed under- iron prevented the most common form of anemia, standing of the interrelation between vitamin A vitamin D cured rickets, vitamin C scurvy, and and iron, it has been concluded that the preven- vitamin A blindness. Although it was understood tion of VAD should also be considered when iron that people on poor diets usually had a variety of supplementation is given in the control of nutri- deficiencies, it was easy to lose sight of this fact tional anemia (IVACG, 1998). when these micronutrients alone were seen to be The previous chapter looked at the effects of doing such a good job at curing their associated vitamin A supplementation on anemia and iron diseases. status. In the present context, most studies relate More recently these severe states of defi- to the results of joint vitamin A and iron supple- ciency have been largely brought under control, mentation. only to reveal much more widespread marginal, Muslimatun, Schultink, West et al. (2001) but important, states of deficiency underneath, as found that weekly supplementation of pregnant it were. Perhaps not surprisingly, such marginal women with iron and vitamin A improved hemo- deficiencies very frequently coexist, and it is globin concentration; serum ferritin fell, however, logical that multiple deficiencies demand a mul- and it was concluded that this resulted from im- tiple approach. Hence the need to devote several proved iron utilization. sections of this book to the interaction between, In Bangladesh it was shown that adding vita- and joint action of vitamin A and other micronu- min A to the weekly supplementation of iron and trients. folic acid given to nonpregnant, anemic teenage In theory, the task has barely begun. The girls significantly improved the response (Ahmed, number of computations and permutations of the Khan, Akhtauzzaman 2005; Ahmed, Khan, Jack- actions and interactions of more than 40 micronu- son 2001). trients is astronomical. It would take courage for When the modified relative dose response any scientist to say that most could be ruled out as test (MRDR) was used to assess vitamin A sta- not significant for public health nutrition. Little is tus in a small group of pregnant Indonesian known about the majority, even at the experimen- women (Tanumihardjo 2002) the MRDR was tal level. Human trials have already been carried depressed by supplementation with both iron out in which diets were supplemented with pre- and vitamin A, but iron status was improved. cise amounts of as many as 13 micronutrients at In a particularly interesting recent study men- a time. Multiple supplementation has not always tal and psychomotor development was tested resulted in enhanced effects over those of single in Indonesian infants whose mothers had been nutrients. Sometimes these effects have even been supplemented with vitamin A and iron dur- adverse. Even though a start has been made in this ing pregnancy (Schmidt, Muslimatun, West et 116 11 Interaction of Vitamin A and Other Micronutrients

al. 2004). The subjects were considered to be and cough with fever, whereas zinc given alone moderately deficient in vitamin A and iron had no effect. There was no report on combined at the beginning of the test, and in this case, no therapy. This subject in which the demonstrated significant improvement resulted from supple- beneficial role of vitamin A supplementation is mentation. It is clear that this important and com- being ignored is discussed further in Chapter 15. plex subject requires much more thorough inves- The whole area of multi-micronutrient sup- tigation. plementation tends to become overcomplicated An interesting approach to the investigation by the use of supplementation with varying com- into the cause(s) of anemia was adopted recently binations of the three micronutrients: vitamin A, in northeastern Thai schoolchildren (Thurlow, iron and zinc. Lymphocyte responsiveness has Winichagoon, Green et al. 2005) which may be been enhanced with vitamin A and zinc (Kramer, beneficially replicated elsewhere. In over 500 Udomkesmalee, Dhanamitta et al. 1993). Growth schoolchildren aged 6–12.9 years a battery of in height and weight was improved by vitamin A serum tests were carried out. Importantly, those but not by zinc (Smith, Makdani, Hegar et al. with the elevated C-reactive protein concentra- 1999). An interesting report claimed a better he- tions indicative of infection were excluded. The matologic response in anemic women when vita- prevalence of anemia was 31%. Hemoglobinopa- min A and zinc were added to iron than when iron thies, retinol concentration <0.70 μmol/L and age was given alone or with vitamin A (Kolsteren, were the major predictors of hemoglobin concen- Rahman, Hildebrand et al. 1999). More recently tration. Only 16% of the HbAA and HbAE chil- Muñoz, Rosado, López et al. (2000) have shown dren were iron deficient. The contribution of iron that supplementation with iron and zinc improved deficiency to anemia was low. serum retinol levels, and it seems likely that there will be many more studies like these in the near future. VITAMIN A AND ZINC A trial of vitamin A and zinc in undernourished children with persistent diarrhea was under- This relationship has been known for some taken in Bangladesh (Khatun, Malek, Black et al., time but has only recently begun to be explored 2001). Those children receiving zinc had lower in the context of VADD. Many enzymes are zinc- stool output, lower cumulative stool weight, dependent and among them is retinol dehydro- greater weight gain, and more rapid clinical re- genase, which is involved in rod function. As a covery. Vitamin A supplementation, however, result, it has been found that some cases of night gave no benefit. blindness that are resistant to vitamin A may re- A similar study in slum children in Bangla- spond to zinc. Zinc deficiency may also interfere desh (Rahman, Vermund, Wahed et al. 2001) with the synthesis of retinol-binding protein. gave rather different results. Combined vitamin A A number of recent studies have demon- and zinc supplementation was more effective strated that supplementation with zinc has a benin reducing persistent diarrhea and dysentery eficial effect in diarrheal and some other infec- than either zinc or vitamin A alone. In the case tious diseases (Bhutta, Black, Brown et al., 1999; of respiratory infection, zinc given alone was Shankar, 1999) very similar to that of vitamin A. found to significantly increase the risk. This The only trial to date of both vitamin A and zinc was not entirely unexpected, as zinc is known to supplementation in diarrheal disease and respira- enhance microbial growth, function and virulence. tory tract infections comes from Mexico (Long, It also inhibits some aspects of cellular immunity Montoya, Hertzmark et al. 2006). and facilitates pulmonary tissue injury. Combined Where vitamin A supplementation was given supplementation with vitamin A and zinc signifi- alone there was an increase in diarrheal disease cantly reduced respiratory tract infections. 11 Interaction of Vitamin A and Other Micronutrients 117

Also in Bangladesh, Mahanlabis, Lahiri, Paul pneumonia when zinc was added to vitamin A et al. (2004) found that in boys (but not girls) supplementation. treated with zinc, recovery rates from severe acute Dijkhuizen, Wieringa, West et al. (2004) lower respiratory infection were significantly supplemented pregnant women with iron and fo- higher than for those receiving vitamin A, which late, in combination with zinc, and/or β-carotene had no effect when compared with a placebo. or neither, and the mother-infant pairs were fol- Further studies in Bangladesh by the same lowed for 6 months after delivery. Supplementa- group failed, perhaps not surprisingly, to dem- tion with β-carotene and zinc significantly im- onstrate an increase in growth after short term proved serum retinol levels of both mothers and supplementation with vitamin A and zinc (Rah- infants up to 6 months postpartum, and reduced man, Tofail, Wahed et al. 2002), but found that the prevalence of vitamin A deficiency in infants dual supplementation improved vitamin A status by >30%. Breast milk retinol concentrations only whilst vitamin A given alone did not (Rahman, increased in women who had received zinc. Wahed, Fuchs et al. 2002). This study may have In children given oral cholera vaccine supple- implications for control programs. In the National mentation with zinc, improved seroconversion to Vitamin A Week campaign 4–6 months previous- vibriocidal antibody was observed. This was not ly, it had been found that 38% of child recipients the case with vitamin A (Albert, Qadri, Wahed et were still vitamin A deficient. al. 2003). Children with diarrhea and respiratory infec- In undernourished children in northeast tions received combinations of vitamin A and/or Brazil it was found that improvement in intesti- zinc or a placebo (Rahman, Vermund, Wahed et nal permeability, as measured by the lactulose/ al. 2001). Combined zinc and vitamin A supple- mannitol ratio, occurred only in those receiving mentation was more effective in reducing persis- both vitamin A and zinc supplementation (Chen, tent diarrhea and dysentery than either vitamin A Soares, Lima et al. 2003). or zinc alone. Zinc given alone increased respira- A recent study of immunoregulatory cyto- tory illness, but the addition of vitamin A reduced kines in Indonesian infants deficient in vitamin A this adverse effect. and zinc (Wieringa, Dijkhuisen, West et al. 2004) More evidence for the beneficial effects of showed that vitamin A deficiency impaired pri- combining vitamin A and zinc supplementation marily Th1 (cell-mediated) response and zinc de- comes from Nepal, Indonesia, and India. In the ficiency mainly Th2 (humoral) response. first, acute diarrhea in children responded to dosing with zinc, and the result was not influenced by concomitant vitamin A administration (Arne, VITAMIN A, IRON, AND ZINC Krisna, Rajiv et al. 2002). In the second, a case- SUPPLEMENTATION control study of tuberculosis, combined supplementation improved the effects of treatment after Amongst nonpregnant women with iron de- 2 months and resulted in earlier sputum smear ficiency anemia who received combined supple- conversion (Karyadi, West, Schultink et al. 2002). mentation of iron, vitamin A and zinc (Kolsteren, In the third, in about 2,500 New Delhi slum chil- Rahman, Hildebrand et al. 1999), only in the dren aged 6–30 months, the incidence of pneu- group receiving zinc and vitamin A as well as iron monia was significantly reduced only in the group did hemoglobin levels rise more than with iron that received zinc supplementation (Bhandari, alone. Bahl, Taneja et al. 2002). The effect of combinations of iron and zinc A small study in Kolkota (Calcutta) India supplementation on indicators of vitamin A (Mahanlabis, Wahed, Khaled, 2002) showed no status was studied in a group of preschool chil- beneficial effect on measles accompanied by dren (Munoz, Rosado, Lopez et al. 2000). After 118 11 Interaction of Vitamin A and Other Micronutrients

6 months, improvements were observed in all OTHER MULTI-MICRONUTRIENT groups, but with certain differences: zinc was COMBINATIONS associated with higher plasma retinol and transthyretin, but not RBP; iron was associated with Selenium is an essential trace element, re- increased plasma retinol, RBP, and transthyretin quired in μg amounts daily, and both deficiency whereas supplementation with zinc and iron was and excess can cause disease. Baeten, Mostad, associated with increased plasma retinol, but not Hughes et al. (2001) found that 11% of wom- RBP or transthyretin. en examined in Kenya were deficient (serum The effect of zinc and vitamin A was studied selenium <85 μg/L). Those deficient in selen- in young children who were receiving iron ther- ium were found to be more than three times apy for anemia and who also had diarrhea (Alar- more prone to genital mucosal shedding of HIV- con, Kolsteren, Prada et al. 2004). Adding zinc 1-infected cells. VAD was also found to increase to iron therapy increased hemoglobin response, shedding. improved iron indices, and had a positive effect Great interest is centered around the possible on diarrhea. Vitamin A produced no additional ef- relationship between low birth weight and disease fect. in later life. In Nepal, Christian, Khatry, Katz et al. (2003) studied a large group of mothers and infants and found that antenatal supplementation VITAMIN A AND IODINE with folic acid and iron modestly reduced the risk of low birth weight. Additional multi- VADD and IDD (iodine deficiency disorders) micronutrients brought no additional benefit. frequently occur concurrently in developing All groups received vitamin A, which served countries. In rats deficient in both of these as a placebo. micronutrients it was shown that vitamin A sup- The same group (Christian, West, Katz et al. plementation, independent of iodine repletion, 2005) found that a multi-micronutrient mix, con- reduced thyroid hyperstimulation and gland size sisting of iron, folic acid, and vitamin A as well (Biebinger, Arnold, Langhans et al. 2007). The as 11 other micronutrients, may reduce risk of mechanism was probably through the effects of some obstetric complications. However, the risk vitamin A on pituitary TSH-β gene expression. of obstructed and prolonged labor, possibly from increased birth size, may rise. Also from the same group (Katz, Christian, VITAMIN A, RIBOFLAVIN AND OTHER Dominici et al. 2006) comes a statistical comment MICRONUTRIENTS that could have important implications for the analysis and interpretation of data from this kind There is some anecdotal evidence that ribo- of study. The importance of examining whether flavin plays a part in both retinal rod function and there is a constant or variable treatment effect in anemia. Inclusion of riboflavin in the supple- across the distribution of birth weight is demon- mentation of deficient pregnant women improved strated. Further results from this group indicate the response of anemia in China (Ma, Schouten, that maternal supplementation with folic acid or Zhang et al. 2008) and of dark adaptation, mea- folic acid, iron and zinc reduced the risk of kid- sured by pupillary threshold (see Chapter 8) in ney dysfunction and, to some extent, metabolic Nepal (Graham, Haskell, Pandey et al. 2007) in syndrome among children at 6-8 years of age the presence of iron and folic acid supplementa- (Stewart, Christian, Schulze et al. 2009). tion. A study from Mexico (Moriaty-Craig, Ra- makrishnan, Neufeld et al. 2004) raises an important question. In nonpregnant women they found 11 Interaction of Vitamin A and Other Micronutrients 119

that iron supplements given alone improved even should, be sustained on a long term basis. hemoglobin status significantly more than did a Fortification may be a more sustainable approach. multi-micronutrient mix containing iron at the It is likely that multi-micronutrient deficien- same level. Micronutrient interactions might have cies are common in vulnerable communities, as had an adverse effect on iron absorption. recently shown in Nepal (Jiang, Christian, Khatry An unusual, but interesting, example of the et al. 2005), and therefore many more studies, in a fortification of salt with multi-micronutrients number of variations, may be expected in the near comes from Switzerland (Zimmermann, Weg- future. When the Johns Hopkins Group pooled mueller, Zeder et al. 2004). Microcapsules of salt the results of its two trials of antenatal micronu- were fortified with 30 μg iodine, 2 mg iron, and trient supplementation in Nepal (Christian, Kha- 60 μg vitamin A/g salt and given to goitrous try, Katz et al. 2003; Osrin, Vaidya, Shrestha et al. school children. Status indices improved signifi- 2005), they showed that birthweight was significantly over a 10 month period of supplementation cantly higher in the supplemented groups, while in those receiving the mixture of vitamin A, iodine neonatal mortality was also significantly higher. and iron. A similar study took place in Thailand Subsequent correspondence in the literature left in which a seasoning powder was supplemented most issues unresolved. Such unexpected results (Winichagoon, McKenzie, Chavasit et al. 2006). posed difficult questions. Recently the results of the largest multi- In a broader context an NIH panel in the US is micronutrients supplementation trial so far were advising against multi-micronutrient supplements published. Infants aged 6–12 months received dif- for treating various chronic diseases, and research ferent combinations of nutrients in four countries: discourages the taking of antioxidant vitamins South Africa, Vietnam, Peru and Indonesia. The for the prevention of preeclampsia (Jeyabalan, results of each country study were published sep- Caritis 2006). More research is needed to inves- arately and were remarkably similar. For simplic- tigate the long-term consequences of low micro- ity, the pooled data are referred to here (Smuts, nutrient intakes. Hooper, Ashton, Harvey et al. Lombard, Benade et al. 2005). Overall the best (2009) very recently have developed a new results were obtained in the group which received methodology for assessing potential biomarkers 11 micronutrients daily with supplements given of the status of a number of micronutrients. in the form of foodlets, which have been defined Much attention has been given recently to the as “a crushable hybrid between a tablet and a wa- significance for an organism’s age of the length ter miscible food.” This group produced greater of the telomere part of the gene structure of cells. gain in weight, but not height, than the placebo Xu, Parks, De Roo et al. (2009) found that on group. Anemia also improved in this group, but average the relative telomere length of leukocyte the results were not clear-cut. Although other sim- DNA was 5.1% longer among daily multivitamin ilar studies are planned, with larger supplements, users than non-users. it is questionable whether daily dosing can, or 120 121

12 Global Occurrence

PREVALENCE OF VADD figures have been reproduced here and the authors are grateful to WHO for their permission to do so. The magnitude of the problem of VADD There is a bewildering array of indicators throughout the world, within nations, and in cer- to choose from in order to assess the vitamin A tain regions of a nation, is clearly of paramount status of populations at risk of VAD across the importance for the implementation of measures world. It should be recognized that these data do aimed at controlling the problem. Identification of not come from some massive study that has been VADD in the first place and monitoring the prog- planned from the center. It is a collection of data ress of control measures require data collection that have been produced by various organizations, systems to be in place and operating. using various methods, frequently not following Up to the 1950s there were reports of en- guidelines that WHO has recommended in the demic xerophthalmia in India, Indonesia and sev- past (WHO Expert Group 1976, WHO Expert eral other countries, but little more was known. Group 1982). These inadequacies are dealt with In the early 1960s, the World Health Organiza- in the next section. They are not apparent from tion (WHO) sponsored a global survey in which a reading of WHO (2009) but they should be ad- three consultants visited about 50 countries where dressed in future. VADD was suspected to be a public health problem (Oomen, McLaren, Escapini 1964). This survey revealed the widespread nature and serious VAD AND BLINDNESS magnitude of the problem, especially in much of South and East Asia, and parts of Africa and Latin The magnitude of the effects of VAD on the America. Estimates of the worldwide magnitude eye was touched on in Chapter 1. Chapter 8 was of the problem were also made (McLaren, 1962; devoted to descriptions of various manifestations Sommer 1982). of xerophthalmia. Chapter 7 gave accounts of In recent years, the Micronutrients Unit of how some of the eye signs are used in the assess- the WHO’s Department of Nutrition for Health ment of vitamin A status. Chapter 13 will include and Development in Geneva, Switzerland, with accounts as to how the eye signs have been used its worldwide contacts, has taken on the role of to increase our understanding of the epidemiol- collecting and disseminating data. The Unit ogy of VADD. claims that its data file, the Vitamin and Mineral It has to be recognized in the present context Nutrition Information System (VMNIS), formerly of global occurrence of VAD that no account known as the The Micronutrient Deficiency Infor- whatsoever is being taken of its contribution to mation System (MDIS), is continuously updated. global blindness. No data have been collected for The previous two editions of this book, published a decade or more to indicate whether or not it is in 1997 and 2001 respectively, used the compi- on the decrease. All data on childhood blindness, lation of data from WHO published in 1995. or sometimes corneal blindness, relating to pre- The latest publication, WHO (2009), which is vention are merged and consequently xerophthal- used here, contains data obtained between 1995 mia is not being recorded. Moreover, in a clinical and 2005. For more detailed information on the sense xerophthalmia does not appear in the latest subject of prevalence the original should be con- data published by WHO (2009). In all the maps sulted. For our purposes several key tables and and tables of this publication there is no mention 122 12 Global Occurrence Category of public health significance (prevalence of night blindness) Mild Mild ( > 0% – < 1%) None Moderate ( ≥ 1% – < 5%) Severe ( > 5%) GDP ≥ US$ 15 000 (countries assumed to be free of vitamin A deficiency of public health significance) No data

Figure 12.1: Night blindness as a public health problem by country 1995–2005: Preschool-age children. Countries and areas with survey data and regression-based estimates. 12 Global Occurrence 123 ≥ 20%) Category of public health significance (prevalence of serum retinol <0.70 µmol/l) None ( < 2%) Mild ( ≥ 2% – < 10%) Moderate ( ≥ 10% – < 20%) ( Severe GDP ≥ US$ 15 000 (countries assumed to be free of vitamin A deficiency of public health significance) No data

Figure 12.2: Biochemical vitamin A deficiency (retinol) as a public health problem by country 1995–2005: Preschool-age children. Countries and areas with survey data and regression-based estimates 124 12 Global Occurrence Category of public health significance (prevalence of night blindness) No public health problem ( < 5%) Public health problem ( ≥ 5%) GDP ≥ US$ 15 000 (countries assumed to be free of vitamin A deficiency of public health significance) No data

Figure 12.3: Night blindness as a public health problem by country 1995–2005: Pregrant women. Countries and areas with survey data and regression-based estimates. 12 Global Occurrence 125 ≥ 20%) Category of public health significance (prevalence of serum retinol <0.70 µmol/l) None ( < 2%) Mild ( ≥ 2% – < 10%) Moderate ( ≥ 10% – < 20%) ( Severe GDP ≥ US$ 15 000 (countries assumed to be free of vitamin A deficiency of public health significance) No data

Figure 12.4: Biochemical vitamin A deficiency (retinol) as a public health problem by country 1995–2005: Pregnant women. Countries and areas with survey data and regression-based estimates. 126 12 Global Occurrence Pregnant women Pregnant 2.2 1 Moderate 2.1 1 Mild 2.1 1 Mild 1.7 4 No 1.8 0 No 2.2 2 Mild 2.3 0 Mild 2.0 14 Mild 2.3 0 Mild 1.4 1 No % No (000) Public health problem 12.2 50 Moderate 16.7 110 Moderate 19.3 9 Moderate 17.0 2 Moderate 18.0 66 Moderate 23.7 941 Severe 23.2 32 Severe 23.3 9 Severe 15.0 122 Moderate 21.2 151 Severe 16.0 213 Moderate WHO ). * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed 6.5 1 Mild * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed 0.6 1 No 7.4 1 Mild * No public health problem assumed * No public health problem assumed

% No (000) Public health problem 11.7 11.7 4 Moderate 27.9 408 Severe 54.3 1415 Severe 18.3 62 Moderate 13.3 2405 Moderate 26.1 57 Severe 13.2 26 Moderate 22.0 13 Severe 70.7 1052 Severe 17.4 79 Moderate 21.7 4112 Severe 32.1 176 Severe 14.3 478 Moderate 64.3 1982 Severe 15.7 505 Moderate 18.6 47 Moderate 64.5 3109 Severe Country Burundi Burkina Faso Bulgaria Brunei Darussalam Brazil Botswana Bosnia and Herzegovia Bolivia (Plurinational State of) 21.8 271 Severe Bhutan Benin Belize Belgium Belarus Bangladesh Barbados Bahrain Bahamas Azerbaijan Austria Australia Armenia Argentina Argentina Antigua and Barbuda Angola Andorra Algeria Albania Afghanistan Preschool-age children children Preschool-age * GDP ≥ US$ 15000 * GDP Country estimates of the prevalence of serum retinol <70 μmol/L in preschool-age children and pregnant women 1995–2005 showing the proportion as well 12.1: Country estimates of the prevalence serum retinol <70 μmol/L Table as the number of individuals. (Adapted from 12 Global Occurrence 127 2.4 0 Mild 1.7 1.7 5 No 3 No 2.2 5 Mild 1.8 0 No 2.3 2 Mild 2.3 3 Mild 2.4 1 Mild 2.2 2 Mild 2.2 0 Mild 2.0 17 Mild 2.4 6 Mild 13.2 424 Moderate 15.7 30 Moderate 21.5 397 16.5 Severe 3 Moderate 18.2 4 Moderate 16.1 510 Moderate 17.8 56 Moderate 19.0 131 Moderate 18.2 24 Moderate 15.1 4 Moderate 22.8 3981 Severe 17.1 85 Moderate 16.8 27 Moderate 21.2 3 Severe 16.5 17.9 64 116 Moderate Moderate 8.7 6 Mild 4.2 0 Mild * No public health problem assumed * No public health problem assumed 5.8 27 Mild * No public health problem assumed * No public health problem assumed 3.6 23 Mild 9.2 19 Mild 8.8 35 Mild 5.9 262 Mild 9.3 7877 Mild 7.9 97 Mild 2.0 1 Mild * No public health problem assumed * No public health problem assumed 11.9 11.9 1027 Moderate 46.1 6195 Severe 14.6 13.9 113 21.4 11 Moderate 173 Moderate Severe 14.7 208 Moderate 13.7 152 Moderate 35.2 38 Severe 27.5 441 Severe 57.3 1633 Severe 10.4 0 Moderate 24.6 144 Severe 21.5 28 Severe 50.1 973 Severe 68.2 455 Severe 38.8 1106 Severe 22.3 377 Severe Ethiopia Estonia Egypt El Salvador Equatorial Guinea Eritrea Ecuador Dominican Republic Dominica Djibouti Denmark Democratic Republic of Congo 61.1 7236 Severe Democratic People̓s Republic of Korea Czech Republic Cyprus Cuba Croatia Côte d᾿Ivoire Costa Rica Cook Islands Congo Comoros Colombia China Chile Chad Central African Republic Central Cape Verde Verde Cape Canada Cameroon Cambodia * GDP ≥ US$ 15000 * GDP 128 Pregnant women Pregnant 2.2 2 Mild 1.5 3 No 2.0 5 Mild 1.0 0 No 1.1 5 No 2.3 0 Mild 2.3 0 Mild % No (000) Public health problem 24.2 38 Severe 14.4 8 Moderate 21.0 196 Severe 15.2 222 Moderate 17.1 748 Moderate 16.4 4438 Moderate 18.0 15 Moderate 18.8 71 Moderate 18.1 127 Moderate 23.6 11 Severe 34.0 21 Severe 20.0 7 Severe WHO ). * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed 0.5 31 No * No public health problem assumed * No public health problem assumed 7.0 33 Mild 4.1 3 Mild * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed

% No (000) Public health problem 15.1 108 Moderate 29.4 81 Severe 29.8 1256 Severe 19.6 4261 Moderate 62.0 78643 Severe 13.8 130 Moderate 32.0 398 Severe 54.7 176 Severe 45.8 707 Severe 15.8 326 Moderate 14.1 1 Moderate 75.8 2422 Severe 30.9 73 Severe 64.0 167 Severe 16.9 27 Moderate 13.6 12 Moderate Country Jordan Japan Jamaica Italy Israel Ireland Iraq Iran (Islamic Republic of) Indonesia India Iceland Hungary Honduras Haiti Guyana Guinea-Bissau Guinea Guatemala Grenada Greece Ghana Germany Georgia Georgia Gambia Gabon France Finland Fiji Preschool-age children children Preschool-age Country estimates of the prevalence of serum retinol <70 μmol/L in preschool-age children and pregnant women 1995–2005 showing the proportion as well 12.1: Country estimates of the prevalence serum retinol <70 μmol/L Table as the number of individuals. (Adapted from 129 0 Mild 2.2 2.2 0 Mild 2.2 0 Mild 1.9 40 No 2.2 0 Mild 2.4 0 Mild 2.4 1 Mild 2.4 1 Mild 1.5 0 No 19.2 10 Moderate 18.0 161 Moderate 14.3 122 Moderate 20.9 135 Severe 19.3 9 Moderate 22.,5 4 Severe 17.1 18 Moderate 16.7 101 Moderate 19.9 1 Moderate 22.2 123 Severe 13.7 79 Moderate 13.8 100 Moderate 21.7 32 Severe 12.0 23 Moderate 14.7 9 Moderate 23.5 18 Severe 16.6 26 Moderate 20.1 23 Severe 17.3 259 Moderate 23.6 72 Severe * No public health problem assumed * No public health problem assumed 9.2 9 Mild 4.0 1 Mild 9.4 3 Mild 3.5 97 Mild * No public health problem assumed * No public health problem assumed 8.0 54 Mild * No public health problem assumed * No public health problem assumed 11.1 11.1 17 Moderate 11.0 11.0 40 Moderate 10.0 0 Moderate 17.5 43 Moderate 36.7 1523 Severe 40.4 1203 Severe 17.2 7 Moderate 19.8 46 Moderate 26.8 2799 Severe 47.7 217 Severe 60.7 4 Severe 58.6 1317 Severe 59.2 1436 Severe 42.1 1323 Severe 52.9 365 Severe 32.7 89 Severe 13.0 13 Moderate 26.3 133 Severe 21.8 2 Severe 84.4 5200 Severe 27.1 340 Severe 6868.0 2525 Severe Nauru Namibia Myanmar Mozambique Morocco Montenegro Mongolia Monaco Micronesia (Federated States of) 54.2 8 Severe Mexico Mauritius Mauritania Marshall Island Malta Mali Maledives Malaysia Malawi Madagascar Lithuania Luxenbourg Libyan Arab Jamahiriya Libyan Liberia LebaNon Lesotho Latvia Lao People̓s Democratic Republic Republic Democratic People̓s Lao 44.7 320 Severe Kyrgyzstan Kyrgyzstan Kuwait Kiribati Kenya Kazakhstan * GDP ≥ US$ 15000 * GDP 130 Pregnant women Pregnant 2.2 0 Mild 0.7 0 No 2.2 0 Mild 2.3 0 Mild 6.2 27 Mild 2.2 33 Mild 2.0 4 Mild 2.2 1 Mild 2.2 8 Mild 1.7 10 No 2.0 3 Mild 1.2 2 No 1.8 1 No 2.3 0 Mild 2.2 0 Mild 1.7 102 No 1.7 2 No % No (000) Public health problem 17.5 401 Moderate 10.0 451 Moderate 23.5 14 Severe 14.7 104 Moderate 31.5 252 Severe WHO ). 7.1 0 Mild 6.4 103 Mild * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed 9.3 164 Mild 9.4 32 Mild 8.9 0 Mild 5.5 15 Mild * No public health problem assumed * No public health problem assumed 3.1 21 Mild * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed

% No (000) Public health problem 11.3 11.3 2 Moderate 11.1 11.1 100 Moderate 16.1 4 Moderate 14.1 1017 Moderate 16.3 173 Moderate 25.6 55 Severe 40.1 4422 Severe 14.9 419 Moderate 14.1 103 Moderate 12.5 2377 Moderate 15.5 0 Moderate 29.5 7228 Severe 67.0 1819 Severe 32.3 1171 Severe Country Samoa Saint Vincent and the Grenadines Vincent Saint 2.1 0 Mild Saint Lucia Saint Kitts and Nevis Rwanda Russian Federation Romania Republic of Moldova Republic of Korea Qatar Portugal Poland Phillipines Peru Paraguay Papua New Guinea Panama Palau Pakistan Oman Norway Niue Nigeria Niger Nicaragua New Zealand Nepal Netherlands Preschool-age children children Preschool-age * GDP ≥ US$ 15000 * GDP Country estimates of the prevalence of serum retinol <70 μmol/L in preschool-age children and pregnant women 1995–2005 showing the proportion as well 12.1: Country estimates of the prevalence serum retinol <70 μmol/L Table as the number of individuals. (Adapted from 131 2.3 0 Mild 2.3 0 Mild 2.1 0 Mild 1.7 16 No 2.1 0 Mild 1.5 0 No 2.2 1 Mild 2.2 3 Mild 20.7 23 Severe 22.8 317 Severe 22.5 39 Severe 19.9 49 Moderate 15.4 0 Mild 18.0 33 Moderate 21.8 118 Severe 17.7 6 Moderate 16.1 198 Moderate 22.7 66 Severe 18.9 205 Moderate 18.8 71 Moderate 17.6 48 Moderate 24.2 1 Severe 19.4 85 Moderate 22.5 140 Severe 17.9 1 Moderate 7.2 7 Mild * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed 8.3 21 Mild * No public health problem assumed * No public health problem assumed 8.0 1 Mild 3.6 104 Mild * No public health problem assumed * No public health problem assumed 21.8 0 Severe 28.0 137 Severe 14.6 120 Moderate 17.0 2 Moderate 35.0 366 Severe 45.8 87 Severe 29.7 35 Severe 15.7 708 Moderate 26.8 230 Severe 12.1 302 Moderate 44.6 66 Severe 18.0 8 Moderate 27.8 1523 Severe 35.3 524 Severe 16.9 890 Moderate 61.7 930 Severe 13.1 9 Moderate 74.8 747 Severe 17.2 104 Moderate 37.0 707 Severe 95.6 22 Severe Tuvalu Tuvalu Turkmenistan Turkmenistan Tunesia Tunesia Trinidad and Tobago Tobago and Trinidad Tonga Tonga Togo Togo Timor-Leste Timor-Leste The former Yugloslav Republic Yugloslav The former of Macedonia Thailand Tajikistan Tajikistan Syrian Arab Republic Syrian Switzerland Sweden Swaziland Suriname Sudan Sri Lanka Spain South Africa South Somalia Solomon Islands Slovenia Slovakia Singapore Sierra Leone Seychelles Serbia Senegal Saudi Arabia Saudi Sao Tome and Prinicipe Tome Sao San Marino * GDP ≥ US$ 15000 * GDP 132 Pregnant women Pregnant 2.0 12 Mild 1.7 0 No 2.1 1 Mild 2.2 0 Mild % No (000) Public health problem 20.0 75 Severe 14.0 86 Moderate 15.8 138 Moderate 17.7 292 Moderate 21.0 131 Severe 14.8 237 Moderate 23.3 342 Severe WHO ). 9.4 271 Mild * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed * No public health problem assumed

% No (000) Public health problem 11.9 11.9 30 Moderate 35.8 610 Severe 54.1 1089 Severe 27.0 984 Severe 12.0 972 Moderate 16.1 5 Moderate 53.1 1519 Severe 23.8 476 Severe 27.9 1629 Severe Country Zimbabwe Zambia Yemen Yemen Viet Nam Viet Venezuela Venezuela Vauatu Vauatu Uzbekistan Uruquay United States of America United States of United Republic of Tanzania Tanzania United Republic of 24.2 1683 Severe United Kingdom of Great Britain and Northern Ireland United Arab Emirates United Ukraine Uganda Preschool-age children children Preschool-age * GDP ≥ US$ 15000 * GDP Country estimates of the prevalence of serum retinol <70 μmol/L in preschool-age children and pregnant women 1995–2005 showing the proportion as well 12.1: Country estimates of the prevalence serum retinol <70 μmol/L Table as the number of individuals. (Adapted from 12 Global Occurrence 133

of xerophthalmia. There is yet another mistake signs of xerophthalmia were being made, it was that has persisted for two decades or so: in this found that about 10% of these led to blindness, period eye disorders attributable to VAD (i.e. that but about 20% of those blinded were dead within part of VADD affecting the eye) have not been a year. Xerophthalmia in young children is now divided, as they were in the past, into those caus- estimated to have a prevalence of about 4.4 mil- ing blindness and into those that do not. lion (see Table 1.3). Reduction for deaths would The result is not only that the world of micro- still leave xerophthalmia blindness many times nutrient deficiency prevention is more than ever greater than the highest estimate for measles misled and bemused by this latest authoritative (McLaren 2004). report on the subject, but also the world of blindness prevention will now have nothing to guide it on xerophthalmia. Already recent (WHO) pub- VITAMIN A IN GLOBAL FOOD SUPPLIES lications on the subject fail to address vitamin A related childhood blindness in a uniform way. It It will be recalled from Chapter 3 and Chap- may just be grouped with other causes of “child- ter 4 that the major underlying cause of VADD is hood blindness”, or its identity may be lost in the inadequacy of some basic diets in vitamin A “corneal blindness in young children”. In effect, sources. Peoples of developing countries are es- nutritional blindness, once generally recognized pecially dependent on provitamin A carotenoid as the most common cause of blindness in young sources. In recent years more thorough investiga- children worldwide, has ceased to exist. The pos- tion of the vitamin A activity of these vegetable sible implications for the work of the many pre- and fruit sources has shown that they are of con- ventive eye care programs are serious. siderably less activity than previously thought. The close association between measles kera- Some of these recent estimates have been more titis and xerophthalmia has been noted in other pessimistic than others. Even when the more lib- chapters, especially 9 and 13. The latter, in par- eral estimates have been made there is still about ticular, draws attention to the frequent concur- a 55% overestimate globally. As a consequence rence, especially in parts of Africa, of measles of recent estimates it has become evident that keratitis, xerophthalmia, herpes simplex, and many people, especially young children, liv- the use of harmful local eye medicine. Measles ing in poverty are unable to meet their vitamin and xerophthalmia both have a high mortality in A requirements through consumption of fruit and malnourished children. In recent years mortality vegetables alone. Some preformed vitamin A from measles has been reduced dramatically by sources appear essential. These estimates also ex- the Expanded Program on Immunization (EPI) plained the occurrence for many years of signs of (see Chapter 15). Between 1999 and 2005 it xerophthalmia in the presence of apparently rich decreased from an estimated 873,000 deaths to provitamin A carotenoid sources.These points are 345,000 deaths (Wolfson, Strebel, Gacic-Dobo made evident in Figure 12.5. et al. 2007). Unfortunately, there is no record as The reader̓s attention should be given to cer- to what part accompanying VAS might have been tain facts about the data that lie behind this impor- playing in EPI and National Death Index (NDI) tant figure. The basic data come from FAO food (see Chapter 15). intake sheets that date back as far as 1979–1981. It has been suggested (Semba, Bloem 2004) Not only have they not been kept updated since, that measles is now the leading cause of child- but the originals have not been given the pride of hood blindness in low income countries, account- place which they deserve. It is no excuse to say ing for an incidence of 15,000–60,000. This range that they are not precise enough to be applied in will be falling with continuing success with EPI. any way; earlier in the chapter other data in this When numerous large field surveys of the eye field have been shown to be similar in this respect. 134 12 Global occurrence

Table 12.3: Preliminary assessment of vitamin A status of a population. (After Sommer, 1995). 1. Interviews, by structured questionnaire, with central and provincial public health officials, clinicians, nutritionists, community health workers, directors and staff of hospitals, feeding and rehabilitation centres, and schools for the blind. 2. Chart reviews at hospitals, clinics, institutions where children suspected to suffer from corneal destruction, or where coded charts do not exist, records from malnutrition, infectious diseases, pediatric and ophthalmic services. 3. Search for clinically active cases among children in high-risk situations – hospitals, clinics, impoverished communities. 4. Search for old, healed disease in schools for the blind etc. 5. Collect existing data on child rearing, including breast and other feeding and hygiene practices.

However, it has to be questioned to what ex- able to employ a preliminary assessment, along tent they may be applied to current situations. the lines of that advised by WHO (WHO Expert Group, 1982). Table 12.3 indicates the kind of background data that should be sought in the first METHODOLOGY instance. This can be done in a simple way that is eco- It is evident that careful attention should be nomical in terms of personnel and resources. It given to the methods used for data collection on will be observed that most of these data relate to the prevalence of VADD. In a country or area the detection of a problem of xerophthalmia. If where there has been no previous study it is advis- the health concern is at the subclinical level of

1200

1000

800

600

400 g RAE/person/day) Food supply of vitamin A ( μ 200 75% 69% 34% 35% 32% 30% 55%

0 Asia Africa South N/C Europe Oceania World America America

Figure 12.5: Vitamin A activity in the regional and world food supplies as provitamin A carotenoids with their percent of total (based on a 12:1 β-carotene:retinol conversion ratio, white segments) and as preformed retinyl esters (black segments). Stippled bar represents the previous estimate of total vitamin A using a 6:1 conversion ratio, reflecting a 55% overestimate globally. (West, Darnton-Hill 2008). 12 Global Occurrence 135

VAD then an assessment may be made using the 1.0% is accepted as the cut-off point for “xeroph- socioeconomic and ecological indicators recom- thalmia” (see Chapter 1) has had the unfortunate mended by WHO (1996). Only if the results are consequence that no figures are now available for highly suggestive of the presence of a significant blindness due to vitamin A deficiency. Experience problem should a full point prevalence survey be suggests that the ratio between nonblinding and embarked upon. This is much more expensive and blinding xerophthalmia is approximately 10:1. time-consuming and must be carried out meticu- This suggests that somewhat less than half a mil- lously if the results are to be relied upon (Sommer lion cases exist worldwide, although the usual fa- 1995). tality rate of about 20% in the first year (McLaren 2004a) will reduce this. It has been customary in the past to estimate TRENDS that approximately 10% of all blindness in young children (about 30 million) could be attributed to A comprehensive document was published xerophthalmia. In the absence of current figures recently (Mason, Rivers, Helwig 2005) detail- the present position is unknown. ing trends in malnutrition in developing regions: Two other groups are also especially suscep- vitamin A deficiency, anemia, iodine deficiency, tible to vitamin A deficiency, namely pregnant and child underweight. With the ever increasing and lactating women and school-age children. emphasis on multi-micronutrients it is especially Provisional estimates suggest that each year there useful to have these data available together. This are nearly 20 million pregnant women with low Manual concentrates on data for vitamin A defi- vitamin A status (serum or breast milk retinol ciency. The maps that appeared in its second edi- <1.05 μmol/L), of whom 7 million are deficient tion are reproduced again here as the data from (<0.70 μmol/L). Six million are night blind (West, which they were constructed dates from October McLaren 2003). 2000, and the data behind the new documents is Singh and West (2004) examined various not more recent. A new feature is the provision of sources of data to calculate the prevalences of information on trends in the prevalence of both VAD (<0.70 μmol/L) and xerophthalmia. The xerophthalmia and vitamin A deficiency for each prevalence of VAD in children aged 5–15 years region over the decade 1990–2000. Figures 12.6 in southeastern Asia was 23.4% or 83 million (an and 12.7 sumarize this information and indicate estimated 127 million worldwide, see above). that, with the exception of xerophthalmia in India, Xerophthalmia (Bitot’s spots and night blindness) trends have followed a steady downward course. was present in 10.9% – a very high percentage Annex 1 of the WHO document should be con- compared with the estimate of 4.4 million world- sulted for data from individual countries and for wide given above. Potentially blinding corneal individual years, where available. xerophthalmia was negligible in this age group. Estimates for cases of xerophthalmia (non- It is very important to recognize that all data blinding and blinding) have fluctuated greatly of a biochemical nature that is based on the esti- over the years, influenced in opposite directions mation of retinol in serum – especially such data mainly by better reporting and by an increasing from developing countries – is subject to the ef- number of control measures. Most recently, the fect of the acute phase response that depresses prevalence figure for young children stands at 4.4 serum retinol. Although efforts are being made to million (West 2002). correct for this effect in data analysis (see Thurn- It is not known how many of those cases ham, McCabe, Northrop-Clewes et al. 2003 and are for blinding xerophthalmia. The simplified Chapter 6), standardization and general accep- procedure where all forms of xerophthalmia are tance are likely to be some time coming. grouped together and the recommended level of 136 12 Global Occurrence

Figure 12.6: Trends in the prevalence of xerophthalmia in children 1990–2000. *India measles coverage (year/coverage): 1990/86, 1995/81, 2000/05 (Mason, Rivers, Helwig 2005).

Figure 12.7: Trends in the prevalence of vitamin A deficiency (VAD) <0.70 µmol/L in children, 1990–2000 (Mason, Rivers, Helwig 2005). 12 Global Occurrence 137

COUNTRY SITUATIONS (<0.70 μmol/L) occurs in 24.9%. These results suggest that the Philippines should have been There is still a surprisingly large number of categorized at the time as having a severe sub- countries from which no data is available and clinical VAD public health problem (Solon, per- where the occurrence of a problem might be sus- sonal communication). More recent data (Solon, pected. The largest block of such countries is that 2002) for 1993 and 1998 for pre-school children of the former Soviet Union, with millions of chil- (Figure 12.9) and pregnant and lactating women dren at risk. (Figure 12.10) in the Philippines shows continu- The situation is of course fluid, and improve- ing evidence of a VADD problem. Rare details ment or deterioration may occur over quite a like this suggest the kind of problems that may be short period of time. For example, there is good encountered in this field. evidence that the situation has improved consider- Xerophthalmia was extremely common in ably in three of the major countries where VADD Vietnam until after the cessation of hostilities. were highly endemic only two decades or so ago In recent years, highly efficient capsule distribu- (see Figure 12.8). tion nationwide has resulted in a pronounced de- A National Nutrition Survey carried out by crease (Khoi, Khan, Thang et al. 1996). Some- the Food and Nutrition Research Institute in the times improvement is more likely to be attribut- Philippines in 1993 showed that the prevalence able to general economic and social development, of clinical VAD among children aged 6 months rather than just to specific vitamin A interven- to 6 years was 0.4% for XN and 0.2% for X1B, tions. and for the total population the prevalence of XN On the other hand, VADD has emerged in was 0.8% and of X1B 0.3%, suggesting that VAD several countries in the Pacific region, as a result may no longer be a clinical public health problem. of rapid changes in lifestyle and the abandonment However, serum retinol <0.35 μmol/L occurs in of traditional dietary and other practices that were more than 5% of children aged 6 months to 6 protective (Lloyd-Puryear, Humphrey, West et al. years in 11 out of 15 regions. Low serum retinol 1989; Schaumberg, O’Connor, Semba 1996).

Figure 12.8: Trends in xerophthalmia in children up to six years (ACC/SCN Consultative Group, 1994). 138 12 Global Occurrence

40 38.0 35.8 1993 30 1998

10.4 10 8.2 Percent prevalence Percent

0 <0.35 µmol/l <0.70 µmol/l w

Figure 12.9: Prevalence of subclinical vitamin A deficiency among pre-school children (1993 and 1998) from the Philippines (Solon 2002).

25 22 1993 20 1998 16 16.4 16.5 15

10 Percent prevalence Percent 5

0 Pregnant women Lactating women

Figure 12.10: Prevalence of subclinical vitamin A deficiency among pregnant and lactating women from the Philippines (Solon 2002). 12 Global Occurrence 139

In China in 1999–2000, 8,681 children aged 0–6 years were examined for vitamin status (Jingxiong, Toschke, von Kries et al. 2006), whereby 12.2% were discovered to have VAD (serum retinol <0.7 μmol/L) and 0.5% to have serum retinol <0.35 μmol/L. No correction was made for the presence of APR and no clinical disease was reported. Serum levels were lowest in minority groups and in the less developed west of the country. A prophylaxis program is planned. A public health problem has long been suspected in certain other countries, but no data were available. This seems to be the situation in Laos (Malyavin, Bouphany, Arouny et al. 1996) and in Afghanistan. 140 141

13 Epidemiology

INTRODUCTION into the young child’s diet if at all possible. On the other hand, it may be found that dietary intake of Epidemiology may be defined as the study of carotenoids is adequate but VADD is nevertheless the distribution and determinants of disease in a a problem. Further investigation may reveal that population. In simple terms various parts of the there is heavy round-worm infestation in the com- Manual may be considered to have dealt in turn munity and it is likely that much of the ingested with the answers to the how, what, where and how carotene is not being absorbed. Deworming and much of VADD. measures to improve sanitation should have prior- The present chapter sets out to give some an- ity in this case. swers to the question “Why do VADD occur?” A There is no special order in which the risk fac- clear understanding of these matters is vital be- tors are considered here, but in general those of fore appropriate measures can be taken for the greater importance are taken first. control of VADD. It is logical, therefore, that this chapter on epidemiology should precede Chapter 15 on “What can be done?” to control AGE VADD. Those factors that are closely associated with Evidence from every source agrees that pre- the occurrence of a disease are known as risk fac- school-age children form the most vulnerable tors. This association may be demonstrated by group (Sommer 1982). This applies to corneal the application of statistical methods. However, xerophthalmia in hospitals (see Figure 13.1) and there is no statistical way of proving that the as- noncorneal xerophthalmia in field studies (see sociation is causal; i.e., that the factor causes the Figure 13.2). From the data taken together it will disease. A logical relationship between the factor be evident that the more serious lesions peak at an and the disease lends support to the association earlier age than the less serious. being causal, but does not provide proof. A combination of factors probably lies behind Consequently, epidemiology can provide use- this phenomenon. The requirements for growth ful indications as to how a disease arises. Vulner- of the younger child tend to be greater, while able groups can be identified towards whom in- the body stores tend to be lower. The diet of the terventions may be targeted. Defective or, on the vulnerable child tends to be more restrictive and other hand, protective aspects of diet and other limited during infancy or just after weaning than aspects of lifestyle may be revealed and as a result later on. Infections have an additional adverse nu- this knowledge may be used in planning interven- tritional impact. tions. Keratomalacia in the infant usually proceeds It is important to recognize that epidemiologi- rapidly without any tendency for xerosis of the cal factors may vary considerably from place to conjunctiva to develop. From some parts of India place. Interventions need to be appropriate for there have been recent reports of increase in the local conditions. For example, in some circum- occurrence of infant keratomalacia. This may be stances it may be found that the weaning diet related to the increasing tendency for women of lacks rich sources of provitamin A in the form of low socioeconomic status to go out to work and dark green leaves or yellow fruits. Ways should to leave their infants at home without adequate at- be sought to ensure that these are incorporated tention (Rahmathullah, Raj, Darling et al. 1994). 142 13 Epidemiology

Figure 13.1: Age distribution of consecutive cases with corneal xerophthalmia (X2 or X3) presenting to the Cicendo Eye Hospital (n=162), Bandung, Indonesia, and to the Lahan Eye Hospital (n=295), Nepal (Sommer, West 1996:337).

Figure 13.2: Age distribution of mild xerophthalmia in selected countries in Asia and Africa: Nepal, Zambia, India and Indonesia (showing data from both countrywide surveys in 1978–79 and 1992) (Sommer, West 1996:338). 13 Epidemiology 143

School age is another period of increased was observed up to 28 weeks of gestation on mor- susceptibility, usually involving the milder de- bidity rates. After 28 weeks symptoms of nausea, grees of xerophthalmia (night blindness or Bitot’s faintness and night blindness were reduced with spots) (Khan, Haque, Khan 1984). This may be vitamin A, but not with β-carotene. Vitamin A due to increased requirements for growth, espe- was associated with shortened labor by 1.5 hours cially during the adolescent growth spurt (Ahmed, in nulliparous, and 50 minutes in multiparous Hasan, Kabir 1997). women. Both interventions were associated with reduced postpartum prevalence of at least 4 loose stools and night blindness. β-carotene was as- PHYSIOLOGICAL STATUS sociated with reduced symptoms of high fever postpartum, vitamin A with reduced number of It has long been recognized that the pregnant days of illness symptoms over the last 12 weeks and lactating woman is especially vulnerable. Re- of pregnancy. Vitamin A or β-carotene reduced cently some very high rates have been reported selected illness symptoms during late pregnancy, especially from the Indian subcontinent (Katz, at the time of birth, and during 6 months post- Khatry, West et al. 1995). In this study from Ne- partum. pal 16.2% of women reported XN at some time Vitamin A deficiency, as assessed by serum during the pregnancy that produced the child they retinol (see Chapter 11) and impaired dark adap- were breastfeeding at the time of the interview. tation, is not infrequently accompanied by iron XN was reported by 8.1% at the time of the inter- deficiency anemia during pregnancy (Radhika, view. The chances of XN in the current pregnancy Bhaskaram, Balakrishna et al. 2002; van den were six times greater for those who reported XN Broek, Letsky 2000). in their previous pregnancy than for those who did not. Only mild xerophthalmia is usually seen and DIET the mother’s health may not be permanently affected. However, low vitamin A content of breast Regular ingestion of the customary dietaries milk will contribute to the increased susceptibility around the world, both now and in the past, has of the infant. The proposal that breast milk retinol provided for vitamin A requirements to be met. be used as a biological indicator of vitamin A sta- It has only been when consumption was limited, tus in the community was discussed in Chapter 7. perhaps at certain seasons of shortage or when More recent studies by the same group in certain food items were omitted altogether, that Nepal have shown that XN in pregnancy may problems have arisen. Young children have to rely be a marker of other nutritional and health risks on others to feed them, and pregnant and lactat- (Christian, West, Khatry et al. 1998). Attitudes ing women are sometimes especially targeted for of women towards XN have been documented food taboos. (Christian, Bentley, Pradhan et al. 1998) as has Staple foods, mostly cereals but including the extent to which it impairs their work activities some starchy roots, provide the bulk of a diet. (Christian, Thorne-Lyman, West et al. 1998). They generally contain small concentrations of As part of a large study of the effects of vi- carotenoids but because they are eaten in bulk tamin A and β-carotene supplementation in preg- the amount supplied may be considerable. Rice nancy in Nepal (see Chapter 9), Christian, West, is the main exception to this rule. It contains no Khatry et al. (2000a) recorded the impact of week- carotenoids in the form in which it is consumed ly supplementation on the prevalence of preg- (see Chapter 15). Furthermore, mothers often nancy and postpartum illness symptoms among consider it to be a complete diet for the young 15,832 women. No impact of either supplement child. Infants find it easy to eat and satisfying. 144 13 Epidemiology

“Rice-dependent” communities, where rice and especially in young children (Hudelson, Dzi- little else forms the daily diet, have proved to kunu, Mensah et al. 1999). Adverse factors are be especially prone to suffer from VADD. It is large family size (Kjolhede, Stallings, Dibley et ironical that this situation occurs mainly in the al. 1995), sharing of the food plate with an adult moist tropics, where carotene is readily available male rather than a female family member (Shan- on every hand. This is truly an instance of nutri- kar, West, Gittelsohn et al. 1996) and lack of tional “poverty in the midst of plenty” (McLaren child care (Gittelsohn, Shankar, West et al. 1998). 1962). Feeding practices in infancy can influence sub- Enquiries about dietary intake should start sequent risk of VAD (Gittelsohn, Shankar, West with identification of the staple food. In this way et al. 1997). Preformed vitamin A intake may be the traditional “home” of xerophthalmia was more important than carotenoid-containing foods found to be the rice-dependent areas of south for the protection of children and others (Shankar, and east Asia. In India the rice-eating south has West, Gittelsohn et al. 1996). been much more vulnerable than the wheat-eating north. On some of the small islands of Indonesia maize is the staple food, and xerophthalmia was BREASTFEEDING virtually confined to communities that consumed white maize (carotene free) and not seen among There is abundant evidence that breastfeed- those that ate the yellow variety (Oomen 1961). ing is highly protective against xerophthalmia Recent studies continue to provide evidence (Sommer, West 1996: 343–345 and Figure 13.3). for the close relationship between VAD and the This may be partly due to the regular supply of details of dietary intake and cultural practices, preformed vitamin A in the milk. Another impor-

Figure 13.3: Association between percent of children breastfeeding and mild xerophthalmia (XN, X1B), by age. Open bar, non-xerophthalmic children; black bar, cases. A, Indonesia, B, Nepal, C, Bangladesh, D, Malawi (Sommer, West 1996). 13 Epidemiology 145

tant factor is the lower rate of infections as com- the society. They are not as open to understand- pared with artificially fed children. ing by outsiders as are less complicated charac- A study in south India (Ramakrishnan, Mar- teristics like others considered here. Groups un- torell, Latham et al. 1999) showed that while non- der study often speak a language not readily un- breastfed children met only 60% of the Indian derstood by the investigators, even if of the same RDA, those breastfed met 90% during the second country. For outsiders, and this is what scientists year of life. The WHO and UNICEF recommend investigating a community inevitably are, under- that all infants be exclusively breastfed for at least standing of the worldview of others is always 6 months and that they continue to be breastfed limited. up to two years or beyond with the addition of More than fifty years ago (McLaren 1956) adequate complementary foods from about 6 the apparent protection of the children of a tribal months of age. group in an area of south India where keratoma- The composition of the weaning diet is obvi- lacia was highly endemic was attributable to the ously of great importance. Infrequent consump- custom of child spacing practiced by this group, tion of dark green leaves or yellow fruits was as- but not practiced by another group among whom sociated with fourfold to sixfold increase in the corneal xerophthalmia was common. In this rice- risk (odds ratio) of xerophthalmia in one study dependent state of Orissa, India, the “deposed (see Figure 13.4). child” was the target of keratomalacia. This phenomenon had first been described in Ghana (Wil- liams 1933) where the victim of kwashiorkor was CULTURAL FACTORS the child who had been abruptly weaned when the mother realized she was pregnant again. It is The customs practiced by a community are of considerable interest that in both instances the usually deeply embedded within the fabric of investigators made their contributions as by-prod-

Figure 13.4: Relative risk (case-control odds ratio) (±95% CI) of mild xerophthalmia (vitamin A deficiency) by type of food reportedly consumed by children daily or every other day during their first twelve months of weaning (Mele, West, Kusdiono et al. 1991). Not eating dark green leaves, for example, increases the risk about sixfold. 146 13 Epidemiology

ucts of their primary work (medical missionary (1.0 μmol/L) (Northrop-Clewes, Mwela, Kankasa and colonial government medical officer respec- et al. 2005). tively), and from long and intimate contact with Two studies conducted on infants, one on the people concerned. healthy infants in Germany (Abraham, Mueller, Grueters et al. 2003), and another in Indonesia on infants with infections (Wieringa, Dijkhuizen, INFECTIOUS DISEASES West 2002) have provided new evidence. The first showed that in this age group APR is triggered at As was noted earlier, when the topic of the very low inflammatory levels and can even occur impact on infections of vitamin A deficiency was in apparently healthy infants. A high-sensitivity considered (see Chapter 9), the relationship is assay was used and the high prevalence of posi- complex. Infections also predispose to the devel- tive results is attributed to this. It was concluded opment of vitamin A deficiency, as will be con- in the Indonesian study that the prevalence of vi- sidered here. tamin A deficiency was ‘overestimated’ by >16%. The acute phase response (APR) needs to In patients with HIV infection low serum reti- be looked at in the present context. The account nol levels are considered to be mainly attributable given in Chapter 7 may be consulted as an in- to the infection itself and the associated APR, troduction to the subject. Thurnham, McCabe, rather than an indication of low vitamin A status Northrop-Clewes (2003) carried out a meta- (Baeten, McClelland, Richardson et al. 2002). analysis from the results of which Figure 7.1 The situation appears to be even more com- was constructed. This suggests approximate plicated in pregnancy (Christian, West, Khatry et levels of effect of subclinical infection on al. 1998a). Physiological changes that take place plasma retinol, and consequently on vitamin A include hemodilution and an increase in acute status. phase proteins. Both of these changes lead to re- An important implication of these observa- duction in serum retinol concentration (Dijkhui- tions is that care must be exercised in the inter- zen, Wieringa, West 2005). Recently the same pretation of serum retinol and RBP levels in re- group has studied other micronutrients for effects lation to the assessment of vitamin A status (see of the acute phase response (Jiang, Xu, Christian, Chapter 7). et al. 2005). The situation is complex. Results dif- As knowledge of this phenomenon has spread fer according to the kind of test used. Nutrients in recent years studies have been carried out that that, like serum retinol, are decreased include increasingly underline the practical importance of β-carotene and lutein (both P<0.0001). taking it into account in studies in communities in Clearly much further research is required be- developing countries. Stephensen and Gilden- fore standards can be set for use in realistically gorin (2000) reinterpreted the NHANES III data assessing vitamin A status of populations. Maq- for serum retinol in the United States. Other similar sood, Dancheck, Gamble et al. (2004) found in a large data bases will presumably also be affected study of preschool-age children in the Marshall in a similar manner. Islands that removing children with evidence of A recent study of preschool children in Zam- raised acute phase proteins introduced sample bia, correcting for the presence of infection by bias for several variables. measuring C-reactive protein (CRP) concentra- In a group of Venezuelan children aged 4–7 tions, raised the mean plasma retinol concentra- years with subclinical vitamin A deficiency, as tions by approximately 10% overall. The retinol measured by conjunctival impression cytology, concentrations of those children with no infec- circulating cytokines IL-10, IL-4, IFN-γ, and tion, as defined by a normal CRP, were 13–20% IL-2 were measured. Only IL-10 was found to be lower than the UK mean for preschool children diminished (Leal, Castejon, Romero et al. 2004). 13 Epidemiology 147

Diarrheal diseases absence of improvements in sanitation and the There is strong evidence from both clinical and breaking of the cycle of transmission. field studies that diarrheal disease is closely as- It is well known that vitamin A supplemen- sociated with xerophthalmia and with impaired tation often improves hemoglobin concentration. vitamin A status. An additional response was observed if deworm- Considerable interest was aroused by the re- ing (for Ascaris lumbricoides and Trichuris trich- port from Peru (Alvarez, Salazar-Lindo, Kohatsu iura) took place at the same time (Tanumihardjo, et al. 1995) that diarrheal disease, especially that Permaesih, Muhilal 2004). due to rotavirus infection and with accompanying high fever, may lead to as much as a tenfold Respiratory diseases increase in the urinary excretion of vitamin A (see As was the case for diarrheal diseases, for respira- also Chapter 14). Alvarez and colleagues reported tory infections there is also good evidence from a similar urinary loss of vitamin A in adults with field and hospital studies that they worsen vita- sepsis and pneumonia in the United States and min A status at all levels. Several possible mecha- have extended their studies to Bangladesh (Mitra, nisms may be at work including loss in urine and Alvarez, Stephensen 1998). In shigellosis urinary impaired absorption of both vitamin A and carot- retinol loss was proportional to the severity of the enoids (Sommer, West 1996:191–220). disease, and impaired tubular reabsorption of low molecular weight proteins like RBP appeared to Measles and associated conditions be the cause. The effect of vitamin A deficiency on the course When Sight and Life published a monograph of measles infection was discussed earlier (see to mark the centenary of the first description of Chapters 9 and 12). Decades ago measles was a large outbreak of xerophthalmia (Mori 1904), recognized to carry a high mortality risk in chil- the nature of diarrhea accompanying xerophthal- dren in Africa with severe PEM, and involvement mia was discussed (McLaren 2004). The weight of the eye was occasionally noted. It took many of evidence seems to support the view that it is years to provide the evidence that was necessary primarily a symptom and sign of the vitamin A to bring about general acceptance of the fact that deficiency state itself. vitamin A deficiency is frequently responsible for Perhaps the most significant recent develop- the blindness that sometimes followed an attack ment that supports the suggestion that diarrhea is of measles. There is evidence from other parts part of VAD is the work of Mayo-Wilson and of the world, as well as from several regions of colleagues (2011). Evidence was obtained to Africa, that about one quarter to one half of all show that vitamin A deficiency induces impair- cases of corneal blindness in young children ment of the gut integrity. are associated with measles (Sommer, West 1996:191–220). Intestinal parasites It is now recognized that often there is not Giardia lamblia, Ascaris lumbricoides, (round- only the combination of VAD and measles but worm) and Ankylostoma duodenale (hookworm) that other factors may also play a part. Sometimes have all been shown to reduce vitamin A ab- the initial eye involvement is treated by the ap- sorption, and in some instances to be associated plication of traditional eye medicines that usually with clinical VAD (Curtale, Pokhrel, Tilden et al. serve to increase the damage. Ulceration in the 1995). The impact may be greater on absorption periphery of the cornea, especially in otherwise of carotenoids than on absorption of preformed quiet eyes, is then characteristic (Lewallen, Court- vitamin A. Most of the population becomes in- right 1995). Severe involvement of the cornea fested where sanitation is lacking. Deworming may be due to superadded infection with the her- programs fail to have a long-term impact in the pes simplex virus (see Table 13.1). In this series 148 13 Epidemiology

of cases of corneal ulceration treated in hospital PROTEIN-ENERGY MALNUTRITION there is a higher proportion of cases following (PEM) measles infection in the group with xerophthalmia as a specific cause than in those in which Impaired growth usually accompanies xe- herpes simplex or traditional eye medicine was rophthalmia and tends to increase pari passu with the cause. This is perhaps not surprising as severe the degree of vitamin A deficiency (see Chapter measles tends to precipitate nutritional deficiency. 10). Severe degrees of PEM, evidenced clinically This combination has been reported from as marasmus, marasmic-kwashiorkor, or kwashi- several parts of Africa but is not known to have orkor, are not necessarily associated with eye the same importance elsewhere. Malaria has also signs of xerophthalmia, although serum retinol been reported to be a precipitating factor in some is usually depressed. Corneal xerophthalmia on places (Yorston, Foster 1992; Genton, Al-Yaman, the other hand is usually accompanied by severe Semba et al. 1994). This merits investigation else- PEM in some form (McLaren, Shirajian, Tchalian where in view of the widespread occurrence of et al. 1965). drug-resistant malaria. Much of the association can be explained by dietary habits and disease patterns that at the same time adversely affect both protein energy HIV / AIDS and vitamin A status. In addition, there is both experimental and clinical evidence that low pro- Only in recent years has the association be- tein status can impair RBP synthesis and its re- tween HIV infection and vitamin A status begun lease from the liver. Therefore the RBP response to be investigated (see Chapter 9). There is no to a large dose of vitamin A is reduced (Som- evidence yet as to whether infection impairs vita- mer, West 1996:207). Low protein status can min A status. It is likely that a serious disease like impair response to vitamin A therapy and delay AIDS would have such an effect; patients with recovery from corneal xerophthalmia (Sommer, AIDS infection, rather than just HIV positive, are West 1996:207). However, in chronic protein usually markedly undernourished. deficiency growth and metabolic demands are in- hibited and when protein is given therapeutically vitamin A demands may be increased and xerophthalmia precipitated.

Table 13.1: Corneal ulceration within one month of measles infection by cause and age as a percentage of all ulcers in Tanzanian children (Foster, Sommer, 1987). Age Xerophthalamia Herpes simplex Traditional eye medicine (years) N n % N n % N n % <2 16 10 62 24 3 12 8 1 12 2–4 12 11 92 10 3 30 5 5 100 5–10 6 3 50 13 4 31 5 2 40 Total 34 24 71 47 10 21 18 8 44

N=total numer of children seen with corneal ulceration due to a specific cause; n = number of children seen with corneal ulceration due to a specific cause which developed within one month of measles infection. 13 Epidemiology 149

Three eminent nutrition scientists (Schuftan, mer pasture. In developing countries yellow fruits Ramalingaswami, Levinson 1998) pointed out like mangoes and papaya are usually consumed that PEM was largely neglected and excessive during their quite short seasons and may not be attention was being given to multi-micronutrient stored. Leafy vegetables have longer, but still lim- (MMN) deficiencies. In more recent years the ited, seasons. Communities often experience an emphasis has grown even stronger. It should not annual dry, “hungry” season when young children be forgotten, however, that MMN is almost in- and their mothers tend to suffer most (Oomen variably accompanied by PEM of some degree; 1969). Failure of the rains or, conversely, excess usually wasting. Control of both states should be flooding for several years in succession can lead concurrent. to serious food shortage or even famine. Under such conditions outbreaks of night blindness and other deficiencies tend to affect a high proportion SEASON of the total population. Among communities that subsist in semi- Fluctuation in the amount of vitamin A and desertic terrain, as in the deserts of Rajasthan carotene available in the diet throughout the year (Desai, Desai, Desai 1992), a single season with is well documented (Moore 1957). Where sources a lack of monsoon rainfall can spell nutritional are limited the change is reflected in variations in disaster (Figure 13.5). plasma levels of vitamin A and the seasonal oc- Frequently superimposed upon the seasonal currence of signs of vitamin A deficiency. Cattle pattern described above there is the added ef- fed on winter fodder produce milk with lower fect of seasonal outbreaks of infectious diseases vitamin A content than when they receive sum- (Oomen, McLaren, Escapini 1964; McLaren,

Figure 13.5: The yearly prevalence of xerophthalmia in a population of children aged less than 10 years shows a peak of 34.3% in 1987, the year of severe drought in western Rajasthan (Desai, Desai, Desai, 1992). 150 13 Epidemiology

Shirajian, Tchalian et al. 1965). The most impor- LOCATION tant of these is the group of diarrheal diseases, gastroenteritis, often known locally as “summer Sommer and West (1996:335–354) sum- diarrhea” because of its peak season. VAD is usu- marize a large body of evidence that shows ally accompanied by PEM and emerges towards that VADD tend to cluster at a variety of levels. the end of the diarrhea season, as the effects of This phenomenon applies to provinces, districts, prolonged and repeated attacks of diarrhea take subdistricts, villages and households (IVACG their toll. Respiratory infections tend to peak in 1996a). Figure 13.8 shows this clustering in re- the winter, which in some developing countries gions in Bangladesh. with a continental climate or at high altitude may In some countries childhood blindness has be bitterly cold. long been known to be a special feature of cer- The most precise documentation of seasonal tain remote regions. The Luapula Valley in Zam- patterns and VADD was made by Sinha, who bia (Sukwa, Mwandu, Kapui et al. 1988) and the resided for two years in the West Bengal village Lower Shire Valley in Malawi (West, Chirambo, area of Ichag (see Figure 13.6 and 13.7). Katz et al. 1986) are two especially well studied examples. Vitamin A deficiency has proved to be the major cause of childhood blindness in both SOCIOECONOMIC STATUS places. The underlying reasons have proved to be complex. Economic deprivation was found in the coun- As for any disease, the occurrence of clusters trywide survey in Indonesia (Sommer, 1982) to may be seen as the probable presence at one time correlate closely with vitamin A status. Other in- and place of a combination in some degree of indices of health have shown the same thing, as did tensity of several of the risk factors considered for example growth of young children in Lebanon here. There might also be other variables that are (Kanawati, McLaren 1973). as yet unrecognized.

Figure 13.6: Seasonal prevalence of vitamin A deficiency signs, Ichag, 1971–73 (Sinha, Bang, 1976). 13 Epidemiology 151

Figure 13.7: Seasonal prevalence of vitamin A deficiency (night blindness and/or Bitot’s spot) according to sex, Ichag, 1971–73 (Sinha, Bang, 1973).

X1B >1% XN >3% and X1B <1% XN <3% and X1B <1% Not surveyed

Figure 13.8: Regional clustering of VAD by severity and district in Bangladesh (Cohen, Rahman, Mitra et al., 1987). 152 13 Epidemiology

Katz, Zeger, West et al. (1993) have studied VAD than females. In healthy human adults plas- clustering of xerophthalmia within households ma retinol and RBP are both about 20% higher in and villages in Malawi, Zambia, Indonesia and males, but the significance of this is unclear (Pilch Nepal. The magnitude of clustering varied be- 1987; Smith, Goodman 1971). Night blindness tween countries and in all places studied it was and Bitot’s spot are almost uniformly reported to much greater within households than within vil- be from 1.2 to 10 times more common in males lages. Infectious diseases appeared not to explain (Paton, McLaren 1960; ten Doesschate 1968; much of the clustering. Other household factors Solon, Popkin, Fernandez et al. 1978). These related to diet and nutrition are suspected to be selected samples were supported by the country- more important. Identification of clustering can wide survey in Indonesia (Sommer 1982) with considerably influence the implementation of an about twice as many Bitot’s spots in males as in intervention. In Nepal it was estimated that mea- females (Figure 13.9). sures to prevent xerophthalmia were 7–34 times Interestingly, no evident sex difference was more efficient in high risk, highly populated ar- found in the detailed longitudinal study of Sinha eas than in remoter communities (West, Pokhrel, and Bang (1973) (see Figure 13.7). Prevalence of Khatry et al. 1992). VAD was extremely high in this area. Not only did it reach an astonishing peak of 15–16% in both sexes at the beginning of the rainy season; it SEX never fell to a rate below that of the WHO criteria (1% and 0.5%). Most of the evidence from animals and hu- Male preponderance was again found in the mans suggests that males are more susceptible to largest hospital series ever reported (Oomen

Figure 13.9: The prevalence of Bitot’s spots increased significantly (P>0.01) at 1, 2, and 3 years of age. Bitot’s spots were more common among boys (12.5/1000) than among girls (7.2/1000) (P<0.001) (Sommer, 1982). 13 Epidemiology 153

Figure 13.10: Age distribution of 6,300 cases of xerophthalmia in the Yap Eye Hospital at Jogjakarta, Indonesia, from 1935 to 1954. The male preponderance throughout increases with age (Oomen, 1961). 154

1961) from the eye hospital of Dr Yap Kie Tiong THE VADD CYCLE in Jogjakarta, Indonesia (Figure 13.10). Among the 6,300 cases the male:female ratio varied The flow diagram (Figure 13.11) has been with age. It was 1.4:1.0 in the preschool-age devised to give some impression of the kind of period, and 6.0:1.0 at around 10 years of age. combinations of risk factors that conspire togeth- In some cultures male children will tend to er at different stages of the life cycle to predispose appear preferentially in hospital studies because to the development and persistence of a VADD medical attention is more likely to be sought for problem in a community. them.

Figure 13.11: The Vitamin A Deficiency Disorders (VADD) cycle. 155

14 Retinoids and Carotenoids in General Medicine

INTRODUCTION potential to play a protective role in a number of common diseases. This is poorly understood at In addition to the various aspects of the global present. There is growing evidence that habitual problem of VADD with which this Manual is pri- consumption of diets that are rich sources of these marily concerned, there are other closely related carotenoids is associated with low incidences of and rapidly expanding areas of interest concer- some of these diseases. ning retinol and other retinoids in the context of human health. The aim of this chapter is to touch on these and to indicate where more detailed in- SECONDARY VAD formation can be obtained if desired. In addition to arising from some deficiency In most instances, the degree of vitamin A in the dietary intake of vitamin A, VADD may deficiency is not severe. As little attention is also occur as a result of some defect in the body’s given to VAD or any other nutritional deficiency physiology or metabolism. This may be called se- in general medicine, it is usually not until some condary or endogenous, as opposed to exogenous clinical manifestation, like night blindness or VAD, and attention will be given here to a variety conjunctival xerosis, occurs that deficiency is su- of forms that it can take. spected. Consequently, it is really not known how There are an increasing number of diseases in common deficiency is in these conditions, espe- which dosing with vitamin A appears to have a cially at the subclinical level. The various causes beneficial effect, even in the absence of any defi- of malabsorption are by no means uncommon and ciency. This area is clearly a part of the vitamin A most cases of secondary VAD have this basis. story and is rightly receiving considerable interest A rare group of disorders affecting the trans- at present. port of lipids in plasma includes two related Vitamin A is one of the vitamins that can conditions called abetalipoproteinemia and hy- lead to disease as the result not only of deficient pobetalipoproteinemia (see Table 14.1). Retinal but also of excessive intake. The latter is usually degeneration is a constant feature of these condi- known as hypervitaminosis A and may result from tions. Patients have usually been treated with vit- prophylactic or therapeutic use of vitamin A. This amin A and vitamin E for prolonged periods. Re- will be discussed later, in Chapter 15. It is more cent long-term assessment of 13 patients who had appropriate to deal with some aspects of an exces- been treated for an average of 11.7 years showed sive intake of retinol or other retinoids here. that visual impairment persisted even in those re- Many retinoids have been synthesized for ceiving treatment soon after diagnosis (Chowers, use in various diseases, especially those in which Banin, Merin et al. 2001). epithelial cell differentiation has been disturbed. On the other hand, only single cases have so Some attention will be paid to this topic here. far been reported of the enzyme defect, the trans- Finally, in recent years, many of the caro- port abnormality, the genetic problem in twins, tenoids, irrespective of whether provitamins or and the mutations in plasma retinol-binding pro- non-provitamins, have been shown to have the tein referred to in Table 14.1. However, there are 156 14 Retinoids and Carotenoids in General Medicine

Table 14.1 summarizes the various different types of secondary VAD that have been reported to date, together with the mechanisms. Disease Mechanism Celiac disease, sprue, obstructive jaundice, Impaired absorption of lipids including ascariasis, giardiasis, partial or total gastrectomy vitamin A Chronic pancreatitis In some cases secondary to zinc deficiency Chronic liver disease, especially cirrhosis Storage impaired by damage to liver cells; zinc deficiency enhances the effect Abetalipoproteinemia and hypobetalipo- Impaired lipid transport in blood proteinemia (1) Severe infection Loss of RBP in urine (2) Cystic fibrosis Excessive fecal loss, unrelated to degree of fat in stools (3) Enzyme defect Failure to cleave β-carotene in small intestine (4) Heterozygotic reduction of plasma RBP One case reported of keratomalacia due to reduced transport (5) Mutations in the gene for RBP Biochemical but not clinical deficiency (6) 141N and G75D mutations In human plasma RBP

(1) Alvarez, De Andres, Yubero et al. (1995) (4) Matsuo, Matsuo, ShiragA et al. (1998) (2) Ahmed, Ellis, Murphy et al. (1990) (5) Biesalski, Frank, Beck et al. (1999) (3) McLaren, Zekian (1971) (6) Folli, Viglioni, Busconi et al. (2005)

indications that genetic predisposition to VAD A recent study of 138 patients with cystic fib- may be of more importance than has been ge- rosis showed that serum retinol was inversely cor- nerally recognized hitherto (Ward, MacGowan, related with C-reactive protein (Greer, Buntain, Hornby et al. 2000). Larger doses than normal of Lewindon et al. 2004). As these patients have re- vitamin A are required to overcome the metabolic spiratory and other infections almost constantly, problems present, and where there is impaired fat the diagnosis of vitamin A deficiency is proble- absorption a water-miscible preparation should matic in relation to serum retinol and APR (see be given by mouth and perhaps initially intra- Chapter 7). Routine vitamin A supplementation muscularly. It is often not appreciated that the would seem to be necessary. usual form of oily vitamin A given intramuscu- larly stays in the muscle and is not released into the circulation to any extent (McLaren 1969). HYPERVITAMINOSIS A Recently, a second report was published of a loss-of-function mutation in CMO1 (see Acute toxicity Chapter 5), the enzyme that catalyzes the first Vitamin A toxicity may be either acute or step in the conversion of dietary provitamin A chronic. Acute toxicity usually follows a large carotenoid to vitamin A in the small intestine. single dose of vitamin A and is recognized by The patient had hypercarotenemia and hypo- symptoms suggesting an acute rise in intracranial vitaminosis A (Lindqvist, Sharvill, Sharvill et al. pressure – nausea, vomiting, and headache. Pro- 2007). viding that no further vitamin A is given, symp- 14 Retinoids and Carotenoids in General Medicine 157

toms rapidly subside and there are no permanent bilateral ovariectomy, which resembled the bone ill effects (see Chapter 15). The subjects have changes seen in postmenopausal women, was in- usually been young children in a supplementa- tensified by daily dosing with vitamin A (Pytlik, tion program or Arctic explorers short of food Cegiela, Folwarczna et al. 2004). The same group who have consumed the liver of polar bear, seal, (Sliwinski, Janiec, Pytlik et al. 2004) found that or their own sled dogs. the preventive action of alendronate on osteopenic changes was lessened by retinol. On a rather Chronic toxicity different note, Uchiyama, Sumida and Yamaguchi This is uncommon but may be very difficult to (2004) showed that dosing aged rats daily with diagnose. This is partly because the physician β-cryptoxanthin had an anabolic effect on bone may neglect to take a careful dietary history and components in femoral tissues, both in vivo and routine clinical laboratories are not able to mea- in vitro. sure serum retinol. Perhaps even more important In elderly human populations, several large is the fact that the symptomatology of hypervita- surveys of vitamin A intake and serum retinol and minosis A may be very varied, affecting a number occurrence of various fractures (Michaelsson, of different systems and sometimes mimicking Lithell, Vessby et al. 2003; Lim, Harnack, Lazo- other diseases. Common symptoms are headache, vich et al. 2004; Barker, McCloskey, Saha et al. vomiting, diplopia, alopecia, dryness of mucus 2005) have not provided consistent results. membranes, desquamation, bone and joint pain, A meta-analysis of vitamin A intake and bone liver damage including cirrhosis, hemorrhages (Crandall 2004) showed an association between into the skin, and coma. If the condition is not dietary and supplemental vitamin A and hip and correctly diagnosed and further excessive vitamin other bone fractures. Adverse effects seemed to A intake prevented, death can result. occur when the intake was about twice the cur- A review of both acute and chronic toxic ef- rent recommendation for adult females. This lat- fects of vitamin A has been published recently ter point is consistent with the data in Table 6.2 (Penniston, Tanumihardjo 2006). Much of what which show serum retinol levels rising constantly is said here has been said before but, importantly, with age. the authors place new emphasis on the possibility A further important contribution to our un- of harmful effects resulting from high and/or pro- derstanding of the problem was recently provided longed dosing with vitamin A in VAD-prevention by analysis of data of the NHANES I follow-up programs in developing countries. study in the United States (Opotowsky, Bilezikian High intake of vitamin A by whatever means 2004). Increased risk of hip fracture was shown to results in the adverse effects of exposure to toxic be associated with both low and high serum reti- metabolites of vitamin A such as retinyl palmita- nol concentrations, as evidenced by the U-shaped te, 13-cis-retinoic acid, and 13-cis-4-oxo-retinoic curve of association. This would be consistent acid (Hartmann, Brors, Bock et al. 2005). The with the animal studies. phototoxicity and photomutagenicity of these No firm conclusions can be drawn at present compounds is well-known (Yan, Xia, Cherng et and larger and better studies are needed. Mean- al. 2005). while, several observations come to mind. The bone lesions of hypervitaminosis A in experi- Vitamin A and bone mental animals do not resemble the osteoporo- It has long been known that both deficiency and sis associated with fractures in elderly humans. toxicity of vitamin A can damage bone in expe- Serum retinol is not by itself a valid indicator of rimental animals (see Chapter 6). Three recent vitamin A status. Moreover, many elderly people animal experiments contribute to this topic. In will have raised acute phase proteins and lowered the rat it was shown that osteopenia caused by serum retinol from infections and inflammation 158 14 Retinoids and Carotenoids in General Medicine

(see Chapter 7). Many other factors, including Surveillance of Congenital Abnormalities for lack of vitamin D, have not hitherto been taken 1980–94. This included 35,727 normal infants into account. and 20,830 infants with congenital abnormalities Governments in some countries advise elder- of all kinds. The mothers of 3,399 (9.5%) of the ly people not to consume more than one helping normal infants received vitamin A supplements, of liver per week – advice also given to pregnant as did 1,642 (7.9%) of those giving birth to infants women. The average vitamin A intake of most el- with malformations. The difference was highly derly people in developed countries tends to be significant (P<0.001). It was suggested that this higher and excessive supplementation is clearly analysis considered low or moderate doses of to be guarded against. In the context of food for- vitamin A to be protective and not teratogenic. An tification the issue of the risk of hypervitaminosis additional analysis, not pursued by the authors, A has been thoughtfully reviewed recently (see would be to group the infants according to whe- also Chapter 15) (Kraemer, Waelti, de Pee et al. ther or not their mothers received vitamin A and 2008). to compare the numbers of those with malformations in each group. Of 5,041 receiving vitamin A, Congenital malformations 1,642 (32.5%) had malformations; of 51,516 not For a long time, both vitamin A deficiency and receiving vitamin A, 19,188 (37.2%) had mal- toxicity have been known to induce congenital formations. This difference is also highly signi- malformations in experimental animals. In se- ficant at the P<0.001 level and also suggests that veral animal models similar malformations have vitamin A might be protective. All of the evidence been regularly produced by synthetic retinoids. available to date is reassuring with regard to the In humans the evidence for the involvement safety of the use of vitamin A supplements accord- of vitamin A deficiency is slender, amounting ing to generally accepted practice. mainly to isolated case reports. In striking cont- Fraunfelder, Fraunfelder and Edwards (2001) rast, there are numerous reports of high inciden- reported on their experience of treating various ces (>20%) of spontaneous abortions and birth resistant skin disorders using isotretinoin, 13- defects in the fetuses of women ingesting thera- cis-retinoic acid (Accutane/Roaccutane), over a peutic doses of 13-cis RA and other retinoids du- period of 20 years. Thirty-eight different signs or ring the first trimester of pregnancy (Ross 1999). symptoms of ocular abnormalities probably attri- The most controversial area has been the pos- butable to isotretinoin usage have been reported. sible teratogenic effect of excess vitamin A intake Those classified as “certainly” attributable to this during early pregnancy. The present consensus is cause include corneal opacities, decreased dark that in communities where VAD is known to be adaptation, decreased vision, myopia, and terato- present, vitamin A supplementation throughout genic ocular abnormalities. pregnancy should not exceed 10,000 IU per day Recent evidence for the involvement of vit- (see also Chapter 15). Where the diet supplies amin A in various congenital malformations has the RDA and more, even this level may be exces- been published. A population-based case control sive. Nevertheless, a recent pan-European study study in the United States (Botto, Loffredo, found no association between high vitamin A Scanlon et al. 2001) showed that levels of intake in early pregnancy and major malforma- maternal dietary intake of vitamin A or carot- tions (Mastroiacovo, Mazzone, Addis et al. 1999). enoids were not related to cardiac outflow tract One hundred and twenty infants exposed to more defects. However, compared with an average than 50,000 IU daily showed no malformations. intake from retinol supplements of less than Another very large study from Europe (Czeizel, 10,000 IU daily, an intake of 10,000 IU or more Rockenbauer 1998) reported on the population- was associated with a ninefold increase in relative based data set of the Hungarian Case-Control risk of transposition of the great arteries, but not of 14 Retinoids and Carotenoids in General Medicine 159

outflow tract defects with normally related arteries. optimal mode and the duration of its administrati- The toxicity of vitamin A during pregnancy on need further investigation. was assessed in the macaque or cynomolgus mon- A recent meta-analysis (Darlow, Graham key (Hendrickx, Peterson, Hartmann et al. 2000). 2000) confirmed the value of vitamin A supple- Extrapolation to humans using a safety factor of mentation, especially in terms of reduced oxygen 10 yielded safe levels of vitamin A during human requirement and mortality rate. In view of the pregnancy in the range of approximately 25,000 very small size of these infants, it was recommen- to 37,000 IU/day – well above the current gene- ded that intravenous vitamin A be considered. rally recommended safe level of up to 10,000 IU/ Another study urged the need for prolonged nu- day. tritional support until catch-up growth has been Recently a large international team of resear- achieved (Atkinson 2001). chers has observed two unrelated consanguineous In a large study of 120 infants with BPD it families with malformation syndromes with some was demonstrated that high-dose or once-per- features in common (Pasutto, Sticht, Hammersen week regimes did not give better results than the et al. 2007). Pathogenic homozygous mutations standard dose, which should continue to be given were identified in STRA6. This is a member of (Ambalavanan, Wu-Tzong, Tyson et al. 2003). a large group of “stimulated by retinoic acid ge- Recently, a large study of vitamin A sup- nes that encode novel transmembrane proteins, plementation for extremely low birth weight transcription factors, and secreted signaling mole- (ELBW) infants with an outcome of 579 (88%) cules or proteins of largely unknown function” surviving infants at 18 to 22 months was repor- (see Chapter 6). ted (Ambalavanan, Tyson, Kennedy et al. 2005). BPD was reduced without increasing mortality or neurodevelopmental impairment. NON-NUTRITIONAL DISEASES The results of a recent study of ELBW in- RESPONSIVE TO VITAMIN A AND/OR fants, in whom C-reactive protein and the RBP/ CAROTENOIDS TTR molar ratio (see Chapter 7) were measured, suggest that inflammation rather than vitamin A Bronchopulmonary dysplasia (BPD) deficiency might be the cause of low serum re- This disease is the leading cause of chronic tinol in these cases (Ambalavanan, Ross, Carlo lung disease in infancy and affects particular- 2005). This may prove to be yet another example ly very low birth weight infants. In the United of misdiagnosis arising from incorrect use of se- States, of the 3.9 million or so births per year, rum retinol. approximately 55,000 newborns weigh less than 1,500 g at birth. These are classified as very low Emphysema birth weight (VLBW). About 49,000 survive hos- Emphysema is another common lung disease, pitalization and about 24% of these suffer from affecting about 2 million Americans and causing BPD. The histological changes in the lungs that the deaths of 17,000 each year. Most cases are accompany this condition are similar to those associated with cigarette smoking but about 5% described in children dying with xerophthalmia. are caused by a deficiency of the enzyme α-1- Most of these infants have serum retinol and RBP antitrypsin. Emphysema has been induced in rats in the deficient range. They are also born with low by intratracheal instillation of a saline solution liver reserves. Several trials have shown benefits of elastase which destroys elastin fibers leading from dosing with vitamin A. The usual regime is to the collapse of alveolar walls in the lungs and 5,000 IU of water-miscible vitamin A intramus- resulting in emphysema. It has been shown that cularly three times a week for four weeks (Shenai retinoic acid can reverse the action of elastase 1999). The precise requirement of vitamin A, the and permits the growth of new alveoli (Massaro, 160 14 Retinoids and Carotenoids in General Medicine

Massaro 1997). This work offers hope for the the large study reported above have given consi- future development of a treatment for the disease derable encouragement to the addition of docosa- in humans. Chytil (1992) has reviewed the role of hexaenoic acid (DHA) to previously instituted retinoids in lung physiology. vitamin A treatment (Berson, Rosner, Sandberg et al. 2004). The course of the disease was slowed.

EYE DISEASES Age-related macular degeneration (AMD) It was mentioned in Chapter 2 that carotenoids Retinitis pigmentosa and some other commonly accumulate in various tissues and retinopathies certain carotenoids appear to have a predilection In addition to VAD, several other diseases af- for certain tissues. β-carotene accumulates prefer- fecting the retina may also have night blindness entially in the human ovary but the significance as part of their symptomatology. Almost all of of this is unclear. In primates there is a specific these are rare genetic disorders. Retinitis pigmen- uptake from the diet of two non-provitamin A tosa (RP) affects about one in 4,000 worldwide carotenoids, lutein and zeaxanthin, by the retina and eventually leads to total blindness. There are and other ocular tissues including the lens. The many different genetic forms of the disease. Over greatest uptake is by the macula lutea or yellow many years claims have been made for improved spot which contains the site of maximal visual vision with vitamin A and other treatments, but acuity, the fovea centralis, where there are only the only large, fully controlled trial with vitamin A cones and no rods (Harnois, Samson, Malenfaut over a prolonged period was carried out in the et al. 1989; Demmig-Adams, Gilmore, Adams III United States (Berson, Rosner, Sandberg et al. 1996). Zeaxanthin and meso-zeaxanthin (presum- 1993). It was claimed that patients receiving vit- ably formed from lutein) predominate in the amin A tended to experience a slowing in the rate fovea, and lutein predominates in the peripheral of deterioration of their electroretinogram com- retina (Bone, Landrum 2005). pared with those not receiving vitamin A. There Age-related macular degeneration (AMD) was no improvement in vision. There was con- is a disease, or group of diseases, that causes a siderable controversy about these findings at the devastating loss of vision in the elderly. It is the time. The publicity the report received has, un- commonest cause of blindness in the 75 years fortunately, meant that many patients have started plus age group. The cause is unknown but inten- self-medication with vitamin A in the absence of sive research, both epidemiological and clinical, any proven treatment. The equivocal nature of the is now underway, in which it is hypothesized that results may have been due at least in part to the diet plays a part. mixed nature of RP cases included in the trial. Lutein and zeaxanthin absorb short (blue) Recent evidence suggests that some forms of RP light. These are the highest energy photons across may be more likely to respond than others (Fariss, the visual spectrum to reach the retina. The ca- Zong-Yi, Millam 2000). rotenoids may serve to filter these, as well as to Bassen-Kornzweig Syndrome and Sorsby Fun- exert their known antioxidant properties. A recent dus Dystrophy are other rare diseases in which article (Schalch 2004) explains how, according to vitamin A has brought about some improvement one hypothesis, the level of concentration of these in rod function (Berson 1999). This whole com- pigments in the macular region might influence plex subject was discussed at greater length re- the ability of cones there to override the more cently (McLaren 2000). sluggish signals from rods in a zone of vision At the time of writing, 32 genes have been known as the “twilight” zone or zone of mesopic shown to be associated with retinitis pigmentosa. vision. Recently, the results of treatment of a subgroup of It has been shown that the macular pigment 14 Retinoids and Carotenoids in General Medicine 161

optical density (MPOD) is directly related to the macular degeneration or advanced age-related dietary intake of lutein and zeaxanthin and even macular degeneration in one eye are recom- more strongly to serum concentrations (Mares, mended to take the formulation proven in the Age- LaRowe, Snodderly et al. 2006). Related Eye Disease Study (AREDS) to prevent Lutein and zeaxanthin are present in the diet the development of advanced AMD by 25%. The mainly in various dark green, leafy vegetables. formulation consists of vitamins C, E, β-carotene One study found that eating spinach and collard and zinc (Coleman, Chew 2007). Another recent greens five or more times a week significantly successful preventive approach in AMD in- reduced the risk of the wet form of macular de- volved the taking of folic acid, vitamin B6 and generation (Landrum, Bone 2001). Another study vitamin B12 for an average of 7.3 years (Christen, (Van Leeuwen, Boekhoorn, Vingerling et al. Glynn, Chew et al. 2009). The subjects in this 2005) found a benefit from a high dietary intake placebo-controlled trial were women with a high of β-carotene, vitamins C and E and zinc. risk of ischemic coronary heart disease; 2,607 con- Recent animal research has contributed pos- trol and 2,598 treatment. The supplements reduced sible clues to our understanding of AMD. Mice the risk of new early AMD by 34%. This study lacking the component collagen XVIII/endostatin was embedded in a larger trial of various vitamin in the retinal pigment epithelium (RPE) develop supplements for the prevention of IHD. Combina- massive deposits under the RPE, abnormal RPE tion of these three B vitamins is known to lower and age-related macular degeneration, as in the serum homocysteine, an amino acid that is known human disease AMD. Progressive loss of visual to be raised in both IHD and AMD, but not shown function is due to decreased retinal rhodopsin to prevent IHD. These are promising results in content (Marneros, Keene, Hansen et al. 2004). early AMD prevention and progression. Gollapalli and Rando (2004) found that It has recently been suggested that there might RPE65 acts as a chaperone for all-trans-retinyl be a causal relationship between the increased oc- esters in the retina. Some forms of retinal and ma- currence of AMD and obesity (Johnson 2005). It cular degeneration in man are caused by the accu- is suggested that in this condition increased oxi- mulation of vitamin A based retinotoxic products. dative stress, changes in lipoprotein profile and The identification of RPE65 as the visual cycle increased inflammation might result in increased target for retinoic acid makes the development of destruction and decreased circulatory delivery to useful drugs to treat retinal and macular degene- the macula of lutein and zeaxanthin. ration feasible. However, there is currently a con- flict over the possible benefit (Owsley, McGwin, Age-related cataract Jackson et al. 2006) or harm (Radu, Han, Bui Cataracts, which consist of a gradual opacifica- et al. 2005) of a substance accumulation in the tion of the lens of the eye leading eventually to retina. Many pharmaceutical companies are mar- blindness, have hitherto only been capable of keting lutein and other carotenoid supplements treatment by removal of the damaged lens. On the for AMD (Stringham, Hammond 2005). basis that oxidative damage may be responsible, In a population-based cohort study it was a possible association of dietary intake of anti- found that a high level of intake of lutein and ze- oxidant nutrients, such as carotenoids, and vita- axanthin was accompanied by up to a 65 percent mins C and E, is being studied. As with AMD, reduced risk of neo-vascular AMD over up to a there is no conclusive evidence at present (Taylor 10-year period of observation. There were 3,654 1999). A recent study of monozygotic and di- participants at the start and 2,454 of those were zygotic twins in the UK showed that almost 50% of re-examined at 10 years (Tan, Wang, Flood et al. the variation in severity of nuclear cataracts could 2008). be explained by heredity, about 38% by age and Persons with intermediate risk of age-related only 14% by environment (Hammond, Snieder, 162 14 Retinoids and Carotenoids in General Medicine

Spector et al. 2000). In other parts of the world, leads to hypercarotenosis in which the blood level where nutritional deficiencies and other environ- is high and deposition occurs in the skin. This was mental factors are very different, results might also shown to be protective in this disease and the re- be very different. However, a recent study in India commended dose is about 180mg per day (Mos- found no evidence of slowing cataract progres- hell, Bjornson 1977). sion with the use of antioxidants (Gritz, Sriniva- san, Smith et al. 2006). In a recent study, spectral fundus reflectance CANCER CHEMOPREVENTION AND was used in a group of 276 subjects between 18 TREATMENT and 75 years of age to assess lens optical density. Lutein and zeaxanthin are the only carotenoids Many studies in experimental animals have found in the human lens. The results suggested shown the ability of both natural and synthetic that the presence of these pigments may retard retinoids to reduce carcinogenesis in organs such aging in the lens and, consequently, cataract as the breast, skin, liver, colon, prostate, lung formation (Berendschot, Broekmans, Klopping- and other sites. Natural retinoids are too toxic Ketelaars et al. 2002). for prolonged use in man, but some synthetic re- The US Food and Drug Administration (FDA) tinoids are well tolerated and have the potential has recently claimed in its evidence-based review for prophylactic use. In recent years, attention in system for health that “no credible evidence this field has turned more to the potential of some exists for a health claim about the intake of lutein carotenoids to act as agents in chemoprevention or zeaxanthin (or both) and the risk of age-related of cancer and some other conditions (see below). macular degenerations or cataracts” (Trumbo, Ellwood 2006).

DERMATOLOGICAL PHARMACEUTICAL USE OF SYNTHETIC RETINOIDS

Natural and synthetic retinoids influence epithelial cell proliferation and differentiation and are increasingly being used in dermatological practice to treat hyperkeratotic disorders. Etreti- nate and acitretin are very effective in psoriasis. 13-cis-retinoic acid and retinoyl-β-glucuronide suppress sebum production and are used in acne and acne-related disorders (Figure 14.1). Tre- tinoin has a beneficial effect in some cases of actinic keratosis, caused by chronic exposure to sunlight. At present only three retinoids are ap- proved for oral administration in many countries: isotretinoin, etretinate, and acitretin. Their teratogenic effects were discussed above. Erythropoietic porphyria is an uncommon disorder of porphyrin metabolism in which severe dermatosis may affect areas of the skin exposed Figure 14.1: Formulae of some highly active synthetic to sunlight. The regular ingestion of β-carotene retinoids. 14 Retinoids and Carotenoids in General Medicine 163

Retinoids have proved to be particularly ef- (CARET) referred to briefly above (see Omenn fective in the treatment of one form of leukemia, et al. 1996) tested the effect of daily β-carotene called acute promyelocitic leukemia (APL). In (30 mg) and retinyl palmitate (25,000 IU) on the this disease chromosome 15 is abnormal, with incidence of lung cancer, other cancers, and death a translocation that affects the gene for RAR-α in 18,314 participants. The participants were at (Norum 1994). The best results have been a high risk of lung cancer because of a history obtained with a regime that consists of all-trans of smoking or asbestos exposure. The trial was retinoic acid/arsenic trioxide-based therapy that is stopped in 1996 because of a significant increase applied as soon as possible after diagnosis (Hu, in lung cancer, death and cardiovascular disease Liu, Wu et al. 2009). In addition, selected RXR mortality. A six-year follow-up without further ligands, known as rexinoids, because they are intervention has shown that adverse effects have involved in the control of a variety of metabolic persisted but are no longer statistically significant. processes, are showing promise in cancer therapy Analyses suggest the excess risk of lung cancer (Nahoum, Perez, Germain et al. 2007 ). was restricted primarily to females, and cardio- As with vitamin A (see above), the types of vascular disease mortality primarily to females and cancer that appear to be most closely related to former smokers (Goodman, Thornquist, Balmes carotenoid intake are those of epithelial cell ori- et al. 2004). This subject has recently been exten- gin – lung, head and neck, esophagus and sto- sively reviewed (Sommerburg, Langhans, Salerno mach, colorectal, breast, cervix, and prostate et al. 2005). cancers. β-carotene has been studied the most, Animal experiments suggest that lycopene but in the case of prostate cancer lycopene ap- may be protective against smoke-induced lung pears to be most promising (Obermueller-Jevic, lesions by means of its inhibiting effect on the Olano-Martin, Corbacho et al. 2003). Trials of most abundant insulin growth promoting (IGF) β-carotene supplementation in high-risk groups binding protein in human serum. This protein, like asbestos workers and cigarette smokers have IGFBP-3, is known to be growth inhibitory and given unexpected results showing either negative apoptosis-inducing (Wang 2005). or even adverse effects of supplementation. It is generally concluded that supplementation with single nutrients, especially in large doses, is not CAROTENOIDS AND CHRONIC advisable. Increased intake of fruit and vegetables DISEASES provides a variety of antioxidant and other beneficial substances (Orfanos, Braun-Falco, Farber et There is a great deal of evidence to show that al. 1981). serum levels of β-carotene are lower than usual From several recent studies the provitamin A in many disease states. It has usually been as- carotenoid cryptoxanthin appears to have a pro- sumed that this indication of what might be called tective effect against the development of certain low “β-carotene status” is an adverse effect and, epithelial cancers. These include lung cancer if corrected, should lead to some health benefit. (Yuan, Stram, Arakawa et al. 2003) and cervical Surprise has been expressed that β-carotene sup- dysplasia (Goodman, Kiviat, McDuffie 1998). plementation under some circumstances, such as in Cellular retinol-binding protein 1 (CRBP-1) cigarette smokers with lung cancer, has been mediates retinol storage, and CRBP-1 down- related to deterioration rather than benefit (The regulation chronically compromises retinoic acid α-Tocopherol, β-Carotene Cancer Prevention receptor (RAR) activity. This leads to loss of cell Study Group 1994; Omenn, Goodman, Thornquist differentiation and progression of some human et al. 1996). The hypothesis has recently been put tumors (Farias, Ong, Ghyselinck et al. 2005). forward (Jandacek 2000) that low β-carotene re- The β-Carotene and Retinol Efficacy Trial flects the result of disease, serves as an indicator 164 14 Retinoids and Carotenoids in General Medicine

of the presence of disease, and is not causally re- a biomarker for tomato consumption and other lated. This might explain some of the anomalous carotenoids, such as phytoene and phytofluene or results obtained, but much more work needs to oxidized metabolites of lycopene, might be bio- be done. active (Petr, Erdman 2005). Modest increase in β-cryptoxanthin dietary intake, equivalent to a single glass of freshly squeezed orange juice daily, has been associated OTHER SYSTEMS with a reduced risk of developing polyarthritis (Pattison, Symmons, Lunt et al. 2005). Because retinoid-sensitive nuclear receptors On the other hand it has been suggested, on have been found in all types of cell examined, it is the basis of work in experimental animals, that likely that retinoids, especially in the form of re- abnormal oxidation products of β-carotene might tinoic acid, are capable of influencing any tissue. be responsible for the adverse effects of large Experimental research reports have provided ini- doses mentioned above. It is noteworthy that tial evidence for tissues and diseases that include the concurrent use of other antioxidants such the following. as vitamins C and E mitigated the formation of β-carotene and vitamin A metabolism appears β-carotene oxidation products and lung damage to be abnormal in patients with abnormal thyroid in experimental animals (Kim, Chongviriyaphan, function (Aktuna, Buchinger, Langsteger et al. Liu et al. 2006). 1993). See Chapter 11 for effects of vitamin A Until recently it has always been assumed that supplementation on subjects with iodine defi- hypercarotenosis had no harmful effects, causing ciency. only high blood levels and pigmentation of the A recent review (Berdanier 2003) on the role skin. It seems that there is still a lot to learn about of vitamin A in diabetes mellitus provides con- the safety as well as the benefits of carotenoid in- vincing evidence for the need for research to be gestion. focused on this problem. Recently it has been shown experimentally that serum retinol binding protein 4 contributes to insulin resistance in CARDIOVASCULAR DISEASE obesity and type 2 diabetes. A synthetic retinoid, fenretinide, increases urinary excretion of RBP4, The results of numerous epidemiological normalizes serum values and improves insulin re- studies involving β-carotene and other micro- sistance and glucose intolerance (Yang, Graham, nutrients in relation to the occurrence of ischemic Mody et al. 2005). Human studies confirm some heart disease are conflicting (Gaziano, Sesso of these findings (Graham, Yang, Bluher et al. 2005; Kabagambe, Furtado, Baylin et al. 2005; 2006). Gasparetto, Malinverno, Culacciati et al. 2005). Sudden infant death syndrome (SIDS) occurs A possible mechanism – if there is an effect – worldwide and occurs in 1.5:1000 live births in is reduced oxidation of low density lipopro- the United States. Peak incidence is between the teins, which seem to play a part in atherogenesis. second and fourth month of life. The exact cause Carotenoids other than β-carotene may be effec- is unknown but various factors are considered to tive or the intervention time may have been too be of importance. A study in Scandinavia found short for improvement to have been seen. A great a significant association between increased risk deal more research is required in this and other of SIDS and infants not being given vitamin A related areas touched on here. Lycopene and its during the first year of life (Alm, Wennergren, richest source in the average diet, tomato, have Norvenius et al. 2003). shown some promise in supplementation trials. Sickle cell disease is a chronic hemolytic However, it might be possible that lycopene is anemia occurring almost exclusively in African- 14 Retinoids and Carotenoids in General Medicine 165

origin populations. It has been shown that a high proportion have vitamin A deficiency. In comparison with people with normal vitamin A status, these have lower BMI, Hb and Hct, and ten times as many hospitalizations, increased pain and fever episodes (Schall, Zemel, Kawchak 2004). Routine vitamin A supplementation appears warranted. The role of vitamin A in mammalian reproduction and embryonic development has been reviewed (Clagett-Dame, DeLuca 2002). There is strong evidence that all-trans retinoic acid generated in situ is the functional form. It has been shown that mild vitamin A deficiency during fetal life can significantly reduce nephron number in the kidney (Burrow, 2000). Retinoids have an essential role in the branching growth of the ureteric bud in early fetal life. Many retinoid target genes involved in kidney development have been identified. 166 167

15 Control

INTRODUCTION rations as capsules or in other forms. Provision of a regular supply of vitamin A in these forms At the World Summit for Children in 1990 where they are needed has yet to be achieved in and at the International Conference on Nutrition vast areas of the world where VADD are a serious in 1992, the goal agreed upon was to “virtually public health problem. In addition to provision of eliminate vitamin A deficiency and all its conse- vitamin A preparations on a routine basis, there quences including blindness by the year 2000.” is widespread need for training in the recogni- 1995 passed without the mid-decade goal of tion of the various stages of VADD and in simple “ensuring that at least 80% of children under 24 measures for preventing recurrence of the prob- months of age receive adequate vitamin A through lem. Achievement of even these modest goals is a combination of strategies” being met. The first an important part of control. In recent years, it has edition of this book noted that “With very little been recognized that children with subclinical de- time remaining the goal for the year 2000 is most ficiency have an increased risk of dying. Many of unlikely to be reached in its entirety.” In Chapter these have severe measles, diarrhea, and/or PEM 12 the latest figures from WHO are given on and have been recommended in the past for what global prevalence of VADD and commented has been termed “targeted prevention” with vita- upon. min A supplementation. In these circumstances, None of the goals of the past has been fully they should be considered to be subjects of treat- met, and while blinding xerophthalmia has ceased ment that includes vitamin A. to be the major public health problem in young children that it once was, it is no longer being Prophylaxis monitored (see Chapter 12). Low vitamin A status Periodic high-dose vitamin A distribution remains at about the same level and is now recog- or supplementation in the community may be nized to be a common problem in some women viewed as a prophylactic extension of treatment and school-age children. in a hospital. The form of vitamin A is usually Several distinct types of intervention have, the same – capsules. The aim here is prevention. however, been designed to control VADD, and The measure, like treatment of the established these will be briefly introduced here before they case, should be superseded or supplemented by are considered at greater length later. something of longer-term effect. Research into multi-micronutrient deficiency disease states, a concept which emerged in the early Control of infections years of the third millennium, began only recently In recent years, the close two-way association (see Chapter 11). It will be some years before the between VADD and infectious diseases has be- control of such complex deficiency disease states come well recognized (see Chapters 9 and 13). can become part of a chapter like the present one. This has led to the view that the combination of vitamin A supplementation with immunization Treatment has distinct theoretical and practical advantages. For the foreseeable future, control of clinical In recent years, this combined approach has been VAD will include treatment of established cases tested and found not to have important adverse in hospitals and clinics. Central to this process is effects. The promotion of National Immuniza- the administration of high-dose vitamin A prepa- tion Days (NID) by many countries has provided 168 15 Control

a novel opportunity for vitamin A supplementa- Disaster relief tion of young children on a large scale. This and Victims of natural and man-made disasters are other aspects of the subject are discussed further especially susceptible to hunger and malnutrition. below. In these unnatural and emergency situations it is important that those who take on their care are Food fortification able to provide an adequate and balanced diet. Food fortification has a long history, starting in industrialized countries and spreading more recently to developing countries. Adding nutrients GENERAL CONSIDERATIONS to widely consumed foodstuffs is justifiable if vulnerable groups are unlikely to obtain their A report by the Administrative Committee nutrient requirements in any other way. Certain on Coordination, Subcommittee on Nutrition conditions need to be met if a program is to be (ACC/SCN Consultative Group 1994), Cont- successfully sustained. In the past few years this rolling Vitamin A Deficiency, set out results of form of VAD control has grown considerably, a comparative evaluation of different inter- and recently multi-micronutrient fortification has ventions, drawing on 46 individual evaluations. been introduced. Twenty-five were of supplementation interventions (vitamin A dosing), thirteen of dietary modifica- Dietary interventions tion, four of fortification, two of public health, Dietary interventions of various kinds would and two of breastfeeding. Supplementation is seem to be the logical approach to the problem easiest to implement and to evaluate. It is difficult in most circumstances. For those who approach to make valid comparisons between the different the problem for the first time it is usually a sur- types of intervention. In addition to this, it should prise and a shock to learn that the vast majority be realized that some studies have been carried of young children who go blind and die from out under carefully controlled research conditions vitamin A deficiency do so surrounded by readily while others have been of a more programmatic available sources of the vitamin. There are indeed nature. What goes on in strictly routine programs “many slips between cup and lip”: problems in is very unlikely to be reported in the scientific getting the vitamin out of the food and into the literature. child. Recent evidence that the bioavailability of In practice, choices will need to be made at provitamin A carotenoids from fruit and vegeta- the local level between the various control measu- bles is usually less than previously thought (Chap- res on offer that have been outlined above. Wher- ter 5) has complicated attempts to control VAD by ever possible, this choice should be made after means of dietary improvement. As a consequence, the situation has been assessed in some way (see greater emphasis is being placed on the value of Table 12.2). A baseline may then be established promoting the much more readily bioavailable and the efficacy of measures introduced evalu- preformed vitamin A sources wherever this is ated to some extent. Treatment of cases and pro- feasible. phylaxis for vulnerable groups can be instituted almost right away. Other measures take much New plants longer to be implemented fully. However, it should Recently plant breeding and genetic modification always be possible to start some nutrition and have been applied to the development of a number health education pari passu with capsule use. of high-carotene foods. The genetic modification It is important to bear in mind that interven- of rice to provide β-carotene (Golden Rice) has tions of this nature are taking place in a situation attracted worldwide attention and requires careful that can never be fully characterized and that is consideration in relation to the control of VADD. constantly undergoing change. It should there- 15 Control 169

fore be evident that it will rarely be possible to route of administration. The full dosing schedule attribute any benefit observed directly to the inter- should be given for all stages of clinical xeroph- vention. Good examples of this are provided by thalmia, not just for the more severe. This should the improvements that have been documented also apply to those eligible for what was termed in Bangladesh, India and Indonesia (see Figure “targeted prevention,” referred to above. 12.8). In rare cases where there is persistent vomit- Those engaged in the control of VADD should ing or severe diarrhea that might prevent inges- constantly keep the wider implications of health tion and absorption of the vitamin, intramuscular interventions in the backs of their minds. Health water-miscible vitamin A may be given. Recently, should be a sustainable state (King 1990). Some an aerosol of retinyl palmitate for inhalation has communities, by population increase, have out- been successfully field tested in preschool chil- grown, or are in the process of outgrowing, the dren in Ethiopia (Biesalski, Reifen, Fürst et al. carrying capacity of their ecosystem. In other 1999). Vitamin A was well absorbed through the words, there are too many (especially young) respiratory tract and might be preferable to injec- mouths to feed. The ethical implications of ap- tion. plying effective health interventions like vitamin Because of the known teratogenic effects of A supplementation under such circumstances large doses of vitamin A, care has to be exercis- should not be shirked (King, Elliott 1993). ed in the treatment of xerophthalmia in pregnant women. Active corneal lesions should receive the full treatment, but XN and X1B are treated with TREATMENT 10,000 IU (3.0 mg retinol, see Table 15.1) daily for two weeks. On several occasions, WHO has made recom- Associated medical conditions, such as PEM, mendations for the treatment of xerophthalmia. measles, and diarrhea, should receive appropriate The latest schedule is shown in Table 15.1. treatment. Frequently, secondary infections of the Retinyl palmitate in oily solution (as contain- eye are also present and should receive local and/ ed in capsules) by mouth is the preferred form and or systemic treatment.

Table 15.1: Treatment schedule for xerophthalmia for all age groups except women of reproductive ageA (WHO, UNICEF, IVACG Task Force, 1997).

Timing Vitamin A dosageB Immediately on diagnosis: <6 months of age 50,000 IU 6–12 months of age 100,000 IU >12 months of ageA 200,000 IU Next day Same age-specific doseC At least 2 weeks later Same age-specific doseD

A Caution: Women of reproductive age with night blindness or Bitot’s spots should receive daily doses ≤ 10,000 IU or weekly doses of ≤ 25,000 IU. However, all women of reproductive age, whether pregnant or not, who exhibit severe signs of active xerophthalmia (i.e. acute corneal lesions) should be treated as above. B For oral administration, preferably in an oil-based preparation. C The mother or another responsible person can administer the next-day dose at home. D To be administered at a subsequent health-service contact with the individual. 170 15 Control

PROPHYLACTIC SUPPLEMENTATION Universal distribution of vitamin A capsules in Bangladesh was reviewed (Bloem, Hye, Wijnroks This is often termed vitamin A supplementa- et al. 1995) and was shown to be an effective stration. In almost all countries where programs of tegy for reducing the incidence of night blindness, this nature are in operation, which numbered 103 but weaknesses were revealed in the rural areas of in 2009 according to UNICEF, the supplement the program. takes the form of a capsule. In India the supple- There are two main areas of concern with ment has been administered as a locally made regard to the safety of use of vitamin A prophy- syrup dispensed by teaspoon. laxis (Sommer, West 1996:394–9). The first of Children under the age of 6 years and preg- these relates to acute adverse effects in children. nant and lactating women constitute the main vul- Nausea, vomiting and headache have been re- nerable groups in communities where VAD has ported in several percent of children taking part been identified as a public health problem. They in large-dose programs (100,000 or 200,000 IU should participate in a supplementation program vitamin A). Severe vomiting (1.2%) was limited wherever this is deemed to be appropriate (Table to the children given 60 mg vitamin A; symptoms 15.2). lasted in almost all cases no longer than 12–24 A number of studies carried out recently in hours. In young infants there may also be ob- several developing countries have shed additional served bulging of the anterior fontanelle of the light on some aspects of vitamin A supplementa- skull, which is still open at that age. Like the tion programs. In Nepal (Christian, Schulze, Stoltz- other symptoms the effect is transient and there fus et al. 1998) supplementation with weekly is no evidence of after-effects (de Francisco, dosage of vitamin A or β-carotene at about the Chakraboty, Chowdhury et al. 1993). Several RDA level failed to eliminate maternal night controlled trials have been carried out and the blindness in about 60% of cases. It is possible that consensus is that the extremely slight risk is some other factors besides vitamin A may also fully justified in view of the potential benefit to play a part, such as zinc deficiency (see Chap- life and health (Florentino, Tanchoco, Ramos ter 11). In a similar study in Bangladesh (Rice, et al. 1990). The effects of neonatal vitamin A Stoltzfus, de Francisco et al. 1999) vitamin A or supplementation were followed over a period β-carotene failed to prevent subclinical vitamin A of many months and no evidence was found for deficiency in lactating mothers and their infants. an adverse effect on development or growth (van

Table 15.2: High-dose universal-distribution schedule of vitamin A for prevention of deficiency (WHO, UNICEF, IVACG Task Force 1997). Infants <6 months of ageA • Non-breastfed infants 50,000 IU orally • Breastfed infants whose mothers have 50,000 IU orally not received supplemental vitamin A Infants 6–12 months of age 100,000 IU orally, every 4–6 monthsB Children >12 months of age 200,000 IU orally, every 4–6 monthsB Mothers 200,000 IU orally, within 8 weeks of delivery A Programs should ensure that infants <6 months of age do not receive the larger dose intended for mothers. It may therefore be preferable to dose infants with a liquid dispenser to avoid possible confusion between capsules of different dosages. B Evidence suggests that vitamin A reserves in deficient individuals can fall below optimal levels 3–6 months following a high dose; however, dosing at 4–6 months intervals should be sufficient to prevent serious consequences of vitamin A deficiency. 15 Control 171

Dillen, de Francisco, Overweg-Plandsoen et al. who subsequently drop out are usually those in 1996; Humphrey, Agoestina, Juliana et al. 1998). greatest need of the service. A recent review of the vitamin A supplemen- Targeted distribution offers the greatest flexi- tation of young infants (Humphrey, Rice 2000) bility and cost-effectiveness, and best utilizes concludes that “a regimen in which mothers existing contacts between health providers and are given 200,000 IU vitamin A at delivery and the community. This requires planning and coor- their infants receive four doses of 50,000 IU (i.e., dination and needs to be sustained if the results at birth and with their immunization contacts) are to be better than those that come from the pas- would allow nearly all infants to enter the second sive targeting of only those children who attend half of infancy in adequate vitamin A status.” health clinics. The second issue concerns vitamin A pro- Universal distribution requires dosing of all phylaxis during and shortly after pregnancy. children of the vulnerable age group in a high-risk Vitamin A and related compounds in large doses area, usually on a semi-annual basis. It is parti- are known to be teratogenic in early pregnancy cularly this type of distribution that often suffers (Nau, Chahoud, Dencker et al. 1994) (see Chap- from low coverage, as mentioned above, due to ter 14). High-dose vitamin A given at or short- all kinds of logistical problems. ly after delivery has been shown to raise breast Before presenting some lessons learned from milk levels considerably for a number of months. individual VAS programs it is necessary to draw Frequently this is a rare opportunity to apply pro- attention to the source of one of the special prob- phylaxis. Present recommendations are that the lems that is currently in dispute and which is dealt high-dose prophylaxis (200,000 IU) would be with below. This concerns the lack of uniformity given only to breastfeeding women at delivery or of response to VAS of children in different age during the infertile postpartum period, currently groups under the age of 72 months. Between thought to last 4–6 weeks. IVACG (1998) has the ages of six and 72 months many studies issued a statement on “Safe doses of vitamin A have shown an increase in survival with VAS of during pregnancy and lactation.” WHO current- 24% (Mayo-Wilson E., Imdad A., Herzer K. et al. ly recommends that the relatively small increa- 2011). Between 1–5 months there has been little sed need for vitamin A during pregnancy should or no response, for unknown reasons (West K.P. be met by diet, or a supplement not exceeding Jr, Katz J., Shrestha S.R et al. 1995, WHO Guide- 10,000 IU daily throughout gestation (WHO, line 2011: Vitamin A supplementation in infants UNICEF, IVACG Task Force 1997). This is a 1–5 months of age). second edition of its guide to the use of vitamin A In newborns dosed within 2–3 days of birth supplements and prevention of vitamin A defici- mortality has been reduced in some settings by ency and xerophthalmia. >15% (Humphrey, Agoestina, Wu et al. 1996, As mentioned above, supplementation does Klemm, Labrique, Christian et al. 2008, Tielsch, not address the underlying cause(s) and is an Klemm, West et al. 2008), but not in others emergency measure. It should be accompanied by (Benn, Diness, Roth et al. 2008, Malaba, Iliff, dietary counseling and/or food fortification. Re- Nathoo et al. 2005). search studies (Sommer, West 1996: 388–409) The Nepal National Vitamin A Program have shown that full implementation of the (NNVAP) has been independently evaluated and schedule is very efficacious. However, expe- provides important lessons for all in this field rience has shown that once a supplementation (Fiedler 2000). The NNVAP commenced in 1993 program has been integrated into the routine, in eight of the country’s 75 districts and plan- primary healthcare system efficiency tends to fall ned to implement nationwide coverage by 2003. to unacceptably low levels (Sommer, West 1984). Table 15.3 summarizes some key aspects of the Those who are not covered in the first place or NNVAP. 172 15 Control

It consists primarily of twice-yearly distri- group. The impact was on mortality between 2 bution of high-dose vitamin A capsules to all weeks and 3 months after treatment, with no ad- children 6–59 months of age. Female community ditional impact after 3 months. Another commendation of extension of the application of periodic vitamin A supplementa- Table 15.3: Nepal National Vitamin A Program (NN- tion comes recently from a demonstrated de- VAP) (Fiedler, 2000). crease in mortality rate in HIV-infected children • VAD responsible for 9,000 child deaths/year in Uganda (Semba, Ndugwa, Perry et al. 2005). Recently in the Indian pediatric literature • VAD responsible for 2,500 blind (Kapil, Prakash 2004), the relative merits of de- • VAS costs US$ 1.25 for 2 capsules/year livery of vitamin A by capsule or by syrup have • Cost/averted death US$ 327 been discussed. In each case, cost is low and effi- • VAS saves US$ 1.5 million/year from cost ciency of delivery is high. However, with millions of diarrhea and measles of doses being dispensed annually these considerations merit further attention. Long-term coverage has long been known to health volunteers play a key role. The annual cost be disappointingly low in many circumstances. A is US$ 1.7 million. The cost per averted death is recent study in central Java, Indonesia (Martini, US$ 327. The incidence and severity of diarrheal Rice, de Pee et al. 2005) found that vitamin A cap- disease and measles have been reduced, saving sule coverage among postpartum women overall the government yearly US$ 1.5 million. was as low as 19.3%. Mothers active as health A recent update on NNVAP presented at the cadres, those whose infants were delivered by a XXII IVACG Meeting stated that coverage rates midwife or a doctor, or those who delivered their of 85% and above had been maintained since in- baby in village maternity clinics had higher rates ception in 1993 and all 75 districts were covered than others. None achieved more than 50%. from October 2002. Bitot’s spot (X1B) prevalence Experience with vitamin A prophylactic sup- was reduced to 0.33% (WHO cut off <0.5%). plementation in India has been more extensive NNVAP was also seen as a major factor in the both over time (the national program started in reported decrease in child mortality. This cost- 1970) and in terms of numbers covered. Detailed effective program, at just US$ 0.74/child/year, accounts of this work have recently been pub- has also contributed significantly to a deworming lished (Reddy 2002; Kapil, Goindi 2004). In the program at almost no extra cost. first of these papers, and in some other sources, dis- Even with such a successful and apparently cussion has taken place concerning the reports in cost-effective supplementation program as this, newspapers in 2001 that vitamin A caused deaths the question must be asked about its continuance. of children taking part in a mass supplementation Other countries, especially India, are in a similar campaign in the state of Assam. Like other similar position and decisions need to be taken. reports over the previous years (see earlier) the In South India (Rahmathullah, Tielsch, Thu- general consensus was reached that vitamin A lasiraj et al. 2003) the impact of supplementing supplementation could not be held responsible for newborn infants with vitamin A on early morta- the deaths that occurred (Sommer, West 2002). lity was assessed. 11,619 newborn infants re- Accounts have been published of a number ceived 24,000 IU vitamin A by mouth or placebo of national vitamin A supplementation programs, on days 1 and 2 after delivery. Mortality at age which report difficulties as well as achievements. 6 months was recorded. Those in the vitamin A As examples, attention may be drawn to the re- group had a significant 22% reduction in total cent India experience (Kapil, Goindi 2004), the mortality compared with those in the placebo vitamin A capsule distribution among infants 15 Control 173

aged 6–11 months in rural Bangladesh (Akhter, THE SPECIAL ISSUES OF Stallkamp, de Pee et al. 2004), and experience 1) NEWBORN VAS AND with enrichment of breast milk with vitamin A 2) EFFECTS OF VACCINES from several Asian countries (Harvey 2002). A recent study of refugee women in Sudan (Strand- 1. Newborn VAS ers, Northrop-Clewes, Thurnham 2004) showed a Following upon a number of very large double- prevalence of low milk retinol concentrations and blind controlled field trials that have been under- elevated mammary epithelial permeability. The taken in several low-income countries in South postpartum megadose of vitamin A alone did not Asia and Africa on the effects of a single dose appear to be sufficient for a substantial and pro- (of 50,000 IU vitamin A) to newborns, no general longed improvement of vitamin A status. agreement concerning widespread implementa- A recent study in Nepal (Grubesic 2004) tion of VAS has been reached at the present time. looked at the relationship between the prevalence The situation was discussed by one of the propo- of health indicators of malnutrition, diarrhea, and nent groups (Tielsch, Klemm, West et al. 2008). acute respiratory infection and the consumption It was suggested that for VAS to reduce mortality of vitamin A rich food and the supplementation significantly, two things need to be present; en- status of three groups of children (regularly sup- demic VAD and high infant mortality rates (most plemented, supplemented once, and never supple- such deaths are attributable to infectious dis- mented). Regardless of the amount of vitamin A eases which are usually responsive to vitamin A). rich foods consumed, those regularly supple- On the whole the countries chosen for study in mented were better protected against all three Asia tend to meet these conditions while those in groups of diseases than others. Africa do not. An investigation of breastfeeding in Kenya Apart from infants whose mothers were given by mothers with low vitamin A status showed vitamin A during pregnancy and infants aged 1–5 that infants frequently remained deficient when months at the time of supplementation, the sug- weaned onto diets low in vitamin A sources gestion holds good: i.e. VAS significantly re- (Ettyang, Oloo, Lichtenbelt et al. 2004). duces mortality of newborn infants and those In the Philippines, the National Vitamin A 6–59 months of age. The reasons for the age- Supplementation Program has been operative related differences in outcome are not known, among preschool children for many years. A although it has been observed that the effect recent report (Pedro, Madriaga, Barba et al. 2004) results from case fatality rather than from in- showed that two aspects of this experience were cidence of infection (Tielsch, Rahmathullah, of major concern. Firstly, the impact of vitamin A Thulasiraj et al. 2007). supplementation on serum retinol was mainly Now, after the event as it were, suggestions confined to those with severe vitamin A defici- are being made and evidence being sought to try ency. Secondly, the effect did not persist for 6 to explain the conflicting results. It has been clear months, which is the interval between doses. It for a long time that we have not been comparing was recommended that in areas of low prevalence like with like here. In general vitamin A status is of VAD, distribution of capsules should be tar- different, as shown by Thurnham (2010) recently. geted to stunted children. In areas of high preva- Also it may be significant that the data previously lence, capsules should be distributed to children given in Figure 12.5 show the Africa region to aged 1–5 years three times a year. be better off for vitamin A supplies than is South The Nutrition Section, UNICEF HQ (2005) Asia, and infant mortality rates tend to be lower. issued a monograph entitled “Vitamin A Supple- In this context and from earlier knowledge, data mentation: Progress for child survival” that sum- in Tables 12.1 and 12.2 are particularly difficult marizes most of the aspects of the subject. to interpret. Africa generally has rates up to 174 15 Control

twice as high as Southern Asia. On the other hand West Africa. The topic is dealt with below. it might be that application of the FAO food in- Meanwhile, as so often happens, things tend take data is misleading and Africa deteriorates to move on and decisions are taken without all while Southern Asia improves. the possible questions to any particular situation The main contribution of Thurnham (2010) being answered. In 2008 the group of research is his analysis of the varying practices regarding workers at the Johns Hopkins Bloomberg School the initiation of breastfeeding in the newborn of Public Health, Baltimore, USA which had VAS studies. The well-known use or discarding been involved in the research in Asia, received of colostrum, of high vitamin A activity, seems to support for what is termed the A2Z Program. It have been hitherto largely ignored and the relevant is anticipated that about 200,000 lives in South data in these studies are very scanty. Vitamin A Asia would be saved per year by high coverage status of the newborn is obviously enhanced of supplementation of newborns with vitamin A and it should not be difficult to encourage the (50,000 IU). feeding of colostrum where appropriate. Thurn- ham rightly draws attention to two papers that 2. Effects of vaccines show that failure to provide human milk to the The group of Aaby, Benn et al. has investigated neonate in the first hour after delivery will in- programs of immunization of young children in crease the risk of neonatal mortality (Edmond, West Africa for many years. They took the op- Zandoh, Quigley et al. 2006, Mullany, Katz, Li portunity of the introduction of VAS to the WHO et al. 2007). In two related Opinions (Klemm program, EPI, referred to below to study the pos- 2010, Benn 2010) it is good to see some common sible effects of the combination. They became ground being shared. aware of the apparent differences between dif- In the same issue of the Sight and Life ferent countries with regard to the effect of VAS Magazine (McLaren 2010), in a personal commu- on newborns and published an hypothesis (Benn, nication a former colleague of mine raises an issue Balé, Sommerfelt et al. 2003). In outline it pro- which those involved would do well to take heed posed ‘that the effect of VAS may depend on the of. He asks whether a detailed comparison has amplification of non-specific effects of vaccines been made of the precise chemical formulation on childhood mortality’. They suggested that of the vitamin A product used, including details different effects in different age groups may be of the possible presence of any isomers. Such in- explained by different factors operating. They formation appears not to be in any related publi- point out that in their own and in two other cations. Differences between studies in this regard studies boys are more benefited than girls, who might easily influence results. The raising of such may actually be harmed (Benn, Fisker, Jorgensen questions and the possible existence of others yet et al. 2008). Subsequent reports of this group have undiscovered imply lack of thoroughness in de- tended to strengthen doubts previously raised sign in this field. A question was asked about about some aspects of recommendations WHO monitoring of quality of vitamin A capsules for continues to make about VAS in the presence of supplementation programs (Newton, Owusu- routine vaccinations of young children. In Open Agyei, Kirkwood 2007) and answered (Mannar, Commentary recently (Prentice, Savy, Darboe et Court 2008) but the issues raised were entirely al. 2009) the complexities of the issues were different. revealed. More recent analysis from the same The question of the possibility of differences group (Savy, Edmond, Fine et al. 2009) shows arising in the effects of VAS supplementation in how the prevailing multinutrient approach is the presence of the use of vaccines, including sex invading this field also. Yet again from Guinea- differences, has been repeatedly raised by the Bissau comes evaluation of BCG vaccination Danish group of researchers in Guinea-Bissau, which includes adverse effects of vitamin A in 15 Control 175

girls (Roth, Benn, Ravn et al. 2010, Shann 2010). extent with seroconversion to measles (Semba, A line has to be drawn somewhere, or two Munasir, Beeler et al. 1995). Subsequent studies lines here, without a prospect of satisfactory re- have failed to confirm these initial concerns for solution. measles (Benn, Aaby, Balé et al. 1997), DPT (van Dillen, de Francisco, Overweg-Plandsoen 1996) or polio (Semba, Muhilal, Mohgaddam et al. PREVENTION AND MANAGEMENT 1999). A large study in Ghana, India and Peru OF INFECTIOUS DISEASES with DPT, polio and measles immunization (WHO/CHD Immunization-Linked Vitamin A After its success in eradicating smallpox Supplementation Study Group 1998) confirmed WHO established its Expanded Program on the safety of the intervention. Unfortunately it Immunization (EPI), employing BCG, OPV, failed to find any sustained benefits in terms of diphtheria, tetanus, pertussis, and measles vac- vitamin A status beyond the age of 6 months, or cines. By 1990 nearly 80% coverage had been in terms of infant morbidity. achieved and other vaccines and measures, such as Local circumstances have varied greatly and vitamin A, have been added. Fifty percent of global socioeconomic and familial characteristics have coverage of vitamin A, one dose per year, was influenced a program’s success (e.g., Gorstein, achieved in 1999. This rose to 68% in 2004 and Bhaskaram, Khanum et al. 2003; Panganbari- 82% in 2008. The greatest gains have been in buyan, Scherbaum, Erhardt et al. 2004). Ching, the least developed countries. The most success- Birmingham, Goodman et al. (2000) made an ful aspect of EPI, or of any other immunization estimate of the cost of integrating vitamin A programs in developing countries, usually ac- supplementation into immunization campaigns. companied by VAS in recent years, has been that During 1998, in 41 countries, more than 94 against measles, and this can be traced throughout million doses of vitamin A were administered. the book through use of the Index. It is estimated that this helped to avert Attention was drawn earlier to the inter-rela- about 169,000 deaths. During 1999, delivery of tionships of infections and vitamin A status (see more than 97 million doses in 50 countries Chapters 9 and 13) that need to be taken into ac- helped to avert an estimated 242,000 deaths. The count in the control of VADD. Immunizations that estimated incremental cost per death averted have been developed for certain infectious disea- was US$ 72 (range 36–142) in 1998 and US$ 64 ses, especially measles, may provide an opportu- (range 32–126) in 1999. The estimated average nity for a joint approach. The Program WHO EPI total cost of providing supplementation per death advises that “any EPI contact after the age of six averted was US$ 310 (range 157–609) in 1998 and months is appropriate for supplementing the in- US$ 276 (range 139–540) in 1999. Costs per fant or young child [with vitamin A]. The visit for death varied by campaign, depending on the measles vaccine at around 9–11 months is espe- number and proportion of the child population cially suitable” (WHO 1994). reached, number of doses received per child, and It has been demonstrated on a number of oc- child mortality rates. casions that integration of vitamin A supplemen- National Immunization Days (NID) were tation into a successful immunization program, introduced in the 1990s and at that time were like EPI, can result in greatly increased coverage particularly focused on the elimination of polio- (Karim, Shahjahan, Begum et al. 1996). There myelitis as a threat to young child health. It was is evidence that immunization against measles soon appreciated that this might be an additional is already having a favorable impact on redu- opportunity to improve vitamin A status where cing corneal blindness in young children. High- considered necessary, although optimal results dose vitamin A was shown to interfere to some appear to require more frequent dosing than 176 15 Control

once a year. While NIDs for polio continued they Potentially, food fortification offers a direct, offered a sound delivery structure and an unparal- effective and sustainable way to correct VAD. leled reach for VAS. In 1997 more than 450 mil- However, in practice it has sometimes proved lion children, almost two-thirds of those under 5 difficult to meet all the necessary criteria. In the years of age, were immunized during polio NIDs. early efforts to fortify staple foods technological For VAS human and financial resources are very obstacles had to be overcome. Today such prob- reasonable (see Chapter 1). In 1998 75% of the lems are not considered to be limiting factors. 118 countries where VAD is a known or suspected Although technologically possible, implementa- public health problem conducted NIDs and over tion of food fortification has proved to be a com- 60 million children received VAS. plicated and long lasting process. Niger became one of the first countries in A food item to be fortified should be con- Africa to effectively integrate vitamin A supple- sumed regularly by most of the target popula- mentation into NIDs for polio eradication. Main- tion in certain quantities. There should be no risk taining high vitamin A supplementation coverage of overdosing for those consuming the highest has always proved to be difficult. In this instance, quantities. Further criteria are that the vitamin A coverage exceeding 80% was achieved. This led should not affect the appearance, color, texture, or in 1999 to the first ever National Micronutrient organoleptic properties of the food, so that it is ac- Days (NMD) in Africa (Aguayo, Baker, Cre- ceptable to the consumer. The stability of the vita- spin et al. 2005). As NIDs are being phased out, min A should remain at an acceptable level during although not as early as expected in some coun- processing, transport, storage and cooking. tries, this new approach might prove to be an im- It is evident that fortification is only possible portant way forward. if the processing of the food in question is to some Vitamin A supplementation has never been extent centralized. This is also necessary in order incorporated into a program of oral rehydra- to provide adequate quality control. Difficulties to tion for the prevention and treatment of diarrheal be overcome in relation to such matters include diseases (see Chapter 9). Beneficial effects of the passing and implementation of food laws and vitamin A were demonstrated years before the regulations, and their continuing enforcement. same were reported for zinc. It is therefore puzz- The long-term financing of all the costs involved ling to see that UNICEF recommends the latter, in fortification has proved to be a stumbling block while no mention is made of vitamin A (Baqui in some programs. 2006). This is also contrary to other statements by Like supplementation, food fortification may the UN System (see Chapter 1 and also below) on be universal or targeted. In the former, fortifica- these micronutrients. tion is applied throughout the population. In the latter, it applies to specific groups (e.g., to supple- mentary feeding programs for pregnant women, welfare recipients, school children, or those re- FOOD FORTIFICATION ceiving complementary foods). Fortification of food aid in emergencies is an important issue (see In industrialized countries, food fortification later). Several research studies (Sommer, West has long been an accepted strategy for improving 1996:410–30) have demonstrated conclusively micronutrient nutriture, including that of vita- that fortification can significantly improve vitamin A. In Denmark, during World War I, an epi- min A status of a whole population. demic of xerophthalmia paralleled the substitu- In developing countries, many foods have tion of butter by margarine, which lacked vitamin A. been fortified with vitamin A or imported as forti- Today, margarine is among those food items most fied products. These include wheat, rice and other frequently fortified with vitamin A in the world. grain products, tea, dairy foods (especially dried 15 Control 177

skim milk), margarine, edible oils, formula foods and specialty items. Sugar in Latin America and monosodium glutamate (MSG, a popular flavor enhancer) in Southeast Asia were the first vehicles for vitamin A fortification to be extensively tested, widely distributed and evaluated for their public health impact (Sommer, West 1996: 411–25). Neither of these food substances is an ideal vehicle from the nutritional point of view, but even so, sugar fortification is in operation at the present time in several countries in Central America and is in the process of being introduced in some countries in Latin America, in Africa, and elsewhere. Figure 15.1: Effect of MSG (monosodium glutamate) From the early 1970s, a number of Central fortification (Muhilal, Permaesih, Idjradinata et al. 1988). and South American countries fortified white, refined sugar with vitamin A; much of the development took place in Guatemala. Surveillance over ter. In some countries there are now more than a number of years demonstrated a positive impact 20 different food items fortified with vitamin A of the program (Arroyave 1986). Adverse internal posing possible problems of excess intake. One and external circumstances halted the program study from Guatemala (Krause, Delisle, Solo- for nearly eight years. In 1988, six months after mons 1998) showed that poor urban toddlers were a program restart there was significant reduction obtaining as much as about half of their recom- in the number of children with low serum reti- mended vitamin A intake from fortified foods. nol (26% to 10%) and of those with an abnormal Other evidence suggests that even prolonged for- RDR (33% to 14%) (Pineda 1993). tification of a single food item may not have been In the 1970s, fortification of MSG was pur- sufficient to make a lasting impact on a problem sued in the Philippines and in Indonesia. The proof VAD. It is being suggested that both fortifica- grams proved to be efficacious in children to raise tion and supplementation may be required. serum retinol levels, reduce the prevalence of xe- Despite the limitations and the drawbacks, rophthalmia, improve linear growth, and reduce the examples given here have demonstrated that mortality (Solon, Latham, Guirriec et al. 1985; food fortification can be a very powerful strate- Muhilal, Permaesih, Idjradinata et al. 1988) (see gy. In order for food fortification to be successful, Figure 15.1). effective collaboration of all parties involved is As a result of the greatly increased interest in essential, including scientists, industry, public ad- micronutrient fortification of foods in developing vocacy groups, legislators and politicians. Som- countries in recent years several trends have be- mer and West (1996:410–30) end their review of come evident that are likely to pose difficulties the subject as follows: “... implementing a natio- in the near future. Fortification of food with a nal fortification program is a major undertaking number of micronutrients at once is rapidly on that requires sound scientific rationale, industrial the increase. For example, biscuits for primary capacity, training, advocacy, adequate legislative schoolchildren in South Africa are being fortified support, economic viability, community accep- with iron, iodine, and β-carotene (van Stuijven- tance and long-term sustainability, monitoring berg, Kvalsvig, Faber et al. 1999). The evaluation and quality control.” of the effects of such studies and comparison of In recent years food fortification programs results from different studies is a complex mat- have commenced in many countries. There have 178 15 Control

been reports on many of these at IVACG meetings of essential vitamins and minerals (such as vit- and in the Sight and Life Magazine. One feature amin A, iron, zinc and iodine) in low doses that has been the successful fortification of novel contribute to its consumers’ daily needs. food items, for example vegetable oil in Morocco (Borghos, Rahmani, Elkari et al. 2002) and cassava in Nigeria (Asonye 2002). DIETARY MODIFICATION As evidence continues to accumulate in support of the low bioavailability of carotenoids (see The ACC/SCN Consultative Group (1994) Chapter 5), the need for the inclusion in most tra- pointed out that there are four types of strategies ditional diets of some vitamin A in its preformed aimed at achieving the goal of dietary modifica- state (i.e. from animal sources) seems unavoid- tion: able (see Chapter 12). In this context mention was 1) Nutrition education or communication, often earlier made (see Chapter 14) of recent discussion using a social marketing approach, to im- on the possibility of excessive vitamin A fortifica- prove practices related to the consumption of tion contributing to bone damage available vitamin A rich food sources. Recently, the case has been made for inclu- 2) Horticultural interventions (or home food pro- sion of selenium and iodine in multi-micronutri- visioning, e.g. home gardening), that aim to ent biofortification programs which already in- increase availability of vitamin A rich foods. clude vitamin A, iron and zinc (Lyons, Stangoulis, 3) Economic/food policies affecting availability, Graham 2004). price and effective demand of vitamin A rich Home fortification refers to a relatively new foods. approach where vitamins and minerals usually 4) Technological advances concerning food pre- used for food fortification are added to the com- servation, plant breeding, etc. monly eaten, prepared food just before consump- Strategies 2–4 aim to improve the availability of tion (Hyder, Zlotkin 2008; de Pee, Kraemer, van vitamin A rich foods. Strategy 1 aims to improve den Briel et al. 2008). Home fortification has been their consumption. applied at increasingly larger scales since 2002. The report goes on to review 13 dietary modi- Through home fortification the micronutrients fication evaluations. Nine of these incorporated can be targeted and specifically dosed (one RNI some education/communication project activities, or less per individual), and as they are added to seven included home gardening (four of which the plate or bowl of food just before eating, they were combined with social marketing activities), are not subject to processing and preparation and one study examined changes in consump- losses. The concept of home fortification has been tion in response to naturally occurring changes in proven to be efficacious for reducing deficiencies prices and availability of vitamin A rich foods. among young children under controlled circum- Most of the evaluations were of pilot projects or stances, and has been piloted in large emergency field trials. and refugee settings. It was possible to demonstrate positive change The concept of home fortification was de- to varying degrees in several projects. These in- veloped in the 1990s with the earliest forms of cluded improvements in knowledge, attitude and single-dose sachets containing micronutrients in practices (KAP) in northeast Thailand (Smita- a powdered form that could be easily sprinkled siri, Attig, Dhanamitta 1992) and West Sumatra on any individual portion of food prepared in the (Pollard, van der Pasch 1990). In Bangladesh household. Designed in such as way as to mini- there was a 40–60% increase in production and mize changes to the taste, color, or texture of the consumption by young children of green leafy food upon addition of the micronutrient powder, vegetables and yellow fruits. There was increased these micronutrient powders contained a mixture awareness of night blindness and prevalence of 15 Control 179

other kinds, it is of special importance to take into account the efficacy or effectiveness of the intervention. An intervention is effective if it can be shown to have produced the desired effect; in this case to have improved vitamin A nutriture. Consequently, to have increased the vitamin A (usually provitamin A) intake does not satisfy this criterion. Improving vitamin A status (measured Figure 15.2: Composite profile of age-specific by serum retinol or some other method) may be protection against xerophthalmia and low serum retinol considered to have done so. levels conferred by dietary intakes of selected types Recent work suggests that provitamin A caro- of foods. Solid bar denotes ages for which epidemio- logic evidence is strong. Dashed bar denotes ages at tenoids, in their natural form in foods, are not as which some evidence of protection exists for a food effective as has often been assumed in the past (de (Sommer, West, 1996: 130–137). Pee, West, Muhilal et al. 1995; Bulux, Quan de Serrano, Giuliano 1994). The nature of the matrix may be important, and some diets of vulnerable the affliction fell (Institute of Nutrition and Food communities are low in fat. Dietary fat appears Science, Bangladesh 1990). However, documen- to be more important for the absorption of caro- tation of the extent of improvement comparable tene than for that of preformed vitamin A. More to the documentation of the efficacy of fortifica- attention is now being given to encouraging even tion or capsule programs is lacking. a small increase in preformed vitamin A intake In general it appears that consumption of from such readily available sources as eggs and more than 40% of total vitamin A in the form of perhaps milk or fortified foods. preformed vitamin A is highly protective (FAO/ Dietary modification can be brought about WHO Expert Consultation 1988). by a variety of means, and if several can be Dietary interventions are usually targeted to- employed so much the better. In Vietnam (Eng- wards vulnerable groups – infants, preschool-age lish, Badcock, Giay et al. 1997), a combination children and their pregnant and lactating mothers. of home garden production and nutrition educa- School-age children might also be included if tion resulted in highly significant improvement shown to be particularly at risk. The food catego- in the incidence of diarrheal disease and respi- ries to be promoted will vary with the vulnerable ratory infections. Social marketing of vitamin A group (Figure 15.2). rich foods in Thailand, with special emphasis on Surveys have shown that young children fre- use of the readily available ivy gourd, could be quently eat less than 15 g of green leafy vegeta- readily applied in the region to bring about be- bles in a day. It was shown in Bangladesh (Rah- havioral change (Smitasiri, Attig, Valyasevi et al. man, Mahalanabis, Islam et al. 1993) that 40 g 1993). of dark green leaves would be readily consumed Recent experience has confirmed that red daily if prepared attractively. This is sufficient to palm oil is a valuable source of both provitamin A provide the daily requirement of vitamin A in one and non-provitamin A carotenoids (Mahapatra, meal. Manorama 1997; Solomons 1998). Other palms Most yellow fruits and some green leaves are have similar potential, for example in Bangladesh usually available only at certain seasons. Village- the palmyra palm (Borassus flabellifer), which is based food processing and preservation can ex- native to tropical Africa as well as being found tend the availability and acceptability (Sommer, in south and southeast Asia, is being promoted West 1996:355–87). (Shamim, Mawla, Islam et al. 2003). Another In dietary interventions, more so than in important contribution comes from Bangladesh. 180 15 Control

Roos, Islam, Thilsted (2003) have advocated the GENERAL ISSUES incorporation into carp culture of some wild, small indigenous fish species to augment intakes of vi- Multiple strategies are now being increas- tamin A, calcium and iron in rural households. ingly employed and this sometimes makes inter- Rodriguez-Amaya (2002) has for long investiga- pretation of results very difficult if not impos- ted and promoted what she calls the “bounty” of sible. Many such interventions are being under- carotenoid sources in Brazil. taken in countries that are undergoing rapid so- Dark green leafy vegetables have received cial and economic change where control, or even bad press in recent years concerning the low identification, of some important variables is not bioavailability and low bioconversion of caro- possible. tenoids (see Chapter 5). The deuterated-retinol- Cost-effectiveness has always been an im- dilution technique was used to demonstrate that portant issue but is only recently being addressed daily consumption of cooked, pureed green leafy by well-designed and executed studies. Targeting vegetables or sweet potatoes has a positive effect vitamin A supplements to high-risk children was on vitamin A stores in Bangaldeshi men (Haskell, found not to be an efficient use of resources. It Jamil, Hassan et al. 2004). was concluded that, like immunization, vitamin A Again from Bangladesh comes the account should be provided to all preschool-age chil- of a success story of a homestead gardening pro- dren in developing countries (Loevinsohn, Sutter, gram, which resulted in household food security Costales 1997). In a comparison of nutrition and empowerment of women (Bushamuka, de education and vitamin A supplementation it was Pee, Talukder et al. 2005). found that the latter was quicker and cheaper, but Attention has been drawn to the use of “green education had longer lasting effects. It was con- rice,” traditional in Vietnam (Vuong, Burgess cluded that both were needed for a comprehen- 2004). The local variety of rice is harvested early, sive national program (Pant, Pokharel, Curtale et when the hulls are still green, and processed into a al. 1996). Under ideal circumstances, vitamin A cereal commonly eaten for breakfast. Analysis of fortification can be cheaper and more efficient fresh green rice showed concentrations of about than either capsule distribution or home food 1.9 μg/g of β-carotene and of 2.7 μg/g of lutein. production with nutrition education (Phillips, Unlike the hulls of paddy rice, which are removed Sanghvi, Suaréz et al. 1996). during processing, those of green rice remain Considerable concern should be expressed attached. Other vitamins and minerals are there- about the mixed messages that are emanating fore available. from WHO and UNICEF relating to some of the It is becoming more widely appreciated that measures being advocated for child survival and better results from interventions may be expected health at this time. For example, “vitamin supple- when more than one adverse factor is counter- mentation coverage with at least one dose in last 6 acted at any one time. Thus, increased fat intake months” is advocated for prevention (Countdown and anthelminthic drug treatment have enhanced to 2015 Conference 2006), without specifying the the effectiveness of food-based programs (Jalal, vitamin and not including iron or zinc. Another Nesheim, Agus et al. 1998). group (Edejer, Aikins, Black et al. 2005) echoes These and many other reports from around most statements on methods to achieve the fourth the developing world provide the hope that Millennium Development Goal (MDG), specific knowledge-based, careful and persistent efforts to to child health, by including supplementation utilize the local resources may go a long way to with vitamin A and zinc. More recently progress bringing about the final eradication of vitamin A has been made towards agreement on the means deficiency in many areas. of meeting MDG requirements, especially as expressed in the Copenhagen Consensus (2008). 15 Control 181

Reducing undernutrition is one of the MDGs and Rice (Oryza sativa) is usually milled to re- is also a key factor underpinning several others. move the oil-rich aleurone layer that turns ran- Recent estimates published in The Lancet Series cid upon storage especially in tropical areas. (see Chapter 1) showed maternal and child un- This layer contains some β-carotene and some dernutrition being responsible as the underlying other nutrients that do not occur in the remaining cause for 3.5 m deaths, 35% of the disease burden edible portion of the grain, the endosperm. In in children <5 years old, and 11% of the total glo- this research recombinant DNA technology, with bal DALYs. The considerable contribution to this a combination of transgenes, was used to enable from VADD was referred to. biosynthesis of provitamin A in the endosperm to A third group (Baqui 2006) advocates addition occur. Immature rice endosperm can synthesize of zinc to fluids for control of diarrhea in young the early intermediate geranylgeranyl diphos- children, but ignores vitamin A. Most recently, phate (GGPP – see also Chapter 2). This can be a fourth proposes fortifying milk with zinc and used to produce the uncolored carotene phytoene iron, but not vitamin A (spelling disaster in the by expressing the enzyme phytoene synthase in form of severe marasmus, as was the case through rice endosperm. For the synthetic pathway to be free government milk distribution in some coun- completed towards β-carotene three additional tries in the past) (Bhutta 2007). The consequent plant enzymes are required – phytoene desaturase confusion is serious, as is the ignoring of all the and ζ-carotene desaturase, each of which cata- evidence of the efficacy of vitamin A from the re- lyzes the introduction of two double bonds, and cent past (see Chapter 9). lycopene β-cyclase, encoded by the lcy gene. Four transgenes were required; all were obtained from daffodil (Narcissus pseudonarcissus). PLANT BREEDING AND GENETIC Earlier studies have shown that the amount of MODIFICATION β-carotene available from “golden” rice is quite low, ranging from 0.4 μg/g to 1.6 μg/g (Datta, Traditional methods of plant breeding have Parkhi, Rai et al. 2005). The consumption of been applied quite extensively to the enhance- 200 g of such rice/day should meet daily requirement of the β-carotene content of foods since this ments, but only providing absorption and conver- topic was raised in the first edition of the Manual sion rates are high. More recently (Paine, Shipton, (Bouis 1996). These include higher-yielding vari- Chaggar et al. 2005), the daffodil gene encoding eties of amaranth in India, tomatoes with a high phytoene synthase has been replaced by one from β-carotene content in Taiwan, high-carotene sweet maize. In this new Golden Rice 2, total caroten- potato in Africa, and red canola oil with a unique oids were increased up to 23-fold and there was blend of several carotenoids (Reddy 2000). preferential accumulation of β-carotene. Genetically modified (GM) rice, aimed at In most cases the transformed endosperms of improving the supply of iron and vitamin A in the rice were yellow in color, indicating the for- the human diet, was first announced at the XVI mation of carotenoid. There was some variation International Botanical Congress on 3 August in different seed lines concerning the carotenoids 1999. The definitive publication of this research produced. Often this was only β-carotene, but was made a few months later (Ye, Al-Babali, occasionally lutein and zeaxanthin were also de- Klöti et al. 2000). The research involved col- tected. From the results so far it seems that the laboration between the Institute for Plant Sciences yield of provitamin A will be at least 2 μg/g of of the Swiss Federal Institute of Technology, rice as consumed. The authors suggest that a Zurich, Switzerland, and the Center for Applied daily intake of 300 g of rice will provide 100 μg Biosciences of the University of Freiburg, Frei- RAE. This would be equivalent to about 25% of a burg, Germany. childʼs RDA. This assumes that the bioavailability 182 15 Control

of β-carotene would be 1/6 that of retinol, but this DISASTER RELIEF remains to be fully established. Recent research indicates a conversion factor of 1:4 (Tang, Qin, Over the past two decades the number of re- Dolnikowski et al. 2009). fugees registered by the UN High Commissioner There are still many technological, sociolo- for Refugees has risen steadily and now approx- gical and other considerations to be addressed imates 20 million. In addition, the estimated before there is any possibility of GM rice be- number of people forced from their homes but coming the staple food of hundreds of millions of not from their countries is now nearly 30 million. the world’s poor. A diet of ordinary rice ade- About half of these groups are children under the quately supplemented with vegetables and fruits as age of 15 years and are especially vulnerable. Ex- well as some meat and fish is capable of providing perience has shown that they are not only particu- a balanced supply of known nutrients and other larly susceptible to infections and protein-energy health-promoting substances. The overriding goal malnutrition but also to vitamin A and other mi- of attaining full nutrition for all by practical and cronutrient deficiencies (McLaren 1987). realistic means should not be allowed to be ob- If agencies involved in disaster relief opera- scured by any “quick fix” approaches. tions are made aware of the problem, there is no Potrykus, the scientist who originally devel- reason why it should not be averted. It should be oped Golden Rice, has written about some of the ensured that the vitamin A content of food rations difficulties being experienced in practice (2003). supplied is adequate. In particular skim milk must Legal complexities had to be overcome to make be fortified, as it usually is, with vitamin A. The Golden Rice freely available to developing most readily implemented single measure is the countries. The International Humanitarian Gol- distribution of vitamin A capsules to all young den Rice Network has been formed, with many children. Mothers should be encouraged to breast- institutes in rice-growing countries. Challenges feed their infants. In more settled circumstances include transfer of safe technology, overcoming the growing of green leaves and yellow fruits consumer prejudice and GMO opposition, and should be encouraged. deregulation. The experienced team of Aaby and colleagues Genetically engineered ‘Golden Rice’ was (Nielsen, Benn, Balé et al. 2005), following the recently shown to be an effective source of vita- WHO recommendation to provide vitamin A min A, contanining 35 μg β-carotene per gram of supplementation in emergency situations, eval- rice in a trial in middle-school children in China uated such a program in the war emergency in (Tang, Qin, Dolnikowski et al. 2009). Guinea-Bissau, 1998–9. They were able to com- The general trend, demonstrated throughout pare wartime with pre-wartime mortality in over this book, of the emphasis in future moving 5,000 children aged 6 months to 5 years in the steadily towards multi-micronutrient, rather than capital. They found a slight, nonsignificant mor- simple nutrient, deficiency control is shown in a tality reduction during the war in those receiving recent study involving corn (Naqvi, Zhu, Farre et supplementation, but this compared with a signi- al. 2009). The three vitamins investigated were ficant 12% reduction during three years pre-war. β-carotene, ascorbate, and folate. The amount of Vitamin A supplementation was associated with a vitamins contained in the transgenic kernels as significant reduction in cultural and socioecono- compared with normal was 169-fold, 6-fold and mic inequalities. double respectively. Where disaster relief is indicated, multiple deficiencies of micronutrients are especially likely to occur and to be sometimes of a severe degree. Concerned UN Agencies recommend supplementation with micronutrient powders 15 Control 183

Table 15.4: The composition of multiple micronutrient supplements for pregnant women, lactating women, and children from 6 to 59 months of age, designed to provide the daily recommended intake of each nutrient (one RNI) (WHO, WFP, UNICEF Joint statement 2007).

Micronutrient Pregnant and lactating Children (6–59 months) women Vitamin A μg 800.0 400.0 Vitamin D μg 5.0 5.0 Vitamin E mg 15.0 5.0 Vitamin C mg 55.0 30.0

Thiamine (vitamin B1) mg 1.4 0.5

Riboflavin (vitamin 2B ) mg 1.4 0.5

Niacin (vitamin B3) mg 18.0 6.0

Vitamin B6 mg 1.9 0.5

Vitamin B12 μg 2.6 0.9 Folic acid μg 600.0 150.0 Iron mg 27.0 10.0 Zinc mg 10.0 4.1 Copper mg 1.15 0.56 Selenium μg 30.0 17.0 Iodine μg 250.0 90.0 containing 15 micronutrients, shown in Table 15.4, which are also being used in multi- micronutrient supplementation research trials (see Chapter 10). 184 185

INDEX

A and iron absorption 107–8 α-Carotene 19, 32–33, 40 bioavailability 37–42 Abetalipoproteinemia 155 cleavage 25 Aceh (Indonesia) 90 concentrations 176–8 Acitretin 162 conversion to vitamin A 38, 41 Acne 162 excentral cleavage of 39 Actinic keratosis 162 in chronic diseases 163–4 Acute lower respiratory tract infections (ALRI) in general medicine 160–4 98–100, 101, 106, 117 in health 43–59 Acute-phase response (APR) 63–9, 97–106, 139, supplementation 117, 143 146, 156 9-cis β-carotene 40 Age 25, 30, 41, 47, 141–3, 157 Bacteriuria 106 Age-related macular degeneration (AMD) Banana 33 19, 28, 160–2 Beriberi 11, 76 AIDS (see HIV/AIDS) Berries 33 All-trans retinol etc, see retinol etc Bile salts 43 Analysis Bioconversion 27, 31, 38, 40–2, 73, 180 HPLC 27–8, 30, 66 Bitot’s spot (X1B) 69, 71, 75–86, 102, 135, 172 In vitro 42 and age 143 Anemia 13, 57, 107–9, 115–9, 117–21, 135, 143 and sex 152 Animal products 30 prevalence of 62, 82 Animal sources 178 unresponsive 82 Annecy Accords 65 Black currant 33 Anthropometric status 109 Body composition 109 Antioxidant 119,160–4 reserves 67–8 activity of carotenoids 19, 31, 33, 96, weight 109 activity of vitamin E 40 Bone 14, 114, 157, 178 and chronic disease 59 abnormalities 114 Apolipoprotein D 47 fractures in the elderly 157 Apricot 31 Breastfeeding 111, 144–5, 168, 171, 173 Assam 168 Breast milk 69, 143, 171, 173 Ascaris infestation 41, 147 and HIV transmission 103 Asia 34, 121, 134, 142, 173–4, 179 retinol in 37, 69, 117, 135, 143, 173 vitamin A in 93, 143 B Bronchitis 98 β-apo-8’-carotenal 34 Bronchopulmonary dysplasia (BPD) 159 β-carotene 11, 78, 79, 96, 143, 146, 168, Bulging of the anterior fontanelle 170 170, 177 Buriti palm 31 activity 30–34 Butter 34, 35, 176 all-trans 19 and immune response 113 186 Index

C remnant 44–5 Cancer 44, 51, 58, 162–3 Cirrhosis of liver 45, 157 Cardiovascular disease 163, 164 Clustering 150–2 Carotenoids (see also individual carotenoids) Cod 35 accumulation of 181 Collagen type III 45 actions 58–9, Collagenases 84 analysis of 28 Colliquative necrosis 84 antioxidant activity of (see Antioxidant) (see also keratomalacia) bioavailability of 27, 31, 37–43, 73, 168, Community trials 98 178, 180, 181 Complement C8γ 47 conversion efficiency of 38, 41 Cone cells 23–4, 52, 75, 160 functions of 19, 20, 25, 58–9 threshold 77 in cancer 44, 58, 162–3 -rod transition time 77 in chronic diseases 59, 163–4 Congenital malformations 55, 158–9 in plasma 20, 38, 40, 48–9 Conjunctiva 54, 61, 63, 69–71, 75, 80–2, 88, 141 metabolism of 27–8, 38, 164 appearances 71, 75, 80 non-provitamin A 17–20, 30, 37, 40, 113, Conjunctival Impression Cytology (CIC) 155, 160, 179 63, 69–71, 87, 102, 106, 108, 146 occurrence of 31 with transfer (ICT) 69, 71 prooxidant properties of 19 Conjunctival xerosis (X1A) 62, 63, 69, 80, protective function of 19 82, 155 provitamin A sources 29, 31–34, 48, 69, Conjunctivitis 71 133, 141, 179, 181 Conversion factors 38, 42, 73, 182 structure of 17–19 Copenhagen Consensus 15, 180 Carrot 31–4, 37–41, 79 Cornea 54, 63, 69, 75, 80, 82, 84–6, 88, 95, 147 Carr-Price reaction 27 Corneal scar (XS) 62, 82, 86 Cassava 33, 178 ulceration 62, 86, 147 Cataract 161–2 ulceration/keratomalacia (X3A, X3B) 62 Celiac disease 156 xerophthalmia 76, 109, 135, 141, Cell differentiation 54, 56, 58, 155, 162–3 145, 147 Cellular retinal-binding protein (CRALBP) xerosis (X2) 62, 82 49, 52 Coronary heart disease 111, 161 Cellular retinoic acid-binding protein (CRABP) C-reactive protein (CRP) 27, 64, 146 49, 55 Cryptoxanthin 19, 31, 157, 163, 164 Cellular retinol-binding protein (CRBP) 41, 44, Cyclic AMP 58 48–50, 163 Cystic fibrosis 156 Cervical dysplasia 163 Cytochrome P450 enzymes 51 Cheese 35 Cytokines 63, 105, 113, 146 “Chicken eyes” 71, 77 (see also night blindness) D Child spacing 145 Dairy produce 34 Chlorophyll 19, 20, 31, 40 DALYs (see disability-adjusted life years) Chloroplast 19, 21 Dark adaptation 25, 71, 77–79, 118, 143, 158 Cholecystokinin 43 Dark green leafy vegetables (DGLV) 31, 37–8, Chromophores 21–6, 27 42, 161, 180 Chylomicron 44–5, 48 Dermatological disorders 114, 162 Index 187

Dermomalacia 84, 86 F Deuterium isotope 27, 39, 68–9 Failure to thrive 11 Diabetes mellitus 111, 164 Famine 149 Diagnosis of FAO 37, 40, 133, 174, 179 accompanying infections 90, 100, 105 Fat 38, 39–41, 43, 69, 179, 180 AMD 28 Fatty acids, essential 11 CIC 69, 71, 80, 87 Fiber 34, 40 hypervitaminosis A 156–7 Field trials 93, 173, 178 MRDR 64 Final rod threshold 63, 78 night blindness 77 Fish 25, 29, 34, 180, 182 RBP 66, 67 fatty 34 RDR 45 liver 34 VAD 88, 114, 156, 159 liver oils 34 vitamin A status 163 scales 71 xerophthalmia 76, 86, 169 Folic acid 112, 115, 118, 161 Diarrheal diseases 98–101, 106, 116, 147, 150, Food fortification 15, 35, 68, 107, 119, 158, 168, 172, 176, 179 171, 176–8 3,4-Didehydroretinal 25 law 176 3,4-Didehydroretinyl 67 shortage 143, 149 Diet 25, 29–30, 39, 42, 48, 111, 141, 143–4, 145, Foodlets 119 152, 158, 160, 168, 181–2 Fortification, see food fortification Dietary counseling 171 Fruit 20, 23, 29–34, 37–42, 59, 60, 72, 133, 141, intake assessment 61, 72 145, 149, 163, 168, 178, 179, 182 interventions 69, 72, 168, 179 Disability-adjusted life years (DALY) G 12–15, 181 γ-carotene 19, 31 Disaster relief 168, 182–3 Gap junction communication 58 Dopaminergic system 58 Gastrectomy 156 Gastroenteritis 150 E (see also Diarrheal diseases) Economic/food policies 178 Gene expression 44, 53, 57, 118 Egg 34, 35, 179 defects and VAD 41 Elaeis sp. 34 modification in rice 168, 181–2 Electroretinography 76 Genetic polymorphisms 38 Embryogenesis 54–6 Genito-urinary tract 98 Endoplasmic reticulum 45, 49 Geranylgeranyl diphosphate 181 Endosome 45, 49 Giardia 41, 147, 156 Enzyme defect 155, 156 Global nutritional index (GNI) 15 Epididymal-binding protein (EBP) 48 Glycoprotein 27, 52, 54, 63, 64 Epithelial cell 42, 48, 63, 69–70, 80, 84, 155, Glycosaminoglycan 54 162, 163 Goblet cell 63, 70, 80 lining 112 Goiter 13, 114, 119 Erythropoietic porphyria 162 Golden Rice 30, 168, 181–2 Erythropoietin 105, 108 Golgi complex 45 Etretinate 162 G protein 21, 23, 25, 52 Ethiopia 169 Grape 33 188 Index

Green vegetables 31 T-cell 56, 103 (see also dark green leafy vegetables) Immunization (see also Measles and Growth 57–58, 61, 104, 109–112, 159, 170 Supplementation) adolescent 143 campaigns 167, 174 failure 103 Expanded Program on (EPI) 100, linear 177 133, 175 retardation 12, 13, 57, 61, 104, 109, 111 programs 113–4 stunting of 12, 13, 61, 148 Infectious diseases 57, 89, 93, 97, 112, 116, zinc and 116–7 146–8, 152, 167, 173 Gut 47, 98, 109, 111, 112, 147 prevention and management 175–6 severe 12, 86, 98, 101, 117, 156, 167 H Intervention, horticultural 178 Hemoglobin 107–9, 115–9, 147 types of 167–9 Hemopoiesis 52, 57, 107 Intracellular retinoid-binding protein 49 Halibut 34 In-vitro screening techniques 42 Height/age 109 Iodine deficiency disorders (IDD) 12, 181 Hemeralopia 71 Iodopsin 23, 52, 75 (see night blindness) β-ionone 18 Hepatocyte 45, 46 Iron 12–13, 52, 57, 104–5, 111, 113, 115, 115–9, Hepcidin 109 176, 177 Herpes simplex 103, 133, 147 absorption 107–9, 119 High-performance liquid chromatography deficiency 12–13, 57, 105,107–9, (HPLC) 27–8, 30, 66 116, 143 HIV-AIDS 67–8, 99–102, 102–4, 105, 106, 146, fortification with 19,1 177–8, 181 148, 172 status 105, 107–9, 115 in mothers 94, 102–4, 118 storage 57, 107 Homeobox genes 55 supplementation 105, 113, 115–9, 183 Human growth hormone 57 utilization 115 Hyderabad (India) 92 Isomerization 24, 40, 52 Hypercarotenemia 156 Isoprene 17 Hypercarotenosis 162, 164 Isotope dilution 38–9, 40, 68 Hyperkeratosis 114 Isotretinoin 158, 162 Hypervitaminosis A 155, 156–7 (see also vitamin A toxicity) J Hypobetalipoproteinemia 155 Jaundice 156 Hypopyon 86 Hypovitaminosis A 156 K Keratinization 71, 80, 85, 113 I Keratomalacia (X3A, X3B) 62, 63, 75, 82, 84–6, Immunity 56, 112–4, 88, 143, 145, 156 cellular 112–4, 116 (see also corneal xerophthalmia) defensive response 112 Kwashiorkor 11, 12, 145, 148 humoral 112–4 natural killer (NK) cell 112, 113 L offensive response 112 Lacrimal gland 75 retinoids and 56 Lactation 59, 104, 109, 113, 171 Index 189

β-Lactoglobulin 47 Millenium Development Goals (MDG) 180 Latin America 15, 121, 177 Mitochondria 21, 58–9 Lecithin:retinol acyltransferase (LRAT) 44, 47, Modified Relative Dose Response (MRDR) 49, 52 67, 115 Leukemia, acute promyelocitic (APL) 163 15,15’-monooxygenase (CMO1) 25 Lipocalin 46 Monosodium glutamate (MSG) 92, 177 Lipoprotein 49, 161, 164 Morbidity 37, 97–106 Liver 34, 43, 44–47, 64, 79, 109, 148, 156, Mortality 11–16, 37, 63, 79, 89–106, 109, 111, 157, 158 119, 147, 159, 163, 171–183 reserve 26, 28, 30, 42, 44–5, 60, 61, and xerophthalmia 89–90, 133 67–69, 156, 159 cumulative 90 sheep and ox 34 prevention trials 92, 93, 95 structure 46 under-five 12, 13, 94 Low responder phenotype 38 Mucin 54, 69, 80, 113 Lutein 18, 19, 31–4, 40, 59, 66, 113, 146, 160–2, Multi-micronutrients 119, 135, 168, 178, 182 180, 181 deficiencies of 14, 19,1 149, 167, 182 Lycopene 18, 19, 30–33, 40, 59, 113, 163–4, 181 supplementation with 104, 116, 119 Lymph 43, 44, 48 Muscle circumference 109

M N Macula 28, 160, 161 Neuronal signaling 17, 58, 159 Maize 144, 181 Niacin 11, 183 Malaria 67, 87, 104–5, 148 Night blindness (XN) 14–15, 71, 75, 76–80, 102, Malformations, congenital 54–5, 114, 158 135, 143, 149, 169, 178 Mango 31, 33, 111 and eye diseases 160 Marasmic-kwashiorkor 148 and sex 152 Marasmus 11, 12, 148, 181 as an indicator of VAD 61–63, 155 Margarine 34, 176, 177 maternal 99, 170 Maternal mortality 11, 78, 96–7 responsive to zinc 116 Maternal stores 93 NNIPS (Nepal) 92, 93 Measles 86, 93, 95–6, 100–2, 106, 114, 117, 133, Noncompliance 91 147–8, 167, 169, 172, 175 Nutritional blindness 133 and associated conditions 147–8 Nutrition education 178–80 case fatality rates 96 Nutritional status 15, 61, 100 complications 95, 100–1 Nyctalopia (see night blindness) keratitis 133 immunization in 95, 100–1, 114, 175 O mortality 95–6 Opsin 24–6, 52, 71 outcome 100–1 Optic nerve 114 Meat 29, 171828 Orange 33, 164 Meningococcal disease 106 Orange fruits 37 Micelle 40 Osteomalacia 11, 14 Micronutrient Deficiency Information System Otitis media 106 (MDIS) 121 Milk 34, 35, 73, 149, 179, 181 P skim 177, 182 Pacific Region 108, 137 190 Index

Palm oil 31, 33, 34, 79, 179 Respiratory disease 98–100, 147 extracts 34 Retina, bleaching 63 species 34 11-cis retinal 18–26, 52–53, 71 Palmyra palm 179 Retinal-binding protein, cellular (CRALBP) Pancreatic lipase 43 49, 52 Pancreatitis, chronic 156 Retinitis pigmentosa 75, 160 Papaya 31, 33, 39, 149 Retinoic acid 18, 26, 44, 47–49, 50–1, 54–57, Parasitic infestation 46 66, 102, 103, 107, 113, 157–65 Pellagra 11, 76 -binding protein, cellular (CRABP) 49 Peru 119, 147, 175 nuclear receptors RAR and RXR 26, 50, Phospholipid 44 57, 58, 164 Photoentrainment 26 9-cis retinoic acid 51, 56 Photoisomerization 24, 52 Retinoids 17–26, 28, 47–58, 66, 102, 112–4 Photosynthesis 19, 20–21, 29, 59 in general medicine 155–165 Phototransduction 21–24 Retinol 17, 26, 31, 37–42, 43–59, 61–73, 113, Physiological status 143 114, 116, 155–164, 169, 182 Plant breeding 33, 168, 178, 181–2 (see also vitamin A) activity Pneumonia 12, 13, 98, 98–106, 117, 147 (see also Retinol Equivalent and Potato 29, 33, 180 Retinol Activity Equivalent) 31 Poverty 133, 144 all-trans- 25–6 Pregnancy 34, 59, 66, 67, 69, 94, 103–5, 112, as indicator of vitamin A status 27, 37 115, 143, 158–9, 173 in breast milk 37, 69, 117, 135, 143, 173 and HIV 94, 103–5, 146 in serum 14, 15, 378, 47, 61, 64–73, and safety of vitamin A 171 77–80, 93–4, 97–107, 109, 112, 113, cytokine responses in 113 114, 116–8, 122–32, 135–9, 143, 146–8, night blindness during 78–9, 143 156–9, 164, 173, 177, 179 related mortality and morbidity 96–97 Retinol-binding protein (RBP) 27, 41, 45–6, Protein status 148 63–8, 118, 146–52 Protein-energy malnutrition (PEM) 11, 46, 61, membrane receptor 48 68, 76, 86, 89, 94, 97, 147, 148–9, 150, 167, Retinol-binding protein/transthyretin molar ratio 169, 182 (RBP/TTR molar ratio) 67, 159 Provitamin A carotenoids (see Carotenoids, Retinol Activity Equivalents (RAE) 31, 42, 59 provitamin A sources) Retinol Equivalents (RE) 37, 42 Psoriasis 162 Retinyl ester 38, 43, 44, 45, 49, 67 Public health 12–14, 41, 61–3, 75, 78, 87, 103–5, Rhodopsin 18, 21–6, 52–4, 71, 75, 161 115, 121, 134, 137, 167, 168, 170, 176–7 Riboflavin 118, 183 Pumpkin 33 Rice 30, 79, 109, 143–5, 168, 176, 180–2 Pupillary contraction 63, 72 (see also Golden Rice) GM 181, 182 R green 180 Rajasthan (India) 149 Rickets 11, 14, 76, 115 Raman spectroscopy 28 Rod cells 71, 75, Red palm oil 31, 33, 34, 79, 179 scotometry 76 Relative Dose Response (RDR) 45, 67, 68, 69, Roots 31, 33, 143 115, 177 Rotavirus infection 147 Reproduction 50, 57, 165 Index 191

S Thiamine 29, 183 Sanitation 141, 147 Thyroid function 56, 114, 118, 164 Scandinavia 164 Tomato 31, 33–4, 39, 164, 181 Scurvy 11, 76, 115 Trachoma 71, 82 Selenium 118, 178, 183 Traditional eye medicine 133, 147–8 Sex differences 174 Transferrin 107 in Bitot’s spot 152 Transthyretin (TTR) 46, 64, 67 Shark, liver 34 Tretinoin 162 Shedding Tuber 32, 33, 39 genital 103 Tuberculosis 100, 105–6, 117 HIV 103, 118 mucosal 103, 118 U vaginal 103 Undernutrition 12, 109, 181 Sjogren’s syndrome 75 UNICEF 11, 145, 170, 173, 176, 180 Skin 28, 50, 84, 114, 157, 158, 162, 164 United Nations Millennium Development Goals Skinfold thickness 109 (MDG) 180 SLAMENGHI mnemonic 39–40 Urinary tract 106 Small intestine 43, 156 Spermatogenesis 114 V Sprue 156 Vaccination 96, 174 Staple foods 30, 143, 176 VAST (Ghana) 92, 93 Starfruit 33 Vegetable oil 34, 178 Stellate cell 45 Vegetables 20, 29, 30, 31–3, 37, 39, 40, 42, 59, Steroid-Thyroid-Retinoid Superfamily 50 60, 73, 79, 133, 163, 168, 182 Stunting (see Growth) (see also dark green leafy vegetables) Subclinical-clinical divide 62–3 Vegetarian 30 Sudden infant death syndrome (SIDS) 164 Very low birth weight (VLBW) 159 Sugar, fortification of 68, 107, 177 Vision 21, 23, 25, 51–54, 71, 75–77, 158, 160 Summer diarrhea 150 restoration 63 (see also diarrheal disease) restoration time (VRT) 71–72 Superficial punctate keratopathy (SPK) 82 Visual cycle 25, 48, 49, 52, 161 (see also corneal xerophthalmia) functions 161 Supplementation (see also β-Carotene, Iron, impairment 86, 155 Multi-micronutrient, Vitamin A) Visual purple (see Rhodopsin) and immunization 95, 101, 114, 167, Vitamin A 171, 174–5 absorption 43–4 maternal 69, 93, 95, 96–7, 104, 118 bone modeling and 114 multi-micronutrient 104, 116 capsule distribution 137, 167, 170–3, neonatal 95, 101, 104 180, 182 newborn 171, 173, 173–5 cellular uptake 28 Sweet potato 31, 33, 39, 180 functions of 26, 50, 51–8 functions unrelated to light 26 T in breast milk 69 Taro 33 in nature 17–26 Telomere length 119 interactions with other micronutrients Teratogenic 34, 54–5, 158, 169, 171 115–119 192 Index

interventions 14–16, 33, 39, 64, 68, Vitamin B2 (see riboflavin)

90–3, 98 Vitamin B3 (see niacin)

-iron interrelationships 115 Vitamin B6 161, 182

metabolism in liver 45–6 Vitamin B12 30, 161, 182 methods of analysis 27–8 Vitamin C 11, 115, 182 occurrence 34 Vitamin D 11, 14, 51, 115, 158, 182 preformed 34 Vitamin E 34, 40, 155, 182 prophylaxis 139, 167, 170–3 Vomiting 156, 157, 169, 170 requirement of 30, 59–60, 133, 143, 179, 181 W reserves 45, 67–8, 159 Wasting 12–13, 106, 109, 149 -rich foods 72, 173, 178–9 Weaning diet 111, 141, 145 safety 158–9, 170, 175 WHO 12–15, 37, 40, 61–2, 67–9, 71, 75, 80, 84, status, assessment of 61–73 100, 121, 133, 134, 135, 145, 152, 167, 169–75, storage form 34 179, 180, 182 supplementation 13, 15, 37, 38, 57, 69, World Bank 15 72, 78, 90–106, 109–116, 117–121, 139, 143, 152–5, 160, 163, 166–9, 170–9 X tissue 41–2, 43, 50, 54, 109 Xanthophyll 27, 34 toxicity 156–9 Xerophthalmia (see also conjunctiva and cornea) transport 44, 46–8, 155 75–88, 89–90, 93 treatment 75, 79, 82, 90, 95, 99, 101–2, and breastfeeding 144–5 105, 109, 155, 160, 167, 169, 172, 176 and diet 144 urinary excretion of 38, 147 and food fortification 176–7 units 30, 42 and growth 109–10, 141, 148

Vitamin A1 25, 34 and disease 98, 106, 147, 159

Vitamin A2 25, 34, 67 and PEM 148 Vitamin A deficiency (VAD) as indicator of physiological status 143 control 14–15, 34–35, 37–38, 117, cause of childhood blindness 102, 121, 167–183 133, 150 elimination 15, 167 classification 14, 61–3, 75, 08 global occurrence of 121–139 clustering 152 in experimental animals 42, 56–7, 67–8, criteria for 62, 80 89, 98, 107–114, 153–160 endemic 121, 145 prophylaxis 155, 167–173 in experimental animals 11 prevalence 14, 61, 65, 110, 119, 134–9, treatment 169 146–52 trends 135–6 public health importance 12, 14, 61–3, Xerophthalmic fundus (XF) 62, 72, 75, 86 121, 137, 170, 176 Xerosis 54, 62–3, 69, 71, 75, 80, 82, 84, 88, season 57, 110–11, 149–50, 152 141, 155 secondary (endogenous) 75, 88, 155–6 subclinical 37, 61–72, 93–4, 108–10, Z 134, 137, 146, 155, 167, 170 Zeaxanthin 19, 34, 160–2, 181 susceptible groups 14, 41, 135, 152, 182 Zinc 12–13, 15, 40–1, 58, 72, 94, 96–8, 105–6, trends 135–6 111, 113, 116–18, 156, 161, 170, 176, 178, 180

Vitamin B1 (see thiamine) interrelation with vitamin A 116–18 193

REFERENCES

Abraham K, Mueller C, Grueters A et al. gladesh. Am J Clin Nutr 74: 105–15. (2003). Minimal inflammation, acute phase re- Akhter N, Stallkamp G, de Pee S et al. (2004). sponse and avoidance of misclassification of vi- Vitamin A capsule distribution among infants in tamin A and iron status in infants – importance of rural Bangladesh aged 6–11 months. Sight and a high-sensitivity C-reactive protein (CRP) assay. Life Newsletter 3: 26–8. Int J Vitam Nutr Res 73: 423–30. Aktuna D, Buchinger W, Langsteger W et al. ACC/SCN Consultative Group (1994). Con- (1993). β-carotene, vitamin A and carrier pro- trolling vitamin A deficiency. ACC/SCN State-of- teins in thyroid diseases. Acta Med Austriaca 20: the-art Series, Nutrition Policy Discussion Paper 17–20. no 14, United Nations, Geneva. Alacorn K, Kolsteren PW, Prada AM et al. ACC/SCN (1997). Third Report on the World (2004). Effects of separate delivery of zinc or zinc Nutrition Situation. and vitamin A on hemoglobin response, growth, Adelekan-Delana A, Northrop-Clewes CA, and diarrhea in young Peruvian children receiv- Owa JA et al. (2003). Use of biomarkers of sub- ing iron therapy for anemia. Am J Clin Nutr 80: clinical infection, nutrition and neonatal maturity 1276–82. to interpret plasma retinol in Nigerian neonates. Albert MJ, Qadri F, Wahed MA et al. (2003). Br J Nutr 90: 353–61. Supplementation with zinc, but not with vitamin A Aguayo VM, Baker SK, Crespin X et al. improves seroconversion to vibriocidal antibody (2005). Maintaining high vitamin A supplemen- in children given an oral cholera vaccine. J Infect tation coverage in children: lessons from Niger. Dis 187: 909–13. Food Nutr Bull 26: 26–31. Alm B, Wennergren G, Norvenius SG et al. Aguayo VM, Kahn S, Ismael C (2005). (2003). Vitamin A and sudden infant death syn- Vitamin A deficiency and child mortality in drome (SIDS) in Scandinavia 1992–1995. Acta Mozambique. Publ Health Nutr 8: 29–31. Paediatrica 92: 162–4. Ahmed F, Hasan N, Kabir Y (1997). Vitamin A Almekinder J, Manda W, Kumwenda N et al. deficiency among adolescent female garment (1999). Use of retinol-binding protein to measure factory workers in Bangladesh. Eur J Clin Nutr vitamin A deficiency in population-based studies. 51: 698–702. Proc XIX IVACG Meeting, Durban, South Africa Ahmed F, Ellis J, Murphy J et al. (1990). Ex- 9–11 March: 97. cessive faecal loss of vitamin A (retinol) in cystic Alsop DH, Brown SB, Van Der Kraak GJ fibrosis. Arch Dis Child 65: 589–93. (2004). Dietary retinoic acid induces hindlimb Ahmed F, Khan MR, Akhtauzzaman M et al. and eye deformities in Xenopus laevis. Environ (2005). Efficacy of twice-weekly multiple micro- Sci Techn 38: 6290–9. nutrient supplementation for improving the hae- Alvarez JO, Salazar-Lindo E, Kohatsu J et al. moglobin and micronutrient status of anaemic (1995). Urinary excretion of retinol in children adolescent schoolgirls in Bangladesh. Am J Clin with acute diarrhea. Am J Clin Nutr 61: 1273–6. Nutr 82: 829–35. Alvarez R, De Andres J, Yubero P et al. Ahmed F, Khan MR, Jackson AA (2001). (1995). A novel regulatory pathway of brown fat Concomitant supplemental vitamin A enhances thermogenesis. J Biol Chem 270: 5666–73. the response to weekly supplementation of iron Alwitry A (2000). Vitamin A deficiency in co- and folic acid in anemic teenagers in urban Ban- eliac disease. Br J Ophthalmol 84: 1079-80. 194 References

Amatayakul K, Underwood BA, Ruckphao- shedding of herpes simplex virus among HIV-in- punt S et al. (1989). Oral contraceptives: effect of fected women: a randomized clinical trial. J Infect long-term use on liver vitamin A storage assessed Dis 189: 1466–71. by the relative dose response test. Am J Clin Nutr Baeten JM, McClelland RS, Richardson BA 49(5): 845–8. et al. (2002). Vitamin A deficiency and the acute Ambalavanan N, Ross AC, Carlo WA (2005). phase response among HIV-1-infected and -unin- Retinol-binding protein, transthyretin, and C- fected women in Kenya. J Acquir Immun Defic reactive protein in extremely low birth weight Syndr 31: 243–9. (ELBW) infants. J Perinatol 25: 714–9. Baeten JM, Mostad SB, Hughes MP et al. Ambalavanan N, Tyson JE, Kennedy KA et (2001). Selenium deficiency is associated with al. (2005).Vitamin A supplementation for ex- shedding of HIV-1-infected cells in the female tremely low birth weight infants: outcome at genital tract. J Acquir Immune Defic Syndr 26: 18–22 months. Pediatrics 115: e249–e254. 360–4. Ambalavanan N, Wu-Tzong J, Tyson JE et al. Baeten JM, Richardson BA, Bankson DD et (2003). A comparison of three vitamin A dosing al. (2004). Use of serum retinol-binding protein regimes in extremely low birth weight infants. J for prediction of vitamin A deficiency: effects Pediatr 142: 656–61. of HIV-1 infection, protein malnutrition, and the Andreozzi VL, Bailey TC, Nobre FF et al. acute phase response. Am J Clin Nutr 79: 218–25. (2006). Random-effects models in investigating Baeten JM, Wener MH, Bankson DD et al. the effect of vitamin A in childhood diarrhoea. (2006). HIV-1 infection alters the retinol-binding Ann Epidemiol 16: 241–7. protein:transthyretin ratio even in the absence of Anon (1972). Nutrition: a review of the WHO the acute phase response. J Nutr 136: 1624–9. programme.WHO Chronicle 26: 160–5. Bahl R, Bhandari N, Taneja S et al. (1997). Apgar J, Makdani D, Sowell AL et al. (1996). The impact of vitamin A supplementation on Reproducibility of relative dose response (RDR) physical growth of children is dependent on sea- test and serum retinol and retinyl ester concen- son. Eur J Clin Nutr 51: 26–9. trations in children after a 2-week interval. J Am Ballew C, Bowman BA, Sowell AL (2001). Coll Nutr 15: 450–7. Serum retinol distribution in residents of the Arne ST, Krisna CR, Rajiv B et al. (2002). Ef- United States: Third National Health and Nutri- fectiveness and efficacy of zinc for the treatment tion Examination Survey, 1988–1994. Am J Clin of acute diarrhea in young children. Pediatrics Nutr 73: 586–93. 109: 898–903. Baqui AH (2006). Diarrhoea and malnutrition Arroyave G (1986). Vitamin A deficiency in children. Replacing fluid and minerals, particu- control in Central America. In: Bauernfeind JC, larly zinc, remains vital. BMJ 332: 378. ed. Vitamin A deficiency and its control. Orlando, Barker DJP (1998). Mothers, babies and Florida: Academic Press Inc: 405–24. health in adult life. Edinburgh: Churchill-Living- Asonye CC (2002). Cassava fortification in stone. Nigeria: the journey so far. Sight and Life News- Barker ME, McCloskey E, Saha S et al. letter 4: 13–5. (2005). Serum retinoids and β-carotene as predic- Atkinson SA (2001). Special nutritional tors of hip and other fractures in elderly women. needs of infants for prevention of and recovery J Bone Min Res 20: 913–20. from bronchopulmonary dysplasia. J Nutr 131: Barua AB (1997). Retinoyl β-glucuronide: a 942S–6S. biologically active form of vitamin A. Nutr Revs 55: 259–67. Baeten JM, McClelland RS, Corey L et al. Batres RO, Olson RA (1987). Separation of (2004). Vitamin A supplementation and genital rat liver hepatocytes with high and low vitamin A References 195

concentration by flow cytometry. J Cell Biol 105: supplementation. Vaccine 33: 4111–12. 303a. Benn CS, Lisse IM, Balé C et al. (2000). No Batres RO, Olson RA (1987). A marginal strong long-term effect of vitamin A supplemen- vitamin A status alters the distribution of vitamin A tation in infancy on CD4 and CD8 T-cell subsets. among parenchymal and stellate cells in rat liver. Ann Trop Pediatr 20: 259–64. J Nutr 117: 874–9. Berdanier CD (2003). Vitamin A needs in Batten ML, Imanishi Y, Maeda T et al. (2004). diabetes mellitus. Sight and Life Newsletter 1: Lecithin-retinol acyltransferase is essential for 3–9. accumulation of all-trans-retinyl esters in the eye Berendschot TTJM, Broekmans WMR, Klop- and in the liver. J Biol Chem 279: 10422–32. ping-Ketelaars IAA et al. (2002). Lens aging in Beach RS, Mantero-Atienza E, Shor-Posner G relation to nutritional determinants and possible et al. (1992). Specific nutrient abnormalities in risk factors for age-related cataract. Arch Oph- asymptomatic HIV-1 infection. AIDS 6: 701–8. thalmol 120: 1732–7. Beaton GH, Martorell R, Aronson KJ et al. Berson EL (1999). Nutrition and retinal de- (1993). Effectiveness of vitamin A supplementa- generations. In: Modern nutrition in health and tion in the control of young child morbidity and disease, 9th edition (eds Shils ME, Olson JA, mortality in developing countries. ACC/SCN Shike M, Ross AC). Baltimore: Williams & State-of-the-art Series, Nutrition Policy Discus- Wilkins: 1491–1501. sion Paper no 13, United Nations, Geneva. Berson EL, Rosner B, Sandberg MA et al. Benn CS (2010). Opinion 2: We need to (1993). A randomized trial of vitamin A and understand more about the factors which modify vitamin E supplementation for retinitis pigmen- the effects of vitamin A. Sight and Life tosa. Arch Ophthalmol 111: 761–72. Magazine 1: 28-29. Berson EL, Rosner B, Sandberg MA et al. Benn CS, Aaby P, Balé C et al. (1997). Ran- (2004). Further evaluation of docosa-hexaenoic domized trial of effect of vitamin A supplementa- acid in patients with retinitis pigmentosa receiv- tion on antibody response to measles vaccine in ing vitamin A treatment. Arch Ophthalmol 122: Guinea-Bissau, West Africa. Lancet 350: 101–5. 1306–14. Benn CS, Balé C, George E et al. (2002). Ef- Bertram JS (2004). Dietary carotenoids, con- fect of vitamin A supplementation on measles- nexins, and cancer: what is the connection? Bio- specific antibody levels in Guinea-Bissau. Lancet chem Soc Trans 32 (Pt 6): 985–9. 359: 1313–4. Bertram JS, Vine AL (2005). Cancer preven- Benn CS, Balé C, Sommerfeldt H et al. tion by retinoids and carotenoids: independent (2003). Hypothesis: vitamin A supplementation action on a common target. Biochim Biophys and childhood mortality: amplification of the non- Acta 1740 (2): 170–8. specific effects of vaccines. Int J Epidemiol 32: Bhandari N, Bahl R, Taneja A et al. (2002). 822–8. Effect of routine zinc supplementation on pneu- Benn CS, Diness BR, Roth A et al. (2008). monia in children aged 6 months to 3 years: ran- Effect of 50,000IU vitamin A given with BCG domized controlled trial in an urban slum. BMJ vaccine on mortality in infants in Guinea-Bis- 324: 1358–61. seau: a randomised placebo controlled trial. BMJ Bhutta ZA (2006). Effect of infections and 336: 1416-1420. environmental factors on growth and nutritional Benn CS, Fisker AB, Diness BR et al. (2006). status in developing countries. J Pediatr Gastro- Neonatal vitamin A supplementation: sex-differ- enterol Nutr 43 Suppl 3: S13–21. ential effects on mortality? J Infec Dis 194: 719. Bhutta ZA (2007). Iron and zinc deficiency in Benn CS, Fisker AB, Jørgensen MJ et al. children in developing countries. BMJ 334: 104. (2008). Conflicting evidence for neonatal vitamin A Bhutta ZA, Black RE, Brown KH et al. 196 References

(1999). Prevention of diarrhea and pneumonia by Bangladesh. Am J Epidemiol 142: 843–55. zinc supplementation in children in developing Blomhoff R (1994). Introduction: overview of countries: pooled analysis of randomized con- vitamin A metabolism and function. In: Vitamin A trolled trials. J Pediatr 135: 689–97. in health and disease. Blomhoff R, ed. New York: Biebinger R, Arnold M, Langhans W et al. Marcel Dekker: 1–35. (2007). Vitamin A repletion in rats with concur- Boileau TWM, Boileau AC, Erdman JW Jr rent vitamin A and iodine defiency affects TSHβ (2002). Bioavailability of all-trans and cis-iso- gene expression and reduces thyroid hyperstimu- mers of lycopene. Exp Biol Med 227: 914–9. lation and thyroid size. J Nutr 137: 573–7. Boileau TWM, Moore AC, Erdman JW Jr Biesalski HK, Frank J, Beck SC et al. (1999). (1999). Carotenoids and vitamin A. In: Antioxi- Biochemical but not clinical vitamin A deficien- dant status, diet, nutrition, and health (ed Papas cy results from mutations in the gene for retinol AM). Boca Raton: CRC Press: 133–58. binding protein. Am J Clin Nutr 69: 931–6. Bone RA, Landrum JT (2005). Macular Biesalski H, Reifen R, Fürst P et al. (1999). carotenoids in eye health. In: Carotenoids and Retinyl palmitate supplementation by inhalation retinoids: molecular aspects and health issues. of an aerosol improves vitamin A status of pre- Packer L, Obermuller-Jevic U, Kraemer K, Sies H, school children in Gondar (Ethiopia). Br J Nutr eds. Champaign, Illinois: AOCS Press, 115–29. 82: 179–82. Borghos L, Rahmani M, Elkari K et al. (2002). Binka FN, Ross DA, Morris SS et al. (1995). Fortification of vegetable oil with vitamin Ain Vitamin A supplementation and childhood malar- Morocco: a stability study. Sight and Life News- ia in northern Ghana. Am J Clin Nutr 61: 853–9. letter 4: 10–2. Bishai D, Kumar KCS, Waters H et al. (2005). Botto LD, Loffredo C, Scanlon KS et al. The impact of vitamin A supplementation on (2001). Vitamin A and cardiac outflow tract de- mortality inequalities among children in Nepal. fects. Epidemiology 12: 491–6. Hlth Pol Plan 20: 60–6. Bouis H (1996). Enrichment of food staples Black R, Allen L, Bhutta Z et al. (2008). through plant breeding: a new strategy for fight- Maternal and child undernutrition: global and ing micronutrient malnutrition. Nutr Revs 54: regional exposures and health consequences. 131–7. Lancet 371: 243–60. Britton G (1995). Structure and properties of Blaner WS (1998). Recent advances in under- carotenoids in relation to function. FASEB J 9: standing the molecular basis of vitamin A action. 1551–8. Sight and Life Newsletter 2: 3–6. Britton G, Liaaen-Jensen S, Pfander H (eds) Blaner WS, Gamble MV, Burger H et al. (2004). Handbook of carotenoids. Compiled by (1997). Maternal serum vitamin A levels are not Mercadante AZ, Egeland ES. associated with mother-to-child transmission of Brown KH, Gaffar A, Alamgir SM (1979). HIV-1 in the United States. Proc XVIII IVACG Xerophthalmia, protein-calorie malnutrition, and Meeting, Cairo, Egypt 22–26 September: 89. infections in children. J Pediatr 95: 651–6. Blaner WS, Piantedosi R, Gamble MV (1999). Brown N, Roberts C (2004). Vitamin A for Use of retinol-binding protein as a surrogate mea- acute respiratory infection in developing coun- sure for serum retinol: possible effects of apo- tries: a meta-analysis. Acta Paediatr 93: 1437–2. retinol-binding protein, transthyretin, lipid and Bryce J, Black RE, Walker N et al. (2005). protein denaturing conditions. Proc XIX IVACG Can the world afford to save the lives of 6 million Meeting, Durban, South Africa, 8–11 March: 92. children each year? Lancet 365: 2193–200. Bloem MW, Hye A, Wijnroks M et al. (1995). Bryce J, Boschi-Pinto C, Shibuya K (2005). The role of universal distribution of vitamin A WHO estimates of the causes of death in young capsules in combating vitamin A deficiency in children. Lancet 365: 1147–52. References 197

Buck J, Ritter G, Dannecker L et al. (1991). (2000). Child mortality and impact and costs of Intracellular signaling by 14-hydroxy-4, 14-ret- integrating vitamin A supplementation into im- roretinol. Science 254: 1654–5. munization campaigns. Am J Public Health 90: Bulux J, Quan de Serrano J, Giuliano A et al. 1526–9. (1994). Plasma response of children to short-term Chintu C, Mudenda V, Lucas S et al. (2002). chronic β-carotene supplementation. Am J Clin Lung diseases at necropsy in African children dy- Nutr 59: 1369–75. ing from respiratory illnesses: a descriptive nec- Burke DS, Smidt CR, Vuong LT (2005). ropsy study. Lancet 360: 985–90. Momordica cochinchinensis, Rosa roxburghii, Chowdhury S, Kumar R, Ganguly NK et al. wolfberry, and sea buckthorn – highly nutritional (1996). Conjunctival impression cytology with fruits supported by tradition and science. Curr transfer (CICT) to detect pre-clinical vitamin A Top Nutraceut Res 3: 259–66. deficiency among slum children in India. BrJ Burri BJ (2001). The formation of vitamin A Nutr 75: 785–90. from β-carotene: good, bad, and variable. Sight Chowdhury S, Kumar R, Ganguly NK et al. and Life Newsletter 2: 3–5. (1997). Dynamics of conjunctival impression cy- Burri BJ, Clifford AJ (2005). β-carotene me- tologic changes after vitamin A supplementation. tabolism measured by accelerator mass spectrom- Br J Nutr 77: 863–9. etry. Sight and Life Newsletter 1: 6–8. Chowers I, Banin E, Merin S et al. (2001). Burrow CR (2000). Retinoids and renal devel- Long-term assessment of combined vitamins A opment. Exp Nephrol 8: 4–5. and E treatment for prevention of retinal degen- Bushamuka VN, de Pee S, Talukder A et al. eration in abetalipoproteinemia and hypolipopro- (2005). Impact of a homestead gardening proteinemia patients. Eye 15: 520–30. gram on household food security and empower- Christen WG, Glynn RJ, Chew EY et al. ment of women in Bangladesh. Food Nutr Bull (2009). Folic acid, pyridoxine, and cyanocobala- 26: 17–25. min combination treatment and age-related macular degeneration in women. Arch Intern Med 169: Chen P, Soares AM, Lima AAM et al. (2003). 335–41. Association of vitamin A and zinc status with al- Christian P, Bentley ME, Pradhan R et al. tered intestinal permeability: analysis of cohort (1998). An ethnographic study of night blindness data from northeastern Brazil. J Hlth Pop Nutr “ratauni” among women in the Terai of Nepal. 21: 309–15. Soc Sci Med 46: 879–89. Chen Q, Ross AC (2005). Vitamin A and Christian P, Khatry SK, Katz J et al. (2003). immune function: retinoic acid modulates popu- Effects of alternative maternal micronutrient sup- lation dynamics in antigen receptor and CD38- plements on low birth weight in rural Nepal: a stimulated splenic B cells. Proc Natl Acad Sci double blind randomized community trial. BMJ USA 102: 14142–9. 326: 571–6. Chen Y, Noy N (1994). Retinoid specificity of Christian P, Schulze K, Stoltzfus RJ et al. interphotoreceptor retinoid-binding protein. Bio- (1998). Hyporetinolemia, illness symptoms, and chemistry 33: 10658–65. acute phase protein response in pregnant women Chew BP, Park JS (2004). Carotenoid action with and without night blindness. Am J Clin Nutr on the immune response. J Nutr 134: S257–S61. 67: 1237–43. Chichili GR, Nohr D, Schaeffer M et al. Christian P, Thorne-Lyman AL, West KP Jr (2005). Beta-carotene conversion into vitamin A et al. (1998). Working after the sun goes down: in human retinal pigment epithelial cells. Invest exploring how night blindness impairs women’s Ophthalmol Vis Sci 46: 3562–9. work activities in rural Nepal. Eur J Clin Nutr 52: Ching P, Birmingham M, Goodman R et al. 519–24. 198 References

Christian P, West KP Jr, Katz J et al. (2005). tion in young South Indian infants. Pediatr Infect Impact of antenatal micronutrient supplementa- Dis J 20: 289–95. tion on self-reported labor and delivery compli- Congdon N, Dreyfuss ML, Christian P et al. cations and puerpural morbidity. IVACG XXII (1999). Dark adaptation testing among pregnant Meeting abstract T60: 60. women in rural Nepal. Proc XIX IVACG Meet- Christian P, West KP Jr, Khatry SK et al. ing, Durban, South Africa: 92. (1998). Night blindness of pregnancy in rural Congdon N, Dreyfuss ML, Christian P et al. Nepal – nutritional and health risks. Int J Epide- (2000). Responsiveness of dark adaptation thresh- miol 27: 231–7. old to vitamin A and β-carotene supplementation Christian P, West KP Jr, Khatry SK et al. in pregnant and lactating women in Nepal. Am (1998). Vitamin A or β-carotene supplementa- J Clin Nutr 72: 1004–9. tion reduces but does not eliminate maternal night Congdon N, Sommer A, Severns M et al. blindness in Nepal. J Nutr 128: 1458–63. (1995). Pupillary and visual thresholds in young Christian P, West KP Jr, Khatry SK et al. children as an index of population vitamin A (2000). Vitamin A or β-carotene supplementa- status. Am J Clin Nutr 61: 1076–82. tion reduces symptoms of illness in pregnant and Copenhagen Consensus (2008). Malnutrition lactating Nepali women. J Nutr 130: 2675–82. and Hunger. www.copenhagenconsensus.com/ Christian P, West KP Jr, Khatry SK et al. Default.aspx?ID=1149. (2000). Night blindness during pregnancy and Cormack DH (1987). Ham’s Histology, 9th subsequent mortality among women in Nepal: edn. JB Lippincot Company, Philadelphia, pp effects of vitamin A and β-carotene supplementa- 273–323. tion. Am J Epidemiol 152: 542–7. Countdown to 2015 Conference (2006). The Christian P, West KP Jr, Khatry SK et al. coming decade for global action in child health. (2001). Maternal night blindness increases risk Lancet 367: 3–5. of mortality in the first six months of life among Courtright P, Fine D, Broadhead RL et al. infants in Nepal. J Nutr 131: 1510–2. (2002). Abnormal vitamin A cytology and mortal- Chung H-Y, Rasmussen HM, Johnson EJ ity in infants aged 9 months and less with measles. (2004). Lutein bioavailability is higher from Ann Trop Pediatr 22: 239–43. lutein-enriched eggs than from supplements and Coutsoudis A, Bobat R, Coovadia HM et al. spinach in men. J Nutr 134: 1887–93. (1994). Vitamin A prophylaxis reduced morbidity Chytil F (1992). The lungs and vitamin A. in HIV-1 infected infants: a controlled trial. XVI Am J Physiol 262 (Lung Cell Mol Physiol 6): IVACG Meeting Report, Chiang Rai, Thailand. L517–27. Washington DC: The Nutrition Foundation. Clagett-Dame M, DeLuca HF (2002). The Coutsoudis A, Broughton M, Coovadia HM role of vitamin A in mammalian reproduction (1991). Vitamin A supplementation reduces mea- and embryonic development. Ann Rev Nutr 22: sles morbidity in young African children: a ran- 347–81. domized, placebo-controlled, double-blind trial. Cohen N, Rahman H, Mitra M et al. (1987). Am J Clin Nutr 54: 890–5. Impact of massive doses of vitamin A on nutri- Coutsoudis A, Kiepiela P, Coovadia HM et tional blindness in Bangladesh. Am J Clin Nutr al. (1992). Vitamin A supplementation enhances 45: 970–76. specific IgG antibody levels and total lymphocyte Coleman H, Chew E (2007). Nutritional sup- numbers while improving morbidity in measles. plementation in age-related macular degenera- Pediatr Infect Dis J 11: 203–9. tion. Curr Opin Ophthalmol 18: 220-3. Cox SE, Arthur P, Kirkwood BR et al. (2006). Coles Cl, Kanungo R, Rahmathullah R et al. Vitamin A supplementation increases ratios of (2001). Pneumococcal nasopharyngeal coloniza- proinflammatory to anti-inflammatory cytokine References 199

responses in pregnancy and lactation. Clin Exp Meeting Report, Chiang Rai, Thailand. Washing- Immunol 144: 392–400. ton DC: The Nutrition Foundation. Cox SE, Staalsoe T, Arthur P et al. (2005). Deming DM, Baker DH, Erdman Jr JW Maternal vitamin A supplementation and immu- (2002). The relative vitamin A value of 9-cis nity to malaria in pregnancy in Ghanaian primi- β-carotene is less and that of 13-cis β-carotene gravids. Trop Med Int Health 10: 1286–97. may be greater than the accepted 50% that of all- Craft NE (1999). Development of a rapid trans β-carotene in gerbils. J Nutr 132: 2709–12. vitamin A field test. Proc XIX IVACG Meeting, Demmig-Adams B, Gilmore AM, Adams III Durban, South Africa, 8-11 March: 96. WA (1996). In vivo functions of carotenoids in Craft NE, Haitema T, Brindle LK et al. (2000). higher plants. FASEB J 10: 403–12. Retinol analysis in dried blood spots by HPLC. De Onis M, Monteiro C, Akre J et al. (1993). J Nutr 130: 882–5. The worldwide magnitude of protein-energy mal- Crandall CJ (2004). Vitamin A intake and nutrition: an overview from the WHO Global osteoporosis: a clinical review. J Women’s Hlth Database on Child Growth. Bull Wld Hlth Org 13: 939–53. 71: 703–12. Curtale F, Pokhrel RP, Tilden RL et al. (1995). de Pee S, Bloem MW, Gorstein J et al. (1998). Intestinal helminths and xerophthalmia in Nepal: Reappraisal of the role of vegetables for vitamin A a case-control study. J Trop Pediat 41: 334–6. status of mothers in central Java, Indonesia. Am J Cusick SE, Tielsch JM, Ramsan M et al. Clin Nutr 68: 1068–74. (2005). Short-term effects of vitamin A and anti- de Pee S, Bloem MW, Halati S et al. (1999). malarial treatment on erythropoiesis in severely 24-VASQ method for estimating vitamin A intake anemic Zanzibari preschool children. Am J Clin reproducibility and relationship with vitamin A Nutr 82: 406–12. status. Proc XIX IVACG Meeting, Durban, South Czeizel AE, Rockenbauer M (1998). Preven- Africa, 9–11 March: 96. tion of congenital abnormalities by vitamin A. de Pee S, Bloem MW, Kiess L et al. (1999). Internat J Vit Nutr Res 68: 219–31. Serum retinol concentration of non-breastfed Bangladeshi children determined by, amongst Darlow BA, Graham PJ (2000). Vitamin A others, mother’s vitamin A intake. Proc XIX supplementation for preventing morbidity and IVACG Meeting, Durban, South Africa, 9–11 mortality in very low birth weight infants. Co- March: 95. chrane Database Syst Rev 2000, 2: 1469–93X. de Pee S, Kraemer K, van den Briel T et al. Datta K, Parkhi V, Rai M et al. (2005). Bio- World Food Programme; Sprinkles Global Health engineered provitamin A enriched tropical rice. Initiative (2008). Quality criteria for micronu- IVACG XXII Meeting, M58: 43. trient powder products: report of a meeting organ- Daulaire NMP, Starbuck ES, Houston RM et ized by the World Food Programme and Sprinkles al. (1992). Childhood mortality after a high dose Global Health Initiative. Food Nutr Bull 29(3): of vitamin A in a high- risk population. BMJ 304: 232–41. 207–10. de Pee S, West CE (1996). Dietary carote- de Francisco A, Chakraborty J, Chowdhury noids and their role in combating vitamin A defi- HR et al. (1993). Acute toxicity of vitamin A ciency: a review of the literature. Eur J Clin Nutr given with vaccines in infancy. Lancet 342: 50, Suppl 3: S38–S53. 526–7. de Pee S, West CE, Muhilal et al. (1995). de Francisco A, Yasui Y, Chakraborty J (1994). Lack of improvement in vitamin A status with Vitamin A supplementation given to mothers increased consumption of dark-green leafy veg- after delivery reduces infant mortality and in- etables. Lancet 346: 75–81. creases symptoms of morbidity. XVI IVACG de Pee S, West CE, Permaesih D et al. (1998). 200 References

Orange fruit is more effective than are dark-green, Dowell SF, Papic Z, Bresee JS et al. (1996). leafy vegetables in increasing serum concentra- Treatment of respiratory syncitial virus infection tions of retinol and β-carotene in schoolchildren with vitamin A: a randomized, placebo-controlled in Indonesia. Am J Clin Nutr 68: 1058-1067. trial in Santiago, Chile. Pediatr Infect Dis J 15: de Pee S, Yuniar Y, West CE et al. (1997). 782–6. Evaluation of biochemical indicators of vitamin A Dräger UC, Wagner E, McCaffery P (1998). status in breast-feeding and non-breast-feeding Aldehyde dehydrogenases in the generation of Indonesian women. Am J Clin Nutr 66: 160–7. retinoic acid in the developing vertebrate: a cen- Desai NC, Desai S, Desai R (1992). Xe- tral role of the eye. J Nutr 128: 463S–6S. rophthalmia clinics in rural eye camps. Int Oph- Dresser DW (1963). Adjuvanticity of vita- thal 16: 139–45. min A. Nature 217: 527–9. Dijkhuizen MA, Wieringa FT, West CE et al. Dreyfuss ML, Stoltzfus RJ et al. (2000). (2004). Zinc plus β-carotene supplementation of Hookworms, malaria and vitamin A deficiency pregnant women is superior to β-carotene supple- contribute to iron deficiency among pregnant mentation alone in improving vitamin A status women in the plains of Nepal. J Nutr 130: in both mothers and infants. Am J Clin Nutr 80: 2527–36. 1299–307. Du X, Tabeta K, Mann N et al. (2005). An Dijkhuizen MA, Wieringa FT, West CE essential role for Rxr (alpha) in the development (2005). Pregnancy affects plasma concentrations of the Th2 responses. Eur J Immunol 35: 3414–23. of retinol and acute phase response proteins. Dunchek B, Nussenblatt B, Ricks MO et al. IVACG XXII Meeting: T56. (2005). Breast milk retinol concentrations are not Dimitrov NV, Meyer C, Ullrey DE et al. associated with systemic inflammation among (1988). Bioavailability of β-carotene in humans. breat-feeding women in Malawi. J Nutr 135: Am J Clin Nutr 48: 298–304. 223–6. Djakoure C, Guibourdeuche J, Porquet D et During A, Hussain MM, Morel DW et al. al. (1996). Vitamin A and retinoic acid stimulate (2002). Carotenoid uptake and secretion by within minutes cAMP release and growth hor- CaCo-2 cells: β-carotene isomer selectivity and mone secretion in human pituitary cells. J Clin carotenoid interactions. J Lip Res 43: 1086–95. Endo Metab 81: 3123–6. Djunaedi E, Sommer A, Pandji A et al. (1988). Edejer T T-T, Aikins M, Black R et al. (2005). Impact of vitamin A supplementation on xeroph- Cost effectiveness analysis of strategies for child thalmia. A randomized controlled community health in developing countries. BMJ 331: 1177– trial. Arch Ophthal 106: 218–22. 80. Dkhissi-Benyaha O, Rieux C, Hut RA et Edmond KM, Zandoh C, Quigley MH et al. (2006). Immunohistochemical evidence of a al. (2006). Delayed breast-feeding initiation in- melanopsin cone in human retina. Invest Ophthal creases risk of neonatal mortality. Pediatrics 117: Vis Sci 47: 1636–41. e380-e386. Donnen P, Dramaix M, Brasseur D et al. English RM, Badcock JC, Giay T et al. (1997). (1998). Randomized placebo-controlled clinical Effect of nutrition improvement project on mor- trial of the effect of a single high dose or daily bidity from infectious diseases in preschool chil- low doses of vitamin A on the morbidity of hos- dren in Vietnam: comparison with control com- pitalized, malnourished children. Am J Clin Nutr mune. BMJ 315: 1122–5. 68: 1254–60. Enwonwu, CO, Phillips RS (2004). Increased Dorosko SM (2005). Vitamin A, mastitis, retinol requirement in acute measles infection and mother-child transmission of HIV-1 through in children: an hypothesis on role of hypercorti- breast feeding. Nutr Revs 63: 332–46. solemia. Nutr Res 24: 223–7. References 201

Erhardt J (2005). Use of dried blood spots ments in relation to mortality among human im- (DBS) – a simple and field-friendly method munodeficiency virus-infected and uninfected of collecting blood samples for the measure- children in Tanzania. Pediat Inf Dis J 18: 127– ment of vitamin A status. Sight and Life News- 33. letter 1: 4–6. Fawzi WW, Mbise WW, Spiegelman D et al. Eskild W, Hansson V (1994). Vitamin A func- (2000). Vitamin A supplements and diarrheal and tions in the reproductive organs. In: Vitamin A in respiratory tract infections among children in Dar health and disease. Blomhoff R, ed. New York: es Salaam, Tanzania. J Pediatr 137: 660–7. Marcel Dekker: 531–59. Fawzi WW, Msamanga GI, Hunter D et al. Ettyang G, Oloo A, van Marken Lichtenbelt (2002). Randomized trial of vitamin supplements W et al. (2004). Consumption of vitamin A by in relation to transmission of HIV-1 through breast feeding children in rural Kenya. Food Nutr breast-feeding and early child mortality. AIDS 6: Bull 25: 256–63. 1935–44. Evain-Brion D, Porquet D, Thérond P et Fawzi WW, Msamanga GI, Spiegelman D et al. (1994). Vitamin A deficiency and noctur- al. (1998). Randomized trials of effects of vita- nal growth hormone secretion in short children. min supplements on pregnancy outcomes and T Lancet 343: 87–8. cell counts in HIV-1-infected women in Tanzania. Ezzati M, Lopez AD, Rodgers A et al. (2002). Lancet 351: 1477–82. Selected major risk factors and global and region- Fawzi WW, Msamanga GI, Spiegelman D al burden of disease. Lancet 360: 1347–60. et al. (2004). A randomized trial of multivitamin supplements and HIV disease progression and FAO/WHO (1976). Requirements of vita- mortality. N Eng J Med 351: 78–80. min A, thiamine, riboflavin, and niacin. WHO Fawzi WW, Msamanga GI, Wei R et al. Technical Report Series no 362. Geneva: WHO. (2003). Effect of providing vitamin supple- FAO/WHO Expert Consultation (1988). Re- ments to human immunodeficiency virus-infected quirements of vitamin A, iron, folate and vitamin lactating mothers on the child’s morbidity and

B12. Food and Nutrition Series no. 23. Rome: CD4 (+) cell counts. Clin Infect Dis 36: 1053– FAO. 62. Farias EF, Ong DE, Ghyselinck et al. (2005). Fiedler JL (2000). The Nepalese National Cellular retinol-binding protein 1, a regulator of Vitamin A Program: prototype to emulate or breast epithelial retinoic acid receptor activity, donor enclave? Hlth Pol Plan 15: 145–56. cell differentiation and tumoricity. J Nat Cancer Fine P (2000). Commentary: an unexpected Inst 97: 21–9. finding that needs confirmation or rejection. BMJ Fariss RN, Zong-Yi L, Millam H (2000). Ab- 321: 1439. normalities in rod photoreceptors, amacrine cells, Florentino RF, Tanchoco CC, Ramos AC and horizontal cells in human retinas with retinitis et al. (1990). Tolerance of preschoolers to two pigmentosa. Am J Ophthalmol 129: 215–23. dosage strengths of vitamin A preparation. Am Fawzi WW (2006).The benefits and concerns J Clin Nutr 52: 694–700. related to vitamin A supplementation. J Infec Dis Flores H, Campos F, Aranjo CRC et al. (1984). 193: 756–9. Assessment of marginal vitamin A deficiency in Fawzi WW, Herrera G, Willett WC et al. Brazil in children using the relative dose response (1997). The effect of vitamin A supplementation procedure. Am J Clin Nutr 40: 1281–9. on the growth of preschool children in the Sudan. Folb PI (2001). WHO responds to Guinea- Am J Publ Hlth 87: 1359–62. Bissau report. BMJ 322: 361. Fawzi WW, Mbise RL, Hertzmark E et al. Folli C, Viglioni S, Busconi M et al. (2005). (1999). Randomized trial of vitamin A supple- Biochemical basis for retinol deficiency induced 202 References

by the 141N and G75D mutations in human shall Islands. Am J Clin Nutr 73: 594–601. plasma retinol-binding protein. Biochem Biophys Ganji V, Kafai MR (2005). Population de- Res Comm 336: 1017–22. terminants of serum lycopene concentrations in Food and Nutrition Board, Institute of Medi- the United States: data from the Third National cine (2001). Dietary reference intakes for vita- Health and Nutrition Examination Survey, 1988- min A, vitamin K, arsenic, boron, chromium, 1994. J Nutr 135: 567–72. copper, iodine, iron, manganese, molybdenum, Garrett DA, Failla ML, Sarama RJ (1999). nickel, silicon, vanadium, and zinc. Washington Development of an in vitro method to assess car- DC: National Academy Press. otenoid bioavailability from meals. J Agri Food Forrester AD, Dick P, McMenami W et al. Chem 47: 4301–9. (1996). The eye: basic sciences in practice. New Gasparetto C, Malinverno A, Culacciati D et York: Saunders: 194. al. (2005). Antioxidant vitamins reduce oxidative Foster A, Sommer A (1987). Corneal ulcer- stress and ventricular remodeling in patients with ation, measles, and childhood blindness in Tanza- acute myocardial infarction. Int J Immunopathol nia. Br J Ophthal 71: 331–3. Pharmacol 18: 487–96. Foster A, Yorston D (1992). Corneal ulcer- Gaziano JM, Sesso HD (2005). Carotenoids ation in Tanganyikan children: relationship be- and cardiovascular disease. In: Carotenoids and tween measles and vitamin A deficiency. Trans R retinoids: molecular aspects and health issues. Soc Trop Med Hyg 86: 454–5. Packer L, Obermuller-Jevic U, Kraemer K, Sies Foster R, David-Gray Z, Lucas R (2001). H, eds. Champaign, Illinois: AOCS Press, 321–37. Eye, light, the eye, and the regulation of biologi- Gellermann W, Zidichouski JA, Smidt CR cal time. Eye News 7: 24–30. et al. (2005). Raman detection of carotenoids in Fraunfelder FT, Fraunfelder FW, Edwards R human tissue. In: Carotenoids and retinoids: mole- (2001). Ocular effects possibly associated with iso- cular aspects and health issues. Packer L, Ober- tretinoin usage. Am J Ophthalmol 132: 299–305. muller-Jevic U, Kraemer K, Sies H, eds. Cham- French AL, Cohen MH, Gange SJ et al. paign, Illinois: AOCS Press, 86–114. (2002).Vitamin A deficiency and genital viral bur- Genaro P de S, Martini LA (2004). Vitamin A den in women infected with HIV-1. Lancet 359: supplementation and risk of skeletal fracture. 1210–2. Nutr Revs 62: 65–72. Fujita M, Brindle E, Rocha A (2009). Assess- Genton B, Al-Yaman F, Semba RD et al. ment of the relative dose-response test based on (1994). Vitamin A status and malaria infection, serum retinol-binding protein instead of serum morbidity and immunity in a highly endemic area retinol in determining low hepatic vitamin A of Papua New Guinea. XVI IVACG Meeting Re- stores. Am J Clin Nutr 90: 217–224. port, Chiang Rai, Thailand. Washington DC: The Funk C (1912). The etiology of the deficiency Nutrition Foundation. diseases. J State Med 220: 341–68. Gerster H (1997). Vitamin A – functions, Furr HC, Barua AB, Olson JA (1992). Reti- dietary requirements and safety in humans. Int J noids and carotenoids. In: Modern chromato- Vit Nutr Res 67: 71–90. graphic analysis of the vitamins, 2nd edition (eds Ghana VAST Study Team (1993). Vitamin A Nelis HJ, Lambert WE, DeLeenheer AP). New supplementation in northern Ghana: effects on York: Marcel Dekker, 1–72. clinic attendances, hospital admissions, and child mortality. Lancet 342: 7–12. Gamble MV, Ramakrishnan R, Palafox NA et Gillespie C, Ballew C, Bowman BA et al. al. (2001). Retinol-binding protein as a surrogate (2004). Intraindividual variation in serum retinol measure for serum retinol: studies in vitamin A- concentrations among participants in the third deficient children from the Republic of the Mar- National Health and Nutrition Examination Sur- References 203

vey, 1988–1994. Am J Clin Nutr 79: 625–32. fortified rice in iron-deficient, pregnant, nightblind Gireesh T, Nair PP, Sudhakaran PR (2004). Nepali women. Am J Clin Nutr 85: 1375–84. Studies on the bioavailability of the provitamin A Graham TE, Yang Q, Bluher M et al. 2006. carotenoid β-carotene, using human exfoliated Retinol-binding protein 4 and insulin resistance colonic epithelial cells. Brit J Nutr 92: 241–5. in lean, obese, and diabetic subjects. N Eng J Med Gittelsohn J, Shankar AV, West KP Jr et al. 354: 2552–63. (1997). Infant feeding practices reflect antecedent Green HN, Mellanby E (1928). Vitamin A as risk of xerophthalmia in Nepalese children. Eur J an anti-infective agent. BMJ October 20: 691–6. Clin Nutr 51: 484–90. Green MH, Green JB (1996). Quantitative Gittelsohn J, Shankar AV, West KP Jr et al. and conceptual contributions of mathematical (1998). Child feeding and care behaviors are as- modeling to current views on vitamin A metabo- sociated with xerophthalmia in rural Nepalese lism, biochemistry, and nutrition. Adv Food Nutr households. Soc Sci Med 47: 477–86. Res 40: 3–24. Giuliano G, Al-Babili S, von Lintig J (2003). Green M, Green JB (2005). Contributions of Carotenoid oxygenases: cleave it or leave it. mathematical modeling to understanding whole- Trends Plant Sci 8: 145–9. body vitamin A metabolism and to the assessment Gollapalli DR, Rando RR (2004). The specific of vitamin A status. Sight and Life Newsletter 2: binding of retinoic acid to RPE65 and approaches 4–10. to the treatment of macular degeneration. Proc Greer RM, Buntain HM, Lewindon PJ et al. Natl Acad Sci USA 101: 10030–5. (2004). Vitamin A levels in patients with cystic Goodman DS, Huang HS, Shiratori T (1966). fibrosis are influenced by the inflammatory pro- Mechanism of the biosynthesis of vitamin A from cess. Cystic Fibrosis 3: 143–9. β-carotene. J Biol Chem 241: 1929-1932. Gregory JR, Collins DL, Davies PSW et al. Goodman GE, Thornquist MD, Balmes J et (1995). Report of the Diet and Nutrition Survey; al. (2004). The β-carotene and retinol efficacy vol 1. London: HMSO. trial: incidence of lung cancer and cardiovascular Gregory JR, Foster K, Tyler H et al. (1990). disease mortality during 6-year follow up after The Dietary and Nutritional Survey of British stopping β-carotene and retinol supplies. J Nat Adults. London: HMSO. Cancer Inst 96: 1743–50. Gregory JR, Lowe S, Bates CJ et al. (2000). Goodman MT, Kiviat N, McDuffie C (1998). National Diet and Nutrition Survey of Children 4 The association of plasma micronutrients with the to 18 Years. vol. I: Report of the Diet and Nutri- risk of cervical dysplasia in Hawaii. Cancer Epi- tion Survey. London: HMSO. demiol Biomarkers Prev 7: 537–44. Gritz DC, Srinivasan M, Smith SD et al. Gorstein J, Bhaskaram P, Khanum S et al. (2006). The antioxidants in prevention of cata- (2003). Safety and impact of vitamin A supple- racts study: effects of antioxidant supplements mentation delivered with oral polio vaccine as on cataract progression in South India. Brit J part of the immunization campaign in Orissa, Ophthal 90: 847–51. India. Food Nutr Bull 24: 319–31. Grotto I, Mimouni M, Gdalevich M et al. Goto R, Panter-Brick C, Northrop-Clewes (2003). Vitamin A supplementation and childhood CA, Manahdhar R, Tuladhar NR (2002). Poor morbidity from diarrhea and respiratory infec- intestinal permeability in mildly stunted Nepali tions: a meta-analysis. J Pediatr 142: 297–304. children: associations with weaning pratices and Grubesic RB (2004). Children aged 6 to 60 Giardia lamblia infection. Br J Nutr 88: 141–149. months in Nepal may require a vitamin A supple- Graham JM, Haskell MJ, Pandey P et al. ment regardless of dietary intake from plant and (2007). Supplementation with iron and riboflavin animal food sources. Food Nutr Bull 25: 248–55. enhances dark adaptation response to vitamin A- Gundersen TE (2006). Methods for detecting 204 References

and identifying retinoids in tissue. J Neurobiol 66: Bangladeshi surgical patients. Am J Clin Nutr 66: 631–44. 67–74. Haskell MJ, Jamil KM, Hassan F et al. (2004). Hadi H, Dibley MJ, West KP Jr (2004). Com- Daily consumption of Indian spinach (Basella plex interactions with infections and diet may alba) or sweet potatoes has a positive effect on explain seasonal growth responses to vitamin A total-body vitamin A stores in Bangladeshi men. in pre-school-aged Indonesian children. Eur J Am J Clin Nutr 80: 705–14. Clin Nutr 58: 990–9. Haskell MJ, Mazumder RN, Peerson JM et Hadi H, Stoltzfus JR, Dibley MJ et al. (2000). al. (1999). Use of the deuterated-retinol-dilution Vitamin A supplementation selectively improves technique to assess total-body vitamin A stores of the linear growth of Indonesian preschool chil- adult volunteers consuming different amounts of dren: results from a randomized controlled trial. vitamin A. Am J Clin Nutr 70: 874–80. Am J Clin Nutr 71: 507–13. Haskell MJ, Pandey P, Graham JM et al. Hail N Jr, Lotan R (2005). Mitochondria as (2005) Recovery from impaired dark adaptation novel targets for proapoptotic synthetic retinoids. in night-blind pregnant Nepali women who re- In: Carotenoids and retinoids: molecular aspects ceive small daily doses of vitamin A as amaranth and health issues. Packer L, Obermuller-Jevic leaves, carrots, goat liver, vitamin A-fortified rice, U, Kraemer K, Sies H, eds. Champaign, Illinois: or retinyl palmitate. Am J Clin Nutr 81: 461–71. AOCS Press, 229–43. Hendrickx AG, Peterson P, Hartmann D et Hakimi M, Dibley MJ, Suryono A et al. al. (2000). Vitamin A teratogenicity and risk (1999). Impact of vitamin A and zinc supplements assessment in the macaque retinoid model. Re- on maternal postpartum infections, in rural cen- prod Toxicol 14: 311–23. tral Java, Indonesia. Proc XIX IVACG Meeting, Herrera MG, Fawzi WW, Nestel P (1996). Durban, South Africa, 9–11 March: 79. Effect of vitamin A supplementation on the inci- Hamilton WD, Brown SP (2001). Autumn dence of cough, diarrhea, and fever. XVII IVACG tree colors as a handicap signal. Proc Roy Soc B Meeting Report, Guatemala City. Washington 268: 1489–93. DC: The Nutrition Foundation, 95. Hammond CJ, Snieder H, Spector TD et al. Herrera MG, Nestel P, El Amin A et al. (1992). (2000). Genetic and environmental factors in Vitamin A supplementation and child survival. age-related nuclear cataracts in monozygotic and Lancet 340: 267–71. dizygotic twins. N Eng J Med 342: 1786–90. Hickenbottom SJ, Follett JR, Lin Y et al. Harnois C, Samson J, Malenfaut M et al. (2002). Variability in conversion of β-carotene to (1989). Canthaxanthin retinopathy: anatomic and vitamin A in men as measured by using a double- functional reversibility. Arch Ophthalmol 107: tracer study design. Am J Clin Nutr 75: 900–7. 538–40. Hickenbottom SJ, Lemke SL, Dueker SR et al. Hartmann S, Brors O, Bock J et al. (2005). (2002). Dual isotope test for assessing β-carotene Exposure to retinoic acids in non-pregnant wom- cleavage to vitamin A in humans. Eur J Nutr 41: en following high vitamin A intake with a liver 141–7. meal. Int J Vitam Nutr Res 75: 187–94. Hix LM, Vine AL, Lockwood SF, et al. (2005). Harvey P (2002). Enriching breast milk with Retinoids and carotenoids as cancer chemopre- vitamin A. Sight and Life Newsletter 3/2002: ventive agents: role of upregulated gap junctional 41– 44. communication. In: Carotenoids and retinoids: Haskell MJ, Handelman GJ, Peerson JM et molecular aspects and health issues. Packer L, al. (1997). Assessment of vitamin A status by the Obermuller-Jevic U, Kraemer K, Sies H, eds. deuterated-retinol-dilution technique and com- Champaign, Illinois: AOCS Press, 182–203. parison with hepatic vitamin A concentration in Ho CC, de Moura FF, Kim S-H et al. (2007). References 205

Excentral cleavage of β-carotene in vivo in a HIV-positive women and their infants on child healthy man. Am J Clin Nutr 85: 770–7. HIV infection, HIV-free survival, and mortality. Hopkins WG, Huener NPA (1995). Intro- J Inf Dis 193: 860–71. duction to Plant Physiology. New York: Wiley. Humphrey JH, Rice AL (2000). Vitamin A Hooper L, Ashton K, Harvey LJ et al. (2009). supplementation of young infants. Lancet 356: Assessing potential biomarkers of micronutrient 422–4. status by using a systematic review methodology: Humphrey JH, West KP Jr, Sommer A (1992). methods. Am J Clin Nutr 89: 1951S–1959S. Vitamin A deficiency and attributable mortality Hoyos B, Hammerling U (2005). The es- among under-5-year-olds. Bull Wld Hlth Org 70: sential role of vitamin A in signal transduction. 225–32. In: Carotenoids and retinoids: molecular aspects Hussain A, Kvale G, Odland M (1995). and health issues. Packer L, Obermuller-Jevic Diagnosis of night blindness and serum vitamin A U, Kraemer K, Sies H, eds. Champaign, Illinois: level: a population-based study. Bull Wld Hlth AOCS Press, 42–60. Org 73: 469–76. Hu J, Liu YE, Wu CF (2009). Long-term ef- Hussey GD, Klein M (1990). A randomized, ficacy and safety of all-trans retinoic acid/arsenic controlled trial of vitamin A in children with trioxide-based therapy in newly diagnosed acute severe measles. N Eng J Med 323: 160–4. polymyelocytic leukaemia. Proc Natl Acad Sci Hyder SMZ, Zlotkin SH (2004). Micronutri- 106: 3342–7. ent Sprinkles: a home fortification strategy for Huang C, Tang Y, Chen C et al. (2000). The combating iron and other micronutrient deficien- bioavailability of β-carotene in stir- or deep-fried cies in infants and young children. Sight and Life vegetables in men determined by measuring the Newsletter 2: 23–5. serum response to a single ingestion. J Nutr 130: 530–40. Ingenbleek Y, Young V (1994). Transthyretin Hudelson P, Dzikunu H, Mensah JD et al. (prealbumin) in health and disease: nutritional (1999). Dietary patterns in a rural area of Ghana implications. Ann Rev Nutr 14: 495–533. and their relevance for vitamin A consumption. Institute of Nutrition and Food Science, Ban- Ecol Food Nutr 38: 183–207. gladesh (1990). Evaluation Report on Worldview Hume EM, Krebs HA (1949). Vitamin A International Foundation (WIF) Nutritional Blind- requirement of human adults: an experimental ness Prevention Programme, Dinajpur, University study of vitamin A deprivation in man. Spec Rep of Dhaka, Bangladesh, mimeo. Ser Med Res Coun no 264. London: HMSO. International Conference on Nutrition (1992). Humphrey JH (2000). Micronutrients in IVACG (1989). Guidelines for the develop- malaria and HIV. Sight and Life Newsletter 3: ment of a simplified dietary assessment to identify 3–8. groups at risk for inadequate intake of vitamin A. Humphrey JH, Agoestina T, Juliana A et al. Washington DC: IVACG. (1998). Neonatal vitamin A supplementation: IVACG (1996). Policy statement on cluster- effect on development and growth at 3 y of age. ing of xerophthalmia and vitamin A deficiency Am J Clin Nutr 68: 109–17. within communities and families. Washington Humphrey JH, Agoestina T, Wu L et al. (1996). DC: IVACG. Impact of neonatal vitamin A supplementation IVACG (1998). Vitamin A and iron inter- on infant mortality and morbidity. J Pediatr 128: actions. Washington DC: IVACG. 489–96. IVACG (1998). Safe doses of vitamin A Humphrey JH, Iliff PJ, Marinda ET et al. during pregnancy and lactation. Washington DC: (2006). Effects of a single large dose of vita- ILSI. min A, given during the postpartum period to 206 References

Jalal F, Nesheim MC, Agus Z et al. (1998). EPI programmme in Bangladesh: an approach Serum retinol concentrations in children are af- to increase coverage of vitamin A administration fected by food sources of β-carotene, fat intake, among infants. XVII IVACG Meeting Report, and anthelminthic drug treatment. Am J Clin Nutr Guatemala City. Washington DC: The Nutrition 68: 623–9. Foundation, 85. Jandacek RJ (2000). The canary in the cell: Karyadi E, West CE, Schultink W et al. a sentinel role for beta-carotene. J Nutr 130(3): (1997). A double-blind, placebo-controlled study 648–51. of vitamin A and zinc supplementation in persons Jeyabalan A and Caritis SN (2006). Antioxi- with tuberculosis in Indonesia: effects on clinical dants and the prevention of preeclampsia – unre- response and nutritional status and healthy “nor- solved issues. N Eng J Med 354: 1841–3. mals”. J Pediatr 131: 925–7. Jiang T, Christian P, Khatry SK et al. (2005). Karyadi E, West CE, Schultink W et al. Micronutrient deficiencies in early pregnancy are (2002). A double-blind, placebo-controlled study common, concurrent, and vary by season among of vitamin A and zinc supplementation in persons rural Nepali pregnant women. J Nutr 135: 1106– with tuberculosis in Indonesia: effects on clinical 12. response and nutritional status. Am J Clin Nutr Jiang T, Xu P, Christian P, et al. (2005). In- 75: 720–7. fection and inflammation may influence multiple Katz J, Christian P, Dominici F et al. 2006. micronutrient status in pregnant females: findings Treatment effects of maternal micronutrient from Nepal. IVACG XXII Meeting T4: 45. supplementation vary by percentiles of the birth Jingxiong J, Toschke AM, von Kries R et al. weight distribution in rural Nepal. J Nutr 136: (2006). Vitamin A status among children in Chi- 1389–94. na. Publ Hlth Nutr 9: 955–60. Katz J, Khatry SK, West KP Jr et al. (1995). Johnson EJ (2005). Obesity, lutein metabo- Night blindness during pregnancy and lactation in lism, and age-related macular degeneration: a rural Nepal. J Nutr 125: 2122–7. web of connections. Nutr Revs 63: 9–15. Katz J, Zeger SL, West KP Jr et al. (1993). Joint FAO/WHO Expert Group (1976). Vita- Clustering of xerophthalmia within households min A deficiency and xerophthalmia. Tech Rep and villages. Int J Epidem 22: 709–15. Ser no 590. Geneva: WHO. Kawaguchi R, Yu J, Honda J et al. (2007). A membrane receptor for retinol binding protein Kabagambe EK, Furtado J, Baylin A et al. mediates cellular uptake of Vitamin A. Science (2005). Some dietary and adipose tissue caro- 315: 820–5. tenoids are associated with the risk of nonfatal Kawaguchi R, Yu J, Wiita P et al. (2008). An myocardial infarction in Costa Rica. J Nutr 135: essential ligand-binding domain in the membrane 1763–69. receptor for retinol-binding protein revealed by Kanawati AA, McLaren DS (1973). Failure large-scale mutagenesis and a human polymor- to thrive in Lebanon II. An investigation of the phism. J Biol Chem 283: 15160–8. causes. Acta Paediat Scand 62: 571–6. Kawakami Y, Raya A, Raya RM et al. (2005). Kapil U, Goindi G (2004). Status of VADD Retinoic acid signalling links left-right asymmetric and strategy for their prevention and control in patterning and bilaterally symmetric somito-gene- India. Sight and Life Newsletter 2: 13–6. sis in the zebrafish embryo. Nature: 435: 165–8. Kapil U, Prakash R (2004). Cost of syrup ver- Kelleher SL, Lonnerdal B (2005). Low sus capsule form of vitamin A. Indian Pediatr 41: vitamin A intake affects milk iron level and iron 864–6. transport in rat mammary gland. J Nutr 135: Karim R, Shahjahan M, Begum S et al. (1996). 27–32. Integration of vitamin A supplementation with Kennedy CM, Kuhn C, Stein Z (2000). Vita- References 207

min A and HIV infection: disease progression, mor- vival in Ghana (ObaapaVitA): a cluster-random- tality and transmission. Nutr Revs 58: 291–303. ised, placebo-controlled trial. Lancet 8; 375 Khachik F (2005). Chemical and metabolic (9726): 1640–9. oxidation of carotenoids. In: Carotenoids and Kjolhede CL, Stallings RY, Dibley MJ et al. retinoids: molecular aspects and health issues. (1995) Serum retinol levels among preschool Packer L, Obermuller-Jevic U, Kraemer K, Sies children in central Java: demographic and socio- H, eds. Champaign, Illinois: AOCS Press, 61–75. economic determinants. Int J Epidemiol 24: 399– Khan MU, Haque E, Khan MR (1984). Nutri- 405. tional ocular diseases and their association with Klemm RD (2010). Opinion 1 Newborn vita- diarrhoea in Matlab, Bangladesh. Brit J Nutr 52: min A supplementation – conflicting results or 1–9. explanatory contextual factors? Sight and Life Khatun UHF, Malek MA, Black RE et al. Magazine 1: 27-28. (2001). A randomized controlled clinical trial of Klemm RD, Labrique AB, Christian P et al. zinc,vitamin A or both in undernourished children (2008). Newborn vitamin A supplementation with persistent diarrhoea in Bangladesh. Acta reduced infant mortality in rural Bangladesh. Paediatr 90: 376–80. Pediatrics 122(1): 180–1. Khoi HH, Khan NC, Thang HV et al. (1996). Kolsteren P, Rahman SR, Hildebrand K et al. Progress of vitamin A deficiency control pro- (1999). Treatment of iron deficiency anaemia with gramme in Vietnam. XVII IVACG Meeting Re- a combined supplementation of iron, vitamin A port, Guatemala City. Washington DC: The Nutri- and zinc in women of Dinajpur, Bangladesh. Eur tion Foundation, 90. J Clin Nutr 53: 102–6. Kim Y, Chongviriyaphan N, Liu C et al. (2006). Kraemer K, Waelti M, de Pee S et al. (2008). Combined antioxidant (β-carotene, α-tocopherol Are low tolerable upper levels for vitamin A and ascorbic acid) supplementation increases undermining effective food fortification efforts? the levels of lung retinoic acid and inhibits the Nutr Rev 66: 517–25. activation of mitogen-activated protein kinase Kramer TR, Udomkesmalee E, Dhanamitta S in the ferret lung cancer model. Carcinogenesis et al. (1993). Lymphocyte responsiveness of chil- 27:1410-9. dren supplemented with vitamin A and zinc. Am King M (1990). Health is a sustainable state. J Clin Nutr 58: 566–70. Lancet 336: 664–7. Krause VM, Delisle H, Solomons NW (1998). King M, Elliott C (1993). Legitimate double- Fortified foods contribute one half of recom- think. Lancet 341: 669–72. mended vitamin A intake in poor urban Guatema- Kirkwood B (1996). Epidemiology of inter- lan toddlers. J Nutr 128: 860–4. ventions to improve vitamin A status in order to Krinsky NI (2005). The biological activity reduce child mortality and morbidity. In: Beyond of carotenoid metabolites. Sight and Life News- nutritional recommendations: implementing sci- letter 2: 10–15. ence for healthy populations. Garza C, Haas JD, Kristensen I, Aaby P, Jensen H et al. (2000). Habicht J-P, Pellettier DL, eds. New York: Cor- Routine vaccinations and child survival: follow- nell University. up study in Guinea-Bissau, West Africa. Brit Med Kirkwood BR, Ross DA, Arthur P et al. J 321: 1435–8. (1996). Effect of vitamin A supplementation on Kritchevsky D (2000). Impact of red palm oil the growth of young children in northern Ghana. on human nutrition and health. Food Nutr Bull Am J Clin Nutr 63: 773–81. 21: 182–8. Kirkwood BR, Hurt L, Amenga-Etego S et al. (2010). Effect of vitamin A supplementation Landrum JT, Bone RA (2001). Lutein, zeax- in women of reproductive age on maternal sur- anthin, and the macular pigment. Arch Biochem 208 References

Biophys 385: 28–40. Health Study. Osteopor Int 15: 552–9. La Rue AC, Argraves WS, Zile MH et al. Lima PF, Oliveira MAL, Goncalves PBD et (2004). Critical role for retinol in the generation/ al. (2004). Effects of retinol on the in vitro de- differentation of angioblasts required for em- velopment of Bos indicus embryos to blastocysts bryonic blood vessel formation. Dev Dyn 230: in two different culture systems. Reprod Domest 666–74. Anim 39: 356–60. Layrisse M, Garcia-Casal MN, Solano L et al. Lindqvist A, Sharvill J, Sharvill D et al. (2000). New property of vitamin A and β-carotene (2007). Loss-of-function mutation in carotenoid in human iron absorption: effect on phytate and 15,15̓-monooxygenase identified in a patient polyphenols as indicators of iron absorption. Arch with hypercarotenemia and hypovitmainosis A. J Latinoam Nutr 50: 243–8. Nutr. 137: 2346–50. Leal JY, Castejon HV, Romero T et al. (2004). Liu C-S, Glahn RP, Liu R (2004). Assessment Serum values of cytokines in children with vita- of carotenoid bioavailability of whole foods us- min A deficiency disorders. Invest Clin 45: 243– ing a Caco-2 cell culture model coupled with an 56. in vitro digestion. J Agri Food Chem 52: 4330–7. Lemke SL, Dueker SR, Follet JR et al. (2003). Lloyd-Puryear M, Humphrey JH, West KP Jr Absorption and retinol equivalence of β-carotene (1989). Vitamin A deficiency and anemia among in humans is influenced by dietary vitamin A in- Micronesian children. Nutr Res 9: 1007–16. take. J Lipid Res 44: 1591–600. Lloyd-Puryear MA, Mahoney J, Humphrey Leung WC, Hessel S, Méplan C et al. (2009). JH et al. (1991). Vitamin A deficiency in Micro- Two common single nucleotide polymorphisms in nesia: a statewide survey in Chuuk. Nutr Revs 11: the gene encoding β-carotene 15.15’-monooxy- 1101–10. genase alter β-carotene metabolism in female Lodish H, Baltimore D, Berk A et al. (1995). volunteers. FASEB J. 23: 1041–1053. An overview of photosynthesis. In: Molecular Levine HM, Pollitt, Galloway R et al. (1993). cell biology, 3rd ed. New York: WH Freeman & Poverty and child development relevance of re- Co Ltd. search in developing countries to the United Loevinsohn BP, Sutter RW, Costales MO States. Child Devel 65: 283–95. (1997). Using cost-effectiveness analysis to Lewallen S, Courtright P (1995). Peripheral evaluate targeting strategies: the case of vitamin A corneal ulcers associated with use of African supplementation. Hlth Pol Plan 12: 30–7. traditional eye medicines. Br J Ophthalmol 79: Long KZ, Montoya Y, Hertzmark E et al. 343–6. (2006). A double-blind, randomized, clinical trial Lewallen S, Taylor TE, Molyneux ME et al. of the effect of vitamin A and zinc supplementa- (1998). Association between measures of vita- tion on diarrheal disease and respiratory tract in- min A and the ocular fundus findings in cerebral fections in children in Mexico City, Mexico. Am malaria. Arch Ophthalmol 116: 293–6. J Clin Nutr 83: 693–700. Li E, Norris AW (1996). Structure/function Long KZ, Santos JI, Garcia TE et al. (2006). of cytoplasmic vitamin A-binding proteins. Annu Vitamin A supplementation reduces the monocyte Rev Nutr 16: 205–34. chemoattractant protein-1 intestinal immune re- Lietman TM, Dhital SP, Dean D (1998). Con- sponse of Mexican children. J Nutr 136: 2600–5. junctival impression cytology for vitamin A defi- Lopez AD, Mathers CD, Ezzati M et al. ciency in the presence of infectious trachoma. Br (2006). The global and regional burden of disease J Ophthalmol 82: 1139–42. and risk factors, 2001: systematic analysis of pop- Lim LS, Harnack LJ, Lazovich D et al. (2004). ulation health data. Lancet 367: 1747–57. Vitamin A intake and the risk of hip fracture in Luzeau R, Carlier C, Ellrodt A et al. (1988). postmenopausal women: the Iowa Women’s Impression cytology with transfer: an easy References 209

method for detection of vitamin A deficiency. Int (2004). Vitamin A deficiency and inflammatory J Vitam Nutr Res 58: 166–70. markers among preschool children in the Repub- Lyons GH, Stangoulis JCR, Graham RD lic of the Marshall Islands. Nutr 20: 1471–81. (2004). Exploiting micronutrient interaction to Mares JA, LaRowe TL, Snodderly DM et al. optimize biofortification programs: the case for (2006). Predictors of optical density of lutein and inclusion of selenium and iodine in the Harvest zeaxanthin in retinas of older women in the Caro- Plus Program. Nutr Revs 62: 247–52. tenoids in Age-Related Eye Disease Study, an ancillary study of the Women’s Health Initiative. Ma AG, Schouten EG, Zhang FZ et al. (2008). Am J Clin Nutr 84: 1107–22. Retinol and riboflavin supplementation decreases Margulis L, Sagan D (1995). What is life? the prevalence of anemia in Chinese pregnant London: Weidenfeld and Nicolson, 69. women taking iron and folic acid supplements. J Marneros AG, Keene DR, Hansen U et al. Nutr 138: 1946–50. (2004). Collagen XVIII/endostatin is essential Maden M (2004). Retinoids in lung devel- for vision and retinal pigment epithelial function. opment and regeneration. Top Devel Biol 61: EMBO J 23: 8999. 153–89. Martini E, Rice A, de Pee S et al. (2005). Low Mahanlabis D, Lahiri M, Paul D et al. (2004). vitamin A capsule coverage rates among post- Randomized, double-blind, placebo-controlled partum females in Central Java, Indonesia. What clinical trial of the efficacy of treatment with zinc are the limiting factors? IVACG XXII Meeting, or vitamin A in infants and young children with W8: 63. severe acute lower respiratory infection. Am J Mason JB, Musgrave P, Habicht J-P (2003). Clin Nutr 79: 430–6. At least one third of poor countries’ disease bur- Mahanlabis D, Wahed MA, Khaled MA den is due to malnutrition. DCIP Working Paper (2002). Zinc supplementation as adjuvant therapy no 1, Bethesda, M: Fogerty Int Center NIH. in children with measles accompanied by pneu- Mason J, Rivers J, Helwig C (2005). Recent monia: a double-blind, randomized controlled trends in malnutrition in developing regions: vita- trial. Am J Clin Nutr 76: 604–7. min A deficiency, anemia, iodine deficiency, and Mahapatra S, Manorama R (1997). The pro- child underweight. Food Nutr Bull 26: 57–162. tective effect of red palm oil in comparison with Massaro GD, Massaro D (1997). Retinoic massive vitamin A dose in combating vitamin A acid treatment abrogates elastase-induced pulmo- deficiency in Orissa, India. Asia Pac J Clin Nutr nary emphysema in rats. Nature (Med) 3: 675–7. 6: 246–50. Mastroiacovo P, Mazzone T, Addis A et al. Malaba IC, Iliff PJ, Nathoo KJ et al. (2005). (1999). High vitamin A intake in early pregnancy Effect of postpartum maternal or neonatal vita- and major malformations: a multicenter prospec- min A supplementation on infant mortality tive controlled study. Teratology 59: 7–11. among infants born to HIV-negative mothers in Mathies RA, Lin SW, Amed JB et al. (1991). Zimbabwe. Am J Clin Nutr 81: 454–60. From femtoseconds to biology: mechanism of Malyavin A, Bouphany V, Arouny A et al. bacteriorhodopsin‘s light-driven proton pump. (1996). National vitamin A survey in Laos PDR. Annu Rev Biophys Biophys Chem 20: 491–518. XVII IVACG Meeting Report, Guatemala City. Matsuo T, Matsuo N, Shiraga F et al. (1988). Washington DC: The Nutrition Foundation, 89. Keratomalacia in a child with familial hypo- Mannar V, Court A (2008). Ensuring the retinol-binding proteinaemia. Jap J Ophthal 32: quality of vitamin A capsules used in supple- 249–54. mentation programs. Am J Clin Nutr 87: 1539- Mayo-Wilson E, Imdad A, Herzer K et al 1540. (2011). Vitamin A supplements for preventing Maqsood M, Dancheck B, Gamble MV et al. mortality, illness, and blindness in children aged 210 References

under 5: systematic review and meta-analysis. (1965). Xerophthalmia in Jordan. Am J Clin Nutr BMJ 343: d5094. 17: 117–30. McCollum EV, Davis M (1915). The essential McLaren DS, Zekian B (1971). Failure of en- factors in the diet during growth. J Biol Chem 23: zyme cleavage of β-carotene: the cause of vita- 231–54. min A deficiency in a child. Am J Dis Child 121: McDevitt TM, Tchao R, Harrison EH et al. 278–80. (2005). Carotenoids normally present in serum Mejia LA, Arroyave G (1982). The effect of inhibit proliferation and induce differentiation of vitamin A fortification of sugar on iron metabo- a human monocyte/macrophage cell line (U937). lism in preschool children in Guatemala. Am J J Nutr 135: 160–4. Clin Nutr 36: 87–93. McLaren DS (1956). A study of the factors Mele L, West KP Jr, Kusdiono et al. (1991). underlying the special incidence of keratomalacia Nutritional and household risk factors for xe- in Oriya children in the Phulbani and Ganjam dis- rophthalmia in Aceh, Indonesia: a case-control tricts of Orissa, India. J Trop Pediat 2: 135–40. study. Am J Clin Nutr 53: 1460–5. McLaren DS (1959). Influence of protein Melikian G, Mmiro F, Ndugwa C et al. deficiency and sex on the development of ocular (2001). Relation of vitamin A and carotenoid sta- lesions and survival time of the vitamin A-defi- tus to growth failure and mortality among Ugan- cient rat. Br J Ophthal 43: 234–41. dan infants with human immunodeficiency virus. McLaren DS (1962). Malnutrition and the Nutr 17: 567–72. eye. New York: Academic Press, 172–9, 213–4. Michaelsson K, Lithell H, Vessby B et al. McLaren DS (1969). Preparations of vita- (2003). Serum retinol levels and risk of fracture. min A. BMJ 1: 782–783. N Eng J Med 348: 287–94. McLaren DS (1980). Nutritional ophthalmol- Miller MF, Stoltzfus RJ, Iliff PJ et al. (2005). ogy. London: Academic Press. A randomized trial of multivitamin supplements McLaren DS (1987). Age-specific xerophthal- and HIV disease progression and mortality. mia rates among displaced Ethiopians. Arch Dis IVACG XXII Meeting T58 59. Child 62: 539–40. Miller MF, Stoltzfus RJ, Iliff PJ et al. (2006). McLaren DS (1991). Vitamin A supplementa- Effect of maternal and neonatal vitamin A supple- tion and mortality. Lancet 338: 1208–9. mentation and other postnatal factors on anemia McLaren DS (1999). Towards the conquest of in Zimbabwean infants: a prospective, random- Vitamin A Deficiency Disorders (VADD). Basel: ized study. Am J Clin Nutr 84: 212–22. Sight and Life. Ministry of Agriculture, Fisheries and Food McLaren DS (2000). Vitamin A and retinitis UK (1995). Manual of Nutrition. Reference book pigmentosa. Sight and Life Newsletter 2: 11–3. no 342, 10th edition. London: HMSO 53. McLaren DS (2004). Measles blindness. Mitra AK, Alvarez JO, Stephensen CB (1998). Sight and Life Newsletter 3: 58. Increased urinary retinol loss in children with McLaren DS (2004). The control of xe- severe infections. Lancet 351: 1033–4. rophthalmia: a century of contributions and les- Moore T (1957). Vitamin A. Amsterdam: sons. Basel: Sight and Life. Elsevier. McLaren DS (2008). Multiple micronutrients Mori M (1904). Ueber den sogenannten Hikan and “hidden hunger”. Sight and Life Magazine 3: (Xerosis conjunctivae infantum ev. Kerato- 6–12. malacie). Jb Kinderheilkd 59: 175–95. McLaren DS (2010). Reminiscences of vita- Moriaty-Craig SE, Ramakrishnan U, Neufeld min A deficiency disorders – a follow up letter to L et al. (2004). Multivitamin-mineral supplemen- Al Sommer’s guest editorial. 1: 63–4 tation is not as efficacious as iron supplementa- McLaren DS, Shirajian E, Tchalian M et al. tion in improving hemoglobin concentration of References 211

nonpregnant anemic women living in Mexico. Nahoum V, Perez E, Germain P et al. (2007). Am J Clin Nutr 80: 1308–11. Modulators of the structural dynamics of the Moshell AN, Bjornson L (1977). Photopro- retinoid X receptor to reveal receptor function. tection in erythropoietic porphyria: mechanism of Proc Natl Acad Sci 104: 17323–8. photoprotection by beta-carotene. J Invest Derm Napoli JL (1994). Retinoic acid homeostasis: 68: 157–60. prospective roles of β-carotene, retinol, CRBP, Mostad SB, Overbaugh J, DeVange DM et and CRABP. In: Vitamin A in health and dis- al. (1997). Hormonal contraception, vitamin A ease. Blomhoff R, ed. New York: Marcel Dekker: deficiency, and other risk factors for shedding of 135–88. HIV-1 infected cells from the cervix and vagina. Napoli JL (2000). A gene knockout corrobo- Lancet 350: 922–7. rates the integral function of Cellular Retinol- Mucida D, Park Y, Kim G et al. (2007). Recip- Binding Protein in retinoid metabolism. Nutr rocal TH17 and regulatory T cell differentiation Revs 58: 230–4l. mediated by retinoic acid. Science 317: 256–60. Naqvi S, Zhu C, Farre G et al. (2009). Trans- Mugusi FM, Rusizoka O, Habib N et al. genetic multivitamin corn through biofortification (2003). Vitamin A status of patients presenting of endosperm with three vitamins representing with pulmonary tuberculosis and asymptomatic three distinct metabolic pathways. Proc Natl Acad HIV-infected individuals. Int J Tuberc Lung 7: Sci 106: 7762–7767. 804–7. Nau H, Chahoud I, Dencker L et al. (1994): Muhilal, Permaesih D, Idjradinata YR et al. Teratogenicity of vitamin A and retinoids. In: (1988). Vitamin A-fortified monosodium glu- Vitamin A in health and disease. Blomhoff R, ed. tamate and health, growth, and survival of chil- New York: Marcel Dekker: 615–664. dren: a controlled field trial. Am J Clin Nutr 48: Newton S, Owusu-Agyei S, Kirkwood B 1271–6. (2007). Is there any monitoring of the quality Mullany LC, Katz J, Li YM et al. (2007). of vitamin A capsules used in supplementation Breast-feeding patterns, time to initiation, and programs? Am J Clin Nutr 86: 1254. mortality risk among newborns in southern Nielsen J, Benn CS, Balé C et al. (2005). Nepal. J Nutr 138: 599-603. Vitamin A supplementation during war-emergency Muñoz EC, Rosado JL, López P et al. (2000). in Guinea-Bissau 1998–1999. Acta Tropica 39: Iron and zinc supplementation improves indica- 275–282. tors of vitamin A status of Mexican preschoolers. Northrop-Clewes CA, Paracha PI, McLoone Am J Clin Nutr 71: 789–94. UJ, Thurnham DI (1996). Effect of improved Muslimatun MK, Schultink W, West CE et vitamin A status on response to iron supplemen- al. (2001). Weekly supplementation with iron and tation in Pakistani infants. Am J Clin Nutr 64: vitamin A during pregnancy increases hemoglo- 694–9. bin concentration but decreases serum ferritin Northrop-Clewes CA, Mwela C, Kankasa C concentration in Indonesian pregnant women. J et al. (2005). Impact of C-reactive protein on Nutr 131: 85–90. retinol concentrations of Zambian children (T40). International Vitamin A Consultative Group Nagendran B, Unnithan UR, Choo YM et al. Meeting, Washington 87 Peru 2004. (2000). Characteristics of red palm oil, a caro- Norum KR (1994). Retinoids and acute tene- and vitamin E-rich refined oil for food uses. myeloid leukemia. In: Vitamin A in Health and Food Nutr Bull 21: 189–94. Disease (ed R Blumhoff) pp 485-501, New York: Nagy NE, Holven KB, Roos N et al. (1997) Marcel Dekker. Storage of vitamin A in extrahepatic stellate cells in normal rats. J Lipid Res 38: 645–58. Obermüller-Jevic UC, Olano-Martin E, 212 References

Corbacho AM et al. (2003). Lycopene in- Ong DE, Kakkad B, MacDonald PN (1987). hibits the growth of normal human prostate Acyl-CoA-independent esterification of retinol epithelial cells in vitro. J Nutr 133(11): 3356–60. bound to cellular retinol-binding protein, type O’Byrne SM, Blaner WS (2005). Introduc- two, by microsomes from rat intestine. J Biol tion to retinoids. In: Carotenoids and retin- Chem 262: 2729–36. oids: molecular aspects and health issues. Packer Oomen HAPC (1961). An outline of xe- L, Obermuller-Jevic U, Kraemer K, Sies H, eds. rophthalmia. Int Rev Trop Med 1: 131–213. Champaign, Illinois: AOCS Press, 1–22. Oomen HAPC (1969). Clinical epidemiol- O’Byrne SM, Wongsiriroj N, Libien J et al. ogy of xerophthalmia in man. Am J Clin Nutr 22: (2005). Retinoid absorption and storage is im- 1098–105. paired in mice lacking lecithin:retinol acyltrans- Oomen HAPC, McLaren DS, Escapini H ferase (LRAT). J Biol Chem 280: 35647–57. (1964). Epidemiology and public health aspects Olmedilla B, Granado F, Blanco I et al. of hypovitaminosis A. A global survey on xe- (1994). Seasonal and sex-related variations in six rophthalmia. Trop Geogr Med 16: 271–315. serum carotenoids, retinol and alpha-tocopherol. Opotowsky AR, Bilezikian JP (2004). Se- Am J Clin Nutr 60: 106–10. rum vitamin A concentration and the risk of hip Olsen SF (1999). Effect of vitamin A and fracture among women 50 to 74 years old in the β-carotene supplementation on women’s health. United States: a prospective analysis of the BMJ 318: 551–2. NHANES 1 follow-up study. Am J Med 117: Olson JA (1994). Vitamin A, retinoids, and 169–74. carotenoids. In: Modern nutrition in health and Orfanos CE, Braun-Falco O, Farber EM et al., disease, 8th edition. Shils ME, Olson JA, Shike eds (1981). Retinoids: advances in basic research M, eds. Philadelphia: Lea & Febiger, 287–307. and therapy. Berlin: Springer-Verlag. Olson JA (1996). Biochemistry of vitamin A Osrin D, Vaidya A, Shrestha Y et al. (2005). and carotenoids. In: Vitamin A deficiency: health, Effects of antenatal multiple micronutrient sup- survival, and vision. Sommer A, West KP Jr. eds. plementation on birth weight and gestational New York: Oxford University Press, 221–50. duration in Nepal: double-blind, randomized con- Olson JA (1997). Isotope-dilution techniques: trolled trial. Lancet 365: 955–62. a wave of the future in human nutrition. Am J Owsley C, McGwin G, Jackson GR et al. Clin Nutr 66: 186–7. (2006). Effect of short-term, high-dose retinol Olson JA (1999). Carotenoids. In: Modern on dark adaptation in aging and early age- nutrition in health and disease, 9th edition ME related maculopathy. Invest Ophthalmol Vis Sci Shils, Olson JA, Shike M, Ross AC, eds. Balti- 47: 1310–8. more: Williams & Wilkins, 525–41. Omenn GS, Goodman GE, Thornquist MD Paine JA, Shipton CA, Chaggar S et al. et al. (1996). Risk factors for lung cancer and for (2005). Improving the nutritional value of Golden intervention effects in CARET, the Beta-Carotene Rice through increased pro-vitamin A content. and Retinol Efficacy Trial. J Natl Cancer Inst Nature Biotech 23: 482–7. 88(21): 1550–9. Palmer AC, Stewart CP (2009). Micronutri- Omenn GS, Goodman GE, Thornquist MD ents, health and development: evidence-based et al. (1996). Effects of a combination of beta programs. Sight and Life Magazine 2: 25-33. carotene and vitamin A on lung cancer and cardio- Panel on Dietary Reference Values (1991). vascular disease. N Engl J Med 334(18): 1150–5. Dietary references values for food energy and Ong ASH, Tee ES (1992). Natural sources of nutrients for the United Kingdom. London: carotenoids from plants and oils. Methods Enzy- HMSO. mol 213: 142–67. Panganbaribuyan R, Scherbaum V, Erhardt References 213

JG et al. (2004). Socioeconomic and familial in child survival are strongly associated with characteristics influence caretakers’ adherence to changes in malnutrition in developing countries. the periodic vitamin A capsule supplementation J Nutr 133: 107–19. program in Central Java, Indonesia. J Trop Pe- Pelletier DL, Frongillo EA Jr, Schroeder DG, diatr 50: 143–8. Habicht JP. (1995). The effects of malnutrition Panozzo G, Babighian S, Bonora A (1998). on child mortality in developing countries. Bull Association of xerophthalmia, flecked retina, and World Health Organ 73(4): 443–8. pseudotumor cerbri caused by hypovitaminosis Penniston KL, Tanumihardjo SA (2006). The A. Am J Ophthalmol 125: 708–10. acute and chronic toxic effects of vitamin A. Am Pant CR, Pokharel GP, Curtale F et al. (1996). J Clin Nutr 83: 191–201. Impact of nutrition education and mega-dose vi- Pepperberg DR, Crouch RK (2001). An il- tamin A supplementation on the health of children luminating new step in visual-pigment regenera- in Nepal. Bull Wld Hlth Org 74: 533–45. tion. Lancet 358: 2098–9. Parker A (2003). In the blink of an eye: the Petr L, Erdman JW Jr (2005). Lycopene and cause of the most dramatic event in the history of risk of cardiovascular disease. In: Carotenoids life. UK: Free Press. and retinoids: molecular aspects and health is- Parker RS (2000). Methodological consider- sues. Packer L, Obermuller-Jevic U, Kraemer K, ations in determining vitamin A and carotenoid Sies H, eds. Champaign, Illinois: AOCS Press, bioavailability in humans. Food Nutr Bull 21: 204–17. 124–9. Pfahl M, Chytil F (1996). Regulation of me- Pascal AA, Liu Z, Broess K et al. (2005). tabolism by retinoic acid and its nuclear recep- Molecular basis of photoreception and control tors. Annu Rev Nutr 16: 257–83. of photosynthetic light-harvesting. Nature 436: Pfander H (2002). Problems analysing caro- 134–7. tenoids? Sight and Life Newsletter 2: 41–2. Pasutto F, Sticht H, Hammersen G et al. Phillips M, Sanghvi T, Suaréz R et al. (1996). (2007). Mutations in STRA6 cause a broad spec- The costs and effectiveness of three vitamin A trum of malformations including anophthalmia, interventions in Guatemala. Soc Sci med 42: congenital heart defects, diaphragmatic hernia, 1661–8. alveolar capillary dysplasia, lung hypoplasia, and Pilch SM (1987). Analysis of vitamin A data mental retardation. Am J Hum Gen 80: 550–60. from the Health and Nutrition Examination Sur- Paton D, McLaren DS (1960). Bitot spots. veys. J Nutr 117: 636–40. Am J Ophthal 50: 568–74. Pillat A (1929). Does keratomalacia exist in Pattison DJ, Symmons DP, Lunt M et al. adults? Arch Ophthal 2: 256–87. (2005). Dietary beta-cryptoxanthin and inflam- Pineda O (1993). Fortification of sugar with matory polyarthritis: results from a population- vitamin A. Nutriview 2: 6–7. based prospective study. Am J Clin Nutr 82(2): Pokhrel RP, Khatry SK, West KP Jr et al. 451–5. (1994). Sustained reduction in child mortality Paul AA, Southgate DAT (1978) (eds). Mc- with vitamin A in Nepal. Lancet 343: 1368–9. Cance and Widdowson’s The composition of Pollard R, van der Pasch N (1990). World- foods, 4th edition. London: HMSO. view International Foundation, Bangladesh, Pedro MRA, Madriaga VR, Barba CVC et al. Nutritional Blindness Prevention Program: Final (2004). The National Vitamin A Supplementation Evaluation for NOVIB. Program and subclinical vitamin A deficiency Potrykus I (2003). Nutritionally enhanced among preschool children in the Philippines. rice to combat malnutrition disorders of the poor. Food Nutr Bull 25: 319–29. Nutr Revs 61: 101–4. Pelletier DL, Frongillo EA (2003). Changes Prentice AM, Savy M, Darboe MK et al. 214 References

(2009). Commentary: Challenging public health double-blind controlled trial. BMJ 323: 314–8. orthodoxies – prophesy or heresy? Int J Epide- Rahman MM, Wahed MA, Fuchs GJ et al. miol 38(2): 591–593. (2002). Synergistic effect of zinc and vitamin A Prior RL (2003). Fruit and vegetables in the on the biochemical indexes of vitamin A nutrition prevention of cellular oxidative damage. Am J in children. Am J Clin Nutr 75: 92–8. Clin Nutr 78: 570S–8S. Rahmathullah L, Raj P, Darling K et al. Puffer RR, Serrano CV (1973). Patterns of (1994). Changing trends in vitamin A deficiency mortality in childhood. PAHO Scientific Publica- in children below 5 years of age. Xero Bull no tion no 262. Washington DC: PAHO. 56: 1–3. Pytlik M, Cegiela U, Folwarczna J et al. Rahmathullah L, Tielsch JM, Thulasiraj RD (2004). Vitamin A intake and the risk of hip frac- et al. (2003). Impact of supplementing newborn ture in postmenopausal women. Osteoporosis Int infants with vitamin A on early mortality: com- 15: 552–9. munity-based randomized trial in southern India. BMJ 327: 254–7. Quadro L, Hamberger L, Gottesmann ME et Rahmathullah L, Underwood BA, Thulasiraj al. (2005). Pathways of vitamin A delivery to the RD et al. (1990). Reduced mortality among chil- embryo: insights from a new tunable model of dren in Southern India receiving a small weekly embryonic vitamin A deficiency. Endocrinology dose of vitamin A. N Eng J Med 323: 929–35. 146: 4479–90. Ramachandran G, Santha T, Garg R et al. (2004). Vitamin A levels in sputum-positive pul- Radhika MS, Bhaskaram P, Balakrishna N et monary tuberculosis patients in comparison with al. (2002). Effects of vitamin A deficiency dur- household contacts. Inter J Tuberc Lung Dis 8: ing pregnancy on maternal and child health. Int J 1130–3. Obstet Gynaecol 109: 689–93. Ramakrishnan U, Latham MC, Abel R et al. Radhika MS, Bhaskaram P, Balakrishna N et (1995). Vitamin A supplementation and morbid- al. (2003). Red palm oil supplementation: a feasi- ity among preschool children in south India. Am ble diet-based approach to improve the vitamin A J Clin Nutr 61: 1295–303. status of pregnant women and their infants. Food Ramakrishnan U, Latham MC, Abel R et al. Nutr Bull 24: 208–17. (1995). Vitamin A supplementation does not im- Radu RA, Han Y, Bui TV et al. (2005). Re- prove growth of preschool children: a random- ductions in serum vitamin A arrest accumulation ized, double-blind field trial in south India. J Nutr of toxic retinal fluorophores: a potential therapy 125: 202–11. for treatment of lipofuscin-based retinal diseases. Ramakrishnan U, Martorell R, Latham MC et Invest Ophthalmol Vis Sci 46: 4393–401. al. (1999). Dietary vitamin A intakes of preschool- Rahman MM, Mahalanabis D, Islam MA et age children in south India. J Nutr 129: 2021–7. al. (1993). Can infants and young children eat Rao BSN (2000). Potential use of red palm oil enough green leafy vegetables from a single tradi- in combating vitamin A deficiency in India. Food tional meal to meet their daily vitamin A require- Nutr Bull 21: 202–11. ments? Eur J Clin Nutr 47: 68–72. Rask L, Valtersson C, Arundi H et al. (1983). Rahman MM, Tofail F, Wahed MA et al. Sub-cellular localization in normal and vitamin A (2002). Short term supplementation with zinc and deficient rat liver of vitamin A serum transport vitamin A has no significant effect on growth. Am proteins, albumin, ceruloplasmin and class I J Clin Nutr 75: 87–91. major histocompatibility antigens. Exp Cell Res Rahman MM, Vermund SH, Wahed MA et al. 143: 91–102. (2001). Simultaneous zinc and vitamin A supple- Ray SK, Mallik S, Munsi AK et al. (2004). mentation in Bangladeshi children: randomized Epidemiological study of measles in slum areas References 215

of Kolkata. Ind J Pediatr 71: 583–6. fied relative-dose response ratio, and breast-milk Reddy V (2000). Comments on the develop- vitamin A as indicators of response to postpartum ment of high-carotene foods with the use of bio- maternal vitamin A supplementation. Am J Clin technology. Food Nutr Bull 21: 247. Nutr 71: 799–806. Reddy V (2002). Vitamin A programme in Rivera JA, González-Cossio T, Flores M et India – why the controversy? Sight and Life al. (2001). Multiple micronutrient supplementa- Newsletter 3: 55–62. tion increases the growth of Mexican infants. Am Reddy CG, Nohr D, Schaeffer M et al. (2005). J Clin Nutr 74: 657–63. β-carotene conversion into vitamin A in human Rivera JA, Hotz C, González-Cossio T et al. retinal pigment epithelium cells. Invest Ophthal- (2003).The effect of micronutrient deficiences mol Vis Sci 46: 3562–9. on child growth: a review of results from com- Ribaya-Mercado JD, Maramag CC, Tengco munity-based supplementation trials. J Nutr 133: LW et al. (2007). Carotene-rich plant foods in- Suppl 2; S4010–20. gested with minimal dietary fat enhance the to- Rodriguez A, Hamar DH, Rivera J et al. tal-body vitamin A pool size in Filipino school- (2005). Effects of moderate doses of vitamin A children as assessed by stable-isotope-dilution as an adjunct to the treatment of pneumonia in methodology. Am J Clin Nutr 85: 1041–9. underweight and normal-weight children: a ran- Ribaya-Mercado JD, Mazariegos M, Tang domized, double-blind, placebo-controlled trial. G et al. (1999). Assessment of total body stores Am J Clin Nutr 82: 1090–6. of vitamin A in Guatemalan elderly by the deu- Rodriguez-Amaya DB (1999). A guide to terated-retinol-dilution method. Am J Clin Nutr carotenoid analysis of foods. Washington, DC: 69: 278–84. Omni Research. Ribaya-Mercado JD, Solomons NW (2003). Rodriguez-Amaya DB (2002). Brazil: a Deuterated-retinol-dilution approach for longitu- bounty of carotenoid sources. Sight and Life dinal monitoring of health and nutritional inter- Newsletter 4: 3–9. ventions involving vitamin A. Sight and Life Rondo PHC, Abbott R, Tomkins AM (2001). Newsletter 2: 3–4. Vitamin A and neonatal anthropometry. J Trop Ribaya-Mercado JD, Solomons NW, Medrano Ped 47: 307–10. Y et al. (2004). Use of deuterated-retinol-dilution Roodenburg AJC, Leenen R, van het Hof KH technique to monitor the vitamin A status of Nica- et al. (2000). Amount of fat in the diet affects bio- raguan schoolchildren 1 year after the Nicaraguan availability of lutein esters but not of α-carotene, national program of sugar fortification with vita- β-carotene, and vitamin E in humans. Am J Clin min A. Am J Clin Nutr 80: 1291–8. Nutr 71: 1187–93. Ribaya-Mercado JD, Solon FS, Fermin LS et Roos N, Islam MM, Thilsted SH (2003). al. (2004). Dietary vitamin A intakes of Filipino Small indigenous fish species in Bangladesh: con- elders with adequate or low vitamin A concen- tribution to vitamin A, calcium and iron intakes. trations as assessed by the deuterated-retinol- J Nutr 133: Suppl 2; S4021–6. dilution method: implications for dietary require- Rosales FJ, Chau KK, Haskell MH et al. ments. Am J Clin Nutr 79: 633–41. (2002). Determination of a cut-off value for the Rice AL, Stoltzfus RJ, de Francisco A et al. molar ratio of retinol-binding protein to trans- (1999). Maternal vitamin A or β-carotene supple- thyretin (RBP:TTR) in patients with low hepatic mentation in lactating women benefits mothers vitamin A stores. J Nutr 132: 3687–92. and infants but does not prevent subclinical defi- Rosales FJ, Jang J-T, Pinero DJ et al. (2000). ciency. J Nutr 129: 356–65. Iron deficiency alters the distribution of vitamin A Rice AL, Stoltzfus RJ, de Francisco A et al. between plasma and liver, and retinol and retinyl (2000). Evaluation of serum retinol, the modi- esters in young rats. Proc XIX IVACG Meeting, 216 References

Durban, South Africa, 9-11 March. Washington Roth AE, Benn CS, Ravn H et al. (2010). DC: ILSI, 34. Effect of revaccination with BCG in early child- Rosales FJ, Kjolhede C, Goodman C (1996). hood on mortality: randomized trial in Guinea- Efficacy of a single oral dose of 200,000 IU of Bissau. BMJ 340: 749. oil-soluble vitamin A in measles-associated morbidity. Am J Epidemiol 143: 413–22. Samal D, Rojanawatsirivet C, Wernsdorfer G Rosales FJ, Ritter SJ, Zolfaghari R et al. et al. (2005). Synergism of desbutyl benflumetol (1996). The mechanism of inflammation-induced and retinol against Plasmodium falciparum. Wien hyporetinolemia. XVII IVACG Meeting Report, Klin Wochenschr 117, Suppl 4: 39–44. Guatemala City. Washington DC: The Nutrition Sanchez AM, Congdon NG, Sommer A et al. Foundation, 98. (1997). Pupillary threshold as an index of popu- Rosales FJ, Ross AC (1998). A low molar lation vitamin A status among children in India. ratio of retinol binding protein to transthyretin Am J Clin Nutr 65: 61–6. indicates vitamin A deficiency during inflamma- Sankaranarayanan S, Suarez M, Taren G et tion: studies in rats and a posteriori analysis of al. (2005). The concentration of free holo-retinol- vitamin A-supplemented children with measles. J binding protein is higher in vitamin A-sufficient Nutr 128: 1681–7. than in deficient Nepalese women in late preg- Rosales FJ, Topping J, Smith JE et al. (1999). nancy. J Nutr 135: 2817–22. Reduced predictability of acute phase proteins Savy M, Edmond K, Fine PEM et al. (2009). versus RBP or TTR on serum retinol during ma- Landscape analysis of interactions between nu- laria-related morbidity. Proc XIX IVACG Meet- trition and vaccine responses in children. J Nutr ing, Durban, South Africa, 9-11 March. Washing- 139: 2154S-2218S. ton DC: ILSI, 93. Scanziani M, Häusser M (2009). Electro- Rosen D, Haselow N, Sloan NL. (1993). How physiology in the age of light. Nature 461: 930– to use the HKI Food Frequency Method to assess 939. community risk of vitamin A deficiency. New Schalch W (2004). Lutein and zeaxanthin York: Helen Keller International Vitamin A Tech- supplementation improves visual performance. nical Assistance Program. Sight and Life Newsletter 3: 4–7. Rosenbloom J, Kaluski D, and Berry E Schall JI, Zemel BS, Kawchak DA (2004). (2008). A Global Nutritional Index. Food Nutr Vitamin A status, hospitalizations, and other out- Bull 29: 266–275. comes in young children with sickle cell disease. Ross AC (1996). The relationship between J Pediatr 145: 99–106. immunocompetence and vitamin A status. In: Vi- Schaumberg DA, O’Connor J, Semba RD tamin A deficiency: health, survival, and vision. (1996). Risk factors for xerophthalmia in the Re- Sommer A, West KP Jr. New York: Oxford Uni- public of Kiribati. Eur J Clin Nutr 50: 761–4. versity Press, 251–73. Schmidt MK, Muslimatun S, West CE et al. Ross AC (1999). Vitamins and retinoids. In: (2004). Mental and psychomotor development in Modern nutrition in health and disease, 9th edi- Indonesian infants of mothers supplemented with tion. Shils ME, Olson JA, Shike M, Ross AC, eds. vitamin A in addition to iron during pregnancy. Br Baltimore: Williams & Wilkins. J Nutr 91: 279–85. Ross AC (2005). Introduction to vitamin A: a Schuftan C, Ramalingaswami V, Levinson J nutritional and life cycle perspective. In: Carote- (1998). Micronutrient deficiencies and protein- noids and retinoids: molecular aspects and health energy malnutrition. Lancet 351: 1812. issues. Packer L, Obermuller-Jevic U, Kraemer Schwartz SJ (2004). Food matrix effects on K, Sies H, eds. Champaign, Illinois: AOCS Press, carotenoid absorption. Sight and Life Newsletter 23–41. 3: 8–13. References 217

Screde B, Lie SO, Blomhoff R, Norum KR (1995). Infant mortality and maternal vitamin A (1994). Uptake and storage of retinol and retinyl deficiency during human immunodeficiency virus esters in bone marrow of children with acute infection. Clin Inf Dis 21: 966–72. myeloid leukemia treated with high-dose retinyl Semba RD, Miotti PG, Chiphangwi JD et al. palmitate. Eur J Haematol 52(3): 140–4. (1997). Maternal vitamin A deficiency and child Sedgh G, Herrera M, Nestel P et al. (2000). growth failure during human immunodeficiency Dietary vitamin A intake and non-dietary factors virus infection. J Acq Immun Def Syns Hum Ret- are associated with reversal of stunting in chil- rov 14: 219–22. dren. J Nutr 130: 2520–6. Semba RD, Miotti P, Chiphangwi JD et al. Seki T, Isono K, Ozkaki K et al. (1998). The (1998). Maternal vitamin A deficiency and infant metabolic pathway of visual chromophore forma- mortality in Malawi. J Trop Pediatr 44: 232–4. tion in Drosophilia melanogaster. Eur J Biochem Semba RD, Muhilal, Mohgaddam NEG et al. 257: 522–7. (1999). Interpretation of vitamin A supplementa- Semba RD (1998). The role of vitamin A and tion with the Expanded Program on Immunization related retinoids in immune function. Nutr Revs does not affect seroconversion to oral poliovirus 56: (11) S38–48. vaccine in infants. J Nutr 129: 2203–5. Semba RD (1999). Vitamin A as “anti-infec- Semba RD, Muhilal, Ward BJ et al. (1993). tive” therapy, 1920-1940. J Nutr 129: 783–91. Abnormal T-cell subset proportions in vitamin A- Semba RD, Bloem MW, eds (2001). Nutrition deficient children. Lancet 341: 5–8. and health in developing countries. Totowa, NJ: Semba RD, Munasir Z, Beeler J et al. (1995). Humana Press. Reduced seroconversion to measles in infants Semba RD, Bloem MW (2004). Measles given vitamin A with measles vaccination. Lancet blindness. Surv Ophthalmol 49: 243–55. 345: 1330–2. Semba RD, Bulterys M, Munyeshuli V et al. Semba RD, Ndugwa C, Perry T et al. (2005). (1996). Vitamin A deficiency and T-cell subpopu- Effect of periodic vitamin A supplementation on lations in children with meningococcal disease. J mortality and morbidity of human immunodefi- Trop Ped 42: 287–90. ciency virus-infected children in Uganda: a con- Semba RD, Caiaffa WT, Graham NMH et al. trolled clinical trial. Nutr 21: 25–31. (1995). Vitamin A deficiency and wasting as pre- Sempertegui F (2001). Commentary: is vi- dictors of mortality in human immunodeficiency tamin A playing a hidden role in children’s lung virus-infected injection drug users. J Infect Dis function? Int J Epidemiol 30: 465–6. 171: 1196–1202. Shamim AA, Mawla MG, Islam Z et al. Semba RD, Kumwenda N, Taha TE et al. (2003). Palmyra palm – an under-exploited car- (2000). Plasma and breast milk vitamin A as indi- otene-rich source. Sight and Life Newsletter 2: cators of vitamin A status in pregnant women. Int 21–4. J Vitam Nutr Res 70: 271–7. Shankar AH (1999). Using micronutrients Semba RD, Kumwenda N, Taha TE et al. as a public health intervention for malaria. Proc (2001). Impact of vitamin A supplementation XIX IVACG Meeting, Durban South Africa, 9–11 anaemia and plasma erythropoietin concentra- March. Washington DC: ILSI, 73. tions in pregnant women: a controlled clinical Shankar AH (2000). Nutritional modulation trial. Eur J Haematol 66: 389–95. and malaria morbidity and mortality. J Infect Dis Semba RD, Miotti PG, Chiphangwi JD et al. 182 (Suppl 1): 537–55. (1994). Maternal vitamin A deficiency and moth- Shankar AH, Genton B, Semba RD et al. er-to-child transmission of HIV-1. Lancet 343: (1999). Effect of vitamin A supplementation 1593–7. on morbidity due to Plasmodium falciparum Semba RD, Miotti PG, Chiphangwi JD et al. in young children in Papua New Guinea: a ran- 218 References

domised trial. Lancet 354: 203–9. Sivaprasadarao A, Findlay JBC (1994). The Shankar AV, West KP Jr, Gittelsohn J et al. retinol-binding protein superfamily. In: Vitamin A (1996). Chronic low intakes of vitamin A-rich in health and disease. Blomhoff R, ed. New York: foods in households with xerophthalmic children: Marcel Dekker, 87–117. a case-control study in Nepal. Am J Clin Nutr 64: Sliwinski L, Janiec W, Pytlik M et al. (2004). 242–8. Effects of retinol on development of osteopenic Shann F (2010). BCG vaccination in develop- changes induced by bilateral ovariectomy in rats. ing countries. BMJ 340: c809. Pol J Pharmacol 56: 817-24. Shenai JP (1999).Vitamin A supplementation Sloan NL, Rosen D, de la Paz T et al. (1997). in very low birth weight neonates: rationale and Identifying areas with vitamin A deficiency: the evidence. Pediatrics 104: 1369–74. validity of a semiquantitative food frequency Shey WI, Brocklehurst P, Sterne JA (2002). method. Am J Publ Hlth 87: 186–91. Vitamin A supplementation for reducing the risk Smitasiri S, Attig GA, Dhanamitta S (1992). of mother-to-child transmission of HIV infec- Participatory action for nutrition education: social tion. Cochrane Database Syst Rev 2002(3) P: marketing of vitamin A-rich foods in Thailand. CD003648. Ecol Food Nutr 28: 199–210. Shyam T, Kim CM, Retherford RD (2005). Smitasiri S, Attig GA, Valyasevi A et al. Effects of vitamin A supplementation on child (1993). Social marketing of vitamin A-rich foods mortality: evidence from Nepal’s 2001 Demo- in Thailand. Thailand: Institute of Nutrition, Ma- graph and Health Survey. Trop Med Int Health hidol University. 10: 782–9. Smith FR, Goodman DS (1971). The effects Siems W, Wiswedel I, Alija A et al. (2005). of diseases of the liver, thyroid, and kidneys on β-carotene cleavage products impair cellular and the transport of vitamin A in human plasma. J mitochondrial functions and may lead to geno- Clin Invest 50: 2426–36. toxicity. In: Carotenoids and retinoids: molecular Smith JC, Makdani D, Hegar A et al. (1999). aspects and health issues. Packer L, Obermuller- Vitamin A and zinc supplementation of preschool Jevic U, Kraemer K, Sies H, eds. Champaign, Il- children. J Am Coll Nutr 18: 213–22. linois: AOCS Press, 130–48. Smuts CM, Lombard CJ, Benade AJS et al. Simpson KL (1990). Use of food composition (2005). Efficacy of a foodlet-based multiple micro- tables for retinol and provitamin A carotenoid nutrient supplement for preventing growth failure, content. Methods Enzymol 190: 237– 41. anaemia and multinutrient deficiences in infants: Simpson KL, Tsou SCS (1986). Vitamin A and the four country IRIS trial pooled data analysis. J provitamin A composition of foods. In: Bauern- Nutr 135: S631S–8. feind JC, Vitamin A Deficiency and its Control, Soderlund U, Fex GA, Nilsson-Ehle P (2005). pp 461–478. New York: Academic Press. Concentrations of retinoids in early pregnancy Singh V, West KP Jr (2004). Vitamin A de- and in newborns and their mothers. Am J Clin ficiency and xerophthalmia among school-aged- Nutr 81: 633–6. children in Southeastern Asia. Eur J Clin Nutr 58: Solomons NW (1998). Plant sources of vita- 1342–9. min A and human nutrition: red palm oil does the Sinha DP, Bang FB (1973). Seasonal varia- job. Nutr Revs 56: 309–11. tions in signs of vitamin A deficiency in rural Solon FS (1993). Personal communication. West Bengal children. Lancet ii: 228–31. Solon FS (2002). Progress and challenges in Sinha DP, Bang FB (1976). The effect of mas- controlling vitamin A deficiency in the Philip- sive doses of vitamin A on the signs of vitamin A pines. Sight and Life Newsletter 3: 69–74. deficiency in preschool children. Am J Clin Nutr Solon FS, Latham MC, Guirriec R et al. 29: 110–5. (1985). Fortification of MSG with vitamin A: References 219

The Philippines experience. Food Technology 39: Syst Rev 2001; P: CD001479. 71–9. Stahl W, Sies H (1998). The role of carote- Solon FS, Popkin BM, Fernandez TL et al. noids and retinoids in gap junctional communica- (1978). Vitamin A deficiency in the Philippines: a tion. Int J Vit Nutr Res 68: 358–9. study of xerophthalmia in Cebu. Am J Clin Nutr Stahl W, Sies H (2005). Gap junctional in- 31: 360–8. tercellular communication. In: Carotenoids and Sommer A (1982). Nutritional blindness: xe- retinoids: molecular aspects and health issues. rophthalmia and keratomalacia. New York: Ox- Packer L, Obermuller-Jevic U, Kraemer K, Sies ford University Press. H, eds. Champaign, Illinois: AOCS Press, 76–85. Sommer A (1995). Vitamin A deficiency and Stephensen CB (2003). Vitamin A, β-carotene, its consequences. 3rd edition. Geneva: WHO. and mother-to-child transmission of HIV. Nutr Sommer A, Davidson FR (2002). Assessment Revs 61: 280–92. and control of vitamin A deficiency: the Annecy Stephensen CB, Gildengorin G (2000). Serum Accords. J Nutr 132: S2845S–50. retinol, the acute phase response, and the apparent Sommer A, Emran N, Tamba T (1979).Vi- misclassification of vitamin A status in the third tamin A-responsive punctate keratopathy in xe- National Health and Nutrition Examination Sur- rophthalmia. Am J Ophthalmol 87: 330–3. vey, 1988–1994. Am J Clin Nutr 79: 625–32. Sommer A, Hussaini G, Muhilal et al. (1982). Stephensen CB, Jiang X, Freytag T (2004). History of night blindness: a simple tool for xe- Vitamin A deficiency increases the in vivo devel- rophthalmia screening. Am J Clin Nutr 33: 887– opment of IL-10-positive Th2 cells and decreases 91. development of Th1 cells in mice. J Nutr 134: Sommer A, Hussaini G, Tarwotjo I et al. 2660–6. (1983). Increased mortality in children with mild Stewart CP, Christian P, Schulze KJ et al. vitamin A deficiency. Lancet ii: 585–8. (2009). Antenatal micronutrient supplementation Sommer A, Sugana T, Djunaedi E et al. (1978). reduces metabolic syndrome in 6- to 8-year-old Vitamin A-responsive panocular xerophthalmia in children in rural Nepal. J Nutr 139: 1575-81. a healthy adult. Arch Ophthalmol 96: 1630-1634. Stoltzfus RJ, Hakimi M, Miller KW et al. Sommer A, Tarwotjo I, Djunaedi E et al. (1993). High-dose vitamin A supplementation of (1986). Impact of vitamin A supplementation on breast-feeding Indonesian mothers: effects on the childhood mortality. A randomised controlled vitamin A status of mother and infant. J Nutr 123: community trial. Lancet i: 1169–73. 666–75. Sommer A, West KP Jr (1996). Vitamin A de- Stranders SK, Northrop-Clewes CA, Thurn- ficiency: health, survival, and vision. New York: ham DI (2004). Vitamin A capsule distribution Oxford University Press. among infants in rural Bangladesh aged 6-11 Sommerburg O, Langhans C-D, Salerno C months. Sight and Life Newsletter 3: 26–8. et al. (2005). Formation of β-carotene cleavage Stringham JM, Hammond BR Jr (2005). Di- products in view of the particular conditions in in- etary lutein and zeaxanthin: possible effects on flamed lung tissue. In: Carotenoids and retinoids: visual function. Nutr Revs 63: 59–64. molecular aspects and health issues. Packer L, Stryer L (1995). Biochemistry 4th ed. New Obermuller-Jevic U, Kraemer K, Sies H, eds. York: Freeman, 318–20. Champaign, Illinois: AOCS Press, 149–67. Sukwa T, Mwandu D, Kapui A et al. (1988). Sonneveld E, van der Saag PT (1998). Me- The prevalence and distribution of xerophthalmia tabolism of retinoic acid: implications for devel- in pre-school age children of the Luapula Valley, opment and cancer. Int J Vit Res 68: 404–10. Zambia. J Trop Med Hyg 34: 12–5. Souza RM, Souza R (2002). Vitamin A for Sumner P (2002). Color vision: why are we treating measles in children. Cochrane Database primates unique? Eye News 8: 48–60. 220 References

Takyi EEK (1999). Children’s consumption night blindness: the Night Vision Threshold Test of dark green, leafy vegetables with added fat and a portable dark room. Sight and Life News- enhances serum retinol. J Nutr 129: 1549–54. letter 4: 3–5. Tan JSL, Wang JJ, Flood V et al. (2008). Taren D, Duncan B, Shrestha K et al. (2004). Dietary antioxidants and the long-term inci- The night vision threshold test is a better predictor dence of age-related macular degeneration: The of low serum vitamin A concentration than self- Blue Mountains Eye Study. Ophthalmology 115: reported night blindness in pregnant urban Nepali 334–41. women. J Nutr 134: 2573–8. Tang AM, Graham NMH, Kirby AJ et al. Tarwotjo I, Sommer A, West KP Jr et al. (1993). Dietary micronutrient intake and risk of (1987). Influence of participation on mortality in progression to acquired immunodeficiency syn- a randomized trial of vitamin A prophylaxis. Am drome (AIDS) in human immunodeficiency virus J Clin Nutr 45: 1466–71. type 1 (HIV-1)-infected homosexual men. Am J Taylor A, ed. (1999). Nutritional and environ- Epidemiol 138: 937–51. mental influences on the eye. Boca Raton: CRC Tang G, Gu X, Hu S et al. (1999). Green and Press. yellow vegetables can maintain body stores of ten Doesschate J (1968). Causes of blindness vitamin A in Chinese children. Am J Clin Nutr 70: in and around Surabaja, East Java, Indonesia. 1069–76. Thesis, University of Indonesia, Jakarta. Tang G, Qin J, Dolnikowski GG et al. (2005). Teng-Khoen-Hing (1959). Fundus changes Spinach or carrots can supply significant amounts in hypovitaminosis A. Ophthalmologica (Basel) of vitamin A as assessed by feeding with intrinsi- 137: 81–5. cally deuterated vegetables. Am J Clin Nutr 82: Thiel MA, Bossart W, Bernauer W (1997). 821–8. Improved impression cytology techniques for Tang G, Qin J, Dolnikowski GG et al. (2009). the immunopathological diagnosis of superficial Golden Rice is an effective source of vitamin A. virus infection. Br J Ophthalmol 81: 984–8. Am J Clin Nutr 89: 1776–1783. Thriemer K, Wernsdorfer G, Rojanawatsiriv- Tang G, Qin J, Hao LY et al. (2002). Use of a et C et al. (2005). In vitro activity of artmesinin short-term isotope-dilution method for determin- alone and in combination with retinol against ing the vitamin A status of children. Am J Clin Plasmodium falciparum. Wien Klin Wochenschr Nutr 76: 413–8. 117 Suppl 4: 45–8. Tanumihardjo SA (2002). Vitamin A and iron Thurlow RA, Winichagoon P, Green T et al. status are improved by vitamin A and iron supple- (2005). Only a small proportion of anaemia in mentation in pregnant Indonesian women. J Nutr northeast Thai schoolchildren is associated with 132: 1909–12. iron deficiency. Am J Clin Nutr 82: 380–7. Tanumihardjo SA, Koellner PG, Olson JA Thurnham DI (2009). Personal communica- (1990). The modified relative-dose response as- tion. say as an indicator of vitamin A status in a popu- Thurnham DI (2010). Newborn vitamin A lation of well-nourished American children. Am J dosing and neonatal mortality. Why do the results Clin Nutr 52: 1064–7. differ? Sight and Life Mag 1: 19-26. Tanumihardjo SA, Permaeish D, Muhilal Thurnham DI, Mburu ASW, Mwaniki DL et (2004). Vitamin A status and haemoglobin con- al. (2005). Micronutrients in childhood and the centrations are improved in Indonesian children influence of subclinical inflammation. Proc Nutr with vitamin A and deworming interventions. Eur Soc 64: 502–9. J Clin Nutr 58: 1223–30. Thurnham DI, Mburu ASW, Mwaniki DL Taren D, Duncan B, Greivenkamp J et al. et al. (2008). Using plasma acute-phase protein (2001). Appropriate technology for measuring concentrations to interpret nutritional biomarkers References 221

in apparently healthy HIV-1-seropositive Kenyan ington DC: The Nutrition Foundation, 26–7. adults. Br J Nutr 100: 174–82. UNICEF (2005). Vitamin A supplementation: Thurnham DI, McCabe GP, Northrop-Clewes progress for child survival. New York. CA (2003). Effects of sub-clinical infection on UNICEF (2009). Global child mortality con- plasma retinol concentrations and assesment of tinues to drop. vitamin A deficiency: meta-analysis. Lancet 362: US Institute of Medicine (2001). Dietary 2052–8. reference intakes for vitamin A, arsenic, boron, Thurnham DI, Northrop-Clewes CA, Mc- chromium, copper, iodine, iron, manganese, moly- Cullough FSW et al. (2000). Innate immunity, gut bdenum, nickel, silicon, vanadium, and zinc. integrity, and vitamin A in Gambian and Indian Washington, DC: National Academy Press. infants. J Infect Dis 182 (Suppl 1): S23–8. USDA-NCC (1998) Carotenoid database for Tielsch JM, Klemm R, West KP Jr et al. U.S. foods. http://www.nal.usda.gov/fnic/food- (2008). Newborn vitamin A dosing and early in- comp/Data/car98/car98.html. fant survival: current evidence and controversy. Sight and Life Mag 2: 22–6. Vahlquist A (1994). Role of retinoids in nor- Tielsch JM, Rahmathullah L, Thulasiraj RD mal and diseased skin. In: Vitamin A in health et al. (2007). Newborn vitamin A dosing reduces and disease. Blomhoff R, ed. New York: Marcel the case fatality but not incidence of common Dekker, 365–424. childhood morbidities in South India. J Nutr 137: van den Broek Nr, Letsky EA (2000). Etiol- 2470–4. ogy of anemia in pregnancy in South Malawi. Am Toshino A, Shiraishi A, Zhang W et al. (2005). J Clin Nutr 72 (Suppl 1): S247–56. Expression of keratinocyte transglutaminase in van den Broek NR, White SE, Flowers C et cornea of vitamin A-deficient rats. Cur Eye Res al. (2006). Randomised trial of vitamin A supple- 30: 731–9. mentation in pregnant women in rural Malawi Traber MC (2000). The bioavailability buga- found to be anaemic on screening by Hemocue. boo. Am J Clin Nutr 71: 1029–30. Brit J Obstet Gynaec 113: 569–76. Trumbo PR, Ellwood KC (2006). Lutein van Dillen J, de Francisco A, Overweg-Pland- and zeaxanthin intakes and risk of age-related soen WCG (1996) Long-term effect of vitamin A macular degeneration and cataracts: an evalua- with vaccines. Lancet 347: 1705. tion using the Food and Drug Administration’s van Leeuwen R, Boekhoorn S, Vingerling JR evidence-based review system for health claims. et al. (2005). Dietary intake of antioxidants and Am J Clin Nutr 84: 971–4. risk of age-related macular degeneration. J Am Med Assoc 294: 3101–7. Uchiyama S, Sumida T, Yamaguchi M (2004). van Lettow M, Harris AD, Kumwenda JJ et Anabolic effect of β-cryptoxanthin on bone com- al. (2004). Micronutrient malnutrition and wast- ponents in the femoral tissues of aged rats in vivo ing in adults with pulmonary tuberculosis with and in vitro. J Hlth Sci 50: 491–6. and without HIV co-infection in Malawi. Infect Udomkesmalee S, Dhanamitta S, Sirisinha S Dis 4: 1471–8. et al. (1992). Effect of vitamin A and zinc supple- van Lieshout M, West CE, van Breeman RB mentation on the nutriture of children in North- (2003). Isotope tracer techniques for studying the east Thailand. Am J Clin Nutr 56: 50–7. bioavailability of dietary carotenoids, particularly Underwood BA (1990). Methods for assess- β-carotene, in humans: a review. Am J Clin Nutr ment of vitamin A status. J Nutr 120: 1459–63. 77:12–28. Underwood BA (1996). Present status and van Stuijvenberg ME, Kvalsvig JD, Faber global progress toward the year 2000 goal. XVII M et al. (1999). Effect of iron-, iodine-, and IVACG Meeting Report, Guatemala City. Wash- β-carotene-fortified biscuits on the micronutrient 222 References

status of primary school children: a randomized response (MRDR) and the relative dose response controlled trial. Am J Clin Nutr 69: 497–503. (RDR) in the assessment of vitamin A status Varandas L, Julien M, Gomes A et al. (2001). in malnourished children. Am J Clin Nutr 61: A randomized, double-blind, placebo-controlled 1253–6. clinical trial of vitamin A in severe malaria in Wake K (1994). Role of perisinusoidal stel- hospitalized Mozambican children. Trop Pediatr late cells in vitamin A storage. In: Vitamin A in 21: 211–22. health and disease. Blomhoff R, ed. New York: Vernet N, Dennefeld C, Rochette-Egly C et Marcel Dekker, 73–86. al. (2006). Retinoic acid metabolism and signal- Wang X-D (2005). Biological activity of ing pathways in the adult and developing mouse lycopene against smoke-induced lung lesions: testis. Endocrinology 147: 96–110. targeting the IGF-1/IGFBP-3 signal transduction Vijayaraghavan K, Radhalah G, Prakasam BS pathway. In: Carotenoids and retinoids: molecular et al. (1990). Effect of massive dose vitamin A on aspects and health issues. Packer L, Obermuller- morbidity and mortality in Indian children. Lan- Jevic U, Kraemer K, Sies H, eds. Champaign, cet 336: 1342–5. Illinois: AOCS Press, 168–81. Villamor E, Fawzi WW (2005). Effects of Wang Z, Yin S, Zhao S et al. (2004). vitamin A supplementation on immune responses β-carotene-vitamin A equivalence in Chinese and correlation with clinical outcomes. Clin adults assessed by an isotope dilution technique. Microbiol Revs 18: 446–64. Br J Nutr 91: 121–31. Villamor E, Msamanga G, Spiegelman D et Ward SJ, MacGowan AL, Hornby SJ et al. al. (2002). Effect of multivitamin and vitamin A (2000). Vitamin A deficiency and genetic predis- supplements on weight gain during pregnancy position. Sight and Life Newsletter 1: 8–13. among HIV-1-infected women. Am J Clin Nutr Wasantwisut E (2002). Application of isotope 76: 1082–90. dilution technique in vitamin A nutrition. Food Vitamin A and Pneumonia Working Group Nutr Bull 23, Suppl 3: S103–6. (1995). Potential interventions for the prevention Wedner SH, Ross DA, Congdon N et al. of childhood pneumonia in developing countries: (2004). Validation of night blindness reports a meta-analysis of data from field trials to as- among children and women in a vitamin A-defi- sess the impact of vitamin A supplementation on cient population in rural Tanzania. Eur J Clin Nutr pneumonia morbidity and mortality. Bull WHO 58: 409–19. 73: 609–19. Wei S, Lai K, Patel S et al. (1997). Retinyl von Lintig J and Vogt K (2000). Filling the ester hydrolysis and retinol efflux from BFC-1B gap in vitamin A research. Molecular identifica- adipocytes. J Biol Chem 272: 1459–65. tion of an enzyme cleaving β-carotene to retinal. West CE, Eilander A, van Lieshout M (2002). J Biol Chem 275: 11915–20. Consequences of revised estimates of carote- von Lintig J and Vogt K (2004). Vitamin A noid bioefficacy of dietary control of vitamin A formation in animals: molecular identification deficiency in developing countries. J Nutr 132: of carotene cleaving enzymes. J Nutr 134: S2920S–6. S251S–6. West KP Jr (2002). Extent of vitamin A defi- Vuong Le T, Burgess ML (2004). Green rice ciency among preschool children and women of – a hidden source of provitamin A and micronu- reproductive age. J Nutr 132, Suppl: S2857S–66. trients for rice-consuming populations. Sight and West KP Jr, Chirambo M, Katz J et al. (1986). Life Newsletter 2: 10–2. Breast-feeding, weaning patterns, and the risk of xerophthalmia in Southern Malawi. Am J Clin Wahed MA, Alvarez JO, Khaled MA et al. Nutr 44: 690–7. (1995). Comparison of the modified relative dose West KP Jr, Darnton-Hill I (2008).Vitamin A References 223

deficiency. In Nutrition and health in developing WHO (1994). Using immunisation contacts countries. Semba RD, Bloem MW, eds. Totowa, as the gateway to eliminating vitamin A deficien- NJ: Humana Press, 377–433. cy. WHO/EPI/GEN/94.9.Rev.1. Geneva: WHO. West KP Jr, Katz J, Khatry SK et al. (1999). WHO (1995). Global prevalence of vitamin A Double blind, cluster randomised trial of low dose deficiency. WHO/NUT/95.3. Geneva: WHO. supplementation with vitamin A or β-carotene on WHO (1996). Indicators for assessing vita- mortality related to pregnancy in Nepal. Brit Med min A deficiency and their application in moni- J 318: 570–5. toring and evaluating intervention programmes. West KP Jr, Katz J, Shrestha SR et al. (1995). Geneva: WHO. Mortality of infants under six months of age sup- WHO (2009). Global prevalence of vitamin A plemented with vitamin A: a randomized, double- deficiency in populations at risk 1995–2005: masked trial in Nepal. Am J Clin Nutr 62: WHO Global database on Vitamin A Deficiency. 143–148. Geneva: WHO. West KP Jr, Khatry SK, Katz L et al. (1997). WHO/CHD Immunisation-Linked Vitamin A Impact of weekly supplementation of women Supplementation Study Group (1998). Random- with vitamin A or β-carotene on fetal, infant, and ised trial to assess benefits and safety of vitamin A maternal mortality in Nepal. Proc XVIII IVACG supplementation linked to immunisation in early Meeting, Cairo, Egypt, 22-26 September: 86. infancy. Lancet 352: 1257–63. West KP Jr, Le Clerq SC, Shrestha SR et al. WHO Expert Group (1976). Vitamin A defi- (1997). Effects of vitamin A on growth of vitamin ciency and xeropthalmia. Tech Rep Ser no. 590. A-deficient children: field studies in Nepal. J Nutr Geneva: WHO. 127: 1957–65. WHO Expert Group (1982). Control of vita- West KP Jr, McLaren DS (2003). The epi- min A deficiency and xerophthalmia. Tech Rep demiology of vitamin A deficiency disorders Ser no. 672. Geneva: WHO. (VADD) pp240-259 In: The epidemiology of WHO Guideline: Vitamin A supplementation eye disease, 2nd ed. Johnson GJ, Minassian DC, in infants 1–5 months of age. http://whqlibdoc. Weale RA, West SK, eds. London: Arnold. who.int/publications/2011/9789241501811_eng. West KP Jr, Pokhrel RP, Katz J et al. (1991). pdf. Efficacy of vitamin A in reducing preschool child WHO, UNICEF, IVACG Task Force (1997). mortality in Nepal. Lancet 338: 67–71. Vitamin A supplements. A guide to their use in the West KP Jr, Pokhrel RP, Khatry SK et al. treatment and prevention of vitamin A deficiency (1992). Estimating the relative efficiency of a and xerophthalmia, 2nd ed. Geneva: WHO. vitamin A intervention from population-based WHO, WFP, UNICEF (2007). Joint statement data. J Nep Med Assoc 30: 159–62. by the World Health Organization, the World West KP Jr, Sommer A (1984). Periodic large Food Programme and the United Nations Chil- oral doses of vitamin A for the prevention of vita- dren’s Fund. Preventing and controlling micro- min A deficiency and xerophthalmia: a summary nutrient deficiencies in populations affected by an of experiences. Washington DC: The Nutrition emergency multiple vitamin A and mineral sup- Foundation. plements for pregnant and lactating women, and West KP Jr, Sommer A (2002). Vitamin A for children aged 6 to 59 months. programme in Assam probably caused hysteria. Wiegand UW, Hartmann S, Hummler H. Brit Med J 324: 791. (1998). Safety of vitamin A: recent results. Int J Whalen C, Semba RD (2001). Tuberculo- Vitam Nutr Res 68(6): 411–6. sis. In: Human nutrition and health in develop- Wieringa FT, Dijkhuizen MA, van der Ven- ing countries. Semba RD, Bloem MW, eds. New Jongekrijg J et al. (2003). Micronutrient defi- Jersey: Humana Press, 209–36. ciency and supplementation in Indonesian in- 224 References

fants: effects on immune function. In: Tropical 283–91. diseases. Marzuki, Verhoel, Snippe, eds. New Wolf G (2005). Function of the protein RPE65 York: Kluver Academic/Plenum, 369–77. in the visual cycle. Nutr Revs 63: 97–100. Wieringa FT, Djikhuizen MA, West CE Wolf G (2006). Is 9-cis-retinoic acid the en- (2002). Estimation of the effect of the acute phase dogenous ligand for the retinoic acid-x receptor? response on indicators of micronutrient status in Nutr Revs 64: 532–8. Indonesian infants. J Nutr 132: 3061–6. Wolf G (2008). Retinoic acid as cause of cell Wieringa FT, Dijkhuizen MA, West CE proliferation or cell growth inhibition depending (2003). Redistribution of vitamin A after iron on activation of one of two different nuclear re- supplementation in Indonesian infants. Am J Clin ceptors. Nutr Revs 66: 55–9. Nutr 77: 651–7. Wolfson L, Strebel PM, Gacic-Dobo M et al. Wieringa FT, Dijkhuizen MA, West CE et al. (2007). Has the 2005 measles mortality reduction (2004). Reduced production of immunoregula- goal been achieved? A natural history modelling tory cytokines in vitamin A- and zinc-deficient study. Lancet 369: 191–200. Indonesian infants. Eur J Clin Nutr 58: 1498– Wongsiriroj N, Blaner WS (2004). Vitamin A 1504. storage and action in the liver. Sight and Life Williams CD (1933). A nutritional disease of Newsletter 2: 4–10. childhood associated with a maize diet. Arch Dis Wongsiriroj N, Paik J, Blaner WS (2005). Childh 8: 423–33. Beta-carotene conversion to vitamin A: regulation Winichagoon P, McKenzie JE, Chavasit V et in humans. IVACG XXII Meeting M56: 43. al. (2006). A multimicronutrient-fortified season- World Bank (1993). World Development ing powder enhances the hemoglobin, zinc, and Report: Investing in Health. Washington DC: iodine status of primary school children in North World Bank. East Thailand: a randomized controlled trial of Wrighting DM, Andrews NC (2006). Inter- efficacy. J Nutr 136: 1617–23. leukin induces hepcidin expression through Wittpenn JR, Tseng SCG, Sommer A (1986). STAT3. Blood 108: 3204–9. Detection of early xerophthalmia by impression Wyss A, Wirtz GM, Woggon W-D et al. cytology. Arch Ophthal 104: 237–9. (2001). Expression pattern and localisation of Wittpenn JR, West KP, Keenum D et al. β,β-carotene 15,15̓ dioxygenase in different tis- (1988). ICEPO Training Manual. Assessment of sues. Biochem J 354: 521–9. vitamin A status by impression cytology. Balti- more: The Johns Hopkins University Press. Xu Q, Parks CG, De Roo LA et al. (2009). Wiysonge CS, Shey MS, Sterne JAC et al. Multivitamin use and telomere length in women. (2005). Vitamin A supplementations for reduc- Am J Clin Nutr 89: 1857–1863. ing the risk of mother-to-child transmission of HIV infection. Cochrane Database Syst Rev no Yan J, Xia Q, Cherng SH et al. (2005). Photo- 4: CD003648. induced DNA damage and photocytotoxicity of Wolf G (1998). Vitamin A functions in the retinyl palmitate and its photodecomposition regulation of the dopaminergic system in the brain products. Toxicol Ind Health 21: 167–75. and pituitary gland. Nutr Revs 56: 354–8. Yang, Graham, Mody et al. (2005). Serum Wolf G (2002). The regulation of the thyroid- retinol binding protein 4 contributes to insulin stimulating hormone of the anterior pituitary resistance in obesity and type 2 diabetes. Nature gland by thyroid hormone and by 9-cis-retinoic 436: 356–62. acid. Nutr Revs 60: 374–77. Ye X, Al-Babali S, Klöti A et al. (2000). Wolf G (2004). The visual cycle of the cone Engineering the provitamin A (β-carotene) bio- photoreceptors of the retina. Nutr Revs 62: synthetic pathway into (carotenoid-free) rice 225

endosperm. Science 287: 303–5. Zhao X, Aldini G, Johnson EJ et al. (2006). Yeum K-J, Booth SL, Sadowski JA (1996). Modification of lymphocyte DNA damage by Human plasma carotenoid response to the inges- carotenoid supplementation in postmenopausal tion of controlled diets high in fruits and vegeta- women. Am J Clin Nutr 83: 163–9. bles. Am J Clin Nutr 64: 594–602. Zile MH (2004). Vitamin A requirement for Yeum K-J, Russell RM (2002). Carotenoid early cardiovascular morphogenesis specification bioavailability and bioconversion. Annu Rev Nutr in the vertebrate embryo: insights from the avian 22: 483–504. embryo. Exp Biol Med 229: 598–606. Yorston D, Foster A (1992). Corneal ulcer- Zimmermann MB, Biebinger R, Rohner F et ation in Tanzanian children: relationship between al. (2006). Vitamin A supplementation in child- malaria and herpes simplex keratitis. Trans Roy ren with poor vitamin A and iron status increases Soc Trop Med Hyg 86: 456–7. erythropoietin and hemoglobin concentrations You C-S, Parker RS, Goodman KJ et al. without changing total body iron. Am J Clin Nutr (1996). Evidence of cis-trans isomerization of 84: 580–6. 9-cis-β-carotene during absorption in humans. Zimmermann MB, Wegmueller R, Zeder C et Am J Clin Nutr 64: 177–83. al. (2004). Triple fortification of salt with micro- Yuan JM, Stram DO, Arakawa K et al. (2003). capsules of iodine, iron and vitamin A. Am J Clin Dietary cryptoxanthin and reduced risk of lung Nutr 80: 1283–90. cancer: the Singapore Chinese Health Study. Epi- Zimmermann MB, Wegmueller R, Zeder C et demiol Biomarkers Prev 12: 890-898. al. (2004). The effects of vitamin A supplementation on thyroid function in goitrous children. J Zago LB, Dupraz H, Sarchi MI et al. (2002). Clin Endocrinol Metab 89: 5441–7. The molar ratio of retinol-binding protein to transthyretin in the assessment of vitamin A status in adults. Proposal of a cut-off point. Clin Chem Lab Med 40: 1301–7. 226

Building bridges for better nutrition.

For more information please visit www.sightandlife.org Donald S. McLaren, Klaus Kraemer Sight and Life Press, Third Edition 2012