nutritional anemia book

1Preface

Nutritional Anemia

SIGHT AND LIFEPress

Edited by Klaus Kraemer

SIGHT AND LIFE, Basel, Switzerland

Michael B. ZimmermannSwiss Federal Institute of Technology, Zurich, Switzerland

NutritionalAnemia

SIGHT AND LIFE Mission StatementSIGHT AND LIFE is a humanitarian initiative of DSM. It aims to ensure a sustainable and significantimprovement in human nutrition and health by encouraging partnerships with universities and intergov-ernmental and governmental agencies, by generating and exchanging scientific information and byforming networks.

Copyright SIGHT AND LIFE 2007All rights reserved. Publications from SIGHT AND LIFE can be obtained from

SIGHT AND LIFE Pressc/o SIGHT AND LIFE / DSM Nutritional Products LtdPO Box 21164002 BaselSwitzerlandPhone: +41 61 68 87494Fax: +41 61 68 81910Email: [email protected]: www.sightandlife.org

Requests for permission to reproduce or translate SIGHT AND LIFE publications should be submittedto the address above.

Opinions, compilations, tables and figures contained in this publication do not necessarily represent the point of view of SIGHT AND LIFE and are the sole responsibility of the authors. The mention ofspecific companies and trademarks does not imply that they are endorsed by SIGHT AND LIFE. Allreasonable precautions have been taken by SIGHT AND LIFE to verify the content of this publication.However, this publication does not constitute or provide scientific or medical advice and is distributedwithout warranty of any kind, either express or implied. The reader shall be solely responsible for anyinterpretation or use of the material contained herein. In no event shall SIGHT AND LIFE be liable forany damages arising from the reader's reliance upon, or use of, these materials.

The paper used in this book is acid-free and falls within the guidelines established to ensure perma-nence and durability.

Cover photo by Ulla Lohmann, GermanyCover illustration by graphic art studio, Grenzach-Wyhlen, GermanyProofreading by transparent translation & text services, Berlin, GermanyTypeset and print by Burger Druck, Waldkirch, Germany

ISBN 3-906412-33-4

IV

For half a billion women in developingregions worldwide, anemia is a life-long bur-den, one which affects most of their infantsand young children as well. Controlling ane-mia in these vulnerable groups could signifi-cantly reduce maternal and infant morbidity.It would also enhance intellectual and workcapacity, thereby improving family, commu-nity and national socioeconomic development.

In May 2002, the General Assembly of the United Nations reemphasized that control of nutritional anemia should be one of theglobal Development Goals to be achieved inthe early years of this new millennium. Despitethis, the global prevalence of anemia hashardly declined in the past decade, althoughconsiderable programmatic experience existsand a vast amount of scientific data has beencompiled on iron metabolism. Much is stillunknown, however, and many new issues continue to emerge from the ongoing research,both basic and programmatic. The reasons forthis lack of improvement include the multi-factorial etiology of anemia, underfunding andpoor program implementation, often designedon the assumption that the sole cause of anemia is iron deficiency.

It is increasingly clear that effective con-trol of anemia requires integrated solutionsthat are tailored to the particular needs andopportunities in each country. Components ofsuch an approach include food fortification,micronutrient supplementation of vulnerablegroups (particularly children and women ofchildbearing age), education, and dietarydiversification, as well as control of diseasessuch as malaria, worm infections, and otherchronic endemic infections. While each of

these can help reduce the burden of anemia,none is capable of doing the job on its own.

The chapters of this book offer an accountof the information that was presented and comprehensively discussed at a workshop onNutritional Anemia in Barcelona, Spain, onSeptember 27, 2006, in which all the contribu-tors to this volume themselves took part. Thisevent was co-organized by the editors of thisvolume. We sought the timely publication ofthis book in order to provide the latest updateon the complex causes and consequences ofnutritional anemia, and the effectiveness ofcurrent control strategies. The field of anemiais clearly of great interest to scientists, policymakers and program mangers. We hope thisvolume will help point the way forward in con-trolling this major global health problem. Theintroductory chapters in this book give anoverview of the global burden of anemiaprevalence, the economic implications andfunctional consequences, and the significanceof these factors for policy makers. Subsequentchapters provide basic scientific informationon iron metabolism and interactions withmacronutrients and micronutrients as well asthe role of infections in fostering anemia.Other chapters address the information needsof program managers, detailing programmaticapproaches and outlining the safety and techni-cal aspects of interventions.

We are particularly grateful to the authors,who are all leading scholars from UN agenciesand the wider academic world, for their excel-lent contributions. A special acknowledgementis also due to all reviewers whose valuablecomments have helped to improve the qualityof the chapters. We would like to give a very

1Preface

PREFACE

special thanks to the SIGHT AND LIFE team,Svenia Sayer-Ruehmann and Anne-CatherineFrey, for assisting with the workshop as well asfor all their technical support in finalizing thisbook. Svenia devoted much of her time andenergies to corresponding with the authors toensure that our tight timeline was met. Wewould also like to express our deep apprecia-tion to Jane Badham for her invaluable assis-

tance with the editing of the chapters. Lastly,we wish to thank SIGHT AND LIFE and DSM,in particular its President and incoming CEO,Feike Sijbesma, for his continuing support ofour work and the publication of this book.

Klaus KraemerMichael B. Zimmermann

VI Preface

Every year hunger and undernutrition claimmore than 10 million lives more than the deathsfrom AIDS, malaria and tuberculosis combined.Many of these people are affected by hiddenhunger, a lack of essential vitamins and miner-als, known as micronutrient deficiencies, whichdo not necessarily result in the swollen bellies andstick-like limbs many associate with serious mal-nutrition.

The effects of poor nutrition are not alwayseasy to spot. Undernourished children are lesslikely to attend school, more likely to have learn-ing difficulties, are more susceptible to disease.They are more likely to die young. Undernour-ished adults are less capable of providing suffi-cient food and other necessities for their families.Their immune systems may be compromised.Their productivity and income is invariablylower.

Poor nutrition impacts the health and devel-opment of individuals. It also retards the eco-nomic and social development of nations. Forinstance, it is estimated that anemia alone causesUS $50 billion in gross domestic product lossesannually.1 Reducing the number of hungry andmalnourished people translates into a better qual-ity of life for individuals, as well as narrowingnational disparities in health, education, andwealth.

Infants, young children and women of child-bearing age are those at greatest risk of nutritional

anemia. This condition, which claims one millionlives each year, is associated with increased childand maternal mortality, stillbirths, low-birth-weight and premature babies. It is one of theworld's leading causes of disability and can causemild mental retardation and decreased work per-formance.

Caused predominantly by iron-deficiency,anemia is inextricably linked with people's nutri-tional status and hunger. In 2000, the UnitedNations Sub-Committee on Nutrition reportedthat 43% of people in developing countries cur-rently suffer from anemia2. In spite of the signifi-cant burden anemia places on health systems andeconomies, it has often been overlooked by theinternational and public health communities.

This book assembles some of the leadingresearch on health and nutrition. The evidenceclearly points to the benefits and feasibility ofreducing the prevalence of anemia and combatingundernutrition. This textbook also serves as aguide for how government, international agen-cies, and non-governmental organizations canwork together to decrease rates of nutritional ane-mia worldwide. It reviews the most effectiveways of measuring and monitoring the prevalenceof nutritional anemia and the most successful pro-gram designs for public health interventions.

International organizations, such as the WorldFood Programme, depend on such research inorder to deliver the best possible assistance to

1Preface

1 MacDonald B, Haddad L, Gross R, McLachlan M. A foundation for development. Administrative Committee on Coordination/Subcommittee on Nutrition. Accessed July 24, 2005. Available at: http://www.ifpri.org/pubs/books/intnut/intnut.pdf

2 ibid

FOREWORD

hungry and malnourished poor people. These arethe people who have least choice in their diet, andwho are at tremendous risk of anemia and othermicronutrient deficiencies. The World Food Pro-gramme and its partners help to provide iron sup-plements and fortified foods in developing coun-tries around the world.

Iron fortification is one of the most cost-effective interventions and nutrition educationprograms have reduced the prevalence of anemiaamong infants and young children by increasingtheir consumption of fortified foods. Presently,the World Food Programme is the world's largestpurchaser of vitamin- and mineral-fortified blendedfoods.

Fortified foods alone are not enough. Suc-cessful delivery of such interventions requires thestrengthening of health systems, increased aware-ness, and financial investment. An integratedapproach including better water and sanitation,

infectious disease control, nutrition education,food security, and poverty-reduction programs is required.

Together with partners like DSM and SIGHTAND LIFE, we can and must do more to reducemicronutrient deficiencies such as nutritionalanemia. This textbook, which focuses on re-ducing the prevalence of nutritional anemia, iskey to reducing overall hunger and malnutritionrates. It is incumbent upon us to use this informa-tion to combat nutritional anemia, improve thenutritional status of those in developing countries,and mitigate one of today's greatest public healthproblems.

The health and well being of millions of hungry women and children depends on it.

James T. MorrisExecutive Director World Food Programme

VIII Foreword

HAROLD ALDERMANAfrica Region of the World Bank, Washington,DC, USA; [email protected]

JANE BADHAMJB Consultancy, Health Communication andStrategy Consultants, Johannesburg, SouthAfrica; [email protected]

HANS-KONRAD BIESALSKIInstitute for Biological Chemistry and Nutritionat the University of Hohenheim, Hohenheim,Germany; [email protected]

MARTIN BLOEMWorld Food Programme (WFP), Rome, Italy;[email protected]

TOMMASO CAVALLI-SFORZANutrition and Food Safety, WHO Regional Officefor the Western Pacific, Manila, Philippines; [email protected]

MARY COGSWELLDivision of Nutrition and Physical Activity, Cen-ters for Disease Control and Prevention, Atlanta;USA

IAN DARNTON-HILLNutrition Section, UNICEF, New York, USA;[email protected]

OMAR DARYA2Z Project, Academy for Educational Develop-ment, Washington, DC, USA; [email protected]

BRUNO DE BENOISTWorld Health Organization (WHO), Geneva,Switzerland; [email protected]

SASKIA DE PEEWorld Food Programme (WFP), Rome, Italy;[email protected]

INES EGLIInstitute of Food Science and Nutrition, SwissFederal Institute of Technology (ETH), Zurich,Switzerland; [email protected]

JRGEN ERHARDTUniversity of Indonesia, SEAMEO-TROPMED,Jakarta, Indonesia; [email protected]

ALISON D. GERNANDBloomberg School of Public Health, Johns HopkinsUniversity, Baltimore, USA; [email protected]

GARY R. GLEASONFriedman School of Nutrition Science and Policy,Tufts University, Boston, USA;[email protected]

EVA HERTRAMPF DAZInstitute of Nutrition and Food Technology(INTA), University of Chile, Santiago, Chile;[email protected]

SUSAN HORTONWilfrid Laurier University, Waterloo, Canada;[email protected]

RICHARD HURRELLInstitute of Food Science and Nutrition, SwissFederal Institute of Technology (ETH), Zurich,Switzerland; [email protected]

ALAN JACKSONInstitute of Human Nutrition, University ofSouthampton, Southampton, UK; [email protected]

AFAF KAMAL-ELDINDepartment of Food Science, Swedish Universityof Agricultural Sciences, Uppsala, Sweden;[email protected]

KLAUS KRAEMERSIGHT AND LIFE, Basel, Switzerland; [email protected]

IXPreface

CONTRIBUTORS

SEAN LYNCHEastern Virginia Medical School, Norfolk, USA;[email protected]

M.G. VENKATESH MANNARThe Micronutrient Initiative, Ottawa, Canada;[email protected]

ERIN MCLEANWorld Health Organization (WHO), Geneva,Switzerland; [email protected]

REGINA MOENCH-PFANNERGlobal Alliance for Improved Nutrition (GAIN),Geneva, Switzerland; [email protected]

CHRISTINE A. NORTHROP-CLEWESNorthern Ireland Centre for Food and Health, Univer-sity of Ulster, Coleraine, UK; [email protected]

MANUEL OLIVARESInstitute of Nutrition and Food Technology(INTA), University of Chile, Santiago, Chile;[email protected]

NEAL PARAGASInstitute of Human Nutrition, Columbia Univer-sity, New York, USA; [email protected]

KLAUS SCHMANNTechnical University of Munich, Freising, Ger-many; [email protected]

JOHN M. SCOTTSchool of Biochemistry & Immunology, TrinityCollege Dublin, Dublin, Ireland; [email protected]

NEVIN SCRIMSHAWInternational Nutrition Foundation, Boston, USA;[email protected]

RICHARD SEMBASchool of Medicine, Johns Hopkins University,Baltimore, USA; [email protected]

NOEL SOLOMONSCenter for Studies of Sensory Impairment, Agingand Metabolism (CeSSIAM), Guatemala City,Guatemala; [email protected]

ALFRED SOMMERBloomberg School of Public Health, Johns Hopkins University, Baltimore, USA;[email protected]

ELISABETH STOECKLINR & D Human Nutrition and Health, DSM Nutri-tional Products Ltd, Kaiseraugst, Switzerland;[email protected]

BRIAN THOMPSONFood and Agriculture Organization (FAO), Rome,Italy; [email protected]

DAVID THURNHAMNorthern Ireland Centre for Food and Health,University of Ulster, Coleraine, UK;[email protected]

MELODY C. TONDEURDivision of Gastroenterology, Hepatology andNutrition, Hospital for Sick Children, Toronto,Canada, [email protected]

MARET G. TRABERLinus Pauling Institute & Department of Nutritionand Exercise Sciences, Oregon State University,Corvallis, USA; [email protected]

RICARDO UAUYInstitute of Nutrition and Food Technology(INTA), University of Chile, Santiago, Chile;[email protected]

KEITH P. WESTBloomberg School of Public Health, Johns HopkinsUniversity, Baltimore, USA; [email protected]

DANIEL WOJDYLAEscuela de Estadistica, Universidad Nacional deRosario, Argentina

MICHAEL ZIMMERMANNLaboratory for Human Nutrition, Swiss FederalInstitute of Technology (ETH), Zurich, Switzer-land; [email protected]

STANLEY ZLOTKINDepartments of Paediatrics and Nutritional Scien-ces and Public Health Sciences, University ofToronto, Canada; [email protected]

X Contributors

Preface V

Foreword VII

Contributors IX

Contents XI

CHAPTER 1 Worldwide prevalence of anemia in preschool aged children, pregnant women and non-pregnant women of reproductive ageErin McLean, Ines Egli, Mary Cogswell, Bruno de Benoistand Daniel Wojdyla 1

CHAPTER 2 The case for urgent action to address nutritional anemiaM.G. Venkatesh Mannar 13

CHAPTER 3 The economics of addressing nutritional anemiaHarold Alderman and Susan Horton 19

CHAPTER 4 Diagnosis of nutritional anemia laboratory assessment of iron statusHans-Konrad Biesalski and Jrgen G. Erhardt 37

CHAPTER 5 An overview of the functional significance of iron deficiencyGary Gleason and Nevin S. Scrimshaw 45

CHAPTER 6 Iron metabolismSean Lynch 59

CHAPTER 7 Optimizing the bioavailability of iron compounds for food fortificationRichard Hurrell and Ines Egli 77

CHAPTER 8 Copper and zinc interactions in anemia: a public health perspectiveManuel Olivares and Eva Hertrampf and Ricardo Uauy 99

CHAPTER 9 Nutritional anemia: B-vitaminsJohn M. Scott 111

CHAPTER 10 Vitamin A in nutritional anemiaKeith P. West, Jr., Alison D. Gernand and Alfred Sommer 133

XIPreface

CONTENTS

CHAPTER 11 Oxidative stress and vitamin E in anemiaMaret G. Traber and Afaf Kamal-Eldin 155

CHAPTER 12 SeleniumRichard D. Semba 189

CHAPTER 13 Interactions between iron and vitamin A, riboflavin, copper, and zinc in the etiology of anemiaMichael B. Zimmermann 199

CHAPTER 14 Anemia in severe undernutrition (malnutrition)Alan A. Jackson 215

CHAPTER 15 Infection and the etiology of anemiaDavid I. Thurnham and Christine A. Northrop-Clewes 231

CHAPTER 16 Making programs for controlling anemia more successfulSaskia de Pee, Martin W. Bloem, Regina Moench-Pfanner and Richard D. Semba 257

CHAPTER 17 Successful approaches: SprinklesStanley H. Zlotkin and Melody Tondeur 269

CHAPTER 18 Safety of interventions to reduce nutritional anemiasKlaus Schmann and Noel W. Solomons 285

CHAPTER 19 The importance and limitations of food fortification for the management of nutritional anemiasOmar Dary 315

CHAPTER 20 Food-based approaches for combating iron deficiencyBrian Thompson 337

CHAPTER 21 Global perspectives: accelerating progress on preventing and controlling nutritional anemiaIan Darnton-Hill, Neal Paragas and Tommaso Cavalli-Sforza 359

CHAPTER 22 Conclusions and research agendaKlaus Kraemer, Elisabeth Stoecklin and Jane Badham 383

Index 389

XII Contributors

1Worldwide prevalence of anemia in pre-

school aged children, pregnant women andnon-pregnant women of reproductive age*

1Department of Nutrition for Health and Development, World Health Organization, Geneva, Switzerland

2Division of Nutrition and Physical Activity, Centers for Disease Control and Prevention, Atlanta, USA

3Human Nutrition Laboratory, Swiss Federal Institute of Technology, Zurich, Switzerland4Escuela de Estadistica, Universidad Nacional de Rosario, Rosario, Argentina

Contact: [email protected]

* Two of the authors are staff member of the World health Organization. They alone are responsible for the views expressed inthis publication and they do not necessarily represent the decisions, policy or views of the World Health Organization. Moreover

the findings and conclusions in this report are those of the authors and do not necessarily represent the views of CDC.

and Mary Cogswell2

Erin McLean1 Ines Egli3

Bruno de Benoist1 Daniel Wojdyla4

ERIN MCLEANErin obtained her PhD in Nutrition from theUniversity of California, Davis, USA. She is cur-rently working as a Technical Officer for theMicronutrient Program in the Department ofNutrition for Health and Development at the WorldHealth Organization in Geneva, Switzerland. Erinsprimary work is with the Micronutrient DeficiencyInformation System, developing regional and globalprevalence estimates for anemia.

MARY COGSWELLMary obtained her PhD. in nutritional epidemiol-ogy from Johns Hopkins University, School ofHygiene and Public Health, Baltimore, USA. Sheis currently an epidemiologist with the Divisionof Nutrition and Physical Activity, Centers forDisease Control and Prevention. Mary collabo-rates with and provides consultation to statehealth departments, universities, internationalorganizations, and other constituents on nutritionassessment, surveillance, and evaluation of pro-grams. She has over 50 publications in peer-reviewed journals and has won awards for herresearch on perinatal nutrition and on the assess-ment and prevention of iron deficiency.

INES EGLIInes has a PhD in Nutrition from the Swiss FederalInstitute of Technology (ETH), Zurich, Switzerland.

She is currently a senior scientist and lecturer at theHuman Nutrition Laboratory at ETH where shesupervises research projects on mineral bioavailabil-ity and infant nutrition, and leads a postgraduatecourse in Nutrition and Health. She has also workedin the Nutrition Department of the World HealthOrganization in the area of micronutrient deficien-cies with a special focus on iron, iodine, and vitaminA.

DANIEL WOJDYLADaniel has an MSc in Biostatistics from theUniversity of North Carolina, Chapel Hill, USA. Heis currently Adjunct Professor in the School ofStatistics of the National University of Rosario,Rosario, Argentina and is also a statistician at theRosario Center for Perinatal Studies (CREP), wherehe is in charge of statistical analyses for clinical tri-als and epidemiologic studies on reproductivehealth.

BRUNO DE BENOISTBruno is a paediatrician. He obtained his MD fromthe University of Paris, France, and an MSc inNutrition from the London School of Hygiene andTropical Medicine, UK. He is currently Coordinatorof the Micronutrient Unit at the World HealthOrganization, Geneva, Switzerland. His special inte-rest is micronutrient disorders and their relation topublic health and food fortification.

E. McLean, M. Cogswell, I. Egli, D. Wojdyla and B. de Benoist2

Anemia is a widespread public health problemassociated with an increased risk of morbidityand mortality, especially in pregnant womenand young children (1). Among the numerousfactors, both nutritional (such as vitamin andmineral deficiencies) and non-nutritional (suchas infection and hemoglobinopathies), that con-tribute to the onset of anemia, iron deficiencyand malaria play an important role. Given the role of iron in oxygen transport and the low levels of available iron in the diets of alarge proportion of the global population, it isassumed that iron deficiency is one of thebiggest contributing factors to the global burdenof anemia. Iron deficiency is considered one ofthe ten leading global risk factors in terms of itsattributable disease burden.

One of the mandates of the World HealthOrganization (WHO) is to inform its MemberStates about the global health situation. Previousestimates of anemia prevalence were made for allpopulation groups in 1985 (2) and 2001 (3), whileestimates of anemia prevalence in women weremade in 1982 (4) and 1992 (5). For these estimates, the data were collected prior to 1990,with the exception of the 2001 estimates, whichincluded data up to 1995, but did not includedetailed description of the methodology on how the estimates were derived (3). Thus, it istime to update the global anemia estimates andprovide a current picture of the global situation,especially in the most affected groups womenand young children. The objective of this paper is to present estimates of anemia prevalence in preschool aged children, pregnant women and nonpregnant women at global and regional levels, using data collected by WHO for the Vitamin and Mineral Nutrition Information Sys-tem (VMNIS) (http://www.who.int/vmnis). Theseestimates are based on the 192 Member States of WHO, which represent 99.8% of the globalpopulation.

METHODOLOGY

Data collectionWe used data from the VMNIS, which are col-lected from scientific literature and from partners,including WHO Regional and Country offices,United Nations organizations, Ministries ofHealth, research and academic institutions, andnon-governmental organizations. We searchedMEDLINE and WHO Regional databases system-atically, and manually searched for articles pub-lished in non-indexed medical and professionaljournals. We included surveys in the VMNIS onlyif hemoglobin was measured from capillary,venous, or cord blood using quantitative photo-metric methods or automated cell counters and ifanemia prevalence or mean hemoglobin concen-trations were reported, while we excluded sur-veys that measured only clinical signs of anemiaor hematocrit. For the VMNIS, we included datarepresentative of any administrative level within acountry, including nationally representative data,surveys representative of a Region, the firstadministrative level boundary, second administra-tive level boundary, or local surveys.

Data selection for the estimatesWe selected survey data from the VMNIS onhemoglobin concentration and/or anemia pre-valence for each country using four criteria: thetime frame in which the survey took place, theadministrative level for which the survey was representative (national or sub-national), the sur-vey sample size, and the population groups sur-veyed.

The time frame for the current estimates is19932005. If no survey date was provided, weused the date of publication. Nationally represen-tative surveys were used preferentially for coun-try estimates. If two or more national surveyswere available, we used the most recent. When a

3Worldwide prevalence of anemia

INTRODUCTION

national survey was unavailable for a country, iftwo or more surveys representative of the firstadministrative boundary within a country (e.g.,state, province, etc.) were available we pooledthem, weighted by the population size of the area they represented and used them as represen-tative of the entire country. A sample size of 100or more was generally required although we madesome exceptions. If the sample size was between50 and 100 and the results were being extrap-olated to fewer than 50,000 people or to preg-nant women, for whom the number of womenincluded is usually small, we used the data. Whena country did not have data that met these criteria,we estimated the prevalence using predictionmodels.

PopulationWe defined the population groups as follows: preschool aged children (04.99 years), pregnantwomen (no age range defined), nonpregnantwomen (15.0049.99 years). Wherever possible,children below 0.5 years of age were excludedfrom the estimate for preschool aged childrensince the cut-off for anemia is not defined in thisage group. However, the estimate was applied tothe entire population of children less than 5 yearsof age. Occasionally, in the nonpregnant womengroup, pregnant women could not be excludedbecause all women were included in the figureprovided by the country report, but pregnantwomen usually made up a small proportion of thegroup and thus their exclusion would not beexpected to change the figure significantly. If asurvey reported results by physiological status,lactating women were combined with other non-pregnant nonlactating women to provide the esti-mate for nonpregnant women.

Hemoglobin thresholdThe hemoglobin concentration cut-offs to defineanemia are the WHOrecommended cut-offs foreach population group (3): 110 g/L for preschool

aged children and pregnant women and 120 g/Lfor nonpregnant women. If anemia prevalencewas adjusted for altitude or smoking in a survey,the adjusted figure was used since there is statisti-cal and physiological evidence to support thathemoglobin distributions vary under these condi-tions (6, 7). However, we did not adjust data notalready adjusted and we did not accept any othercorrections.

Anemia prevalence from survey dataWhen the anemia prevalence was reported usingthe appropriate hemoglobin threshold, we usedthe data provided in the survey. However, if sur-veys provided mean hemoglobin concentration oronly anemia prevalence for an alternative cut-off,we derived the prevalence by assuming a normaldistribution of hemoglobin and utilizing otherinformation provided about the populationshemoglobin concentration. We utilized, in orderof preference: the mean and standard deviationprovided in the survey, the mean hemoglobin concentration alongside the prevalence for analternative cut-off to derive a standard deviation,or the prevalence for an alternative cut-off and an average standard deviation derived for the population group from data in the VMNIS. Since hemoglobin concentrations are likely to be skewed towards lower values in a populationwith a high prevalence of deficiency, we mayhave slightly overestimated anemia prevalence insome populations.

In cases where disaggregated data were provided or where subnational data were used, wepooled the data. For data disaggregated by age,physiological status or any other classification,we derived anemia prevalence by weighting eachprevalence estimate by its sample size. For subna-tional data, we weighted the data by the generalpopulation estimate for that area using the mostrecently available census data for the countrybetween 1993 and 2005.


We considered each estimate representativeof the entire country whether from national orsubnational data. For each estimate, we calculatedthe variance in the logit scale using the samplesize and generated a 95% confidence interval as ameasure of uncertainty, which was back-trans-formed to the original scale (8, 9). We used adesign effect of two to calculate the confidenceinterval, since the majority of the surveysemployed cluster sampling, but did not provide anestimate of their design effect.

Models to estimate anemia prevalence for countries with no eligible dataThe level of development and the health of a population are closely related. For this reason, wedeveloped regression models to predict anemiaprevalence in countries with no eligible data,using health and development indicators. Westarted with the 2002 Human Development Index(HDI) score, a numerical reflection of develop-ment produced by the United Nations Develop-ment Programme (UNDP) and comprising indica-tors of life expectancy, education and wealth (10). For the 17 WHO Member States with no HDIscore produced by UNDP, an estimate of the HDI score was generated using a regressionmodel and the same indicators for life expectancyand wealth, but a proxy indicator for education(11-13). HDI explained 22.448.9% of the varia-tion in anemia prevalence in countries with sur-vey data for the three population groups. To fur-ther improve the model, we included other healthindicators available from WHO statistics (avail-able for M190 of 192 WHO Member States) aspotential explanatory variables. Some of theseadditional indicators significantly improved theanemia prediction model and we kept them in the model. For preschool aged children (n=82),the additional variables were expenditure onhealth and adult female mortality, which togetherwith HDI explained 55% of the variation in anemia prevalence. In pregnant women (n=60),immunization for DTP3 before 1 year of age,

expenditure on health and adult male mortalitywere utilized alongside HDI to predict anemia,explaining 32.3% of the variation in prevalence.To predict anemia prevalence for nonpregnantwomen, population growth rate and expenditureon health with HDI explained 45.3% of the varia-tion. For these estimates, we calculated the vari-ance based on the regression equations and pro-duced 95% confidence intervals as a measure ofuncertainty.

Classification of anemia as a public healthproblemThe prevalence of anemia as a public health prob-lem is categorized as follows:


Populationgroup2 Global Africa Asia Europe LAC NA Oceania

PreSAC 76.1 76.7 82.1 19.2 70.5 92.4 5.1

PW 69.0 65.3 80.9 0.9 38.4 92.8 4.7

NPW 73.5 63.6 88.8 23.9 37.5 89.9 16.5

Table 1.1: Percentage of the population covered by actual survey data globally and by UN Region.1

1 WHO Member States are stratified by United Nations Regions: Africa, Asia, Europe, Latin America and the Caribbean (LAC),Northern America (NA), and Oceania.

2 Population groups: PreSAC, preschool aged children (0.004.99y); PW, pregnant women (no age range defined); NPW, non-pregnant women (15.0049.99y).

Area

PreSAC1 PW NPW

Prevalence(%)3

# affected(millions)

Prevalence(%)3


Prevalence(%)3


Global 47.4(45.749.1)293.1

(282.8303.5)41.8

(39.943.8)56.4

(53.859.1)30.2

(28.731.6)468.4

(446.2490.6)

UN Region2

Africa 64.6(61.767.5)93.2

(89.197.4)55.8

(51.959.6)19.3

(18.020.7)44.4

(40.947.8)82.9

(76.589.4)

Asia 47.7(45.250.3)170.0

(161.0178.9)41.6

(39.044.2)31.7

(29.733.6)33.0

(31.334.7)318.3

(302.0334.6)

Europe 16.7(10.523.0)6.1

(3.88.4)18.7

(12.325.1)1.4

(0.91.8)15.2

(10.519.9)26.6

(18.434.9)

LAC 39.5(36.043.0)22.3

(20.324.3)31.1

(21.840.4)3.6

(2.54.7)23.5

(15.931.0)33.0

(22.443.6)

NA 3.4(2.04.9)0.8

(0.41.1)6.1

(3.48.8)0.3

(0.20.4)7.6

(5.99.4)6.0

(4.67.3)

Oceania 28.0(15.840.2)0.7

(0.41.0)30.4

(17.043.9)0.2

(0.10.2)20.2

(9.530.9)1.5

(0.72.4)

Table 1.2: Anemia in preschool aged children, pregnant women and non-pregnant women globally and byRegion.

1 Population groups: PreSAC, preschool aged children (0.004.99y); PW, pregnant women (no age range defined); NPW,nonpregnant women (15.0049.99y).

2 WHO Member States are stratified by United Nations Regions: Africa, Asia, Europe, Latin America and the Caribbean (LAC),Northern America (NA), and Oceania.

3 95% CI

We estimated the number of individuals withanemia for each country and grouping of coun-tries for all population groups based on the esti-mated proportion of the population with anemiafor every country and presented the 95% confi-dence intervals as a measure of uncertainty.

Combining national estimatesWe combined country estimates to provide esti-mates at the global level as well as by UnitedNations Region by pooling the data and weight-ing it by the population that each estimate repre-sented. We constructed a 95% confidence intervalby using the estimated variance of the weightedaverage. For one country without data, no proxyindicators were available and so no country esti-mate was generated, but the UN subregional esti-mate had to be applied to that country to makeregional and global estimates.

RESULTS

All three population groups were covered by asignificant amount of actual data, which coveredbetween 69.076.1% of the population in allgroups (Table 1.1). Coverage varied by UNRegion and was highest in Northern America,Asia, and Africa, while it was lower in Europeand Oceania. Only preschool children had eligible subnational data, which covered a smallproportion of the population (3.7%) and the coverage of this group by national data remainedhigh at 72.3%.

The global prevalence of anemia in preschoolaged children, pregnant women and nonpregnantwomen is 47.4%, 41.8%, and 30.2% respectively.These estimates, the number of individualsaffected, and the information from UN Regionsare presented in Table 1.2. Globally, 818 million(95% CI; 751885) women (both pregnant andnonpregnant) and young children suffer from ane-mia and over half of these, approximately 520

million (95% CI: 493547), live in Asia. Thehighest prevalence of anemia is in Africa for allthree population groups, but the greatest numberof people affected are in Asia, where 58.0%,56.1%, and 68.0% of the global anemia burdenexists in preschool aged children, pregnantwomen and nonpregnant women respectively.

Anemia as a public health problemAnemia is a worldwide public health problem.More than half the world's population of pre-school aged children and pregnant women residein countries where anemia is a severe publichealth problem (56.3% and 57.5% respectively)(Table 1.3). The proportion is lower for nonpreg-nant women of childbearing age, but still signifi-cant (29.6%). The degree of severity of the publichealth problem by country for preschool agedchildren, pregnant women, and nonpregnantwomen is presented in Figures 1.11.3. Countrieswith anemia as a severe public health problemwere grouped in Africa, Asia, and Latin Americaand the Caribbean.

DISCUSSION

Globally, almost half of preschool aged childrenand pregnant women and close to one third ofnonpregnant women suffer from anemia. Since alarge segment of the population is covered byactual survey data (69.076.1%), these estimatesare likely to reflect the actual global prevalence of anemia for these population groups. However,UN regional estimates may be more accurate for some populations and in some areas since the coverage varies significantly among regions.For all three groups, the coverage is the greatest in three UN regions: North America, Asia, andAfrica. North America is the best covered, but the number of countries in this region is muchlower (two countries) than in the African (53countries) and Asian (47 countries) UN regions.The coverage in the European and Oceania


regions is low, where data are available for lessthan one quarter of the population in all groups. InLatin America and the Caribbean, coverage forpreschool aged children is similar to coverage inAsia or Africa, but for pregnant and nonpregnantwomen it is about half the coverage found in Asiaand Africa.

The pattern of anemia prevalence by region issimilar for the three groups, Africa and Asia beingthe most affected. These regions are the poorestand this may therefore reflect the link betweenanemia and development. Compared to NorthAmerica, anemia is three times more prevalent inEurope. One reason may be that the Europeanregion includes countries with a range of socialand economic profiles, especially in the Easternsubregion. However, the difference remains whenNorth America is compared to Western or North-ern Europe, where the countries have more simi-lar economic profiles to those in North America(data not shown). It may also be the result of thelow coverage of anemia survey data in Europecompared to North America. Finally, it could bethat in North America foods are widely fortified

with iron and a high proportion of iron intakecomes from fortified foods (15).

The current estimates are the first to utilizenationally representative data for China, whichaccounts for 20% of the global population. Fur-thermore, the majority of the surveys used arenationally representative, which was not the casefor previous estimates. Surveys are also based onlarger sample sizes than many of the previousestimates. For example, the median sample size inour estimates was 2,580 preschool aged children,611 pregnant women, and 4,265 nonpregnantwomen; while in the DeMaeyer estimates, themedian number of subjects was 500 for all popu-lation groups (2). Finally, in these estimates, weused regression-based equations to generate esti-mates for countries with no eligible data, consid-ering the country's health and development situa-tion. In previous estimates, neighbouring countryinformation or regional estimates were applied tocountries without data.

These estimates are not quantitatively compa-rable to previous estimates since the methodolo-


Level ofpublic health

problem

PreSAC2 PW NPW

# of countries

Total population1000's (%)3

# of countries

Total population1000's (%)

# of countries

Total population1000's (%)

None 2 20570(3.3) 00

(0.0) 14171(0.3)

Mild 40 40921 (6.6) 3311156 (8.3) 59

647857(41.7)

Moderate 81 208472 (33.7) 9146162(34.2) 78

441285(28.4)

Severe 69 348322 (56.3) 6877466(57.5) 54

459518 (29.6)

192 618285 192 134783 192 1552831

Table 1.3: Anemia as a public health problem1 in WHO Member States.

1 The prevalence of anemia as a public health problem is categorized as follows:


Figure 1.1: Anemia as a public health problem in preschool aged children.

Figure 1.2: Anemia as a public health problem in pregnant women.

Normal (_40.0%) No Data


The boundaries and names shown and the designations used on this map do not imply the expression of any opinion whatsoeveron the part of the World Health Organization concerning the legal status of any country, territory, city or area or of its authorities,or concerning the delimitation of its frontiers or boundaries. Dotted lines on maps represent approximate border lines for whichthere may not yet be full agreement. WHO 2005. All rights reserved


gies used are so different. However, it is interest-ing to note that DeMaeyers anemia estimates(which excluded China) were 43% for preschoolaged children, 51% in pregnant women, and 35%in all reproductive aged women. When weexclude China from our estimates, the prevalenceof anemia is respectively 52%, 44%, and 34%.The variation in methods and a larger number ofnationally representative surveys in the currentestimates compared with previous estimates maybe responsible for these differences.

In 1992, WHO published the anemia preva-lence estimates of 37%, 51%, and 35% for allwomen, pregnant and nonpregnant respectively.These estimates included subnational data forChina. Current estimates of 31%, 42%, and 30%are lower, but this change may be accounted forby the considerable difference in methods andcoverage of national surveys.

The current estimates do have some limita-tions. Firstly, we treated all surveys as equal

despite the fact that quality varies greatly andadjustments are made in some surveys for popula-tion representativeness, smoking or altitude, butnot in others. Secondly, some surveys coveredonly a segment of the population group (e.g.,3.004.99 years for preschool aged children), butwe used these data to make an estimate for theentire population group. Similarly, the estimatesfor pregnant women do not take into account thetrimester assessed by the surveys since it is rarelyreported. However, this could affect the estimateof anemia prevalence, since prevalence is likelyto be lower early in pregnancy. Also, subnationaldata were treated equally to national data eventhough they may actually under- or over-estimatethe prevalence of anemia depending on the rea-sons for which the survey was conducted. Finally,we had to adjust hemoglobin concentrations forcountry estimates which did not present theprevalence of anemia for the appropriate thresh-old and we based this adjustment on normalhemoglobin distribution. In fact, the distributionmay be negatively skewed in populations with a


Figure 1.3: Anemia prevelance in non-pregnant women.



high prevalence of anemia and we may haveslightly over-estimated the anemia prevalence inthese populations.

These data on anemia prevalence are based onthe best available information and they are a goodstarting point to track progress in eliminating ane-mia. However, additional information wouldallow interventions to be more targeted and spe-cific.

Anemia in children less than two years of ageis of greatest concern since their rapid growthrequires a high intake of iron which is frequentlynot covered by their diet. It was not possible toestimate the anemia prevalence in this group sep-arately because of insufficient data. However,given that almost half the global population ofpreschool aged children suffer from anemia, witha prevalence as high as 64.6% in Africa and47.7% in Asia, and that we know anemia preva-lence is higher in the group of children less thantwo years old, we would expect that anemia inthis age group is a major global public healthproblem, especially in low income countries.

Based on these estimates of anemia preva-lence, the magnitude of nutritional anemia oreven of iron deficiency anemia is difficult toassess since most anemia surveys in the WHOVMNIS do not address the causes of anemia andare restricted solely to measuring hemoglobin.More specifically, few surveys provide informa-tion on iron deficiency, on the relative proportionsof anemia with concomitant iron deficiency, or oniron deficiency with concomitant anemia. Thesurveys that do provide information on iron defi-ciency often use different indicators and thresh-

olds. The assumption in designing anemia sur-veys is that iron deficiency is the main cause ofanemia and therefore anemia prevalence can beused as a proxy for iron deficiency prevalence.Previously, the US NHANES 197680 data wereutilized to estimate iron deficiency from theprevalence of anemia (3, 16) and it has been suggested that when anemia prevalence is 20%,iron deficiency exists in 50% of the population,and when anemia prevalence is greater than 40%,the entire population suffers from some degree of iron deficiency. This assumption may apply tocountries with a high prevalence of anemia andiron deficiency where the primary cause of ane-mia is iron deficiency, but does not necessarilyhold in situations where the prevalence of anemiaand iron deficiency are low or where factors otherthan iron deficiency (other nutritional deficien-cies, malaria, infections) cause anemia (17).

In spite of its limitations, anemia prevalencedata remains an important indicator of publichealth since anemia is related to morbidity and mortality in the population groups usuallyconsidered the most vulnerable preschool agedchildren and pregnant women. At a global level,anemia prevalence is a useful indicator to assessthe impact of widespread or highly effective interventions and to track the progress madetowards the goal of reducing anemia in pregnantwomen and preschool children by one third thatwas adopted by the UN Special Session on Children in 2002 (18). However, in order to makefull use of these prevalence data, information on the cause of anemia should be collected in any anemia survey so that interventions for anemia control can be better adapted to the localsituation and can therefore be more effective.


1. World Health Organization. The world healthreport 2002: reducing risks, promoting healthy life.Geneva: WHO, 2002.2. DeMaeyer E, Adiels-Tegman M. The preva-lence of anaemia in the world. World Health Stat Q1985;38:30216.3. UNICEF/UNU/WHO. Iron deficiency anaemia:assessment, prevention, and control. A guide for pro-gramme managers. WHO/NHD, 2001 [report no. 01.3].4. Royston E. The prevalence of nutritionalanaemia in women in developing countries: a criti-cal review of available information. World HealthStat Q 1982;35:5291.5. World Health Organization. The prevalence ofanaemia in women: a tabulation of available infor-mation. WHO/MCH/MSM,1992 [report no. 92.2]. 6. Hurtado A, Merino C, Delgado E. Influence ofanoxemia on haematopoietic activities. Arch IntMed 1945;75:284323.7. Nordenberg D, Yip R, Binkin NJ. The effect ofcigarette smoking on hemoglobin levels and anemiascreening. JAMA 1990;264:15569.8. Lohr SL. Sampling: design and analysis.Pacific Grove, CA: Duxbury Press, 1998.9. Wackerly D, Mendenhall W, Scheaffer RL.Mathematical statistics with applications. PacificGrove, CA: Duxbury Press, 2001.10. United Nations Development Programme.Human development indicators. In: Murphy C,Ross-Larson B, eds. Human development report2004. New York: UNDP, 2004:139250.

11. World Health Organization. The world healthreport 2004 changing history. Geneva: WHO,2004.12. World Health Organization. The world healthreport 2000 health systems: improving perfor-mance. Geneva: WHO, 2000.13. Mathers CD, Loncar D. Projections of globalmortality and burden of disease from 2002 to 2030.PLoS Med. 2006;3:e442.14. United Nations Population Division. WorldPopulation Prospects - the 2004 revision. New York:UN, 2005.15. Hurrell RF, Jacob S. Role of the food industryin iron nutrition: iron intake from industrial foodproducts. In: Hallberg L, Asp NG, eds. Iron nutritionin health and disease. London: John Libbey & Co.,1996:33947.16. Dallman PR, Yip R, Johnson C. Prevalence andcauses of anemia in the United States, 1976 to 1980.Am J Clin Nutr 1984;39:43745.17. Asobayire FS, Adou P, Davidsson L, Cook JD,Hurrell RF. Prevalence of iron deficiency with andwithout concurrent anemia in population groupswith high prevalences of malaria and other infec-tions: a study in Cote d'Ivoire. Am J Clin Nutr2001;74:77682.18. United Nations Children's Fund. A world fit forchildren: millennium development goals; specialsession on children documents; the convention onthe rights of the child. New York: UNICEF, 2004.


REFERENCES

2The case for urgent action to address nutritional anemia

M.G. Venkatesh Mannar

The Micronutrient Initiative, Ottawa, CanadaContact: [email protected]

VENKATESH MANNARVenkatesh has a Masters degree in chemical engineering and extensive experience with planning and implementing programs to eliminatemicronutrient deficiencies. He is currentlyPresident of the Micronutrient Initiative (MI), aninternational nonprofit agency based in Canada.His role is overseeing the implementation of MIsglobal mandate to support national actions to eliminate micronutrient malnutrition. He has aspecial interest in salt iodization and serves on the International Council for the Control of Iodine Deficiency Disorders, and the Network forthe sustained Elimination of Iodine Deficiency, aswell as being a member of the Leadership Groupof the Flour Fortification Initiative.

The continued persistence of anemia in manyparts of the world at high levels is a challenge thatneeds to receive the highest priority for attentionand action. Despite the magnitude of the anemiaproblem, and the constantly expanding body ofresearch findings relating to pathogenesis, riskfactors, and efficacious interventions, coverage ofinterventions remains poor.

In part this is because there is no easy solutionand the approaches that are effective have consid-erable drawbacks. However, lack of priority mayalso be an underlying reason for the lack ofprogress. For many decades iron deficiency hasbeen seen as a womens problem. Moreover, itis so common, affecting up to three-quarters ofpregnant women in some parts of Asia for exam-ple, that it has come to be regarded as normal.Among policy makers the unspoken view hasseemed to be that women somehow cope and thatiron deficiency is not enough of a problem to jus-tify a major national effort to reduce it.

This view, never justifiable, must surely nowgive way under the weight of evidence linking irondeficiency in early childhood to substantial effectson physical and intellectual development. In mostdeveloping countries today, iron deficiency is esti-mated to be preventing 40 to 60% of the populationfrom reaching their full mental potential.

Several recent global reviews have under-scored the urgency to act to address iron defi-ciency and anemia:

In the World Health Report 2002, which quanti-fied the most important risks to health, iron defi-ciency was identified among the 10 most seriousrisks in countries with high infant mortality cou-pled with high adult mortality (1). The samestudy found that particular measures to addressiron deficiency anemia are among the most cost-effective public health interventions.

The Copenhagen Consensus 2004 panel ofeminent economists, including several NobelLaureates, concluded that the returns of invest-ing in micronutrient programs (includingiron), among a list of 17 possible developmentinvestments, are second only to those of fight-ing HIV/AIDS. The benefit-to-cost ratio ofiron interventions based on resource savings,improvement in cognitive development andschooling, and physical productivity was esti-mated to be as high as 200:1 (2).

The Disease Control Priorities Project (3) hasalso highlighted the cost-effectiveness of ironinterventions (66 to 70 USD per DALYaverted for iron fortification).

THE PROGRESS

Interventions to address nutritional anemias tradi-tionally have focused on providing iron and folicacid supplements principally to pregnant womenand, to a much lesser extent, children under 2 yearsof age. Iron supplementation in controlled experi-ments has proven highly efficacious where anemiais not exacerbated by parasitic infection or malaria.However, those analyses which have attempted toassess effectiveness in field settings have generallyfailed to show significant reductions in anemiaprevalence. Further the effectiveness of large-scalefood fortification programs has not been systemati-cally documented so far, notwithstanding the factthat some of these programs have been ongoing for more than 50 years. Data on the conditionsunder which food fortification can reduce iron deficiency is still lacking. Overall the goal of reduc-ing by one-third the prevalence of anaemia,including iron deficiency, by 2010 (4) in womenand children is unlikely to be met unless westrengthen field application of supplementationefforts coupled with other creative means ofincreasing the iron content of diets and enhancingits absorption.

14 M.G. Venkatesh Mannar

INTRODUCTION

Fortunately over the past 10 years there hasbeen a significant scale-up and intensification ofefforts in several countries to address anemia and particularly iron deficiency through supple-mentation and food fortification. These impartconfidence that systematic application of knowninterventions can significantly reduce anemia infield settings and be sustained on a population-wide basis. Key developments are summarizedbelow:

There is as we will hear today a better techni-cal consensus on key issues. There is a betterunderstanding of the conditions under whichsupplementation programs can be effective.There is sufficient knowledge and experiencewith iron supplementation for pregnant womento design and implement effective national pro-grams. Programmatic and technical guidelines(5) reflect a much larger body of experience toguide effective programming.

In food fortification we have much better in-formation and guidelines today regarding ironcompounds that are stable and bioavailable (6).Work continues to refine our knowledge onwhat iron compounds work best under a specificset of conditions. WHO has just published com-prehensive guidelines for food fortification withmicronutrients, including iron (7).

The food industry (especially the cereal flourindustry) is aware of the problem and large-scale fortification has commenced in severalparts of the world, especially Latin America,the Middle East/North Africa, and South Eastand Central Asia. More than 63 countries arefortifying all or some of their flour with iron,folic acid and other nutrients, covering 25% ofthe flour that is milled in large roller mills (8).Fortification with iron is also being extendedon a large scale to cereal flour derivativesincluding processed baked products, comple-mentary foods, noodles, and pastas.

Workable strategies and interventions nowexist to provide iron to children under twoyears of age through fortified complementary

foods as well as multinutrient premixes foraddition in the home (9).

Considerable efforts have been invested by theMicronutrient Initiative, ETH Zurich, and otherorganizations in developing and recently takingto scale the double-fortification of salt. Technol-ogy is now available for a stable encapsulatediron premix that can be easily added to iodizedsalt. By using the capacity and delivery systemsalready established during the push for universalsalt iodization, double-fortified salt is alreadybeing produced and distributed through com-mercial channels and also through public pro-grams to reach economically weaker sections ofthe population. Double-fortified salt couldpotentially reach more than 1 billion peoplearound the world (10).

Since 1999 the Chilean National SupplementaryFeeding Program (PNAC) has been providing 2kg/month of iron-fortified full-fat powderedmilk to 70% of infants 12 to 18 months of age,resulting in a significant decrease in anemia pre-valence. Lactating mothers consume it untiltheir infants are weaned.

Condiments fortified with iron (soy and fishsauces) are emerging as a major vehicle inparts of Asia (China, Vietnam, and Thailand).In China, through a project supported by theGlobal Alliance for Improved Nutrition(GAIN), fortified soy sauce is already reach-ing nearly 12 million people.

Technologies to fortify rice with iron and folicacid have been developed and successfullytested for acceptability and efficacy (11).

There is a growing body of work led by the Inter-national Food Policy Research Institute (IFPRI)and HarvestPlus related to improved varieties ofstaple crops such as rice, wheat, and beans withhigh iron content. Polished rice with 8 ppm iron(16 times the level in current commercial vari-eties) and 20 ppm zinc has been identified ingermplasm collections at the International RiceResearch Institute in the Philippines. Initial stud-ies using the rice showed efficacy in improvingiron stores of women with iron-poor diets (12).


High-iron beans with iron levels up to 127 mg/ghave been identified. The first set of selectedmaterials combining high iron with superioragronomic attributes, e.g. drought tolerance,was recently distributed to African partners forlocal evaluation (13).

The intersection between iron status and infec-tion is also being given more attention today andpublic health measures are beginning to have animpact on anemia levels in a few parts of theworld.

These efforts must be supported by anincrease in the amount of attention given to com-munication, for example in the form of com-pelling advocacy to those policy makers capableof mobilizing resources. It is equally important tofoster strategic alliances, reach agreement on pri-ority target groups, and support behavioralchange. The social marketing perspective is alsocritical for iron, particularly when new andaffordable iron-containing products are intro-duced in a society or group.

WHAT WOULD IT TAKE TO FIGHTIDA MORE EFFECTIVELY?

A basic yet formidable challenge is putting iron on the agenda of policy makers and development agencies nationally and globally. Creating aware-ness, building alliances, and mobilizing actors at alllevels and sectors is critical. Iron needs global cham-pions to communicate the need and urgency foraction.

While the starting point is to address keyissues and propose consensus statements that pro-vide clarity to policy makers and program plan-ners in order to strengthen and expand programs,the information that a policy maker needs doesnot flow automatically from scientific consensusand technological feasibility. We need to buildmuch better bridges between those that have thescience and technology, those that deliver the

services, and those that have the power to makethe political and financial decisions.

The main challenge in food fortification is toenable universal fortification of staple foods orcondiments with meaningful levels of iron. Thisinvolves systematic planning and collaborationwith the food processing industry in addressingissues of coverage, cost, effectiveness, benefits,and risk in relation to gains, performance of thetechnology within the environmental, socioeco-nomic and cultural context of the recipient popu-lation, the identification of intended beneficiaries,their needs, and their social and economic cir-cumstances. Parallel with improving bioavailabil-ity of iron compounds (including encapsulatedforms of iron), development of other strategies toeffectively improve iron utilization from the dietalso needs increased attention. Gaining betterknowledge of interactions among various micro-nutrients (e.g. iron and zinc, vitamins limitingiron absorption and mobilization) and otherdietary compounds, and of other nutritional andnon-nutritional causes of anemia, pose as chal-lenges to be overcome for enhanced effectivenessof interventions.

In the area of iron supplementation, in addi-tion to timely supply of good quality supplementsand effective delivery systems, there is also the challenge of ensuring high compliance for the supplementation programs to be successful.Ensuring effectiveness through improved pro-gramming and assured provision of higher qualitysupplements to target groups and addressing fac-tors that limit motivation to take supplementshave the largest potential to improve programeffectiveness.

The potential of the food industry to createnutritious complementary foods for young chil-dren has hardly been tapped and has yet to bedeveloped. Industrially produced fortified com-plementary foods are recommended by pediatri-cians worldwide as an important part of a nutri-


tionally adequate diet for infants (complementaryto breast milk and home-prepared complementaryfoods)especially to meet the micronutrient re-quirements of infants for iron and zinc. Beyondhaving a superior micronutrient content to that ofhome-prepared rice porridge and other traditionalinfant foods, industrially fortified complementaryfoods also have the advantages of deliveringhigher bioavailability of micronutrients, higherenergy density, and higher protein quality, all in asafe and convenient manner. From the food tech-nology perspective, the challenge is to increaseboth the energy density of complementary foodsand levels of iron and other nutrients (and elimi-nate absorption-inhibiting factors) at an afford-able price. From the public health perspective, weneed a combination of proper regulation that pro-tects infant health yet supports industrial innova-tion, and strong public education on appropriatepractices of feeding and caring for infants andyoung children. Large and rapid growth in theproduction and consumption of fortified comple-mentary foods will be possible only through aneffective public-private social marketing partner-ship to increase the percentage of infants andyoung children who are fed fortified complemen-tary foods and promote the use of fortified com-plementary foods only in the latter half of infancyand the second year of life.

While recent studies have questioned the safety of iron supplementation for young childrenin malaria-endemic areas, a preliminary statementfrom WHO has recommended that the con-clusions not be extrapolated to fortification orfood-based applications where patterns of ironabsorption and metabolism may be substantiallydifferent (14).

THE WAY FORWARD

Clearly a more effective anemia strategy willinvolve concurrent efforts to address inadequateiron intake and to reduce concurrent infections,

particularly parasitic and malarial. Given thehighly regulated nature of iron absorption, no sin-gle intervention will revert and prevent IDA inany given population. A combination of interven-tions need to be universally advocated and imple-mented including supplementation of at-risk groups,universal and targeted fortification, dietary modi-fication, parasitic disease and malaria control, andvitamin A interventions, in addition to overalleducation of policy makers, professionals and the public. The proper combination of effectivestrategies will vary according to each countrysepidemiological, socioeconomic, political andcultural context.

Programmatically, the priority in globalefforts to increase the iron intakes of vulnerablepopulations almost certainly should be given tonational scale programs for:

Fortification of staple foods, condiments, andcomplementary foods with bioavailable formsof iron (care is needed in selecting the com-pound and the level of fortification);

Iron supplementation programs for the highestpriority population groups: pregnant women,children under two and adolescent girls.

While biofortification of crops with iron is not yet ready for widespread application, it isclearly one of the interventions of the future andneeds to be aligned with crop improvementefforts for key staples and promotion/adoption atthe regulatory, farmer and consumer level.

Country strategies must be tailored to suitspecific national paradigms, combining comple-mentary and effective interventions that willresult in the most cost-effective model. In thewider picture there is certainly need to establishand scale up effective technologies (especially inusing iron compounds that do not impart color orreact with food matrices while offering a goodbioavailability). Beyond the technology we needto tackle on a parallel track the operational con-siderations related to making programs work in


1. World Health Organization. The world healthreport 2002: reducing risks, promoting healthy life.Geneva, Switzerland: World Health Organization,2002.2. Behrmann JR, Alderman H, Hoddinott J.Hunger and malnutrition. Copenhagen ConsensusChallenge Paper, 2004.3. Disease Control Priorities in DevelopingCountries. 2nd ed. New York: Oxford UniversityPress, 2006.4. United Nations. Report of the 27th session ofthe UN General Assembly Special Session onChildren. 2002.5. USAID. Anemia guidance manual. 2003.6. Flour Fortification Initiative. Wheat flour forti-fication: current knowledge and practical applica-tions. Summary report of an International TechnicalWorkshop: Cuernevaca (Mexico), 2004.7. World Health Organization. Guidelines for foodfortification with micronutrients. Geneva, Switzerland:World Health Organization, Department of Nutritionfor Health and Development. 2006.8. Flour Fortification Initiative. Tabular overviewof progress in flour fortification. In press 2006.9. Zlotkin SH, Schauer C, Christofides A, SharieffW, Tondeur MC, Hyder SMZ. Micronutrient sprinklesto control childhood anemia. PLoS Med 1 (1):e1(2004)10. The Micronutrient Initiative. Double fortifica-tion of salt: a technical breakthrough to alleviate iron

and iodine deficiency disorders around the world.Ottawa: The Micronutrient Initiative, 2005.11. Moretti D, Lee TC, Zimmerman MB, Nuessli J,Hurrell RE. Development and evaluation of iron-fortified extruded rice grains. J Food Sci 2005;70:4.12. Haas Jere D, Beard JL, Murray-Kolb LE, delMundo AM, Felix A, Gregorio GB. Iron-biofortifiedrice improves the iron stores of non-anemic Filipinowomen. J Nutr 2005;135:28232830.13. HarvestPlus, www.harvestplus.org.14. World Health Organization. Iron supplementa-tion of young children in regions where malariatransmission is intense and infectious disease highlyprevalent. WHO Statement. Geneva: World HealthOrganization, 2006.15. Beaton GH, McCabe GP. Efficacy of intermit-tent iron supplementation in the control of iron defi-ciency anaemia: an analysis of experience in devel-oping countries. Ottawa: The Micronutrient Ini-tiative, 1999.16. Horton S, Ross J. The economics of iron defi-ciency. Ottawa: The Micronutrient Initiative, 2000. 17. Committee on Micronutrient Deficiencies,Institute of Medicine. Prevention of micronutrientdeficiencies: tools for policymakers and publichealth workers. Howson CP, Kennedy ET, HorwitzA, eds. Washington, DC: National Academy Press,1998.

communities where iron-deficient people live.Issues of demand, supply and logistics, communi-cations and community participation, partnershipbuilding across a wide spectrum of players pub-lic and private are equally important to ensure

the success and sustainability of efforts to elimi-nate anemia and iron deficiency in large popula-tions. Many of these needs interact and are mutu-ally reinforcing.


REFERENCES

3The economics of addressing

nutritional anemia

Harold Alderman1

1World Bank, Washington DC, USA2Wilfrid Laurier University, Ontario, Canada

Contact: [email protected]

Susan Horton2

HAROLD ALDERMANHarold has a PhD in economics from HarvardUniversity and an MS in Nutrition from CornellUniversity, both in the USA. He is currentlySocial Protection Advisor for the Africa Region of the World Bank where he has worked for thelast 15 years. Harolds research focus has been onthe economics of nutrition interventions and foodpricing policies. Recent studies include estimatesof the economic returns from investment in nutri-tion and evaluations of program impacts.

SUSAN HORTONSusan has a PhD in economics from HarvardUniversity, USA. She is currently Vice President:Academic at Laurier University in Canada.Susans main area of specialization is in healthand labor market issues in developing countries.These include economics of health, nutrition, house-hold use of time, labor markets, and poverty andshe has worked and researched extensively indeveloping countries across the world.

There is abundant evidence that iron deficiencyanemia is associated with a range of health conse-quences, as reviewed in the various chapters of thisvolume. Similarly, evidence from various clinicaland field trials shows the potential for practicalinterventions to reduce some of these undesirablehealth outcomes. This chapter addresses thequestion of how programs aiming to diminish therisks of poor health associated with anemia can beevaluated in economic terms.

The first step in such an evaluation is to statethe costs of the iron deficiency anemia in dollarterms so as to have the consequences in a unit ofmeasurement that is in common with other claimson public resources. This differs from the calcula-tion of effectiveness of an intervention in terms ofnatural units (such as increases in life expectancy)or in terms of disability adjusted life years(DALYs), a composite measure that combinesyears lived with disability and years lost to pre-mature death in a single metric (1).

Using an economic denominator facilitatesaggregation of different benefits from the sameintervention, for example, with a deworming program that reduces anemia as well as facilitatesponderal growth and vitamin A absorption. Moreimportant in many contexts is the fact that such a metric permits the comparison of health pro-grams with interventions outside the health area(2).

There are two general approaches to suchestimates of economic benefits. First, one can cal-culate what the expected gains in economic termswould be if a case of anemia were prevented.Alternatively, one could estimate the impact onGNP if anemia could be reduced. The latter dif-fers from the former in that it scales the individualgains by the prevalence rate of anemia, and hasthe strong advantage of motivating political will.However, the former is most amenable to com-

parisons of intervention costs and expected bene-fits, as discussed below.

The economic gains from reducing anymicronutrient deficiency can come from both costreductions (say, by reducing the costs associatedwith mortality or morbidity) or from enhancedproductivity. At least six distinct categories ofeconomic benefits from improved nutrition canbe identified:

1) Reduced infant and child mortality. 2) Reduced costs of health care for neonates,

infants, and children.3) Productivity gain from improved physical

capacity.4) Productivity gain from increased cognitive

ability.5) Reduction in costs of chronic diseases.6) Intergenerational benefits through improved

health.

Placing precise numbers on the economic valuefrom any one of these benefits involves a range ofassumptions and adaptations to the country context.A particularly vexing problem is how to quantifythe economic cost of early mortality. Most simply,this can be based on the expected lifetime earningsof the individual. Other approaches are linked to therevealed behavior of either individuals or govern-ments. For example, the value of a statistical life(VSL) can be derived from the premium paid to aworker to induce him or her to accept employmentthat incrementally increases the risk of mortality(3). The magnitude of the higher wage relative tothe decrease in life expectancy provides an estimateof how the employee values the risk. Typically,such estimates in dollar terms are in the seven dig-its range for a lifetime or over $100,000 (US) peryear of life expectancy. Estimates are higher inwealthier countries, although the proportionalincrease in the VSL is generally less than theincrease in national income. An alternative

H. Alderman and S. Horton20

INTRODUCTION

approach to valuing reduced mortality is based onthe behavior of governments. In particular, theresources actually used in a society to avert adeath provide an estimate of the average valuethat the public places on averting a death (4).

These two approaches generally result in esti-mates that are far apart, but as they measure dif-ferent things they cannot be directly compared.Due in part to the inherent limitations of suchmethodologies, many approaches to estimatingthe economic benefits of nutrition interventionsonly indicate costs in terms of productivity andthey can be considerable in their own right, as dis-cussed in Section 2. Other studies report sensitiv-ity estimates and details on the underlyingassumptions, so that it is possible to see if the eco-nomic rationale for an investment changes over areasonable range of presumed values for deferredmortality (5). A further and flexible alternative isto provide results in terms of DALYs and then toconvert the DALYs into dollar terms using arange of estimates. One study suggests a rangebetween $1,000 and $5,000 (6, see also 7).

As with any analysis of causality, it is neces-sary to distinguish the specific consequences ofanemia from its correlates when determining theexpected benefits from a specific intervention.This is less of an issue with respect to contempo-raneous impacts of anemia on productivity, sincethere are experimental approaches that have beenused to directly assess changes in productivity. Insuch cases, however, it is still important to deter-mine the incentive structure a beneficiary faces; acapacity for increased work does not necessarilytranslate into increased effort unless there areincentives for the worker to increase perfor-mance. Still, it is comparatively straightforwardto assign a value to the output from increasedeffort, controlling for economic context.

The impact of improved iron status duringchildhood on subsequent adult productivity, how-ever, is seldom obtained directly from experimen-tal evidence. In the absence of longitudinal stud-ies that track experimental interventions overdecades, in order to estimate the economic impactof increased cognitive development due to sup-plementation or fortification in childhood, it isnecessary to draw upon the general literature onproductivity enhancement. For a given change inan indicator of cognitive ability, the change onsubsequent earnings must be estimated using amaintained assumption that the relationshipbetween IQ or similar cognitive measures andearnings is not affected by the type of interventionthat influences the indicator.1

One such estimate of the impact of IQ onearnings, conditional on years of schooling, usesdata from the United States (8). For men, theimpact of a half standard deviation decline in IQon the logarithm of wages was 0.05, or slightlymore than the impact of an additional year ofpost-secondary schooling. Using the same dataset, but a different measure of ability, anotherstudy shows that the net impact of ability is boththe direct impact on wages as well as the impactthat works through schooling choices (9). Usingthe methodology of the latter study disaggregatedby gender and ethnic group, as well as includingother background variables, but without school-ing, a half standard deviation decline in cognitiveability leads to 812% lower wages.

A different measure of cognitive ability per-formance on Ravens matrices was used in astudy of wages in rural Pakistan (10) that found asimilar pattern in the effect of ability on wageswith and without controlling for the impact ofability on schooling. This study found that a halfstandard deviation decline in this measure leads

21The economics of adressing nutritional anemia

1 Non cognitive skills may have as much, or more, impact on earnings. However, while these may be influenced by developmentalprograms, it is less clear that they are malleable to micronutrient interventions.

to a 6.5% reduction in wages, in estimates that do not include schooling in the regression. Thepoint estimate drops by two thirds in estimatesthat include both years of schooling as well as achievement in school, both of which are indirectly affected by ability. Overall, these and similar studies cited in Behrman, Alderman andHoddinott (2) imply an impact of a half standarddeviation change in IQ on earnings in the neigh-borhood of 5%.

Since improvements are the result of inter-vention in childhood, for any comparison of theprogram costs of such benefits, it is necessary toaccount for the time lag between the interventionand the stream of benefits. That is, unlike thedirect productivity effects due to increased workcapacity following iron fortification or supple-mentation, there is typically a 1015 year lagbetween interventions that increase the cognitivecapacity of children and the stream of benefits,which commences only when they enter the labormarket and which continues for their whole work-ing life. Benefits incurred at different times haveto be given different economic values, due to thefact that monetary intervention carries a greaterlong-term impact if it takes effect early in liferather than later. This is also because the sooner itis obtained, then economic benefits can be rein-vested and further productive returns gained.

Unfortunately, although the rationale for dis-counting future benefits is not in dispute, there isno unambiguous way to determine this discountrate. For example, DALYs are typically calcu-lated using a fairly low discount of 3% per year(1) while rates of 10% have been proposed forWorld Bank investments (11). All discount ratesreduce the present value of future benefits relativeto current costs; the larger the discount rate, thegreater the importance given to immediatereturns.

A final consideration before discussing thepotential economic benefits as well as the cost of

interventions in addressing iron deficiency ane-mia is whether low income countries can expectthat robust income growth will address the needfor interventions at all. Even though it is currentlywidespread, if anemia were to decline rapidly inconcurrence with progress in providing equitablegrowth, it might be a moot point to prioritize pro-grams which tackle iron deficiency. Unfortu-nately, while such a viewpoint is often expressedin regards to malnutrition in general, the Millen-nium Development Goal (MDG) of reducingunderweight is unlikely to be met with incomegrowth (12): Iron deficiency anemia is even lessresponsive to economic growth.

For example, for every 10% increase inincome per capita , the percentage of underweightchildren declines by 5%; a similar estimate under-taken for this paper using data from Mason,Rivers and Helwig (13) shows that a 10%increase in income per capita is associated with adecline of only 2.5% in prevalence rates of ane-mia in children 059 months old. This estimatealso reveals a slight trend in lower rates of anemiaover time, after controlling for national incomes.

Table 3.1 illustrates a similar difference be-tween the rates at which underweight and anemiadecline, using the example of India. As indicated,the share of children in the upper income quintilethat are underweight is less that half the share inthe lower income group. However, the percentageof children that are anemic declines by only 20%. There are few multivariate estimates of the deter-minants of anemia in the literature. One study,from Indonesia (14), looked at hemoglobin con-centration, which admittedly is not synonymouswith anemia rates. Still, this analysis found thatfor every 10% increase of household per capitaexpenditures, hemoglobin concentration in chil-dren increased by only 0.2%. Thus, it would takea fivefold increase of income to raise hemoglobinconcentration by 1 g/dL, an improvement that isfeasible in controlled trials of supplementationand fortification.


ESTIMATING THE COST OF ANEMIA

Section 1 identified six categories of costs of anemia. Of these, there are more estimates forthree of the categories: 1) - infant and child andmaternal mortality; 2) productivity gains fromimproved physical capacity; and 3) productivitygains from increased cognitive ability. The literature examining possible effects of iron onmorbidity is very recent; there is little on chronicdisease. We simply do not have results of ironintervention studies on young girls, which requireus to track them longitudinally in order to look at pregnancy outcomes and intergenerational benefits.

There are two key groups for whom iron hasbeen identified as affecting mortality, namelywomen in childbirth, and infants during the peri-natal period. In one study it is estimated that 20%of perinatal mortality, and 10% of maternal mor-tality in developing countries, is attributable toiron deficiency. The study suggests that 0.8 mil-lion (1.5%) of deaths worldwide are attributableto iron deficiency, 1.3% of male deaths and 1.8%of female deaths. Attributable DALYs are evengreater, amounting to the loss of about 35 millionhealth life years (2.4% of global DALYs). Ofthese DALYs, 12.5 million (36%) occurred inSEAR D, 4.3 million (12.4%) in WPR B, and 10.1

million (29%) in Africa (15). (Note that SEARrefers to Southeast Asia Region, and WPR refers to Western Pacific Region, using the WHOcategories). These estimates are based on theknown relation between maternal and perinatalmortality rates and hemoglobin levels in blood.Thus nutritional anemia is associated with a significant health burden. This could be convertedto an economic burden using methods suggestedin Section 1.

The cost of lost adult productivity associatedwith anemia has been investigated intensively.Intervention studies both in the laboratory and inthe field date back to the 1970s. Iron interventionsin well-controlled studies have clear impacts onphysical capacity (both maximal work capacityand endurance) and have been demonstrated tohave productivity impacts in the order of 5% onlight manual labor, and as high as 17% on heavymanual labor (16). A recent ongoing study inIndonesia confirms effects on the income of theself-employed as being as high as 20% for menand 6% for women, confirming the potentiallylarge impact on productivity and, by inference, onincome as well (17).

Productivity changes of this size have poten-tially important economic impacts in poor coun-tries where anemia levels are high, and manual

23The economics of adressing nutritional anemia

Income quintilesPercentage of children with

weight-for-age below 2 standard deviations

Percentage of children with anemia (HB

labor is a significant proportion of employment.Estimates of the annual costs per person of the current levels of anemia in nine developingcountries range from $0.62 to almost $4.00, usingdata from 1994. One caveat is that these are typi-cally labor-surplus economies. Arguably, in-creased productivity might not be as important asin a labor-scarce country; on the other hand theloss of productivity in physical activity outsidethe market economy (carrying water, collectingwood, etc.) might have further effects on the quality of life that market statistics on GDP do not capture.

Estimates of the effects of iron interventionon childrens cognitive ability and therefore sub-sequent adult economic productivity are morespeculative, relying on linking iron interventionsto mental and motor abilities in children under 2 and to cognitive ability in children over 2, andinferring the effects on adult productivity. Inter-vention studies suggest that there are adverseeffects of iron deficiency in children, with defi-ciencies in the second year of life perhaps beingvery critical: When anemia levels are particularlyhigh while growth is very rapid, there may poten-tially be irreversibilities. One can infer that theanemia potentially reduces adult earnings (due tocognitive effects) by 2.5%, based on a series ofassumptions.

A recent study of an iron intervention with a lon-gitudinal follow-up over a period of 18 years con-firms that these estimates may be reasonable (18).Children were enrolled in the study between theages of 1 and 2. Anemia was corrected with ironsupplements for those children in the interventiongroup. All children received cognitive tests inearly childhood (below age 2), and again at ages58, 1114, 1518, and 19. (The specific testsincluded Bayley tests in infancy, and age-appro-priate Wechsler tests and other cognitive tests atlater ages.) The results were not reported for theintervention group versus the control group, butrather for those without anemia at the end of the

intervention (either because anemia was cor-rected, or because they had not been anemic), ver-sus those whose anemia had not been corrected(i.e., a subset of the control group). The nonane-mia group contained a higher proportion of highersocioeconomic status (SES) households (sinceanemia status is somewhat correlated with SES). The results are quite striking. In the middle SESgroup, those with adequate iron status scored 89points above those who remained anemic in earlychildhood, a difference that was maintained up toage 19. In the lower SES group, those with ade-quate iron status scored 10 points higher in earlychildhood, a gap which widened to 25 points atage 19. These results suggest that iron deficiencybelow age 2 does indeed have significant effectson life course: a difference in cognitive score ofthis magnitude is known to have effects onschooling and on earnings.

The results are not as powerful as resultswhich directly compare intervention and control,since the measure of SES used does not perfectlycontrol for all other household factors whichaffect ability independent of iron status. Never-theless, the results are intriguing.

The estimated effect of a 2.5% loss of earn-ings due to cognitive losses in childhood for ninedeveloping countries implies an additional loss ofper capita of GDP which ranges from $0.211.88(these represent present values of discountedfuture lifetime losses (16). The total losses percapita (due to physical productivity as well ascognitive losses) amount to $0.834.81, or0.371.86% of GDP. The range is from approxi-mately 12% in the poor countries in South Asia,and 0.41% in the other developing countriesconsidered (in sub-Saharan Africa, Latin Amer-ica, and the Middle East). These losses amount tobillions of dollars annually, which is very consid-erable, especially when compared to the modestcosts of decreasing nutritional anemia discussedin the next section.


ECONOMIC COSTS OF ACHIEVINGREDUCTIONS IN NUTRITIONAL ANEMIA

In this section we first discuss five conceptual andpractical issues involved in calculating cost-effec-tiveness of interventions, before turning to theevidence. We focus largely on interventionsaffecting iron, as the most common nutrientwhose deficiency is associated with anemia. Lit-erature on the cost-effectiveness of other nutrientsaffecting anemi

nutritional anemia book

Documents

life publications

life pressco sight

life mission statementsight

copyright sight

lifelong burden

control of nutritional

scientific data

scientific information