45086 Environmental Health and Child Survival Epidemiology, Economics, Experiences E N V I R O N M E N T A N D D E V E L O P M E N T Environmental Health and Child Survival E N V I R O N M E N T A N D D E V E L O P M E N T A fundamental element of sustainable development is environmental sustain- ability. Hence, this series was created in 2007 to cover current and emerging issues in order to promote debate and broaden the understanding of environ- mental challenges as integral to achieving equitable and sustained economic growth. The series will draw on analysis and practical experience from across the World Bank and from client countries. The manuscripts chosen for publi- cation will be central to the implementation of the World Bank's Environment Strategy, and relevant to the development community, policy-makers, and academia. Topics addressed in this series will include environmental health, natural resources management, strategic environmental assessment, policy instru- ments, and environmental institutions. Also in this series: International Trade and Climate Change: Economic, Legal, and Institutional Perspectives Poverty and the Environment: Understanding Linkages at the Household Level Strategic Environmental Assessment for Policies: An Instrument for Good Governance Environmental Health and Child Survival Epidemiology, Economics, Experiences © 2008 The International Bank for Reconstruction and Development/The World Bank 1818 H Street, NW Washington, DC 20433 Telephone 202-473-1000 Internet www.worldbank.org E-mail feedback@worldbank.org All rights reserved. 1 2 3 4 :: 11 10 09 08 This volume is a product of the staff of the International Bank for Reconstruction and Development / The World Bank. The findings, inter- pretations, and conclusions expressed in this volume do not necessarily reflect the views of the Executive Directors of The World Bank or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other informa- tion shown on any map in this work do not imply any judgement on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. R I G H T S A N D P E R M I S S I O N S The material in this publication is copyrighted. Copying and/or transmit- ting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923, USA; telephone: 978-750- 8400; fax: 978-750-4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street NW, Washington, DC 20433, USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. ISBN-13: 978-0-8213-7236-4 eISBN-13: 978-0-8213-7237-1 DOI: 10.1596/978-0-8213-7236-4 Library of Congress Cataloging-in-Publication Data Environmental health and child survival : epidemiology, economics, expe- riences. p. ; cm. -- (Environment and development) Includes bibliographical references and index. ISBN 978-0-8213-7236-4 1. Environmentally induced diseases in children--Developing countries. 2. Malnutrition in children--Developing countries. I. World Bank. II. Series: Environment and development (Washington, D.C.) [DNLM: 1. Child, Preschool. 2. Environmental Health. 3. Cost of Illness. 4. Developing Countries. 5. Disorders of Environmental Origin. 6. Malnutrition. WA 30.5 E605 2008] RJ383.E583 2008 618.92'98--dc22 2008022136 Cover photo: World Bank Photo Library Cover design: Auras Design, Silver Spring, Maryland C O N T E N T S xi Acknowledgments xiii Abbreviations and Acronyms C H A P T E R 1 1 Introduction 4 Objectives 5 Audience 5 A Primer on Environmental Health 6 A Primer on Malnutrition 7 Content and Organization P A R T I 15 Epidemiology C H A P T E R 2 17 Environmental Health, Malnutrition, and Child Health 18 Environmental Factors, Exposure, and Transmission Pathways 19 Vicious Cycle of Infections and Malnutrition 23 Environmental Role in Early Childhood Health 28 Averting Cognition and Learning Impacts 30 Key Messages 30 Note C H A P T E R 3 31 How Environmental Health Supplements Other Child Survival Strategies 32 Adding Value to Health Systems 38 Adapting Environmental Management Programs 39 Adjusting Infrastructure Strategies 43 Key Messages 43 Notes P A R T I I 45 Economics C H A P T E R 4 47 How Large Is the Environmental Health Burden? 48 Burden of Disease 51 Environmental Health Burdens 58 Areas for Future Research v vi CONTENTS 59 Key Messages 59 Notes C H A P T E R 5 61 Estimating the Environmental Health Burden and Costs at the Country Level 61 Existing Practice in Environmental Health Valuation 62 Building New Estimates for Environmental Health Costs 64 Case Studies of Ghana and Pakistan 66 Results for Ghana and Pakistan 77 Conclusion 79 Next Steps 81 Key Messages 81 Notes P A R T I I I 83 Experiences C H A P T E R 6 85 Approaches to Environmental Health 86 History of Environmental Health 88 Agenda Falling through the Cracks 90 Environmental Health Experiences in Developing Countries 97 Understanding the Enabling Environment 100 Governance and Institutional Implications 104 Institutional Requirements for Successful Environmental Health Governance 109 A Critical Moment 109 Key Messages 110 Note C H A P T E R 7 111 Conclusion 111 Contributions of This Report 113 Next Steps Appendixes 117 APPENDIX A: Technical Review of Cohort Studies 117 Background 118 Search Strategy and Selection Criteria 118 Findings and Discussion 122 Experimental Evidence from Deworming 123 Conclusions 125 Notes 141 APPENDIX B: Review of Studies on Nutritional Status and Education 145 Diarrhea and Education CONTENTS vii 145 Conclusions 146 Note 147 APPENDIX C: New Estimates for Burden of Disease from Water, Sanitation, and Hygiene 151 APPENDIX D: Computing Country-Level Environmental Health Burden of Disease 151 Mortality 160 Education 170 Notes 173 APPENDIX E: Methodological Aspects of Assessing Environmental Health Burden of Disease 173 From Relative Risks to Attributable Fractions 176 Dealing with Biased Estimates of Relative Risk 177 Notes 179 APPENDIX F: Monetary Valuation of the Cost of Environmental Health Risks 181 Note 183 References 201 Index Boxes 2 1.1 What Is Environmental Health? 22 2.1 Impact of Diarrhea on Child Malnutrition: Evidence from Research 25 2.2 Overweight Mothers Carrying Underweight Children 50 4.1 Why 50 Percent? Supporting Evidence from Recent Cohort Studies 54 4.2 Revisiting the "Asian Enigma" 56 4.3 The Mills-Reincke Phenomenon 65 5.1 Basic Indicators for Ghana and Pakistan 75 5.2 Attributable Fractions and Burden of Disease When Multiple Risk Factors Are Present 80 5.3 How Policy-Makers Should Interpret These Results 87 6.1 Combating Disease through Improved Milk 92 6.2 Mexico: Multisectorality through a Diagonal Approach 93 6.3 Thailand's National Nutrition Program 95 6.4 Ethiopia: The Toilet Revolution 97 6.5 Vietnam's Dengue Program 101 6.6 Atrophy of Environmental Health Functions in India 102 6.7 Institutional Evolution of Environmental Health: The Case of Ethiopia Figures 19 2.1 The F-Diagram: Transmission Routes for Infection 20 2.2 Relationship between Nutrition and Infection viii CONTENTS 24 2.3 Environmental Health Inputs and Health Outcomes in the Child's Life Cycle 27 2.4 The Window of Opportunity for Addressing Undernutrition 32 3.1 Range of Preventive Activities in Child Survival 52 4.1 The Health Effects of Environmental Risks Factors 53 4.2 Water-Related (WSH plus WRM) Burden of Disease in Children under Five Attributable to Environmental Risk Factors by WHO Region, 2002 55 4.3 Mills-Reincke Ratios for Subregions 63 5.1 Cost of Environmental Health Risks 70 5.2 Weight-for-Age Distribution of Children in Ghana and Pakistan 71 5.3 Two-Week Diarrheal Prevalence Rate by Age and Underweight Status in Ghana and Pakistan 73 5.4 Underweight Malnutrition Rates in Children with and without Diarrheal Infections in Ghana and Pakistan 74 5.5 Calculating Revised Estimates (Indirect and Direct Effects) 78 5.6 Final Results of Ghana and Pakistan Case Studies 152 D.1 Summary of the Methodology 155 D.2 Exposure Categories 159 D.3 Exposure Categories, Population Shares, and Relative Risks of ALRI in Ghana Tables 3 1.1 Millennium Development Goals and Environmental Health 10 1.2 Annual Cost of Direct and Indirect Impact of Environmental Risk Factors in 2005 18 2.1 Water-Related Transmission Routes and Disease Outcome 21 2.2 Impact of Infection on Nutritional Status 33 3.1 Role of Environmental Health in Supplementing Health System Strategies 49 4.1 Environmental Risk Factors and Related Diseases Included in the Comparative Risk Assessment 67 5.1 Environmentally Attributable Fractions of Child Mortality, Keeping Malnutrition Unchanged 67 5.2 Estimated Mortality in Under-Five Children from Environmental Risk Factors, 2005 68 5.3 Malnutrition Rates in Children under the Age of Five 70 5.4 Malnutrition-Attributable Fractions of Child Mortality 74 5.5 Environmentally Attributable Fractions and Child Mortality with Malnutrition-Mediated Effects 76 5.6 Effects of Malnutrition on Education 79 5.7 Annual Cost of Direct and Indirect Effect of Environmental Risk Factors in 2005 126 A.1 Cohort Follow-up Studies Relating Infectious Disease and Nutritional Status of Children in Developing Countries 142 B.1 Studies of the Effects of Malnutrition on Educational Outcomes CONTENTS ix 148 C.1 Burden of Disease (in DALYs) in Children under Five Years Attributable to Water, Sanitation, and Hygiene, by World Health Organization Subregions, 2002 152 D.1 Causes of Death and Risk Factors Considered in this Study 154 D.2 Estimating the Cost of Environmental Health Risks: Information Types and Sources 156 D.3 Relative Risks by Exposure Categories, Assuming Cox Hazard Model 157 D.4 Weight Gain Retardation Factors by Age and z-Score 158 D.5 Weight for Age in Children under Five: Current Rates and Estimated Rates in the Absence of Diarrheal Infections in Ghana 159 D.6 Estimated Mortality in Children under Five from Environmental Risk Factors, Ghana 161 D.7 Estimated Annual Cost of Education Outcomes from Stunting and Share from Environmental Factors in Ghana 162 D.8 Height Growth Retardation Factors by Age and z-Score 162 D.9 Height-for-Age Rates in Children under Five: Current Rates and Estimated Rates in the Absence of Diarrheal Infections in Ghana 165 D.10 Parameter Values Applied in Estimation of Income Losses 167 D.11 Income Distribution across Malnutrition Categories and Wealth Quintiles in Ghana 168 D.12 Annual Cost of Environmental Factors (Percentage of GDP in 2005), Using 3 Percent Discount Rate 169 D.13 Annual Cost of Environmental Factors (Percentage of GDP in 2005), Using 5 Percent Discount Rate 174 E.1 Environmental Risk Factors and Related Diseases Included in the WHO Comparative Risk Assessment Acknowledgments This book is a product of the Environmental Health Anchor Program in the Environment Department of theWorld Bank.The book was prepared by a team led byAnjaliAcharya (Environmental Specialist,ENV/World Bank) and Mikko K.Paunio (Sr. Environmental Specialist, ENV/World Bank) under the guidance of Kulsum Ahmed (Lead Environmental Specialist and Team Leader, Environmental Health Anchor Program,ENV/World Bank) and Laura Tlaiye (Sector Manager,ENV/World Bank). The core writing team also included Maria Fernanda Garcia (Consultant, ENV/World Bank), Monica Das Gupta (Sr. Social Scientist, DECRG/World Bank), Peter Kolsky (Sr. Water Sanitation Specialist, ETWWA/World Bank), Bjorn Larsen (Consultant,ENV/World Bank),and Giovanni Ruta (Economist,ENV/World Bank). Special thanks go to the peer reviewers for this study, who included Harold Alderman (Advisor, AFTHD/World Bank), Enis Baris (Sr. Health Specialist, MNSHD/World Bank), Sandy Cairncross (Professor of Environmental Health, London School of Hygiene and Tropical Medicine), and Maureen Cropper (Professor of Economics, University of Maryland). Sandy's continuous and passionate engagement on the content and tone of this book is highly appreci- ated, while Maureen's role in providing substantial inputs to the economic costing methodology developed for part of this report is especially recognized. Additional comments, inputs and guidance are gratefully acknowledged from Douglas Barnes (Sr.Energy Specialist,ETWES/World Bank),Caroline van den Berg (Sr.Economist,ETWWA/World Bank),Jan Bojö (Lead Environmental Economist, ENV/World Bank),Sandra Cointreau (SolidWaste ManagementAdviser,FEU/World Bank),James Listorti (Consultant,FEU/World Bank),Richard Seifman (Consultant, AFTHV/World Bank), and Kate Tulenko (Public Health Specialist, WSP/World Bank). The team would also like to thank Maria Neira (Director), Jamie Bartram, Carlos Corvalán and Annette Prüss-Üstün, from the World Health Organization's Department of Public Health and Environment, for sharing data relating to their new estimates of burden of disease from water, sanitation, and hygiene. The support of the Bank-Netherlands Partnership Program in the preparation of this book is also gratefully acknowledged. xi Abbreviations and Acronyms AF attributable fraction AIDS acquired immune deficiency syndrome ALRI acute lower respiratory infection ARI acute respiratory infection DALY disability-adjusted life year DDT dichloro-diphenyl-trichloroethane DHS Demographic and Health Survey GDP gross domestic product HAZ height for age z-score HIV human immunodeficiency virus IAP indoor air pollution IMCI Integrated Management of Childhood Illness (strategy) ITN insecticide-treated net IUGR intrauterine growth restriction LSMS Living Standards Measurement Survey MAL malnutrition MDG Millennium Development Goals MICS Multiple Indicator Cluster Survey NGO nongovernmental organization NISP National Improved Stove Program (China) RR relative risk SD standard deviation UNICEF United Nations Children's Fund WAZ weight for age z-score WHO World Health Organization WRM water resource management WSH water, sanitation, and hygiene WTP willingness to pay xiii C H A P T E R 1 Introduction INTEREST IN ENVIRONMENTAL HEALTH has mounted in recent years, spurred by concern that the most vulnerable groups--including children under five years of age--are disproportionately exposed to and affected by health risks from environmental hazards (see box 1.1).More than 40 percent of the global burden of disease attributed to environmental factors falls on children below five years of age, who account for only about 10 percent of the world's population (WHO 2007b). In large,populous areas in SouthAsia and Sub-SaharanAfrica,where environmental health problems are especially severe,malnutrition in young children is also rampant. Malnutrition is an important contributor to child mortality. Today, in low- income countries, more than 147 million children under the age of five remain chronically undernourished or stunted, and more than 126 million are under- weight (Svedberg 2006; World Bank 2006c). Children in the developing world continue to face an onslaught of disease and death from largely preventable factors. These children are especially susceptible to these environmental factors, which put them at risk of developing illness in early life. Acute respiratory infections annually kill an estimated 2 million children under the age of five; 800,000 of those deaths are from indoor air pollution (WHO 2007b). Diarrheal diseases claim the lives of nearly 2 million children every year; most of those deaths are attributed to contaminated water and inadequate sanitation and hygiene (WHO 2007b). Each year, approximately 300 million to 500 million malaria infections 1 2 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL BOX 1.1 What Is Environmental Health? Environmental health is defined as those health outcomes that are a result of environmental risk factors. The World Health Organization has defined environmental health as "all the physical, chemical, and biological factors external to a person, and all the related factors impacting behaviours. It encompasses the assessment and control of those environmental factors that can potentially affect health. It is targeted towards preventing disease and creating health-supportive environments" (WHO 2008). This study incorporates only those environmental health issues that relate to children-- primarily water, sanitation, and hygiene; indoor air pollution; and malaria. These problems cause the top three diseases that affect children in developing countries. Sources: Breman, Alilio, and Mills 2004; Ezzati, Rodgers, and others 2004; WHO 2008. lead to more than 1 million deaths, of which more than 75 percent occur in African children under five years of age. Malnutrition and environmental infections are inextricably linked; however, over time, these links have been forgotten or neglected by policy-makers in their formulation of strategies aimed at child survival and development. Persistent malnutrition and rampant environmental health problems are contributing to the widespread failure among developing countries to meet several of their commit- ments toward the Millennium Development Goals (MDGs), including not only the goal to halve poverty and hunger (MDG 1), but also the potential to halve maternal and child mortality (MDGs 4 and 5), to achieve universal primary educa- tion (MDG 2), to promote gender equality (MDG 3), and to combat malaria and confront the HIV/AIDS pandemic (MDG 6) by 2015 (see table 1.1). Research indicates that globally under-five mortality has fallen from 100 per 1,000 live births in 1980 to 72 per 1,000 in 2005. It is expected that the under-five death rate for the world will fall by 37 percent from 1990 to 2015, substantially less than the MDG 4 target of a 67 percent decrease (Murray and others 2007). Environmental health can contribute to many of the MDGs, as is shown in table 1.1. In many developing countries, programs to improve child health have focused on improved feeding practices, micronutrient supplementation, national immu- nization campaigns, and measures to strengthen health systems (such measures include improving the availability of drugs, ensuring better treatment of cases, and hiring more trained personnel). However, with continued exposure to contam- inated water, inadequate sanitation, smoke and dust, and mosquitoes, children in developing countries are still falling sick, a problem that imposes a sustained and heavy burden on the health system.And with the recognition of the environment's INTRODUCTION 3 TABLE 1.1 Millennium Development Goals and Environmental Health Millennium Development Goal Environmental Health Determinants Relating to Child Health 1. To eradicate extreme poverty · Expenses incurred for informal sector delivery of and hunger water and sanitation services, as well as costs of medical treatment, impose a burden on family budgets (including food budgets). Lack of adequate water and sanitation services leads to diarrhea. These problems affect children's nutritional status adversely and indirectly add to a vicious cycle of poverty. · In urban areas, time spent fetching or queuing for water limits earning capacity. 2. To achieve universal primary The environmental health burden has significant education effects on school performance and attendance. 3. To promote gender equality · Women disproportionately suffer from and empower women (a) exposure to smoke from use of biomass for cooking, (b) drudgery and inconvenience from poor access to water, and (c) privacy and dignity issues related to inadequate sanitation. · Time spent collecting water and firewood impinges on time to care for sick children or to seek livelihood opportunities. 4. To reduce child mortality Leading causes of child mortality include diarrhea, acute respiratory infections, and malaria. Indoor air pollution adversely affects young children (exposure to smoke from biomass use). Sickness and deaths result from inadequate hygiene, water supply, and sanitation. 5. To improve maternal health · Inadequate hygiene and lack of availability of clean water results in poor health outcomes related to delivery and birthing. · Malaria and helminths affect pregnant women and can lead to malnutrition of the fetus. 6. To combat HIV/AIDS, malaria, · HIV-infected children especially need clean and other diseases environments. · Environmental conditions related to mosquito breeding (such as lack of irrigation, poor drainage, and stagnant water) point to the need for adequate water resource management practices. 7. To ensure environmental · Access to water and sanitation is a goal in itself. sustainability · Slum dwellers (including children) face dismal living conditions, congested settlements, and poor access to environmental services. 8. To establish a global Multisectoral coordination on environmental partnership for development health issues is lacking. Both horizontal and vertical links are needed. Source: Compiled by World Bank team. 4 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL contribution to malnutrition, there is an urgent need to broaden the spectrum of interventions beyond the health sector. Objectives The World Bank (2006c) study titled Repositioning Nutrition as Central to Development placed nutrition as a central issue to achieving the MDGs and estab- lished that malnutrition is not only due to lack of food but also the result of environmental risk factors. This report complements that study by looking at environmental health issues that affect child health broadly, while also exploring the links through malnutrition. This report argues that environmental health interventions are preventive measures that are imperative to improve child survival with sustainable results in the long term. Preventive measures--such as improving environmental conditions--are effective in reducing a child's expo- sure to a disease agent and thereby averting infection (Murphy, Stanton, and Galbraith 1997). The overall aim of this report is to provide information to decision-makers on the optimal design of policies to help reduce premature deaths and illness in children under five years of age. To protect the health, development, and well- being of young children, decision-makers must identify and reduce environmental risk factors by providing appropriate interventions that prevent and diminish exposures. This study is intended to advance the understanding of what those risk factors are, when and how to reduce children's exposure to them, and how to mitigate their consequent health impact.Accordingly, the study has the following objectives: To provide an improved understanding of the links between environmental health risks and malnutrition through a review of literature and research. Moreover, the study discusses the role of environmental health inputs in a child's survival and growth. To analyze new data for the environmental health burden of disease (at a subre- gional level) that relates to children under five. These data, which are from a World Health Organization (WHO) report (Fewtrell and others 2007), include the total effects of environmental risk factors on health outcomes (including those mediated through malnutrition). Using two country examples, the study calculates the associated economic costs (including the costs of cognitive and learning impacts and of future work productivity). To highlight--through illustrative examples--how environmental health inter- ventions are being delivered in developing countries through a variety of health, infrastructure, and environmental programs. The study also discusses the insti- tutional and governance implications of delivering such multisectoral interventions. INTRODUCTION 5 Audience The main audience for this report is senior policy-makers (and their technical staffs) who work in the ministries of planning, finance, health, environment, rural development, and infrastructure in developing countries and who are involved in designing policies for and allocating resources to programs that contribute toward improving child health. The study will also be useful to state- and local- level governments, because the actual implementation of programs and initiatives on child health is at the level of communities and households. Furthermore, donors and other organizations financing child health improvement initiatives and proj- ects will benefit from a discussion of how interventions addressing environmental risks are important complements to health sector programs such as micronu- trient supplementation and vaccination campaigns. Finally, health, environment, and infrastructure specialists working in developing countries will also gain from understanding the importance of working on children's health from different angles in a harmonized, constructive, and collaborative way. A Primer on Environmental Health Environmental health relates to human activity or environmental factors that have an impact on socioeconomic and environmental conditions with the poten- tial to reduce human disease, injury, and death, especially among vulnerable groups--mainly the poor, women, and children under five (Listorti and Doumani 2001; Lvovsky 2001). The top killers of children under five are acute respiratory infections (from indoor air pollution); diarrheal diseases (mostly from poor water, sanitation, and hygiene); and malaria (from inadequate environmental management and vector control). This report concentrates on three specific envi- ronmental risk factors that influence a child's health: (a) poor water,sanitation,and hygiene; (b) indoor air pollution; and (c) inadequate malaria vector control. Poor Water and Sanitation Access With 1.1 billion people lacking access to safe drinking water and 2.6 billion without adequate sanitation, the magnitude of the water and sanitation problem remains significant (WHO and UNICEF 2005). Each year contaminated water and poor sanitation contribute to 5.4 billion cases of diarrhea worldwide and 1.6 million deaths, mostly among children under the age of five (Hutton and Haller 2004). Intestinal worms--which thrive in poor sanitary conditions--infect close to 90 percent of children in the developing world and, depending on the severity of the infection, may lead to malnutrition, anemia, or retarded growth, which, in turn, leads to diminished school performance (see Hotez and others 2006; UNICEF 2006). About 6 million people are blind from trachoma, a disease caused by the lack of water combined with poor hygiene practices. 6 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Indoor Air Pollution Indoor air pollution--a much less publicized source of poor health--is respon- sible for more than 1.6 million deaths per year and for 2.7 percent of global burden of disease (Smith, Mehta, and Maeusezahl-Feuz 2004; WHO 2006). It is estimated that half of the world's population, mainly in developing countries, uses solid fuels (biomass and coal) for household cooking and space heating (Rehfuess, Mehta, and Prüss-Üstün 2006). Cooking and heating with such solid fuels on open fires or stoves without chimneys lead to indoor air pollution, which, in turn, results in respiratory infections. Exposure to these health-damaging pollutants is particularly high among women and children in developing countries, who spend the most time inside the household. As many as half of the deaths attributable to indoor use of solid fuel are of children under the age of five (Smith, Mehta, and Maeusezahl-Feuz 2004). Malaria Approximately 40 percent of the world's people--mostly those living in the world's poorest countries--are at risk from malaria. Every year, more than 500 million people become severely ill with malaria, with most cases and deaths found in Sub- Saharan Africa. However,Asia, Latin America, the Middle East, and parts of Europe are also affected.Pregnant women are especially at high risk of malaria.Nonimmune pregnant women risk both acute and severe clinical disease, resulting in fetal loss in up to 60 percent of such women and maternal deaths in more than 10 percent, including a 50 percent mortality rate for those with severe disease. Semi-immune pregnant women with malaria infection risk severe anemia and impaired fetal growth, even if they show no signs of acute clinical disease. An estimated 10,000 women and 200,000 infants die annually as a result of malaria infection during pregnancy (WHO 2007d). A Primer on Malnutrition Malnutrition remains an underlying cause of death in half of the 10.5 million deaths globally in children under five (Bryce and others 2005). In low-income countries, more than 147 million (or 27 percent) children under the age of five remain chronically undernourished or stunted, and more than 126 million (or 23 percent) are underweight. South Asia, where about one-fifth of the world popu- lation lives, still has both the highest rates and the largest numbers of malnourished children in the world. In Afghanistan, Bangladesh, India, and Pakistan, the preva- lence rate varies from 38 to 51 percent and is only gradually declining, whereas in Sub-Saharan Africa,while the rate is lower at 26 percent,it is on the rise (Svedberg 2006; World Bank 2006c). Although lack of food is obviously an important reason for malnutrition, recent reports and studies ever more consistently suggest that much of malnutrition is INTRODUCTION 7 actually caused by bad sanitation and disease, especially in young children (WHO 2007e; World Bank 2006c). Thus, contrary to popular perception, in many coun- tries where malnutrition is widespread, insufficient food production is often not the determining factor of malnutrition (Prüss-Üstün and Corvalán 2006; World Bank 2006c). A recent collective expert opinion stated that about 50 percent of the consequences of malnutrition are in fact caused by inadequate water and sani- tation provisions and poor hygienic practices (Prüss-Üstün and Corvalán 2006), thus highlighting the need to mainstream environmental health into the devel- opment agenda. Nutrition in early childhood--starting from the womb--is critical for child health and, consequently, for adult health. Maternal anemia in pregnant women-- caused from a combination of malaria and hookworm infections--leads to malnourishment of the fetus, a condition called intrauterine growth restriction (IUGR). Babies suffer from low birth weight in developing countries mostly because of IUGR, whereas in developed countries, the condition is far more often attributable to preterm birth. Repeated infections--especially diarrhea and helminths--caused by poor environmental conditions lead to underweight (low weight for age) and stunted (low height for age) children. These growth-faltering effects, in turn, make individuals more predisposed to infections and even to chronic diseases later in life. Commonly used indicators of malnutrition are underweight, stunting, and wasting. Underweight is measured as the child's weight for age relative to an inter- national reference population. Stunting is measured as the child's height for age, and wasting is measured as the weight for height. Underweight is an indicator of chronic or acute malnutrition or a combination of both. Stunting is an indicator of chronic malnutrition, and wasting an indicator of acute malnutrition. How far a child's measure is from the mean of the reference population--measured in stan- dard deviations (SDs) from the mean--determines the extent of malnutrition: mild (­1 SD to ­2 SD), moderate (­2 SD to ­3SD), or severe (greater than ­3SD). Childhood malnutrition is associated with increased susceptibility to disease and with poor mental development and learning ability. In the long term, those outcomes are a significant cost to countries (Alderman and others 2006).Although research and mainstream debate have revolved around how malnourished chil- dren are more susceptible to infectious diseases (including diarrhea and acute respiratory infections), the extent to which environmental risk factors contribute to malnutrition is not widely acknowledged. Content and Organization This report is organized into three main sections: the first looks at the epidemi- ology (science and research evidence), the second presents the economics (costs of 8 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL the burden of disease and costs related to learning deficits and productivity losses), and the third describes the experiences of environmental health actions in devel- oping countries. Each section strives to present the latest information and data and highlights the reasons environmental health is so critical in the context of child survival and development. Epidemiology Chapter 2 argues that improvements in environmental health are very important for child survival and development, especially considering its links through malnu- trition. The epidemiological underpinnings of the infections-malnutrition cycle are important because repeated infections cause a decrease in dietary intake, pro- ducing, for example,malabsorption of nutrients,which in effect causes malnutrition, thereby making children weak in resisting disease and likely to fall sick again. Until recently, the impact of diseases such as diarrhea and respiratory infec- tions on malnutrition in children was relatively ignored. Over the past several decades, dozens of studies--many of them long-term cohort studies--have inves- tigated the causal relationship between disease and malnutrition. These cohort studies have provided strong evidence of how almost all infections influence a child's nutritional status. A review of the studies was carried out for this report and served to provide further corroboration of the impacts of environmental infections on child growth, including through malnutrition. Evidence from several of the studies demonstrates how exposure to environmental health risks in early infancy leads to permanent growth faltering, lowered immunity, and increased morbidity and mortality. Environmental health inputs--both at the household and the community levels--play a critical role in a child's survival and growth. In the life cycle of a child, environmental health interventions are critical, especially in the period from the womb to the age of about two years. This period is the so-called window of opportunity. Pregnant women in developing countries are often exposed to envi- ronmental risks such as malaria and hookworm infections, which contribute to poor fetal growth and result in babies with low birth weights. Smoky kitchens from use of biomass fuels have anecdotally revealed impacts on low birth weight and perinatal mortality. In early infancy, improper feeding practices and poor sanitation have a pernicious synergistic effect on the child's nutritional status. Many of these impacts on a child's growth have also been seen to result in cogni- tion and learning impacts as well as chronic diseases later in life. Current child survival strategies in developing countries mostly adopt a more treatment-oriented perspective, relying mainly on case management and focusing primarily on reducing mortality. Most of these strategies, while intended to increase the ability of the host to resist or reduce infection once exposure has occurred, do not attempt to reduce the exposure to environmental determinants of ill health. INTRODUCTION 9 Chapter 3 explores how appropriate environmental health actions can comple- ment and supplement strategies that focus on child health by adding value to health systems, by assisting in the adaptation of environmental management programs, and by promoting adjustments to infrastructure strategies. Economics Chapter 4 provides key information and data relating to the burden of disease from environmental factors and to the associated economic costs. Measuring the burden of disease and subsequent economic costs from environmental health risks is important in helping policy-makers better integrate environmental health into economic development and, specifically, into their decisions relating to the allocation of resources among various programs and activities to improve child health. Building on previous estimates and taking into consideration the links between environmental health, malnutrition, and disease, WHO recently revised the burden of disease estimates taking into account malnutrition-mediated health impacts associated with inadequate water and sanitation coverage and improper hygienic practices (Fewtrell and others 2007). The new WHO estimates reveal that the environmental health burden in chil- dren under five years is substantially higher when all links through malnutrition are incorporated. This finding is especially apparent in subregions such as Sub- Saharan Africa and South Asia, where malnutrition and poor environmental conditions coexist. In Sub-Saharan Africa, despite much poorer living standards, fewer babies are born with low birth weight than in South Asia. This enigma may in part be explained by the poor survival rate of both fetuses and children in Sub- Saharan Africa as a result of unhealthy environmental conditions. Furthermore, even when conservatively estimated, a multiplier effect exists for environmental health interventions: investments addressing environmental risks (such as lack of water and sanitation) not only reduce diarrheal mortality but also reduce mortality from malnutrition-related diseases and its consequences on education attainment. Using case studies from Ghana and Pakistan, chapter 5 translates the burden into economic costs at a country level. In doing so, it updates earlier estimates by providing measures of the total effects of environmental risks, including those through malnutrition. Also, the report for the first time attempts to estimate the longer-term impacts of these environmental health risks on cognition and learning and on future work productivity. These revised estimates show that when malnutrition-mediated health effects attributed to environmental health risks are included, the total costs for Ghana and Pakistan range from 4 to 6 percent of a country's gross domestic product (GDP) (see table 1.2). These costs are at least 40 percent higher than when malnutrition-mediated effects are not included. In the longer term, malnutrition (which is partly attributed to environment- related infections) is found to affect a child's cognitive function, school enrollment, 10 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL TABLE 1.2 Annual Cost of Direct and Indirect Impact of Environmental Risk Factors in 2005 Ghana Pakistan Cost Cost Cost Cost Cost Annual Cost (US$ (% of Annual (PRs (US$ (% of Deaths (C/million) million) GDP) Deaths billion) million) GDP) Estimation Excluding Malnutrition-Mediated Effects Mortality effects 24,712 371 412 3.84 131,611 195 3,250 2.90 Estimation Including Malnutrition-Mediated Effects Mortality effects 35,702 537 595 5.55 187,429 278 4,633 4.13 Education effects 367 407 3.79 317 5,281 4.71 Total effects 904 1,002 9.34 595 9,914 8.84 Source: Compiled by World Bank team. Note: C/ Ghanaian new cedi. grade repetition, school dropout rate, grade attainment, and future income- earning potential. For Ghana and Pakistan, the annual cost of stunting attributable to early childhood diarrheal infections is estimated to be 4 to 5 percent of the country's GDP. To estimate malnutrition-mediated costs, these analyses often rely on param- eters from global and regional studies when corresponding country-level data are unavailable. Overall, wherever assumptions are required, the parameters have been conservatively chosen. Thus, when all effects through malnutrition are consid- ered (including education costs), the total estimated annual costs may be as high as 9 percent of a country's GDP (see table 1.2). This social and economic burden is not trivial. It highlights the urgent need for policy-makers to position environ- mental health at the center of all child survival strategies. Experiences Chapter 6 begins with a historical review of environmental health, outlining the trends in the evolution of environmental health functions in developed countries and highlighting how circumstances have led to the unfortunate neglect of envi- ronmental health in the development agenda.Environmental health actions are the earliest public health activities on record.Lessons from history have shown the enor- mous benefits of multisectoral environmental health actions,with today's developed countries having undergone an evolution in environmental health functions.However, both institutionally and conceptually, environmental health has fallen through the cracks in the development agenda in the world's poorest countries. INTRODUCTION 11 On the one hand, there has been a growing environmental movement, with the creation of ministries of environment and accompanying policies and regu- lations. On the other hand, health ministries and state health departments have been engaged in scaling up vertical health sector programs, which focus mainly on treatment. This artificial separation of environment from traditional public health functions allows only limited multisectoral action that is needed to tackle environmental risks facing children under five. Chapter 6 then presents illustrative examples of how different developing coun- tries have incorporated environmental health activities within other health, nutrition, and infrastructure programs. Developing countries vary considerably in terms of institutional capacity, political will, and socioeconomic development. Environmental health interventions therefore need to be customized to the specific enabling environment in a developing country. Recognizing those differences, rather than providing specific recommendations, the chapter presents some illus- trative examples of ways in which some developing countries are beginning to mainstream environmental health components and objectives within existing child survival programs, nutrition initiatives, and infrastructure projects (water and sanitation or rural energy projects). Some common elements for successful environmental health actions in devel- oping countries have included garnering high-level political commitment,involving and empowering communities, allocating responsibilities and resources at the local level, and finding a balance between private and public sector roles. Furthermore, successful environmental health governance requires strong insti- tutional underpinnings, with clearly articulated roles at all levels of administration within a country. The study provides a discussion of the roles that national and local governments, as well as the international community, can play in delivering and managing environmental health interventions. Now is a critical time for this agenda to take the forefront in developing countries, with governments, donors, and civil society beginning to strengthen measures to address environmental health, especially in the context of child survival. Chapter 7 highlights several key conclusions of this report: Diseases from environmental risk factors--diarrheal diseases, acute respira- tory infections, and malaria--remain the top killers of children under five in developing countries. Research evidence of the cycle of disease (infections) and malnutrition implies a larger role for environmental health, because it also indirectly contributes to other diseases by weakening the child's immunity (through malnutrition). Environmental health actions also improve the effec- tiveness of other child health strategies. Specific subregions of the world--such as Sub-Saharan Africa and South Asia, where poor environmental conditions and high malnutrition prevalence coexist--should be especially targeted to fund and implement environmental 12 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL health interventions. The multiplier effect of such interventions points to the potential of their significant health externalities.At a country level, the burden of disease associated with environmental risk factors is as high as 9 percent of a country's GDP--significant enough for policy-makers to consider environ- mental health programs in resource allocation decisions. Developing countries can learn from the experience of developed countries in addressing public health risks. More than 150 years ago, today's developed countries made deliberate attempts to improve environmental health condi- tions by specifically addressing sanitation and air pollution issues with relatively cost-effective interventions. Recent experience from, for example, the Ethiopian sanitation revolution has shown that improvements in rural settings in devel- oping countries can also be achieved with modest fiscal inputs (WSP 2007), but such improvements need to be backed by political support and commu- nity involvement. The child health agenda remains unfinished in the developing world, with millions of children continuing to fall sick and die from preventable environ- mental health causes.Although considerable progress has been made, the potential of environmental health actions to complement existing health, infrastructure, and environment management strategies remains largely untapped in the devel- oping world. In many ways, this report represents a first step toward providing policy-makers with the epidemiological, economic, and experiential evidence to incorporate environmental health into the child survival agenda. However, additional research and studies will help donors and governments in developing countries choose appropriate environmental health interventions. Such research efforts should include the following: Further research on environmental health impacts during pregnancy, on addi- tional disease transmission pathways, and on better relative risk estimates will help improve disease burden and costing estimates. At a country level, cost-effectiveness and cost-benefit analyses are important follow-up exercises that will help guide decision-makers to prioritize among the various available interventions. A more in-depth country-level institutional analysis is required of the coordi- nation mechanisms between ministries and of the ways mandates and budgets are assigned. Such analysis would help guide the roles and responsibilities of different agencies for better environmental health governance. In the longer term, environmental health concerns are expected to grow. As the world's climate changes, diseases such as diarrhea and malaria, among other important health burdens that are the result of environmental risk factors, are likely to worsen, particularly for the poor in developing countries (Campbell- Lendrum, Corvalán, and Neira 2007; IPCC 2007). Thus, scaling up preventive INTRODUCTION 13 environmental health interventions to reduce the current burden of disease is a prudent investment (Campbell-Lendrum, Corvalán, and Neira 2007). Given the multisectoral nature of environmental health issues, the advocacy and regulatory roles of the health sector and the supporting roles of other sectors (such as the environmental, infrastructure, agricultural, and education sectors) in promoting and delivering environmental health actions need to be revitalized. Ultimately, good environmental health governance will require policy-makers to develop signaling mechanisms to identify environmental risks, to translate these signals into appropriate interventions, to adjust their policies to better address environmental health outcomes, and to set up institutional mechanisms to success- fully implement interventions. A concerted and continuous effort is needed on behalf of both developed and developing countries to ensure that environmental health is placed high on the development agenda, and corresponding interventions must be financed and undertaken to improve children's survival and development potential. P A R T I Epidemiology C H A P T E R 2 Environmental Health, Malnutrition, and Child Health MALNUTRITION, POOR ENVIRONMENTAL CONDITIONS, and infec- tious diseases are highly associated geographically and take their heaviest tolls on children under five years of age in Sub-Saharan Africa, South Asia, and certain countries in the Eastern Mediterranean region (Ezzati,Vander Hoorn, and others 2004; Ezzati, Rodgers, and others 2004). Malnutrition is an underlying cause of child mortality that contributes to between 34.5 percent and 52.5 percent of the 10.5 million deaths globally in children under five (Caulfield and others 2006; Fishman and others 2004). In addition, childhood malnutrition is associated with disease, poor mental development, and reduced learning ability (Alderman and others 2006). Because research and mainstream debate have revolved around how malnourished children are more susceptible to infectious diseases--including diarrhea and acute respiratory infection (ARI)--the links between environmental risk factors and malnutrition are less acknowledged. This chapter revisits the links between malnutrition and environment-related infections and seeks to demonstrate the importance of environmental health in child survival and growth.An overview of past and recent research that shows the importance of repeated disease episodes (such as diarrhea and malaria) in the development of malnutrition is presented (for a detailed analysis see appendix A). 17 18 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Environmental Factors, Exposure, and Transmission Pathways Environmental health focuses on disease transmission routes rather than on how people are treated when they are sick. The identification of transmission routes, rather than the diseases themselves, is the important conceptual framework, and because diseases can be transmitted by more than one route, environmental health interventions often make more sense at a community level than at the level of individuals (Yacoob and Kelly 1999). Table 2.1 gives examples of the different transmission routes that various water-related diseases can take. Such transmis- sion routes have largely been blocked in the developed world.In developing countries, the poor continue to be exposed to many transmission routes at one time. Understanding how different transmission routes affect disease outcomes-- especially for diarrhea--is important because even when an intervention may aim at blocking one transmission route, the effect on the disease may be limited because the population is still exposed through another transmission route (see figure 2.1).This concept is known as residual transmission (Briscoe 1987; Cairncross 1987; Eisenberg, Scott, and Porco 2007).As Cairncross and Valdmanis (2006: 775) point out, "practically all potentially waterborne infections that are transmitted by the feco-oral route can potentially be transmitted by other means (contami- nation of fingers, food, fomites, field crops, other fluids, flies, and so on) all of which are water-washed routes." The pervasive nature of fecal pollution in developing countries (see, for example, Kimani-Murage and Ngindu 2007) makes effective prevention of disease by blocking just one transmission route difficult (Eisenberg, Scott, and Porco TABLE 2.1 Water-Related Transmission Routes and Disease Outcome Transmission Route Description Disease Group Examples Waterborne Pathogen is ingested Feco-oral infections Diarrhea, dysentery, in drinking water typhoid fever Water-washed Transmission by Most feco-oral, Diarrhea, dysentery, inadequate water for oro-oral, acute typhoid fever, acute hygiene conditions respiratory, skin lower respiratory and practices and eye infections infections, scabies, and trachoma Water-based Transmission by means Water-based Schistosomiasis, of aquatic invertebrate infections guinea worm host Water-related Transmission by insect Water-related Dengue, malaria, insect vector vector that breeds in insect vector trypanosomiasis or near water infections Source: Adapted from Cairncross and Valdmanis 2006, table 41.1. ENVIRONMENTAL HEALTH, MALNUTRITION, AND CHILD HEALTH 19 FIGURE 2.1 The F-Diagram: Transmission Routes for Infection fluids sanitation water quality fields new feces food host flies fingers water quantity hand washing Source: Hunt 2006. 2007). The provision of clean water (one transmission route) has often produced less-than-anticipated outcomes because water may be scarce or hygiene practices poor (thus exposure remains through another transmission route). The emphasis on drinking water possibly occurs because those living in affluent conditions (with substantial water quantity1 and proper sanitation in their homes) often ignore the other water-washed transmission routes for diseases that poorer households (with inadequate water for proper hygiene practices) face. This idea also has its roots in the historical drama of single-source epidemics rather than in the long- term tragedy of endemic diarrhea. Vicious Cycle of Infections and Malnutrition Infections and malnutrition operate in a vicious cycle to affect child health.Though the effect of malnutrition on disease is generally recognized, the role of infections in the worsening of nutritional status has been relatively neglected. Effect of Malnutrition on Disease Poor nutritional status, especially in infants and young children, makes infections worse and often more frequent. Data from a number of studies reviewed by Scrimshaw, Taylor, and Gordon (1968) provide evidence that moderate and severe undernutrition increases the seriousness of infections such as diarrhea and acute lower respiratory infection. Increased mortality is an effect of malnutrition, which makes individuals susceptible to infectious disease; when illness occurs, it is more severe and prolonged and carries an increased risk of death (Scrimshaw, Taylor, and Gordon 1968). As predicted in 1968, malnutrition was convincingly estab- lished as a potentiator of mortality in young children,with the risk of death from all 20 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL infections increasing exponentially with decreasing nutritional status (Caulfield and others 2004; Fishman and others 2004; Pelletier 1994; Pelletier and others 1994). Malnutrition can increase a child's susceptibility to infection by negatively affecting the barrier protection afforded by the skin and mucous membranes and by inducing alterations that reduce the child's immunity (Brown 2003) (see figure 2.2). For example, in a malnourished child, diarrhea can quickly result in life-threatening dehydration caused by loss of water and minerals (Thapar and Sanderson 2004). Malnutrition also increases the duration of many infections: the more severe the level of malnutrition, the longer the illness lasts, and the longer the child takes to recover (Thapar and Sanderson 2004). Effect of Infections on Malnutrition Up until the middle of the 20th century,nutrition textbooks hardly ever mentioned the role of infections in the worsening of nutritional status, which, in turn, reduces growth in children (Keusch 2003; Scrimshaw 2003). Even though, historically, vitamin deficiencies were known to be aggravated by infections, the effect of diseases such as diarrhea and respiratory infections on malnutrition in children was not recognized, and poor diets were considered to be predominantly respon- sible for poor growth in children (Scrimshaw 2003). Over the past several decades, dozens of studies--many of them long-term cohort studies--have investigated the causal relationship between disease and malnutrition and have provided conclusive evidence of how almost all infections influence a child's nutritional status (see appendix A). Table 2.2 shows how infec- tions adversely affect nutritional status in young children through reductions in food intake caused by loss of appetite as well as changes in intestinal absorption, changes in metabolism, and excretion of specific nutrients (Scrimshaw, Taylor, FIGURE 2.2 Relationship between Nutrition and Infection decreased infection dietary intake impaired immune malabsorption function catabolism, impaired nutrient disposal barrier protection nutrient malnutrition sequestration Source: Brown 2003. ENVIRONMENTAL HEALTH, MALNUTRITION, AND CHILD HEALTH 21 TABLE 2.2 Impact of Infection on Nutritional Status Infection Effect Nutritional Impact ARI, diarrhea, measles Anorexia Decrease in food (especially) intake Acute and chronic diarrhea Malabsorption Decrease in food (from bacteria, viruses, or intake intestinal helminths) Diarrhea Effusion of serous fluid to gut; Direct nutrient loss, traditionally "treated" by mostly to gut or urine food withdrawal Hookworms Intestinal bleeding HIV infection or inflammation Increased nutrient Increase in resting (especially in smokers) requirements energy expenditure Many infections Several mechanisms among Diminished transport infected individuals leading of micronutrients to reduced serum retinol, (vitamin A, iron, and iron, and zinc concentrations zinc) (relevance remains uncertain) Source: Based on Stephensen 1999. Note: For all the diseases, the stress, fever, or tissue damage caused by disease increases the food requirement while also reducing the intake and absorption of nutrients, which leads to anorexia. and Gordon 1968; Stephensen 1999). These nutrient losses have implications for tissue synthesis and growth in young children--and lead to growth faltering (Brown 2003). The effect of infections on the nutritional status of young children appears to be directly proportional to the severity of the infection (Powanda and Beisel 2003), which means that children with more serious infections, such as dysentery, measles, or pneumonia, are more likely to become stunted than those with acute diarrhea (see appendix A). Numerous cohort studies have researched the impact of infections (mostly diarrheal disease) on weight and linear growth (see box 2.1). A recent Peruvian study found that children ill with diarrhea 10 percent of the time during the first 24 months of life were 1.5 centimeters shorter than children who never had diar- rhea (Checkley and others 2003).Another study from Brazil found that,on average, 9.1 diarrheal episodes before two years of age were associated with a 3.6 centimeter growth shortfall at age seven years (Moore and others 2001). Similarly, Moore and colleagues (2001) found that early childhood helminthiasis (infections caused by parasitic worms such as hookworms) led to a further 4.6 centimeter growth reduction by age seven. Because early childhood growth faltering is known to predict height in adulthood (Martorell 1995), the effects of infections on linear growth are considered irreversible (Checkley and others 2003). In general, effects on weight have been easier to demonstrate than effects on linear growth, even in shorter follow-up studies (Stephensen 1999). 22 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL BOX 2.1 Impact of Diarrhea on Child Malnutrition: Evidence from Research Among the various infections, diarrhea is one of the most prevalent in developing countries and is responsible for a high proportion of sickness and death in children under five years of age (Scrimshaw 2003). Because of its high occurrence and its involvement with the malabsorption of nutrients, diarrhea has been a key issue in child malnutrition (Mata 1992; Scrimshaw 2003). The effects of different types of malnutrition on diarrheal illness have been studied over the past several decades (Guerrant and others 1992) (see accompanying figure). Several studies in developing countries support the argument that an increased risk of diarrheal mortality is associated with low weight for age; however, the effect on diarrheal incidence has been less clear. Evidence now supports the idea that incidence is increased (Fishman and others 2004). Effects of Diarrhea and Malnutrition incidence diarrhea duration, severity weight gain (short term) malnutrition height gain (long term) Source: Guerrant and others 1992. Scrimshaw, Taylor, and Gordon (1959, 1968) summarized the evidence on how infections have a deleterious, although often ignored, effect on nutritional status. By 1968, the metabolic consequences of infections were already well established (Powanda and Beisel 2003). A major problem in the review was that long-term human follow-up information was lacking (Scrimshaw 2003). After 1968, about 38 cohort follow-up studies have produced information on this topic. Practically all the cohort studies (see appendix A) favor the idea that not only diarrheal disease but also other infections (for example, helminths, measles, and acute lower respiratory infections) cause growth to falter. This view was strengthened because investigators could show larger effects on nutritional status with increasing frequency and severity of infections. Some inconsistencies and weak observed effects can be explained by the mitigating effects of breastfeeding and other factors that effectively mask effects of infections on growth faltering in these studies (see appendix A). Experimental clinical trial information on deworming of (continued) ENVIRONMENTAL HEALTH, MALNUTRITION, AND CHILD HEALTH 23 children further supports a causal inference between infection and poor growth (Taylor-Robinson, Jones, and Garner 2007). The debate in the 1990s focused on whether these effects could be reversed, although even those who suggested that subsequent catch-up growth corrects much of the growth faltering recognize the failure in children with persistent diarrhea (that is, those who have diarrhea more than 10 percent of the time) (Bairagi and others 1987; Briend and others 1989; Moy and others 1994). Several studies during the past decade have shown that if infection burden begins before six months of age, lagging linear growth effects that are observed are likely to be irreversible (Adair and others 1993; Brush, Harrison, and Waterlow 1997; Checkley and others 2003; Moore and others 2001), which diminishes the possibility that malnourishment, in the first place, made the host more susceptible and that observed effects would result from reverse causality rather than true causality (see appendix A). A WHO collective expert opinion (Prüss-Üstün and Corvalán 2006) states that 50 percent of consequences of maternal and childhood underweight is attributable to lack of water, sanitation, and hygiene. The idea that infections affect a child's nutritional status has faced some skepticism. Even while skeptics have agreed that, in the short term, the adverse metabolic effects from infections do lead to growth faltering in children, they have argued that subsequent "catch-up" growth fully compensates for these adverse effects of infection in the majority of children (Bairagi and others 1987; Briend and others 1989; Moy and others 1994; see also appendix A). Part of the reason for this skepticism and controversy about the impact of infections on growth faltering in children lay in the paucity of direct long-term human observations indicating the irreversibility of these effects (Scrimshaw 2003).Compelling evidence of this irreversibility was not gained until early 2000 (see appendix A). Environmental Role in Early Childhood Health Environmental health inputs--at both the household and the community levels-- play a critical role in a child's survival and growth (see figure 2.3 and appendix A for further detail). In the life cycle of a child, from the womb to the age of about two years, environmental health interventions--such as access to water and sani- tation, proper hygiene practices, proper vector control, and the use of cleaner fuels for cooking and heating--are especially critical for preventing growth faltering in the fetus and infant, which has consequences for a child's subsequent health. These impacts on a child's growth have also been seen to result in cognition and learning problems as well as chronic diseases later in life, an issue discussed in more detail later in this chapter. 24 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL FIGURE 2.3 Environmental Health Inputs and Health Outcomes in the Child's Life Cycle pregnancy early infancy early childhood (0­2 years) (2­5 years) birth and postneonatal perinatal postweaning s peri- or postneonatal infant and under-5 acute anemia, placental deaths, number of deaths; omec insufficiency, infectious disease incidence of fetal loss, out episodes, diarrhea, ARI, stillbirths underweight malaria; malnutrition school performance, alth work productivity, he chronic restricted growth stunting, cognitive and learning impairments chronic diseases such as coronary disease and obesity vector control, safe safe water, improved safe water, improved s environ- water, improved sanitation, vector sanitation, vector input mental sanitation, control, indoor air control, indoor air fo health indoor air pollution mitigation, pollution mitigation, pollution mitigation hygiene promotion hygiene promotion amplexe other prenatal care proper breastfeeding, immunizations, immunizations, micronutrient micronutrient supple- supplementation, mentation, better better case case management management Source: Adapted from World Bank 2008a. The following sections discuss the specific roles of environmental health actions in the life cycle of a child from pregnancy and protection of the fetus; to early infancy, when growth-faltering effects are irreversible; to early childhood, when both household- and community-level actions can make a difference. Finally, the longer-term effects, specifically in terms of cognition and learning, are discussed. Pregnancy: Protecting the Fetus Nutrition plays a crucial role in the growth and the development of the fetus. During pregnancy, the mother's own nutritional status and exposure to infec- tions have an important effect on the fetus (Fishman and others 2004). The impact of infections on malnourishment of the fetus and subsequent growth faltering has been inadequately studied (Breman, Alilio, and Mills 2004). Whereas some infectious diseases (for example, rubella) can infect the fetus through the placenta, other infections (for example, malaria and hookworm) can induce fetal death, stillbirths, and perinatal deaths, as well as contribute to poor fetal growth, without infecting the fetus (van Geertruyden and others 2004). In addition to experiencing micronutrient deficiencies, pregnant women in developing countries are exposed to numerous environmental risks. Malaria is ENVIRONMENTAL HEALTH, MALNUTRITION, AND CHILD HEALTH 25 endemic across the tropics and subtropics,and it thrives in areas with poor drainage and stagnant water.Areas with bad sanitation provide prime conditions for hook- worm infections (Hotez and others 2006). In many developing countries, and especially among the poor, malaria and hookworm infections coexist--both syner- gistically affecting the health of the pregnant woman and her unborn child (Watson-Jones and others 2007). Anemia in pregnancy, which is associated with increased risks of premature labor and low birth weight (Watson-Jones and others 2007), is a multifactoral condition and is often caused by a combination of malaria, hookworm infections, and dietary deficiencies (Menendez, Fleming, and Alonso 2000; Watson-Jones and others 2007; see also box 2.2). Malaria. Every year approximately 50 million pregnant women worldwide are exposed to malaria; 30 million of these women live in the African region (Crawley and others 2007). Malarial infection during pregnancy is an important, and preventable, environmental cause of low birth weight (Allen and others 1998; van BOX 2.2 Overweight Mothers Carrying Underweight Children It is not uncommon in developing countries to see overweight mothers carrying underweight babies. These images effectively convey the message that not food security but repeated infections from poor environmental health conditions often contribute to undernutrition among children (World Bank 2006c). Moreover, food requirements during early childhood--the "window of opportunity"--are small (World Bank 2006c). Even where food scarcity is a concern, the lack of affordable environmental health services such as clean water only aggravates the situation, diverting resources from the family's food budget and thus contributing indirectly to the child's poor nutritional status (Cairncross and Kinnear 1992). A key feature in the "fetal-programming" or Barker hypothesis is that being underweight in early infancy or even earlier, in the womb, makes the child more susceptible to rebound growth if food becomes available in abundance. Thus, those girls who were growth restricted in the womb or became stunted in early infancy could be susceptible to overweight and related chronic disease later in life (Eriksson 2005; Sachdev and others 2005). Environmental health conditions have either more distal effects (that is, the mother's own irreversibly developed underweight status during early infancy or in womb) or more proximate effects (that is, malaria and hookworms that effectively reduce birth weight). Although reduced birth weight is a strong predictor of subsequent underweight status or stunting, adverse effects of environmental health conditions on fetal growth cannot be incorporated in cohort studies that have studied the effects of infections on growth faltering (see box 2.1 and appendix A). 26 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Geertruyden and others 2004). Malaria is thought to affect a child's birth weight in two ways: first, through placental infection, and second, through malaria- induced maternal anemia (Menendez, Fleming, and Alonso 2000). Placental infection appears to be associated with a significant reduction in birth weight and with increased risk of "low"(that is, under 2 kilograms) birth weight (Menendez, Fleming, and Alonso 2000). One review estimates that in areas where malaria is endemic, one-fifth of babies with low birth weight are underweight because of malarial infection of the placenta while the mother was pregnant (Menendez, Fleming, and Alonso 2000). Malaria also clearly contributes to maternal anemia; approximately 400,000 pregnant women develop moderate or severe anemia each year in Sub-Saharan Africa because of malaria infection (Guyatt and Snow 2004). Hookworm infections. Hookworm infections are common worldwide but thrive in poor communities in the tropics where poor water supply and poor sanitation are common (Steketee 2003). Data from the early 1990s suggested that about a third of the pregnant women in the developing world harbored hookworm infec- tion (Steketee 2003).Few studies have looked at the impact of hookworm infections on pregnant women (Steketee 2003). Some studies in Nepal, however, showed that hookworm infection exacerbated iron deficiency and anemia in pregnant women. Indoor air pollution. Apart from parasitic infections such as malaria and hook- worms, indoor air pollution adversely affects the health of pregnant women. Biomass is often used for cooking and heating in developing countries, leading to smoky kitchens and high levels of particulate matter. These conditions have been shown to lead to various respiratory infections in women and young chil- dren--who are the most exposed because of time spent near the stove (Bruce, Perez-Padilla, and Albalak 2002). Two studies have implicated indoor air pollu- tion in low birth weight and perinatal mortality.A study in rural Guatemala found that birth weight was 63 grams lower for babies born in households using wood than for those born in households using cleaner fuels, and a study in Zimbabwe showed an association between perinatal mortality and exposure to indoor air pollution (WHO 2000). The association of indoor air pollution with growth faltering, however, is far less clear (Smith, Mehta, and Maeusezahl-Feuz 2004) than the association with malaria (Watson-Jones and others 2007). Early Infancy: The "Window of Opportunity" Several studies that looked at the impact of infections on child growth have shown that exposure to environmental health risks in early infancy leads to permanent growth faltering, lowered immunity, and increased mortality (figure 2.4; see also appendix A). Averting repeated disease episodes, especially in the first two years of life--the"window of opportunity"--prevents the more permanent and devas- tating wasting and stunting, which have longer-term implications for a child's health and prognosis (World Bank 2006c). ENVIRONMENTAL HEALTH, MALNUTRITION, AND CHILD HEALTH27 FIGURE 2.4 The Window of Opportunity for Addressing Undernutrition 0.50 0.25 0.00 ­0.25 -scorez ­0.50 age ­0.75 for ­1.00 ­1.25 weight ­1.50 ­1.75 ­2.00 0 3 6 9 12 15 18 21 24 27 30 33 36 39 42 45 48 51 54 57 60 age months Africa Asia Latin America and Caribbean Source: Shrimpton and others 2001. Neonates and infants up to about six to nine months of age acquire immunity from their mothers, which often prevents ARIs (Sato and others 1979). The imma- turity of the infants' immune responses (Marodi 2006; Nair and others 2007; Wilson 1986), however, helps explain why early exposure to diarrheal pathogens, especially in early infancy, is so critical in the development of malnutrition. Thus, the same infections that cause disease in the developed world only among those with impaired immune systems (for example,AIDS patients) lead to diarrhea and growth faltering in infants in developing countries (Checkley and others 1998; Molbak and others 1997). Breastfeeding is considered an effective means of protecting infants from diar- rheal diseases (Dai and Walker 1999;VanDerslice, Popkin, and Briscoe 1994). The stress and fever caused by disease increase the child's food requirements while also reducing the intake and absorption of nutrients. Proper feeding practices during diarrheal episodes are critical for a child's health and recovery. In most cultures, the common practice is to restrict food from sick children, often for too long a time (Brown 2003). Only in recent years has the medical profession emphasized the need for continuous feeding during diarrheal episodes (Brown 2003).Reducing the level of environmental contamination similarly reduces the risk of diarrhea (VanDerslice, Popkin, and Briscoe 1994). Good sanitation practices protect infants by creating barriers to keep pathogens out of their environment. These practices 28 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL include properly disposing of excreta to isolate human wastes, improving water supplies to protect drinking water from fecal contamination, and hand washing and personal hygiene to reduce the transmission of enteric pathogens in the home (VanDerslice, Popkin, and Briscoe 1994).Analysis of data in Bangladesh, Malaysia, and the Philippines has provided evidence that lack of breastfeeding and poor envi- ronmental sanitation have a pernicious synergistic effect on infant mortality (Habicht, DaVanzo, and Butz 1988; Pelletier 1994). The protective effect provided by good-quality drinking water and improved community sanitation is greatest for nonbreastfed infants and completely weaned infants (VanDerslice, Popkin, and Briscoe 1994). For the latter, reducing exposures (especially foodborne exposures) remains important. Strategies for reducing exposures include improvements in choice of supplemental foods, better preparation practices and storage methods, and better personal hygiene (VanDerslice, Popkin, and Briscoe 1994). Clean Environments, Healthy Children In early childhood (two to five years of age), the growth-faltering effects of repeated disease episodes are considered largely reversible, in contrast to the irreversibility of such effects in early infancy. Still, the environmental risk factors associated with poor access to water, improper sanitation, and bad household and community hygiene remain a threat--especially given the child's increased mobility (walking) and associated ever-increasing peri-household activities. Action should be taken at both the household level and the community level: Household level. In the late infancy and toddler periods, hand-washing practices, proper disposal of children's feces, and safe storage of milk and weaning foods are critical activities to cut diarrheal transmission. In addition, with increased mobility of the child (from crawling on the floor, to walking and running), addressing general household hygiene in terms of surfaces (including bedclothes and floors) becomes crucial, as do cleaner yards outside (soil helminths or parasites caused by lack of sanitation or animal wastes from domestic animals) (Cairncross and others 1996). Community level. Community action and control of the public domain can be seen as an important step to enable improved household and personal hygienic practices (the private domain). Both are needed, but in some cases (for example, with hookworms) collective sanitation efforts are key to success (Cairncross and others 1996). In addition, young children will benefit from indoor residual spraying or other personal-level protection against malaria even if the problem is not addressed collectively. Averting Cognition and Learning Impacts Cognitive function in children--reflecting an ability to learn--is affected by environmental and health-related factors (Berkman and others 2002). Risk factors ENVIRONMENTAL HEALTH, MALNUTRITION, AND CHILD HEALTH 29 that interfere with cognition are especially important in the first two years of a child's life, which marks a period of rapid growth and development (Berkman and others 2002). In early childhood, diseases attributed to environmental factors, such as diarrhea and helminth infections, also have the potential to affect a child's later cognitive functions. Over the past several years, studies have begun to inves- tigate the impact of diarrheal illness and helminth infections during early childhood on verbal fluency, cognitive function, and school performance. Several studies of children in Brazil and Peru have investigated the impact of diarrheal illness in the first two years of life and subsequent deficits in cognitive function as well as school performance. Among Brazilian children, the incidence of diarrheal illnesses in infancy was found to be associated with impaired cogni- tive performance 4 to 7 years later (Guerrant and others 1999; Niehaus and others 2002), as well as with disproportionate reduction in verbal fluency 5 to 10 years later (Patrick and others 2005). However, one large study in Peru that controlled for confounders saw no independent association with IQ at nine years of age (Walker and others 2007). The number of overt episodes of diarrhea has also been associated with indicators of school performance (Lorntz and others 2006). Helminth infections have the highest prevalence and intensity in school-age children, and analyses of schoolchildren in Jamaica and Kenya have shown the positive effects of treatment on cognition function (Guerrant and others 1999). Only a few studies show the effect of helminths on development in children under five years of age (Walker and others 2007). In rural Nicaragua, the presence of intestinal parasites was associated with poor language performance among children (see Walker and others 2007). In recent years, several new longitudinal studies, often enrolling close to 2,000 participants, have investigated the effect of early childhood malnutrition on educational outcomes such as school grade attainment (Alderman and others 2006; Grantham-McGregor and others 2007; Maluccio, Hoddinott, and Behrman 2006); learning productivity (Behrman and others 2006); school dropout or grade repetition (Lorntz and others 2006); and primary school enrollment (Alderman and others 2006; Lorntz and others 2006). Earlier studies had already consistently shown that an increasing level of stunting has negative effects on the studied variables (Grantham-McGregor and others 2007). Alderman and others (2006) have recently shown how one standard deviation drop in height for age leads to a 0.68-year lower school grade attainment or a 0.4-year delayed enrollment. (For a review of these studies, see appendix B.) In addition to diarrheal illness and helminth infections,severe or cerebral malaria in children under five years of age is associated with subsequent neurological and cognitive impairments (Walker and others 2007). Data from a study of Kenyan children suggest severe falciparum malaria may be the single most common cause of acquired language disorder in the tropics (Carter and Mendis 2006). 30 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Key Messages Malnutrition, poor environmental conditions, and infectious diseases are highly associated geographically and take their heaviest tolls on children under five years of age in Sub-Saharan Africa, South Asia, and certain countries in the Eastern Mediterranean region. Infections and malnutrition operate in a vicious cycle to affect child health. Until recently, however, the role of infections in the worsening of nutritional status, which, in turn, reduces growth in children, was not addressed. Over the past several decades, several cohort studies have provided strong evidence of how almost all infections influence a child's nutritional status. The impact of infections on child growth has shown that exposure to environ- mental health risks in early infancy leads to permanent growth faltering,lowered immunity, and increased mortality. In the life cycle of a child, the period from the womb to the age of about two years--the so-called window of opportunity-- is critical in terms of environmental health interventions. The infections that adversely affect a child's growth have also been seen to result in cognition and learning impacts as well as chronic diseases later in life. Note 1 The World Health Organization defines 100 liters per capita per day as the amount of water required to meet all consumption and hygiene needs. Additional information on lower service levels and potential effects on health are described in Howard and Bartram (2003). C H A P T E R 3 How Environmental Health Supplements Other Child Survival Strategies THE BIGGEST EFFECTS ON MORTALITYcan be achieved by simultaneously improving health status and nutritional status of high-mortality populations (Pelletier 1994). However, the current child survival strategies in developing coun- tries focus mostly on case management and treatment. From the health sector's perspective, primary prevention comprises vaccinations, micronutrient supple- mentation, promotion of breastfeeding, and measures to decrease low birth weight, including birth spacing (Murphy, Stanton, and Galbraith 1997) (see figure 3.1). These strategies all are intended to increase the ability of the host to resist or reduce infection after exposure has occurred, but they do not attempt to reduce exposure to the environmental determinants of ill health, which constitute another aspect of primary prevention. Additional activities relating to environmental health (sanitation, clean water, and vector control) are largely ignored because they fall outside of the current health sector mandate (see figure 3.1) even though environmental health legis- lation is often in place and such activities are known to be extremely effective. Within the World Bank's new Health, Nutrition, and Population strategy, the Multisectoral Constraints Assessment attempts to address this issue by looking at the multisectoral determinants of health outcomes, including those from the environmental and infrastructure sectors. The new strategy's focus on strength- ening health systems--including disease surveillance systems--will also help 31 32 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL FIGURE 3.1 Range of Preventive Activities in Child Survival prevention traditional child survival sanitation immunization diagnosis clean water breastfeeding treatment (case reduction of air food supplementation management) pollution micronutrients vector control birth spacing household and community hygiene health exposure disease death Source: Adapted from Favin, Yacoob, and Bendahmane 1999, figure 1-1. better track environment-related infections and exposures. Given the evidence showing how environmental risks and subsequent infections affect child malnu- trition, correcting the current neglect of environmental health in child survival and child health strategies in developing countries becomes even more impera- tive. This chapter highlights how appropriate environmental health actions can be an important complement and supplement to other existing strategies addressing child health. Adding Value to Health Systems Health system frameworks are designed around improving case management of childhood illnesses.Through broad frameworks such as Integrated Management of Childhood Illness,efforts are being made to strengthen case management,to benefit from synergies across vertical programs, and to capitalize on externalities in health care delivery (Mason and others 2006).Nevertheless,several areas where case manage- ment through health systems is not enough--in the case of childhood illnesses such as diarrhea,acute respiratory infections,and helminth infections,for example--call for better primary prevention of environmental risks (see table 3.1 for description of how environmental health can supplement the activities of the health sector). With the current case management strategies of child survival programs, several issues limit the effectiveness of these interventions: First is the safety of administering drugs for parasitic infections during preg- nancy. Where possible, priority should be given to treating or preventing TABLE 3.1 Role of Environmental Health in Supplementing Health System Strategies Health Systems Cause of Death (%) Main Strategies Main Limitations Role of Environmental Health Neonatal causes (27%), omitting Evidence-based prenatal care Lack of professional obstetric and · Because of malaria and sepsis and diarrhea but including (tetanus toxoid important) prenatal care capacity hookworms, role is potentially neonatal tetanus Counseling and professional help very large in prenatal period, during and after delivery which could ease pressure on scarce obstetric capacity. · Neonatal tetanus control policy may rely mainly on clean delivery (that is, on hygiene). Acute lower respiratory infections Improved case management Antibiotic resistance · Role is large through indoor air (ALRIs) (19%) pollution mitigation and becomes greater when the effects through malnutrition are considered. · There is increasing reason to believe that hygiene may play a role in preventing ALRIs. Diarrhea (18%), including · Oral and intravenous · Low compliance with oral · Role in primary prevention of neonatal diarrhea rehydration therapy rehydration therapy diarrhea is very large through · Appropriate breastfeeding · Diarrhea incidence unaltered, but improved water and sanitation practices chronic and bloody diarrhea facilities and improved hygienic · Micronutrient supplementation emerge as problem in an practices. increasing proportion of the remaining deaths · Inappropriate use of antibiotics and emerging anthelminth resistance 33 (continued) 34 TABLE 3.1 Role of Environmental Health in Supplementing Health System Strategies (continued) Health Systems Cause of Death (%) Main Strategies Main Limitations Role of Environmental Health · Relatively low (about 20%) efficacy of micronutrient supplementation (especially with rampant infections) · Breastfeeding by HIV-positive mothers Neonatal sepsis (10%) · Antibiotic therapy · Antibiotic resistance and lack of · Clean delivery in household access (especially in rural areas) settings is augmented. to hospitals Malaria (8%) · Case finding and case · Antimalarial resistance · Horizontal community management · Low efficacy of bednets involvement is needed. · Bednets · Compliance issues, leading to · Better environmental · Indoor residual spraying retreatment costs management, especially in urban areas, results. Sources: Annual proportions for main under-five killer diseases: Bryce and others 2005; Jefferson and others 2007; Luby and others 2005; Rabie and Curtis 2006. Note: Neonatal causes include birth asphyxia, low birth weight, and neonatal tetanus. HOW ENVIRONMENTAL HEALTH SUPPLEMENTS OTHER CHILD SURVIVAL STRATEGIES 35 infections either before the first pregnancy or between pregnancies. For conditions that are diagnosed during pregnancy or that must be treated or prevented during pregnancy (for example,anemia,malaria,hookworm,or some combination of these conditions), the safety of the treatment must be balanced with the adverse consequences of the disease (Steketee 2003). Currently, reproductive health programs encourage public health approaches during pregnancy for anemia, malaria, and hookworm prevention and treatment because the substantial adverse consequences outweigh any risk associated with prevention and treatment (Steketee 2003). Second, there is growing drug resistance, as well as widespread misuse of antibi- otics to treat children for infections such as diarrhea (Thapar and Sanderson 2004). In older children, the World Health Organization (WHO) is closely monitoring resistance to the drugs for helminth treatment, while problems relating to reinfection remain (Awasthi, Bundy, and Savioli 2003; Sur and others 2005). Antimalarial drug resistance is weakening the effect of chemotherapy malaria control in Africa and constraining the drug options for the early treat- ment strategy recommended by WHO. Third are problems related to compliance. For example, the uptake and use of oral rehydration therapy has possibly led to significant reductions in diar- rhea mortality; however, use of oral rehydration therapy remains very low, at around 20 percent, in countries such as Bangladesh and India (Thapar and Sanderson 2004). Childbirth and Infant Care Programs The components of safe motherhood strategy usually fall primarily under the health promotion activities (care packages) organized and provided by health ministries (Graham and others 2006). They consist of, for example, community education,evidence-based prenatal care,and counseling,including primary preven- tion of peri- or neonatal infections (tetanus and HIV). In many developing countries, unclean delivery practices, traditional umbil- ical applications, and the use of unclean cloths to swaddle the baby can have detrimental health effects on the newborn (Bennett and others 1996; Mullany and others 2006; Quddus and others 2002). Additionally, in many developing countries, the prevalence of multiresistant bacteria that cause neonatal infec- tions (a major cause of neonatal deaths) is rapidly increasing, even while broad-spectrum antibiotics are losing their effectiveness (Bang and others 1999; Vergnano and others 2005).Also, difficulties remain in maintaining high coverage of tetanus immunizations for pregnant women (to prevent neonatal sepsis). In these areas, hygienic and clean delivery practices, supplemented by environmental health interventions approaches, become even more important (Mullany and others 2006; Quddus and others 2002). 36 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL By advocating hygienic practices for delivery and infant care, several health sector programs in developing countries are already adopting environmental health principles related to clean household hygiene. Additional environmental health interventions, such as providing clean water and sanitation, play an augmentive--yet important--role in these health promotion efforts. Only with adequate quantities of water and proper sanitation facilities can households with infants maintain clean surfaces, practice proper hand washing, and properly dispose of feces. Another important tenet of infant care programs is appropriate feeding prac- tices. Established evidence shows that exclusive breastfeeding can prevent malnutrition (Adair and others 1993; Alvarado and others 2005; Chisti and others 2007; Cohen and others 1995). Consequently, infant care programs through the health sector are encouraging breastfeeding practices as well as the postpone- ment of the weaning period among severely malnourished infants to prevent death (Pelletier 1994). Additionally, as mentioned earlier, inadequate sanitation compounds poor breastfeeding practices and synergistically affects child health. This need for improved sanitation to reduce diarrheal transmission to infants is seen even more starkly among countries with high rates of HIV prevalence.Medical research has shown that for HIV-positive mothers, as many as 9.5 percent of their infants can acquire HIV through breast milk over a three-year period (Bertozzi and others 2006). These mothers are faced with a very difficult dilemma--whether to breastfeed or not--because their children may face severe malnourishment (caused by diarrhea from bad water and poor sanitation) or alternatively might acquire HIV if breastfed. Exclusive breastfeeding, even though uncommon, might be safer than mixed breastfeeding for reducing HIV transmission risk (Coovadia and others 2007). Even in these circumstances, however, healthy environments remain a high priority. Immunization Programs The global Expanded Program on Immunization has been very effective in reducing mortality in children under five years of age (Brenzel and others 2006)--with sustainable reductions especially through measles, whooping cough (pertussis), and tetanus vaccines (Brenzel and others 2006). There is still potential to save an additional 1.8 million children under the age of five annually by improving coverage with current licensed vaccines in South Asia and Sub-Saharan Africa (Brenzel and others 2006); however, economic costs as well as logistical considerations (for example, the need to maintain a cold chain for live vaccines) constrain this possi- bility. A very recent study from the WHO and United Nations Children's Fund, reporting a 91 percent drop in the number of measles deaths in Africa, has attrib- uted the decrease largely to better coverage in routine immunization programs and targeted campaigns (Zaracostas 2007). HOW ENVIRONMENTAL HEALTH SUPPLEMENTS OTHER CHILD SURVIVAL STRATEGIES 37 In recent years, a number of new vaccines against rotavirus, pneumococcal infections, hookworms, and malaria have been developed and tested. The rotavirus vaccine has shown promise but still faces several technical hurdles before it can be used widely against diarrheal disease in developing countries (Banerjee and others 2006; Thapar and Sanderson 2004). Also important, rotavirus--unlike other intestinal pathogens--apparently is not strongly related to the lack of clean water, sanitation, and appropriate hygiene (Thapar and Sanderson 2004), perhaps because a low number of viruses is needed to infect an individual and because the virus tends to survive outside a host a long time--that is, it is difficult to"wash away" (Eisenberg and others 2006). Some progress has also been achieved in hookworm vaccines (Chu and others 2004),and several malaria vaccines are currently being tested in clinical trials,where they have shown to provide some protection (Komisar 2007). Although progress continues, with some setbacks (Cutts and others 2005; Singleton and others 2007), in the field of new vaccines (for example, bacterial vaccines against neonatal sepsis and meningitis or pneumococcal infections and diarrhea for children under five), the improvement of environmental health conditions--especially water and sani- tation--should continue to be emphasized (Thapar and Sanderson 2004). Micronutrient Supplementation Programs Along with underweight, micronutrient deficiencies--such as in vitamin A, iron, and zinc--contribute to the global burden of disease in young children. Like underweight,micronutrient deficiencies increase susceptibility to infections,which, in turn, may further aggravate the deficient state (West 2003). A number of micronutrient supplementation programs are being undertaken in developing countries. Several studies have shown that vitamin A and zinc play key roles in determining the duration, frequency, and severity of infectious diseases such as diarrhea (Guerrant, Lima, and Davidson 2000). However, evidence also shows that the burden of infections, in turn, often constrains the effectiveness of these supplementation programs. Likely, this effect is caused by malabsorption of key antimicrobial drugs, resulting in the emergence of drug resistance in impov- erished areas (Guerrant, Lima, and Davidson 2000). Research has revealed that infectious diseases, such as diarrhea and respira- tory infections, and intestinal worm burdens may modify the growth response to vitamin A supplementation (Hadi and others 1999; West 2003). With a heavy burden of respiratory infections, supplemental vitamin A may not be completely absorbed and can be expected to be excreted in high quantity (Hadi, Dibley, and West 2004). In a research trial of 1,405 Indonesian children, those who experi- enced respiratory infections for more than 26 days during a four-month observation period had a 74 percent reduction (0.17 centimeters in four months) in their linear growth response to vitamin A supplementation compared with children 38 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL with no respiratory infection (Hadi and others 1999). Results from trials of iron supplementation have not found significant growth effects, even in anemic chil- dren (Iannotti and others 2006). Iron supplementation, in addition, may actually lead to increased morbidity and consequently to reduced dietary intakes and poor nutrient absorption (Iannotti and others 2006).Also, studies have found that the degree of infectious disease treatment is an important factor in determining who benefits and who is harmed in terms of serious infectious disease outcomes when low-dose iron supplements are provided (Iannotti and others 2006). Interventions to prevent or decrease malnutrition and infectious disease are expected to decrease child mortality, and interventions that accomplish both will have the greatest effect (Pelletier 1994; Pelletier and Frongillo 2003). Environmental health interventions have an important complementary as well as supplementary role to play in addressing nutritional deficiencies. Water, sanitation, and hygiene have the potential to reduce zinc deficiencies that occur from losses during diar- rheal illnesses. Similarly, efforts to reduce or control respiratory infections will help increase the effectiveness of vitamin A supplementation programs in improving growth in children (Hadi, Dibley, and West 2004). Adapting Environmental Management Programs Recent evidence from verbal autopsy data provides concrete evidence that the mortality rates of children under five years of age caused by malaria in Sub-Saharan Africa remain very high (Adjuik and others 2006). Key strategies in the control of malaria have relied on early diagnosis and treatment, on the use of insecticide- treated nets (ITNs),and--to a lesser extent--on indoor residual spraying (Mabaso, Sharp, and Lengeler 2004; WHO 2003). The gains in malaria control in Sub- Saharan Africa achieved in the 1940s to 1970s from indoor residual spraying with DDT (dichloro-diphenyl-trichloroethane) and other insecticides were eroded in the mid-1980s--with malarial epidemics becoming frequent and more severe (Mabaso, Sharp, and Lengeler 2004). ITNs have been shown to reduce all-cause mortality by about 20 percent in community-wide trials in several African settings (WHO 2003), although a recent observational community-based intervention study suggests a substantially higher efficacy of about 44 percent (Fegan and others 2007). Poor compliance with net retreatment, combined with the additional cost of the insecticide and a lack of understanding of its importance, however, pose difficult barriers to full imple- mentation of ITNs in endemic countries (Biran and others 2007; National Center for Infectious Diseases 2006; WHO 2003). In the case of communities with longer- lasting nets, the problem of retreatment is more or less obsolete, but the problem of adherence to best practices with use of these nets remains an issue. With few options for malaria, many rural households in Sub-Saharan Africa continue to rely on smoke from indoor cooking as a mosquito repellent (Kristensen HOW ENVIRONMENTAL HEALTH SUPPLEMENTS OTHER CHILD SURVIVAL STRATEGIES 39 2005). A recent review of research evidence investigated the association between indoor air pollution and malaria transmission (Biran and others 2007). The main findings were that smoke from domestic fuel use probably does not have much effect on mosquito feeding, while the presence of eaves spaces makes houses more permeable to mosquitoes and has been associated with an increased risk of malaria (Biran and others 2007). Soot from domestic fires may impair the effectiveness of ITNs because the nets are perceived to be dirty, possibly resulting in an increase in the frequency of washing them and a consequent loss of insecticide (Biran and others 2007). Conversely, it can be concluded that a reduction in indoor air pollu- tion might have beneficial effects on malaria control by allowing ITNs to remain unwashed for longer periods. Strong evidence indicates that environmental management has been effective in controlling malaria, although mosquito ecology in many tropical areas often makes this approach impractical (Fewtrell and others 2007; Keiser, Singer, and Utzinger 2005).1 Malaria control methods that involve environmental manage- ment--that is, a modification or manipulation of the environment to reduce malaria transmission (for example, through the installation, cleaning, and main- tenance of drains or the systematic elimination of standing pools of water)--currently receive far less attention (Keiser, Singer, and Utzinger 2005). Adjusting Infrastructure Strategies Infrastructure investments (such as in water supply, sanitation, and rural energy) are typically designed with the primary objective of improving access or efficiency rather than maximizing public health effects, although such interventions do sometimes recognize the importance of public health.Yet with growing emphasis on increasing access to water supply systems and sanitation facilities and on moving poorer populations up the energy ladder, there remains a huge opportunity to appropriately incorporate public health considerations into the design and delivery of these interventions. Improving Water Supply More than 1.1 billion people still do not have access to safe drinking water. Service is poor even in many countries that have water supply systems. For many consumers, piped water is often intermittent, and when available, it is unsafe for drinking.With growing demands for water, governments in developing countries are making some progress in extending coverage to improved water supply2 with large-scale water supply projects, especially in urban areas. From a health perspective, water quantity is generally more important than water quality, because increased quantities of water promote good hygiene and can prevent fecal-oral transmission by a number of different routes; increased quantities of water also reduce skin and eye infections (Boeston, Kolsky, and Hunt 40 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL 2007). Furthermore, household water connections are considered optimal because consumption of water doubles or triples as households use it for improved hygiene practices (Cairncross and Valdmanis 2006), and far greater health benefits result. In both urban and rural areas in developing countries, people demand water supplies to reduce the time and cost of fetching water. To meet demands for water, governments have been increasing efforts to extend the coverage of piped water systems. The major investments in urban water focus on network solu- tions and thus reflect the needs of utilities, which are concentrated on providing satisfactory service to their existing customers; managing large amounts of water safely and efficiently; meeting environmental standards for their wastewater discharges; and recovering costs to permit continued operation, maintenance, replacement, and expansion. Accordingly, engineering, financial, and environ- mental perspectives dominate the design of network interventions, and public health is reflected only in technical standards to reduce the risk of common- source outbreaks from the water supply (Boeston, Kolsky, and Hunt 2007). In rural areas,the time-saving benefits of water access are even more pronounced, and the"most obvious benefit is that water is made available closer to where rural households need it" (Cairncross and Valdmanis 2006: 773). Accordingly, govern- ment investments in rural water supply aim at improving access; however, the smaller size of rural communities means that piped systems, in general, and house connections, in particular, tend to be more expensive per capita (Cairncross and Valdmanis 2006). Microbial contamination of drinking water in rural and urban areas can occur either at the source, through seepage of contaminated runoff water, or within the piped distribution system. Moreover, unhygienic handling of water during transport or within the home can contaminate previously safe water (WHO 2007c). Household water treatment technologies (such as chlorination and solar disinfection) and safe storage can be important interventions in situations where access to water supplies is secure but household water quality is not assured (WHO 2007c). Improving Sanitation Roughly 2.6 billion people in the developing world are without adequate sanita- tion,3 and facilities are often overloaded, in disrepair, or unused. Even though the sanitation gap is twice as large as that of water supply, investments in sanitation and hygiene have lagged far behind those in water and other"social"sectors, such as health and education. The main costs of urban sanitation services are those of sewers and sewage treatment.Whereas sewers contribute to public health through reducing everyday contact with sewage (especially by children), wastewater treatment is designed largely to meet ecological objectives and not those of public health (Boeston, HOW ENVIRONMENTAL HEALTH SUPPLEMENTS OTHER CHILD SURVIVAL STRATEGIES 41 Kolsky, and Hunt 2007). Urban utilities, by and large, are not well designed or staffed to address off-network solutions for water supply or sanitation, yet those solutions are likely to be the most important first steps of progress in environ- mental health for many of the urban poor. Rural sanitation is frequently a low priority relative to food, shelter, schooling, and employment. Families that invest in rural sanitation usually do so for other reasons--for cleanliness, dignity, moder- nity, comfort, and convenience--more than for public health. For a variety of historical, political, and institutional reasons, public health plays little role in project design, despite a large role in project justification of both water supply and sanitation investments in developing countries (Kolsky and others 2005). The roles for the infrastructure sector have largely been related to improving access (for example, increasing the number of tap connections), while those of the health sector have traditionally focused on improved treatment strate- gies. Opportunities remain for the health sector to take on a larger role in improving the health outcomes of infrastructure investments through promotion (for example, hygiene promotion or support to the marketing of sanitation) and advocacy (for example, persuading utilities to lower their off-putting connection charges and to recover the connection costs through the monthly tariff so that house connec- tions are more affordable to the poor). Relatively easy improvements in public health outcomes can be achieved in a number of ways: Relatively low-cost hygiene promotion programs that attempt behavior change in household hygiene through a wide range of social-marketing tools can be implemented.Neither network utilities nor rural water engineering staff members are naturally inclined to assign high priority to hygiene promotion, although it is highly cost-effective as a health intervention (Laxminarayan, Chow, and Shahid-Salles 2006). Urban water and sewerage network expansions can often be sequenced in different ways, and a relatively easy way to speed public health improvement is to stress the need to extend services to new areas as quickly as possible. Trade-offs between ideal environmental objectives and the priorities of public health can be explored.Wastewater treatment systems are,by and large,designed to achieve ecological objectives, not public health ones. If strict environmental standards are adopted in poor countries, the often high costs of such systems means that wastewater treatment comes at the expense of depriving the poor of basic access to sanitation. The costs and benefits of wastewater treatment need to be balanced against the public health benefits and costs of basic access. Moving up the Energy Ladder Worldwide, more than 3 billion people continue to rely on solid fuels, including biomass and coal, to meet their most basic energy needs: cooking, boiling water, and heating (WHO 2006). To address this issue, developing countries have made 42 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL various rural energy investments to move these people up the energy ladder toward cleaner fuels. Nevertheless, progress since 1990 has been negligible because the small gains made are lagging population growth (WHO 2006). Recent decades have witnessed many household energy initiatives, ranging from ambitious government-run improved cookstove programs to small-scale community-led projects. Cleaner fuels and improved cookstove programs that reduce indoor air pollution can mitigate diseases such as respiratory and other illnesses (World Bank 2007d).Studies in India and Nepal have revealed that women who were exposed to biomass smoke--but who did not smoke themselves--had chronic respiratory disease death rates comparable to those of heavy male smokers (Modi and others 2005; World Bank 2007d). Improved cookstove programs have an enormous potential to affect respira- tory health. They can especially benefit women and young children, who are disproportionately affected by respiratory disease. In the past, however, such programs have more often focused on energy-efficiency criteria rather than on health considerations.The first improved cookstove programs in India,for example, were more concerned with the potential to burn biomass energy more cleanly and efficiently than with health outcomes. Secondary benefits generally included conserving fuel, reducing smoke emission in the cooking area, keeping kitchens clean, reducing deforestation (Agarwal 1983, 1986; World Bank 2007d), lessening the drudgery of women, reducing cooking time, and improving employment opportunities for the rural poor. The most organized intervention of rural house- hold stoves--China's National Improved Stove Program (NISP)--focused primarily on fuel efficiency. However, a major independent review of NISP, including a 3,500-household survey in three provinces at different levels of coverage, found that the program improved air quality and health in rural households, but not sufficiently to meet Chinese indoor air quality standards (Zhang and Smith 2005). Whereas the supply of technology is more concerned with energy efficiency and costs, the demand for cookstoves has reflected limited awareness of the health effect of improved cookstoves. In India, for example, perceived benefits cited by users of improved cookstoves included reduction in time spent in cooking, reduced fuel consumption, health benefits such as reduction of eye irritation, and cleaner kitchens. Yet only 6 percent of the households cited "health bene- fits" as a perceived benefit. This finding suggests that users over the years had accepted the burning sensation in the eyes and coughing as daily facts of life and considered only medical issues that required a visit to the rural doctor as"health" issues (World Bank 2007d). In Guatemala, similarly, respondents from three improved cookstove programs were found to value the benefits from lower fuel- wood use, reduced cooking times, and less smoke. Significantly, the advantages of reduced eye irritation and better respiratory health were perceived to be the lowest benefits under the program (Ahmed and others 2005). This evidence of HOW ENVIRONMENTAL HEALTH SUPPLEMENTS OTHER CHILD SURVIVAL STRATEGIES 43 households undervaluing the health benefits associated with improved cookstoves highlights opportunities for raising awareness through a variety of public media or social-marketing approaches. From a health perspective, behavior-change components are also important supplements in clean stove programs. In Bangladesh, for example, pollution exposure for children in a typical household can be halved by adopting two simple measures: increasing children's outdoor time from three to five or six hours a day and concentrating outdoor time during peak cooking periods (Dasgupta and others 2004b). Including such health promotion activities in rural energy projects is important in raising awareness of health benefits associated with improved indoor air and creating a demand for improvement of cookstoves and changes in behavior. Key Messages Current child survival strategies in developing countries focus on case manage- ment and treatment, to the neglect of primary prevention--especially relating to reducing exposure to infections. These strategies typically intend to increase the ability of the host to resist or reduce infection after exposure has occurred, but they do not attempt to reduce the exposure to environmental determinants of ill health. Environmental health actions incorporated into existing child health strategies can provide a bigger"bang for the buck"in terms of health improvements, and they are an important complement and supplement to other existing strate- gies addressing child health by adding value to health systems, adapting environmental management programs, and adjusting infrastructure strategies. Notes 1 Conversely, a neglect of environmental health action through vector habitat control has been responsible for the resurgence of dengue virus in almost all continents (Cattand and others 2006; Kay and Nam 2005) and a forceful comeback of African trypanosomiasis spread by tsetse flies in Central and East Africa in the past few decades (Remme and others 2006). 2 Improved water supply is defined to encompass reasonable access to protected water resources, which include rainwater collection, protected springs and dug wells, boreholes, public standpipes, and household connections. In addition, it involves the application of measures to protect the water source from contamination (Hutton and Haller 2004). Reasonable access means at least 20 liters per person per day, accessible within 1 kilometer of that person's dwelling (WHO and UNICEF 2005). 3 Improved sanitation involves access to sanitation facilities that allow for safe disposal of excreta, which may include connection to a sewer or septic tank system, pour-flush latrine, simple pit, or ventilated improved pit latrine (WHO and UNICEF 2005). P A R T I I Economics C H A P T E R 4 How Large Is the Environmental Health Burden? THE PREVIOUS CHAPTERS REVISITED THE RESEARCH evidence on malnutrition-infection links to demonstrate the importance of environmental health interventions for child survival and development and pointed toward the complementary role of environmental health interventions in other child survival strategies. Given this evidence, quantitatively measuring the importance of envi- ronmental health issues for better child health becomes important. Furthermore, to make decisions regarding investments in environmental health actions, policy- makers need to get a sense of the averted health costs. Hence, this chapter seeks to answer the following questions: How large is the environmental health burden for children under five, after considering its links through malnutrition? Which subregions have the highest environmental health burdens? What are the economic costs associated with these burdens at a country level? Answering these questions will help policy-makers appreciate the magnitude of the environmental health burden and better integrate environmental health into decision making for economic development. Furthermore, translating health effects into economic costs--such as a percentage of a country's gross domestic product--has been a powerful means of raising awareness about environmental health issues and facilitating progress toward sustainable development. 47 48 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL This chapter looks at the methodology used by the World Health Organization (WHO) to estimate the "burden of disease" within and across various subre- gions of the world, specifically in the context of environmental health risks. This methodology--dynamic and evolving over the past decade--has been applied to assess and compare the health status of populations in countries and subregions. Results from the latest WHO analysis of environmental health burdens incorporate the complex links between malnutrition and infections into the calculations and provide a basis for comparisons across different subregions. The next chapter extends this analysis to estimate the corresponding burden at the country level and the consequent economic costs to society. Burden of Disease In the field of health, a challenge one faces when trying to choose between different health sector interventions is deciding how to compare different health outcomes (or end points). For example, how does one compare environmental interven- tions that differ in terms of how many lives of children are saved (mortality) and how many days of various illnesses (morbidity) are avoided? How would one choose between an intervention that saved 20 child lives and 200 cases of illness, for example, and another that saved 15 lives and 1,000 cases of illness? Also, how would one compare a case of diarrheal illness averted to a case of respiratory infec- tion averted? The Global Burden of Disease Project, initiated by WHO in 1996, aims at answering these questions and, in doing so, provides useful guidance for policy making (WHO 2007a). To assist governments in setting priority actions in health, WHO developed a methodology to evaluate the burden of disease and to quantify the health status of a population. One of the most commonly used summary measures of popu- lation health is the disability-adjusted life year (DALY), which creates a common metric to measure both the potential years of life lost to premature death and the years of healthy life lost because of disability or illness. One DALY can be thought of as one lost year of healthy life. Environmental Burden of Disease A 1999 paper assessing the global environmental burden of disease reported that about 43 percent of the total burden of disease attributable to environmental risks falls on children under five years of age, even though they make up only 12 percent of the population (Smith, Corvalán, and Kjellström 1999).1 Since then, impor- tant reports on environmental health burden estimation were undertaken in the framework of the Global Burden of Disease Project. In 2002, a section of the World Health Report (WHO 2002: chapter 4) quantified the significance of environ- mental health risks, which were revisited in depth in 2004 by the Comparative HOW LARGE IS THE ENVIRONMENTAL HEALTH BURDEN? 49 Risk Assessment project2 (Ezzati and others 2003; Ezzati, Lopez, and others 2004; Ezzati, Rodgers, and others 2004; Ezzati, Vander Hoorn, and others 2004). The method used to calculate the environmental burden of disease is based on an exposure approach, supported by a comprehensive analysis of evidence on specific health risks. Exposure-response relationships for a given risk factor--such as poor water, sanitation, and hygiene (WSH)--are obtained from epidemiolog- ical studies, and the derived attributable fractions are then applied to disease burden (in deaths or DALYs) associated with the same risk factor (see WHO n.d.). WHO's (2004) Comparative Risk Assessment considers the following six envi- ronmental risk factors and examines them in relation to the diseases identified in table 4.1 (Ezzati, Lopez, and others 2004). The risk-factor approach quantified WSH and indoor air pollution as having the greatest effect on children under five years of age (Ezzati, Vander Hoorn, and others 2004; Ezzati, Rodgers, and others 2004). This risk-factor approach to assessing health status is particularly useful to Policy-makers because it enables them to think of specific interventions to address specific environmental risks. In 2006, WHO published Prüss-Üstün and Corvalán's environmental health report, Preventing Disease through Healthy Environments. Using a different approach from earlier studies, the report elicited expert opinions on the fractions of different diseases attributable to environmental factors. Apart from water, sanitation, and indoor air pollution, malaria emerged as an important environmental health problem among children under five--especially in Sub-Saharan Africa. The report estimated that among children up to 14 years of age, the proportion of deaths attributable to the environment was as high as 36 percent (Prüss-Üstün and Corvalán 2006), which compares to an attributable disease burden of 24 percent TABLE 4.1 Environmental Risk Factors and Related Diseases Included in the Comparative Risk Assessment Risk Factor Related Diseases Water, sanitation, hygiene Diarrheal diseases, trachoma, schistosomiasis, ascariasis, trichuriasis, hookworm disease Indoor air pollution Chronic obstructive pulmonary disease, lower respiratory infections, lung cancer Outdoor air pollution Respiratory infections, selected cardiopulmonary diseases, lung cancer Lead Mild mental retardation, cardiovascular diseases Climate change Diarrheal diseases, malaria, unintentional injuries, protein-energy malnutrition Selected occupational Unintentional injuries, hearing loss, cancers, risk factors asthma, chronic obstructive pulmonary disease, lower back pain Source: WHO 2002. 50 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL for all age groups. A number of earlier studies had led the way toward these important findings. In the same report, expert opinion indicated that poor WSH contributed to 50 percent of the consequences of childhood and maternal underweight (see box 4.1). For example, when environmentally attributable deaths from acute lower respiratory infection (ALRI) were reported, they were caused by indoor air pollution. Nevertheless, perhaps up to 50 percent of these deaths would not have occurred if the child had not been malnourished (Ezzati, Vander Hoorn, and others 2004). The implications of this link were mainly reported through the consequences of different environmental risks as deaths in different disease categories. Current literature on risk factors has internalized a significant portion of these consequences of malnutrition, often without much discussion. Although the WHO report (Prüss-Üstün and Corvalán 2006) did include a relatively small category (measles and meningitis deaths) of child deaths associated with BOX 4.1 Why 50 Percent? Supporting Evidence from Recent Cohort Studies In Preventing Disease through Healthy Environments (Prüss-Üstün and Corvalán 2006), expert opinion was cited as indicating that poor water, sanitation, and hygiene and inadequate water resources management contributed to 50 percent of the consequences of childhood and maternal underweight. This expert opinion builds on evidence that emerged toward the end of the 1950s on the impacts of environmental infections on a child's growth (Keusch 2003; Scrimshaw, Taylor, and Gordon 1959). In this report, a technical review of 38 recent cohort studies further corroborates this 50 percent (confidence interval 39 to 61 percent) figure used to estimate the consequences of malnutrition attributable to environ- mental risk factors. Evidence from several of the studies demonstrates how exposure to environmental health risks in early infancy leads to permanent growth faltering, lowered immunity, and increased mortality. Using Bradford Hill's causality criteria as a framework, the results from the studies were scrutinized to evaluate whether infections cause growth faltering and, if so, to what extent (for details, see appendix A). For example, a recent relatively large study from Bangladesh reveals that dysentery and watery diarrhea together could retard weight gain by 20 to 25 percent compared with those periods where no infections occurred (Alam and others 2000). This weight gain retardation is likely to be signifi- cantly higher when compared with international standards, and simulations that reveal 35 percent weight gain retardation result in approximately similar environmentally attributable health burden as the estimate in Prüss- Üstün and Corvalán (2006). HOW LARGE IS THE ENVIRONMENTAL HEALTH BURDEN? 51 malnutrition, it did not include other deaths that would have been avoided had the child not been malnourished. Including the Malnutrition-Infection Link As discussed in chapter 2, there is now substantial evidence that repeated infec- tions in early childhood--mainly diarrheal disease from inadequate WSH--is a significant contributor to child malnutrition (Prüss-Üstün and Corvalán 2006). Moreover, the evidence indicates that malnutrition, in turn, increases the risk of child mortality and the incidence of morbidity3 and impairs cognitive develop- ment with consequences for educational attainment and performance (see Fishman and others 2004). A WHO analysis in 2007 sought to incorporate this evidence of malnutrition- infection links in its environmental burden of disease methodology (Fewtrell and others 2007). For the first time, this methodology not only incorporated direct health risks from environmental factors (such as diarrheal disease burden from poor water and sanitation) but also sought to include indirect risks (concentrating on WSH and its indirect effect on mortality through malnutrition). Thus, while a traditional burden of disease calculation would associate WSH with only diar- rheal diseases (see figure 4.1), the inclusion of the indirect path implies the need to include all diseases attributable to malnutrition (because 50 percent of the consequences of malnutrition are, in turn, attributed to poor WSH).4 These additional diseases include ALRIs, measles, malaria, protein-energy malnutrition, and other infectious diseases that cannot be directly attributed to environmental risk factors. Figure 4.1 illustrates these direct and indirect effects. To quantify the health effects of malnutrition, Fewtrell and others (2007) adopted the methodology of Blössner and de Onis (2005), who estimated fractions attrib- utable to malnutrition for diarrheal diseases, low birth weight, lower respiratory infections, malaria, measles, and protein-energy malnutrition. Note that to avoid possible double-counting, Fewtrell and others (2007) did not include as an indirect effect diseases such as diarrhea that are a direct effect of poor WSH, even though diarrhea is one of the consequences of malnutrition. Therefore, the estimates in Fewtrell and others (2007) are conservative. Environmental Health Burdens Results from the latest WHO analysis (Fewtrell and others 2007) show the water- related (WSH and water resource management, or WRM) environmental health burden for children under five years of age in various subregions (see figure 4.2). From a point of view of child health, three subregions in Sub-Saharan Africa and South Asia stand out in terms of the environmental health burden of disease. An interesting feature of figure 4.2 is that regions differ widely from each other depending on whether one looks at DALYs per capita (panel a) or total DALYs 52 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL FIGURE 4.1 The Health Effects of Environmental Risks Factors poor water inadequate sanitation improper personal and household hygiene indoor air pollution (IAP) inadequate water resource management (WRM) Diarrhea diarrhea Acute lower respiratory acute direct infections (ALRIs) infections malaria Malaria malnutrition (50% from bad sanitation direct and repeated infections diarrhea (WSH) 50% from food availability) ALRI (IAP) malaria (inadequate WRM) tuberculosis (IAP) indirect protein-energy malnutrition (WSH) diarrhea (food) ALRI (non IAP) infections not environmentally related (excluding HIV) Source: Compiled by World Bank team. (panel b).Although the Southeast Asian subregion SEARD has a similar number of DALYs as the African subregion AFRD,the two subregions differ quite substantially in terms of DALYs per capita--with the latter having a value more than two times higher. Incidentally, diarrheal disease in SEARD is, in relative terms, more impor- tant as a percentage of total DALYs than diarrheal disease inAFRD.When compared with the most developed subregion, EURA, which is in Europe, the environmental health burden (in DALYs per 1,000 children under five) in the two subregions of Africa (AFRB and AFRD) is about 150 to 175 times higher. For SEARD, this ratio is about 70 times higher. Important risk factors such as indoor air pollution and environmental-attributable perinatal mortality are not included in figure 4.2. In the public's imagination, the home of the malnourished child is Sub- Saharan Africa. Nevertheless, some researchers showed evidence that the HOW LARGE IS THE ENVIRONMENTAL HEALTH BURDEN? 53 FIGURE 4.2 Water-Related (WSH plus WRM) Burden of Disease in Children under Five Attributable to Environmental Risk Factors by WHO Region, 2002 a. DALYs per capita WPRB SEARD SEARB diarrheal diseases EMRD malnutrition and its consequences region EMRB malaria WHO AMRD other AMRB AFRE AFRD 0.00 0.10 0.20 0.30 0.40 0.50 0.60 DALYs per 1000 children under 5 b. Total DALYs WPRB SEARD SEARB EMRD diarrheal diseases malnutrition and its consequences region EMRB malaria WHO AMRD other WSH AMRB AFRE AFRD 0 5,000 10,000 15,000 20,000 25,000 30,000 DALYs Sources: Adapted from Fewtrell and others 2007; Prüss-Üstün and Corvalán 2006. Note: Calculations are based on deaths attributable to perinatal conditions published in Graham and others (2006: 507) and on estimating the regional aggregate number of live births using live birth per 1,000 data population data from the World Bank Development Data Platform and Live Database time series for 2000. The WHO regions are defined as follows: WPRB (Western Pacific subregion with low child and low adult mortality); SEARD (Southeast Asian subregion with high child and high adult mortality); SEARB (Southeast Asian subregion with low child and low adult mortality); EMRD (Eastern Mediterranean subregion with high child and high adult mortality); EMRB (Eastern Mediterranean subregion with low child and low adult mortality); AMRD (subregion of the Americas with high child and high adult mortality); AMRB (subregion of the Americas with low child and low adult mortality); AFRE (African subregion with high child and very high adult mortality); AFRD (African subregion with high child and high adult mortality). For a detailed list of the countries that belong to each subregion, see http://www.who.int/choice/demography/regions/en/index.html. 54 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL worst-affected region is not Africa but South Asia (Ramalingaswami, Jonsson, and Rohde 1996). Over the past few decades, other explanations have been suggested for these surprising results, which have been labeled the"Asian enigma" (see box 4.2). BOX 4.2 Revisiting the "Asian Enigma" As many as 30 percent of children in India are born with low birth weight, as compared with only 16 percent in Sub-Saharan Africa. Furthermore, continuing up to five years of age, underweight rates remain far lower in Sub-Saharan Africa (approximately 30 percent) compared with those in South Asia (close to 50 percent). Although the depth of hunger is clearly greatest in Sub-Saharan Africa (FAO 2006), and the gross domestic product and living standards are higher in South Asian countries, this nutritional status difference is considered to be somewhat of an enigma (Gragnolati and others 2006). Using multivariate analyses of household survey data, Ramalingaswami, Jonsson, and Rohde (1996) suggested that this difference was explained by the better decision-making power of women in Sub-Saharan Africa compared with those in Asia, but they also stated that sanitation differences (use of latrine) between the regions clearly contribute to the nutritional status gap as well. Scientific inference from household surveys is often hampered by issues of confounding (other related factors ultimately may be the culprit) as well as by inconsistencies in definitions (what is meant by latrine in different countries?) (Cairncross and Kolsky 1997). Also, for example, high latrine use might significantly threaten the integrity of sources of drinking water--again making inferences difficult (Kimani-Murage and Ngindu 2007). Often, serious problems with data quality exist; for example, in a recent Kenya National Census, the enumerators missed Kibera, the largest shantytown in Africa, which led to overestimating latrine use in Kenya. Finally, mechanisms of causality often remain obscure in studies relying on indicator data (Cairncross and Kolsky 1997; Kolsky and Blumenthal 1995). If sanitation and hygiene conditions were poorer in South Asia, one might logically expect more diarrhea per child--which is not the case (see figure 4.2). In attempting to explain this enigma, researchers have not adequately considered the intergenerational dimensions of malnutrition--that is, malnourished fetuses result in babies with low birth weight. These babies become underweight mothers and, in turn, give birth to more babies with low birth weight. The prevalence of fetal undernutrition (or restricted growth in utero) is directly related to the duration and incidence of malnutrition (which is increased by mother's low weight before pregnancy and by infections during pregnancy, such as hookworms and malaria). (continued) HOW LARGE IS THE ENVIRONMENTAL HEALTH BURDEN? 55 A number of miscarriages, stillbirths, and perinatal deaths (owed in large part to rampant malaria) likely are preferentially harvesting the underweight fetuses and children from the African population (Adjuik and others 2006; van Geertruyden and others 2004). These early deaths, then, do not add to the statistics of underweight children, and therefore the region appears to have fewer underweight babies than does South Asia. The Multiplier Effect for Environmental Health Interventions Results from these subregional data show the multiplier effect of environmental health interventions on mortality--where investments addressing environmental risks (for example, lack of water and sanitation) not only reduce diarrhea mortality but also reduce mortality from malnutrition-related diseases and its consequences on educational attainment. In these subregions, this multiplier ranges from 1.5 to 1.9 (figure 4.3), which means for every direct death avoided FIGURE 4.3 Mills-Reincke Ratios for Subregions 2.0 1.9 AFRD SEARD 1.8 1.7 EURC WPRB AFRE ratio 1.6 EMRD EURB SEARB 1.5 AMRD 1.4 AMRB Mills-Reincke 1.3 EMRB 1.2 1.1 EURA 1.0 0.00 0.05 0.10 0.15 0.20 diarrheal mortality Source: Compiled by World Bank team on the basis of Fewtrell and others 2007. Note: The World Health Organization regions are defined as follows: WPRB (Western Pacific subregion with low child and low adult mortality); SEARD (Southeast Asian subregion with high child and high adult mortality); SEARB (Southeast Asian subregion with low child and low adult mortality); EMRD (Eastern Mediterranean subregion with high child and high adult mortality); EMRB (Eastern Mediterranean subregion with low child and low adult mortality); AMRD (subregion of the Americas with high child and high adult mortality); AMRB (subregion of the Americas with low child and low adult mortality); AFRE (African subregion with high child and very high adult mortality); AFRD (African subregion with high child and high adult mortality); EURA (Europe subregion with very low child and very low adult mortality); EURB (Europe subregion with low child and low adult mortality); EURC (Europe with low child and high adult mortality). 56 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL from an environmental health intervention, at least half to almost one additional death is averted from indirect links. A recent research paper using household survey data for rural China similarly found evidence of a 2.0 multiplier, implying that for every diarrheal death averted in children under five from household water connections, an additional ALRI death was also averted (Jacoby and Wang 2004). This multiplier effect is not new: it was first empirically observed in the late 19th century and came to be known as the Mills-Reincke phenomenon (see box 4.3). BOX 4.3 The Mills-Reincke Phenomenon In the late 19th century, impressed by the beneficial effects of clean water supply on mortality (especially child mortality), a German sanitary engineer (Johann Julius Reincke) and an American sanitary engineer (Hiram F. Mills) independently found that for every prevented death from waterborne diseases, additional pulmonary tuberculosis, pneumonia, and infant deaths were averted. This curiosity came to be called the Mills-Reincke phenomenon (Sedgwick and Macnutt 1908). A recent analysis of several U.S. cities for 1900 to 1936 has conserva- tively estimated that, for one diarrhea death, three other deaths were averted by clean water supply systems (Cutler and Miller 2005). Mortality declined steeply without major advances in curative and individually targeted preventive interventions (for example, vaccines) (Cutler and Miller 2005). Another important source of typhoid fever, killing especially children under five years of age from unclean milk, was also eliminated in the early decades of the 21st century (Rosen 1958). The explanations behind the Mills-Reincke phenomenon have remained enigmatic (Cairncross 2003). Contemporaries (Sedgwick and Macnutt 1908) considered two alternatives: (a) that the "other" category would ultimately also be waterborne or (b) that clean water supply somehow enhanced vitality--it was known that after convalescence, typhoid fever patients suffered greatly from secondary infections (Ferrie and Troesken 2004, as cited in Cutler and Miller 2005). Evidence to support the first explanation still has not been found. The results derived from Fewtrell and others (2007) are in line with the second explanation and bring together current competing nutrition and infection-based models (Cutler and Miller 2005) for explaining the dramatically lowered mortality rates in the developed countries over the past century. Recent research has revealed that unintentional consequences of improved hygienic conditions include a steep decline in stomach cancer (Kivi and Tindberg 2006; Parkin 2006) and potential for decreased susceptibility to coronary heart disease (Eriksson 2005). HOW LARGE IS THE ENVIRONMENTAL HEALTH BURDEN? 57 Conservative Estimates The Mills-Reincke ratios (the ratio of consequences of malnutrition-diarrhea deaths in subregional WHO estimates) are plotted on a graph with decreasing level of development (using increased risk of diarrheal deaths as a proxy), reflecting the close correlation between diarrheal disease and malnutrition (see figure 4.3). These ratios are small (ranging from 1.5 to 1.9) when compared to those conser- vatively derived (3.0) for the United States during the early 20th century. One explanation for the relatively more conservative ratios might be current child survival interventions such as measles vaccinations. Although current methodo- logies to estimate burdens attributable to environmental health use the best available information, the impacts of environmental risk factors are considerably under- estimated for several reasons. The health impacts from environmental risk factors are potentially under- estimated in several areas. The first is in the area of perinatal deaths. The 11 percent of these deaths considered to be attributable to the environment (in developing countries) is likely to be a substantial underestimate (Prüss-Üstün and Corvalán 2006). Breman, Alilio, and Mills (2004) suggest that as many as about 60 percent of perinatal deaths could be attributable to rampant malaria in Sub-SaharanAfrica.5 Second,growing evidence (Jefferson and others 2007) shows that ALRIs can be prevented by improving personal and household practices,6 implying that the WSH-attributable health burden becomes substantially larger (Cairncross and Valdmanis 2006). For example, a randomized clinical trial in Karachi, Pakistan, showed that hand washing reduces ALRIs by 50 percent (Luby and others 2005). Third, a substantial proportion of ALRI deaths associated with indoor air pollution would not have occurred had the children not been malnourished (Ezzati, Vander Hoorn, and others 2004).7 However, only some of these deaths are included in the consequences of malnutrition. Finally, malaria deaths that are attributable to inadequate mosquito habitat control (for example, 42 percent in Sub-Saharan Africa) do not include the fraction that can be effec- tively prevented by other environmental management measures, such as indoor residual spraying and bednets. Similarly, there are several indications that the malnutrition-mediated effects of environmental risk factors ("consequences of malnutrition") are underesti- mated. First, the consequences of malnutrition in Fewtrell and others (2007) consider only attributable fractions of moderate and severe maternal and child- hood underweight--less than or equal to 2 standard deviations (SD)--but do not include the consequences of mild malnutrition (from ­2SD to ­1SD). Mild malnutrition, the health burden of which is included in other estimates, is the most prevalent form of malnutrition and increases death risk almost twofold (Caulfield and others 2004; Fishman and others 2004). Second, 42 percent of malaria deaths in Sub-Saharan Africa are caused by poor WRM. Of the remaining 58 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL 58 percent of malaria deaths caused from other factors, however, at least a portion should be attributable to environmental factors through the effect of diarrheal infections on malnutrition and increased malaria fatality rates (see figure 4.1). They are not included because some of these deaths would overlap with those directly attributable to inadequate habitat control. Third, given the price and wage inelasticity among the poorest, income spent on environmental health services (such as water) often takes away from food budgets, thereby contributing to malnu- trition. In Khartoum, Sudan, for example, residents of one squatter settlement were found to pay as much as 55 percent of their income to purchase water (Cairncross and Kinnear 1992)--sacrificed at the cost of their food budgets. A final point on why these current estimates are conservative involves the longer-term perspective. The Barker or fetal-programming hypothesis, mentioned in box 2.2, postulates that malnourishment in utero or very early in infanthood makes adults more susceptible to chronic diseases (like coronary heart disease) later in life (Eriksson 2005). This hypothesis means that current poor environ- mental health conditions in early childhood will result in higher future health effects in terms of chronic diseases (Eriksson 2005). Areas for Future Research Given the importance of protecting the fetus from malnutrition and subsequent growth faltering, researchers need to pay more attention to quantifying environ- mental health problems during pregnancy. This research is particularly important in Sub-Saharan Africa,where the malaria-hookworm association results in maternal anemia and subsequent growth restrictions in utero. Understanding the role of environmental factors during pregnancy also has important implications for helping reduce the 4 million perinatal deaths that occur every year. More clinical trials are needed to test whether ALRIs are water washable, using hand washing or more comprehensive household hygiene interventions (Ahmed and others 1993).Additionally,clinical trials on indoor air pollution are also needed. Studies using deaths as end points would be important when health effects of both WSH and indoor air pollution are established,to overcome current methodological issues that lead to underestimating effects of these factors. Recognizing the links between infections and malnutrition, future studies should properly account for correlated multiple risk factors. Current studies on the effect of malnutrition on respiratory infections (for example, Fishman and others 2004) do not control for indoor air pollution exposure. Similarly, studies on the effect of indoor air pollution or poor water and sanitation rarely control for malnutrition. In the absence of such controls, the burden from a specific disease (for example, diarrheal infections) attributable to various risk factors (such as WSH or malnutrition) may exceed 100 percent. HOW LARGE IS THE ENVIRONMENTAL HEALTH BURDEN? 59 Key Messages In terms of child health, the environmental health burden is substantially higher than previously estimated if links through malnutrition are incorporated, espe- cially in those subregions, such as Sub-Saharan Africa and South Asia, where both malnutrition and poor environmental conditions coexist. In Sub-Saharan Africa, environmental health conditions which cause poorer fetus and child survival, may explain the counterintuitively lower malnutrition rates (when compared with South Asia) from poorer fetus and child survival. Even when the environmental health burden is conservatively estimated,a multi- plier effect exists for environmental health interventions, where investments addressing environmental risks (for example, lack of water and sanitation) not only reduce diarrhea mortality but also reduce mortality from malnutrition- related diseases and its consequences on educational attainment. Future research is needed on environmental health effects during pregnancy, additional disease transmission pathways, and better relative risk estimates to more accurately estimate the environmental health burden in children under five years of age. Notes 1 The authors' definition was fairly broad and covered 22 conditions, including war, violence, and alcohol. 2 For more details on Comparative Risk Assessment project publications, see http://www.who.int/healthinfo/boddocscra/en/index.html. 3 Fishman and others (2004) cite a recent meta-analysis that shows diarrheal incidence is 23 percent higher (12 to 35 percent, with a 95 percent confidence interval) in children with moderate to severe malnutrition. 4 The distinction between direct and indirect effects of environmental risk factors is concep- tual. In practice, when calculating the burden of disease related to these risk factors, the two effects may be difficult to separate. 5 According to a recent meta-analysis of 117 studies by van Geertruyden and others (2004), malaria deaths globally are underestimated by 2 million deaths. 6 This is an alternative complementary explanation to the Mills-Reincke phenomenon (Cairncross 2003) in addition to those provided in box 4.3. 7 All together, about 800,000 deaths annually are associated with indoor air pollution (Smith, Mehta, and Maeusezahl-Feuz 2004). C H A P T E R 5 Estimating the Environmental Health Burden and Costs at the Country Level IN PART I, new results from the World Health Organization (WHO) on the environmental health burden of disease were presented,and findings were discussed (Fewtrell and others 2007). These new calculations take into account the effects of environmental risk factors (including their effects on malnutrition). In this part, the analysis is taken to the next stage: estimating and translating this burden into economic costs at a country level. With the use of case studies from Ghana and Pakistan, earlier estimates are updated by calculating the effects of environmental risks (including the impacts through malnutrition) and by attempting to estimate the impacts on cognition and learning and on future work productivity and income (see appendix D for a detailed methodology). Existing Practice in Environmental Health Valuation The World Bank has made extensive use of valuation techniques to measure the costs of environmental degradation at the country level. Studies of the cost of environmental health effects are mostly based on quantification and valuation of disease-specific mortality and morbidity caused by environmental risk factors. In developing countries, these studies have focused mainly on environmental risks, such as inadequate supply of water, sanitation, and hygiene (WSH); indoor air pollution from household solid fuels; urban (outdoor) air pollution; and sometimes exposure to lead. To value health risks, these analyses have typically 61 62 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL looked at different health end points, such as deaths or years of life lost, years lived with disability, incidence of diarrheal and respiratory illness, number of hospital admissions, number of doctor visits, restricted-activity days, number of days of caregiver, and so on. In general, child mortality has been valued using the human capital approach-- that is, the present value of the average individual's income stream--and adult mortality has been valued using both the human capital approach and the value of statistical life, reflecting individuals' willingness to pay for a reduction in the risk of death. Morbidity has been valued by the cost of illness, including the cost of hospital admissions, doctor visits, restricted-activity days, time and income losses from illness, and days of caregiving. A fraction of the average wage is usually taken as the unit value of time lost to illness and caregiver's time (see appendix E for more details). The cost of environmental health risks at the country level has ranged from 1.2 percent of gross domestic product (GDP) in Tunisia to more than 4.0 percent of GDP in China (see figure 5.1). These costs have typically represented between 40 and 60 percent of total cost of environmental degradation (environmental health and natural resources). Lack of access to water supply and sanitation and poor hygiene often take the largest share of environmental health risks. In Bolivia, El Salvador, Guatemala, and Nigeria, indoor air pollution is particularly important. Results from these valuation studies--often quoted in various newspapers-- have been used to raise awareness on environmental issues in many countries, as well as to attempt to catch the attention of policy-makers. In Algeria, for example, the results were used at the highest political level and proved instrumental in increasing the budget for environmental protection by US$450 million. The studies have also sparked interest in additional work. For example, the governments of Algeria, the Arab Republic of Egypt, and Morocco have commissioned in-depth analyses of the costs of environmental degradation in the water sector. Tunisia is in the process of setting up a unit of environmental economics and policy in the Ministry of Environment to build capacity to undertake such analyses. Building New Estimates for Environmental Health Costs Research evidence discussed in chapter 2 highlights the importance of the vicious cycle of malnutrition and infections. This new valuation exercise seeks to estimate previously unaccounted costs associated with how environmental risk factors affect mortality through their effects on malnutrition. In the short term, this calcu- lation means including a portion of the deaths not usually attributed to environmental factors (for example, measles) that would not have occurred if the child had not been malnourished. In the longer term, it means assessing the impact (and subsequent costs) associated with impaired cognition and learning caused by malnutrition (which in turn was affected by environmental factors). ESTIMATING THE ENVIRONMENTAL HEALTH BURDEN AND COSTS AT THE COUNTRY LEVEL 63 FIGURE 5.1 Cost of Environmental Health Risks Tunisia Lebanon Colombia Peru Guatemala Egypt, Arab Rep. of Nigeria Iran, Islamic Rep. of Ghana Pakistan China 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 percent urban air pollution indoor air pollution water, sanitation, and hygiene Source: China (World Bank 2007a); Colombia, Ghana, Guatemala, Nigeria, Pakistan, and Peru (Country Environmental Analysis); Arab Republic of Egypt (World Bank 2002), Islamic Republic of Iran (World Bank 2005); Lebanon and Tunisia (Sarraf, Larsen, and Owaygen 2004). Notes: Economic burden includes the burden from mortality, morbidity, and cost of illness. Morbidity is usually valued using the human capital value approach. Adult mortality is valued by averaging the value of a statistical life approach and the human capital value approach. With the exception of China, the Islamic Republic of Iran, and Nigeria, child mortality is valued using human capital value only. With the exception of China, the value of a statistical life is obtained through benefit transfer of international studies. Water, sanitation, and hygiene mortality is estimated only for children; in China, such mortality excludes lack of sanitation and hygiene costs. Estimating and Valuing Health Impacts Accounting for Malnutrition Repeated infections in early childhood contribute to malnutrition. Malnutrition, in turn, makes children more vulnerable to repeated infections. As discussed in chapter 2, numerous studies have found that infections, especially diarrheal disease, in the first few years of children's life impair weight gains, often in the range of 20 to 50 percent (Adair and others 1993; Bairagi and others 1987; Black, Brown, and Becker 1984; Checkley and others 1997; Condon-Paoloni and others 1977; Molbak and others 1997; Rowland, Cole, and Whitehead 1977; Rowland, Rowland, and Cole 1988; Zumrawi, Dimond, and Waterlow 1987). Valuing Education and Cognition Costs Children exposed to poor environmental conditions often end up malnourished, which, in turn, impairs their cognitive development and educational performance. 64 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL This impairment affects labor productivity, thus reducing individual earnings. Furthermore, malnutrition can lead to delayed primary school enrollment and grade repetition, which, in turn, may affect lifetime income in terms of delayed labor-force entry. As discussed in chapter 2, a number of recent longitudinal studies (see appendix B) have investigated the effect of early childhood malnutrition on educational outcomes. These studies use data that follow children over time, usually from birth to well into primary or secondary school level, and they contain various anthropometric, schooling, and socioeconomic status indicators. The studies analyzed show that the change in grade attainment ranges from 0.12 years (mild stunting in the Philippines), to 0.91 years (moderate malnutrition in Brazil) and 0.7 years (in Zimbabwe) for each z-score (one standard deviation from the mean). The effect of malnutrition on learning productivity or achievement appears to be even larger, ranging from 0.8 years of grade equivalent in Guatemala and the Philippines for each z-score to 1.8 years for each z-score in severely malnour- ished children. Delayed primary school entry may vary between two and four months per z-score in the studies reviewed. Case Studies of Ghana and Pakistan The previous subsections mention the progress in the methodology for estimating costs associated with environmental health risks--that is, extending the analysis to incorporate the malnutrition-mediated effects of WSH-related infections, as well as estimating the longer-term costs of lower school performance and work productivity from impaired cognitive development and learning capacity. In this section, this adapted methodology is applied to two country case studies: Ghana and Pakistan. These countries were chosen for two reasons: (a) because each belongs to one of the two subregions that have the highest environmental health burden of disease in children under five and (b) because in each country the methodology estimating the effects of environmental factors had been applied earlier, providing an opportunity to compare the "old" (excluding malnutrition) and "new" (including malnutrition) estimates. In Ghana (see box 5.1)--a country of 22 million people--one in every nine children dies before reaching age five. Under-five mortality in the rural areas is 118 per 1,000 live births as compared with 93 for the urban areas. The country has an inadequate supply of piped water in terms of both population coverage rates and quality of services. On average, access to improved sanitation is also very low at 18 percent (World Bank 2006b), while in the largest cities about 30 percent of the population has household latrines. About 95 percent of households use solid fuels for cooking in Ghana (Ghana Statistical Service 2003). Results from household surveys reveal that among children under five years of age, 10 percent ESTIMATING THE ENVIRONMENTAL HEALTH BURDEN AND COSTS AT THE COUNTRY LEVEL 65 BOX 5.1 Basic Indicators for Ghana and Pakistan Ghana Population: 22.1 million Children under five years of age: 3.1 million GDP: US$10.7 billion Prevalence of malnutrition in children under five: 55 percenta Percentage of population having access to an improved water source: 75 percent Percentage of population having access to improved sanitation: 18 percent Energy from biomass products and waste (percentage of total): 69.1 percent Under-five mortality (per 1,000 live births): 112 Acute respiratory infection prevalence in children under five: 10 percentb Diarrhea prevalence in children under five: 17.9 percentb Malaria incidence per 1,000 (2003): 169.81b Pakistan Population: 155.8 million Children under five years of age: 19.0 million GDP: US$110.7 billion Prevalence of malnutrition in children under five: 70 percenta Percentage of population having access to an improved water source: 91 percent Percentage of population having access to improved sanitation: 59 percent Energy from biomass products and waste (percentage of total): 35.6 percent Under-five mortality (per 1,000 live births): 99 Acute respiratory infection prevalence in children under five: 24 percentb Diarrhea prevalence in children under five: 26 percentb Malaria incidence per 1,000 (2003): 0.80b Sources: Population figures for children under five are from UNICEF country data; malaria incidence figures are from RBM (2005); other information is from World Bank (2007h). a. Percentage of mildly, moderately, and severely underweight children under five years of age. b. Two-week (14 days) prevalence. were reported to have had symptoms of acute respiratory illness, and 18 percent had diarrhea in the two weeks preceding the survey (Ghana Statistical Service 2003).Vectorborne or water-based diseases, such as malaria and schistosomiasis, are rampant because of inadequate environmental management programs,although some success has occurred in the increased use of bednets. Ghana is the only Sub- Saharan country to have dramatically improved food security, according to its international pledges (FAO 2006). 66 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Pakistan is a country with 156 million inhabitants (see box 5.1). In Pakistan, population growth is coupled with trends in industrialization and urbanization. The under-five mortality rate for Pakistan is estimated to be 99 per 1,000 live births. Both rural and urban areas suffer from underinvestment in water supply and sanitation. Sewerage systems, where they exist, are usually poorly maintained and often contribute to the cross-contamination of water systems and the ground- water. In rural areas, more than 40 million people lack access to safe drinking water, while 60 million lack adequate sanitation. Inadequate quantity and quality of potable water and poor sanitation facilities and hygiene practices are associated with a host of illnesses, such as diarrhea, typhoid, intestinal worms, and hepatitis. About 70 percent of households use solid fuels for cooking in Pakistan (Federal Bureau of Statistics 2005). Dehydration attributable to diarrhea is a major cause of mortality among children in Pakistan. In 2001 to 2002, some 12 percent of children under five were reported to have suffered from diarrhea in the previous 30 days (Krupnick, Larsen, and Strukova 2006). Results for Ghana and Pakistan This part of the chapter presents and discusses the new estimates for the economic costs attributed to environmental risk factors in Ghana and Pakistan. The first subsection presents the results of the costs from environmental risk factors (excluding malnutrition), alluding also to estimates provided in previous studies. The next subsection presents the context and results for the total effects--including malnutrition--of these environmental risk factors. The final subsection presents estimates for the economic costs of poor school performance and lost work produc- tivity that are attributed to impaired cognitive development and learning from the infection-malnutrition links.1 The Burden of Environmental Risk Factors If No Change in Malnutrition Earlier studies have reported the cost of environmental health effects in Ghana (Larsen 2006) and the cost of environmental degradation in Pakistan (Krupnick, Larsen, and Strukova 2006). In this study, the same methodology is applied to recently published baseline data to estimate mortality from indoor air pollution and inadequate water supply, sanitation, and hygiene. This report uses WHO country estimates of cause-specific mortality for 2002, adjusted to reflect under- five child mortality rates in 2005. Solid fuel use prevalence is also updated from the Pakistan Social and Living Standards Measurement Survey 2004­05 (Federal Bureau of Statistics 2005). Estimated environmentally attributable fractions of child mortality in Ghana and Pakistan are presented in table 5.1. These fractions are estimated as follows. The attributable fraction for acute lower respiratory infection (ALRI) is related ESTIMATING THE ENVIRONMENTAL HEALTH BURDEN AND COSTS AT THE COUNTRY LEVEL 67 TABLE 5.1 Environmentally Attributable Fractions of Child Mortality, Keeping Malnutrition Unchanged Disease Ghana (%) Pakistan (%) Pneumonia/ALRI 55.2 47.8 Diarrhea 88.0 88.0 Malaria 42.0a 40.0 Sources: Compiled by World Bank team on the basis of Federal Bureau of Statistics 2005; Ghana Statistical Service 2003; Prüss-Üstün and Corvalán 2006. a. In the case of Ghana, the subregional environmentally attributable fraction for malaria of 42 percent may be an overestimate because of mosquito ecology, which demands that environmental manipu- lation be supplemented by bednets and indoor residual spraying. TABLE 5.2 Estimated Mortality in Under-Five Children from Environmental Risk Factors, 2005 Ghana Pakistan Mortality from Mortality from Under-5 Environmental Risk Under-5 Environmental Risk Child Factors (Excluding Child Factors (Excluding Disease Mortality Malnutrition) Mortality Malnutrition) Pneumonia/ALRI 11,800 6,508 97,100 46,435 Diarrhea 9,900 8,712 96,200 84,656 Malaria 22,600 9,492 1,300 520 Total 75,800a 24,712 402,600a 131,611 Source: Compiled by World Bank team on the basis of WHO 2004. a. Includes all under-five child mortality, not just mortality from ALRI, diarrhea, and malaria. to indoor air pollution and is estimated from the share of households using solid fuels and the relative risk of ALRI in children from such fuel use--that is, relative risk equals 2.3--as reported by Desai, Mehta, and Smith (2004).About 95 percent of households use solid fuels for cooking in Ghana (Ghana Statistical Service 2003), and 70 percent use such fuels in Pakistan (Federal Bureau of Statistics 2005). The attributable fraction for diarrhea is related to WSH and is based on regional estimates by WHO. The attributable fraction for malaria is related to water resource management and is from Prüss-Üstün and Corvalán (2006). Under-five child mortality in 2005 was 112 and 99 per 1,000 live births in Ghana and Pakistan, respectively (World Bank 2007h). Those figures imply about 76,000 annual deaths in Ghana and about 400,000 in Pakistan.2 Accordingly, the mortality attributed to different diseases (cause-specific mortality) is estimated from WHO (2004). The results show that when malnutrition is not fully accounted for, environment-related mortality is estimated at about 25,000 children per year in Ghana and 132,000 in Pakistan, or about one-third of all child mortality in each country (table 5.2). 68 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL The Burden of Environmental Risk Factors and Malnutrition This subsection presents the revised estimates in a country context for assessing the effect on mortality relating to environmental risk factors, such as poor WSH, indoor air pollution, inadequate water resource management, and malnutrition. First, the subsection presents child malnutrition status in the two countries to provide a sense of the prevalence of stunting and underweight in children (see table 5.3). Second, it examines the relationship between diarrheal infections and malnutrition status and estimates the malnutrition rates in the absence of infections. Third, the subsection estimates the impact of environmental factors--including the effect of infections on malnutrition--on health outcomes in children under five (mortality). Finally, a separate subsection addresses the effect of infection- malnutrition links on cognition and learning, as well as the accompanying effects on education and costs of lifetime income losses. Malnutrition in Ghana and Pakistan. Malnutrition prevalence is high in many developing countries. In Ghana, about 55 percent of children under the age of five years were mildly, moderately, or severely underweight in 2003,3 which is still 15 percentage points less than the average in Sub-Saharan Africa (Fishman and others 2004). About 70 percent of children under five in Pakistan were mildly, moderately, or severely underweight in 1991, with minimal change in 2001 to 2002.4 An important feature of malnutrition rates in developing countries such as Ghana and Pakistan is that it tends to rise very quickly in the first stages of life and then levels off or declines. More than 90 percent of the children in the Ghana Demographic and Health Survey (Ghana Statistical Service 2003) were not under- weight in their first three months of life; moderate and severe underweight was TABLE 5.3 Malnutrition Rates in Children under the Age of Five Malnutrition Rate (%) Indicator Ghana Pakistan Underweight (weight-for-age) Mild (-1 to -2 SD) 33 29 Moderate (-2 to -3 SD) 17 27 (26) Severe (less than -3 SD) 5 14 (12) Total (less than -1 SD) 55 70 Stunting (height-for-age) Mild (-1 to -2 SD) 28 22 Moderate (-2 to -3 SD) 19 20 (19) Severe (less than -3 SD) 11 30 (18) Total (less than -1 SD) 58 72 Sources: Compiled by World Bank team on the basis of data from Ghana Statistical Service 2003; National Institute of Population Studies 1992. Figures in parentheses are based on PIDE 2003. ESTIMATING THE ENVIRONMENTAL HEALTH BURDEN AND COSTS AT THE COUNTRY LEVEL 69 practically nonexistent in this age group. The prevalence of moderate and severe underweight peaked at over 35 percent in children 10 to 13 months of age, however, and then gradually declined to the age of five years. The prevalence of mild under- weight reached about 40 percent in children age 18 to 21 months and then leveled off to the age of five years.A similar evolution can be inferred from Pakistani data. More than 97 percent of the children were not underweight in their first month of life, but by 9 to 10 months of age, the prevalence of moderate and severe under- weight rose rapidly to 40 percent, stabilizing to a mean of 45 percent for children age 18 to 59 months. In global comparisons, too, children 1 to 3 months of age in Ghana and Pakistan have similar weight-for-age scores (indicator for underweight) to the interna- tional reference population, but the scores fall considerably below the international reference during the first year of life (see figure 5.2). By the age of 10 to 15 months, children from both countries are significantly underweight, with more than 70 percent malnourished. Health effects of malnutrition. The malnutrition-attributable fractions for different diseases are calculated for Ghana and Pakistan (see table 5.4) using estimates of increased risk of cause-specific mortality and all-cause mortality in children under five years of age with varying levels of malnutrition. The results indicate that, in Ghana, 44 to 63 percent of under-five mortality from pneumonia-ALRI, diarrhea, measles,malaria,and other causes is due to malnutrition.The corresponding attrib- utable fractions for Pakistan are in the range of 50 to 68 percent because of higher underweight malnutrition rates in Pakistan. In addition to these malnutrition- related mortalities,Fishman and others (2004) include 100 percent of protein-energy malnutrition mortality and morbidity and a share of mortality from perinatal conditions (low birth weight) associated with mothers' low pre-pregnancy body mass index (less than 20 kilograms per square meter). About 16 percent of infants had low birth weight (less than 2,500 grams) in Ghana in 2003, as did about 19 percent of infants in Pakistan in the past decade (UNICEF 2005). Infections and malnutrition. Household survey data from Ghana and Pakistan provide some insights into the close links between diarrheal prevalence and under- weight in children.In Ghana,the two-week diarrheal prevalence was 15 to 18 percent in children under five years and 20 to 22 percent in children under three years, according to the three most recent demographic and household surveys. This diarrheal prevalence differs dramatically by degree of underweight in Ghanaian children age 3 to 23 months (panel a of figure 5.3), with prevalence in severely underweight children age 3 to 11 months reaching 3.5 times higher than that in children who are not underweight. A similar pattern, although with less drastic differences across malnutrition categories, exists in Pakistan. With a two-week diarrheal prevalence of 15 percent in children under five years, the prevalence in 70 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL FIGURE 5.2 Weight-for-Age Distribution of Children in Ghana and Pakistan a. Ghana -4.5 -3.5 -2.5 -1.5 0 1.5 2.5 3.5 4.5 standard deviation from the mean weight of the international reference population age 0-3 months age 10-15 months b. Pakistan -5.5 -4.5 -3.5 -2.5 -1.5 0 1.5 2.5 3.5 4.5 5.5 standard deviation from the mean weight of the international reference population age 0-3 months age 10-15 months Sources: Compiled by World Bank team on the basis of data from Ghana Statistical Service 2003; National Institute of Population Studies 1992. TABLE 5.4 Malnutrition-Attributable Fractions of Child Mortality Cause of death Ghana (%) Pakistan (%) Pneumonia-ALRI 54.1 60.1 Diarrhea 63.2 67.8 Malaria 57.6 63.1 Measles 43.8 50.5 Protein-energy malnutrition 100.0 100.0 Other causes (excluding perinatal conditions) 55.8 61.5 Source: Compiled by World Bank team on the basis of Fishman and others 2004. ESTIMATING THE ENVIRONMENTAL HEALTH BURDEN AND COSTS AT THE COUNTRY LEVEL 71 FIGURE 5.3 Two-Week Diarrheal Prevalence Rate by Age and Underweight Status in Ghana and Pakistan 60 a. Ghana en 50 a childrfo 40 diarrhe 30 age ntec with 20 per 10 0 age 3-11 months age 12-23 months not underweight mildly underweight moderately underweight severely underweight b. Pakistan 30 en 25 a childrfo 20 diarrhe 15 age ntec 10 with 5 per 0 age 4-11 age 12-23 age 24-35 age 36-47 age 48-59 months months months months months not underweight mildly underweight moderately underweight severely underweight Source: Compiled by World Bank team on the basis of data from Ghana Statistical Service 2003; National Institute of Population Studies 1992. severely underweight children is 50 percent higher than in children who are not underweight (panel b of figure 5.3). Because these data are cross-sectional (both indicators measured at the same point in time), no causal relationship between underweight and diarrheal preva- lence can be determined. However, the pattern is consistent with research findings in other countries that show that diarrheal infections lead to weight gain retar- dation in children under five years. This research literature finds that the weight gain retardation often is in the range of 20 to 50 percent. Diarrheal infections are 72 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL most frequent during ages 6 to 11 months and 12 to 23 months, so weight gain retardation from diarrheal infections can be expected to be greatest in this period; in Ghana and Pakistan, it is particularly great up to the age of 12 months, during which underweight malnutrition increases rapidly. Malnutrition in the absence of infections: the counterfactual. Results from the research literature are used to estimate the retarding effect that diarrheal infections have on weight gain. This information sheds light on an important question: what would be the counterfactual level of malnutrition in children who were not exposed to diarrheal infections? Current underweight malnutrition rates in children under five years from the 2003 demographic and health survey in Ghana (Ghana Statistical Service 2003) and the 1991 survey in Pakistan (National Institute of Population Studies 1992), as well as the estimated malnutrition rates in the absence of diarrheal infections that retard weight gain are presented in figure 5.4. Basically, the figures for both Ghana and Pakistan starkly show that in the absence of diarrheal infections, chil- dren under five years in both countries would climb up the nutritional status ladder--with practically no severely underweight children in either country and with the number of moderately underweight also substantially reduced.5 Calculating the burden of environmental risk factors, including malnutrition- mediated effects. Information from the previous subsections provides the necessary building blocks for estimating the effects of environmental health risk factors. The counterfactual population exposure rates estimated are used to compare the current world with a world without environmental risk factors in order to compute the environmental health risk--that is,the attributable fraction (table 5.5).Appendix D (equation D.2) is used. In making these estimates, the study takes a conservative approach. In fact, the parameters on which attributable fractions are based are likely to be affected by omitted variable bias (see appendix E). To deal with this problem, the study attempted to correct such parameters by using formulas from econometrics literature. The methodology is graphically represented in figure 5.5. The indirect effect of environmental health risks is equivalent to the difference in mortality between the malnutrition-attributable fractions of different diseases (right bar) and the consequences of malnutrition in the absence of diarrheal infec- tions (center bar). Notice that the joint direct and indirect effects of environmental health risks (left bar) are not necessarily equal to the sum of the two effects computed separately. This problem is common in the presence of risk factors that interact with each other (see box 5.2). Figure 5.5 shows that additional pneumonia-ALRI, diarrhea, malaria, measles, protein-energy malnutrition, and other deaths repre- sented in the light-gray area have not been accounted for previously. The results (shown in table 5.5) demonstrate that when both direct and indi- rect effects from diarrheal infections on malnutrition are included, the mortality ESTIMATING THE ENVIRONMENTAL HEALTH BURDEN AND COSTS AT THE COUNTRY LEVEL 73 FIGURE 5.4 Underweight Malnutrition Rates in Children with and without Diarrheal Infections in Ghana and Pakistan a. Ghana 120 (%)eatr 5 0 100 2 17 30 80 60 33 malnutrition 40 eightw 68 45 20 under 0 with infections without infections severe moderate mild not underweight b. Pakistan 120 0 (%)eatr 100 14 8 80 27 42 60 malnutrition 30 40 eightw 50 under 20 30 0 with infections without infections severe moderate mild not underweight Sources: Compiled by World Bank team on the basis of data from Ghana Statistical Service 2003; National Institute of Population Studies 1992. 74 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL TABLE 5.5 Environmentally Attributable Fractions and Child Mortality with Malnutrition-Mediated Effects Ghana Pakistan Revised Effects Revised Effects (Number of Attributable (Number of Attributable Disease Deaths) Fraction (%) Deaths) Fraction (%) Pneumonia-ALRI 7,375 62.5 58,260 60.0 Diarrhea 9,009 91.0 88,504 92.0 Malaria 14,035 62.1 826 63.5 Measles 58 28.8 8,316 30.8 Protein-energy malnutrition 400 40.0 1,359 45.3 Other causes (excluding perinatal conditions) 4,826 38.3 30,164 39.9 Total 35,702 47.1 187,429 46.6 Source: Compiled by World Bank team. FIGURE 5.5 Calculating Revised Estimates (Indirect and Direct Effects) indirect effect of environmental consequences health risks of malnutrition (actual) consequences of malnutrition direct effect (counterfactual) of environmental health risks disease or illness disease or illness caused by caused by malnutrition environmental risks Source: Compiled by World Bank team. from environmental factors is estimated at 36,000 in Ghana and 187,000 in Pakistan annually, or about 47 percent of all child mortality in both countries. Translating those deaths into economic costs will help quantify the magnitude of environmental health risks in real terms. In this analysis, the human capital approach (a conservative approach) is used to value this mortality. For Ghana, the earlier estimates of costs of environmental factors add up to C/ 370 million (US$412 million), which is equivalent to 3.8 percent of the country's GDP.6 ESTIMATING THE ENVIRONMENTAL HEALTH BURDEN AND COSTS AT THE COUNTRY LEVEL 75 BOX 5.2 Attributable Fractions and Burden of Disease When Multiple Risk Factors Are Present The concept of an attributable fraction (AF) is the essential building block of the burden of disease methodology. An AF tells what proportion of a disease is caused by any given risk factor or combination of risk factors. To take into account the total effects of environmental risks on child health--including those effects mediated through malnutrition--one needs to allow for the simultaneous presence of multiple risk factors. Whether one or more risk factors are being controlled should be clearly stated when one calculates AFs. Tables 5.1, 5.4, and 5.5 present the fraction of deaths that are in turn attributed to (a) environmental health risks alone (table 5.1), (b) malnutrition alone (table 5.4), and (c) all relevant environmental health risks, including their effect through malnutrition (table 5.5). This calculation is made for specific causes of death (for example, ALRI, diarrhea, malaria, measles, and protein-energy malnutrition). AFs in tables 5.1 and 5.4 measure the cases of death from a disease that would be eliminated by removing a risk factor from the population while leaving the exposure to other risk factors unchanged. Take Ghana as an example. The fraction of ALRI attributable to indoor air pollution is 55.2 percent (table 5.1), and the fraction of ALRI attributable to malnutrition is 54.1 percent (table 5.4). One could also compute the fraction of a disease that could be avoided if one eliminated both risk factors. In the case of Ghana, eliminating both indoor air pollution and malnutrition would reduce ALRI by nearly 80 percent. Notice that the sum of AFs calculated for individual risk factors can be more than 100 percent and can differ considerably from the AF calculated for the risk factors jointly. This result is common, as shown in Chatterjee and Hartge (2003). This study controls for risk factors jointly (for example, environmental health risks and malnutrition) but allows malnutrition to be reduced only to the level that would be present without diarrheal infections. It answers this question: by how much would a disease (for example, ALRI) be reduced if there were no environmental health risk (for example, indoor air pollution) and if the burden of ALRIs that was caused by lack of water, sanitation, and hygiene and was not related to indoor air pollution were avoided? A problem one faces is the difficulty in saying what fraction of the resulting AF is attributable to each risk factor. The "new" (including malnutrition) costs amount to C/ 540 million (or US$595 million), which is equivalent to 5.6 percent of Ghana's GDP--a 47 percent increase over previous estimates.7 In Pakistan, the earlier estimated costs of environmental factors amount to PRs 195 billion (US$3.2 billion), which is equivalent to 2.9 percent of GDP. The"new"(including malnutrition) costs raise that figure to 4.1 percent of the country's GDP--a 41 percent increase over previous estimates. 76 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL The lower cost in Pakistan results from its lower child mortality rate and its lower birth rate, which gives a lower share of children under five in the total population. Impact on Education: Affecting School Performance and Future Work Productivity Chapter 2 discussed another indirect effect of the link between repeated infections in early childhood and malnutrition: the impairment of cognitive development, resulting in poor school performance and, subsequently, in a loss of lifetime earn- ings. Many studies have documented the effect of early childhood malnutrition on cognitive function, school enrollment, grade repetition, school dropout rates, grade attainment, and test scores among school-age children (see appendix B). The majority of these studies use stunting (low height for age) as a measure of malnutrition, which is considered an indicator of chronic nutritional deficit. Here, for the first time, this study attempts to estimate the loss in lifetime earnings from stunting in Ghana and Pakistan and the share of this loss that could be avoided if children were not exposed to repeated diarrheal infections in early childhood. Effects of malnutrition on education outcomes applied to Ghana and Pakistan for estimation of the cost of malnutrition are presented in table 5.6, stratified by height-for-age deficit. The magnitude of these effects is based on the review of longitudinal studies. Effects analyzed include those on (a) grade attainment, (b) school achievement (learning productivity) in terms of grade equivalents, (c) delayed primary school enrollment, and (d) grade repetition. The latter two effects result in delayed labor-force entry. Psacharopoulos and Patrinos (2004) provide a review of studies of returns to investment in education. Private returns are estimated on the basis of private cost of schooling and lifetime income. Social returns include social cost of providing TABLE 5.6 Effects of Malnutrition on Education Height-for-age z-Score Effects of Stunting -1 to -2 -3 to -2 Lower than ­3 Loss in grade attainment (school years) 0.50 1.0 1.5 Loss in school achievement (grade equivalents) 0.8 1.6 3.6 Total grade equivalent losses (grade attainment and achievement) 1.30 2.60 5.10 Delayed enrollment (years) 0.4 0.8 1.2 Grade repetition (percentage of students with one repetition) 2 5 9 Delayed labor-force entry (years) caused by delayed enrollment and grade repetition 0.42 0.85 1.29 Sources: Compiled by World Bank team on the basis of review of longitudinal studies. For more details see appendix B. Note: Stunting is defined as height-for-age z-score lower than -1. ESTIMATING THE ENVIRONMENTAL HEALTH BURDEN AND COSTS AT THE COUNTRY LEVEL 77 schooling, which is most often significantly higher than private cost because of subsidized provision of education. Social returns are therefore lower than private returns to investment in education.8 Returns to one additional year of schooling are calculated estimates from Psacharopoulos and Patrinos (2004). On average, the return to one additional year of schooling is 9.7 percent. The rate of return is 11.7 percent in Sub-Saharan Africa and 9.9 percent in Asia. These rates of return are applied to Ghana and Pakistan, respectively, to estimate the cost of impacts of malnutrition on educa- tion outcomes. To estimate the lifetime income for children in Ghana and Pakistan, the analysis considered existing income differentials between children in different nutritional status (from not malnourished to severely malnourished). This differentiation is important because malnourished children live disproportionately in poorer house- holds; given that poverty is correlated with lower socioeconomic status, such as lower educational attainment of parents, it is, in turn, correlated with lower educa- tional attainment of their children. Malnourished children are more likely to come from lower-income households than are children who are not malnourished. Therefore, the income potential of malnourished children is affected not only by the action of malnutrition on educational performance but also by the socioeco- nomic status of the household they grew up in (see appendix D for methodology on calculations).9 About 630,000 children are born annually in Ghana.10 Thirty-five percent of these children are moderately or severely stunted by the time they reach one to two years of age, and an additional 30 percent are mildly stunted (height-for-age z-score ­1 to ­2). On the basis of these prevalence rates, the effect of stunting on educational outcomes, income losses from loss in education, and a labor-force participation rate of 75 percent, the annual cost of stunting that is attributable to early childhood diarrheal infections is estimated to be about 3.8 percent of GDP. About 3.66 million children are born annually in Pakistan.11 According to the Pakistan National Nutrition Survey 2001­2002 (PIDE 2003),prevalence of moderate and severe stunting in children under five years was 19 and 18 percent,respectively. No prevalence rate is reported for mild stunting. A rate of 25 percent is therefore assumed,in accordance with cross-country comparison.On the basis of these preva- lence rates, the effect of stunting on educational outcomes, income losses from educational losses, and a labor-force participation rate of 60 percent, the annual cost of stunting attributable to early childhood diarrheal infections is estimated to be about 4.7 percent of GDP (see appendix D for methodology on calculations). Conclusion Malnutrition-mediated effects of environmental risk factors are estimated to have a considerable social and economic burden in Ghana and Pakistan (see figure 5.6 78 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL FIGURE 5.6 Final Results of Ghana and Pakistan Case Studies 10.0 9.3 8.8 9.0 8.0 7.0 GDPfo 6.0 5.6 age 5.0 4.1 ntec 3.8 4.0 2.9 per 3.0 2.0 1.0 0.0 direct effect only direct effect + indirect direct + indirect + (malnutrition-mediated) education effect Ghana Pakistan Source: Compiled by World Bank team. and table 5.7). For Ghana, the cost of total health effects (including malnutrition) amounts to C/ 537 million (US$595 million).When longer-term effects on educa- tion and income are included, this figure may reach as high as C/ 894 million (US$1 billion)--equivalent to 9.3 percent of GDP. Similarly, for Pakistan, the cost of total health effects from environmental risk factors amounts to PRs 278 billion (US$4.6 billion). And with longer-term effects on education and income, this figure may reach as high as PRs 595 billion (US$9.9 billion), or equivalent to 8.8 percent of the country's GDP. Put in perspective, the costs of these total effects are nearly three times higher than the cost of the effect of environmental risk factors only (excluding malnutrition). The issue of temporality is important to recognize while analyzing these results. These figures indicate that the health benefits from reducing environmental risks are high even if such benefits will accrue far into the future. With the focus on children under five years of age, environmental health interventions implemented today to avoid deaths and illness and to increase education attainment will generate returns only a few years from now. Conventional decision making in developing countries often lacks this longer-term perspective when policy-makers choose appropriate interventions to benefit child health. By providing a better under- standing of the links between environmental risks (the causes) and child health ESTIMATING THE ENVIRONMENTAL HEALTH BURDEN AND COSTS AT THE COUNTRY LEVEL 79 TABLE 5.7 Annual Cost of Direct and Indirect Impact of Environmental Risk Factors in 2005 Ghana Pakistan Cost Cost Cost Cost Cost Annual Cost (US$ (% of Annual (PRs (US$ (% of Deaths ( million) million) GDP) Deaths billion) million) GDP) Estimation excluding malnutrition-mediated effects Mortality effects 24,712 371 412 3.84 131,611 195 3,250 2.90 Estimation including malnutrition-mediated effects Mortality effects 35,702 537 595 5.55 187,429 278 4,633 4.13 Education effects 367 407 3.79 317 5,281 4.71 Total effects 904 1,002 9.34 595 9,914 8.84 Source: Compiled by World Bank team. Note: Ghanaian new cedi. (the effect), as well as the timing of these effects, this study hopes to influence the way that governments choose between short-term and long-term interventions. Given the importance of time in these results, the methodology relating to discounting becomes very relevant. In this report, the calculations used a 3 percent discount rate, which is consistent with the rate used in WHO burden of disease calculations. The purpose of discounting is to recognize a general preference for benefits that happen earlier. Different discount rates are used in different situa- tions. When one looks at public sector decisions, a social discount rate is preferred to a private one (that is, the rate a commercial bank would apply on a loan). Pearce and Ulph (1999) report an estimated social rate of return on investments for industrial countries between 2 and 4 percent. In appendix D, a sensitivity analysis applied to Ghana varies the discount rate from 3 percent to 5 percent (a more conservative discount rate) to show how the results would vary. In the case of Ghana, a 5 percent discount rate would reduce the total cost estimates from 9.2 percent to 4.6 percent of GDP, and in Pakistan it would reduce them from 8.6 percent to 4.3 percent of GDP. The considerable difference is explained by the fact that a large part of the benefits from interventions (for example, higher incomes owing to education) will happen several years from now. Box 5.3 provides guidelines for policy-makers in interpreting the results. Next Steps For policy-makers to prioritize among various environmental health interven- tions, further analyses and studies are required to guide their decisions along two 80 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL BOX 5.3 How Policy-makers Should Interpret These Results Using Ghana and Pakistan as examples, this study estimates the annual costs associated with environmental risks for children under five years of age. These estimations, carried out at a national level, provide a sense of the magnitude of these costs (about 9 percent of a country's GDP) to policy- makers and highlight the urgent need to position environmental health at the center of all child survival strategies. These figures help carry the message that environmental risks to child health are considerable and that such risks impose a significant economic burden. That being said, these figures do not help policy-makers prioritize (in terms of costs and capacity to implement) among the various types of environmental health interventions that are available to them. For example, if policy-makers in Pakistan choose to address diarrhea in children, should they invest in sanitation, in hygiene promotion, in increased water availability, or in some combination of these interventions? These choices, as well as the underlying framework that helps determine their priority, are important because they help policy-makers translate the environmental health "agenda" into concrete programs and actions addressing environmental health. major paths: one relating to costs and benefits of specific interventions, which has implications for budgetary and resource allocation decisions, and the second relating to institutional capacity, which helps ascertain the required roles and responsibilities of various agencies in implementing these interventions. These analyses and studies need to be carried out separately for specific countries because the priorities may differ according to the country's current level of development, its population health profile and the relative exposures to different disease, and its capacity to implement specific interventions. In terms of economic approaches, two main methods are typically used: cost- effectiveness analysis and cost-benefit analysis. Cost-effectiveness analysis allows researchers to compare policy options that provide outcomes measurable with a common metric (for example,per child life saved).An example would be an assess- ment of how costs of providing latrines to a village compare with costs of providing access to safe drinking water per child life saved. Cost-benefit analysis, in contrast, can be applied to compare options that do not necessarily provide outcomes meas- ured using the same metric by reducing the benefits of each option to a common numeraire (that is, by assigning a monetary value for the various policy options). The choice of method used to carry out an economic analysis at a country level must be made after examining the availability of relevant data. For Ghana and Pakistan, for example, this choice means looking at specific environmental risks,identifying possible interventions that would address them,and then assessing ESTIMATING THE ENVIRONMENTAL HEALTH BURDEN AND COSTS AT THE COUNTRY LEVEL 81 how much data on the costs and benefits of these interventions are available. In other country-level studies, cost-benefit analyses have helped governments prior- itize among available interventions. In Peru, for example, the World Bank's country environmental analysis reported that drinking-water disinfection and hand- washing programs had the highest cost-benefit ratio for addressing water- and sanitation-related risks, while shifting households from unimproved to improved cookstoves was the best recommendation for interventions to address indoor air pollution (World Bank 2007g). Policy-makers in developing countries also need to choose programs to address environmental risks in accordance with their country's institutional capacity to implement these interventions. The next chapter provides some discussion of the generic roles that national and local governments can play in delivering and managing environmental health interventions. A more in-depth country-level institutional analysis is required about what coordination mechanisms exist between ministries and how mandates and budgets are assigned. Such an insti- tutional analysis would provide guidance on how to better define the roles and responsibilities of different agencies related to addressing public environmental health services and how to correct incentive structures to raise the profile of the environmental health agenda. Key Messages When malnutrition-mediated health effects attributed to environmental risk factors are included, the total costs of the environmental health burden are at least 40 percent higher for both Ghana and Pakistan.When longer-term effects of malnutrition (partly attributed to environment-related infections) on educa- tion and income are considered, the estimated annual cost may be as high as 9 percent of a country's GDP. This social and economic burden is not trivial, and it highlights the urgent need for policy-makers to position environmental health at the center of all child survival strategies. Next steps to translate these messages into concrete actions include carrying out cost-effectiveness and cost-benefit analyses at a country level to guide decision- makers in prioritizing among the various available interventions. An in-depth country-level institutional analysis would also be needed to assess the capacity of the country to implement these interventions. Notes 1 Details on the steps in the methodology to calculate these costs for Ghana can be found in appendix D. 2 These estimates of mortality are based on crude birth rates and population in 2005. 82 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL 3 The Ghana Demographic and Health Survey 2003 (Ghana Statistical Service 2003) does not provide figures for mild underweight (­1 to ­2 standard deviation). This figure was therefore calculated from the survey's household data. 4 Underweight malnutrition rates in Pakistan were practically the same in 1991 (National Institute of Population Studies 1992) and 2001 to 2002 (PIDE 2003). The original survey data from the nutrition survey (PIDE 2003) were not readily available for this study, so data from the Pakistan Demographic and Health Survey 1990/91 (National Institute of Population Studies 1992) were used instead. The survey for 2006 was being compiled at the time of this study. The 2006 survey may show malnutrition rates and diarrheal disease that differ from some of the patterns found in the earlier survey. 5 This finding is based on the fact that 35 percent of children's weight deficit, relative to the international reference population, is from infections, which is well within the range found in research studies. 6 Ghanaian currency is denominated in new cedi (C/ ). The Ghanaian new cedi--U.S. dollar exchange rate is from 2007 (US$1 = C/ 9,375). 7 A working life from 15 to 65 years is applied. Annual income is approximated by GDP per capita. Income is discounted at 3 percent. Real income growth is 2 percent per year. 8 Social returns do not (in almost all studies) include possible external benefits of education. 9 Income is discounted at a social rate of time preference of 3 percent per year. Social cost of providing a year of schooling is assumed to be 25 percent to 50 percent of average annual income per person. 10 By the time the children turn five years old, 11 percent of them will have died. Under- five child mortality rate in Ghana is 112 per 1,000 live births in 2005 according to World Bank (2007h). 11 By the time the children turn five years old, 10 percent of them will have died. Under-five child mortality rate in Pakistan is 99 per 1,000 live births in 2005 according to World Bank (2007h). P A R T I I I Experiences C H A P T E R 6 Approaches to Environmental Health THE PRECEDING CHAPTERS PROVIDE EVIDENCE of the importance of environmental health interventions in addressing child health. Chapter 2 explained the science and the epidemiological underpinnings of the malnutrition- infection cycle and its implications for a child's health both in utero and in early childhood. Chapter 3 illustrated how environmental health interventions can appropriately complement and supplement other health and environmental programs that address child health. Chapter 4 then went on to demonstrate the scale of the environmental health burden in children under five and how that burden is especially important in some subregions with high levels of malnutri- tion and poor environmental conditions. Using two country examples--Ghana and Pakistan--chapter 5 then analyzed the consequent economic burden that these environmental health risks pose. Clearly, the issue of environmental health is critical for child health--with the "why,""to what extent," and "where" explained in earlier chapters. What remains is the "how": how can these environmental health issues be addressed success- fully in developing countries? This chapter begins with a historical review of environmental health, outlining the trends in the evolution of environmental health functions in developed countries and highlighting how circumstances have led to the relative neglect of environmental health in the development agenda. The chapter provides illustrative examples of how different developing countries 85 86 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL have incorporated environmental health activities within other health, nutrition, and infrastructure programs. Identifying some common elements from these examples, the chapter then discusses--through an institutional lens--the roles that national and local governments can play in delivering and managing envi- ronmental health interventions. History of Environmental Health Environmental health is among the earliest public health activities on record. Through the ages, today's developed countries have invested in environmental health interventions, focusing on sanitation, cleanliness, and hygiene. The history of environmental health can be loosely divided into four generations--with shifts in the nature and focus of interventions. Looking at the evolution of environ- mental health functions over the past few centuries helps to understand the context behind today's environmental health agenda in developing countries. First Generation: Origins of "Miasma" Theory In Classical Greece, the source of disease was attributed to bad air or "miasmatic odors,"and interventions sought to fight disease through cleanliness and hygiene (Rosen 1958). Subsequently, as the Roman Empire conquered the Mediterranean world, it incorporated the ideas of cleanliness and hygiene into infrastructure projects, which included the great sewerage system (the cloaca maxima), baths, fresh water supplies, health facilities, suitable latrines, proper ventilation, and the organization of the health sector (Rosen 1958). For the English sanitary reformers of the mid-19th century, miasma continued to prevail in explanations of disease transmission. In 1854, however, John Snow traced an outbreak of cholera in London to a common drinking-water source, refuting the miasma hypothesis. The 1871 Report of the Royal Sanitary Commission listed 11 essential health services, mostly environmental in nature, that covered access to improved water and sanitation, provision of a sewerage system, health- iness of dwellings, and inspection of food (Kotchian 1997; WHO 2007e). Second Generation: Sanitation, Vaccinations, Multisectoral Interventions Refuting the dominant theory of miasma, Louis Pasteur, by the late 19th century, had developed the germ theory that proposes that microorganisms are the cause of many diseases. This theory paved the way for more targeted and cost-efficient environmental and public health interventions, such as chlorination of water supplies, pasteurization of milk (see box 6.1), vaccinations, improvements in food safety and vector control, and introduction of a more effective sewerage system (Berg 2005). Many of the newly adopted health protection policies were APPROACHES TO ENVIRONMENTAL HEALTH 87 BOX 6.1 Combating Disease through Improved Milk In the early 20th century, one avenue of attack on preventable deaths of children was through improvements in milk. In New York, cross-sectoral collaboration between different government and nongovernmental agencies began as they worked together to combat disease. Nurses visited mothers of newborn babies and promoted hygiene and breastfeeding of babies (Moser Jones 2005). Organized milk stations were set up to provide clean and preheated milk to inner-city inhabitants. Inspectors began to scrutinize milk production facilities, farms, and associated workers to ensure hygienic conditions for milk (Moser Jones 2005). Railway companies were advised to have their milk cars properly refrigerated. The city authority began to supervise the distribution and sale of milk. The New York City Health Department took the lead in coordinating the efforts of its mandate to secure a safe and hygienic food chain, from farm to child. In many cases, it worked closely with the courts to enforce regulations. Today, different mandates in the legal framework perform the same task in the different stages of the production process (Rosen 1958). based on field evidence and the result of multisectoral collaboration. After local health protection activities had proved successful and enough local level expe- rience had been accumulated, the state played a key role in scaling up the interventions (Rosen 1958). Third Generation: Hands-on Scaling Up In the beginning of the 20th century, much of environmental health work focused on providing rural sanitation (Berg 2005). This period was characterized by a community-based approach in which public health nurses made visits to homes and then informed sanitarians, who would address issues related to food safety, garbage collection, sanitation, vector control, and access to drinking water (Berg 2005). In the United States, the result was the most rapid reduction in morbidity and mortality (especially among children) in history (Berg 2005; Cutler and Miller 2005).With time,scientific knowledge evolved,containment measures became more sophisticated,and some infectious disease outbreaks were gradually brought under control with improved sanitation and the discovery of new vaccines (WHO 2007e). Fourth Generation: The Invisible Regulator By the middle of the 20th century, the institutional framework that regulates envi- ronmental health issues had become well established in the developed world (Berg 2005). Mandates and responsibilities for the different agencies involved in envi- ronmental health had been established. With dramatic improvements in 88 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL environmental health conditions after World War II, the work of sanitarians evolved from more visible hands-on type of work to less visible regulation-based functions as health inspectors (Berg 2005). Agenda Falling through the Cracks The middle of the 20th century saw a rise in ecological awareness in the devel- oped world, heralded by the publication and widespread reading of Rachel Carson's (1962) Silent Spring and the subsequent occurrence of numerous industrially related public health outbreaks. Growing concern about environmental exposures to industrial emissions led to the passage of numerous environmental laws and the creation of the independent environmental agencies. Increased emphasis on environmental laws and compliance ensued, and environmental health staff members were moved into these new environmental agencies. Over time, these staff members identified more with environmentalists, often forgetting their public health roots as they morphed into their new roles relating to environmental regu- lation (Kotchian 1997). Their counterparts in state-level health agencies--epidemiologists and other health personnel--also began to lose their public health roles as budgets and agendas began to emphasize treatment over prevention. With a growing atten- tion to health care delivery and hospital regulation, public health programs lost visibility and influence in the agenda of state health agencies (Kotchian 1997). Decades later, the same pattern has been followed in developing countries. At one end is a growing environmental movement with the creation of ministries of environment and accompanying policies and regulations, while at the other end are health ministries and state health departments engaged in strengthening health systems and scaling up health sector programs that focus mainly on patient care. Environmental and infrastructure investments are not usually considered to be key determinants of human health by governments and health ministries-- the latter tend to think of health investments in terms of number of hospital beds and doctors per person (Lvovsky 2001). In the process, the environmental health agenda has fallen between the cracks. It has become an institutional orphan as it is increasingly neglected in the programs of both the health and environment ministries. Numerous other factors compound this relative neglect of the environmental health agenda in developing countries. High-level political commitment is a crit- ical issue that is often lacking or, if it exists, does not get translated into effective policy discussions and actions. Additionally, both donor agencies and govern- ment investments work in isolated silos where investments in agriculture, nutrition, infrastructure, education, the environment, and health are often not well linked. Providing water and sanitation in urban areas, for example, has never enjoyed APPROACHES TO ENVIRONMENTAL HEALTH 89 high priority for donors (Satterthwaite 2007), and funding for child health programs often includes only actions that lie within the health systems framework of the country. Country financing is usually allocated by sectors or ministries, so unless one sector takes the lead, no large-scale action can follow (Shekar and others 2006). Part of the problem stems from the fact that health ministries tend to focus on interventions in the health care system itself, rather than on primary preven- tion of environmental risks through infrastructural and behavioral interventions. Institutional coordination at both the national and local levels is often weak. At the national level, poor intersectoral coordination across ministries means that environmental health issues--which are inherently cross-sectoral in nature--are not sufficiently addressed, and efforts remain ad hoc and often uncoordinated. Many of the environmental health interventions important for child health should be carried out at both the household and the community levels. In many devel- oping countries, local government authorities--who are the best positioned to deliver environmental services and proper hygiene awareness--fall short in their efforts because of inadequate funding, weak revenue bases, and lack of staff members with expertise in environmental health. Local governments are also impoverished, powerless, and suffering from weak management capacity. In part, this state can be explained by the colonial legacy of imported codes and regula- tions for planning, land-use management, and buildings that were applied only to areas of the colonial city inhabited by "nonnatives." Also, however, the inade- quacies of service provision can be attributed to a very rapid growth experienced around the time of independence (Satterthwaite 2007). Finally, in many developing countries, the crumbling or inadequate infrastruc- ture, underfinancing, and lack of personnel have eroded health systems even as the burden of disease has increased. Governments in developing countries find imple- menting and rehabilitating water infrastructure without donor support difficult. For example, in Central Asia, government budgets for operation and maintenance dropped from US$60 per hectare to less than US$8 per hectare. The result is that infrastructure is rapidly deteriorating (World Water Forum 2006). In the case of sanitation, more work needs to be done because 2.6 billion people lack adequate sanitation. Rapid urbanization and uncontrolled growth of urban slums are now creating a double environmental health burden for the urban poor. Poor urban people are exposed to health risks associated with dirty cooking fuels, traditional stoves, crowding, and poor access to water and sanitation (Satterthwaite 2007). This neglect of environmental health within the national and local agendas of developing countries is critical--especially because the major environmental health concerns (poor water and sanitation access, inappropriate hygienic practices, lack of vector control, and continued indoor air pollution) are the same ones that devel- oped countries faced about a century ago (Rosen 1958).Whatever the root causes for the poor environmental conditions and neglect of the environmental health 90 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL arm of the primary prevention agenda by the health sector, the current reality in developing countries demands a revitalized approach to implementing environ- mental health interventions, given its importance for child survival. Many governments in developing countries have increasingly begun to recog- nize the importance of environmental health in child survival, poverty reduction, and economic growth, and they are beginning to incorporate environmental health activities into existing integrated child development and nutrition programs. Additionally, successful community-level initiatives, such as the Total Sanitation Campaign in Bangladesh, are beginning to take root and to spread to other coun- tries and regions (WSP 2005). In the following section, this chapter provides an illustrative list of different projects that have either recognized the need or incor- porated environmental health into current policies, programs, and projects, such as child survival, nutrition, and water and sanitation. Environmental Health Experiences in Developing Countries Developing countries themselves vary considerably in terms of institutional capacity, political will, and socioeconomic development. Environmental health interventions therefore need to be customized to the specific enabling environ- ment in a developing country. Recognizing these differences, rather than providing specific recommendations, this section presents examples of how some devel- oping countries are beginning to mainstream environmental health components and objectives within existing child survival programs, nutrition initiatives, and infrastructure projects (water and sanitation, rural energy). As countries develop and their institutional capacity strengthens, they move toward more truly multi- sectoral programs that seek to address environmental health. Integrated Child Survival Programs In developing countries, several child survival strategies, such as immunization campaigns and improved case management of specific diseases, have proven to be effective. Nevertheless, accumulating evidence has suggested that child health programs need to move beyond addressing single diseases to addressing the overall health and well-being of the child. In response, during the mid-1990s, the World Health Organization (WHO), in collaboration with the United Nations Children's Fund (UNICEF) and many other agencies, institutions, and individuals, devel- oped a strategy known as Integrated Management of Childhood Illness (IMCI). Although the IMCI strategy mostly focuses on better treatment practices (in terms of diagnosis, medicine administration, and the like), elements of hygiene are already found within its key practices. Innovative projects in Cuzco, Peru, and Chinandega, Nicaragua, that have promoted key hygiene behaviors within the APPROACHES TO ENVIRONMENTAL HEALTH 91 programmatic framework of their community-level IMCI strategies have resulted in significant improvements in health outcomes (Favin 2004). In another example, the Honduras Integrated Community-Based Child Care Program included a specific component on hygiene education, in addition to advice for improving home care, feeding practices, and related health behaviors (Griffiths and McGuire 2005). Similarly, the Integrated Child Development Services in India has included hygiene awareness (such as hand-washing practices for children) within a multi- sectoral approach that incorporates health, education, and nutrition interventions (Gragnolati and others 2006). A few integrated child survival programs have been even more comprehensive-- including components that traditionally lie outside the health sector but that have important health effects. For example, a subcomponent of the Eritrea Integrated Early Childhood Development Project included environmental health interven- tions to control childhood illnesses through promotion, maintenance, and use of safe latrines; improved food safety and hygiene; water source protection and handling; prevention of acute respiratory infections through indoor air pollution reduction interventions (for example, promotion of improved kitchen and house- hold ventilation); and improved collaboration in the prevention of malaria and other vectorborne diseases (see World Bank 2000). Other child survival programs in low-income countries, such as Bangladesh, Benin, Brazil, Cambodia, Eritrea, Haiti, Malawi, Nepal, and Nicaragua, have included vector-control measures (use of insecticide-treated mosquito nets) as part of their preventive interventions (Victora and others 2005). Some middle-income countries have begun to address child mortality more systematically by moving beyond health system interventions to a more integrated multisectoral program with strong political commitment from the government. Costa Rica, for example, added 15 to 20 years of life expectancy and significantly reduced child mortality in over two decades through a number of multisectoral policy actions, including a targeted program to control infectious diseases through improved sanitation, immunization, and the extension of primary health care to a wider population (World Resources Institute and others 1998). In Mexico, the government brought about significant improvements in under-five mortality by systematically rolling out programs on vaccinations, clean water, and health reform (see box 6.2). Nutrition Programs Improving the nutritional status of young children requires not only health sector interventions but also appropriate actions in agriculture, rural development, water supply and sanitation, social protection, education, gender programs, and community-driven development (Shekar and others 2006). Reviews of different nutrition programs in Central America, Ecuador, Thailand, Ghana, and India 92 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL BOX 6.2 Mexico: Multisectorality through a Diagonal Approach Mexico is one of seven countries on track to achieve the goal of reducing child mortality by 2015 (Sepúlveda and others 2006). Improvement in children's health is attributed to two programs that worked together: the Clean Water Program and the Universal Vaccination Program. The Clean Water Program focused on appropriate water chlorination and regulations that, for example, banned the use of sewage for crop irrigation. Important steps in the program were to improve sewage-treatment plants, to provide adequate disposal of waste, and to monitor the maintenance of drainage systems. The Universal Vaccination Program was created as a result of the measles epidemic in 1989 and 1990 (see the accompanying figure). The outbreak caused more than 70,000 cases and 6,000 deaths in mostly malnourished children under five years of age. Within three years of the project, immunization coverage exceeded 92 percent. By introducing national health weeks and improving the health monitoring system, the authorities focused on immunization, vitamin A supplementation, oral rehydration salts, and deworming--reaching sustained high population coverage. Evidence suggests that high levels of coverage of public health interventions as well as investments in women's education, social protection, water, and sanitation all contributed to the success (Frenk 2006; Sepúlveda and others 2006). This recent Mexican health reform included an unprecedented effort to strengthen environmental health services, regulatory actions to protect the public, and, more generally, a set of intersectoral interventions that define a health policy capable of modifying the broader determinants of disease (Frenk 2006). Mortality in Children under Five Years, under One Year, and Neonates, Mexico, 1980­2005 oral rehydration salts 70 64.3 national vaccination days ths measles vaccination campaigns bire 60 Universal Vaccination Program 51 51.2 Clean Water Program liv 50 national health weeks 41.2 44.2 0001, 40 Progresa Arranque Parejo 33.7 pereatry 28.8 36.2 en la Vida 30 27.6 23 20 25.6 21.9 19.3 23.3 18.8 alitt 10 17.0 12.9 mor 0 8182 83 84 85 86 87 88 89 1 2 3 4 5 6 7 8 02 03 19 801919 19 19 19 19 19 19 19 1990199 199 199 199 199 199199 199 19992000200120 20 200420052006 year under 5 years under 1 year neonatal Sources: Data from the Mexican Ministry of Health Epidemiological and Statistical Mortality System and the Mexican National Council on Population. APPROACHES TO ENVIRONMENTAL HEALTH 93 illustrate the opportunities that exist to include or supplement environmental health activities and hygiene promotion in these programs. Nutrition projects in developing countries also require complementary invest- ments in water and sanitation; without these, children may lose the benefits of additional food intake to parasite infestation and diarrhea (Horton 1999). A review of health reforms in Central America found that programs addressing malnutrition should focus on preventive measures such as (a) ensuring access to safe water and sanitation; (b) educating mothers about nutrition, hygiene, and health practices; (c) using growth monitoring to ensure early detection of malnutrition; and (d) establishing links with health care specialists for at-risk populations, including the rural indigenous (World Bank 2007e). An analysis of India's main child survival program concluded that even with improvements in the Integrated Child Development Services, the country will not meet the Millennium Development Goals in nutrition without investing more in indirect interventions, such as water and sanitation infrastructure and women's educa- tion (Gragnolati and others 2006). In the late 1980s, Thailand's nutrition program was able to effectively address malnutrition by adopting a more integrated approach (see box 6.3). In Ghana, the Community-Based Poverty Reduction Project (1999­2005, US$5.0 million) used an integrated community-based approach for basic health and nutrition service delivery. In addition to monitoring child growth, promoting pre- and BOX 6.3 Thailand's National Nutrition Program Thailand's National Nutrition Program successfully reduced moderate malnutrition from about 25 percent in the under-five population in 1986 to about 15 percent in 1995. Although the health sector led the nutrition program's outreach into villages, the fundamental assumption was that malnutrition was a symptom of poverty and, therefore, that a multisectoral attack on poverty was needed to deal with malnutrition. Accordingly, the Rural Development Plan and the Poverty Alleviation Program supple- mented the nutrition program by funding a wide range of programs, including the following: ·Basic primary health and nutrition services ·Latrines and clean water ·A literacy program ·Agricultural production projects ·Village development projects, such as creation of fishponds or development of water sources ·Employment schemes during the dry season in disadvantaged, nonirrigated areas Source: Heaver and Kachondam 2002. 94 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL postnatal care and family planning, and immunizing children, community workers advised on environmental sanitation and hygiene (World Bank 2007b). An evaluation of programs in Ecuador (Sistema Integral de Alimentación y Nutrición) suggested that nutritional programs in rural areas be complemented with improved water and sanitation, community-based growth promotion, improved health service access, and nutritional supplements for pregnant mothers and young children, among other interventions. The study recommended that, at the national level, programs should include behavior change communications programs and work closely with the authorities responsible for water and sani- tation (World Bank 2007f). Infrastructure Programs Development is often seen as synonymous with improvements in infrastructure, and accordingly, donor funding and government investments have been geared toward improving access to water supply, expanding coverage for sanitation and sewerage, and extending rural electrification. The result has been improvements in the access and coverage indicators for these environmental services. Despite these improvements in infrastructure, however, the trends in the incidence of diarrhea have not fallen correspondingly in developing countries (Kosek, Bern, and Guerrant 2003). Over the years, donor agencies and country governments have increasingly realized that simply providing new infrastructure--be it latrines, piped water systems, or improved cookstoves--is not the"magic bullet"for achieving improved health. Elements of behavior change--the so-called software--are a critical part of the equation. For example, Peru's national government expressed interest in piloting a hand-washing campaign after US$2 billion of investments in water supply infrastructure failed to significantly reduce the prevalence of water-related diseases (Public-Private Partnership for Handwashing with Soap 2005). These behaviors--ranging from hand-washing practices to proper feces disposal to venti- lation practices--are what make the difference between simply "having access" and actually "improving health." One program that has brought together elements of infrastructure and behavior change has been the Total Sanitation Campaign in rural Bangladesh. This comprehensive program has been successful in eradicating the practice of open defecation by promoting behavior change and sanitation facilities in rural areas. Going beyond providing affordable sanitation options (in terms of tech- nologies and costs), this program--as well as some similar programs--has been effective in ensuring high toilet use through a combination of participatory processes; hygiene promotion; institutional incentives, such as financial rewards for achieving universal toilet coverage; and community bans on practices such as open defecation (WSP 2005). The Total Sanitation Campaign has seen more APPROACHES TO ENVIRONMENTAL HEALTH 95 than 400 villages achieve 100 percent sanitation coverage without a household subsidy for infrastructure (WSP 2005). With the widely regaled success of the Bangladesh story, Total Sanitation Campaigns began to spread to Ethiopia, India, Nigeria, and Pakistan. The Ethiopian story was an equally successful one, with a strong multisectoral approach at the local level that emphasized education and community participation (see box 6.4). The effectiveness of the campaign varied across different states in India.Although these programs were meant to be based on the same principles of community involvement and attention to hygiene and social intermediation, a review showed that excessive attention to coverage targets overshadowed the goals of improved community health. Surveys of some villages showed that although they had 100 percent latrine coverage, several households were, in fact, using the latrines to store rice (WSP 2005). With rural populations still largely dependent on biomass fuels for cooking and heating (and the consequent health impacts from indoor air pollution), country governments, donor agencies, and nongovernmental organizations (NGOs) have been working to help rural populations move up the energy ladder toward improved cookstoves and cleaner fuels (for example, liquefied petroleum gas). Although considerations of energy efficiency and costs have dominated rural energy technology choices, recent studies have pointed out the need to pay attention to changes in behavior. A research study in Bangladesh, for example, suggested changes in cooking arrangements and location, construction materials, BOX 6.4 Ethiopia: The Toilet Revolution Another example that drew from the Total Sanitation Campaign is in the Southern Nations Region of Ethiopia. The Regional Health Bureau, with support from the regional cabinet, shifted from a broad primary health care model to a focus on preventable diseases through a limited number of high- impact, broad-reach, low-cost public health interventions in rural areas. This change was facilitated by a strong regionwide commitment to empower households, resulting in a wave of household latrine building. At the village level, after consultation meetings highlighted the need to build latrines, community health promoters visited each household to persuade it to join the community in building latrines. Within less than two years, pit latrine ownership rose from 13 percent to 78 percent (WSP 2007), although it may not necessarily have resulted in a corresponding increase in use. The long- term success will depend on many factors, including durability, technical standards, design, and behavior (with regard to toilet use, and especially children's compliance with latrine use). 96 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL and ventilation practices that may help improve indoor air. The analysis also suggests that poor families may not have to wait for clean fuels or clean stoves to enjoy significantly cleaner air when basic modifications in cooking behaviors may produce cleaner conditions even when "dirty" biomass fuels are used (Dasgupta and others 2004a). Vector-Control Programs Malaria continues to be a major public health challenge, especially in Sub-Saharan Africa, where it is responsible for at least 700,000 deaths among children under five years of age (Snow and others 2003). Although key strategies in the control of malaria have often focused on early diagnosis and treatment, indoor residual spraying, and the use of insecticide-treated nets, environmental management interventions can play a major role within integrated vector management by adding resilience to the results of individual control strategies and by reducing costs as well as the likelihood of development and spread of drug and insecticide resistance (Keiser, Singer, and Utzinger 2005). A review of 40 studies on malaria interventions that emphasized environ- mental management interventions reported very high protective efficacies. Often, a package of multiple interventions is used, guided by a multisectoral staff with expertise in malaria epidemiology and entomology, vector ecology, and land and water engineering (Keiser, Singer, and Utzinger 2005). In Malaysia, for example, the incidence of malaria was radically reduced by environmental management methods devised by interdisciplinary teams of medically quali- fied personnel across many different departments, such as agriculture, health, transportation, urban planning, and the army (Flemming and others 2004). Although environmental vector-control measures may work in Asia, they are much less likely to succeed in rural Africa. The African vectors are far more efficient, so their numbers must be reduced to much lower levels to reduce transmission significantly. For this reason, mosquito ecology demands that environmental manipulation be supplemented by bednets and indoor residual spraying (Touré 2001). Lessons learned from successful vector-control programs show the need for interdisciplinary collaborations among the health, agricultural, water, and infra- structure development sectors in the design of integrated intervention packages. Collaboration, in this context, means that the higher expenses of environmental management are not borne by the health sector and that vector control per se is not the driving force (Keiser, Singer, and Utzinger 2005). Effective vector-control programs should encourage community participation (Keiser, Singer, and Utzinger 2005). In the case of dengue control, a novel effort carried out in several communes in Vietnam demonstrates how a community- based approach can be highly effective against this mosquito (see box 6.5). APPROACHES TO ENVIRONMENTAL HEALTH 97 BOX 6.5 Vietnam's Dengue Program The dengue-control case in Vietnam shows that a community-based approach can be highly effective against this disease. The approach involves a smooth interface among technical experts, health workers, and communities. Dengue control was driven by the chair of each commune and by other leaders from the women's union and youth union; it was implemented by communal health personnel, health collaborators, schoolteachers, and their pupils. Health collaborators were each responsible for monthly inspection of water-storage containers in about 100 houses, delivery of health education messages to make sure that the householder knew how to manage the mesocyclops (a tiny one-eyed shrimplike creature used for biological control of larvae to combat mosquitoes successfully) that inhabited those containers, and reporting of any suspected dengue cases to the communal health center. The program was not simply a matter of the health volunteers eliminating the mosquitoes for the householders--the success of the mosquito-control efforts depended on the close involvement of the house- holders. Schoolchildren participated by cleaning up discarded containers; supporting the aged and infirm in mosquito-control measures; and dissemi- nating information about mosquito control through plays, songs, quiz nights, and--in one district--a Meso football cup. This control strategy has already been implemented in 46 communes in Vietnam. Mosquitoes have been eliminated in 40 of those communes, and the remaining 6 communes show very low densities of mosquitoes. Sources: Das Gupta and others 2006; Ha and Huan 1997; Kay and Nam 2005; Vu and others 2005. Understanding the Enabling Environment As the preceding section shows, developing countries have used a range of ways to mainstream environmental health actions into existing child survival and nutri- tion programs, infrastructure projects, and vector-control strategies. Given the varying socioeconomic, institutional, and developmental contexts of developing countries, knowing what combination of factors contributed to the successful inclusion of environmental health considerations within these programs is often difficult. Looking at the big picture, however, can help identify some common elements that provided the appropriate enabling environment. Some of these common elements have included obtaining high-level political commitment, involving and empowering communities, allocating responsibilities and resources at the local level, and finding a balance between private and public sector roles. 98 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Political Will and Effective Consensus The existence of a positive policy environment is a central principle for successful interventions. High-level political commitment is critical to successful child survival and environmental health programs. The Ethiopian example shows the need for a well-connected leadership that is well informed and willing to take risks and challenge conventional primary health care approaches. The right context is required to implement interventions. In the case of Ethiopia, the intervention was driven by population expansion and deforestation, which reduced the private open defecation option. The success of the story came from the willingness of senior management to listen to the people and put their needs first (WSP 2007). An important first step in the program was facilitating consensus around the need for community ownership of health behavior--especially with respect to vaccination, family planning, and hygiene and sanitation. Consultation and advo- cacy meetings at all levels and covering the whole region were conducted to bring sustained massive response (Knapp 2006). As Shiferaw Teklemariam, state health minister in Ethiopia, points out,"[N]one of these achievements would have been possible without close inter-sectoral collaboration and strong leadership, committed to universal water supply and sanitation access" (WSP 2007: 2). China's improved cookstoves programs have been found to be broadly successful in delivering better stoves to a majority of households in targeted counties, and the success was based on strong administrative, technical, and outreach compe- tence and resources situated at the local level, motivated by sustained national-level attention (Sinton and others 2004). In Brazil, similarly, favoring a democratic and continuous process in education on health and disease processes has been found necessary in stimulating the participation of the population in health promotion programs (Mahendradhata and Moerman 2004). Involving and Empowering Communities Participatory governance mechanisms that encourage the involvement of poorer groups are an important factor for improving environmental health (Satterthwaite 2007). Several programs that have been successful in achieving child health outcomes through environmental health actions have included community partic- ipation as an integral component. In Ethiopia, community health promoters visited households, promoting simple and action-based health messages to under- take actions at individual, household, and community levels that would lead to behavioral change in practices (Knapp 2006). Lessons learned from many programs highlight the need to cultivate a sense of local ownership. Community groups are potentially quite influential for behavior change and for maintaining infrastructure, as the Bangladesh sanitation story demonstrates (WSP 2005). In Thailand, effective consensus and commitment building in the multisectoral nutrition program was instrumental in reducing APPROACHES TO ENVIRONMENTAL HEALTH 99 malnutrition among children (Heaver and Kachondam 2002). In Vietnam, the success of the dengue-control program also lay in the extensive community involve- ment (Das Gupta and others 2006). Another example is the large program of toilet blocks in urban slums in India that were designed, built, and managed by the communities. The National Slum Dwellers Federation develops and promotes these blocks through a women's saving group (Mahila Milan) that is supported by local authorities, thus improving sani- tation for millions of slum dwellers (Burra, Patel, and Kerr 2003). A key feature in Mumbai was the involvement of the slum community in project implementa- tion at the planning stage. Their involvement facilitated collaboration between the NGO, contractors, and community-based organizations (WSP 2006). Balancing Private and Public Sector Roles The nature and complexity of environmental health issues require intersectoral actions not only across various levels of government but also through partner- ships between the public and private sectors. While the public sector retains its power over setting regulations and standards, the private sector is increasingly getting involved in the delivery of environmental services such as water, sanita- tion, and energy. However, the question of who provides and pays for an intervention is important. Improved stoves are an example of private gains to public health measures. To be sure, the absence of credit is a possible constraint for individuals to obtain these stoves, but if the stoves save more on fuel than they cost, one would expect widespread adoption. In contrast, if the benefits are more general than individual, the scope for subsidies is clearer. An evaluation of costs and benefits is necessary to determine the role of the private and public sectors. In the water sector, the privatization of water companies not only can lead to efficiency gains but also can have beneficial health effects. In Argentina, for example, the privatization of public water companies resulted in both increases in connections and improvements in quality (financed from user fees and connec- tion charges). These improvements included the speed of repairs of water leaks and sewer blockages and the percentage of clients with appropriate water pres- sure. Child mortality fell by 8 percent in the areas that privatized their water services, and that effect was largest (26 percent) in the poorest areas, especially in poorer municipalities (Galiani, Gertler, and Schargrodsky 2005; Wagstaff and Claeson 2004). In the field of sanitation, concepts of "sanitation marketing"are proving to be extremely effective. In a bid to harness market power for rural sanitation, IDE-- an international NGO--developed a range of low-cost sanitation options and stimulated a network of local masons to market and deliver them to the rural population in two Vietnamese provinces. As a result, the sanitation access rate increased markedly in the area, even among the poor (Frias and Mukherjee 2005). 100 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL In hygiene behavior, the private sector has begun to play an important role in advocating for and supporting hand-washing campaigns through the Global Public-Private Partnership for Handwashing with Soap. In campaigns such as those in Ghana, Peru, and Senegal, the private sector has engaged in hand-washing promotion and educational activities mainly by contributing its marketing expertise to publicly funded generic campaigns not tied to any specific soap brands. In providing rural energy technologies, too, the private sector is beginning to play an important role in collaboration with the public sector. In Nepal, for example, the Alternative Energy Promotion Center, a quasi-governmental body, is working closely with private sector contractors who deliver improved cook- stove technologies to rural households (World Bank 2008b). This section has provided examples of interventions in different sectors that have provided some key lessons; however, contexts, actors, and situations are different, which makes comparison difficult. Governance and Institutional Implications A final key issue is that environmental health efforts will not be sustainable unless explicit attention is paid to institutional issues. Therefore, the ultimate design and delivery of environmental health interventions must depend on the institu- tional and governance structure of each country, an issue that is elaborated on in this section. Unraveling the Environmental Health Agenda in the Developed World Earlier, this report discussed how, across the developed world, countries sought to establish strong and coherent systems for ensuring environmental sanitation to protect the public's health. As a result of those efforts, mechanisms to ensure environmental sanitation became well established, the burdens of associated disease dropped, and importantly, the public's expectations regarding the acceptability of poor sanitation and disease altered so that providing poor services has become politically very difficult. As mentioned earlier, recent decades have seen some significant unraveling of the coherence of the systems put in place over the previous decades. Environmental health has become disconnected from the mainstream agendas of the health agencies (which now focus more on a case management perspective) and the newly formed environmental protection agencies (which focus more on environmental regulation and compliance) (Kotchian 1997). Correspondingly, donor initiatives related to environmental health issues have also become highly compartmentalized. With donor funding being chan- neled through separate sector silos (of environment, water and sanitation, and health), the possibility of generating the intersectoral coordination that is so APPROACHES TO ENVIRONMENTAL HEALTH 101 crucial for environmental health initiatives was more or less preempted. This development is also evidenced in the way that international goals and develop- ment initiatives addressing health give very little attention to the role of environmental sanitation. Sanitation was excluded from the list of the original Millennium Development Goals and was only subsequently introduced (Satterthwaite 2007; von Schirnding 2005). Developing Countries Follow Suit As the world's poorer countries have moved along the development path with increasing investments in infrastructure, education, health, and other growth- oriented sectors, a similar trend has been seen, with government investments and donor funding increasingly programmed along sectoral lines. The role for the health sector is primarily that of regulation (for example, urban drinking-water quality); promotion (for example, hygiene promotion); and advocacy (for example, persuading utilities to lower their connection charges to the poor). In some cases, as in India (see box 6.6), attention by the health department to environmental health functions has atrophied over the years, as increased emphasis on treatment and delivery of health services led to a dwindling role for public health issues BOX 6.6 Atrophy of Environmental Health Functions in India The role of health departments in the stewardship of sanitation and public health activities has atrophied in India since the 1950s. Before independence, sanitary departments existed at national and provincial levels, with sanitary inspectors and other staff members on the ground in each district. In the districts, sanitary inspectors worked with the village officials and village panchayats, which had been set up to extend the bare bones of government to rural areas. The municipal areas also had extensive machinery for environmental sanitation and public health activities, including the management of water, solid waste, and liquid waste. Sanitary departments were accountable directly to the government and were admin- istratively separate from the Indian Medical Service, which provided medical services. In the 1950s, public health services were merged with the medical services into a unified health department. Interest in specializing in public health dwindled as specialty curative skills became far better rewarded. This atrophy was further fueled by strong electoral demand for services to cure the sick, while by contrast, public health services remained low profile because their success depended on ensuring no outbreaks. Source: Das Gupta and others 2006. 102 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL (Das Gupta and others 2006). One cause for atrophy is the lack of demand for environmental health services. The public demands curative care over preventive care, and environmental health is the most invisible type of preventive care. The consequences of such compartmentalization are much more severe than in the developed world, where intensive environmental sanitation measures had become fundamental to public administration by the middle of the 20th century. This subsequent compartmentalization may create inefficiencies and the reemer- gence of some threats, but the fundamental protections for public health remain firmly in place. By contrast, the era of compartmentalization arrived for many developing countries at a time when levels of exposure to environment-related infections were still very high and before a public health consciousness was intrin- sically found in communities. Box 6.7 illustrates the evolution of environmental health institutions in Ethiopia, an example of such compartmentalization. BOX 6.7 Institutional Evolution of Environmental Health: The Case of Ethiopia Traditional Organization (1908­46) The development of environmental health interventions in Ethiopia dates back to 1908, when it was organized under the Ministry of Interior. The Ministry of Interior consolidated decision-making power over health matters by proclaiming a series of legal notices following its mandate in sanitation. During this period, environmental health activities had an urban bias, the institutionalized aspect did not reach the grassroots level to address environmental health, and interventions were not significant because of limited human resources and unclear vision. New Organization, New Tasks (1947­74) In 1947, the Ministry of Public Health was established with a clear mandate, and rural service provision began to promote medical and sanitation services. The ministry set up a hygiene and environmental health unit for the control and prevention of communicable diseases, with a senior sanitarian guiding sanitation activities. A series of sanitation regulations was enacted, and a training center was set up to train sanitarians, who were largely involved in urban inspections pursuant to sanitation regulations. They were key actors in the eradication of smallpox in Ethiopia. (continued) APPROACHES TO ENVIRONMENTAL HEALTH 103 Primary Health Care Emerges (Late 1970s­Late 1980s) By the late 1970s, a socialist regime set up primary health care, which included drinking water and sanitation as formal components. A 10-year plan was made, with rural and urban programs used as entry points for sanitation efforts. Organized committees at various levels coordinated and ran environ- mental health interventions. The Ministry of Public Health was renamed the Ministry of Health, and in 1983, its Department of Hygiene and Environmental Health was organized into four divisions responsible for major sanitation activities: water and sanitation, food hygiene, industrial hygiene, and quarantine hygiene services. Restructuring of Health Policy and Shift in Strategy (1990­Present) The beginning of the 1990s brought a decentralization of powers down to the grassroots level and expansion of the private sector in environmental health services. Although the Department of Hygiene and Environmental Health maintained its status, regional states developed teams of environ- mental health professionals organized under the Department of Disease Prevention and Control or Department of Health Programs. Environmental health personnel worked in administrative units that coordinated and guided socioeconomic development at the grassroots levels. Health extension packages targeted households with a strong focus on community-based approaches and sustained preventive and promotional health care. Ethiopia is a clear case of how an environmental health institutional framework has evolved over the years. In the most recent past, success has been largely due to the use of local government as a platform for environ- mental health, including hygiene promotion. Source: Kumie and Ahmed 2005. Compartmentalization at the national level means that the basic institutional mechanisms for successful governance of environmental sanitation are lacking at local levels. The consequences can be devastating. In Ghana, a government program for expanding irrigation through the construction of dams and canals, supported by the U.S.Agency for International Development, started as a well-designed effort to make water available for use by humans and livestock, as well as for agricul- ture and fishing (Hunter 1992, 2003). However, no measures were in place to protect against the spread of water-related diseases. Poor maintenance and absence of community extension services to encourage dredging, repairs, and vector control resulted in an increase in cases not only of schistosomiasis in children but also of filariasis among adults. The water supply project is described as a"classic example 104 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL of un-coordinated unisectoral intervention by the Ministry of Agriculture and its international collaborators" (Hunter 2003: 231). Institutional Requirements for Successful Environmental Health Governance Successful environmental health governance requires strong institutional under- pinnings, with clearly articulated roles at all levels of administration within a country. Because developing countries differ in their level of institutional complexity, this section talks in more general terms about the typical roles and responsibilities that should exist in countries for effective governance, but it caters to the more specific case of environmental health governance.1 Roles of National Governments Successful environmental health governance is not possible without strong involvement at the national level. Provide the regulatory framework. An appropriate legal and regulatory frame- work should underpin effective environmental health action. Many countries have such regulations, but many do not have adequate implementation mechanisms. In several African countries, for example, an examination of their legislation reveals an encouraging picture, with many countries having environmental health poli- cies and sanitation codes (WHO Regional Office for Africa 2006), but problems with implementation remain. For example, in Ethiopia, issues with competing mandates (but also other institutional weaknesses) hinder implementation of reasonable policies. These regulations need to be backed by at least a rudimen- tary implementation framework, including mechanisms for building stakeholder compliance, monitoring mechanisms, and transparent enforcement mechanisms that minimize reliance on the overburdened judiciary. These implementation frameworks, such as frameworks for ensuring the quality of public and private sources of drinking water, will serve to guide local governments to engage in appropriate environmental health actions. When necessary, local governments can amend model bylaws framed by the national governments, which can provide a regulatory basis for local environmental health activities, such as the maintenance and cleaning of drains and other civic amenities. Provide the necessary platforms for intersectoral coordination. Environmental health is intrinsically multisectoral and requires intersectoral coordination across health,environment,infrastructure,and other sectors.In an era of sectoral ministries and silo mentality, intersectoral coordination mechanisms (such as intersectoral committees) need to be established at each level of the administrative hierarchy down to local governments. These mechanisms must be set up to work on a APPROACHES TO ENVIRONMENTAL HEALTH 105 routine basis rather than in response to an emergency. They must involve all key departments whose actions pertain to environmental sanitation and public health outcomes. In the Philippines, for example, the Interagency Committee on Environmental Health was established in mid-1980s; the Department of Health serves as the chair, and the Department of Environment and Natural Resources is the vice chair. The committee has addressed priority environmental health problems in the country, such as water quality and sanitation, among other issues (Blössner and de Onis 2005). Ensure interjurisdictional coordination. Environmental health exposures have significant public health externalities--so inadequate attention to clean water and proper sanitation, for example, results in infections such as diarrhea, which can cross jurisdictional boundaries. Interjurisdictional coordination therefore becomes essential, because local governments' efforts to reduce environmental exposures to infections can be seriously undermined if neighboring local authorities do not take similar action. Thus if a peri-urban area is poorly serviced, it can be a recur- ring source of disease outbreaks that can spread to the nearby metropolis. Another set of issues arises because urban and rural local bodies have different regulatory, organizational, and implementation structures, and they typically have very low levels of coordination with each other. This situation greatly increases the probability of differential public health performance in neighboring locali- ties and resultant externalities. Even within urban areas, ignoring the growing slum settlements with dismal environmental conditions and almost negligent access to environmental services can derail attempts by city governments to provide healthy environments. Clarify mandates and appropriate devolution of authority and finances. Inappropriate devolution of authority is a perennial issue in any country, devel- oped or developing, and becomes an issue in the implementation of environmental health interventions. For example, Dandoy (1997: 83) quotes a former U.S. state health officer as saying,"The federal government has most of the money, the state government has most of the legal authority, and the local government has most of the responsibility for protecting the health of the people." Various institutional problems typically hamper effective environmental service delivery and reduce the scope for serious accountability. Problems relating to the decentralization of functions can result in delegation of environmental health functions to local governments without adequate administrative, financial, and technical support from line agencies or higher levels of government. In addition, insufficient clarity and differentiation of mandates and functions between the different layers of the government hierarchy and line agencies creates confusion, undermines accountability, and leads to duplication of efforts and waste of resources. Given the multisectoral nature of environmental health interventions, 106 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL clear mapping of the specific roles of environmental, health, and infrastructural actors at each administrative level becomes even more critical. Finally, in many countries, funds may reside in higher levels of government than the one respon- sible for service delivery--creating shortfalls in financing. Provide technical and policy inputs. Agencies at the national level play an impor- tant role in providing technical and policy inputs related to environmental health actions. Local governments typically lack the sectoral expertise on the wide range of issues they need to address to ensure environmental health, and they can benefit from technical advice and guidance from national-level agencies. One set of issues relates to providing the performance standards,norms,guidelines,training modules, and technical support for environmental health services. These require inputs primarily from the department of health,but also from other departments concerned with environmental management for health, such as the department of environ- ment. In addition, national-level health agencies can provide critical information gathered from disease surveillance, assessment of health threats, and research to help local governments respond appropriately to environmental health risks. Foster communication, advocacy, and promotion. Health ministries in devel- oping countries have an important role to play in health promotion (Cairncross and Valdmanis 2006) to raise awareness among the public as well as in advocacy to raise the profile among policy-makers.Although nationwide campaigns on the need for immunizations (polio drives, for example) and public health messages relating to treatment of diseases (use of oral dehydration salts to treat diarrhea, for example) are commonplace, in only a few instances do national-level agen- ciesplayaroleinpromotingbehaviorchange.Someexamplesof nationalhand-washing campaigns--such as Peru's"Lávate las manos con jabón"("Wash your hands with soap")--have met with mass popularity and success in the adoption of better hygiene. Other examples include national hand-washing campaigns in Ghana, Nepal, and Senegal (for more information, see http://www.globalhandwashing.org). The health ministry can also play an advocacy role in raising the awareness of these issues within local and national governments. At ministerial meetings, it could link health outcomes with infrastructure projects, thus raising awareness of environmental health issues. Roles of Local Governments With guidance from national-level legal and regulatory frameworks, accompanied by devolution of functions and funding, local governments can play an important role in environmental health at the community and household levels. Promote an integrated approach to local level health. Given the nature of envi- ronmental risks and transmission mechanisms, the most appropriate levels for carrying out actions are the household and community levels. Local governments APPROACHES TO ENVIRONMENTAL HEALTH 107 are therefore best placed to ensure the intersectoral and interjurisdictional coordi- nation needed to ensure public health, because they are the convergent point at the appropriate level of all the departments whose actions affect environmental health. The benefits of intersectoral coordination at the local level are illustrated by a successful malaria-control program in Kheda district in Gujarat, India. Conducted with the active participation of local communities, it involved intensified case detection and treatment to reduce the number of mosquitoes carrying malaria parasites, coupled with eliminating vector breeding habitats in homes and in irrigation canals and the use of larvivorous fish. The active participation of the district and local governments and the high level of intersectoral coordination, including those responsible for the health, fisheries, and irrigation sectors, contributed to the success of this effort in reducing the number of mosquito breeding places (Malaviya and others 2006). Provide civic amenities related to environmental health. A core task of any local government is to provide basic civic amenities, and local governments have at their disposal a range of funds and other resources that they can use for these tasks, including locally generated revenues. These basic amenities include the construction, maintenance, and cleaning of public resources, such as roads, drains, water supply systems, sewers, and waste disposal sites; measures to improve the sanitary condition of the locality, such as solid waste removal; and monitoring activities, such as the protection of water sources. Mobilize citizens and support health activities. With their strong grassroots pres- ence, local governments can do a great deal to enhance the effectiveness of services and programs aimed at improving health. For example, they can help collect disease surveillance information by reporting easily identifiable conditions, such as an episode of severe vomiting and diarrhea. They can disseminate public health messages and mobilize citizens to improve their own health behaviors and envi- ronmental hygiene and to pressure others around them to help maintain better sanitary conditions in their neighborhoods. Additionally, local governments can mobilize community-based groups to help improve environmental services. The total sanitation program in West Bengal illustrates how successful local governments can be in inducing widespread adop- tion of personal latrines and in using NGOs to set up for-profit "sanitary marts," where latrines are produced to given standards and retailed for an agreed price (Majumder 2004; also see discussion in Mavalankar and Manjunath 2004). Under the leadership of local governments, citizens' groups such as self-help groups are active in the rural areas of West Bengal and have proved to be a powerful force in the spread of pit latrines. These successful efforts to mobilize citizens can be extended to a larger range of environmental health issues, such as solid waste management, vector control, and drainage. 108 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Strengthen roles in communication and advocacy. A critical component of envi- ronmental health programs is to raise citizen awareness of the benefits of clean water, proper sanitation, and good hygiene behavior, as well as to stimulate their demand for these services. Although national-level campaigns play an important role in raising awareness, more community-specific and customized messages relating to behavior change for better environmental health should be advocated by local health units and schools. Local governments can reach citizens more effec- tively by engaging communities and individuals in a host of face-to-face forms of communication. Reorient the role of an environmental health workforce. Whether under the aegis of the health department or the local governments, a minimum complement of public health personnel is needed at each administrative level to implement and ensure good environmental health. Unlike a century ago, the all-encompassing role of the environmental health officer may no longer be realistic. The role will have to be reoriented through in-service training at the local level to address the changing priorities in environmental health. In South Africa, for example, envi- ronmental health officers form the backbone of environmental health services and are employed at the local and state levels. Because of the need for multiple skills in response to the crosscutting nature of environmental health interventions, training of environmental health officers has changed fundamentally, with the focus moving beyond law enforcement to community participation and develop- ment (Mathee, Swanepoel, and Swart 1999; Thomas, Seager, and Mathee 2002). Establish and enforce regulations. Local government can play an important role in establishing, monitoring, and enforcing regulation. This activity involves building codes, land tenure bylaws, and tenancy regulation. Regulations can require that no house be occupied without prior construction of a latrine, can prevent land- lords from raising rent if the tenant has a latrine built, and can even show the mason how to build it. Local government is usually responsible for outbreak inves- tigation and the ensuing environmental health measures, with the cooperation and technical knowledge of experts in the national governments. Environmental health regulations are necessary but not sufficient, because many environmental health issues are the result of informal settlements or slums, where regulations are not enforced. The roles and responsibilities at the local and national levels discussed here are generic and broadly represent the key institutional requirements for better environmental health governance. An important next step would be to carry out an in-depth institutional analysis at country level to better understand the coor- dination mechanisms between ministries and how mandates and budgets are assigned. Such an understanding is critical for environmental health, an issue that is inherently multisectoral, requiring collaborations across health, environment, APPROACHES TO ENVIRONMENTAL HEALTH 109 and infrastructure ministries. For a specific country, this type of analysis would help governments better define the roles and responsibilities of different agen- cies, correct the incentive structures, and improve the implementation of environmental health programs and actions. A Critical Moment A critical moment has arrived for this neglected story to take the forefront. As this chapter shows, governments in several developing countries are beginning to incorporate environmental health actions in their sectoral strategies. Similar changes are occurring in the donor community, with a growing recognition of the crucial role of environmental health. Multilateral development banks such as the World Bank--acknowledging the need for multisectoral collaborations to improve health outcomes--are revisiting the role of the health sector, while acknowledging the multisectoral nature of such outcomes. The Bank's newly released Health Sector Strategy states: One of the clearest cases for strong government intervention in the [health, nutrition, and population] sector can be made when there are large exter- nalities (the benefits to society are greater than the sum of benefits to individuals). This is true in the case of clean water, sanitation services, vector control, food safety measures. . . . (World Bank 2007c: 5) Additionally, other international organizations such as WHO are strength- ening measures to evaluate the disease burden from environmental risks in developing countries. Recognizing the effect of environmental health on malnu- trition, new WHO burden of disease estimates are now including the consequences of malnutrition that are due to poor sanitation in developing countries (Fewtrell and others 2007). The Pan-American Health Organization stresses that a multisectoral community health approach is vital to reduce many of the"neglected"diseases in Latin America, which are often preventable through improved environmental health (Holveck and others 2007). UNICEF, with its focus on child survival and development, has been consistent in its approach emphasizing water and sanitation measures alongside nutritional and health system strategies. Key Messages Lessons from history have shown the enormous benefits of multisectoral envi- ronmental health actions. Today's developed countries have undergone an evolution in environmental health functions. Both institutionally and concep- tually, however, environmental health has fallen through the cracks in the development agenda in the world's poorest countries. 110 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Several governments in developing countries are beginning to mainstream environmental health components and objectives within existing child survival programs, nutrition initiatives, and infrastructure projects (water and sanita- tion, rural energy). Some common elements for successful environmental health actions in devel- oping countries have included obtaining high-level political commitment, involving and empowering communities, allocating responsibilities and resources at the local level, and finding a balance between private and public sector roles. Furthermore, successful environmental health governance requires strong insti- tutional underpinnings,with clearly articulated roles at all levels of administration within a country. Now is a critical moment for this agenda to take the forefront in developing coun- tries, with governments,donors,and civil society beginning to strengthen measures to address environmental health, especially in the context of child survival. Note 1 The framework provided in this section is based on Das Gupta and others' (2006) analysis of environmental sanitation and public health in India and was adapted to discuss the broader issue of environmental health. C H A P T E R 7 Conclusion THE CHILD HEALTH AGENDA REMAINS UNFINISHED in the developing world, with millions of children continuing to fall sick and die from preventable environmental health causes. Malnutrition, poor environmental conditions, and infectious diseases are highly associated geographically and take their heaviest tolls on children under five years of age in South Asia, Sub-Saharan Africa, and certain countries in the Eastern Mediterranean region (Fewtrell and others 2007). The contribution of repeated infections, especially diarrheal episodes, to a child's nutritional status and subsequent mortality has only recently been incor- porated in assessments of the environmental burden of disease (Fewtrell and others 2007). A recent collective expert opinion stated that about 50 percent of the consequences of malnutrition are in fact caused by inadequate water and sani- tation provisions and poor hygienic practices (Prüss-Üstün and Corvalán 2006). Evidence from a review of recent cohort studies undertaken for this report further corroborates this link between environmental infections and malnutrition. Contributions of This Report This report reemphasizes the window of opportunity highlighted in the World Bank's 2006 nutrition report--pointing to a critical period in the life cycle of a child, from the womb to the age of about two years (World Bank 2006c). With 111 112 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL research evidence from various studies, this report demonstrates how exposures to environmental health risks jeopardize the health and nutritional status of young children. Poor environmental management and bad sanitation expose pregnant women to both malaria and hookworm infections. Left untreated, these infec- tions lead to permanent growth faltering, lowered immunity, and increased mortality for these women's children. In early infancy, improper feeding practices and poor sanitation have a pernicious synergistic effect on the child's nutritional status. Many of these impacts on a child's growth also result in negative cogni- tion and learning impacts as well as chronic diseases later in life. Although considerable progress has been made, this report calls attention to the potential of untapped environmental health actions to complement existing health, infrastructure, and environmental management strategies in the devel- oping world. Current child survival strategies in developing countries focus on case management and treatment, neglecting primary prevention, especially as it relates to reducing exposure to infections. The evidence of the impacts of envi- ronmental risks and subsequent infections on child malnutrition makes correcting the current neglect of environmental health in child survival and child health strategies even more imperative in developing countries. Environmental health actions supplement existing strategies addressing child health. An analysis of results from Fewtrell and others (2007) shows that envi- ronmental health interventions have a multiplier effect on mortality: investments addressing environmental risks (for example, lack of water and sanitation) pay off with more than just their direct effect on disease (for example, reducing diar- rhea). This finding has important consequences for developing countries seeking to invest in environmental health interventions; their budgets will stretch further because of the nature of the externalities associated with such interventions. Therefore, governments in developing countries have an enormous opportunity to incorporate environmental health interventions in existing strategies that affect child health and to provide a bigger"bang for the buck"in terms of health improve- ments. As has been discussed in this report, environmental health actions add value to existing programs for child care, micronutrient supplementation, and immunization. They can also be used to adapt environmental management programs in developing countries, such as those relating to vector control, as well as to form the basis of adjustments to strategies in the water sanitation and rural energy sectors to enhance health outcomes. Revised country-level estimates show that when malnutrition-mediated health effects attributed to environmental health risks are included, the total costs are at least 40 percent higher for both Ghana and Pakistan. The longer-term effects of malnutrition (partly attributed to environment-related infections) on education and income are estimated to add another 3 to 4 percent in terms of annual costs. The total annual cost attributed to environmental health risks, including all CONCLUSION 113 malnutrition effects (such as higher education costs), are estimated to be as high as 9 percent of Ghana's or Pakistan's gross domestic product. Given this consid- erable social and economic burden, policy-makers in developing countries where malnutrition and poor environmental conditions coexist should place environ- mental health interventions high on the policy agenda. Furthermore, as developing countries strive to meet several of their commitments for the Millennium Development Goals, these interventions will remain critical for addressing poverty reduction and child mortality. Growing evidence indicates the need to mainstream environmental health interventions in developing countries' child survival strategies. With environ- mental health having fallen through the cracks in the development agenda, however, policy-makers in those countries will have to make a renewed effort to revitalize the environmental health content within existing child survival programs, nutri- tion initiatives, and infrastructure projects. Anecdotal evidence of such efforts is beginning to emerge in developing countries. Common criteria for successful implementation include obtaining high-level political commitment, involving and empowering communities, allocating responsibilities and resources at the local level, and finding a balance between private and public sector roles. Finally, successful environmental health governance requires strong institutional under- pinnings. This report discusses--through an institutional lens--the roles that national and local governments can play in delivering and managing environ- mental health interventions. Next Steps In many ways, this report represents a first step toward providing policy-makers with the epidemiological, economic, and experiential evidence to incorporate environmental health in the child survival agenda. However, in each of the three main sections of the report, additional research and studies will help donors and governments in developing countries choose to invest in appropriate environ- mental health interventions. The epidemiological evidence presented in this report reaffirms the importance of environmental health for child survival, especially when considering the links through malnutrition.Further research on the environmental health impacts during pregnancy, additional disease transmission pathways, and better relative risk esti- mates will help improve disease burden and costing estimates while informing governments about appropriate programs. For example, in Sub-Saharan Africa, malaria and hookworm infections coexist to cause anemia and subsequent growth retardation of the fetus in pregnant women.Communitywide investments to improve water resources management, use of insecticide-treated nets and indoor residual spraying, and improved sanitation facilities would help improve birth weight. 114 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL This report demonstrates the substantial burden and consequent economic costs associated with environmental health risks, but important questions remain: How should governments prioritize among the different environmental health, infrastructure, nutrition, and child survival interventions to improve child health? What are the cost-benefit ratios and the levels of cost-effectiveness of individual interventions? Some research on this subject has provided answers at the global level--with hygiene promotion measures emerging as the most cost-effective intervention (Laxminarayan, Chow, and Shahid-Salles 2006). Carrying out such cost-effectiveness and cost-benefit analyses at a country level is the next step in guiding decision-makers in specific countries in prioritizing among the various available interventions. Some recent analyses in Colombia and Peru have begun to explore the ranking environmental health interventions through cost-benefit analyses (World Bank 2006a, 2007g). In addition to conducting economic analyses, governments in developing coun- tries need to look toward implementing specific environmental health actions. The effectiveness of such interventions ultimately depends on the enabling environment in the country, which highlights the need to assess (a) institutional mandates and capacities for addressing environmental health issues, (b) regulations on environ- mental health, and (c) availability of budgets (Poverty Environment Partnership 2008). However, a one-size-fits-all solution will not work, and policy-makers will need to look beyond the generic roles and responsibilities for better environmental health governance discussed in this report. An important next step would be to carry out an in-depth institutional analysis at a country level to better understand the coordination mechanisms between ministries and the ways mandates and budgets are assigned. Such an understanding is critical for environmental health-- an issue that is inherently multisectoral and that requires collaborations across health, environment, and infrastructure ministries. Temporality is also an important consideration in how governments choose between various interventions that address child survival, because the costs of programs are often incurred in the short term and are front-loaded, whereas the benefits--especially those relating to cognition and learning--are spread over the long term. In the absence of appropriate economic analysis at the country level, governments may be tempted to choose interventions that are cheaper in the short term, such as oral rehydration programs, rather than the more expensive water and sanitation programs, even though the latter may have greater benefits over the long term. Additionally, over the longer term, environmental health concerns are expected to grow. As the world's climate changes, diseases such as diarrhea and malaria, among other important health burdens that are the result of environmental risk factors, are likely to worsen, particularly for the poor and in developing countries (Campbell-Lendrum, Corvalán, and Neira 2007; IPCC 2007). Changing CONCLUSION 115 temperature and precipitation will also affect agricultural production and threaten food security, thus having implications for malnutrition. Therefore, scaling up preventive environmental health interventions (such as clean water and sanitation) to reduce the current burden of disease is a prudent investment (Campbell-Lendrum, Corvalán, and Neira 2007). Given the multisectoral nature of environmental health issues, the advocacy and regulatory role of the health sector and the supporting roles of other sectors (for example,environment,infrastructure,agriculture,and education) in promoting and delivering environmental health actions need to be revitalized.Without focused and targeted attention to improving environmental health, governments and donors are losing out on the significant opportunities to make dramatic changes in the child health scenario in the world's poorest countries. Ultimately, good environmental health governance is about how policy-makers in the poorest countries develop mechanisms to pick up signals on environ- mental risks and then find ways to translate these findings into appropriate and well-targeted actions. These governments will also need to adjust their policies to address environmental health outcomes and set up institutional mechanisms to successfully implement interventions. Finally, the creation of long-term constituencies within a country will help to continually raise attention to issues related to the environment, health, and poverty and to promote social account- ability among public officials for effective action on these issues (Poverty Environment Partnership 2008). A concerted and continuous effort is needed on behalf of both developed and developing countries to ensure that environmental health is placed high on the development agenda and that corresponding interventions are financed and under- taken to improve children's survival and development potential. A P P E N D I X A Technical Review of Cohort Studies Background The World Health Organization (WHO) monograph by Scrimshaw, Taylor, and Gordon (1968) initiated a lengthy discussion on whether repeated infections in childhood resulted in permanent growth faltering and stunting in adulthood. Results from observational cohort follow-up studies on effects of infections over time--and ultimately over generations--have provided evidence to stimulate the debate. By the time the WHO monograph was published, extensive microbiological, immunological, and physiological studies--partly based on human experiments-- favored the idea that all infections contribute to growth faltering, but the long-term effects of infections on nutritional status were still lacking (Scrimshaw 2003). Diarrheal diseases are very common from infancy up to early childhood. Unlike upper respiratory infections, which are relatively uncommon in early infancy because of maternal antibodies (Nair and others 2007; Sato and others 1979), diarrheal diseases affect the gastrointestinal tract of the child from birth onward, causing malabsorption. For this reason, diarrheal diseases have been particularly studied to investigate their impact on growth faltering (Scrimshaw 2003; Stephensen 1999). 117 118 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Search Strategy and Selection Criteria For this study, an initial review of cohort studies published on Pubmed that linked infections and malnutrition (underweight, wasting, and stunting) was under- taken. In all, 38 cohort studies conducted in developing countries (table A.1) were identified; one recent and very large case-control study in Bangladesh (Chisti and others 2007) was also included. Studies that reported the effects of infec- tions on underweight status (weight for age) only--that is, those that lacked information on stunting (measured by height for age)--were also included. One cohort study that was included described only the natural growth history of patients with persistent diarrhea (Valentiner-Branth and others 2001) and had no control group. One study merely descriptively associated diarrhea episodes and development of underweight (Schorling and others 1990). These studies were included whether or not associations of diarrheal and other diseases with malnutrition were separated. Published experimental studies (and their systematic reviews) that report results from deworming trials undertaken after a recent meta-analysis (Dickson and others 2000) were separately included to allow inferences to be drawn about whether the elimination of soil-transmitted helminths that cause loss of appetite, malabsorption, or nutrient losses affect the nutritional status of individuals (Taylor- Robinson, Jones, and Garner 2007). Findings and Discussion Epidemiological studies typically examine associations between an exposure vari- able and a health outcome. Austin Bradford Hill recognized the importance of moving from association to causation as a necessary step for taking preventive action against environmental causes of disease (Lucas and McMichael 2005). Hill set out nine viewpoints, or guidelines, against which an epidemiological associ- ation might be assessed when attempting to reach an appropriate conclusion on causation: strength of association, consistency of results, specificity (dose-response), temporality, biological gradient, plausibility, coherence, experiment, and analogy. In this review of the literature, some of the more important of these criteria are used as a framework for evaluating whether infections cause growth faltering. Strength of Association The cohort studies use heterogeneous analyses and different parameters to study infection-attributable growth faltering. One often finds somewhat weak associ- ations in these studies. One should bear in mind the following issues, which lead to difficulties in establishing an infection-attributable fraction of poor growth: 1. The internal comparison group--or merely a diarrhea-free period--in the reviewed studies is often not infection free. Thus, asymptomatic children may TECHNICAL REVIEW OF COHORT STUDIES 119 have endemic diarrheagenic bacteria (12 to 37 percent, according to Nguyen and others 2005). Schorling and others 1990; Steiner and others 1998), as well as carry parasites and protozoa (Checkley and others 1997; Prado and others 2005). These children have been shown to present with pathological damage in the gut, which results in subsequent growth faltering (Checkley and others 1997; Steiner and others 1998). A cohort follow-up study of 71 Gambian infants has provided evidence that such damage--which results from a malabsorption disease likely caused by unhygienic conditions (Lim 2001)--causes 55 percent of the loss in weight gain in developing countries, even if infants were breastfed (Campbell, Elia, and Lunn 2003). In an earlier study in rural Gambia, infants between the ages of 3 and 15 months were found to have this malabsorption in the gut for 75 percent of the time but showed diarrheal symptoms only 7.3 percent of the time (Lunn, Northrop- Clewes, and Downes 1991). 2. In earlier studies in Guatemala and Nigeria (Martorell, Yarbrough, and others 1975; Morley,Bicknell,andWoodland 1968) and in a more recent study in Malawi (Maleta and others 2003),the comparison group was also exposed to a relatively high morbidity burden, which obviously diminishes the association between growth faltering and infections. 3. In most studies, the children being studied are subject to nutritional inter- ventions (especially breastfeeding promotion) or receive appropriate medical case management. These interventions and treatments are also likely to mask the effects of diarrhea on growth because it becomes difficult to observe the natural history of growth faltering attributable to infections under these circumstances. 4. Another reason for difficulties in attributing growth faltering to infections lies in the methodology for research studies. Many studies (for example, Briend and others 1989) do not start follow-up from early infancy. This lessens the ability of investigators to find the effects of infections on growth faltering (Checkley and others 2003; see also the discussion of early infancy in chapter 2). 5. Furthermore, the environmental effects in the fetal period are not adequately considered in these cohort studies. Many infections can cause intrauterine growth restriction (Graham and others 2006), but recent robust ecological (van Geertruyden and others 2004) and analytic (Watson-Jones and others 2007) evidence reveals that malaria during pregnancy has dramatic adverse effects on fetal and newborn survival and related nutritional status at birth, especially in Sub-Saharan Africa. Consistency Infections contribute to growth faltering. This proposition is supported by 35 of the reviewed cohort studies. In the three studies in which the effects were largely 120 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL not found (Alvarado and others 2005; Cohen and others 1995; Kolsteren, Kusin, and Kardjati 1997), the results can be explained by the well-established masking effects of breastfeeding (Brown 2003).1 Studies by Briend and others (1989) and later by Moy and others (1994) paid attention to the timing of infections within the follow-up periods. (Growth periods were analyzed as statistical units rather than by individual children.) The results demonstrated that the effect on growth faltering was observed to be corrected by catch-up growth among children experiencing watery nonchronic diarrhea if enough time had elapsed since diarrhea episodes. However, even those studies have agreed that chronic diarrhea (that is, diarrhea that persists for greater than 10 percent of time) has irreversible effects. The study by Briend and others (1989) also suggested that dysentery adversely affected linear growth after a three- month lag. Studies that started follow-up from birth (see also the specificity and tempo- rality arguments later) have shown with increasing consistency that diarrheal episodes lead to irreversible growth faltering. The Gambia study (Campbell, Elia, and Lunn 2003: 1337) concluded that "the lack of association between diarrheal disease and long-term growth as described by Briend et al. (1989) does not preclude the possibility of the presence of a growth retarding gastrointestinal enteropathy." Hence, when the gut of young infants is damaged from malab- sorption disease, growth faltering will result even in the absence of diarrheal symptoms (and may not be attributed to the diarrheal episodes). Investigators have already questioned how children in the Bangladesh study (Briend and others 1989) could have caught up in growth despite the extra nutritional require- ments generated by this common malabsorption disease (Lunn, Northrop-Clewes, and Downes 1991). There are few follow-up data concerning helminths. One study from a north- eastern Brazilian shantytown, where subjects were followed from birth to seven years of age, showed that helminths, independently and additively, caused 4.6 centimeters of growth faltering, on average. This result was on top of the effects of diarrhea (Moore and others 2001) and is consistent with an early quasi-experimental study from rural St. Lucia, West Indies. That study revealed that the introduction of household connections and latrines resulted in lower levels of Ascaris and Trichuris, lower diarrhea prevalence rates, and improved growth in infanthood (Henry 1981). Specificity (Dose-Response) Researchers have shown dose-response relationships between infections and growth faltering in several ways that closely relate to the specificity criteria: 1. Four studies (Adair and others 1993;Brush,Harrison,andWaterlow 1997;Checkley and others 2003; Moore and others 2001) have demonstrated that diarrheal TECHNICAL REVIEW OF COHORT STUDIES 121 episodes (the dose), with increasing period prevalence, show lagged or lasting effects in terms of growth faltering (the response).This finding is consistent with many older studies that performed regression analyses for effects that did not lag (Baumgartner and Pollitt 1983; Black,Brown,and Becker 1984; Lutter and others 1989;Martorell,Yarbrough,and others 1975;Rowland,Cole,andWhitehead 1977; Zumrawi, Dimond, and Waterlow 1987). 2. Several studies have demonstrated dose-response effects relating the nature of infection bouts to growth faltering: the more severe the infection bouts, the more significant the growth-faltering effects in children. Especially in the case of invasive infections (such as enterotoxigenic E. coli, Shigella, measles, and acute lower respiratory infection), significant metabolic (or merely dose-response) effects have been observed (Alam and others 2000; Black, Brown, and Becker 1984; Briend and others 1989; Chisti and others 2007; Henry and others 1987; Victora and others 1990; Villamor and others 2004), as expected by Powanda and Beisel (2003) and Scrimshaw, Taylor, and Gordon (1968). Even the study that has most strongly challenged the irreversible effects of infections on growth makes an exemption with respect to dysentery (Briend and others 1989).2 Also, infections (for example, protozoa) pose special immunological challenges in early, preweaning infancy (Marodi 2006; Wilson 1986) appear to contribute to permanent growth faltering (Checkley and others 1997, 1998; Molbak and others 1997). 3. Prado and others (2005) show that among Giardia-infected children, the more time the child remained untreated during the follow-up period, the more he or she experienced lagged linear growth faltering. Temporality The proposition that the associations between infections and growth faltering are merely transitory has been given importance in several relatively new studies, which emphasize that there are lagged or lasting effects that continue until two to seven years of age if the burden of infection began in early infancy (Adair and others 1993; Checkley and others 2003; Maleta and others 2003; Moore and others 2001; Prado and others 2005). This interpretation is supported by many earlier as well as more contemporary studies (Baumgartner and Pollitt 1983; Brush, Harrison, and Waterlow 1997; Checkley and others 1997; Chisti and others 2007; Fikree, Rahbar, and Berendes 2000; Martorell, Habicht, and others 1975; Martorell, Yarbrough, and others 1975; Molbak and others 1997; Morley, Bicknell, and Woodland 1968; Rowland, Cole, and Whitehead 1977; Victora and others 1990). As follow-up data from Central America indicate, the growth status of chil- dren at two years of age strongly predicts their height in adulthood (Martorell 1995). The observed lagged or lasting effects are considered irreversible (Checkley and others 2003). These lagged effects are also important because they reduce the 122 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL possibility of reverse causality (that is, that malnutrition increased the risk of infection in the first place) (Prado and others 2005). Biological Plausibility The proposition that infections cause undernutrition is biologically plausible. There are many mechanisms, both proven (anorexia, malabsorption, loss of nutri- ents) and still hypothetical (decreased nutrient transport to periphery),that interact with the host's immune responses (table 2.2; see also Keusch 2003; Scrimshaw 2003; Scrimshaw, Taylor, and Gordon 1959, 1968). An important but surprisingly little-studied mechanism is the acute phase response to infection in children. The withholding of food (because of anorexia) in the acute phase of an infection such as diarrhea appears initially to have some beneficial effects. However, long-lasting anorexia delays recovery and is ultimately deleterious to a child's health (Langhans 2000). This mechanism probably under- lies widespread traditional cultural practices relating to the deliberate restriction of food intake in children during disease episodes.It was only in 1997 that continued feeding during diarrhea was included as an important component of diarrheal patient management in the American Academy of Pediatrics' Pediatric Nutrition Handbook (Brown 2003). Experimental Evidence from Deworming Deworming with anthelminthics is often safe and highly effective in reducing or eliminating such soil-transmitted helminths as roundworm (Ascaris lumbricoides), hookworm (Necator americanus or Ancylostoma duodenale),or whipworm (Trichuris trichiuria) (Awasthi and Bundy 2007; Dickson and others 2000). WHO recom- mends3 the widespread use of anthelminthics among children, in an effort to cut transmission that occurs in tropical countries (Awasthi and Bundy 2007). More and more, in large areas of developing countries, almost all children-- especially after weaning--are being infected (Awasthi, Bundy, and Savioli 2003). Trials carried out in the late 20th century that tried to establish the effective- ness of anthelminthics in improving nutritional status were conducted largely on schoolchildren infected with roundworm. The trials showed that, despite method- ological problems, a single anthelminthic dose may lead to a significant weight gain of 0.24 kilograms (95 percent confidence interval, 0.15 to 0.32 kilograms; fixed-effect model) to 0.38 kilograms (95 percent confidence interval, 0.01 to 0.70 kilograms; random-effect model) over a period of one year or more (Dickson and others 2000). Subsequent rapid reinfection often masks the beneficial effects of deworming on weight gain. A recent comprehensive meta-analysis of randomized trials supports the idea that a single dose of a deworming drug has a significant effect on the mean change TECHNICAL REVIEW OF COHORT STUDIES 123 in weight in children who have an initial high worm prevalence or intensity, compared with no effect in a group with low worm prevalence or intensity (Taylor- Robinson, Jones, and Garner 2007). Conclusions Consistent and biologically plausible results from the reviewed cohort studies, coupled with evidence of specificity, reveal how increments in infection load-- number, severity, and duration of episodes--are associated with stronger effects on nutrition, which appear after a time lag in terms of stunting. The results support the idea that infections--especially if acquired in early infancy--cause irreversible and often highly prevalent undernutrition among children under five years of age in developing countries.4 Experimental anthelminthic trials on humans also support the idea that soil-transmitted helminths and other pathogens associated with similar pathophysiological effects (malabsorption and loss of nutrients) cause undernutrition. In summary, several strong masking elements are important in estimating how much of undernutrition is attributable to infections: 1. Enteric infections cause growth faltering in children in developing countries, even in the absence of diarrhea. These infections make children with diarrhea and without diarrhea appear to be more alike with respect to growth. 2. In some studies, there is a relatively high disease burden in the control group, which again makes children with and without infections more alike. 3. The active case management and supporting nutritional therapies (including breastfeeding) that are often targeted at study participants mitigate the effects of infections on nutritional status. 4. Not enough infants are enrolled in many of these cohort studies; therefore, the studies miss the long-lasting effects in the most vulnerable period (0 to 6 months) on nutritional status. 5. Environmentally attributable growth faltering (especially malaria and hook- worms) during the fetal period that manifests itself in intrauterine growth retardation is not adequately captured in these studies. Independent of this review, a group of experts chosen by WHO (Prüss-Üstün and Corvalán 2006) estimated on the basis of converging expert opinion (using the Delphi method) that 50 percent (with a confidence interval of 39 to 61 percent) of the consequences of maternal and childhood underweight are attributable to environmental issues, especially to poor water, sanitation, and hygiene. Fewtrell and others (2007) use this attributable fraction in estimating the regional burden of disease from poor water resources management (including lack of water, sani- tation, and hygiene). The regional estimates of disease burden presented in chapter 4 are based on those two studies. 124 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL For reasons such as heterogeneous inclusion criteria and analysis techniques and differing follow-up times, masking effects, and the like, it is virtually impos- sible to assess the literature systematically to find out how much (what percentage) of growth faltering is attributable to diarrheal episodes. Furthermore, not all studies present their results; hence, the variance of growth faltering that is explain- able by infections is not readily available. It is important to note that because the prevention of infections more than proportionately moves children from a severe to a moderate and mild malnutri- tion state (see table D.5), far less than 50 percent of weight gain attributed to environmental infections explains 50 percent of the consequences of maternal and childhood underweight. The idea that a substantial proportion of children with severe undernutrition can be lifted to the moderate and mild categories is also supported by experimental data on humans (Ahmed and others 1993). Findings from this review of 38 cohort studies support the 50 percent figure used by WHO in its reports. Examples of cohort studies from this review provide some idea of how experts have come up with this figure. For example, a recent relatively large study in Bangladesh reveals that dysentery and watery diarrhea together could retard weight gain by 20 to 25 percent compared with the results in periods when no infections occurred (Alam and others 2000). The weight gain retardation is likely to be significantly higher when compared with international standards. Simulations that use a weight gain retardation of 35 percent result in environmentally attributable health burdens that are approximately similar to the point estimate from WHO experts. Choice of Parameters The results of the review informed the choice of parameters used in this study to estimate the revised costs of environmental infections through malnutri- tion. The literature reviewed shows that diarrhea and other infections cause a weight gain retardation of 20 to 50 percent. Limited research studies how the retardation factor may vary with z-score and age. This study assumes an average loss of 35 percent in terms of weight gain (see table D.4 and step 4 in appendix D) as attributable to infections in early childhood in developing countries. This percentage is used in studies in Ghana and Pakistan to compute counterfac- tual underweight z-scores and prevalence rates in order to estimate deaths in children under five years of age attributable to infections and malnutrition (presented in chapter 5). For estimating the effect of stunting on educational performance, this study assumes an average loss of 25 percent in height gain from diarrheal infections in early childhood. The literature reviewed in this appendix suggests that the effect on stunting is somewhat less than on weight gain (see table D.11 and step 10 in appendix D). TECHNICAL REVIEW OF COHORT STUDIES 125 Notes 1 Breastfeeding not only diminishes transmission by various mechanisms but also substan- tially supports energy balance during disease bouts (Brown 2003). 2 Briend and others (1989) observed that the effects of dysentery did not become visible until three months had elapsed. 3 World Health Assembly resolution 54.19, passed in May 2001. 4 Recent expert opinion (Prüss-Üstün and Corvalán 2006) suggests that 50 percent of the health burden attributable to maternal and childhood underweight results from rampant diarrheal disease. 126 TABLE A.1 Cohort Follow-up Studies Relating Infectious Disease and Nutritional Status of Children in Developing Countries Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition Nigeria Rural 104 6­12 months Up to Symptom of diarrhea Monthly A child whose weight Age In groups A and B, Group A was characterized by (Morley, 4 years accepted only when a weights fell below the 10% 8 and 3 children died, lower birthweights, diarrheal Bicknell, and clinical examination percentile at 6, 9, or 12 respectively. bouts before 6 months, and Woodland confirmed diagnosis and months) was selected measles and pertussis bouts at 1968) illness was considered of to group A (the under- 6­12 months; marked distinction sufficient severity to nourished group); a between the groups in weight warrant specific child who remained remained until children were treatment between the 50th and 5 years old. 75th percentile (at 9 or 12 months) was selected to group B. Guatemala Rural 716 0­7 23 Retrospective Every Used low versus high Age, sex, dietary Children were Found on average 6.3% (Martorell, months months disease reporting 6 months diarrhea groups; Used supplement subject to (3.5 centimeter) length and 11.0% Habicht, and every 2 weeks linear regression. supplemental (1.5 kilogram) weight relative others 1975; nutrition. difference; Martorell, Yarbrough, Martorell, and others' (1975) regression Yarbrough, techniques revealed dose- and others response and 10% of growth 1975) deficits would be explainable by diarrhea. Gambia Rural 152 3 months to On Self-reported period During clinic Period prevalence Age, season Treatment was Mean difference between 100% (Rowland, less than 3 average, prevalence (percentage visits (3- to was regressed with encouraged and and 0% period prevalence of Cole, and years 80 weeks of sick days) of disease 10-week height and weight facilitated. diarrhea in height and weight gain Whitehead between clinic visits intervals) gain. was ­4.2 millimeters and ­746 1977) (3­10 weeks); diarrhea grams per month; regression period prevalence 14% techniques revealed dose- response. Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition Mexico Rural 276 0 months 3 years Retrospective probing of Monthly, but Mean growth in Age, sex, 19 children who Diarrhea was the only infection (Condon- disease by a health weight upper and lower "growth failure" developed clinical associated with decreased Paoloni worker every second measured diarrhea quartiles protein calorie malnu- weight gain, and high-diarrhea and others week to establish period every 2 weeks were compared in a trition were excluded group attained 95% of the 1977) prevalence factorial design. weight of the low-diarrhea group; height was not affected by diarrhea. St. Lucia, West Rural 229 Less than 6 up to 18 Examination of stools Monthly Quasi- Comparability by Ascaris infected In the nonintervention village, Indies (Henry months months every 6 months for experimental design for children were actively children remained shorter and 1981) Ascaris and Trichuris; community trial; confounders: treated with weighed less and had mothers instructed to fill testing took place in food intake, piperazine phosphate. helminths, diarrhea, and other out diarrhea cards a factorial design of socioeconomic water-washed disease more the effectiveness of conditions, often than in the village, where household water general both household connection and connections and household sanitation were provided; when water-seal latrines to conditions interventions were introduced prevent helminths. they led to catch-up growth. Taiwan, China Rural 118 0 months More than Daily collection by nurses Monthly or Twelve disease Nutritional Nutritional supple- Negative effect of diarrhea on (Baumgartner 6 years to record frequency, bimonthly categories were treatment, mentation (800 growth rate was the most and Pollitt duration, and total days regressed with treatment kilocalories) was important finding, though it 1983) per symptom growth. timing, sex provided in a placebo- explained only 3% of variance; controlled regression techniques revealed trial. dose-response. Northeastern Small 297 study 0­6 years 3 months Daily diary by mother or Quarterly Regression Age, breast- Antidiarrheal Multiple different Brazil (Guerrant town participants, legal guardian plus (height); analysis of 482 feeding, growth treatments pathogens were found in and others plus of which weekly visits by a weekly (weight) child-periods of status before the (iodochlorhydroxyquin 50% of episodes, and height 1983) rural 53 subjects surveillance team 3 months were period and kaolin pectin) faltering attributable to adjacent were used recording clinical regressed against were given. episodes was especially areas to study symptoms to the number of pronounced (41%) during with effects on establish period diarrhea episodes the second year of life; 127 varying nutrition prevalence; Breastfeeding found to be highly (continued) TABLE A.1 128 Cohort Follow-up Studies Relating Infectious Disease and Nutritional Status of Children in Developing Countries (continued) Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition levels of collection of stool protective, especially in first environ- specimens by health 6 weeks. Regression techniques mental worker revealed dose-response. health services Bangladesh Rural 157 6­48 12 months Period prevalence Monthly Growth in zero- Age, All cases of Children with no diarrhea gain on (Black, Brown, months of diarrhea within period prevalence weight and diarrhea average 0.42 centimeter more in and Becker 60-day periods as periods was length at were medically length per year compared with 1984) actively established compared with base-line treated. children suffering from diarrhea; by a health worker who growth in nonzero diarrhea also adversely affected visited homes every periods. weight gain; invasive Shigella 2 days and enterotoxigenic E. coli might especially cause irreversible growth faltering; regression techniques revealed dose- response. Bangladesh Rural On 5­24 6­19 Period (60-day) Five Mean duration of Age No association of watery (Henry and villages average, months months prevalence and anthropometric watery diarrhea and diarrhea with growth faltering others 1987) 300 children attack rates based surveys every dysentery in different was noted, but stunting was on weekly visits by 6 months anthropometric associated with dysentery. health assistants categories were tabulated. Bangladesh 12 rural Approxi­ 1­4 years 9 months Weekly visits by field Bimonthly Multiple regression Age, dwelling Several Those with greater than 10% (Bairagi and villages in mately 1,000 workers to probe analysis was used to floor space observational diarrhea period prevalence had others 1987) Matlab (7-day recall) diarrhea investigate diarrhea- (proxy for epidemiological 71% (weight) and 75% (height) episodes and duration; nutrition relationship. socioeconomic studies led by growth velocity throughout the no distinction between status), International Centre follow-up period, compared with watery diarrhea and mother's for Diarrhoeal Disease those who did not have diarrhea; dysentery education Research were catch-up growth among those carried out in who had less than 10% diarrhea Matlab region. period prevalence prevented growth faltering. Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition Sudan Urban 439, but 0 months 1 year Morbidity probed Every 2 weeks Number of days ill Age Weight gain deficits from 1 day of (Zumrawi, falling to at visits every were regressed with diarrhea and cough or cold was Dimond, and 262 due to 2 weeks visits on the weight gain; weight 32 and 16 grams, respectively; Waterlow drop-outs and basis of 14-day recall; loss due to infections External comparison revealed 1987) families period prevalence 6% were externally (WHO that almost all weight gain deficit moving away reference) and was due to infections, though internally between internal comparison revealed quartiles assessed little weight gain deficit attributable to infections; regression techniques revealed dose-response. Gambia Urban 126 0 months 24 months Number of illness Monthly Illness growth Age, season Study infants had Combined effect of acute lower (Rowland, days (10 categories) periods were ready access to a respiratory infection (­320) and Rowland, and probed 1 month compared with no- pediatrician whenever diarrhea (­610) was 930 grams, Cole 1988) retrospectively illness periods. they were ill. but there was no effect on during every clinic height; acute lower respiratory visit to establish infection impact varied by period prevalence season; effects were exclusively found among non- breastfed babies; regression techniques revealed dose- response. Bangladesh Rural On average 6­35 months 42 months Height and weight Quarterly (varying) Growth rates of Effective case Only transient changes in (Briend and 230a `fourteen measured quarterly; diarrhea-free periods management with oral growth were observed, and were others 1989) quarterly diarrhea morbidity were compared with rehydration therapy caught-up later, though rounds' collected by means of growth rates of was organized within dysenteric patients were weekly recall by the periods with varying the study. 40.7% observed to experience stunting children's mothers prevalence of diarrhea, of periods excluded with a 3-month lag. with attention to because of missing timing of disease data within studied periods. 129 (continued) 130 TABLE A.1 Cohort Follow-up Studies Relating Infectious Disease and Nutritional Status of Children in Developing Countries (continued) Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition Colombia (Lutter Urban 148 0 36 Semimonthly visits Semimonthly Linear Age, Throughout Each day ill with diarrhea is and others 1989) setting (control), months months to establish incidence regression and treatment the study, families associated with a reduction 0.03 with 140 and duration of diarrhea, 2-way analysis of group, received free obstetric cm (p < 0.001) in attained length at least (supplement) vomiting, variance were size at and pediatric care, at age 36 mo; food supplements half of and respiratory used to model either birth including medicine; the effectively mitigated growth siblings and skin infections increment supplement group faltering (4.9 centimeters by age under 5 who to establish of weight and height received 600 36 months), especially among were less period prevalence; since birth or kilocalories and children who experienced greater than 85% diarrhea prevalence absolute height and 30 grams of protein. than 10% diarrhea period weight for 4.5­12.2% according weight until prevalence; the difference age. to age 36 months. between this group and the WHO (World Health Organization) standard was 13 centimeters at 36 months (that is, the treatment group experienced growth faltering of about 8 centimeters); regression techniques revealed dose-response. Brazil Urban 175 0­5 28 Three times a week; Monthly Descriptive Age Authors noted that the greatest (Schorling slum in years months during third year, time-series analysis contribution to the declines came and others northeastern extensive stool was made. from a 50% increase in the 1990) Brazil, where specimens collected prevalence of persistent diarrhea an excep- from acute, persistent, and that this decline closely tionally high and control children; followed a decrease in the rate of specimens subject to prevalence of severe malnutrition endemic intensive microbiological (weight for age) (p = 0.01); diarrhea testing control children also had high persists rates (37%) of diarrheagenic agents. Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition Bangladesh Rural 66 5­18 months Up to 14 Home visits every Monthly Three different Energy Matlab is a special If diarrhea and fever were (Becker, Black, villages in months second day by health random effects intake, age observation region, reduced to the level of a U.S. and Brown Matlab worker, who recorded models were used, where many study children would gain 226 1991) area diarrhea, vomiting, nasal with morbidity period observational grams per month more; discharge, cough, prevalence fitted as epidemiological regression techniques revealed anorexia, and fever independent studies are carried out dose-response. variables. by the International Centre for Diarrhoeal Disease Research. Brazil (Victora Relatively 5,914 0 months 36­52 During two surveys Every second Comparisons were Age, sex, income Effect of early infancy hospital- and others wealthy months every second year, year made of mean level, birthweight izations strongly predicted 1990) urban mothers probed about standard z-scores growth faltering (at 36­52 setting hospitalizations caused by, among months), though pneumonia-- for example, pneumonia hospitalized and unlike diarrhea--was statistically and diarrhea; validation nonhospitalized significantly linked with stunting done by reviewing records children. (for example, early infanthood of 120 hospitalizations diarrhea and pneumonia hospi- talizations decreased mean height-for-age z-score by about 50% and 33%, respectively). Zimbabwe Rural 204 0­12 months 22 months Weekly maternal Monthly Comparisons of Children were Authors found very little (Moy and impover- recall to establish growth rates between exclusively breastfed difference in growth rates with others 1994) ished period prevalence diarrhea-free periods until 3 months; all the most diarrhea and with setting and periods with children were fully least; study showed that a where varying period immunized, and were diarrheal episode led to 51 caloric prevalence of diarrhea, mothers trained to grams and 0.18 centimeters of intake was with attention paid to treat children with growth loss, which could not 65% of timing of disease oral rehydration explain the significant growth recom- within studied faltering in the in the cohort; mended 131 (continued) 132 TABLE A.1 Cohort Follow-up Studies Relating Infectious Disease and Nutritional Status of Children in Developing Countries (continued) Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition periods; in addition, supplements. high timing of diarrhea within the 3- growth rates were (more than 9) month periods suggested that compared between and low (fewer growth faltering was transient, than 4) occurrence though with respect to linear of diarrhea in growth it was not clear cut. 22 months. Philippines Urban 2,186 0 months 24 months Bimonthly maternal Bimonthly Lagged weight Breast-feeding Recent diarrhea had a substantial (Adair and and (+375) recall (prior week) of gain was and energy intake significant negative effect on others 1993) rural infants acute respiratory infection regressed against from other weight of children after 6 months areas and diarrhea episodes morbidity; an instru- sources, of life, whereas acute respiratory around mental variables preventive health infections significantly negatively and within technique was care, maternal affected growth up to 6 months Cebu used as explanatory weight, season of of life. Regression coefficients: variables were measurement approximately 250­400 grams correlated. lagged or immediate effects of diarrhea or acute respiratory infections in either 0­6-month or 6­24-month period; results are suggestive of lasting irreversible effects; breastfeeding was strongly protective. Sudan Poor 120 0 months 12 months Biweekly recall Monthly Comparison was made Housewives' children were on (Harrison, peri- mother-infant of mean duration of average 3.8 centimeters longer Brush, and urban pairs illness and growth of at 12 months although they were Zumrawi 1993) Khartoum children born to shorter at birth; on average housewives and illness duration at 2­12 months nonhousewives was up to 5 times onger among children born to working mothers despite the fact that such children tended l to have Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition less access to water supply, sewage systems, and general conditions of hygiene. Honduras Low- 141 mother- 0 months 12 months Morbidity data (upper Monthly Comparison was Sex; birth- Willingness to There was an initial rapid catch- (Cohen and income child pairs respiratory infection, between made of growth weight; breastfeed for 26 up compared with the U.S. data others 1995) com- participating in fever, diarrhea) collected 7 and 12 velocities maternal height, weeks was an at less than 3 months; after 6 munities a randomized at less than 4 months months; before between age, and inclusion criterion, months and especially at greater in San trial of 3 during weekly visits that, either weekly treatment groups education; and during weekly than 9 months, rapid growth- Pedro treatment (7-day recall); mothers or more irregularly and U.S. group, income, visits this habit faltering; intrauterine growth Sula, groups kept stool frequency and at less than standards at potable water;, was encouraged; restriction was a strong where consistency records 16 weeks 0­4 months, intervention hygienic food was predictor of stunting; fever, but environ- to establish period 4­6 months, group, breast provided during post- not diarrhea, predicted growth mental prevalence (about 3.2% 6­12 months. milk energy weaning faltering between 6 and 12 sanitation for diarrhea); at 4 months, period for those months; the low morbidity rates was poor at greater than income, (2 arms) who discon- and lack of association of 6 months, monthly visits maternal BMI at tinued exclusive diarrhea with growth may be probed 30-day recall. 4 months breastfeeding, attributable to the study's according to the inter- emphasis on vention protocol. breastfeeding. Sudan Peri-urban 167 0 months 12 months Biweekly mother's Monthly Infancy was Mother's socioe- Duration of illness in the first 6 (Brush, Khar-toum recall of the duration of divided into conomic status months was strongly inversely Harrison, illness periods of greater and income, age related to length and weight and than 6 months of weaning, and growth at 12 months. Waterlow 1997) and less than disease duration 6 months, and at under 6 biweekly period months prevalence was regressed with growth parameters. 133 (continued) 134 TABLE A.1 Cohort Follow-up Studies Relating Infectious Disease and Nutritional Status of Children in Developing Countries (continued) Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition Guinea- Peri- 1,064 0­? 36 months Biweekly visits by Approximately Growth faltering was Preinfection Long-lasting growth-faltering Bissau urban followed months field workers; stool 3 months studied in different weight, sex, age, effect was observed in both (Molbak district of for 1,441 (462 specimens collected the age groups (less season, and age weight and height velocities, and others Bandim II child-years recruited same day if child was sick than 1 year, 1­2 years) at infection effects being most pronounced 1997) at birth) with three different at under 1 year. At less than 1 lag periods and a year, cryptosporidiosis lowered common estimate. weight by 392 and 294 grams, among boys and girls respec- tively; respective results for height were ­4.7 and ­8.3 millimeters. Indonesia Rural 141 0­24 months 36 months Weekly visits by Monthly Period Duration of a Sick children were Weight indexes were signifi- (Kolsteren, Indonesia, health worker; ill children prevalence of specific disease referred to the cantly affected, especially by Kusin, and where revisited after 3 days to acute respiratory episode, sex, project clinic; acute respiratory infections after Kardjati childhood establish period infections and age virtually all were 6 months of age, but no effect 1997) infections prevalence of acute diarrhea was breastfed on linear growth was observed; are relati- respiratory infection, associated with until 12 months. duration of diarrhea, acute vely rare diarrhea, and fever growth velocity respiratory infections, and fever in 4-week periods. did not contribute to length velocity, but weight velocity was adversely affected by acute respiratory infection duration. Peru Peri- 207 0­3 23 Stool specimens Monthly Comparison was Sex, weight, and On average, symptomatic and (Checkley urban months months collected weekly made of growth stunting indicator asymptomatic children gained and others shanty- irrespective of velocity in at the start of 342 and 162 grams less after the 1997) town symptoms; diarrhea 1-month growth period; interaction first Cryptosporidium infection; symptoms probed periods among of C. parvum and because asymptomatic infections biweekly by a health those infected by diarrhea; presence are twice as common, they might worker Cryptosporidium of G. lamblia; have a bigger public health effect (first time infected seasonality through malnutrition. Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition included and first month after infection covered) and those not infected. Brazil (Steiner Shanty- 186 0 months 5 years Stool samples collected Quarterly Growth of E. coli Presence of Cohort was well Regardless of presence of and others town in periodically from healthy (EAggEC) positive pathogens described. diarrhea, EAggEC positive 1998) north- children and during each children with other than E coli children showed significantly eastern episode of diarrhea persistent and EAggEC increased growth faltering (more Brazil acute diarrhea than1.5 drop in Z-score) 3 and children without months after infection. EAggEC diarrhea were children without diarrhea (n = 6) compared with growth showed intestinal inflammation of children who were and also had growth faltering not carrying enteric (­0.64 in z-score; p = 0.064); pathogens. EAggEC remains dramatically underreported because showing its presence requires special techniques. 15­31% of asympto- matic children in Brazil and 6.5­9.9% in India are carriers. Peru Peri- 185 0­3 24 months Weekly collection of Monthly measure- Natural cubic- Age prior C. parvum when acquired very (Checkley urban months stool specimens to ments by health B-spline was stunting, food early after birth--and and others Lima diagnose C. parvum workers used to model expenses, especially if child is stunted 1998) and biweekly growth, with special number of prior to infection--likely leads diagnosis of diarrhea to attention paid to diarrheal to irreversible stunting; C. establish period 63 children who episodes, parvum contributes to about 1 prevalence became infected persistent centimeter height loss after 1 with C. parvum diarrhea, water year of infection. during the study. source and quality 135 (continued) 136 TABLE A.1 Cohort Follow-up Studies Relating Infectious Disease and Nutritional Status of Children in Developing Countries (continued) Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition Brazil (Lima Shanty- 39­42 0 months Up to Examination 3 Quarterly Growth at Age, sex, feeding Statistically significant short- and others town in (first-ever 48 months times weekly by a nurse; 3 months before and practices term growth faltering (weight for 2000) north- persistent extensive sampling of after first-ever age and height) was found, but eastern diarrheas diarrheal pathogens from persistent diarrhea there was no effect on height for Brazil found among those who developed episode was compared age; Cryptosporidium, Giardia, 179 children) persistent diarrhea and with that of children enteric adenoviruses, and matched controls who did not develop enterotoxigenic E. coli were persistent diarrhea. significantly associated with persistent diarrhea; even partial breastfeeding was found to be protective against diarrheal disease; persistent diarrhea was strongly associated with lack of toilet (pit or flush). Bangladesh Rural 412 6­48 months 6­48 months Home visits by Monthly Weight and Age, sex, Matlab is a special Strong evidence was found that (Alam Matlab trained health height gains parental observation region, watery diarrhea, but especially and others region workers every in no infection, education, where many dysentery, causes growth 2000) 4 days; and watery watery diarrhea, household observational faltering in the short and long diarrhea and and dysentery income, land epidemiological terms. Watery diarrhea and dysentery episodes groups were ownership, studies are carried dysentery resulted in 300 recorded to establish compared by baseline out by the grams (0.68 centimeters) and period prevalence (approxi- using 3-month nutritional International 516 grams (1.34 centimeters) of mately 10%) or 1-year periods. status, Centre for statistically significant growth diarrhea Diarrhoeal faltering, respectively, in a period Disease Research. 1-year period. prevalence Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition Pakistan Urban 994 pregnant n.a. Children Previous month 1-, 3-, 6-, 9-, 12-, Logistic Socioeconomic Intrauterine growth restriction (Fikree, squatter women visited at one, recall (acute lower and 24-month regression technique indicators was the strongest predictor of Rahbar, and settle- initially three, six, respiratory infection) or 2 visits used to estimate (income, house, stunting and wasting; water Berendes ments recruited; nine, twelve, weeks (diarrhea) during relative risk of water supply); supply and diarrheal episodes 2000) of 738 eighteen visits; for child to be becoming malnour- maternal predicted stunting, but risk Karachi newborns and twenty- classified as having ished indicators factors for wasting were limited recruited for four months diarrhea he or she had to when child was (education, only to socioeconomic status. follow-up have diarrhea in at least exposed to infections weight, parity, Stunting rose from 27.5% at 3 2 of 3 contact times (that first child); months to 67.9% at 18 months, is, exposure relied on the feeding practice which the authors noted could number of episodes) (breast or be due to feeding practices and nonhuman milk) exposure to a more contami- nated environment than found in Western countries; breast- feeding was protective. Brazil (Moore Shanty- 119 0 months 3 years, Home visits 2 or 3 times Quarterly Repeat measure Nutritional Results represent a Study found early childhood and others town in 9 months weekly to establish modeling was status in infancy, best-case scenario as diarrhea and intestinal helminths 2001) north- clinical period prevalence used to explain other parasitic this shantytown has were independently associated eastern of diarrhea and stool growth faltering infections, been subject to with profound and lasting linear Brazil samples to establish caused by early diarrhea in later substantial nutritional growth faltering. By age 7, an enteric pathogens, ova, childhood diarrhea childhood, interventions; Ascaris average child experiencing 9.1 and parasites and intestinal socioeconomic infections were diarrhea episodes before age 2 helminths. status treated. had a 3.6-centimeter growth deficit; early helminth infection added 4.6 centimeters to this deficit. Guinea-Bissau Peri- 107 0­2 years 6.6 months Morbidity recorded Monthly (knee-to- Natural growth history Children were examined Authors found that the negative (Valentiner- urban (median) during weekly visits; heel length) and of children with by a physician and effect of persistent diarrhea on Branth and setting period prevalence 10.3% weight, total persistent diarrhea offered fluid therapy linear growth was sustained others 2001) length quarterly was studied. and antibiotics, during the follow-up period, but according to WHO some catch-up in weight gain guidelines; breast occurred in the treatment group. feeding was encouraged; 137 half of children received nutritional interventions. (continued) 138 TABLE A.1 Cohort Follow-up Studies Relating Infectious Disease and Nutritional Status of Children in Developing Countries (continued) Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition Egypt Peri- 143 0­24 12­22 months Twice-weekly home visits Approximately Study analyzed odds Age, socioeco- Oral rehydration There was a ­0.16 (p < 0.05) (Wierzba urban to establish number of quarterly ratio to fall below ­2 nomic status, therapy was provided, reduction of the standard z- and others Alex- diarrhea episodes standard deviations mother's and referrals for score; when periods extended to 2001) andria, greater than or equal education, days hospital care were 6 months and diarrhea occurred where to 1 episode versus of exclusive made when needed; during the first half of the malnu- none during interval; breastfeeding, exclusive breast- period, the association remained trition additional special presence of feeding was negative in models but was not rates emphasis was placed sanitary latrine ascertained during statistically significant; authors remain on 6-month periods visits. noted that catch-up may below where diarrhea compensate growth faltering, 20% occurs in the first but those with diarrhea in the half, to study catch- first half when compared to up growth in a nondiseased individuals had disease-free slower growth velocity environment. (p > 0.05). Peru Peri- 224 0 months 35 months Active surveillance of Monthly CAR1 error term Age, sex, breast- Study does not Diarrheal period prevalence in the (Checkley urban diarrhea to establish modeling was used feeding, water specifically mention, cohort explained 2­27 % of the and others Lima clinical period prevalence based on 156,436 supply and but as children were growth retardation, and the 2003) and stool samples to child-days with 3,335 sanitation attended daily, case average child was 2.5 establish pathogens days of diarrhea maternal stature management must centimeters shorter than interna- (2.1%). have been included. tional peers; diarrhea, especially before age 6 months, predisposes children to irreversible stunting; with extensive resolution of collected data and elaborate modeling, this study showed decelerations on linear growth later in life as a result of early infancy exposure to diarrhea; a 10 cm difference in maternal stature corresponded to a 1.5 cm height deficit at 24 months Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition Malawi (Maleta Rural 767 0 months 36 months Data on infant Monthly until 18 Odds ratio of Average size High attendance rates Infancy disease episodes were and others (covered feeding and morbidity months; thereafter falling below at 1­3 months, (128 died and only 62 strongest determinants of 2003) 95% of collected monthly quarterly until severe distance to dropped) were a growth faltering, followed by deliveries until 18 months; no 36 months (­3 standard health center, feature throughout intrauterine growth restriction in the morbidity data collected deviations) gestation less the study; during the (using a proxy measure); catchmen thereafter underweight and than 37 weeks, first prenatal visit disease episodes remained t area) stunting and maternal HIV, mothers were offered strong predictors of moderate (?2 delivery by medical examinations underweight and wasting, but standard deviations) trained assistant and laboratory adjustment with intrauterine wasting during or not, sex measurements. growth restriction (which follow-up, using remained a highly significant greater than or equal predictor of stunting) dropped to 2.4 versus less infancy disease episodes from than 2.4 disease 3.4 to 1.1 (adjustment with episodes per month itself). in infancy as the explanatory variable. Tanzania Urban Dar 524 6­60 12 Morbidity episodes Monthly Mixed-effect models Age, index HIV was strongly associated-- (Villamor es Salaam months months (diarrhea, on average were estimated to episodes, HIV, especially in the 6- to 11-month and others hospital 3.4 episodes per child see how much HIV maternal period with growth faltering (1.3 2004) (children per year) established infection causes education, centimeters and 1.3 kilograms), admitted during monthly visits growth faltering; treatment arm, whereas diarrhea, though it recruited using pictorial diary; adjusted various anemia decreased length in the 6­11- to receive blood samples from all morbidity effects on month period by 1.4­2 a series of children at hospital; growth faltering were centimeters, did not show linear A-vitamin blood collected from also estimated. dose-response; dysentery shots) subsamples later appeared to be related to lower weight gain (0.32 kilograms) in children in the 12­23-month group. 139 (continued) 140 TABLE A.1 Cohort Follow-up Studies Relating Infectious Disease and Nutritional Status of Children in Developing Countries (continued) Control Country Number of Age at Follow-up Exposure Data Anthropometric Analysis Variables Notes Effect of Disease on (study) Setting Children Entry Period Collection Data Collection Strategy (ok if blank) (ok if blank) Malnutrition Brazil (Prado Urban 597 6­45 months 16 months Stool samples collected Every 6 months Variance and Sex, age, house All children were Both symptomatic and asympto- and others once during the first year, covariance were floods during treated to eradicate matic Giardia infections had a 2005) but diarrhea symptoms analyzed. rain, presence of Giardia lamblia, lasting effect on linear growth; probed twice a week to rubbish near Ascaris lumbricoides, investigators were able to establish period house, duration and Trichuris trichiura exclude reverse causality (that prevalence of interval during the is, the mere susceptibility of the between first year of study. malnourished measurements, to Giardia infection as the prevalence of explanation for the short stature diarrhea during of those infected) by showing that interval and the effect on stunting with a lag. intensity of infection, Ascaris lumbricoides and Trichuris trichiura baseline anthropometry Colombia Remote 133 5­7 12 Morbidity recorded Approximate 3- Hierarchical linear Initial status, Continued breast- Study showed the substantial (Alvarado rural months months by mothers recorded daily month intervals model was used to gender, preterm feeding (more than 6 protective effect of breast- and others Afro- (diarrhea, cough, fever, or model time-varying or term delivery, months) was feeding, which diminished with 2005) Colombian other nonspecific and fixed variables on maternal height, widespread at 15 increasing socioeconomic status village symptoms); data collected weight and height mother's months; women in- and which probably mitigated (Guapi) by a health worker weekly gain. education, child's creased breastfeeding growth faltering caused by age, breast- during disease diarrhea morbidity; respiratory feeding episodes, unlike in infections were associated with many other cultures. growth faltering. Source: Compiled by World Bank team. a. Not a fixed cohort, because surveys were made in an "open cohort." A P P E N D I X B Review of Studies on Nutritional Status and Education THIS APPENDIX REVIEWS RECENT LITERATURE on the effects of low height for age in childhood on subsequent educational outcomes.1 One group of studies investigates the effects of malnutrition on cognitive function, measured in terms of IQ scores or performance on cognitive tests (Fishman and others 2004).A second group assesses the effect on educational outcomes. This appendix concentrates on the latter. Several longitudinal studies have recently become available that investigate the effect of early childhood malnutrition on educational outcomes (table B.1). The majority of longitudinal studies use stunting (low height-for-age z-score, or HAZ) as a measure of nutritional status, because HAZ is considered an indicator of chronic nutritional deficit. These studies use data that follow children over time, usually from birth well into primary or secondary schooling, and contain various anthropometric, schooling, and socioeconomic status indicators. Such longitu- dinal data have important advantages in terms of investigating causal relationships between malnutrition and educational outcomes compared with the cross-sectional data often used in earlier studies. Grantham-McGregor and others (2007) present an analysis of longitudinal data of more than 2,000 male youths in Brazil. They find that stunting (HAZ < ­2) in early childhood is associated with a loss of 0.9 years of grade attainment by age 18. Using the stratified data in their study, Grantham-McGregor and others (2007) 141 142 TABLE B.1 Studies of the Effects of Malnutrition on Educational Outcomes Source Country Data Children Indicator Impact on Education School grade attainment Grantham-McGregor Brazil Longitudinal, 2,041 males HAZ < ­2; HAZ ­2 to ­1 versus > ­1 0.91 lower grade and others 2007 1982­2001 attainment; 0.44 lower grade attainmenta Maluccio, Hoddinott, Guatemala (rural) Longitudinal, 1,469 Nutrition supplement 1.0 higher grade and Behrman 2006 1969­2004 attainment (women) Daniels and Adair Philippines (Cebu) Longitudinal, 1,970 HAZ < ­3; HAZ ­3 to ­2; HAZ ­2 0.80 lower grade attainment; 2004 1983­2002 0.33 lower grade attainment; 0.12 to ­1 versus > ­1 lower grade attainment Alderman and others Uganda (rural) Longitudinal, 570 HAZ 0.68 lower grade attainment 2006 1983­2000 per 1 HAZ decline Learning productivity (grade equivalents) Behrman and others Guatemala (rural) Longitudinal, 1,448 HAZ 0.84 school year equivalent 2006 1969­2004 per 1 HAZ Glewwe, Jacoby, and Philippines (Cebu) Longitudinal, 1,016 HAZ; HAZ < ­2.5 0.8 school year equivalent King 2001 1983­95 per 1 HAZ; 1.8 school year equivalent per 1 HAZ School dropout and grade repetition Maluccio, Hoddinott, Guatemala (rural) Longitudinal, 1,469 Nutrition supplement 0.1 grade per year faster and Behrman 2006 1969­2004 school progression (women) Walker and others Jamaica (Kingston) Longitudinal, 167 HAZ < ­2 versus HAZ > ­1 100% higher school dropout 2005 1986­2003 rate by age 17­18 Daniels and Adair Philippines (Cebu) Longitudinal, 1,970 HAZ < ­3; HAZ ­3 to ­2; HAZ ­2 to ­1 18% grade repetition; 11% 2004 1983­2002 grade repetition; 5% grade repetitionb Primary school enrollment Glewwe and Jacoby Ghana Cross-sectional 1,757 HAZ 0.34 year delayed school 1995 1988­89 enrollment per 1-unit decrease in HAZ Alderman and others Pakistan (rural) Longitudinal, 534 0.5 HAZ increase 18% increase in girls' 2001 1986­91 enrollment; 4% increase in boys' enrollment Glewwe, Jacoby, and Philippines (Cebu) Longitudinal 1,016 HAZ 2 months delayed enrollment King 2001 1983­95 per 1-unit HAZ decline Alderman and others Uganda (rural) Longitudinal, 570 HAZ 0.4 year delayed enrollment 2006 1983­2000 per 1-unit HAZ decline Effect of diarrhea Lorntz and others Brazil Longitudinal 77 Number of diarrheal episodes 0.7 month delayed school 2006 (shantytown) 1993­2001 (0­2 years) enrollment per episode of diarrhea Source: Compiled by World Bank team. a. Estimated here from table 5 in Grantham-McGregor and others (2007). b. Repetition rates are relative to less than ­1 standard deviation; estimated here from odds ratios and data presented in Daniels and Adair (2004). 143 144 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL estimate that mild stunting (HAZ ­1 to ­2) is associated with a loss of 0.4 years of schooling. Alderman and others (2001), using longitudinal data relating to 560 chil- dren in rural Zimbabwe, find that a decline in HAZ of 1 is associated with a loss of 0.68 years of grade attainment and 0.4 years of delayed primary school enrollment. Grade attainment is defined as completed grades in primary and secondary school. Behrman and others (2006), using longitudinal data of 1,448 children followed into adulthood in rural Guatemala, assess the effect of preschool, schooling, and postschooling experiences on adult cognitive skills (reading comprehension and nonverbal skills). They find that preschool nonstunting (HAZ ­2) increases reading comprehension test scores by the same amount as 4.2 additional years of schooling attainment. Using sample and subsample mean values of HAZ reported in the study, an improvement in HAZ of 1 has an effect on reading comprehen- sion equivalent to nearly one additional year of schooling attainment. Stunting is also found to have a profound negative effect on nonverbal test scores. To the extent that nonverbal cognitive skills may affect learning productivity per year of schooling, this estimated effect of a change in HAZ of 1 on schooling attainment is conservative. Maluccio, Hoddinott, and Behrman (2006) use the same data set to evaluate the effect of early childhood nutrition supplementation in Guatemala. The supple- ment was found to decrease the prevalence of severe stunting by 50 percent, increase the height of three-year-old children by 2.5 centimeters, and increase school grade attainment by one full year for women. No significant effect on schooling was found for men. The effect for women comes through a reduction in women who never attended school and an increase in women who progressed into secondary school. Women who received supplementation progressed faster through school than women who did not receive supplementation, at a rate of 0.1 grade per year in school. Walker and others (2005), using longitudinal data of 167 children in Kingston, Jamaica, find that 17- to 18-year-olds, stunted by the age of two (HAZ < ­2), had a school dropout rate twice as high as children who were not stunted (HAZ > ­1). Daniels and Adair (2004) use longitudinal data of 1,970 children in the Cebu metropolitan area (urban and rural households) in the Philippines. They estimate odds ratios for completing high school, delayed enrollment, and grade repetition in relation to early childhood stunting. Using the odds ratios and data presented in their article, one can infer that severe stunting in early childhood (HAZ < ­3) is associated with 0.80 years of lower grade attainment, that moderate stunting (HAZ ­2 to ­3) is associated with 0.33 years of lower attainment, and that mild stunting (HAZ ­1 to ­2) is associated with 0.12 years of lower attainment. Daniels and Adair (2004) also find that severe stunting is associated with repeating a grade REVIEW OF STUDIES ON NUTRITIONAL STATUS AND EDUCATION 145 at least once (18 percent of severely stunted children) and that 5 percent of mildly stunted children repeat a grade at least once as a result of stunting. Glewwe, Jacoby, and King (2001), using longitudinal data relating to more than 1,000 children in the Cebu metropolitan area, investigate the effect of stunting on learning productivity per year of schooling through primary school using achievement test scores (sum of math and English). They then convert the effect of stunting on test scores into the extra months of schooling it takes to compen- sate for the difference in test scores. They find that the effect on test scores of a change in HAZ of 1 is equivalent to 0.8 year of school attendance, and for the more severely stunted children (HAZ < ­2.5), it is equivalent to 1.8 years of schooling. Glewwe, Jacoby, and King (2001) also find that a decline in HAZ of 1 delays primary school enrollment by two months and increases the probability of first-grade repetition by 9 percent. Using the baseline data in Daniels and Adair (2004) on frequency of grade repetition by stunting status, they find that a grade repetition of 9 percent is equivalent to a 0.5-month delay in labor force entry for a severely stunted child and a 0.4-month delay for a mildly stunted child. Alderman and others (2001), using longitudinal data on more than 500 chil- dren in rural Pakistan, find that a 0.5 increase in HAZ is associated with an 18 percent increase in girls' school enrollment and a 4 percent increase in boys' school enrollment. Glewwe and Jacoby (1995) find in a study of children in Ghana, albeit using cross-sectional data, that a decline in HAZ of 1 is associated with a 0.34-month delay in primary school enrollment. Diarrhea and Education Lorntz and others (2006), using longitudinal data on children in a shantytown in northern Brazil, estimate the effect of early childhood diarrheal episodes on subse- quent timing of primary school enrollment. They find a 0.7-month delay in enrollment per diarrheal episode in the first two years of life. The median number of episodes in the first two years of life was 4.5 per year. On average, this finding corresponds to a two-week prevalence rate of nearly 22 percent, which is the same as reported in the 2003 Ghana Demographic and Health Survey for children under two years of age. Reducing this prevalence rate by half (that is, to 11 percent) would suggest a reduction in delayed enrollment by about three months. Conclusions The literature reviewed and presented in table B.1 provides substantial evidence of a significant effect of malnutrition on children's educational performance. The single most important effect is reduced learning productivity during schooling, followed by reduced grade attainment. Delayed primary school enrollment and 146 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL grade repetition, while important in their own rights, have a much lesser conse- quence for lifetime income (see appendix D and chapter 5). The effects of malnutrition on educational performance, as applied to Ghana and Pakistan to estimate the impact on lifetime income, are presented in table 5.6 (see also table D.7 in appendix D). The attributable fraction of malnutrition (stunting) from diarrheal infections associated with water, sanitation, and hygiene is discussed in appendixes A and D. The relation between educational perform- ance and future income is discussed in chapter 5. Note 1 This appendix is based on a review of several recent longitudinal studies that looked at the effect of low height for age in childhood on educational outcomes. This review was carried out by Bjřrn Larsen (consultant) specifically for this report. A P P E N D I X C New Estimates for Burden of Disease from Water, Sanitation, and Hygiene THE ESTIMATES IN TABLE C.1 ARE FROM the new study by the World Health Organization (Fewtrell and others 2007). The study incorporates the effects (through malnutrition) of water, sanitation, and hygiene on mortality and disease burden. 147 148 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL TABLE C.1 Burden of Disease (in DALYs) in Children under Five Years Attributable to Water, Sanitation, and Hygiene, by World Health Organization Subregions, 2002 Mortality stratum (thousands) Africa Americas DALYs High Attributable High Child, Very Low Low High to WSH- Developed Developing Child, Very Child, Child, Child, Number Related Risk Countries Countries High High Very Low Low High (thousands) (%)d (thousands) (thousands) Adult Adult Adult Adult Adult Population of children under 5 years 616,908 81,206 535,702 51,925 60,004 22,986 45,433 9,509 Total DALYs 433,832 13,004 420,828 82,044 85,893 2,299 15,358 5,063 Total WSH- related DALYs 107,881 786 107,095 26,568 26,178 39 1,397 990 Percentage of total DALYs 25 6 25 32 30 1.7 9.1 19.5 Diarrheal diseasesa 47,520 44.0 432 47,088 9,283 9,390 22 901 565 Intestinal nematode infectionsb 531 0.5 1 530 117 58 0 15 38 Malnutrition (only protein- energy malnutrition) 7,104 6.6 70 7,034 1,239 1,412 0 244 104 Consequences of malnutritiona 28,475 26.4 166 28,309 6,853 5,744 0 95 177 Trachomab 4 0.0 0 4 0 1 0 1 0 Schistosomiasisb 14 0.0 0 14 5 6 0 0 0 Lymphatic filariasis 146 0.1 1 146 17 18 0 0 0 Subtotal, WSH 83,794 77.7 669 83,125 17,516 16,629 22 1,256 884 Malaria 17,482 16.2 2 17,480 7,736 8,011 0 24 10 Onchocerciasis 5 0.0 0 5 3 2 0 0 0 Dengue 137 0.1 0 137 0 1 0 6 7 Japanese encephalitis 247 0.2 0 247 0 0 0 0 0 Subtotal, water resources management 17,871 16.6 2 17,868 7,739 8,014 0 30 17 Drownings 1,473 1.4 62 1,411 141 125 13 53 13 Subtotal, safety of water environments 1,473 1.4 62 1,411 141 125 13 53 13 Other infectious diseasesc 4,743 4.4 52 4,690 1,172 1,411 5 59 76 Source: Data based on literature review/expert survey. Not a formal WHO estimate but based on Prüss-Üstün and Corvalán 2006. Notes: DALY = disability-affected life year; WSH = water, sanitation, and hygiene. Subtotals may not add because of rounding. a. Data further validated by Comparative Risk Assessment methods. b. Included in the Comparative Risk Assessment study. c. Not attributable to one group alone. d. Percentage of all DALYs attributable to WSH-related risks. NEW ESTIMATES FOR BURDEN OF DISEASE FROM WATER, SANITATION, AND HYGIENE 149 Mortality stratum (thousands) Eastern Mediterranean Europe Southeast Asia Western Pacific Low High Very Low Low Low Low High Very low Low Child, Child, Child, Child, Child, Child, Child, Child, Child, Low High Very Low Low High Low High Very Low Low Adult Adult Adult Adult Adult Adult Adult Adult Adult 15,213 53,234 21,974 17,576 10,859 28,449 150,232 7,812 121,704 5,086 53,622 1,719 5,717 2,670 10,738 117,012 599 45,507 614 12,911 37 600 88 2,038 28,390 22 8,010 12.1 24 2 10.5 3.3 19 24 4 18 412 6,500 21 342 39 1,120 14,896 8 4,020 0 40 0 0 0 21 108 0 134 81 834 0 48 22 316 1,938 0 853 52 3,482 0 163 3 333 9,774 0 1,784 0 1 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 2 0 1 0 5 96 0 9 546 10,861 21 554 64 1,794 26,810 9 6,800 18 665 0 2 0 158 715 0 142 0 0 0 0 0 0 0 0 0 2 3 0 0 0 17 53 0 48 0 50 0 0 0 7 51 0 139 20 719 0 2 0 182 819 0 329 24 123 5 24 17 35 246 4 652 24 123 5 24 17 35 246 4 652 24 1,208 11 21 7 26 514 9 229 A P P E N D I X D Computing Country-Level Environmental Health Burden of Disease THIS APPENDIX PROVIDES A DESCRIPTION OF THE METHODS used to estimate the environmental health burden of disease,including effects on educa- tion, and relative monetary costs.1 An overview of the results for Ghana and Pakistan has been provided in chapter 5.The methodology (summarized graphically in figure D.1) has two main components. The first involves the valuation of mortality for each cause of death (left side of figure D.1), and the second relates to the estima- tion of educational costs of disease (right side of figure D.1). Mortality This section details the steps used to estimate mortality from environmental health risks (figure D.1, steps 1 to 6). The environmentally attributable disease in Ghana is estimated in reference to diarrhea, acute lower respiratory infections (ALRI), malaria, measles, protein-energy malnutrition, and a residual group of other causes of mortality (table D.1). It is important to note that diarrhea constitutes both a cause of death and a risk factor through its effect on malnutrition and consequent mortality. Deaths related to urban air pollution are not included because of the small effects on child mortality that can be calculated from current estimation methods. 151 152 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL FIGURE D.1 Summary of the Methodology Mortality Education Step 7: Step 8: Step 1: Data collection on Assessment of Data collection current child counterfactual child population, population, prevalence of prevalence of Step 2: Step 4: malnutrition malnutrition Creation of exposure Assessment of (stunting) (stunting) categories; assessment of counterfactual population at risk population at risk if Step 9: no diarrheal Assessment of attributable loss in school infections and no Step 3: performance from malnutrition Assessment of relative risk in other environmental each exposure category health risk exposure Step 10: Valuation · Estimation of current and counterfactual lifetime income of Step 5: children Estimation of mortality attributable to environmental health risks · Calculation of difference between (direct + indirect effect) current and counterfactual lifetime income for the whole population Step 6: Valuation Source: Compiled by World Bank team. TABLE D.1 Causes of Death and Risk Factors Considered in this Study Cause of Death Direct Environmental Risk Factor Indirect Risk Factor Diarrhea Inadequate water, sanitation, Malnutrition through diarrheal and hygiene infections Acute lower respiratory Indoor air pollution Malnutrition through diarrheal infection infections Malaria Poor management of water Malnutrition through diarrheal resources infections Measles Nonea Malnutrition through diarrheal infections Protein-energy Nonea Malnutrition through diarrheal malnutrition infections Other causes of Nonea Malnutrition through diarrheal mortality (excluding infections perinatal conditions) Source: Compiled by World Bank team. a. This report focuses on water, sanitation, and hygiene; indoor air pollution; and water resource management. None of these environmental health risks are considered to contribute significantly to these causes of death. COMPUTING COUNTRY-LEVEL ENVIRONMENTAL HEALTH BURDEN OF DISEASE 153 Total observed mortality--denoted as M0--in any given population is the result of exposure to different risk factors. If the fraction of mortality caused by environmental risk factors is denoted as FEH, then mortality attributable to envi- ronmental health risks, MEH, is given by the formula j=m MEH = F EH 0 j Mj , (D.1) j=1 0 where Mj is current annual cases of cause-specific mortality; j is ALRI, diarrhea, EH malaria, measles, and"other causes"of deaths; Fj is the fraction of cause-specific deaths j attributable to environmental health risks; and MEH is the annual cases of mortality caused by environmental health risk factors. The methodology for estimating MEH relies on the estimation of the impact EH fraction Fj for each cause of death j.A general formula for the fraction of deaths attributable to a set of risk factors is PRR P n n i i- ' i RRi F = i=1 i=1 , n PRR i i i=1 (D.2) where the subscript i denotes exposure category, Pi is the population's current exposure distribution, P'i is the population's counterfactual exposure distribution (that is, exposure distribution in the absence of environmental risks), and RRi is the risk of mortality in each exposure category.2 Step 1: Data Collection The estimation of the environmental health burden of disease and its subsequent monetary valuation uses data from a variety of sources.Typical sources of informa- tion are listed in table D.2. The more data are available from country sources, the better.As an alternative,one can use information from international sources,making the necessary assumptions to adapt the parameters to the country in question. Step 2: Assessment of the Population at Risk To compute the health burden of each cause of death or disease from environmental risk factors, one needs estimates of the population exposed to the relevant risk factors. Multiple environmental risk factors may contribute to a particular disease and mortality from this disease, and an environmental risk factor is often associ- ated with several levels of risk.It is therefore desirable to create exposure categories, allowing for as many combinations of risk factors and levels of risk as is practically 154 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL TABLE D.2 Estimating the Cost of Environmental Health Risks: Information Types and Sources Type of Information Source Source Level Exposure to risk factors Demographic and Health Surveys Country-specific (for example, access to (DHSs), Multiple Indicator Cluster sources water, malnutrition, Surveys (MICSs), national nutrition indoor air pollution) surveys, national censuses, Living Standards Measurement Surveys (LSMSs) Baseline health end Vital records, health information Country-specific points (for example, systems, and statistical sources crude mortality rate, estimates of cause-specific disease-specific mortality; DHSs and MICSs; mortality, disease WHO national estimates of prevalence) cause-specific mortality Relative risks of illness Example: the relative risk of Typically international and mortality: dying from different levels of studies and epidemiology literature malnutrition is summarized in meta-analyses Fishman and others (2004) Links between Example: the effect of diarrheal International studies and malnutrition and diseases on weight gain in meta-analyses infections: epidemiology children is analyzed in literature Checkley and others (1997) and Molbak and others (1997) Value of mortality and For the human capital approach: For the human capital morbidity national accounts statistics (for approach: country- example, gross domestic specific sources product), economic indicators (for example, wage rate) For willingness-to-pay approach: For the willingness-to-pay value of a statistical life studies approach: international (benefit transfer approach) studies and meta-analyses Cost of illness Economic indicators, public Country-specific sources health statistics, interviews with hospitals and practitioners; DHSs and MICSs (treatment rates and type of treatment) Source: Compiled by World Bank team. feasible,and to compute the population shares that fall in each category.Figure D.2 presents a simplified illustration for ALRI mortality from indoor air pollution and malnutrition (see table D.1). In the case of diarrheal mortality, relevant risk factors include (a) lack of access to water, sanitation, or both and (b) malnutrition. Exposure data can often be obtained from the Demographic and Health Surveys (DHSs). Ghana's 2003 DHS reports malnutrition levels for each child, and this information can be overlaid with household characteristics to obtain exposure to indoor air pollution and lack of water and sanitation. COMPUTING COUNTRY-LEVEL ENVIRONMENTAL HEALTH BURDEN OF DISEASE 155 FIGURE D.2 Exposure Categories Indoor air pollution No indoor air pollution Malnutrition Exposure category 1 Exposure category 2 No malnutrition Exposure category 3 Exposure category 4 Source: Compiled by World Bank team. Step 3: Assessment of Relative Risks Relative risks express the probability of an event (for example, developing a disease or dying from a disease) occurring in the exposed group versus the probability of the same event occurring in the group that was not exposed. Relative risks typi- cally are available from the epidemiology literature. However, current research literature does not contain estimates of the combined relative risk of multiple environmental risk factors, such as the risk of death from exposure to indoor air pollution and from malnourishment, with the latter partially caused by diarrheal infections from inadequate water, sanitation, and hygiene. In this analysis, the problem is overcome by assuming a Cox proportional hazard model in the calcu- lation of relative risks.3 If that model is used, the relative risk is calculated as RRi = exp(g1ENV + g2MAL), (D.3) where RR is relative risk; ENV is a binary variable (that is, one that takes a value of either 1 or 0) indicating the presence of an environmental risk; MAL is a binary variable indicating the presence of malnutrition; and g1 and g2 are the parameters linking the presence of risk factors to the relative risk of a disease. It is easy to see that equation D.3 implies that the relative risk for exposure category 1 in figure D.2 is equal to the product of the relative risks for exposure categories 2 and 3 (that is, RR1 = RR2 * RR3). The relative risks for exposure categories 2 and 3 are revealed in the epidemiology literature, which has been adjusted for potential biases.4 Table D.3 summarizes.5 The relative risks discussed in literature could, overestimate the mortality from environmental risk factors if the effects of underweight on mortality in Fishman and others (2004) did not control for environmental factors or if the relative risks of mortality from environmental factors did not control for underweight status. An alternative approach used to obtain the results in chapter 5 is therefore applied to assess the potential magnitude of overestimation.6 The approach uses relationships between least-squares coefficients when a variable is omitted from an equation. In the case of Ghana,the following two equations are estimated by using the 2003 DHS: MAL = b0 + b1 ENV (D.4) ENV = a0 + a1 MAL (D.5) 156 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL TABLE D.3 Relative Risks by Exposure Categories, Assuming Cox Hazard Model Environmental Health No Environmental Health Exposure Exposure Malnourished RR1 = RR2 * RR3 RR2 from literature, adjusted (for example, Fishman and others 2004) Not malnourished RR3 from literature, RR4 = 1 adjusted Source: Compiled by World Bank team. The estimated coefficients b1 and a1 are then used in the following equations to solve for gi, where bi represents the biased estimators in equation D.3: b2=g2+g1*a1 (D.6) b1=g1+g2*b1 (D.7) Step 4: Calculation of Counterfactual Exposure Categories The next step is to estimate the population in each exposure category in the absence of environmental risk factors. This step is important because child malnu- trition depends not only on environmental health risks, but also on various nutritional factors such as maternal malnutrition and low protein-energy intake. Therefore, one needs to separate out the effects of nonenvironmental factors on malnutrition. Because the main environmental health risk influencing malnu- trition acts through diarrheal infections, this analysis estimates malnutrition prevalence in a counterfactual child population that has not suffered from repeated diarrheal infections. Estimation of the effect of diarrheal infections on nutritional status and conse- quent disease burden involves the following steps: 1. A weight gain retardation factor, bt , is specified for each child i at age t and is i a function of diarrheal infection dose from initial age. 2. The counterfactual weight,Wt of child i at age t is estimated (that is, the weight Ni of the child if the child had not suffered weight gain retardation from repeated diarrheal infections). 3. The weight-for-age z-score (WAZ) for Wt and corresponding underweight Ni malnutrition rates are calculated. 4. Mortality from environmental factors is then estimated. The relationship between a child's observed weight, the weight gain retarda- tion factor, and the counterfactual weight is Wt -W0 = Di i (1- t x Wt -W0 , i) ( Ni i) (D.8) COMPUTING COUNTRY-LEVEL ENVIRONMENTAL HEALTH BURDEN OF DISEASE 157 which gives i Wt = 1 Ni Di (1 - t )Wt - t i, (D.9) i (1 - t )W0 i where Wt is the observed weight of child i at age t who has suffered weight gain Di i retardation from diarrheal infections, and W0 is the initial weight of the child. Table D.4 presents the weight gain retardation factors (for a subset of age groups) used in this study.7 This factor is the reduction in weight gain from repeated infections over a period of a child's life in his or her first five years. For instance, if a child's initial weight was 5 kilograms at the age of 3 months and was 10 kilograms at the age of 40 months, and if the child had suffered a weight gain retardation of 25 percent (bt = 0.25) during that period, then the child's weight i would have been 11.67 kilograms in the absence of the weight gain retardation. In contrast, the median child at 40 months of age in the international reference population has a weight of 14.6 kilograms. The share of the weight retardation from repeated infections is therefore 36 percent.8 The remaining weight retarda- tion (64 percent) is from other factors. The factors in table D.4 are used to calculate Wt for each child in the DHS Ni data set in Ghana, along with a new weight-for-age z-score for each child. The results imply that weight gain retardation from diarrheal infections is about 35 percent of total weight retardation relative to the median child in the interna- tional reference population (WAZ = 0) (see appendix A). This result is equivalent to approximately a 40 percent change in the standard deviation. This change in standard deviation is applied to each of the children in the Ghana DHS (for example, the new z-score is ­1.8 for an old z-score of ­3.0).9 TABLE D.4 Weight Gain Retardation Factors by Age and z-Score Age (Months) WAZ = --1 WAZ = --2 WAZ = --3 WAZ = --4 12 0.15 0.37 0.45 0.55 18 0.11 0.27 0.34 0.43 24 0.09 0.22 0.29 0.37 30 0.09 0.19 0.26 0.34 36 0.08 0.18 0.25 0.33 42 0.08 0.18 0.24 0.32 48 0.08 0.17 0.24 0.32 54 0.08 0.17 0.23 0.31 60 0.08 0.17 0.23 0.31 Source: Compiled by World Bank team. Note: WAZ = weight-for-age z-score. 158 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL TABLE D.5 Weight for Age in Children under Five: Current Rates and Estimated Rates in the Absence of Diarrheal Infections in Ghana Prevalence Rates in the Absence Prevalence of Weight Gain Retardation Weight for Age Score Rates in 2003 (%) from Diarrheal Infections (%) > --1 WAZ (not underweight) 45.3 67.70 --1 to --2 WAZ (mild underweight) 32.9 30.00 --2 to --3 WAZ (moderate underweight) 17.3 2.30 < --3 WAZ (severe underweight) 4.6 0.04 Source: Compiled by World Bank team. Note: WAZ = weight-for-age z-score. Table D.5 presents current underweight prevalence rates in children under five from the 2003 DHS and estimated rates in the absence of weight gain retardation from diarrheal infections. In the absence of retardation from diarrheal infections, the estimates indicate that there would be practically no severely underweight children in Ghana.The prevalence of moderate underweight would also be substan- tially reduced, while mild underweight would be about the same as reported in the 2003 DHS. Step 5: Estimation of Deaths Attributable to Environmental Health Risk Factors The fraction of cause-specific deaths attributable to environmental risk factors (for example, the fraction of ALRI mortality attributable to indoor air pollution and diarrhea-related malnutrition) are obtained by combining in equation D.2 the parameters obtained in steps 2 and 3. This procedure is undertaken for each cause of death for which environmental risks are a contributing factor. Figure D.3 explains, using ALRI as an example. The left side of the figure presents child population shares and relative risks for the current situation, and the right side presents the counterfactual situation (that is,no environmental health risks). Combining this information in equation D.2 gives the impact fraction at the bottom of the figure. Table D.6 presents the cause-specific mortality, values of impact fraction F, and mortality attributable to environmental risk factors for each cause of death in Ghana. Step 6: Valuation of Attributable Health End Points The last step consists of multiplying each end point by its unit cost. Mortality can be valued by first estimating a stream of income that the individual would have generated if he or she had been alive and then discounting the stream to obtain COMPUTING COUNTRY-LEVEL ENVIRONMENTAL HEALTH BURDEN OF DISEASE 159 FIGURE D.3 Exposure Categories, Population Shares, and Relative Risks of ALRI in Ghana Current Situation Situation with no Environmental Health Risks Not Not Exposed to Exposed to exposed to exposed to Population environmental environmental environmental Population environmental health risk health risk health risk health risk Malnourished 53.0% 1.7% Malnourished 0.0% 32.3% Not malnourished 41.6% 3.7% Not malnourished 0.0% 67.7% Not Not Exposed to Exposed to exposed to exposed to Relative risks environmental environmental environmental Relative risks environmental health risk health risk health risk health risk Malnourished 5.30 3.07 Yes n.a. 2.10 Not malnourished 1.72 1.00 No n.a. 1.00 Impact fraction 0.625 Source: Compiled by World Bank team. TABLE D.6 Estimated Mortality in Children under Five from Environmental Risk Factors, Ghana Mortality Attributable Mortality Impact Fraction to Environmental Cause of Mortality (2005) (%) Health Risks Pneumonia-ALRI 11,800 62.5 7,375 Diarrhea 9,900 91.0 9,009 Malaria 22,600 62.1 14,035 Measles 200 28.8 58 Protein-energy 1,000 40.0 400 malnutrition Other malnutrition-related 12,600 38.3 4,826 causes of mortality (excluding perinatal conditions) Total 58,100 35,703 Source: Compiled by World Bank team. Note: Not including perinatal conditions. ALRI = acute lower respiratory infection. a measure of lost productivity. This method--the human capital approach--is often considered to provide a lower-bound estimate of the value of avoiding mortality (see appendix F).Applying the human capital approach to valuation of mortality suggests a cost in Ghana equivalent to 5.55 percent of gross domestic product (GDP) for the effect of environmental factors.10 160 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Education The health links between malnutrition and the environment provide further insight into the issue of costing environmental health risks.11 Repeated infections in early childhood contribute to child malnutrition, which, in turn, impairs cogni- tive development and educational performance. This effect has an impact on labor productivity, reducing individual earnings. Malnutrition can also lead to delayed primary school enrollment and grade repetition, which, in turn, may have an impact on lifetime income in terms of delayed labor-force entry. To value the impacts of malnutrition through education, one needs to esti- mate the loss in education performance attributable to malnutrition, estimate the relationship between education and lifetime income, and combine this informa- tion to obtain the loss in lifetime income attributable to malnutrition. The discussion here begins with the data collection issues. Step 7: Data Collection Stunting, an indicator of chronic malnutrition, is often used when measuring the effects of malnutrition on education. Stunting is expressed as deviations from the median height for age z-score (HAZ) in an international reference population. Mild stunting means an HAZ between ­1 and ­2 standard deviations from the median child in the reference population (­2 < HAZ < ­1). Moderate stunting is defined as ­3 < HAZ < ­2 and severe stunting as HAZ < ­3. The current stunting prevalence rate in Ghana is presented in table D.7, row 1. Step 8: Assessment of Counterfactual Malnutrition-Attributable Loss in School Performance To assess the loss in school performance attributable to malnutrition, one first needs to establish the counterfactual situation. Table D.7 provides the counter- factual prevalence rates achievable by controlling for diarrheal infections. The share of severely malnourished children would go down from 10.5 percent (current situation) to 2.9 percent (situation without infections). Changes in prevalence rates are estimated by assuming a 25 percent change in standard deviation (consis- tent with the literature analyzed in appendix B and somewhat more conservative than the change used for underweight rates in table D.5). The 25 percent change in standard deviation is arrived at in a way similar to that described in step 4. The effect of nonenvironmental factors on malnutrition first needs to be separated--through computations of the malnutrition prevalence of a counterfactual child population that has not suffered from repeated diarrheal infections. The estimation of the effect of diarrheal infections on stunting involves the actions described in step 4. A smaller retardation factor is used for height for age than for weight for age because it appears from the research literature COMPUTING COUNTRY-LEVEL ENVIRONMENTAL HEALTH BURDEN OF DISEASE 161 TABLE D.7 Estimated Annual Cost of Education Outcomes from Stunting and Share from Environmental Factors in Ghana Standard Deviations > ­1 -1 to ­2 ­3 to ­2 < ­3 Nutritional status (prevalence rate) 1. Percentage of children (<5 years) stunted (height for age ­1 standard deviation), current situation 42.4 28.20 19.00 10.50 2. Percentage of children (<5 years) stunted (height for age <­1 standard deviation), counterfactual situation 51.90 32.30 12.90 2.90 3. Standard deviation of median child in each nutrition bracket n.a. ­1.50 ­2.40 ­3.50 Effects of malnutrition on education outcomes (years per malnourished child) 4. Loss in grade attainment 0 0.50 1.00 1.50 5. Loss in learning productivity (grade equivalents) 0 0.80 1.60 3.60 6. Delayed labor-force participation a 0 0.42 0.85 1.29 Present value of lifetime income (US$ thousands per child) b 7. Income at current malnutrition rates 42.3 21.6 12.6 6.5 8. Counterfactual income 42.3 23.0 13.8 8.4 Present value of lifetime income for birth cohort (US$ million) b 9. Labor-force participation (% of population age 15­65) 75 75 75 75 10. Present value of lifetime income for birth cohort (US$ million), current malnutrition rates 6,985 2,999 1,207 373 11. Present value of lifetime income for birth cohort (US$ million), counterfactual income 6,985 3,183 1,325 479 12. Present value of lifetime income for birth cohort (US$ million), current malnutrition rates 11,564 13. Present value of lifetime income for birth cohort (US$ million), counterfactual income 11,972 14. Total cost of environmental factors (US$ million) 407 15. Total cost of environmental factors (% of gross domestic product in 2005) 3.79 Source: Compiled by World Bank team. Note: n.a. = not applicable. a. From delayed primary school enrollment and grade repetition. b. Present value of lifetime income losses of an annual national birth cohort, adjusted for child mortality (discounted to childhood age = 1 year). 162 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL (see appendix B) that height growth retardation from infections is somewhat less than weight gain retardation (table D.8). Using the factors in table D.8,one can calculate the counterfactual height (in the absence of growth retardation from infections) for each child in the DHS data sets in Ghana, along with a new height-for-age z-score for each child. The results imply that height growth retardation from diarrheal infections is 25 percent of total height retardation relative to the median child in the international reference population (HAZ = 0) (see appendix A).The new z-scores are 25 percent higher than observed z-scores (for example, the new z-score is ­2.25 for an old z-score of ­3.0). Table D.9 presents current height-for-age prevalence rates in children under five from the 2003 DHS and estimated rates in the absence of height growth retar- dation from diarrheal infections.In the absence of height retardation from diarrheal infections, the estimates indicate a more than 70 percent reduction in the number TABLE D.8 Height Growth Retardation Factors by Age and z-Score Age (Months) HAZ = --1 HAZ = --2 HAZ = --3 HAZ = --4 12 0.05 0.13 0.19 0.24 18 0.04 0.09 0.13 0.18 24 0.03 0.07 0.11 0.15 30 0.03 0.07 0.10 0.14 36 0.03 0.06 0.10 0.13 42 0.03 0.06 0.09 0.13 48 0.03 0.06 0.09 0.12 54 0.03 0.06 0.09 0.12 60 0.03 0.06 0.09 0.12 Source: Compiled by World Bank team. Note: HAZ = height-for-age z-score. TABLE D.9 Height-for-Age Rates in Children under Five: Current Rates and Estimated Rates in the Absence of Diarrheal Infections in Ghana Prevalence Rates in the Absence Prevalence of Height Growth Retardation Height-for-Age (Malnutrition) Rates in 2003 (%) from Diarrheal Infections (%) > ­1 standard deviation (not stunted) 42.4 51.9 ­1 to ­2 standard deviation (mild stunting) 28.2 32.3 ­2 to ­3 standard deviation (moderate stunting) 19.0 12.9 < ­3 standard deviation (severe stunting) 10.5 2.9 Source: Compiled by World Bank team. COMPUTING COUNTRY-LEVEL ENVIRONMENTAL HEALTH BURDEN OF DISEASE 163 of severely stunted children and a more than 30 percent reduction in the number of moderately stunted children. The prevalence of mild stunting would be about the same as reported in the 2003 DHS. Step 9: Assessment of Malnutrition-Attributable Loss in School Performance A number of recent longitudinal studies (see appendix B) have investigated the effect of early childhood malnutrition on educational outcomes. These studies use data that follow children over time, usually from birth well into primary or secondary schooling, containing various anthropometric, schooling, and socioeconomic status indicators. The educational and cognitive impacts of malnutrition are expressed as changes in Final grade attainment Learning productivity or school achievements (often measured by cognitive and math and reading achievement tests) School grade repetition Years of delayed primary school entry The studies reviewed show that the change in grade attainment ranges between 0.12 years (mild stunting in the Philippines) and 0.91 years (moderate malnutrition in Brazil) and 0.7 years (for each z-score in Zimbabwe). The effect of malnutrition on learning productivity or achievement appears to be even larger, ranging from 0.8 years of grade equivalent in Guatemala and the Philippines for each z-score to 1.8 years for each z-score in severely malnourished children in those countries.Delayed primary school entry varies between two to four months per one z-score in the studies reviewed. The parameters used for Ghana are summarized in table D.7. The magnitude of the effect of a child's nutritional status on grade attainment applied to Ghana (row 4, table D.7) is somewhat lower than that in the study of Zimbabwe, about the same as that in the study of Brazil, but significantly higher than that in the study in the Philippines. The rationale for applying a magnitude higher than that found in the Philippines is the substantially higher primary school completion rate there. One important reason for that higher rate is the higher emphasis on school completion in the Philippines, which likely reduces losses in grade attainment of children with lower nutritional status.12 Primary school competition and years of children's school life expectancy in Zimbabwe are also higher than in Ghana. Choosing a lower effect on grade attainment in Ghana than that found in Zimbabwe may therefore be considered conservative and possibly an underestimate of actual effect. The magnitude of effect of nutritional status on school achievement or learning productivity applied to Ghana (row 5, table D.7), measured in grade equivalents, is about the same as found in Guatemala and the Philippines for children with mild and moderate height-for-age deficit or stunting. For children with severe 164 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL stunting, the magnitude applied is about the midpoint of that found in the Philippines and Guatemala. Delayed primary school enrollment is derived from the study in Zimbabwe. Grade repetition is half of that found by Daniels and Adair (2004) in the Philippines. The combined effect of delayed enrollment and grade repetition is reported in row 6, table D.7. The cost of delayed enrollment and grade repetition in terms of delayed labor- force participation is, however, very minimal compared with the cost of lower grade attainment and learning productivity. So variations in the magnitude of effects, within the range found in studies, have minimal influence on the esti- mated total cost of nutritional status. Step 10: Valuation To arrive at costs related to educational and cognitive impacts, one has to value the impact of lower grade attainment, learning productivity, and delay in primary school entry on lifetime income. To do this, one has to 1. Estimate private marginal returns to schooling--that is, a measure of the incre- ment in returns that one extra year of schooling allows. This measure helps to estimate the lost returns for one less year of schooling (owing to lower educa- tional attainment, for example). 2. Estimate the social cost of providing a year of schooling. 3. Using these estimates, calculate the marginal increase in income that schooling allows. 4. Estimate the lifetime income in the current situation and in the absence of infections (shown in rows 7 and 8, table D.7). The present value of lifetime income of a child in the current situation is t =65 PV(I) = I (D.10) 0[(1+ g)/(1+ r)]t , t =15 where I0 is initial annual income of adults at the time the child is one year old, g is annual rate of growth in real per capita income, and r is a constant annual discount rate. It is assumed that an individual has a productive working life starting at the age of 15 (t = 15) and ending at the age of 65 (t = 65) (see the Ghana DHS). The present value of counterfactual lifetime income of a child in the absence of infections is j=3 PV(IC)= PV(I) 1+Z i j (D.11) j=1 Ej , where dZ is the counterfactual change in height-for-age z-score; ij is the percentage change in present value of lifetime income from an additional year of schooling COMPUTING COUNTRY-LEVEL ENVIRONMENTAL HEALTH BURDEN OF DISEASE 165 (j = 1), improved learning productivity equivalent to a year of schooling (j = 2), and an additional year of labor force participation (j = 3); and dEj is the change in education outcome and labor-force participation, for j = 1, . . . ,3, per one z-score. Specifically, dZ is the absolute value of 25 percent of the current z-score of each child. The percentage change in present value of lifetime income from an additional year of schooling and from improved learning productivity equivalent to a year of schooling is estimated from table D.10. For change in lifetime income from an additional year of school (j = 1)--that is, increased grade attainment-- the cost of this additional year of schooling (C) is subtracted from income at age 15.13 The percentage change in lifetime income from the change in entry into the labor force is estimated on the basis of lifetime income flows. Change in educational outcome and labor-force participation for one z-score is as follows (see table D.7): 0.5 years of grade attainment (row 4, table D.7) 0.8 years of grade equivalent from learning productivity (school achievement) for children with mild and moderate stunting (row 5, table D.7) TABLE D.10 Parameter Values Applied in Estimation of Income Losses Parameters Ghana Pakistan Remarks Average annual 4,370,000 PRs 57,800 Ghana: GDP per capita in 2005 income per (a proxy for per capita income); income earner, Pakistan: income in 2001/02 2005 (Household Integrated Economic Survey 2001/02), adjusted to 2005 r 3% 3% Same as used by WHO for health outcomes g 2% 2% Assumed per capita annual real income growth over the next 65 years Ct 0.37 I0(1 + g)t ­1 0.37 I0(1 + g)t ­1 Assumption that cost of a year of schooling is 37% of per capita income at time = t i1 9.5% 7.2% Calculated from rate of return to a year of schooling; that is, 11.7% in Sub-Saharan Africa and 9.9% in Asia (see Psacharopoulos and Patrinos 2004), less the cost of an additional year of schooling i2 13.2% 10.9% Calculated from rate of return to a year of schooling; that is, 11.7% in Sub-Saharan Africa and 9.9% in Asia (Psacharopoulos and Patrinos 2004) Source: Compiled by World Bank team. 166 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL 2 years of grade equivalent from learning productivity (school achievement) for children with severe stunting, for the first one-unit improvement in z-score (row 5, table D.7) 0.42 to 0.44 years' change in entry into the labor force (row 6, table D.7) Additional parameter values are presented in table D.10. Psacharopoulos and Patrinos (2004) reviewed studies of returns to investment in education. Returns to one additional year of schooling are given by the coefficient on years of schooling. On average, the return to one additional year of schooling is 9.7 percent. The return is 11 to 12 percent in Sub-Saharan Africa and the group of low-income countries. This figure is used to estimate the percentage change in the present value of lifetime income from improved educational performance in the absence of early childhood diarrheal infections. Four educational outcomes are considered: 1. A one-year lower grade attainment caused by poor nutritional status (stunting) results in an additional year of income (leaving school early) at t = 15, loss of income in all years of working life (lower income per year) for each t 16, and a saving in cost of a year of schooling at t = 15.14 Ct = 0 for t 15. 2. Lower learning productivity equivalent to one year of schooling results in a loss of income in all years of working life for each t 15. Ct = 0 for all years.15 3. Delayed primary school enrollment results in delayed labor-force entry.A one- year delay results in lost income at t = 15. Ct = 0 for all years.16 4. Grade repetition has the same effect as delayed enrollment.17 The cost of these educational outcomes is implicit in the difference between current and counterfactual lifetime incomes in rows 7 and 8, table D.7.18 In fact, counterfactual lifetime income in each malnutrition category rises as a conse- quence of reducing infections and subsequent higher educational outcomes. In estimations of current and counterfactual lifetime income for children of different nutritional status (from not malnourished to severely malnourished), assumptions about baseline income (I0 in equation D.10) are important. Malnourished children live in disproportionately poorer households, and poverty is correlated with lower socioeconomic status, such as lower educational attain- ment of parents.This factor,in turn,is correlated with lower educational attainment of children as they grow up. Malnourished children are therefore likely to have lower income than children who are not malnourished, not only because of the effect of malnutrition on educational performance and income but also because of the socioeconomic status of the household they grew up in. Therefore, instead of using GDP per capita as a proxy for per capita income for all children, this analysis attempts to develop a matrix of income differentials in Ghana across the range of nutritional status, from not malnourished to severely malnourished. These figures represent a more accurate estimate of lifetime income of children as they move across nutritional categories. Table D.11, panel a, shows the distribution of children by nutritional status and household wealth in Ghana. COMPUTING COUNTRY-LEVEL ENVIRONMENTAL HEALTH BURDEN OF DISEASE 167 TABLE D.11 Income Distribution across Malnutrition Categories and Wealth Quintiles in Ghana Mild Moderate Severe Not (Height-for- (Height-for- (Height- for- Malnourished Age -1 to -2 Age -2 to -3 Age < -3 Wealth (Height-for- Standard Standard Standard Quintile Age > -1) Deviation) Deviation) Deviation) a. Height-for-age prevalence rates (Ghana 2003 DHS) Totalb Poorest 30.04 29.1 25.47 15.39 100 Poorer 37.68 30.85 19.12 12.36 100 Middle 40.77 29.27 20.45 9.51 100 Richer 49.49 28.12 15.84 6.55 100 Richest 65.24 20.96 8.9 4.9 100 Total 42.42 28.15 18.95 10.49 100 b. Per capita income (expenditure) shares (%)a Totalb Poorest 1.7 1.6 1.4 0.9 5.6 Poorer 3.8 3.1 1.9 1.2 10.1 Middle 6.1 4.4 3.0 1.4 14.9 Richer 11.3 6.4 3.6 1.5 22.9 Richest 30.4 9.8 4.1 2.3 46.6 Total 53.3 25.3 14.2 7.3 100.1 c. Per household income (expenditure) shares (%)a Totalb Poorest 2.8 2.7 2.4 1.4 9.4 Poorer 6.4 5.2 3.2 2.1 16.9 Middle 7.8 5.6 3.9 1.8 19.2 Richer 12.7 7.2 4.0 1.7 25.6 Richest 18.9 6.1 2.6 1.4 28.9 Total 48.6 26.8 16.2 8.5 100.0 Source: Compiled by World Bank team. a. Estimated from the height-for-age prevalence rates and total income (expenditure) shares. b. Total income (expenditure) shares by quintile in Ghana in 1998/99 from the Ghana Living Standard Survey (World Bank 2007h; Trades Union Congress 2004). Panels b and c show the income (expenditure) shares in relation to household wealth quintile and children's nutritional status. Total income (expenditure) shares for each group of children (not malnourished and mild, moderate, and severely malnourished) are then used to estimate baseline annual income (I0) for each of these four groups, which is applied in equations D.10 and D.11 to estimate life- time income (rows 7 and 8, table D.7). By combining the information on current stunting prevalence (row 1,table D.7), lifetime income (rows 7 and 8),and labor-force participation (row 9),one can obtain total current and counterfactual lifetime income per malnutrition category (rows 10 and 11) and for the whole population birth cohort (rows 12 and 13). The difference between current and counterfactual lifetime income for an annual population birth cohort as a whole represents the educational and cognitive 168 TABLE D.12 Annual Cost of Environmental Factors (Percentage of GDP in 2005), Using 3 Percent Discount Rate Ghana Pakistan Cost Cost Cost Cost Cost Cost Annual Deaths ( million) (US$ million) (% of GDP in 2005) Annual Deaths (PRs billion) (US$ million) (% of GDP) Estimates excluding malnutrition-mediated effects Mortality effects 24,712 371 412 3.84 131,611 195 3,250 2.90 Estimates including malnutrition-mediated effects Mortality effects 35,702 537 595 5.55 187,429 278 4,633 4.13 Education effects 367 407 3.79 317 5,281 4.71 Total annual costs 904 1,002 9.34 595 9,914 8.84 Source: Compiled by World Bank team. Note: = Ghanaian new cedi. TABLE D.13 Annual Cost of Environmental Factors (Percentage of GDP in 2005), Using 5 Percent Discount Rate Ghana Pakistan Cost Cost Cost Cost Cost Cost Annual Deaths ( million) (US$ million) (% of GDP in 2005) Annual Deaths (PRs billion) (US$ million) (% of GDP) Estimates excluding malnutrition-mediated effects Mortality effects 24,712 188 208 1.94 131,611 99 1,650 1.46 Estimates including malnutrition-mediated effects Mortality effects 35,702 271 300 2.80 187,429 140 2,333 2.09 Education effects 180 199 1.86 156 2,592 2.31 Total annual cost 451 499 4.66 296 4,925 4.40 Source: Compiled by World Bank team. Note: = Ghanaian new cedi. 169 170 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL cost of malnutrition from diarrheal infections. This cost is estimated at 3.8 percent of GDP in 2005. The cost attributable to environmental factors (that is, diarrheal infections from inadequate water supply, sanitation, and hygiene) is about one- third of the total cost of stunting on lifetime income. Summary of Results for Ghana and Pakistan The methodology described in this appendix has also been applied to Pakistan. The results, summarized in chapter 5, are reported here for convenience (table D.12.). In total, the cost of malnutrition-mediated effects (mortality and educa- tion effects) of environmental risk factors on children in Ghana and Pakistan is estimated at an equivalent of 9.2 percent and 8.6 percent, respectively, of GDP per year. This finding is in stark contrast to a cost of 3.8 percent and 2.9 percent, arrived at when considering only the estimates excluding malnutrition-mediated effects of environmental risks. The annualized monetary cost of mortality and impaired education is sensi- tive to the discount rate of income losses. Annual costs are presented in tables D.12 and D.13 using discount rates of 3 and 5 percent. In both Ghana and Pakistan, at an annual discount rate of 5 percent, the annual cost is half of the cost at a 3 percent discount rate. Total costs in Ghana and Pakistan are, however, nearly two and one-half and nearly three times higher, respectively, than the cost that does not include malnutrition effects, regardless of the discount rate used. Notes 1 This appendix is based on the analysis of country-specific data and surveys for Ghana. Results are arrived at using a methodology developed specifically for this report by Björn Larsen (consultant). Useful input was provided by Maureen Cropper. The analysis esti- mates costs that are attributable to environmental risks in children under five years of age for Ghana and Pakistan and for the first time includes the malnutrition-mediated effects, as well as the longer-term impacts on cognition, school performance, and future work productivity. 2 See appendix E for more detail on impact fractions or attributable fractions. 3 The use of the Cox hazard model has a drawback if one wants to calculate the attribut- able fractions of individual risks (as opposed to calculating joint attributable fractions, as is done here). In fact, in such model specifications, the effect of each risk factor is augmented by the multiplicative interaction with other risk factors. The result is that the fractions attributable to individual risks may add to more than 1. If one imposes additive relative risks (so that the sum of the attributable fraction equals 1), the results do not differ much from those presented here. 4 The estimate of the impact of a single risk factor on a disease (for example, the impact of malnutrition on deaths caused by ALRI) is often biased because other risk factors were not controlled for in the epidemiological study. So if the relative risk of ALRI death attrib- utable to malnutrition is estimated ignoring indoor air pollution, MAL will pick up its effects, because malnutrition and exposure to indoor air pollution are correlated in most populations. Appendix E describes a method to deal with biased relative risks. COMPUTING COUNTRY-LEVEL ENVIRONMENTAL HEALTH BURDEN OF DISEASE 171 5 This framework for calculating relative risks can be expanded to a larger set of exposure categories that reflect different risk levels associated with ENV and MAL. For instance, indoor air pollution exposure categories can be households using fuelwood, charcoal, unimproved rather than improved stoves, and liquefied petroleum gas. Malnutrition can be mild, moderate, and severe. 6 For the alternative approaches, please see appendix E. 7 Although the research literature finds that the retardation factor often is in the range of 0.2 to 0.5, there is limited research on how the retardation factor may vary with z-score and age. It is postulated here that the factor declines with age and is greater the more underweight the children are. Diarrheal infections are most frequent during the ages of 6 to 11months and 12 to 23 months. It may therefore be expected that weight gain retarda- tion from diarrheal infections is greatest in this period and, at least in Ghana and Pakistan, is particularly great up to the age of 12 months, during which underweight increases rapidly. The retardation factors in table D.4 for WAZ = ­2 to ­4 are within the bounds often found in the research literature. 8 The calculation is as follows: (11.67 ­ 10)/(14.6 ­ 10) = 36%. 9 The calculation is as follows: ­3 + (3 * 0.4) = ­1.8. 10 A working life from 15 to 65 years is applied. Annual income is approximated by GDP per capita. Income is discounted at 3 percent. Real income growth is 2 percent per year. 11 The methodology for this part of the analysis, which estimates effects of stunting (partly attributed to repeated infections) on cognition and learning, has benefited from comments by Harold Alderman (senior adviser, Africa Region Human Development Department, World Bank). 12 On-time enrollment, grade attainment, and high school completion rates in Cebu are rela- tively high: 94 percent on-time (less than 8 years) enrollment, 9.5 years of schooling, and 55 percent high school completion rate. 13 An average child in Ghana starts school at the age of about seven and has seven to eight years of schooling (World Bank 2007h). The average age of leaving school is therefore around 15 years. 14 Cost of schooling is the private and public costs of schooling. 15 Notice that there is a possibility of underestimation of damages linked to impaired cogni- tive development. In fact, income earners with impaired cognitive development are more likely to have lower income than others with the same educational achievement. The nega- tive wage effect for income earners with impaired cognitive development as a consequence of early childhood malnutrition may therefore be larger than that reflected in the studies reviewed by Psacharopoulos and Patrinos (2004), if those studies did not control for cogni- tive functioning. 16 Delayed school enrollment may affect grade attainment and thus schooling cost. However, including such an effect here would involve double counting because grade attainment is dealt with separately. 17 Increased grade repetition as a consequence of poor nutrition status is equivalent to an average of one month of a school year per child with severe stunting status and one week per child with mild stunting status. The additional schooling cost of grade repetition is there- fore ignored here because it is minimal compared with the costs related to grade attainment. 18 These amounts are estimated from the Ghana 2003 DHS. A P P E N D I X E Methodological Aspects of Assessing Environmental Health Burden of Disease THIS APPENDIX CONSISTS OF TWO SECTIONS that provide details on certain methodological aspects of the analysis to estimate the environmental health burden of disease.1 The first section relates to how one gets from relative risks (RRs) to attributable fractions (AFs). Environmental risk factor AFs for Ghana and Pakistan are discussed in chapter 5, and a detailed methodology is described in appendix D. The second section discusses a possible way to get around the biased estimates of RRs. Unbiased estimates of RRs are used to calculate the AFs in table 5.5. From Relative Risks to Attributable Fractions The burden of disease methodology aims at estimating the effects of selected risk factors on mortality and disease. It applies disability weights to combine deaths and illness into a single metric: disability-adjusted life years (DALYs). The envi- ronmental burden of disease is obtained by limiting the analysis to environmental risk factors.2 The World Health Organization's Comparative Risk Assessment (WHO 2002) considers six environmental risk factors and puts them in relation to the diseases identified in the table E.1. For each pair of risk factors and diseases, the Comparative Risk Assessment defines a relative risk--that is, the probability of a disease or death from that disease (for example, diarrhea) occurring in a group exposed to a risk factor 173 174 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL TABLE E.1 Environmental Risk Factors and Related Diseases Included in the WHO Comparative Risk Assessment Risk Factor Related Diseases Outdoor air pollution Respiratory infections, selected cardiopulmonary diseases, lung cancer Indoor air pollution Chronic obstructive pulmonary disease, lower respiratory infections, lung cancer Lead Mild mental retardation, cardiovascular diseases Water, sanitation, Diarrheal diseases, trachoma, schistosomiasis, ascariasis, and hygiene trichuriasis, hookworm disease Climate change Diarrheal diseases, malaria, selected unintentional injuries, protein-energy malnutrition Selected occupational risk factors: Injuries Unintentional injuries Noise Hearing loss Carcinogens Cancers Airborne particulates Asthma, chronic obstructive pulmonary disease Ergonomic stressors Lower back pain Source: WHO 2002. (for example, inadequate water, sanitation, and hygiene) versus the probability of the same event occurring in a group that was not exposed. Pr Disease RR = ( Exposed ) (E.1) Pr Disease ( Nonexposed ) So, for example, a relative risk of 2.3 of mortality from ALRI from indoor air pollution (IAP) means that the probability of a person contracting ALRI after being exposed is 2.3 times higher than the probability of the same person contracting ALRI if not exposed. RRs, which are typically obtained from epidemi- ology studies and expert forums, are reported for both mortality and morbidity and constitute the basic building block of the estimation of burden of disease. Information on exposure to a risk factor (for example, the number of children under five years of age exposed to IAP) is combined with the RRs to estimate the risk AF of death or illness. For example, suppose one wants to attribute deaths from acute respiratory infections (ARI) in children under five to exposure to IAP. The standard definition of the fraction of ARI deaths attributable to IAP is AF = (Excess deaths attributable to IAP)/(Total ARI deaths) (E.2) Excess deaths attributable to IAP = (Deaths in exposed population based on RR) ­ (Deaths in exposed population if no exposure) (E.3) METHODOLOGICAL ASPECTS OF ASSESSING ENVIRONMENTAL HEALTH BURDEN OF DISEASE 175 The fraction of deaths attributable to IAP is the ratio of excess deaths caused by that pollution to total ARI deaths. Excess deaths attributable to IAP is the difference between deaths in the exposed population based on estimated RR, minus deaths in the exposed population if there were no exposure--that is, if the RR were 1. The AF defined in equations E.2 and E.3 can be written as follows. Let N equal the population of interest (for example, children under five in Ghana) and P be the proportion of the population exposed to the risk factor (for example, IAP). Let m be the death rate from ARI in the unexposed population; RR*m is the death rate from ARI in the exposed population. Then, from equation E.3, Excess deaths attributable to IAP = NP(RR*m) ­ NPm = NP(RR ­ 1)m (E.4) Total deaths attributable to ARI = NP(RR*m) + N(1 ­ P)*m = NP(RR ­ 1)m + Nm Dividing the right-hand side of equation E.4 by total deaths gives AF from IAP = [NP(RR ­ 1)m]/ [NP(RR ­ 1)m + Nm] (E.5) Canceling N and m from the numerator and denominator yields the standard formula: AF from IAP = P(RR ­ 1)/[P(RR ­ 1) + 1] (E.6) Equation E.6 generalizes when there is more than one exposure category (for example, exposure to indoor smoke in an unventilated kitchen and exposure to indoor smoke in ventilated kitchen) to PRR n i i -1 AF = i=1 , n PRR i i (E.7) i=1 where Pi is the population exposed to level i of the risk factor and RRi is the rela- tive risk of dying from the selected disease by being exposed to level i of the risk factor. It is important to note that equation E.7 also applies when there are multiple risk factors and when exposure categories represent different combinations of exposure to risk factors. A generalized version of equation E.7 is PRR P n n i i - iRRi AF = i=1 i=1 , (E.8) n PRR i i i=1 176 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL where the subindex i denotes exposure category; Pi is the population's current expo- sure distribution; Piis the population's exposure distribution after controlling one or more risk factors (that is, exposure distribution in the absence of environ- mental risks); and RRi is the risk of mortality in each exposure category. A major advantage of equation E.8 is that it allows estimation of the AF of a target risk factor (such as IAP) while leaving other risk factors (such as malnutrition that is not environmentally attributable) unchanged. This advantage is exploited in esti- mating the burden of disease for Ghana and Pakistan in this report. Dealing with Biased Estimates of Relative Risk There is a second problem relating to this study in estimating both the direct and indirect effects of environmental risk factors. The RRs of mortality from expo- sure to IAP and from malnutrition are usually calculated in separate epidemiological studies, in which the other risk factor is not controlled for. For example, most of the studies of ALRI that is caused by IAP that are reviewed in Desai, Mehta, and Smith (2004) do not control for nutritional status. In the study by Fishman and others (2004) of the impact of malnutrition on ALRI, there is no control for IAP. In econometric jargon,this problem is referred to as the omitted variable bias problem. If there is positive correlation between malnutrition and exposure to IAP, the RR in each of the studies is likely to overstate the impact of the risk factor studied, relative to a study in which both risk factors are included. This problem is diffi- cult to correct; a second-best approach is suggested here, which is based on the properties of least-squares estimators in the presence of omitted variables. The estimates in Fishman and others (2004) of the effect of malnutrition on mortality from various diseases are very likely biased because they fail to control for other risk factors.3 If the estimate of the relative risk of death from ALRI attrib- utable to malnutrition ignores IAP, malnutrition will pick up its effects, because malnutrition and exposure to IAP are correlated in most populations. For policy purposes, there are at least four risk factors whose impact on ALRI deaths might need to be estimated: exposure to IAP,4 malnutrition, crowding, and poor medical treatment received when ill. These factors are all important, and all have policy significance. Because poor children are likely to be at risk for all four factors, omitting three of them makes it likely that malnutrition will pick up their effects, thus overstating the impact of malnutrition. In most studies, one could reduce this bias by using family income (or assets) or mother's education as a proxy for the omitted factors, even if the factors could not be estimated directly. The only way to solve this omitted vari- able bias problem is to find studies that have controlled for other risk factors (or have proxied for them) in addition to malnutrition. METHODOLOGICAL ASPECTS OF ASSESSING ENVIRONMENTAL HEALTH BURDEN OF DISEASE 177 To understand the nature of the bias, suppose that in the correct model y = g 0 + g 1IAP + g 2MAL, (E.9) where y = death rate attributable to ALRI or the logarithm of the death rate. It is well known that if malnutrition (MAL) is dropped from this equation and y = b0 + b1IAP (E.10) is estimated instead, that the least-squares estimator of b1 --that is, ^1--is related to the least-squares estimators of the coefficients in equation E.9 as follows: ^1 = ^1 + ^2 ^ , (E.11) where ^ is the coefficient of IAP in a regression of MAL on IAP. This equation implies that if ^2>0 and ^ >0, ^1 will be biased upward: indoor air pollution will capture the effects of malnutrition. This situation is likely in practice, since indoor air pollution and malnutrition are positively correlated in most populations. A similar argument can be made for a model in which IAP is dropped from equation E.9. In such a case, the coefficient for MAL will be biased as it captures the effect of IAP. In the calculation of the total cost of environmental health risk in chapter 5,relative risks are corrected by using these econometric results.Although this solution is second best, it is done in order to provide conservative estimates. Notes 1 This appendix is based on detailed comments and guidance from Maureen Cropper. 2 To distinguish environmental risk factors from nonenvironmental risk factors, one has to define environment. This report follows Prüss-Üstün and Corvalán (2006: 22) by defining environment as "all the physical, chemical, and biological factors external to a person, and all the related behaviors." 3 Pelletier (1994) has demonstrated that estimates are robust when a wide array of variables is controlled for. 4 Notice that IAP could also contribute to malnutrition through repeated respiratory infec- tions. Multicollinearity will cause the estimated effect to be less robust (although unbiased). A P P E N D I X F Monetary Valuation of the Cost of Environmental Health Risks CHAPTER 5 PROVIDES MONETARY ESTIMATES of the environmentally attributable burden of disease in Ghana and Pakistan. The World Health Organization's methodology for estimating burden of disease is a powerful tool for decision-makers. Knowing the burden of disease makes it possible to compare policy options that act on different causes of death and have different mortality and morbidity impacts. A major advantage of using burden of disease is that it requires no monetary valuation of the benefits of an intervention (for example, there is no need to know the value of a life). The main limitation of burden of disease is that it cannot be applied when one compares policies with outcomes measured in different metrics (for example, a case in which one policy reduces diarrhea mortality while another policy increases children's cognition and, thus, future income). In such cases, the monetary valuation of burden of disease can be very useful. Monetary valuation has also been used extensively to express burden of disease in a metric that is easy for policy-makers with no background in public health to understand. Several approaches are available to value the impact of health risks. Traditionally, valuation of health damages has been based on the calculation of the observed costs caused by death or illness. Death, incapacity, or illness is typically valued using the human capital approach, which calculates the present value of forgone earnings.1 For the sake of simplicity, 179 180 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL many studies of the costs of environmental degradation have valued a lost year of healthy life using gross domestic product per capita as a proxy of the forgone annual earnings. Illness also results in expenses other than lost wages. The cost-of-illness approach measures such costs by estimating the change in costs incurred as a result of a change in the incidence of a particular illness. Both direct costs (such as cost of doctor visits and treatment) and indirect costs (such as loss of wages) are included in the estimation. In cases where some of the costs are borne by medical insurance, cost-of-illness measures will not be limited to a patient's out-of- pocket expenses but should include the additional costs borne by the insurance company or treatment facility, to capture the social benefits of the reduced risk. It is important to keep in mind that unit values should be expressed net of taxes and subsidies. Taxes and subsidies are transfers within the economy and should not be taken as a component of the true economic value of an item. So, for example, admissions in public hospitals that are free (or subsidized) can be valued by using a comparable estimate from the private sector. Likewise, meas- ures of wage are taken net of labor taxes. More recently, economists have started adopting estimates of willingness to pay (WTP) or willingness to accept compensation to value health damages. The human capital and cost-of-illness approaches take into account only the out-of- pocket expenses paid as a consequence of illness or death. They are based on the notion that individuals should at least be willing to pay those out-of-pocket expenses to avoid illness. But individuals' willingness to pay to avoid illness or death may be higher than their out-of-pocket expenses. The WTP approach, when correctly applied, can capture people's preferences to avoid pain and discomfort. It does so by looking at how they respond to risk of illness or death. Typically, valuation is determined either by observing behavior or by eliciting responses from a questionnaire. Responses are then analyzed using econometric techniques to estimate willingness to pay. Despite its advantages, the WTP approach produces a much higher uncer- tainty regarding values and requires much closer scrutiny. The WTP approach is also data intensive: proper estimates require large and costly surveys and careful application of econometric methods. For this reason, studies in developing coun- tries have commonly transferred WTP estimates from U.S. and European studies by accounting for differences in purchasing power and, occasionally, the income elasticity of consumption. An important limit to the application of the WTP approach is with regard to children's life. It is conceptually and practically very difficult to observe or measure monetarily children's preferences. For this reason, in this study, valuation relies solely on the human capital approach. For simplicity, the study does not look at cost of illness and it concentrates on the valuation of mortality. Deaths in Ghana MONETARY VALUATION OF THE COST OF ENVIRONMENTAL HEALTH RISKS 181 are valued using a per capita income of US$485 (2005 dollars). In Pakistan, a per capita income of US$714 (2005 dollars) is used. Per capita income values are also used to estimate the present value of income losses from lower education attain- ment, together with an assumed real rate of income growth of 2 percent. Note 1 Some have argued that there is a conceptual problem with the human capital approach, in that most people value safety not out of concern for preserving current and future income levels but primarily because they have an aversion to pain, suffering, and death. References Adair, L., B. Popkin, J.VanDerslice, J. Akin, D. Guilkey, R. Black, J. Briscoe, and W. Flieger. 1993. "Growth Dynamics during the First Two Years of Life: A Prospective Study in the Philippines." European Journal of Clinical Nutrition 47 (1): 42­51. Adjuik, M., T. Smith, S. Clark, J. Todd, A. Garrib, Y. Kinfu, K. Kahn, M. Mola, A. Ashraf, H. Masanja, U. Adazu, J. Sacarlal, N. Alam, A. Marra, A. Gbangou, E. Mwageni, and F. Binka. 2006. "Cause-Specific Mortality Rates in Sub-Saharan Africa and Bangladesh." Bulletin of the World Health Organization 84 (3): 181­88. Agarwal, B. 1983. "Diffusion of Rural Innovations: Some Analytical Issues and Case Study of Improved Woodburning Stoves." World Development 11: 359­76. ------. 1986. Cold Hearths and Barren Slopes: Woodfuel Crisis in the Third World. New Delhi: Allied Press. Ahmed, K.,Y. A. Awe, D. F. Barnes, M. L. Cropper, and M. Kojima. 2005. Environmental Health and Traditional Fuel Use in Guatemala. Washington, DC: World Bank. Ahmed, N., M. Zeitlin, A. Beiser, C. Super, and S. Gershoff. 1993."A Longitudinal Study of the Impact of Behavioural Change Intervention on Cleanliness, Diarrhoeal Morbidity, and Growth of Children in Rural Bangladesh." Social Science and Medicine 37 (2): 159­71. Alam, D., G. Marks, A. Baqui, M. Yunus, and G. Fuchs. 2000. "Association between Clinical Type of Diarrhoea and Growth of Children under 5 Years in Rural Bangladesh."International Journal of Epidemiology 29 (5): 916­21. Alderman, H., J. Behrman,V. Lavy, and R. Menon. 2001."Child Health and School Enrollment: A Longitudinal Analysis." Journal of Human Resources 36 (1): 185­205. 183 184 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Alderman, H., J. Konde-Lule, I. Sebuliba, D. Bundy, and A. Hall. 2006."Effect on Weight Gain of Routinely Giving Albendazole to Preschool Children during Child Health Days in Uganda: Cluster Randomised Controlled Trial." British Medical Journal 333 (7559): 122. Allen, S. J., A. Raiko, A. O'Donnell, N. D. E. Alexander, and J. B. Clegg. 1998."Causes of Preterm Delivery and Intrauterine Growth Retardation in a Malaria Endemic Region of Papua New Guinea." Archive of Disease in Childhood 79 (2): F135­40. Alvarado, B.-E., M. V. Zunzunegui, H. Delisle, and J. Osorno. 2005. "Growth Trajectories Are Influenced by Breast-Feeding and Infant Health in an Afro-Colombian Community." Journal of Nutrition 135 (9): 2171­78. Awasthi, S., and D. Bundy. 2007. "Intestinal Nematode Infection and Anaemia in Developing Countries." British Medical Journal 334 (7603): 1065­66. Awasthi, S., D. Bundy, and L. Savioli. 2003. "Helminthic Infections." British Medical Journal 327 (7412): 431­33. Bairagi, R., M. K. Chowdhury, Y. J. Kim, G. T. Curlin, and R. H. Gray. 1987. "The Association between Malnutrition and Diarrhoea in Rural Bangladesh." International Journal of Epidemiology 16 (3): 477­81. Banerjee, I., S. Ramani, B. Primrose, P. Moses, M. Iturriza-Gomara, J. J. Gray, S. Jaffar, M. Bindhu, J. P. Muliyil, D. W. Brown, M. K. Estes, and G. Kang. 2006."Comparative Study of the Epidemiology of Rotavirus in Children from a Community-Based Birth Cohort and a Hospital in South India." Journal of Clinical Microbiology 44 (7): 2468­74. Bang, A. T., R. A. Bang, S. B. Baitule, M. H. Reddy, and M. D. Deshmukh. 1999. "Effect of Home-Based Neonatal Care and Management of Sepsis on Neonatal Mortality: Field Trial in Rural India." Lancet 354 (9194): 1955­61. Baumgartner, R., and E. Pollitt. 1983. "The Bacon Chow Study: Analyses of the Effects of Infectious Illness on Growth of Infants." Nutrition Research 3: 9­21. Becker, S., R. E. Black, and K. H. Brown. 1991."Relative Effects of Diarrhea, Fever, and Dietary Energy Intake on Weight Gain in Rural Bangladeshi Children."American Journal of Clinical Nutrition 53 (6): 1499­503. Behrman, J., J. Hoddinott, J. Maluccio, E. Soler-Hampejsek, E. L. Behrman, R. Martorell, M. Ramirez-Zea, and A D. Stein. 2006."What Determines Adult Cognitive Skills? Impacts of Pre-school, Schooling, and Post-schooling Experiences in Guatemala." PSC Working Paper, Population Studies Center, University of Pennsylvania, Philadelphia. Bennett, J., M. Schooley, H. Traverso, S. B. Agha, and J. Boring. 1996. "Bundling, a Newly Identified Risk Factor for Neonatal Tetanus: Implications for Global Control."International Journal of Epidemiology 25 (4): 879­84. Berg, R. 2005. "Bridging the Great Divide: Environmental Health and the Environmental Movement." Journal of Environmental Health 67 (6): 39­52. Berkman, D. S., A. G. Lescano, R. H. Gilman, S. L. Lopez, and M. M. Black. 2002. "Effects of Stunting, Diarrhoeal Disease, and Parasitic Infection during Infancy on Cognition in Late Childhood: A Follow-up Study." Lancet 359 (9306): 564­71. Bertozzi, S., N. Padian, J. Wegbreit, L. DeMaria, B. Feldman, H. Gayle, J. Gold, R. Grant, and M. Isbell. 2006. "HIV/AIDS Prevention and Treatment." In Disease Control Priorities in Developing Countries, 2nd ed., ed. D. T. Jamison, J. G. Breman, A. R. Measham, G. Alleyne, M. Claeson, D. B. Evans, P. Jha, A. Mills, and P. Musgrove, 331­69. Washington, DC: World Bank; New York: Oxford University Press. Biran, A., L. Smith, J. Lines, J. Ensink, and M. Cameron. 2007. "Smoke and Malaria: Are Interventions to Reduce Exposure to Indoor Air Pollution Likely to Increase Exposure REFERENCES 185 to Mosquitoes?" Transactions of the Royal Society of Tropical Medicine and Hygiene 101 (11): 1065­71. Black, R. E., K. H. Brown, and S. Becker. 1984. "Effects of Diarrhea Associated with Specific Enteropathogens on the Growth of Children in Rural Bangladesh."Pediatrics 73 (6): 799­805. Blössner, M., and M. de Onis. 2005. Malnutrition: Quantifying the Health Impact at National and Local Levels. WHO Environmental Burden of Disease 12. Geneva: World Health Organization. Boeston, K., P. Kolsky, and C. Hunt. 2007. "Improving Urban Water and Sanitation Services: Health, Access, and Boundaries." In Scaling Urban Environmental Challenges: From Local to Global and Back, ed. P. Marcotullio and G. McGranahan, 106­31. London: Earthscan. Breman, J. G., M. S. Alilio, and A. Mills. 2004."Conquering the Intolerable Burden of Malaria: What's New, What's Needed--A Summary." American Journal of Tropical Medical Hygiene 71 (2 Suppl.): 1­15. Brenzel, L., L. J. Wolfson, J. Fox-Rushby, M. Miller, and N. A. Halsey. 2006."Vaccine-Preventable Diseases." In Disease Control Priorities in Developing Countries, 2nd ed., ed. D. T. Jamison, J. G. Breman, A. R. Measham, G. Alleyne, M. Claeson, D. B. Evans, P. Jha, A. Mills, and P. Musgrove, 389­411. Washington, DC: World Bank; New York: Oxford University Press. Briend, A., K. Hasan, K. Aziz, and B. Hoque. 1989."Are Diarrhoea Control Programmes Likely to Reduce Childhood Malnutrition? Observations from Rural Bangladesh."Lancet 2 (8658): 319­22. Briscoe, J. 1987. "The Author Replies." American Journal of Epidemiology 125 (5): 922­25. Brown, K. H. 2003."Symposium: Nutrition and Infection, Prologue and Progress since 1968-- Diarrhea and Malnutrition." Journal of Nutrition 133 (1): 328S­32S. Bruce, N., R. Perez-Padilla, and R. Albalak. 2002. "The Health Effects of Indoor Air Pollution Exposure in Developing Countries."WHO/SDE/OEH/02.05. World Health Organization, Geneva. http://whqlibdoc.who.int/hq/2002/WHO_SDE_OEH_02.05.pdf. Brush,G.,G.Harrison,and J.Waterlow.1997."Effects of Early Disease on Later Growth,and Early Growth on Later Disease, in Khartoum Infants." Annals of Human Biology 24 (3): 187­95. Bryce, J., C. Boschi-Pinto, K. Shibuya, and R. Black. 2005. "WHO Estimates of the Causes of Death in Children." Lancet 365 (9465): 1147­52. Burra, S., S. Patel, and T. Kerr. 2003."Community Designed, Built and Managed Toilet Blocks in Indian Cities." Environment and Urbanization 15 (2): 11­32. Cairncross, S. 1987. "Ingested Dose and Diarrhea Transmission Routes." American Journal of Epidemiology 125 (5): 921­22. ------. 2003. "Handwashing with Soap: A New Way to Prevent ARIs." Tropical Medicine and International Health 8 (8): 677­79. Cairncross, S., U. Blumenthal, P. Kolsky, L. Moraes, and A. T. Tayeh. 1996. "The Public and Domestic Domains in the Transmission of Disease." Tropical Medicine and International Health 1 (1): 27­34. Cairncross, S., and J. Kinnear. 1992. "Elasticity of Demand for Water in Khartoum, Sudan." Social Science and Medicine 34 (2): 183­89. Cairncross, S., and P. Kolsky. 1997. "Water, Waste, and Well-Being: A Multicountry Study." American Journal of Epidemiology 146 (4): 359­61. Cairncross, S., and V.Valdmanis. 2006."Water Supply, Sanitation, and Hygiene Promotion." In Disease Control Priorities in Developing Countries, 2nd. ed., ed. D. T. Jamison, J. G. Breman, 186 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL A. R. Measham, G. Alleyne, M. Claeson, D. B. Evans, P. Jha, A. Mills, and P. Musgrove, 771­92. Washington, DC: World Bank; New York: Oxford University Press. Campbell, D., M. Elia, and G. Lunn. 2003. "Growth Faltering in Rural Gambian Infants Is Associated with Impaired Small Intestinal Barrier Function, Leading to Endotoxemia and Systemic Infection." Journal of Nutrition 133 (5): 1332­38. Campbell-Lendrum, D., C. Corvalán, and M. Neira. 2007."Global Climate Change: Implications for International Public Health Policy." Bulletin of the World Health Organization 85 (3): 235­37. Carson, Rachel. 1962. Silent Spring. New York: Houghton Mifflin. Carter, R., and K. Mendis. 2006. "Measuring Malaria." American Journal of Tropical Medical Hygiene 74 (2): 187­88. Cattand, P., P. Desjeux, M. Guzman, J. Jannin, A. Kroeger, A. Medici, P. Musgrove, M. B. Nathan, A. Shaw, and C. Schofield. 2006. "Tropical Diseases Lacking Adequate Control Measures: Dengue, Leishmaniasis, and African Trypanosomiasis." In Disease Control Priorities in Developing Countries, 2nd ed., ed. D. T. Jamison, J. G. Breman, A. R. Measham, G. Alleyne, M. Claeson, D. B. Evans, P. Jha, A. Mills, and P. Musgrove, 451­66. Washington, DC: World Bank; New York: Oxford University Press. Caulfield, L. E., M. de Onis, M. Blössner, and R. E. Black. 2004. "Undernutrition as an Underlying Cause of Child Deaths Associated with Diarrhea, Pneumonia, Malaria, and Measles." American Journal of Clinical Nutrition 80 (1): 193­98. Caulfield, L. E., S. A. Richard, J. A. Rivera, P. Musgrove, and R. E. Black. 2006."Stunting, Wasting, and Micronutrient Deficiency Disorders." In Disease Control Priorities in Developing Countries, 2nd ed., ed. D. T. Jamison, J. G. Breman, A. R. Measham, G. Alleyne, M. Claeson, D. B. Evans, P. Jha, A. Mills, and P. Musgrove, 551­68. Washington, DC: World Bank; New York: Oxford University Press. Chatterjee, N., and P. Hartge. 2003."Apportioning Causes, Targeting Populations, and Predicting Risks: Population Attributable Fractions."European Journal of Epidemiology 18 (10): 933­35. Checkley, W., L. D. Epstein, R. H. Gilman, R. E. Black, L. Cabrera, and C. R. Sterling. 1998. "Effects of Cryptosporidium parvum Infection in Peruvian Children: Growth Faltering and Subsequent Catch-up Growth." American Journal of Epidemiology 148 (5): 497­506. Checkley, W., L. D. Epstein, R. H. Gilman, L. Cabrera, and R. E. Black. 2003. "Effects of Acute Diarrhea on Linear Growth in Peruvian Children." American Journal of Epidemiology 157 (2): 166­75. Checkley, W., R. H. Gilman, L. D. Epstein, M. Suarez, J. F. Diaz, L. Cabrera, R. E. Black, and C. R. Sterling. 1997."Asymptomatic and Symptomatic Cryptosporidiosis: Their Acute Effect on Weight Gain in Peruvian Children." American Journal of Epidemiology 145 (2): 156­63. Chisti, M. J., M. I. Hossain, M. A. Malek, A. S. Faruque, T. Ahmed, and M. A. Salam. 2007. "Characteristics of Severely Malnourished Under-Five Children Hospitalized with Diarrhoea and Their Policy Implications." Acta Paediatrica 96 (5): 693­96. Chu, D., R. D. Bungiro, M. Ibanez, L. M. Harrison, E. Campodonico, B. F. Jones, J. Mieszczanek, P. Kuzmic, and M. Cappello. 2004. "Molecular Characterization of Ancylostoma ceylan- icum Kunitz-Type Serine Protease Inhibitor: Evidence for a Role in Hookworm-Associated Growth Delay." Infection and Immunity 72 (4): 2214­21. Cohen, R. J., K. H. Brown, J. Canahuati, L. L. Rivera, and K. G. Dewey. 1995."Determinants of Growth from Birth to 12 Months among Breast-Fed Honduran Infants in Relation to Age of Introduction of Complementary Foods." Pediatrics 96 (3): 504­10. REFERENCES 187 Condon-Paoloni, D., J. Cravioto, F. Johnston, E. De Licardie, and T. Scholl. 1977. "Morbidity and Growth of Infants and Young Children in a Rural Mexican Village." American Journal of Public Health 67 (7): 651­56. Coovadia, H. M., N. C. Rollins, R. M. Bland, K. Little, A. Coutsoudis, M. L. Bennish, and M.-L. Newell. 2007."Mother-to-Child Transmission of HIV-1 Infection during Exclusive Breastfeeding in the First 6 Months of Life: An Intervention Cohort Study." Lancet 369 (9567): 1107­16. Crawley, J., J. Hill, J. Yartey, M. Robalo, A. Serufilira, A. Ba-Nguz, E. Roman, A. Palmer, K. Asamoa, and R. Steketee. 2007."From Evidence to Action? Challenges to Policy Change and Programme Delivery for Malaria in Pregnancy."Lancet Infectious Diseases 7 (2): 145­55. Cutler, D. M., and G. Miller. 2005. "The Role of Public Health Improvements in Health Advances: The Twentieth-Century United States." Demography 42 (1): 1­22. Cutts, F., S. Zaman, G. Enwere, S. Jaffar, O. Levine, J. Okoko, C. Oluwalana, A. Vaughan, S. Obaro, A. Leach, K. McAdam, E. Biney, M. Saaka, U. Onwuchekwa, F. Yallop, N. Pierce, B. Greenwood, R. Adegbola, and the Gambian Pneumococcal Vaccine Trial Group. 2005. "Efficacy of Nine-Valent Pneumococcal Conjugate Vaccine against Pneumonia and Invasive Pneumococcal Disease in The Gambia: Randomised, Double-Blind, Placebo-Controlled Trial." Lancet 365 (9465): 1139­46. Dai, D., and W. Walker. 1999."Protective Nutrients and Bacterial Colonization in the Immature Human Gut." Advances in Pediatrics 46: 353­82. Dandoy, S. 1997."The State Public Health Department." In Principles of Public Health Practice, ed. F. D. Scutchfield and C. W. Keck, 83. Albany: Delmar. Daniels, M. C., and L. S. Adair. 2004. "Growth in Young Filipino Children Predicts Schooling Trajectories through High School." Journal of Nutrition 134 (6): 1439­46. Das Gupta, M., D. Ghosh, K. Datta, and S. Chakrabarti. 2006."Harnessing Local Governments and Communities to Improve Environmental Sanitation and Public Health in India."Paper presented at the Workshop on Fiscal Decentralization and Local Governance in India, organized by the Indian Institute of Public Administration in partnership with the govern- ment of Japan and the World Bank, Goa, India, December 11­15, 2006. Dasgupta, S., M. Huq, M. Khaliquzzaman, K. Pandey, and D. Wheeler. 2004a. Indoor Air Quality for Poor Families: New Evidence from Bangladesh. Washington, DC: World Bank. ------. 2004b. Who Suffers from Indoor Air Pollution? Evidence from Bangladesh. Washington, DC: World Bank. Desai, M., S. Mehta, and K. Smith. 2004. Indoor Smoke from Solid Fuels: Assessing the Environmental Burden of Disease at National and Local Levels. Environmental Burden of Disease Series 4. Geneva: World Health Organization. Dickson, R., S.Awasthi, P.Williamson, C. Demellweek, and P. Garner. 2000."Effects of Treatment for Intestinal Helminth Infection on Growth and Cognitive Performance in Children: Systematic Review of Randomised Trials." British Medical Journal 320 (7251): 1697­701. Eisenberg, J. N. S., W. Cevallos, K. Ponce, K. Levy, S. J. Bates, J. C. Scott, A. Hubbard, N. Vieira, P. Endara, M. Espinel, G. Trueba, L. W. Riley, and J. Trostle. 2006."Environmental Change and Infectious Disease: How New Roads Affect the Transmission of Diarrheal Pathogens in Rural Ecuador." Proceedings of the National Academy of Sciences 103 (51): 19460­65. Eisenberg, J. N. S., J. C. Scott, and T. Porco. 2007. "Integrating Disease Control Strategies: Balancing Water Sanitation and Hygiene Interventions to Reduce Diarrheal Disease Burden." American Journal of Public Health 97 (5): 846­52. 188 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Eriksson,J.2005."The Fetal Origins Hypothesis: 10Years On."British Medical Journal 330: 1096­97. Ezzati, M., A. D. Lopez, A. Rodgers, and C. J. L. Murray, ed. 2004. Comparative Quantification of Health Risks: Global and Regional Burden of Disease Attributable to Selected Major Risk Factors. Geneva: World Health Organization. Ezzati, M., A. Rodgers, A. D. Lopez, S.Vander Hoorn, and C. J. L. Murray. 2004."Mortality and Burden of Disease Attributable to Individual Risk Factors." In Comparative Quantification of Health Risks: Global and Regional Burden of Disease Attributable to Selected Major Risk Factors, vol. 2, ed. M. Ezzati, A. D. Lopez, A. Rodgers, and C. J. L. Murray, 2141­65. Geneva: World Health Organization. Ezzati, M., S.Vander Hoorn, A. Rodgers, A. D. Lopez, C. D. Mathers, and C. J. L. Murray. 2004. "Potential Health Gains from Reducing Multiple Risk Factors."In Comparative Quantification of Health Risks: Global and Regional Burden of Disease Attributable to Selected Major Risk Factors, vol. 2, ed. M. Ezzati, A. D. Lopez, A. Rodgers, and C. J. L. Murray, 2167­90. Geneva: World Health Organization. Ezzati, M., S. Vander Hoorn, A. Rodgers, A. D. Lopez, C. D. Mathers, C. J. L. Murray, and the Comparative Risk Assessment Collaborating Group. 2003. "Estimates of Global and Regional Potential Health Gains from Reducing Multiple Major Risk Factors." Lancet 362 (9380): 271­80. FAO (Food and Agriculture Organization). 2006. The State of Food Insecurity in the World 2006: Eradicating World Hunger--Taking Stock Ten Years after the World Food Summit. Rome: FAO. Favin, M. 2004."Promoting Hygiene Behavior Change within C-IMCI: The Peru and Nicaragua Experience."Activity Report 143, Environmental Health Project, Office of Health, Infectious Diseases,and Nutrition,Bureau for Global Health,U.S.Agency for International Development, Washington, DC. http://www.ehproject.org/PDF/Activity_Reports/AR-143CIMCI%20- format.pdf. Favin, M., M. Yacoob, and D. Bendahmane. 1999. "Behavior First: A Minimum Package of Environmental Health Behaviors to Improve Child Health." Applied Study 10, Environmental Health Project, Office of Health, Infectious Diseases, and Nutrition, Bureau for Global Programs, U.S. Agency for International Development, Washington, DC. http://pdf.usaid.gov/pdf_docs/PNACF961.pdf. Federal Bureau of Statistics. 2005. Pakistan Social and Living Standards Measurement Survey 2004­05. Islamabad: Federal Bureau of Statistics, Government of Pakistan. Fegan, G. W., A. M. Noor, W. S. Akhwale, S. Cousens, and R. W. Snow. 2007."Effect of Expanded Insecticide-Treated Bednet Coverage on Child Survival in Rural Kenya: A Longitudinal Study." Lancet 370 (9592): 1035­39. Ferrie, J., and W. Troesken. 2004."Death in the City: Mortality and Access to Public Water and Sewer in Chicago, 1880." Northwestern University, Evanston, IL. Fewtrell, L., A. Prüss-Üstün, R. Bos, F. Gore, and J. Bartram. 2007. Water, Sanitation, and Hygiene: Quantifying the Health Impact at National and Local Levels in Countries with Incomplete Water Supply and Sanitation Coverage. Environmental Burden of Disease Series 15. Geneva: World Health Organization. Fikree, F., M. Rahbar, and H. Berendes. 2000."Risk Factors for Stunting and Wasting at Age Six, Twelve, and Twenty-Four Months for Squatter Children of Karachi, Pakistan." Journal of the Pakistan Medical Association 50 (10): 341­48. Fishman, S., L. E. Caulfield, M. de Onis, M. Blössner, A. Hyder, L. Mullany, and R. E. Black. 2004. "Childhood and Maternal Underweight." In Comparative Quantification of Health REFERENCES 189 Risks: Global and Regional Burden of Disease Attributable to Selected Major Risk Factors, vol. 1, ed. M. Ezzati A. D. Lopez, A. Rodgers, and C. J. L. Murray, 39­61. Geneva: World Health Organization. Flemming, K., W. van der Hoek, F. P. Amerasinghe, C. Muteroc, and E. Boelee. 2004. "Engineering and Malaria Control: Learning from the Past 100 Years." Acta Tropica 89 (2): 99­108. Frenk, J. 2006. "Bridging the Divide: Global Lessons from Evidence-Based Health Policy in Mexico." Lancet 368 (9539): 954­61. Frias, J., and N. Mukherjee. 2005. Private Sector Sanitation Delivery in Vietnam: Harnessing Market Power for Rural Sanitation. Jakarta: World Bank. Galiani, S., P. Gertler, and E. Schargrodsky. 2005."Water for Life: The Impact of the Privatization of Water Services on Child Mortality." Journal of Political Economy 113 (1): 83­120. Ghana Statistical Service. 2003. The Ghana Demographic and Health Survey 2003. Accra: Ghana Statistical Service. http://www.measuredhs.com/aboutsurveys/search/metadata.cfm? surv_id=235&ctry_id=14&SrvyTp=type. Glewwe, P., and H. Jacoby. 1995."An Economic Analysis of Delayed Primary School Enrollment in a Low-Income Country: The Role of Early Childhood Nutrition." Review of Economics and Statistics 77 (1): 156­69. Glewwe, P., H. Jacoby, and E. King. 2001."Early Childhood Nutrition and Academic Achievement: A Longitudinal Analysis." Journal of Public Economics 81 (3): 345­68. Gragnolati, M., M. Shekar, M. Das Gupta, C. Bredenkamp, and Y.-K. Lee. 2006. India's Malnourished Children: A Call for Reform and Action. Washington, DC: World Bank. Graham, W., J. Cairns, S. Bhattacharya, C. H. W. Bullough, Z. Quayyum, and K. Rogo. 2006. "Maternal and Perinatal Conditions." In Disease Control Priorities in Developing Countries, 2nd ed., ed. D. T. Jamison, J. G. Breman, A. R. Measham, G. Alleyne, M. Claeson, D. B. Evans, P. Jha, A. Mills, and P. Musgrove, 499­529. Washington, DC: World Bank; New York: Oxford University Press. Grantham-McGregor, S., Y. B. Cheung, S. Cueto, P. Glewwe, L. Richter, and B. Strupp. 2007. "Child Development in Developing Countries: Developmental Potential in the First 5 Years for Children in Developing Countries." Lancet 369 (9555): 60­70. Griffiths, M., and J. S. McGuire. 2005. "A New Dimension for Health Reform: The Integrated Community Child Health Program in Honduras." In Health System Innovations in Central America: Lessons and Impact of New Approaches, ed. G. La Forgia, 173­96. Washington, DC: World Bank. Guerrant, D. I., S. R. Moore, A. A. M. Lima, P. D. Patrick, J. B. Schorling, and R. L. Guerrant. 1999. "Association of Early Childhood Diarrhea and Cryptosporidiosis with Impaired Physical Fitness and Cognitive Function Four­SevenYears Later in a Poor Urban Community in Northeast Brazil." American Journal of Tropical Medicine and Hygiene 61 (5): 707­13. Guerrant, R. L., L. V. Kirchhoff, D. S. Shields, M. K. Nations, J. Leslie, M. A. De Sousa, J. G. Araujo, L. L. Correia, K. T. Sauer, K. E. McClelland, F. L. Trowbridge, and J. M. Hughes. 1983. "Prospective Study of Diarrheal Illnesses in Northeastern Brazil: Patterns of Disease, Nutritional Impact, Etiologies, and Risk Factors." Journal of Infectious Diseases 148 (6): 986­97. Guerrant, R. L., A. A. M. Lima, and F. Davidson. 2000. "Micronutrients and Infection: Interactions and Implications with Enteric and Other Infections and Future Priorities." Journal of Infectious Diseases 182 (1 Suppl.): S134­38. 190 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Guerrant, R. L., J. B. Schorling, J. F. McAuliffe, and M. A. D. Souza. 1992."Diarrhea as a Cause and Effect of Malnutrition: Diarrhea Prevents Catch-up Growth and Malnutrition Increases Diarrhea Frequency and Duration." American Journal of Tropical Medicine and Hygiene 47 (1): 28­35. Guyatt, H. L., and R. W. Snow. 2004."Impact of Malaria during Pregnancy on Low Birth Weight in Sub-Saharan Africa." Clinical Microbiology Reviews 17 (4): 760­69. Ha, D. Q., and T. Q. Huan. 1997. "Dengue Activity in Viet Nam and Its Control Programme, 1997­98." Dengue Bulletin 21 (December): 35­40. http://www.searo.who.int/en/ Section10/Section332/Section519_2384.htm. Habicht, J.-P., J. DaVanzo, and W. P. Butz. 1988."Mother's Milk and Sewage: Their Interactive Effects on Infant Mortality." Pediatrics 81 (3): 456­61. Hadi, H., M. Dibley, and K. West. 2004. "Complex Interactions with Infection and Diet May Explain Seasonal Growth Responses to Vitamin A in Preschool Aged Indonesian Children." European Journal of Clinical Nutrition 58 (7): 990­99. Hadi, H., R. Stoltzfus, L. Moulton, M. Dibley, and K. West Jr. 1999. "Respiratory Infections Reduce the Growth Response to Vitamin A Supplementation in a Randomized Controlled Trial." International Journal of Epidemiology 28 (5): 874­81. Harrison, G., G. Brush, and F. Zumrawi. 1993."Motherhood and Infant Health in Khartoum." Bulletin of the World Health Organization 71 (5): 529­33. Heaver, R., and Y. Kachondam. 2002. Thailand's National Nutrition Program: Lessons in Management and Capacity Development. Washington, DC: World Bank. Henry, F. 1981."Environmental Sanitation Infection and Nutritional Status of Infants in Rural St. Lucia, West Indies." Transactions of the Royal Society of Tropical Medicine and Hygiene 75 (4): 507­13. Henry, F., N. Alam, K. Aziz, and M. Rahaman. 1987. "Dysentery, Not Watery Diarrhoea, Is Associated with Stunting in Bangladeshi Children."Human Nutrition and Clinical Nutrition 41 (4): 243­49. Holveck, J., J. Ehrenberg, S. Ault, R. Rojas, J. Vasquez, M. Cerqueira, J. Ippolito-Shepherd, M. Genovese, and M. Periago. 2007. "Prevention, Control, and Elimination of Neglected Diseases in the Americas: Pathways to Integrated, Inter-Programmatic, Inter-Sectoral Action for Health and Development." BMC Public Health 7 (1): 6. Horton, S. 1999. "Opportunities for Investments in Nutrition in Low-Income Asia." Asian Development Review, 17 (1­2): 246­73. http://www.adb.org/documents/periodicals/ adr/pdf/ADR-Vol17-Horton.pdf. Hotez, P. J., D. H. Molyneux, A. Fenwick, E. Ottesen, S. E. Sachs, and J. D. Sachs. 2006. "Incorporating a Rapid-Impact Package for Neglected Tropical Diseases with Programs for HIV/AIDS, Tuberculosis, and Malaria: A Comprehensive Pro-Poor Health Policy and Strategy for the Developing World." PLoS Medicine 3 (5): 576­84. Howard, G., and J. Bartram. 2003. "Domestic Water Quantity, Service Level, and Health." WHO/SDE/WSH/03.02. World Health Organization, Geneva. http://www.who.int/ water_sanitation_health/diseases/WSH0302exsum.pdf. Hunt, Caroline. 2006. "Sanitation and Human Development." Occasional Paper 26, Human Development Report Office, United Nations Development Programme, New York. Hunter, J. M. 1992. "Elephantiasis: A Disease of Development in North East Ghana." Social Science and Medicine 35 (5): 627­49. ------. 2003."Inherited Burden of Disease: Agricultural Dams and the Persistence of Bloody Urine (Schistosomiasis hematobium) in the Upper East Region of Ghana, 1959­1997."Social Science and Medicine 56 (2): 219­34. REFERENCES 191 Hutton, G., and L. Haller. 2004. Evaluation of the Cost and Benefits of Water and Sanitation Improvements at the Global Level. WHO/SDE/WSH/04.04. Geneva: World Health Organization. http://www.who.int/water_sanitation_health/wsh0404/en/index.html. Iannotti, L. L., J. M. Tielsch, M. M. Black, and R. E. Black. 2006."Iron Supplementation in Early Childhood: Health Benefits and Risks." American Journal Clinical Nutrition 84 (6): 1261­76. IPCC (Intergovernmental Panel on Climate Change). 2007. "Climate Change 2007: Impacts, Adaptation, and Vulnerability."In Contribution of Working Group II to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change, ed. M. L. Parry, O. F. Canziani, J. P. Palutikof, P. J. van der Linden, and C. E. Hanson, 79­131. Cambridge, U.K.: Cambridge University Press. Jacoby, H., and L. Wang. 2004. "Environmental Determinants of Child Mortality in Rural China: A Competing Risks Approach." Policy Research Working Paper 3241, World Bank, Washington, DC. Jefferson, T., R. Foxlee, C. Del Mar, L. Dooley, E. Ferroni, B. Hewak, A. Prabhala, S. Nair, and A. Rivetti. 2007."Interventions for the Interruption or Reduction of the Spread of Respiratory Viruses (Review)." Cochrane Database of Systematic Reviews (4). Kay, B., and V. Nam. 2005."New Strategy for Dengue Control in Vietnam." Lancet 365 (9459): 613­17. Keiser, J., B. Singer, and J. Utzinger. 2005. "Reducing the Burden of Malaria in Different Eco- Epidemiological Settings with Environmental Management: A Systematic Review." Lancet Infectious Diseases 5 (11): 695­708. Keusch, G. 2003. "The History of Nutrition: Malnutrition, Infection and Immunity." Journal of Nutrition 133 (1 Suppl.): 336S­40S. Kimani-Murage, E. W., and A. Ngindu. 2007. "Quality of Water the Slum Dwellers Use: The Case of a Kenyan Slum." Journal of Urban Health 84 (6): 829­38. Kivi, M., and Y. Tindberg. 2006. "Helicobacter pylori Occurrence and Transmission: A Family Affair?" Scandinavian Journal of Infectious Diseases 38 (6): 407­17. Knapp, A. 2006. "Mainstreaming Hygiene and Sanitation into the Health Sector: Experience from Local `at-Scale Sanitary Revolutions' and Recent Sector Developments in Ethiopia." Paper presented by W. Gebreselassie, Ethiopia Ministry of Health, at Water Week 2007, February 27­March 2. http://siteresources.worldbank.org/INTWRD/Resources/ Worku_Gebreselassie_Ethiopia_Mainstreaming_Hygiene_and_Sanitation_Into_Preventive_ Health_Care_Programs.pdf. Kolsky, P., and U. Blumenthal. 1995."Environmental Health Indicators and Sanitation-Related Disease in Developing Countries: Limitations to the Use of Routine Data Sources." World Health Statistics Quarterly 48 (2): 132­39. Kolsky, P., E. Perez, W.Vandersypen, and L. Jensen. 2005."Sanitation and Hygiene at the World Bank." Working Note No. 6, October 6, 2005. http://www-wds.worldbank.org/external/ default/WDSContentServer/WDSP/IB/2005/11/30/000090341_20051130115025/Rendered/ PDF/344690Sanitation0and0hygiene0at0wb.pdf. Kolsteren, P., J. Kusin, and S. Kardjati. 1997. "Morbidity and Growth Performance of Infants in Madura, Indonesia." Annales Tropical Pediatrics 17 (3): 201­8. Komisar, J. L. 2007. "Malaria Vaccines." Frontiers in Bioscience 12 (May): 3928­55. Kosek, M., C. Bern, and R. L. Guerrant. 2003. "The Global Burden of Diarrhoeal Disease, as Estimated from Studies Published between 1992 and 2000." Bulletin of the World Health Organization 81 (3): 197­204. 192 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Kotchian, S. 1997."Perspectives on the Place of Environmental Health and Protection in Public Health and Public Health Agencies." Annual Review of Public Health 18: 245­59. Kristensen, P. 2005. "Sub-Saharan Africa Region." Environment Matters at the World Bank: Annual Review 2005, 32­35. Krupnick,A., B. Larsen, and E. Strukova. 2006."Cost of Environmental Degradation in Pakistan: An Analysis of Physical and Monetary Losses in Environmental Health and Natural Resources." Background paper, World Bank, Washington, DC. Kumie, A., and A. Ahmed. 2005. "An Overview of Environmental Health Status in Ethiopia with Particular Emphasis to Its Organization, Drinking Water, and Sanitation: A Literature Survey." Ethiopia Journal of Health Development 19 (2): 89­103. Langhans, W. 2000. "Anorexia of Infection: Current Prospects." Nutrition Research 16 (10): 996­1005. Larsen, B. 2006. "Ghana: Cost of Environmental Damage--An Analysis of Environmental Health." Background paper, World Bank, Washington, DC. Laxminarayan, R., J. Chow, and S. Shahid-Salles. 2006. "Intervention Cost-Effectiveness Overview of Main Messages." In Disease Control Priorities in Developing Countries, 2nd ed., ed. D. T. Jamison, J. G. Breman, A. R. Measham, G. Alleyne, M. Claeson, D. B. Evans, P. Jha, A. Mills, and P. Musgrove, 35­86. Washington, DC: World Bank; New York: Oxford University Press. Lim, M. 2001."A Perspective on Tropical Sprue."Current Gastroenterology Reports 3 (4): 322­27. Lima, A., S. Moore, M. Barboza, A. Soares, M. Schleupner, R. Newman, C. Sears, J. Nataro, D. Fedorko, T. Wuhib, J. Schorling, and R. L. Guerrant. 2000."Persistent Diarrhea Signals a Critical Period of Increased Diarrhea Burdens and Nutritional Shortfalls: A Prospective Cohort Study among Children in Northeastern Brazil." Journal of Infectious Diseases 181 (5): 1643­51. Listorti, J., and F. Doumani. 2001. Environmental Health: Bridging the Gaps. Washington, DC: World Bank. Lorntz, B., A. M. Soares, S. R. Moore, R. Pinkerton, B. Gansneder, V. E. Bovbjerg, H. Guyatt, A. M. Lima, and R. L. Guerrant. 2006."Early Childhood Diarrhea Predicts Impaired School Performance." Pediatric Infectious Disease Journal 25 (6): 513­20. Luby, S., M. Agboatwalla, D. Feikin, J. Painter, W. Billhimer, and R. Hoekstra. 2005. "Effect of Handwashing on Child Health: A Randomised Controlled Trial."Lancet 366 (9481): 225­33. Lucas, R. M., and A. J. McMichael. 2005."Association or Causation? Evaluating Links between `Environment and Disease.'" Bulletin of the World Health Organization 83 (10): 792­95. Lunn, P., C. Northrop-Clewes, and R. Downes. 1991."Intestinal Permeability, Mucosal Injury, and Growth Faltering in Gambian Infants." Lancet 338 (8772): 907­10. Lutter, C., J. Mora, J. Habicht, K. Rasmussen, D. S. Robson, S. Sellers, C. Super, and M. Herrera. 1989. "Nutritional Supplementation: Effects on Child Stunting because of Diarrhea." American Journal of Clinical Nutrition 50 (1): 1­8. Lvovsky, K. 2001. Health and Environment. Washington, DC: World Bank. Mabaso, M. L. H., B. Sharp, and C. Lengeler. 2004. "Historical Review of Malarial Control in Southern Africa with Emphasis on the Use of Indoor Residual House-Spraying." Tropical Medicine and International Health 9 (8): 846­56. Mahendradhata, Y., and F. Moerman. 2004."Integration and Disease Control: Notes from the Prince Leopold Institute of Tropical Medicine Colloquium 2002." Tropical Medicine and International Health 9 (6): A5­10. REFERENCES 193 Majumder, R. 2004. "Total Sanitation Campaign: Changing the Face of Rural Burdwan." In India Infrastructure Report 2004: Ensuring Value for Money, ed. S. Morris, 325­26. New Delhi: Oxford University Press. Malaviya, V. S., R. Kant, C. S. Pant, H. C. Srivastava, and R. S. Yadav. 2006."Community Based Integrated Malaria Control with Reference to Involvement of Social Forestry Activities: An Experience." Indian Journal of Community Medicine 31 (4): 234­36. Maleta, K., S. M.Virtanen, M. Espo, T. Kulmala, and P. Ashorn. 2003."Childhood Malnutrition and Its Predictors in Rural Malawi." Paediatric and Perinatal Epidemiology 17 (4): 384­90. Maluccio, J., J. Hoddinott, and J. Behrman. 2006. "The Impact of an Experimental Nutrition Intervention in Childhood on Education among Guatemalan Adults." IFPRI and Food Consumption and Nutrition Division Discussion Paper, International Food Policy Research Institute, Washington, DC. Marodi, L. 2006."Innate Cellular Immune Responses in Newborns." Clinical Immunology 118 (2­3): 137­44. Martorell, R. 1995. "Results and Implications of the INCAP Follow-up Study." Journal of Nutrition 125 (4 Suppl.): 1127S­38S. Martorell, R., J. P. Habicht, C. Yarbrough, A. Lechtig, R. E. Klein, and K. A. Western. 1975. "Acute Morbidity and Physical Growth in Rural Guatemalan Children." American Journal of Disabled Children 129 (11): 1296­301. Martorell, R., C.Yarbrough, A. Lechtig, J. Habicht, and R. Klein. 1975."Diarrheal Diseases and Growth Retardation in Preschool Guatemalan Children." American Journal of Physical Anthropology 43 (3): 341­46. Mason, J., D. Sanders, P. Musgrove, and R. Galloway. 2006."Community Health and Nutrition Programs." In Disease Control Priorities in Developing Countries, 2nd ed., ed. D. T. Jamison, J. G. Breman, A. R. Measham, G. Alleyne, M. Claeson, D. B. Evans, P. Jha, A. Mills, and P. Musgrove, 1053­76. Washington, DC: World Bank; New York: Oxford University Press. Mata, L. 1992. "Diarrheal Disease as a Cause of Malnutrition." American Journal of Tropical Medicine and Hygiene 47 (1 Suppl.): 16­27. Mathee, A., F. Swanepoel, and A. Swart. 1999. "Environmental Health Services." In The South African Health Review, ed. N. Crisp and A. Ntuli, 277­88. Durban, South Africa: Health Systems Trust. http://www.healthlink.org.za/uploads/files/chapter20_99.pdf. Mavalankar, D., and S. Manjunath. 2004. "Sanitation and Water Supply: The Forgotten Infrastructure."In India Infrastructure Report 2004: Ensuring Value for Money, ed. S. Morris, 314­24. New Delhi: Oxford University Press. Menendez, C., A. F. Fleming, and P. L. Alonso. 2000. "Malaria-Related Anaemia." Parasitology Today 16 (11): 469­76. Modi, V., S. McDade, D. Lallement, and J. Saghir. 2005. "Energy Services for the Millennium Development Goals." Paper prepared for the Millennium Project of the United Nations Development Programme and World Bank, Washington, DC. Molbak, K., M. Andersen, P. Aaby, N. Hojlyng, M. Jakobsen, M. Sodeman, and A. P. J. d. Silva. 1997. "Cryptosporidium Infection in Infancy as a Cause of Malnutrition: A Community Study from Guinea-Bissau,West Africa."American Journal Clinical Nutrition 65 (1): 149­52. Moore, S., A. Lima, M. Conaway, J. Schorling, A. Soares, and R. L. Guerrant. 2001. "Early Childhood Diarrhoea and Helminthiases Associate with Long-Term Linear Growth Faltering." International Journal of Epidemiology 30 (6): 1457­64. 194 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Morley, D., J. Bicknell, and M. Woodland. 1968. "Factors Influencing the Growth and Nutritional Status of Infants and Young Children in a Nigerian Village." Transactions of the Royal Society of Tropical Medicine and Hygiene 62 (2): 164­99. Moser Jones, M. 2005. Protecting Public Health in New York City: 200 Years of Leadership (1805­2005). New York: New York City Department of Health and Mental Hygiene. Moy, R., T. Marshall, R. Choto, A. McNeish, and I. Booth. 1994. "Diarrhoea and Growth Faltering in Rural Zimbabwe." European Journal of Clinical Nutrition 48 (11): 810­21. Mullany, L. C., G. L. Darmstadt, S. K. Khatry, J. Katz, S. C. LeClerg, S. Shrestra, R. Adhikari, and J. M. Tielsch. 2006. "Topical Applications of Chlorhexidine to the Umbilical Cord for Prevention of Omphalitis and Neonatal Mortality in Southern Nepal: A Community- Based, Cluster-Randomized Trial." Lancet 367 (9514): 910­18. Murphy, H., B. Stanton, and J. Galbraith. 1997. Prevention: Environmental Health Interventions to Sustain Child Survival. Washington, DC: U.S. Agency for International Development. Murray, C., T. Laakso, K. Shibuya, T. Laakso, K. Hill, and A. D. Lopez. 2007. "Can We Achieve Millennium Development Goal 4? New Analysis of Country Trends and Forecasts of under-5 Mortality to 2015." Lancet 370 (9592): 1040­54. Nair, N., H. Gans, L. Lew-Yasukawa, A. C. Long-Wagar, A. Arvin, and D. E. Griffin. 2007."Age- Dependent Differences in IgG Isotype and Avidity Induced by Measles Vaccine Received during the First Year of Life." Journal of Infectious Diseases 196 (9): 1339­45. National Center for Infectious Diseases. 2006. "Malaria: Vector Control." Centers for Disease Control and Prevention, Atlanta. http://www.cdc.gov/malaria/control_prevention/ vector_control.htm. National Institute of Population Studies. 1992. Pakistan Demographic and Health Survey 1990/91. Islamabad: National Institute of Population Studies. Nguyen, T. V., P. Le Van, C. Le Huy, K. N. Gia, and A. Weintraub. 2005. "Detection and Characterization of Diarrheagenic Escherichia coli from Young Children in Hanoi, Vietnam." Journal of Clinical Microbiology 43(2): 755­60. Niehaus, M. D., S. R. Moore, P. D. Patrick, L. L. Derr, B. Lorntz, A. A. Lima, and R. L. Guerrant. 2002. "Early Childhood Diarrhea Is Associated with Diminished Cognitive Function 4 to 7 Years Later in Children in a Northeast Brazilian Shantytown."American Journal of Tropical Medicine and Hygiene 65 (5): 590­93. Parkin, D. M. 2006. "The Global Health Burden of Infection-Associated Cancers in the Year 2002." International Journal of Cancer 118 (12): 3030­44. Patrick, P. D., R. B. Oriá, V. Madhavan, R. C. Pinkerton, B. Lorntz, A. A. M. Lima, and R. L. Guerrant. 2005."Limitations in Verbal Fluency Following Heavy Burdens of Early Childhood Diarrhea in Brazilian Shantytown Children." Child Neuropsychology 11 (3): 233­44. Pearce,D.W.,and D.Ulph.1999."A Social Discount Rate for the United Kingdom."In Environmental Economics: Essays in Ecological Economics and Sustainable Development, ed. D. W. Pearce, 268­85. Cheltenham, U.K.: Edward Elgar. Pelletier, D. L. 1994. "The Relationship between Child Anthropometry and Mortality in Developing Countries: Implications for Policy, Programs, and Future Research." Journal of Nutrition 124 (10 Suppl.): 2047S­81S. Pelletier, D. L., and E.A. Frongillo. 2003."Changes in Child Survival Are Strongly Associated with Changes in Malnutrition in Developing Countries." Journal of Nutrition 133 (1): 107­19. Pelletier, D. L., E. A. Frongillo, D. G. Schroeder, and J.-P. Habicht. 1994. "A Methodology for Estimating the Contribution of Malnutrition to Child Mortality in Developing Countries." Journal of Nutrition 124 (10 Suppl.): 2106S­22S. REFERENCES 195 PIDE (Pakistan Institute of Development Economics, Micronutrient Laboratories Aga Khan University, Medical Centre). 2003. National Nutrition Survey 2001­2002. Islamabad: Government of Pakistan, Planning Commission. Poverty Environment Partnership.2008."Poverty,Health,and Environment: Placing Environmental Health on Countries' Development Agendas."Joint agency paper, World Bank, Washington, DC. Powanda, M. C., and W. R. Beisel. 2003."Metabolic Effects of Infection on Protein and Energy Status." Journal of Nutrition 133 (1 Suppl.): 322S­27S. Prado, M., S. Cairncross, M. Barreto, A. Oliviera-Assis, and S. Rego. 2005. "Asymptomatic Giardiasis and Growth in Young Children: A Longitudinal Study in Salvador, Brazil." Parasitology 131 (1): 51­56. Prüss-Üstün, A., and C. Corvalán. 2006. Preventing Disease through Healthy Environments: Towards an Estimate of the Environmental Burden of Disease. Geneva: World Health Organization. Psacharopoulos, G., and H. Patrinos. 2004. "Returns to Investment in Education: A Further Update." Education Economics 12 (2): 111­34. Public-Private Partnership for Handwashing with Soap. 2005."Peru."World Bank, Washington, DC. http://www.globalhandwashing.org/Country%20act/Peru.htm. Quddus, A., S. Luby, M. Rahbar, and Y. Pervaiz. 2002. "Neonatal Tetanus: Mortality Rate and Risk Factors in Loralai District, Pakistan." International Journal of Epidemiology 31 (3): 648­53. Rabie, T., and V. Curtis. 2006."Handwashing and Risk of Respiratory Infections: A Quantitative Systematic Review." Tropical Medicine and International Health 11 (3): 258­67. Ramalingaswami, V., U. Jonsson, and J. Rohde. 1996. "Commentary: The Asian Enigma." In The Progress of Nations, ed. P. Adamson, 11­17. New York: United Nations Children's Fund. RBM (Roll Back Malaria Partnership). 2005. World Malaria Report 2005. Geneva: United Nations Children's Fund and World Health Organization. http://www.rbm.who.int/ wmr2005/index.html. Rehfuess, E., S. Mehta, and A. Prüss-Üstün. 2006."Assessing Household Solid Fuel Use: Multiple Implications for the Millennium Development Goals." Environmental Health Perspectives 114 (3): 373­78. Remme, J. H. F., P. Feentstra, P. R. Lever, A. Medici, C. Morel, M. Noma, K. D. Ramaiah, F. Richards, A. Seketeli, G. Schmunis, W. H. van Brakel, and A. Vassall. 2006. "Tropical Diseases Targeted for Elimination: Chagas Disease, Lympahtic Filariasis, Onchocerciasis, and Leprosy." In Disease Control Priorities in Developing Countries, 2nd ed., ed. D. T. Jamison, J. G. Breman, A. R. Measham, G. Alleyne, M. Claeson, D. B. Evans, P. Jha, A. Mills, and P. Musgrove, 433­49. Washington, DC: World Bank; New York: Oxford University Press. Rosen, G. 1958. A History of Public Health. Baltimore, MD: Johns Hopkins University Press. Rowland, M., T. J. Cole, and R. Whitehead. 1977. "A Quantitative Study into the Role of Infection in Determining Nutritional Status in Gambian Village Children." Journal of Nutrition 37 (3): 441­50. Rowland, M., S. G. Rowland, and T. J. Cole. 1988. "Impact of Infection on the Growth of Children from 0 to 2 Years in an Urban West African Community." American Journal of Clinical Nutrition 47 (1): 134­38. Sachdev, H. S., C. H. D. Fall, C. Osmond, R. Lakshmy, S. K. D. Biswas, S. D. Leary, K. S. Reddy, D. J. P. Barker, and S. K. Bhargava. 2005. "Anthropometric Indicators of Body 196 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Composition in Young Adults: Relation to Size at Birth and Serial Measurements of Body Mass Index in Childhood in the New Delhi Birth Cohort." American Journal of Clinical Nutrition 82 (2): 456­66. Sarraf, M., B. Larsen, and M. Owaygen. 2004 "Cost of Environmental Degradation: The Case of Lebanon and Tunisia." Environment Department Paper 97, World Bank, Washington, DC. Sato, H., P. Albrecht, D. W. Reynolds, S. Stagno, and F. A. Ennis. 1979. "Transfer of Measles, Mumps, and Rubella Antibodies from Mother to Infant: Its Effect on Measles, Mumps, and Rubella Immunization." American Journal of Diseases of Children 133 (12): 1240­43. Satterthwaite, D. 2007. "In Pursuit of a Healthy Urban Environment in Low- and Middle- Income Nations." In Scaling Urban Environmental Challenges: From Local to Global and Back, ed. P. Marcotullio and G. McGranahan, 69­105. London: Earthscan. Schorling, J., J. F. Mcauliffe, M. A. De Souza, and R. L. Guerrant. 1990. "Malnutrition Is Associated with Increased Diarrhoea Incidence and Duration among Children in an Urban Brazilian Slum." International Journal of Epidemiology 19 (3): 728­35. Scrimshaw, N. S. 2003. "Historical Concepts of Interactions, Synergism, and Antagonism between Nutrition and Infection." Journal of Nutrition 133 (1 Suppl.): 316S­21S. Scrimshaw, N. S., C. E. Taylor, and J. E. Gordon. 1959."Interactions of Nutrition and Infection." American Journal of Medical Sciences 237 (3): 367­72. ------. 1968. Interactions of Nutrition and Infection. Geneva: World Health Organization. Sedgwick, W. T., and S. Macnutt. 1908."An Examination of the Theorem of Allen Hazen That for Every Death from Typhoid Fever Avoided by the Purification of Public Water Supplies Two or Three Deaths Are Avoided from Other Causes." Science 28 (711): 215­16. Sepúlveda, J., F. Bustreo, R. Tapia, J. A. Rivera, R. Lozano, G. Oláiz,V. Partida, L. García-García, and J. L. Valdespino. 2006. "Improvement of Child Survival in Mexico: The Diagonal Approach." Lancet 368 (9551): 2017­27. Shekar, M., R. Heaver, Y.-K. Lee, and World Bank. 2006. Repositioning Nutrition as Central to Development: A Strategy for Large-Scale Action. Washington, DC: World Bank. Shrimpton, R., C. Victora, M. de Onis, R. Costa Lima, M. Blössner, and G. Clugston. 2001. "The Worldwide Timing of Growth Faltering: Implications for Nutritional Interventions." Pediatrics 107 (5): e75. Singleton, R., T. Hennessy, L. Bulkow, L. Hammitt, T. Zulz, D. Hurlburt, J. Butler, K. Rudolph, and A. Parkinson. 2007. "Invasive Pneumococcal Disease Caused by Nonvaccine Serotypes among Alaska Native Children with High Levels of 7-Valent Pneumococcal Conjugate Vaccine Coverage." Journal of the American Medical Association 297 (16): 1784­92. Sinton, J., K. Smith, J. Peabody, Y. Liu, X. Zhang, and R. Edwards. 2004. "An Assessment of Programs to Promote Improved Household Stoves in China." Energy for Sustainable Development 8 (3): 33­52. Smith, K. R., C. F. Corvalán, and T. Kjellström. 1999. "How Much Global Ill Health Is Attributable to Environmental Factors?" Epidemiology and Infection 10 (5): 573­84. Smith, K. R., S. Mehta, and M. Maeusezahl-Feuz. 2004. "Indoor Air Pollution from Household Use of Solid Fuels." In Comparative Quantification of Health Risks: Global and Regional Burden of Disease Attributable to Selected Major Risk Factors, vol. 2, ed. M. Ezzati, A. D. Lopez, A. Rodgers, and C. J. L. Murray, 1435­93. Geneva: World Health Organization. REFERENCES 197 Snow, R., M. Craig, C. Newton, and R. Stekete. 2003. The Public Health Burden of Plasmodium falciparum Malaria in Africa: Deriving the Numbers. Bethesda, MD: Fogarty International Center, National Institutes of Health. Steiner, T., A. Lima, J. Nataro, and R. L. Guerrant. 1998. "Enteroaggregative Escherichia coli Produce Intestinal Inflammation and Growth Impairment and Cause Interleukin-8 Release from Intestinal Epithelial Cells." Journal of Infectious Diseases 177 (1): 88­96. Steketee, R. W. 2003. "Pregnancy, Nutrition and Parasitic Diseases." Journal of Nutrition 133 (5): 1661S­67S. Stephensen, C. B. 1999."Burden of Infection on Growth Failure." Journal of Nutrition 129 (2): 534­38. Sur, D., D. Saha, B. Manna, K. Rajendran, and S. Bhattacharya. 2005. "Periodic Deworming with Albendazole and Its Impact on Growth Status and Diarrhoeal Incidence among Children in an Urban Slum of India." Transactions of the Royal Society of Tropical Medicine and Hygiene 99 (4): 261­67. Svedberg, P. 2006."Child Malnutrition in Shining India: A X-state Empirical Analysis." Institute for Economic Studies, Stockholm. http://www.bvsde.paho.org/bvsacd/cd56/svedberg- 070906.pdf. Taylor-Robinson, D., A. Jones, and P. Garner. 2007. "Deworming Drugs for Treating Soil- Transmitted Intestinal Worms in Children: Effects on Growth and School Performance." Cochrane Database of Systematic Reviews 17 (4). Thapar, N., and I. Sanderson. 2004."Diarrhoea in Children: An Interface between Developing and Developed Countries." Lancet 363 (9409): 641­53. Thomas, E. P., J. R. Seager, and A. Mathee. 2002. "Environmental Health Challenges in South Africa: Policy Lessons from Case Studies." Health and Place 8 (4): 251­61. Touré, Y. 2001. "Malaria Vector Control in Africa: Strategies and Challenges." Report from a symposium held at the 2001 American Association for the Advancement of Science Annual Meeting, San Francisco, February 17. http://www.aaas.org/international/africa/ malaria/toure.html. Trades Union Congress. 2004. "Incomes in Ghana: Policy Discussion Paper." Trades Union Congress, Accra. UNICEF (United Nations Children's Fund). 2005. "Global Database on Low Birthweight." UNICEF, New York. http://www.childinfo.org/areas/birthweight/database.php. ------. 2006. "Water." UNICEF, New York. http://www.childinfo.org/areas/water/. ------. 2008. "Info by Country." UNICEF, New York. http://www.unicef.org/infobycountry/ ghana_statistics.html. ------. 2008. "Info by Country." UNICEF, New York. http://www.unicef.org/infobycountry/ pakistan_pakistan_statistics.html#47. Valentiner-Branth, P., H. Steinsland, G. Santos, M. Perch, K. Begtrup, M. K. Bhan, F. Dias, P. Aaby, H. Sommerfelt, and K. Molbak. 2001. "Community-Based Controlled Trial of Dietary Management of Children with Persistent Diarrhea: Sustained Beneficial Effect on Ponderal and Linear Growth." American Journal of Clinical Nutrition 73 (5): 968­74. VanDerslice, J., B. Popkin, and J. Briscoe. 1994."Drinking-Water Quality, Sanitation, and Breast- Feeding: Their Interactive Effects on Infant Health."Bulletin of the World Health Organization 72 (4): 589­601. van Geertruyden, J., F. Thomas, A. Erhart, and U. D'Alessandro. 2004. "The Contribution of Malaria in Pregnancy to Perinatal Mortality." American Journal of Tropical Medicine and Hygiene 71 (2 Suppl.): 35­40. 198 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL Vergnano, S., M. Sharland, P. Kazembe, C. Mwansambo, and P. T. Heath. 2005."Neonatal Sepsis: An International Perspective." Archives of Disease in Childhood: Fetal and Neonatal Edition 90 (3): F220­24. Victora, C. G., F. C. Barros, B. R. Kirkwood, and P.Vaughan. 1990."Pneumonia, Diarrhea, and Growth in the First 4 Years of Life: A Longitudinal Study of 5,914 Urban Brazilian Children." American Journal Clinical Nutrition 52 (2): 391­96. Victora, C. G., B. Fenn, J. Bryce, and B. R. Kirkwood. 2005. "Co-coverage of Preventive Interventions and Implications for Child-Survival Strategies: Evidence from National Surveys." Lancet 366 (9495): 1460­66. Villamor, E., M. Fataki, R. Bosch, R. Mbise, and W. Fawzi. 2004. "Human Immunodeficiency Virus Infection, Diarrheal Disease, and Sociodemographic Predictors of Child Growth." Acta Paediatrica 93 (3): 372­79. von Schirnding, Y. 2005. "Short Report: The World Summit on Sustainable Development: Reaffirming the Centrality of Health." Globalization and Health 1 (8): 1­6. Vu, S. N., T. Y. Nguyen, T. V. Phong, T. U. Ninh, L. Q. Mai, L. V. Lo, L. T. Nghia, A. Bektas, A. Briscombe, J. G. Aaskov, P. A. Ryan, and B. H. Kay. 2005. "Elimination of Dengue by Community Programs Using Mesocyclops (Copepoda) against Aedis Aegypti in Central Vietnam." American Journal of Tropical Medicine and Hygiene 72 (1): 67­73. Wagstaff, A., and M. Claeson. 2004. The Millennium Development Goals for Health: Rising to the Challenges. Washington, DC: World Bank. Walker, S. P., S. M. Chang, C. A. Powell, and S. M. Grantham-McGregor. 2005."Effects of Early Childhood Psychosocial Stimulation and Nutritional Supplementation on Cognition and Education in Growth-Stunted Jamaican Children: Prospective Cohort Study." Lancet 366 (9499): 1804­7. Walker, S. P., T. D. Wachs, J. Meeks Garder, B. Lozoff, G. A. Wasserman, E. Pollitt, and J. Carter. 2007."Child Development: Risk Factors for Adverse Outcomes in Developing Countries." Lancet 369 (9556): 145­57. Watson-Jones, D., H. A. Weiss, J. M. Changalucha, J. Todd, B. Gumodoka, J. Bulmer, R. Balira, D. Ross, K. Mugeye, R. Hayes, and D. Mabey. 2007. "Adverse Birth Outcomes in United Republic of Tanzania: Impact and Prevention of Maternal Risk Factors." Bulletin of the World Health Organization 85 (1): 9­18. West, K. P. 2003."Vitamin A Deficiency Disorders in Children and Women."Food and Nutrition Bulletin 24 (4): S78­S90. WHO (World Health Organization). 2000. Addressing the Links between Indoor Air Pollution, Household Energy and Human Health. Geneva: WHO. ------. 2002. The World Health Report 2002: Reducing Risks, Promoting Healthy Life. Geneva: WHO. http://www.who.int/whr/2002/en/index.html. ------. 2003. Insecticide-Treated Mosquito Net Interventions: A Manual for National Control Programme Managers. Geneva: WHO. ------. 2004. World Health Report 2004. Geneva: WHO. ------. 2006. Fuel for Life: Household Energy and Health. Geneva: WHO. ------. 2007a. "Burden of Disease Project." WHO, Geneva. http://www.who.int/healthinfo// bodproject/en/index.html. ------. 2007b."Children's Environmental Health."WHO, Geneva. http://www.who.int/ceh/en/. ------. 2007c. Combating Waterborne Disease at the Household Level. Geneva: WHO. ------. 2007d."Malaria." Fact Sheet 94, WHO, Geneva. http://www.who.int/mediacentre/fact sheets/fs094/en/index.html. REFERENCES 199 ------. 2007e. World Health Report 2007-- A Safer Future: Global Public Health Security in the 21st Century. Geneva: WHO. ------. 2008. "Environmental Health." WHO, Geneva. http://www.who.int/topics/environ mental_health/en/. ------. n.d. "Quantification of the Burden of Disease Attributable to Environmental Risk Factors." WHO, Geneva. http://www.who.int/quantifying_ehimpacts/summaryEBD.pdf. WHO (World Health Organization) Regional Office for Africa. 2006. "Country Profiles on Environmental Health Policy." WHO Regional Office for Africa, Brazzaville. http://www.afro.who.int/des/phe/ country_profiles/index.html. WHO and UNICEF (World Health Organization and United Nations Children's Fund). 2005. Water for Life: Making It Happen. Geneva: WHO and UNICEF Joint Monitoring Programme for Water Supply and Sanitation. http://www.wssinfo.org/en/40_ wfl_2005.html. Wierzba, T., R. Abu El-Yazeed, S. Savarino, A. Mourad, M. Rao, M. Baddour, A. N. El-Deen, A. Naficy, and J. Clemens. 2001. "The Interrelationship of Malnutrition and Diarrhea in a Periurban Area outside Alexandria, Egypt." Journal of Pediatric Gastroenterology and Nutrition 32 (2): 189­96. Wilson, S. 1986."Immunologic Basis for Increased Susceptibility of the Neonate to Infection." Journal of Pediatrics 108 (1): 1­9. World Bank. 2000."Eritrea Integrated Early Childhood Development Project."Report 20373-ER. Washington,DC,World Bank.http://www-wds.worldbank.org/external/default/WDSContent Server/WDSP/IB/2000/07/29/000094946_0006020538277/Rendered/PDF/multi_page.pdf. ------. 2002. "Arab Republic of Egypt: Cost Assessment of Environmental Degradation-- Sector Note." Report 25175-EGT, World Bank, Washington, DC. ------. 2005. "Islamic Republic of Iran: Cost Assessment of Environmental Degradation-- Sector Note." Report 32043-IR, World Bank, Washington, DC. ------. 2006a. Environmental Priorities and Poverty Reduction: A Country Environmental Analysis for Colombia. Washington, DC: World Bank. ------. 2006b. Ghana: Country Environmental Analysis. Washington, DC: World Bank. ------. 2006c. Repositioning Nutrition as Central to Development: A Strategy for Large-Scale Action. Washington, DC: World Bank. ------. 2007a. Cost of Pollution in China: Economic Estimates of Physical Damages. Washington, DC: World Bank. ------. 2007b. "Ghana--Nutrition and Malaria Control for Child Survival Project." Project appraisal document, World Bank, Washington, DC. ------. 2007c. Health, Nutrition, and Population Sector Strategy Paper. Washington, DC: World Bank. ------. 2007d. Improving Traditional Hearths: Cleaner Air and Better Health in India. Washington, DC: World Bank. ------. 2007e. Key Issues in Central America Health Reforms: Diagnosis and Implications. Report 36426-LAC, World Bank, Washington, DC. ------. 2007f. Nutritional Failure in Ecuador: Causes, Consequences, and Solutions. Washington, DC: World Bank. ------. 2007g. "Republic of Peru Environmental Sustainability: A Key to Poverty Reduction in Peru: Country Environmental Assessment."Report 40190-PE, World Bank, Washington, DC. http://go.worldbank.org/LDDPJN2TU0. ------. 2007h. World Development Indicators. Washington, DC: World Bank. 200 ENVIRONMENTAL HEALTH AND CHILD SURVIVAL ------. 2008. Poverty and the Environment: Understanding Linkages at the Household Level. Washington, DC: World Bank. World Resources Institute, United Nations Environment Programme, United Nations Development Programme, and World Bank. 1998. World Resources, 1998­99: A Guide to the Global Environment: Environmental Change and Human Health. New York: Oxford University Press. World Water Forum. 2006."Water Infrastructures for Sustainable and Equitable Development." Framework Theme 1.34, Fourth World Water Forum, Mexico City, March 16­22. http://www.worldwatercouncil.org/index.php?id=1196. WSP (Water and Sanitation Program). 2005. Lessons Learned from Bangladesh, India, and Pakistan: Scaling-Up Rural Sanitation in South Asia. New Delhi: WSP, World Bank. ------. 2006. The Mumbai Slum Sanitation Program: Partnering with Slum Communities for Sustainable Sanitation in a Megalopolis. New Delhi: WSP, World Bank. ------. 2007. From Burden to Communal Responsibility: A Sanitation Success Story from Southern Region in Ethiopia. New Delhi: WSP, World Bank. Yacoob, M., and M. Kelly. 1999. "Secondary Cities in West Africa: The Challenge for Environmental Health and Prevention." Comparative Urban Studies Occasional Paper 21, Woodrow Wilson International Center for Scholars, Washington, DC. Zaracostas, J. 2007. "Mortality from Measles Fell by 91% in Africa and by 68% Worldwide from 2000 to 2006." British Medical Journal 335: 1173. Zhang, J., and K. Smith. 2005. "Indoor Air Pollution from Household Fuel Combustion in China: A Review." Presentation at Indoor Air 2005, Beijing, September 4­9. Zumrawi, F., H. Dimond, and J. Waterlow. 1987."Effects of Infection on Growth in Sudanese Children." Human Nutrition and Clinical Nutrition 41 (6): 453­61. I N D E X Boxes, figures, notes, and tables are indicated by b, f, n, and t, respectively. A environmental health interventions in, 11­12 acute lower respiratory infection (ALRI) immunization in, 36 health systems and, 32, 33t infrastructure budgets in, 89 immunity, 27 malaria in, 6 indoor air pollution and, 1, 42, 50, 57, 75b, malnutrition in, 1, 6, 17, 27f, 30, 111 174­77 vector-control programs in, 96 malnutrition and, 17, 19, 21t attributable fractions (AF) micronutrient supplementation and, 37­38 burden of disease when multiple risk factors multiplier effect on, 56 present and, 75b prevention of, 57, 58 of child mortality, 66­67, 67t, 151­59 Adair, L.S., 144­45 Cox hazard model and, 170n3 advocacy, 13, 106, 108, 115 malnutrition-attributable fractions for AF. See attributable fractions different diseases, 69, 70t, 71­72, 74t, Afghanistan, malnutrition in, 6 173­76 Africa See also specific regions and countries B environmental health interventions in, 104 malnutrition in, 27f Bangladesh trypanosomiasis in, 43n1 case-control study in, 118, 124 AIDS. See HIV/AIDS indoor air pollution, 43 air pollution. See indoor air pollution; outdoor malnutrition in, 6 air pollution mortality in, 28 Alderman, H., 29, 144, 145 nutritional status and infectious disease in, Algeria and budget increases for 128t, 129t, 131t, 136t environmental protection, 62 oral rehydration therapy in, 35 Alilio, M.S., 57 Total Sanitation Campaign, 90, 94­95, ALRI. See acute lower respiratory infection 95b, 98 Alternative Energy Promotion Center vector-control measures in, 91 (Nepal), 100 weight gain retardation in, 50b American Academy of Pediatrics, 122 Barker hypothesis, 25b, 58 anemia, 7, 25, 26, 35, 58, 113 See also fetal programming antibiotics, 35 bednets. See insecticide-treated nets Argentina, privatization of public water Behrman, J., 144 companies in, 99 Beisel, W.R., 121 ARI. See acute lower respiratory infection bias (statistical), 170n4, 176­77 Asia biomass, 8, 26, 41­42, 95, 96 See also specific countries See also fuels birth weight in, 9 birth weight, 9, 25b, 25­26, 31, 54b, 113 burden of disease in, 51­52, 54, 54b, 59 See also underweight education in, 77 Blössner, M., 51 201 202 INDEX Bolivia, indoor air pollution in, 62 China Brazil cost of environmental health risks in, growth stunting in, 141 62, 63f helminth infections in, 21, 120 National Improved Stove Program (NISP), malnutrition and educational outcomes 42, 98 in, 64 nutritional status and infectious disease nutritional status and infectious disease in, in, 127t 127­28t, 130t, 131t, 135­37t, 140t water-related deaths in rural areas of, 56 public participation in health programs citizen mobilization, 107 in, 98 Clean Water Program (Mexico), 92b school enrollment and performance in, climate change 29, 145 related diseases, 49t, 174t vector-control measures in, 91 cognition breastfeeding, 22b, 27­28, 31, 36, 120 See also school performance Breman, J.G., 57 childhood growth and, 23 Briend, A., 120, 125n2 diarrhea and, 29 burden of disease, 47­59, 115 environmental impacts on, 28­29, 30, areas for further research, 58, 59 61, 112 environmental burdens, 48­51 malnutrition and, 28­30, 51, 63­64, 68, environmental health burdens, 9, 51­58, 76, 112 52f, 53f, 59 valuations of, 63­64 estimates of, 57­59 Colombia country-level environmental health, cost-benefit analysis in, 114 computation of, 151­71 cost of environmental health risks in, 63f methodological aspects of, 173­77 nutritional status and infectious disease in, monetary values, 179­81 130t, 140t from water, sanitation, and hygiene, Community-Based Poverty Reduction Project 147­49, 148t­149t (Ghana), 93 malnutrition-infection link, 51 community empowerment, 11, 98­99 multiplier effect of interventions, Comparative Risk Assessment (WHO), 48­49, 55f, 55­56 49t, 173, 174t cookstoves, 42­43, 95, 98, 99, 100 C See also fuels Cairncross, S., 18 coordination. See intersectoral and interjuris- Carson, Rachel, 88 dictional coordination Central America Corvalán, C., 49, 67, 177n2 See also specific countries Costa Rica, targeted environmental health growth and height indicators in, 121 programs in, 91 nutrition programs in, 91, 93 cost-benefit analysis, 80­81, 114 Chatterjee, N., 75b Cox hazard model, 155, 156t, 170n3 childbirth, 35­36 D child survival strategies, 31­43, 32f, 112­13 DALYs. See disability-adjusted life years environmental management programs, 9, Dandoy, S., 105 38­39, 43 Daniels, M.C., 144­45 health systems, 9, 32­38, 33t­34t, 43 Das Gupta, M., 110n1 childbirth and infant care programs, death rates. See mortality 35­36 dehydration, 20 immunization programs, 36­37 micronutrient supplementation See also oral rehydration therapy programs, 37­38 Demographic and Health Surveys (DHSs), infrastructure strategies, 9, 39­43 154, 155, 157­58, 162 fuel supply, 41­43 dengue virus, 43n1, 97b, 99 sanitation, 40­41 de Onis, M., 51 water supply, 39­40 Desai, M., 67 INDEX 203 developing countries Egypt See also specific countries cost of environmental health risks in, 62, 63f environmental health experiences in, 10­13, nutritional status and infectious disease 83­110 in, 138t health agendas in, 89­90, 101­4 El Salvador, indoor air pollution in, 62 infectious disease studies in, 126t­140t empowerment, 11, 98­99, 110, 113 infrastructure programs in, 94­96, 109, 113 energy and energy supplies. See fuels integrated child survival programs in, environmental health, defined, 2b, 5­6 90­91, 109 environmental health approaches, 85­110 nutrition programs in, 91­94, 109, 113 critical moments in, 109, 110 vector-control programs in, 96, 112 decline in agenda for, 10, 88­90, 109, 113 devolution of authority, 105­6 developing country experiences, 10­13, deworming, 22b­23b, 92, 118, 122­23 90­97, 109, 112­13 See also helminth infections enabling environment, 97­100 DHS. See Demographic and Health Surveys governance and institutional implications, diarrhea 100­104 antibiotics for treating, 35 history, 86­88 burden of disease, 51, 156 institutional requirements, 104­9, 110, 113 cognition and, 29 environmental health valuations, 61­82, dehydration and, 20 179­81 food restriction and, 122 cognition and, 63­64 growth faltering and, 120­21, 124 current practices, 61­62, 63f health systems and, 32, 33t education and, 63­64, 76t, 76­77, 81, 112, household prevention of, 28 141­46, 164­70 in infancy, 22b, 27, 36, 117 malnutrition and, 9­10, 63, 66­78, 68t, 74f, 74t, 76t, 81 infections and, 17, 21, 123 new estimates for, 62­64, 77­79, 79t, 80b, malnutrition and, 21t, 22b­23b, 68 151­71, 154t, 179­81 micronutrient supplementation and, 37, 38 next steps, 79­81 mortality and, 1, 55f, 55­56, 56b, 57, 59 environmental management programs, 9, transmission routes, 18­19 38­39 underweight and, 50b, 118, 125n4, 158, 158t epidemiology, 8­9, 12, 15­43 disability-adjusted life years (DALYs), 48, Eritrea Integrated Early Childhood 51­52, 53f, 148t­149t, 173 Development Project, 91 diseases. See specific diseases by name Ethiopia drug resistance, 35, 37 environmental health institutions in, 102, dysentery, 21, 50b, 120, 124, 125n2 102b­103b environmental health policies in, 104 E political commitment in, 98 sanitation in, 12, 95, 95b, 102b early infancy, 26­28, 27f, 30 Europe economic costs, 9­10, 12 developing countries compared to, 52, cost-benefit analysis, 80­81, 114 55f, 180 country-level costs of environmental health malaria in, 6 burden, 51­58 Expanded Program on Immunization, 36 environmental health valuations, 61­82 See also environmental health valuations F Ecuador, nutrition programs in, 91, 94 education fecal pollution, 18, 19f, 28, 39 malnutrition and, 76t, 76­77, 81, 141­46, "fetal-programming," 25b, 58 142t­143t, 160­70 See also Barker hypothesis school performance, 29, 160, 162­64 fetuses, 7, 24­26, 25b, 58, 59 stunting and, 161t See also pregnancy valuation of, 63­64, 76t, 76­77, 81, 112, Fewtrell, L., 51, 56b, 57, 112, 123 141­46, 164­70 Fishman, S., 69, 176 204 INDEX food restriction and diarrhea, 122 growth faltering. See linear growth and growth food security, 25b, 115 faltering fuels See also IUGR See also biomass; cookstoves Guatemala indoor air pollution and, 6, 26, 61, 95­96 cost of environmental health risks in, 63f infrastructure improvements and, 41­43 indoor air pollution in, 26, 42, 62 nutritional status and infectious disease in, 126t G school performance in, 64, 144, 163­64 strength of association study in, 119 Gambia growth retardation in, 119, 120 nutritional status and infectious disease in, H 126t, 129t handwashing. See hygiene GDP. See gross domestic product Hartge, P., 75b Ghana health inspectors, 88 ALRI in, 57, 66­67, 158, 159f health ministries and departments basic indicators for, 64­65, 65b advocacy role of, 106 childhood weight in, 68­69, 70f, 71f, 71­72, in developing countries, 88­89 73f, 124, 158, 158t environmental health and, 100 Community-Based Poverty Reduction Project, 93 health programs and, 11 cost of environmental health risks in, 63f public health personnel, 108 Demographic and Health Survey, 154, 155, technical and policy inputs, 106 157, 162 health systems, 9, 32­38, 33­34t, 43 diarrhea in, 69, 71f, 71­72, 73f height and growth of children. See linear environmental health valuations in, 9, 61, growth and growth faltering 64­80, 78f, 81, 112, 168­69t helminth infections fuels in, 66­67 See also hookworms GDP in, 9­10, 10t, 74­75, 78, 113, 159 cognitive functions and, 29 height-for-age rates in, 162, 162t deworming, 118, 122­23 hygiene in, 100, 106 See also deworming indoor air pollution in, 66­67, 75b, 154, drug resistance and, 35 155f growth faltering and, 21, 120 malaria in, 65b health systems and, 32 malnutrition in, 166­67, 167t, 170 Hill, Austin Bradford, 50b, 118 mortality in, 66­67, 67t, 70t, 82nn10­11, HIV/AIDS, 2, 21t, 36 151, 159, 159t, 180­81 Hoddinott, J., 144 nutritional programs in, 91 Honduras school enrollment and education in, 145, Integrated Community-Based Child Care 146, 161t, 163, 171n13 Program, 91 water in, 103 nutritional status and infectious disease in, Ghana Demographic and Health Survey, 133t 68, 82n3 hookworms Glewwe, P., 145 See also helminth infections Global Burden of Disease Project, 48 deworming of, 122 Global Public-Private Partnership for environmental risk factors and, 112 Handwashing with Soap, 100 malnutrition and, 21t Gordon, J.E., 19, 22b, 117, 121 during pregnancy, 35 Grantham-McGregor, S., 141 sanitation and, 25, 26 gross domestic product (GDP) vaccines against, 37 burden of disease and, 12, 180 human capital approach, 159, 179, 180, 181n1 diarrheal infections and, 77 hygiene health costs and, 47 burden of disease, 49, 147­49, 148t­149t malnutrition's effect on, 81 child delivery and, 35­36 INDEX 205 child survival programs and, 90­91 linear growth in, 23, 25b diarrhea and, 38 malnutrition and, 19, 26­28, 27f, 30, 68t, environmental risk factors and, 61 68­76 infrastructure improvements and, 40, 41 infections maternal and childhood underweight, 50, 50b diarrhea and, 17, 21, 123 miasma theory and, 86 in early infancy, 26­28, 35­36 pathogen transmission and, 28 growth faltering and, 117, 119­20 private sector involvement in, 100 malnutrition and, 19­23, 20f, 30, 51, 63, related diseases, 49t, 174t 118, 123­24, 126t­140t water quantities and, 39­40 micronutrient supplementation and, 37­38 during pregnancy, 24­26, 25b I primary prevention of, 43 infrastructure IAP. See indoor air pollution child survival strategies and, 9, 39­43 IMCI. See Integrated Management of developing countries' programs, 94­96, 109, Childhood Illness 113 immunity, 20, 20f public health agenda and, 89 immunizations, 36­37, 106, 112 insecticide-treated nets (ITNs), 38­39, See also vaccinations 96, 113 income losses, 68, 77, 165, 165t, 180, 181 institutional capacity, 80 India Integrated Child Development Services birth weight in, 54b (India), 91, 93 environmental health functions in, 101b, Integrated Community-Based Child Care 101­2 Program (Honduras), 91 indoor air pollution in, 42 Integrated Early Childhood Development Integrated Child Development Services, 91, Project (Eritrea), 91 93 Integrated Management of Childhood Illness malaria in, 107 (IMCI), 32, 90­91 malnutrition in, 6 intersectoral and interjurisdictional nutritional programs in, 91 coordination, 13, 104­5, 107, 108­9, oral rehydration therapy in, 35 114­15 sanitation in, 95, 99, 107, 110n1 interventions Indonesia, 37, 134t burdens of disease and, 48 indoor air pollution (IAP) childhood growth and, 23 ALRI and, 1, 42, 50, 57, 75b, 174­77 costs and benefits of, 80 burden of disease, 49, 52 customizing of, 11 clinical studies of, 58 by health ministries, 89 cookstoves and, 42­43, 95­96 in history of environmental health, 86­87 environmental risk factors of, 6, 61 multiplier effect for environmental health, 55­56 malaria and, 39 intrauterine growth restriction (IUGR), 7 mortality, 59n7 invisible regulator, 87­88 pregnancy and, 26 ITNs. See insecticide-treated nets related diseases, 49t, 174t IUGR (intrauterine growth restriction), 7 indoor residual spraying, 28, 38, 57, 96, 113 infant care programs, 35­36 J infants and young children See also specific diseases and interventions Jacoby, H., 145 breastfeeding. See breastfeeding Jamaica burden of disease, 48­49, 53f correlation between education deficiency cognition and, 28­29 and stunting, 144 diarrhea and, 22b, 27, 36, 117 helminth infections in, 29 environmental health risks and, 1, 2b, 8, Jonsson, U., 54b 26­28 jurisdictional collaborations. See intersectoral infections in, 20­21 and interjurisdictional coordination 206 INDEX K early infancy and, 26­28, 27f, 30 economic impact of, 81 Kenya education and, 76t, 76­77, 81, 141­46, census omissions in, 54b 142t­143t, 160­70 helminth infections in, 29 environmental factors, 4, 11, 18­19, 30, 50b, malaria and cognitive impairments in, 29 50­51, 159t King, E., 145 environmental health valuations and, 9­10, 63, 66­78, 68t, 74f, 74t, 76t, 81 L environmental role in childhood health, Latin America 23­28, 24f See also specific countries infections and, 8, 19­23, 30, 51, 126t­140t community health approach in, 109 infections affecting malnutrition, 20­23, 21t, 118, 123­24 malaria in, 6 malnutrition effecting disease, 19­20, 20f malnutrition in, 27f mortality and, 1, 70t, 75b latrines. See sanitation pregnancy and, 24­26 lead-related diseases, 49t, 174t Maluccio, J., 144 linear growth and growth faltering measles, 21, 21t, 36, 51, 57, 92b, 159t diarrhea and, 22b­22b Mediterranean (region), 17, 30, 86, 111 infections and, 21, 23, 117, 119­21 Mehta, S., 67 rates of, 123­24, 162t Mexico retardation factor, 160­62, 162t Clean Water Program, 92b local government roles, 106­9 environmental health programs in, 91 Lorntz, B., 145 nutritional status and infectious disease M in, 127t Universal Vaccination Program, 92b malabsorption, 37, 117, 118, 119, 120 MDGs. See Millennium Development Goals malaria miasma theory, 86 burden of disease, 148­49t micronutrient supplementation, 20, 31, 37­38, cognition and, 29 112 control strategies, 38­39, 96 Middle East, malaria in, 6 drug resistance and, 35 milk, 86, 87b environmental risk factors and, 6, 51, 112 See also breastfeeding health systems and, 34t Millennium Development Goals (MDGs), mortality, 1­2, 38, 57­58, 59n5, 96, 159t 2, 3t, 4, 93, 101, 113 pregnancy and, 24­26 Mills, A., 57 treatment and prevention, 28, 37 Mills-Reincke phenomenon and ratios, 55f, young children and, 49 56b, 56­57 Malawi mobilization, 107 child survival programs in, 91 Moore, S., 21 nutritional status and infectious disease in, Morocco, costs of environmental degradation 139t in water sector in, 62 strength of association study in, 119 mortality Malaysia See also attributable fractions (AF) infant mortality in, 28 diarrheal, 1, 55f, 55­56, 56b, 57, 59 malaria in, 96 indoor air pollution, 59n7 malnutrition, 17­43, 111­13, 119 malaria and, 1­2, 38, 57­58, 59n5, 96 ALRI and, 17, 19 malnutrition and, 1, 17, 19, 51, 70t, 75b areas for future research, 58 multiplier effect and, 55, 112 background on, 6­7 valuation of, 62­63, 151­59, 152f, 152t burden of disease, 148­49t mosquito nets. See insecticide-treated nets cognition and learning impacts, 28­30, 51, mosquitoes. See vector controls 63­64, 68, 112 Moy, R., 120 diarrhea and, 22b­23b, 68 multiplier effects, 12, 55­56, 112 INDEX 207 multisectoral collaborations. See intersectoral mortality valuation in, 181 and interjurisdictional coordination nutritional status and infectious disease in, Multisectoral Constraints Assessment, 137t 31­32 sanitation campaigns in, 95 school enrollment in, 143t, 145, 146 N United Nations Children's Fund, 70f Pakistan Demographic and Health Survey national government roles, 104­6 1990/91, 82n4 National Improved Stove Program (NISP, Pakistan Social and Living Standards China), 42, 98 Measurement Survey 2004-05, 66 National Nutrition Survey 2001­2002 Pan-American Health Organization, 109 (Pakistan), 77 parasites. See helminth infections National Slum Dwellers Federation (India), 99 Pasteur, Louis, 86 Nepal Patrinos, H., 76­77, 166 Alternative Energy Promotion Center, 100 Pediatric Nutrition Handbook (American child survival programs in, 91 Academy of Pediatrics), 122 hookworm infections in, 26 Peru national hand-washing campaign in, 106 cost-benefit analysis in, 114 smoke exposure of women in, 42 cost of environmental health risks in, 63f New York City milk regulation, 87b diarrhea in, 21 Nicaragua hygiene in, 81, 100, 106 integrated child survival programs in, infrastructure improvements in, 94 90­91 Integrated Management of Childhood intestinal parasites in, 29 Illness in, 90­91 vector-control measures in, 91 nutritional status and infectious disease in, Nigeria 134t, 135t, 138t cost of environmental health risks in, 63f school performance in, 29 indoor air pollution in, 62 Philippines nutritional status and infectious disease in, infant mortality causes in, 28 126t Interagency Committee on Environmental strength of association study in, 119 Health, 105 Total Sanitation Campaign in, 95 malnutrition and educational outcomes in, nutrition programs, 91­94, 109, 113 64, 142­43t, 144­45, 163­64, 171n12 nutritional status and infectious disease in, O 132t pneumococcal infections. See pneumonia occupational risk factors, 49t, 174t pneumonia, 21, 37 oral rehydration therapy, 35, 92, 114 political will and consensus, 11, 98 outdoor air pollution, 49t, 174t Powanda, M.C., 121 Prado, M., 121 P pregnancy Pakistan See also fetuses ALRI in, 57, 66­67, 158, 159f environmental risk factors during, 58, 59, basic indicators for, 65b, 66 112 childhood weight in, 69, 70f, 71f, 71­72, 73f, fetal protection, 24­26 82n4, 124 health systems and, 32, 35 cost of environmental health risks in, 63f malaria and, 6 diarrhea in, 69, 71f, 71­72, 73f Preventing Disease through Healthy environmental health valuations in, 9, 61, Environments (Prüss-Üstün & Corvalán), 64­80, 78f, 81, 112, 168t­169t 49, 50, 50b fuels in, 66­67 prevention. See primary prevention GDP in, 9­10, 10t, 78, 113 primary prevention, 31, 35, 43, 89, 90, 112 malaria in, 67 private sector, 11, 99­100, 110, 113 malnutrition in, 6, 70t, 170 Prüss-Üstün, A., 49, 67, 177n2 208 INDEX Psacharopoulos, G., 76­77, 166 Snow, John, 86 public health personnel, 108 soap. See hygiene; handwashing public sectors, 11, 99­100, 110, 113 South Africa, environmental health officials in, Pubmed, 118 108 spraying. See indoor residual spraying R stoves. See cookstoves strategies. See child survival strategies Ramalingaswami, V., 54b strength of association studies, 118­19 regulations stunting. See birth weight; linear growth and Clean Water Program in Mexico, 92b growth faltering; underweight colonial history of, 89 Sub-Saharan Africa health sector, 108­9, 114, 115 See also specific countries interjurisdictional coordination, 105 birth weight in, 9, 54b local government, 106­7, 108 burden of disease in, 51­52, 54, 54t, 59 national government framework for, 104 education in, 77, 166 need to develop, 13 environmental health interventions in, New York City milk regulation, 87b 11­12 sanitation in Ethiopia, 102b hookworms in, 113 trend of enacting, 11, 87­88 immunization in, 36 Report of the Royal Sanitary Commission malaria in, 6, 25­26, 27f, 38­39, 49, 57­58, (UK), 86 113 Repositioning Nutrition as Central to malnutrition in, 1, 6, 17, 30, 111, 119 Development (World Bank), 4 vector-control programs in, 96 residual spraying. See indoor residual spraying weight of children in, compared to Asian residual transmission. See transmission routes children, 54b retardation factor. See linear growth and Sudan growth faltering; weight retardation nutritional status and infectious disease in, factors 129t, 132t, 133t Rohde, J., 54b water and food budgets in, 58 rotavirus, 37 supplements. See micronutrient supplementation S St. Lucia, 120, 127t T sanitation Taylor, C.E., 19, 22b, 117, 121 burden of disease, 49, 147­49, 148t­149t Teklemariam, Shiferaw, 98 child delivery and, 36 tetanus, 36 diarrhea and, 27­28, 38 Thailand environmental risk factors and, 5, 61, 112 community empowerment in, 98­99 hookworm infections and, 25, 26 nutrition programs in, 91, 93, 93b infrastructure improvements to, 40­41, Total Sanitation Campaign (Bangladesh), 90, 43n3, 94­95, 95b 94­95, 95b, 98 maternal and childhood underweight, transmission routes, 18t, 18­19, 19f 50, 50b trypanosomiasis, 43n1 as Millennium Development Goal, 101 Tunisia related diseases, 49t, 174t cost of environmental health risks in, 63f in rural areas, 87, 99 plan to develop unit of environmental as second generation health intervention, economics and policy, 62 86­87 typhoid fever, 56b vaccines and, 37 school performance. See cognition; education U Scrimshaw, N.S., 19, 22b, 117, 121 Senegal, 100, 106 underweight Silent Spring (Carson), 88 See also birth weight Smith, K., 67 defined, 7 INDEX 209 diarrhea and, 50b, 118, 125n4, 158, 158t during pregnancy, 25b factors for, 157, 157t privatization of supply, 99 See also weight retardation factors related diseases, 49t, 174t infections and, 118 vaccines and, 37 rates of, 158, 158t weight. See birth weight; underweight United Nations Children's Fund (UNICEF), weight retardation factors, 157t, 157­58, 90, 109 158t, 171n7 United States WHO. See World Health Organization correlation between diarrheal disease and whooping cough, 36 malnutrition in, 57 willingness to pay (WTP) approach, 180 devolution of authority in, 105 window of opportunity, 8, 25b, 26­28, rural sanitation initiatives in, 87 27f, 30, 111 Universal Vaccination Program (Mexico), 92b World Bank environmental degradation valuation, 61 V Health, Nutrition, and Population strategy, 31­32 vaccinations, 31, 36­37, 86­87, 92b Health Sector Strategy, 109 See also immunizations nutrition studies, 111 Valdmanis, V., 18 Repositioning Nutrition as Central to valuations. See environmental health Development, 4 valuations World Health Organization (WHO) van Geertruyden, J., 59n5 antihelminthics recommended by, 122 vector control, 91, 96, 97b, 112 burden of disease studies, 48, 51, See also dengue virus 61, 109, 124, 147, 148t­149t Vietnam, vector-control programs in, 96, Comparative Risk Assessment, 48­49, 49t, 97b, 99 173, 174t vitamins and vitamin deficiencies. See drug resistance studies, 35 micronutrient supplementation environmental health defined by, 2b environmental risk factor studies, 4 W hygiene assessment, 30n1 wage losses. See income losses infection studies, 117 Walker, S.P., 144 Integrated Management of Childhood wastewater treatment, 40, 41 Illness (IMCI), 90­91 wasting. See birth weight; linear growth and malnutrition studies, 9 growth faltering; underweight underweight assessment, 23b water World Health Report (WHO), 48 burden of disease, 49, 53f, 147­49, worms. See deworming; helminth infections 148t­149t WSH. See hygiene; sanitation; water child delivery and, 36 WTP (willingness to pay approach), 180 Clean Water Program (Mexico), 92b cost of, 58 Y diarrhea and, 38, 56b young children. See infants and young children environmental risk factors of, 5, 61 fecal pollution and, 28 Z hookworm infections and, 26 improvement of supply, 39­40, Zimbabwe 41, 43n2 indoor air pollution in, 26 infrastructure improvements to, 94 malnutrition and educational outcomes in, malaria and, 25 64, 144, 163­64 maternal and childhood underweight, nutritional status and infectious disease in, 50, 50b 131­32t ECO-AUDIT Environmental Benefits Statement The World Bank is committed to preserving Saved: endangered forests and natural resources. The · 48 trees Office of the Publisher has chosen to print · 33 million BTUs of Environmental Health and Child Survival: total energy Epidemiology, Economics, Experiences on · 4,216 lbs. of CO2 recycled paper with 100 percent postconsumer equivalent of green- fiber in accordance with the recommended house gases standards for paper usage set by the Green Press · 20,650 gallons of waste Initiative, a nonprofit program supporting pub- water lishers in using fiber that is not sourced from · 2,247 pounds of solid endangered forests. For more information, visit waste www.greenpressinitiative.org. "This rigorous study is a godsend to anyone involved in advocacy for water and sanitation in developing countries. Until now, proving that environmental health measures make good sense economically has been a tricky business. Now this rigorous and detailed study shows that inadequate environmental health has huge costs to the economy (about 9 percent of the GDP of typical developing countries) in addition to the pain and suffering they cause. For politicians who are unmoved by arguments that failure to invest in water and sanitation will make their people poor, this study offers a clincher; it shows how lack of investment will also negatively affect their children's educational and cognitive performance, because of the effects of malnutrition, exacerbated by frequent episodes of illness." Sandy Cairncross Professor of Environmental Health London School of Hygiene and Tropical Medicine "Recent research has highlighted once more that improved environmental services and conditions can play an essential role in reducing child mortality. Additional insights in the economic evaluation at country level therefore provide timely information for the design of policy action." Maria Neira Director, Department of Public Health and Environment The World Health Organization "Child survival is probably the most important and the most elusive development goal, particularly in Sub-Saharan Africa and South Asia. One reason is that it has traditionally been treated only as a health issue, whereas--as this book clearly shows--the determinants often lie outside health, in water, sanitation, and other environmental factors. To treat this multi-sectoral problem, we need a multi-sectoral effort. This book provides guidance and inspiration for that effort." Shantayanan Devarajan Chief Economist, Africa Region (former Chief Economist, South Asia Region) The World Bank ISBN 978-0-8213-7236-4 SKU 17236