CHILD STUNTING, HIDDEN HUNGER and HUMAN CAPITAL in SOUTH ASIA Implications for Sustainable Development Post 2015 Contributors

CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA Implications for sustainable development post 2015 Contributors:

Aguayo, Victor M.: United Nations Children's Fund (UNICEF), New York, USA.

Corsi, Daniel: Harvard Center for Population and Development Studies, Cambridge, MA, USA.*

Harding, Kassandra L.: Friedman School of Nutrition Science and Policy, Tufts University, Boston, MA, USA.

Kim, Rockli: Harvard Center for Population and Development Studies, Cambridge, MA, USA.

Krishna, Aditi: Harvard Center for Population and Development Studies, Cambridge, MA, USA.*

McGovern, Mark E.: Harvard Center for Population and Development Studies, Cambridge, MA, USA.*

Mejía-Guevara, Iván: Harvard Center for Population and Development Studies, Cambridge, MA, USA.*

Perkins, Jessica M.: Harvard Center for Population and Development Studies, Cambridge, MA, USA.*

Rasheed, Rishfa: South Asian Association for Regional Cooperation (SAARC) Secretariat, Kathmandu, Nepal.

Subramanian, S.V.: Harvard Center for Population and Development Studies, Cambridge, MA, USA.

Torlesse, Harriet: UNICEF Regional Office for South Asia (ROSA), Kathmandu, Nepal.

Webb, Patrick: Friedman School of Nutrition Science and Policy, Tufts University, Boston, MA, USA.

*Contributor’s affiliation at the start of the research and/or when the bulk of the research was conducted

The findings, interpretations and conclusions expressed in this report are those of the authors and do not necessarily reflect the policies and views of UNICEF.

Permission to copy, disseminate or otherwise use the information in this report is granted as long as appropriate acknowledgement is given.

Suggested citation: UNICEF (2018). Child Stunting, Hidden Hunger and Human Capital in South Asia: Implications for Sustainable Development Post 2015. UNICEF: Kathmandu, Nepal.

Design by: EKbana Solutions

Photograph credits: Cover page: © UNICEF/2010/Pirozzi Pages vi: © UNICEF/2014/Qadri Page x, 21, 22: © UNICEF/2016/Giacomo Page 36: © UNICEF/2017/Brown CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA Implications for sustainable development post 2015

CONTENTS

Forward V Executive summary VII Child stunting and human capital in South Asia 1 Introduction 1 Child stunting and human capital in South Asia: a review of the evidence 3 Child stunting and national development: a review of the evidence 7 Trends and inequalities in child stunting in South Asia 9 Drivers of child stunting in South Asia 14 Implications for action 19

Hidden hunger and human capital in South Asia 23 Introduction 23 Hidden malnutrition and the development of nations: a review of evidence 24 Women’s education and hidden malnutrition: a review of evidence 25 Trends in hidden malnutrition in South Asia 28 Drivers of hidden malnutrition in South Asia: a focus on anaemia 31 Implications for action 33 References 37

LIST OF FIGURES

Figure 1. Summary of the pathways linking stunting to economic growth 10 Figure 2. Prevalence of stunting (95% confidence intervals) by dietary diversity score groups (limited data for India and Pakistan) 12 Figure 3. Prevalence of stunting (95% confidence intervals) by mother’s education 12 Figure 4. Prevalence of stunting (95% confidence intervals) by household wealth quintile 13 Figure 5. Fully adjusted odds ratios and 95% confidence intervals for risk factors of stunting among children aged 6-8 months in South Asia (pooled sample) 16 Figure 6. Fully adjusted odds ratios and 95% confidence intervals for risk factors of stunting among children aged 6-23 months in South Asia (pooled sample) 17 Figure 7. Prevalence of micronutrient deficiencies by average years of women’s schooling at the country level 26 Figure 8. Prevalence of micronutrient deficiencies globally and in South Asia and Sub-Saharan Africa. 28 Figure 9. Prevalence of anaemia among women of reproductive age across time (1990 to 2012), by country 29 Figure 10. Prevalence of vitamin A deficiency (VAD) among children 6-59 months old across time (1991 to 2013), by country 30 Figure 11. Median urinary iodine excretion (µg/L) by country (2015) 31

III ACRONYMS

AARR Average annual rate of reduction BMI Body mass index CI Confidence interval DHS Demographic and Health Survey GDP Gross domestic product HAZ Height for age z-score IFPRI International Food Policy Research Institute IQ Intelligence quotient LMIC Low and middle income countries OR Odds ratio SD Standard deviation SDG Sustainable Development Goal SUN Scaling Up Nutrition UIE Urinary iodine excretion UNICEF United Nations Children’s Fund VAD Vitamin A deficiency WHO World Health Organization YLD Years lived with disability

IV CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA FOREWORD

South Asia’s policy decision makers today have vital choices for the economic growth and prosperity of the region. With inequality rising, the demographic dividend diminishing and middle income traps threatening, the social and economic options have never been more daunting, or more promising. Today, the most prominent source of economic growth lies in cognitive and human capital.

Cognitive and human capital cannot be mined or traded, but rather must be carefully cultivated by forward-looking policies. Investments in children, particularly in the earliest years, yield dividends that help pave the way to an economic prosperity characterized by the achievement of cognitive and human capital.

Events in early life influence cognitive and human capital. During the first one thousand days between conception and a child’s second birthday, the brain is developing at an enormous rate. Vitamins, minerals and other nutrients are essential for healthy brain development. If they are in short supply to the mother during pregnancy or the child following birth, there can be considerable consequences for growth, brain development, learning and productivity.

This regional report raises a serious concern that poor nutrition in early life is very common in South Asia and is holding back children, their families and nations from prosperity. The region has a higher percentage (35%) and number (59.4 million) of stunted children than any other region in the world. Stunted children are short for their age, and stunting signals that they have been deprived of nutrients for linear growth and for the growth of essential organs, including the brain. There has also been insufficient progress in reducing vitamin and mineral deficiencies (known as ‘hidden hunger’) in women and children in the region.

Poor nutrition in early life is not only very common, it is also extremely costly. The annual economic losses from low weight, poor growth, and vitamin and mineral deficiencies average 11% of GDP in Asia and Africa. Yet, the returns on investments to prevent stunting and hidden hunger are considerable. Every one centimetre increase in height is associated with a 4% increase in wages for men and 6% increase in wages for women, and every US$ 1 invested in interventions to reduce stunting will have an average return of US$ 18.

These findings highlight that investments in cognitive and human capital must begin in early life if children are to achieve their learning potential and benefit from education and training opportunities in later life. These investments should address the main drivers of stunted growth and hidden hunger, including the poor nutritional status of adolescent girls and women, both before and during pregnancy, poor diets of children in the first two years of life, and poor sanitation and hygiene. Cost-effective interventions are available but they need to reach all children, particularly those who are most marginalized and disadvantaged, as articulated in SAARC’s South Asia Action Framework for Nutrition.

Amjad Hussain B. Sial Jean Gough Secretary General Regional Director South Asian Association for Regional Cooperation UNICEF Regional Office for South Asia

V VI CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA EXECUTIVE SUMMARY

Harriet Torlesse, Rishfa Rasheed, and Victor M. Aguayo

Introduction and burden (59.4 million) of stunted children of all regions in the world (UNICEF et al., 2018) and where The early years of a child’s life provide the best micronutrient deficiencies, known colloquially as opportunity to nourish physical and brain development. hidden hunger, continue to be a public health concern. There is mounting evidence that poor nutrition in early life can have long lasting consequences for learning The aim of this report is to review the evidence on the and future productivity (Victora et al., 2008). When a human and economic impacts of stunting and hidden girl or boy becomes stunted, it signals that they have hunger and the implications for South Asia. Chapter 1 been deprived of nutrients for linear growth and for the reviews the current evidence linking stunting with child growth of essential organs, including the brain. These development, economic outcomes in adulthood and links explain why nutrition has the potential to transform economic growth in low- and middle-income countries. the lives of the world’s most vulnerable citizens and why It then examines where to focus policy and programme it is so central to the achievement of the Sustainable attention in South Asia by exploring recent temporal Development Goals (Development Initiatives, 2017). trends, socioeconomic inequities and predictors of stunting in children aged 6- 23 months in the five Policy makers require evidence on the impact of poor South Asian countries that account for over 95% of the nutrition on human capital and the long-term benefits region’s stunting burden: Afghanistan, Bangladesh, associated in investments to improve nutrition so that India, Nepal and Pakistan. Chapter 2 presents an they can take informed decisions on the allocation of overview of the evidence on the human and economic resources. Nowhere is this more crucial than South impacts of micronutrient deficiencies, and recent Asia, which bears the highest prevalence (35.0%) patterns and trends in micronutrient deficiencies across

VII South Asia. It draws on the most recent surveys and over time and in some cases increased for the most datasets to highlight where gains have been made and deprived children. In the pooled analysis for the five where a lack of progress continues to be of concern countries, short maternal stature and the delayed introduction of complementary foods are associated Key findings with a significantly higher odds of stunting in infants aged 6-8 months. In children aged 6-23 months, short There is considerable evidence from many low- and maternal stature, low household wealth, maternal middle-income countries that support a plausible underweight, failure to meet children’s minimum link between stunting and poor child development. dietary diversity, lack of maternal education and The effect sizes vary by development domain, with mother’s low age at marriage were the strongest greater effects indicated for cognitive and schooling correlates of child stunting. In addition, children who outcomes. There is also evidence of positive are not fully vaccinated or are not fed with a minimum synergistic interactions between nutrition and early frequency are at a higher risk of being severely stunted. stimulation on child development. These effects on child development in early life appear to have long- In addition, there has been little progress across South lasting economic impacts. For example, two long-run Asia in resolving hidden malnutrition. Several countries evaluations found that child nutrition interventions have seriously high levels of micronutrient deficiencies increased adult wages by 25% (Hoddinott et al., 2008) and others have pockets of unusual severity. The and 46% (Gertler et al., 2014), while a one centimetre region has the highest burden of vitamin A deficiency increase in stature is associated with a 4% increase in in children and anaemia in children and women in the wages for men and 6% increase in wages for women world (Stevens et al., 2013, Stevens et al., 2015), and (McGovern et al., 2017). the per capita supply of zinc is highly inadequate (Marc et al., 2016). The underlying and proximate drivers of The most common micronutrient deficiencies in South deficiencies are numerous, and their respective roles Asia – iron, vitamin A, iodine and zinc – cause so vary by country context. However, lower household much impairment to intellectual development, physical wealth, lower maternal education, pregnancy status growth, health and productivity that their hidden and increasing parity, and poor quality diets are impacts on the economy can also be considerable. common drivers of anaemia in women. Iron-deficiency anaemia costs low-income countries an estimated 4% of national income (Horton & Ross, Implications for South Asia 2003), while low vitamin A status is associated with an annual loss of 1% of gross national product South Asia’s insufficient progress in resolving stunting (Meenakshi et al., 2010). More often than not, multiple and hidden hunger affects the survival, growth, micronutrient deficiencies combine together in the development and future livelihoods of children, same individual, and with other nutrition deficits. and continues to hamper the region’s attempts to Anemia and iodine deficiency was associated with accelerate human capital formation. Improvements a 1% annual fall in national economic growth rates in child linear growth and micronutrient nutrition will (Madsen, 2016), and annual economic losses from low allow young children to reach their full growth and weight, poor growth, and micronutrient deficiencies development potential and expand their economic average 11% of GDP in Asia and Africa (IFPRI, 2016). opportunities as adults, generating substantial economic returns at the individual and society level, as The returns on investments to prevent stunting and well as significant increases in economic growth. micronutrient deficiencies are considerable. An analysis of country-specific cost-benefit ratios for Preventing child stunting and micronutrient interventions that reduce stunting had a mean value of deficiencies should be moved to the forefront of 1:18 (Hoddinott et al., 2013a), while a separate study advocacy, policy, programme, and research agendas in found that a combination of investments to prevent the region. This renewed attention should focus on the micronutrient deficiencies could generate annual following priorities: benefits of US$ 15.3 billion at a cost:benefit ratio of 1:13 (FAO, 2013). South Asian countries need to universalize This body of evidence has immense implications for evidence-based policies and programmes South Asia. Although stunting has declined in the designed to improve the nutrition situation of region, the current prevalence remains unacceptably young children and their mothers as a national high. Socio-economic disparities have not declined development priority. Nutrition investments that

VIII CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA focus on adolescence and the 1000 days from complementary foods and feeding, care and hygiene conception to age two years are cost-beneficial. practices, adolescent girls’ and maternal nutrition, Research evidence indicates that packages that micronutrient supplementation, food fortification, combine context-specific nutrition, care, and and the prevention and treatment of severe acute stimulation interventions may bring about the malnutrition and infectious diseases, should be greatest impact on stunting. combined with nutrition-sensitive programmes that address the more distal determinants of Greater attention needs to be paid to ensure undernutrition through agriculture, water, sanitation that economic growth reduces persistent and hygiene, education and social protection. inequities and addresses the drivers of child stunting and hidden hunger among the most There is a need for research and programmatic vulnerable population groups. There is increasing evidence to better understand how to drive consensus that economic growth will only be faster and larger declines in stunting and hidden effective in addressing child stunting if increases in hunger in South Asian countries, particularly national income are directed at improving the diets among the most vulnerable children: the of children, strengthening the status of women, youngest, poorest, and the socially excluded. addressing inequities, and reducing poverty. Understanding how best to scale up programmes to improve children’s growth and development is Nutrition-specific and nutrition-sensitive key area of further inquiry. programmes should be implemented jointly to address the main drivers of stunted growth and hidden hunger in South Asian children. Nutrition- specific interventions, such as the protection, promotion and support of breastfeeding, appropriate

IX X CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA CHILD STUNTING AND HUMAN CAPITAL IN SOUTH ASIA

Rockli Kim, Aditi Krishna, Mark E. McGovern, Iván Mejía-Guevara, Jessica M. Perkins, Daniel Corsi, S.V. Subramanian, and Victor M. Aguayo

Introduction Özaltin et al., 2010), and increased risk of chronic diseases, including elevated blood pressure, renal In 2017, there were 151 million stunted children dysfunction and altered glucose metabolism (Huxley worldwide. Importantly, about 24% of children younger et al., 2000; Victora et al., 2008; Whincup et al., than five years in low and middle income countries 2008), especially when accompanied by rapid weight (LMICs) had stunted growth (UNICEF et al., 2018). gain in later childhood (Gluckman et al., 2007). In Stunting is a linear growth failure that has severe short- addition, evidence shows that stunting has functional term and long-term consequences (Aguayo & Menon, consequences that hamper the development of entire 2016; Onis & Branca, 2016). For instance, stunting is societies. Stunting in early childhood is associated with associated with increased morbidity and mortality from poor cognition (Prendergast & Humphrey, 2014) and infections, in particular pneumonia and diarrhea (Black worse educational performance in school-age children et al., 2008; Kossmann et al., 2000; Olofin et al., 2013), (Adair et al., 2013; Martorell et al., 2010a), as well as and is the cause of about one million child deaths lower productivity and reduced earnings in adult men annually (Aguayo & Menon, 2016). Furthermore, and women (Hoddinott et al., 2013a). growth failure in infancy and early childhood is associated with reduced stature in adulthood (Coly et To understand the full impact of stunting in early al., 2006; Stein et al., 2010), increased risk of maternal, childhood, we review current evidence linking stunting perinatal and neonatal mortality (Lawn et al., 2005; to child development and individual and national

1 economic outcomes. In Section 1, we summarize we provide a comprehensive analysis of social and the evidence on stunting and child development in economic inequalities in stunting prevalence, within low- and middle-income countries and discuss the and among South Asian countries and stunting rates issues for further research. We focus on studies over time. We also investigate inequalities in stunting measuring low height-for-age caused by experiences due to multiple dimensions of deprivation (poor child of chronic nutritional deprivation among children less diets, low maternal education, and household poverty). than 5 years old as the exposure and gross/fine motor Based on a time-series analysis using multiple, cross- skills, psychosocial competencies, cognitive abilities, sectional Demographic and Health Surveys (DHS) on or schooling and learning milestones as outcomes. children aged 6 to 23 months in Bangladesh, India, This review highlights three key findings. First, the Nepal, and Pakistan from 1991-2014, we find that all variability in child development tools and metrics used countries had markedly high stunting prevalences and among studies and the differences in the timing and that, although stunting rates declined in all countries, frequency of the assessments complicate comparisons Bangladesh and Nepal recorded the largest reductions. across study findings. Second, despite methodological Moreover, we find that there were heterogeneous differences and differential associations, considerable changes over time between countries and among evidence from across many countries supports groups defined by multiple dimensions of deprivation an association between stunting and poor child within countries. For instance, there was greater development. Effect sizes differ by development domain stunting among children who had poorer diets, had with greater effects shown for cognitive/schooling mothers with low educational attainment, and/or lived outcomes. How stunting influences child development, in poor households within all South Asian countries. which domains of child development are more affected, Although there were marked declines in stunting in and how the various domains of child development the most deprived groups in some countries, socio- influence one another require further research. Finally, economic differences in the prevalence of stunting there is some evidence of positive synergistic interaction were largely preserved over time, and in some cases between nutrition and stimulation on child development. worsened, namely among wealth quintiles. However, understanding how best to improve child developmental outcomes – either through nutrition In Section four, we assess the relative importance of 13 programs or integrated nutrition and psychosocial correlates of child stunting and severe stunting selected stimulation – is a key area of further inquiry. based on a collective review of existing multi-factorial frameworks: breastfeeding, complementary feeding Focusing on the costs of stunting to children and (timely introduction, feeding frequency and dietary societies supports the case for investing in childhood diversity), maternal height, body mass index (BMI), nutrition. In Section 2, we review the literature on the education, and age at marriage, child vaccination, association between stunting in early childhood and access to improved drinking source and sanitation economic outcomes in adulthood. At the national facilities, household indoor air quality, and household level, we also evaluate the evidence linking stunting wealth. We use the most recent DHS from Bangladesh to economic growth. Two long-run evaluations of (2014), India (2006), Nepal (2011), and Pakistan (2013), randomized childhood nutrition interventions indicate and the National Nutrition Survey from Afghanistan substantial returns to the programs (a 25% and 46% (2013), and find that in South Asia (combined sample), increase in wages for those affected as children). short maternal stature and delayed introduction of Cost-benefit analyses of nutrition interventions using complementary foods are associated with a significantly calibrated return estimates have been found to range higher risk of stunting in infants aged 6-8 months. In from 3.5:1 to 42.7:1 per child, with a median of children aged 6-23 months, failure to meet children’s 17.9:1. Assessing the wage premium associated with minimum dietary diversity, mothers’ short stature, adult height, we find that the median return to a 1 underweight, lack formal education, or underage at centimetre increase in stature is associated with a 4% marriage, and poor household wealth are statistically increase in wages for men and a 6% increase in wages significant risk factors for child stunting. In addition, for women in the set of studies that have attempted to children who are not fully vaccinated or are not fed address unobserved confounding and measurement with a minimum frequency are at a higher risk of being error. In contrast, the evidence on the impact of severely stunted. Some differences were found in the economic growth on stunting is mixed. relative ordering and statistical significance of the South Asia bears 40 percent of the global burden of correlates in country-specific analyses. child stunting. Hence, the third and fourth sections specifically focus on trends in stunting rates and risk Our findings make important contributions to the existing factors of stunting in South Asia. In Section three, evidence demonstrating that eliminating child stunting

2 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA needs to be moved to the forefront of the advocacy, Identifying peer-review evidence on policy, programmatic and research agendas. In Section stunting and child development 5, we conclude with a summary of the implications for sustainable development (universal education, poverty We conducted a comprehensive review of key reduction, improved equity and shared prosperity), databases in public health, economics, social science, policy formulation, strategy design, programme scale and psychology (PubMed, Web of Science, PsycInfo up, and research post 2015 in South Asia. and Embase) to identify studies on stunting and child development to include in this review. We searched for studies or reviews of studies assessing height (or length) among 0-5 year old children as an exposure Child stunting and human for child development. Height could be measured as capital in South Asia: a review (a) height-for-age, which is a linear growth measure of the evidence standardized into z-scores using international growth standards or (b) stunting, which is a widely used binary indicator of chronic undernutrition in infancy Introduction and early childhood (WHO, 2006). We further required studies and reviews to have focused on at least one Establishing a causal relationship between stunting measure of child development as the outcome. Most and child cognitive development is complicated. child development assessment tools focus on major Stunted children are more likely to grow up in educational milestones, measure a specific individual conditions of overall deprivation (Grantham-McGregor developmental domain (e.g., gross motor skills, et al., 2007) that affect physical growth and child fine motor skills, or psychosocial development), or development, potentially confounding the relationship represent a battery of skills tests. The following search between stunting and child cognition. Yet, if stunting terms were used: cognition, cognitive, awareness, causes developmental deficits, the consequences consciousness, cognitive disorder, intelligence, at the population-level are immense considering the intelligence tests, mental tests, achievement, 151 million children who are stunted (UNICEF et al., achievement tests, school readiness, psychosocial, 2018). Moreover, global development priorities (e.g., behavior, attitude, height, stunting, stunt, height-for- the 2015 Sustainable Development Goals) and experts age, anthropometry, anthropometric, growth, growth in child development have set the stage for further disorders, infant, infancy, child, childhood, adolescent, prioritizing research and interventions to address the adolescence, teenage, youth. Only studies published in global challenge of poor child development (Black et English were included. al., 2016; Britto et al., 2016; Dua et al., 2016; Richter et al., 2016). Therefore, if there is a strong evidence Challenges with child development base linking stunting and child development, this assessments strengthens the case for investing in programmes targeted at reducing the prevalence of stunting. Most child development assessments were designed for use with English-speaking populations in high- We provide a substantive review of the association income countries. Thus, there are challenges with between stunting and child development across using them with children in low- and middle-income the world, citing both correlational evidence as countries (Grantham-McGregor, 1984, 1993). Although well as studies that aim to address causality, many of the metrics have been adapted for local including interventions. Our intention is to create a contexts, they may not be appropriate for children useful repository of information for policymakers, living in poverty and lead to biased testing results practitioners, and program managers, or researchers (Isaacs & Oates, 2008; Prado & Dewey, 2014). In new to these fields, who seek to address stunting and addition, studies on child development often evaluate child development challenges and consequences, learning and schooling outcomes such as grade particularly in low- and middle-income countries. completion or years of schooling in lieu of cognitive Thus, this review summarizes the literature on tests as the more specific tests are difficult to adapt stunting and child development with the objective of and administer in diverse contexts and even more understanding the conceptual relationships between challenging to compare across these different settings. stunting as a measure of nutritional deprivation and Yet, the quality of education, the selection of material child development, which can generally be defined as that is taught, and the amount of learning that occurs the attainment of gross motor and fine motor skills, during years of education, vary both within and psychosocial competencies, and cognitive abilities. between countries (Behrman & Birdsall, 1983).

3 Moreover, the use of different child development tools HAZ at 3 years old or younger and gross motor and metrics across studies and differences in the scores at age 5-8 years and also between HAZ at two timing and frequency of the assessments complicate years old or younger and cognition at age 5-11 years comparisons of findings in the literature (Grantham- (Sudfeld et al., 2015). In addition, evidence across McGregor & Baker-Henningham, 2005; Sudfeld et al., several studies suggested that stunting was associated 2015). In addition, implementation of assessments with poor psychosocial development among children varies in practice. For cognitive assessments, children who were stunted at age 9-24 months. Most studies most often engage in the testing. In contrast, to indicated an association between stunting or child assess motor and psychosocial development, height and child development across each of the parents, teachers, or an observer often assess a developmental domains. (Adair et al., 2013; Alberto child’s skills. Therefore, children’s performance on Camargo-Figuera et al., 2014; Aubuchon-Endsley et these assessments is not only a measure of their al., 2011; Aurino & Burchi, 2014; Berkman et al., 2002; abilities, but also reflects their interactions with the Casale et al., 2014; Chang et al., 2002; Chang et al., test administrator and the testing situation (Isaacs & 2010; Cheung & Ashorn, 2010; Cheung et al., 2008; Oates, 2008). Lastly, there are also differences in how Cheung et al., 2001; Crookston et al., 2013; Hamadani test results are analyzed. Some studies use number et al., 2012; Hamadani et al., 2014; Kuklina et al., 2004, of successes as measures of cognitive ability (e.g., 2006; Lima et al., 2004; Meeks Gardner et al., 1999; number of objects correctly identified by children Mendez & Adair, 1999; Niehaus et al., 2002; Pollitt et taking the Peabody Picture Vocabulary Test) while al., 1993; Sanchez, 2013; Walker et al., 2007a; Walker others transform the raw numbers of successes into et al., 2000; Whaley et al., 1998; Yang et al., 2011). age-standardized z-scores or percentiles (Cheung et al., 2008). Given this background, these differences Quasi-experimental studies in testing approaches and methodology need to be considered when interpreting findings in the literature Quasi-experimental studies have provided some (Cheung et al., 2008). evidence that stunting causes cognitive impairments. These studies used instrumental variables, such as rainfall or food price shocks, or natural experiments in Evidence of a link between stunting the form of social welfare programs (i.e., the Protective and child development Safety Nets Program, the National Rural Employment Guarantee Scheme, or similar initiatives), which are not Cross-sectional observational studies randomly implemented but are plausibly exogenous to stunting and cognitive status. Other quasi-experimental A recent review found that across studies, height-for- studies use data from siblings to account for age z-scores (HAZ) among children three years old or unmeasured confounding at the household level. Most younger were positively associated with gross motor of these studies have found negative effects of stunting scores (Sudfeld et al., 2015). Similarly, HAZ scores on cognitive development with varying effect sizes were positively associated with cognitive scores (Dercon & Porter, 2014; Glewwe et al., 2001; Glewwe & among children 0-23 months old, and, to a lesser King, 2001; Leight et al., 2015; Outes-Leon et al., 2011; extent, also with cognitive scores among children aged Umana-Aponte, 2011). two and older (Sudfeld et al., 2015). Due to differences in how psychosocial development was measured and Experimental studies on social welfare the small number of studies included, generalized findings about the association between HAZ and Some experimental studies assess how program- psychosocial development could not be established. facilitated changes in nutritional status influence child Overall, these studies show a positive relationship development by studying the effects of social welfare between child height and child development. programs, which enroll participants randomly. Two (Abubakar et al., 2008; Avan et al., 2010; Bogale et out of five studies included in this review found that al., 2013; Crookston et al., 2011; Fernald et al., 2006; improvements in cognitive development due to cash Handal et al., 2007; Kariger et al., 2005; Ketema et al., transfer programs were mediated by increases in height 2003; Kordas et al., 2004; Kuklina et al., 2006; Mohd (Fernald et al., 2008; Fernald et al., 2009). In contrast, Nasir et al., 2012; Olney et al., 2009; Olney et al., 2007; the other three studies indicated that cash transfers Siegel et al., 2005; Taneja et al., 2005) had no effect on height, but independently improved Longitudinal observational studies cognition (Fernald & Hidrobo, 2011; Macours et al., 2012; Paxson & Schady, 2010). However, follow-up The same recent review study described above also periods in these studies may not have been long enough reported positive prospective associations between to observe changes in height or stunting status.

4 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA Experimental studies on nutrition found no interaction between the two interventions on supplementation stunting and child development (Chang et al., 2010; Meeks Gardner et al., 1999; Walker et al., 2007a). Only two out of seven studies included in this review found improved motor development among children Potential pathways who received nutrition supplementation and no study found psychosocial development benefits. In contrast, The associations between stunting and child several studies included in this review found that development present in some of the papers we early nutrition supplementation appeared to influence reviewed may be due to a variety of different processes cognitive development (Chang et al., 2010; Grantham- linking the exposure and outcome. Although it is McGregor et al., 1991; Grantham-McGregor et al., a nascent area of work, several studies suggest 1997; Grantham-McGregor et al., 1996; Meeks Gardner various mechanistic pathways, including neurological et al., 1999; Nahar et al., 2012; Pollitt et al., 1993; pathways (Black, 1998; John et al., 2017; Lozoff, Pollitt et al., 1995; Pollitt et al., 1997; Vazir et al., 2012; 2007; Vohr et al., 2017), hormonal (Berger, 2001; Waber et al., 1981; Walker, 2006). A recent review of Le Roith, 1997; van Pareren et al., 2004), functional more than 20 postnatal nutrition intervention studies isolation pathways (Grantham-McGregor et al., 2007), among children under two years of age in low- and stress (Fernald & Grantham-McGregor, 1998; Soeters middle-income countries found small effects on mental & Schols, 2009; Wachs et al., 2013), stigma (Currie development (Larson & Yousafzai, 2017). Although & Vogl, 2013), and pathways related to infectious some evidence suggests that nutrition supplementation diseases (Black et al., 2013; UNICEF, 2013; Walker et improves developmental outcomes, there is significant al., 2007b). The research to date, however, is often heterogeneity in the design of the studies. Moreover, unclear on whether such factors act as mediators or as the effects of supplementation may not be uniform precursors to stunting. and may instead depend on many factors which non- experimental studies have been found to be relevant In addition, there may be dynamic interplays between for child development, such as the timing and duration these processes. For example, impaired motor of the intervention (Martorell, 1995; Pollitt et al., 1995), development may mediate the relationships between child sex (Martorell, 1995; Waber et al., 1981) and stunting and cognitive development (Larson & socioeconomic status (Pollitt, 2009; Pollitt et al., 1993; Yousafzai, 2017). Stunted children with lower motor Pollitt et al., 1995). activity are more likely to be carried by caregivers, further handicapping motor development and Interventions integrating nutrition inhibiting cognitive and psychosocial development supplementation and stimulation attained through independent exploration of environments (Adolph et al., 2003; Kariger et al., 2005; Comprehensive summaries of existing studies that Meeks Gardner et al., 1995; Olney et al., 2007; Siegel examined integrated programs are provided in recent et al., 2005). Greater apathy as well as distress in reviews (Black et al., 2015; Grantham-McGregor stunted children (Grantham-McGregor, 1995; Pollitt, et al., 2014). In addition, one paper reviewed the 2000) increase the likelihood that caregivers treat implementation of integrated nutrition and psychosocial stunted children as if they were younger, resulting in stimulation (Yousafzai & Aboud, 2014) and a more a lack of age-appropriate stimulation (Kuklina et al., recent study reviewed the benefits and challenges of 2004). This evidence supports the idea that children’s implementing integrated interventions to address early motor, psychosocial, and cognitive development occur childhood nutrition and development (Hurley et al., in an interactive and dynamic manner with significant 2016). Of the six studies included in the present review, influence from their social environments (Wachs et al., two found that the combined treatment effects (i.e., 2013). However, the lack of study designs permitting nutrition supplementation plus psychosocial stimulation) rigorous mechanistic study prohibits a definitive were significantly more effective than either alone layout of a complete framework for pathways and for motor skills and cognitive development among mediators. Research on these pathways (as well as children in Bangladesh (Nahar et al., 2012) and Jamaica other potential mechanisms) is still in its early stages, (Grantham-McGregor et al., 1991). However, the additive and future work may be able to shed light on this effect of combined treatments was no longer significant important issue, which will be of particular interest to four years after the end of the study in Jamaica policy makers trying to identify potential interventions (Grantham-McGregor et al., 1997). Other studies designed to improve child development.

5 Further questions about stunting and assessments for gross and fine motor development child development and psychosocial development do not adequately reflect the positive impact of nutrition interventions. Timing of nutrition interventions Alternatively, there may be dynamic interactions not captured by analyses. Advanced statistical techniques Nutrition intervention effects on cognition may be such as structural equation models, which allow for greatest in the first two years of life (Pollitt et al., an investigation into interactions between multiple 1993). This claim is supported by findings from several determinants, may help illuminate the pathways by observational studies (Pongcharoen et al., 2012; Sudfeld which different developmental domains influence one et al., 2015). However, it is difficult to draw strong another and how nutritional status may directly or conclusions on this issue, at least in respect to the pre indirectly influence developmental outcomes. and postnatal periods, as some studies have found either no difference in associations between prenatal Integrating interventions and postnatal linear growth and cognition (Yang et al., 2011) while others have found that postnatal growth Child development occurs in a dynamic fashion in is more salient (Adair et al., 2013; Kuklina et al., 2004). which children’s characteristics interact with the social Children who experience developmental impairments environment to influence their development. Therefore, due to nutritional deprivation in the first 1,000 days may impact of integrated interventions may be stronger be able to recover later on (Levitsky & Strupp, 1995; than any nutrition supplementation alone (Black et Strupp & Levitsky, 1995). Moreover, brain development al., 2015). Although evidence on synergy between continues into adulthood (Thompson & Nelson, 2001) nutrition and stimulation interventions is inconsistent with important periods of growth in adolescence and and inconclusive, some evidence supports the addition early adulthood (Isaacs & Oates, 2008; Wachs et al., of stimulation to nutrition programs to improve child 2013). Furthermore, growth faltering can also happen growth as well as developmental outcomes in the long- after the first two years (Lundeen et al., 2013; Prentice run (Alderman et al., 2014; Black et al., 2015; Christian et al., 2013), suggesting that children may also be et al., 2015; DiGirolamo et al., 2014; Fernandez-Rao et vulnerable to impaired development later in later life al., 2013; Grantham-McGregor et al., 2014). Moreover, (Crookston et al., 2011). while studies have largely focused on integrated interventions that combine macro- or micronutrient Long-run associations and the role of catch- supplementation with psychosocial stimulation to up growth improve child development outcomes, there is little evidence on the interaction between child development The degree to which associations between early and other exposure variables such as sanitation, hygiene, stunting and later cognition are mediated by later access to health care, and early responsive learning growth remains unclear. While some studies find interventions. Thus, there is a need for comprehensive that children who catch-up in growth are also able evaluations of a variety of integrated programs, including to recover in development (Cheung & Ashorn, 2010; those addressing macronutrient deficiencies combined Crookston et al., 2010a; Crookston et al., 2010b; with child stimulation and hygiene and sanitation Crookston et al., 2013; Gandhi et al., 2011; Mendez intervention (Larson & Yousafzai, 2017), and especially & Adair, 1999), other work finds that catch-up concerning the long-term sustainability of benefits. growth does not help children recover cognitive Despite inconsistent evidence on synergistic interaction deficits (Sokolovic et al., 2014). Moreover, there is no between nutrition and stimulation on child development, conclusive evidence about the best timing for catch-up simultaneous implementation of cross-functional policies growth (Cheung & Ashorn, 2010; Gandhi et al., 2011). and programs at the population level aimed at improving Further research is required to resolve the ambiguity food security, reducing poverty and social inequalities, about critical windows for physical and cognitive and improving maternal education has exhibited success development as well as the role of catch-up growth. in reducing undernutrition in some countries in South East Asia and Latin America (DiGirolamo et al., 2014). Differential effects on development domains Finally, a recent review concluded that multi-sectoral Many of the nutrition studies report improvements in approaches combining nutrition, health, education, child cognitive development but no impact on gross or fine protection, and social protection are critical for building motor development or psychosocial development. successful and sustainable interventions, and that While it is possible that there are heterogeneous families and caregivers should be supported in providing effects which differ by domain, perhaps current nurturing care (Britto et al., 2016).

6 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA Conclusion interventions targeted at preventing stunting and assist in identifying priorities for national and international Our comprehensive review of stunting and child efforts (Halim et al., 2015). Despite the potential development has three salient findings. First, there is substantial returns of programs aimed at reducing considerable heterogeneity in the methodology used stunting (Gertler et al., 2014; Hoddinott et al., 2008), to examine the relationship between child stunting progress in reducing stunting has been sub-optimal and child development. Such heterogeneity makes in many countries (Stevens et al., 2012). The aim of comparisons of study findings difficult. Creating Section 2 is to provide a comprehensive review of the a harmonized way to measure and evaluate the evidence of the long run economic impact of child relationship between stunting and child development is undernutrition, as measured by stunting. Assessing one solution. Second, there is evidence that stunting is the potential economic benefits of investments in child plausibly linked to poor developmental outcomes among nutrition will inform whether these programs are an children despite variability across studies. How stunting efficient use of public funds. We also consider the influences child development, which domains of child evidence linking child undernutrition and economic development are more affected, and how the various growth at the national level, and conclude with a domains of child development influence one another discussion of the implication of these findings for all require further research. Finally, understanding countries with high stunting prevalence in the context how best to improve developmental outcomes – either of Sustainable Development Goals (SDGs) post-2015. through nutrition programs or integrated nutrition and psychosocial stimulation – is a key area of further Methods inquiry. This process requires a better understanding of the complex relationship between stunting and the We conducted a literature search for published studies different domains of child development and scaling up up to July 2015 which examined whether childhood proven interventions to target these multiple, interacting stunting was associated with economic outcomes in factors. Improving multiple aspects of the early life later life. We used the Pubmed and Econlit databases environment that are critical for child development may to search for keywords in abstracts and titles related have synergistic effects. However, more research is to the following economic outcomes; wages, needed to rigorously evaluate the effects of these types income, salary, pay, earnings, productivity, capital, of integrated programs, particularly in the long-run. resources, work, employment, industry, hours worked, occupation, labour, sector, job, socioeconomic, Given that nearly 40% of children under age five savings, economic, returns, make ends meet, welfare, suffer from loss of developmental potential - for which poverty; and the following measures of childhood stunting is likely one of the key risk factors - reductions undernutrition; stunting, child development, growth in stunting could have tremendous implications for child retardation, growth trajectory, linear growth, linear development and human capital formation, particularly growth retardation, growth faltering, growth failure, in low- and middle-income countries. If interventions are early life growth failure, child undernutrition, child able to address the etiology of stunting, or preferably malnutrition, child nutrition. the etiologies of stunting and other determinants of poor child development such as poor child stimulation Findings for stunting in early or poverty simultaneously, all of which are concentrated childhood and economic outcomes in in resource-poor settings, they are likely to be more effective in targeting the root causes of developmental adulthood deficits in young children than interventions which focus There are two long term follow ups to nutrition solely on educational outcomes. intervention studies. The first, the Institute of Nutrition of Central America and Panama (INCAP) study (Engle & Fernandez, 2010; Stein et al., 2008), allocated nutritional Child stunting and national supplements to pregnant women and infants in two Guatemalan villages in the late 1960s and 1970s. development: a review of the While not a pure randomized design, two nearby evidence comparison villages did not receive the supplements. Studies which follow the INCAP children into adulthood Introduction find positive effects of nutritional supplementation in early childhood on economic outcomes in adulthood, Assessments of the effects of child stunting on including increased productive capacity, wages, and economic outcomes provide an evidence base for expenditure, and lower probability of living in poverty

7 (Haas et al., 1995; Haas et al., 1996; Hoddinott et al., Brazil and Guatemala, and a 21% increase in household 2005; Martorell et al., 2010b). Intention to treat analysis assets in India (Victora et al., 2008). (ITT) found that men who received the supplements as children had, on average, a 46% increase in wages Cost benefit analysis of childhood (Hoddinott et al., 2008). The association was not found nutrition interventions to be statistically significant for women. An instrumental A number of publications focus on the cost-benefit variables analysis found that a one standard deviation analysis of interventions aiming to reduce child (SD) increase in HAZ at age 2 years was associated with stunting and estimate the economic impact of stunting a subsequent 21% increase in household per capita reduction. A micronutrient supplementation and expenditure and a 10 percentage point decrease in the early childhood stimulation program in Nicaragua probability of reporting living in poverty at ages 25-42 was found to have a cost-benefit ratio of 1.5 (Boo et years (Hoddinott et al., 2013b). A second intervention al., 2014). An intervention in Indonesia was found to study in Jamaica in the mid-1980s randomly allocated have a cost-benefit ratio of 2.08 based on productivity undernourished children to a nutrition supplementation enhancements from reduced undernutrition, earnings intervention, psycho-social stimulation, combined from deaths averted, and household savings from nutrition and stimulation, and a control group (Gertler et health care costs avoided (Qureshy et al., 2013). In a al., 2014). These children were subsequently interviewed review of cost-benefit ratios of nutrition interventions, in 2007-2008 at age 22 years. Those who benefited estimates ranged from 3.5:1 to 42.7:1 per child, with a from stimulation or combined nutrition and stimulation median of 17.9:1 (Hoddinott et al., 2013a). interventions as children were found to have 25% increased earnings relative to the control group. A similar methodology has been used to estimate the aggregate burden of stunting at a national level and Prospective analyses evaluate the association between worldwide. Productivity gains due to reductions in child measured height in childhood and economic measured stunting in China over the period 1991 to 2001 were outcomes in adulthood. Many of these analyses are estimated at US$ 12 billion in 2001 (Ross et al., 2003) based on cohort studies in the UK, Brazil, and the while the productivity-related costs of stunting in Peru Philippines (Adair et al., 2011; Elliott & Shepherd, 2006; were estimated at 2.2% of gross domestic product Power & Elliott, 2006; Victora & Barros, 2006). Only (GDP) (Alcázar et al., 2013). The cost of undernutrition one study in Barbados (1967-1972) includes a matched in Cambodia has been estimated at more than US$ control group and has followed a cohort of children 400 million annually, or approximately 2.5% of yearly who were hospitalized because of undernutrition GDP, of which 57% is a result of undernutrition in early during the first year of life and a matched cohort of childhood (Bagriansky et al., 2014). Based on a model children who did not experience undernutrition (Galler originally developed for Latin America (Fernández & et al., 2012). Children who experienced undernutrition Martínez, 2007), the Cost of Hunger in Africa (COHA) as children were found to be lower on social position study aims to provide economic estimates of the costs and standard of living indices in adulthood. An early of malnutrition in 13 countries. A recent report on study in India found that HAZ at age 5 years was Malawi indicated that the costs associated with stunting associated with lower work capacity at age 14-17 years were of the order of 10% of GDP per year (African Union (Satyanarayana et al., 1978; Satyanarayana et al., 1979). Commission/World Food, 2015). The annual GDP loss In the Philippines, better HAZ at age 2 years was found due to undernutrition has been estimated to be up to to be associated with an increase in the probability 12% in LMICs (Horton & Steckel, 2011). of being engaged in formal work at age 20-22 years, both in men and women (Carba et al., 2009). Research Stunting and economic growth from US and UK cohort studies has found that height at various stages in childhood and adolescence is Individual level estimates of the costs of stunting do positively related to work status, wages, and measures not take into account the potential spillover effects on of socioeconomic status in later life (Case et al., 2005; aggregate capital formation, labour markets, savings Case & Paxson, 2008; Montgomery et al., 1996; Persico and investment behavior, and other factors that make et al., 2004; Sargent & Blanchflower, 1994). In a cohort up the determinants of economic growth. Relevant of men born in Helsinki, Finland, between 1934 and pathways, summarized in Figure 1, include increases 1944, length in the first year of life was associated with in morbidity, mortality and health expenditure, and subsequent earnings (Barker et al., 2005). A review of reductions in human capital investment (for example, three cohort studies in LMICs found that HAZ at age 2 education), physical capital investment, and labour years was associated with an 8% increase in income in supply. Productivity may suffer because of ill health or

8 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA reduced work capacity in adulthood, and reductions of stunting remains high in some regions, and is in human capital and technological progress (due to currently at 38% in South Asia and 37% in sub-Saharan lower cognitive development or levels of educational Africa. Assessing the wage premium associated with attainment and infrastructure). Other channels include adult height, we find that the median return to a 1 the impact of undernutrition on chronic disease, which centimetre increase in stature is associated with a would involve the diversion of productive savings, 4% increase in wages for men and a 6% increase in which boost investment, towards unproductive wages for women in our preferred set of studies which treatment costs (Bloom et al., 2014). While the have attempted to address unobserved confounding available evidence indicates a strong association and measurement error. In contrast, the evidence on between economic growth and proxies for nutritional the impact of economic growth on stunting is mixed. status in childhood (Arora, 2001; Chakraborty et al., Further evidence needs to inform how best to prioritize 2010; Dalgaard & Strulik, 2015; María-Dolores & effective and cost effective interventions to reduce Martínez-Carrión, 2011; Piper, 2014; Weil, 2007), it is stunting in the most adversely affected areas. hard to quantify the causal magnitude of these effects. One recent paper finds that each centimetre increase in height at the population level is expected to raise income per capita by 6% (Akachi & Canning, 2015). Trends and inequalities in child stunting in South Asia There are numerous pathways through which poverty, and therefore income per capita, are expected to Introduction impact the nutritional status of children. For example, capacity to purchase food, raising both the quantity Recognizing the importance of early child nutrition, and the quality of food and nutrient intake by women much global attention has focused on averting child and young children. In addition, net nutritional intake is undernutrition through the Millennium Development also likely to increase because of reductions in the risk Goals, the Scaling Up Nutrition (SUN) movement, and of infection and poor health (Shekar et al., 2006). At the SDGs, which endorse the World Health Assembly’s the national level, education, infrastructure and health target of reducing the number of stunted children services are all positively associated with GDP. However, aged 0-59 months by 40% by 2025 (UN, 2015b). A economic growth may not necessarily translate into key recommendation from these initiatives is the need improvements in household income due to unequal for better quality data, information and analyses to sharing of growth gains and lack of access to public understand who is most vulnerable to stunting (Black and private services (Haddad, 2015; Lawrence et al., et al., 2013). Following on this recommendation, 2015). The impact of a 10% increase in GDP per capita Section 3 seeks to provide a comprehensive in the short-run is generally estimated to be in the range assessment of the patterns in child stunting in South of 0-2 percentage point absolute decrease in stunting Asia with particular attention to temporal trends and prevalence (Bershteyn et al., 2015; Harttgen et al., 2013; variations in the prevalence of stunting among different Headey, 2013; Heltberg, 2009; Ruel et al., 2013; Smith socioeconomic groups. & Haddad, 2002; Vollmer et al., 2014; Webb & Block, 2012), although some estimates, particularly those There is substantial evidence that child stunting follows which focus on long run changes in national income, a socioeconomic gradient (Akhtar, 2015; Di Cesare et are found to be larger (Haddad et al., 2003; Ruel et al., al., 2015; Gaiha & Kulkarni, 2005; Kanjilal et al., 2010; 2013; Smith & Haddad, 2002). In contrast, a recent Kumar & Kumari, 2014; Kumar et al., 2014; Menon, analysis of Indian states found that economic growth 2012). In this section, we look at the patterning of child was not associated with reductions in stunting or other stunting along three domains of deprivation: 1) children’s measures of undernutrition (Subramanyam et al., 2011). access to food, represented by dietary diversity; 2) social, One paper examined whether the timing of economic represented by maternal educational attainment; and 3) growth is important for attained height and found that economic, represented by household wealth. Together, economic growth in the first year of life and puberty these three domains represent three levels of the were most associated with the adult height of women in determinants of child nutrition – child-level, mother-level, sub-Saharan Africa (Moradi, 2010). and household-level – in UNICEF’s conceptual model of nutrition (UNICEF, 2013). While much work considers Conclusion the second two aspects of deprivation (Gaiha & Kulkarni, 2005; Kumar & Kumari, 2014; Kumar et al., 2014), to our Welcome progress has seen stunting fall from 50% in knowledge, only a few studies consider all three domains 1970 to 30% in 2010 in LMICs. However, the prevalence of deprivation in South Asia (Di Cesare et al., 2015;

9 Figure 1. Summary of the pathways linking stunting to economic growth

Stunting

Health Morbidity Mortality Expenditure

Human Capital Labour Physical Capital Investment Supply/Productivity Investment

Economic Growth

Adapted from Bloom et al. (2014)

Kanjilal et al., 2010). We focus on children ages 6-23 standards. Children who were two SD or more below months as the first two years of life are critical for linear median HAZ score were classified as stunted. growth and growth faltering (Victora et al., 2010), and the existing population surveys assess dietary diversity in children aged 6-23 months, after the initial six-month We considered three domains of deprivation as key exclusive breastfeeding period. We identify patterns in explanatory variables: child dietary diversity, mother’s the prevalence of stunting by children’s dietary diversity, education, and household wealth. Dietary diversity mother’s education, and household wealth and examine was measured using a score (Corsi et al., 2016; WHO, temporal changes in the prevalence of stunting in groups 2008; WHO, 2010) ranging from 0 to 7, based on the defined by these social and economic dimensions of number of food groups a child was fed during the 24 deprivation. Furthermore, we consider both absolute hours preceding the survey. The seven food groups and relative changes in the prevalence of stunting by are: 1) grains, roots and tubers; 2) legumes and nuts; categories of deprivation to understand inequalities in the 3) dairy products (milk, yogurt, cheese); 4) flesh foods prevalence of child stunting more comprehensively. (meat, fish, poultry and liver/organ meats); 5) eggs; 6) vitamin-A rich fruits and vegetables; 7) other foods and Methods vegetables. Dietary diversity scores were categorized in three groups - low, medium, and high - depending We used cross-sectional data from 15 DHS, collected in of the level of dietary diversity and using the following Bangladesh, India, Nepal, and Pakistan between 1991 cut-offs from the dietary diversity score: 0-1, 2-3, and and 2014. The final analytic sample comprised 55,459 4-7, respectively. Children 6-23 months old with a children aged 6 to 23 months. In the dietary diversity dietary-diversity score between 4 and 7 were classified analyses for India and Pakistan, only 13,570 children as meeting the minimum dietary diversity (WHO, from the latest survey were included due to lower data 2010). Mother’s education was stratified in 3 levels: availability and/or quality in the previous surveys. no education, primary, and secondary and higher. Linear growth was measured as length in younger Wealth quintiles were provided by the DHS (Rutstein & children (6-23 months) using Shorr measuring boards, Johnson, 2004). which are adjustable to the nearest millimetre (WHO, 2006). Measurements were standardized into HAZ We estimated the weighted prevalence of stunting over using WHO age- and sex-specific child growth time and by children’s dietary diversity group, mother’s

10 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA education level, and household wealth quintile. We Gradients in stunting by mother’s education estimated the percentage change in the prevalence of stunting between the earliest and latest estimates The prevalence of child stunting was lower among from each country within each group. We also more educated mothers (Figure 3). Stunting estimated Average Annual Rate of Reduction (AARR), prevalences among children whose mothers’ had which is annualized change in stunting prevalence secondary education or higher were 27.2% (95% CI for each subgroup, to account for differences in time 24.4-30.1), 32.3% (95% CI 30.7-34.0), 16.4% (95% intervals between the earliest and latest surveys within CI 11.4-22.9), and 25.4% (95% CI 18.9-33.2) in the countries. Absolute and relative differences were most recent surveys in Bangladesh, India, Nepal, and calculated between the two categories in dichotomous Pakistan respectively. In comparison, about 40-50% social groups, and between the lowest and highest of children whose mothers were uneducated were categories in categorical groups, for both the earliest stunted in the most recent surveys. All sub-groups and latest survey years in every country. Standard experienced declines in stunting prevalence in all design-based confidence limits were approximated countries; however, there was some country variability. using the logit transformation procedure, which In Bangladesh, children of mothers with no education guarantees that the endpoints of the estimated and primary education experienced similar declines proportion lie between 0 and 1. Adjusted models were – around a 1.9-2.1 AARR in stunting while children also estimated using logistic regression to assess of mothers with secondary and more education associations between stunting and different degrees experienced no significant decline in stunting. of deprivation, controlling for age, sex, birth order, Meanwhile, in India and Nepal, the greatest reductions and place of residency. We accounted for the cluster in stunting occurred among children with mothers who design of the survey in our analyses. had secondary or more education although significant reductions were also recorded among children whose Results mothers had no education. In Pakistan, no significant annual reductions were observed in any education Country-specific analyses group. The education gradient in stunting was preserved in all four countries despite changes over Approximately half of children were stunted at baseline time among education groups. in all countries. From the 1990s, when the earliest DHS surveys were administered, to the most recent survey, Patterning in stunting by household wealth which took place between 2006-2014, all countries experienced reductions in stunting prevalence with the In both the earliest and latest surveys, the prevalence largest declines recorded by Bangladesh and Nepal, of stunting was significantly higher among children where stunting prevalence declined annually by 2.9 from poorer households, with half to nearly two thirds AARR and 4.1 AARR respectively compared to 1.3 of children in the lowest two quintiles being stunted AARR in India and 0.6 AARR in Pakistan. At 43.5% (Figure 4). Stunting rates declined in all quintiles in (95% CI 42.3-44.8) and 39.7% (95% CI 35.4-44.2), India all countries; however, reductions were not uniformly and Pakistan had the highest prevalence of stunting distributed. In all four countries the largest reductions in the most recent DHS survey (2006 and 2013 in the prevalence of stunting occurred in the richest respectively). quintiles while the poorest quintiles recorded the smallest declines. In the richest quintile, the prevalence Trends in stunting by dietary diversity of stunting prevalence declined by 2.6, 2.7, 7.0, and 0.8 AARRs in Bangladesh, India, Nepal and Pakistan, In general, the prevalence of stunting was on average respectively. In comparison, in the poorest quintile, the higher among children whose dietary diversity was prevalence of stunting declined by 2.4, 0.5, 2.1, and -0.5 poorer in both the earliest and latest surveys, where AARRs for the same order of countries. Absolute and data were available, yet this trend was significant relative differences between the poorest and richest only for India (Figure 2). In Bangladesh and Nepal, a quintiles increased in all four countries with the largest marked decrease in the prevalence of stunting was increases in Nepal and Pakistan. observed among the most deprived group between the Association of stunting with different domains earliest and most recent surveys, both experiencing a of deprivation similar percent annual decline: the AARR was 3.4 in Bangladesh and 3.3 in Nepal. A similar reduction was After controlling for child age, sex, birth order, and also observed among the medium dietary diversity place of residence, results from fully-adjusted models, group in Nepal, with a 3.6 AARR. using pooled data from the latest survey years, revealed

11 Figure 2. Prevalence of stunting (95% confidence intervals) by dietary diversity score groups (limited data for India and Pakistan*)

Highest Earliest Most recent t a n

i s Second

P a k Lowest

Highest a l Second N e p Lowest

Highest i a

d Second n I Lowest

h Highest e s

l a d Second g a n

B Lowest

10 20 30 40 50 60 70

Stunting prevalence (%)

* Data from Pakistan (2013), Nepal (1996; 2011), India (2006), Bangladesh (1997, but data for the highest dietary diversity group not available; 2011).

Figure 3. Prevalence of stunting (95% confidence intervals) by mother’s education*

Earliest Most recent Secondary or Higher

Primary t a n i s

P a k None

Secondary or Higher

a l Primary N e p None

Secondary or Higher

Primary i a d n I None

Secondary or Higher h

e s Primary l a d g a n

B None

0 10 20 30 40 50 60 70 80

Stunting prevalence (%)

* Data from Pakistan (1991; 2013), Nepal (1996; 2011), India (1993; 2006), Bangladesh (1997; 2011).

12 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA Figure 4. Prevalence of stunting (95% confidence intervals) by household wealth quintile*

Richest Earliest Most recent

Fourth t a n

i s Third

P a k Second

Poorest

Richest

Fourth

a l Third

N e p Second

Poorest

Richest

Fourth i a

d Third n I Second

Poorest

Richest

h Fourth e s Third l a d g

a n Second B Poorest

0 10 20 30 40 50 60 70 80

Stunting prevalence (%)

* Data from Pakistan (1991; 2013), Nepal (1996; 2011), India (1993; 2006), Bangladesh (1997; 2011).

significant associations between stunting and children’s In country-specific models, we observed mixed results. dietary diversity, mother’s education, and household In India, the odds of stunting was significantly higher wealth quintile. The odds of stunting was significantly among children from the most deprived groups in higher for children in the most deprived groups; for terms of child’s dietary diversity, maternal education, instance, a higher odds of stunting was observed for and household wealth. In Bangladesh, the odds of children in the group with the lowest dietary diversity stunting was significantly higher across wealth quintiles compared to children in the group with the highest groups, but not by children’s level of dietary diversity and maternal education. No significant differences were dietary diversity (odds ratio (OR)=1.47, 95% confidence observed in any domain in Nepal. Lastly, in Pakistan the interval (CI) 1.28-1.69); for children of uneducated odds of stunting was significantly higher for wealthier mothers compared to children of the most educated and more educated groups, but not by dietary diversity mothers (OR=1.51, 95% CI 1.34-1.69); and for children group. Differences in the findings from pooled models in the poorest wealth quintile compared to children and the country-specific models can be attributed to in the richest wealth quintile (OR=3.01, 95% CI 2.48- the influence of the data from India, which is the most 3.64). In an extended model using the pooled sample, populous country in the pooled analyses. Indeed, the we included interactions between dietary diversity and findings from the India-specific model are similar to the levels of mother’s education, and found significant pooled analyses. interactions (p-value <0.05) between these two domains of deprivation. We conducted the same interaction Conclusion tests in subsamples of the poorest and richest children in separate models; however, we found no evidence of Our analysis of the prevalence and trends in child multiplicative effects of maternal education and dietary stunting in South Asia had three key findings. First, diversity among either the poorest or richest children. all countries had high stunting prevalence with nearly

13 half of the children suffering from stunted growth at 2008; Kanjilal et al., 2010; Smith & Haddad, 2015). baseline. Stunting declined in all four countries with These findings have important policy and programme greater reductions in Bangladesh and Nepal. Second, implications at the national level, but the relative in all four countries, the prevalence of stunting was significance of the same set of risk factors may greater among children who had poor diets, who had substantially vary when the focus is on the outcome mothers with low educational attainment, or who lived of stunting, a measure of chronic undernutrition, as in poor households. Pooled, adjusted models showed opposed to general nutritional status; the first two higher rates of stunting in children who experienced years of life, when most child stunting happens in all three dimensions of deprivation. However, country- LMICs, as opposed to children under age of five; specific models showed varying relationships between and by different countries. In Section 4, we use domains of deprivation and stunting. Third, the largest the most recent nationally representative data from declines in stunting were recorded in higher wealth Afghanistan, Bangladesh, India, Nepal, and Pakistan quintiles while the lower wealth quintiles recorded the (home to over 95% of stunted children in South Asia) smallest declines in stunting. Disparities in stunting to investigate the relative and joint importance of a rates were largely preserved over time, and in some set of 13 correlates of stunting and severe stunting in cases worsened between children from poorer vs. infants and young children aged 6-23 months old. richer households and between children whose mothers had low vs. high educational attainment. Methods Future efforts to avert child stunting should target groups that are most vulnerable. Data for Afghanistan (2013) come from the National Nutrition Survey (Afghanistan, 2013) while data for Bangladesh (2014), India (2006), Nepal (2011), and Drivers of child stunting in Pakistan (2013) come from the latest DHS (Measure DHS, 2016). Both National Nutrition Survey and DHS South Asia are nationally representative household surveys that collect detailed information of health and nutrition Introduction indicators from children and their mothers, including anthropometric measurements and demographic The conceptual models for child undernutrition characteristics. We stratified the sample and describe immediate, underlying and basic performed separate analyses for two different model determinants of undernutrition at the individual, specifications, depending on the age of children and household, environmental, socioeconomic, structural the risk factors that are relevant for each age group. In and cultural domains, and support the need for multi- the first model (Model 1 hereafter) we analyzed data factorial interventions (Bhutta et al., 2008; Black et al., for 3,159 children aged 6-8 months. In the second 2013; UNICEF, 2013). Nutrition-specific interventions model (Model 2), we included data from 18,586 that address the immediate and underlying children aged 6-23 months. determinants of fetal and child nutrition are less likely to be effective in sustainably reducing undernutrition Children were considered stunted if their HAZ was unless the household, social and structural below -2 SD of the median for their sex according to determinants of maternal and child nutrition, including the WHO Child Growth Standards. Additionally, severe household poverty, access to health services, stunting was considered as a secondary outcome health environment, women’s education and social for analysis of Model 2. Children were classified as exclusion, are also addressed (Black et al., 2013; severely stunted if their HAZ was <-3 SD of the median Bryce et al., 2008; Ruel et al., 2013). of the WHO Child Growth Standards (de Onis, 2006).

Empirical research on a selective set of risk factors We considered a comprehensive set of 13 correlates on child growth and development are valuable for that has shown varying degree of association with assessing the role of specific determinants, but do stunting in different settings (Bhutta et al., 2008; not allow an examination of the relative importance Bhutta et al., 2013; UNICEF, 2013). Child’s timing of multiple factors on children’s health and nutrition of the introduction of complementary foods and outcomes (Bhutta et al., 2008). Existing studies minimum dietary diversity were defined closely focused on LMICs have identified child feeding adhering to the UNICEF and WHO infant and young practices, maternal nutrition, and household wealth child feeding indicators (WHO, 2008, 2010). Timely as key determinants of the nutritional status for pre- introduction of complementary foods was defined for school age children (Espo et al., 2002; Jones et al., infants 6-8 months old as having received solid, semi-

14 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA solid or soft foods the previous day of the interview. The following logistic regression models were Continued breastfeeding was defined for all children conducted to evaluate the association between selected aged 6-23 months. The definition of minimum feeding variables and stunting by pooling the data from five frequency varied by age and breastfeeding status: for countries and then for each country separately. A breastfed children, it was defined as two times if 6-8 different set of correlates was considered for 6-8 months, and three times for 9-23 months; for non- months old infants (Model 1) and 6-23 months old breastfed, it was defined as four times for all children children (Model 2) due to clinical relevance and 6-23 months old. Based on the dietary diversity score technical reasons. For instance, the timely introduction used in the previously published literature (Ruel & of complementary foods was included only in Model Menon, 2002), minimum dietary diversity was defined 1 since data on this indicator are only obtained for for children aged 6-23 months as having received infants aged 6-8 months. On the other hand, the child’s foods from four or more of the following food groups: vaccination status was included only in Model 2 since 1) grains, roots and tubers; 2) legumes and nuts; 3) full vaccination data are applicable for children aged dairy products (milk, yogurt, cheese); 4) flesh foods 12-23 months. For technical reasons, minimum feeding (meat, fish, poultry and liver/organ meats); 5) eggs; frequency was excluded from Model 1 due to high 6) vitamin-A rich fruits and vegetables; and 7) other correlation with timely introduction of complementary fruits and vegetables. foods; and continued breastfeeding and minimum dietary diversity were excluded from Model 1 because Several maternal variables were also considered. of the small sample sizes and stunting cases for this Mother’s education was categorized in three levels: no age group in country-specific models that resulted in schooling, primary, and secondary or higher. Mother’s lack of statistical power and unstable estimation. For height was measured by trained field investigators and Afghanistan-specific analysis, mother’s age at marriage was categorized in four groups: <145, 145-149.9, 150- and child’s full vaccination were not available. For this 154.9, and 155+ cm. Of note, height measures below reason, the pooled analyses are presented for results 100 cm or above 200 cm were excluded (Özaltin et al., including and excluding Afghanistan. For Model 2, 2010). Mother’s BMI was calculated from the objectively we also estimated the fully-adjusted models for the measured height and weight, and was grouped in <18.5, secondary outcome of severe stunting. Based on 18.5-25, and 25+ kg/m2. Mothers’ age at marriage these models, the relative significance of each of the was defined dichotomously for married or cohabitating correlates was compared and ordered by magnitude. As mothers, using the age of 18 years as cutoff. a sensitivity analysis, we re-estimated the fully-adjusted models after removing type of drinking water, sanitation Finally, the following proxy variables for household facility, and indoor air quality since these assets were socioeconomic conditions were included. Child’s full considered in the estimation of DHS wealth quintiles vaccination was defined as having received 8 vaccines and may have issues of multicollinearity. All analyses (measles, BCG, DPT 3, and Polio 3). Household indoor were accounted for the cluster survey design, and were pollution was defined as high air quality if non-solid further adjusted for age and sex of the child, birth order, fuels were used for cooking and poor air quality if and place of residency (urban/rural). The country fixed solid fuels were used (Corsi et al., 2015). The source effects was included in the pooled analysis. All statistical of drinking water was defined as safe if water sources tests were two-sided and p<0.05 was considered were available in the household (piped into dwelling or to determine statistical significance. We used Stata yard/plot, public tap/standpipe, tube well or borehole, (version 13.1) for all analyses procedures. protected well or spring, rain water, and bottled water), and unsafe otherwise. Household sanitation facility Results was defined as improved if households had access to flush to piped sewer system, septic tank, or pit latrine, South Asia ventilated improved pit latrine, pit latrine with slab, and composting toilet, and unimproved otherwise. We Among 3,159 children aged 6-8 months in South Asia, included the household wealth index, as reported by the 25.9% (n=696) were stunted. About two in five (37.5%) NNS and DHS, which was constructed as a weighted of these 3,159 children were not fed complementary sum of household assets by using the first factor foods indicating late initiation of complementary resulted from a principal component analysis on those feeding. Among 18,586 children aged 6-23 months, assets as weights. The resulting wealth index was used 38.8% (n=7,140) were stunted and 18.3% (n=3,362) to construct cut offs to rank households into wealth were severely stunted. 70.1% of the children aged 6-23 quintiles, based on the weighted frequency distribution months did not meet the minimum dietary diversity of households (Rutstein et al., 2004). requirements. Mother’s nutrition and status varied

15 Figure 5. Fully adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for risk factors staged 6-8 months in South Asia* (pooled sample)

5.0

4.5

4.0

3.5

3.0

2.5 Odds Ratio

2.0

1.5

1.0

0.5

Low BMI

Unsafe water Shorts Stature No education Child marrige Poorest HH wealth HH indoor pollution not improved sanitation No Time inrod comp foods * Afghanistan was excluded due to quality of data

substantially across the countries. Overall, 2.8% of effect sizes, were: maternal short stature (OR: 3.37, the mothers of children aged 6-23 months had short 95% CI: 2.82-4.03), poorest household wealth stature (height <145 cm), 12.4% were underweight quintile (OR: 2.25, 95% CI: 1.72-2.94), maternal (BMI <18.5 kg/m2), and 82% had no formal education. underweight (OR: 1.59, 95% CI: 1.27-2.00), failure In terms of household level characteristics, 25.8% of to meet the minimum dietary diversity (OR: 1.48, the children aged 6-23 months did not have access to 95% CI: 1.27-1.72), no maternal education (OR: improved water source, 61.8% did not have access to 1.36, 95% CI: 1.18-1.56), and mother’s young age improved sanitation facility, and 16.1% were from the at marriage (<18 years) (OR: 1.17, 95% CI: 1.05- lowest wealth quintile. 1.30) (Figure 6). When Afghanistan was included in the pooled analysis, lack of continued breastfeeding In the fully adjusted Model 1 (excluding Afghanistan), (OR: 2.02, 95% CI: 1.37-2.97) was also found maternal stature and timely introduction of to be statistically significant, while maternal complementary foods were statistically significant underweight and young age at marriage were no risk factors of stunting in children aged 6-8 months. longer significant. Finally, in the pooled analysis for Children with short mothers (<145cm) had 2.93 times severe stunting among children aged 6-23 months higher odds of being stunted (95% CI: 1.93-4.46) than (both including and excluding Afghanistan), the those with taller mothers (155+ cm). Children who statistically significant predictors were: maternal had delayed introduction of complementary foods short stature (OR: 3.15, 95% CI: 2.55-3.90), poorest had 1.47 times higher odds of stunting (95% CI: 1.12- wealth quintile (OR: 3.07, 95% CI: 2.20-4.28), no 1.93) compared to their counterparts (Figure 5). When maternal education (OR: 1.48, 95% CI: 1.25-1.76), Afghanistan was included in the pooled analysis, no maternal underweight (OR: 1.41, 95% CI: 1.05- factor remained statistically significant. 1.90), no minimum dietary diversity (OR: 1.37, 95% CI: 1.11-1.70), not fully vaccinated (OR: 1.22, 95% For children aged 6-23 months (fully adjusted Model CI: 1.06-1.41), mother’s young age at marriage (OR: 2, excluding Afghanistan), the statistically significant 1.16, 95% CI: 1.01-1.33) and no minimum feeding risk factors of stunting, ranked from the strongest frequency (OR: 1.14, 95% CI: 1.01-1.29).

16 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA Figure 6. Fully adjusted odds ratios (ORs) and 95% confidence intervals (CIs) for risk factors staged 6-23 months in South Asia* (pooled sample)

5.0

4.5

4.0

3.5

3.0

2.5 Odds Ratio

2.0

1.5

1.0

0.5

Low BMI

Unsafe water Shorts Stature No education Child marrige No feeding freq No Breasfeeding Poorest HH wealth Not fully vaccinated HH indoor pollution No Min diet diversity Not Improved sanitation * Afghanistan was excluded due to quality of data

Afghanistan to be statistically not significant, probably due to the small sample size. Among children aged 6-23 In Afghanistan, there were 341 children aged 6-8 months, 31.9% were stunted and 10.8% were months and 26% of them were stunted. None of the severely stunted. The following five risk factors risk factors were found to be statistically significantly were found to be statistically significant for stunting associated with risk of stunting for children aged 6-8 among children aged 6-23 months (Model 2): short months in Afghanistan. Of 3,644 children aged 6-23 maternal stature (OR: 3.42, 95% CI: 2.03-5.76), maternal underweight (OR: 2.05, 95% CI: 1.27-3.30), months, 38.7% were stunted and 18.3% were severely lowest wealth quintile (OR: 1.98, 95% CI: 1.08- stunted. In this age group, lack of formal education for 3.63), no education for mothers (OR: 1.78, 95% CI: mothers (OR: 3.01, 95% CI: 1.53-5.93), short maternal 1.17-2.70), and no minimum dietary diversity (OR: stature (OR: 2.61, 95% CI: 1.41-4.84), poorest wealth 1.58, 95% CI: 1.11-2.24). For severe stunting, only quintile (OR: 2.28, 95% CI: 1.48-3.50), lack of continued maternal stature and wealth quintile remained to be breastfeeding (OR: 2.08, 95% CI: 1.35-3.21), and failure statistically significant predictors. to meet the minimum dietary diversity (OR: 1.34, 95% CI: 1.04-1.72) were statistically significant risk factors India for stunting. Consistent results were found for severe stunting in this age group, but dietary diversity and Among the five South Asian countries considered maternal education were not statistically significant. in this study, India had the highest prevalence of stunting for children aged 6-8 months (25.1%) and for Bangladesh those aged 6-23 months (43.5%). In the fully adjusted model for children aged 6-8 months (Model 1), short The prevalence of stunting in Bangladesh was 16.9% maternal stature (OR: 3.39, 95% CI: 2.10-5.45) and among children aged 6-8 months. For this age group, failure to introduce complementary foods in a timely all the risk factors for stunting (Model 1) were found manner (OR: 1.50, 95% CI: 1.11-2.04) were the

17 only statistically significant risk factors for stunting. 1.05-2.42) were statistically significant risk factors Maternal underweight, household wealth, maternal for stunting. For severe stunting, only short maternal education, and other factors were not statistically stature was statistically significant (OR: 4.68, 95% CI: significant. These findings were consistent with 1.57-13.9). those from the pooled analysis. For children aged 6-23 months (Model 2), there were five statistically Conclusion significant risk factors for stunting: short maternal stature (OR: 3.43, 95% CI: 2.82-4.18), lowest wealth Our findings provide evidence on the relative importance quintile (OR: 2.26, 95% CI: 1.67-3.05), maternal of multiple correlates of child stunting, and enriches underweight (OR: 1.56, 95% CI: 1.20-2.03), poor the existing assessments of determinants of nutritional dietary diversity (OR: 1.52, 95% CI: 1.26-1.82), and status among pre-school age children in LMICs (Espo mothers with no education (OR: 1.36, 95% CI: 1.17- et al., 2002; Jones et al., 2008; Kanjilal et al., 2010; 1.57). In addition to these risk factors, not being fully Smith & Haddad, 2015). For instance, child feeding vaccinated (OR: 1.29, 95% CI: 1.11-1.51) and failure has been increasingly recognized as a key determinant to meet the minimum feeding frequency (OR: 1.15, of child nutrition, with greater than two-thirds of 95% CI: 1.01-1.31) were also found to statistically malnutrition-related deaths in children estimated to be significantly increase the odds of severe stunting for associated with inappropriate feeding practices (Black children aged 6-23 months. et al., 2008). Young children should be introduced to complementary foods at six months of age (WHO, Nepal 2002), and failure to do so has been shown to leave children vulnerable to irreversible outcomes of stunting In Nepal, 18.8% of the children aged 6-8 months (Gross et al., 2000; Saha et al., 2008). Almost 40% of and 29.2% of those aged 6-23 months were stunted. infants aged 6-8 months in our pooled sample were The sample size and number of stunted children not given complementary foods, and failure to do so were very small in Nepal, and therefore both Models was significantly associated with increased odds of 1 and 2 could not be reliably estimated due to low stunting. Moreover, not meeting the minimum dietary statistical power. Among 122 children aged 6-8 diversity was also found be important for stunting months, only 23 cases (18.9%) of stunting were among children aged 6-23 months, which is consistent detected, and all the estimates associated with with a prior review reporting that dietary diversity was stunting, except for short maternal height, were associated with height-for-age in many LMICs (Arimond found to be statistically not significant. Among & Ruel, 2004). In addition, maternal height, a marker children aged 6-23 months, short maternal stature of mother’s early life nutritional and environmental (OR: 5.58, 95% CI: 2.52-12.4) and lowest wealth conditions that has intergenerational consequences on quintile (OR: 5.26, 95% CI: 1.35-20.5) were found to nutrition and health (Özaltin et al., 2010; Subramanian be statistically significant predictors of stunting, but et al., 2009), was also found to be one of the strongest the associated confidence intervals were very wide, predictors of child stunting and severe stunting in which indicated unstable estimation. The same was South Asia. Moreover, the consistent findings on true for severe stunting among children aged 6-23 the importance of maternal undernutrition, maternal months. education and household wealth index further support that the overall environmental and socioeconomic Pakistan conditions influence child nutrition through pathways involving food insecurity and inadequate access to care The prevalence of stunting was 25% and 39.6% and dietary resources for children (Ruel et al., 2013). for children aged 6-8 months and 6-23 months, Overall, our analysis found that the top five risk factors respectively, in Pakistan. Due to small number of identified for stunting and underweight for children aged stunted children (n=62) aged 6-8 months, all the risk 6-59 months in a recent study in India (short maternal factors for stunting were found to be statistically not stature, mothers having no education, households in significant, expect for no maternal education (OR: lowest wealth quintile, poor dietary diversity in young 7.39, 95% CI: 1.71-31.9), but the wide confidence children, and maternal underweight) (Corsi et al., 2015) interval indicated unstable estimation. For children are also applicable for infants and children aged 6-23 aged 6-23 months (Model 2), short maternal stature months across five South Asian countries, albeit the (OR: 3.56, 95% CI: 1.24-10.2), child marriage (OR: notable variations in the ranking and significance of 1.71, 95% CI: 1.10-2.67), unimproved drinking correlates by country. An important policy implication of water source (OR: 1.63, 95% CI: 1.00-2.68) and no our research is to further emphasize that interventions minimum feeding frequency (OR: 1.59, 95% CI: focused on specific risk factors alone may be

18 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA inadequate, and that a multi-factorial framework productivity, investments targeted at reducing child approach should be employed to address child stunting bring about substantial returns to society, stunting in South Asia. For instance, comprehensive including significant increases in economic growth strategies focused on a broader progress on maternal (Akachi & Canning, 2015; Bloom et al., 2014). and household socioeconomic factors as well as investments in nutrition specific programs to promote In South Asia, economic growth ‘per se’ will dietary diversity and appropriate complementary not address child stunting. Greater attention feeding are needed to improve stunting rates at the needs to be paid to ensure that economic population level (Bhutta et al., 2008; Bhutta et al., 2013; growth reduces inequities and addresses Black et al., 2008; Ruel et al., 2013). the drivers of child stunting among the most vulnerable population groups. Economic growth as a policy instrument to reduce stunting Implications for action is not enough. Greater attention needs to be paid to addressing the social, economic, and political drivers of child stunting, with targeted Stunting in childhood has immense implications as efforts towards the populations at higher risk of stunted children experience worse survival, growth, growth faltering and stunting in children. There development, and livelihoods than non-stunted is increasing consensus that economic growth children (Black et al., 2013; Grantham-McGregor et al., will only be effective in addressing child stunting 2007), leading to significant losses of human capital if increases in national income are directed at and productivity (Behrman et al., 2015; Black et al., improving the diets of children, strengthening 2013; Grantham-McGregor et al., 2007; Horton & the status of women, addressing inequities, Steckel, 2011). In this chapter, we provide evidence to and reducing poverty. Therefore, the literature further support that eliminating child stunting should has begun to emphasize the importance of the be moved to the forefront of national and international “quality of economic growth”, or how income policy, programme, advocacy, and research agendas. gains translate in reductions in child undernutrition More specifically, we argue that renewed attention to (Haddad et al., 2015). The impact of intermediary four main priorities is needed: factors, such as growth in food production, sanitation, access to health services, education, South Asian countries, which have the largest fertility, governance, and the roll-out of program proportion and burden of stunted children interventions have all been highlighted (Alderman globally, need to universalize evidence-based et al., 2006; Fink et al., 2011; Headey, 2013; Smith policies and programmes designed to improve & Haddad, 2015; Webb & Block, 2012). Therefore, the nutrition situation of young children and focusing on promoting an economic growth model their mothers as a national development that does not address these intermediary factors priority. We provide evidence that proven nutrition among the most vulnerable populations is not investments that focus on the 1,000 days from likely to be effective in addressing child stunting conception to age two years and the 10-18 years (Haddad, 2015). The disproportionate burden adolescence period (i.e. the two windows of of stunting experienced by the most deprived opportunity to address stunting) are cost-beneficial. population groups in South Asia and the persistent Research evidence indicates that packages that and worsening inequalities among socio-economic combine context-specific nutrition, care, and groups is of great concern. To unleash the potential stimulation interventions may bring about the of nutrition policies and programmes, progress is greatest impact (Alderman et al., 2014; Gertler et needed in addressing persistent socio-economic al., 2014; Yousafzai et al., 2014). Improvements in inequities (Oruamabo, 2015). child linear growth and reductions in child stunting allow young children to reach their full growth and Nutrition-specific and nutrition-sensitive development potential and expand their economic programmes should be implemented jointly opportunities as adults, generating substantial to address the main drivers of stunted growth economic returns at the individual level (Barker and development in South Asian children. et al., 2005; Hoddinott et al., 2013b; Martorell Programming efforts to prevent and reduce et al., 2010b; Victora et al., 2008). Furthermore, stunting, especially those focused on the critical because of the spillover effects of reductions in 1,000 day window of opportunity, should be the prevalence of stunting on labour markets and guided by the multi-factorial conceptual framework

19 of maternal and child undernutrition (UNICEF, inquiry. Recent reviews indicate that integrated 2013). This framework recognizes the interaction interventions have great promise (Black et al., between the underlying and basic causes of 2015; Christian et al., 2015; Grantham-McGregor undernutrition in determining optimal nutrition, et al., 2014; Yousafzai & Aboud, 2014; Yousafzai growth, and development in infancy and early et al., 2014). However, more research is needed to childhood (Bhutta et al., 2008; Maternal & Child evaluate rigorously the effects of these programs, Group, 2013). In our analyses, when we considered particularly in the long-run. If these interventions three levels of determinants of child nutrition (child- are able to address the joint etiologies of stunting level, mother-level, and household-level), delayed and poor child development, they may be more introduction of complementary foods, failure to effective in promoting child health and well-being. meet children’s minimum dietary diversity, mother’s High child stunting rates in South Asia challenge short stature, underweight, lack of education, progress towards the SDGs, notably ending poverty in and early marriage, and poor household wealth all its forms (SDG 1); promoting sustained, inclusive, were found to be the most significant drivers of and sustainable economic growth, full productive child stunting in South Asia. Nutrition-specific employment and decent work for all (SDG 8); and interventions, such as the protection, promotion, reducing inequality within and among countries (SDG and support of breastfeeding, appropriate 10) (UN, 2015b). Indeed the SDGs explicitly note the complementary foods and feeding, care, and importance of nutrition with SDG 2 aiming to end hygiene practices, adolescent girls’ and women’s hunger, achieve food security and improved nutrition nutrition, macro and micronutrient supplementation and endorsing the WHA target to reduce the number for vulnerable population groups, food fortification of stunted children by 40% by 2025 (UN, 2015b). programmes, and the prevention and treatment of The achievement of this global target requires that infectious diseases and severe acute malnutrition, South Asian countries accelerate efforts to reduce the should be combined with nutrition-sensitive prevalence of stunting in the region and contribute programmes that address the more distal socio- to achieve the SDGs related to child survival, growth, economic determinants of maternal and child development, education, participation, and equity. undernutrition to prevent stunting in the first two years of life (Bhutta et al., 2008; Bhutta et al., 2013; Black et al., 2008; Pinstrup-Andersen, 2013; Ruel et al., 2013). We argue that neither approach in isolation is adequate to address the challenge of widespread stunted growth and development among South Asia’s children.

In parallel to national and regional efforts to reduce child stunting, there is a need for research and programmatic evidence to better understand how to drive faster and larger declines in South Asian countries, particularly among the most vulnerable children: the youngest, poorest, and the socially excluded. Understanding what drives child stunting at the national and subnational levels (i.e. the importance of context), how stunting influences child growth and development, and which domains of growth and development are more affected require further research. Relatedly, understanding how best to implement and scale up programmes aiming at improving children’s growth and development and reducing stunting – either through information, counselling and support on optimal child feeding and care, through dietary diversification, nutrition supplementation and/or fortification programs, or through integrated child nutrition and psychosocial stimulation programmes are key areas of further

20 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA 21 22 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA

Patrick Webb, Kassandra L Harding, and Victor M. Aguayo

Introduction alone (Black et al., 2013). Yet even mild or moderate forms of hidden malnutrition have wide-ranging Much of the world’s burden of malnutrition is visible. implications for human growth and development. Children and adults who are wasted or obese can often Deficiencies in iodine and iron, for example, contribute be identified without special training or equipment. to physical growth impairment in utero and early However, there are other equally important forms childhood, as well as to cognitive deficits. In 2013, out of malnutrition that are less obvious to the naked of 289 diseases and injuries contributing to the global eye; namely, deficiencies of essential vitamins and burden of disease, the combination of just three - iron, minerals. According to the World Health Organization iodine and vitamin A deficiencies - contributed 5% of an estimated two billion or more people suffer the world’s burden of ‘years lived with disability’ (YLD), what is called micronutrient malnutrition or more which is a metric of lost wellbeing associated with a colloquially ‘hidden hunger’ (WHO, 2016). This form poor state of health (Global Burden of Disease Study of malnutrition tends to be hidden both to the person 2013 Collaborators, 2015). suffering the deficiency and to the outside world until it becomes so severe that clinical signs emerge. Such nutrition losses are associated with impaired human capital. The term human capital is used to The most serious deficiencies of some micronutrients, reflect the combination of skills, knowledge, and such as vitamin A and zinc, cause death: roughly experience accumulated by individuals throughout the 270,000 annual deaths in children less than 5 years old life-course based on their physical attributes and brain are attributed to these two micronutrient deficiencies development. Growing well as an infant, learning in

23 school, working productively as an adult, and providing A second major global hidden deficiency is zinc. for the next generation all contribute to the social Assessments of national diets suggest that more than and economic wealth of nations. By contrast, threats 17% of the world’s population is at risk (Black et al., to human capital represent a drag on sustainable 2013). Deficiency in zinc (a serum concentration of development. South Asia is one of the regions of the <60 µg/dl) is linked to poor child growth outcomes, world carrying the highest burden of hidden malnutrition impaired immune function, the incidence and severity and associated health burdens (Vos et al., 2012). of diarrhoeal disease, and numerous other nervous and reproductive system problems (Wessels et al., 2012). This paper presents an overview of what is empirically Indeed, an estimated 116,000 child deaths are ascribed known about the human and economic impacts of to zinc deficiency or to elevated risk of deficient dietary various micronutrient deficiencies, and about recent intake of zinc each year (Black et al., 2013). Conversely, patterns and trends in micronutrient deficiencies zinc supplementation to pregnant women has been across South Asia. It draws on the most recent surveys shown to significantly reduce pre-term births in low and datasets to highlight where gains have been income countries (Mori et al., 2012). made and where a lack of progress continues to be of concern. A final section lays out policy and programme The third micronutrient deficiency of significance is implications, calling for an urgent prioritization by vitamin A. Globally, one third of all children under five South Asia’s governments and development partners years suffer vitamin A deficiency (VAD) which causes of this largely unseen threat to future development. blindness or death when associated with severe bouts of diarrhoea or measles. An estimated 157,000 children Hidden malnutrition and the under five die annually due to VAD (Black et al., 2013). Around 15% of pregnant women also experience development of nations: a review VAD, susceptibility being at its highest during the of evidence third trimester due to accelerated foetal development. Overall (across all ages), VAD is associated with over There are many form of micronutrient malnutrition, 800,000 YLD (Vos et al., 2012). but a few essential vitamins and minerals stand out in terms of the scale of global deficiencies and for the Around 30% of people suffer from iodine deficiency significance of their public health impact. Arguably the disorders, which represent the fourth nutrient most pernicious and widespread condition is anaemia. deficiency of critical importance globally (Andersson Defined as a low blood haemoglobin concentration, et al., 2012). Iodine deficiency in pregnant women is anaemia affects an estimated 2 billion people (WHO, responsible for birth defects, including in utero foetal 2015). An estimated 50% of cases of anaemia are due brain development and impaired cognitive potential, to iron deficiency; other causes include deficiencies in while in children it impairs cognitive ability; in the most other micronutrients such as folate, riboflavin, vitamins severe cases, iodine deficiency results in the growth of A or B12 and infections such as malaria, tuberculosis goitre, in cretinism or even in death (Vos et al., 2012). or HIV/AIDS. These ‘big four’ micronutrients cause so much damage Iron deficiency anaemia adversely affects birth weight, to health, education, physical and mental growth, labour children’s cognitive and motor development, and adult productivity, and longevity that their hidden impacts on labour productivity (Stevens et al., 2013)., Almost one the economy can be enormous. Good nutrition supports in five children under the age of five years* have some high mental acuity and individual earnings, which in form of iron deficiency anaemia, while severe anaemia turn support macroeconomic and societal growth. is associated with around 90,000 deaths annually Conversely, malnutrition impairs individual output, (Stevens et al., 2013;WHO, 2014). which acts cumulatively as a drag on national growth.

While delivery of supplements to pregnant women and For example, Meenakshi et al. (2010) calculated that fortification of staple foods have been widely pursued low vitamin A status among mothers and children for decades, anaemia remains a major challenge can be associated with an annual loss of 1% of gross marked out for the “lack of progress” made in reducing national product. Similarly, Horton & Ross (2003) prevalence rates since the early 1990s (IFPRI, 2015). showed that the cumulative economic impact of

* The term “children under 5 years old” refers to children 6 to 59 months old unless otherwise specified.

24 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA cognitive impairment and lower labour productivity due know about nutrition, how well educated and literate to iron-deficiency anaemia is on average 4% of national they are, and how they are able to act on knowledge income in low income countries. Indeed, children is a major determinant of hidden malnutrition and the who were iron deficient in Costa Rica during infancy ability of societies and nations to interrupt the cycle of had 10% lower cognitive test scores than those who hidden hunger. were not deficiency, but also 26% lower test scores at the age of 19 years; in other words, the impairment Women’s education and hidden persisted throughout their childhood (Lozoff et al., 2006). Similarly, roughly 10% of children born to malnutrition: a review of iodine-deficient mothers are likely to suffer severe evidence mental retardation (Hunt, 2005). This has extraordinary implications for countries of South Asia, where almost Formal educational attainment and informally-obtained a third of households do not consume adequately knowledge held by mothers have both been shown to iodized salt on a regular basis (UNICEF, 2014). be significantly linked to improved nutrition among their children (Smith and Haddad, 1999). Cross-country time What is more, where there is one nutrient deficiency series as well as studies using natural experiments have there are often others. It has been estimated that confirmed that maternal education is a key determinant annual economic losses from low weight, poor growth, of birthweight, neonatal survival, and children’s and clusters of micronutrient deficiencies average 11% attained height. For example, using Demographic and of gross domestic product (GDP) in Asia and Africa Health Survey (DHS) data from 19 countries, Ruel (IFPRI, 2016). Indeed, Madsen (2016) showed that and Alderman (2013) showed that the risk of child from 1960 through 2006, the combination of anaemia stunting was significantly lower among mothers with and iodine deficiency was linked to lower intelligence primary or secondary education even after controlling quotients (IQ), and that a 10% fall in IQ among for wealth, residence, and child’s age and sex. The individuals was associated with a 1% fall in national same study found a significant yet smaller effect of economic growth rates each year. paternal education. Single country studies, such as one in Bangladesh, note that as the proportion of women In terms of hidden malnutrition, the main drivers of with some secondary education rises (in Bangladesh such negative economic impacts are lower learning it doubled between 1994 and 2005), the prevalence and productivity capabilities due to iron deficiency of child stunting tends to fall -- as it did in Bangladesh anaemia, diarrhoeal episodes associated with zinc from 70% to 48% over the same period (Heady, 2013). deficiency, low immunity to diseases linked to a lack Similarly, Cunningham et al. (2016) argue that maternal of vitamin A, and compromised intellectual potential education in Nepal was a key factor in underpinning due to iodine deficiency. Many individuals suffer more what they describe as “remarkable improvements” in than one of these deficiencies simultaneously, and child undernutrition from 1996 to 2011. many deficiencies persist through a person’s lifespan. Micronutrient deficiencies are also associated with the While correlations between education levels and growing global problem of overweight and obesity, undernutrition are robust across regions and over which carry high risks of non-communicable diseases time, the links between education and micronutrient (Via, 2012). The association of obesogenic conditions deficiencies have received less attention. Although, and vitamin and mineral deficiencies is underpinned maternal education is a commonly identified predictor by low quality diets (Aitsi-Selmi, 2014). Monotonous, for risks of anaemia, serum retinol, serum zinc, nutrient-poor diets affect maternal health and nutrition, and other measures of micronutrient status among but they can also lead to in utero malnutrition that has children, few studies have explicitly considered how effects on the newborn that last into adulthood and maternal education interacts with micronutrient can impact the next generation. One study in Thailand deficiencies in the household, and vice versa. Such that followed children from birth to age 9 years found links are bi-directional in that knowledge may that “growth between birth and four months, whether allow families to prevent or deal with micronutrient in length or weight, was the variable most consistently deficiencies, while such deficiencies impact and strongly related to IQ.” (Pongcharoen et al., 2012). individuals’ ability to obtain and act on knowledge. As Block (2007) points out, the ability to understand Since much of a child’s risk of malnutrition is and identify micronutrient content and quality in foods determined by the mother’s nutrition and health status may require more knowledgeable and/or educated at conception, conditions during her pregnancy, birth consumers, thus maternal education could be an outcomes, and subsequent childcare, what women important determinant of a child’s micronutrient status.

25 Figure 7. Prevalence of micronutrient deficiencies by average years of women’s schooling at the country level (Harding et al., 2018)

40 High income

Lower middle income

Low income

20 Upper middle income Prevalence of VAD among <5 yr (%) VAD of Prevalence

0 5 10 15 Years of schooling a. Prevalence of vitamin A deficiency among children <5 years (2013) by women’s schooling (2010)

60 k e (%) a t n i c n z i

t 40 a u q e d a n i

High income d e t Lower middle income m a i t

s 20 Low income e

f o Upper middle income c e n e l v a P r e

0 5 10 15 Years of schooling

b. Prevalence of estimated inadequate zinc intake in the population (2005) by women’s schooling (2000)

26 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA 500

400

300 High income

Lower middle income 200 Low income

Upper middle income

100 Median urinary iodine excretion (µg/L)

0 5 10 15 Years of schooling

c. Median urinary iodine excretion (UIE) of a country (µg/L) (2012) by women’s schooling (2010)

A recent analysis using cross-country data (roughly that all governments should act upon; ii) while 200 countries depending on the outcome of focus) for women’s anaemia and children’s vitamin A improve the period 1990 to 2013 confirmed such relationships along with women’s education, improvements in (Harding et al. 2018). The number of years of women’s zinc and child anaemia are mediated by the effects schooling was significantly associated with all of poverty; which means that complementary actions micronutrient deficiencies, with the exception of iodine (beyond) education are needed to address multiple (Figures 7a-c). forms of hidden malnutrition; and iii) where poverty, overt (not hidden) malnutrition and education of girls Importantly, when controlling for national income per all lag (as in parts of South Asia), there is an urgent capita, women’s years of schooling was no longer need for governments to prioritize actions to address associated with child anaemia in the upper middle such problems simultaneously. income countries, but it remained a significant predictor of child anaemia in the lower income categories, Additionally, it is important to invest in appropriate especially in the lowest income group, while the lack informal knowledge acquisition (beyond schooling) by of formal education among women was not predictive girls and mothers relating to quality diets and nutrition. of zinc deficiency in the lower income countries, but For example, Webb and Block (2004) assessed the remained a significant predictor of zinc deficiency in impact of a mass media campaign in Central Java the upper income countries. Thus, in the case of child that promoted population-wide understanding of anaemia, education appears to matter relatively more good dietary sources of vitamin A. Using seven survey in the context of greater poverty, whereas in the case rounds involving almost 32,000 rural households, of zinc deficiency, measured as inadequate zinc intake, the authors correlated with child nutrition outcomes there may be a base level of resources needed before mothers’ responses to questions about the messages education can impact zinc status (i.e. in poor context, that had formed the basis of the campaign. The regardless of education or knowledge, it may not be messages included information about the value of feasible to purchase foods rich in zinc). dietary vitamin A to child health and growth, and the importance of feeding vitamin A-rich foods to young These findings, complemented by others, strongly children. They showed that while maternal schooling support the conclusions that i) girls’ education plays was, as expected, an important determinant of child an important role in improving nutrition across growth, even ‘uneducated mothers’ (i.e. mothers generations; it represents a nutrition sensitive action lacking formal education) benefitted from well-crafted

27 nutrition information that they could act on. That is, is huge. South Asia alone accounts for almost 40% knowledge gained informally (rather than just the of the entire world’s YLDs attributed to anaemia years spent in school) offers a strong basis for mothers (Kassebaum et al., 2014). In 2013, roughly 2% of all to improve the diversity and quality of the food that preventable child deaths in low and middle income they provide to their children (Webb and Block, countries (LMICs) were attributable to vitamin A 2004). As such, it becomes all the more important for deficiency, and South Asia has some of the highest governments to seek to equip girls and women with prevalence of vitamin A deficiency in the world: the resources (informational as well as educational) exceeding 30% in countries like Nepal and Sri Lanka they need to be able to do all they can to tackle hidden and exceeding 60% in India (Stevens et al., 2015). At hunger through their own dietary choices. the same time, South Asia has the highest rates of inadequate zinc intake, at 30% compared with the Trends in hidden malnutrition in global average of 17% (Wessels et al., 2012)

South Asia This matters for South Asia’s future. Despite some important progress made across the region in While hidden malnutrition is still global in its reach, recent years, at least in terms of economic growth, its prevalence and impacts area disproportionately agricultural output and exports, poverty reduction concentrated in poorer regions of the world such as and even in improved health, South Asia’s burden of South Asia. The scale of the problem facing South micronutrient deficiencies is still very large: Pakistan Asia’s national governments and development partners has long been classified as a nation with ‘severe iodine

Figure 8. Prevalence of micronutrient deficiencies globally and in South Asia and Sub-Saharan Africa*.

100

World South Asia Sub-Saharan Africa 90

Prevalence of micronutrient de ciences (%) Prevalence 20

Vitamin A Iodine Zinc Anemia among Anemia among de ciency de ciency De ciency children < 5 pregnant among women children <5

* Vitamin A deficiency: Serum retinol <0.70 µmol/L; data from 2013 (Stevens et al., 2015); Iodine deficiency: Urine iodine excretion <100 µg/L among school aged children; South Asia includes Southeast Asia and Sub-Saharan Africa includes all of Africa; data from 2011 (Andersson et al., 2012); Zinc deficiency: Weighted average of country mean of estimated inadequate zinc intake at population level; data from 2005 (Wessells et al., 2012); Anaemia: Defined as haemoglobin <11 g/dL; data from 2011; Sub-Saharan Africa includes West and Central Africa (Stevens et al., 2013).

28 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA deficiency’; almost two-thirds of children under 5 in region, detailed knowledge of drivers and differential Afghanistan suffer vitamin A deficiency and India outcomes is a key to effectively-tailored action. For continues to report that around 53% of women of example, South Asia had the highest prevalence reproductive age are anaemic (India, 2016), compared, of anaemia in 1990, and still had the world’s 4th for example, with only 19% in China (IFPRI, 2016) highest prevalence among males in 2010, and 3rd highest among females (Kassebaum et al., 2014). At the same time, improvements remain patchy However, important differences exist across the region. and slow. The prevalence of inadequate zinc intake In Afghanistan, Nepal, and Sri Lanka, anaemia is in South Asia has changed little between 1990 and considered to be a ‘moderate’ public health problem 2005: Afghanistan, Nepal, and Sri Lanka, for example, for non-pregnant women, whereas it is a ‘severe’ recorded less than a one percentage point change problem in Bangladesh, Bhutan, India, and Pakistan over 15 years (Wessels et al., 2012). Between 1995 (Kassebaum et al., 2014; Harding et al., 2017a). and 2011, the prevalence of anaemia among non- pregnant women in South Asia fell only slightly from With around 50 percent of women of reproductive 53% to 47%, but with little change among pregnant age suffering some form of anaemia in the mid-2000s, women (Stevens et al., 2013). Similarly, the estimated India has one of the highest prevalences of anaemia prevalence of vitamin A deficiency among children in South Asia; further, this prevalence was largely under 5 years of age across South Asia declined very remained unchanged at a national level between the slowly and only slightly from 47% in 1991 to 44% in late 1990s and 2006 (India 2006; Bharati et al., 2015) 2013 (Stevens et al., 2013). The lack of progress in (Figure 9). In recent years (2016), there has been a reducing these deficiencies puts most of the countries substantial decreased in anaemia among women in the region off-course in their attempt to reach global of reproductive age in some states in India and not nutrition targets set for the coming couple of decades. others, though the prevalence remains high. For example, the prevalence in Tripura dropped from 65% The details of country-level patterns and trends in 2006 to 55% in 2016 and the prevalence in Sikkim matter a lot to the design of appropriate policy and dropped remarkably from 60% to 35%. On the other programme responses by location. That is, since hand, the prevalence increased or did not change micronutrient problems are not equally felt by all in states such as Haryana, Tamil Nadu and West populations in the same ways everywhere in the Bengal (India, 2016). Anaemia rose in Afghanistan,

Figure 9. Prevalence of anaemia among women of reproductive age across time (1990 to 2012), by country [data from Stevens et al., 2013 and Harding et al., 2017a]

Prevalance of anemia among non-pregnant women (%) Prevalance 1990 1995 2000 2005 2010

Afghanistan Bangladesh Bhutan India

Nepal Pakistan Sri Lanka

29 from 40% of children under 5 in 2004 to 45% in 2011 Similar differences in rates across locations and (Afghanistan, 2013), and in Pakistan, from 51% in populations apply to other micronutrients as well. 2001 to 62% in 2011(WHO, 2007; Pakistan, 2011). Yet, In the case of zinc, about 39% of children under five countries like Nepal and Sri Lanka have seen declines in Pakistan are deficient (Pakistan 2011), while in in anaemia among pregnant women since the 1990s. Bangladesh the rate is 45% (ICDDRB et al., 2013). Current rates are still too high (they remain public Wessels et al. (2012) proposed that Sri Lanka had one health problems), but the downward trend over 20 of the highest population-wide rates of inadequate years from 1990 to around 2010 is a bright spot to be dietary intakes of zinc in the world at 48% in the mid- noted so that lessons might be learned on how this 2000s (2003-2007). But a more recent assessment positive trend has been achieved (Stevens et al., 2013). of dietary zinc deficiency using 2009 data suggests that India tops the South Asia ranking with a risk of Of course, even within countries there are important inadequacy of 46% (Mark et al., 2016). differences. In India, Bangladesh and Nepal, anaemia is more prevalent in rural than in urban populations, Again, there is large intra-nation variability; in a sample whereas in Pakistan there is little difference by location. of states, serum zinc deficiency (cut-off level ≤65 μg/ In neighbouring Nepal, the prevalence of anaemia dL) was highest in Orissa (>51%) and Uttar Pradesh among children is highest in the low-lying plains (50%), (48.1%), but below 40% in Madhya Pradesh (39%), and but only slightly lower in the hills (41%) (Bhandari and Karnataka (36%) (Akhtar, 2013). After India, the next Banjara, 2015). Some Indian states such as Puducherry, highest risk of inadequate dietary intake of zinc was Goa, Meghalaya and Tripura have a relatively low reported for Pakistan at over 39%, while the lowest prevalence of anaemia in children aged 6 to 59 months in the region was Nepal (less than 16%) (Mark et al., below 50%, while other states such as Haryana have an 2016). In Bangladesh, zinc deficiency is higher in slums anaemia prevalence rate exceeding 70% (India, 2016). (52%) and rural areas (49%) than in urban areas (30%) Trend lines also differ: Bharati et al., (2015) reported that (ICDDRB et al., 2013). Yet, rates do not differ between in India’s southern states, mean haemoglobin levels in urban and rural Pakistan (Pakistan 2011). women of reproductive age fell from roughly 11.9 g/dL in 1998/99 to slightly above 11.6 g/dL in 2005/2006 but Where vitamin A deficiency (VAD) is concerned, rose over the same period in the north-east (from 11.6 g/ South Asia is still the region with the greatest number dL to 11.8 g/dL) (Bharati et al., 2015). of affected children; while all LMICs together have

Figure 10. Prevalence of vitamin A deficiency (VAD) among children 6-59 months old across time (1991 to 2013), by country [data from Stevens et al., 2015 and Harding et al., 2017b]

10 Prevalance of VAD among children (%) among children VAD of Prevalance

1990 1995 2000 2005 2010 2015

Afghanistan Bangladesh Bhutan India

Maldives Nepal Pakistan Sri Lanka

30 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA seen VAD decline from almost 40% to less than 30% iodine deficiency (median UIE <100 µg/L) is not between 1991 and 2013, South Asia’s average remains a public health problem for any country in South in the order of 44% (Stevens et al., 2015). Sri Lanka Asia (Andersson et al., 2012) (Figure 11) However, and the Maldives are two positive outliers, as it appears some countries like Afghanistan (where only 20% of that these countries have made huge improvements households consumed adequately iodized salt) still in the vitamin A status of its children since the early show rates almost double those of other parts of the 1990s (Thompson and Amoroso, 2014) (Figure 10). same region, such as Nepal (UNICEF 2014).

Where adult women are concerned, a relatively low Drivers of hidden malnutrition in 11% of women of reproductive age were vitamin A deficient in Afghanistan in 2011/12 (Afghanistan South Asia: a focus on anaemia 2013). This may in part be due to the effective roll out of vitamin A supplementation and associated There has been little progress across South Asia prenatal care since the early 2000s. For example, in resolving hidden malnutrition. Furthermore, two-dose coverage rose in that country from only 58% some countries have seriously high levels of in 2000 to 95% in 2014 (UNICEF, 2015). By contrast, micronutrient deficiencies and others have pockets the prevalence of vitamin A deficiency in women of unusual severity. Therefore, much more needs was recently over 40% in Bangladesh and Pakistan to be understood about the context-specific drivers (Pakistan 2011; ICDDRB et al., 2013). Pregnant women of various micronutrient deficiencies and how to in Pakistan recorded rates around 46% (Pakistan 2011). formulate interventions appropriately to accelerate The average for Pakistan in 2011 masked a wide range, positive change. The underlying and proximate drivers running from around 16% of pregnant women in Sindh of deficiencies are numerous, and their respective with severe VAD (serum level <0.35 μmol/L), to almost roles may vary by country context. 48% in Khyber Pakhtunkhwa, and almost all regions of the country reporting 20% to 30% mild VAD (0.35-0.70 Take anaemia as an example. An analysis of recent μmol/L) (Akhtar et al., 2013). datasets suggests that the factors that are most significantly associated with anaemia in women Finally, iodine deficiency is also unevenly distributed of reproductive age vary by country. In Nepal and across and within the region. At national level, Pakistan, there is wide variation across agro-ecological

Figure 11. Median urinary iodine excretion (µg/L) (population level) by country (Iodine Global Network 2017)

300

250

Adequate iodine status: 100-299 μg/L 200

150 L ) / E (μ g I

U 100

ia s al a an esh tan d ive p an k ist ad u In d Ne kist Lan an gl Bh al a ri gh M P S Af Ban

31 zones: around 40% of Nepalese women in the plains other challenges pertaining to the nutrition situation of are anaemic compared with 25% in the hills and adolescent girls in South Asia (Aguayo & Paintal, 2017). mountains; in Pakistan, anaemia rates range from 67% in sub-tropical Sindh to 30% in mountainous Of course, diets vary widely across the region based Gilgit (Nepal 2011; Pakistan 2011). on wealth, location, culture and religion. Self-professed religion is a household-level factor that was statistically In India and Bangladesh, household wealth is strongly correlated with anaemia among women in Bangladesh, associated with the presence or absence of anaemia Nepal and India. In Bangladesh, a predominately in women. Balarajan et al. (2013) note that anaemia Muslim country, being non-Muslim was associated in India is “socially patterned”, meaning that it is with a higher prevalence of anaemia compared with higher among women from poorer households, being Muslim (>49% vs. 40%) (Kamruzzaman et al., who also tend to have less schooling and to belong 2015), while in India, which is predominately Hindu, to scheduled castes and scheduled tribes. Level of being Muslim (56%) or Christian (51%) was associated household wealth - be it income or assets-based with a slightly lower risk of anaemia than being Hindu - is also correlated with anaemia among children (57%) (Balarajan et al., 2013). In Nepal, which is largely in Pakistan, yet in Nepal, wealth is not statistically Hindu, but also Buddhist, being of the Kirat religion correlated with anaemia for either women or children. (the Rai people of eastern Nepal) was associated with By contrast, in Nepal access to improved sanitation a lower prevalence of anaemia (prevalence ratio =0.61; facilities and to improved drinking water sources are CI 95% 0.39-0.97) compared with being Hindu, after associated with lower anaemia rates. What is more, in controlling for other potential factors (Harding et al., Nepal children in from homes with a de facto female 2017). Perhaps surprisingly, Balarajan et al. (2013) head of household were at a lower risk of anaemia. reported that having any or severe anaemia among In Pakistan, low egg consumption by women and Indian women of reproductive age was lower among current breastfeeding were both linked to higher risk vegetarians than non-vegetarians. If confirmed, the of anaemia (Harding et al., 2017a). latter would indicate that diet alone is not the sole driver of anaemia in South Asia. Other factors must also be taken into account. According to Bhutan’s National Nutrition Survey Education and literacy are additional important 2015, the prevalence of anaemia was lower among predictors of anaemia among adult women, as noted pregnant women (27.3%) than among non-pregnant in the section above. In Bangladesh, women with women (34.9%), probably due to the iron and folate higher education had lower rates of anaemia (33%) supplementation programme during pregnancy than those with secondary (38%) or no education (46%) (Bhutan, 2015). In Sri Lanka, the prevalence of (Kamruzzaman et al., 2015). In India, women with ≥ 13 anaemia among pregnant and non-pregnant women years of education had a lower prevalence of anaemia was similar (both groups at roughly 25%) while in than those who had none (43% vs. 60%) (Balarajan et al., Afghanistan it was higher among pregnant women 2013), and in Pakistan, women’s literacy was associated than among non-pregnant women (44% v. 31% with a lower prevalence of anaemia (Pakistan 2011). respectively) (WHO, 2015). The fact that parity was associated with anaemia in Bangladesh (OR=11.34; Iron depletion was a co-morbid condition with anaemia CI 95% 1.94-66.1) and India (54% anaemia prevalence in Pakistan. Indeed, roughly 27% of non-pregnant among women who have yet to give birth compared women and 44% of children from the 2011 National with over 59% among mothers with five or more Nutrition Survey had low ferritin levels, with a higher children) suggests that more attention needs to be prevalence of low ferritin concentrations among paid to reaching women during pregnancy, but also children in urban than in rural areas, and variation to improving their diets and health over the long-run across provinces (Pakistan 2011). In a recent survey (Merrill et al., 2012; Balarajan et al., 2013). in rural Bangladesh, a high prevalence of anaemia was reported (57%) where iron deficiency was absent Finally, the nutritional status of pre-pregnant adolescent (Merrill et al., 2012). It was found that women exhibited girls cannot be overlooked. Akhtar et al. (2013) high levels of thalassemia, and that they had high iron reported that 90% of adolescent girls in 16 districts of intake from groundwater. Similarly, a recent survey in India were iron deficient (Akhtar et al. 2013), which sets Nepal of pregnant women in the western district of in motion the problem of poor maternal nutrition when Banke found that women of Tharu ethnicity (who are adolescents become pregnant leading to compromised also known to exhibit thalassaemias) had a higher odds birth outcomes and nutrient deficiencies in infants. ratio for anaemia than Brahmin women (OR=2.28; CI A recent comprehensive review highlights these and 95% 1.06-4.90) (Balarajan et al., 2011; Ghosh et al.,

32 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA 2016). Such findings underline the importance of This does not mean that governments should remove understanding the local causes of anaemia to know supplementation from the tool-kit available to how to prevent and treat it. policymakers and other development practitioners where it is cost-effective. Quite the contrary. Supplementation with vitamin A, iron and folic acid, Implications for action and sometimes multiple micronutrients has proven to be very cost effective in appropriate contexts. South Asia’s lack of progress in resolving hidden Currently, LMICs are “doing well” on providing malnutrition over recent decades continues to hamper vitamin A supplements to pre-school age children, the region’s attempts to accelerate development. The with a median population coverage of 79 percent, median cost of iron deficiency for 10 low income and in providing iron and folic acid supplementation countries was found to be approximately US$ 2.32 to pregnant women, with a median coverage of per capita (0.57% of GDP) based on annual physical 78 percent (IFPRI, 2016). Hoddinott et al. (2012) productivity losses alone; when both physical and estimate that the benefit:cost ratio of vitamin A cognitive losses are accounted for this increased to supplementation for children under five in low income US$ 16.78 lost per capita (4% of GDP) (Horton and countries was 12.5, while the figure for iron and folic Ross, 2003). When one considers countries with larger acid supplementation for children aged 6 to 23 months economies, such as India, the economic impact of and mothers was double that at 24 months. In the hidden malnutrition escalates rapidly. For example, case of salt iodization the benefit:cost ratio was 81. Stein & Qaim (2007) calculated the costs just for India, where an estimated 9.3 million DALYs were lost to iron Vitamin A supplementation efforts have been ongoing deficiency anaemia, and zinc, vitamin A and iodine across the region for decades. High dose biannual deficiency in the mid-2000s, equivalent to losing no supplementation of children under five in Pakistan less than 0.8% of annual GDP. Based on India’s 2007 from the 1990s has been linked to a roughly 24% GDP of US$ 1.2 trillion, this translates in 2014 dollars reduction in VAD as a result of effective program to more than US$ 16 billion (based on 2014 GDP of administration and targeting (Siddiqi and Iqbal 2008; US$ 2 trillion). Such a huge figure from a single country Haider and Bhutta, 2011). In Sri Lanka, a vitamin A contributes about one-third of the estimated global cost supplementation programme has been associated with of malnutrition of US$ 3.5 trillion (FAO, 2013). fewer school days missed due to illness compared to the placebo group (Mahawithanage et al., 2007). A significant proportion of such economic cost is due to the effects of poor child nutrition on educational However, with an average coverage of vitamin A performance and subsequent lifetime earnings. That supplementation of only 62% in 2014 (largely driven by is, impaired child growth and intellectual development India’s lower performance), South Asia still has one of across South Asia hinders school enrolment at an the lowest coverage rates globally – the world average appropriate age, causes absenteeism or early drop-out being 69% (UNICEF, 2016a). Even current rates of due to ill-health and/or poor learning, and prevents coverage are at risk of declining in some countries as optimal learning and skills development. Thus, a result of the phasing out of National Immunization there remains a lot to be done to stem the losses at Days. Horton et al. (2008) point out that Bangladesh individual and national levels, and to bolster good relied on accessing hard-to-reach populations with nutrition as an input to faster human and economic vitamin A supplements by combining routine health development in the future. As the International Food services with Child Health Days that linked vitamin Policy Research Institute (2016) puts it, countries of A distribution with immunization. But Child Health the region “…so long synonymous with the problem of Days are increasingly questioned by governments malnutrition, can become a major part of the solution.” seeking places to save on health budget outlays. The implication is that rates of supplementation coverage The solutions will have to come from a stronger more cannot be guaranteed into the future where nation consistent national level political and policy commitment mechanisms for delivery are not secure. to urgently addressing micronutrient malnutrition from multiple directions at once. A nutrient-by-nutrient Iron (plus folic acid) supplementation relies on approach will not suffice and neither will a focus on clinic distribution or community health workers to some but not all micronutrient problems. Similarly, reach remote populations. For example, Nepal has reliance on the distribution of supplements and other demonstrated the feasibility of achieving wide coverage targeted nutrition-specific actions by themselves will by empowering community health workers and by achieve some but not all the gains required. focusing resources on antenatal and postnatal care.

33 In response to a 75% prevalence of anaemia among at 75% in 2014, South Asia was still lagging at 69% pregnant women in 1998, the government began the (UNICEF, 2014). That said, a renewed push is underway Intensification of Maternal and Neonatal Micronutrient among governments of the region towards the universal Program in 2004 with financial support from UNICEF goal of 90% coverage. In India, for example, this and the Micronutrient Initiative. The program targeted included the formation of state level coalitions, the antenatal care visits, iron and folic acid supplementation provision of technical support to enhance the capacity and deworming medication during pregnancy using of salt producers, and the sensitization of wholesalers female community health volunteers. Monitoring and retailers to procure only adequately iodized salt. indicators suggested successful implementation, and Recent data show that conditions are improving, with the prevalence of anaemia among pregnant women household use of adequately iodized salt increasing fell to 42% by 2006 (Pokharel et al., 2011). The rate of from 71% in 2009 to 78% in 2014 (Rah et al., 2015). anaemia among pregnant women continued to fall, reaching 34% by 2010/11 (Nepal 2011). Actions to improve the dietary intake of essential nutrients in early life include approaches to protect, However, supplementation is not enough to achieve promote and support breastfeeding and higher quality sustained full micronutrient nutrition to entire complementary foods and feeding practices. Exclusive populations. In addition to high supplementation breastfeeding immediately after delivery until six months coverage where appropriate, additional actions are of age provides infants with nutrients and protects needed. This requires other interventions focused on them from diseases that may lead to nutrient loss. Only i) additional non-supplement micronutrient delivery 64% of infants in South Asia benefit from exclusive mechanisms, including food fortification, biofortification breastfeeding, and 55% are not breastfed from within and diet diversification, ii) actions to promote the the first hour of life (UNICEF, 2016b). The introduction nutrition security of infants and young children, and iii) of solid, semi-solid or soft foods should occur at the complementary nutrition-sensitive interventions that age of 6 months, yet half of infants (46%) receive foods will support gains in human capital development. too late for optimal growth and development. Only 47% of children aged 6-23 months are fed the minimum The non-supplement approaches to delivering number of meals a day and only 23% are eating the micronutrients have also been widely implemented minimum number of food groups, leaving the vast in the region. For example, vegetable oil and ghee majority vulnerable to micronutrient deficiencies. are fortified with vitamin A and D in Afghanistan since 2014, and there have been several attempts to All countries in South Asia recognize infant and young implement iron fortification of wheat flour. India and child feeding as a strategic development priority and Pakistan have pursued the fortification of wheat flour have a range of policy and legislation support for with iron, folic acid and vitamin B12, while Nepal has breastfeeding (Thow et al., 2017). However, gaps in also implemented mandatory wheat flour fortification the design and implementation of these policies and with iron and folic acid. Several countries in the region, legislation remain, and there is need to strengthen such as Nepal are testing the potential of point-of-use health system support for counselling and education or ‘home’ fortification; that is, the distribution of small on breastfeeding, as well as workplace, community packets of multiple micronutrient powders mixed into and family support to breastfeeding mothers. There the complementary food offered to infants after 6 is relatively little policy emphasis on complementary months of age (Ip et al., 2009; Sazawal et al., 2014). feeding in the region (Thow et al., 2017), and greater An additional means of delivering micronutrients into government leadership, prioritization and financing is the local food matrix at the point of consumption is needed to address the nutrient gap in children’s diets through small dose lipid nutrient supplements which (UNICEF, 2016b). Progress on complementary feeding are highly nutrient-dense oil-based spreads that can be requires actions by multiple sectors and stakeholder blended into an infant’s meal (Lopriore et al., 2004). groups that can collectively ensure a household has access to affordable and nutritious food, and whether Salt iodization is another widely adopted approach to caregivers have the skills, knowledge and other tackling iodine deficiency. As mentioned above, the resources to prepare frequent nutritious and safe foods. benefit:cost ratio of salt iodization is very high at 81 (Hoddinott et al., 2012). However, policy and regulatory/ The promotion of dietary diversification is a food- monitoring mechanisms are required to ensure that based approach to increasing intake of micronutrients. adequate iodization is actually occurring and that Diet diversification represents a large-scale but as families have access to such salt. While the global yet unfulfilled opportunity. While the consumption average for adequately iodized salt consumption stood of foods high in beta carotene by vitamin A deficient

34 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA consumers has been shown to have a significant It is the complementary among a set of actions that positive association with serum retinol concentrations offers the best promise for accelerated change. (Fujita et al., 2012; Webb and Kennedy, 2014), there Such complementarity has been underlined by FAO, is limited empirical evidence of cost-effective policy or which calculated that an annual investment of US$ programme interventions that succeed in sustaining 1.2 billion in the combination of a) micronutrient diet diversity across a range of nutrients (Thompson supplementation, b) staple food fortification and c) and Amoroso, 2014). What is more, Mark et al. (2016) biofortification of staple crops, would result in “better calculated that the current food supply in most of health, fewer deaths and increased future earnings” South Asia is not adequately composed to deliver valued at over US$ 15 billion per year: a 13-to-1 benefit seven of the most important nutrients required for to cost ratio (FAO, 2013). sound nutrition and health. For example, India has the lowest supply (in foods) of vitamin B-12 and The second set of actions needed to complete the zinc, Pakistan has the lowest supply of thiamine, and work of resolving micronutrient deficiencies in South Bangladesh the lowest supply of vitamin A, riboflavin Asia involve nutrition sensitive approaches that and calcium. This does not mean that promotion support human capital formation. As noted by Ruel of agricultural diversity and market access for all et al. (2013) “nutrition-sensitive programmes can help consumers cannot achieve such a goal, but more scale up nutrition-specific interventions and create a needs to be done to demonstrate effective approaches stimulating environment in which young children can that can be adopted by governments across the region grow and develop to their full potential.” This means (Sibhatu et al., 2015). appropriately supporting agriculture, social safety nets, early child development activities, water, hygiene and The most recent food-based approach to delivering sanitation (WASH) activities, health promotion and micronutrients is biofortification. An expert consultation universal education. It is critically important that all suggested in 2003 that “probably as many as 30 children in South Asia gain access to the right kinds biologically distinct types of foods, with the emphasis of care, health services and diets; that agricultural on plant foods, are required for healthy diets” (WHO programmes and social safety nets be designed to and FAO, 2003). Achieving access to a diverse diet is reach the most vulnerable and protect them effectively not easy for poor people, and it represents a major from negative effects of food price and other shocks. policy challenge for developing country governments. It is equally important that national and local policies Many consumers are constrained by income level support women’s empowerment, literacy and formal or distance to markets to eating a monotonous education. nutritionally-inadequate diet. Biofortification represents a way to ‘deliver more’ in the context of dietary It is also essential that policymakers should promote constraints. It is pursued through conventional crop poverty- and inequity-reducing economic growth. breeding techniques that are used to identify varieties Gross national income growth is critical to enabling with high concentration of desired micronutrients. governments to make the investments in the nutrition Those varieties are then cross-bred with high-yielding specific and nutrition sensitive programmes outlined varieties to develop biofortified varieties that have high above. Yet, while economic growth does tend to reduce levels of, for example, zinc or betacarotene, in addition most forms of malnutrition over time, the decrease to other productivity traits desired by farmers (Global is generally far slower and less equitably distributed Panel, 2015). than the corresponding poverty reduction might have predicted (Ruel et al., 2013). In other words, malnutrition Rice biofortified with zinc was released to farmers in general, and hidden malnutrition in particular, “does in Bangladesh in 2013 (Chowdhury, 2014). These not take care of itself” (Webb and Block, 2004). biofortified varieties have a zinc content that is 30% higher than local varieties, and it matures faster than South Asia’s future depends on ensuring that policies some traditional strains (HarvestPlus, 2014). The across the many sectors with responsibility for capacity to scale up adoption of high-zinc rice has promoting good nutrition are mutually-supportive, that still to be demonstrated; but since poor consumers public programming is evidence-based and measurably in Bangladesh rely heavily on large amounts of rice cost-effective, and that the benefits of accelerated daily, such an approach could significantly improve improvements in nutrition are equitably shared. Human dietary intake of zinc. The same applies to biofortified capital has to be preserved, protected and nurtured pearl millet, with higher iron and zinc content, which as a South Asia-wide goal. For this to happen, hidden is already being grown in Maharashtra state of India malnutrition has also to be embraced as a South Asia- (HarvestPlus, 2014). wide problem to be resolved as a matter of urgency.

35 36 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA REFERENCES

Abubakar, A., Van de Vijver, F., Van Baar, A., Akachi, Y., & Canning, D. (2015). Inferring the economic Mbonani, L., Kalu, R., Newton, C., et al. (2008). standard of living and health from cohort height: Socioeconomic status, anthropometric status, and Evidence from modern populations in developing psychomotor development of Kenyan children from countries. Economics and Human Biology, 19, 114-128. resource-limited settings: a path-analytic study. Early Human Development, 84, 613-621. Akhtar, S. (2013). Zinc status in South Asian populations – an update. Journal of Health, Population Adair, L.S., Fall, C.H.D., Osmond, C., Stein, A.D., and Nutrition, 31 (2), 139-149. Martorell, R., Ramirez-Zea, M., et al. (2013). Associations of linear growth and relative weight Akhtar, S. (2015). Malnutrition in South Asia - a critical gain during early life with adult health and human reappraisal. Critical Reviews in Food Science and capital in countries of low and middle inhcome: Nutrition, 56, 2320-2330. findings from five birth cohort studies. Lancet, 382, 525-534. Akhtar, S., Ahmed, A., Randhawa, M.A., Atukorala, S., Arlappa, N., Ismail, T., & Ali, Z. (2013). Prevalence of Adair, L.S., Popkin, B.M., Akin, J.S., Guilkey, D.K., vitamin A deficiency in South Asia: causes, outcomes, Gultiano, S., Borja, J., et al. (2011). Cohort profile: and possible remedies. Journal of Health Population the Cebu longitudinal health and nutrition survey. and Nutrition, 31 (4), 413-423. International Journal of Epidemiology, 40, 619-625. Alberto Camargo-Figuera, F., Barros, A.J.D., Santos, Adolph, K.E., Vereijken, B., & Shrout, P.E. (2003). I.S., Matijasevich, A., & Barros, F.C. (2014). Early life What changes in infant walking and why. Child determinants of low IQ at age 6 in children from the Development, 74, 475-497. 2004 Pelotas Birth Cohort: a predictive approach. BMC Pediatrics, 14, 308. doi: 10.1186/s12887-014-0308-1, Afghanistan (Government of Afghanistan & UNICEF) (2013). National Nutrition Survey Afghanistan Alcázar, L., Ocampo, D., Huamán-Espino, L., & Pablo 2013. Kabul, Afghanistan: Ministry of Public Health, Aparco, J. (2013). Economic impact of chronic, acute Government of Afghanistan. and global malnutrition in Peru. Revista Peruana de Medicina Experimental y Salud Pública, 30, 569-574. African Union Commission/World Food Programme. (2015). The Cost of Hunger in Malawi: Social Alderman, H., Behrman, J.R., Grantham-McGregor, and Economic Impacts of Child Undernutrition in S., Lopez-Boo, F., & Urzua, S. (2014). Economic Malawi: Implications on National Development and perspectives on integrating early child stimulation Vision 2020. Addis Ababa. with nutritional interventions. Annals of the New York Academy of Sciences, 1308, 129-138. Aguayo, V.M., & Menon, P. (2016). Stop stunting: improving child feeding, women’s nutrition and Alderman, H., Hoogeveen, H., & Rossi, M. (2006). household sanitation in South Asia. Maternal and Reducing child malnutrition in Tanzania: Combined Child Nutrition, 12, 3-11. effects of income growth and program interventions. Economics and Human Biology, 4, 1-23. Aguayo, V.M., Paintal, K. (2017). Nutrition in adolescent girls in South Asia. British Medical Andersson, M., Karumbunathan, V., & Zimmermann, M. Journal, 357, j1309. doi: 10.1136/bmj.j1309. (2012). Global iodine status in 2011 and trends over the past decade. Journal of Nutrition, 142, 744-750. Aitsi-Selmi, A. (2014). Households with a stunted child and obese mother: trends and child feeding Arimond, M., Ruel, M.T. (2004). Dietary diversity is practices in a middle-income country. Maternal and associated with child nutritional status: evidence Child Health Journal, 19 (6), 1284-91. doi: 10.1007/ from 11 demographic and health surveys. Journal of s10995-014-1634-5. Nutrition 134, 2579e2585.

37 Arora, S. (2001). Health, human productivity, and long- Behrman, J.R., & Birdsall, N. (1983). The quality of term economic growth. Journal of Economic History, schooling: quantity alone is misleading. American 61, 699-749. Economic Review, 73, 928-946.

Aubuchon-Endsley, N.L., Grant, S.L., Berhanu, G., Behrman, J.R., Bhalotra, S., Deolalikar, A.B., Thomas, D.G., Schrader, S.E., Eldridge, D., et al. Laxminarayan, R., & Nandi, A. (2015). Human (2011). Hemoglobin, growth, and attention of infants in capital and productivity benefits of early childhood Southern Ethiopia. Child Development, 82, 1238. nutritional interventions. In: Disease Control Priorities, 3rd edition (DCP3). Seattle: University of Aurino, E., & Burchi, F. (2014). Children’s Washington, Department of Global Health. Mmultidimensional health and medium-run cognitive skills in low-and middle-income countries. Young Lives Berger, A. (2001). Insulin-like growth factor and Working Paper, 129. Oxford: Young Lives. cognitive function. British Medical Journal, 322, 203-203. Avan, B., Richter, L.M., Ramchandani, P.G., Norris, S.A., & Stein, A. (2010). Maternal postnatal depression Berkman, D.S., Lescano, A.G., Gilman, R.H., Lopez, and children’s growth and behaviour during the early S.L., & Black, M.M. (2002). Effects of stunting, years of life: exploring the interaction between physical diarrhoeal disease, and parasitic infection during and mental health. Archives of Disease in Childhood, infancy on cognition in late childhood: a follow-up 95, 690-695. study. Lancet, 359, 564-571.

Bagriansky, J., Champa, N., Pak, K., Whitney, S., & Bershteyn, A., Lyons, H.M., Sivam, D., & Myhrvold, Laillou, A. (2014). The economic consequences of N.P. (2015). Association between economic growth malnutrition in Cambodia, more than 400 million US and early childhood nutrition. Lancet Global Health, dollar lost annually. Asia Pacific Journal of Clinical 3, e79-e80. Nutrition, 23, 524. Bhandari, S., & Banjara, M. (2015). Micronutrients Balarajan, Y., Fawzi, W., & Subramanian, S. (2013). deficiency, a hidden hunger in Nepal: prevalence, Changing patterns of social inequalities in anaemia causes, consequences, and solutions. International among women in India: cross-sectional study using Scholarly Research Notices, 276469. doi: nationally representative data. British Medical Journal 10.1155/2015/276469. Open, 3, e002233. doi: 10.1136/bmjopen-2012-002233. Bharati, S, Pal, M., Som, S., & Bharati, P. (2015). Balarajan, Y., Ramakrishnan, U., Özaltin, E., Shankar, Temporal trend of anaemia among reproductive- A., & Subramanian, S. (2011). Anaemia in low-income aged women in India. Asia Pacific Journal of Public and middle-income countries. Lancet, 378, 2123–35. Health, 27 (2), 1193-1207.

Bangladesh (National Institute of Population Research Bhutan (Ministry of Health) (2013). National and Training) (1996). Demographic and Health Survey Nutrition Survey. Thimphu, Bhutan: Department of 1996-7. Dhaka, Bangladesh: National Institute of Public Health, Ministry of Health. Population Research and Training. Bhutta, Z.A., Ahmed, T., Black, R.E., Cousens, S., Bangladesh (National Institute of Population Research Dewey, K., Giugliani, E., et al. (2008). What works? and Training) (2011). Demographic and Health Survey Interventions for maternal and child undernutrition 2011. Dhaka, Bangladesh: National Institute of and survival. Lancet, 371, 417-440. Population Research and Training. Bhutta, Z.A., Das, J.K., Rizvi, A., Gaffey, M.F., Bangladesh (National Institute of Population Research Walker, N., Horton, S., et al. (2013). Evidence-based and Training) (2014). Demographic and Health Survey interventions for improvement of maternal and 2014. Dhaka, Bangladesh: National Institute of child nutrition: what can be done and at what cost? Population Research and Training. Lancet, 382, 452-477.

Barker, D.J.P., Eriksson, J.G., Forsen, T., & Osmond, Black, M.M. (1998). Zinc deficiency and child C. (2005). Infant growth and income 50 years later. development. American Journal of Clinical Nutrition, Archives of Disease in Childhood, 90, 272-273. 68, 464S-469S.

38 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA Black, M.M., Perez-Escamilla, R., & Fernandez-Rao, S. Carba, D.B., Tan, V.L., & Adair, L.S. (2009). Early (2015). Integrating Nutrition and Child Development childhood length-for-age is associated with the Interventions: Scientific Basis, Evidence of Impact, work status of Filipino young adults. Economics and Implementation Considerations. Advances in and Human Biology, 7, 7-17. Nutrition: An International Review Journal, 6, 852-859. Casale, D., Desmond, C., & Richter, L. (2014). The Black, M.M., Walker, S.P., Fernald, L.C.H., Andersen, association between stunting and psychosocial C.T., DiGirolamo, A.M., Lu, C., et al. (2016). Early development among preschool children: a study childhood development coming of age: science using the South African Birth to Twenty cohort through the life course. Lancet, 389, 77-90. data. Child: Care, Health and Development, 40, 900-910. Black, R.E., Victora, C.G., Walker, S.P., Bhutta, Z.A., Christian, P., de Onis, M., et al. (2013). Maternal and Case, A., & Paxson, C. (2008). Stature and Status: child undernutrition and overweight in low-income Height, Ability, and Labor Market Outcomes. and middle-income countries. Lancet, 382 (9890), Journal of Political Economy, 116, 499-532. 427–451. doi: 10.1016/S0140-6736(13)60937-X. Case, A., Fertig, A., & Paxson, C. (2005). Black, R.E., Allen, L.H., Bhutta, Z.A., Caulfield, L.E., De The lasting impact of childhood health and Onis, M., Ezzati, M., et al. (2008). Maternal and child circumstance. Journal of Health Economics, 24, undernutrition: global and regional exposures and 365-389. health consequences. Lancet, 371, 243-260. Chakraborty, S., Papageorgiou, C., & Sebastián, Block, S. (2007). Maternal nutrition knowledge versus F.P. (2010). Diseases, infection dynamics, and schooling as determinants of child micronutrient development. Journal of Monetary Economics, 57, status. Oxford Economic Papers, 59, 330-353. 859-872.

Bloom, D., Cafiero-Fonseca, E.T., McGovern, M.E., Chang, S.M., Walker, S.P., Grantham-McGregor, S., Prettner, K., Stanciole, A., Weiss, J., et al. (2014). & Powell, C.A. (2002). Early childhood stunting and The Macroeconomic Impact of Non-Communicable later behaviour and school achievement. Journal Diseases in China and India: Estimates, Projections, of Child Psychology and Psychiatry, and Allied and Comparisons. Journal of the Economics of Disciplines, 43, 775-783. Ageing, 4, 100-111. Chang, S.M., Walker, S.P., Grantham-McGregor, S., Bogale, A., Stoecker, B.J., Kennedy, T., Hubbs-Tait, & Powell, C.A. (2010). Early childhood stunting and L., Thomas, D., Abebe, Y., et al. (2013). Nutritional later fine motor abilities. Developmental Medicine status and cognitive performance of mother-child and Child Neurology, 52, 831-836. pairs in Sidama, Southern Ethiopia. Maternal and Child Nutrition, 9, 274-284. Cheung, Y.B., & Ashorn, P. (2010). Continuation of linear growth failure and its association with Boo, F.L., Palloni, G., & Urzua, S. (2014). Cost–benefit cognitive ability are not dependent on initial length- analysis of a micronutrient supplementation and early for-age: a longitudinal study from 6 months to 11 childhood stimulation program in Nicaragua. Annals years of age. Acta Paediatrica, 99, 1719-1723. of the New York Academy of Sciences, 1308, 139-148. Cheung, Y.B., Gladstone, M., Maleta, K., Duan, X., Britto, P.R., Lye, S.J., Proulx, K., Yousafzai, A.K., & Ashorn, P. (2008). Comparison of four statistical Matthews, S.G., Vaivada, T., et al. (2016). Nurturing approaches to score child development: a study care: promoting early childhood development. Lancet, of Malawian children. Tropical Medicine and 389, 91-102. International Health, 13, 987-993.

Bryce, J., Coitinho, D., Darnton-Hill, I., Pelletier, Cheung, Y.B., Yip, P., & Karlberg, J. (2001). D., Pinstrup-Andersen, P., Maternal, et al. (2008). Fetal growth, early postnatal growth and motor Maternal and child undernutrition: effective action at development in Pakistani infants. International national level. Lancet, 371, 510-526. Journal of Epidemiology, 30, 66-72.

39 Chowdhury, M. (2014). Address by Ms. Matia Currie, J., & Vogl, T. (2013). Early-Life Health and Adult Chowdhury, Honorable Minister for Agriculture, Circumstance in Developing Countries. Annual Review Government of People’s Republic of Bangladesh, of Economics, 5, 1-36. November 2014, Rome, Italy. http://www.harvestplus. org/sites/default/files/Bangladesh%20Statement%20 Dalgaard, C.-J., & Strulik, H. (2015). The physiological at%20ICN2.pdf foundations of the wealth of nations. Journal of Economic Growth, 20, 37-73. Christian, P., Mullany, L.C., Hurley, K.M., & Katz, J. (2015). Nutrition and maternal, neonatal, and child de Onis, M. (2006). WHO child growth standards: health. Seminars in Perinatology, 39, 361-371. length/height-for-age, weight-for-age, weight-for- length, weight-for-height and body mass index-for-age. Coly, A.N., Milet, J., Diallo, A., Ndiaye, T., Bénéfice, Geneva: World Health Organization. E., Simondon, F., et al. (2006). Preschool stunting, adolescent migration, catch-up growth, and adult Dercon, S., & Porter, C. (2014). Live aid revisited: long- height in young Senegalese men and women of rural term impacts of the 1984 Ethiopian famine on children. origin. Journal of Nutrition, 136, 2412-2420. Journal of the European Economic Association, 12, 927-948. Corsi, D.J., Mejía-Guevara, I., & Subramanian, S.V. (2016). Risk factors for chronic undernutrition among Development Initiatives (2017). Global Nutrition Report children in India: Estimating relative importance, 2017: Nourishing the SDGs. Bristol, UK: Development population attributable risk and fractions. Social Initiatives. Science and Medicine, 157, 165-185. Di Cesare, M., Bhatti, Z., Soofi, S.B., Fortunato, L., Crookston, B.T., Dearden, K.A., Alder, S.C., Merrill, Ezzati, M., & Bhutta, Z.A. (2015). Geographical and R.M., Dickerson, T.T., Stanford, J.B., et al. (2010a). socioeconomic inequalities in women and children’s Peruvian children who recover from early stunting and nutritional status in Pakistan in 2011: an analysis of well-nourished children demonstrate similar cognition. data from a nationally representative survey. Lancet Federation of American Societies for Experimental Global Health, 3, e229-e239. Biology Journal, 24 (1 Supplement), 734-6. DiGirolamo, A.M., Stansbery, P., & Lung’aho, M. Crookston, B.T., Dearden, K.A., Alder, S.C., Porucznik, (2014). Advantages and challenges of integration: C.A., Stanford, J.B., Merrill, R.M., et al. (2011). Impact opportunities for integrating early childhood of early and concurrent stunting on cognition. Maternal development and nutrition programming. Annals of the and Child Nutrition, 7, 397-409. New York Academy of Sciences, 1308, 46-53.

Crookston, B.T., Penny, M.E., Alder, S.C., Dickerson, Dua, T., Tomlinson, M., Tablante, E., Britto, P., Yousfzai, T.T., Merrill, R.M., Stanford, J.B., et al. (2010b). Children A., Daelmans, B., et al. (2016). Global research who recover from early stunting and children who are priorities to accelerate early child development in the not stunted demonstrate similar levels of cognition. sustainable development era. Lancet Global Health, 4, Journal of Nutrition, 140, 1996-2001. e887-e889.

Crookston, B.T., Schott, W., Cueto, S., Dearden, K.A., Elliott, J., & Shepherd, P. (2006). Cohort profile: 1970 Engle, P., Georgiadis, A., et al. (2013). Postinfancy growth, British birth cohort (BCS70). International Journal of schooling, and cognitive achievement: Young Lives. Epidemiology, 35, 836-843. American Journal of Clinical Nutrition, 98, 1555-1563. Engle, P.L., & Fernandez, P.D. (2010). INCAP studies of Cunningham, K., Singh, A., Headey, D., Pandey Rana, malnutrition and cognitive behavior. Food and Nutrition P., & Karmacharya, C. (2016). Reaching new heights: Bulletin, 31, 83-94. 20 years of nutrition progress in Nepal. In: Nourishing Millions: Stories of Change in Nutrition, Gillespie, S., Espo, M., Kulmala, T., Maleta, K., Cullinan, T., Salin, Hodge, J., Yosef, S., & Pandya-Lorch, R. (editors). M.L., & Ashorn, P. (2002). Determinants of linear Washington, DC: International Food Policy Research growth and predictors of severe stunting during infancy Institute. in rural Malawi. Acta paediatrica, 91, 1364-1370.

40 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA FAO (2013). State of Food and Agriculture. Rome, Gaiha, R., & Kulkarni, V. (2005). Anthropometric failure Italy: Food and Agriculture Organization. and persistence of poverty in rural India. International Review of Applied Economics, 19, 179-197. Fernald, L., & Grantham-McGregor, S.M. (1998). Stress response in school-age children who have Galler, J.R., Bryce, C., Waber, D.P., Zichlin, M.L., been growth retarded since early childhood. American Fitzmaurice, G.M., & Eaglesfield, D. (2012). Journal of Clinical Nutrition, 68, 691-698. Socioeconomic outcomes in adults malnourished in the first year of life: a 40-year study. Pediatrics, 130, e1-e7. Fernald, L., & Hidrobo, M. (2011). Effect of Ecuador’s cash transfer program (Bono de Desarrollo Humano) Gandhi, M., Ashorn, P., Maleta, K., Teivaanmaki, T., on child development in infants and toddlers: a Duan, X., & Cheung, Y.B. (2011). Height gain during randomized effectiveness trial. Social Science and early childhood is an important predictor of schooling Medicine, 72, 1437-1446. and mathematics ability outcomes. Acta Paediatrica, 100, 1113-1118. Fernald, L., Gertler, P.J., & Neufeld, L.M. (2008). Role of cash in conditional cash transfer programmes for Gertler, P., Heckman, J., Pinto, R., Zanolini, A., child health, growth, and development: an analysis of Vermeersch, C., Walker, S., et al. (2014). Labor market Mexico’s Oportunidades. Lancet, 371, 828-837. returns to an early childhood stimulation intervention in Jamaica. Science, 344, 998-1001. Fernald, L., Gertler, P.J., & Neufeld, L.M. (2009). 10- year effect of Oportunidades, Mexico’s conditional Ghosh, S., Andrews-Trevino, J., Davis, D., Shrestha, cash transfer programme, on child growth, cognition, R., Bhattarai, A., KC A, Paudel, M., et al. (2016). language, and behaviour: a longitudinal follow-up Anemia status in pregnant women in Banke, Nepal. study. Lancet, 374, 1997-2005. Poster for 4th Scientific Symposium, July 18-19, 2016. Kathmandu, Nepal. Fernald, L., Neufeld, L.M., Barton, L.R., Schnaas, L., Rivera, J., & Gertler, P.J. (2006). Parallel deficits in Glewwe, P., & King, E.M. (2001). The impact of early linear growth and mental development in low-income childhood nutritional status on cognitive development: Mexican infants in the second year of life. Public does the timing of malnutrition matter? World Bank Health Nutrition, 9, 178-186. Economic Review, 15, 81-113.

Fernández, A., & Martínez, R. (2007). Model for Glewwe, P., Jacoby, H.G., & King, E.M. (2001). Early Analysing the Social and Economic Impact of childhood nutrition and academic achievement: a Child Undernutrition in Latin America: Economic longitudinal analysis. Journal of Public Economics, 81, Commission for Latin America and the Caribbean. 345-368. New York: United Nations. Global Burden of Disease Study 2013 Collaborators Fernandez-Rao, S., Hurley, K.M., Nair, K.M., (2015). Global, regional, and national incidence, Balakrishna, N., Radhakrishna, K.V., Ravinder, P., prevalence, and years lived with disability for 301 acute et al. (2013). Integrating nutrition and early child- and chronic diseases and injuries in 188 countries, development interventions among infants and 1990–2013: a systematic analysis for the Global Burden preschoolers in rural India. Annals of the New York of Disease Study 2013. Lancet, 386 (9995), 743–800. Academy of Sciences, 1308, 218-231. doi: 10.1016/S0140-6736(15)60692-4.

Fink, G., Günther, I., & Hill, K. (2011). The effect of Global Panel (2015). Biofortification: An Agricultural water and sanitation on child health: evidence from Investment for Nutrition. Policy Brief No. 1. London, the Demographic and Health Surveys 1986–2007. UK: Global Panel on Agriculture and Food Systems for International Journal of Epidemiology, 40, 1196-1204. Nutrition.

Fujita, M., Lo, Y., & Baranski, J. (2012). Dietary Gluckman, P.D., Hanson, M.A., & Beedle, A.S. (2007). diversity score is a useful indicator of vitamin A status Early life events and their consequences for later of adult women in Northern Kenya. American Journal disease: a life history and evolutionary perspective. of Human Biology, 24, 829-834. American Journal of Human Biology, 19, 1-19.

41 Grantham-McGregor, S.M. (1984). Chronic Haas, J.D., Murdoch, S., Rivera, J., & Martorell, R. undernutrition and cognitive abilities. Human (1996). Early nutrition and later physical work capacity. Nutrition-Clinical Nutrition, 38, 83-94. Nutrition Reviews, 54, S41-S48.

Grantham-McGregor, S.M. (1993). Assessments of Haddad, L. (2015). Equity: not only for idealists. the effects of nutrition on mental development and Development Policy Review, 33, 5-13. behavior in Jamaican studies. American Journal of Clinical Nutrition, 57, 303S-309S. Haddad, L., Alderman, H., Appleton, S., Song, L., & Yohannes, Y. (2003). Reducing child malnutrition: Grantham-McGregor, S.M. (1995). A review of How far does income growth take us? World Bank studies of the effect of severe malnutrition on mental Economic Review, 17, 107-131. development. Journal of Nutrition, 125, 2233S-2238S. Haddad, L., Masset, E., & Smith, L. (2015). Does Grantham-McGregor, S.M., & Baker-Henningham, The Quality of Income Growth Affect Child Nutrition H. (2005). Review of the evidence linking protein Status? In L. Haddad, K. Hiroshi, & M. Nicolas (Eds.), and energy to mental development. Public Health Growth is Dead, Long Live Growth: The Quality of Nutrition, 8, 1191-1201. Economic Growth and Why it Matters. Tokyo, Japan: Agence Française de Développement (AFD), Institute Grantham-McGregor, S.M., Cheung, Y.B., Cueto, of Development Studies (IDS), Japan International S., Glewwe, P., Richter, L., & Strupp, B. (2007). Cooperation Agency (JICA). Developmental potential in the first 5 years for children in developing countries. Lancet, 369, 60-70. Haider, B., & Bhutta, Z. (2011). Neonatal vitamin A supplementation for the prevention of mortality and Grantham-McGregor, S.M., Fernald, L.C.H., Kagawa, morbidity in term neonates in developing countries. R.M.C., & Walker, S. (2014). Effects of integrated Cochrane Database Systematic Reviews, CD006980. child development and nutrition interventions on child doi: 10.1002/14651858.CD006980.pub2. development and nutritional status. Annals of the New York Academy of Sciences, 1308, 11-32. Halim, N., Spielman, K., & Larson, B. (2015). The economic consequences of selected maternal and early Grantham-McGregor, S.M., Powell, C.A., Walker, S.P., childhood nutrition interventions in low-and middle- & Himes, J.H. (1991). Nutritional supplementation, income countries: a review of the literature, 2000-2013. psychosocial stimulation, and mental development of BioMed Central Women’s Health, 15, 33. stunted children: the Jamaican Study. Lancet, 338, 1-5. Hamadani, J.D., Tofail, F., Cole, T., & Grantham- McGregor, S. (2012). The relation between age of Grantham-McGregor, S.M., Walker, S.P., Chang, attainment of motor milestones and future cognitive S.M., & Powell, C.A. (1997). Effects of early childhood and motor development in Bangladeshi children. supplementation with and without stimulation on later Maternal and Child Nutrition, 9, 89-104. development in stunted Jamaican children. American Journal of Clinical Nutrition, 66, 247-253. Hamadani, J.D., Tofail, F., Huda, S.N., Alam, D.S., Ridout, D.A., Attanasio, O., et al. (2014). Cognitive Grantham-McGregor, S.M., Walker, S.P., Himes, deficit and poverty in the first 5 years of childhood in J.H., & Powell, C.A. (1996). Stunting and mental Bangladesh. Pediatrics, 134, e1001-e1008. development in children. Nutrition Research, 16, 1821-1828. Handal, A.J., Lozoff, B., Breilh, J., & Harlow, S.D. (2007). Sociodemographic and nutritional correlates Gross, R., Giay, T., Sastroamidjojo, S., Schultink, W., & of neurobehavioral development: a study of young Lang, N.T. (2000). Premature complementary feeding children in a rural region of Ecuador. Revista is associated with poorer growth of Vietnamese Panamericana de Salud Pública, 21, 292-300. children. Journal of Nutrition, 130, 2683-2690. Harding, K., Aguayo, V., Namirebe, G., & Webb, P. Haas, J.D., Martinez, E.J., Murdoch, S., & Conlisk, (2017a). Determinants of anaemia among women and E. (1995). Nutritional Supplementation during the children in Nepal and Pakistan: An analaysis of recent Preschool Years and P. Journal of Nutrition, 125, 1078. national survey data. Maternal and Child Nutrition, doi: 10.1111/mcn.12478

42 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA Harding, K., Aguayo, V. & Webb, P. (2017b). Hidden Horton, S., & Ross, J. (2003). The economics of iron Hunger in South Asia: a review of recent trends deficiency. Food policy, 28 (1), 51-75. and persistent challenges. Public Health Nutrition, doi:10.1017/S1368980017003202 Horton, S., & Steckel, R.H. (2011). Malnutrition: Global Economic Losses Attributable to Malnutrition Harding, K., Aguayo, V.M., Masters, W.A. & Webb, 1900–2000 and Projections to 2050. Copenhagen: P. (2018). Education and micronutrient deficiencies: Copenhagen Consensus on Human Challenges. an ecological study exploring interactions between women’s schooling and children’s micronutrient Horton, S., Begin, F., & Greig, A., & Lakshman, A. status. BMC Public Health 18, 470. doi: 10.1186/ (2008). Best practice Paper: Micronutrient Supplements s12889-018-5312-1. for Child Survival (Vitamin A and Zinc). Paper prepared for 2008 Global Copenhagen Consensus. Washington, Harttgen, K., Klasen, S., & Vollmer, S. (2013). D.C.: International Food Policy Research Institute. Economic Growth and Child Undernutrition in sub‐ Saharan Africa. Population and Development Review, Hunt, J. (2005). The potential impact of reducing 39, 397-412. global malnutrition on poverty reduction and economic development. Asia Pacific Journal of Clinical Nutrition, HarvestPlus (2014). Biofortification Progress Brief: 14(S), 10-38. August 2014. Washington, D.C.: HarvestPlus. http:// www.harvestplus.org/sites/default/files/Biofortification_ Hurley, K.M., Yousafzai, A.K., & Lopez-Boo, F. (2016). Progress_Briefs August2014_WEB_2.pdf Early child development and nutrition: a review of the benefits and challenges of implementing integrated Headey, D.D. (2013). Developmental drivers of interventions. Advances in Nutrition, 7, 357-363. nutritional change: a cross-country analysis. World Development, 42, 76-88. Huxley, R.R., Shiell, A.W., & Law, C.M. (2000). The role of size at birth and postnatal catch‐up growth in Heltberg, R. (2009). Malnutrition, poverty, and determining systolic blood pressure: a systematic review economic growth. Health Economics, 18, S77-S88. of the literature. Journal of Hypertension, 18, 815-831.

Hoddinott, J., Alderman, H., Behrman, J.R., Haddad, ICDDR,B (International Centre for Diarrhoeal Disease L., & Horton, S. (2013a). The economic rationale for Research, Bangladesh), GAIN (Global Alliance investing in stunting reduction. Maternal and Child for Improved Nutrition), UNICEF (2013). National Nutrition, 9, 69-82. Micronutrient Status Survey 2011-2012. Dhaka, Bangladesh: International Centre for Diarrhoeal Hoddinott, J., Behrman, J.R., & Martorell, R. (2005). Disease Research, Bangladesh. Labor force activities and income among young Guatemalan adults. Food and Nutrition Bulletin, 26, IFPRI (International Food Policy Research Institute) 98S-109S. (2015). Global Nutrition Report 2015: Actions and Accountability to Advance Nutrition and Sustainable Hoddinott, J., Behrman, J.R., Maluccio, J.A., Melgar, Development. Washington, D.C: International Food P., Quisumbing, A.R., Ramirez-Zea, M., et al. (2013b). Policy Research Institute. Adult consequences of growth failure in early childhood. American Journal of Clinical Nutrition, 98, IFPRI (International Food Policy Research Institute) 1170-1178. (2016). Global Nutrition Report 2016: From Promise to Impact: Ending Malnutrition by 2030. Washington, DC: Hoddinott, J., Maluccio, J.A., Behrman, J.R., Flores, R., International Food Policy Research Institute. & Martorell, R. (2008). Effect of a nutrition intervention during early childhood on economic productivity in India (Government of India) (1992). National Family Guatemalan adults. Lancet, 371, 411-416. Health Survey, 1992-3. Bombay: International Institute for Population Sciences. Hoddinott, J., Rosegrant, M., & Torero, M. (2012). Investments to reduce hunger and undernutrition. India (Government of India) (2006). National Paper prepared for 2012 Global Copenhagen Family Health Survey (NFHS-3), 2005–6. Mumbai: Consensus. Washington, D.C.: International Food International Institute for Population Sciences, Ministry Policy Research Institute. of Health and Family Welfare.

43 Iodine Global Network (2017). Global Scorecard of with mixed crop-livestock production systems, Ghinchi, Iodine Nutrition in 2017 in the general population. http:// central Ethiopia. Ethiopian Medical Journal, 41, 151-161. www.ign.org/cm_data/IGN_Global_Scorecard_AllPop_ and_PW_May2017.pdf, last accessed February 5, 2018. Kordas, K., Lopez, P., Rosado, J.L., Garcia Vargas, G., Alatorre Rico, J., Ronquillo, D., et al. (2004). Blood Ip, H., Hyder, S., Haseen, F., Rahman, M., & Zlotkin, S. lead, anemia, and short stature are independently (2009). Improved adherence and anaemia cure rates associated with cognitive performance in Mexican with flexible administration of micronutrient Sprinkles: school children. Journal of Nutrition, 134, 363-371. a new public health approach to anaemia control. European Journal of Clinical Nutrition, 63 (2), 165-72. Kossmann, J., Nestel, P., Herrera, M., El Amin, A., & Fawzi, W. (2000). Undernutrition in relation to Isaacs, E., & Oates, J. (2008). Nutrition and cognition: childhood infections: a prospective study in the Sudan. assessing cognitive abilities in children and young European Journal of Clinical Nutrition, 54, 463-472. people. European Journal of Nutrition, 47, 4-24. Kuklina, E.V., Ramakrishnan, U., Stein, A.D., Barnhart, John, C.C., Black, M.M., & Nelson, C.A. (2017). H.H., & Martorell, R. (2004). Growth and diet quality Neurodevelopment: The Impact of Nutrition and are associated with the attainment of walking in rural Inflammation During Early to Middle Childhood in Low- Guatemalan infants. Journal of Nutrition, 134, 3296- Resource Settings. Pediatrics, 139, S59-S71. 3300.

Jones, L.L., Griffiths, P.L., Adair, L.S., Norris, S.A., Kuklina, E.V., Ramakrishnan, U., Stein, A.D., Barnhart, Richter, L.M., & Cameron, N. (2008). A comparison of H.H., & Martorell, R. (2006). Early childhood growth the socio-economic determinants of growth retardation and development in rural Guatemala. Early Human in South African and Filipino infants. Public Health Development, 82, 425-433. Nutrition, 11, 1220-1228. Kumar, A., & Kumari, D. (2014). Decomposing the Kamruzzaman, M., Rabbani, M., Saw, A., Sayem, M.A., Rural-Urban Differentials in Childhood Malnutrition in & Hossain, M.G. (2015). Differentials in the prevalence India, 1992–2006. Asian Population Studies, 10, 144- of anaemia among non-pregnant, ever-married women 162. in Bangladesh: multilevel logistic regression analysis of data from the 2011 Bangladesh Demographic and Kumar, A., Kumari, D., & Singh, A. (2014). Increasing Health Survey. BMC Womens’ Health 15 (54). doi: socioeconomic inequality in childhood undernutrition 10.1186/s12905-015-0211-4. in urban India: trends between 1992-93, 1998-99 and 2005-06. Health Policy and Planning, 30 (8), 1003-16. Kanjilal, B., Mazumdar, P.G., Mukherjee, M., & doi: 10.1093/heapol/czu104. Rahman, M.H. (2010). Nutritional status of children in India: household socio-economic condition as the Larson, L.M., & Yousafzai, A.K. (2017). A meta-analysis contextual determinant. International Journal for Equity of nutrition interventions on mental development in Health, 9, 19. doi: 10.1186/1475-9276-9-19. of children under-two in low- and middle-income countries. Maternal and Child Nutrition, 13 (1), e12229. Kariger, P.K., Stoltzfus, R.J., Olney, D., Sazawal, S., doi: 10.1111/mcn.12229. Black, R., Tielsch, J.M., et al. (2005). Iron Deficiency and Physical Growth Predict Attainment of Walking Lawn, J.E., Cousens, S., Zupan, J., & Team, L.N.S.S. but Not Crawling in Poorly Nourished Zanzibari Infants. (2005). 4 million neonatal deaths: When? Where? Journal of Nutrition, 135, 814-819. Why? Lancet, 365, 891-900.

Kassebaum, N., Jasrasaria, R., Naghavi, M., Lawrence, H., Hiroshi, K., & Nicolas, M. (2015). Growth Wulf, S.K., Johns, N., Lozano, R., et al. (2014). A is dead, long live growth: The quality of economic systematic analysis of global anaemia burden from growth and why it matters. In: Growth is Dead, Long 1990 to 2010. Blood, 123 (5), 615-24. doi: 10.1182/ Live Growth: The Quality of Economic Growth and blood-2013-06-508325. Why it Matters, Lawrence, H., Hiroshi, K., Nicolas, M. (editors). Tokyo, Japan: Agence Française de Ketema, L., Abate, G., & Jabar, M. (2003). Correlates of Développement (AFD), Institute of Development Studies children’ss cognitive skills in an Agrarian community (IDS), Japan International Cooperation Agency (JICA).

44 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA Le Roith, D. (1997). Insulin-like growth factors. New status and absenteeism of school children in Sri Lanka. England Journal of Medicine, 336, 633-640. Asia Pacific Journal of Clinical Nutrition, 16, 94-102.

Leight, J., Glewwe, P., & Park, A. (2015). The impact María-Dolores, R., & Martínez-Carrión, J.M. (2011). of early childhood rainfall shocks on the evolution of The relationship between height and economic cognitive and non-cognitive skills. Gansu Survey of development in Spain, 1850–1958. Economics and Children and Families Papers. Philadelphia: University Human Biology, 9, 30-44. of Pennsylvania Scholarly Commons. Mark, H., Houghton, L., Gibson, R., Monterrosa, E., & Levitsky, D.A., & Strupp, B.J. (1995). Malnutrition and Kraemer, K. (2016). Estimating dietary micronutrient the brain: changing concepts, changing concerns. supply and the prevalence of inadequate intakes from Journal of Nutrition, 125, 2212S-2220S. national Food Balance Sheets in the South Asia region. Asia Pacific Journal of Clinical Nutrition, 25 (2), 368- Lima, M.C., Eickmann, S.H., Lima, A.C.V., Guerra, 376. M.Q., Lira, P.I.C., Huttly, S.R.A., et al. (2004). Determinants of mental and motor development at 12 Martorell, R. (1995). Results and Implications of the months in a low income population: a cohort study in INCAP Follow-up Study. Journal of Nutrition, 125, northeast Brazil. Acta Paediatrica, 93, 969-975. 1127S-1138S.

Lopriore, C., Guidoum, Y., Briend, A., & Branca, F. Martorell, R., Horta, B.L., Adair, L.S., Stein, A.D., (2004). Spread fortified with vitamins and minerals Richter, L., Fall, C.H., et al. (2010a). Weight gain in induces catch-up growth and eradicates severe the first two years of life is an important predictor anaemia in stunted refugee children aged 3-6 y. of schooling outcomes in pooled analyses from five American Journal of Clinical Nutrition, 80 (4), 973-81. birth cohorts from low-and middle-income countries. Journal of Nutrition, 140, 348-354. Lozof, B. (2007). Iron Deficiency and Child Development. Food and Nutrition Bulletin, 28, Martorell, R., Melgar, P., Maluccio, J.A., Stein, A.D., S560-S571. & Rivera, J.A. (2010b). The nutrition intervention improved adult human capital and economic Lozoff, B., Jimenez, E., & Smith, J. (2006). Double burden productivity. Journal of Nutrition, 140, 411-414. of iron deficiency in infancy and low socioeconomic status: a longitudinal analysis of cognitive test scores Maternal & Child Group, C.N.S. (2013). Executive to age 19 years. Archives of Pediatrics and Adolescent Summary of The Lancet Maternal and Child Nutrition Medicine, 160 (11), 1108-1113. Series. Oxford, UK: Elsevier. http://download. thelancet. com/flatconten tassets/pdfs/nutrition-eng. pdf. Lundeen, E.A., Behrman, J.R., Crookston, B.T., Dearden, K.A., Engle, P., Georgiadis, A., et al. (2014). McGovern, E., Krishna, A., Aguayo, V.M., & Growth faltering and recovery in children aged 1–8 Subramanian, S.V. (2017). A review of the evidence years in four low-and middle-income countries: Young linking child stunting to economic outcomes. Lives. Public Health Nutrition, 17 (9), 2131-7. doi: International Journal of Epidemiology, 46 (4), 1171-91. 10.1017/S1368980013003017. doi: 10.1093/ije/dyx017.

Macours, K., Schady, N., & Vakis, R. (2012). Cash Measure DHS (2016). MEASURE DHS: Demographic transfers, behavioral changes, and cognitive and Health Surveys. Available: http://dhsprogram.com/. development in early childhood: evidence from a Accessed July 2016. randomized experiment. American Economic Journal: Applied Economics, 4, 247-273. Meeks Gardner, J.M., Grantham-McGregor, S.M., Chang, S.M., Himes, J.H., & Powell, C.A. (1995). Madsen, J. (2016). Barriers to prosperity: parasitic and Activity and Behavioral Development in Stunted and infectious diseases, IQ, and economic development. Nonstunted Children and Response to Nutritional World Development, 78, 172-187. Supplementation. Child Development, 66, 1785-1797.

Mahawithanage, S., Kannangara, K., Wickremasinghe, Meeks Gardner, J.M., Grantham-McGregor, S.M., R., Chandrika, U., Jansz, E., Karunaweera, N. et al. Himes, J., & Chang, S. (1999). Behaviour and (2007). Impact of vitamin A supplementation on health development of stunted and nonstunted Jamaican

45 children. Journal of Child Psychology and Psychiatry, Nepal (Government of Nepal) (1996). Family Health and Allied Disciplines, 40, 819-827. Survey 1996. Kathmandu, Nepal: Ministry of Health.

Meenakshi, J.V., Johnson, N.L., Manyong, V.M., Nepal (Government of Nepal) (2006). Demographic and DeGroote, H., Javelosa, J.M Yanggen, D.R., et al. Health Survey 2006. Kathmandu, Nepal: Ministry of (2010). How cost-effective is biofortification in Health and Population. combating micronutrient malnutrition? An ex ante assessment. World Development, 38 (1), 64-75. Nepal (Government of Nepal) (2011). Demographic and Health Survey 2011. Kathmandu, Nepal: Ministry of Mendez, M.A., & Adair, L.S. (1999). Severity and Health and Population. timing of stunting in the first two years of life affect performance on cognitive tests in late childhood. Niehaus, M.D., Moore, S.R., Patrick, P.D., Derr, L.L., Journal of Nutrition, 129, 1555-1562. Lorntz, B., Lima, A.A., et al. (2002). Early childhood diarrhea is associated with diminished cognitive Menon, P. (2012). Childhood Undernutrition in South function 4 to 7 years later in children in a northeast Asia: Perspectives from the Field of Nutrition. CESifo Brazilian shantytown. American Journal of Tropical Economic Studies, 58, 274-295. Medical and Hygiene, 66, 590-593.

Merrill, R., Shamim, A., Ali, H., Labrique, A., Olney, D.K., Kariger, P.K., Stoltzfus, R.J., Khalfan, S.S., Schulze, K., Christian, P., & West, K. Jr. (2012). High Ali, N.S., Tielsch, J.M., et al. (2009). Development prevalence of anaemia with lack of iron deficiency of nutritionally at-risk young children is predicted by among women in rural Bangladesh: a role for malaria, anemia, and stunting in Pemba, Zanzibar. thalassemia and iron in groundwater. Asia Pacific Journal of Nutrition, 139, 763-772. Journal of Clinical Nutrition, 21 (3), 416-24. Olney, D.K., Pollitt, E., Kariger, P.K., Khalfan, S.S., Mohd Nasir, M.T., Norimah, A.K., Hazizi, A.S., Ali, N.S., Tielsch, J.M., et al. (2007). Young Zanzibari Nurliyana, A.R., Loh, S.H., & Suraya, I. (2012). Child Children with Iron Deficiency, Iron Deficiency Anemia, feeding practices, food habits, anthropometric Stunting, or Malaria Have Lower Motor Activity Scores indicators and cognitive performance among and Spend Less Time in Locomotion. Journal of preschoolers in Peninsular Malaysia. Appetite, 58, Nutrition, 137, 2756-2762. 525-530. Olofin, I., McDonald, C.M., Ezzati, M., Flaxman, S., Montgomery, S.M., Bartley, M.J., Cook, D.G., Black, R.E., Fawzi, W.W., et al. (2013). Associations of & Wadsworth, M.E. (1996). Health and social suboptimal growth with all-cause and cause-specific precursors of unemployment in young men in Great mortality in children under five years: a pooled analysis Britain. Journal of Epidemiology and Community of ten prospective studies. PloS One, 8, e64636. Health, 50, 415-422. Onis, M., & Branca, F. (2016). Childhood stunting: a Moradi, A. (2010). Nutritional status and economic global perspective. Maternal and Child Nutrition, 12, development in sub-Saharan Africa, 1950–1980. 12-26. Economics and Human Biology, 8, 16-29. Oruamabo, R.S. (2015). Child malnutrition and the Mori, R., Ota, E., Middleton, P., Tobe-Gai, Millennium Development Goals: much haste but R., Mahomed, K. & Bhutta, Z.A. (2012). Zinc less speed? Archives of Disease in Childhood, 100, supplementation for improving pregnancy and infant S19-S22. outcome. Cochrane Database Systematic Reviews, 7, CD000230. doi: 10.1002/14651858.CD000230.pub4. Outes-Leon, I., Porter, C., & Sanchez, A. (2011). Early Nutrition and Cognition in Peru: a Within-Sibling Nahar, B., Hossain, M.I., Hamadani, J.D., Ahmed, T., Investigation. Washington, D.C.: Inter-American Huda, S.N., Grantham-McGregor, S.M., et al. (2012). Development Bank. Effects of a community-based approach of food and psychosocial stimulation on growth and development Özaltin, E., Hill, K., & Subramanian, S. (2010). of severely malnourished children in Bangladesh: Association of maternal stature with offspring mortality, a randomised trial. European Journal of Clinical underweight, and stunting in low-to middle-income Nutrition, 66, 701-709. countries. Journal of the American Medical Association,

46 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA 303, 1507-1516. Pollitt, E., Gorman, K.S., Engle, P.L., Rivera, J.A., & Martorell, R. (1995). Nutrition in early life and the Pakistan (Government of Pakistan and UNICEF) fulfillment of intellectual potential. Journal of Nutrition, (2011). National Nutrition Survey 2011. Islamabad, 125, 1111S-1118S. Pakistan: Planning Commission of the Planning and Development Division. Pollitt, E., Watkins, W.E., & Husaini, M.A. (1997). Three-month nutritional supplementation in Indonesian Pakistan (National Institute of Population Studies) infants and toddlers benefits memory function 8 y later. (1990). Demographic and Health Survey 1990-1. American Journal of Clinical Nutrition, 66, 1357-1363. Islamabad, Pakistan: National Institute of Population Studies. Pongcharoen, T., Ramakrishnan, U., DiGirolamo, A.M., Winichagoon, P., Flores, R., Singkhornard, J., et al. Pakistan (National Institute of Population Studies) (2012). Influence of prenatal and postnatal growth (2013). Demographic and Health Survey 2012-3. on intellectual functioning in school-aged children. Islamabad, Pakistan: National Institute of Population Archives of Pediatrics and Adolescent Medicine, 166 Studies. (5), 411-416.

Paxson, C., & Schady, N. (2010). Does Money Matter? Power, C., & Elliott, J. (2006). Cohort profile: 1958 The Effects of Cash Transfers on Child Development in british birth cohort (national child development study). Rural Ecuador. Economic Development and Cultural International Journal of Epidemiology, 35, 34-41. Change, 59, 187-229. Prado, E.L., & Dewey, K.G. (2014). Nutrition and brain Persico, N., Postlewaite, A., & Silverman, D. (2004). development in early life. Nutrition Reviews, 72, 267-284. The e affect of adolescent experience on labor market outcomes: The case of height. Journal of Political Prendergast, A.J., & Humphrey, J.H. (2014). The Economy, 112 (5), 1019-1053. stunting syndrome in developing countries. Paediatrics and International Child Health, 34, 250-265. Pinstrup-Andersen, P. (2013). Nutrition-sensitive food systems: from rhetoric to action. Lancet, 382, 375-376. Prentice, A.M., Ward, K.A., Goldberg, G.R., Jarjou, L.M., Moore, S.E., Fulford, A.J., et al. (2013). Critical Piper, B. (2014). A production function examination windows for nutritional interventions against stunting. of the aggregate effects of nutrition. Journal of American Journal of Clinical Nutrition, 97, 911-918. Macroeconomics, 40, 293-307. Qureshy, L.F., Alderman, H., Rokx, C., Pinto, R., Wai- Pokharel, R., Maharjan, M., Mathema, P., Harvey, Poi, M., & Tandon, A. (2013). Positive returns: cost- P.W.J. (2011). Success in Delivering Interventions to benefit analysis of a stunting intervention in Indonesia. Reduce Maternal Anemia in Nepal: A Case Study of the Journal of Development Effectiveness, 5, 447-465. Intensification of Maternal and Neonatal Micronutrient Program. Washington, DC: A2Z: USAID Micronutrient Rah, J., Anas, A.M., Chakrabarty, A., Sankar, R., Pandav, and Child Blindness Project. C. & Aguayo, V.M. (2015). Towards universal salt iodization in India: Achievements, challenges and future Pollitt, E. (2000). Developmental sequel from early actions. Maternal and Child Nutrition, 11, 483–496. nutritional deficiencies: conclusive and probability judgements. Journal of Nutrition, 130, 350S-353S. Richter, L.M., Daelmans, B., Lombardi, J., Heymann, J., Boo, F.L., Behrman, J.R., et al. (2016). Investing in the Pollitt, E. (2009). Timing and Vulnerability in Research foundation of sustainable development: pathways to on Malnutrition and Cognition. Nutrition Reviews, 54, scale up for early childhood development. Lancet, 389, S49-S55. 103-118.

Pollitt, E., Gorman, K.S., Engle, P.L., Martorell, Ross, J., Chen, C.-M., He, W., Fu, G., Wang, Y.Y., Fu, R., Rivera, J., Wachs, T.D., et al. (1993). Early Z.-Y., et al. (2003). Effects of malnutrition on economic supplementary feeding and cognition: effects over two productivity in China as estimated by PROFILES. decades. Monographs of the Society for Research in Biomedical and Environmental Sciences: BES, 16, 195- Child Development, 58(7, Serial No. 253): 1–118. 205.

47 Ruel, M.T., Alderman, H., Maternal, & Child Nutrition in Pakistan? Journal of the Pakistan Medical Association, Study, G. (2013). Nutrition-sensitive interventions and 58, 265-6. programmes: how can they help to accelerate progress in improving maternal and child nutrition? Lancet, 382, Siegel, E.H., Stoltzfus, R.J., Kariger, P.K., Katz, J., 536-551. Khatry, S.K., LeClerq, S.C., et al. (2005). Growth indices, anemia, and diet independently predict motor milestone Rutstein, S.O., & Johnson, K. (2004). The DHS wealth acquisition of infants in South Central Nepal. Journal of index. DHS Comparative Reports 6. Calverton, Nutrition, 135, 2840-2844. Maryland: ORC Macro. Smith, L., & Haddad, L. (1999). Explaining Child Saha, K.K., Frongillo, E.A., Alam, D.S., Arifeen, S.E., Malnutrition in Developing Countries: A Cross-Country Persson, L.Å., & Rasmussen, K.M. (2008). Appropriate Analysis. Food Consumption and Nutrition Division infant feeding practices result in better growth of Discussion Papers. Washington D.C.: International Food infants and young children in rural Bangladesh. Policy Research Institute.. American Journal of Clinical Nutrition, 87, 1852-1859. Smith, L.C., & Haddad, L. (2002). How Potent Is Sanchez, A. (2013). The structural relationship between Economic Growth in Reducing Undernutrition? What Are nutrition, cognitive and non-cognitive skills. Young the Pathways of Impact? New Cross-Country Evidence. Lives Working Paper 111. Oxford: Young Lives. Economic Development and Cultural Change, 51, 55-76.

Sargent, J.D., & Blanchflower, D.G. (1994). Obesity Smith, L.C., & Haddad, L. (2015). Reducing child and stature in adolescence and earnings in young undernutrition: past drivers and priorities for the post- adulthood: analysis of a British birth cohort. Archives MDG era. World Development, 68, 180-204. of Pediatrics and Adolescent Medicine, 148, 681-687. SMS (2010). Scaling Up Nutrition: A Framework for Satyanarayana, K., Nadamuni, N.A., & Narasinga, Action. Geneva: Scaling Up Nutrition Movement R.B.S. (1978). Nutrition, physical work capacity and Secretariat (SMS). work output. Indian Journal of Medical Research, 68, 88-93. Soeters, P.B., & Schols, A.M. (2009). Advances in understanding and assessing malnutrition. Current Satyanarayana, K., Naidu, A.N., & Rao, B.S.N. (1979). Opinion in Clinical Nutrition And Metabolic Care, 12, Nutritional deprivation in childhood and the body size, 487-494. activity, and physical work capacity of young boys. American Journal of Clinical Nutrition, 32, 1769-1775. Sokolovic, N., Selvam, S., Srinivasan, K., Thankachan, P., Kurpad, A.V., & Thomas, T. (2014). Catch-up growth does Sazawal, S., Dhingra, P., Dhingra, U., Gupta, not associate with cognitive development in Indian school- S., Iyengar, V., Menon, V., et al. (2014). Compliance age children. European Journal of Clinical Nutrition, 68, with home-based fortification strategies for delivery of 14-18. iron and zinc: its effect on haematological and growth markers among 6-24 months old children in North Stein, A.D., Melgar, P., Hoddinott, J., & Martorell, R. India. Journal of Health, Population and Nutrition, 32 (2008). Cohort profile: The Institute of Nutrition of Central (2), 217–226. America and Panama (INCAP) nutrition trial cohort study. International Journal of Epidemiology, 37, 716-720. Shekar, M., Heaver, R., & Lee, Y.-K. (2006). Repositioning Nutrition as Central to Development: Stein, A.J., & Qaim, M. (2007). The human and economic A Strategy for Large Scale Action. Washington D.C.: cost of hidden hunger. Food and Nutrition Bulletin 28(2), World Bank Publications. 125-34.

Sibhatu, K., Krishna, V., & Qaim, M. (2015). Production Stein, A.D., Wang, M., Martorell, R., Norris, S.A., Adair, diversity and dietary diversity in smallholder farm L.S., Bas, I., et al. (2010). Growth patterns in early households. Proceedings of the National Academy of childhood and final attained stature: Data from five Sciences, 112(34), 10657–10662. birth cohorts from low- and middle-income countries. American Journal of Human Biology, 22, 353-359. Siddiqi, N., & Iqbal, R. 2008. Maternal postpartum vitamin A supplementation programme: is there a need Stevens, G., Finucane, M., De-Regil, L., Paciorek, C.,

48 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA Flaxman, S., Branca, F. et al. (2013). Global, regional, American Psychologist, 56, 5-15. and national trends in haemoglobin concentration and prevalence of total and severe anaemia in children and Thow, A.M., Karn, S., Devkota, M.D., Rasheed, S., pregnant and non-pregnant women for 1995–2011: Roy, S.K., Suleman, Y. et al. (2017). Opportunities for a systematic analysis of population-representative strengthening infant and young child feeding policies data. Lancet Global Health, 1 (1): e16-25. doi: 10.1016/ in South Asia: Insights from the SAIFRN policy S2214-109X(13)70001-9. analysis project. BMC Public Health, 17(Suppl 2), 5-14. doi: 10.1186/s12889-017-4336-2. Stevens, G.A, Bennett, J., Hennocq, Q., Lu, Y., De- Regil, L.M., Roger, L., et al. (2015). Trends and Umana-Aponte, M. (2011). Long-term effects of mortality effects of vitamin A deficiency in children in a nutritional shock: the 1980 famine of Karamoja, 138 low-income and middle-income countries between Uganda. Working Paper 11/258. Bristol: Center for 1991 and 2013: a pooled analysis of population-based Market and Public Organisation. surveys. Lancet Global Health, 3, e528-e536. UN (2015a). The Millennium Development Goals Stevens, G.A., Finucane, M.M., Paciorek, C.J., Report. New York: United Nations. Flaxman, S.R., White, R.A., Donner, A.J., et al. (2012). Trends in mild, moderate, and severe stunting UN (2015b). Sustainable Development Goals. New and underweight, and progress towards MDG 1 in York: United Nations. 141 developing countries: a systematic analysis of population representative data. Lancet, 380, 824-834. UNICEF (2013). Improving Child Nutrition: The Achievable Imperative for Global Progress. New York: Strupp, B.J., & Levitsky, D.A. (1995). Enduring United Nations Children’s Fund. cognitive effects of early malnutrition: a theoretical reappraisal. Journal of Nutrition, 125, 2221S-2232S. UNICEF (2014). Global Database on Iodine Deficiency, October 2014 update. Accessed July 26, 2016. http:// Subramanian, S., Ackerson, L.K., Smith, G.D., & John, data.unicef.org/nutrition/iodine.html N.A. (2009). Association of maternal height with child mortality, anthropometric failure, and anemia in India. UNICEF (2015). Global Database on Vitamin A Journal of the American Medical Association, 301, Supplementation, November 2015 update. Accessed 1691-1701. July 26, 2016. http://data.unicef.org/nutrition/ vitamin-a.html Subramanyam, M.A., Kawachi, I., Berkman, L.F., & Subramanian, S.V. (2011). Is economic growth UNICEF (2016a). State of the World’s Children 2016. associated with reduction in child undernutrition in New York City, USA.2015. Global Database on India? PLoS Medicine, 8, 448. Vitamin A Supplementation, November 2015 update. Sudfeld, C.R., Charles McCoy, D., Danaei, G., Fink, Available: http://data.unicef.org/nutrition/vitamin-a. G., Ezzati, M., Andrews, K.G., et al. (2015). Linear html. Accessed July 26, 2016. growth and child development in low- and middle- income vountries: a meta-analysis. Pediatrics, 135, UNICEF (2016b). From the First Hour of Life: Meeting e1266-e1275. the Case for Improved Infant and Young Child Feeing Everywhere. New York: United Nations Children’s Taneja, S., Bhandari, N., Bahl, R., & Bhan, M.K. Fund. (2005). Impact of zinc supplementation on mental and psychomotor scores of children aged 12 to 18 months: UNICEF, WHO, & World Bank Group (2018). Joint A randomized, double-blind trial. Journal of Pediatrics, Child Malnutrition Estimates 2018 Edition. Available: 146, 506-511. http://www.who.int/nutgrowthdb/estimates2017/en/. Accessed 23 May 2018. Thompson, B., & Amoroso, L. (editors). (2014). Improving Diets and Nutrition: Food-Based van Pareren, Y.K., Duivenvoorden, H.J., Slijper, Approaches. Wallingford, UK: Food and Agriculture F.S.M., Koot, H.M., & Hokken-Koelega, A.C.S. (2004). Organization & CAB International. Intelligence and psychosocial functioning during long- term growth hormone therapy in children born small Thompson, R.A., & Nelson, C.A. (2001). Developmental for gestational age. Journal of Clinical Endocrinology science and the media: Early brain development. and Metabolism, 89, 5295-5302.

49 Vazir, S., Engle, P., Balakrishna, N., Griffiths, P.L., Walker, S.P. (2006). Effects of psychosocial stimulation Johnson, S.L., Creed-Kanashiro, H., et al. (2012). and dietary supplementation in early childhood on Cluster-randomized trial on complementary and psychosocial functioning in late adolescence: follow-up responsive feeding education to caregivers found of randomised controlled trial. British Medical Journal, improved dietary intake, growth and development 333, 472-470. among rural Indian toddlers. Maternal and Child Nutrition, 9, 99-117. Walker, S.P., Chang, S.M., & Powell, C.A. (2007a). Early childhood stunting is associated with poor Via, M. (2012). The malnutrition of obesity: psychological functioning in late adolescence and micronutrient deficiencies that promote diabetes ISRN effects are reduced by psychosocial stimulation. Endocrinology, 103472. doi: 10.5402/2012/103472. Journal of Nutrition, 137, 2464-2469.

Victora, C.G., & Barros, F.C. (2006). Cohort profile: the Walker, S.P., Grantham-McGregor, S.M., Powell, C.A., 1982 Pelotas (Brazil) birth cohort study. International & Chang, S.M. (2000). Effects of growth restriction Journal of Epidemiology, 35, 237-242. in early childhood on growth, IQ, and cognition at age 11 to 12 years and the benefits of nutritional Victora, C.G., Adair, L., Fall, C., Hallal, P.C., Martorell, supplementation and psychosocial stimulation. Journal R., Richter, L., et al. (2008). Maternal and child of Pediatrics, 137, 36-41. undernutrition: consequences for adult health and human capital. Lancet, 371, 340-357. Walker, S.P., Wachs, T.D., Gardner, J.M., Lozoff, B., Wasserman, G.A., Pollitt, E., et al. (2007b). Child Victora, C.G., de Onis, M., Hallal, P.C., Blossner, M., development: risk factors for adverse outcomes in & Shrimpton, R. (2010). Worldwide Timing of Growth developing countries. Lancet, 369, 145-157. Faltering: Revisiting Implications for Interventions. Pediatrics, 125, e473-e480. Webb, P., & Block, S. (2004). Nutrition information versus formal schooling as inputs to child nutrition. Vohr, B.R., Poggi Davis, E., Wanke, C.A., & Krebs, N.F. Economic Development and Cultural Change, 52 (4), (2017). Neurodevelopment: the ompact of nutrition and 801-820. inflammation during preconception and pregnancy in low-resource settings. Pediatrics, 139, S38-S49. Webb, P., & Block, S. (2012). Support for agriculture during economic transformation: Impacts on poverty Vollmer, S., Harttgen, K., Subramanyam, M.A., Finlay, and undernutrition. Proceedings of the National J., Klasen, S., & Subramanian, S.V. (2014). Association Academy of Sciences, 109, 12309-12314. between economic growth and early childhood undernutrition: evidence from 121 Demographic and Webb, P., & Kennedy, E. (2014). Impacts of agriculture Health Surveys from 36 low-income and middle- on nutrition: nature of the evidence and research gaps. income countries. Lancet Global Health, 2, e225-e234. Food and Nutrition Bulletin, 35 (1), 126-33. Weil, D.N. (2007). Accounting for the Effect Of Health Vos, T., Flaxman, A.D., Naghavi, M., Lozano, R., on Economic Growth. Quarterly Journal of Economics, Michaud, C., Ezzati, M., et al. (2012). Years lived with 122, 1265-1306. disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990–2010: a systematic analysis for the Global Wessells, K., Singh, G., & Brown, K. (2012). Estimating Burden of Disease Study 2010. Lancet, 380, 2163–96. the global prevalence of inadequate zinc intake from national food balance sheets: effects of methodological Waber, D.P., Vuori-Christiansen, L., Ortiz, N., Clement, assumptions. PLoS One 7, e50565. J.R., Christiansen, N.E., Mora, J.O., et al. (1981). Nutritional supplementation, maternal education, and Whaley, S.E., Sigman, M., Espinosa, M.P., & Neumann, cognitive development of infants at risk of malnutrition. C.G. (1998). Infant predictors of cognitive development American Journal of Clinical Nutrition, 34, 807-813. in an undernourished Kenyan population. Journal of Developmental and Behavioral Pediatrics, 19, 169-177. Wachs, T.D., Georgieff, M., Cusick, S., & McEwen, B.S. (2013). Issues in the timing of integrated Whincup, P.H., Kaye, S.J., Owen, C.G., Huxley, R., early interventions: contributions from nutrition, Cook, D.G., Anazawa, S., et al. (2008). Birth weight and neuroscience, and psychological research. Annals of risk of type 2 diabetes: a systematic review. Journal of the New York Academy of Sciences, 1308, 89-106. the American Medical Association, 300, 2886-2897.

50 CHILD STUNTING, HIDDEN HUNGER AND HUMAN CAPITAL IN SOUTH ASIA WHO (2002). The World Health Report 2002: Reducing WHO (2016). Micronutrient Deficiencies. Available: Risks, Promoting Healthy Life. Geneva:World Health http://www.who.int/nutrition/topics/micronutrients/en/. Organization. Accessed June 25, 2016.

WHO (2006). WHO Child Growth Standards: Length/ WHO & FAO (2003). Diet, Nutrition and the Prevention Height-for-Age, Weight-for-Age, Weight-for-Length, of Chronic Diseases: Report of a Joint WHO/FAO Weight-for-Height and Body Mass Index-for-Age: Expert Consultation, Geneva, 28 January-1 February Methods and Development. Geneva: World Health 2002. WHO Technical Report Series No. 916. Geneva, Organization. Switzerland: World Health Organization.

WHO (2007). Global Database on Anaemia, Last Yang, S., Tilling, K., Martin, R., Davies, N., Ben-Shlomo, updated for Pakistan: October 25, 2007. Accessed Y., & Kramer, M.S. (2011). Pre-natal and post-natal accessed June 28, 2016. http://www.who.int/vmnis/ growth trajectories and childhood cognitive ability and database/anaemia/countries/en/ mental health. International Journal of Epidemiology, 40, 1215-1226. WHO (2008). Indicators for Assessing Infant and Young Child Feeding Practices. Part 1: Definitions. Yousafzai, A.K., & Aboud, F. (2014). Review of Conclusions of a Consensus Meeting held 6-8 implementation processes for integrated nutrition and November 2007 in Washington DC, USA. Geneva: psychosocial stimulation interventions. Annals of the World Health Organization. New York Academy of Sciences, 1308, 33-45.

WHO (2010). Indicators For Assessing Infant and Yousafzai, A.K., Rasheed, M.A., Rizvi, A., Armstrong, Young Child Feeding Practices. Part 2: Measurement. R., & Bhutta, Z.A. (2014). Effect of integrated Geneva: World Health Organization. responsive stimulation and nutrition interventions in the Lady Health Worker programme in Pakistan on WHO (2014). Global health estimates 2014 summary child development, growth, and health outcomes: a tables: deaths by cause, age and sex, by WHO region, cluster-randomised factorial effectiveness trial. Lancet, 2000–2012. Geneva: World Health Organization. 384, 1282-1293. WHO (2015). The global prevalence of anaemia in 2011. Geneva: World Health Organization.

SAARC Secretariat UNICEF Regional Office for South Asia Tridevi Marg, Kathmandu 44600, Nepal Leknath Marg, Kathmandu 44600, Nepal Email: [email protected] E-mail: [email protected] Website: www.saarc-sec.org Website: www.unicef.org/rosa