Early life mortality and height in Indian states☆
Introduction
Economists have often used height as a measure of human development. Recent research has documented important relationships between height and survival, height and cognitive achievement, and height and productivity.1 The importance of height as a measure of population health and well-being has led to the question: what determines height?
Several studies have used data from Europe to establish correlations between height and early life conditions, as captured by mortality levels in infancy (Schmidt et al., 1995, Crimmins and Finch, 2006, Bozzoli et al., 2009, Hatton, 2011, Hatton, 2013). Postneonatal mortality (PNM), the number of infants per 1000 live births who die between the first and the twelfth months of life, emerges as a strong predictor of height. In these studies, postneonatal mortality is understood to proxy for the disease environment to which infants are exposed. Infants exposed to more disease early in life experience poor growth, leaving them stunted. In Bozzoli et al. (2009), neonatal mortality (NNM), which was largely determined by the availability of advanced medical care for neonates, had no correlation with adult heights in European and American cohorts born from 1950 to 1980, controlling for other factors.
As in developed countries, height in developing countries is determined in large part by early life conditions, especially from conception to two years of age, but these conditions tend to be more varied, and more damaging than in developed countries. Also as in developed countries, the disease environment is likely an important determinant of height, but other factors, such as economic well-being and pre-natal nutrition may play a more important role in developing countries than in developed ones.
This paper examines correlates of height in India, a large developing country with one of the shortest populations in the world, and with significant differences in health outcomes across states (Deaton, 2008, Dréze and Sen, 1997). It finds that state level variation in mortality levels from a cohort's infancy predict both the heights of Indian adults born between 1970 and 1983, and the heights of Indian children born between 1995 and 2005. Consistent with prior literature on mortality and the disease environment, I find that state level variation in pre-adult mortality and postneonatal mortality correlate with adult height and child height, respectively. A novel finding of the paper is that, in contrast to in developed countries, neonatal mortality, or mortality between birth and one month of life, is a robust predictor of height in India. This is true both for adults born between 1970 and 1983, and for children born between 1995 and 2005.
The relationship between state level neonatal mortality and height is robust to controlling for postneonatal mortality, which proxies for the disease environment, as well as economic circumstances, measured by state domestic product, in the case of the adults, and household asset wealth, in the case of the children. Unlike in literature on the correlates of height from developed countries, but consistent with prior research on height in developing countries, I find that these measures of economic well-being indeed predict height (Steckel, 1983). Regressions of children's height on early life mortality rates also control for mother's height, suggesting that there is not a spurious correlation between mortality in the state where a baby is born and her genetic height potential.
The finding that state level variation in neonatal mortality predicts Indians’ heights raises the question: for what early life conditions that shape height does neonatal mortality proxy? I review the literature and existing cause of death data and propose that low birth weight, caused by poor pre-natal nutrition, could be driving the correlation between neonatal mortality and height. Vital statistics data suggest that low birth weight has long been, and continues to be, a leading cause of neonatal mortality in India. Indeed, UNICEF (2004) estimate that about a third of infants in India are born at a low birth weight, and that India is home to forty percent of the world's low birth weight babies. Although no representative state or national-level data on birth weight exist, the National Family Health Survey data reveal important differences in women's nutrition across Indian states.
These findings are important for researchers seeking to understand why, despite its high rates of economic growth, India continues to have one of the shortest populations in the world. Although Indians’ heights are certainly determined by many factors,2 these findings suggest that continuing neglect of women's health, and in particular pre-natal nutrition, will have continuing consequences for heights in India, and for all of the health and human development indicators that height reflects.
The paper proceeds as follow: Section 2 provides contextual information on the early life determinants of height in developed and developing countries, as well as on causes of early life mortality in India. Section 3 presents the data sources and modeling strategy. Section 4 presents the results of three analyses: the first regresses the heights of adult cohorts born between 1970 and 1983 in different states of India on neonatal and postneonatal mortality rates in their years of birth; the second regresses the heights of these adults on pre-adult mortality rates in their year of birth; and the third regresses the height-for-age z-scores of individual children from two rounds of India's Demographic and Health Survey on state-survey round level measures of neonatal and postneonatal mortality. Section 5 discusses the results, as well as the hypothesis that variation in neonatal mortality proxies for state level variation in maternal net nutrition. Section 6 concludes.
Section snippets
Early life determinants of height in developed countries
In Europe, the relationships between height and the early life environment have been examined in the context of select groups, as well as at the population level. Schmidt et al. (1995) document a relationship between the adult heights of men who were conscripted in European armies and postneonatal mortality in the year of birth. Bozzoli et al. (2009) provide evidence for a similar relationship in the general populations of European countries and the United States for cohorts born between 1950
Adult analyses
Mortality indicators. Mortality indicators from the 1970s and 1980s used for the analysis of adult heights come from the Sample Registration System (SRS), a vital statistics system run by the Office of the Registrar General at the Indian Ministry of Home Affairs. Trained SRS enumerators register vital events in sample localities in order to estimate demographic rates at the state and national levels. The SRS data cover the 17 major Indian states;9
Neonatal mortality, postneonatal mortality & adult height
I use ordinary least squares (OLS) regression to estimate the association between early life mortality and the mean heights of adult cohorts. The variation exploited in this analysis comes primarily from large variation in mortality across states. While other papers that have looked at the relationship between measures of early life conditions and adult height have done so during a period of declining infant mortality, and have found effects of mortality on adult height controlling for year,
Discussion
The correlations between height and postneonatal mortality and pre-adult mortality documented here likely reflect relationships between height and infectious diseases that both stunt growth and cause early death. However, prior papers have not found relationships between height and neonatal mortality. In these data, what are the early life conditions that stunt growth for which neonatal mortality likely proxies?
One factor that is likely to be be important is poor pre-natal nutrition, which has
Conclusion
Using data from state cohorts born between 1970 and 1983 and from children in two rounds of the more recent NFHS surveys, this article has documented a negative relationships between height and measures of early life mortality in India. These findings contribute to the literature the first evidence from within a developing country of the associations between neonatal mortality and height. This finding is novel, since other authors who have looked at the relationships between early life
References (72)
- et al.
Health trends in Sub-Saharan Africa: conflicting evidence from infant mortality rates and adult heights
Econ. Hum. Biol.
(2010) - et al.
Tall claims: mortality selection and the height of children in India
Econ. Hum. Biol.
(2011) fatal fluctuations? Cyclicality in infant mortality in India
J. Dev. Econ.
(2010)Gender bias among children in India in their diet and immunisation against disease
Soc. Sci. Med.
(2004)- et al.
Stunting and selection effects of famine: a case study of the Great Chinese Famine
J. Dev. Econ.
(2012) Child undernutrition, tropical enteropathy, toilets, and handwashing
Lancet
(2009)- et al.
Environmental enteropathy: critical implications of a poorly understood condition
Trends Mol. Med.
(2012) - et al.
Energy supplementation during pregnancy and postnatal growth
Lancet
(1992) - et al.
The hidden penalties of gender inequality: fetal origins of ill-health
Econ. Hum. Biol.
(2003) Height and cognitive achievement among Indian children
Econ. Hum. Biol.
(2012)
The relative contribution of prematurity and fetal growth retardation to low birthweight in developing and developed societies
Am. J. Obstet. Gynecol.
Size at birth and growth trajectories to young adulthood
Am. J. Hum. Biol.
Famine in China, 1958–1961
Popul. Dev. Rev.
Is home-based diagnosis and treatment of neonatal sepsis feasible and effective? Seven years of intervention in the Gadchiroli field trial (1996 to 2003)
J. Perinatol.
Child gender and parental investments in India: are boys and girls treated differently?
Am. Econ. J. Appl. Econ.
Comparison of the causes and consequences of prematurity and intrauterine growth retardation: a longitudinal study in southern Brazil
Pediatrics
Intrahousehold allocation of nutrients in rural India: are boys favored? Do parents exhibit inequality aversion?
Oxford Econ. Pap.
Returns to birthweight
Rev. Econ. Stat.
On the trail of ‘missing’ Indian females: I: Illusion and reality
Econ. Polit. Wkly
Vital rates in India, 1961–1981. Report 24
Maternal and child health: Is South Asia ready for change?
BMJ
Birth weight and childhood growth
Pediatrics
Adult height and childhood disease
Demography
Bootstrap-based improvements for inference with clustered errors
Rev. Econ. Stat.
Stature and status: height, ability, and labor market outcomes
J. Polit. Econ.
Multi-country analysis of the effects of diarrhoea on childhood stunting
Int. J. Epidemiol.
Infection, inflammation, height and longevity
Proc. Natl. Acad. Sci.
Perspectives on women's autonomy and health outcomes
Am. Anthropol.
WHO child growth standards: Length/height-for-age, weight-for-age, weight-for-length, weight-for-height and body mass index-for-age: Methods and development. Technical report
Global patterns of income and health: Facts, interpretations, and policies
Height, health & inequality: the distribution of adult heights in India
AEA Pap. Proc.
Nutrition in India: facts and interpretations
Econ. Polit. Weekely
Indian Development: Selected Regional Perspectives
Neonatal pneumonia in developing countries
Arch. Dis. Childhood-Fetal Neonatal Ed.
Domestic product of states of India: 1960–61 to 2006–07
The Escape from Hunger and Premature Death, 1700–2100: Europe, America, and the Third World
Cited by (0)
- ☆
I would like to thank Angus Deaton, Jean Drèze, Aashish Gupta, Jeffrey Hammer, Laura Nolan, Germán Rodríguez, Dean Spears, Elizabeth Sully, Sangita Vyas, and the editors and reviewers at Economics & Human Biology for comments, as well as Office of Population Research librarians Elana Broch and Joann Donatiello, and librarians at the Institute for Economic Growth and the Delhi School of Economics, in New Delhi, India for help assembling the data for this project. All errors are my own. Partial support for this research was provided by a grant from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (5R24HD047879).