Abstract

India has one of the highest incidence of childhood undernutrition as well as related adverse
child health outcomes like anemia and low birthweight. These indicators can be correlated
in a particular population which, if not accounted for in an analytical framework can lead
to inaccurate estimates of their spread and penetration. In this study, we assess the spa-
tial heterogeneity and risk factors of the triple burden of childhood stunting, anemia and
low birthweight in India based on nationally representative data from the 2019-21 National
Family Health Survey 5 (NFHS 5). As our empirical strategy, we use a trivariate copula
geoadditive modeling framework that enables simultaneous specification of the marginal dis-
tributions of the disease indicators as well as their dependence structures with respect to
complex functions of covariates. The joint modeling approach is corroborated by the signifi-
cant positive correlation between the ailment-pairs obtained from initial exploratory analysis
of the data. Model based estimates were indicative of considerable spatial heterogeneity in
their individual and collective burdens as well as in their pairwise correlations across India.
Various child, maternal and household specific attributes like child gender, type of delivery,
prenatal checkup of mother, maternal health and educational attainment, household wealth
and sanitation facilities had significant association with the incidence of the three ailments.
The study reaffirms the importance of joint modeling of various dimensions of childhood
undernutrition and related ailments towards effective policy formulation and targeted imple-
mentation to improve the state of child health in India.