Food price inflation and child stunting in South Asia: Evidence from panel data

Abstract

Child stunting reflects chronic nutritional deprivation, while food-price increases can reduce household purchasing power and diet quality. This ecological panel study examined whether sustained food price inflation was associated with modeled stunting prevalence among children younger than five years in seven South Asian countries during 2003-2023. Annual food inflation was calculated as the percentage change in annual mean national Food Consumer Price Index from FAOSTAT. The primary exposure was the mean of current, one-year-lagged, and two-year-lagged inflation. UNICEF-WHO-World Bank Joint Child Malnutrition Estimates were linked to GDP per capita, current health expenditure, sanitation, and urbanization. Two-way fixed-effects models used Driscoll-Kraay standard errors because residual diagnostics indicated serial correlation and cross-sectional dependence. The primary model included 147 country-year observations. A one-percentage-point increase in three-year average food inflation was associated with a 0.120-percentage-point lower stunting prevalence (95% CI -0.245 to 0.005; p=0.058), which did not support the hypothesized positive association. Contemporaneous and one-year-lagged inflation were not statistically distinguishable from zero, whereas the two-year-lagged estimate was negative and narrowly excluded zero. Given the secondary nature of these exposure-window analyses, the isolated two-year-lagged result should be interpreted cautiously. The negative direction persisted in leave-one-country-out analyses, but precision and statistical significance varied. These findings do not support a positive country-level association between sustained food inflation and modeled annual stunting prevalence. They should not be interpreted causally; strong secular improvements in stunting, limited between-year outcome variation, modeled outcomes, and residual confounding may obscure household-level harms documented elsewhere.