Abstract
BACKGROUND: Stunting remains a considerable public health problem globally, and sex differences in prevalence have been documented. While many risk factors for stunting have been identified, few studies examine how these factors may contribute to sex differences. We test whether: (1) boys and girls are differentially exposed to stunting risk factors, (2) boys and girls respond differently to similar exposures, and (3) these associations are sensitive to the growth measures used. METHODS: Data comes from 7486 Zambian infants, aged 0-23 months, participating in the SUN (Scaling Up Nutrition) 2.0 Program baseline survey. Surveys and dietary recalls were collected from primary caregivers, and anthropometry was collected for caregivers and children. Stunting was defined as height-for-age z-score (HAZ) <-2. Descriptive statistics and adjusted multilevel logistic regression models controlling for age and province were used to identify sex differences in exposures and risk factors for stunting, respectively. Interaction terms between exposure and infant sex were added to test for sex differences in response. Sensitivity testing with alternate measures of infant size, including height-for-age difference (HAD), was conducted. RESULTS: Boys were more likely to be stunted than girls (34.5% vs. 25.7%, respectively). Numerous maternal, care giving, diet, and household characteristics were associated with the risk of stunting, but few sex differences were seen in exposure to these factors. Only one risk factor-maternal marital status-showed evidence of moderation by sex at the p < 0.05 level, while a limited number of risk factors did so at the p < 0.10 level. At all ages, boys were longer than girls, and there were no sex-specific differences in the patterns of HAZ or HAD with age. Results were robust to alternate specifications. CONCLUSIONS: Our results show that Zambian boys have lower mean HAZ scores than girls and a greater prevalence of stunting throughout the first two years We do not find strong evidence that infant feeding practices, environmental exposures, or care giving differ consistently between boys and girls or that boys and girls respond differently to these exposures. Our results instead indicate that further investigation of prenatal factors and/or measurement issues is needed.