Variation in Anthropometric Status and Growth Failure in Low- and Middle-Income Countries.

Abstract

Addressing anthropometric failure in low- and middle-income countries can have 2 targets of inference: addressing differences between individuals within populations (W_pop) or differences between populations (B_pop). We present a multilevel framework to apply both targets of inference simultaneously and quantify the extent to which variation in anthropometric status and growth failure is reflective of undernourished children or undernourished populations. Cross-sectional data originated from the Demographic and Health Surveys program, covering children under age 5 from 57 countries surveyed between 2001 and 2015. A majority of variation in child anthropometric status and growth failure was attributable to W_pop-associated differences, accounting for 89%, 83%, and 85% of the variability in <i>z</i> scores for height for age, weight for age, and weight for height. B_pop-associated differences (communities, regions, and countries combined) were associated with 11%, 17%, and 15% of the variation in height-for-age <i>z</i> score, weight-for-age <i>z</i> score, and weight-for-height <i>z</i> score. Prevalence of anthropometric failure was closely correlated with mean levels of height and weight. Approximately 1% of W_pop variability, compared with 30% to 50% of the B_pop variability, was explained by mean values of maternal correlates of anthropometric status and failure. Although there is greater explanatory power B_pop, this varied because of modifiability of what constitutes population. Our results suggest that universal strategies to prevent future anthropometric failure in populations combined with targeted strategies to address both the impending and existing burden among children are needed.