Hidden Undernutrition: How universal cutoffs can fail to capture stunting in low and middle income countries
share
Stunting, or impaired child growth due to undernutrition, has multiple negative health effects, making it a top global health priority. The current benchmark for classifying stunting assumes a universal model of growth with height-for-age z-score (HAZ) cutoffs set by the WHO. However, this universal model may hide hotspots of stunting if populations differ in HAZ in ways that are independent of undernutrition. We assess the potential magnitude of this problem by decomposing variation in HAZ from 1,406,609 children from 63 low-and middle-income countries into two components: (1) a component shaped by environmental inputs, such as poverty, infectious disease, inadequate sanitation, and healthcare access, and (2) a country-specific starting point that is independent of such inputs. After removing the effects of environmental inputs, we find that different countries have reliably and substantially different starting points in average HAZ scores even before considering environmental inputs (a range of 1.7 SD). As expected from a two-component model, these country-specific starting points (basal HAZ) are not associated with key indicators of undernutrition (e.g., infant mortality and average calorie deficit). By contrast, increases in HAZ above a country's basal estimate (accrued HAZ) show strong correlations with these same variables, suggesting that accrued HAZ captures standard definitions of stunting as impaired growth due to undernutrition. Using these two components, we show how universal cutoffs can underestimate stunting in specific world regions (e.g., sub-Saharan Africa and the Caribbean), where children on average start off taller. As stunting is a high priority global health problem, standards that are sensitive to such population variation in healthy growth should improve efforts to target those most vulnerable to childhood undernutrition.
READ FULL TEXT