Anthropometry relationship with duodenal histologic features of children with environmental enteric dysfunction: a multicenter cross-sectional study

Background: Environmental enteric dysfunction (EED) is a precursor of growth faltering in children living in impoverished conditions who are frequently exposed to environmental toxins and enteropathogens, leading to small bowel inflammatory, malabsorptive, and permeability derangements and low-grade chronic systemic inflammation. Objectives: We explored the association between anthropometrics and duodenal histologic features of EED among children from 3 lower middle-income country centers. Methods: In this cross-sectional study, Pakistani children (n = 63) with wasting, Bangladesh children (n = 116) with stunting or at risk for stunting (height-for-age Z score [HAZ] <−1 but ≥−2), and Zambian children (n = 108) with wasting or stunting received nutritional intervention. Children with anthropometric status refractory to intervention underwent endoscopy. Linear regression models included anthropometric around endoscopy, scores of histology parameters, and a global index score of EED—the total score percent-5 (TSP-5). Multivariable models were adjusted for center, age, sex, and histology slide quality. Results: Intersite variation was observed while exploring the association between anthropometrics and the TSP-5; for example, Pakistani children had the worst HAZ, yet their median TSP-5 score was lower than that of the other 2 centers. Even within each site, no overall pattern of higher TSP-5 score was observed with worsening HAZ. During univariate analysis, TSP-5 (coefficient: 0.01; 95% confidence interval [CI]: 0, 0.02), goblet cell depletion (coefficient: 0.22; 95% CI: 0.06, 0.37), and Paneth cell depletion (coefficient: 0.14; 95% CI: 0.01, 0.27) were associated with HAZ scores; however, they lost statistical significance in the multivariable models, with study center most strongly confounding the relationships seen in univariate models between anthropometry and histology. Conclusions: This study contributes a crucial negative finding that duodenal morphological features did not associate with anthropometric phenotypes; hence, anthropometric measurements may not be a suitable outcome measure for use in EED trials. Trial outcomes may need to be defined by combining the functional and structural elements of the gut to monitor EED.