Small-Quantity Lipid-Based Nutrient Supplements Do Not Affect Plasma or Milk Retinol Concentrations Among Malawian Mothers, or Plasma Retinol Concentrations among Young Malawian or Ghanaian Children in Two Randomized Trials

Abstract

ABSTRACT Background Vitamin A (VA) deficiency is prevalent in preschool-aged children in sub-Saharan Africa. Objectives We assessed the effect of small-quantity lipid-based nutrient supplements (SQ-LNS) given to women during pregnancy and lactation and their children from 6 to 18 mo of age on women's plasma and milk retinol concentrations in Malawi, and children's plasma retinol concentration in Malawi and Ghana. Methods Pregnant women (≤20 wk of gestation) were randomized to receive daily: 1) iron and folic acid (IFA) during pregnancy only; 2) multiple micronutrients (MMN; 800 μg retinol equivalent (RE)/capsule), or 3) SQ-LNS (800 μg RE/20g) during pregnancy and the first 6 mo postpartum. Children of mothers in the SQ-LNS group received SQ-LNS (400 μg RE/20 g) from 6 to 18 mo of age; children of mothers in the IFA and MMN groups received no supplement. Plasma retinol was measured in mothers at ≤20 and 36 wk of gestation and 6 mo postpartum, and in children at 6 and 18 mo of age. Milk retinol was measured at 6 mo postpartum. VA status indicators were compared by group. Results Among Malawian mothers, geometric mean (95% CI) plasma retinol concentrations at 36 wk of gestation and 6 mo postpartum were 0.97 μmol/L (0.94, 1.01 μmol/L) and 1.35 μmol/L (1.31, 1.39 μmol/L), respectively; geometric mean (95% CI) milk retinol concentration at 6 mo postpartum was 1.04 μmol/L (0.97, 1.13 μmol/L); results did not differ by intervention group. Geometric mean (95% CI) plasma retinol concentrations for Malawian children at 6 and 18 mo of age were 0.78 μmol/L (0.75, 0.81 μmol/L) and 0.81 μmol/L (0.78, 0.85 μmol/L), respectively, and for Ghanaian children they were 0.85 μmol/L (0.82, 0.88 μmol/L) and 0.88 μmol/L (0.85, 0.91 μmol/L), respectively; results did not differ by intervention group in either setting. Conclusions SQ-LNS had no effect on VA status of mothers or children, possibly because of low responsiveness of the VA status indicators.