Maternal and child factors associated with child body fatness in a Ghanaian cohort

Abstract Objective: We aimed to identify factors (child diet, physical activity; maternal BMI) associated with body composition of Ghanaian pre-school children. Design: Longitudinal analysis of the International Lipid-Based Nutrient Supplements (iLiNS)-DYAD-Ghana randomized trial, which enrolled 1320 pregnant women at ≤20 weeks’ gestation and followed them and their infants until 6 and 18 months postpartum, respectively. At follow-up, child age 4–6 years, we collected data on body composition (by 2 H dilution), physical activity and diet, extracted dietary patterns using factor analysis, and examined the association of children’s percentage body fat with maternal and child factors by regression analysis. Setting: Eastern Region, Ghana. Participants: Children 4–6 years of age. Results: The analysis included 889 children with percentage body fat and dietary data at follow-up. We identified two major dietary patterns, a snacking and a cooked foods pattern. Percentage body fat was positively associated (standardized β ( se )) with maternal BMI at follow-up (0·10 (0·03); P = 0·003) and negatively associated with physical activity (−0·15 (0·05); P = 0·003, unadjusted for child gender), but not associated with the snacking (0·06 (0·03); P = 0·103) or cooked foods (−0·05 (0·07); P = 0·474) pattern. Boys were more active than girls (1470 v. 1314 mean vector magnitude counts/min; P < 0·0001) and had lower percentage body fat (13·8 v. 16·9 %; P < 0·0001). Conclusions: In this population, maternal overweight and child physical activity, especially among girls, may be key factors for addressing child overweight/obesity. We did not demonstrate a relationship between the dietary patterns and body fatness, which may be related to limitations of the dietary data available.