Abstract MP43: Breast Milk-dependent Associations Of Infant Gut Microbiota With Childhood BMI Z Score

Abstract

Introduction: The prevalence of overweight and obesity in children ages 2-5 years continues to rise in the US. Experimental germ-free animal models indicate gut microbiota can cause excess weight gain. Observational human studies, mostly cross-sectional, also suggest gut microbiota is associated with obesity, but these studies have largely been conducted in older children and adults. Infants have a unique gut microbiota composition and function, under strong influence by human milk. Bifidobacteria, in particular, is hypothesized to be beneficial in the presence of human milk oligosaccharides. To our knowledge, no longitudinal studies have examined the association of the infant gut microbiota with childhood BMI, taking into account intake of human milk. Hypothesis: We hypothesized that infant gut microbiota composition, in particular relative abundance of Bifidobacteria, in the first year is prospectively associated with differences in child BMI from ages 2-5 years and associations are modified by duration of human milk feeding. Methods: We examined longitudinal data from mother-child dyads in the New Hampshire Birth Cohort, which began enrolling pregnant women from New Hampshire in 2009. We measured the infant gut microbiota using 16S rRNA sequencing at 6 weeks and 12 months of age. We estimated alpha diversity using the Shannon diversity index. Child BMI z scores (BMI-z) at 2-5 years of age were calculated using sex- and age-specific WHO growth charts. We used unadjusted and multivariable adjusted linear mixed models. We adjusted for pre-pregnancy BMI, birth weight, delivery mode, and infant BMI-z at 12 months (in 12-month microbiota models). We considered effect measure modification by breastfeeding duration. Results: Our analytic sample comprised 148 and 146 infants with microbiota data at 6 weeks and 12 months, respectively, and at least 1 BMI-z from ages 2-5 years. Shannon diversity at 6 weeks and 12 months of age, and top genera at 6 weeks were not significantly associated with child BMI-z. Abundance of 12-month Bifidobacterium was associated with lower BMI-z (-0.12; 95% CI (-0.25, 0.006)) and interacted with breastfeeding duration (p interaction <0.01); among infants consuming breast milk ≥6 months Bifidobacterium was associated with a 0.22 (95% CI: 0.06, 0.37) lower child BMI-z. Lower abundance of 12-month Prevotella , a bacteria linked to obesity in adults, was suggestively associated with lower child BMI-z and this association was also modified by breastfeeding duration (p interaction=0.01), such that it was only significant among infants breastfed < 6 months (0.34; 95% CI: 0.09, 0.60). Both interactions were consistent when breastfeeding duration was dichotomized at 12 months. Conclusion: Higher percent Bifidobacterium and lower Prevotella at 12 months was prospectively associated with lower childhood BMI-z, and both associations were modified by breastfeeding duration.