Objective: To investigate whether the intestinal microbiota composition in early infancy is associated with subsequent weight development in children.

Methods: Analyses were conducted within the KOALA Birth Cohort Study (n = 2834). This cohort originates from two recruitments groups: pregnant women with a conventional lifestyle (no selection based on lifestyle) and pregnant women recruited through alternative channels (organic shops, anthroposophic clinicians/midwives, Steiner schools and relevant magazines). From 909 one-month-old infants, fecal samples were collected and analyzed by quantitative PCR targeting bifidobacteria, Bacteroides fragilis group, Clostridium difficile, Escherichia coli, Lactobacilli and total bacteria counts. Between the ages of 1 and 10 years, parent-reported weight and height was collected at 7 time points. Age- and gender-standardized body mass index (BMI) z-scores were calculated. Data were analyzed using generalized estimating equation.

Results: Colonization with B. fragilis group was borderline significantly associated with a higher BMI z-score of 0.15 (95% confidence interval (CI): -0.02 to 0.31), in the conventional subcohort. After stratification for fiber intake (P(forinteraction) = 0.003), colonization with B. fragilis group was associated with a 0.34 higher BMI z-score among children with a low-fiber intake in this subcohort (95% CI: 0.17-0.53). Higher counts among colonized children were positively associated with BMI z-score only in children within the conventional subcohort and a high-fiber diet (BMI z-score 0.08; 95% CI: 0.01-0.14), but inversely associated in children with a low-fiber diet (BMI z-score -0.05; 95% CI: -0.10 to 0.00), and in children recruited through alternative channels (BMI z-score -0.10; 95% CI: -0.17 to -0.03). The other bacteria were not associated with BMI z-scores, regardless of subcohort.

Conclusion: Using a targeted approach, we conclude that the intestinal microbiota, particularly the B. fragilis group, is associated with childhood weight development. To identify the potential impact of additional bacterial taxa, further prospective studies applying an unconstrained in-depth characterization of the microbiota are needed.