Exercise during pregnancy and infant body mass index during the first year of life: a secondary per-protocol analysis of a randomized clinical trial

Introduction Maternal lifestyle during pregnancy may influence infant growth trajectories. Prenatal exercise improves maternal outcomes, but associations with infant Body Mass Index (BMI) beyond early infancy remain uncertain. The aim was to examine the association between participation in a supervised prenatal exercise program and infant BMI during the first year of life. Methods This secondary per-protocol analysis was derived from a randomized clinical trial (NCT04563065) conducted in three public hospitals. Pregnant women were allocated to a supervised exercise program performed throughout pregnancy (IG = Intervention Group) or to standard obstetric care (CG = Control Group). Infant weight, length, BMI, and feeding type were assessed at 1, 2, 4, 6, and 12 months postpartum. Primary analyses followed a per-protocol approach restricted to mother–infant dyads meeting predefined adherence and follow-up criteria, with additional intention-to-treat sensitivity analyses. Results Among 229 randomized participants, 126 mother–infant dyads were included in the per-protocol analyses. Infant BMI trajectories were similar between groups through 6 months of age. At 12 months, mean infant BMI was lower in the IG compared with controls in both per-protocol (16.49 ± 1.26 vs. 17.10 ± 1.40; p = 0.006) and intention-to-treat analyses (16.50 ± 1.25 vs. 17.07 ± 1.37; p = 0.007), although the absolute difference was modest. Infants in the CG had higher odds of overweight during mid and late infancy across both analytic approaches. Mothers in the IG gained less weight during pregnancy and had higher rates of exclusive breastfeeding during early infancy. Discussion In this secondary per-protocol analysis, participation in a supervised prenatal exercise program was associated with lower infant BMI at 12 months and more favorable early feeding patterns. Given substantial attrition and the per-protocol analytic approach, findings should be interpreted cautiously and considered exploratory and hypothesis-generating.