Protein intake, breastfeeding frequency and breastfeeding duration affect BMI from infancy to mid‐childhood

Background Protein intake (PI) is an important determinant of growth, but excess protein in infancy has been associated with increased risk of obesity. Few studies have explored age‐specific associations of PI with BMI from infancy to adulthood. Objective To analyze how PI and breastfeeding relate to infant and toddler BMI, and how breastfeeding history and PI relate to BMI from age 2 to 22 yr using 24‐hour dietary recalls and anthropometric data from the Cebu Longitudinal Health and Nutrition birth cohort. Methods Random‐effects longitudinal regression models were used to estimate 1) the joint association of daily breastfeeding frequency (BF) and dietary protein residuals with concurrent BMI Z‐scores measured bi‐monthly from 2 to 24 months (n=2899), and 2) the association of breastfeeding history and protein residuals with concurrent BMI measured at ages 2, 11, 15, 19 and 22 years (n=2435). Where appropriate, models included statistical interactions (PI*energy, energy*age, BF*energy, BF*age), and were adjusted for socio‐demographic and anthropometric confounders (maternal education, maternal height and maternal age at baseline, offspring sex, birth weight and age‐2 BMI, and time‐varying household assets, urbanicity, offspring age and offspring education). Results 1) Relative to not being breastfed the previous day, breastfeeding was associated with higher zBMI, though this effect ameliorated with time (Beta=0.524 at 6mo & 0.138 at 18mo). An additional 5 breast milk feedings in the previous day contributed to higher zBMI in younger infants (B=0.030 at 6mo) and lower zBMI in later infancy (B=−0.075 at 18mo). Compared to those at the 25 th percentile of protein intake, being at the 75 th percentile of protein intake was positively associated with zBMI regardless of age, however, the magnitudes of these associations were highest in non‐breastfed and older infants (at 6 mo: B=0.019 vs 0.005, at 18 mo: B=0.024 vs 0.012 for non‐breastfed vs breastfed infants respectively). 2) In post‐infancy analyses, 15 months of breastfeeding duration was inversely associated with BMI at age 11 (Beta=−0.145). Protein residuals were not significantly associated with BMI in the post‐infancy analyses. (All shown betas were statistically significant, p<0.05.) Implications Breastfeeding frequency, dietary protein intake and breastfeeding duration are significant contributors to concurrent BMI in infancy and childhood. Support or Funding Information Howard Hughes Medical Institute International Student Research Fellowship