Background The link between high protein intake during infancy and obesity later in childhood has been much debated, and the association with differing protein sources remains unclear.

Objective This study aimed to examine the associations between total protein intake and protein from different sources (ie, nondairy animal, dairy, and plant) reported at age 9 months and development in body mass index (BMI) z scores until age 5 years.

Design This study involved a secondary data analysis of the Melbourne InFANT (Infant Feeding, Activity and Nutrition Trial) program, an observational prospective cohort study that was conducted from 2008 to 2013.

Participants/setting Participants were children (n = 345) who completed both the 9-month and 5-year follow-up visits within the Melbourne InFANT program.

Main outcome measures BMI z score was measured at age 5 years.

Statistical analyses performed Linear mixed models with a random effect for clusters of mother's group and with adjustment for baseline child and maternal covariates were conducted.

Results With adjustment for covariates, every 1 g or 1% energy increase in total protein intake at age 9 months was associated with a 0.016-unit (95% CI 0.003 to 0.029) or 0.034-unit (95% CI 0.005 to 0.063) increase in BMI z score at age 5 years, respectively. With respect to protein sources, associations of similar magnitude were found for nondairy animal protein. No evidence of an association with BMI z score was found for dairy (including milk, yogurt, cheese, breast milk, and infant formula) and plant proteins.

Conclusions High intakes of total protein, nondairy animal protein, but not dairy or plant proteins, during infancy were associated with higher BMI z score in early childhood. These findings can inform dietary recommendations regarding protein intakes during infancy.