Association Between Diet Patterns of Human Infants and Growth Differs Across Randomized Formula Groups in the First 18.5 Months

Background: The composition of the formula fed to infants and the velocity of weight gain during infancy can impact later weight status. However, whether patterns of food and beverage intake during the complementary feeding period affect this relationship remains unclear. Objectives: To identify patterns of complementary feeding (primary aim) and to examine how these diet patterns, along with infant formula composition and early weight gain velocity, influence weight status over time (secondary aim). Methods: Diet records were collected monthly from 0.5 to 12.5 mo and again at 18.5 mo (n = 80) as part of a trial involving healthy infants randomly assigned to receive either cow milk formula (CMF) or an isocaloric extensively hydrolyzed protein formula (EHF). Foods and beverages were assigned to mutually exclusive What We Eat in America food categories, and intake at each age was standardized for body weight (g/kg). Diet patterns were derived using principal component analysis. Generalized estimating equations examined the independent and interactive effects of formula type (fixed effect), rapid weight gain (RWG) category (rapid/nonrapid; fixed effect), and infant diet patterns (time-varying effect) on weight-for-length Z-score (WLZ) over time. Results: Over the first 18.5 mo, 4 diet patterns were identified: (PC1) “infant formula;” (PC2) “low-sugar cereal, not fried potatoes, grain mixed dishes, fruit, vegetables, meat mixed dishes, 100% fruit juice;” (PC3) “milk, pizza, water, high-sugar cereal and snack/meal bars;” (PC4) “fried poultry, bakery products, fried potatoes, starchy vegetables.” Over time, the effect of RWG category on WLZ was significant for the CMF (χ² = 127.15, P ≤ 0.001) and EHF (χ² = 140.71, P ≤ 0.001) groups. However, the main effect of PC1 (β = 0.189, P = 0.007) and subsequent greater WLZ score over time (χ² = 17.95, P < 0.05) were significant only in the CMF group. Conclusions: The prolonged feeding of CMF, but not EHF, during complementary feeding was associated with greater WLZ in the first 18.5 mo. This trial was registered at clinicaltrials.gov as NCT01700205 (https://clinicaltrials.gov/study/NCT01700205).