Estimated Causal Effects of Complementary Feeding Behaviors on Diet Quality in Early Childhood Using Birth Cohort Data

Abstract

Abstract Objectives We aimed to examine effects of 3 complementary feeding behaviors on later diet quality using a causal inference framework. Methods Using data from 1041 mother-child pairs from the Boston, MA-area Project Viva cohort, we estimated effects on the mean Youth Healthy Eating Index (YHEI) score, a measure of diet quality in early childhood, of hypothetical interventions that expose (vs. do not expose) to the following: 1) introduce sweets and fruit juice at ≥12 m and provide no daily fruit juice at age ∼1 y; 2) introduce varied, non-sweet flavors (at least 2/3 of fish, eggs and peanut butter) at <12 m; and 3) continue offering foods initially refused by the child. Mothers reported complementary feeding behaviors (exposures) at 1y and completed food frequency questionnaires for children in early childhood (median age 3.2 y). We estimated average treatment effects (ATEs) using inverse probability weighted linear regression analysis with stabilized weights to adjust for both confounding and selection bias due to censored outcomes. For each effect, we adjusted for child race/ethnicity, maternal socio-demographics, BMI, and diet quality, and the other 2 exposures, and examined effect modification by child sex and infant feeding mode (breastfed at 6 m [BF group] vs. not [FF group]). Results Of 1041 pairs, 12% delayed introducing sweets/fruit juice and provided no daily fruit juice at 1y, 68% introduced fish, eggs and peanut butter before 12 m, and 93% continued to offer refused foods. We estimated that mean YHEI score was higher under “delay sweets and fruit juice” (1.8 points, 95% CI: −1.0, 5.5), particularly among the BF group (ATE 5.1 points, 95% CI: 0.0, 8.7) and females (ATE 4.4 points, 95% CI: −2.0, 9.0). ATEs for the “early flavor variety” exposure were 1.7 points (95% CI: 0.2, 3.0) and similar among males and females but stronger (2.6 points, 95% CI: 0.4, 4.8) among the FF group. Mean YHEI score was also higher (2.6 points, 95% CI: −1.1, 6.5) under the “continue offering refused foods” exposure, particularly among females (ATE 5.7 points, 95% CI: 1.2, 9.6). Conclusions Delayed introduction of sweets and fruit juice, early introduction of varied, non-sweet flavors, and continued offering of initially refused foods may result in better diet quality in early childhood. Effects may differ by child sex and breastfeeding status. Funding Sources US NIH.