Abstract P212: Higher Rates of Breastfeeding Mediate the Association Between the 2009 Wic Food Package Change and BMI Z-score at Age 4

Abstract

Background: The Special Supplemental Nutrition Program for Women, Infants and Children (WIC) is a federal nutrition assistance program which provides supplemental food and nutrition services to children 0-5 years and their mothers, living in low-income households. In October of 2009, WIC food packages were changed to be better aligned to the Dietary Guidelines of Americans. Objective: The objective of this study was to evaluate whether observed differences in children’s BMI z-scores (BMIz) at age 4 years between WIC participants enrolled before and after the 2009 policy change could be attributable to changes in the type of infant feeding package they were exposed to. Methods: Our sample is based on children enrolled in WIC in Los Angeles County, CA between 2003 and 2016, from birth to age 5, with infant package issuance data for 11 or more of the first 13 months of life. Infant package exposure was quantified by the number of months receiving the fully breastfeeding infant package (a proxy for exclusive breastfeeding). A mediation analysis was performed in which BMIz at age 4 was the outcome, WIC package (new vs. old) was the exposure, and the number of months receiving the fully breastfeeding infant package was the mediator. Mediation was assessed by a product of coefficients method, and confidence limits for the mediated effect were based on the non-normal distribution of the mediated effect, using the PRODCLIN program in SAS. Models were gender-stratified and adjusted for child race/ethnicity, maternal education and language preference, household poverty status, initial weight status, and whether the child had a sibling in the sample. Results: The sample consisted of 140,204 children. Boys and girls exposed to the new WIC package received the fully breastfeeding infant package for 1.40 and 1.45 months longer than boys and girls exposed to the old WIC package. Among boys, the new WIC package was associated with a 0.12 standard deviation lower BMIz at age 4 (SE 0.01, p <0.0001), whereas the corresponding number for girls was 0.07 (SE 0.01, p<0.0001). Number of months exposed to the fully breastfeeding infant package partially mediated the association between exposure to the new WIC food package (vs. old) and BMIz at age 4, explaining 22.1% of the total effect (95% CI 19.1-25.1) for boys and 34.4% of the total effect (95% CI 29.9-38.9) for girls. Conclusions: Longer exposure to the fully breastfeeding infant package seemed to explain part of the association between BMIz at age 4 and receiving the new (vs. old) WIC food package. The percent of the total effect mediated by months of fully breastfeeding exposure was greater in girls than in boys. Future revisions to the WIC packages that continue to support women to fully breastfeed are likely to have significant effects on childhood obesity.