Differences by race and ethnicity in the relationship between breastfeeding and obesity in preschool children.

OBJECTIVE To test the hypothesis that the relationship between breastfeeding and later obesity would differ by race/ethnicity. DESIGN Data were obtained from the Fragile Families and Child Wellbeing Study, a prospective birth cohort study. SETTING Twenty large US cities in 15 states. PARTICIPANTS The 2146 three-year-old children in the study were born between 1998 and 2000 and were either non-Hispanic White, non-Hispanic Black, or Hispanic (any MAIN OUTCOME MEASURES Obesity was defined as having a BMI > or =95th percentile. RESULTS Fifty-two percent of the children were ever breastfed and 18% were obese at 3 years of age. After adjustment for covariates (maternal BMI, smoking, age, relationship status, and education, plus the child's birth weight and the household income-to-poverty ratio), the relationship between breastfeeding and the prevalence of obesity was significantly different between White, Black, and Hispanic children (P = .02). The adjusted prevalence of obesity was lower in Hispanic children who were ever breastfed compared to those who were never breastfed (23.3% vs. 33.0%, P=.01), but there was no significant association between breastfeeding and obesity in either White or Black children (16.6% vs. 11.3%, P = .18 for Whites and 18.0% vs. 14.5%, P = .13 for Blacks). CONCLUSIONS In this study we found that the relationship between breastfeeding and obesity differed by race/ethnicity. Future studies examining the relationship between breastfeeding and later adiposity should consider the possibility that this relationship may differ by race/ethnicity.