Early Childhood Obesity in the United States: An Assessment of Racial/Ethnic Disparities and Risk Factors.

Abstract

This dissertation focused on childhood obesity among preschool aged children in the United States, using data from the Early Childhood Longitudinal Study Birth Cohort. Chapter 1 examined racial/ethnic differences in preschool-aged children’s weight trajectories and identified sensitive periods at which disparities emerge, using mixed growth models and nonparametric LOESS curves. Racial/ethnic disparities in US children’s weight-status and growth trajectories emerge at different ages for different racial groups, but they are generally well established by kindergarten age. Our findings indicate that interventions designed to prevent early childhood overweight/obesity should be implemented early in the life-course. Chapter 2 assessed the contribution of behavioral and environmental risk factors to racial/ethnic disparities in preschool children’s weight status, using decomposition analyses to estimate the percent of disparity explained by individual obesity risk factors. Gaps in the prevalence of socio-economic-status (SES) accounted for a substantial part (ranging from 24.4% to 63.3%) of the explained disparities in BMI z-scores between racial/ethnic minority children and their white peers. Apart from SES and its correlates, infant weight gain during the first 9-months of life, lack of breastfeeding, early introduction of solids, and sugar sweetened beverage consumption were additional factors that played important roles in explaining racial/ethnic differences. Interventions implemented early in the life-course that target these key contributory risk factors could potentially help reduce the magnitude of racial/ethnic disparities in early childhood obesity Chapter 3 examined the effect of attending childcare on children’s BMI z-scores, employing OLS regression, as well as two quasi-experimental approaches designed to minimize the effect of selection bias and unmeasured confounding. In linear regression models, compared to children in parental care, children in non-parental childcare at 24 months had higher BMI z-scores at kindergarten entry. However, both quasi-experimental approaches revealed no significant effect of childcare attendance on children’s BMI z-score, suggesting that the link between non-parental childcare and obesity may not be causal. Previously reported associations may be confounded by unobserved family circumstances resulting in selection into different types of childcare arrangement.