Testing an Ecological Model of Obesity Among School-Age Children: Identifying Targets for Tailored Intervention

Childhood obesity rates remain historically high in the US. One way to conceptualize the many factors that contribute to obesity is through the use of an ecological model. There is a particular need to adapt and test this type of comprehensive model among vulnerable racial/ethnic and socioeconomic groups. Using a large sample of US youth drawn from the ECLS-K:2011 (N=8,225), this project first investigated an ecological model of childhood obesity from kindergarten to second grade, including factors such as child physical activity, child screen time, child bedtime, family physical activity, family food insecurity, family meals, and neighborhood safety. Then, it compared the contributions of each individual factor across racial/ethnic, socioeconomic, and income-to-needs groups, concurrently and longitudinally. Among the full sample, the largest standardized effect on weight was for income-to-needs ratio. Moving from above to below 200% of the poverty line resulted in an increase of 0.12 standard deviations in zBMI. Multigroup analyses indicated that there were no differences in model fit by socioeconomic status or sex. However, there was a significant difference in model fit based on race/ethnicity. Among Latino youth, income-to-needs ratio was a significant negative predictor of kindergarten zBMI; however, this effect was not significant among Black youth. Overall, findings highlighted the impact of income-to-needs ratio, as children of families who fell below 200% of the poverty line were more likely to weigh more