A Meta-Analysis of Obesity Interventions Among U.S. Minority Children

Abstract

To quantitatively evaluate the efficacy of interventions designed to prevent or treat obesity among U.S. minority children using meta-analytic techniques. A total of 40 intervention trials involving 10,725 children aged 6-19 years were examined. Interventions with more components showed a higher mean effect size than those with fewer components: among 32 controlled trials, d = .07 for one-component (n = 6); d = .08 for two-component (n = 15); d = .33 for three-component (n = 10); and d = .71 for four-component (n = 1) interventions. Interventions with parental involvement (n = 22, d = .21) and lifestyle interventions (n = 14, d = .34) showed a greater mean effect size than those without parental involvement (n = 10, d = .05) or lifestyle interventions (n = 18, d = .04), despite the fact that their 90% confidence intervals overlapped. Among uncontrolled trials (n = 8), two-component interventions (n = 5) yielded d = .86 and three-component interventions (n = 3) yielded d = .96. Evidence indicates that, among U.S. minority children, obesity interventions with three or more components might be more efficacious than those using fewer components. Parental involvement, lifestyle change, culturally-based adaptation, and interactive computer programs seem to show promise in the reduction of obese minority children.