The Effect of Pediatric Obesity Interventions on BMI Among Hispanic Youth: A Meta-Analysis

Introduction: Pediatric obesity is a major public health issue in the U.S. that disproportionately affects Hispanics. Previous systematic reviews of obesity-related interventions in Hispanic youth have had narrow inclusion criteria and have failed to calculate an overall quantitative effect on youth weight status. This meta-analysis provides the first overall quantitative summary of intervention effects in this population. Our aim was to summarize how well obesity-related interventions for Hispanic youth work in terms of changing their body mass index (BMI) and to determine if, and how, intervention effects vary as a function of potential moderators (e.g., developmental stage, study setting).

Method: We conducted a systematic review of the databases PubMed, PsycInfo, and Scopus using PRISMA guidelines. We included studies written in the English language and published between 2000-2018 that focused on either the prevention or treatment of obesity in 0-18 year old Hispanics/Latinos. Based on the Cohen’s d effect size with sample size bias-correction, we calculated Hedges’ g and subsequently conducted the overall test of homogeneity for BMIz using a Q_total statistic. If Q_total was significant, we then ran a random-effects model or a mixed-effects model with the moderators of interest.

Results: Sixty-one studies met inclusion criteria. The Q_total statistic (Q(60)=184.91, p<.0001) indicated a significant amount of heterogeneity between studies. Under the random effects model, the overall estimated intervention effect was -0.19 (SE=.03, 95% CI: -0.25-^-.12, p<.0001), indicating that interventions produced slight reductions in BMI compared to controls. With regard to moderators, developmental stage, youth weight status, income, and study setting all significantly explained variation in effect sizes. Specifically, studies conducted during infancy (g= -0.52, SE= 0.14) and early-mid adolescence (g= -0.33, SE= 0.07) showed larger effects than those conducted in early childhood (g=-0.15, SE=0.06) and childhood (g=-.12, SE=0.04). Studies conducted with predominantly overweight/obese (g=-.22, SE=0.05) versus normal or mixed populations (g=-.16, SE=.04) and those with higher-income (g=-.29, SE=0.07) versus lower-income (g=-.16, SE=0.03) populations also showed stronger effects on BMI. Studies conducted in the family (g=-.17, SE=0.06), school (g=-.19, SE=0.04), and primary care (g=-.49, SE=0.11) settings showed significant reductions in BMI versus those conducted at the individual and community levels.

Conclusion: Overall, existing obesity prevention and treatment interventions for Hispanic youth, particularly those conducted with higher weight, less disadvantaged youth in family, school, and primary care settings, show promise in terms of reducing youth BMI. We plan to conduct similar analyses for physical activity and dietary outcomes.