Abstract 3709: The Memphis Girls health Enrichment Multisite Studies (GEMS): An Evaluation of the Long Term Efficacy of A Weight Gain Prevention Intervention in African American Girls
Obesity in children has reached epidemic levels, and obesity prevalence in African American girls has increased even more dramatically. This study tested the efficacy of a carefully tailored and previously pilot-tested intervention to prevent excessive weight gain in African American girls. The Memphis GEMS (Girls health Enrichment Multisite Studies) trial randomly assigned 303 predominantly low income 8 –10 year old African American girls at high risk for obesity to a 2-year intervention either to prevent excessive weight gain or to an alternative control intervention promoting self esteem. The weight gain prevention intervention was a family based program that focused on improving dietary intake and increased physical activity. Girls in both interventions were offered weekly meetings for 14 weeks, followed by monthly contacts. The primary outcome was body mass index (BMI) at two years. Secondary outcomes included changes in body fat (bioelectric impedance), dietary intake (3-day diet recalls), and physical activity (3 days with accelerometer). Repeated measures, fixed-effects ANCOVA, using intention-to-treat principles with imputation for missing data, with adjustment for baseline BMI, was used to test for treatment differences in BMI. At two years, those assigned to the weight gain prevention intervention had significantly lower increases (46.9%) in BMI (p< .03; effect size [ES]= .226, 95%, CI: .06–.44) compared to those assigned to the alternative self-esteem intervention. Increase in fat mass was also less (p= .009, ES= .216, 95%, CI: .05–.43) whereas changes in fat free mass did not differ significantly (p= .67). Greater changes in dietary intake in the desired direction were observed: reduced total kilocalories (p= .03, ES= .324, CI: .13–.57), increased vegetable intake (p= .03, ES= .236, CI: .04–.49), increased water consumption (p= .02, ES= .431, CI: .23–.67), and decreased sweetened beverage intake (p= .03, ES= .32, CI: .13–.56). No significant differences in physical activity were observed. We conclude that a behavioral intervention can reduce weight gain in African American girls during a developmental period with high risk for weight gain.