Abstract 3710: Stanford GEMS (Girls health Enrichment Multisite Studies): Long-Term Efficacy of After-School Dance and Screen Time Reduction in Low-Income African-American Girls
Objective: To test the efficacy of a 2-year, community- & family-based intervention to reduce weight gain in low-income, preadolescent African-American girls
Design: 2 arm, parallel group, randomized controlled trial Setting: Low-income areas of Oakland, CA Participants: 8 –10 year old African-American girls
Interventions: Girls were randomized to a 2 year, culturally-tailored after school dance program and a home/family-based intervention to reduce screen media use (dance/TV) versus an information-based community health education (HE) active-placebo comparison intervention Primary
Outcome: Change in body mass index measured up to 5 times over the course of the study
Results: 261 girls were randomized (134 dance/TV, 127 HE) and 225 (86.2%) completed 1 or more follow-up measures (118 dance/TV, 107 HE; mean ± SD follow-up = 25.7 ± 8.5 mos and 25.2 ± 9.6 mos, respectively). Intent-to-treat analysis found no statistically significant differences between groups in BMI change (mean difference in changes= .04 kg/m2 per year, 95% confidence interval [95% CI] =−.19 to .27; P=.72). Similar results were found for waist circumference and triceps skinfold. Compared to girls in HE, girls in the dance/TV group significantly reduced fasting total cholesterol levels (difference in changes = −3.49 mg/dl per year, 95% CI −5.28 to −1.70; P<.001), fasting LDL-cholesterol (−3.02, 95% CI −4.74 to −1.31; P<.001) and depressive symptoms (P<.05). Changes between groups were not statistically significant for HDL-cholesterol, triglycerides and insulin, resting blood pressure and heart rate, dietary energy and fat, accelerometer-measured physical activity, and screen time. Moderator analysis indicated the dance/TV reduction intervention was significantly more effective than HE in reducing BMI gain for girls with single parent families, more screen time and more depressive symptoms at baseline.
Conclusions: A culturally-tailored after school dance and screen time reduction intervention did not slow BMI increases more than a health education intervention in low-income African-American girls. The dance/TV intervention resulted in greater reductions in fasting total and LDL-cholesterol and depressive symptoms, but not in other cardiovascular disease risk factors.