Reducing children's television viewing to prevent obesity - A randomized controlled trial

Context Some observational studies have found an association between televi sion viewing and child and adolescent adiposity, Objective To assess the effects of reducing television, videotape, and vide o game use on changes in adiposity, physical activity, and dietary intake. Design Randomized controlled school-based trial conducted from September 19 96 to April 1997. Setting Two sociodemographically and scholastically matched public elementa ry schools in San Jose, Calif. Participants Of 198 third- and fourth-grade students, who were given parent al consent to participate, 192 students (mean age, 8.9 years) completed the study. Intervention Children in 1 elementary school received an 18-lesson, 6-month classroom curriculum to reduce television, videotape, and video game use. Main Outcome Measures Changes in measures of height, weight, triceps skinfo ld thickness, waist and hip circumferences, and cardiorespiratory fitness; self-reported media use, physical activity, and dietary behaviors; and pare ntal report of child and family behaviors. The primary outcome measure was body mass index, calculated as weight in kilograms divided by the square of height in meters. Results Compared with controls, children in the intervention group had stat istically significant relative decreases in body mass index (intervention v s control change: 18.38 to 18.67 kg/m(2) vs 18.10 to 18.81 kg/m(2), respect ively; adjusted difference -0.45 kg/m2 [95% confidence interval {CI}, -0.73 to -0.17], P = .002), triceps skinfold thickness (intervention vs control change: 14.55 to 15.47 mm vs 13.97 to 16.46 mm, respectively; adjusted diff erence, -1.47 mm [95% CI, -2.41 to -0.54]; P = .002), waist circumference ( intervention vs control change: 60.48 to 63.57 cm vs 59.51 to 64.73 cm, res pectively; adjusted difference, -2.30 cm [95% CI, -3.27 to -1.33]; P < .001 ), and waist to-hip ratio (intervention vs control change: 0.83 to 0.83 vs 0.82 to 0.84, respectively; adjusted difference, -0.02 [95% CI, -0.03 to -0 .01]; P < .001). Relative to controls, intervention group changes were acco mpanied by statistically significant decreases in children's reported telev ision viewing and meals eaten in front of the television. There were no sta tistically significant differences between groups for changes in high-fat f ood intake, moderate-to-vigorous physical activity, and cardiorespiratory f itness. Conclusions Reducing television, videotape, and video game use may be a pro mising, population-based approach to prevent childhood obesity.