Abstract
OBJECTIVE: To measure the relationship between screen exposure and obesity in a large, urban sample of children and to examine whether the relationship is moderated by sociodemographics. METHODS: We asked parents of 11 141 children visiting general pediatrics clinics if the child had a television (TV) in the bedroom and/or watched more than 2 hours of TV/computer daily. We measured children's height and weight, then used logistic regression to determine whether screen exposure indicators predicted obesity (body mass index ≥ 85th percentile) and interacted with race/ethnicity, sex, age, and health care payer. RESULTS: Having a TV in the bedroom predicted obesity risk (P = .01); however, watching TV/computer for more than 2 hours a day did not (P = 0.54). There were no interactions. CONCLUSIONS: Asking whether a child has a TV in the bedroom may be more important than asking about duration of screen exposure to predict risk for obesity.