Twenty-two potential correlates of children's physical activity were examined. Two hundred and one Mexican-American and 146 Anglo-American families with 4-year-old children were studied. Children's physical activity was directly observed in the evening at home on 4 visits for 1 hr each time. Anglo-American children and male children were found to be more active. Demographic variables explained 11% of the variance in children's physical activity. After adjusting for demographics, 3 children's variables and 6 social-family variables did not account for significantly more variance. Five environmental variables accounted for 11% additional variance. Variables observed concurrently with physical activity, such as time spent outdoors and prompts to be active, were highly associated with children's physical activity.

We investigated physical activity and encouragement for activity in a bi-ethnic cohort during recess. Activity and associated interactions of 287 children were recorded at preschool and again 2.2 years later. Children expended nearly twice as much energy at preschool recess than at elementary recess. Activity levels declined as recess time elapsed. At preschool, European-American children engaged in more moderate to vigorous activity than Mexican-Americans. As participants moved to elementary school, teachers' prompts to be active decreased and prompts from peers increased. Boys and girls received similar amounts of activity prompts at preschool, but prompts to boys increased over time. The findings suggest that school environments could be altered to promote healthful physical activity among young children.

Identifying correlates of children's television (TV) watching could help identify strategies to reduce children's TV watching and impact children's weight status. Children's TV time and home environment factors were assessed longitudinally among 169 families with children followed from 6 to 12 years of age. TV watching increased with age, as did the number of in-home TVs and the presence of videocassette recorders (VCRs), the frequency of meal eating while watching TV, and the percentage of children with bedroom TVs. Greater meal eating while watching TV was consistently related to overall TV watching and increases in TV watching over time. The number of in-home TVs and the presence of a bedroom TV were less consistent correlates. TV watching was related to weight status when children were younger, and when children were older, more than 2 hours of TV daily was a risk factor for higher weight. Modifying TV access, particularly by reducing the frequency of meals eaten while watching TV, could promote lower children's TV watching and adiposity.