Purpose: To investigate the association between the likelihood of frequent vigorous physical activity (VPA), characterized by 7 hours/week or more, and various aspects of the built environment that promote opportunities for exercise (intersection density, number of street intersections, total length of exercise and walking paths, sports facilities and urban green space). Methods: We measured self-reported frequency of VPA in boys and girls using the Health Behavior in School-Aged Children 2006 survey (n = 6046) in 80 schools. We conducted multi-level logistic regression analyses to examine the relationship between the frequency of VPA and aspects of the built environment (intersection density, number of street intersections, total length of exercise and walking paths, sports facilities and urban green space) within 2 km surrounding each school. We controlled for confounding variables at the individual (grade, family social class) and neighborhood levels (urbanicity) and employed the use of interaction terms between family social class and the built environment variables. Results: Of all built environment variables examined, urban green space devoted to recreational use was overall the strongest predictor of frequent VPA while adjusting for individual characteristics. Boys from low social class families or those attending schools with little urban green space had significantly lower odds (OR = .66 and .67, respectively) of achieving frequent VPA compared to high social class or high access referents. Similarly, girls from low social class families had lower odds (OR = .63) of achieving frequent VPA compared to high income referents. However, the likelihood was greater (OR = 1.42) for girls attending schools with low urban green space. Interactions between family social class and urban green space showed even decreased likelihood of achieving frequent VPA for boys and girls from low social class backgrounds attending schools with little urban green space (OR = .49 and .28 respectively) . The synergistic effect of exercise paths, the number of sites devoted to urban green space, and the number of sports facilities showed significant reductions in frequent VPA in boys (OR = .59), but no effect in girls. Results from the interaction between family social class and the combined variable showed boys and girls from low social class backgrounds attending schools with the least exposure were significantly least likely to achieve frequent VPA (OR = .44 and .19, respectively). Conclusions: The likelihood of frequent vigorous exercise was strongly correlated with family social class for both boys and girls. However, the association between frequent VPA and urban green space was greater for boys than girls. In combination with family social class, the likelihood of frequent VPA was signficantly reduced in boys and girls from low social class backgrounds attending schools with low exposure to urban green space. Findings were similar for boys and girls even when a combination of built environment exercise exposures were employed. Results indicate the need to explore interactions between the built environment and socioeconomic status when examining individual physical activity, especially concerning adolescents of low SES with limited access to opportunities for physical activity.
schools; physical activity; built environment; youth