Determinants of activity-friendly neighborhoods for children: results from the SPACE study

PURPOSE: To examine the association between children's physical activity and factors of the built environment. DESIGN: Cross-sectional study. Setting. Ten neighborhoods in six cities in the Netherlands. SUBJECTS: Four hundred twenty-two children (age range, 6-11 years; 49% male). MEASURES: Physical activity diary, neighborhood observations, and anthropometric measures. ANALYSIS: Univariate and multivariate linear regression analyses. RESULTS: According to univariate analyses adjusted for age, sex, body mass index, and highest level of maternal education, physical activity (> or = 3 metabolic equivalents) was significantly (p < .05) associated with the proportion of green space, with the residential density, with the general impression of activity-friendliness of the neighborhood, and with the frequency of certain types of residences (e.g., terraced houses), sports fields, water, dog waste, heavy traffic, and safe walking and cycling conditions (e.g., cycle tracks and 30-km speed zones) in the neighborhood. According to adjusted multivariate analyses, physical activity was best predicted by the frequency of parallel parking spaces in the neighborhood and by the general impression of activity-friendliness of the neighborhood (I2 = 0. 193). CONCLUSIONS: Children's physical activity is associated with certain modifiable factors of the built environment. Longitudinal studies should examine whether there is a causal relationship