Built environment and physical activity: domain- and activity-specific associations among Brazilian adolescents

Inácio Crochemore Mohnsam da Silva, Adriano Akira Hino, Adalberto Lopes, Ulf Ekelund, Soren Brage, Helen Gonçalves, Ana B Menezes, Rodrigo Siqueira Reis, Pedro Curi Hallal

Physical activity is likely to be determined as a complex interplay between personal, interpersonal, and environmental factors. Studying the built environment involves expanding the focus from the individual perspective to a public health one. Therefore, the objetive of this study was to examine the association between the built environment and objectively-measured physical activity among youth. Cross-sectional analysis of data from of a Brazilian birth cohort during adolescence. Physical activity was measured using accelerometers (GENEActiv) and self-report (International Physical Activity Questionnaire, long version). Participants' home addresses were geocoded and built environment characteristics such as streets' pattern and quality, and public open spaces attributes for physical activity practice were evaluated in a 500-m circular buffer surrounding their homes. A total of 3379 participants were included. Street lighting (β = 2.2; 95%CI: 0.5; 3.9) was positively associated with objectively-measured moderate-vigorous physical activity (MVPA) and proportion of paved streets and buffer's average family income were associated with lower MVPA. Living near the beach increased the odds of leisure-time MVPA practice by 3.3 (95%CI: 1.37; 8.02) times. There was a built environment-by-socioeconomic status (SES) interaction for the associations with commuting physical activity; street lighting [Odds ratio (OR) = 1.22; 95%CI: 1.01; 1.47] and presence of cycle lanes (OR = 1.77; 95%CI: 1.05; 2.96) were positively associated with commuting physical activity only among the intermediate SES tertile. Beachfront, street lighting, paved streets and cycle lanes were associated with physical activity patterns. This suggests that infrastructure interventions may influence physical activity levels of Brazilian adolescents.