Longitudinal associations between the school built environment and adolescents? physical activity: Evidence from the COMPASS study.

Purpose: Examine adolescent moderate-to vigorous-intensity physical activity (MVPA) and active school travel over a four-year period; Examine associations between the surrounding school built environment and adolescent MVPA and active school travel, and whether adolescent MVPA and active school travel are moderated by the built environment over time. Methods: Data from the COMPASS project (2013/14 to 2016/17) were used. Participants were adolescents aged 13-18 (n = 20,221) from 91 schools. MVPA and active school travel were selfreported via questionnaire. The built environment (park-, recreation-, retail-densities, Walk Score (R)) was objectively measured. Multilevel modeling was conducted. Results: Adolescent square root transformed MVPA (SQRT-MVPA) and the likelihood of active school travel decreased over time. Several positive associations were observed between the surrounding school built environment and SQRT-MVPA and active school travel when time was held constant. Adolescents attending schools in very walkable areas had an increased likelihood of active school travel over time. Higher park-, retail-, and recreation-densities appeared to help maintain the likelihood of active travel to school over time, though the same pattern was not apparent for active travel from school or SQRT-MVPA. Conclusion: SQRT-MVPA and active school travel declined over time. However, the downward trend for the likelihood of active school travel appeared to be opposite for students attending schools in very walkable areas. Schools and municipalities should work together in creating built environments around schools that are more walkable and have greater park and recreation densities. Schools in unfavourable built environments may need additional physical activity programming.