Spatial classification of youth physical activity patterns.

BACKGROUND Physical activity is an essential element in reducing the prevalence of obesity, but much is unknown about the intensity and location of physical activity among youth-this is important because adolescent health behaviors are predictive of behaviors in adults. PURPOSE This study aims to identify the locations where youth moderate-to-vigorous physical activity (MVPA) occurs, and to examine how MVPA varies according to urbanicity (urban, suburban, rural). METHODS Participants included adolescent students (N=380, aged 12-16 years) from Halifax, Nova Scotia. Locations of MVPA were measured using accelerometers and GPS data loggers for up to 7 days. Specialized software was developed to integrate and process the data. Frequencies of MVPA by location were determined, and differences in MVPA were assessed for association with urbanicity. RESULTS Active commuting accounted for the largest proportion of time in MVPA among urban and suburban students. Rural students achieved most MVPA at school. Other residential locations, shopping centers, and green spaces accounted for a majority of the remaining MVPA. Minutes in MVPA varied significantly overall (196.6 ± 163.8, 84.9 ± 103.2, 81.7 ± 98.2); at school (45.7 ± 45.2, 18.6 ± 28.0, 29.8 ± 39.7); while commuting (110.3 ± 107.1, 31.5 ± 55.2, 19.5 ± 39.7); and at other activity locations (19.7 ± 27.1, 14.8 ± 26.8, 12.0 ± 22.1) and by urbanicity. CONCLUSIONS Findings reveal that the journeys between locations are as important as home and school settings in contributing to greater MVPA in adolescent youth. The relative importance of context as a contributor to MVPA varies with urbanicity. Combining actimetry and GPS data provides a precise link between physical activity measurements and contexts of the built environment.