Neighborhood and home environments and GPS-derived location-specific physical activity in adolescents

Purpose: To compare relations of neighborhood walkability and home equipment to (1) total and (2) location-specific physical activity.

Methods: 290 adolescents (M age=14.1±1.4 years; 45% girls; 74.7% White non-Hispanic) wore GPS and accelerometers for 5.4±1.9 days set to record at 30-second epochs. GIS was used to calculate a walkability index (sum of z-scores for density, mixed-use, connectivity, and retail floor-area-ratio) and classify GPS points as in-neighborhood (1km street network) or in-home (50m radius). MVPA was assessed using Evenson cut points, and home equipment (sum items of sports equipment) was self-reported. Three-level mixed effects linear regression (days, participants, neighborhoods) was used; Bs are MVPA minutes/day.

Results: Participants had 5.7±3.9 hours/day of wear time and 5.5±1.8 minutes/day of MVPA in their home. In-neighborhood (out of home) wear time was 21±4 minutes/day and in-neighborhood MVPA was 1.9±2.2 minutes/day. Total MVPA was 30.2±1.3 minutes/day. Walkability was not associated with total or in-neighborhood MVPA until adjusting for time spent in-home and in neighborhood(Bs=0.9). Home equipment was associated with total and in-home MVPA (Bs=1.6; 2.2). Adjusting for time spent in home strengthened the association between equipment and total MVPA (B=1.9) and attenuated the association between equipment and in-home MVPA (B=1.5).

Conclusions: Assessing physical activity and time spent in specific locations with GPS provides improved specificity for investigating relations of environments to physical activity. Proportional effect sizes were much stronger for location-specific than total MVPA. Assessing environmental exposure in multiple locations should lead to an improved understanding of environmental influences on total physical activity.