Modeling individual and physical environmental factors with adolescent physical activity

Background: Recent calls have been made to consider both environmental factors and individual-level factors in the explanation of physical activity (PA). The present study tested a conceptual model that integrated past PA, relevant environmental-level and individual-level factors and their associations with adolescent PA, using the tenets of the Theory of Planned Behavior (TPB). Methods: Data were gathered in an adolescent sample (n=221; 60% girls) using questionnaires. PA was assessed in May 2003 and November 2003. Theory, of Planned Behavior cognitions regarding PA and the environmental factors under study were assessed in November 2003. Confirmatory factor analyses and path analyses were performed using AMOS software. Results: The initial structural model did not provide an acceptable fit to the data. Including a direct path from past PA to current PA significantly improved model fit to an acceptable fit. Including a direct path from past PA to environmental perceptions did not significantly improve model fit. Including a direct path from the environmental variables to current PA did not significantly improve model fit. Current PA was most strongly influenced by past PA, while environmental aesthetics and distance to PA opportunities were indirectly related to adolescents' intention to be physically active. Significant standardized path coefficients ranged from 0.14 to 0.34 and explained 17% variance in current PA. Conclusions: Combining past PA, Theory of Planned Behavior cognitions, and environmental factors increased our understanding of their relative influences on adolescent PA. Implications for future research on physical activity are discussed.