Abstract
BACKGROUND: Several narrative reviews have been conducted on the literature examining environmental correlates of physical activity (PA). To date these reviews have been unable to provide definitive summaries of observed associations. This study utilizes meta-analytical techniques to calculate summaries of associations between selected environmental characteristics and PA. METHODS: Published studies were identified from electronic databases and searches of personal files. Studies were examined to determine the environmental constructs most frequently studied. Included studies (N = 16) examined at least one identified construct and determined associations between perceived environmental constructs and PA using logistic regression. Data were analyzed separately for crude and adjusted ORs using general-variance based fixed effect models. RESULTS: No significant associations emerged between environmental characteristics and PA using crude OR. The perceived presence of PA facilities (OR 1.20, 95% 1.06-1.34), sidewalks (OR 1.23, 95% 1.13-1.32), shops and services (OR 1.30, 95% 1.14-1.46) and perceiving traffic not to be a problem (OR 1.22, 95% 1.08-1.37) were positively associated with activity using adjusted ORs. Variance in PA accounted for by significant associations ranged from 4% (heavy traffic not a problem) to 7% (presence of shops and services). CONCLUSION: Results of the meta-analysis support the relevance of perceived environmental characteristics for understanding population PA. These results should encourage the use of comprehensive ecological models that incorporate variables beyond basic demographic information.