Abstract
BACKGROUND: Limited research has been conducted on gender differences in perceived environmental correlates of physical activity (PA). The purpose of this study was to explore the potential role of gender in the link between perceived environment and PA. METHODS: Using a telephone-administered survey, data was collected on leisure time physical activity (LTPA), perceptions of the neighbourhood environment, and self-efficacy in a representative sample of 1209 adults from the province of Alberta, Canada. LTPA was regressed on ten measures of perceived neighbourhood environment and self-efficacy in a series of logistic regressions. RESULTS: Women were more likely than men to perceive their neighbourhood as unsafe to go for walks at night (chi2 = 67.46, p < 0.001) and to report seeing people being active in their neighbourhood (chi2 = 6.73, p < 0.01). Conversely, women were less likely to perceive easy access to places for PA (chi2 = 11.50, p < 0.01) and availability of places to buy things within easy walking distance from home (chi2 = 4.30, p < 0.05). Adjusting for age, education, income, and place of residence, access to places for PA (OR = 2.49) and interesting things to look at in the neighbourhood (OR = 1.94), were associated with higher levels of LTPA in men. Access to places for PA (OR = 2.63) and reporting seeing people being active (OR = 1.50) were associated with increased LTPA among women. After controlling for sociodemographic variables and self-efficacy, the presence of shops and places to buy things within easy walking distance from home (OR = 1.73), interesting things to look at in the neighbourhood (OR = 1.65), and access to places for PA (OR = 1.82) were associated with higher levels of LTPA in men. Among women, no significant relationships were observed between perceived environment and LTPA after adjusting for self-efficacy. CONCLUSION: The results provide additional support for the use of models in which gender is treated as a potential moderator of the link between the perceived environment and PA. Further, the results suggest the possibility of differential interventions to increase PA based on factors associated with gender.