Safety in numbers: does perceived safety mediate associations between the neighborhood social environment and physical activity among women living in disadvantaged neighborhoods?

OBJECTIVE The aim of this study is to examine associations between the neighborhood social environment and leisure-time physical activity (LTPA)(1) and walking among women, and whether these associations are mediated by perceived personal safety. METHODS Women (n = 3784) living in disadvantaged urban and rural neighborhoods within Victoria, Australia completed a self-administered survey on five social environment variables (neighborhood crime, neighborhood violence, seeing others walking and exercising in the neighborhood, social trust/cohesion), perceived personal safety, and their physical activity in 2007/8. Linear regression analyses examined associations between social environment variables and LTPA and walking. Potential mediating pathways were assessed using the product-of-coefficients test. Moderated mediation by urban/rural residence was examined. RESULTS Each social environment variable was positively associated with engaging in at least 150 min/week of LTPA (OR = 1.16 to 1.56). Only two social environment variables, seeing others walking (OR = 1.45) and exercising (OR = 1.31), were associated with ≥ 150 min/week of walking. Perceived personal safety mediated all associations. Stronger mediation was found in urban areas for crime, violence and social trust/cohesion. CONCLUSION The neighborhood social environment is an important influence on physical activity among women living in disadvantaged areas. Feelings of personal safety should not be included in composite or aggregate scores relating to the social environment.