Associations between individual- and neighborhood-level characteristics and physical inactivity among adults in Los Angeles County

From an epidemiologic perspective, physical inactivity is a strong risk factor for a variety of chronic conditions. Its pervasiveness in populations makes it one of the most prevalent modifiable risk factors for poor health. The purpose of this study was to examine individual-level and neighborhood-level predictors of physical inactivity. Using cross-sectional data on adults 18-65 years old from the 2002-2003 and 2005 Los Angeles County Health Surveys, this study first assessed the racial/ethnic disparities in the prevalence of physical inactivity. Asians, recent U.S. immigrants in particular, had the highest levels of physical inactivity compared to all other racial/ethnic groups. Integrating these data with neighborhood-level data, matched by census tract of residence, from the 2000 Census and other sources, neighborhood-level characteristics of the built and social environments were then assessed to estimate their independent effects on physical inactivity. There was considerable disparity in the exposure to neighborhood characteristics by race/ethnicity. Among the built and social environment characteristics studied, intersection density and residential density were weakly associated with physical inactivity in weighted logistic regression models, controlling for all other individual-level predictors of physical inactivity. For females, low intersection and residential density were positively associated with physical inactivity. However, the reverse was seen for males. Individual-level perceptions of the social environment, safety and belonging, were also examined. Low safety and low belonging were both positively associated with physical inactivity for both genders. Race/ethnicity and acculturation did not modify any of these observed associations.; Because it is a socially desirable behavior, recall is difficult, and there are variations over both short and long periods of time, measuring levels of physical activity through questionnaire carries a high risk of misclassification. Sensitivity analysis showed that even small levels of misclassification result in biased associations between variables. Without further knowledge of the validity of the physical inactivity assessment tool, it is difficult to draw strong conclusions based on the observed associations. Further studies are needed to better describe the causal nature of the hypothesized predictors of physical inactivity.