Multilevel analysis of physical activity among us adults across US Census regions: The role of environmental contexts.

摘要

This study uses nationally representative data to explain the variation in physical activity among U.S. adults across four geographic regions of Northeast, Midwest, South, and West. The purpose of the study is to investigate whether environmental conditions inclusive of natural amenity, built environment, urbanization, crime, and social economic neighborhood are associated with physical activity and whether these associations are modified by geographic region. The study also investigates whether the influences of environmental contexts vary by subgroup and whether they are modified by geographic region. Determining regional factors that influence physical activity is essential to planning appropriate physical activity interventions.;The dependent variable is adult physical activity (PA). The measure of PA is dichotomized as either meeting current physical activity recommendations or not. A host of independent variables are tested to identify empirical relationships with the dependent variable. Multilevel logistic regression is used to examine the environmental factors and their cross-level interaction effects that explain variation in PA across the four regions.;In Northeast, built environments and urban status are the environmental contexts that associate with PA. There are no significant predictors associated with PA in the Midwest. Three significant PA predictors in the South include built environments, natural amenities, and socio economic environments. Significant predictors in the West include urban status, built environments, and socio economic environments.;The study has five major contributions. Firstly, the study finds that living in a more built environment gives a consistent advantage for whites across the regions, whereas no advantage of built environment appears among minorities. Secondly, the study finds that age is a moderator between environments and PA. Older people are consistently sensitive to immediate environments. Thirdly, unexpected, the study finds that some minorities such as Asians in high SES counties are more likely to be physically inactive relative to Asians in low SES counties. Fourthly, the study demonstrates that Blacks in non-metro South and West are faced with physically inactive. Finally, this study finds that some sub-population such as Blacks in Northeast residing in high crime areas are more physically active than those living in less crime areas.