BACKGROUND
Prior research has identified micro-level features of the built environment, such as mixed land uses and connected multi-modal transportation infrastructure, which may promote activity-friendly environments. However, there is a paucity of research examining the role of macro-level policies, such as urban containment policies, that may facilitate development patterns supportive of these attributes. Adopted at the state, metropolitan, county, or municipal levels, urban containment policies attempt to manage the location and timing of urban growth to support goals such as compact development, preservation of open space, promotion of social equity, and efficient use of infrastructure. Increasingly popular in the U.S., containment policies include a variety of tools such as growth boundaries, infrastructure service areas, and the delineation of greenbelts. Despite previous studies evaluating the effects of urban containment policies on urban density, housing, and transportation outcomes, the implications of these policies on physical activity remain unexplored.

OBJECTIVE
This study longitudinally examines the relationship between metropolitan urban containment policies, state adoption of growth management legislation, transportation indicators, and physical activity in 63 large U.S. metropolitan statistical areas (MSAs) from 1990-2002.

METHODS
Multiple data sources were combined, including previously published surveys and studies of predominant urban containment policy frameworks in the U.S., self-reported physical activity data from the Behavioral Risk Factor Surveillance System, the U.S. Census, the National Resources Inventory, and the Texas Transportation Institute Urban Mobility database. Urban containment policies were defined as the presence of a formally adopted urban growth boundary, urban service limit, or greenbelt in one or more jurisdictions within the MSA. Policies were further classified as weak or strong according to criteria published by the American Planning Association. Mixed models were used to examine whether specific types of urban containment policies and state-level growth management legislation were associated with a) leisure-time physical activity at the MSA-level, and b) walking/bicycling to work; while controlling for MSA sociodemographic and transportation factors over time.

RESULTS
Stronger urban containment policies and state growth management programs (e.g., those mandating urban growth boundaries) were associated with significantly higher prevalences of leisure-time physical activity and active commuting from 1990-2002. Sociodemographic factors including the proportion of African American residents, the proportion of the population with less than a high school education, and the proportion age 65 years and over were consistently associated with lower leisure-time physical activity prevalences. After adjustment for covariates, residents of MSAs with stronger policy frameworks averaged 27-43 additional minutes of leisure-time physical activity per week compared to residents of MSAs without policies during the study period (p<0.05). Additionally, the proportion of residents reporting no leisure-time physical activity was lower in MSAs with containment policies that included a regional planning component compared to MSAs without such policies. Analyses suggested that a substantial proportion of the variance in leisure-time physical activity (44%-49%) and active commuting (82%) was due to variation between MSAs. Models including policy variables, transportation indicators, and sociodemographic factors explained between 60% and 88% of the between-MSA variance in physical activity outcomes over time.

CONCLUSION
This study provides preliminary empirical evidence regarding relationships between urban containment policies and physical activity, and underscores the importance of considering the role of local, regional, and state-level policies in national physical activity trends. These findings suggest that areas which have adopted policies supportive of active lifestyles have maintained higher population prevalences of both leisure-time physical activity and active commuting over the past decade.