BACKGROUND
Obesity has become an epidemic in the United States, and it is poised to become the nation’s leading health problem. Individuals of varied ages, sex and all ethnic groups are affected, but unequally. Although in the United States, sixty-five percent of the population is either overweight or obese, the prevalence of adult overweight and obesity is higher among Hispanics and African-Americans than among non-Hispanic Whites. Women are more likely to be overweight and obese than men among Hispanics, Whites, and especially African Americans. Rates of overweight are higher for adults of low socioeconomic status. Men with incomes below poverty are slightly less likely than men with higher incomes to be overweight, and slightly more likely to be obese. Women living below the poverty level, on the other hand, are much more likely to be overweight or obese than women with higher incomes.

While there are proposed biological and social pathways that contribute to obesity, its manifestation results from a chronic imbalance between energy intake and energy expenditure. Environmental factors by which obesity results include increased availability of high caloric food or the decreased physical activity in day-to-day life.  Spatial separation of population groups along racial/ethnic and socioeconomic lines is a central feature of the social organization of U.S. urban areas. While segregation is comprised of multifaceted dimensions, different aspects of its measure may be more relevant to understanding obesity as a health outcome. Access to open space for physical activity and healthy food comprise two major pathways which link dimensions of the built environment to obesity. Neighborhood-level socioeconomic status and residential stability are features related to access to services within neighborhoods. Limited research has been able to combine neighborhood level variables of the built and social environment, especially considering both diet (i.e. the food environment) and physical activity (i.e. green space environment) concurrently, alongside of individual-level health behavioral and obesity (BMI) data.

OBJECTIVE
We sought to examine the built environment, specifically food purchasing venues, green space location and density, neighborhood socio-demographic characteristics and body mass index in Pittsburgh, Pennsylvania.

METHODS
Food purchasing venues, green space data and socioeconomic and racial/ethnic composition of neighborhoods were attained respectively from the Allegheny County Health Department, Allegheny County Parks Department, and the United States Census 2000. These tract-level data from Pittsburgh, Pennsylvania were merged with Allegheny County Behavioral Risk Surveillance Survey (BRFSS) body mass index (BMI) data specific to Pittsburgh. We used Geographic Information Systems (GIS) for management and visual display of features and attributes of places. We performed an ecological multivariate analysis to examine associations between the type and density of food purchasing and green space venues, alongside of neighborhood characteristics and relationships between the built environment and obesity.

RESULTS
Of 90 municipally defined neighborhoods (corresponding with census-defined tracts) in the City of Pittsburgh, 1335 individuals were sampled through the BRFSS. There were 41 neighborhoods where at least half of the individuals sampled within the neighborhood were overweight or obese, 13 neighborhoods where at least half of the individuals sampled were obese and 3 neighborhoods where 75% or more of the individuals sampled were obese. These neighborhoods differed significantly in acres of green space, socioeconomic and racial composition characteristics. Total park acres showed averages of 24.3, 18.4 and .84 acres respectively, while proportion of the neighborhood population with household incomes under the poverty line was .19, .26 and .35 respectively. Finally, the proportion of blacks within the neighborhood varied from .32, .47 and .61 respectively. Similarly, types and quality of food purchasing venues within neighborhoods differed systematically across neighborhoods. However, analyses showing average BMI of individuals across neighborhoods did not show significant associations with respect to the proportion of the population under poverty, racial composition and green space acreage.

CONCLUSIONS
Our results demonstrate a picture of a complex associations between characteristics of the built and social environment and prevalence of obesity in the City of Pittsburgh. These preliminary analyses show differences in prevalence of overweight and obesity, neighborhood characteristics and built environmental characteristics across Pittsburgh, yet also present a picture of how additional features of the built environment might contribute toward both physical activity and food purchasing of the neighborhood population. A visual understanding of the city design, its amenities and its population can assist policymakers on the city level incorporate programs and policies toward a healthier Pittsburgh. Our merged national (Reference USA, Census 2000) City (Department of Parks), and local (BRFSS) data demonstrate how local built and social environments play out with respect to health outcomes on a national level.