Regular physical activity has been associated with prevention of overweight and obesity, 1 as well as numerous other diseases 2. Epidemiological studies have linked area of residence to physical activity 345. Given consistently low rates of physical activity in the US, 2 it is important to develop an understanding of neighborhood factors that influence physical activity.

Residence in areas with high levels of material deprivation, an indicator of low area socioeconomic status (e.g., low median household income, and low educational attainment) has been associated with low levels of physical activity 3567 as well as poor overall health 58 that may indicate low levels of physical activity. Despite the consistency of these findings, it is difficult to pinpoint the physical characteristics of material deprivation that contribute to physical inactivity. Deprived neighborhoods often have a lower proportion of owned homes compared to non-deprived neighborhoods, resulting in a low tax base for municipal improvements, have higher crime rates and reduced collective efficacy, 5910 and have fewer goods and services available 711. Although studies in two countries (UK, USA) have found that physical activity resources 1213 vary by the socioeconomic status of the neighborhood, with lower SES areas having fewer physical activity resources, another study in Australia found greater access to physical activity resources in more deprived areas 14. Although lack of access may be a driving factor to lower rates of physical activity in some deprived areas, there are likely additional qualitative elements of the physical activity resources that have not been well described or documented. For example, one study reported that attractiveness of public open space was associated with higher rates of walking 15. However, there remain many elements of physical activity resources that have yet to be described or investigated in detail.

Although research in this area has been limited by access to geocoded databases to provide environmental information, the widespread adoption of Internet-based "user friendly" formats has resulted in a proliferation of research. Wider availability of these kinds of data have produced several studies investigating urban development constructs including walkability, a measure of how convenient and pleasant a neighborhood is for walking, 1617 along with other street-scale elements, such as protected pedestrian pathways and availability of goods and services 18.

Despite these innovations, there remains no widely accepted system or protocol for describing or evaluating physical activity resources. The potential for endless permutations in type, equipment, size, shape, and condition of resources has hampered development of a common protocol. Further, until recently, little interest had been devoted to investigating the role that the physical environment plays in physical activity 19. Recognition of physical activity as a universal good health recommendation, associated with the prevention of many chronic diseases, and success of policy and environmental approaches to manage health behaviors in other domains (e.g., tobacco control) has led researchers and funding agencies alike to pursue environmental approaches 19.

Available evidence suggests that there is a link between neighborhood factors and physical activity. Before this link can be clearly defined, it is important to develop strategies to assess neighborhood factors that may influence physical activity. The goals of this study were threefold. First, we aimed to develop an assessment protocol to describe physical activity resources based on existing literature and extensive pilot testing. Second, we assessed the type, quantity, features, amenities, and quality of all publicly accessible physical activity resources in urban neighborhoods available to public housing residents. Last, we compared the resources in the public housing neighborhoods to less deprived, comparison neighborhoods matched on age of housing and urban design (as measured by street network connectivity) and population density.

This study (1) developed the PARA instrument that provides a brief, reliable and effective strategy to objectively assess neighborhood factors that may influence physical activity by describing the type, quantity, features, amenities and incivilities of physical activity resources, and (2) used the PARA to evaluate the physical activity resources in thirteen lower income, high ethnic concentration neighborhoods that surrounded public housing developments in comparison to physical activity resources in higher income, low ethnic minority concentration neighborhoods.

Lower income, higher ethnic concentration, housing development neighborhoods varied widely in the number and type of resources that were available for physical activity. All but one of the neighborhoods had access to parks, and most had accessible public school yards and community centers. The high number of parks found in housing development neighborhoods likely reflects city planners and landscape architects from over a century ago, such as Frederick Law Olmstead 30 who advocated for park development for the civilizing influence of "neighborly recreation" and its benefits to human health. As was posited then, more recent data suggest that people who live near attractive, public open spaces may be almost twice more likely to walk at moderately active levels than were those who do not have access to public open spaces 31. Higher income, low ethnic concentration, comparison neighborhoods also varied widely in the number and type of physical activity resources available. The neighborhoods were selected to be similar in age and urban design; thus, the housing development neighborhoods and comparison neighborhoods might have similar amounts of parks, schoolyards and other public structures resulting from similar urban planning strategies. Although efforts were made to select appropriate comparison neighborhoods for the purposes of the current study, these neighborhoods do not represent the entire universe of possible neighborhoods.

Although the net number of physical activity resources did not vary by neighborhood income and ethnic concentration, the overall environment of physical activity resources was strikingly different in the neighborhoods, and suggests that evaluating merely the presence or absence of physical activity resources may be an overly simplistic way to investigate access to resources. In this sample, incivilities were consistently present and conspicuously bad and offensive at physical activity resources in lower income, higher ethnic concentration neighborhoods. In comparison, we found no incivilities at the resources in half of the comparison neighborhoods, and very few incivilities at the resources of the other half. It would seem to make sense that the presence of litter and debris, as well as lack of apparent maintenance might be important detractors from using a physical activity resource for physical activity, because people who walk more frequently typically rate their environment more positively 31. In fact, a high proportion of incivilities might suggest lack of attention to an area, and might even encourage less desirable behavior (e.g., drug trafficking, prostitution) clearly not promoting favorable conditions for recreational physical activity 25.

Perhaps most distressing about these findings is the suggestion that the enduring relationship between social inequalities and poor health in an ostensibly wealth country, the US, is complex and appalling. The results suggest that merely building a park in a deprived area may be insufficient for insuring its intended use and maintenance. It is critical to provide ongoing support for maintenance and civic improvements. There is a deep need for policy makers and political leaders to work with communities to improve the quality of publicly available physical activity resources to improve the quality of life for all.

Previous work suggests that convenient neighborhood resources are critical for meeting physical activity recommendations 3132. An important contribution of this study is the inclusion of quality and incivility ratings in our instrument. To date, investigations that have examined physical activity resources have been limited to investigating the type and accessibility of resources 1325. This study is among the first to comprehensively evaluate a broad range of physical activity resources on a number of dimensions. At the same time, the PARA instrument was developed to be easily and rapidly administered, promoting broad scale dissemination and relatively, easily translatable outcomes for policy implementation. Future research is needed that tests the associations among these physical activity resource dimensions and individual physical activity levels to guide individual and environmental intervention development.

REL conceived of the study, and directed all aspects of the study including development, assessment and analyses, and lead the writing of the manuscript. KMB, JRS, GR all helped to develop measures, conduct assessment and analyses and assisted with the writing of the manuscript. HH provided geographic support including neighborhood maps, location of resources and assistance with variable construction, as well as assisting with the writing of the manuscript. All authors read and approved the final manuscript.