Given the emphasis and emerging importance of the built environment on influencing physical activity [10–17], this study provides important insights for informing development of the CBPR walking intervention and informing local recreational and environmental policies in the Hattiesburg region. In this study, we found that proximity to trails as well as higher perceptions of pedestrian safety, trail safety, and trail amenities predicted frequency of trail use. Further, we found that weekly trail use positively predicted the odds of meeting walking recommendations, physical activity recommendation, and total MET/minutes per week. Our findings support previous findings that indicate a relationship between trail proximity, frequency of trail use, and physical activity [48–50].
We did not find a relationship between objective audits and meeting walking or PA recommendations, nor did we find that perceptions of the environment predicted meeting walking or PA recommendations. While previous studies have found that perceived and objective environmental measures relate to physical activity differently [51–53], our study provides no evidence that either was a good predictor of physical activity. Despite the increased focus on the physical environment in recent years, relatively few studies have used objective, in-person audits to understand the environmental influence on meeting physical activity recommendations [27, 54, 55]. For examples, Zenk and colleagues found no association between objectively measured environmental variables and adherence to a 12-month walking intervention . However, Lee and colleagues found that lower speed limits were most commonly associated with increased physical activity among both women and men . Yet, contrary to hypotheses, greater segment connectivity was not associated with more physical activity . To add to the list of inconclusive research findings, evidence regarding the role of perceived safety, maintenance and aesthetics in relation to physical activity outcomes is mixed [18, 47, 49, 50, 56]. Taken as a whole, our results combined with other studies highlight the need to further investigate the role and importance of perceived and objective measures to inform intervention development related to physical activity promotion.
The objective PEAT audits indicated numerous high quality walking trails in Hattiesburg, Mississippi. The number, excellence, and proximity of walking trails to residents' homes are conflicting to several studies that indicate a shortage of physical activity resources (e.g. parks and walking trails) in low income communities [22–24]. Rather, our study parallels those by Abercrombie and colleagues, which did not find deprivation of recreation facilities among low-income and high-minority neighborhoods in Maryland . In our study, the availability and objective quality of trails rated generally high; however, evidenced by findings that the average use was 1.5 (S.D = 1.6) times per week and about 44% of respondents used the trails less than 0.5 times/week, overall trail use was relatively low. This is particularly noteworthy given that walking was the most preferred mode of physical activity for 61% of respondents, and that fewer respondents met PA recommendations by walking (35%) as compared to meeting PA recommendations (57%). Since 'meeting PA recommendations by walking' is defined in this study as walking 5 or more days a week for at least 30 minutes [38, 39], it is conceivable that many of those indicating walking as a preferred activity walked less than the defined amount. Further, some may have chosen other walking locations besides trails for convenience (e.g. neighborhood streets), for protection from the weather (e.g. indoor facilities like a mall), or for other reasons that were not explored in this study.
Unlike prior findings, we did not find that trail use varied by demographic characteristics such as gender, age, race, or socioeconomic level . Importantly, our study provides benchmark data related to health and physical activity patterns as well as baseline data for tracking future environmental changes. Based on self-reported height and weight, 70% of respondents were classified as overweight and obese. These rates far exceed national averages and HP 2020 goals, indicating that efforts to promote walking and other physical activity behaviors remain a high priority in Hattiesburg. Although reporting on quality indicators of individual trails was beyond the scope of this paper, individual trail characteristics also provide important data for specific improvements that are needed at each walking trail.
The evidence-based Guide to Community Preventive Services recommends four key strategies to promote physical activity in adults including enhanced access for physical activity along with informational outreach, community-wide campaigns, social-support interventions in community settings, and individually adapted health behavior change interventions [57, 58]. In a simulated cost-effectiveness study, each of these four strategies offered substantial gains in quality-adjusted life years or good value for money, and no one strategy appeared more cost-effective than the others . Our study provides critical formative data for program planning and implementation. For example, findings indicated that less investment is needed to build new walking trails or improve existing trails, and more efforts should be focused on maximizing the use of existing trails through community marketing campaigns and social support programming to decrease trail use barriers.
In the context of CBPR, key strategies are to build collaborative community-academic partnerships, develop a community's research skills, and promote data sharing and local dissemination of findings. This study signifies engagement of the community in an influential phase of research. Most of the community members hired and trained through this project have maintained leadership roles as peer-support coaches and assistance coaches in H.U.B. City Steps walking intervention [28, 29]. Furthermore, to highlight and promote the characteristics, location, and amenities of each trail, results of the PEAT assessment were used to help develop a community resource guide for distribution via hard copy brochure and the H.U.B. City Steps website. Through on-going dissemination efforts, findings are being promoted through a written policy brief, and presentation to local government and civic groups. Within the CBPR framework, these lay methods of data dissemination are equally as important as hypothesis-testing and data-driven scientific dissemination. Creating a sense of community power and promoting community ownership of the individual, social, and environmental health problems and solutions will be key to improving the modifiable CVD risk factors and the long-term health of residents of Hattiesburg, Mississippi. Importantly, a recent paper that highlights opportunities to address disparities pertaining to recreation environments recommends the CBPR process to engage communities of color in research .
A few limitations should be considered when interpreting the findings of this study. First, a random sample promotes the greatest generalizability of study findings, yet was beyond the scope of this project. While use of the proportional sampling plan to match the targeted enrollment matrix for the larger CBPR walking intervention assures that our findings are sufficiently applicable to study population and intended use of data, the participants surveyed were much more active (i.e. 57% meeting physical activity recommendations) than what is typically reported by surveillance data (i.e. 37.5% of Mississippians meeting physical activity recommendations according to 2009 Behavioral Risk Factor Surveillance System data) . It is not entirely clear if participants who met physical activity recommendations are overrepresented in our sample, as it was not our intent to recruit physically active participants or trail users, or if this discrepancy is a function of differences in methodology for reporting physical activity behaviors. While the validated IPAQ is one of the most frequently used estimates of self-reported physical activity , there are concerns for overestimation of physical activity with this instrument [42, 61, 62]. Regardless, the representativeness of our study population should be considered when interpreting the findings. Finally, scoring individual items into component scales and statistical modeling for items in the PEAT audit was challenging. While the PEAT proved to be a user friendly audit instrument and an appropriate tool for the broader objectives of this CBPR study[44, 45], no known published papers have reported or modeled summary scales. Nonetheless, our methodological approach of creating standardized mean scores across four content areas is congruent with other standard approaches of modeling community audits, such as the Physical Activity Resource Assessment (PARA) instrument .
Future research efforts should address the limitations of this study. For examples, an objective indicator of physical activity, such as use of accelerometers, would provide a more accurate indication of physical activity patterns . Second, future research should capitalize on understanding the broader built environment. Objective measurements of the built environment have made great advancements in recent years and gold standard measures are still emerging [11, 16]. Of particular interest in Hattiesburg is to better understand the walkability and general environment for lifestyle physical activity, beyond just walking trails. For instance, GIS-derived measures could help better understand access to walking trail and recreational facilities, as well as street patterns, land-use mix, and population density . While there is certainly room to utilize more sophisticated methods to advance the understanding of the physical activity patterns and environment in Hattiesburg, this study sufficiently informs intervention and policy development. Furthermore this study begins to fill an important void in the current literature as few identified papers have examined the built environment for physical activity in Mississippi , despite the fact that it consistently ranks among the least healthy states .
In conclusion, the modifiable risk factors related to CVD place a significant health burden on Mississippi's residents and health care systems. In addition to benefits for managing CVD risk factors, physical activity provides a constellation of health benefits including weight loss/maintenance, reduced cholesterol, stress management, and psychological well-being.
CBPR studies such as this one that engages communities and considers multiple levels of influence on physical activity are critical to identifying cost-effective and culturally relevant intervention approaches that have the potential for long-term sustainability, as well as the potential to inform local policies. Contrary to assumptions, numerous high quality walking trails/tracks were available within four city wards of a community with high minority population and low median income. In order to improve physical activity levels and decrease CVD risk factors, continued intervention efforts are needed to promote and maximize the use of existing trails, as well as improve residents' perceptions related to incivilities, safety, conditions of trail, and amenities of the walking trails. Furthermore, to best guide future programming efforts in Hattiesburg and similar communities, our findings should be mutually interpreted within the context of evidence-based physical activity recommendations and alongside approaches that have demonstrated cost-effectiveness potential for physical activity promotion [21, 49, 58, 59].