Next Article in Journal
Athletes Drive Distinctive Trends of COVID-19 Infection in a College Campus Environment
Next Article in Special Issue
¿Qué Pasa Con Papá? Exploring Paternal Responsibilities and Physical Activity in Mexican-Heritage Families
Previous Article in Journal
Significant Decrease in Heavy Metals in Surface Sediment after Ten-Year Sustainable Development in Huaxi Reservoir Located in Guiyang, Southwestern China
Previous Article in Special Issue
Physical Activity Barriers and Assets in Rural Appalachian Kentucky: A Mixed-Methods Study
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Environmental Factors Associated with Physical Activity in Rural U.S. Counties

1
Department of Social and Behavioral Sciences, School of Public Health, West Virginia University, Morgantown, WV 26506, USA
2
WVU Office of Health Affairs, Robert C. Byrd Health Sciences Center, West Virginia University, Morgantown, WV 26505, USA
3
Department of Public Health, Robbins College of Health and Human Sciences, Baylor University, Waco, TX 76798, USA
4
Department of Parks, Recreation and Tourism Management, College of Natural Resources, North Carolina State University, Raleigh, NC 27695, USA
5
School of Kinesiology and Physical Education, College of Education and Human Development, University of Maine, Orono, ME 04469, USA
6
School of Nursing, Oregon Health Science University, Portland, OR 97239, USA
7
US Department of Agriculture, Agricultural Research Service, Grand Forks, ND 58201, USA
*
Author to whom correspondence should be addressed.
Present Address: Texas AM AgriLife Extension Service, College Station, TX 77843, USA.
Int. J. Environ. Res. Public Health 2021, 18(14), 7688; https://0-doi-org.brum.beds.ac.uk/10.3390/ijerph18147688
Submission received: 30 June 2021 / Revised: 15 July 2021 / Accepted: 16 July 2021 / Published: 20 July 2021

Abstract

:
Background: Rural U.S. adults’ prevalence of meeting physical activity (PA) guidelines is lower than urban adults, yet rural-urban differences in environmental influences of adults’ PA are largely unknown. The study’s objective was to identify rural-urban variations in environmental factors associated with the prevalence of adults meeting PA guidelines. Methods: County-level data for non-frontier counties (n = 2697) were used. A five-category rurality variable was created using the percentage of a county’s population living in a rural area. Factor scores from Factor Analyses (FA) were used in subsequent Multiple Linear Regression (MLR) analyses stratified by rurality to identify associations between environmental factor scores and the prevalence of males and females meeting PA guidelines. Results: FA revealed a 13-variable, four-factor structure of natural, social, recreation, and transportation environments. MLR revealed that natural, social, and recreation environments were associated with PA for males and females, with variation by sex for social environment. The natural environment was associated with PA in all but urban counties; the recreation environment was associated with PA in the urban counties and the two most rural counties. Conclusions: Variations across the rural-urban continuum in environmental factors associated with adults’ PA, highlight the uniqueness of rural PA and the need to further study what succeeds in creating active rural places.

1. Introduction

Rural U.S. communities experience a multitude of impactful health disparities [1], including a lower prevalence of engaging in preventive health behaviors such as physical activity (PA) [2]. Specifically, the percentage of adults meeting combined aerobic and muscle-strengthening PA guidelines is 31% lower among rural residents compared to urban residents [3]. This deficit exists despite evidence of greater relative increases of meeting PA guidelines from 2008 to 2017 among rural adults (47.4%) compared to urban adults (30.4%) overall. Analyses by Whitfield and colleagues also revealed greater relative increases in rural adults across multiple sociodemographic categories, including males and females, all age groups except 25–34 years, all Census regions except the South, those with a high school or some college educational attainment, and in Black and White non-Hispanic adults [3]. While these relative improvements are promising, the prevalence of meeting the PA guidelines is still greater among urban adults than rural adults of all races and ethnicities, for both sexes, and across all regions and education levels.
Two decades of PA research has accumulated evidence that multiple “environments” influence PA based on a social-ecological perspective [4], emphasizing the primacy of the physical, or built, environment [5] in the multiple places that individuals interact with daily (e.g., work, school, home), and the transportation system that connects those places. As illustrated by Sallis and colleagues [6] built environments are influenced by political, natural, and social/cultural environments. Collectively, these environments influence—and are influenced by—individual decisions in a cycle of reciprocal determinism [7] or mutual influence [4]. Much of what is known about the environmental influences on PA is based on studies of urban and suburban built environments that may have little applicability to rural places. Recent literature reviews consistently highlight this gap, identifying the near-exclusive reliance on urban-derived evidence to inform practices in rural communities and making strong recommendations for research specific to rural communities, given the health and PA disparities observed in rural communities. Despite these recommendations, there has yet to be an empirical analysis of objective measures on a national level examining the rural-urban variation of environmental associations with PA [8,9,10].
Rural-urban comparisons are important so that appropriate interventions can be developed and undertaken. Recent research highlights some of the rural-urban differences in walking behaviors and perceived walkability [3,11,12]. For example, Whitfield and colleagues [11] found that walking is associated with the number of environmental supports (e.g., sidewalks) and destinations in both urban and rural areas of the U.S., but the most influential types of supports and destinations may differ between urban and rural communities. Additionally, even if the associations between walkability supports and behaviors are similar across rural and urban communities, the implementation of supports for walking may face unique challenges in rural communities (e.g., longer block lengths, lower development density, faster-moving traffic, the availability of ample parking) [13]. While access to natural amenities that facilitate PA like walking or biking (e.g., multiuse trails) may be associated with higher PA in rural communities [14,15], it is also possible that natural environments such as weather and climate may present meaningful barriers to PA among rural residents in comparison to urban residents [16]. Research further suggests that sociocultural factors (e.g., crime, safety, social support) may be associated with PA prevalence in rural communities, though few studies have compared the relationship between these factors and PA across rural and urban settings [10,17]. A childhood obesogenic environment index (COEI) inclusive of social-ecological variables hypothesized to be associated with sedentary, nutrition, and PA behaviors suggests that rural areas are more conducive to obesogenic behaviors of children than urban and micropolitan counties in all regions of the U.S. [18,19]. However, an analysis of national-level data to identify the social, natural, built, and economic environment factors associated with rural adults’ PA in the U.S. has yet to be undertaken.
The objective of this study was to answer the following question: how do environmental factors associated with the prevalence of adults meeting PA guidelines differ along the rural-urban continuum for males and females?

2. Materials and Methods

2.1. Procedures

Data from 3140 counties and county-equivalents (e.g., Baltimore City) (hereafter “counties”) were utilized for analyses. Rurality was defined as the county population living in rural areas according to 2010 Census data. Frontier counties (n = 443), defined as counties with a population density of six or fewer people per square mile [20], were excluded from analyses. The 2697 counties in the analytic sample were classified into five groups. The most rural group (“Exclusively Rural”) was defined as counties in which 100% of the population was living in rural areas. The remaining 2308 counties were then divided equally into quartiles based on the percentage of the population living in rural areas. The final sample resulted in 2697 observations from all 50 U.S. States, categorized as Exclusively Rural (n = 389), Rural (n = 577), Somewhat Rural (n = 577), Somewhat Urban (n = 577), and Urban (n = 577).

2.2. Data Sources

This study included a series of indicators from eight publicly available national data sources. For brevity, only a brief description for each data source has been provided, with URLs where additional methodologies, questionnaires, metrics, and data information may be found (see Table 1). Data collection years ranged from 2010 to 2011 to match the timeframe of the collection of the primary dependent variable (see Section 2.3.1).

2.2.1. The American Community Survey (ACS)

The ACS is a critical element in the Census Bureau’s reengineered decennial census program. The ACS collects and produces population and housing information every year and publishes both one-year and five-year estimates [21].

2.2.2. The Behavioral Risk Factor Surveillance (BRFSS)

The BRFSS is a state-based surveillance system that is operated by state health departments in collaboration with the Centers for Disease Control and Prevention (CDC). The BRFSS collects self-reported information by telephone on an assortment of behaviors and health outcomes among adults [22].

2.2.3. The FBI Uniform Crime Reporting (UCR) Program

The UCR Program consists of four data collections from roughly 17,000 law enforcement agencies in the U.S.: (1) the National Incident-Based Reporting System (NIBRS), (2) the Summary Reporting System (SRS), (3) the Law Enforcement Officers Killed and Assaulted (LEOKA) Program, and (4) the Hate Crime Statistics Program. The UCR Program publishes annual reports for each of these data collections, a preliminary semiannual report of summary data each winter, and statistical publications, such as the County-Level Detailed Arrest and Offense Data report that is utilized to create the County Health Rankings [23].

2.2.4. Fatality Analysis Reporting System (FARS)

The FARS is a census of fatal motor vehicle crashes with a set of data files documenting all qualifying fatalities that occurred within the 50 States, the District of Columbia, and Puerto Rico since 1975 [24].

2.2.5. National Environmental Public Health Tracking Network (EPH)

For more than a decade, the EPH Program has collected, integrated, and analyzed non-infectious disease and environmental data from a nationwide network of partners. The purpose of this program is to deliver information and data to protect the nation from health issues arising from or directly related to environmental factors [25].

2.2.6. 2010. Decennial United States Census

The decennial U.S. Census gathers data from all households in the United States and Island Areas related to name, gender, age, race, ethnicity, relationship, and home-ownership. The 2010 questionnaire was one of the shortest in history—asking just 10 questions—and is augmented by the U.S. Census Bureau’s annual ACS [21].

2.2.7. CDC Wide-Ranging Online Data for Epidemiologic Research System (CDC WONDER)

CDC WONDER manages nearly 20 collections of public-use data for U.S. births, deaths, cancer diagnoses, tuberculosis cases, vaccinations, environmental exposures, and population estimates, among other topics. These data collections are available as online databases, which provide public access to ad-hoc queries, summary statistics, maps, charts, and data extracts. Most of the data are updated annually; some collections are updated monthly or weekly [26].

2.3. Measures

Table 1 provides a description, data source and collection year, and definition of each county-level variable and PA prevalence.

2.3.1. Prevalence of Physical Activity

For the current study, the county-level prevalence of meeting the PA guidelines by males and females over 18 years old, assessed via the 2011 BRFSS, was the outcome of interest. Individual respondents were asked about the frequency and time of each type of PA that they engaged in outside of work in the prior month if they engaged in any activity. The relative intensity of each type of activity engaged in was classified as moderate or vigorous based on the age and sex of the respondent. Each respondent was classified as meeting PA guidelines, insufficiently active, or inactive based on their responses. At the time of data collection for the 2011 BRFSS, the U.S. PA guidelines were for adults to achieve 150 min of moderate intensity PA per week, the equivalent in vigorous PA, or a combination of moderate and vigorous PA where one minute of vigorous PA is equivalent to two minutes of moderate PA. At the county level, the prevalence estimates (valid percent) of adult females and males meeting the PA guideline were acquired from the Dwyer-Lindgren et al. online supplemental material [28]. Specifically, the dependent variables were the 2011 BRFSS estimates for females and males, predicted using a small area estimation model that included race/ethnicity, education, poverty, unemployment, air pollution, rural-urban status covariates, and a geospatial term. These data are the most recent county-level estimates of meeting PA guidelines available for the entire U.S. For further information on the small area estimation methodology see Dwyer-Lindgren et al. [28] and Srebotnjak, Mokdad, and Murray [29].

2.3.2. Environment Factors

The independent/predictor variables were categorized into natural, social, recreation, and transportation environments in alignment with Sallis and colleagues’ [6] framework. Some variables are measures of the built environment (e.g., park proximity), some are behavioral proxies for environmental measures (e.g., commuting by bike), and some are objective measures that may impact the perceived barriers to engaging in PA (e.g., crime, heat).

2.4. Statistical Analysis

All analyses were performed in SAS 9.4© (SAS Institute, Cary, NC, USA) [30] with missing observations handled using pairwise techniques (missing patterns 1%). Univariate and bivariate analyses were used to examine measures of central tendency and correlational associations between study variables. Analyses were conducted in two steps: (1) developing a parsimonious model, and (2) testing for associations of model factors and the dependent variable.
Factor Analyses (FA) were used to establish model parsimony, variable redundancy, and obtain weighted factor scores prior to regression analyses on the dependent variables. First, all variables were standardized into z-scores, reviewed for non-normality, and log-transformed (when appropriate) to ensure statistical assumptions for inferential analyses were met. FA criteria used squared multiple correlations (SMC) as priors for communality estimation, an orthogonal (varimax) rotation, and loadings ≥40 considered “large” for factor extraction. Lastly, estimated factor scores were assigned to each observation using the linear composite from the observed variables [31].
Multiple Linear Regression (MLR) analyses were used to identify associations between principal components (factor scores) and county-level prevalence of males and females meeting PA guidelines. MLR models were stratified by quintile and accounted for clustering effects by state. Standardized (β) beta coefficients and standard errors (SE) represent model association. Lastly, model assumptions as outlined by [32] were performed prior to MLR with alphas of ≤0.05 considered statistically significant.

3. Results

3.1. Descriptive County Characteristics

Univariate analyses are presented in Table 2. Across all counties, the average prevalence of meeting PA guidelines was 52.8% among males and 49.0% among females, ranging from 56.2% for male residents of urban counties to 46.8% among female residents of rural counties. Bivariate correlations for male and female PA for the indicators and each of the environment factors (natural, social, recreation, and transportation) are presented in Supplementary Tables S1–S4. Those results demonstrate that male and female PA correlate similarly with all 18 of the indicators in each of the four environment factors across the rural-urban continuum.

3.2. Factor Loadings and Model Parsimony

Factor loadings are reported in Table 3. FAs extracted a single factor for each environmental construct. Variable redundancy and loading criteria resulted in a reduction from 18 to 13 variables. The FA solution suggests factors comprised of natural environment (Air Temperature, Sun, and Heat Index), social environment (Single Parent Households, Violent Crime, and Vacant Housing), recreation environment (Proximity to a Park, Access to Exercise Opportunities, and Proximity to an Elementary School), and transportation environment (Walk to Work, Works in County, Works in Place, and Long Commute). Of note are the low factor loading (0.40) of Access to Exercise Opportunities and Bike to Work in the Somewhat Rural, Rural, and Exclusively Rural counties. For this reason, and to further support model parsimony, these variables were not included in the final model.

3.3. Associations between PA and Factor Scores by Sex

Results from the MLR analyses by sex are presented in Table 4. For males, the analyses reveal associations between natural environment (β = −32, p 0.01), social environment (β = −11, p 0.05), and recreation environment (β = 0.34, p 0.01). Female PA was associated with the natural environment (β = −31, p 0.01), social environment (β = −24, p 0.01), and recreation environment (β = 0.33, p 0.01). The transportation environment was not associated with PA for either sex (β = 0.01, p 0.05).

3.4. Associations between PA, and Factor Scores by Rurality and Sex

Results from the MLR analyses by rurality and sex are also presented in Table 4. Overall, the pattern of associations across the rural-urban continuum is similar between the sexes. Negative associations are seen for natural environment factors in all levels of rurality (β −36 to −46; all p 0.01) except the Urban counties. The recreation environment was associated with meeting PA guidelines in the Urban category of counties and the two most rural categories, with a greater difference in β between rural and urban counties among males than females. The negative association of the social environment and PA was significant among men except in the two most rural categories of counties; among females, it was also significant in the Exclusively Rural counties (β = −20, p 0.05). The transportation environment was associated with PA only in females in Somewhat Urban counties (β = 0.12, p 0.05).

4. Discussion

Findings from this study address a critical research gap highlighted in reviews of the literature [9,10] by identifying variations in the strength of environmental factors associated with meeting PA guidelines across the rural-urban continuum, using objective measures for—and an inclusive definition of—“environment” elucidated by Sallis and colleagues in their seminal work [6]. The natural environment factor, including air temperature, heat index, and precipitation, emerged as a barrier to PA, especially in rural areas. While long recognized by public health professionals for its role in communicable disease [33], this study highlights a mechanism by which climate may limit PA and lead to chronic disease, especially in rural areas of the southern and southeastern U.S. that have a lower prevalence of PA than other regions [3] and increasing experience of extreme climate events [34]. These events, such as wildfires, tornadoes, and flooding, may be particularly impactful in rural areas. For example, poor air quality from regional wildfires was found to have a greater negative impact on a walking intervention among rural participants than urban ones [35]. In addition to such acute effects, extreme weather events may cause lasting damage or destruction to the limited number of outdoor recreation facilities available in rural areas, where financial resources for reconstruction may be limited.
The association of the recreation environment with meeting PA guidelines was strongest in the Rural and Exclusively Rural counties. Within this factor, access to exercise opportunities (including private recreation facilities) did not load significantly in the two most rural categories of counties, suggesting the other variables—proximity to parks and elementary schools—are particularly important as rurality increases, in concordance with existing literature [8,9,10]. This is important because access to and funding of these facilities is a function of local decision making (e.g., shared use of school facilities, parks and recreation department budgets) [36]. In the Urban counties, access to places to exercise was of particular importance, demonstrating the primacy that private and/or indoor facilities may have in high-density urban areas. In combination with the natural environment results, these findings highlight the importance of creating opportunities for low-to-no-cost indoor activity, providing shaded outdoor places, and/or utilizing natural water assets (e.g., lakes, rivers, swimming holes) or swimming pools for rural PA, in concurrence with existing research [9].
The social environment was negatively associated with meeting PA guidelines, with greater regression coefficients in urban counties than rural ones, and a coefficient for females twice that of males across all levels of rurality. This factor was composed of variables serving as proxy measures for safety and social barriers to PA, with the prevalence of violent crime and single-parent households loading most significantly on the factor across all levels of rurality and for both sexes. Our findings, using objective measures, add to the preponderance of evidence observing that safety is associated with PA in rural settings [8]. The literature regarding the differential impact of social role constraints such as single parenting on female’s PA, especially rural females, is less clear. For example, Dlugonski and colleagues [37] identified chronic stress from being a caregiver and from meeting children’s basic needs as barriers to PA in focus groups of low-income Black single mothers of unknown rurality. Additionally, Eyler and colleagues [38] found that having fewer social role strains (e.g., child/eldercare, household tasks, work) was associated with meeting PA guidelines among rural white females (15.6% unpartnered, parenting status unknown) but not rural African American females (40.5% unpartnered, parenting status unknown) [38,39,40]. Our results add to this evolving aspect of the literature, suggesting that sociocultural PA barriers such as crime and single parenting (1) are more impactful on a female’s PA than on a male’s, and (2) may reduce in influence on PA as rurality increases until reaching the Most Rural counties; additional understanding of these relationships is needed.
The transportation environment factor had a nearly zero correlation with the adult prevalence of meeting PA guidelines, except for a negative association with PA among females in Somewhat Urban counties. This suggests that commuting behaviors and living near work were not influential in meeting PA guidelines when considering a comprehensive model of multiple environmental factors. This is consistent with other data from rural areas where transportation PA is less likely [9,10], but is somewhat unexpected for urban areas where transportation PA is more likely [41]. However, this lack of association should be considered with caution because the dependent variable in our analyses was leisure time PA as measured by the BRFSS and did not include occupational, household, or transportation PA. Recent analyses of 2011–2016 National Health and Nutrition Examination Survey data by Whitfield and colleagues [41], using domain-specific PA for rural and urban adults, demonstrated that a lower prevalence of meeting PA guidelines among rural adults is attenuated when considering occupational and household PA, in addition to leisure and transportation PA that are more common in urban adults.
There are multiple limitations to the current study worth noting. First, the PA data utilized are 10 years old, based on a small area estimation, and include only aerobic and leisure-time PA. Though PA guidelines have since been revised to incorporate muscle-strengthening PA [42], and research now highlights the importance of household [43] and occupational PA [41,43] for rural adults, these are the most recent data for meeting PA guidelines for every county in the U.S. Second, we chose to exclude frontier counties from the analyses. We excluded frontier counties primarily because the environmental measures we used in our analyses were largely missing from these counties or had extremely wide variation because of small sampling in each county. We encourage in-depth study of PA in those 443 frontier counties which have such unique characteristics. Third, because factors known to be associated with PA were included in the Dwyer-Lindgren small area estimation of PA prevalence (e.g., individual-level race/ethnicity, age group, county-level race, poverty, education), we could not assess their influence in our analysis [28]. PA prevalence data at the county level for all counties in the US continues to be a necessity and should be a priority in future national surveillance (e.g., BRFSS) and localized data reporting (e.g., CDC PLACES).
The findings of this study highlight multiple future needs, to (1) improve the granularity of PA surveillance in rural areas of the U.S., (2) identify the facilitators of and barriers to adults’ PA in rural areas of the US, and (3) utilize evidence-based strategies from rural research to address the PA disparities evident in rural areas of the U.S. Building this rural PA evidence will require a more detailed understanding of the variation of environmental factors associated with adults’ PA within rural communities. Positive Deviance methods have been used in under-resourced communities to identify promising practices in places that have better-than-expected outcomes (“Positive Outliers”) than peers so that the practices can be disseminated to peer communities [44,45]. Used most often to identify promising nutrition practices in developing countries, it has only recently been applied to identify promising PA practices among African American females [46]. The first step in this approach is identifying the places that are outperforming peers on the measures of importance. The data used for the analyses presented herein provide a promising first step in that process. Additional studies of PA in rural U.S. areas focusing on specific age groups are also warranted because rural areas of the U.S. tend to have a larger proportion of older adults than urban and suburban areas [47]. Further, though our study focused exclusively on the U.S., comparative international rural PA research is also warranted. Like the U.S.-based built environment PA literature, comparative international studies, such as those conducted by the International Physical Activity and Environment Network (IPEN), have initially focused on urban areas where large sample sizes and datasets are more readily accessible [48,49,50,51].

5. Conclusions

This comprehensive, national-level comparison of environmental factors associated with adults’ PA in rural and urban areas in the U.S. addresses a gap previously identified in the literature [9,10]. Despite the limitations noted, the findings highlight the variation of natural, recreation, and social environment factors associated with PA across the rural-urban continuum by sex. Replication of our analyses with more recent data is needed to determine whether these associations have persisted or changed over time.

Supplementary Materials

The following are available online at https://0-www-mdpi-com.brum.beds.ac.uk/article/10.3390/ijerph18147688/s1: Table S1: Bivariate Correlations for Male and Female Physical Activity by Natural Environment Indicators, Table S2: Bivariate Correlations for Male and Female Physical Activity by Social Environment Indicators, Table S3: Bivariate Correlations for Male and Female Physical Activity by Recreation Environment Indicators, Table S4: Bivariate Correlations for Male and Female Physical Activity by Transportation Environment Indicators.

Author Contributions

Conceptualization, C.G.A., M.R.U.M., M.B.E., C.K.P. and J.N.R.; methodology, C.G.A. and S.M.D.; software, S.M.D.; formal analysis, S.M.D.; data curation, C.G.A. and S.M.D.; writing—original draft preparation, C.G.A., S.M.D., M.R.U.M., and M.B.E.; writing—review and editing, L.J., M.M., C.K.P., and J.N.R.; visualization, S.M.D.; supervision, C.G.A. and M.R.U.M.; project administration, C.G.A. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding. J.N.R. is funded by the United States Department of Agriculture, Agricultural Research Service, project 3062-51000-051-00D.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

No new data were created or analyzed in this study. Data sharing is not applicable to this article.

Acknowledgments

This work is a product of the Physical Activity Policy Research and Evaluation Network (PAPREN). PAPREN is supported by Cooperative Agreement Number U48DP006381 from the Centers for Disease Control and Prevention and is a product of the Physical Activity Policy Evaluation and Research Network, a thematic network of the Prevention Research Center network. All authors except S.M.D. and J.N.R. are members of the PAPREN Rural Active Living Work Group. PAPREN is an applied research and evaluation network focused on identification and implementation of local, state, and national policy approaches that influence opportunities for physical activity and built environment strategies. The findings and conclusions in this paper are those of the author(s) and do not necessarily represent the official position of the Centers for Disease Control and Prevention. We would also like to acknowledge the technical support of Jeff Whitfield of the CDC’s National Center for Chronic Disease Prevention and Health Promotion, Physical Activity and Health Branch.

Conflicts of Interest

The authors declare no conflict of interest.

References

  1. Eberhardt, M.S.; Pamuk, E.R. The Importance of Place of Residence: Examining Health in Rural and Nonrural Areas. Am. J. Public Health 2004, 94, 1682–1686. [Google Scholar] [CrossRef] [PubMed]
  2. Matthews, K.; Croft, J.B.; Liu, Y.; Lu, H.; Kanny, D.; Wheaton, A.G.; Cunningham, T.J.; Khan, L.K.; Caraballo, R.S.; Holt, J.B.; et al. Health-Related Behaviors by Urban-Rural County Classification—United States, 2013. MMWR. Surveill. Summ. 2017, 66, 1–8. [Google Scholar] [CrossRef] [PubMed]
  3. Whitfield, G.P.; Carlson, S.A.; Ussery, E.N.; Fulton, J.E.; Galuska, D.A.; Petersen, R. Trends in Meeting Physical Activity Guidelines Among Urban and Rural Dwelling Adults—United States, 2008–2017. MMWR. Morb. Mortal. Wkly. Rep. 2019, 68, 513–518. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  4. Stokols, D. Establishing and maintaining healthy environments. Toward a social ecology of health promotion. Am. Psychol. 1992, 47, 6–22. [Google Scholar] [CrossRef]
  5. Ding, D.; Gebel, K. Built environment, physical activity, and obesity: What have we learned from reviewing the literature? Health Place 2012, 18, 100–105. [Google Scholar] [CrossRef] [Green Version]
  6. Sallis, J.F.; Cervero, R.B.; Ascher, W.; Henderson, K.A.; Kraft, M.K.; Kerr, J. An ecological approach to creating active living communities. Annu. Rev. Public Health 2006, 27, 297–322. [Google Scholar] [CrossRef] [Green Version]
  7. Bandura, A. Social Learning Theory; Prentice-Hall: Englewood Cliffs, NJ, USA, 1977. [Google Scholar]
  8. Frost, S.S.; Goins, R.T.; Hunter, R.H.; Hooker, S.; Bryant, L.L.; Kruger, J.; Pluto, D. Effects of the Built Environment on Physical Activity of Adults Living in Rural Settings. Am. J. Health Promot. 2010, 24, 267–283. [Google Scholar] [CrossRef]
  9. Hansen, A.Y.; Meyer, M.R.U.; Lenardson, J.D.; Hartley, D. Built Environments and Active Living in Rural and Remote Areas: A Review of the Literature. Curr. Obes. Rep. 2015, 4, 484–493. [Google Scholar] [CrossRef]
  10. Meyer, M.R.U.; Moore, J.; Abildso, C.; Edwards, M.B.; Gamble, A.; Baskin, M.L. Rural Active Living: A Call to Action. J. Public Health Manag. Pract. 2016, 22, E11–E20. [Google Scholar] [CrossRef] [Green Version]
  11. Whitfield, G.P.; Carlson, S.A.; Ussery, E.N.; Watson, K.B.; Berrigan, D.; Fulton, J.E. National-level environmental perceptions and walking among urban and rural residents: Informing surveillance of walkability. Prev. Med. 2019, 123, 101–108. [Google Scholar] [CrossRef] [PubMed]
  12. Carlson, S.A.; Whitfield, G.P.; Peterson, E.L.; Ussery, E.N.; Watson, K.B.; Berrigan, D.; Fulton, J.E. Geographic and Urban–Rural Differences in Walking for Leisure and Transportation. Am. J. Prev. Med. 2018, 55, 887–895. [Google Scholar] [CrossRef]
  13. King, K.E.; Clarke, P.J. A Disadvantaged Advantage in Walkability: Findings from Socioeconomic and Geographical Analysis of National Built Environment Data in the United States. Am. J. Epidemiol. 2014, 181, 17–25. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  14. Brownson, R.C.; Housemann, R.; Brown, D.R.; Jackson-Thompson, J.; King, A.C.; Malone, B.R.; Sallis, J.F. Promoting physical activity in rural communities: Walking trail access, use, and effects. Am. J. Prev. Med. 2000, 18, 235–241. [Google Scholar] [CrossRef]
  15. Librett, J.J.; Yore, M.M.; Schmid, T.L. Characteristics of Physical Activity Levels among Trail Users in a U.S. National Sample. Am. J. Prev. Med. 2006, 31, 399–405. [Google Scholar] [CrossRef] [PubMed]
  16. Edwards, M.B.; Jilcott, S.B.; Floyd, M.F.; Moore, J.B. County-level disparities in access to recreational resources and associations with adult obesity. J. Park Recreat. Adm. 2011, 29, 39–54. [Google Scholar]
  17. Moore, J.B.; Beets, M.W.; Morris, S.F.; Kolbe, M.B. Comparison of Objectively Measured Physical Activity Levels of Rural, Suburban, and Urban Youth. Am. J. Prev. Med. 2014, 46, 289–292. [Google Scholar] [CrossRef]
  18. Kaczynski, A.T.; Eberth, J.M.; Stowe, E.W.; Wende, M.E.; Liese, A.D.; McLain, A.C.; Breneman, C.B.; Josey, M.J. Development of a national childhood obesogenic environment index in the United States: Differences by region and rurality. Int. J. Behav. Nutr. Phys. Act. 2020, 17, 83. [Google Scholar] [CrossRef] [PubMed]
  19. Wende, M.E.; Stowe, E.W.; Eberth, J.M.; McLain, A.C.; Liese, A.D.; Breneman, C.B.; Josey, M.J.; Hughey, S.M.; Kaczynski, A.T. Spatial clustering patterns and regional variations for food and physical activity environments across the United States. Int. J. Environ. Health Res. 2020, 1–15. [Google Scholar] [CrossRef]
  20. U.S. Department of Health and Human Services, Health Resources and Services Administration (HRSA). Methodology for Designation of Frontier and Remote Areas; U.S. Department of Health and Human Services, Health Resources and Services Administration (HRSA): Rockville, MD, USA, 2012; pp. 66471–66476. Available online: https://www.federalregister.gov/d/2012-26938 (accessed on 12 May 2021).
  21. United States Census Bureau. American Community Survey. Available online: https://www.census.gov/data.html (accessed on 23 February 2020).
  22. Centers for Disease Control and Prevention (CDC). Behavioral Risk Factor Surveillance System (BRFSS) Prevalence and Trends Data. Available online: https://www.cdc.gov/brfss/data_tools.htm (accessed on 23 February 2020).
  23. University of Wisconsin Population Health Institute. County Health Rankings Roadmaps. Available online: https://www.countyhealthrankings.org/ (accessed on 23 February 2020).
  24. United States Department of Transportation. National Highway Traffic Safety Administration. In Fatality Analysis Reporting System (FARS). Available online: https://www.transportation.gov/sites/dot.gov/files/2021-03/Privacy%20-%20NHTSA%20-%20CDAN%20-%20PIA%20-%20Approved%20-%2003021.pdf (accessed on 23 February 2020).
  25. Centers for Disease Control and Prevention National Center for Environmental Health Environmental Health Tracking Branch. National Environmental Public Health Tracking Network. Available online: http://ephtracking.cdc.gov/ (accessed on 28 July 2020).
  26. Centers for Disease Control and Prevention (CDC). North America Land Data Assimilation System (NLDAS) Daily Air Temperatures and Heat Index, Years 1979–2011 on CDC WONDER Online Database; Centers for Disease Control and Prevention (CDC): Atlanta, GA, USA, 2012. Available online: https://wonder.cdc.gov/nasa-nldas.html (accessed on 7 June 2021).
  27. Centers for Disease Control and Prevention. Behavior Risk Factor Surveillance System: 2011 Prevalence Data. Available online: http://apps.nccd.cdc.gov/brfss/ (accessed on 31 January 2020).
  28. Dwyer-Lindgren, L.; Freedman, G.; Engell, R.; Fleming, T.D.; Lim, S.S.; Murray, C.J.; Mokdad, A.H. Prevalence of physical activity and obesity in US counties, 2001–2011: A road map for action. Popul. Health Metrics 2013, 11, 7. [Google Scholar] [CrossRef] [Green Version]
  29. Srebotnjak, T.; Mokdad, A.H.; Murray, C.J. A novel framework for validating and applying standardized small area measurement strategies. Popul. Health Metrics 2010, 8, 26. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  30. SAS Institute. Base SAS 9.4® Procedures Guide: Statistical Procedures, 2nd ed.; SAS Institute, Inc.: Cary, NC, USA, 2013. [Google Scholar]
  31. Hatcher, L.; O’Rourke, N. Principal Components Analysis. In A Step-by-Step Approach to Using the SAS System for Factor Analysis and Structural Equation Modeling, 2nd ed.; SAS Institute, Inc.: Cary, NC, USA, 2013; pp. 1–42. [Google Scholar]
  32. Lusk, A.C.; Morency, P.; Miranda-Moreno, L.F.; Willett, W.C.; Dennerlein, J.T. Bicycle Guidelines and Crash Rates on Cycle Tracks in the United States. Am. J. Public Health 2013, 103, 1240–1248. [Google Scholar] [CrossRef] [PubMed]
  33. Bell, M.; Greenberg, M.R. Climate Change and Human Health: Links between History, Policy, and Science. Am. J. Public Health 2018, 108, S54–S55. [Google Scholar] [CrossRef] [PubMed]
  34. U.S. Department of Commerce, National Oceanic and Atmospheric Administration (NOAA), U.S. Climate Extremes Index (CEI). Available online: https://www.ncdc.noaa.gov/extremes/cei/ (accessed on 13 May 2021).
  35. Perry, C.K.; Campbell, L.P.; Currier, J.; Farris, P.E.; Wenzel, E.S.; Medysky, M.E.; Zell, A.; McDonell, M.; Shannon, J.; Winters-Stone, K. An Evidence-Based Walking Program in Oregon Communities: Step It Up! Survivors. Prev. Chronic Dis. 2020, 17, E156. [Google Scholar] [CrossRef] [PubMed]
  36. American Public Health Association. Improving Health and Wellness through Access to Nature; Policy Statement 20137; American Public Health Association: Washington, DC, USA, 2013. [Google Scholar]
  37. Dlugonski, D.; Martin, T.R.; Mailey, E.L.; Pineda, E. Motives and Barriers for Physical Activity among Low-Income Black Single Mothers. Sex Roles 2016, 77, 379–392. [Google Scholar] [CrossRef]
  38. Eyler, A.A.; Matson-Koffman, D.; Young, D.R.; Wilcox, S.; Wilbur, J.; Thompson, J.L.; Sanderson, B.K.; Evenson, K.R. Quantitative Study of Correlates of Physical Activity in Women from Diverse Racial/Ethnic Groups: Women’s Cardiovascular Health Network Project Introduction and Methodology. Am. J. Prev. Med. 2003, 25, 5–14. [Google Scholar] [CrossRef]
  39. Eyler, A.A. Personal, social, and environmental correlates of physical activity in rural Midwestern white women. Am. J. Prev. Med. 2003, 25, 86–92. [Google Scholar] [CrossRef]
  40. Sanderson, B.K.; Foushee, H.R.; Bittner, V.; Cornell, C.; Stalker, V.; Shelton, S.; Pulley, L. Personal, social, and physical environmental correlates of physical activity in rural African-American women in Alabama. Am. J. Prev. Med. 2003, 25, 30–37. [Google Scholar] [CrossRef]
  41. Whitfield, G.P.; Ussery, E.N.; Carlson, S.A. Combining Data From Assessments of Leisure, Occupational, Household, and Transportation Physical Activity Among US Adults, NHANES 2011–2016. Prev. Chronic Dis. 2020, 17, E117. [Google Scholar] [CrossRef]
  42. Piercy, K.L.; Troiano, R.P.; Ballard, R.M.; Carlson, S.A.; Fulton, J.E.; Galuska, D.A.; George, S.M.; Olson, R.D. The Physical Activity Guidelines for Americans. JAMA 2018, 320, 2020–2028. [Google Scholar] [CrossRef] [PubMed]
  43. Fan, J.X.; Wen, M.; Kowaleski-Jones, L. Rural–Urban Differences in Objective and Subjective Measures of Physical Activity: Findings from the National Health and Nutrition Examination Survey (NHANES) 2003–2006. Prev. Chronic Dis. 2014, 11, E141. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  44. Marsh, D.R.; Schroeder, D.G. The Positive Deviance Approach to Improve Health Outcomes: Experience and Evidence from the Field—Preface. Food Nutr. Bull. 2002, 23, 3–6. [Google Scholar] [CrossRef] [Green Version]
  45. Marsh, D.R.; Schroeder, D.G.; Dearden, K.; Sternin, J.; Sternin, M. The power of positive deviance. BMJ 2004, 329, 1177–1179. [Google Scholar] [CrossRef] [PubMed] [Green Version]
  46. Kinsey, A.W.; Segar, M.L.; Barr-Anderson, D.J.; Whitt-Glover, M.C.; Affuso, O. Positive Outliers Among African American Women and the Factors Associated with Long-Term Physical Activity Maintenance. J. Racial Ethn. Health Disparities 2019, 6, 603–617. [Google Scholar] [CrossRef]
  47. Pew Research Center. What Unites and Divides Urban, Suburban and Rural Communities. May 2018. Available online: https://www.pewresearch.org/social-trends/2018/05/22/demographic-and-economic-trends-in-urban-suburban-and-rural-communities/ (accessed on 13 July 2021).
  48. Cain, K.L.; Salmon, J.; Conway, T.L.; Cerin, E.; Hinckson, E.; Mitáš, J.; Schipperijn, J.; Frank, L.D.; Anjana, R.M.; Barnett, A.; et al. International Physical Activity and Built Environment Study of adolescents: IPEN Adolescent design, protocol and measures. BMJ Open 2021, 11, e046636. [Google Scholar] [CrossRef]
  49. Kerr, J.; Sallis, J.F.; Owen, N.; De Bourdeaudhuij, I.; Cerin, E.; Sugiyama, T.; Reis, R.; Sarmiento, O.; Frömel, K.; Mitás, J.; et al. Advancing Science and Policy Through a Coordinated International Study of Physical Activity and Built Environments: IPEN Adult Methods. J. Phys. Act. Health 2013, 10, 581–601. [Google Scholar] [CrossRef]
  50. Sallis, J.F.; Cerin, E.; Conway, T.L.; Adams, M.A.; Frank, L.D.; Pratt, M.; Salvo, D.; Schipperijn, J.; Smith, G.; Cain, K.L.; et al. Physical activity in relation to urban environments in 14 cities worldwide: A cross-sectional study. Lancet 2016, 387, 2207–2217. [Google Scholar] [CrossRef] [Green Version]
  51. Sallis, J.F.; Cerin, E.; Kerr, J.; Adams, M.A.; Sugiyama, T.; Christiansen, L.B.; Schipperijn, J.; Davey, R.; Salvo, D.; Frank, L.D.; et al. Built Environment, Physical Activity, and Obesity: Findings from the International Physical Activity and Environment Network (IPEN) Adult Study. Annu. Rev. Public Health 2020, 41, 119–139. [Google Scholar] [CrossRef] [PubMed] [Green Version]
Table 1. List of Data Sources and Variables by Environment Type.
Table 1. List of Data Sources and Variables by Environment Type.
VariableSourceDefinition (Scale)
Dependent Variable: Prevalence of meeting PA Guidelines
Female PA2011 BRFSS [27]150 min per week of MPA, 75 min per week of VPA, or combination of MVPA (1 min VPA = 2 min MPA), (% of females) [28]
Male PA2011 BRFSS [27]150 min per week of MPA, 75 min per week of VPA, or combination of MVPA (1 min VPA = 2 min MPA), (% of males) [28]
Natural Environment
Air Temperature2011 WONDER [26]Average Daily Max Air Temperature (F°)
Heat Index2011 WONDER [26]Average Daily Max Heat Index (F°)
Sunlight2011 WONDER [26]Average Daily Sunlight (KJ/m2)
Precipitation2011 WONDER [26]Average Daily Precipitation (mm)
Water area2010 Census [21]Total Area of County that is Water (%)
Social Environment
Alcohol Vehicle Deaths2012, 5 year FARS (from 2014 CHR) [23]Motor vehicle crash deaths with alcohol involvement (%)
Violent Crime2011, 3 year UCR (from 2014 CHR) [23]Number of violent crimes per 100,000 population
Single Parent Households2012, 5 year ACS (from 2014 CHR) [23]Children 18 years of age living in family households that are headed by a single parent (%)
Vacant Housing2011, 5 year ACS (from 2011 EPH) [25]Housing units that are vacant (%)
Recreation Environment
Access to Exercise Opportunities2014 CHR (using 2010 and 2012 data) [23]Population within ½ mile of a park or 1 mile (urban) or 3 miles (rural) of a recreation facilities (%)
Highway proximity2010 EPH [25]Population living within 150 m of a highway (%)
Park proximity2010 EPH [25]Population living within Half-Mile of Park (%)
School proximity2010 EPH [25]Population aged 5 to 9 years living within a half mile of a public elementary school (%)
Transportation Environment
Bike to Work2011, 5 year ACS (from 2011 EPH) [25]Workers over 16 years that bike to work (%)
Walk to Work2011, 5 year ACS (from 2011 EPH) [25]Workers over 16 years that walk to work (%)
Works in County2012, 5 year ACS (from 2014 CHR) [23]Works in County of Residence (%)
Works in Place2012, 5 year ACS (from 2014 CHR) [23]Works in Place of Residence (%)
Long Commute2012, 5 year ACS (from 2014 CHR) [23]Drive Alone-Long Commute to Work (%)
Cluster and Stratification
Rural Percent2010 Census [21]Population that lives in a rural area (%)
Rural Quintile (group)
State (name)2010 Census [21]
Notes: F°, Fahrenheit; KJ/m2, kilojoules per square meter; mm, millimeters; MPA, moderate physical activity; VPA, vigorous physical activity; MVPA, moderate-to-vigorous physical activity; ACS, American Community Survey, (https://www.census.gov/data.html, accessed on 23 February 2020); BRFSS, Behavioral Risk Factor Surveillance System, (https://www.cdc.gov/brfss/, accessed on 23 February 2020); CDC, Centers for Disease Control and Prevention; CHR, County Health Rankings, (https://www.countyhealthrankings.org/, accessed on 23 February 2020); WONDER, Wide-ranging Online Data for Epidemiologic Research, (https://wonder.cdc.gov/nasa-nldas.html, accessed on 28 July 2020); EPH, National Environmental Public Health Tracking Network, (https://ephtracking.cdc.gov/, accessed on 28 July 2020); UCR, The Uniform Crime Reporting Program, (https://www.fbi.gov/services/cjis/ucr, accessed on 23 February 2020).
Table 2. Measures of Central Tendency for 2011 Male and Female County-level Prevalence of Meeting Physical Activity Guidelines.
Table 2. Measures of Central Tendency for 2011 Male and Female County-level Prevalence of Meeting Physical Activity Guidelines.
Male PAFemale PA
% RuralGroupn%MSDMinMaxn%MSDMinMax
0–25Urban57721.456.215.936.872.757721.452.621.633.174.2
25–48Somewhat Urban57721.454.113.738.473.257721.450.219.730.774.7
48–67Somewhat Rural57721.452.216.337.572.257721.448.021.929.168.8
67–99Rural57721.450.620.233.771.457721.446.825.529.067.8
100Exclusively Rural38914.450.121.533.170.438914.446.926.031.371.6
All2697100.052.819.233.173.22697100.049.024.429.074.7
Notes: The data source was Dwyer-Lindgren and colleagues’ [28] county-level small area estimation using 2011 Behavioral Risk Factor Surveillance System data. The US PA guidelines at the time were for adults to achieve 150 min of moderate intensity PA per week, the equivalent in vigorous PA, or combination of moderate and vigorous PA where one minute of vigorous PA is equivalent to two minutes of moderate PA.
Table 3. Exploratory Factor Analysis Factor Loadings.
Table 3. Exploratory Factor Analysis Factor Loadings.
All GroupsUrbanSomewhat UrbanSomewhat RuralRuralExclusively Rural
(0–25% Rural)(25–48% Rural)(48–67% Rural)(67–99% Rural)(100% Rural)
Natural Environment (n)(2681)(567)(573)(577)(577)(387)
Air0.980.960.970.980.990.99
Sun0.870.900.840.890.900.85
Heat0.660.520.640.720.760.70
Water−0.080.00−0.07−0.19−0.19−0.08
Precipitation−0.21−0.42−0.22−0.24−0.130.12
Social Environment (n)(2577)(571)(561)(554)(554)(337)
Single Parent Households0.670.800.680.730.610.47
Violent Crime0.590.750.670.680.570.53
Vacant Housing0.240.490.360.480.340.42
Alcohol Vehicle Deaths0.070.000.050.140.050.16
Recreation Environment (n)(2686)(576)(576)(576)(577)(381)
Live within Half-Mile of Park0.760.750.670.610.650.58
Access to Exercise Opportunities0.590.580.410.290.250.10
Elementary Half-Mile0.550.570.540.510.570.56
Live within 150 Miles of Highway0.460.320.120.080.010.14
Transportation Environment (n)(2697)(577)(577)(577)(577)(389)
Works in County0.820.710.820.820.820.79
Works in Place0.750.700.740.690.650.68
Walk to Work0.500.530.600.470.530.56
Bike to Work0.490.470.520.370.360.35
Long Commute−0.81−0.71−0.82−0.80−0.76−0.82
Notes: Factor scores ≥ 0.4 are bolded for emphasis.
Table 4. Standardized Regression Coefficients and Standard Errors for Multiple Linear Regression of 2011 Male and Female Prevalence of Meeting Physical Activity Guidelines by Environmental Factor Scores.
Table 4. Standardized Regression Coefficients and Standard Errors for Multiple Linear Regression of 2011 Male and Female Prevalence of Meeting Physical Activity Guidelines by Environmental Factor Scores.
Male Physical Activity
All Groups (n = 2560)Urban (n = 564)Somewhat Urban (n = 557)Somewhat Rural (n = 553)Rural (n = 554)Exclusively Rural (n = 332)
Factorβseβseβseβseβseβse
Natural Environment−0.32 a0.07−0.170.10−0.41 a0.08−0.46 a0.09−0.46 a0.11−0.43 a0.14
Social Environment−0.11 b0.07−0.32 a0.08−0.19 a0.08−0.16 b0.080.010.11−0.070.10
Recreation Environment0.34 a0.060.12 b0.140.010.080.070.100.24 a0.090.28 b0.14
Transportation Environment−0.010.03−0.070.060.050.05−0.050.060.010.05−0.060.09
Female Physical Activity
All Groups (n = 2560)Urban (n = 564)Somewhat Urban (n = 557)Somewhat Rural (n = 553)Rural (n = 554)Exclusively Rural (n = 332)
Factorβseβseβseβseβseβse
Natural Environment−0.31 a0.08−0.160.13−0.36 a0.09−0.42 a0.11−0.46 a0.13−0.41 a0.13
Social Environment−0.24 a0.06−0.43 a0.08−0.31 a0.08−0.30 a0.08−0.090.12−0.20 b0.11
Recreation Environment0.33 a0.070.20 b0.150.070.100.110.090.27 a0.090.27 b0.14
Transportation Environment0.010.03−0.030.060.12 b0.05−0.010.05−0.020.06−0.020.09
Notes: a: p ≤ 0.01, b: p ≤ 0.05. Regression models were adjusted for counties clustered by state.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Share and Cite

MDPI and ACS Style

Abildso, C.G.; Daily, S.M.; Meyer, M.R.U.; Edwards, M.B.; Jacobs, L.; McClendon, M.; Perry, C.K.; Roemmich, J.N. Environmental Factors Associated with Physical Activity in Rural U.S. Counties. Int. J. Environ. Res. Public Health 2021, 18, 7688. https://0-doi-org.brum.beds.ac.uk/10.3390/ijerph18147688

AMA Style

Abildso CG, Daily SM, Meyer MRU, Edwards MB, Jacobs L, McClendon M, Perry CK, Roemmich JN. Environmental Factors Associated with Physical Activity in Rural U.S. Counties. International Journal of Environmental Research and Public Health. 2021; 18(14):7688. https://0-doi-org.brum.beds.ac.uk/10.3390/ijerph18147688

Chicago/Turabian Style

Abildso, Christiaan G., Shay M. Daily, M. Renée Umstattd Meyer, Michael B. Edwards, Lauren Jacobs, Megan McClendon, Cynthia K. Perry, and James N. Roemmich. 2021. "Environmental Factors Associated with Physical Activity in Rural U.S. Counties" International Journal of Environmental Research and Public Health 18, no. 14: 7688. https://0-doi-org.brum.beds.ac.uk/10.3390/ijerph18147688

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop