Multilevel correlates of domain-specific physical activity among rural adults – a cross-sectional study

Physical activity (PA) has myriad benefits including improved weight status, reduced risk of various cancers and cardiovascular disease [1]. Most Americans do not meet the guidelines of at least 150-minutes of moderate intensity PA per week, 75-minutes of vigorous intensity PA per week, or a combination of the two [2]. Further, fewer rural residents meet PA guidelines when compared to their urban/suburban counterparts [3]. The PA divide between rural and urban areas should be delineated in order to reduce health disparities and improve health for all.

Physical activity is a complex behavior with a multitude of influences, and is strongly influenced by environmental and policy influences. Multilevel factors encourage engagement in PA such as the environment (e.g., parks, trails, sidewalks [4]) and social support (e.g., friends, family [5]) in rural communities. Ecological models of health behaviors suggest targeting multiple levels of influence (e.g., individual, interpersonal) [6]. Intervening at multiple levels takes into consideration both the individual and the environment in which the individual resides.

In 2016, a specific call to action for rural active living promoted the application of a refined ecological approach to determine the correlates of domain-specific PA in rural communities [7]. Further, the call urges the development of evidence-based interventions to improve PA in rural communities. Recent community-based, multilevel trials have been implemented in rural areas – with limited success [8, 9]. In particular, applications of multilevel trials geared toward PA improvement in rural communities often encounter issues due to small sample sizes and implementation challenges inherent to rurality (e.g., culture) [8]. Additionally, many multilevel trials tend to focus on increasing weekly total PA as the outcome of interest, with special attention paid to leisure-time PA – especially in high-income countries [10]. However, other domains of PA in addition to leisure-time PA provide opportunities for individuals from rural communities to meet the recommended guidelines. The four common domains of PA include occupational, household, transportation-related, and recreational [11‐14]. Several studies have examined multilevel correlates of domain-specific PA in urban/suburban communities [15]; however, less is understood about how these domains manifest in rural areas. One study of rural Midwestern adults found the domains of active living (i.e., walking and biking) and sport (i.e., athletic activities and exercises) had domain-specific correlates in policy, neighborhood characteristics, and social support among rural Midwestern adults while the domains of occupational and house work PA did not [5]. In the domains of active living and sport, rural residents were more likely to engage in active living if they agreed with government funding for PA resources, and they were more likely to engage in sport if they had social support to exercise [5].

In another study among rural Midwestern adults, unique correlates were found for the previously discussed domains of PA and other domains such as active transportation, yard work in conjunction with housework, and recreation [16]. The results further demonstrated the social and environmental correlates of PA are specific to the domain being measured. For example, female participants have been socialized to engage in more PA through housework when compared to male participants, while male participants have been socialized to engage in more leisure-time PA when compared to single female participants [16]. Additionally, the authors assessed correlates of intensity-specific PA and found greater awareness and use of PA-related resources encouraged engagement in vigorous PA [16].

The literature is rich in evidence supporting the impact of individual and environmental level correlates of recreational and overall PA, yet lacks in providing evidence in the context of rural communities. Furthermore, the understanding of individual and environmental level correlates of domain-specific PA in rural communities remains limited. Therefore, to better understand the correlates of overall and domain-specific (i.e., transportation, occupational, and recreational) MVPA in rural communities in order to inform PA interventions, the objectives of the present study were to (1) provide estimates of overall and domain-specific total weekly MVPA, and (2) examine associations of individual and environmental factors with overall and domain-specific weekly MVPA minutes.

The analytical sample dropped from 1252 participants to 1241 participants due to relevant missing data. The sample (mean age = 57.0 years, SD = 16.7 years) was predominantly female and white (Table 1). When examining education levels of the sample, 33.5% had a high school diploma or lower. Overall, participants reported engaging in an average of 616.9 (SD = 1141.5) minutes of weekly MVPA. When broken down further by domain, occupational MVPA contributed the most weekly minutes (mean = 445.8, SD = 1024.6), followed by recreational MVPA (mean = 119.5, SD = 222.6), and lastly transportation related MVPA (mean = 56.4, SD = 209.9), see Fig. 1.

When examining individual and environmental correlates of overall weekly MVPA, only individual correlates, including age, gender, and education were statistically significant (Table 2). Compared to younger adults, older adults’ overall weekly MVPA minutes lowered by 58% (95% CI = 48, 66%) with all other variables held constant. When compared to males, females had 35% (95% CI = 22, 46%) fewer weekly minutes of overall MVPA. Finally, when compared to those of higher education, participants with lower education levels had 44% (95% CI = 16, 79%) more weekly minutes of overall MVPA.

This study was among the first to elucidate multilevel correlates of overall and domain-specific PA in a rural population. The findings of the study demonstrate the influence of age, gender, race, and education on overall and domain-specific PA. Further, this sample of rural residents had high levels of occupational PA and conversely, low levels of transportation PA. Findings from this study contribute to the gap in literature regarding multilevel correlates of domain-specific PA in rural communities, and will benefit the design and adaptation of interventions seeking to improve levels of domain-specific PA, ultimately reducing adverse health outcomes, among rural community members.

Consistently across the domains and overall PA, increased age was found to be associated with decreasing PA. PA is well known to decrease as age increases across multiple domains [10]. The transportation domain was the one domain in which age did not reduce PA. Transportation PA in rural areas is limited due to the disperse nature of rural communities, low population density, and ample parking [23‐26]. These rural communities are car-centric as the walkability and connectivity are low, and various destinations of interest (e.g., pharmacy, grocery store) are often not near one’s home, or are located in a different community. Therefore, it is likely since all ages depend on automobiles for transportation, transportation PA did not decrease with age. While some have found maintenance of leisure and recreational activities among older adults [27], we found an inverse relationship between recreational PA and age.

In addition to being influenced by age, overall MVPA and every MVPA domain were associated with various individual level correlates. In particular, the gender of a participant determined the extent to which a participant engaged in overall MVPA and each MVPA domain. The trend maintained consistency across the different MVPA outcomes, where female participants engaged in significantly fewer weekly minutes of overall MVPA, occupational MVPA, transportation MVPA, and recreational MVPA than male participants. In general, women are known to be less physically active than their male counterparts in rural areas [1, 12, 16, 28]. Although the gender divide in MVPA has been empirically noted previously, research has been limited to overall MVPA and the recreational domain of MVPA.

Our findings highlight a significant difference in weekly minutes of occupational and transportation MVPA in addition to overall and recreational MVPA based on gender. Females in this sample of rural communities engaged in fewer minutes of weekly recreational MVPA than males, a finding aligning with previous conclusions of gender differences in the recreational MVPA domain in rural communities [16]. Unlike the age variable, a difference by gender was detected in the transportation MVPA domain. Due to the disperse layout of rural communities, walking or cycling to destinations in rural communities is uncommon [7, 29]; however, providing an environment conducive to transportation MVPA, for example, sidewalks in good repair leading to points of interest, may aid in increasing MVPA in rural communities, especially among female residents. Our findings reveal the importance of designing PA domain-specific interventions to account for gender disparities in MVPA, particularly in rural areas.

Educational attainment and PA have historically shown a direct relationship [11, 12, 30‐32]; however, the current study demonstrated an inverse relationship between education and PA, specifically in the overall, occupational, and transportation domains. The occupational domain was the main driver of overall MVPA in this study, likely due to more manual labor job participation (e.g., manufacturing, construction) leading to more occupational MVPA, and subsequently, overall MV|PA. Higher education is oftentimes not a prerequisite for manual labor jobs, therefore, it is also feasible those with lower education could not afford a vehicle and associated costs (e.g., fuel, insurance, maintenance). Therefore, they may have relied on other modalities (i.e., walking, cycling) to get to their destination – thereby increasing their transportation MVPA [31].

Lastly, trail use and perceived indoor recreation access were predictive of weekly minutes of recreational and transportation MVPA among rural community members. More walking trail use has been associated with more weekly total PA [4]; however, this study contributes to the literature by demonstrating the domain-specific impact of trail use on weekly minutes of MVPA among participants from rural communities in the Midwest. Perceived indoor recreation accesses association with improved PA has been noted previously [33, 34]; however, the current finding is specific to rural residents and the recreational domain of MVPA. Further, as indoor recreational access perceptions increased, transportation MVPA went down. The finding of increased recreational access and decreased transportation MVPA may mean the car-centric rural area residents were inclined to drive to local facilities (e.g., gyms, recreation centers) to exercise. By raising awareness about the availability and proximity to local trails and recreation facilities, individuals and families in rural communities may attain more minutes of recreational PA and improve their personal health outcomes as well as the health of their communities.

It is imperative, when measuring PA, to include the domain of the activity along with a broad array of contextual variables. Commonly, PA is viewed as an overall concept where various domains of PA are summed to create a dichotomous variable of meeting guidelines; however, actionable information may be lost when coalescing PA across domains. Physical activity comes in many forms and this study found occupational and recreational PA as important domains for rural people. By examining multilevel correlates of specific domains of PA, we can have a deeper understanding on how to best promote physical activity for different groups in different environments. Ultimately, more targeted interventions could help increase PA and reduce chronic disease at the population level in rural communities.