Elsevier

Preventive Medicine

Volume 39, Issue 6, December 2004, Pages 1115-1125
Preventive Medicine

Age, gender, and urban–rural differences in the correlates of physical activity

https://doi.org/10.1016/j.ypmed.2004.04.024Get rights and content

Abstract

Background. The majority of the population is inactive, and strategies to date for promoting regular physical activity have been limited in their effectiveness. Further research is needed to identify correlates of physical activity in different subgroups to design more efficacious interventions. This study sought to identify correlates of physical activity across men and women, urban and rural geographical locations, and four distinct age groups (18–25; 26–45; 46–59; and 60+).

Methods. This study employed data from a large provincial household random sample (N = 20,606) of Canadians. Analyses were utilized to examine the amount of variance explained in self-reported physical activity by a number of demographic and/or biological, psychological, behavioral, social, and environmental variables within each subgroup.

Results. Proportion of friends who exercise, injury from past physical activity, educational level, perceived health status, and alcohol consumption were among the strongest correlates across subgroups.

Conclusions. A number of correlates were identified as being significant across all subgroups examined. Most differences in the correlates of physical activity were found within different age groups rather than among urban and rural residents and gender.

Introduction

Physical activity is associated with a reduction in all-cause mortality, fatal and nonfatal total cardiovascular disease, a reduction in the incidence of obesity, type 2 diabetes mellitus, colon cancer, and osteoporosis [1]. As a result, regular physical activity is strongly recommended by numerous organizations for its health benefits, including the Surgeon General's Report on physical activity and health [2].

Despite such recommendations, fewer than 40% of adults in the Western world currently participate in regular physical activity [3], suggesting an urgent need to implement interventions and programs to promote physical activity in the population. While numerous physical activity interventions or programs have been developed and evaluated, such initiatives have produced varied results, with limited impact on overall population rates of physical activity. Further research is therefore required to gain a better understanding of the predictors of physical activity [4], [5] as these factors may then be targeted in programs designed to promote physical activity [6]. Recent reviews of physical activity correlates in adults indicate that this is a multifactorial behavior influenced by demographic, biological, psychological, behavioral, social and/or cultural, and physical environmental factors [5], [6], [7].

Among the most common correlates of physical activity are ‘self-efficacy’ (positive association), ‘social support’ (positive association), ‘gender’ (males engage in higher levels of physical activity than females), and ‘age’ (younger people are more active than older people) [7]. Early correlate studies of adult physical activity indicate that self-efficacy is consistently associated with physical activity participation [8], [9]. For example, self-efficacy was reported to have the highest correlation (r = 0.48), with physical activity among other social–cognitive constructs in a cross-sectional study of 2,053 adults [8]. In a more recent study, self-efficacy accounted for 10% of change in objectively measured postintervention physical activity [10]. Social support is another consistent correlate of physical activity cited in the literature [11], [12]. De Bourdeaudhuij and Sallis [11] found that social support variables explained the most variance in physical activity, ranging from 3% to 11% across three age groups for males and females (with the exception of 50- to 65-year-old females). Perceived barriers to physical activity participation are also a frequently cited correlate, with different barriers being important for different age groups. Barriers include lack of time for young people and poor health for the elderly [13].

Environmental and demographic factors are also associated with physical activity. Although recent attention has been given to environmental factors [7], they are the least studied variables in the physical activity domain [6]. However, access to exercise facilities in close proximity seems to be a significant predictor of physical activity [14]. By contrast, the demographic characteristics of age and gender are among the most studied correlates of physical activity. It is well established that physical activity declines with age [15], [16], and males are more active than females [7].

These findings demonstrate the importance of examining physical activity correlates, as subgroups of inactive adults may be targeted for tailored intervention studies. Subgroups are likely to differ in the factors that influence their physical activity and therefore require further investigation [6]. Examples of subgroups include men and women, urban and rural residents, and adults of different ages.

Whereas gender and age group are commonly assessed correlates of physical activity, minimal data exist comparing correlates of physical activity between men and women across age groups. Two recent studies conducted in Belgium [11] and Australia [13] have attempted to identify correlates of physical activity for specific age groups for both genders. Findings from these studies suggest that the pattern of associations varies by different age and gender groups. Minimal research has examined whether the predictors of physical activity differ for those living in urban versus rural settings. However, a recent study conducted in the United States [17] found a different pattern of determinants between urban and rural women including diverse barriers and enablers with leisure time physical activity. No study has yet compared both men and women living in urban and rural settings. Furthermore, in their recent review, Trost et al. [7] concluded that leisure time physical activity levels are significantly lower among adults living in rural areas than in urban settings. Consequently, differences between urban and rural residents on the correlates of physical activity participation may therefore be expected.

The purpose of the present study was therefore to compare potential correlates of physical activity across men and women, urban and rural residents, and different age groups. To our knowledge, no other study has examined correlates of physical activity across different subgroups in a large random population-based sample. Findings from the present study will also assist in the corroboration of results from other investigations assessing correlates of physical activity in smaller published study samples. The unique results of this population-based study along with consensus of findings with such other samples will strengthen public and population health research, practice, and policy to target specific factors and subgroups of the population for physical activity promotion interventions and programs.

Section snippets

Sampling procedures

Data from this study were obtained from the 1990 Ontario Health Survey (1990 OHS) [18], [19], a random survey of 54,466 individuals. Objectives of the 1990 OHS survey included the following: (1) measuring the health status of the population, (2) collecting data on the determinants of major causes of morbidity and mortality in the Canadian province of Ontario, and (3) collecting data related to social, economic, demographic, and geographic variations in health [18]. The 1990 OHS comprised a

Physical activity assessment

Leisure time physical activity in the 1990 OHS was assessed by the validated physical activity component of the U.S. National Health Interview Survey [22]. The validity of this instrument is reported elsewhere [23], [24]. The slightly modified physical activity instrument of the 1990 OHS (to reflect physical activities in the past month) has also been employed by other researchers and corresponds with the Minnesota Leisure time Physical Activity Questionnaire [12], [25]. This self-report

Results

Due to the numerous independent variables used in the analyses, only those with the highest association with physical activity are presented in the text. As variables were measured at the nominal level, parentheses next to each variable indicate the level with which the highest significant association was obtained. Significant positive and negative associations found with physical activity are indicated by (+) and (−) signs, respectively.

Discussion

The present study identified correlates of physical activity across gender, urban and rural location, and age subgroups to contribute to current literature on the activity needs of different population groups. In this study, between 16% and 30% of the variance in physical activity was explained by the independent variables. Few discrepancies in the correlates of physical activity were noted in the subgroups of gender and strata. However, some notable discrepancies were observed between the four

Acknowledgements

The authors would like to acknowledge the editorial assistance of Kylie Hugo and Tricia Prodaniuk in preparing this manuscript. Ronald C. Plotnikoff is supported by Salary Awards from the Alberta Heritage Foundation for Medical Research and the Canadian Institute of Health Research.

References (32)

  • T. Stahl et al.

    The importance of the social environment for physically active lifestyle: results from an international study

    Soc. Sci. Med

    (2001)
  • M.S. Kaplan et al.

    Demographic and psychosocial correlates of physical activity in later life

    Am. J. Prev. Med

    (2001)
  • Y.A. Kesaniemi et al.

    Dose–response issues concerning physical activity and health: an evidence-based symposium

    Med. Sci. Sports Exerc

    (2001)
  • Physical Activity and Health: A Report of the Surgeon General

    (1996)
  • V. Seefeldt et al.

    Factors affecting levels of physical activity in adults

    Sports Med

    (2002)
  • J.F. Sallis et al.

    Physical Activity and Behavioral Medicine

    (1999)
  • Cited by (95)

    • Rural-urban differences in individual and environmental correlates of physical activity in Canadian adults

      2022, Preventive Medicine Reports
      Citation Excerpt :

      Regular participation in physical activity is important for overall well-being including noncommunicable disease prevention, social connectedness, and mental health (Chekroud et al., 2018; Lee et al., 2012; Penedo and Dahn, 2005). Despite well-established benefits, there are substantial inter-individual variations in physical activity behavior which are associated with sociodemographic (e.g., gender, age, and socioeconomic status; Trost et al, 2002; Plotnikoff et al., 2004; Abichahine and Veenstra, 2017) and environmental factors (e.g., rurality; Pelletier et al., 2021a; Lim et al., 2021). Unequal distribution of opportunities for engagement in physical activity and for physical activity promotion contribute to inequalities in physical activity behavior (Chastin et al., 2020; Althoff et al., 2017; Cleland et al., 2010).

    • Cross-sectional and prospective associations between active living environments and accelerometer-assessed physical activity in the EPIC-Norfolk cohort

      2021, Health and Place
      Citation Excerpt :

      Our study is the first to estimate the cross-sectional and prospective associations of objectively-assessed population density, junction density and land use mix and perceptions of ALEs with accelerometer-assessed ST, LPA, MVPA, and overall PA among older adults. Our finding that more mixed/dense neighbourhoods are associated with less optimal activity profiles among older adults is consistent with evidence that the correlates of PA differ across age subgroups (Plotnikoff et al., 2004), and that denser and more mixed ALEs are associated with lower diabetes incidence in adults <65 years but not among adults ≥65 years (Booth et al., 2019). While living in neighbourhoods that are denser and more mixed may facilitate PA among young and middle-aged adults (Smith et al., 2017), our study suggests that older adults may not be disadvantaged by living less dense and less mixed-use neighbourhoods.

    View all citing articles on Scopus
    View full text