Individual social capital and physical activity in Riley County, KS



Journal Title

Journal ISSN

Volume Title



BACKGROUND: Resource-deficient social networks can have detrimental health effects on individuals. In part, this may result from limited engagement in health-related behaviors such as physical activity (PA). PA can lower one’s risk for a variety of chronic diseases and promoting this behavior is a public health priority. Understanding the relationship between network-owned resources and PA behavior can help policymakers and health advocates develop comprehensive programs to promote health among those who are disadvantaged. Measurements of resources available through an individual's social network have previously been referred to as social capital (SC). This study set out to explore the relationship between SC and PA by testing the following hypotheses: (1) Individuals with low household income will report low levels of both SC and PA participation, (2) SC has a positive relationship with PA participation, (3) Individuals who engage in leisure-time PA with others will report higher levels of SC than those who do it alone, (4) Social support for PA mediates the relationship between SC and PA.

METHODS: Data were obtained from residents of Riley County, KS (n=828) who completed a survey. One-way analyses of variance (ANOVA) were used to examine differences in SC (bonding, bridging, total) and PA (moderate, vigorous, total, leisure-time) between different income groups. Relationships between SC and PA variables were examined using Pearson product-moment correlations, while independent sample t-tests were used to identify differences in SC between those who did or did not meet recommended PA guidelines. ANOVAs were also used to identify differences in SC among those who primarily engage in leisure-time PA by themselves, with a partner, or with a group. Finally, the Sobel test was used to examine the mediating effect of social support for PA in all significant linear relationships between SC and PA variables.

RESULTS: Social capital was highest among high-income groups (p <.05). Leisure-time PA was lowest in the poorest income group (p <.05). Weak linear relationships were identified between bonding SC and leisure-time PA [r(786) = 0.09, p = 0.01], as well as between total SC and leisure time PA [r(775) = 0.08, p = 0.02]. Individuals who met PA guidelines reported higher levels of SC than those who did not (bridging and total: p <.001; bonding: p = <.01). No differences in SC were found between individuals who primarily engaged in leisure-time PA by themselves, with a partner, or with a group. The Sobel test found that social support for PA fully mediated the relationships between leisure-time PA and both bonding SC (z = 5.61, p <.001) and total SC (z = 6.01, p = <.001).

CONCLUSIONS: Social capital has a positive relationship with leisure-time physical activity and influences this behavior through provisions of behavior-targeted social support. Further examination should be afforded to the relationship between social capital, income, and leisure-time physical activity. While physical activity promotion efforts that target social capital may achieve marginal improvements in physical activity levels at the population level, they are unlikely to result in significant improvements at the individual level.



Social capital, Social networks, Social support, Physical activity

Graduation Month



Master of Public Health


Department of Kinesiology

Major Professor

Brandon Irwin