Assessing the association of walking with health services use and costs among socioeconomically disadvantaged older adults

A. J. Perkins, D. O. Clark

Research output: Contribution to journalArticle

36 Scopus citations


Background. The costs of physical inactivity are beginning to be recognized. Research to pinpoint these costs will provide needed information for researchers and policy-makers to develop cost-effective physical activity promotion programs. We present the association of walking with health services use and costs within a sample of 695 older, urban primary care patients. Methods. A survey provided most data, but utilization and cost data were obtained from a medical records system. Multivariate models were developed to assess the association of walking with health services use and costs, adjusting for sociodemographic characteristics, chronic disease, health status, and previous utilization. Results. Thirty-eight percent of respondents reported walking 0 minutes per week, 49% walked 1 to 119 minutes, and 13% walked 120 minutes or more. In the multivariate analyses, a report of walking 120 or more minutes was associated with a lower risk of emergency room visit and hospital stay in the subsequent year. No association was found between walking and primary care visits and total cost. Conclusion. These analyses suggest an association of walking 120 minutes or more with decreased emergency room visits (OR = 0.5, P = 0.046) and hospital stays (OR = 0.6, P = 0.034). This suggests that physical activity promotion among socioeconomically disadvantaged older adults has the potential to provide cost savings. This will not be known, however, until physical activity can be promoted and maintained among these adults.

Original languageEnglish (US)
Pages (from-to)492-501
Number of pages10
JournalPreventive Medicine
Issue number6
StatePublished - Jan 1 2001



  • Health expenditures
  • Health services use
  • Older adults
  • Socioeconomic factors
  • Walking

ASJC Scopus subject areas

  • Medicine(all)

Cite this