Social status, physical environment and walking for exercise in late life: An ecological model.

Abstract

Most studies of exercise behavior among the elderly focus predominately on the micro- or individual-level. Specifically, many researchers have examined the impact of education and economic well-being on exercise without considering community-level factors. These broader community or environmental factors, in turn, also affect individual-level characteristics such as walking as a source of exercise. This study uses cross-sectional data collected from 746 65 to 99 year old U.S. residents to examine walking as a source of exercise in the context of educational, economic, and physical environmental resources. Latent variable modeling using maximum likelihood estimation and hierarchical logistic regression were used to estimate the interrelationships among demographic characteristics, family income, financial strain, perceived risk of neighborhood crime, neighborhood deterioration, and walking behavior. This research supported the following theoretical sequence as suggested by a modified risk interpretation framework (Ferraro, 1995): (1) low educational attainment led to lower economic well-being; (2) elderly people who had economic problems were more likely to live in deteriorated neighborhoods; (3) older women who lived in deteriorated neighborhoods, in turn, perceived greater risk of crime in their neighborhoods; and (4) older women with a heightened perception of neighborhood risk were less likely to walk for exercise. Although gender differences were predicted for each link in the conceptual model, older men and women appear to differ only in how resources affected walking for exercise. Specifically, deteriorated neighborhood conditions were more likely to heighten women's perceptions of neighborhood risk rather than men's perceptions, and women who believed there is a higher risk of crime were far more likely than men to not walk for exercise. In addition, low educational attainment and economic problems reduced the likelihood that older women, but not older men, walked for exercise. Substantial support was found for the interplay between individual- and physical environmental-level characteristics in predicting the likelihood of walking as a source of exercise among the elderly, which suggests a two-tiered approach to exercise promotion is critical.