TESTING THE EXERCISE AND SELF-ESTEEM MODEL IN OLDER ADULTS

Abstract

The Exercise and Self-Esteem Model (EXSEM; Sonstroem et al., 1992) has proven to be a useful theoretical approach to understanding self-esteem and physical activity relationships in a hierarchical and multidimensional fashion. Specifically, physical activity is theorized to influence global self-esteem through its effect on self-efficacy. Efficacy then influences sub-domains of physical esteem (e.g., conditioning, strength, attractive body) that, in turn, are related to physical self-worth at the domain level. Some support for this model exists but few empirical tests of the model have been made on older adults. PURPOSE To test the structural fit of the EXSEM in a sample (N = 106) of older (M age = 70 yrs) adults at baseline of a randomized controlled trial. METHODS Participants completed a measure of physical activity (PASE; Washburn et al., 1993), a self-efficacy scale, a general measure of self-esteem (Rosenberg, 1965), and the Physical Self-Perception Profile (Fox & Corbin, 1989) at entry into the trial. Structural equation modeling was employed to test the theorized relationships. RESULTS The data fit the proposed EXSEM model very well, χ2 (10) = 15.79, p = .11; Comparative Fit Index = .99; Root Mean Square Error of Approximation = .07. Specifically, participants reporting greater physical activity were more efficacious and this confidence was significantly related to higher perceptions of physical conditioning, strength, and attractive body. In turn, the effects of these latter variables on general levels of self-esteem were mediated by physical self-worth. All paths were significant (p <.05). CONCLUSION These findings support a multidimensional model of exercise self-esteem and provide further evidence for the underlying role that self-efficacy plays in the self-perceptions of older adults. The degree to which the proposed relationships of the EXSEM hold together across the course of an exercise intervention remains to be determined. Funded by grant #AG12113 from the National Institute on Aging