Exercise Self-Efficacy in Older Adults: Social, Affective, and Behavioral Influences Edward McAuley, Ph.D., Gerald J. Jerome, M.S., David X. Marquez, M.S., and Steriani Elavsky, M.S. University of Illinois at Urbana-Champaign Bryan Blissmer, Ph.D. University of Rhode Island ABSTRACT A 6-month randomized controlled trial examined the effect of walking and stretching/toning activity on changes in self-efficacy to overcome barriers and engage in incremental periods of activity in older, formerly sedentary adults (N = 174, M age = 65.5 years). Additionally, we were interested in the extent to which social, affec- tive, and behavioral influences contributed to self-efficacy at the end of the 6-month program. Multiple sample latent growth curve analyses revealed a nonsignificant curvilinear growth pattern for barriers efficacy with increases in efficacy occurring from base- line to 2 months and then declining at 4 and 6 months. In the case of efficacy related to continued activity participation, there was a significant growth pattern demonstrating declines in efficacy over the 4 time points. Structural modeling analyses revealed sig- nificant direct effects of physical activity, affect experienced dur- ing activity, and exercise social support on both types of self-effi- cacy. These relationships were not significantly different between modes of activity. The findings are discussed in terms of the need to target sources of efficacy information prior to program end and the implications that such an approach might have for long-term maintenance of physical activity in older adults. (Ann Behav Med 2003, 25(1):1–7) INTRODUCTION Self-efficacy expectations—individuals’ beliefs relative to their capabilities to execute necessary courses of action (1,2)— have consistently been identified as important determinants of health behavior in older adults. Self-efficacy has been related to adherence to exercise regimens in asymptomatic (3) and clinical (4) populations, recovery from cardiovascular disease (5), and survival from chronic obstructive pulmonary disease (6). In ad- dition, older individuals with low perceptions of personal capa- bilities demonstrate slower gait speed (7) and general declines in physical and social function (8). Moreover, there is evidence to suggest that interventions designed to enhance self-efficacy can effectively reduce attrition from exercise programs (9). However, it is important to realize that self-efficacy effects on physical activity behavior may vary depending on the phase of the exercise process. For example, early studies were able to demonstrate that self-efficacy was related to different types of activity at different stages of the exercise process (10). Further, exercise trials have shown that self-efficacy plays a more promi- nent role in the prediction of exercise behavior in the early adop- tion and adaptation stages of exercise programs but is less potent in the maintenance phases (11,12). However, self-efficacy as- sessed at the end of structured exercise programs has been shown to be a significant determinant of activity, independent of previous exercise participation and fitness, at long-term fol- low-up (13). Clearly, the challenges associated with beginning an exercise regimen and continuing activity beyond program termination offer a greater opportunity for mediation by cogni- tive control than when exercise becomes habituated (2). Although there has been research to document how efficacy changes over the course of a structured program (e.g., 14), typi- cally the measures of efficacy employed reflect perceptions of capabilities for a single bout of activity rather than measures de- signed to predict exercise behavior over prolonged periods of time or in the face of barriers to exercise participation. Indeed, it is quite possible that these latter types of self-efficacy may de- cline in strength prior to program end as participants reevaluate their capabilities in light of program termination. It is therefore important to examine how efficacy expectations change over the course of exercise interventions as well as to identify which fac- tors are important sources for enhancing efficacy at the end of the program. Unfortunately, very little effort has been directed at deter- mining which aspects of the exercise experience influence self-efficacy at this juncture in the exercise process. Bandura (2) identified several important sources of efficacy information that are integral components of any exercise program. Relative to physical activity itself, performance attainment or regular exer- cise participation is likely to act as the most powerful source of efficacy relative to continued participation. In addition, the so- cial network and support that are created within the exercise group are often considered important correlates of exercise be- havior. However, it is clear that social support effects on behav- ior come about through the mediation of self-efficacy (15,16). A further source of self-efficacy, rarely examined in the context of predicting exercise behavior, is the affective experience associ- ated with physical activity participation. Emotional responses to particular situations can serve either to strengthen or to debili- tate perceptions of capabilities (2), which influence subsequent behavior (17). It might also be argued that the social aspects of group exercise may have an indirect effect on efficacy through their influence on affect generated from physical activity partic- ipation. That is, one might experience positive (or negative) af- 1 Funding for this study was provided by the National Institute on Aging (Grant No. AG 12113). Reprint Address: E. McAuley, Ph.D., University of Illinois, Depart- ment of Kinesiology, 215 Freer Hall, Urbana, IL 61801. E-mail: a-mc3@uiuc.edu © 2003 by The Society of Behavioral Medicine.