Despite the beneficial effects of exercise on physical and psychosocial health, few older adults engage and maintain regular exercise, and correlates of regular exercise have only been established in healthy older adults. This study was a secondary analysis of a clinical trial of exercise, “The Effects of Low Intensity Exercise on the Functional Health of Acutely Ill Elderly”, to identify predictors of regular exercise among elderly adults recovering from a hospitalization and who were dependent in some activities of daily living. The 77 experimental subjects completed the study and were involved in a 12-week daily muscle strengthening program. Indicators of regular exercise were weekly frequency and duration of exercise, and correlates of exercise were physical performance (muscle strength, gait), psychological factors (depressive symptoms, self-efficacy for mobility) social support, and experience with exercise within the last 6 months. The controls were comorbidity and gender.
Of the experimental subjects, 77.9% exercised at least 3 times per week—a frequency known to benefit health. The mean exercise frequency was 51 sessions (4.25 sessions/week), and 70.1% exercised at a duration known to be associated with health benefits (20 minutes/week). The mean exercise duration was 1166.76 minutes (97.23 minutes/week or 22.88 minutes/session). In logistic regression analysis, 16% of the variance in exercising at a frequency known to benefit health was significantly explained by muscle strength, self-efficacy and previous exercise experience, but only self-efficacy was a significant predictor (OR=1.05; 95% CI = 1.00-1.11). Depressive symptoms and muscle strength explained 13% of the variance in exercising at a duration known to benefit health, and only depressive symptoms was a significant predictor (OR=.73; 95% CI = 0.56-0.94). Physical performance factors did not predict exercise behavior, suggesting that physical capability is not a factor in exercise behavior. Although physical performance factors must be present to do exercise, depressive symptoms and self-efficacy for mobility are more important in the decision of older adults to exercise. Interventions to enhance exercise behavior may be most effective if they reduce depression and increase the confidence of the older adult in their ability to exercise. Self-efficacy and depressive symptoms most strongly predicted exercise behaviors.