Background: Regular exercise is an important secondary prevention strategy proven to slow disease progression, reduce mortality, and increase functional capacity and quality of life in patients with coronary artery disease (CAD). Yet, many patients do not meet current exercise recommendations. Therefore, new interventions are needed to enhance activity levels in this population. To develop more effective interventions, a better understanding of the determinants of exercise over time is needed.
Purpose: The aim of the present research was to investigate key variables that explain exercise behavior in cardiac patients. Specifically, this study tested a social cognitive model for predicting exercise behavior 6 and 12 months after a cardiac-related hospitalization. We examined the relationships between previous exercise, physiological feedback, positive and negative social environments, supportive physical environments, self-efficacy, outcome expectations, and exercise intentions and behavior over time.
Method: A longitudinal cohort design was employed. Patients with documented CAD (N = 801, 75% male, mean age: 61.5 +/- 10.1 years) completed a survey measuring the above variables at hospital discharge and 2, 6, and 12 months later.
Results: Structural equation modeling indicated a good fit of the social-cognitive model to the empirical data at 6 months ([chi]2(df = 191, N = 770) = 559.86, P < .001; CFI = 0.97; RMSEA = 0.05) and 12 months ([chi]2(df = 191, N = 770) = 528.97, P = .001; CFI = 0.97; RMSEA = 0.05). Overall, the model accounted for 22% and 34% of exercise behavior at 6 and 12 months posthospitalization, respectively. Results showed that previous exercise exerted the largest total effect on exercise at both time points ([beta] = .34 and .55, respectively), followed by self-efficacy ([beta] = .24 and .16, respectively). Supportive physical environments ([beta] = .15) and exercise intentions ([beta] = .17) also had significant and direct links to exercise behavior at 6 months. Outcome expectations were influential modifiable predictors for exercise at 12 months ([beta] = .12).
Conclusion: The results indicate that both personal and environmental factors influence exercise outcomes over time. In the months immediately following cardiac care in hospital, interventions should focus on enhancing patients' confidence in their ability to be active despite barriers and help them become more aware of exercise resources in their environment (eg, provide lists of convenient exercise facilities in their neighborhood). As patients become more active over time, self-efficacy and outcome expectations should be targets for intervention. Social support was shown to enhance these personal variables. Finally, considering the influence of previous exercise on future behavior, it may be possible to "fast-track" participants with experience through interventions and make these programs more cost-effective.