Abstract
PURPOSE: Suboptimal adherence is a major limitation to achieving the benefits of exercise interventions, and our ability to predict and improve adherence is limited. The purpose of this analysis was to identify baseline clinical and demographic characteristics predicting exercise training adherence in the HF-ACTION study cohort. METHODS: Adherence to exercise training, defined by the total duration of exercise performed (min/wk), was evaluated in 1159 participants randomized to the HF-ACTION exercise intervention. More than 50 clinical, demographic, and exercise testing variables were considered in developing a model of the min/wk end point for 1-3 mo (supervised training) and 10-12 mo (home-based training). RESULTS: In the multivariable model for 1-3 mo, younger age, lower income, more severe mitral regurgitation, shorter 6-min walk test distance, lower exercise capacity, and Black or African American race were associated with poorer exercise intervention adherence. No variable accounted for >2% of the variance and the adjusted R2 for the final model was 0.14. Prediction of adherence was similarly limited for 10-12 mo. CONCLUSIONS: Clinical and demographic variables available at the initiation of exercise training provide very limited information for identifying patients with heart failure who are at risk for poor adherence to exercise interventions.