Long-Term Exercise Training in Heart Failure

Abstract & Commentary

By Michael H. Crawford, MD, Editor

Sources: Belardinelli R, et al. 10-year exercise training in chronic heart failure: A randomized controlled trial. J Am Coll Cardiol 2012;60:1521-1528. Whellan D. Long-term exercise training and adherence: It is not just exercise. J Am Coll Cardiol 2012;60:1529-1530.

Exercise training is associated with short-term improvements in functional capacity in heart failure patients, but its effect on mortality and heart failure readmissions have been mixed. Thus, these investigators from Italy and New York studied 135 stable heart failure patients who were divided into a supervised exercise training group (70% peak oxygen consumption, two times a week for 10 years) and a non-trained group. Inclusion criteria included left ventricular ejection fraction < 40% and the ability to exercise. The etiology of heart failure was ischemic in 80% and their average age was about 60 years at intake. Each patient underwent a formal re-evaluation with cardiopulmonary exercise testing every 12 months by observers blinded to the study groups. The primary outcomes were peak oxygen consumption, quality of life, mortality, heart failure exacerbation, and cardiac ischemic events. Any of the cardiovascular events ended the patients’ participation in the study. Of the 135 enrolled, 123 completed the 10-year study (63 trained and 60 non-trained). Peak oxygen consumption was not different between the two groups at intake, but increased in the training group by 15% and decreased 2.5% in the non-trained group at 1 year. This difference persisted during the study and was mirrored by a slower resting heart rate in the training group. Ejection fraction was not different at intake between the two groups, but after 4 years was significantly higher in the training group (41% vs 34%, P < 0.01 at 5 years). Quality of life also improved significantly in the training group and was sustained for the duration. The safety of training was excellent. Clinical events were less common in the training group (12 vs 35; hazard ratio [HR] = 0.55, 95% confidence interval [CI], 0.26-0.72; P < 0.0001). Specifically, cardiac death was less common in the training group (4 vs 10; HR = 0.68, 95% CI 0.30-0.82; P < 0.001). Multivariate analysis showed that peak oxygen consumption and resting heart rate were the only independent predictors of events. The authors concluded that moderate, supervised exercise training performed twice a week for 10 years conferred sustained improvement in exercise tolerance, quality of life, and left ventricular systolic performance in patients with heart failure due to systolic dysfunction. These improvements were associated with lower cardiac morbidity and mortality rates.


The results of this study are remarkable and very different from the recently reported Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training (HF-ACTION) trial, which showed no reduction in cardiac mortality with exercise training, despite a small but significant increase in peak oxygen consumption in the training group.1 There are several differences between the two studies that may explain the results. First, this trial involved exercise training sessions supervised by a cardiologist, whereas HF-ACTION involved self-directed home-based exercise. Adherence to the program was 88% in this trial vs about 60% in HF-ACTION. Second, peak oxygen consumption increased 4% in HF-ACTION and 15% the first year in this study. Third, HF-Action was a 2.5 year study vs 10 years in this study. Some parameters, such as ejection fraction, took 5 years to improve in this study. Fourth, there may have been crossovers to supervised exercise training in HF-ACTION. Fifth, there were differences in medical and device therapy in the two studies. Beta-blocker use was 94% in HF-ACTION and 46% in this study. ICD use was 40% in HF-ACTION and 7% in this study. The authors focus on the supervision aspect as the key difference between the studies. The accompanying editorial suggests that the community aspects of the training sessions, which also included education, may have played a large role in the favorable outcomes.

Importantly, there is no way to be certain that the reduction in events was causally related to the exercise training. The study was underpowered for outcomes assessment; it was powered for changes in peak oxygen consumption. However, the only independent predictors of the outcomes were related to exercise training: peak oxygen consumption and resting heart rate. It is noteworthy that HF-ACTION had 2331 subjects vs 123 in this study. So perhaps the reduction in outcomes in this study was an alpha or type I statistical error, which is more common in small studies. Clearly, a larger trial would have to be done to confirm the outcome results.

Despite these weaknesses, this is a compelling study. Exercise training was safe and seemed highly effective. So why not adopt it? The feasibility and cost of such a program in the united States for almost all heart-failure patients would not be favorable under our current health care system. However, we can certainly refer most patients to cardiac rehabilitation programs and encourage them to continue these programs on their own if possible. Also, it may be worth encouraging community groups to continue this effort if health care agencies can’t fund it. At a minimum, it uniformly improves exercise performance and quality of life.


1. O’Connor CM, et al. Efficacy and safety of exercise training in patients with chronic heart failure: HF-ACTION randomized controlled trial. JAMA 2009;301:1439-1450.