Exercise-based cardiac rehabilitation for heart failure


People with heart failure (HF) experience fatigue and shortness of breath. This negatively affects their activities of daily living and health-related quality of life. They are at increased risk of hospital admission and death.

Study characteristics

We searched the scientific literature for randomised controlled trials (experiments in which two or more interventions, possibly including a control intervention or no intervention, are compared by randomly allocating participants to study groups). We looked at the effectiveness of exercise-based rehabilitation compared with no exercise in adults (over 18 years of age) with heart failure. We considered HF due to reduced ejection fraction (HFrEF) (i.e. the chambers of the heart contract poorly, and, as a result, a smaller volume of blood is pumped around the body). We also considered HF due to preserved ejection fraction (HFpEF) (i.e. the chambers of the heart contract normally but do not relax efficiently, resulting in a smaller volume of blood pumped around the body). Our search is current to January 2018.

Key results

We found 44 studies that included 5783 people with HF, mainly HFrEF. The findings of this update are broadly consistent with those of the previous (2014) version of this Cochrane Review. They show important benefits of exercise-based rehabilitation that include a probable reduction in the risk of overall hospital admissions in the short term, as well as the potential for reduction in heart failure admissions. The effect of exercise-based rehabilitation on health-related quality of life is uncertain due to very low-quality evidence. Exercise-based rehabilitation may make little or no difference in all-cause mortality in trials with follow-up less than 12 months. Further evidence is needed to better show the effects of exercise rehabilitation among people with HFpEF and the impact of alternative models of delivery, such as home-based programmes.

Quality of evidence

Generally, recent trials have been better reported and are at low to moderate risk of bias. Using the GRADE method, we assessed the quality of evidence to range from high to very low across measured outcomes. Common reasons for downgrading outcomes include that results were inconsistent and/or imprecise.

Authors' conclusions: 

This updated Cochrane Review provides additional randomised evidence (11 trials) to support the conclusions of the previous version (2014) of this Cochane Review. Compared to no exercise control, CR appears to have no impact on mortality in the short term (< 12 months' follow-up). Low- to moderate-quality evidence shows that CR probably reduces the risk of all-cause hospital admissions and may reduce HF-specific hospital admissions in the short term (up to 12 months). CR may confer a clinically important improvement in health-related quality of life, although we remain uncertain about this because the evidence is of low quality. Future ExCR trials need to continue to consider the recruitment of traditionally less represented HF patient groups including older, female, and HFpEF patients, and alternative CR delivery settings including home- and using technology-based programmes.

Read the full abstract...

Chronic heart failure (HF) is a growing global health challenge. People with HF experience substantial burden that includes low exercise tolerance, poor health-related quality of life (HRQoL), increased risk of mortality and hospital admission, and high healthcare costs. The previous (2014) Cochrane systematic review reported that exercise-based cardiac rehabilitation (CR) compared to no exercise control shows improvement in HRQoL and hospital admission among people with HF, as well as possible reduction in mortality over the longer term, and that these reductions appear to be consistent across patient and programme characteristics. Limitations noted by the authors of this previous Cochrane Review include the following: (1) most trials were undertaken in patients with HF with reduced (< 45%) ejection fraction (HFrEF), and women, older people, and those with preserved (≥ 45%) ejection fraction HF (HFpEF) were under-represented; and (2) most trials were undertaken in the hospital/centre-based setting.


To determine the effects of exercise-based cardiac rehabilitation on mortality, hospital admission, and health-related quality of life of people with heart failure.

Search strategy: 

We searched the Cochrane Central Register of Controlled Trials (CENTRAL), MEDLINE, Embase, and three other databases on 29 January 2018. We also checked the bibliographies of systematic reviews and two trial registers.

Selection criteria: 

We included randomised controlled trials that compared exercise-based CR interventions with six months' or longer follow-up versus a no exercise control that could include usual medical care. The study population comprised adults (> 18 years) with evidence of HF - either HFrEF or HFpEF.

Data collection and analysis: 

Two review authors independently screened all identified references and rejected those that were clearly ineligible for inclusion in the review. We obtained full papers of potentially relevant trials. Two review authors independently extracted data from the included trials, assessed their risk of bias, and performed GRADE analyses.

Main results: 

We included 44 trials (5783 participants with HF) with a median of six months' follow-up. For this latest update, we identified 11 new trials (N = 1040), in addition to the previously identified 33 trials. Although the evidence base includes predominantly patients with HFrEF with New York Heart Association classes II and III receiving centre-based exercise-based CR programmes, a growing body of studies include patients with HFpEF and are undertaken in a home-based setting. All included studies included a no formal exercise training intervention comparator. However, a wide range of comparators were seen across studies that included active intervention (i.e. education, psychological intervention) or usual medical care alone. The overall risk of bias of included trials was low or unclear, and we downgraded results using the GRADE tool for all but one outcome.

Cardiac rehabilitation may make little or no difference in all-cause mortality over the short term (≤ one year of follow-up) (27 trials, 28 comparisons (2596 participants): intervention 67/1302 (5.1%) vs control 75/1294 (5.8%); risk ratio (RR) 0.89, 95% confidence interval (CI) 0.66 to 1.21; low-quality GRADE evidence) but may improve all-cause mortality in the long term (> 12 months follow up) (6 trials/comparisons (2845 participants): intervention 244/1418 (17.2%) vs control 280/1427 (19.6%) events): RR 0.88, 95% CI 0.75 to 1.02; high-quality evidence). Researchers provided no data on deaths due to HF. CR probably reduces overall hospital admissions in the short term (up to one year of follow-up) (21 trials, 21 comparisons (2182 participants): (intervention 180/1093 (16.5%) vs control 258/1089 (23.7%); RR 0.70, 95% CI 0.60 to 0.83; moderate-quality evidence, number needed to treat: 14) and may reduce HF-specific hospitalisation (14 trials, 15 comparisons (1114 participants): (intervention 40/562 (7.1%) vs control 61/552 (11.1%) RR 0.59, 95% CI 0.42 to 0.84; low-quality evidence, number needed to treat: 25). After CR, a clinically important improvement in short-term disease-specific health-related quality of life may be evident (Minnesota Living With Heart Failure questionnaire - 17 trials, 18 comparisons (1995 participants): mean difference (MD) -7.11 points, 95% CI -10.49 to -3.73; low-quality evidence). Pooling across all studies, regardless of the HRQoL measure used, shows there may be clinically important improvement with exercise (26 trials, 29 comparisons (3833 participants); standardised mean difference (SMD) -0.60, 95% CI -0.82 to -0.39; I² = 87%; Chi² = 215.03; low-quality evidence). ExCR effects appeared to be consistent different models of ExCR delivery: centre vs. home-based, exercise dose, exercise only vs. comprehensive programmes, and aerobic training alone vs aerobic plus resistance programmes.