Patients with heart failure (HF) have a reduced exercise tolerance as the main result of the disease. This exercise intolerance is due to heart conditions, but also to dysfunction of the respiratory and peripheral muscles. Various factors such as chronic hypoxia, oxidative stress, nutritional depletion, peripheral muscle disuse, effects of medications, and sympathetic-vagal imbalance are major contributors to deconditioning. In this scenario, the use of non-invasive ventilatory support (NIV) arises as an adjunct to cardiac rehabilitation in the attempt to improve the functional capacity of patients, since NIV reduces work of breathing, improves oxygenation and increases lung compliance associated with improved ejection volume, due to increased intrathoracic pressure.

The main objective is to evaluate whether the use of NIV during Constant Treadmill Load Tests (CTLT) enhances exercise performance in Cardiac Heart Failure (CHF) patients on a treadmill.

Each recruited patient will be evaluated on 4 occasions, with two different evaluations, an incremental test (IT) to find out the maximum aerobic speed reached and then, in randomized order, three CTLT at 85% of the maximum reached in the IT . Two of the three CTLT will be performed with the use of NIV, with PS (IPAP 10cmH2O and EPAP 5cmH2O) and CPAP (5cmH2O) mode, and the rest will be without the use of NIV. The main study variable will be the time tolerated on the treadmill during CTLT.

Participants will have 5 minutes to adapt to NIV, 5-minute warm-up at 50% and then will start with the CTLT at 85% of the maximum aerobic speed reached in the IT.

In total, each patient will be evaluated four times on four different visits.

Sample size: For the sample's calculation, the maximum time in CTLC was chosen as the main study variable because it is a sensitive measure to detect changes after interventions (both pharmacological and non-pharmacological) and widely used in clinical trials. A minimum difference to detect of 100 seconds from previous literature was determined. Taking this data into account, and using a level of α of 0.05, a 1-β power of 0.80, and an effect size of 0.8, the sample size calculation for a paired t-test investigators found that at least 12 participants are needed. The calculation was made using the Gpower 3.1.9.7 software.

Inclusion Criteria:

• Ejection fraction demonstrated by echocardiography ≤ 40% (in the last 12 months).
• NYHA functional class I - III.
• Clinical stability, defined as four weeks prior to the start of the study without hospitalizations.
• Signed informed consent.
• ≥ 18 years of age.

Exclusion Criteria:

• Chronic obstructive pulmonary disease (COPD), unstable angina or significant arrhythmias.
• Myocardial infarction in the last 3 months, primary valve disease and anemia (hemoglobin under 13 g/dl for men or 12 g/dl for women).
• Patients who are smokers or ex-smokers of less than one year.
• Patients with cognitive impairment that prevents the correct understanding of the evaluations.
• Any neuromuscular or osteoarticular condition that limits test performance.
• Refusal to participate.