ABC | Volume 114, Nº4, April 2020

Original Article Marchese et al. Inspiratory exercise in heart failure Arq Bras Cardiol. 2020; 114(4):656-663 Furthermore, the CO for the different resistive loads increased by 15% with the 60% load and decreased by 3% with the 30% load. It is known that the increase in CO can occur due to an increase in HR alone, only in SV, or both. In our study, the increase in CO in IE with a 60% load occurred mainly due to the significant increase in HR, with a small participation of SV, since this variable also increased, but on a smaller, non-significant scale. On the other hand, when the 30% load is used, the CO had an inverse behavior and presented a small decrease, even with an increase in HR. In this case, what seems to have been decisive for the non-increase in output was the 12.5% drop in SV. Some researchers report that, during exercise in patients with HF, a small increase in SV occurs. Others demonstrate that there is no increase in this variable. 14 In this study, the response to a load of 60% was a small increase of 4.5% and a decrease in IE with a 30% load. The decrease in SV and the increase in HR with a 30% load reported in this study are similar to the hemodynamic repercussions of the Muller maneuver, which also causes negative intrathoracic pressure. Orban et al. 15 studied the hemodynamic effects of the Muller maneuver sustained for 12 seconds in 20 healthy young adults and, among other results, found a decrease in SV and an increase in HR. Hall et al. 16 evaluated the effect of the Muller maneuver sustained for 15 seconds in 8 patients with congestive heart failure and concluded that, during the maneuver, there is an increase in left ventricular afterload and a decrease in systolic volume, but HR did not show significant changes. However, the pressure required to perform the Muller maneuver is around -40 mmHg (-54 cm/H 2 O), and the average load used during IE sessions was -31 cm/H 2 O with 30% load and -61 cm/H 2 O with 60% load. Thus, the load that came closest to the value for performing the Muller maneuver was not the one with a behavior similar to the maneuver, except for the increase in HR. McConnell and Griffiths 17 assessed the acute response of HR, BP and MAP to different loads of inspiratory muscle training (50%, 60%, 70%, 80%, and 90%) in 8 athletes. All loads caused an increase in HR, but only a 60% load caused a sustained increase in MAP, SBP and diastolic blood pressure (DBP). In conclusion, the authors suggest an evidence of a response to the activation of the metaborreflex in this load. The results found by the authors cited above are similar to the findings in this study, where both loads increased HR, but only the load of 60% caused a significant increase in SBP and MAP, which may have occurred due to the activation of the inspiratory metaborreflex. This hypothesis is in agreement with other studies, in which the authors state that the activation of the inspiratory metaborreflex is manifested by the increase in HR and MAP. 18,19 The activation of the metaborreflex by inspiratory muscle work is a factor that contributes to exercise intolerance in patients with HF. During the increase in the respiratory work, there is a redistribution of blood flow from the peripheral muscles to the diaphragm, about 14 to 16% of flow theft of CO, causing an exacerbation of fatigue in peripheral muscles. 20 Corroborating the findings of the present study, Moreno et al. 11 evaluated the effect of respiratory muscle fatigue on oxygenation and perfusion of the intercostal and forearm muscles in patients with HFREF. After inspiratory exercise with a 60% load until fatigue, the authors reported decreased perfusion and oxygenation in both the intercostal muscle and forearm, and suggested that this leads to a reduction in the muscle perfusion reflex of peripheral muscles, activting the inspiratory metaborreflex. However, in the long run, Chiappa et al. 21 demonstrated that 4 weeks of IMT with a 60% load is able to attenuate the inspiratory metaborreflex in patients with HF and muscle weakness. The authors also reported a significant increase in the Borg score, which did not occur in the control group, whose exercise was performed with only 2% of MIP. This result is similar to that of our study, where only the highest load significantly increased Borg's score, in addition to raising the score on the subjective dyspnea scale; however, this last scale was not evaluated in the study by Chiappa et al. 21 High IMT loads (60-70%) are recommended to promote a better effect in patients with HF, while lower loads (20-40%) are indicated for patients with a higher functional class. 3 The present study demonstrated a higher degree of fatigue and dyspnea, as well as greater effects on CHR, during a single session of IE with a 60% load, which corroborates this recommendation and highlights a potential risk for individuals with ischemia and with recent decompensation of HF. Crisafulli et al. 22 were the first to assess the acute hemodynamic response to the activation of the metaborreflex in humans with HF and to compare it to the response of healthy individuals. For this, nine patients with HFREF and nine healthy volunteers were selected. All were submitted to post-exercise ischemia. The hemodynamic response, as in the present study, was assessed by cardiographic impedance. As a result, the authors reported that the increase in SBP was similar in both groups, but the control group obtained an increase in SBP due to the increase in CO; in the group of patients with HF, this increase occurred due to the increase in systemic vascular resistance (SVR). There was also an increase in SV in the group of healthy individuals and a decrease in this variable in patients with HF. The authors suggest that the increase in SVR occurs due to the inability of patients with HF to improve cardiac performance and SV. Correlation between CO and MIP In our study, a moderate correlation between CO and MIP was found. A similar result was found by Nishimura et al. 23 after evaluating 23 patients with HF. However, the correlation found was between cardiac index and MIP. At that time, the authors already suggested that the inspiratory muscles could be dependent on cardiac function. 661