2.1

Muscle sympathetic nerve activity

Muscle sympathetic nerve activity (MSNA) is a neurophysiological method (microneurography) that allows for recording sympathetic nerve traffic. MSNA is markedly increased in patients with CHD (45 ± 2.3 vs. 31 ± 1 bursts/min; p < 0.001), hypertension (33.3 ± 1.7 vs. 23.9 ± 1.6 bursts/min; p < 0.01 respectively), and CHF (62 ± 4 vs. 39 ± 4 bursts/min; p < 0.01) as compared with healthy subjects, but MSNA can be decreased by ET . The research team from the Heart Institute of University of Sao Paulo has spent more than 10 years investigating the effects of ET on MSNA in CHF and CHD patients . The team demonstrated that regular ET can normalize the basal overactivation of the sympathetic nerve ( Table 1 ). After 4–6 months of ET (3 supervised 60-min exercise sessions/week of cycling and strengthening), baseline values of MSNA (about 45 bursts/min for CHF patients) decreased significantly to within normal values relative to healthy participants (about 30 bursts/min) , with no change in untrained groups. Furthermore, ET had no gender- or age-specific effect on MSNA.

2.2

Heart rate variability (HRV)

HRV is a non-invasive reproducible measure of ANS function corresponding to the balance between sympathetic and parasympathetic effects on the sinoatrial node rate . HRV indexes are highly decreased in CVD patients and predict poorer outcomes, such as reduced left-ventricular function and sudden cardiac death . The risk of all-cause and progressive heart failure death was increased with a standard deviation of normal to normal R-R intervals (SDNN) of <67 ms (relative risk [RR] 2.5; 95% CI 1.5–4.2) . In a retrospective analysis of 1284 CHD patients, SDNN values <70 ms significantly and independently predicted cardiac mortality (RR 3.2; 95% CI 1.6–6.3) . According to Bilchick et al. , each increase of 10 ms in SDNN conferred a 20% decrease in risk of mortality ( p = 0.0001) with an increase in vagal tone and a decrease in sympathetic activity . In a recent randomized controlled, single-blinded trial, Murad et al. included 66 CHF patients (mean age 69 years, New York Heart Association [NYHA] class II–III) with preserved or reduced ejection fraction. After 16 weeks of follow-up, the supervised ET group showed significantly greater increase in both SDNN and root mean square successive difference (a time domain measure of heart period variability) as compared to controls (+15.46 vs. +2.37 ms, p = 0.016, and +17.53 vs. +1.69 ms, p = 0.003, respectively). This finding may indicate a favorable effect of ET on prognosis for CHF patients. Larsen et al. evaluated the correlation between HRV index and survival. For this, 12 CHF patients (mean age 67 years; NYHA class III) underwent a 12-week rehabilitation program. After 87 months of follow-up, survivors and non-survivors showed borderline significant differences in temporal variables of HRV after training (+10.4 ± 8.45 vs. −2.7 ± 10.2 ms, p = 0.053). Only survivors showed a significant increase in SDANN-i after ET (107.9 ± 40.6 vs. 118.3 ± 48.1 ms, p = 0.029). In CHD patients, the results were heterogeneous and the resting HRV index often remained unchanged after the cardiac rehabilitation program. For example, Duru et al. found unchanged basal HRV indexes after 8 weeks of aerobic exercise (at 70% of HR reserve) in 25 patients with post-acute coronary syndrome (post-ACS). La Rovere et al. studied HRV indexes in 22 trained and untrained CHD patients by the head-up tilt test. Frequency domain HRV indexes at rest were not changed after a 4-week training program. Nevertheless, during the head-up tilt test, trained patients showed significantly greater increases in low-frequency power (LFnu) (84 ± 3% vs. 69 ± 5%) and decreases in high-frequency power (HFnu) (7 ± 1% vs. 19 ± 4%) than controls. With orthostatic stress (such as a tilt test), the baroreceptors are stimulated to drive an increase in sympathetic vasoconstrictor outflow and a reduction in vagal tone. Therefore, in post-ACS trained patients, the reflex activity of the autonomic pathways may have been improved. Of note, the HRV index could be improved early in post-ACS patients with ET. In a very short ET program, 5 days, in phase I cardiac rehabilitation, Santos-Hiss et al. highlighted that trained patients in the resting position showed increased HFnu after training (35.9% ± 19.5% to 65.19% ± 25.4%, p = 0.002) and decreased LFnu (58.9% ± 21.4% to 32.5% ± 24.1%, p = 0.024) and LF/HF ratio (3.12 ± 4.0 to 1.0 ± 1.5, p = 0.004), with no changes in controls.

2.3

Arterial baroreflex function

Arterial baroreceptors (located in the aortic arch and carotid sinuses) are the starting point of nervous afferences playing a constant inhibitory role in decreasing SNS activity . Defective arterial baroreflex is well known to contribute to sympathetic overactivity . The studies of La Rovere et al. underlined that arterial baroreflex is a powerful predictor of cardiovascular death, with <3.0 ms/mmHg associated with increased risk of cardiac mortality, by 2.8-fold (95% CI 1.24–6.16) . Several mechanisms could explain this impairment: reduced sensitivity of the baroreceptor afferent fibers mediated in part by reduced expression and activation of the voltage-gated sodium channels and in part by elevated plasma aldosterone content, which reduces afferent discharge sensitivity , and deterioration of the central autonomic pathways that mediate the baroreflex . Baroreflex function is improved with ET in animal studies . In CHF patients, ET seems to improve arterial baroreflex function or prevent its deterioration . In 12 patients with CHF (mean age 58 years, left-ventricular ejection fraction 36%, NYHA class II-III), Pietila et al. studied the effects on baroreflex sensitivity (among other things) of a 6-month ET protocol comprising light-intensity circuit muscle training and aerobic cycling once a day for 6 days/week. At the end of the program, baroreflex sensitivity increased by 74%, from 5.83 ± 0.82 to 10.15 ± 1.66 ms/mmHg ( p < 0.05). The most recent study investigating the effect of ET on the arterial baroreflex control of MSNA (ABR _MSNA ) in CHF was conducted by the Heart Institute of the University of Sao Paulo . The authors studied the impact of 4 months of exercise on the magnitude and latency of the arterial baroreflex response. For this, 26 CHF patients (NYHA class II-III, left-ventricular ejection fraction ≤40%) were randomized to undergo no training or an ET program. The ET protocol was in line with the group’s other studies (4 months, 60 min/day, 3 times/week). The gain and time delay of ABR _MSNA were unchanged with training, and with no training, the baseline values worsened at 4-month follow-up (gain and time delay values of ABR _MSNA before and after the 4-month follow with no training were 3.5 ± 0.7 vs. 1.8 ± 0.2, a.u./mmHg, p = 0.04, and 4.6 ± 0.8 vs. 7.9 ± 1.0 s, p = 0.05, respectively). The authors concluded that the reduced sympathetic nerve activity was mediated by an increase in arterial baroreflex sensitivity but also may be modulated by chemoreflex control and/or ergoreflex control.

In the same way, baroreflex gain was decreased after myocardial infarction. In 2011, Martinez et al. showed that at 3 months after ACS (2 months of ET), baroreflex control increased from 6.0 to 15.6 ms/mmHg with ET. This benefit was even greater at 7 months after ACS (6 months of ET), when baroreflex control in this group was 18.5 ms/mmHg and similar to that in healthy controls (16.4 ms/mmHg). Untrained post-ACS patients did not show any improvement in the baroreflex sensitivity variable during 7 months of follow-up, despite the same clinical management. In this context, ET based on 3 times/week for 6 months restored baroreflex sensitivity in patients with myocardial infarction, for a long-term protective effect of ET in these patients.

2.4

Resting HR and HR recovery

Epidemiological studies have confirmed that an elevated resting HR reflects greater neurohormonal activation and is an independent predictor of cardiovascular and overall mortality in the general population and in patients with CVD . In addition, a high resting HR affects ischemic episodes that may trigger arrhythmias . Lowering the HR to about 60 beats/min with pharmacological treatments reduced overall and cardiovascular-related mortality . ET in patients with CVD appears to be efficient in lowering resting HR and increasing chronotropic reserve . For example, after a 2-month residential rehabilitation program, resting HR decreased by 11 beats/min in a CHF exercise group . HR recovery after a maximal exercise stress test is considered a vagal tone indicator. The increase in HR is first due to withdrawal of parasympathetic activity with lower intensity, then, with moderate and high intensity, is due to a sympathetic activation . Just after peak exercise, HR drops during the first seconds and minutes because of parasympathetic reactivation with the decrease, then sympathetic inactivation . HR recovery is considered an indicator of vagal tone and a strong prognostic factor of cardiovascular events and death in healthy people and CHF patients as well as patients with coronary artery disease . In CHF, HR recovery is altered, with parasympathetic and baroreflex dysfunction . Nonetheless, ET studies reported improvements in HR recovery in patients with CHF and CHD . In 2007, Myers et al. showed that HR recovery was significantly faster in the exercise group from minutes 2 to 6 after a 8-week training program as compared to the sedentary control group (ANOVA main effect in trained subjects 12.6 beats/min, p < 0.001; main effect among controls 2.6 beats/min, p = 0.27; between-group interaction p = 0.005). Resting HR and HR recovery were enhanced with an ET, presumably because of impaired vagal tone . In consequence, with a lower resting HR and a better HR recovery after exercise, HR reserve is increased, for better sympathovagal balance and improved survival for patients with CVD.

2.1

Muscle sympathetic nerve activity

Muscle sympathetic nerve activity (MSNA) is a neurophysiological method (microneurography) that allows for recording sympathetic nerve traffic. MSNA is markedly increased in patients with CHD (45 ± 2.3 vs. 31 ± 1 bursts/min; p < 0.001), hypertension (33.3 ± 1.7 vs. 23.9 ± 1.6 bursts/min; p < 0.01 respectively), and CHF (62 ± 4 vs. 39 ± 4 bursts/min; p < 0.01) as compared with healthy subjects, but MSNA can be decreased by ET . The research team from the Heart Institute of University of Sao Paulo has spent more than 10 years investigating the effects of ET on MSNA in CHF and CHD patients . The team demonstrated that regular ET can normalize the basal overactivation of the sympathetic nerve ( Table 1 ). After 4–6 months of ET (3 supervised 60-min exercise sessions/week of cycling and strengthening), baseline values of MSNA (about 45 bursts/min for CHF patients) decreased significantly to within normal values relative to healthy participants (about 30 bursts/min) , with no change in untrained groups. Furthermore, ET had no gender- or age-specific effect on MSNA.

2.2

Heart rate variability (HRV)

HRV is a non-invasive reproducible measure of ANS function corresponding to the balance between sympathetic and parasympathetic effects on the sinoatrial node rate . HRV indexes are highly decreased in CVD patients and predict poorer outcomes, such as reduced left-ventricular function and sudden cardiac death . The risk of all-cause and progressive heart failure death was increased with a standard deviation of normal to normal R-R intervals (SDNN) of <67 ms (relative risk [RR] 2.5; 95% CI 1.5–4.2) . In a retrospective analysis of 1284 CHD patients, SDNN values <70 ms significantly and independently predicted cardiac mortality (RR 3.2; 95% CI 1.6–6.3) . According to Bilchick et al. , each increase of 10 ms in SDNN conferred a 20% decrease in risk of mortality ( p = 0.0001) with an increase in vagal tone and a decrease in sympathetic activity . In a recent randomized controlled, single-blinded trial, Murad et al. included 66 CHF patients (mean age 69 years, New York Heart Association [NYHA] class II–III) with preserved or reduced ejection fraction. After 16 weeks of follow-up, the supervised ET group showed significantly greater increase in both SDNN and root mean square successive difference (a time domain measure of heart period variability) as compared to controls (+15.46 vs. +2.37 ms, p = 0.016, and +17.53 vs. +1.69 ms, p = 0.003, respectively). This finding may indicate a favorable effect of ET on prognosis for CHF patients. Larsen et al. evaluated the correlation between HRV index and survival. For this, 12 CHF patients (mean age 67 years; NYHA class III) underwent a 12-week rehabilitation program. After 87 months of follow-up, survivors and non-survivors showed borderline significant differences in temporal variables of HRV after training (+10.4 ± 8.45 vs. −2.7 ± 10.2 ms, p = 0.053). Only survivors showed a significant increase in SDANN-i after ET (107.9 ± 40.6 vs. 118.3 ± 48.1 ms, p = 0.029). In CHD patients, the results were heterogeneous and the resting HRV index often remained unchanged after the cardiac rehabilitation program. For example, Duru et al. found unchanged basal HRV indexes after 8 weeks of aerobic exercise (at 70% of HR reserve) in 25 patients with post-acute coronary syndrome (post-ACS). La Rovere et al. studied HRV indexes in 22 trained and untrained CHD patients by the head-up tilt test. Frequency domain HRV indexes at rest were not changed after a 4-week training program. Nevertheless, during the head-up tilt test, trained patients showed significantly greater increases in low-frequency power (LFnu) (84 ± 3% vs. 69 ± 5%) and decreases in high-frequency power (HFnu) (7 ± 1% vs. 19 ± 4%) than controls. With orthostatic stress (such as a tilt test), the baroreceptors are stimulated to drive an increase in sympathetic vasoconstrictor outflow and a reduction in vagal tone. Therefore, in post-ACS trained patients, the reflex activity of the autonomic pathways may have been improved. Of note, the HRV index could be improved early in post-ACS patients with ET. In a very short ET program, 5 days, in phase I cardiac rehabilitation, Santos-Hiss et al. highlighted that trained patients in the resting position showed increased HFnu after training (35.9% ± 19.5% to 65.19% ± 25.4%, p = 0.002) and decreased LFnu (58.9% ± 21.4% to 32.5% ± 24.1%, p = 0.024) and LF/HF ratio (3.12 ± 4.0 to 1.0 ± 1.5, p = 0.004), with no changes in controls.

2.3

Arterial baroreflex function

Arterial baroreceptors (located in the aortic arch and carotid sinuses) are the starting point of nervous afferences playing a constant inhibitory role in decreasing SNS activity . Defective arterial baroreflex is well known to contribute to sympathetic overactivity . The studies of La Rovere et al. underlined that arterial baroreflex is a powerful predictor of cardiovascular death, with <3.0 ms/mmHg associated with increased risk of cardiac mortality, by 2.8-fold (95% CI 1.24–6.16) . Several mechanisms could explain this impairment: reduced sensitivity of the baroreceptor afferent fibers mediated in part by reduced expression and activation of the voltage-gated sodium channels and in part by elevated plasma aldosterone content, which reduces afferent discharge sensitivity , and deterioration of the central autonomic pathways that mediate the baroreflex . Baroreflex function is improved with ET in animal studies . In CHF patients, ET seems to improve arterial baroreflex function or prevent its deterioration . In 12 patients with CHF (mean age 58 years, left-ventricular ejection fraction 36%, NYHA class II-III), Pietila et al. studied the effects on baroreflex sensitivity (among other things) of a 6-month ET protocol comprising light-intensity circuit muscle training and aerobic cycling once a day for 6 days/week. At the end of the program, baroreflex sensitivity increased by 74%, from 5.83 ± 0.82 to 10.15 ± 1.66 ms/mmHg ( p < 0.05). The most recent study investigating the effect of ET on the arterial baroreflex control of MSNA (ABR _MSNA ) in CHF was conducted by the Heart Institute of the University of Sao Paulo . The authors studied the impact of 4 months of exercise on the magnitude and latency of the arterial baroreflex response. For this, 26 CHF patients (NYHA class II-III, left-ventricular ejection fraction ≤40%) were randomized to undergo no training or an ET program. The ET protocol was in line with the group’s other studies (4 months, 60 min/day, 3 times/week). The gain and time delay of ABR _MSNA were unchanged with training, and with no training, the baseline values worsened at 4-month follow-up (gain and time delay values of ABR _MSNA before and after the 4-month follow with no training were 3.5 ± 0.7 vs. 1.8 ± 0.2, a.u./mmHg, p = 0.04, and 4.6 ± 0.8 vs. 7.9 ± 1.0 s, p = 0.05, respectively). The authors concluded that the reduced sympathetic nerve activity was mediated by an increase in arterial baroreflex sensitivity but also may be modulated by chemoreflex control and/or ergoreflex control.

In the same way, baroreflex gain was decreased after myocardial infarction. In 2011, Martinez et al. showed that at 3 months after ACS (2 months of ET), baroreflex control increased from 6.0 to 15.6 ms/mmHg with ET. This benefit was even greater at 7 months after ACS (6 months of ET), when baroreflex control in this group was 18.5 ms/mmHg and similar to that in healthy controls (16.4 ms/mmHg). Untrained post-ACS patients did not show any improvement in the baroreflex sensitivity variable during 7 months of follow-up, despite the same clinical management. In this context, ET based on 3 times/week for 6 months restored baroreflex sensitivity in patients with myocardial infarction, for a long-term protective effect of ET in these patients.

2.4

Resting HR and HR recovery

Epidemiological studies have confirmed that an elevated resting HR reflects greater neurohormonal activation and is an independent predictor of cardiovascular and overall mortality in the general population and in patients with CVD . In addition, a high resting HR affects ischemic episodes that may trigger arrhythmias . Lowering the HR to about 60 beats/min with pharmacological treatments reduced overall and cardiovascular-related mortality . ET in patients with CVD appears to be efficient in lowering resting HR and increasing chronotropic reserve . For example, after a 2-month residential rehabilitation program, resting HR decreased by 11 beats/min in a CHF exercise group . HR recovery after a maximal exercise stress test is considered a vagal tone indicator. The increase in HR is first due to withdrawal of parasympathetic activity with lower intensity, then, with moderate and high intensity, is due to a sympathetic activation . Just after peak exercise, HR drops during the first seconds and minutes because of parasympathetic reactivation with the decrease, then sympathetic inactivation . HR recovery is considered an indicator of vagal tone and a strong prognostic factor of cardiovascular events and death in healthy people and CHF patients as well as patients with coronary artery disease . In CHF, HR recovery is altered, with parasympathetic and baroreflex dysfunction . Nonetheless, ET studies reported improvements in HR recovery in patients with CHF and CHD . In 2007, Myers et al. showed that HR recovery was significantly faster in the exercise group from minutes 2 to 6 after a 8-week training program as compared to the sedentary control group (ANOVA main effect in trained subjects 12.6 beats/min, p < 0.001; main effect among controls 2.6 beats/min, p = 0.27; between-group interaction p = 0.005). Resting HR and HR recovery were enhanced with an ET, presumably because of impaired vagal tone . In consequence, with a lower resting HR and a better HR recovery after exercise, HR reserve is increased, for better sympathovagal balance and improved survival for patients with CVD.

4.1

High-intensity interval training (HIIT)

The latest recommendations suggest interval training (IT) for heart failure patients and CHD . HIIT is defined as repeated short-intensity bouts (e.g., 30–60 s at 90–100% peak exercise capacity) interspersed with a recovery period (30–120 s at 50% peak exercise capacity or passive recovery). HIIT may be more effective than moderate intensity and continuous exercise (MICE) for improving exercise capacity, quality of life, maximal oxygen consumption (VO _2max ) and cardiac remodeling in CHD and CHF patients . In a recent review of HIIT in cardiac rehabilitation programs (phase II and III), Gayda et al. discussed the interest of developing progressive models of ET based on HIIT combined with other forms of exercise sessions such as MICE, resistance training or inspiratory muscle training.

In healthy subjects, ANS responses differ by exercise modality (duration, frequency, intensity, recovery, volume). Therefore, our team studied an optimized and safe HIIT model . In 2013 , we studied the effect of one session of HIIT on ANS activity while ensuring security and patient comfort . Our hypothesis was that HIIT would enhance vagal tone and thus reduce the likelihood of arrhythmic events in CHF. Eighteen CHF patients underwent a baseline assessment (control condition) and were randomized to a single session of HIIT (repeated 30 s of exercise at 100% peak power alternating with phases of 30 s of passive recovery) and to isocaloric MICE. As compared to control and MICE conditions, a single session of HIIT significantly increased parasympathetic tone. The normalized HF power measured by 24-h electrocardiography for the 3 conditions was 31.56%, 24.61% and 35.95%, respectively ( p < 0.01). After HIIT, the number of premature ventricular contractions decreased significantly (531 vs. 1007 and 1671 for control and MICE, respectively, p < 0.01). We found a correlation between changes in premature ventricular contraction and LF/HF ratio ( r = 0.66, p < 0.01) in patients exposed to HIIT. Passive recovery with each 30 s may have led to “a vagal training stimulation” and the beneficial effects on HRV and PVC are probably related to this specific mode of HIIT, which suggests sympathovagal balance resetting in the postexercise period. The clinical importance of premature ventricular contractions (and HRV) in CHF is a powerful predictor of cardiovascular mortality . Furthermore, according to patients, HIIT was the most preferred protocol, associated with lower perceived exertion as compared with MICE, and no adverse event was reported by health care providers or patients. Other studies reported improvements in HRV after HIIT in cardiac patients and in patients with type 2 diabetes mellitus . However, the study of Currie et al. reported no improvement in HRV index and HR recovery in CHD patients after a 12-week ET of HIIT or MICE despite an increase in VO _2peak in the 2 groups (+20% p < 0.001, with no difference between them). The authors suggested that the length of recovery between ACS and the beginning of training (5–6 months), the optimal medical management, and the normative baseline values contributed to this lack of significant change in autonomic nervous activity.

These results highlight the positive benefit–risk ratio associated with HIIT and suggest that this type of intervention could reduce the cardiovascular risk. Nevertheless, the long-term effects of HIIT remain to be studied.

4.2

Breathing exercises or relaxation

Respiratory sinus arrhythmia–biofeedback or heart rate variability–biofeedback involves the lowering of the breathing rate to the frequency at which the amplitude of HRV is maximized. This breathing exercise stimulates the baroreceptor , thereby modulating the sinus rhythm to enhance sympathovagal balance and cardiovascular risk factors , including in CVD . In the Bernardi et al. study , 81 patients with CHF and 21 healthy controls underwent electrocardiography, respiration, and blood pressure measurement during 5 min of spontaneous breathing, 4 min of controlled breathing at 15 breaths/min (corresponding to spontaneous breathing) and 4 min of controlled breathing at 6 breaths/min. The slow breathing rate in the CHF group increased the mean RR interval to 20 ms, decreased both systolic and diastolic blood pressure (systolic, from 117 to 110 mmHg, p < 0.009; diastolic, from 62 to 59 mmHg, p < 0.02) and significantly increased the baroreflex sensitivity (from 5.0 to 6.1 ms/mmHg, p < 0.0025). The recent first systematic review of the effect of relaxation and meditation on symptom management strategies in CHF found symptom-related quality of life improved with this approach, as well as pain, dyspnea, fatigue, and sleep disturbance.

Breathing techniques, relaxation exercises or some types of meditation have a favorable impact on parasympathetic and sympathetic activity (for review ), with increased HF power and decreased LF/HF ratio. In the Curiati et al. study , 19 older patients with optimally treated CHF were randomized into 2 groups: a meditation group (who listened to a 30-min audiotape twice a day for 12 weeks and attended a weekly meeting) or a control group (who just attended a weekly meeting). Norepinephrine level was reduced in the meditation group alone (from 677.7 to 387.1 pg/mL, p = 0.008), which suggests reduced SNS activity, and were unchanged in the control group.

4.3

Transcutaneous electrical nerve stimulation (TENS)

TENS is classically used to stimulate large myelinated afferent fibers, used in pain treatment, whereas neuromuscular electrical stimulation (NMES) stimulates efferent fibers, causing muscular contraction. For the first time, our group provided evidence that TENS could directly reduce sympathetic activity (measured by MSNA) in patients with CHF in a randomized, sham-controlled, double-blind study (EMSICA Study ). We recruited 22 CHF patients (NYHA class III): 11 underwent TENS, and 11 NMES. Each of the protocols was cross-over, randomized and sham-controlled. MSNA was recorded immediately after electrical stimulation cessation on the opposite stimulated limb. Compared to sham stimulation, both TENS and NMES reduced MSNA (69.7 vs. 63.5 bursts/min, p < 0.01 after TENS and 56.7 vs. 51.6 bursts/min, p < 0.01 after NMES). These findings highlight the clinical importance of this non-pharmacological therapy based on ET for long-term treatment of patients with myocardial infarction. Reduced MSNA by TENS could be attributable to enhanced spontaneous baroreflex sensitivity; As well, the ergoreflex induced by TENS could stimulate the nucleus tractus solitarius and the release of substance P, which could interact with baroreflex sensitivity and decreased sympathetic outflow .