Adaptations in autonomic function during exercise training in heart failure
- First Online:
- Cite this article as:
- Negrao, C.E. & Middlekauff, H.R. Heart Fail Rev (2008) 13: 51. doi:10.1007/s10741-007-9057-7
- 401 Views
Although neurohumoral excitation is the hallmark of heart failure (HF), the mechanisms underlying this alteration are not entirely known. Abnormalities in several systems contribute to neurohumoral excitation in HF, including arterial and cardiopulmonary baroreceptors, central and peripheral chemoreceptors, cardiac chemoreceptors, and central nervous system abnormalities. Exercise intolerance is characteristic of chronic HF, and growing evidence strongly suggests that exercise limitation in patients with chronic HF is not due to elevated filling pressures or inadequate cardiac output during exercise, but instead due to skeletal myopathy. Several lines of evidence suggest that sympathetic excitation contributes to the skeletal myopathy of HF, since sympathetic activity mediates vasoconstriction at rest and during exercise likely restrains muscle blood flow, arteriolar dilatation, and capillary recruitment, leading to underperfused areas of working muscle, and areas of muscle ischemia, release of reactive oxygen species (ROS), and inflammation. Although controversial, either unmyelinated, metabolite-sensitive afferent fibers, and/or myelinated, mechanosensitive afferent fibers in skeletal muscle underlie the exaggerated sympathetic activity in HF. Exercise training has emerged as a unique non-pharmacological strategy for the treatment of HF. Regular exercise improves functional capacity and quality of life, and perhaps prognosis in chronic HF patients. Recent studies have provided convincing evidence that these benefits in chronic HF patients are mediated by significant reduction in central sympathetic outflow as a consequence of improvement in arterial and chemoreflex controls, and correction of central nervous system abnormalities, and increase in peripheral blood flow with reduction in cytokines and increase in mass muscle.
KeywordsHeart failureSympathetic excitationAutonomic reflex controlExercise intoleranceExercise training
Concepts of heart failure
Despite the improvement of the pharmacologic treatment of heart failure (HF) in the last decades, this syndrome is one of the major causes of death in industrialized countries. In the United States, the rate of (HF) reaches approximately 1.5–2% of the total population and as much as 6–10% of the elderly . This situation is more critical since as many as 20 million additional people who have asymptomatic cardiac dysfunction are likely to become symptomatic in 1–5 years. Moreover, the prevalence of HF will rise as the mean age of the population increases. Thus, the development of new strategies for the treatment of HF is among the great challenges facing medicine in today.
Abnormal autonomic neural regulation
Heart failure is characterized by neurohumoral excitation. Neurohumoral excitation, which initially helps stabilize the individual with systolic dysfunction, becomes deleterious as cardiac dysfunction persists. Thus, with time, neurohumoral excitation leads to worsening of the clinical syndrome of cardiac failure, and is directly associated with poor prognosis of the HF patients.
Sympathetic hyperactivity is typical of HF. Plasma norepinephrine levels are significantly increased in HF , which contributes to the vasoconstrictor state in HF. Cardiac norepinephrine spillover is significantly augmented in HF patients when compared with healthy individuals . In contrast, parasympathetic activity is significantly reduced in chronic HF patients. In addition, loss of parasympathetic heart rate modulation observed in the time and frequency domain of cross-spectral analysis have been found in humans with HF. Direct measures of sympathetic nerve activity by means of a microneurography technique have shown that muscle sympathetic nerve activity progressively increases across the spectrum of normal controls to moderate and severe HF patients .
Angiotensin II, aldosterone, and vasopressin all contribute to salt and water retention in HF. Angiotensin II has been implicated in the progression of compensated and decompensated HF. More recently, it has been shown that aldosterone plays an important role in HF. It causes sodium retention and loss of potassium and, in addition, is implicated in myocardial fibrosis . Vasopressin provokes fluid retention, which aggravates the cardiac congestion and the worsening of HF .
Mechanisms of neurohumoral excitation
The mechanisms of neurohumoral excitation in HF are not entirely certain. It is likely that abnormalities in several systems contribute to neurohumoral excitation in HF, including arterial and cardiopulmonary baroreceptors, central and peripheral chemoreceptors, cardiac chemoreceptors, and central nervous system abnormalities. Impaired arterial and cardiopulmonary baroreflex control has been reported in animals and humans with chronic HF [7–9]. Ferguson and colleagues provided important evidence for the impairment in baroreflex control of heart rate and muscle sympathetic nerve activity in patients with chronic HF . More recently, Al-Hesayen and Parker  demonstrated that renal sympathetic activity responses to sodium nitroprusside, which are mediated by the arterial baroreceptors, were blunted in chronic HF patients.
Cardiopulmonary baroreflex control of sympathetic nerve activity is also blunted in patients with HF. The reflex forearm vasoconstrictor responses to phlebotomy are attenuated in HF patients . In addition, marked attenuation in reflex muscle sympathetic nerve activity responses to stimuli of cardiac filling pressure have been reported in HF patients . Interestingly, skin sympathetic nerve activity, which is not governed by the baroreceptors, is not elevated in HF patients in whom muscle sympathetic nerve activity is increased. This finding further supports the notion that the baroreceptors play an important role in mediating the sympathetic excitation in HF .
Results of animal and human studies have shown that the increase in sympathetic activity in HF may in part be mediated by chemoreceptor hypersensitivity [14, 15]. Hypercapnia associated with hyperoxia causes exaggerated increase in minute ventilation and muscle sympathetic nerve activity in patients with HF [14, 16, 17]. Similarly, pulmonary ventilation and muscle sympathetic nerve activity during hypoxia and isocapnia are exaggerated in chronic HF patients [17, 18]. Di Vanna et al. reported exaggerated sympathetic nerve activity and blunted muscle vasodilatation responses to hypercapnia or hypoxia in humans with chronic HF . An interaction between enhanced chemoreflexes and blunted arterial baroreflex control has been suggested, since carotid body denervation and suppression of chemoreflex control restores arterial baroreflex control in animals and humans with chronic HF, respectively [18, 19].
Another source of sympathoexcitation in HF may be enhanced cardiac reflex control. This reflex consists of fibers in the myocardium that are activated by chemical substances, such as bradykinin, prostaglandins, potassium, and lactate [20–22]. Further studies demonstrated that application of capsaicin and electrical stimulation of the left ventricular epicardial impairs arterial baroreflex control in rats . These investigators also showed that application of losartan in the intracerebroventricular prevents the cardiac sympathetic afferents from inhibiting baroreflex control. These latter findings suggest that the increase in sympathetic outflow in HF may be mediated by cardiac reflex enhancement, mediated at least in part through activation of angiotensin AT1 receptors.
Abnormalities in the central nervous system have been described in chronic heart failure . Angiotensin II, via stimulation of AT1 receptors, located in the paraventricular nucleus of hypothalamus, may reduce arterial and cardiopulmonary reflexes in chronic heart failure. Angiotensin II increases excitatory reflex sensitivity as chemoreflex, cardiac sympathetic afferent reflex, and somatic reflex. The same investigators provided evidences for normalization of these alterations by blockade of the AT1 receptors . In addition, angiotensin II and AT1 receptors operate in concert with nitric oxide, since blockade of nitric oxide synthase increases sympathetic nerve activity only after increasing plasma angiotensin II levels .
Acute exercise responses in heart failure: autonomic implications
Exercise intolerance is characteristic of chronic HF, and our understanding of the mechanisms underlying this exercise dysfunction is evolving. Studies in chronic HF have shown that resting ejection fraction does not correlate with exercise tolerance, including symptoms of fatigue and exhaustion, or peak oxygen consumption [27–29]. It has been proposed that impaired cardiac reserve, rather than resting cardiac function, may be more directly correlated with impaired exercise ability in HF. However, even intracardiac hemodynamic abnormalities during exercise do not correlate with measures of exercise dysfunction [27–29]. Additionally, when cardiac output and intracardiac filling pressures are acutely improved pharmacologically, exercise performance does not acutely improve . Thus, the concept that exercise limitation in patients with chronic HF is only due to elevated filling pressures or inadequate cardiac output during exercise has been revised to acknowledge the central role of the skeletal myopathy of chronic HF [31, 32].
Histological and metabolic changes in skeletal muscle
Contribution of sympathetic excitation for the skeletal myopathy
Several lines of evidence suggest that sympathetic excitation contributes to the skeletal myopathy of HF. First, sympathetic activation contributes to the chronic vasoconstriction in patients with HF, both at rest, and during exercise [4, 45–48].
Exaggerated sympathetic activation during exercise likely restrains muscle blood flow, arteriolar dilatation, and capillary recruitment, leading to underperfused areas of working muscle, and localized areas of muscle ischemia, release of reactive oxygen species (ROS), and inflammation. During exercise in healthy humans, arterioles dilate, thereby opening additional capillary beds and shortening the diffusion distance that oxygen and substrate must travel to meet the increased energy requirements in exercising muscle. Increased resting sympathetic nerve activity directed to muscle in patients with HF leads to increased muscle vasoconstriction, and decreased muscle—specifically capillary—blood flow. In healthy humans during exercise, sympathetic activation during exercise is offset by the generation of metabolic byproducts that produce vasodilatation. However, when sympathetic activation is exaggerated, as in the case of HF, the balance swings in favor of vasoconstriction [4, 45–48].
Studies in our laboratories have consistently demonstrated neurovascular alteration during physiological maneuvers in HF. Muscle sympathetic nerve activity during exercise and mental challenge is significantly greater in HF patients when compared with healthy individuals. In contrast, muscle blood flow responses to mental challenge and exercise are blunted in chronic HF patients. The vasoconstrictor state in HF during physiological maneuvers can be also observed in other vascular beds; renal cortical vascular resistance is exaggerated during mental challenge in HF patients when compared with age-matched healthy individuals. This neurovascular alteration is more prominent as the severity of cardiac dysfunction increases. The question that emerges from these findings is whether the sympathetic nerve activity is, in fact, implicated in the muscle blood flow regulation during exercise in HF. Investigations from our laboratory provide relevant evidence on this matter. Resting forearm blood flow significantly improves after brachial intra-arterial infusion of phentolamine, an agent which blocks the alpha receptors. In addition, forearm blood flow responses to exercise and mental challenge are substantially increased with phentolamine administration.
Sympathetic activation may also trigger skeletal muscle inflammation, further contributing to the skeletal myopathy. Sympathetic activation may initiate inflammation through oxidative stress and the generation of ROS. Oxidative stress is known to activate nuclear factor-kappaBeta, a necessary transcription factor for cytokine gene expression . Release of tumor necrosis factor (TNF)-alpha and interleukin (IL)-6, has been linked to increased skeletal muscle catabolism, and apoptosis [50–52]. In experimental models, increased levels of TNF-alpha are correlated with the degree of apoptosis in skeletal muscle in HF . Interleukin-6 levels are inversely correlated to muscle fiber thickness in HF patients . These findings are consistent with a role for inflammation in the pathogenesis of the skeletal myopathy.
Mechanisms underlying sympathetic overactivity during exercise
What causes the exaggerated sympathetic activity during exercise in chronic HF? Although controversial, investigators have proposed that either unmyelinated, metabolite-sensitive afferent fibers, and/or myelinated, mechanosensitive afferent fibers in skeletal muscle underlie the exaggerated sympathetic activity in HF . In support of the concept that muscle metaboreceptor control of sympathetic nerve activity is impaired in HF is the observation that during post-exercise circulatory arrest, when muscle metabolites are trapped in the muscle, reflex-mediated muscle sympathetic nerve activity responses are blunted in HF patients. This reduction in muscle metaboreflex control is related to severity of cardiac dysfunction. The attenuation in muscle sympathetic nerve activity during post-exercise muscle circulatory arrest is progressively increased from healthy individuals to moderate and severe HF patients. It has become evident that the sympathetic hyperactivity during exercise in HF is mediated by muscle mechanoreceptors. Two pieces of information support this concept. Low level exercise, which preferentially stimulates muscle mechanoreceptors over metaboreceptors, increases muscle sympathetic nerve activity within the first minute of exercise in HF patients, but only in the third minute of exercise in healthy individuals . Passive exercise, when mechanoreceptors are isolated from the central command, muscle sympathetic nerve activity increases significantly in HF patients, but not in healthy individuals .
Abnormal vasodilatation to a number of stimuli has been reported in HF. Previous studies have consistently demonstrated that chronic HF patients have blunted muscle vasodilatation in response to hyperemia and exercise [55, 56]. More recently, it has become clear that the blunted muscle vasodilatory responses during exercise in chronic HF are endothelially-mediated. Intra-arterial infusion of acetylcholine, an endothelial-dependent vasodilator, does not increase forearm blood flow to the same levels in healthy individuals and in chronic HF patients . Moreover, acetylcholine does not change forearm blood flow responses to mental stress [58, 59] and exercise (observational data from our laboratory) in chronic HF patients.
In contrast, the responses to an endothelial-independent vasodilator are preserved in chronic HF . In attempting to better understanding a possible regulation of the sympathetic activation in the endothelially-mediated vasodilatation in chronic HF, we studied the effects of the association of acetylcholine and phentolamine on forearm blood flow during exercise in chronic HF patients . In that study, we observed that phentolamine administered in addition to acetylcholine increased forearm blood flow levels towards normal levels. These findings strongly suggest that the exaggerated sympathetic outflow during physiological maneuvers may restrain the endothelial-mediated muscle blood flow during physiological maneuvers in chronic HF patients.
In summary, the exercise intolerance in HF patients can be attributed to the exaggerated sympathetic outflow, which may be an important instigator of the skeletal myopathy of HF.
Effects of exercise training on autonomic control in heart failure
Improvement in functional capacity
Accumulating evidence in the last two decades has shown that exercise training is a remarkable non-pharmacological strategy for the treatment of chronic HF patients. A number of studies have demonstrated that exercise training increases peak oxygen consumption in a range from 12% to 31% . These previous studies also suggest that most of the improvement in oxygen consumption occurs in the first 3 months, although it can continue if exercise training is maintained and the compliance is good (see relevant review by Barbour and Houston-Miller in this issue). Results from our laboratory, in which exercise training consisted of 5 min of stretching exercise, 25 min of cycling on an ergometer bicycle in the first month and up to 40 min in the last three months, 10 min of local strengthening exercise, 5 min of cool down with stretching exercises, at an intensity established by heart rate levels that corresponded to anaerobic threshold (AT) up to 10% below the respiratory compensation point obtained in the cardiopulmonary exercise test, provokes a 15%–20% increase in peak oxygen consumption [61–63].
Exercise training leads to a reduction in sympathetic nerve activity
Other investigators reported that even exercise training that involved only a minor muscle mass markedly reduced norepinephrine levels at rest and during submaximal exercise in chronic HF patients .
Mechanisms underlying reduction in sympathetic nerve activity
What are the mechanisms involved in the reduction of sympathetic activity in exercise-trained HF patients? So far, there is no definitive answer to this question. However, because afferent autonomic controls coordinated by arterial baroreceptors, cardiopulmonary receptors and chemoreceptors play an important role in the efferent sympathetic control, it would be reasonable to speculate that the reduction in sympathetic nerve activity would be mediated by some or all of these peripheral receptors. In fact, results of recent studies in animal models have provided convincing evidence for the mechanisms involved in the attenuation of sympathetic nerve activity after exercise training in chronic HF. Liu and colleagues, studying the baroreflex control, found that exercise training restored arterial baroreflex control on heart rate and renal sympathetic nerve activity in a rabbit model of pacing-induced HF . The same group of investigators reported in a subsequent study that the increase in baroreflex control after training occurred by means of a vagal mechanism, since atropine administration eliminated the improvement in baroreflex control of heart rate in exercise-trained HF rabbits . In a recent study from our laboratory, we observed that the increase in baroreflex control was mediated by an improvement in the aortic depressor nerve sensitivity, since exercise training in ischemic model of HF in rats increased spontaneous aortic depressor nerve sensitivity towards normal levels (Trained HF = 1791 ± 215, Untrained HF = 1150 ± 158, and Normal Control = 2064 ± 327 au/mmHg, P = 0.05) .
The effects of exercise training on chemoreflex control in HF have also been documented. The enhancement in peripheral chemoreflex control was normalized after exercise training in a rabbit model of HF . In addition, these investigators suggest that the improvement in the chemoreflex control involved nitric oxide synthesis, since blockade of nitric oxide synthesis by L-NNA enhanced peripheral chemoreflex function in exercise-trained, but not in untrained animals with HF .
Exercise training also exerts important effects on cardiopulmonary reflex control in HF. A four-month exercise regimen provoked a significant improvement in cardiopulmonary reflex control of renal sympathetic nerve activity .
Since chemoreflex, baroreflex, and cardiopulmonary reflex control all integrate in the central nervous system and exercise training normalizes these reflex functions, it would be reasonable to conclude that chronic exercise would lead to changes in central mechanisms of HF. In fact, exercise training reduces the expression of the angiotensin II type I (AT1) receptors in the paraventricular nucleus of the hypothalamus and in the rostral ventrolateral medulla and nucleus tractus solitarius . This non-pharmacological strategy has been also shown to reduce plasma angiotensin II levels in animal models of chronic HF . More recently, the same group of investigators demonstrated that exercise training prevented angiotensin II-induced arterial baroreflex impairment and decreased superoxide production and the expression of AT1 receptors in rats with HF. These findings suggest that exercise training causes central nervous system alteration that counteracts the sympathetic overactivity in HF.
There is evidence that nitric oxide production plays a role in the reduction of sympathetic activity after exercise training in HF. Zheng and colleagues  have recently reported that exercise training restored the number of neuronal nitric oxide synthase (nNOS)-positive neurons in the paraventricular nucleus in rats with HF. In addition, exercise training increased nNOS mRNA expression and protein levels in the paraventricular nucleus. In this study, the functionality of nitric oxide within the paraventricular nucleus was tested by microinjections of L-NMMA. The blunted responses of renal sympathetic nerve activity in rats with HF when compared with sham-operated rats were restored by exercise training. Finally, exercise training significantly improved the renal sympathetic nerve activity responses to sodium nitroprusside microinjected in the paraventricular nucleus. More recently, Mueller reported  that either direct excitation or inhibition of the rostral ventrolateral medulla was blunted in exercise-trained rats, which suggests that treadmill exercise training limits sympathoexcitation at this level of central nervous system.
Clinical implications of the excitatory and inhibitory autonomic changes by exercise training in heart failure
Sympathetic nerve activation has been associated with poor prognosis in HF patients. In a classical study, Cohn et al.  reported that plasma norepinephrine concentration was related to poor prognosis in HF patients. Brouwer et al.  found that heart rate variability has prognostic value, not only for identification of patients at high risk for all-cause cardiac mortality, but also for the prediction of sudden cardiac death. Kinugawa et al.  observed that patients with supramedian level of exercise plasma norepinephrine concentration had a significantly lower survival rate than those with inframedian level. La Rovere and colleagues  showed that low-frequency variability during controlled breathing is a powerful predictor of sudden death in patients with chronic HF that is independent of other parameters. More recently, we reported that muscle sympathetic nerve activity predicts mortality rate in patients with HF .
Since accumulating evidence has shown that exercise training significantly reduces sympathetic nerve activity, it is logical to assume that this non-pharmacological strategy would improve prognosis in HF. However, so far there is only one study in a limited sample size that suggests favorable outcome during exercise training in chronic HF patients. Belardinelli and colleagues  reported that long-term exercise training increased functional capacity in chronic HF patients, which was translated into a favorable outcome. A significant lower rate of hospitalization and cardiac mortality were observed in patients who increased functional capacity after training. More definitive information regarding the therapeutic benefit of chronic exercise training in HF will be available soon in the ongoing study called HF-ACTION .
Arterial and cardiopulmonary baroreceptors, central and peripheral chemoreceptors, cardiac chemoreceptors, and central nervous system abnormalities contribute to the neurohumoral excitation in HF. Growing evidence suggests that mechanosensitive afferent fibers in skeletal muscle play a role in the exaggerated sympathetic activity during exercise in chronic HF patients. Sympathetic excitation contributes to skeletal muscle myopathy, which explains, in great part, the exercise intolerance in chronic HF. Exercise training dramatically reduces central sympathetic outflow in chronic heart failure, which, in turn, improves peripheral blood flow and skeletal muscle myopathy. These physiological changes obtained by exercise training lead to improved functional capacity and quality of life in chronic heart failure patients.
Dr. Carlos E Negrao has been supported by the Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP # 2005/59740-7) and Conselho Nacional de Pesquisa (CNPq # 304304/2004-2), Brazil, and Dra. Holly R Middlekauff by the National Institutes of Health (Grant RO1 HL084525), USA.