Clinical Benefits of Exercise Training in Heart Failure
Few randomized controlled trials have examined the effect of exercise training on clinical events including mortality and hospital readmission. The largest trial of exercise training to date, HF-ACTION (Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training), randomized 2,331 participants with HFrEF to usual care or 36 sessions of a structured aerobic exercise training program on a treadmill or bike over 30 months, with heart rate reserve targets to guide exercise intensity.6 There was no significant difference in the primary outcome (death or hospitalization from any cause) between the training and usual care groups. However, in a prespecified supplementary analysis adjusting for characteristics known to be prognostic of the primary endpoint (HF etiology, duration of cardiopulmonary exercise test, left ventricular EF (LVEF), Beck Depression Inventory II score [a questionnaire assessing depressive symptoms], and atrial fibrillation or flutter), exercise training was associated with lower risk of all-cause mortality or hospitalization (HR 0.89, 95% CI 0.81–0.99, p=0.03). The second study examining clinical outcomes, EJECTION-HF (Exercise Joins Education: Combined Therapy to Improve Outcomes in Newly-discharged Heart Failure), evaluated supervised center-based exercise training compared to a home exercise program among patients with HFrEF and HFpEF recently discharged after an acute HF admission.7 There was no difference in all-cause one-year mortality or readmission between groups (60% in the intervention group vs 65% in the control group). In a prespecified subgroup analysis by age, there was a trend toward greater benefit in participants under 70 years, although this did not reach statistical significance (odd ratio: 0.56 [95% CI: 0.30 to 1.02] vs. 1.56 [95% CI: 0.67 to 3.64] in younger vs older individuals, p for interaction = 0.05). While primary analyses of these two trials did not suggest benefit of exercise on clinical events, meta-analyses suggest that exercise interventions in the context of cardiac rehabilitation likely reduce both all-cause and HF hospitalizations, with mixed data on mortality.8 One recognized limitation across exercise trials relevant to long-term impact on outcomes is decreased adherence to the exercise training protocols over time. For example, in HF-ACTION median duration of training per week was 95 minutes during months 4 through 6 compared to 74 minutes for the last two months of the study period.Studies of patients with both HFrEF and HFpEF have demonstrated a consistent improvement in cardiorespiratory fitness as measured by peak oxygen uptake (VO2) with exercise training. Among patients with HFrEF, results from HF-ACTION demonstrate greater improvement in peak VO from baseline to 12 months among the exercise training group (0.7 mL/kg/min (−1.0, 2.5) in the training group vs 0.1 mL/kg/min (−1.5, 1.18) in the usual care group, p<0.001).6 Demonstrating similar results, the SMART-EX study (Study of Myocardial Recovery After Exercise Training in Heart Failure) compared a 12-week program of either high intensity interval training (HIIT) or moderate continuous training (MCT) to a recommendation of regular exercise among 231 patients with HFrEF.9 At 12 weeks, the adjusted change in VO peak was higher in both the HIIT and MCT groups compared to the control group (adjusted between group differences HIIT vs control 1.4mL/kg/min [0.2, 2.6], MCT vs control: 1.8 mL/kg/min [0.5, 3.0]), with similar results between exercise modalities. However, at 52 weeks, there was no significant difference in VO peak between groups. The ARISTOS-HF study (Aerobic, Resistance, InSpiratory Training OutcomeS in Heart Failure) also compared different exercise modalities including aerobic training (AT), resistance training (RT), and inspiratory muscular training (IMT) in a 12-week program for 88 participants with HFrEF.10 They found a significant improvement in peak VO from baseline to 12 weeks in all exercise groups (AT, AT/IMT, AT/RT, and AT/RT/IMT), yet no significant between group differences. Overall, among those with HFrEF, exercise training improves VO in the short term regardless of exercise modality, yet long term benefits remain unclear. While the change in peak VO from baseline in the training groups in these studies may seem modest (1.01 mL/kg [95% CI −0.42-2.44] per minute mean difference between treatment and control groups at 12 months in a meta-analysis of exercise-based cardiac rehabilitation studies), the magnitude is likely clinically meaningful.11 For instance, in HF-ACTION, a 6% increase in peak VO, corresponding to approximately 1 mL/kg/min, was associated with a 5% lower risk of time to all-cause mortality or all-cause hospitalization (HR 0.95, 95% CI 0.93, 0.98, p <0.001) and an 8% lower risk of cardiovascular mortality or HF hospitalization (HR 0.92, 95% CI 0.88, 0.96).6,12 Thus, albeit modest, these improvements in peak VO achieved through exercise training may translate to improved patient outcomes.
2
2
2
2
2
2
2
2
Regarding HFpEF, the largest trial examining the effect of training on exercise capacity in this population was OptimEx-Clin (Optimizing Exercise Training in Prevention and Treatment of Diastolic Heart Failure), which evaluated HIIT versus MCT compared to usual care in 180 sedentary patients.13 After 12 months of training, there were significantly greater changes in peak VO2 in the both the HIIT group (difference in mean changes: 1.5 mL/kg/min [95% CI, 0.4 to 2.7], P = .01) and MCT group (2.0 mL/kg/min [95% CI, 0.9 to 3.1], P = .001) compared to the control group, with no significant difference between training modalities. The Ex-DHF study (Exercise training in Diastolic Heart Failure) pilot study of 58 participants with HFpEF demonstrated a similar significant improvement in peak VO2 at 3 months, yet no change in 6MWD.14 Given the greater risk of HFpEF among those with obesity, two trials evaluated the effect of combined diet and exercise training on exercise capacity. They found a significant increase in peak VO2 and 6MWD after both caloric restriction and exercise, with evidence of additive benefits with both interventions.15,16 These results were similar to the PARIS (Prospective Aerobic Reconditioning Intervention Study) study of older adults (mean age 70 years ± 6) with HFpEF which also demonstrated a significant improvement in peak VO2 (increase of 2.7 mL/kg/min [95% CI 1.4 to 4.0] in the exercise group compared to control group) and 6MWD in the exercise group compared to the control group.17 A meta-analysis across eight randomized controlled trials of exercise training in HFpEF showed a significant improvement in peak VO2 with a mean difference of 1.66 ml/min/kg (95% CI 0.97, 2.35, prandom < 0.001) compared to the control group.18
Taken together, exercise training across diverse programs appears to improve exercise capacity similarly among individuals with both HFpEF and HFrEF, without major differences across type of training program.
While the majority of studies have focused on the effect of exercise training on peak VO2, ventilatory efficiency as measured by the VE/VCO2 slope is associated with poor outcomes in both HFrEF and HFpEF, yet remains less well studied in exercise trials.19–22 In the ARISTOS-EF study of HFrEF patients, only combined RT/AT/IMT led to a significant decrease in VE/VCO2 slope at 12 weeks, whereas other combinations of exercise modalities did not. In the OptimEx-Clin evaluation of HFpEF patients, exercise training did not lead to a change in VE/VCO2 slope at 3 or 12 months.13 By contrast, a study of supervised training in a combined HFrEF and HFpEF sample reported a decrease in VE/VCO2 at 12 weeks only among those with HFpEF.23 These mixed results highlight the need for future studies evaluating effects of exercise on ventilatory efficiency as an important prognostic measure in HF.
Quality of life (QOL) improvements have been observed across multiple studies in both the HFpEF and HFrEF groups. Among the HFrEF population, studies suggest significant improvement with training in the short term that is not consistently maintained in the long term. Indeed, HF-ACTION demonstrated a clinically significant improvement in Beck Depression Inventory II at 3 and 12 months, yet improvement in Kansas City Cardiomyopathy Questionnaire and Euro-QOL visual analog scale at 3 months did not persist at 12 months.24–26 Similarly, in ARISTOS-HF, the Minnesota Living with Heart Failure Questionnaire score improved in all groups at 3 months.10 Regarding HFpEF patients, results have been more favorable, demonstrating an improvement in MLHFQ and KCCQ scores in both the short and long term.13,14,16,27 Taken together, exercise training has a meaningful positive impact on quality of life and mood across the HF population, although results which may be attenuated in the long term among those with HFrEF.
While excessive exercise beyond routine moderate-intensity physical activity may have adverse cardiovascular consequences, including cardiac injury and remodeling and arrhythmia risk,28 it is important to acknowledge that prior studies on excessive exercise have focused on non-HF samples and extrapolation to individuals with HF is challenging. In the context of exercise regimens studied in HF trials, there have been no excessive adverse events. Specifically, among the HFrEF population in HF-ACTION, 3.2% of the exercise group had at least one hospitalization due to an event occurring during or within 3 hours after exercise, compared to 1.9% in the usual care group.6 In SMART-EX there was no statistically significant difference in adverse events between groups (HIIT 39%, usual care 34%, MCT group 25%).9 Similarly, among patients with HFpEF there were no differences between exercise groups in serious adverse events, including Optimex-Clin (31% in the HIIT group, 31% in the MCT group, and 27% in the control group) and PARIS, with low frequency of acute coronary syndrome, atrial fibrillation, and worsening HF (≤5 patients per group for each in OptimEx-Clin).13 17 Taken together, HF trials have not shown excess adverse events with exercise training.
Several limitations in current data deserve mention. First, findings from exercise training trials in patients with HF need to be considered in the context of selection bias, potentially limiting generalizability. Second, adherence varied across trials and likely affected long-term outcomes. Lastly, the optimal exercise regimen likely differs between patients and a “one-size fits all” exercise program may not be appropriate. This is partially addressed in trials that prescribed heart rate and VO2 targets and compared different exercise modalities, but likely is an oversimplification. As addressed below in the “Future Directions” section, a more refined understanding of mechanisms by which exercise confers cardiovascular benefits using multi-omic approaches and other studies has the potential to guide future exercise recommendations including personalized exercise prescriptions in patients with HF.
Exercise training in Early vs Advanced Heart Failure
While not the primary focus of this review, it is relevant to note two studies that have examined exercise training in the pre-heart failure population. Among patients with Stage A HF (with risk factors), one-year of HIIT led to an improved peak VO2 and LV mass.29 These results mirrored a study of patients with Stage B HF (cardiac remodeling), deemed at risk for HFpEF based on left ventricular hypertrophy and elevated cardiac biomarkers. One-year of exercise led to a significant increase in peak VO2 and reduction in myocardial stiffness.30 These results suggest that early exercise training may reverse the pathophysiologic changes that ultimately lead to overt disease, and should be further explored as a HF preventive strategy.
Patients with left ventricular assist devices (LVAD) represent a unique population with a paucity of data regarding efficacy of exercise training in improving clinical outcomes. The Exercise training in patients with LVAD trial (Ex-VAD) randomized 64 patients to 12 weeks of exercise versus usual care.31 While there was no improvement in peak VO2, there was a significant increase in 6MWD and quality of life measures at 12 weeks. Importantly, there was no difference in adverse events in this group, suggesting that exercise in patients with LVAD is safe, and may improve patient reported outcomes.
Physiologic and Molecular Effects of Exercise Training in Heart Failure
While there are a broad range of potential clinical benefits of exercise training in HF, the mechanism of these benefits is less clear. We will review physiological and molecular effects of exercise training in HF by using the oxygen pathway as a guide, spanning oxygen uptake through the pulmonary system, delivery via the cardiovascular system, and utilization through peripheral oxygen extraction and mitochondrial respiration in skeletal muscle. In addition to organ-specific effects, global effects of exercise on key pathophysiologic pathways germane to HF including inflammation, immunomodulatory and oxidative effects, and neurohormonal regulation are discussed (Figure 2). We conclude this section with a discussion of the effect of exercise on inter-organ crosstalk, an area requiring additional studies in the heart failure population.
Respiratory muscle weakness, ventilatory inefficiency, and diffusion impairment are key pulmonary factors contributing to exercise limitation in HF.32 Inspiratory muscle strength is independently associated with prognosis in patients with HF and both targeted inspiratory muscle training and aerobic training can improve exercise capacity in this population.33–35 Indeed, in a murine coronary artery ligation model of HF, aerobic exercise training protected against diaphragm contractile dysfunction when compared with sedentary mice.36 Histological analyses of diaphragm contractile fibers showed a significant reduction in oxidative modification of critical contractile proteins including actin and creatine kinase and reduced activity of proteolytic pathways. Interestingly, the benefit of exercise training on respiratory muscle weakness may be attenuated in the setting of obesity HFpEF models in contrast to hypertensive HFpEF models, where exercise training prevented diaphragmatic mitochondrial and functional impairments.37,38 These data suggest exercise may reduce oxidative stress and protein degradation leading to diaphragmatic dysfunction, yet may not be applicable across all HFpEF phenotypes.
Among HF patients, reduced diffusing capacity of the lung for carbon monoxide (DLCO, comprised of alveolar-capillary conductance and capillary blood volume), and inability to augment diffusion during exercise via capillary recruitment has been shown to contribute to worse peak VO2 in HF.39 These defects were improved with MCT in one study, demonstrating a 25% increase in resting DLCO, a 15% increase in alveolar-capillary conductance and a 10% increase in pulmonary capillary blood volume after exercise training, improvements which reversed with detraining.32 These changes were concomitant with improvement in flow-medial brachial artery dilation with exercise, suggesting that improved systemic endothelial function may underlie potential pulmonary vasodilation and more efficient capillary recruitment, leading to increased pulmonary capillary blood volume and subsequent DLCO improvement. The mechanism for improved alveolar-capillary conductance is less clear. The authors posit that downregulation of sodium and fluid transport from the capillary to the alveolar space may reduce the diffusion path for gas exchange, but further studies are required to confirm this mechanism. Overall, while further investigations are required to fully understand the underlying mechanisms, exercise training in patients with HF promotes favorable changes in respiratory muscle strength and diffusion capacity.32–35
Concomitant pulmonary hypertension is common in patients with HF, seen in up to 40–75% in patients with HFrEF and 52–83% in HFpEF.40,41 While the pathophysiology is complex, pulmonary vasodilation to accommodate increased cardiac output and venous return during exercise is a key element to exercise tolerance in HF, and worsens as pulmonary dysfunction progresses.42 The effect of exercise training on pulmonary vascular function in individuals with HF remains unclear, though studies focused on individuals with pulmonary arterial hypertension suggest similar improvements with exercise training in peak VO2, in parallel with decreased pulmonary artery systolic pressures.43 In PAH, anti-inflammatory and other immune effects of exercise training have been postulated as possible mechanisms underlying potential pulmonary vascular benefit, though this remains understudied in both PAH and pulmonary vascular dysfunction seen in HF.44,45
Exercise training has consistently been found to reverse cardiac remodeling in HFrEF, as evidenced by decreased left ventricular (LV) dimensions and volumes after 6 months.46 By contrast, no significant differences were found in LV volumes or LV mass with exercise training among individuals with HFpEF.15–17
With respect to systolic function, LVEF has been shown to improve in trials with long term (at least 6 months) training interventions, yet not change in trials with a short intervention period (3 months or less).10,23,46 For example, in study of patients with chronic HF enrolled in a 6 month exercise program both LVEF (30% prior, 35% post) and peak exercise stroke volume improved (increase by 14mL +/− 22 in exercise training vs 1mL +/− 19 in control group).46 A meta-analysis including 1,486 patients with HFrEF also demonstrated a concomitant improvement cardiac output (+21.3%) with exercise training.47 By contrast, even though patients with HFpEF display subtle abnormalities in systolic function on echocardiography and exercise limitations in stroke volume reserve,48 meta-analyses have shown that exercise training does not affect LVEF or cardiac output in patients with HFpEF.18,47
LV diastolic function, estimated non-invasively by echocardiography via the early transmitral flow velocity-to-early diastolic mitral annulus velocity ratio (E/e’), deceleration time of the E-wave, and isovolumic relaxation time, may improve with exercise training.49,50 In a meta-analysis including trials featuring populations with varying degrees of diastolic dysfunction (from mild to severe) and training duration (12–24 weeks), exercise training significantly reduced E/e’ ratio in both HFpEF and HFrEF patients.51 Review of specific trials in this meta-analysis, in addition to more recent trials, demonstrates variable improvement in E/e’ with exercise training among patients with HFpEF, likely related to the varied baseline demographics and severity of baseline diastolic dysfunction. Among OptimEx-Clin participants (mean age 70 years, baseline E/e’ 15.8), there were no significant changes in E/e’ after 12 months of exercise training.13 Similarly, two studies of exercise training in symptomatic HFpEF patients (mean age 67 years, baseline E/e’ ~13) found no improvement in diastolic function after 20 weeks.15,16 Conversely, aerobic training in participants with HFpEF (mean age 60 years, baseline E/e’ ~21) led to a significant decrease in E/e’ compared to the usual care in a moderately sized studies.23 While diastolic dysfunction seems to improve in certain subsets of patients with HFpEF, further studies are needed to identify the patients and training program that leads to the greatest benefit.
Exercise training and especially aerobic exercise has been found to decrease circulating levels of natriuretic peptides, a prognostic biomarker and indicator of cardiac stress in HF. This decline in NT-proBNP with exercise training in turn has been found to be associated with improved exercise capacity in a meta-analysis comprising 2,563 patients with HF.50,52,53 Several mechanisms likely contribute to the salutary effects of exercise training in HF, including reduced afterload due to lower peripheral resistance driven by enhanced endothelial function, potential anti-inflammatory effects, and restoration of equilibrium between sympathetic and parasympathetic activity. Each of these global effects are discussed separately in sections below.
Improved diastolic function with exercise is thought to be secondary to improvement in passive myocardial stiffness. Procollagen levels significantly decreased after exercise training among HFpEF patients.54 This mechanism was further investigated in a genetic mouse model of HFpEF in which diastolic stiffness is induced by deletion of the IA junction of titin (TtnΔIAjxn ), a sarcomeric protein that is a key determinant of myocardial stiffiness.55–57 After voluntary exercise, passive myocardial stiffness improved, attributed to increased titin compliance from decreased phosphorylation of the PEVK domain via protein kinase C, which is known to increase passive tension, and increased phosphorylation of the N2B domain, known to reduce passive tension.56 Further, nitric oxide (NO) production activates soluble guanylate cyclase leading to an increase in downstream production of cyclic guanosine monophosphate and protein kinase G, also involved in titin phosphorylation, contributing to decreased diastolic stiffness.58 Restoration of NO production with exercise, as discussed below, may reverse these effects.
Improved myocardial stiffness with exercise may also be mediated by immunomodulatory effects. Myeloid-derived suppressor cells (MDSCs) are a population of immature cells with immunosuppressive function with a paradoxical cardioprotective effect in heart failure.59 In a murine HF model, the MDSC population expanded after exercise training, mediated by the IL-10/STAT3 pathway which plays a protective role in ischemia-reperfusion injury, whereas MDSC depletion worsened isoproterenol-induced cardiac hypertrophy, fibrosis, and dysfunction.60 Increased production of IL-10 in macrophages with exercise training and subsequent expansion in the MDSC population is one potential mechanism by which cardiomyocyte modification and immunomodulatory alterations can improve myocardial stiffness with exercise training.
Exercise training has beneficial effects on the peripheral vasculature via reducing total peripheral resistance (TPR) at both rest and peak exercise.61 This is likely mediated by improved endothelial function with exercise training among patients with HF, given enhanced acetylcholine-mediated endothelium-dependent vasodilation of the skeletal musculature after training.46,61,62 Specifically, after 6 months of exercise training, peripheral blood flow increased in the femoral artery by 203%, and was directly correlated with a 26% increase in peak VO2 among patients with chronic HF.61 The improvement in endothelial cell function may be attributable to beneficial effects of shear stress associated chronic exercise training, including increased NO synthase production and increased expression of cytosolic copper/zinc-containing superoxide dismutase, potentiating locally-produced NO.63 Indeed in the latter study, exercise training led to increased basal nitric oxide (NO) production in resistance vessels, demonstrated by a 174% increase in the inhibitory effect of NG-monomethyl-L-arginine (L-NMMA), an NO synthase inhibitor, on peripheral blood flow.61
Another mechanism by which exercise training may promote endothelial health is via effects on endothelial progenitor cell (EPC) levels, bone marrow-derived cells essential for vascular repair and regeneration that are known to be decreased in patients with HF.64 Exercise training may expand the circulating EPC population via shear stress-induced NO production and hypoxia-inducible factor-1 (HIF-1) production in the setting of exercise-induced hypoxia in skeletal muscles, increasing plasma levels of vascular endothelial growth factor (VEGF) and stromal cell-derived factor-1 (SDF-1), two critical mediators of EPC mobilization.64–66 This may support endothelial repair and restore exercise-mediated peripheral vasodilation.
Cardiopulmonary exercise testing in patients with HF demonstrates impaired peripheral O2 extraction, which is significantly more common among patients with HFpEF (40%) compared to HFrEF (2%).67 Notably, exercise training improves peripheral oxygen extraction among patients with HFpEF.68 This is likely multifactorial in the setting of greater exercise-mediated skeletal vasodilation, improved oxygen diffusion via morphologic skeletal muscle changes, along with other potential global anti-inflammatory, anti-oxidant, and neurohormonal effects, which are discussed in turn below.
Multiple skeletal muscle abnormalities, including a shift from oxidative (type I) fibers to glycolytic (type II) fibers, reduced mitochondrial density and enzyme activity, and muscle atrophy contribute to exercise intolerance in patients with HF and may be reversed with exercise training.69,70 In a small study of patients with chronic HF, muscle biopsies after 6 months of exercise training showed an increase in slow-twitch type I fibers and a decrease in type II fibers, a change that would be expected to enhance fatigue resistance by improving the aerobic properties of muscle fibers.70 Exercise training additionally enhanced mitochondrial density and function of skeletal muscle in this same study. This was evidenced by increased surface density of cytochrome c oxidase-positive mitochondria (+41%), mitochondrial cristae (+43%), and mitochondrial inner border membranes (+92%).70 Other studies have also demonstrated enhanced activity of mitochondrial enzymes including citrate synthase and cytochrome c oxidase, leading to more efficient adenosine triphosphate (ATP) production.71–73 Lastly, mitochondrial biogenesis occurs via the peroxisome-proliferator-activated γ co-activator-1α (PGC)-NRF1-TFAM pathway, which is known to be dysregulated in HF and can be restored with exercise training.73,74
In addition, patients with HF have elevated levels of the muscle-specific E3 ligase MuRF-1, an enzyme contributing to proteolysis and muscle wasting.75 Four weeks of endurance training reduced MuRF-1 expression by 33% in HF patients regardless of age, improving the muscle anabolic-catabolic balance via decreased activity of the ubiquitin-proteasome system.75 In murine models, both resistance and aerobic training led to upregulation of IGF-1 and subsequent PI3K/Akt pathway activation, which leads to downstream inhibition of FoxO transcription factors and downregulation of MuRF1 expression.76 Interestingly, Akt acts as an upstream regulator of PGC, which promotes the transcription of mitochondrial genes, highlighting the pathway’s potentially multifaceted role in the restoration of muscular function through exercise.77,78 Lastly, exercise training induces angiogenesis within skeletal muscles, which triggers an increase in capillary density in muscle fibers, facilitating oxygen delivery and waste product removal during exercise and contributing to better matching of oxygen supply to metabolic demand in skeletal muscle.79 Upregulation of pro-angiogenic factors including VEGF, HIF-1, and angiopoietin-2 found in the skeletal muscle of HF patients after training supports involvement of these molecular pathways.79 Taken together, there are multiple mechanisms by which exercise training improves skeletal muscle health, including enhanced mitochondrial adaptions, decreased muscular atrophy, and muscular angiogenesis.
In patients with chronic HF who underwent a 20-week training program with subsequent increase in peak VO2 compared to controls, exercise training was found to reduce proinflammatory cytokines including tumor necrosis factor-alpha (TNF α), IL-6, and soluble IL-6 receptor, and reduce soluble apoptotic mediators including Fas and Fas ligand.80 These pro-inflammatory and pro-apoptotic mediators are associated with myocardial dysfunction, skeletal myopathy, and endothelial dysfunction and thus contribute to several facets of exercise limitation.80 Specifically, TNF-α is associated with negative inotropic effects, LV dilation, myocyte dysfunction, skeletal muscle dysfunction, and vascular endothelial dysfunction, all contributing to exercise intolerance.80–82 Reduced IL-6 is associated with decreased muscle breakdown and improved mitochondrial function, linked to enhanced aerobic metabolism and exercise capacity.80 Finally, exercise training preserves Fas/FasL activity, leading to attenuation of Fas-induced tissue apoptosis.80 Further investigation is required to better define these molecular pathways and whether they may be exploited for therapeutic use.
Oxidative stress is intimately linked to inflammation and may also be modulated by exercise training. For example, in a study of 23 patients with HF randomized to 6 months of training vs a sedentary lifestyle, skeletal muscle biopsies showed that baseline radical scavenger enzyme levels, including glutathione peroxidase and catalase, were lower compared to those of healthy controls and were significantly augmented by exercise training compared to HF controls.83 Further, levels of local oxidative stress measured by nitrotyrosine formation were similarly increased at baseline in HF patients compared to healthy controls, and reduced after exercise training compared to controls. This decreased skeletal muscle oxidative stress was associated with lower skeletal muscle apoptosis compared to controls. This study also found reduced local TNF-α levels in skeletal muscle after exercise training. Since TNF-α impairs an essential cofactor of glutathione peroxidase, reduced levels after exercise may explain the combined anti-inflammatory and antioxidative effects of exercise training.
Lastly, as discussed above, exercise training augments mitochondrial biogenesis and enhances mitochondrial function in skeletal muscle. This leads to increased mitochondrial coupling (ATP production relative to O2 uptake) which may reduce generation of reactive oxygen species and subsequent mitochondrial DNA damage, potentially interrupting the cycle of mitochondrial dysfunction.84
Patients with HF have chronic sympathetic stimulation and concomitant overactivation of the renin-angiotensin-aldosterone system.85 Exercise training in turn has been shown to improve autonomic balance by normalizing baseline sympathetic overactivation, evidenced by improved heart rate variability and heart rate recovery.86,87 For example, in a study of 66 patients with HFrEF and HFpEF, 4 months of exercise training lead to improvement in heart rate variability in both groups.87 While mechanisms for this are not fully described, a study of rabbits with pacing-induced HF demonstrated central anti-oxidative mechanisms including upregulation of medullary NAD(P)H expression and increased copper/zinc superoxide dismutase activity contributed to beneficial effects of exercise training on autonomic activity.88
Further, muscle sympathetic nerve activity (MSNA) is increased in HF.89–91 This is controlled by muscle afferents including metabolically sensitive unmyelinated group IV fibers (“metaboreflex”) and mechanically sensitive thinly myelinated group III fibers (“mechanoreflex”) which modulate the arterial baroreflex. Animal studies have demonstrated that sympathetic overactivation in HF is mediated by mechanoreflex overactivity whereas the metaboreflex is blunted, changes which can be normalized by exercise training.92,93 In a study of patients with HF, exercise training was shown to increase muscle metaboreflex control and decrease muscle mechanoreflex control, leading to decreased resting MSNA levels.89 The former was associated with increased expression of transient receptor potential vanilloid type-1 (TRPV1) receptor and cannabinoid receptor 1, which are co-located on muscle metaboreceptor group IV afferent neurons, suggesting a mechanism by which exercise training improves MSNA .89,91 The latter was associated with decreased COX-2 expression, a key enzyme in the conversion of arachidonic acid to prostaglandin and thromboxane, and a decreased NFκB/IκB ratio, supporting the anti-inflammatory role of exercise in attenuating sympathetic stimulation. While these alterations highlight the role of skeletal muscle abnormalities in reflex control of the sympathetic nervous system during exercise, further investigations into the interplay between these mechanisms is required.
Arterial baroreceptors located in the aortic arch and carotid sinuses inhibit sympathetic nervous system activity and are dysfunctional in HF. Among 20 patients with post-infarction HF on optimal medical therapy, baroreflex sensitivity improved after 12 weeks of training.94 Further, exercise training prevented the deterioration in the arterial baroreflex control of MSNA in HF.95 Overall, exercise training appears to counteract the maladaptive chronic sympathetic stimulation commonly seen in HF patients, allowing for improved hemodynamic responses to exercise.
Prolonged neurohormonal hyperactivity in HF patients, resulting in elevated levels of angiotensin II, aldosterone, vasopressin, and atrial natriuretic peptide (ANP), contributes to vasoconstriction, fluid retention, and myocardial remodeling.96,97 Exercise training attenuates neurohormonal hyperactivity, which is closely associated with improved prognosis in patients with HF.98 Sixteen weeks of endurance training resulted in decreased resting levels of angiotensin II (−26%), aldosterone (−32%), vasopressin (−30%), and ANP (27%) among patients with HF compared to controls.97 These changes are mirrored in experimental models, including a rabbit model of pacing-induced HF where exercise training led to a reduction in renal sympathetic nerve activity, angiotensin II levels, and improved baroreflex function.98 To assess causation, the study was repeated with an angiotensin II “clamp” during exercise, maintaining plasma levels at that of non-exercising HF rabbits, preventing downstream sympathetic activation via angiotensin II. This prevented downstream reduction in renal sympathetic activity and normalization of the baroreflex function, demonstrating the importance of reduced angiotensin II in mediating sympathetic activation and baroreflex function. A favorable decrease in neurohormonal activity with exercise via decreased plasma angiotensin II likely occurs via multiple mechanisms including angiotensin receptor expression from increased renal blood flow in the setting of improved endothelial function, cytokine mediated regulation of angiotensin receptors, nitric oxide-mediated changed in renal sympathetic nerve activity, and reduction in oxidative stress.98–100
The complex pathophysiology of HF including contributions of extra-cardiac organ systems has become increasingly recognized. Specifically, complex and often bidirectional feedback between the heart and other organs can worsen HF, including heart-kidney, heart-liver, and heart-adipose tissue interactions.101 The mechanisms facilitating crosstalk between organs are complex and not fully elucidated, and by extension our knowledge of the effect of exercise on inter-organ crosstalk in HF is in nascent stages.
One avenue to shed light on effects of exercise on inter-organ crosstalk has been via the study of exerkines, which are molecules released into the circulation in response to exercise and play a critical role in the mediation of inter-organ communication.102 Specifically, skeletal muscles release myokines such as IL-6, IL-15, and irisin, which enhance glucose metabolism, decrease systemic inflammation, and improve lipid profiles.102,103 Additionally, exerkines secreted by the liver, bone, nervous system, and adipose tissue complement and amplify these benefits.103 For example, lipokine 12,13 di-HOME, released by brown adipose tissue after exercise, led to improved cardiac function, reversal of the deleterious effects of a high fat diet on cardiac remodeling, and increased mitochondrial respiration in mice.104
Furthermore, extracellular vesicles, which transport enzymes, proteins, and microRNAs, are thought to play an important role in interorgan crosstalk with exercise, given their presence in the circulation increases after exercise.105 Studies are ongoing to determine the distribution of these extracellular vesicles, the identification of bioactive particles they contain, and whether they contribute to the beneficial effects of exercise.105 Together, exercise training promotes systemic metabolic health, reducing inflammation, and enhancing inter-organ communication, yet further studies are needed in the heart failure population to determine the role of exerkines and extracellular vesicles in disease pathophysiology. While prior studies have largely focused on the effect of exercise on interorgan cross-talk in a healthy state, further studies are needed in individuals with HF.
Summary and Future Directions
In sum, current clinical guidelines broadly support exercise training in patients with HF. These recommendations are based on numerous clinical trials demonstrating benefits across clinical outcomes, exercise capacity, and quality of life. These clinical outcomes are paralleled by known beneficial physiologic effects of exercise training that broadly span multiple organ systems and molecular pathways known to be dysregulated in HF. Despite these benefits, there remain critical knowledge gaps in our current understanding of clinical benefits and mechanisms thereof, precluding widespread programmatic interventions to harness the salutary benefits of exercise. Specific future areas of study include the following:
Exercise training benefits across diverse HF samples: It is important to acknowledge wide phenotypic heterogeneity in HF, and the critical need to enroll diverse samples in future trials to enable greater insights into benefits of exercise training across important HF subgroups. A notable example is the current lack of event-driven HFpEF trials. While not focused on exercise training, the ongoing REHAB-HFpEF trial is examining the effect of a multi-domain rehabilitation intervention on clinical outcomes in this important subgroup.106 Similarly, among patients with Stage A and Stage B heart failure, existing studies have demonstrated improved exercise capacity and reversal of cardiac remodeling with exercise training, however additional studies are need to evaluate for long-term outcomes including development of overt, Stage C disease.
Personalized training regimens: With the growing recognition that HF, particularly HFpEF, is a heterogenous disease, it is increasingly clear that a ‘one-size fits all’ exercise prescription may be inadequate to optimize clinical benefits for each patient. Studies focused on elucidating physiologic and molecular effects of specific exercise modalities and training goals across diverse HF samples will provide guidance for future exercise recommendations including personalized exercise prescriptions in patients with HF.
Exercise training in the context of evolving weight loss therapies for HF: With evolving therapies targeting obesity including incretin-based treatment of obesity-HFpEF, the role and potential benefits of concomitant exercise training must be elucidated.107 In particular, weight loss therapy is known to lead to concomitant muscle loss – whether incorporation of exercise training (particularly resistance training) may attenuate concomitant muscle loss in HF patients remains to be studied.
Deep phenotyping studies to elucidate molecular mechanisms: Large-scale molecular profiling platforms and multi-omic approaches across different tissue types are currently being leveraged to study molecular benefits of exercise in healthy individuals.108 Similar deep phenotyping approaches among individuals with HF are needed to enable a more comprehensive understanding of molecular mechanisms, including inter-organ crosstalk, the role of extracellular vesicles, and anti-inflammatory mechanisms that underlie the benefits exercise training in this especially vulnerable sample.
Sex differences in mechanisms of clinical benefit of exercise training: Sex differences in HF are pervasive, and affect predisposition to HF, HF subtypes, and physiologic exercise responses.109,110 Current data suggest that the benefits of exercise training in HF also display sexual dimorphism, with men showing greater improvements in diastolic reserve and oxygen extraction, and women displaying enhanced vasodilatory responses, primarily due to estrogen-related NO signaling.111 In addition, skeletal muscle adaptations and autonomic effects of exercise training also differ by sex.112,113 These findings underscore the need to better understand sex-specific benefits of exercise training to potentially tailor training and optimize outcomes for both men and women with HF.
In summary, this review highlights the clinical benefits of exercise training in HF, including improved clinical outcomes, functional capacity, and enhanced quality of life, while elucidating underlying mechanisms of exercise-induced physiological changes such as favorable modulation of cardiac remodeling, improved endothelial function, reduced inflammation, and antioxidant effects. We highlight key areas for future research to deepen understanding of the benefits and mechanisms of exercise to potentially enable future implementation of guideline-directed therapeutic exercise training.
EL
Dr. Emily Lau
Cardiology · General Non-Invasive CardiologyMassachusetts General HospitaldremilylauBoston, MA
