Exercise training as a treatment for heart failure: potential mechanisms and clinical implications

Despite advances in pharmacological treatment and device therapy, heart failure (HF) continues to be a major public health problem, with high morbidity and mortality. Consequently, additional therapeutic strategies are being explored, including lifestyle changes such as exercise training. First considered harmful and thus prohibited in HF patients, moderate physical activity is now known not only to be safe, but to improve exercise tolerance, maximal oxygen uptake and quality of life, with modest but significant mortality and hospitalization-rate reducing effects (Pina et al. 2003; O’Connor et al. 2009; Patwala et al. 2009). It is important to understand the mechanisms by which exercise training improves the clinical status of HF patients, in order to exploit optimally the potential benefits. Three types of mechanisms could account for cardiovascular health improvements resulting from exercise: (1) general training effects; (2) cardiovascular training effects; and (3) favourable interactions with underlying remodelling. General training effects that enhance skeletal muscle/cardiovascular efficiency and performance in healthy individuals (Heckman & McKelvie, 2008) would also be expected in HF patients, improving indices like exercise time, and quality of life. Cardiovascular training effects like increased dynamic heart rate range (Hautala et al. 2004), haemodynamic changes (Naylor et al. 2008) and altered vascular function (Tinken et al. 2008) occur in healthy individuals but also improve cardiovascular performance in HF patients. Favourable interactions with remodelling processes would be of particular interest, because they have the potential to prevent the development of, or even reverse, the HF state. Exercise characteristically produces adaptive or physiological hypertrophy, in contrast to the maladaptive hypertrophy associated with HF. Calcineurin, a central player in maladaptive remodelling (Wilkins et al. 2004), is a Ca2+–calmodulin-dependent phosphatase that dephosphorylates nuclear factor of activated T-cells (NFAT) transcription factors, causing their nuclear translocation and induction of hypertrophic gene programs. Calcineurin/NFAT signalling impairs Ca2+ handling, promotes apoptosis and induces re-expression of fetal genes (Selvetella et al. 2004). Cardiac-specific calcineurin activation induces severe hypertrophy followed by important cardiac dysfunction (Molkentin et al. 1998). Moreover, calcineurin expression and activity are increased in both hypertrophied and failing human hearts (Lim & Molkentin, 1999). Ca2+–calmodulin-dependent protein kinase-II (CaMKII) mediates important adrenergic effects on the heart. CaMKII overexpression, common in HF, impairs Ca2+ homeostasis and cardiac function (Zhang et al. 2003). In contrast, the PI3K (phosphatidylinositol 3-kinase)/Akt–protein kinase B (PKB) cascade is a major component of physiological or adaptive remodelling. Akt enhances protein synthesis, inhibits apoptosis and stimulates angiogenesis (Selvetella et al. 2004). Akt overexpression causes concentric hypertrophy with preserved systolic function; Akt knockout prevents swimming-induced cardiac hypertrophy but not pathological hypertrophy induced by transverse aortic constriction (DeBosch et al. 2006). In a recent issue of The Journal of Physiology, Oliveira et al. presented data suggesting that aerobic training prevents the development of maladaptive hypertrophy in a genetically defined mouse HF model (Oliveira et al. 2009). Mice with knockout of α2A and α2C receptors (ARKO mice) develop severe cardiac dysfunction, exercise intolerance and increased mortality by 7 months of age (Brum et al. 2002). Oliveira et al. compared ARKO mice exposed to an aerobic exercise training program between 5 and 7 months of age to sedentary ARKO mice. Exercise training prevented functional deterioration, along with calcineurin/NFAT activation, while leaving unchanged CaMKII and Akt signalling, neither of which were altered in control ARKO mice. The authors suggest that exercise training may prevent pathological hypertrophy by deactivating calcineurin/NFAT pathways. This concept presents the exciting possibility of preventing or reversing HF by exercise training-induced suppression of adverse remodelling. Several other studies point to benefits from exercise training in experimental HF. Miyachi et al. (2009) noted that exercise training attenuated HF and dramatically improved survival in Dahl salt-sensitive hypertensive rats. Akt signalling was enhanced, while ERK and p38 mitogen-activated protein kinase phosphorylation was reduced. Garciarena et al. (2009) similarly showed improved cardiac function, reduced fibrosis, apoptosis and calcineurin activity, but no change in Akt, with endurance training of spontaneously hypertensive rats (SHRs). In post-myocardial infarction rat models, exercise training improved contractility (albeit marginally), reduced left-ventricular dimensions and restored β-adrenergic signalling (Leosco et al. 2008), while reducing cardiomyocyte hypertrophy and restoring Na+–Ca2+ exchange expression to control values (Wisloff et al. 2002). Does exercise training improve experimental HF by reducing calcineurin activation, as suggested by Oliveira et al.? While this notion is entirely plausible and quite attractive, other possibilities must be considered. ARKO mice have prominent sympathetic overactivity (Brum et al. 2002), probably due to loss of negative feedback from presynaptic α2A and α2C-adrenoceptors that control the rate of noradrenaline release from sympathetic nerve terminals (Hein et al. 1999). Chronic sympathetic overstimulation of cardiac adrenoceptors, particularly β1 receptors, induces progressive maladaptive remodelling and HF (Engelhardt et al. 1999). Exercise training alters cardiac autonomic balance, reducing sympathetic outflow to the heart (Mueller 2007). Thus, the primary mechanism by which exercise training prevents adverse remodelling in ARKO mice may be attenuation of the inducing stimulus, cardiac sympathetic overstimulation. This notion is consistent with the beneficial actions of β-adrenoceptor antagonists in ARKO mice (Bartholomeu et al. 2008). If reduced sympathetic outflow is the mechanism by which exercise training benefits ARKO mice, reduced calcineurin activation may simply reflect upstream attenuation of adverse remodelling, as would the suppression of angiotensin-converting enzyme and angiotensin II upregulation observed in exercise-trained ARKO mice (Pereira et al. 2009). Sympathetic overactivity is also prominent in SHRs (Nagase et al. 1996) and Dahl salt-sensitive rats (Peuler et al. 1989), potentially implicating favourable autonomic effects in the benefits of exercise training in these models as well. A number of studies have demonstrated the value of exercise training in HF patients. The most significant benefits have been seen in functional capacity, symptoms and exercise performance (Conraads et al. 2004; O’Connor et al. 2009; Patwala et al. 2009), consistent with general training effects. Improvements in cardiac function have been limited and statistically non-significant. A large randomized controlled trial observed small (10–15%) decreases in mortality and hospitalization rates only after adjusting for prognostic baseline factors (O’Connor et al. 2009). Thus, any potential benefits on the natural history of clinical HF seem small. Sympathetic hyperactivity does appear to play a role in clinical HF (Cohn et al. 1984) and exercise training clearly reduces sympathetic outflow in man (Roveda et al. 2003). Most of the same signalling pathways operative in animal models are implicated in clinical HF. Why then have the promising results of experimental studies like those of Oliveira et al. not translated into comparable improvements in cardiac remodelling and outcome with exercise training in man? One possibility is that the pathophysiology of clinical HF is more complex than that of single intervention-based animal models, and that exercise training is simply mechanistically incapable of preventing or reversing adverse remodelling in man. There are, however, other more encouraging possibilities. One is that earlier exercise intervention may be needed to obviate progression to irreversible damage in clinical HF. A second possibility is that more intense exercise programs may be needed to optimize the effect (Nilsson et al. 2008). Finally, different mechanisms may be involved in different clinical forms of HF, so that exercise training may need to be customized and targeted to the type of HF, according to specific characterizing clinical features. More research is needed to define better the mechanisms underlying different types of HF, to identify biomarkers of predictive and mechanistic significance, and to establish better the effects of different forms of exercise at the molecular level and their interactions with mechanisms of cardiac remodelling. With properly conceived and executed research, we may eventually be able to use exercise training not only to improve performance and well-being of HF patients, but also to improve their long-term outlook. This work was supported by the Canadian Institutes of Health Research (MGP-6957, MOP-68929), the Leducq Foundation (ENAFRA, 07/CVD/03), the Plan Nacional de Investigación Científica, Desarrollo e Innovación Tecnológica, Spanish Health Ministry (CM06/00189); and a research-abroad fellowship from the Sociedad Española de Cardiología: Beca para formación en centros extranjeros (Spain).

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.