During exercise, the sympathetic nervous system plays a pivotal role in the maintenance of blood pressure (BP) and distribution of blood flow towards active skeletal muscle. Specifically, exercise-induced increases in sympathetic vasoconstrictor discharge promotes venous return to the heart which supports increases in left ventricular (LV) cardiac output (Q˙_c), and vasoconstriction in non-active tissues serves to redistribute blood flow toward active skeletal muscle to meet the metabolic demands of exercise (1). Conversely, locally released metabolic by-products within active skeletal muscle offset sympathetically-mediated vasoconstriction (i.e., functional sympatholysis) to ensure blood supply matches the metabolic requirements of active muscle (2, 3). This phenomenon is particularly important in the resistance arterioles of the muscle microcirculation (4). However, sympathetic nervous system hyper-reactivity and impaired peripheral vascular responses to exercise create an imbalance between vasoconstriction in non-active muscles and vasodilation in active muscles, ultimately resulting in a mismatch between oxygen supply and demand (5–8). Indeed, aberrant sympathetic neural reactivity and peripheral vascular responses to exercise are directly related to exercise intolerance and a reduced peak aerobic capacity or peak oxygen uptake (V˙O_2peak) in certain populations characterized by chronic disease (5–8).

Exercise intolerance is a hallmark of heart failure (HF) with preserved ejection fraction (LVEF ≥ 50%: HFpEF), and is directly related to morbidity, hospitalization, and mortality rates (9). Despite, HFpEF representing ∼50% of all HF cases worldwide (10), most studies have focused on the mechanisms underpinning exercise intolerance in patients with HF with reduced ejection fraction (LVEF ≤ 40%: HFrEF). Thus, the determinants of exercise intolerance in patients with HFpEF remain poorly understood. As a result, pharmacological therapies targeted toward improving exercise capacity are increasingly utilized in patients with HFrEF; however, few evidence-based therapies currently exist for patients with HFpEF (11, 12).

Emerging evidence indicates that the underlying mechanisms of exercise intolerance differ between patients with HFrEF and HFpEF, despite both forms of HF having a lower V˙O_2peak than age-matched controls (5, 7, 13–15). Borlaug et al. (15, 16) found that V˙O_2peak in patients with HFpEF was negatively correlated with peripheral vascular reserve (e.g., systemic vascular resistance, and microvascular function) and positively correlated to cardiac function [e.g., lower peak heart rate (HR), and reduced myocardial contractility]. Bhella et al. (13) observed no significant difference in cardiac function at peak exercise between patients with HFpEF and age-matched controls, as assessed by cardiac power output (Q˙_c × mean BP) and stroke work (stroke volume (SV) × mean BP), which are further indicators of overall cardiac function and major determinants of exercise capacity (17–19), despite patients with HFpEF exhibiting a lower V˙O_2peak. These results indicate that unlike in patients with HFrEF who exhibit cardiac limitations, peripheral, but not central (cardiac), factors may primarily explain the lower V˙O_2peak in patients with HFpEF (7, 20). A potential mechanism underpinning the greater vasoconstrictive response to exercise in patients with HFpEF, might be sympathetic nervous system hyperreactivity, which would lead to attenuate muscle blood flow via excessive vasoconstriction, thus impairing oxygen delivery to the contracting skeletal muscles, ultimately lowering V˙O_2peak.

Assessments of autonomic and vascular function during exercise are typically conducted using either dynamic exercise such as cycling or running, or static (i.e., isometric) exercise such as sustained handgrip. Notably, the physiological responses to these types of exercise are distinct (1, 21). The autonomic mechanisms contributing to the maintenance of BP during exercise, called pressor response (e.g., central command, skeletal muscle metaboreflex, arterial baroreflex, chemoreflex, hormonal response), are engaged in both dynamic and static exercise. However, the pressor response during contractions not only depends on muscle mass but the types of contraction (21): Dynamic exercise involves larger central (cardiac) and peripheral (limb vascular) hemodynamic responses, whereas static exercise better isolates peripheral hemodynamic limitations to exercise. To date, there is a paucity of information summarizing the sympathetic and hemodynamic responses to exercise including types of muscle contraction in patients with HFpEF, and how these neuro-cardiovascular responses relate to exercise intolerance in this patient population.

The current mini review focuses on the sympathetic neural and hemodynamic responses during dynamic and static exercise in patients with HFpEF compared to patients with HFrEF and non-HF controls. Further, we explored the relationship between sympathetic nerve activity and vasoconstriction in patients with HFpEF to advance our understanding of the mechanisms underpinning exercise intolerance in this group of individuals (see Table 1 for a summary of the references).

Sympathetic outflow during exercise can be directly measured from peripheral nerves innervating the muscle vasculature as sympathetic neural discharge (muscle sympathetic nerve activity; MSNA) via microneurography (31, 32), or indirectly via plasma concentrations of norepinephrine (NE) (33–35).

A recent case report (22) assessed MSNA in one patient with HFpEF and one non-HF control (both 74 years, female) during dynamic single-leg cycling [0 W (mild) and 50% of V˙O_2peak (moderate), 2 min/stages], finding that MSNA burst frequency (BF) increased (Δ) in the patient with HFpEF (ΔBF, mild: +13 and moderate: +16 bursts/min), yet was reduced in the non-HF control (mild: −6 and moderate: −7 bursts/min). Notably, previous study has demonstrated that patients with HFrEF also exhibit this paradoxical increase in MSNA BF [mild: +5 ± 4 and moderate: +7 ± 7 bursts/min ± standard deviation (SD)] during the same single-leg cycling protocol (36). These studies collectively indicate that the MSNA responses to dynamic exercise tend to be higher in patients with HFpEF compared to non-HF controls, and possibly patients with HFrEF, but studies comprising a larger cohort are needed to confirm this.

Contrary to the MSNA differences, patients with HFpEF exhibit similar increases in plasma NE concentrations to non-HF controls immediately after maximal-effort upright cycling, despite having a shorter exercise duration (15). Furthermore, Tsuchida et al. (24) compared the plasma NE responses to peak cycling in patients with preserved (≥45%) vs. reduced (<45%) LVEF, finding that patients with reduced LVEF had a lower peak exercise workload and duration than HF patients with preserved LVEF, yet no group differences in plasma NE concentrations were observed. Taken together, these two studies indicate that patients with HFpEF may exhibit greater sympathetic neurohormonal reactivity than non-HF controls, yet lower sympathetic outflow than patients with HFrEF for a given absolute workload. Importantly, plasma NE concentration provides a crude index of overall sympathetic activity. Circulating NE concentrations are dependent on both NE release from adrenergic nerve terminals and eventual removal from the circulation (33–35).

To our knowledge, only one assessment of the sympathetic responses to static handgrip exercise in one patient with HFpEF has been published. Bunsawat et al. (23) measured MSNA from the peroneal nerve in a patient with HFpEF (65 years, female) and a healthy individual (70 years, female) during static handgrip exercise at 30% and 40% of their maximal voluntary contraction (MVC) force for 2 min each. They reported that absolute MSNA BF in the patient with HFpEF tended to be higher at rest and throughout exercise, whereas the increases in MSNA BF (30% MVC: 7 vs. 10; 40% MVC: 22 vs. 24 bursts/min) seemed to be similar between the patient with HFpEF and the healthy control participant, respectively, during the last minute of exercise. Notably, though patients with HFrEF also exhibit greater absolute MSNA BF compared to controls (37), ΔMSNA during 30% MVC handgrip was higher in patients with HFrEF (ΔBF at 1-min: ∼5 ± 12, 2-min: ∼10 ± 12 bursts/min ± SD) than non-HF controls (ΔBF at 1-min: ∼1 ± 6, 2-min: ∼2 ± 4 bursts/min ± SD) (37). These responses in patients with HFrEF appears to be greater compared to that observed by Bunsawat et al. (23) in a patient with HFpEF (ΔBF at 1-min: 2, 2-min: 7 bursts/min). Taken together, these studies suggest that patients with HFpEF may exhibit smaller sympathetic neural reactivity to static exercise than patients with HFrEF, but similar reactivity to non-HF controls. Future studies with larger sample sizes are needed to confirm these preliminary observations.

In summary, patients with HFpEF and HFrEF exhibit greater sympathetic neural reactivity than non-HF controls during dynamic exercise, which may contribute to blunted increases in skeletal muscle blood flow and therefore exercise intolerance. Further, patients with HFpEF perhaps exhibit higher sympathetic responses to dynamic exercise than patients with HFrEF. Conversely, patients with HFpEF seem to exhibit smaller sympathetic responses than patients with HFrEF but similar responses to non-HF controls during static exercise. Sympathetic hyperactivity during exercise in patients with HFrEF is thought to be mainly attributed to exaggerated exercise pressor reflex activation (38, 39) including greater metabo- (37) and mechano-reflex activity (40) due to skeletal muscle abnormalities like muscle atrophy (41–43). Other compensatory responses include attenuated sympathoinhibition by cardiopulmonary (44) and arterial baroreceptors (45), heightened humoral activation (46), and enhanced chemoreflex activity (47). However, not all of these mechanisms may contribute to the heightened sympathetic activity in patients with HFpEF during exercise. The impact of the muscle metaboreflex in HFpEF remains equivocal with some studies demonstrating augmented (48) or unchanged (26) metaboreflex activation compared to non-HF controls, despite the observed muscle atrophy and biomolecular changes in patients with HFpEF (49). Moreover, Seravalle et al. (50) reported little to no baroreflex dysfunction in HFpEF, suggesting that this mechanism for sympathoexcitation in HFrEF may be different in HFpEF. Therefore, more work in larger samples is needed to confirm the augmented sympathoexcitation during dynamic exercise and the maintained sympathetic responses during static exercise in patients with HFpEF, and further clarify the underlying mechanisms.

Although no studies have directly assessed the relationship between peripheral vascular and sympathetic neural responses during exercise in patients with HFpEF, pharmacological studies have indicated that altered excessive sympathetically-mediated vasoconstriction primarily contributes to augmented BP responses during dynamic exercise, leading to exercise intolerance in patients with HFpEF. For instance, Kato et al. (8) treated patients with HFpEF and hypertension using candesartan, an angiotensin II receptor blocker (ARB), for 4 weeks and found that patients with HFpEF, but not hypertensive controls, had reduced peak systolic BP and longer time to exhaustion during maximal dynamic exercise. Further, following 1 year of perindopril [peripherally acting angiotensin converting enzyme (ACE) inhibitor] treatment, patients with HFpEF exhibited improvements in the 6-minute walk test (30). One conceivable mechanism relates to an ARB or ACE-inhibitor mediated suppression of excess sympathetic discharge, which could decrease peripheral vascular resistance and increase oxygen supply to the muscle (55) as well as functional capacity in patients with HFpEF. Although a previous systematic review and meta-analysis by Zheng et al. (12) did not demonstrate improvements in aerobic or functional capacity in patients with HFpEF following the use of ACE-inhibitors and ARBs, they did not include the two studies mentioned here. This highlights the issue that there are only a few studies examining therapies that may help mitigate the aberrant cardiovascular responses to exercise observed in patients with HFpEF. Nonetheless, these data provide some insight that sympathetic overexcitation coupled with exaggerated peripheral vasoconstriction during exercise may be one of the primary factors underlying exercise intolerance in patients with HFpEF.

The concept of functional sympatholysis is the ability of active skeletal muscle to override sympathetic vasoconstrictor drive, ensuring adequate oxygen delivery to the muscle (2, 3). While this physiological response is well-established in healthy humans, it is attenuated in diseases like elevated sympathetic activity (3). To our knowledge, only one study has assessed functional sympatholysis in patients with HF mildly reduced EF (45%), finding that the responsiveness to sympathetic vasoconstrictor (tyramine) during exercise was not altered while the peak vasodilatory capacity of ATP [a counter of sympathetic vasoconstriction elicited by exercise (56)] infusion was reduced compared to the non-HF controls (57). Whether sympatholysis is impaired in patients with HFpEF remains unknown and is an important area to clarify in the future

The current data indicate that during dynamic exercise, patients with HFpEF exhibit higher peripheral vascular resistance, likely attributed to sympathetic hyperreactivity, than non-HF controls and perhaps patients with HFrEF, which may explain the larger increases in exercise BP and limited muscle blood flow. Conversely, during static exercise, patients with HFpEF exhibit smaller pressor responses than non-HF and HFrEF, despite the similar sympathetic responses compared to non-HF, suggesting that mechanisms other than sympathetic vasoconstriction may cause the lower workloads and shorter exercise durations observed in patients with HFpEF (see Table 2 for a summary of the conclusion). Altogether, this review highlights that our understanding about sympathetic neural control of the circulation and hemodynamic exercise in patients with HFpEF are still in its infancy and further work is needed to confirm the current understanding about sympathetic neural and hemodynamic responses to exercise in larger sample sizes. The research is also urgently required to clarify the pathophysiology lying between sympathetic neural activity and vasculature to develop treatment for exercise intolerance in this chronic disease that is growing in prevalence.