Sympathetic Neural Overdrive, Aortic Stiffening, Endothelial Dysfunction, and Impaired Exercise Capacity in Severe COVID-19 Survivors: A Mid-Term Study of Cardiovascular Sequelae

Background:

COVID-19 has become a dramatic health problem during this century. In addition to high mortality rate, COVID-19 survivors are at increased risk for cardiovascular diseases 1-year after infection. Explanations for these manifestations are still unclear but can involve a constellation of biological alterations. We hypothesized that COVID-19 survivors compared with controls exhibit sympathetic overdrive, vascular dysfunction, cardiac morpho-functional changes, impaired exercise capacity, and increased oxidative stress.

Methods:

Nineteen severe COVID-19 survivors and 19 well-matched controls completed the study. Muscle sympathetic nerve activity (microneurography), brachial artery flow-mediated dilation and blood flow (Doppler-Ultrasound), carotid-femoral pulse wave velocity (Complior), cardiac morpho-functional parameters (echocardiography), peak oxygen uptake (cardiopulmonary exercise testing), and oxidative stress were measured ~3 months after hospital discharge. Complementary experiments were conducted on human umbilical vein endothelial cells cultured with plasma samples from subjects.

Results:

Muscle sympathetic nerve activity and carotid-femoral pulse wave velocity were greater and brachial artery flow-mediated dilation, brachial artery blood flow, E/e′ ratio, and peak oxygen uptake were lower in COVID-19 survivors than in controls. COVID-19 survivors had lower circulating antioxidant markers compared with controls, but there were no differences in plasma-treated human umbilical vein endothelial cells nitric oxide production and reactive oxygen species bioactivity. Diminished peak oxygen uptake was associated with sympathetic overdrive, vascular dysfunction, and reduced diastolic function in COVID-19 survivors.

Conclusions:

Our study revealed that COVID-19 survivors have sympathetic overactivation, vascular dysfunction, cardiac morpho-functional changes, and reduced exercise capacity. These findings indicate the need for further investigation to determine whether these manifestations are persistent longer-term and their impact on the cardiovascular health of COVID-19 survivors.

Novelty and Relevance.

What Is New?

• Severe COVID-19 survivors experience greater sympathetic neural activity.

• Aortic stiffening is associated with greater sympathetic neural activity in severe COVID-19 survivors.

• Endothelium-dependent dysfunction in severe COVID-19 survivors is related to reduced antioxidant activity, but not by increases in endothelial cell reactive oxygen species bioactivity or reductions in nitric oxide production.

• Diminished aerobic exercise capacity in severe COVID-19 survivors is related to higher sympathetic neural activity, vascular dysfunction, and attenuated cardiac diastolic function.

What Is Relevant?

• COVID-19 survivors are at increased risk for cardiovascular diseases; this outcome may be related to exaggerated sympathetic activity, aortic stiffening, endothelial dysfunction, attenuated vascular conductance, reduced cardiac diastolic function, and marked reduction in aerobic exercise capacity.

• Therapeutic strategies are strongly needed to restore or alleviate cardiovascular damages induced by SARS-CoV-2 infection.

Clinical/Pathophysiological Implications?

• Elevated sympathetic neural activity, reduced vascular function, and attenuated exercise capacity seem to have an important role in the pathophysiology of cardiovascular sequelae in COVID-19 survivors. These findings indicate that therapeutic strategies to restore autonomic and vascular dysfunctions and attenuated exercise capacity should be pursued to design clinical interventions for the care of patients who have survived COVID-19.

Summary

• Severe COVID-19 survivors exhibited greater sympathetic neural activity, aortic stiffening, vascular endothelial dysfunction, reduced vascular conductance, and marked reduction in exercise capacity. Also, these patients have reduced antioxidant activity, despite no changes to reactive oxygen species and nitric oxide. Finally, the reduced exercise capacity in severe COVID-19 survivors is significantly correlated to sympathetic neural overdrive, vascular dysfunction, and reduced cardiac diastolic function.

COVID-19, caused by SARS-CoV-2, has become one of the more dramatic health problems in this century.¹ COVID-19 has proven to be a multiorgan disease with long-term deleterious effects on the cardiovascular system.² Notably, COVID-19 survivors are at increased risk for cardiovascular diseases within 1-year after SARS-CoV-2 infection onset.³

Recent findings show that COVID-19 survivors have inappropriate sinus tachycardia⁴ and postural orthostatic tachycardia syndrome,⁵ indicating autonomic dysfunction. Importantly, young adults recovering from mild COVID-19 infection exhibit high resting muscle sympathetic nerve activity (MSNA),⁶ which is an independent predictor of mortality in several disease states.^7,8 However, these patients were tested early after the COVID-19 diagnosis (3–8 weeks). Thus, it remains unknown whether these autonomic changes are present in patients who required hospitalization to treat severe COVID-19 and if they persist beyond 8 weeks.

SARS-CoV-2 infects endothelial cells through the viral protein (S) that binds to the ACE2 (angiotensin-converting enzyme 2) receptor, leading to endothelial injury, characterized as endotheliitis.⁹ Findings from autopsies show endothelial cell damage in COVID-19 patients, suggesting that vascular alterations are present beyond the acute phase of the illness.⁹ A translational study revealed that alterations in the erythrocyte function induces vascular dysfunction during acute infection in COVID-19 patients, but not 4 months after infection.¹⁰ Endothelial dysfunction is mediated, in part, by alterations in circulating factors in plasma,¹¹ elevated reactive oxygen species (ROS), and reduced nitric oxide (NO) bioavailability, which all may contribute to vascular dysfunction in patients post COVID-19.

COVID-19 survivors have reduced peak oxygen uptake (peak V̇O₂).¹² This diminished aerobic exercise capacity is traditionally related to a decline in the function and integration of pulmonary, circulatory, and musculoskeletal systems.¹³ Because COVID-19 is a multisystem disease, reduced peak V̇O₂ in survivors may be associated with changes in MSNA, vascular function, peripheral muscle blood flow, and cardiac function. However, to date, these physiological responses have not been explored comprehensively in severe COVID-19 survivors.

With this background in mind, we sought to determine whether severe COVID-19 survivors exhibit sympathetic neural overdrive, impaired vascular function (in vivo and ex vivo), cardiac function alterations, and reduced aerobic exercise capacity. We tested the hypothesis that severe COVID-19 survivors would have greater neural sympathetic activity, vascular dysfunction, subclinical cardiac morpho-functional alterations, and marked reduction in exercise capacity.

Methods

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request. All experimental procedures and measurements were conducted in accordance with the Declaration of Helsinki and were approved by the Research Committee of the D’Or Institute for Research and Education (IDOR, CAAE 31468020.1.0000.524). The nature, benefits, and risks of all study procedures were explained to participants, and their written informed consent was obtained before participation in the study.

Participants

COVID-19 survivors who were hospitalized (wards or intensive care units) during acute SARS-CoV-2 infection and control subjects without SARS-CoV-2 infection were included in the study. Our study was performed between September 2020 and October 2021, in which the predominant variants in Brazil were Beta and Gama.¹⁴ A detailed description is available in the Supplemental Material.

Measurements

Detailed descriptions of all measurements mentioned below are available in the Supplemental Material.

COVID-19 Symptom Severity

Severe COVID-19 survivors completed a COVID-19 symptom severity survey on the study day as described previously.⁶

Muscle Sympathetic Nerve Activity

MSNA was measured for 15 minutes using the microneurography technique as described previously.^15,16

Arterial Stiffness

Aortic arterial stiffness was evaluated by carotid-femoral pulse wave velocity (CFPWV), using tonometry.¹⁷ Carotid stiffness and intima media thickness were evaluated using vascular ultrasonography.¹⁸

Endothelial-Dependent Function

Brachial artery flow-mediated dilation (BAFMD) was evaluated according to the current guidelines.^19,20

Brachial Artery Blood Flow, Vascular Conductance, and Vascular Resistance

After 20 minutes at rest, brachial artery blood flow, vascular conductance, and vascular resistance were evaluated with ultrasonography in the left arm of the supine patient.²¹

Plasma Exposure Experiments for Endothelial Cell NO Production and ROS Bioactivity

Human umbilical vein endothelial cells (Lonza; used after 2–4 passages) were treated with 10% human plasma (in media), and acetylcholine-stimulated NO production and basal ROS bioactivity were assessed as previously described.^22–24

Cardiac Morpho-Functional Evaluation

Cardiac morpho-functional evaluation of all patients was performed using transthoracic echocardiography (Vivid E9, General Electric, Horten, Norway), according to the current guidelines.²⁵

Cardiac Autonomic Function and Baroreflex Sensitivity

Finger photoplethysmography (FinometerPro, Finapress Medical Systems, Amsterdam-NL) was used to assess cardiac autonomic function and baroreflex sensitivity.^26–29

Cardiopulmonary Exercise Testing

Maximal exercise capacity was determined by means of cardiopulmonary exercise testing on a cycle ergometer with a ramp protocol.³⁰

Oxidative Stress

Circulating oxidative stress markers were measured in the blood samples.^31–33

Nitrite Concentration

Nitrite, a metabolite of NO, was determined in plasma.³⁴

Data Analysis

All data were evaluated by experienced investigators blinded to the study patients (see Supplemental Material).

Statistical Analysis

All analyses were performed and figures were created using GraphPad Prism 8.0. Data are presented as mean±SD. Significance was set at P≤0.05 (see Supplemental Material).

Results

Participants

Forty-nine participants (control subjects=27; COVID-19 survivors=22) were enrolled in the study. Eight control subjects were excluded, 6 for having positive serology for IgG or IgM and 2 for dropping out. Three COVID-19 patients were excluded, 1 for pregnancy and 2 for dropping out. Thus, 38 participants completed the study, 19 controls and 19 COVID-19 survivors.

The physical and clinical characteristics of participants who completed the study are reported in Table. The groups did not differ in age, body mass index, systolic blood pressure, diastolic blood pressure, heart rate (HR), total cholesterol and fractions, glucose, urea, sodium, potassium, pro-BNP, and troponin I (P>0.05 for all variables). In controls, 1 patient had hypertension,³⁵ and in COVID-19 survivors, 1 patient had both hypertension and type 2 diabetes. Moreover, COVID-19 survivors were tested approximately 3 months after hospital discharge. All control subjects had negative serology for immunoglubulin G and immunoglubulin M and were free of signs and symptoms of COVID-19.

Table.

Clinical and Physical Characteristics in Control Subjects and COVID-19 Survivors

Muscle Sympathetic Nerve Activity

To investigate whether severe COVID-19 survivors exhibit exaggerated sympathetic neural activity, we evaluated MSNA by microneurography. MSNA was not obtained in 1 control and 1 COVID-19 survivor. Thus, in this study MSNA data are presented for 18 controls and 18 COVID-19 survivors. Examples of original nerve and BP recordings per 60 seconds from 1 control and 1 COVID-19 survivor are shown in Figure 1A. MSNA burst frequency was 66% greater in COVID-19 survivors than in controls (P=0.0001, Figure 1B), and MSNA burst incidence was 61% greater (P=0.002, Figure 1C).

Figure 1.

Sympathetic neural overdrive in severe COVID19 survivors. A, Original recordings of muscle sympathetic neve activity (MSNA) and blood pressure (BP) in one control and one COVID-19 survivor. MSNA burst frequency (B) and incidence in controls and COVID-19 survivors (C).

Aortic Arterial Stiffness and Central Blood Pressure

Aortic arterial stiffness and central blood pressure were determined by PWV. CFPWV was not obtained in 1 control and 2 COVID-19 survivors. Thus, Figure 2A and Table S2 present data for 18 controls and 17 COVID-19 survivors. CFPWV was 16% higher in COVID-19 survivors than in controls (P=0.04, Figure 2A). Importantly, our data revealed an association between CFPWV and MSNA burst frequency (r=0.46, P=0.006, Figure S1A) or MSNA burst incidence (r=0.36, P=0.04, Figure S1B). After controlling for confounding factors (mean blood pressure and HR), MSNA burst frequency remained significantly correlated (r=0.42, P=0.01 and r=0.38, P=0.03, respectively, Figure S1A). Similar results were obtained for MSNA burst incidence (r=0.38, P=0.047 and r=0.31, P=0.052, respectively, Figure S1B). There were no differences between groups for central systolic blood pressure, central diastolic blood pressure, and AIX (P>0.05, Table S2).

Figure 2.

Vascular dysfunction in severe COVID19 survivors. Carotid-femoral pulse wave velocity (CFPWV; A), brachial artery flow-mediated dilation (BAFMD; B), BAFMD adjusted to area under the curve of shear rate (BAFMD/ area under the curve of shear rate; C), and brachial artery vascular conductance (BAVC) in control subjects and COVID-19 survivors (D).

Carotid Artery Vascular Parameters

Carotid artery vascular parameters were evaluated by ultrasonography. COVID-19 survivors versus controls did not have differences in intima media thickness, diameter, wall:lumen ratio, distension, Young’s modulus, distensibility, compliance, and stiffness (P>0.05 for all variables, Table S3).

Endothelial Function (In Vivo)

To assess the impact of SARS-CoV-2 on endothelial function (in vivo), BAFMD was assessed. BAFMD% was 45% lower in COVID-19 survivors compared with controls (P=0.0013, Figure 2B), and BAFMD/AUCSR was 48% lower (P=0.004, Figure 2C). Also, BAFMD was 41% lower in COVID-19 survivors than in controls (P=0.003, Table S4), and BAFMD adjusted for baseline diameter was 41% lower (P=0.01, Table S4). No changes were observed between groups in resting diameter, peak diameter, time to peak, AUCSR (P>0.05 to all variables, Table S4).

Endothelial Function (Ex Vivo)

To assess the role of factors in circulation as potential mechanisms of endothelial dysfunction in COVID-19 survivors, we assessed NO production and ROS bioactivity in human umbilical vein endothelial cells in culture treated with plasma of both controls and COVID-19 survivors. Examples of original fluorescent images of NO production and ROS activity from 1 control and 1 COVID-19 survivor are presented in Figure 3A and 3B, respectively. Neither NO production (P=0.75, Figure 3C) nor ROS bioactivity (P=0.70, Figure 3D) in human umbilical vein endothelial cells treated with plasma were different between groups.

Figure 3.

Ex vivo vascular function in severe COVID-19 survivors. Original fluorescent images of human umbilical vein endothelial cell nitric oxide (NO) production (A) and reactive oxygen species (ROS) activity in 1 control and 1 COVID-19 survivor, following a 24-h incubation with plasma from participants (B). Acetylcholine-stimulated NO production (C) and ROS activity in controls and COVID-19 survivors (D).

Hemodynamic Control

Hemodynamic control parameters are presented in Figure 2D and Table S5. As expected, brachial artery vascular conductance was 34% lower in COVID-19 survivors than in controls (P=0.02, Figure 2D). Brachial artery blood flow tended to be 29% lower in COVID-19 survivors compared with controls (P=0.056, Table S5), and brachial artery vascular resistance was 46% greater in COVID-19 survivors (P=0.03, Table S5). There were not differences between groups for systolic blood pressure, diastolic blood pressure, mean blood pressure, and HR (P>0.05 to all variables, Table S5).

Cardiac Morpho-Functional Parameters

Cardiac morpho-functional data assessed by echocardiography are shown in Table S6. COVID-19 survivors compared with controls had greater left atrium dimension, left ventricular (LV) mass index, myocardial performance index, and E/e′ ratio (P<0.05 to all variables). Also, COVID-19 survivors had a smaller E/A ratio than in controls (P=0.05). No changes were observed in left atrium volume, LV end-diastolic volume, LV end-systolic volume, LV ejection fraction, LV global longitudinal strain, myocardial work, RV dimension, tricuspid annular plane systolic excursion, RV fractional area change, S′, RV longitudinal strain, and pulmonary artery systolic pressure (P>0.05 for all variables).

Cardiac Autonomic and Arterial Baroreflex Functions

Cardiac autonomic and arterial baroreflex functions are displayed in Table S7. No changes were observed between COVID-19 survivors versus controls regarding total power (TP), low frequency, high frequency, low-frequency in normalized units, high-frequency in normalized units, low frequency/high frequency, and α index (P>0.05 for all variables).

Cardiopulmonary Response to Exercise Testing

To test whether infection by SARS-CoV-2 is associated with reduced aerobic exercise capacity in COVID-19 survivors, the participants were assessed by cardiopulmonary exercise testing. All patients were affirmed to have performed a maximal volitional effort up to their limit. At peak exercise, relative V̇O₂ was 29% lower in COVID-19 survivors than in controls (P=0.0006, Figure 4A), despite similar respiratory ratio between groups (P=0.11, Figure 4B). Moreover, COVID-19 survivors had lower V̇CO₂, HR, and load reached compared with that in controls (P<0.05 for all variables, Table S8). No differences between groups were observed in systolic blood pressure, diastolic blood pressure, V̇_E, V̇_E/V̇O₂, V̇_E/VCO₂, V̇_E/V̇CO₂ slope, and exercise time (P>0.05 for all variables, Table S8).

Figure 4.

Aerobic exercise capacity in severe COVID-19 survivors. Peak oxygen uptake (peak V̇O₂; A) and respiratory rate (RR) during cardiopulmonary exercise testing in controls and COVID-19 survivors (B).

Reduced Exercise Capacity: Role of Neurovascular Dysfunction and Echocardiographic Changes

Next, to assess the role of autonomic, vascular, cardiac dysfunctions in the reduced exercise capacity of COVID-19 survivors, we used the Pearson correlation. Correlation analyses revealed associations between peak V̇O₂ and MSNA frequency (r=−0.43, P=0.007, Figure 5A), peak V̇O₂ and MSNA incidence (r=−0.36, P=0.05, Figure 5B), peak V̇O₂ and CFPWV (r=−0.45, P=0.008, Figure 5C). To our surprise, no association existed between peak V̇O₂ and BAFMD (r=0.22, P=0.18, Figure 5D) and peak V̇O₂ with BAFMD/area under the curve of shear rate, brachial artery blood flow, or brachial artery vascular conductance (P>0.05 to all variables, data not presented). Finally, peak V̇O₂ was significantly correlated with E/e′ ratio (r=−45, P=0.006, Figure 5F), but not with LV mass index (r=−31, P=0.07, Figure 5E).

Figure 5.

Diminished aerobic exercise capacity is related to neurovascular and cardiac function changes in severe COVID-19 survivors. Correlations between peak oxygen uptake (peak V̇O₂; A) with muscle sympathetic nerve activity (MSNA) frequency, MSNA incidence (B), carotid-femoral pulse wave velocity (CFPWV; C), brachial artery flow-mediated dilation (BAFMD; D), left ventricular mass index (LV mass index; E), and E/e′ ratio (F).

Oxidative Stress and Nitric Oxide

Systemic circulating markers of oxidative stress were measured in plasma. Catalase activity and ferric reducing/antioxidant power were significantly lower in COVID-19 survivors than in controls (P=0.04, Figure S2A and P=0.028, Figure S2B, respectively). There were no differences in superoxide dismutase, hydrogen peroxide (H₂O₂), NADPH oxidase (nicotinamide adenine dinucleotide phosphate oxidase), and thiobarbituric acid reactive substances (P>0.05 to all variables, Figure S2C through S2E). Also, nitrite, an NO metabolite, measured in plasma was not different between groups (P=0.16, Figure S3).

COVID-19 Symptom Severity

The symptom scores for COVID-19 survivors are presented in Table S1 (see the Supplemental Material for details). No relationship between total severity score and outcome variables (MSNA, CFPWV, BAFMD, E/e′ ratio, and peak VO₂) were observed (P>0.05 to all relationships, Figure S4).

Discussion

The main findings of the present study are that severe COVID-19 survivors have greater sympathetic neural activity, elevated central artery stiffness, vascular endothelial dysfunction, reduced vascular conductance, subclinical cardiac morpho-functional alterations, reduced exercise capacity, and decreased antioxidant activity. Moreover, the reduced exercise capacity was associated with sympathetic neural overdrive, vascular dysfunction, and cardiac diastolic function alterations.

In a previous study, Stute et al⁶ reported that nonhospitalized young adults recently recovered from a mild SARS-CoV-2 infection have increased resting MSNA. Our study extends these observations and shows that MSNA is substantially increased 3 months after hospitalization in severe COVID-19 survivors. Moreover, MSNA burst frequency and incidence levels in COVID-19 survivors are greater than those observed in healthy subjects aged >60 years.^36,37 In particular, MSNA values were similar to those previously observed using the same methodology in patients with heart failure with reduced ejection fraction treated with beta-blockers.²⁶ This autonomic dysfunction may be critical to COVID-19 survivors because a high MSNA is an independent predictor of mortality in many disease states.^7,8

Because COVID-19 survivors have sympathetic overactivity, an elevated resting HR would be expected. However, this was not the case, and HR levels were not different between COVID-19 survivors and controls. Explanations for these findings are outside of the scope of this study. However, it is possible that cardiac norepinephrine spillover (and thus cardiac sympathetic drive) was not elevated in the patients of the present study, especially, because they had overweight/obesity.³⁸

Other mechanisms may be behind this higher sympathetic activity in COVID-19 survivors. Based on the hypothesis that SARS-CoV-2 invades carotid body via ACE-2 and that carotid body may be the route through which the virus enters the central nervous system, it is possible to suspect that COVID-19 survivors have an impairment in chemosensitivity. Autopsy data of COVID-19 patients confirm the direct invasion of SARS-CoV-2 in the carotid body, provoking microthrombosis, blood congestion, and microhemorrhages, which could potentially affect chemoreception.³⁹ Furthermore, high levels of angiotensin II may stimulate the carotid body, leading to an increase in peripheral hypoxic chemosensitivity and sympathetic output.⁴⁰ Alternatively, angiotensin II also activates macrophages and other immune cells to produce an inflammatory cytokine storm, increasing centrally sympathetic outflow through the hypothalamic-pituitary-adrenal axis.⁴¹ angiotensin II also works in the peripheral sympathetic nerve terminals to elicit sympathetic neurotransmission.⁴² However, involvement of the carotid body is speculative, but is the subject of ongoing investigation.

It has been shown that an ~1-m/s elevation in PWV is associated with a 15% greater risk of cardiovascular events.⁴³ Our data show that the CFPWV was 1.12 m/s greater in COVID-19 survivors compared with controls, suggesting that increases in arterial stiffness may contribute to the increase in cardiovascular risk in COVID-19 survivors. Several mechanisms are associated with aortic stiffening, including the loss of elastin proteins and increased collagen deposition.⁴⁴ However, these structural changes to arteries are unlikely to occur in the short term. More recently, it has been demonstrated that α-adrenergic-mediated constriction of vascular smooth muscle increases arterial stiffness.⁴⁴ In our study, we found a positive correlation between CFPWV and MSNA burst frequency and incidence, indicating that increase in vascular smooth tone α-adrenergic-mediated may be a potential mechanism of aortic stiffening in COVID-19 survivors.

Our findings are consistent with recent reports showing that endothelium-dependent dysfunction is present after resolution of COVID-19.^45,46 Because BAFMD is a measure that is endothelium-dependent and largely mediated by NO, we hypothesized that reduced BAFMD in COVID-19 survivors is linked to decreases in NO bioavailability and increases in oxidative stress. Using an innovative ex vivo cell culture model to determine the effects of circulating factors in subject plasma on endothelial cell function, we did not find differences in endothelial cell NO production or ROS bioactivity with exposure of plasma from COVID-19 survivors versus control subjects. In accordance with our findings, Mahdi et al¹⁰ also demonstrated that persistent impaired endothelial function at 4 months post COVID-19 was not mediated by changes in circulating factors in plasma. There is no definitive explanation for endothelial dysfunction in COVID-19 survivors in our study, but sympathetic overactivation may play a role. In fact, it has been shown that increases in sympathetic nervous system activity (assessed by MSNA) decreases endothelial function in healthy subjects⁴⁷ and autonomic blockade reverses endothelial dysfunction in obesity-associated with hypertension.⁴⁸ Also, others have shown that endothelial injury markers occur, such as endothelial cell activation, sustained inflammation, and coagulation activation. Of note, recovered COVID-19 patients have greater circulating endothelin-1, elevated IL-6 levels, and increased coagulation activity (eg, von Willebrand factor).⁴⁹ Likewise, damage to the glycocalyx can also occur. This luminal surface layer on the vascular endothelium is a key regulator of endothelial cell homeostasis.⁵⁰ Therefore, the injured glycocalyx might compromise endothelial function. In fact, recent works have reported that endothelial dysfunction 4 and 12 months after COVID-19 is linked to a great perfused boundary region of the sublingual arterial microvessels, indicating impaired endothelial glycocalyx.^51,52

The findings of the current study revealed a subclinical alteration in cardiac diastolic function in COVID-19 survivors compared with controls. Also, we observed that COVID-19 survivors have higher myocardial performance index and some subclinical morphological changes, characterized by greater left atrium dimension and LV mass index. It is known that the presence of risk factors and cardiovascular diseases are associated with diastolic dysfunction.^53,54 For example, arterial hypertension progressively alters diastolic function by cardiac hypertrophic remodeling.⁵³ Our study was conducted on COVID-19 survivors free of cardiovascular disease and the groups were blood pressure-matched. In contrast with previous reports, we did not find changes in morpho-functional parameters of RV. These differences may be due to the COVID-19 severity. Although our study was conducted in patients who had a severe illness, in previous studies most patients had critical COVID-19 and needed invasive mechanical ventilation, suggesting impaired pulmonary structure and function.^55,56 Thus, it is possible that persistent RV dysfunction in recovered COVID-19 patients is not simply attributed to direct myocardial injury. In fact, we did not observe differences between groups in cardiac biomarkers (pro-BNP and Troponin I). Furthermore, recent works have shown that cardiac changes induced by SARS-CoV-2 in survivors persist for a short period and should be resolved after acute infection.⁵⁷ Because in the present study, the patients were tested mid-term after hospital discharge, it is possible that cardiac changes were transient and returned to baseline values.

Recent reports have observed that diminished exercise capacity in COVID-19 survivors is linked mainly to skeletal muscle deconditioning rather than to alterations in ventilation or circulation.^58–60 In accordance with these findings, our data clearly show that COVID-19 survivors have attenuated peak V̇O₂ on cardiopulmonary exercise testing compared with controls, despite similar physical exertion and preserved ventilatory and circulatory responses. Importantly, our findings extend the previous findings by showing that reduced peak V̇O₂ was associated with both elevated MSNA and vascular dysfunction, evoking vasoconstriction, and compromised muscle blood flow. These responses may have caused a dysfunction in O₂ transport, attenuating O₂ delivery to skeletal muscle and therefore exercise capacity in COVID-19 survivors. This is consistent across different populations, such as those with heart failure⁶¹ and chronic obstructive pulmonary disease,⁶² where reduced exercise capacity is linked to dysfunction in the convective transport of O₂ to skeletal muscle during exercise.

The current study has some limitations. Our results show cardiovascular alterations in severe COVID-19 survivors approximately 3 months after hospital discharge. However, it is necessary to confirm whether they are transient or sustained per long-term postacute infection. We cannot consider that the findings observed in the present study are a unique phenomenon to SARS-CoV-2 infection once that other viral infections (eg, influenza) have been associated with cardiac involvement.⁶³ However, previous investigations have shown that cardiovascular sequelae are more evident in COVID-19 than in influenza.⁶³ The study sample was predominantly composed of men. Thus, our findings may not identify whether post-COVID-19 cardiovascular manifestations are more pronounced in women than in men. However, recent evidence suggests that SARS-CoV-2 effects on the cardiovascular system are sex-independent.^45,46 Our study have assessed cardiovascular variables at rest, and we believe that assessment of the cardiovascular responses during excitatory stimulation (eg, mental stress and tilt test) may provide a better understanding of SARS-CoV-2-induced cardiovascular alterations in patients with long COVID-19. However, it has been shown that COVID-19 survivors who have manifested the mild form of the disease already have higher MSNA during the tilt test than controls have.⁶ Thus, it is possible that the severe COVID-19 survivors of the current study also have abnormal cardiovascular responses to stressor stimulus. We did not assess whether participants had concurrent sleep apnea, known to produce similar autonomic and vascular changes and linked to greater body weight and age.^64,65 However, when our results were adjusted for weight, body mass index, and age using an ANCOVA model, the differences between groups were maintained (data not shown). Finally, we did not measure other mediators, such as pro-inflammatory profile, endotehetin-1, ANG II, and coagulation factors that could provide more information about mechanisms to explain our findings.

Taken together, our findings revealed that severe COVID-19 survivors have greater sympathetic neural activity, augmented central arterial stiffness, reduced endothelium-dependent function, decreased vascular conductance, subclinical cardiac morpho-functional changes, and a marked reduction in aerobic exercise capacity. Furthermore, the reduced exercise capacity was associated with sympathetic neural overdrive, vascular dysfunction, and attenuated cardiac diastolic function. These findings strongly indicate the need for further investigation to determine whether these manifestations are long-term persistent and their impact on the cardiovascular health of COVID-19 survivors.

Perspectives

Sympathetic neural overdrive, vascular dysfunction, and reduced exercise capacity are independent predictors of cardiovascular disease-related mortality.^7,43,66 We demonstrated that survivors of severe COVID-19 who required hospitalization have an aberrant sympathetic activity, endothelium-dependent dysfunction, increased arterial stiffness, decreased vascular conductance, subclinical cardiac morpho-functional changes, reduced exercise capacity, and attenuated antioxidant activity. Diminished exercise capacity was significantly associated with sympathetic neural overdrive, vascular dysfunction, and reduced cardiac diastolic function. Collectively, our results provide support for the idea that COVID-19 survivors may be at increased risk for cardiovascular diseases. Thus, it is an urgent biomedical research priority to establish therapeutic strategies (eg, inspiratory muscle strength training)⁶⁷ that can restore or alleviate these cardiovascular manifestations.

Article Information

Acknowledgments

The authors are grateful for the time and effort of all participants. Elaine Lagonegro, Fabiana Panham, Anna Lopes, and NAPES TEAM (IDOR-SP) for the administrative support of the work.

Sources of Funding

A. Sales is supported by Fundação de Amparo à Pesquisa do Estado do Rio de Janeiro (FAPERJ, E-26/211.526/2021) and D’Or Institute for Research and Education. C. Negrão and M. Rondon are supported by Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq, 304697/2020-6 and 313152/2020-9, respectively). M. Rossman is supported by U.S. National Institutes of Health Award K01DK115524. D. Creaghead is supported by U.S National Heart Lung and Blood Institute Award K01HL153326.

Disclosures

None.

Supplemental Material

Figures S1–S4

Tables S1–S8

Supplementary Material

hyp-80-470-s001.pdf ^{(920.2KB, pdf)}