key: cord-0285981-hyvxusnc
authors: Durstenfeld, M. S.; Peluso, M. J.; Kaveti, P.; Hill, C.; Li, D.; Sander, E.; Swaminathan, S.; Arechiga, V. M.; Sun, K.; Ma, Y.; Zepeda, V.; Lu, S.; Goldberg, S. A.; Hoh, R.; Win, S.; Kelly, J. D.; Henrich, T. J.; Martin, J. N.; Lee, Y. J.; Aras, M. A.; Long, C. S.; Grandis, D. J.; Deeks, S. G.; Hsue, P. Y.
title: Reduced Exercise Capacity, Chronotropic Incompetence, Inflammation and Symptoms in Post-Acute COVID-19
date: 2022-05-19
journal: nan
DOI: 10.1101/2022.05.17.22275235
sha: 1902bd5360a95d65bf3203ddc9d24bb32b9da6fd
doc_id: 285981
cord_uid: hyvxusnc

BACKGROUND Mechanisms underlying persistent cardiopulmonary symptoms following SARS-CoV-2 infection (post-acute sequelae of COVID-19 "PASC" or "Long COVID") remain unclear. The purpose of this study was to elucidate the pathophysiology of cardiopulmonary PASC using multimodality cardiovascular imaging including cardiopulmonary exercise testing (CPET), cardiac magnetic resonance imaging (CMR) and ambulatory rhythm monitoring. METHODS In the Long-Term Impact of Infection with Novel Coronavirus (LIINC) Cohort, we performed CMR, CPET, and ambulatory rhythm monitoring among adults > 1 year after PCR-confirmed SARS-CoV-2 infection. We used logistic and linear regression to compare those with and without cardiopulmonary symptoms (dyspnea, chest pain, palpitations) adjusting for confounders. RESULTS One hundred twenty individuals were studied, among whom 46 participants (unselected for symptom status) had at least one advanced test performed at median 17 months (IQR 15-18). Median age was 52 (IQR 42-61), 18 (39%) were female, and 6 (13%) were hospitalized for severe acute infection. On CMR (n=39), smaller RV volume and stroke volume and higher extracellular volume were present among those with symptoms, but no evidence of late-gadolinium enhancement or differences in T1 or T2 mapping were demonstrated. We did not find arrhythmias on ambulatory monitoring. In contrast, on CPET (n=39), 13/15 (87%) participants with reduced exercise capacity (<85% predicted) reported cardiopulmonary symptoms or fatigue (p=0.008). Adjusted peak VO2 was 2.7 ml/kg/min lower among those with cardiopulmonary symptoms (95%CI -6.9 to 1.5; p=0.20) or -11% predicted (95%CI -27 to 5, p=0.17). Including fatigue along with cardiopulmonary symptoms, the adjusted difference in peak VO2 was -5.9 ml/kg/min (-9.6 to -2.3; p=0.002) or -21% predicted (-35 to -7; p=0.006). Chronotropic incompetence was the primary abnormality among 9/15 with reduced peak VO2. Adjusted heart rate reserve <80% was associated with reduced exercise capacity (OR 15.6, 95%CI 1.30-187; p=0.03). Those with chronotropic incompetence had higher hsCRP, lower heart rate recovery, and lower heart rate variability suggestive of autonomic dysfunction. CONCLUSIONS Reduced exercise capacity and reduced heart rate response to exercise, and hsCRP are associated with persistent cardiopulmonary symptoms more than 1 year following COVID-19. Chronic inflammation and autonomic dysfunction may underlie cardiopulmonary PASC.

Following acute SARS-CoV-2 infection, some individuals suffer from persistent symptoms called "Long COVID" or post-acute sequelae of COVID-19 (PASC). Although some studies estimate that >30% of individuals experience persistent symptoms after SARS-CoV-2 infection, 1-3 population-based estimates range from 3-12%. 4 Understanding PASC represents a major public health issue given that more than half the US population has been infected. 5 Mechanisms of PASC remain poorly understood, but chronic inflammation, aberrant immune activation, and endothelial dysfunction have been implicated. 6 Characterizing phenotypes of cardiopulmonary PASC with multi-modality cardiac testing may yield insights into potential mechanisms. We previously demonstrated that inflammatory markers and possibly pericardial effusions were associated with symptoms 6 months following SARS-CoV-2 infection. 7 We and others have demonstrated normal echocardiographic cardiac function 3-6 months after COVID- 19 , suggesting that other techniques are needed to identify mechanisms of symptoms. [7] [8] [9] [10] [11] [12] Evaluation with cardiac magnetic resonance imaging (CMR) has also confirmed normal cardiac function and revealed changes in parametric mapping and late gadolinium enhancement but without clear associations with symptoms or consistent differences from controls. [13] [14] [15] [16] [17] Studies using cardiopulmonary exercise testing (CPET) have consistently demonstrated reduced exercise capacity at 3-6 months after SARS-CoV-2 infection with mixed results regarding classification of the etiology of exercise limitation. [17] [18] [19] [20] [21] Finally, to our knowledge, whether arrhythmias explain palpitations in PASC is unknown beyond 3 months. 22, 23 The Long-Term Impact of Infection with Novel Coronavirus (LIINC) study (NCT 04362150) was implemented to study COVID-19 recovery in individuals with confirmed SARS-COV-2 and includes individuals with asymptomatic to severe acute disease. 24 The purpose of this study was to elucidate mechanisms of cardiopulmonary symptoms more than 1 year following SARS-CoV-2 infection by comparing symptomatic and recovered individuals using a multi-modality tests including blood-based markers, echocardiography, cardiac magnetic resonance imaging, cardiopulmonary exercise testing, and ambulatory rhythm monitoring.

As previously reported, the Long-term Impact of Infection with Novel Coronavirus (LIINC) study is a COVID recovery cohort that includes longitudinal symptom assessment. 24 In a subset, we measured biomarkers and performed echocardiograms which we have previously reported. 7 Here we report findings from the subset who have undergone additional cross-sectional cardiopulmonary testing including CPET, CMR, and ambulatory rhythm monitoring.

Participants were included if they had PCR-confirmed SARS-CoV-2 infection, enrolled in LIINC, and agreed to participate in additional cardiopulmonary testing. We invited participants who met the inclusion and exclusion criteria who had completed the echocardiogram study visit in order of date of infection (earliest first) irrespective of symptom status. Participants were excluded who were pregnant (to reduce confounding due to expected changes during pregnancy) or had significant cardiopulmonary disease including congenital heart disease, heart failure, myocardial infarction, or heart surgery. Additionally, those with non-MRI compatible implants or claustrophobia were excluded from CMR; those with estimated GFR <30ml/min/1.73m 2 were excluded from receiving gadolinium contrast. Those unable to exercise on cycle ergometer were excluded from CPET. Demographics including gender, race, ethnicity, income, and education were self-reported by participants. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 19, 2022. ;  https://doi.org/10.1101/2022.05.17.22275235 doi: medRxiv preprint Symptoms Individuals were queried regarding 32 symptoms from the Centers for Disease Control list and from the Patient Health Questionnaire Somatic Symptom Scale. 24 At enrollment, participants completed a structured interview about medical history, characteristics of acute infection, cardiopulmonary diagnoses, and symptoms within the previous two weeks, which were considered COVID-19-related only if new or worsened since infection. We systematically asked about fatigue, shortness of breath, chest pain, palpitations, syncope, edema, and positional symptoms.

A cardiac sonographer performed echocardiograms with a GE VIVID E90 machine at the first study visit using a standardized protocol. Post-processing and measurements were performed by a single echocardiographer using GE EchoPAC software according to ASE guidelines, as we have previously described. 7 

Participants had venous blood collected and processed for serum and plasma on the day of the echocardiogram. Samples were batch processed for measurement of high sensitivity c-reactive protein (hs-CRP; ADVIA® Chemistry CardioPhase™ High Sensitivity C-Reactive Protein assay). Samples were assayed blinded with respect to patient and clinical information, and assay performance was consistent with the manufacturer's specifications. Additionally, immediately prior to the CMR, blood was drawn for measurement of hematocrit for calculation of extracellular volume.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 19, 2022. ;  https://doi.org/10.1101/2022.05. 17.22275235 doi: medRxiv preprint Multiparametric, sequence-standardized, blinded (technician and reader) cardiac magnetic resonance imaging (CMR) was performed at UCSF China Basin with a 3 Tesla system (Premier, General Electric). The CMR protocol consisted of acquisition of the following sequences after multiplane localizers prior to gadolinium injection: fast imaging employing steady state acquisition (FIESTA) cine in the axial short axis planes, pre-contrast T1 mapping sequences using MOLLI 5-(3)-3 technique, as well as pre-contrast T2 mapping at the basal, mid, and apical short axis planes and T2 fat-saturated weighted black blood spin-echo images in the short axis plane. 8-10 minutes after intravenous gadolinium injection, phase sensitive inversion recovery (PSIR) late gadolinium enhancement imaging in short axis full stack, 4 chamber full stack, three slices of 2 chamber images, and post-contrast T1 mapping sequences were obtained. Inversion times were individualized to null the myocardium. MRI measurements were performed by a single reader at a dedicated workstation using Medis (Leiden, Netherlands) and AI assisted Arterys (San Francisco, CA) software under supervision of a senior cardiac imager, both blinded to all clinical variables, and in accordance with Society for Cardiovascular Magnetic Resonance task force recommendations. Artey software was used for T1 and T2 mapping and ECV calculation using pre-and post-contrast MOLLI sequences.

Cardiopulmonary exercise tests were performed by an experienced exercise physiologist and noninvasive cardiology nurse practitioner blinded to participant data according to standard protocol using a metabolic cart (Medical Graphics Corporation Ultima CardiO2) and supine cycle ergometer (Lode Corival CPET) with continuous 12 lead ECG monitoring (GE CASE) and noninvasive blood pressure and pulse oximetry measurement. First, baseline ECG, blood pressure, and rest spirometry including maximum voluntary ventilation (MVV) were measured. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. We determined the rate of work increase per 1 minute step based on the expected peak VO2 from the MVV for a goal 10 minute test, rounded up or down to the nearest 5 Watts/min based on reported exercise (range 10-30W/min). Participants underwent a 2-minute rest phase, 2-minute no resistance warm up, and then 1-minute steps. Breath-by-breath oxygen consumption (VO2) and carbon dioxide production (VCO2) were measured continuously with best 5/7 averaging.

Participants were blinded to time, wattage, and peak VO2 during the test. Participants were encouraged to exercise to their maximum ability, with the test stopped prematurely for severe hypertension, relative hypotension, moderate to severe angina, ventricular tachycardia or couplets, ST elevations or ST depressions >2 mm. Reason for stopping was recorded. Exercise effort quality was assessed by Borg Scale and by the respiratory exchange ratio (RER); tests with RER <1.05 were excluded from analysis. Anaerobic threshold was determined manually by the exercise physiologist using the slope method. CPETs were interpreted separately by two independent readers and differences in interpretation were discussed and resolved through consensus.

For CPET analyses, we evaluated measured peak VO2 (in ml/kg/min) and used the Wasserman equations to estimate the percent predicted VO2 for sedentary adults achieved. [25] [26] [27] We classified participants as having reduced exercise capacity if observed exercise capacity was <85% predicted. We considered ventilatory limitation if rest spirometry was abnormal, tidal volume did not double over baseline reaching at least 50% of FVC, breathing reserve did not reach < 30%, VD/VT was greater than 0.2 with exercise, or desaturation occurred during exercise. We classified participants as having a cardiac limitation if there were ischemic ECG changes, oxygen pulse was reduced and plateaued early, AT occurred at less than 40% of predicted maximal VO2, and if ventilatory efficiency slope (VE/VCO2) was <32. We defined an abnormal chronotropic All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. response to exercise as achieving <80% of the adjusted heart rate reserve (AHRR) calculated as (HRpeak-HRrest)/(220-age-HRrest); as a sensitivity analysis, achieving <85% of the age predicted heart rate (220-age) was also used. 28 We considered exercise limitation to be likely due to deconditioning, obesity or impaired gas exchange if the exercise capacity was reduced and there was no cardiac or ventilatory limitation.

After completion of CMR (or CPET if excluded from CMR), an ambulatory rhythm monitor (Carnation Ambulatory Monitor, BardyDx) was placed on the chest. Participants were instructed to press the button for any cardiopulmonary symptoms and record symptoms in a diary.

Participants were instructed to remove the device after 2 weeks if it had not spontaneously fallen off and mail it back for processing with the symptom diary. Monitors were processed according to BardyDx standard operating procedures and reports were reviewed and overread by a cardiologist.

As we have previously reported, we defined a composite symptom variable for cardiopulmonary PASC including chest pain, dyspnea, or palpitations in the preceding 2 weeks prior to the study visit; 7 all participants had new symptoms, were more than 3 months after SARS-CoV-2 infection, and did not have alternative cardiac disease to explain their symptoms, consistent with the WHO definition of PASC. We also considered self-reported exercise limitation, reduced exercise capacity <85% predicted, and chronotropic incompetence defined as reduced exercise capacity <85% predicted and AHRR<80%. 28 Adjusted models are adjusted for potential confounders including age, sex, time since SARS-CoV-2 infection, hospitalization for acute All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. SARS-CoV-2 infection, and body mass index. We used linear regression to estimate adjusted differences between those with and without symptoms in quantitative parameters and logistic regression to estimate the association with symptoms based on differences in quantitative parameters. For count outcomes (number of button pushes on rhythm monitoring), we used Poisson regression. Non-normally distributed variables (hsCRP, T1 mapping times, premature atrial contraction burden, for example) were log-transformed and findings are reported per doubling or per 10-fold change depending on the range of values. We conducted sensitivity analyses considering other definitions of symptoms and additionally adjusting for potentially relevant past medical history and echocardiographic parameters. We also conducted additional sensitivity analyses to better understand the association between HR and peak VO2 as detailed in the supplemental materials.

The study was approved by the UCSF Committee on Human Research (IRB 20-33000) and all participants provided written informed consent. REDCap was used for data entry. Statistical analyses were performed using STATA version 17.1 and additional visualization was done using R version 4.2.0 using the ggplot2 package.

Among 120 participants, median age was 53 (IQR 42-61), 45 (38%) were female, and 29 participants (24%) had severe acute infection requiring hospitalization. Other demographics and past medical history are shown in Table 1 . Among 46 participants who had at least one advanced test performed, median age was 52 (IQR 42-61), 18 (39%) were female, and 6 (13%) were hospitalized (2 in ICU). The only notable difference between the subset who underwent All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Table 2 ).

Reporting reduced exercise capacity at the second visit was strongly associated with reporting other cardiopulmonary symptoms (OR 8.04, 95%CI 1.94-33.3; p=0.004); 78% of our sample who reported reduced exercise capacity reported other cardiopulmonary symptoms compared to 32% who reported preserved or improved exercise capacity (p=0.003).

Out of 40 participants who attended a CPET visit, 39 completed CPET at a median 17.5 months (IQR 15.5-18.4) after SARS-CoV-2 infection (one excluded due to hypertensive urgency). No participants were taking beta blockers, non-dihydropyridine calcium channel blockers, or ivabradine at the time of CPET. Two tests (one with and one without symptoms) were stopped early for hypertensive response after predicted normal heart rate and exercise capacity were reached but before the participants felt they needed to stop. All other tests were completed to maximal exertion, and all tests had a respiratory exchange ratio (RER) >1.05 with no difference in RER achieved by symptom status.

Overall, 15/39 (38%) participants had reduced exercise capacity (peak VO2 <85% predicted).

Among those with reduced exercise capacity, 9/15 (60%) reported chest pain, dyspnea, or All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. palpitations and 13/15 (87%) reported cardiopulmonary symptoms or fatigue. The odds of reduced exercise capacity were 3.0 times higher among those reporting cardiopulmonary symptoms (95%CI 0.8-11.4; p=0.11) and 9.1 times higher among those reporting cardiopulmonary symptoms or fatigue (95%CI 1.7 to 49.6; p=0.01). As shown in Figure 1 , exercise capacity was reduced among those with cardiopulmonary symptoms, although it was not statistically significant. After adjustment, peak VO2 was 2.7 ml/kg/min lower among those reporting chest pain, dyspnea, or palpitations (95%CI -6.9 to 1.5; p=0.20) or -11% predicted (95%CI -27 to 5, p=0.17). The adjusted difference was -5.9 ml/kg/min among those with chest pain, dyspnea, palpitations, or fatigue (-9.6 to -2.3; p=0.002) or -21% (-35 to -7; p=0.006). Other ways of considering symptoms are shown in Supplemental Table 2 .

Most CPET parameters were not significantly different among those with and without symptoms (Table 2) , except anaerobic threshold. Anaerobic threshold occurred early (at a lower % predicted maximum VO2) among those with symptoms, and the odds of symptoms were 2.1-fold higher per 5% decrease in anaerobic threshold relative to predicted peak VO2 (95%CI 1.08 to 4.20; p=0.03). There were no significant differences on rest echocardiography (Supplemental Table 3 ) or spirometry (Supplemental Table 4 ) or associations between these parameters and peak VO2 (results not shown).

In classifying the etiology of reduced exercise capacity, no participants had a pattern consistent with a ventilatory limitation (and none with desaturation), one participant had a cardiac limitation (reduced oxygen pulse pressure and ischemic changes on ECG), one participant's peak VO2 was slightly below 85% predicted with no other abnormalities, and three participants had findings most consistent with deconditioning/obesity or impaired gas exchange. Among All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. participants with reduced exercise capacity (<85% predicted), 9/15 (56%) had chronotropic incompetence with an adjusted heart rate reserve (AHRR) <80% and peak heart rate <85% age predicted maximum. One participant with obesity and deconditioning had AHRR<80% but APMHR=85%, and we did not classify that participant as having chronotropic incompetence.

Nineteen participants had an AHRR<80%, which was associated with lower peak VO2 ( Figure   2 ). We classified participants into three categories of heart rate response to exercise by whether they reached 85% predicted maximum VO2 and 80% AHRR ( Figure 3 and Supplemental Figure   1 ). Not reaching AHRR >80% during CPET was associated with 15.6x higher odds of having reduced exercise capacity <85% predicted (95%CI 1.30-187; p=0.03) and 4.9 ml/kg/min lower peak VO2 (95CI CI -9.3 to -0.5; p=0.03). Participants with chronotropic incompetence (peak VO2 <85% predicted, AHRR <80%, and considered the primary explanation for reduced exercise capacity, n=9) had lower peak heart rate and lower AHRR achieved ( Table 5, p<0.0001 for all). Chronotropic incompetence was associated with 6.4 ml/kg/min lower peak VO2 (95%CI -11.3 to -1.6; p=0.01), a large and clinically significant difference. Heart rate recovery at 1 minute after cessation of exercise was also reduced among those with chronotropic incompetence. Respiratory exchange ratio, a measure of reaching peak exercise capacity, was not significantly lower among those with chronotropic incompetence (difference 0.04, 95%CI -0.02 to 12; p=0.19). We conducted additional sensitivity analyses to exclude the possibility that observed differences in heart rate were not explained by differences in achieved VO2

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. One participant was not administered gadolinium due to eGFR<30, and 3 participants could not complete the protocol due to claustrophobia. CMR demonstrated normal LV and RV volumes and ejection fraction in all participants (n=39, Table 4 ). There were statistically significant associations between smaller RV end diastolic volume index and smaller stroke volume and symptoms. No participants had late gadolinium enhancement suggestive of myocardial scar.

Other markers suggestive of cardiac inflammation including native T1 and T2 parametric mapping values and calculated extracellular volume were not significantly different by symptom status or associated with symptoms. A high proportion of participants (11/39, 28%) had trace or small pericardial effusions with no difference by symptom status.

Those with chronotropic incompetence had 19ml lower stroke volume at rest (95%CI 3-34;

p=0.02). Those with chronotropic incompetence also had higher extracellular volume (+6.2%, 95%CI 0.7 to 11.7; p=0.03).

There were no clinically significant arrhythmias including atrial fibrillation or atrial flutter noted on ambulatory rhythm monitoring among those with or without symptoms (Table 5 ). There were no statistically significant differences in premature atrial contractions (PAC), premature ventricular contractions (PVC), or supraventricular tachycardias (SVT) by symptom status. After adjustment, cardiopulmonary symptoms were not associated with an increased burden of PAC, PVC, or supraventricular tachycardia episodes. We could not exclude meaningful increases in PVC burden given wide confidence intervals since most participants had no PVCs. Average HR and HR variability as measured by standard deviation n to n (SDNN) were lower, but not statistically significantly different between those with and without symptoms. Results were similar after adjusting for echocardiographic and CMR parameters including LVEF, LV strain, All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. and LA volume index. Button pushes correlated with brief episodes of SVT in one participant with PASC, but most button pushes were associated with sinus rhythm, sinus tachycardia, or supraventricular ectopy (Supplemental Figure 5 ).

Peak heart rate on CPET was significantly positively correlated with maximum sinus heart rate on ambulatory rhythm monitoring (Pearson's r=0.73; p<0.001). Those with chronotropic incompetence had a higher average heart rate on ambulatory rhythm monitoring, higher minimum heart rate, lower maximum heart rate, and lower heart rate variablility (Table 5) . No individuals had 2 nd or 3 rd degree heart block or evidence of sinus node dysfunction on ambulatory rhythm monitoring. PR intervals were not significantly longer among those with chronotropic incompetence (174ms vs 162ms, +12ms, 95%CI -4 to +27; p=0.13).

After adjustment, the odds of symptoms at the second visit were 1.36 times higher per doubling of hsCRP measured approximately 1 year prior (0.70-2.64; p=0.36), which was not statistically significant. Per each doubling of hsCRP, peak VO2 was 1.8 ml/kg/min lower (95%CI 0.2 to 3.3; p=0.03) or -6.3% predicted (95%CI 0.6 to 12; p=0.03). hsCRP was not statistically significantly associated with chronotropic incompetence (OR 1.15, 95%CI 0.43-3.1; p=0.78) or AHRR (-2.5% per doubling, 95%CI -9 to 4; p=0.44). Lastly, the odds of having a pericardial effusion on CMR were 1.8 times per doubling of hsCRP (95%CI 1.09-2.92; p=0.02).

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. We demonstrate that reductions in objective metabolic capacity during exercise as assessed by CPET are present among individuals with cardiopulmonary symptoms at 18 months following acute SARS-CoV-2 infection. Peak oxygen consumption was lower among those with cardiopulmonary symptoms, especially when defined to include fatigue. Reduced exercise capacity among those recovering from COVID-19 was strongly associated with a blunted heart rate response to exercise, and our findings suggest that chronotropic incompetence is a likely mechanism of reduced exercise capacity for some individuals with PASC. Additionally, we found that chronotropic incompetence on CPET was associated with ambulatory rhythm findings that suggest autonomic dysfunction including changes in heart rate and heart rate variability.

Reduced VO2 was also associated with elevated hsCRP. Other than pericardial effusions (also associated with elevated hsCRP), we did not find evidence of persistent cardiac inflammation or dysfunction on echocardiography or cardiac MRI, and also no evidence of arrhythmias on ambulatory rhythm monitoring. Furthermore, our study exemplifies the value in assessing disease mechanisms through a comprehensive evaluation of cardiac imaging (echo, CMR), cardiopulmonary exercise testing, and rhythm evaluation in the setting of individuals with and without symptoms 18 months after acute COVID-19 infection.

To date, six studies have compared symptomatic and recovered individuals at 3-6 months after COVID. All reported lower peak VO2 among those with symptoms despite differences in defining symptoms. 18-21, 29, 30 Most assessed mostly previously hospitalized individuals at 3-6 months after infection. These studies have not reached consistent conclusions regarding the etiology of reduced exercise capacity, which may be related to different study populations, exercise protocols and proportions with sub-maximal tests. Two studies attributed lower peak All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. VO2 to inadequate augmentation of stroke volume 30 or chronotropic incompetence and poor stroke volume augmentation. 21 Others have attributed exercise limitations to ventilatory inefficiency from pulmonary vasculopathy, 18 respiratory abnormalities, 19 obesity and differences in cardiorespiratory fitness 29 and deconditioning. 20, 21 Impaired oxygen extraction at the peripheral level (assessed using invasive CPET) has also been suggested to explain reduced exercise capacity among those with PASC at 11 months after infection by Singh et al. 31 However, they describe stopping CPETs once RER was greater than 1.10 or heart rate was >85% predicted, which is below maximal exercise for most individuals and does not allow evaluation of chronotropic incompetence. They hypothesize that autonomic regulation of microcirculatory function may be one mechanism by which SARS-CoV-2 may alter oxygen extraction during exercise. 32 Others have hypothesized that IL-6 (which may be elevated in PASC) 33 functions as a myokine that regulates energy allocation during exercise. 34 We did not find differences in the VO2 work slope, a noninvasive correlate of peripheral oxygen extraction, but we did not measure peripheral oxygen extraction since we did not perform invasive CPET. Mancini et al and others have found that dysfunctional (rapid, erratic) breathing or exercise hyperventilation may contribute to symptoms of persistent dyspnea in PASC. [35] [36] [37] In contrast, we did not observe dysfunctional breathing in any participant.

Although not a universal finding, several other groups have also reported chronotropic incompetence in PASC. Szekely et al found that 75% of PASC participants had chronotropic incompetence compared to 8% of control participants; they reported that chronotropic incompetence was the most likely explanation of reduced exercise capacity in PASC. 21 They also found reduced stroke volume augmentation consistent with Brown et al 30 and higher peripheral oxygen extraction consistent with Singh et al. 21 , 31 Abdallah et al found that chronotropic All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. incompetence, present in 38% of their sample, contributed to exercise limitation, with greater limitations in exercise and heart rate among hospitalized individuals. 38 Jimeno-Almazán et al found chronotropic incompetence on CPET using a cycle ergometer and confirmed by treadmill testing among 4/32 individuals (13% of study sample) with reduced exercise capacity after COVID-19. 39 Ladlow et al found that dysautonomia as defined by heart rate parameters (resting HR>75, HR increase <89 bpm, and HR recovery at 1 minute <25 bpm) 40 was associated with lower peak VO2 among individuals with PASC. 41 Our findings build upon these earlier studies by demonstrating chronotropic incompetence at 18 months following COVID-19 and also including evaluation of cardiac inflammation, structural heart disease and cardiovascular arrhythmias in study participants along with including recovered individuals as a comparator group and adjusting for confounders in our models.

Higher resting heart rate and blunted heart rate response during exercise testing are associated with development of cardiovascular disease, sudden death, and all-cause mortality among men without known coronary artery disease. 40, 42, 43 Chronotropic incompetence is also associated with impaired endothelial function assessed by brachial flow-mediated vasodilation 44 and abnormal endothelial function has been demonstrated to be present in the setting of early PASC. 45, 46 Abnormal endothelial function assessed with non-invasive brachial artery flowmediated dilation may be associated with reduced VO2 in PASC. 47 Coronary microvascular dysfunction, present in some people recovering from COVID-19, 48 is also associated with chronotropic incompetence in the general population with microvascular dysfunction. 49, 50 Autonomic Function in PASC All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 41 Alterations in autonomic function could explain chronotropic incompetence, blunted heart rate recovery, and reduction in heart rate variability. Perturbed autonomic function may also explain altered peripheral oxygen extraction, preload failure, and disordered breathing. SARS-CoV-2 may lead to damage or dysfunction of the peripheral autonomic nerve fibers as sequalae of direct viral infection, secondary inflammation of the nerves, or an autoimmune neuropathy. Two groups have found evidence of a small fiber neuropathy among those with COVID-19 associated with abnormal autonomic testing. 51, 52 Autonomic dysfunction could also occur via direct infection of autonomic regulatory regions of the brainstem tone, which are anatomically close to the cribiform plate and the olfactory nerves.

A meta-analysis of autopsy studies of individuals who died from COVID-19 demonstrated that brainstem abnormalities were commonly reported, with detection of viral RNA and proteins. 53 Our group has previously demonstrated abnormal markers of neurologic injury are present in serum 54 and CSF. 55 Chronic inflammation in other conditions is associated with an imbalance between parasympathetic and sympathetic activation and changes in heart rate variability. 56, 57 Young adults recovering from SARS-CoV-2 have elevated sympathetic activation at rest compared to healthy controls. 58 Consistent with our findings that elevated hsCRP is associated with reduced exercise capacity, one other group found that hsCRP, IL-6, and TNF-alpha are associated with reduced peak VO2 3 months after hospitalization for COVID-19. 59 Elevated inflammatory markers are associated with reduced exercise capacity and chronotropic incompetence among adolescents and adults with obesity. 44, 60 In heart failure, resting heart rate is positively associated All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. with levels of hsCRP, but chronotropic incompetence during exercise is negatively correlated with copeptin (C-terminal portion of arginine) and norepinephrine levels. 61 Inflammation in the setting of stress (as evidenced in the association between amygdalar activity and bonemarrow/arterial inflammation) 62 has also been implicated in the pathophysiology of health disparities 63, 64 and post-traumatic stress disorder. 65 Vascular inflammation has been reported in early PASC. 66 Thus, inflammatory responses to stress from direct viral infection, chronic immune activation, or from the distressing nature of persistent symptoms could potentially alter the ability to augment heart rate during exercise in PASC even without direct damage to the autonomic nervous system or the heart.

Sinus node remodeling has been hypothesized to reduce sinus node reserve among patients with heart failure. 67 Apart from autonomic function, SARS-CoV-2 may directly affect sinus node function. Han et al elegantly demonstrated that SARS-CoV-2 can infect the sinoatrial node in hamsters and human embryonic stem cell-derived sinoatrial-like pacemaker cells resulting in sinoatrial node-dysfunction including changes in calcium handling, activated inflammatory pathways, and induced ferroptosis. 68 Although we did not find evidence of localized or diffuse fibrosis on cardiac MRI or evidence of sinus node dysfunction on ambulatory rhythm monitoring or ECG monitoring during CPET (ie sinus pauses, Wenckebach, blocked PACs), sinus node dysfunction could explain chronotropic incompetence.

Multiple case series have reported that deconditioning is the primary cause of exercise limitation in PASC. 20, [69] [70] [71] [72] Our findings argue against deconditioning as the primary explanation as the All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 19, 2022. ; https://doi.org/10.1101/2022.05.17.22275235 doi: medRxiv preprint majority of individuals had normal oxygen pulse and normal VO2 to work slope which suggest normal augmentation of cardiac output. We found earlier anaerobic thresholds which can be seen in deconditioning and identified three individuals (8% of sample, 20% with reduced exercise capacity) whose pattern of findings is most consistent with deconditioning, obesity, or alterations in peripheral oxygen utilization. Our findings of smaller LV and RV volumes could be consistent with deconditioning, although we did not find a lower LV mass index among those with symptoms or reduced exercise capacity (possibly confounded by pre-existing hypertension). One of the challenges with identifying deconditioning as a cause is that it is also a natural effect of reduced exercise, such that those with symptoms limiting exercise capacity will be expected to have deconditioning.

Consistent with other studies that used CMR to look for structural damage, our study reassuringly did not find evidence of abnormal cardiac function or late gadolinium enhancement suggestive of scar. Our findings extend those of Joy et al and Cassar et al from 6 months after infection by demonstrating that over approximately 18 months after infection there remains no evidence for significant scar. 14, 17 Neither study found differences in late gadolinium enhancement or parametric mapping compared to controls at 6 months suggestive of myocardial scar or ongoing inflammation.

Our study also extends three prior studies examining arrhythmias after COVID-19 at 3 months using ambulatory monitoring. Ingul et al found that having more than 200 PVCs in 24 hours was common 3 months after hospitalization, but they did not adjust for cardiac comorbidities. 22 Dewland et al demonstrated no significant arrhythmias nor increased burden of ectopy on 2 week ambulatory rhythm monitoring among outpatients at 3 months. 23 Another study estimated that All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 73 Nonetheless, our findings regarding heart rate variability among those with chronotropic incompetence are consistent with theirs regarding heart rate variability. Our findings suggest that arrhythmias or inappropriate sinus tachycardia are unlikely to explain symptoms among those with PASC more than 1 year after infection.

Investigation into mechanisms of PASC may benefit from proof-of-concept approaches to identify potential targets for intervention. Given associations with elevated inflammatory markers, targeting inflammatory or anti-viral pathways are intriguing possibilities worthy of investigation, although whether such interventions can restore autonomic nervous system function remains unknown. In those without cardiac implantable devices, exercise is the only intervention demonstrated to improve chronotropic response to exercise and improve peak VO2 among those with chronotropic incompetence. 74 Exercise improves cardiac markers of autonomic function including heart rate recovery and heart rate variability among people with heart failure. 75, 76 Exercise has also been demonstrated to improve outcomes in POTS, which may also be related to autonomic responses to stress. 77, 78 Limitations The main limitations from this observational study arise from the convenience sampling, which may be prone to selection bias, and the cross-sectional nature of these advanced cardiovascular measures at a single time point. The first challenge is the classification of PASC-the statistical All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. significance and magnitude of the difference in peak VO2 was sensitive to how we defined cardiopulmonary PASC. We do not have pre-COVID exercise tests and pre-COVID fitness is likely associated with post-COVID peak VO2; if fitness is associated with risk of PASC, it would be an unmeasured confounder. Including individuals with a wide range of self-reported baselinefitness widened our confidence intervals and reduced the precision of our estimates. Although we have adjusted for important measured confounders (ie age, sex, BMI) and conducted sensitivity analyses adjusting for other potential confounders (ie hypertension, asthma/COPD), there are likely residual confounders (including unmeasured ones, like pre-COVID fitness) that may bias our results. Although participants were not selected based on symptoms, volunteer bias may overestimate the prevalence of reduced exercise capacity and possibly the magnitude of the difference. We would not expect selection bias to affect the classification of those with limitations or the association with chronotropic incompetence among those with reduced peak VO2. We did not include an uninfected comparator group, which could have strengthened our ability to make inferences about regardless of symptom status and definition. As others using cardiopulmonary exercise testing have previously noted, selecting an appropriate control group can be challenging. In terms of measurement, we did not perform invasive CPET, which would have allowed for assessment of peripheral oxygen extraction and preload response to exercise, nor did we perform stress echocardiography, stress CMR, or stress ventriculography; to do so would have required altering our exercise protocol which may have reduced our ability to detect chronotropic incompetence.

In conclusion, we found that reduced exercise capacity is common among individuals with prior SARS-CoV-2 infection and persistent cardiopulmonary symptoms at 18 months following All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. infection. Reduced exercise capacity in PASC is associated with chronotropic incompetence, ambulatory rhythm monitoring findings suggestive of autonomic dysfunction, and higher levels of hsCRP. While a significant proportion had trace or small pericardial effusions (also associated with hsCRP), we did not find evidence of prior or ongoing myocarditis. Further investigation into mechanisms of cardiopulmonary PASC should include study of the autonomic nervous system to identify potential therapeutic targets.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Benes J, Kotrc M, Borlaug BA, Lefflerova K, Jarolim P, Bendlova B, Jabor A, Kautzner J and Melenovsky V. Resting heart rate and heart rate reserve in advanced heart failure have distinct pathophysiologic correlates and prognostic impact: a prospective pilot study. JACC Heart Fail. 2013;1:259-66.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Figure 1 Legend: On the left are box and whisker plots of unadjusted peak oxygen consumption (VO2 in ml/kg/min on the left and percent of predicted on the right) among those without (blue) and with chest pain, dyspnea, or palpitations (pink). Mean peak VO2 was 22.1 ml/kg/min among those with cardiopulmonary symptoms compared to 26.0 ml/kg/min among those without symptoms, a non-statistically significant difference of -3.9 ml/kg/min (95%CI -1.7 to 9.6; p=0.17) or 92% vs 103% percent predicted (difference -10.5, 95%CI -5.0 to 26.1; p=0.18). After adjustment for age, sex, hospitalization for acute COVID, BMI category, and months since SARS-CoV-2 infection, peak VO2 was 2.7 ml/kg/min lower among those reporting cardiopulmonary symptoms (95%CI -6.9 to 1.5; p=0.20) which is equivalent to 11% lower than predicted (95%CI -27 to 5, p=0.17), neither of which were statistically significant differences. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Figure 2 Legend: Boxplots of peak VO2 (percent predicted since both are confounded by age) by reaching an AHRR >80% (normal chronotropic response to exercise) or <80% (blunted chronotropic response to exercise). P value is for unadjusted t-test. Not reaching adjusted heart rate reserve >80% during CPET was associated with 15.6x higher odds of having reduced exercise capacity <85% predicted (95%CI 1.30-187; p=0.03).

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Figure 3 Legend: These lines represent the average heart rate at a given percentage of exercise completed classified by normal exercise capacity and chronotropic response during exercise on the top in purple (peak VO2≥85% predicted and AHRR ≥80%; R 2 0.89), normal exercise capacity with reduced chronotropic response in teal (peak VO2≥85% predicted and AHRR <80%; R 2 0.90), and reduced exercise capacity with chronotropic incompetence in yellow (peak VO2<85% predicted and AHRR <80%; R 2 0.75). Gray bars represent 95% confidence intervals for each fitted line. Results are similar when plotting %APMHR or AHRR instead of absolute heart rate (Supplemental figure 1 ).

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Legend. Boxplots of peak VO2 and adjusted heart rate reserve (AHRR) by hsCRP less than or greater than median (1.2 mg/L). Peak VO2 decreases by 1.8 ml/kg/min per doubling of hsCRP (95%CI 0.2 to 3.3; p=0.03) or 6.3% reduction per doubling (95%CI 0.6 to 12; p=0.03); when treated as a dichotomous variable, the odds of reduced peak VO2 <85% predicted were 1.96 per doubling of hsCRP (0.84-4.56; p=0.12). Adjusted heart rate reserve is 2.5% lower per doubling of hsCRP (95%CI -9 to 4; p=0.44), which was not statistically significant.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Table 1 Legend: There were no significant differences between the entire cohort, shown here, and the subset who underwent advanced cardiopulmonary testing (Supplemental Table 1 ), with the exception of earlier date of infection. Hospitalization, elevated hsCRP, and possibly female sex were associated with symptoms. Abbreviations: BMI=body mass index hsCRP=high sensitivity c-reactive protein, LVEF=Left Ventricular Ejection Fraction, OR=odds ratio All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Table 2 Legend: We present both the odds ratios for the association between CPET parameters and cardiopulmonary symptoms estimated using logistic regression with adjustment for age, sex, time since COVID, hospitalization for acute COVID, BMI category; we also estimated mean differences between those with and without symptoms using linear regression and adjusting for the same covariates. Sensitivity analysis incorporating history of hypertension, diabetes, and lung disease had no substantive changes in effect sizes or confidence intervals.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Table 4 Legend. Values are reported as mean±SD or median (interquartile range) for non-normally distributed variables assessed by histogram. The only statistically significant difference between those with and without symptoms was the number of button pushes, which was 2.4 time more among those with symptoms (95CI% 1.7-3.4; p<0.001), and lower maximum sinus heart rate consistent with chronotropic incompetence among those with symptoms. One participant had SVT that correlated with palpitations by patient diary, but no episodes were sustained longer than 30 second. Results were similar in sensitivity analysis when only palpitations were considered (n=10).

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Table 5 Legend: The first row of each measure is the mean±SD for those with chronotropic incompetence ("CI", VO2 <85%, AHRR<80%, and no alternative findings, n=9); the second row is the mean±SD for those with a reduced chronotropic response (VO2 ≥85% and AHRR<80%, n=8) and the third row is those with peak VO2 ≥85% and AHRR≥80% (n=16). Unadjusted and adjusted differences are compared to those with normal exercise capacity and heart rate response during exercise. Abbreviations: HR=heart rate, bpm=beats per minute, CI=chronotropic incompetence, AHRR=adjusted heart rate reserve, SDNN=standard deviation n-to-n All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Among all participants, peak VO2 is associated with peak heart rate during exercise with no difference in slope by the presence of cardiopulmonary symptoms (Supplemental Figure 2) . To exclude the possibility that the peak heart rate is lower because participants did not reach their predicted maximum VO2 (as opposed to the measured peak VO2), we conducted a sensitivity analysis in which we plotted the heart rate at rest, anaerobic threshold, and peak exercise by the percentage of peak VO2 achieved and generated models to estimate the differences in slope of heart rate to percent VO2, using the interaction terms to estimate the difference in heart rate response (Supplemental Figure 3 ). The adjusted difference in peak heart rate was -14.4% among those with chronotropic incompetence accounting for differences in the percentage of predicted exercise capacity (95CI% -21 to -7.7; p= <0.001) and there was a highly statistically significant interaction between chronotropic incompetence and percent predicted peak VO2 on peak heart rate (pinteraction<0.001). We also plotted heart rate based on the % of predicted exercise capacity they achieved and the AHRR based on the percent of achieved VO2 (Supplemental Figure 4) . Together, these findings suggest that the reduced heart rate response to exercise is not simply a manifestation that those participants did less exercise and therefore did not need to augment their heart rate as much.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Table 5 Legend. N listed for the number with that symptom finding.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted May 19, 2022. ; https://doi.org/10.1101/2022.05.17.22275235 doi: medRxiv preprint All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Figure 2 ), adjusted HRR (Middle), and APMHR (Bottom) with clear differences in the heart rate response during exercise regardless of which measure of heart rate is used.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Figure 2 : Relationship between Peak HR and VO2 by Symptoms Supplemental Figure 2 Legend: Scatter plot and linear fits of VO2 by peak HR with blue dots and green line representing those without symptoms and red dots and orange line representing those with symptoms. There is a linear increase in peak VO2 as heart rate increases, with a peak VO2 that was 2.3ml/kg/min lower among those with symptoms accounting for heart rate (95%CI -5.6 to 1.1; p=0.18); the interaction term between symptoms and heart rate on exercise was not significant (coefficient 0.0; 95%CI -0.15 to 0.12; p=0.80). Sensitivity analysis incorporating stroke volume as measured on cardiac MRI did not significantly change the model and the coefficient for stroke volume was 0 (95%CI -0.05 to 0.20; p=0.29).

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Figure 3 : HR by VO2 and Chronotropic Incompetence Supplemental Figure 3 Legend: Scatter plots and linear fits for (A) heart rate (beats per minute), (B) heart rate (% predicted) and (C) adjusted heart rate reserve achieved as a function of VO2 as a percent of peak VO2 achieved. There was an interaction between chronotropic incompetence and heart rate slope on peak heart rate, %APMHR, and AHRR. Chronotropic incompetence (CI) is defined as those with reduced exercise capacity, reduced HR response and no alterative explanation for reduced exercise capacity. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. Figure 4 Legend. On the left, heart rate is plotted on the y axis as a function of the percent of predicted VO2 to demonstrate that the peak HR achieved and slope of the peak HR are similarly reduced among those with AHRR<80% regardless of whether they reached a peak VO2 <85% (yellow) or >85% (teal) predicted compared to those who had normal chronotropic response to exercise (purple). This suggests that the lower achieved peak heart rate is not due to not reaching >85% of predicted exercise capacity. Secondly, on the right, AHRR is plotted as a function of the percent of achieved VO2 (100% for all at the end of their test). Here it is clear that AHRR is the lowest among those with chronotropic incompetence (yellow), intermediate among those with a blunted heart rate response but peak VO2 >85% (teal) and highest among those with normal exercise capacity and heart rate response (purple). All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 Sinus with Ventricular Ectopy All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 

Short-term and Long-term Rates of Postacute Sequelae of SARS-CoV-2 Infection: A Systematic Review

Population-Based Estimates of Post-acute Sequelae of Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) Infection (PASC) Prevalence and Characteristics

Incidence, cooccurrence, and evolution of long-COVID features: A 6-month retrospective cohort study of 273,618 survivors of COVID-19

Technical article: Updated estimates of the prevalence of post-acute symptoms among people with coronavirus (COVID-19) in the UK

Seroprevalence of Infection-Induced SARS-CoV-2 Antibodies -United States

Postacute COVID-19: An Overview and Approach to Classification

Role of antibodies, inflammatory markers, and echocardiographic findings in post-acute cardiopulmonary symptoms after SARS-CoV-2 infection

Cardiac performance in patients hospitalized with COVID-19: a 6 month follow-up study

Reduced Cardiac Function by Echocardiography in a Minority of COVID-19 Patients 3&#xa0;Months after Hospitalization

Echocardiographic Comparison of COVID-19 Patients with or without Prior Biochemical Evidence of Cardiac Injury after Recovery

Cardiac Involvement in Recovered Patients From COVID-19: A Preliminary 6-Month Follow-Up Study

Short-term cardiac outcome in survivors of COVID-19: a systematic study after hospital discharge

Outcomes of Cardiovascular Magnetic Resonance Imaging in Patients Recently Recovered From Coronavirus Disease

Prospective Case-Control Study of Cardiovascular Abnormalities 6 Months Following Mild COVID-19 in Healthcare Workers

Patterns of myocardial injury in recovered troponin-positive COVID-19 patients assessed by cardiovascular magnetic resonance

Symptom Persistence Despite Improvement in Cardiopulmonary Health -Insights from longitudinal CMR, CPET and lung function testing post-COVID-19

Exercise Ventilatory Inefficiency in Post-COVID-19 Syndrome: Insights from a Prospective Evaluation

Exercise Testing in Patients with Post-COVID-19 Syndrome

Cardiopulmonary exercise capacity and limitations 3 months after COVID-19 hospitalisation

Cardiorespiratory Abnormalities in Patients Recovering from Coronavirus Disease

Cardiac Dysfunction and Arrhythmias 3 Months After Hospitalization for COVID-19

Prospective arrhythmia surveillance after a COVID-19 diagnosis

Magnitude, and Patterns of Postacute Symptoms and Quality of Life Following Onset of SARS-CoV-2 Infection: Cohort Description and Approaches for Measurement. Open Forum Infectious Diseases

Predicted values for clinical exercise testing

Principles of Exercise Testing and Interpretation

Determining the preferred percentpredicted equation for peak oxygen consumption in patients with heart failure

Cardiorespiratory physiology, exertional symptoms, and psychological burden in post-COVID-19 fatigue

Ongoing Exercise Intolerance Following COVID-19: A Magnetic Resonance-Augmented Cardiopulmonary Exercise Test Study

Persistent Exertional Intolerance After COVID-19: Insights From Invasive Cardiopulmonary Exercise Testing

Insights From Invasive Cardiopulmonary Exercise Testing of Patients With Myalgic Encephalomyelitis/Chronic Fatigue Syndrome

Markers of Immune Activation and Inflammation in Individuals With Postacute Sequelae of Severe Acute Respiratory Syndrome Coronavirus 2 Infection

Interleukin 6 as an energy allocator in muscle tissue

Use of Cardiopulmonary Stress Testing for Patients With Unexplained Dyspnea Post-Coronavirus Disease

Dysfunctional breathing diagnosed by cardiopulmonary exercise testing in 'long COVID' patients with persistent dyspnoea

Hyperventilation as one of the mechanisms of persistent dyspnoea in SARS-CoV-2 survivors

Pulmonary Function, and Functional Capacity Four Months after COVID-19

Chronotropic Incompetence in Non-Hospitalized Patients with Post

Heart-rate profile during exercise as a predictor of sudden death

Dysautonomia following COVID-19 is not associated with subjective limitations or symptoms but is associated with objective functional limitations

Heart rate response during exercise test and cardiovascular mortality in middleaged men

Impaired Chronotropic Response to Exercise Stress Testing as a Predictor of Mortality

Comparison of endothelial vasodilator function, inflammatory markers, and N-terminal pro-brain natriuretic peptide in patients with or without chronotropic incompetence to exercise test

Altered Vascular Endothelium-Dependent Responsiveness in Frail Elderly Patients Recovering from COVID-19 Pneumonia: Preliminary Evidence

In-vivo evidence of systemic endothelial vascular dysfunction in COVID-19

Cardiopulmonary Exercise Performance and Endothelial Function in Convalescent COVID-19 Patients

A cardiovascular magnetic resonance imaging-based pilot study to assess coronary microvascular disease in COVID-19 patients

Relation between resting amygdalar activity and cardiovascular events: a longitudinal and cohort study

Stress-Associated Neurobiological Pathway Linking Socioeconomic Disparities to Cardiovascular Disease

Heart Rate Variability and Inflammatory Stress Response in Young African American Men: Implications for Cardiovascular Risk

Symptom severity impacts sympathetic dysregulation and inflammation in post-traumatic stress disorder (PTSD)

Vasculitis changes in COVID-19 survivors with persistent symptoms: an [18F]FDG-PET/CT study

Remodeling of sinus node function in patients with congestive heart failure: reduction in sinus node reserve

SARS-CoV-2 Infection Induces Ferroptosis of Sinoatrial Node Pacemaker Cells

Cardiopulmonary exercise testing in COVID-19 patients at 3 months follow-up

Deconditioning as main mechanism of impaired exercise response in COVID-19 survivors

Cardiopulmonary exercise test in patients with persistent dyspnea after COVID-19 disease

Exercise capacity impairment after COVID-19 pneumonia is mainly caused by deconditioning. The European respiratory journal

Inappropriate sinus tachycardia in post-COVID-19 syndrome

Effects of exercise training on chronotropic incompetence in patients with heart failure

Reversal of autonomic derangements by physical training in chronic heart failure assessed by heart rate variability

Effects of exercise training on heart rate recovery in patients with chronic heart failure

Exercise and non-pharmacological treatment of POTS

The international POTS registry: Evaluating the efficacy of an exercise training intervention in a community setting

Supplemental Table 1 Demographics and Medical History of Participants who Underwent Advanced Cardiopulmonary Testing (CMR and/or CPET)

We would like to thank the research participants and the members of the