key: cord-0963835-g4ve91g5
authors: Larsen, N. W.; Stiles, L. E.; Shaik, R.; Schneider, L.; Muppidi, S.; Tsui, C. T.; Miglis, M. G.
title: Characterization of Autonomic Symptom Burden in Long COVID: A Global Survey of 2,314 Adults
date: 2022-04-28
journal: nan
DOI: 10.1101/2022.04.25.22274300
sha: fd3108e8494a9ba485f38a987e08cec36b937706
doc_id: 963835
cord_uid: g4ve91g5

Background: Autonomic dysfunction is a common complication of post-acute sequalae of SARS-CoV-2 (PASC)/long COVID, however prevalence and severity rates are unknown. Objective: The primary goal of this study was to assess the frequency and severity of autonomic symptoms in PASC. We also aimed to assess symptom burden in PASC though well-validated questionnaires, evaluate which pre-existing conditions are associated with an increased risk of developing autonomic dysfunction, and determine whether the severity of acute COVID-19 illness is associated with the severity of autonomic dysfunction in this population. Methods: We conducted an online survey of 2,314 adults with PASC using several validated questionnaires including the COMPASS-31 to evaluate for autonomic dysfunction. We included both participants who had tested positive for COVID-19 (test-confirmed) and participants who were diagnosed with COVID-19 based on clinical symptoms alone (test-unconfirmed). Additional analyses were performed on test-confirmed participants, comparing hospitalized to non-hospitalized participants. Results: 67% of PASC patients had a COMPASS-31 score >20, suggestive of moderate to severe autonomic dysfunction. COMPASS-31 scores did not differ between test-confirmed hospitalized and non-hospitalized participants (28.9+/-30.98 vs 26.4+/-28.35, p=0.06). Both hospitalized and non-hospitalized participants reported significant functional disability across all quality-of-life domains. Conclusions: Moderate to severe autonomic dysfunction was seen in all PASC groups in our study, independent of hospitalization status, suggesting that autonomic dysfunction is highly prevalent in the PASC population and not necessarily dependent on the severity of acute COVID illness.

While the rapid growth of coronavirus disease cases has generally subsided globally, many individuals, regardless of symptoms and disease severity, have been left with long-lasting symptoms, a condition referred to as post-acute sequalae of SARS-CoV-2 (PASC), or long COVID. Although PASC is not universally defined, it is most often identified as symptoms of COVID-19 lasting > 30 days [1] . PASC may include a diverse constellation of symptoms, many of which appear to be autonomic in nature [2] .

The most common autonomic diagnosis associated with PASC is that of postural orthostatic tachycardia syndrome (POTS), an autonomic nervous system disorder strongly associated with a preceeding viral infection [3] . In fact, many of the symptoms of PASC have been documented in individuals with POTS prior to the COVID-19 pandemic, suggesting a shared pathophysiological mechanism and prompting our interest in PASC as a framework for better understanding post-viral dysautonomia. Less commonly reported autonomic conditions associated with PASC include small fiber neuropathy [4] [5] , orthostatic intolerance [6] , orthostatic hypotension [7] , and neurallymediated syncope [7] [6] . The prevalence of POTS and other autonomic disorders in those with PASC is unknown.

While there have been online surveys of patients with PASC [4] , none have specifically addressed autonomic symptom burden in relation to other features of the illness. In this context, we designed and conducted a global online survey of individuals with PASC with the goal of assessing the frequency and severity of autonomic symptoms in this population. We also aimed to: 1) assess symptom burden in PASC though well-validated questionnaires; 2) evaluate which pre-existing conditions are associated with an increased risk of developing autonomic dysfunction in PASC; and 3) determine whether the severity of acute COVID-19 illness is associated with more severe autonomic dysfunction in PASC.

Adults ≥ 18 years of age who had self-suspected, clinician-diagnosed or test-confirmed SARS-CoV-2 infection were recruited through long COVID support groups and social media channels between October 2020 and August 2021. Self-suspected and cliniciandiagnosed COVID-19 was categorized as "test-unconfirmed" and antibody, antigen and PCR-confirmed COVID-19 was categorized as "test-confirmed" for analytic purposes. All participants completed an online, English-language survey consisting of demographic information, medical history, severity of acute COVID-19 illness, and a series of validated-questionnaires. The Composite Autonomic Symptom Score-31 (COMPASS- 31) was used to assess autonomic dysfunction. We also administered the Orthostatic Hypotension Questionnaire (OHQ), Fatigue Severity Scale (FSS), Epworth Sleepiness Scale (ESS), Neuropathic Pain Scale (NPS), General Anxiety Disorders Assessment (GAD- 7) , and the RAND 36-Item Health Survey (RAND-36). Data were collected via the online . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

Research Electronic Data Capture (REDCap) platform. The study was approved by the Stanford University and Stony Brook University Institutional Review Boards, and all participants gave digital informed consent before starting the survey.

The COMPASS-31 is a widely-utilized patient questionnaire that provides a quantitative assessment of the severity and distribution of autonomic symptoms [5] . This questionnaire generates a weighted score from 0 to 100, and questions fall into one of six domains: orthostatic intolerance, vasomotor, secretomotor, gastrointestinal, bladder, and pupillomotor function. A COMPASS-31 score of ≥ 20 suggests moderate-tosevere autonomic dysfunction [6] . The OHQ is used to evaluate the severity of orthostatic intolerance and the functional impact of neurogenic orthostatic hypotension [7] , and, by extension, other disorders of orthostatic intolerance including POTS. It consists of two components, a 6-item symptoms assessment scale and a 4-item daily activity scale. Scores range from 0 to 100, with higher scores representing higher symptom burden. The FSS is a 9-item instrument to assess the effect that fatigue has on daily functioning [8] . The FSS scores range between 9-63 points, with a score of 36 or more suggesting abnormal levels of fatigue. The ESS is a 24-point scale that quantifies the likelihood of dozing in various situations over the preceding weeks. Scores >10 suggest excessive daytime sleepiness [9] . The NPS scale measures 10 specific qualities associated with neuropathic pain. Scores range from 0 to 100, with higher scores suggesting greater disability [10] . The GAD-7 is a valid and efficient tool to screen for generalized anxiety disorder [11] . The scores range from 0-24 with the cut-offs of 5, 10, . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) and 15 representing mild, moderate, and severe levels of anxiety. The RAND 36-Item Health Survey is a quality of life measure that incorporates eight health concepts including physical functioning, bodily pain, role limitations due to physical health problems, role limitations due to personal or emotional problems, emotional well-being, social functioning, energy/fatigue, and general health perceptions [12] . Each health concept is scored on a 0-to-100 range, with lower scores representing greater disability.

In total, 4649 participants responded to the survey (figure 1). Surveys were removed from the dataset based on the following exclusion criteria: incomplete dataset, symptom duration < 30 days, symptom onset before November 2019 and age ≥ 65 years. Participants ≥ 65 years of age were excluded due to concern of survivor bias secondary to disproportionately high mortality in this age range. Ultimately, 2314 survey responses were analyzed, including 1249 test-confirmed participants and 1065 test-unconfirmed participants. Both cohorts were included but analyzed separately.

A tiered approach to analyses was used to first determine comparability of the testconfirmed and test-unconfirmed groups, with subsequent analyses exploring the associations with clinical severity (i.e., hospitalized vs non-hospitalized) being assessed only in the diagnostically confirmed group. Categorical variables are presented as count and percentages, and continuous variables as mean±SD for Gaussian variables or median and interquartile range for non-Gaussian variables, as confirmed by the Shapiro-. CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 28, 2022 Fisher's exact test (when counts fell below 5 in any category) were used to compare categorical variables between groups, and Cramér's V provided an estimation of effect size. Spearman correlation coefficients were calculated to assess strength and alignment of associations between COMPASS-31 scores and other survey measures. A correlation coefficient (r) of >0.5 or <-0.5 was interpreted as a meaningful correlation. A statistical threshold of α=0.05 was set and Bonferroni correction for multiple comparisons was performed for each major analysis. All methods were implemented in python v3.7.3 [13] . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 28, 2022 Relationship of pre-existing conditions with autonomic symptom burden All pre-existing conditions were associated with greater autonomic symptom burden as measured by COMPASS-31 total scores, except for high blood pressure (both groups), obesity (test-confirmed group only) and former smoker/vaper (test-unconfirmed group only) (table 3) . Participants with a history of asthma (p<0.01), obesity (test-unconfirmed group only, p<0.01), vitamin D deficiency (p<0.01), autoimmune disease (p<0.01), food or environmental allergies (p<0.01), anxiety (p<0.01), depression (p<0.01), and smoking/vaping (test-confirmed group only, p<0.01) were more likely to have higher COMPASS-31 scores compared to participants who did not report these conditions. These associations were true in both test-confirmed and test-unconfirmed cohorts, unless otherwise noted.

Overall, 67 % of participants had a median COMPASS-31 score of 20 or above, suggestive of moderate to severe autonomic dysfunction. Test-unconfirmed participants had significantly higher median total COMPASS-31 scores than test-confirmed participants (test-confirmed 26.8±28.9 vs test-unconfirmed 34.7±27.8; p<0.0018) as . CC-BY-NC-ND 4.0 International license It is made available under a 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 April 28, 2022 CC-BY-NC-ND 4.0 International license It is made available under a 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 April 28, 2022 [ 5, 14] ). The NPS did not differ between these two groups (p=0.007). Neither group had ESS scores suggestive of excessive daytime sleepiness (9 [5, 14] , 8 [4, 13] 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 April 28, 2022. ; https://doi.org/10.1101/2022.04.25.22274300 doi: medRxiv preprint 0.5 or -0.5, the only questionnaire with a meaningful correlation with total COMPASS-31 scores was the OHQ (test-confirmed r=0.55, test-unconfirmed; r=0.56).

In this study we report the results of an international survey of 2,314 individuals with PASC. The percentage of female participants in this study (87%) was higher than in other studies where 68-75% percent of patients with PASC were female [14] [15] . The higher percentage of female participants in this study may also be the result of females being more likely to participate in online surveys than males [16] . Most of the PASC participants in this adult population study were between the age of 31 to 65 years (87%), which is consistent with the age distribution seen in other studies [4] .

We assessed 53 different symptoms across eight symptom domains, highlighting the heterogenous nature of PASC. The most common symptoms were fatigue, brain fog, headache, shortness of breath with exertion, body aches, palpitations, lightheadedness, and tachycardia, consistent with prior studies that have noted many of the these same symptoms in PASC patients [2] [17] [4] [18] [19] . An online survey of 3,762 individuals with suspected or confirmed COVID-19 estimated the prevalence of 203 symptoms over 10 organ systems [4] . Similar to our study, some of the most commonly reported symptoms included fatigue (90-100%), brain fog (80-90%), headaches (70-80%), muscle aches (60-70%), palpitations (60-70%), dizziness or vertigo (60-70%), and tachycardia . CC-BY-NC-ND 4.0 International license It is made available under a 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 April 28, 2022. ; https://doi.org/10.1101/2022.04.25.22274300 doi: medRxiv preprint (60-70%). In our study, the test-confirmed group was more likely to report loss of taste (57.6% vs. 38.7%) and loss of smell (61.1% vs. 35.6%) than the test-unconfirmed group.

We suspect that this is because individuals who experienced anosmia and ageusia were more likely to be offered testing, or more likely to pursue testing, as these symptoms were widely reported by the media and public health agencies as key indicator symptoms of acute COVID.

A pre-existing diagnosis of autonomic dysfunction was reported in 5.1% of testconfirmed and 8.3% of test-unconfirmed participants in our study, and the most common autonomic disorder reported was POTS (4.1%, 6.8%), with a prevalence that is much greater than the estimated US prevalence of 0.3-1% [20]. The most reported preexisting conditions in our cohort were anxiety, depression, prior history of smoking or vaping, asthma, food or environmental allergies, obesity, hypertension, and a history of autoimmune disease. The prevalence of pre-COVID asthma and autoimmunity reported in our cohort are well above the rates of these medical conditions in the general US population, raising the possibility that these medical conditions could be risk factors for developing PASC [23, 24] . In comparison, the prevalence of pre-COVID anxiety and depression reported in our cohort were similar to general US population rates, suggesting that anxiety and depression are not necessarily risk factors for developing PASC [21] . All medical conditions analyzed except for hypertension (both groups), obesity (test-confirmed group), and current and former smoker/vaper (testunconfirmed group) were associated with higher COMPASS-31 scores. The COMPASS-31 . CC-BY-NC-ND 4.0 International license It is made available under a 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 April 28, 2022. ; https://doi.org/10.1101/2022.04.25.22274300 doi: medRxiv preprint was originally designed to identify autonomic failure, which is associated with orthostatic hypotension and a reduction in sympathetic nervous system activity.

Hypertension, obesity and tobacco use are all associated with sympathetic activation, [25] . Participants in the cluster with elevated COMPASS-31 scores reported more depression, chills, weakness, diarrhea, musculoskeletal pain, palpitations/tachycardia, cognitive involvement, headache, dizziness, and tinnitus. In another study from Germany, 42 patients with PASC who had persistent fatigue six months after mostly mild acute infection were assessed with the COMPASS-31 [6] . A majority (76%) had COMPASS-31 scores suggestive of moderate (score 20-40, n = 21) or severe (score 40-60, n = 11) autonomic dysfunction. Our study confirms the findings of these smaller studies and supports the concept that autonomic dysfunction is highly prevalent in PASC.

. CC-BY-NC-ND 4.0 International license It is made available under a 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 April 28, 2022. ; https://doi.org/10.1101/2022.04.25.22274300 doi: medRxiv preprint

To evaluate whether the severity of acute COVID-19 infection increases the risk of developing autonomic dysfunction in PASC, we compared test-confirmed hospitalized to test-confirmed non-hospitalized participants. The hospitalized group was older and had a higher BMI than the non-hospitalized group, consistent with prior observations that increasing age and BMI are known risk factors for more severe COVID-19 infection [26] .

The hospitalized group had higher scores on the OHQ (P<0.0018) and FSS (P<0.0018), as well as lower scores on several subdomains of the RAND-36, compared to the nonhospitalized group, suggesting that more severe COVID-19 illness may lead to increased orthostatic burden, worse fatigue and impaired quality of life. However, there was no difference in total COMPASS-31 scores between hospitalized and non-hospitalized participants (p=0.06), suggesting that total autonomic symptom burden in PASC is independent of the severity of the acute SARS-CoV-2 infection.

Test-unconfirmed participants had a substantially longer duration from symptom onset to survey completion compared to the test-confirmed group (P<0.0018). One explanation for this is that the test-unconfirmed group were more likely to have had COVID-19 earlier on in the pandemic, when access to testing and treatment was limited.

Furthermore, hospitalized patients early in the pandemic would have been more likely to have been tested than those with milder disease, which might explain why participants in the test-unconfirmed group had lower rates of hospitalization (14.4% vs 7.8%). Interestingly, the test-unconfirmed group also had significantly greater total COMPASS-31 scores than the test-confirmed group (P<0.0018). In addition, all . CC-BY-NC-ND 4.0 International license It is made available under a 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 April 28, 2022. Furthermore, the acute sequelae of COVID-19 infection may have been underrecognized early in the pandemic, leading to worse outcomes. Another possibility is that testunconfirmed participants experienced lower quality care due to a lack of a unifying diagnosis and discounting of poorly understood symptoms by clinicians, resulting in delayed access to treatment. Additionally, it is possible that different variants present different risks of developing PASC. Lastly, it is possible that some of the testunconfirmed participants did not actually have COVID, however this would not explain why the test-unconfirmed group had higher COMPASS-31 scores. It is worth noting that 63.6 % of the test-unconfirmed participants were diagnosed with COVID-19 by a clinician.

. CC-BY-NC-ND 4.0 International license It is made available under a 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 April 28, 2022 [32] . Similarly, the high FSS scores seen in our cohort is also consistent with prior studies that have demonstrated that fatigue is very common in PASC [18] [19] [17] [4] . Interestingly, none of the groups had excessively high ESS scores, suggesting that fatigue is more common than sleepiness in PASC. The mildly elevated GAD-7 scores seen in all groups is in line with data demonstrating an increase in anxiety in the general population during the pandemic [21] . All groups reported substantial impairment across all subdomains of the RAND-36, highlighting the significant impact that PASC has on quality of life, which is particularly profound in the domains of "role limitations to physical health" and "role limitations to emotional problems."

Of all the questionnaires evaluated in this study, the only questionnaire that was strongly correlated with COMPASS-31 scores was the OHQ. It is not surprising that the OHQ scores correlated with higher COMPASS-31 scores given that disorders of orthostatic intolerance such as POTS and OH are the most common forms of dysautonomia presenting in PASC [2] . The FSS, ESS, and GAD-7 were likely not strongly associated with autonomic dysfunction because fatigue, sleepiness, and anxiety can occur, but are not specific to autonomic dysfunction.

. CC-BY-NC-ND 4.0 International license It is made available under a 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 April 28, 2022 We were surprised that the RAND-36 subdomains did not have a stronger correlation with COMPASS-31 scores. We expected that the presence of autonomic dysfunction in PASC would have had a greater impact on quality of life. These findings do not exclude the possibility that autonomic dysfunction in PASC is associated with a reduced quality of life, but rather suggest that PASC, even in the absence of autonomic dysfunction, has a very high burden of disability.

The limitations of our study include the retrospective nature, which could lead to participants either overreporting or underreporting symptoms due to recall bias. In addition, the study population was limited to those who were English speakers and had access to the PASC support groups and social media channels that our study utilized for recruitment. Finally, lack of confirmed-positive COVID test results in all patients represents a limitation inherent to the design of our survey-based study, however this is counterbalanced by the benefits of utilizing an online survey to access a large, global population of PASC patients, as well as the opportunity to include those infected early in the pandemic without access to COVID testing.

In conclusion, this is the largest study to date that has utilized validated autonomic questionnaire scores to demonstrate that autonomic dysfunction is highly prevalent in PASC. COVID severity did not correlate with the degree of autonomic dysfunction in our cohort, suggesting that even mild cases of COVID can result in significant alterations of autonomic function. Given the role of the autonomic nervous system in regulating . CC-BY-NC-ND 4.0 International license It is made available under a 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 April 28, 2022. ; https://doi.org/10.1101/2022.04.25.22274300 doi: medRxiv preprint immune function and coagulation pathways, both of which appear to play a role in long COVID [33] [34] , future studies should focus on mechanisms of autonomic dysfunction in PASC, their relationship to immune and coagulation biomarkers, and potential interventions that can improve autonomic function.

A conflict of interest disclosure is provided.

We acknowledge Dysautonomia International who helped with the recruitment of participants.

. CC-BY-NC-ND 4.0 International license It is made available under a 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 April 28, 2022. ; https://doi.org/10.1101/2022.04.25.22274300 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a 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 April 28, 2022 *Some respondents checked more than 1 category **Some data was incomplete (n=1234 testing confirmed, n=1049 testing unconfirmed) . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 28, 2022 . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.25.22274300 doi: medRxiv preprint is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.25.22274300 doi: medRxiv preprint . CC-BY-NC-ND 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 28, 2022. ; https://doi.org/10.1101/2022.04.25.22274300 doi: medRxiv preprint

Post-acute COVID-19 syndrome

Preparing for the long-haul: Autonomic complications of COVID-19

The face of postural tachycardia syndrome -insights from a large cross-sectional online community-based survey

Characterizing long COVID in an international cohort: 7 months of symptoms and their impact

COMPASS 31: a refined and abbreviated Composite Autonomic Symptom Score

Chronic COVID-19 Syndrome and Chronic Fatigue Syndrome (ME/CFS) following the first pandemic wave in Germany -a first analysis of a prospective observational study

The Orthostatic Hypotension Questionnaire (OHQ): Validation of a novel symptom assessment scale

The fatigue severity scale: Application to patients with multiple sclerosis and systemic lupus erythematosus

A new method for measuring daytime sleepiness: the Epworth sleepiness scale

Development and preliminary validation of a pain measure specific to neuropathic pain: The Neuropathic Pain Scale

A brief measure for assessing generalized anxiety disorder: The GAD-7

36-Item Short Form Survey (SF-36) | RAND

Python tutorial

Women differ distinctly from men -Mayo Clinic News Network

Post-COVID-19 Syndrome (Long Haul Syndrome): Description of a Multidisciplinary Clinic at Mayo Clinic and Characteristics of the Initial Patient Cohort

Does gender influence online survey participation? A recordlinkage analysis of university faculty online survey response behavior

2021) 6-month consequences of COVID-19 in patients discharged from hospital: a cohort study

Symptom Duration and Risk Factors for Delayed Return to Usual Health Among Outpatients with COVID-19 in a Multistate Health Care Systems Network -United

Post-discharge persistent symptoms and health-related quality of life after hospitalization for COVID-19

Beyond Symptoms: Race and Gender Predict Anxiety Disorder Diagnosis

Post-COVID syndrome. A case series and comprehensive review

Underlying Medical Conditions and Severe Illness Among 540,667 Adults Hospitalized With COVID-19

Association between vaccination status and reported incidence of post-acute COVID-19 symptoms in Israel: a crosssectional study of patients infected between

A case report of postural tachycardia syndrome after COVID-19

New-onset Postural Orthostatic Tachycardia Syndrome Following Coronavirus Disease

Autonomic dysfunction in 'long COVID': rationale, physiology and management strategies

Long-Haul Post-COVID-19 Symptoms Presenting as a Variant of Postural Orthostatic Tachycardia Syndrome: The Swedish Experience

Neurological Outcome and Quality of Life Three Months after COVID-19: A Prospective Observational Cohort Study

A short corticosteroid course reduces symptoms and immunological alterations underlying long-COVID

Persistent clotting protein pathology in Long COVID/Post-Acute Sequelae of COVID-19 (PASC) is accompanied by increased levels of antiplasmin