While most patients recover within weeks of SARS-CoV-2 infection, others experience debilitating symptoms that persist beyond the acute period (1). The overall global prevalence of post-COVID-19 conditions is estimated at 0.43 of acute cases (hospitalized 0.54 and non-hospitalized 0.36) (2). These post-COVID conditions, collectively known as a Post-Acute Sequelae of SARS-CoV-2 infection (PASC), or long COVID, are increasingly recognized even in patients who experience asymptomatic or mild SARS-CoV-2 infection (3). The Center for Disease Control and Prevention (CDC) defines post-COVID conditions as symptoms persisting beyond 28 days after infection (4), while the UK National Institute for Health and Care Excellence (NICE) (5) and the World Health Organization (WHO) define this syndrome as symptoms persisting beyond 12 weeks after infection (6). The American Academic of Physical Medicine and Rehabilitation (AAPMR&R) estimates that there are more than 24 million cases of long COVID as of May, 2022 (7).

PASC is most often characterized by extreme fatigue exacerbated by exertion, referred to as post-exertional malaise, difficulty with concentration and memory often referred to as brain fog, sleep disturbances, headaches, chest pain, and shortness of breath (2, 8). PASC can range from mild to severe and incapacitating, interfering with patients' daily activities and work requirements (2, 8).

In an online, multi-national survey of 3,762 participants with confirmed or suspected COVID-19, fatigue, post-exertional malaise, and brain fog were the most frequently reported symptoms six months after SARS-CoV-2 infection (8). This cluster of symptoms shares similar features with Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS), an often-debilitating disease that has a worldwide prevalence close to one percent, is also believed to frequently arise following a viral infection (9, 10). Women are affected 1.5 to 2 times more often than men (2). In two small cohorts of 41 and 42 COVID-19 patients ~45% met the ME/CFS diagnostic criteria (11, 12). The pathobiological mechanism of ME/CFS is not known, but the large number of PASC patients presenting with features of this syndrome may allow us to better understand both conditions. The goals of our study are 2-fold: 1. Characterize our clinical cohort in terms of demographics and clinical presentations. 2. Estimate the prevalence of ME/CFS phenotype in PASC based on the IOM ME/CFS criteria (9).

Among the 140 patients referred to the Stanford PASC clinic, six patients were excluded because of lack of a diagnostic test for SARS-CoV-2 infection or an incomplete questionnaire. For the remaining 134 patients, the median age was 47 years old (IQR: 30–60; range: 20–88). Seventy-nine (59%) were female, 66 self-identified as White (49.3%), 23 as Hispanic (17.2%), 20 as Asian (14.9%), and 9 as Black (6.7%). Seventeen (12.7%) required hospitalization for their acute infection and were considered to have had severe disease. Two of these patients required Intensive Care Unit (ICU) admission. Sixty-two (46.3%) patients had a comorbid condition known to predispose to severe disease including obesity (BMI > 30 kg/m²) 47 (35.1%), hypertension 24 (17.9%), chronic lung disease 16 (11.9%), diabetes 9 (6.7%), immunosuppressive therapy 5 (3.7%) and/or cardiovascular disease 4 (3%). One hundred and nine (82%) patients had some degree of functional limitation (i.e., FSS of 3, 4, or 5), including 45 (33.9%) who exhibited significantly compromised wellbeing (FSS of 4 or 5) (Table 1).

The duration of the symptoms at the time of the initial office visit ranged from 34 to 792 days with a median duration of 285.5 days. Of the 29 symptoms assessed in the study, the median number per patient was 12 (range: 1–25 symptoms). The most common symptoms were fatigue (86.5%), post-exertional malaise (82.8%), brain fog (81.2%), unrefreshing sleep (76.7%), and daytime sleepiness (74.6%). The median number of symptoms in female patients was greater than the median number in male patients (13 vs. 10, p = 0.005). The most common symptoms significantly higher in the female population were fatigue, insomnia, and change in taste or dysgeusia (P-values = 0.001, 0.044, and 0.01, respectively) (Table 2). There was a significant correlation between the frequency and the median severity of the symptom (Correlation coefficient is 0.86, P < 0.001).

In the PCA analysis, the direction of arrows in the correlation circle illustrates the extent to which symptoms were likely to occur in the same patient while the factor map indicates the contribution of different symptoms to the first five components (Figure 3). The analysis suggests that fatigue, PEM, daytime sleepiness (lethargy), brain fog, and unrefreshing sleep were likely to occur together. These symptoms were major contributors to the first component while anosmia and nasal congestion were major contributors to the second component. Similar analysis was performed on 72 patients without pre-COVID comorbidities two main groups were identified one with anosmia, headache, and congestion and a second that includes unrefreshing sleep, fatigue, and lethargy.

We assessed the prevalence of the ME/CFS phenotype in 105 patients (69%) who had persistent symptoms longer than 6 months. The most common symptoms were clustered in the ME/CFS (Fatigue, PEM, brain fog, unrefreshing sleep, and lethargy) and autonomic dysfunction categories (lightheadedness and gastrointestinal) (Figure 3). The top six most common and severe symptoms were fatigue, PEM, brain fog, unrefreshing sleep, daytime sleepiness, and insomnia (Figure 2). Forty-five (43%) of the study cohort fulfill the IOM criteria for ME/CFS (Table 1). Like our PASC cohort, the ME/CFS cohort was predominantly female, non-hospitalized, and healthy individuals and obesity was the most common risk factor (BMI > 30 Kg/m²) (Table 1). But obesity and worse FSS (FFS 4 and 5) were significantly higher in the ME/CFS population (P = 0.00377895; P = 0.0110474, respectively) (Table 1).

This manuscript presents the characteristics of 134 patients referred to a long COVID clinic who had who have history of SARS-CoV-2 infection and more than 28 more days of symptoms. The median age in our cohort was 47 years and ranged from 20 to 88 years, with a female predominance of 59%. Similar age and sex distribution was reported in several other studies (2, 16). In contrast to hospitalized patients, females PASC were represented in a lower proportion, less severe symptoms, and lower mortality than males (17–19). These differences in sex distribution could be explained by the variations in the immune response between males and females, and the fact that female patients have more robust inflammatory, antiviral, and humoral immune responses, biological differences on sex hormones, and expression and regulation of angiotensin-converting enzyme 2 (ACE2) (18, 20). Our study cohort, similar to other PASC studies, the subjects were predominantly white females, with obesity (17, 21, 22) characteristics associated with lower likelihood for full recovery (23). In contrast, some racial and ethnic minority groups, such as Native American Indians, Alaska Natives, Hispanic and Black, have been shown to have a disproportionately higher risk for infection, severity of illness, hospitalization, and deaths. Those groups were underrepresented in our study (24).

The large majority (87%) of our PASC patients had not been hospitalized or required oxygen for COVID-19 disease. The prevalence of PASC in asymptomatic patients with the mild disease is reported between 30 and 60% (3, 25). We can postulate that unknown factors besides hypoxemia and hospitalization are the drivers of PASC symptoms such as virus persistence, overactivation of the immune system, amyloid fibrin microclots, auto-antibodies, virus reactivation and others, may play a role in the pathobiology of this illness (26).

In our study, the duration of symptoms ranged from 34 to 792 days, with a median duration of 285.5 days. Females experienced more symptoms than males. The most common symptoms were fatigue, post-exertional malaise, brain fog, unrefreshing sleep, and daytime sleepiness; and the frequency of the symptoms correlated with severity (Figure 2, Table 2) and the ME/CFS symptoms clustered with one another (Figure 3). A reliable and straightforward long COVID score system is needed in post-COVID research and clinical care. In a population-based cohort study of confirmed SARS-CoV-2 infection, higher severity scores correlate with lower health quality of life (27). In a meta-analysis which found that fatigue/weakness, myalgia/arthralgia, depression, anxiety, memory loss, concentration difficulties, dyspnea, and insomnia, were the most prevalent symptoms (2, 28). Fatigue was the most prevalent symptom across the PASC studies (2, 28). Post-viral fatigue was commonly reported after a viral infection such as Influenza, Severe acute respiratory syndrome/Middle East respiratory syndrome coronavirus (SARS/MERS), and Ebola (29).

When we compare the frequency of the symptoms and severity, we found a significant correlation between the frequency and the severity of the symptom (Likert scale 4 and 5). A reliable and simple long COVID score system is tool need in post COVID research and clinical care. It will be important to access clinical outcomes and evaluated therapeutics interventions. In a population-based cohort study confirmed SARS-CoV-2 infection, Bahmer et al., the higher score in severity of COVID symptoms correlated with lower health quality of life (27).

PCA is often used to transform a large set of variables into a smaller one that contains most of the information in the large set. Smaller data sets are easier to explore and visualize and facilitate analyzing data much easier and faster (15). We used it to cluster symptoms and severity. We were able to visualize a group that resembles ME/CFS.

In a study on 233 SARS survivors, 40.3 % reported having chronic fatigue, and 27.1% met the criteria for ME/CFS; after influenza with H1N1, ME/CFS has reported 2.08 cases/100,000 person-month (29). In our study, 43% of the selected cohort fulfilled all the ME/CFS criteria, which is similar to 45% reported by Mancini et al. (11) and Kedor et al. (12). Like ME/CFS, the ME/CFS-PASC phenotype was more prevalent among the non-hospitalized female population (Table 1). The clinical similarities in our study cohort between ME/CFS and ME/CFS-PASC allow us to suggest common pathobiology. Those similarities include a preceding a viral illness (30), increase in inflammatory cytokines, neuroinflammatory change, mitochondria dysfunction, and alteration in NK cell function (31). Obesity (BMI > 30 Kg/m²) was in our PASC study cohort, and the PASC-ME/CFS is the most common risk factor for this illness that has also been recognized in multiple studies as a risk factor for Long COVID (32, 33). In addition, PASC-ME/CFS was an indicator worse functional status. However, we were not able to confirm that all symptoms resulted from prior SARS-CoV-2 infection and that it is possible that many of the reported symptoms, including chronic fatigue had other contributing causes. We believe there is an important relationship between Obstructive Sleep Apnea (OSA) and COVID-19, with significant symptomatic overlap between these two conditions and emerging data suggesting COVID is a risk factor for OSA, increasing the risk for severity and hospitalization (34, 35). For these reasons, we agree that the relationship between long COVID and OSA deserves to be further explored, though this exploration is beyond the intended scope of this study.

Our study has several limitations. First, this study only represents the experience of a single center in Northern California located in a generally affluent area with a bias toward specific populations. The selection of this cohort may skew our population in the follow ways: (1) this is a referred population with multiple and more severe symptoms (2) our clinics have a lower proportion of underrepresented minority populations. Therefore, a multicenter study that includes a more diverse and larger population is necessary to corroborate our findings.

In summary, about half of the PASC patients with more than 6 months of symptoms fulfilled the ME/CFS criteria, and PASC-ME/CFS is an indicator of worse functional status; the majority of patients seen in our PACS clinic were female, with a median age of 47 years old, with obesity as the most common comorbidity. The majority of the patients had mild to moderate acute infection and were healthy prior to their COVID infection. Fatigue, post-exertional malaise, brain fog, unrefreshing sleep, and daytime sleepiness were the most prevalent and severe symptoms. This commonality between ME/CFS and ME/CFS-PASC may suggest a shared pathobiology. Therefore, defining specific subtypes within the umbrella of PASC/post-COVID conditions can help us understand different pathogenic mechanisms to tailor treatment.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

The studies involving human participants were reviewed and approved by Stanford University IRB. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

HB: conception and design of the study, data collection, analysis, draft, editing of the manuscript, full access to the whole data, and final version approval. TQ, AT, and AB: data collection, full access to the whole data, analysis, draft, and editing of the manuscript. MM, PY, LE, AS, LS, NS, HH, ZP, HS, AH, OS-p, and AK: conception and design of the study, full access to the whole data, and editing of the manuscript. RS and LG: conception and design of the study, full access to the whole data, data collection, analysis, draft, editing of the manuscript, and final version approval. All authors contributed to the article and approved the submitted version.