Introduction

Fatigue is recognised as one of the most common presenting complaints in individuals infected with SARS-CoV-2, the cause of the current COVID-19 pandemic. In early reports on the clinical characteristics of those infected, fatigue was listed as a presenting complaint in 44–69.6% [1–3]. Further studies were followed by meta-analyses, with 34–46% of those infected presenting with fatigue [4–7]. Whilst the presenting features of SARS-CoV-2 infection have been well-characterised, with symptoms ranging from mild taste and smell disturbance to dyspnoea and respiratory failure, the medium and long-term consequences of SARS-CoV-2 infection remain unexplored [1, 8, 9]. In particular, concern has been raised that SARS-CoV-2 has the potential to trigger a post-viral fatigue syndrome [10, 11].

Patients acutely infected with SARS-CoV-2 demonstrate decreased lymphocyte counts, higher leukocyte counts with an elevated neutrophil-to-lymphocyte ratio (NLR) in addition to decreased percentages of monocytes, eosinophils and basophils. It has also been reported that both helper and suppressor T cells are decreased in those with SARS-CoV-2 [12]. In severe cases, elevated C-reactive protein (CRP), ferritin, d-dimers in addition to pro-inflammatory factors such as IL-6 and soluble CD25 (sC25), and an increase in intermediate (CD16+ CD14+) monocytes have been reported [13, 14]. Whether or not the immunological alterations seen in SARS-CoV-2 have any relationship to the potential development of medium and long-term symptoms following infection is an area which has not been researched to date. The persistence of these changes following resolution of initial infection have also not been examined.

In one of the few reports to assess the long-term consequences of the severe acute respiratory syndrome (SARS) epidemic (caused by the novel coronavirus SARS-CoV), a subset of patients in Toronto experienced persistent fatigue, diffuse myalgia, weakness and depression one year after their acute illness and could not return to work [15]. In a similar follow-up study amongst 233 SARS survivors in Hong Kong, over 40% of respondents reported a chronic fatigue problem 40 months after infection [16]. In those affected by the subsequent Middle-Eastern Respiratory Syndrome Coronavirus (MERS-CoV) outbreak, prolonged symptoms and fatigue were reported up to 18 months after acute infection [17]. Similarly, prominent post-infectious fatigue syndromes have been reported following Epstein-Barr Virus (EBV), Q-Fever and Ross River Virus (RRV) infections, as well as rickettsiosis [18–22]. It has also been reported following HHV-6 infection and in HIV infection [23, 24]. Whether or not infection with SARS-CoV-2 has the potential to result in post-viral fatigue, both in the medium and long-term, is currently unknown.

Persistent fatigue lasting 6 months or longer without an alternate explanation is termed chronic fatigue syndrome (CFS). This may be observed after several viral and bacterial infections [11]. There have also been links between CFS and depression, although it remains unclear whether one diagnosis precedes the onset of the other [25–27]. Whilst infections are thought to precipitate CFS, the pathophysiology remains controversial. Studies of post-viral fatigue and CFS often focus on immune system alterations, but robust data to indicate causation or association is absent. There are a plethora of studies examining immune dysregulation and activation in CFS; however, none of these have provided a consistent finding or biologically plausible answer; rather, there are contrasting findings across studies concerning both immune population changes and cytokine levels [28–30]. The heterogenous findings in immune populations in CFS include changes in both lymphoid and myeloid populations [31–34]. The disparate findings of prior CFS studies may be due to the variety of aetiologies that ultimately lead to CFS. Whether alterations in immune system activity has any relationship to the potential post-viral fatigue experienced with the novel SARS-CoV-2 is an important question for future research. Prospectively examining patients following SARS-CoV-2 infection provides a well characterised population with identical index infection, allowing for more accurate descriptors of both disease state and disease characteristics.

We sought to establish whether patients recovering from SARS-CoV-2 infection remained fatigued after their physical recovery, and to investigate whether there was a relationship between severe fatigue and a variety of clinicopathological parameters. We also sought to examine persistence of markers of disease beyond clinical resolution of infection.

Materials and methods

Results

Participant characteristics

223 patients were offered an outpatient appointment, of which 128 (57%) attended for assessment. The remaining 43% declined an outpatient appointment. These were consecutively enrolled (mean age: 49.5 ± 15 years; 53.9% female). Just over half (71/128; 55.5%) were admitted to hospital for treatment of COVID-19, with the remainder managed as outpatients (57/128; 44.5%). Of the admitted cohort, 35 (49%) received hydroxychloroquine, while 6 (8.5%) of these also received prednisolone. The remaining 36 inpatients did not receive any targeted therapy. No outpatients received targeted therapy. Administration of targeted therapy was at the discretion of the treating physician, in line with the guidelines in our institution at that time. Administration of COVID-directed therapy had no association with fatigue status. Just over half (66/128; 51.6%) were healthcare workers. Baseline characteristics are detailed in Table 1.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Baseline characteristics of study participants by fatigue case status (“caseness”).

https://doi.org/10.1371/journal.pone.0240784.t001

The median interval between study assessment and discharge from hospital or a timepoint 14 days following diagnosis if managed as an outpatient was 72 days (IQR: 62–87). Fifty-four patients (54/128; 42.2%) reported feeling back to their full health, whilst the majority did not. Prior to COVID-19 illness, the majority (82%;105/128) had been employed, of whom 33 (31%) had not returned to work at time of study participation.

COVID-19 disease characteristics and fatigue

Overall, there was no association, either using unadjusted models, or models adjusted for age and sex, between COVID-19 disease related characteristics (days since symptom onset, need for inpatient admission/supplemental oxygen/critical care, length of hospital stay) and either fatigue “caseness” (using logistic regression) or total CFQ-11 score (using linear regression) (See Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Association of COVID-19, laboratory values and circulating pro-inflammatory cytokines with fatigue case status (Fatigue vs non-fatigued/non-severe fatigue) and total fatigue score (CFQ-11).

https://doi.org/10.1371/journal.pone.0240784.t002

Discussion

The current study represents, to our knowledge, the first report in the literature examining the prevalence of fatigue following SARS-CoV-2 infection. There is a significant burden of fatigue at median follow up of 10 weeks, with half of the patient cohort reporting severe fatigue. This has several profound implications. Firstly, 50% of the participants do not feel back to full health, despite being medically deemed recovered from their primary illness. Secondly, the impact of this fatigue on daily function is already evident, with almost one third (31%) having not returned to employment. This is of particular concern, given that it is recommended that post-viral infection return to work should take place after four weeks to prevent deconditioning [43]. The large number of healthcare workers in our cohort is reflective of the overall demographics of Irish data and our institution, where 50% of positive SARS-CoV-2 cases involved healthcare workers [44]. The high proportion of healthcare workers infected by COVID-19, not just in our cohort but internationally, means that this will have a significant impact on healthcare systems [44–46]. These findings are independent of ageing effects, as age is known to be associated with increasing fatigue [47].

The rates of post-COVID fatigue appear much higher than those previously reported following EBV, Q fever or RRV infection at a similar interval [21]. However, post-SARS fatigue has been reported in 40% of individuals one year after initial infection, with 1 in 4 meeting CFS diagnostic criteria at that timepoint [16]. The levels of both physical and psychological fatigue seen post-COVID are higher than those of the general population, but do not reach the levels of those seen in chronic fatigue syndrome [48–50]. Rates of fatigue seen in our cohort are roughly equivalent to those reported in chronic disease states [51, 52]. Given that this cohort has no enduring evidence of active infection, the rate of fatigue is noteworthy. This is particularly important in relation to the 52% of the cohort that meet the diagnostic criteria for fatigue, as their CFQ-11 scores approach those seen in CFS cohorts [53, 54].

The findings concerning correlates of SARS-CoV-2-related fatigue are also notable. The absence of association with severity of initial infection has major implications on both the potential number of patients that may be affected and the burden this will place on healthcare services. Previous studies on SARS have generally focused on function and fatigue in post-ICU patients [55]. Our findings would suggest that all patients diagnosed with SARS-CoV-2 will require screening for fatigue. Our results also show a distinct female preponderance in the development of fatigue. This is in keeping with previous CFS findings [56]. We also noted significant association with pre-existing diagnosis of depression and use of anti-depressant medications and subsequent development of severe fatigue. While depression and CFS have previously been associated, there has been some debate as to the temporal relationship [21, 27]. Longitudinal studies will be needed to assess subsequent development of depression in the aftermath of post-COVID fatigue, as well as assessing the trajectory and persistence of fatigue.

The absence of a specific immune signature associated with persistent fatigue is a striking positive finding. As alluded to previously, CFS has been associated with a large number of differing changes in the inflammatory markers and immune cell populations. However, no consistent change has been reported across multiple studies [30]. This, in combination with our results, leads us to speculate that the pathological changes associated with CFS and post-COVID fatigue are more subtle. CFS may be the end point of a variety of distinct pathways or may be the consequence of pathological changes that are no longer systemically detectable. Despite a lack of distinct immunological findings, it is accepted that CFS can occur in the absence of demonstrable disease [57, 58]. The lack of distinct immune signature, coupled with the association with depression, lends credence to the multifactorial aetiology of CFS [59]. It also supports the use of non-pharmacological interventions for fatigue management and provides no basis for the use of immunomodulation in treating post-COVID fatigue.

Our study concerned findings around COVID-19-related fatigue in the medium-term. In line with published data on viral dynamics, infectivity, and duration of infection, all our participants had recovered from their acute COVID-19 illness [60–62]. The median period between symptom onset and fatigue assessment was ten weeks, with no participant being recruited earlier than six weeks after their last COVID-19 symptoms or hospital discharge. Studies on CFS and post-viral fatigue have commonly assessed individuals at least 6 months after their viral illness. Post-SARS fatigue was described in 22 patients between 1 and 3 years post disease resolution; these patients were chosen due to their symptoms and may therefore not be representative of the overall cohort [15]. We feel that the short interval reported here is relevant due to the burden of fatigue seen and that COVID-19 patients were seen irrespective of post-disease symptoms, minimising the risk of selection bias. We also believe the effect fatigue has on self-perceived health and return to work is profound and worthy of reporting, especially in light of the number of patients that will be affected by this and the potential impact on individuals, employers and governments.

Management of fatigue states requires multi-disciplinary input and will not be appropriately addressed if follow-up is by treating medical physicians alone. A suite of interventions, including graded exercise and cognitive behavioural therapy, are needed to manage CFS and may be relevant to post infectious fatigue [63–65]. Furthermore, successful return to work will require ongoing input from occupational health departments and employers [66].

Our single centre study in a predominantly white Irish population has several limitations worthy of discussion. Our study is cross-sectional in nature and only assessed participants at a single timepoint. As previously mentioned, we are also reporting at a medium time point. As such, we would recommend that longitudinal studies are designed to assess patients at multiple time points and to examine the changes in immune markers and immune cell populations over time. It will also be illustrative to describe the persistence of fatigue at six months and beyond. It is important to note that there is no consensus on the nature of fatigue and its evaluation. However, the use of the widely applied Chalder Fatigue Scale is appropriate in this context. Further studies in large cohorts will be required to tease out fatigue subgroups and the potential complex factors at play. We also suggest that it is now time to consider the management of this post-COVID syndrome and advocate early analysis of multi-disciplinary fatigue management strategies.

Conclusions

We present the first report, to our knowledge, of post-viral fatigue in those recovered from the acute phase of COVID-19 illness. In a similar fashion to previous coronavirus pandemics, COVID-19 appears to result in symptoms of severe fatigue that outlast the initial acute illness. Over half of individuals in the current study demonstrated symptoms consistent with severe fatigue a median of 10 weeks after their initial illness, while almost one-third of those previously employed had not returned to work. Most interestingly, fatigue was not associated with initial disease severity, and there were no detectable differences in pro-inflammatory cytokines or immune cell populations. Pre-existing diagnosis of depression is associated with severe post-COVID fatigue. This study highlights the burden of fatigue, the impact on return to work and the importance of following all patients diagnosed with COVID, not merely those who required hospitalisation. There are enormous numbers of patients recovering from SARS-CoV-2 infection worldwide. A lengthy post-infection fatigue burden will impair quality of life and will have significant impact on individuals, employers and healthcare systems. These important early observations highlight an emerging issue. These findings should be used to inform management strategies for convalescent patients and allow intervention to occur in a timely manner.

Acknowledgments

We would like to acknowledge the teams involved in the care of these patients, in particular the intensive care, immunology, psychological medicine and infectious diseases teams in St James’s Hospital, as well as the contributions of the immunology diagnostic laboratory team. We would specifically like to acknowledge the contributions of the following people to the acquisition of data for this study: Dr Tara Kingston, Ms Patricia Byrne, Dr Katie Ridge, Dr Fionnuala Cox, Dr Catherine King.