COVID-19 sequelae in adults aged less than 50 years: A systematic review

Background There is emerging evidence of long-term sequelae in a considerable proportion of COVID-19 patients after recovery and the spectrum and severity of such sequelae should be systematically reviewed. This review aims to evaluate the available evidence of all intermediate and long-term COVID-19 sequelae affecting formerly healthy adults. Methods A systematic literature search of Embase, WHO, Scopus, Pubmed, Litcovid, bioRxiv and medRxiv was conducted with a cutoff date of the 17th September 2020 according to PRISMA guidelines and registered in PROSPERO (CRD42020208725). Search terms included “COVID-19”, “coronavirus disease 2019”, “SARS-CoV-2”, “sequelae” and “consequence*”. Publications on adult participants, with a confirmed SARS-CoV-2 infection were included. Elderly (>50 years old) and children (<18 years old) were excluded. Bias assessment was performed using a modified Newcastle-Ottawa Scale. Results A total of 31 papers were included. Study types included prospective and retrospective cohort studies, cross-sectional studies and case reports. Sequelae persistence since infection spanned 14 days to three months. Sequelae included persistent fatigue (39–73% of assessed persons), breathlessness (39–74%), decrease in quality of life (44–69%), impaired pulmonary function, abnormal CT findings including pulmonary fibrosis (39–83%), evidence of peri-/perimyo-/myocarditis (3–26%), changes in microstructural and functional brain integrity with persistent neurological symptoms (55%), increased incidence of psychiatric diagnoses (5.8% versus 2.5–3.4% in controls), incomplete recovery of olfactory and gustatory dysfunction (33–36% of evaluated persons). Conclusions A variety of organ systems are affected by COVID-19 in the intermediate and longer-term after recovery. Main sequelae include post-infectious fatigue, persistent reduced lung function and carditis. Careful follow-up post COVID 19 is indicated to assess and mitigate possible organ damage and preserve life quality.


Introduction
Although SARS-CoV-2 predominantly affects the respiratory system, several studies and interim reports indicate that COVID-19 is a multisystem infection with both overt and subtle health consequences. Data on how long symptoms persist, the impact of such symptoms on everyday life and the short-term, intermediate and long-term sequelae of this infection have scarcely been researched. Most original data publications focus on patients who have been hospitalized, on individuals who have suffered severe illness or on older patients with comorbidities. There are few studies following up on young people who have tested SARS-CoV-2 positive and who have experienced mild illness or indeed, been asymptomatic while testing positive. Such data are particularly important in terms of projecting the future burden of COVID-19 sequelae on healthcare systems and also in the context of "fitness to work" of young population groups as young adults constitute a large proportion of any country's workforce.
In the ongoing pandemic, evidence is emerging of individuals suffering from debilitating symptoms weeks and even months after COVID-19 diagnosis. This is of great concern to the medical community. The large number of such cases becomes apparent when looking at the size of groups such as the so called "COVID long haulers" on social media, the members of such groups often number thousands [1]. Several small studies have followed up on recovered COVID-19 cases, particularly on severe cases. However, a systematic and evidence-based review on long-term COVID-19 sequelae in previously healthy adults is missing.
In this systematic review, we identified and summarized original data published until mid-September 2020. The focus was on formerly healthy young to middle-aged adults. Children and elderly were excluded. Speculations on possible consequences, that may be anticipated in the future based on experiences from other viral infections, complemented this review in the discussion section.

Material and methods
Details of the protocol for this systematic review were registered on PROSPERO and can be accessed at www.crd.york.ac.uk/PROSPERO /display_record.asp?ID=CRD42020208725.

Data sources and searches
A systematic search of published papers and preprints was performed in the databases Embase, PubMed, Scopus, WHO, LitCovid, bioRxiv and medRxiv. The search terms were selected to find papers on sequelae to a SARS-CoV-2 infection and included the terms "COVID-19", "coronavirus disease 2019", "SARS-CoV-2", "sequelae" and "consequence*". The terms "child*" and "pregnant" served as negative qualifier. The online searches were performed between 15th and 17th September 2020. The complete search strategies are included in the Supplement (Appendix 1).

Data management
References found through the systematic search were imported into Mendeley, a reference management tool. After a first elimination of duplicates, a single reference file was generated and uploaded into the Covidence online tool, a non-profit service for screening and facilitating the systematic review process [2]. The assessment of inclusion or exclusion of every reference was performed by two out of eight independent reviewers (N.B., N.H., A.H., R.L., C.S., D.S., K.S., S.W.) according to predefined criteria. Only studies including original data and systematic or narrative reviews published up to 15th September 2020 were included. Only English language papers were included. References including children (below the age of 18 years) or only elderly (age 50 years and above) were excluded. The goal of this systematic review was to focus on young adults aged 18-50 years as they make up the majority of the working population. COVID-19 in this group was initially considered to be a mild illness. In the interim, evidence is emerging of persistent sequelae even in previously healthy young populations. Other systematic reviews have looked at COVID-19 sequelae in the elderly, the polymorbid or those who were hospitalized with severe illness. In such groups, it is more difficult to distinguish sequelae of COVID-19 from other disease progression. Children were excluded due their differing anatomy and physiology and their seemingly lower susceptibility to SARS-CoV-2 virus which makes it difficult to compare them to adults directly. Additionally, study subjects had to have a confirmed SARS-CoV-2 infection. Both sexes and formerly healthy participants as well as participants at risk (as defined by WHO [3]) were included. The study period had to include the follow-up after usual recovery and/or discharge from hospital. Animal, laboratory or in vitro studies were excluded. Any discrepancies were resolved by a third reviewer. The full text screening was performed by two independent reviewers (either K.S. and S. W. or N.B. and S. W.). Conflicts in the full text screening were resolved by discussion. The included studies were carefully screened for other references containing original data that fit the inclusion criteria and were missed by the systematic search.
All the studies finally included were summarized in tables according to organ system. A bias assessment of all cohort and cross-sectional studies was performed by two independent reviewers (R.L. and S.W.) using a modified version of the Newcastle-Ottawa Scale (mNOS) for assessing the quality of non-randomized studies [4] as presented in the Supplement. Discrepancies were resolved by discussion. In addition, an analysis of the evidence level according to the JAMA network (modified from Oxford Centre for Evidence-based Medicine) was carried out for each study [5].

Results
A total of 5229 references were retrieved during the search and imported into Mendeley where 2478 duplicates were removed. Covidence recognized another 30 duplicates, leaving 2721 references for title and abstract screening. A total of 2540 references were deemed irrelevant during the title and abstract screening. Of the remaining 181 references, 18 references containing original data were finally included during the full text review. Additionally, there were 13 studies identified through screening of the included references and through other sources. Fig. 1 shows the PRISMA flow chart diagram [6]. Included studies comprised of 11 prospective and 11 retrospective cohort studies, 4 cross-sectional studies and 5 case reports.
All the references included in this review were summarized in Table 1 including the study quality ratings from the modified Newcastle-Ottawa Scale. Scores were classified as unsatisfactory (0-3 points), satisfactory (4-5 points), good (6-7 points) and very good (8-9 points). A total of four studies were awarded "very good", while five studies received a "good" rating. A majority of 12 studies were deemed "satisfactory". Five studies only received three points and therefore were rated as "unsatisfactory". As the modified Newcastle-Ottawa Scale is not suitable for rating case reports, the remaining five case reports were only rated according to the evidence level as defined by the JAMA network [5]. The detailed rating of each study can be found in eTable 1 in the Supplement.
Only information and subpopulations relevant for this review were listed.
The follow-up intervals lasted between 14 days and 12 weeks. Five out of these six studies reported persistent fatigue in 39-72% of study participants [7,[9][10][11][12]. Breathlessness or shortness of breath was reported by four out of six, ranging from 39 to 74% [7,8,10,11]. In five studies, a reduction in life quality or general health status was observed [7,8,[10][11][12]. In one study, 31% of formerly employed participants have not returned to work at approximatively 72 days post-discharge [9]. Another study reported functional restrictions (e.g. changes in life-style, sports and social activities) in 80% of participants, albeit 63% thereof was considered negligible [13].

Cardiovascular and hematological system
References classified into the cardiovascular and hematological category included two prospective cohort studies, one cross-sectional study and one case report [25][26][27][28].
Time to follow-up examination varied from 11 days after quarantine to approximatively three months post-diagnosis. Three studies performed cardiac magnetic resonance imaging and observed evidence of peri-, myoperi-and myocarditis in 3-26% of the participants [25][26][27]. One study showed altered immune cell counts 9-11 weeks after symptom onset [25], while a case report described isolated thrombocytopenia in a patient approximatively one month after symptom onset [28].

Neurological system and mental health
Two studies, a prospective cohort study and a case report, described neurological complications [29,30] while one prospective and one retrospective cohort study investigated psychiatric sequelae [31,32]. In one study, a MRI follow-up after three months showed disruptions to microstructural and functional brain integrity and 55% of participants complained about persistent neurological symptoms [29]. A case report showed persisting lower limb weakness one month after symptom onset [30]. An increased incidence of 5.8% of newly diagnosed psychiatric diseases 14-90 days after diagnosis of COVID-19 infection was shown in a large cohort study in comparison to a control group where 2.5-3.4% of participants received a new psychiatric diagnosis [31]. The other reference described a prevalence of at least one psychopathological dimension approximately one month post-discharge or after evaluation at the emergency department in 56% of study participants [32].

Otorhinolaryngological system
A total of three prospective cohort studies examined olfactory and gustatory dysfunction after COVID-19 [33][34][35]. The follow-up intervals ranged from 12 to 39 days after symptom onset. There was a lack of recovery of gustatory and/or olfactory function in 3-13.9% of study participants. Incomplete recovery was observed in 33.6-36% of participants at follow-up.

Endocrinological system
There were two case reports describing subacute thyroiditis (SAT) 15 days and six weeks after symptom onset, respectively [36,37]. In both cases there was a clinical and laboratory resolution of SAT within two weeks of adequate therapy.

Discussion
There is an ever-expanding body of knowledge regarding the acute effects of a SARS-CoV-2 infection on several organs. So far, it is known that SARS-CoV-2 can infect lung, heart, liver and kidney tissue, gastrointestinal mucosa, vascular endothelium, macrophages, T-lymphocytes and neurons [38]. In addition to the direct damage through the viral infection, in some patients the infection leads to a massive release of cytokines, a cytokine storm, which indirectly results in extensive tissue damage [39]. As a consequence, in every affected organ, directly or

Respiratory System
(continued on next page)   indirectly, there is the possibility of persistent damage with specific sequelae.
In this systematic review we provide an evaluation of data concerning COVID-19 intermediate and long-term sequelae published to date. This paper focuses on sequelae in previously healthy individuals. As COVID-19 is a relatively new disease entity the follow-up time in studies identified in our systematic search does not exceed 12 weeks. It is currently not possible to predict the longer-term impact of several months to years. It is conceivable that a lot of outcomes seen in the more short-term frame of two to six weeks are of transient nature. The WHO proposes a recovery time of up to two weeks for mild and up to six weeks for severe disease courses [40].
Our review, however, shows that a significant proportion of formerly healthy adults aged 18-50 years is affected by COVID-19 sequelae for longer periods. As they make up the majority of the working population, there will be longer term consequences and burdens not only for the health system but also for the economy.

General health
Regarding general health, there seems to be mounting evidence that a considerable number of COVID-19 convalescents will suffer from fatigue even weeks after the acute infection has subsided. Post-infectious fatigue months and even years after recovery has been previously observed in a variety of viral infections. There is evidence of chronic fatigue as a long-term consequence, especially in people younger than 30 years old, after epidemics of influenza A(H1N1) virus, SARS-CoV, Ebolavirus and West Nile virus. In some of these cases, the diagnostic criteria for Chronic Fatigue Syndrome were fulfilled. A miscommunication in the inflammatory response pathways, especially cytokine networks, might be the underlying cause [41]. It has been hypothesized that a COVID-19 infection might lead to a higher predisposition for a number of different cancer types. Hays et al. proposed an increased tumorigenesis through the activation of the MAPK and JAK-STAT pathway upon a COVID-19 infection and the weakened immune system following a cytokine storm [39].

Respiratory system
As seen in the results section, abnormalities in pulmonary function were observed in the lung function examination (e.g. as decrease in aerobic capacity or reduction in diffusion capacity) as well as radiologically. Even though a majority of patients recovered fully or were in the process of radiological and/or clinical resolution, it would not be surprising if lung damage in the form of residual pulmonary fibrosis persisted in the long-term. This would also be a plausible explanation as to why some patients experienced shortness of breath even at up to 12 weeks after hospital admission [10]. Again, there is evidence from the SARS and MERS pandemic that some patients experienced lung damage up to 15 years later [42].

Cardiovascular and hematological system
According to the studies found in this review, cardiac damage might be a significant long-term sequela. Myocarditis, perimyocarditis and pericarditis were diagnosed as late as 11 weeks after symptom of infection onset [25]. Rajpal et al. showed that cardiac consequences were also pronounced in younger athletes [27]. This is especially concerning as myocarditis can lead to sudden cardiac death. Even with apparently recovered cardiac function, there might still be a risk of coronary artery disease, atrial fibrillation or ventricular arrhythmias as a consequence of myocardial injury [43].

Neurological system and mental health
To date, there are few original data on neurological sequelae.
Radiologically, there seem to be distinct changes in cerebral microstructure, especially in those areas related to memory and smell loss, compared to healthy controls [29]. However, there is widespread speculation concerning neurological long-term consequences [38,[44][45][46]. Several authors speculate that SARS-CoV-2 might trigger neurodegenerative diseases like multiple sclerosis, Parkinson's disease and narcolepsy in predisposed individuals in the long-term [44][45][46]. The occurrence of anosmia and ageusia in some patients has even been associated with prodromal features of Parkinson's disease [44].
With over 44 ′ 000 study subjects, Taquet et al. performed the most extensive study in this review in terms of numbers [31]. They observed a significant increase in newly diagnosed psychiatric diseases like anxiety, depression, insomnia and dementia compared to an unaffected control group. Likewise, Mazza et al. found indications of at least one psychiatric disease (post-traumatic stress disorder, depression, anxiety, obsessive-compulsive disorder, insomnia) in 56% of study participants approximatively 30 days post-discharge. Extrapolated from experiences with SARS and MERS, where the prevalence of depression, anxiety and post-traumatic stress disorder was high even after 39 months, a considerable impact on mental health post COVID-19 should be anticipated [47].

Otorhinolaryngological system
All three included studies investigating olfactory and gustatory dysfunction found a lack of or incomplete recovery in a substantial part of participants. However, they have a rather short follow-up period of only two to five weeks. Imam et al. believe that COVID-19 related olfactory dysfunction follows a similar mechanism as the post-viral olfactory dysfunction seen with other viral infections such as influenza virus, rhinovirus, metapneumovirus or parainfluenza virus. So these sequelae may reduce over longer time periods [48].

Endocrinological and reproductive system
Concerning the endocrinological system, only two case reports were found of whom both described a case of subacute thyroiditis following a COVID-19 infection. As this constitutes only sparse evidence, a statement regarding the scope of long-term sequelae in this organ system cannot be made. Nevertheless, Mongioì et al. hypothesize a role of SARS-CoV-2 in pancreatic damage and subsequent development of diabetes, in hypothalamic-pituitary-adrenal axis dysfunction and adrenal insufficiency and in hypothalamic-pituitary-thyroid axis dysfunction with thyroid damage, as seen in SARS-CoV infection [49]. As ACE2 is highly expressed by the human testis, an infection might lead to testicular damage and associated male infertility [49,50].

Strengths and limitations
To our knowledge, this is the first systematic review on long-term sequelae of COVID-19. The strengths of our systematic review include the careful and thorough searches according to PRISMA guidelines and the thorough extraction and interpretation of selected papers. A further strength is the quality assessment of articles using the modified Newcastle-Ottawa Scale as well as the pyramid of evidence. A limitation of our work, common to all research on COVID-19 consequences, is the limited time frame to assess sequelae as the interval since the beginning of the COVID-19 pandemic and this systematic review is just seven months, allowing scant time for the evolution of longer-term implications. Due to the fact that there is a limited number of references as of yet, we did not investigate geographic differences in type of COVID-19 sequelae.

Conclusions
Even though our current knowledge suggests that most COVID-19 sequelae in young, previously healthy adults, are of transient nature, there are indications of multi-organ impact. The large global burden of cases worldwide suggests that we will most likely face an ongoing wave of COVID-19 sequelae. Close attention should be paid to residual impairments in multi-organ function, especially persistent reduced lung function and carditis, and to mental health and neurological sequelae including post-viral fatigue syndrome. We suggest that further research should include lung function tests and sensitive test batteries to detect long-lasting structural and functional damage to the cardiopulmonary and the neurological system. Patients suffering from post-viral fatigue syndrome and mental health impairments could be followed-up routinely with questionnaires to monitor disease course. In general, study test batteries to follow up on sequelae should be carefully designed to detect subtle, long-term sequelae.

Funding
NB and KS thank the Mäxi Foundation (Zurich) for financial support. PS is supported in part by MilMedBiol, a research collaboration between the Swiss Armed Forces and the University of Zurich.

Declaration of competing interest
All authors declare no conflict of interest.