Hostname: page-component-cc8bf7c57-77pjf Total loading time: 0 Render date: 2024-12-10T03:01:04.177Z Has data issue: false hasContentIssue false

Epigenetic changes in patients with post-acute COVID-19 symptoms (PACS) and long-COVID: A systematic review

Published online by Cambridge University Press:  22 October 2024

Madhura Shekhar Patil
Affiliation:
Centre for Environment and Health, Department of Public Health and Primary Care, KU Leuven, Leuven, Belgium
Emma Richter
Affiliation:
Centre for Environment and Health, Department of Public Health and Primary Care, KU Leuven, Leuven, Belgium
Lara Fanning
Affiliation:
Centre for Environment and Health, Department of Public Health and Primary Care, KU Leuven, Leuven, Belgium
Jolien Hendrix
Affiliation:
Centre for Environment and Health, Department of Public Health and Primary Care, KU Leuven, Leuven, Belgium Pain in Motion (PAIN) Research Group, Department of Physiotherapy, Human Physiology, and Anatomy, Vrije Universiteit Brussel, Brussels, Belgium Research Foundation, Flanders (FWO)
Arne Wyns
Affiliation:
Pain in Motion (PAIN) Research Group, Department of Physiotherapy, Human Physiology, and Anatomy, Vrije Universiteit Brussel, Brussels, Belgium
Laura Barrero Santiago
Affiliation:
Department of Cell Biology, Genetics, Pharmacology and Histology – University of Valladolid, Spain
Jo Nijs
Affiliation:
Pain in Motion (PAIN) Research Group, Department of Physiotherapy, Human Physiology, and Anatomy, Vrije Universiteit Brussel, Brussels, Belgium Department of Health and Rehabilitation, Institute of Neuroscience and Physiology, University of Gothenburg, Gothenburg, Sweden
Lode Godderis
Affiliation:
Centre for Environment and Health, Department of Public Health and Primary Care, KU Leuven, Leuven, Belgium External Service for Prevention and Protection at Work (IDEWE), Leuven, Belgium
Andrea Polli*
Affiliation:
Centre for Environment and Health, Department of Public Health and Primary Care, KU Leuven, Leuven, Belgium Pain in Motion (PAIN) Research Group, Department of Physiotherapy, Human Physiology, and Anatomy, Vrije Universiteit Brussel, Brussels, Belgium Research Foundation, Flanders (FWO)
*
Corresponding author: Andrea Polli; Email: andrea.polli@kuleuven.be
Rights & Permissions [Opens in a new window]

Abstract

Background

Up to 30% of people infected with SARS-CoV-2 report disabling symptoms 2 years after the infection. Over 100 persistent symptoms have been associated with Post-Acute COVID-19 Symptoms (PACS) and/or long-COVID, showing a significant clinical heterogeneity. To develop effective, patient-targeted treatment, a better understanding of underlying mechanisms is needed. Epigenetics has helped elucidating the pathophysiology of several health conditions and it might help unravelling inter-individual differences in patients with PACS and long-COVID. As accumulating research is exploring epigenetic mechanisms in PACS and long-COVID, we systematically summarized the available literature on the topic.

Methods

We interrogated five databases (Medline, Embase, Web of Science, Scopus and medXriv/bioXriv) and followed PRISMA and SWiM guidelines to report our results.

Results

Eight studies were included in our review. Six studies explored DNA methylation in PACS and/or long-COVID, while two studies explored miRNA expression in long-COVID associated with lung complications. Sample sizes were mostly small and study quality was low or fair. The main limitation of the included studies was a poor characterization of the patient population that made a homogeneous synthesis of the literature challenging. However, studies on DNA methylation showed that mechanisms related to the immune and the autonomic nervous system, and cell metabolism might be implicated in the pathophysiology of PACS and long-COVID.

Conclusion

Epigenetic changes might help elucidating PACS and long-COVID underlying mechanisms, aid subgrouping, and point towards tailored treatments. Preliminary evidence is promising but scarce. Biological and epigenetic research on long-COVID will benefit millions of people suffering from long-COVID and has the potential to be transferable and benefit other conditions as well, such as Myalgic Encephalomyelitis/Chronic Fatigue Syndrome (ME/CFS). We urge future research to employ longitudinal designs and provide a better characterization of included patients.

Type
Review
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
Copyright © The Author(s), 2024. Published by Cambridge University Press

Introduction

COVID-19 is caused by Severe Acute Respiratory Syndrome Coronavirus 2 (SARS-CoV-2) (Ref. Reference Seyed Hosseini1). Most individuals infected with SARS-CoV-2 suffer from a mild disease but some patients develop moderate-to-severe disease and require hospitalization (Refs Reference Lundgren2, Reference Zhang3). Though the majority of cases recover within the first 2-3 weeks, accumulating clinical observations show that a significant proportion of patients with COVID-19 develop persistent symptoms (Refs Reference Lundgren2, Reference Zhang3, Reference Wahlgren4). Up to 30% of people infected with SARS-CoV-2 during the first wave reported disabling symptoms, affecting everyday activities, 2 years after the infection (Ref. Reference Wahlgren4). Over 100 persistent symptoms have been associated with COVID-19, which led to coining the term post-acute COVID-19 symptoms (PACS) and long-COVID. PACS refers to symptoms persisting over 4 weeks after infection, while long-COVID is used when symptoms persist for 12 weeks or longer, and cannot be explained by alternative diagnosis (Ref. Reference Soriano5). The most common symptoms are fatigue, post-exertional symptom exacerbation (PESE), pain, and cognitive problems – e.g. concentration difficulties and brain fog (Refs Reference Taquet6, Reference Frontera7). Long-COVID represents a huge burden for these patients and a significant cost for healthcare systems worldwide (Refs Reference Tene8, Reference Sykes9, Reference Crook10).

One crucial advancement to the understanding of long-COVID is the observation that different subgroups likely exist. Subgroups are probably not linked to the characteristics of the virus or its variants, but rather to individual characteristics. While some factors such as obesity, depression and the presence of comorbidities are risk factors for both acute COVID-19 and long-COVID (Refs Reference Dessie and Zewotir11, Reference Fernandez-de-Las-Penas12, Reference Fernandez-de-Las-Penas13, Reference Subramanian14) others are specific to either the acute or chronic phase. Severity and mortality in acute COVID-19 was associated with older age and male sex (Ref. Reference Dessie and Zewotir11). On the contrary, risk factors for long-COVID include female sex and younger age, suggesting that gender differences can be very relevant (Refs Reference Fernandez-de-Las-Penas12, Reference Fernandez-de-Las-Penas13, Reference Subramanian14). Importantly, long-COVID can develop after very mild symptoms at onset (Ref. Reference van Kessel15). A study on 545 patients found that over 70% of people with persistent symptoms started with mild or moderate initial infection (Ref. Reference Fischer16). This suggests that mechanisms of long-COVID might be different from the ones in the acute phase. Findings from research on COVID-19 should not be automatically generalized to long-COVID, and specific research is needed.

The acute phase of COVID-19 has been extensively studied, in the quest for mechanisms that can then be targeted by specific treatments. Hyperinflammation and coagulopathy as a result of the dysregulated immune response play a role in the severity of the disease (Ref. Reference Thompson17). Epigenetics is the study of phenotype changes without changes in the DNA sequence and includes alterations in DNA methylation, histone modifications, chromatin reorganization, miRNAs and lncRNAs (Ref. Reference Polli18). It has been shown to be involved in the pathophysiology of many diseases and led to breakthrough findings in understanding a treatment of neurological disorders and cancer (Ref. Reference Berdasco and Esteller19). Epigenetic modifications have been linked to the aberrant immune response and have been suggested to predict disease severity, symptoms persistence and poor prognosis (Ref. Reference AbdelHamid20). For instance, Genome-wide DNA methylation in peripheral blood mononuclear cells from severe COVID-19 patients showed hypermethylation of interferon-related genes and hypomethylation of cytokine genes, suggesting a potential role for sustained inflammation (Ref. Reference Corley21). Other epigenome-wide DNA methylation association studies of peripheral blood samples of COVID-19 patients revealed an association between the increased severity of COVID-19 and altered DNA methylation at genes involved in inflammasome-related pathways, major histocompatibility factor, and HLA-C involved in the interferon-response pathway (Ref. Reference Castro de Moura22). Similarly, non-coding RNA interference exerted by long non-coding RNAs (lncRNAs) or micro RNAs (miRNAs) (another set of mechanisms responsible for post-transcriptional regulation of gene expression) also show changes during COVID-19 infection (Refs Reference Roustai Geraylow23, Reference Zhang24). Hundreds of lncRNAs and miRNAs show significant differential expression which in turn predict severity of the infection and/or mortality (Refs Reference Srivastava25, Reference Fernandez-Pato26, Reference Wu27). Regulation of miRNAs and lncRNAs seem to help to relieve acute symptoms via downregulation of pro-inflammatory cytokines (Refs Reference Sebastian-delaCruz28, Reference O'Brien29). Taken together, the aforementioned observations show clear changes in epigenetic signatures of patients with acute COVID-19.

These findings suggest that epigenetic biomarkers might help unravelling inter-individual differences, disease presentation and prognosis in patients with PACS and long-COVID. Research on epigenetic changes and non-coding RNA interference is scarce in long-COVID, but preliminary observations are accumulating. These observations suggest that long-COVID can be explained by an intricate set of epigenetic mechanisms. For all these reasons, a systematic summary of the current evidence is warranted, but is currently unavailable and represents an important knowledge gap. Here we report a systematic review aiming to summarize the available literature exploring epigenetic changes associated with PACS or long-COVID.

Materials and methods

The protocol for this systematic review was registered on PROSPERO (registration ID: CRD42023393690) and the review was conducted in accordance with the 2020 Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) guidelines. The search was performed on Medline, Embase, Scopus, Web of Science and medRxiv on 30th December 2022. The terms used in the search strings were: ‘Post covid symptom*’, ‘post covid-19 syndrome’, ‘long-COVID’, ‘post-acute covid-19 sequelae’, ‘COVID-19’, ‘SARS-CoV-2’, ‘severe acute respiratory syndrome coronavirus 2’, ‘coronavirus disease 2019’, ‘chronic covid’, ‘post-acute covid syndrome’, ‘post-acute sequelae of SARS-CoV-2 infection’, ‘covid long haulers’, ‘acute COVID-19’, ‘post-COVID conditions’, ‘persistent-covid’, ‘covid complications’, ‘epigenetic’, ‘epigenome’, ‘epigenome-wide association study’, ‘epigenetic modifications’, ‘epigenetic repression’, ‘epigenetic clock’, ‘epigenetic regulation’, ‘miRNA’, ‘microRNA*’, ‘small interfering RNA*’, ‘micro interfering RNA*’, ‘siRNA’, ‘long noncoding RNA*’, ‘noncoding RNA’, ‘lncRNA’, ‘histone modifications’, ‘histone methylation’, ‘histone acetylation’, ‘histone*’, ‘acetylati*’, ‘chromatin remodelling’, ‘chromatin dynamics’, ‘chromatin*’, ‘chromatin reorganization’, ‘nucleosome remodelling’, ‘global methylation’, ‘DNA methylation’, ‘gene methylation’, ‘methylati*’.

The search strings for all five databases and the number of results obtained can be found in the supplementary material (see Table S1 in Supplementary Material). The searches on Medline and Embase were limited to human studies only. No such limit was applied for searches on Web of Science and Scopus. The same keywords were combined and entered in pre-print platform such as medXriv and bioXriv. No publication date and language restrictions were used. We only included articles published in English, Italian and Spanish. Cohort studies, case–controls studies and cross-sectional studies were included. Studies were not selected if they did not include a control group. Only human studies assessing epigenetic biomarkers such as chromatin accessibility, histone modifications, DNA methylation, and miRNA and lncRNA expression in patients with PACS or long-COVID were included. PACS included patients experiencing symptoms persisting more than three weeks from onset; long-COVID included patients experiencing symptoms for at least 12 weeks from onset. Studies assessing epigenetic biomarkers only in acute COVID-19 patients were excluded. In-vitro observations and preclinical studies were also excluded.

Study selection

After removing the duplicates, the initial screening of studies was done based on titles and abstracts by two reviewers independently (MP and ER). This was done using the Rayyan platform. Any disagreement between the two was resolved by a third reviewer (AP). Afterwards, the full text of the relevant studies was checked for eligibility by the two reviewers independently (MP and ER). In case some of the studies did not report potentially relevant information (e.g. a precise characterization of clinical symptoms reported by patients), we tried to contact the corresponding authors and asked for additional information.

Data extraction

Studies meeting inclusion criteria were divided between the two reviewers (MP and ER) for independent data extraction, after which both authors checked the extracted data of all the studies and any discrepancies were resolved through discussion. The following data were extracted: first author's name along with the year of publication, study design, assessed epigenetic biomarkers, blood fraction assessed (cells, serum or plasma), type of participants population (PACS or long-COVID), sample characteristics (patients and controls characteristics, sample size, hospitalization status), findings pertaining to the epigenetics of patients and the quality of the study. The following epigenetic outcomes were considered: chromatin accessibility, DNA methylation, miRNA and lncRNA profiling, and histone modifications. For each outcome, studies enrolling patients with either Post-Acute COVID-19 or long-COVID were grouped together.

Quality assessment

The quality assessment for all the studies was done independently by the two authors (MP and ER) and disagreements were resolved through discussion. Methodological quality was assessed using the NIH study quality assessment tools (https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools). We included three additional questions to the ones present in the tool, to include questions regarding confounding variables particularly important for the study of epigenetic biomarkers in the target population and a few other questions regarding the reporting of methods employed. The final criterion for different study designs can be found in the supplementary material (see Table S2). The quality assessment of each study can be found in the supplementary material. An overall rating was given to each study depending on the number of questions with an affirmative answer. Studies with affirmative answers for ⩾ 75% of questions were labelled as good, those with 50–75% were labelled as fair and those with less than 50% were labelled as poor.

Results

The initial search on all five databases yielded 7500 potentially relevant studies. Following the removal of duplicates, 3435 studies were screened by title and abstract. After a first screening based on the title and abstract, we excluded 3300 results and screened the full text of 135 studies for inclusion. In total, eight case–control studies matched our inclusion criteria and were included in the final systematic review (Refs Reference Mongelli30, Reference Huoman31, Reference Yin32, Reference Lee33, Reference Nikesjo34, Reference Balnis35, Reference Ali36, Reference García-Hidalgo37). The flow diagram of the screening and selection procedure can be found in Figure 1.

Figure 1. Prisma flowchart of the systematic review.

Characteristics of the included studies

The included articles explored in total 455 patients and 380 controls. Participants were either individuals infected with SARS-CoV-2 who then recovered or healthy controls, with no evidence of infection. The smallest study included 14 participants (Ref. Reference Yin32), while the largest study included 261 participants (Ref. Reference Mongelli30). Five out of eight studies were of fair quality (Refs Reference Huoman31, Reference Lee33, Reference Nikesjo34, Reference Balnis35, Reference García-Hidalgo37) while three studies were of poor quality (Refs Reference Mongelli30, Reference Yin32, Reference Ali36). Studied populations were rather heterogeneous. Two studies included participants who were infected at least four weeks before (PACS). The other six studies included participants who were infected three months or more before (long-COVID). In two of these six, persistent symptoms were related to an acute respiratory distress syndrome (ARDS). Two other studies included both subjects with and without evidence of lung complications, in the patient group. No studies explored chromatin accessibility or histone modifications. Six studies explored DNA methylation, (Refs Reference Mongelli30, Reference Huoman31, Reference Yin32, Reference Lee33, Reference Nikesjo34, Reference Balnis35) while two studies measured circulating miRNAs, both of them in ARDS-related symptoms (Refs Reference Ali36, Reference García-Hidalgo37). Figure 2 summarizes the overall quality of the present systematic review. For details on the risk of bias assessment for each study, information can be found in Tables S3 (see Supplementary Material).

Figure 2. Overall quality of the present systematic review. Risk of bias assessed according to the NIH Risk of bias tools (see supplementary material).

Synthesis of the evidence

Given the nature of the data extracted, the heterogeneity of the included samples, the methodological differences, and the different outcome measures, quantitative synthesis using meta-analysis was not possible. Hence, guidelines for synthesis without meta-analysis (SWiM) (Ref. Reference Campbell38) were used to report results in narrative form as an alternative to meta-analysis.

The results are reported in Table 1. Table 1 shows the characteristics of the included studies, and specifies the population included, the findings pertaining to the epigenetics of patients, and the overall study quality, with a comment describing the most relevant criticism of the study.

Table 1. Result summary of the systematic review

DNA methylation

In total, six studies focussed on DNA methylation changes in people with persistent symptoms after COVID-19. Three of them enrolled patients with a combination of PACS (<12 weeks post-infection) and long-COVID (>12 weeks post infection). Mongelli and colleagues enrolled a group of patients who reported symptoms for at least 4 weeks after infection. They measured DeltaAge – a marker of biological age acceleration based on DNA methylation of specific targeted genes (Ref. Reference Bekaert39), telomere length and ACE2 expression in peripheral blood. They found evidence for age acceleration, especially in younger individuals, significant telomere shortening and reduced ACE2 expression (Ref. Reference Mongelli30). Huoman et al. (2021) (Ref. Reference Huoman31) used a genome-wide approach and identified an overrepresentation of the signalling pathways of Wnt, muscarinic and acetylcholine receptors, and the gonadotropin-releasing hormone pathway (Ref. Reference Huoman31). Yin and colleagues discovered DMRs in transposable elements and genes related to immune and stress responses indicating mild activation of TEs and incomplete recovery of immune-related genes (Ref. Reference Yin32).

The study by Lee and collegues (Ref. Reference Lee33) started from a large cohort and performed multiple comparison. One comparison is between patients with PACS (8–12 weeks post infection, with evidence of positive RT-PCR, n = 110), and controls (n = 73). The control group included both healthy controls and symptomatic patients, who all had a negative RT-PCR and no evidence of COVID-19 persistence. They measured DNA methylation using a genome-wide approach, and epigenetic age acceleration (EAA), according to a published algorithm (Ref. Reference Horvath40). They found three hypomethylated CpGs located in IFI44L and ANKRD9 genes, that are involved in innate immune activation and Class I MHC-mediated antigen processing and presentation, respectively. However, contrary to the findings by Mongelli and colleagues (Ref. Reference Mongelli30), they did not find any evidence of age acceleration (Ref. Reference Lee33). Two main differences should be reported between the two studies. First, Mongelli et al. included patients at an earlier stage (from 4 weeks on), while Lee et al. included patients with persistent symptoms of at least 8 weeks. Second, some of the patients included in Mongelli et al.'s study showed lung involvement. On the contrary, people with respiratory tract infection – not due to COVID-19, were in the control group in Lee's study.

Focussing on long-COVID, Lee et al. also compared patients who were symptomatic at least 3 months after infection with participants who were infected but recovered. No between-group differences were found. Of note, patients were assigned to the long-COVID or control group based on one question only of the SF-36 questionnaires, namely Self-reported health somewhat or much worse than a year-ago.

The other two studies measuring DNA methylation enrolled patients with long-COVID (>12 weeks after infection). Nikesiö et al. (2022) (Ref. Reference Nikesjo34) found significant differences in DNA methylation in three genes: SNORD3B, CETP and DLGAP1. Such genes are involved in several biological pathways, including ‘Angiotensin II-stimulated signalling through G proteins and β-arrestin’, ‘Histamine H1 receptor mediated signalling pathway’, ‘Heterotrimeric G-protein signalling pathway-Gq alpha and Go alpha mediated pathway’, ‘PI3 kinase pathway’ and ‘Muscarinic acetylcholine receptor 1 and 3 (CHRM1&3) signalling pathway’. Balnis and colleagues found 71 DMRs in immune and inflammation-related pathways in long-COVID individuals one year after infection (Ref. Reference Balnis35).

miRNA expression

Two studies explored miRNA expression and focused on lung and pulmonary abnormalities in long-COVID. One of them found increased levels of miR-21 and gene transcripts such as TGF-β, Col1A2, Col3A1 and α-SMA involved in profibrotic pathways (Ref. Reference Ali36). miR-21 had high diagnostic value, as it was able to differentiate between patients and controls >72% sensitivity and >80% specificity. The second one discovered distinct miRNA profiles associated with radiologic features (miR-9, miR-24, miR-221, miR-21) and low CO diffusion (miR-17, miR-27a, miR-126, miR-495, miR-146a) capacity in ARDS patients (Ref. Reference García-Hidalgo37).

Discussion

To our knowledge, the present systematic review is the first attempt to summarize findings on epigenetic changes associated with PACS and long-COVID. The first and most important finding is lack of homogeneity in the patient populations included. The World Health Organisation specifies that long-COVID occurs when symptoms persist for over three months from initial infection and an alternative explanation for those symptoms cannot be found (Ref. Reference Soriano5). However, other groups describe PACS and long-COVID as a multi-organ disease, where evidence of pulmonary, cardiovascular or nervous system alterations are considered part of the disease (Ref. Reference Crook10). This is important, as some of the included studies included patients with evidence of lung involvement such as ARDS (Refs Reference Ali36, Reference García-Hidalgo37). Though ARDS was a consequence of Sars-CoV-2 infection and persisted for over 12 weeks, it represents a possible explanation for symptoms. Therefore, in such cases the diagnosis of PACS or long-COVID should not be made. In other cases, authors included individuals with and without evidence of pneumonia and/or upper respiratory tract involvement in the same patient group (Refs Reference Mongelli30, Reference Yin32). In these cases, drawing a valid conclusion becomes difficult. A precise reporting of the clinical presentation is crucial – different mechanisms can be at play in different clinical presentations. We urge future research to apply strict inclusion criteria and provide a comprehensive clinical characterization of participants. This would minimize heterogeneity and provide a tailored approach based on mechanism-based subgroups.

Similarly, controls should be precisely defined. In some studies, controls were people who got infected by the virus, but then recovered. In others, controls were healthy participants with no evidence of SARS-CoV-2 infection, or recruited before the COVID-19 pandemic. We argue that the latter control groups are less relevant for studies investigating the development of long-COVID.

The second relevant result is that research focussing on the epigenetics of PACS and long-COVID is scarce and mostly of low quality. Sample sizes are generally quite small, and the clinical description of the included populations is rather superficial. In addition, different studies used different biological matrices and/or methods to explore the same mechanism (e.g. Illumina array-based assays vs. whole-genome bisulphite sequencing for DNA methylation). In fact, different studies showed different results, and firm conclusions cannot be drawn on the exact underlying epigenetic mechanisms.

However, the present synthesis of the literature does point towards some epigenetic mechanisms that can be relevant in PACS and long-COVID. With the exception of Mongelli et al (Ref. Reference Mongelli30), which used pyrosequencing in targeted genes to assess biological age of the participants, the other five studies measuring DNA methylation used array-based technology to study whole-genome DNA methylation. Whole-genome DNA methylation allows for a broad, explorative analyses of the DNA methylome. They require high-level bioinformatic analyses and the results depend at least partially by the algorithm employed for analyses. In addition, the sample size needed to detect small–medium effect sizes is normally very large, and it is not the case for the studies included in the present systematic review. The studies included here were able to only detect changes with large effect size, at best. Precise replication between different studies is unlikely in this scenario. However, the included studies provide some results that are in line with one another. All studies identified significant differential DNA methylation patterns between patients and controls in several genes. The exact patterns differ from study to study but they can be summarized as related to the immune system (IFI44L and ANKRD9 genes, histamine and T-cell activation pathways, etc.), the autonomic nervous system (muscarinic acetylcholine receptor and angiotensin II signalling), and cell metabolism (WNT pathway, DLGAP1 and PI3 kinases, ERK signalling and mitochondrial function). Other changes in DNA methylation are less related to specific mechanisms, but rather represent widespread changes in the epigenome. These are changes in biological age and differential methylation in LINE-1 regions. Finally, patients with PACS might show evidence of biological age acceleration, but this seems be more relevant in PACS associated with pulmonary involvement (Ref. Reference Mongelli30), rather than long-COVID (Ref. Reference Lee33). Taken together, these findings are promising but not conclusive and more research should attempt to replicate them.

One last finding is that individuals with long-COVID due to lung abnormalities show upregulation of miRNAs involved in fibrotic, cell death, vascularization, thrombosis and inflammation-related pathways. ARDS increases pulmonary vascular permeability, inflammation and loss of aerated lung tissue. Dysregulated inflammation along with coagulation and profibrotic pathway activation has been implicated in the pathophysiology of ARDS (Ref. Reference Grasselli41). Both studies that investigated long-COVID-associated lung injury assessed miRNA expression (Refs Reference Ali36, Reference García-Hidalgo37). Ali and colleagues focused specifically on circulating levels of miR-21 – a miRNA involved in fibroblast activation and lung fibrosis (Ref. Reference Liu42) and found it to be significantly increased. Their observation was also corroborated by the study by Garcia-Hidalgo and colleagues in their sample. Though they employed a more explorative approach, including 41 miRNAs in their analyses, they found that higher levels of miR-21 were associated with lung function.

Such findings point towards the importance of patient subgrouping and stratification. When studies focus on homogenous subgroups (in this case patients with long-COVID associated with lung abnormalities), results can be replicated and inform future research. Over 100 symptoms have been reported by patients with long-COVID, making long-COVID a very heterogenous syndrome (Ref. Reference Sykes9). Subgrouping patients based on clinical presentation or putative pathophysiological mechanisms will likely increase the power of statistical analyses and the chances to identify targetable underlying mechanisms. Observations from other populations already showed that epigenetic changes can successfully contribute to stratify patients into more homogeneous subgroups (Ref. Reference Polli43).

Lastly, we believe it is relevant to note that symptoms of most patients with long-COVID very closely resemble the ones described in another condition – myalgic encephalomyelitis/chronic fatigue syndrome (ME/CFS). ME/CFS is in fact characterized by unexplained, profound fatigue, sleep and cognitive disturbances and PESE (Ref. Reference Clayton44). Many patients suffering from ME/CFS report their symptoms to have started after a viral infection (Ref. Reference Clayton44). The similarity is striking. Research on long-COVID can learn a lot from the knowledge obtained from studies on ME/CFS over the past decades (Ref. Reference Friedman45). In turn, current research on long-COVID will benefit patients with ME/CFS (Ref. Reference White46). Given the expected huge prevalence of long-COVID (and thus the possibility to recruit patients for research purposes), results from studies on patients with long-COVID should be widely transferred to the ME/CFS field (Ref. Reference White46).

The findings of this review should be considered in light of its limitations. As previously mentioned, sample sizes were often small and study quality was generally low. Heterogeneity in methods and sample characteristics exist among the included studies. The timing of the follow-up and assessment is also crucial, especially for studying DNA methylation, which is time-dependent and decreases with age. Individual variabilities in disease presentation, the stage of recovery, and variability in the follow-up measurements make it hard to understand pathophysiological mechanisms and their association to disease progression.

In addition, there are various important covariates that might affect both the disease trajectory and epigenetic mechanisms which are not always controlled for in the studies. Age, sex, medication used, comorbidities such as diabetes, hypertension, smoking, obesity, etc. may alter epigenetics, especially DNA methylation and should be controlled for (Refs Reference Samblas, Milagro and Martinez47, Reference Michels and Binder48, Reference Lee and Pausova49). Moreover, the treatment patients received during their acute infection might play a role in deciding the trajectory of these patients and must be considered during statistical analysis (Ref. Reference Niknam50). Finally, all included studies were case–control studies, and cross-sectional in nature. While such designs can provide relevant information, longitudinal studies are better at characterizing the dynamics of epigenetics and should be preferred.

Conclusion

Epigenetic changes might help the understanding of PACS and long-COVID, elucidate their underlying mechanisms, aid subgrouping and point towards tailored treatments. Preliminary evidence is promising but scarce. We urge future research to employ longitudinal, repeated-measures designs, aim for the best possible methodological quality (following published guidelines on conducting and reporting clinical research, e.g (Ref. Reference von Elm51) or https://www.equator-network.org/), and provide a comprehensive, precise clinical characterization of the included population in order to provide valid and reliable findings. A better understanding of PACS and long-COVID is urgently needed, given the high prevalence of these conditions and the benefits that such research can transfer to other high-burden diseases such as ME/CFS.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/erm.2024.32

Acknowledgements

Not applicable.

Author contributions

M.S.P. contributed to the review process and article selection, as well as in the data synthesis and writing the first draft of the manuscript. E.R. contributed to the review process and article selection, as well as in the data synthesis. L.F. contributed to the review process and article selection. J.H. contributed to the review process and article selection. A.W. contributed to the review process and article selection. L.B.S. contributed to data synthesis. J.N. contributed to the idea and the interpretation of the results. L.G. contributed to the idea and interpretation of the results. A.P. contributed to the idea and design of the study as well as supervising the writing process. All authors contributed to revising the first draft and to the writing of the present manuscript.

Funding statement

Andrea Polli is a postdoctoral research fellow funded by the Research Foundation – Flanders (FWO). Jolien Hendrix is a PhD researcher fellow funded by the Research Foundation – Flanders (FWO). Jo Nijs holds the Berekuyl Academy Chair, funded by the European College for Lymphatic Therapy, the Netherlands. No other financial or personal relationship with organizations that could potentially be perceived as influencing the described research is present.

Competing interests

All authors have read the journal's policy on disclosure of potential conflicts of interest. The authors declare that they have no conflict of interest.

Ethical standards

Not applicable.

Consent for publication

Not applicable.

Availability of data and materials

Not applicable.

References

Seyed Hosseini, E et al. (2020) The novel coronavirus disease-2019 (COVID-19): mechanism of action, detection and recent therapeutic strategies. Virology 551, 19. doi: 10.1016/j.virol.2020.08.011CrossRefGoogle ScholarPubMed
Lundgren, J (2023) Patients hospitalized with coronavirus disease 2019: a diverse population. Clinical Infectious Diseases 77, 558559. doi: 10.1093/cid/ciad280CrossRefGoogle ScholarPubMed
Zhang, H et al. (2021) Clinical characteristics of coronavirus disease 2019 (COVID-19) in patients out of Wuhan from China: a case control study. BMC Infectious Diseases 21, 207. doi: 10.1186/s12879-021-05897-zCrossRefGoogle ScholarPubMed
Wahlgren, C et al. (2023) Two-year follow-up of patients with post-COVID-19 condition in Sweden: a prospective cohort study. Lancet Reg Health Eur 28, 100595. doi: 10.1016/j.lanepe.2023.100595CrossRefGoogle ScholarPubMed
Soriano, JB et al. (2022) A clinical case definition of post-COVID-19 condition by a Delphi consensus. The Lancet. Infectious Diseases 22, e102e107. doi: 10.1016/S1473-3099(21)00703-9CrossRefGoogle ScholarPubMed
Taquet, M et al. (2021) Incidence, co-occurrence, and evolution of long-COVID features: a 6-month retrospective cohort study of 273,618 survivors of COVID-19. PLoS Medicine 18, e1003773. doi: 10.1371/journal.pmed.1003773CrossRefGoogle ScholarPubMed
Frontera, JA et al. (2022) Post-acute sequelae of COVID-19 symptom phenotypes and therapeutic strategies: a prospective, observational study. PLoS One 17, e0275274. doi: 10.1371/journal.pone.0275274CrossRefGoogle ScholarPubMed
Tene, L et al. (2023) Risk factors, health outcomes, healthcare services utilization, and direct medical costs of patients with long COVID. International Journal of Infectious Diseases: IJID: Official Publication of the International Society for Infectious Diseases 128, 310. doi: 10.1016/j.ijid.2022.12.002Google ScholarPubMed
Sykes, DL et al. (2021) Post-COVID-19 symptom burden: what is long-COVID and how should we manage it? Lung 199, 113119. doi: 10.1007/s00408-021-00423-zCrossRefGoogle Scholar
Crook, H et al. (2021) Long COVID-mechanisms, risk factors, and management. BMJ 374, n1648. doi: 10.1136/bmj.n1648CrossRefGoogle Scholar
Dessie, ZG and Zewotir, T (2021) Mortality-related risk factors of COVID-19: a systematic review and meta-analysis of 42 studies and 423,117 patients. BMC Infectious Diseases 21, 855. doi: 10.1186/s12879-021-06536-3CrossRefGoogle ScholarPubMed
Fernandez-de-Las-Penas, C et al. (2022) Female sex is a risk factor associated with long-term post-COVID related-symptoms but not with COVID-19 symptoms: the long-COVID-EXP-CM multicenter study. Journal of Clinical Medicine 11, 413. doi: 10.3390/jcm11020413CrossRefGoogle Scholar
Fernandez-de-Las-Penas, C et al. (2022) Prevalence and risk factors of musculoskeletal pain symptoms as long-term post-COVID sequelae in hospitalized COVID-19 survivors: a multicenter study. Pain 163, e989e996. doi: 10.1097/j.pain.0000000000002564CrossRefGoogle ScholarPubMed
Subramanian, A et al. (2022) Symptoms and risk factors for long COVID in non-hospitalized adults. Nature Medicine 28, 17061714. doi: 10.1038/s41591-022-01909-wCrossRefGoogle ScholarPubMed
van Kessel, SAM et al. (2022) Post-acute and long-COVID-19 symptoms in patients with mild diseases: a systematic review. Family Practice 39, 159167. doi: 10.1093/fampra/cmab076CrossRefGoogle ScholarPubMed
Fischer, A et al. (2022) Long COVID classification: findings from a clustering analysis in the predi-COVID cohort study. International Journal of Environmental Research and Public Health 19, 16018. doi: 10.3390/ijerph192316018CrossRefGoogle ScholarPubMed
Thompson, RC et al. (2023) Molecular states during acute COVID-19 reveal distinct etiologies of long-term sequelae. Nature Medicine 29, 236246. doi: 10.1038/s41591-022-02107-4CrossRefGoogle ScholarPubMed
Polli, A et al. (2019) When environment meets genetics: a clinical review of the epigenetics of pain, psychological factors, and physical activity. Archives of Physical Medicine and Rehabilitation 100, 11531161. doi: 10.1016/j.apmr.2018.09.118CrossRefGoogle ScholarPubMed
Berdasco, M and Esteller, M (2019) Clinical epigenetics: seizing opportunities for translation. Nature Reviews Genetics 20, 109127. doi: 10.1038/s41576-018-0074-2CrossRefGoogle ScholarPubMed
AbdelHamid, SG et al. (2021) Deciphering epigenetic(s) role in modulating susceptibility to and severity of COVID-19 infection and/or outcome: a systematic rapid review. Environmental Science and Pollution Research International 28, 5420954221. doi: 10.1007/s11356-021-15588-6CrossRefGoogle ScholarPubMed
Corley, MJ et al. (2021) Genome-wide DNA methylation profiling of peripheral blood reveals an epigenetic signature associated with severe COVID-19. Journal of Leukocyte Biology 110, 2126. doi: 10.1002/JLB.5HI0720-466RCrossRefGoogle ScholarPubMed
Castro de Moura, M et al. (2021) Epigenome-wide association study of COVID-19 severity with respiratory failure. EBioMedicine 66, 103339. doi: 10.1016/j.ebiom.2021.103339CrossRefGoogle ScholarPubMed
Roustai Geraylow, K et al. (2022) miRNA expression in COVID-19. Gene Reports 28, 101641. doi: 10.1016/j.genrep.2022.101641CrossRefGoogle ScholarPubMed
Zhang, Y et al. (2022) Comparative analysis of long non-coding RNA expression and immune response in mild and severe COVID-19. Frontiers in Molecular Biosciences 9, 835590. doi: 10.3389/fmolb.2022.835590CrossRefGoogle ScholarPubMed
Srivastava, S et al. (2023) Evaluation of altered miRNA expression pattern to predict COVID-19 severity. Heliyon 9, e13388. doi: 10.1016/j.heliyon.2023.e13388CrossRefGoogle ScholarPubMed
Fernandez-Pato, A et al. (2022) Plasma miRNA profile at COVID-19 onset predicts severity status and mortality. Emerging Microbes & Infections 11, 676688. doi: 10.1080/22221751.2022.2038021CrossRefGoogle ScholarPubMed
Wu, Y et al. (2021) A study of differential circRNA and lncRNA expressions in COVID-19-infected peripheral blood. Scientific Reports 11, 7991. doi: 10.1038/s41598-021-86134-0CrossRefGoogle ScholarPubMed
Sebastian-delaCruz, M et al. (2021) The role of lncRNAs in gene expression regulation through mRNA stabilization. Non-Coding RNA 7, 3. doi: 10.3390/ncrna7010003CrossRefGoogle ScholarPubMed
O'Brien, J et al. (2018) Overview of MicroRNA biogenesis, mechanisms of actions, and circulation. Front Endocrinol (Lausanne) 9, 402. doi: 10.3389/fendo.2018.00402CrossRefGoogle ScholarPubMed
Mongelli, A et al. (2021) Evidence for biological age acceleration and telomere shortening in COVID-19 survivors. International Journal of Molecular Sciences 22, 6151. doi: 10.3390/ijms22116151CrossRefGoogle ScholarPubMed
Huoman, J et al. (2021) Epigenome-wide DNA methylation profiling of healthy COVID-19 recoverees reveals a unique signature in circulating immune cells. medRxiv. https://doi.org/10.1101/2021.07.05.21260014Google Scholar
Yin, Y et al. (2022) Transcriptome and DNA methylome analysis of peripheral blood samples reveals incomplete restoration and transposable element activation after 3-months recovery of COVID-19. Frontiers in Cell and Developmental Biology 3, 1001558. doi: 10.3389/fcell.2022.1001558CrossRefGoogle Scholar
Lee, Y et al. (2022) EWAS of post-COVID-19 patients shows methylation differences in the immune-response associated gene, IFI44L, three months after COVID-19 infection. Scientific Reports 12, 11478. doi: 10.1038/s41598-022-15467-1CrossRefGoogle ScholarPubMed
Nikesjo, F et al. (2022) Defining post-acute COVID-19 syndrome (PACS) by an epigenetic biosignature in peripheral blood mononuclear cells. Clinical Epigenetics 14, 172. doi: 10.1186/s13148-022-01398-1CrossRefGoogle ScholarPubMed
Balnis, J et al. (2022) Persistent blood DNA methylation changes one year after SARS-CoV-2 infection. Clinical Epigenetics 14, 9494. doi: 10.1186/s13148-022-01313-8CrossRefGoogle ScholarPubMed
Ali, M et al. (2022) Role of circulatory miRNA-21 and associated signaling pathways in the pathogenesis of pulmonary fibrosis among individuals recovered after COVID-19 infection. Human Gene 34, 201093. doi: 10.1016/j.humgen.2022.201093CrossRefGoogle ScholarPubMed
García-Hidalgo, MC et al. (2022) Identification of circulating microRNA profiles associated with pulmonary function and radiologic features in survivors of SARS-CoV-2-induced ARDS. Emerging Microbes & Infections 11, 15371549. doi: 10.1080/22221751.2022.2081615CrossRefGoogle ScholarPubMed
Campbell, M et al. (2020) Synthesis without meta-analysis (SWiM) in systematic reviews: reporting guideline. BMJ 368, l6890. doi: 10.1136/bmj.l6890CrossRefGoogle ScholarPubMed
Bekaert, B et al. (2015) Improved age determination of blood and teeth samples using a selected set of DNA methylation markers. Epigenetics 10, 922930. doi: 10.1080/15592294.2015.1080413CrossRefGoogle ScholarPubMed
Horvath, S (2013) DNA methylation age of human tissues and cell types. Genome Biology 14, R115. doi: 10.1186/gb-2013-14-10-r115CrossRefGoogle ScholarPubMed
Grasselli, G et al. (2020) Pathophysiology of COVID-19-associated acute respiratory distress syndrome: a multicentre prospective observational study. The Lancet. Respiratory Medicine 8, 12011208. doi: 10.1016/S2213-2600(20)30370-2CrossRefGoogle ScholarPubMed
Liu, G et al. (2010) miR-21 mediates fibrogenic activation of pulmonary fibroblasts and lung fibrosis. Journal of Experimental Medicine 207, 15891597. doi: 10.1084/jem.20100035CrossRefGoogle ScholarPubMed
Polli, A et al. (2019) Linking lifestyle factors to complex pain states: 3 reasons why understanding epigenetics may improve the delivery of patient-centered care. The Journal of Orthopaedic and Sports Physical Therapy 49, 683687. doi: 10.2519/jospt.2019.0612CrossRefGoogle ScholarPubMed
Clayton, EW (2015) Beyond myalgic encephalomyelitis/chronic fatigue syndrome: an IOM report on redefining an illness. JAMA 313, 11011102. doi: 10.1001/jama.2015.1346CrossRefGoogle ScholarPubMed
Friedman, KJ et al. (2021) Our evolving understanding of ME/CFS. Medicina (Kaunas) 57, 200. doi: 10.3390/medicina57030200CrossRefGoogle ScholarPubMed
White, P (2020) Long COVID: don't consign ME/CFS to history. Nature 587, 197. doi: 10.1038/d41586-020-03136-0CrossRefGoogle ScholarPubMed
Samblas, M, Milagro, FI and Martinez, A (2019) DNA methylation markers in obesity, metabolic syndrome, and weight loss. Epigenetics 14, 421444. doi: 10.1080/15592294.2019.1595297CrossRefGoogle ScholarPubMed
Michels, KB and Binder, AM (2018) Considerations for design and analysis of DNA methylation studies. Methods in Molecular Biology 1708, 3146. doi: 10.1007/978-1-4939-7481-8_2CrossRefGoogle ScholarPubMed
Lee, KW and Pausova, Z (2013) Cigarette smoking and DNA methylation. Frontiers in Genetics 4, 132. doi: 10.3389/fgene.2013.00132CrossRefGoogle ScholarPubMed
Niknam, Z et al. (2022) Potential therapeutic options for COVID-19: an update on current evidence. European Journal of Medical Research 27, 6. doi: 10.1186/s40001-021-00626-3CrossRefGoogle ScholarPubMed
von Elm, E et al. (2007) The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet (London, England) 370, 14531457. doi: 10.1016/S0140-6736(07)61602-XCrossRefGoogle ScholarPubMed
Figure 0

Figure 1. Prisma flowchart of the systematic review.

Figure 1

Figure 2. Overall quality of the present systematic review. Risk of bias assessed according to the NIH Risk of bias tools (see supplementary material).

Figure 2

Table 1. Result summary of the systematic review

Supplementary material: File

Shekhar Patil et al. supplementary material

Shekhar Patil et al. supplementary material
Download Shekhar Patil et al. supplementary material(File)
File 44.3 KB