Abstract
Long COVID, or post-acute sequelae of SARS-CoV-2 infection (PASC), is increasingly understood as a heterogeneous biological syndrome rather than the nonspecific residue of an acute respiratory illness. Among the mechanisms under investigation, reactivation of persistent human viruses has emerged as a particularly intriguing possibility. Epstein–Barr virus (EBV), cytomegalovirus (CMV), herpes simplex virus, human herpesviruses 6 and 7, and anelloviruses can persist within healthy individuals under continuous immune surveillance. SARS-CoV-2 infection can profoundly perturb that surveillance system. The resulting reactivation of latent or persistent viruses may, in some individuals, generate a second wave of antigenic stimulation, inflammatory signaling, endothelial dysfunction, metabolic disturbance, autonomic abnormalities and, potentially, autoimmunity.
Recent longitudinal multi-omic research has substantially strengthened this hypothesis. In a large cohort studied with viral transcriptomics, host transcriptomics, immune profiling, proteomics and metabolomics, investigators found temporally distinct reactivation of herpesviruses and anelloviruses during and after acute COVID-19. Persistent viral activity during convalescence was associated with some measures of long-COVID illness.¹ Nevertheless, association is not causation. Viral reactivation may be a consequence of severe illness or immune dysfunction rather than its cause. The critical unanswered question is whether eliminating a reactivated virus improves the clinical phenotype in which it is detected.
The emerging picture is therefore neither that “EBV causes long COVID” nor that viral reactivation is incidental. Rather, SARS-CoV-2 may disturb the equilibrium between the human host and its persistent virome, creating a subset of patients in whom secondary viral activity becomes an amplifier—or potentially a driver—of chronic disease.
Introduction: The Infection That May Not Be Alone
The conventional picture of viral infection is deceptively simple.
A virus enters the body, replicates, produces symptoms, encounters the immune system, and is eliminated. Recovery then represents the restoration of physiologic equilibrium.
That model works reasonably well for many acute respiratory infections.
It is less satisfactory for long COVID.
Months after SARS-CoV-2 infection, some patients continue to experience fatigue, post-exertional malaise, cognitive dysfunction, neuropathic symptoms, dyspnea, palpitations, gastrointestinal disturbance, sleep abnormalities and autonomic dysfunction. The syndrome can follow severe COVID-19, but it can also follow an apparently minor infection. Contemporary reviews now recognize several interacting mechanisms, including persistent SARS-CoV-2 antigen, immune dysregulation, autoimmunity, endothelial dysfunction, metabolic abnormalities, autonomic dysfunction, microbiome disturbance and reactivation of latent viruses.²,³
The viral-reactivation hypothesis introduces an important conceptual change.
Perhaps the problem is not simply that SARS-CoV-2 remains.
Perhaps SARS-CoV-2 changes the ecological conditions under which other viruses live inside us.
The human body is already host to a remarkable collection of persistent viruses. EBV, CMV, HSV and other herpesviruses can remain for decades, usually held in check by immune surveillance. Anelloviruses are extraordinarily common and can persist without producing recognizable disease.
SARS-CoV-2 may disturb this equilibrium.
The result could be a form of postviral virome dysregulation in which the original infection creates conditions allowing previously suppressed viruses to become transcriptionally or replicatively active.
The implications are considerable. If reactivation is merely a marker of immune dysfunction, antiviral therapy may accomplish little. If reactivation is a driver of disease, however, it creates an entirely different therapeutic target.
I. The Human Virome: A Permanent Biological Companion
Humans are not sterile ecosystems.
The human virome includes viruses that establish chronic or latent relationships with their hosts. Some remain almost completely silent; others periodically express viral genes without producing clinically apparent infection.
Herpesviruses are the classic examples.
Epstein–Barr virus
EBV infects most humans and establishes lifelong latency principally in B lymphocytes.³,⁴
The virus passes through several biologic states, including:
- latency;
- restricted latent gene expression;
- B-cell proliferation;
- lytic reactivation;
- productive replication;
- and renewed establishment of latency.
The latent-to-lytic transition is controlled by a complex viral transcriptional program. The immediate-early proteins BZLF1 and BRLF1 initiate a cascade that activates viral DNA replication and subsequent structural-gene expression.⁴,⁵
Cytomegalovirus
CMV establishes persistence primarily within cells of the myeloid lineage and can influence endothelial, epithelial and immune-cell biology.
Its reactivation is particularly associated with severe physiologic stress and impaired cellular immune surveillance.
Herpes simplex virus
HSV-1 and HSV-2 establish latency within sensory neurons.
That anatomical location makes HSV fundamentally different from EBV and CMV. Reactivation can involve viral transcription and axonal transport without necessarily producing substantial circulating viremia.
HHV-6 and HHV-7
These viruses establish persistent infections and can reactivate during immune perturbation.
Anelloviruses
Anelloviruses are among the most prevalent persistent viruses in humans. Their clinical significance remains uncertain, but their abundance makes them potentially important indicators of the state of the host immune system.
The important point is that viral persistence is normal biology.
The question is what happens when SARS-CoV-2 disrupts the equilibrium.
II. How SARS-CoV-2 Could Permit Reactivation
Latent viruses do not ordinarily require eradication because the immune system keeps them under control.
That surveillance depends on:
- CD8+ cytotoxic T cells;
- CD4+ T cells;
- natural-killer cells;
- antibody-mediated immunity;
- interferon signaling;
- antigen presentation;
- tissue-resident immune cells;
- and intact cellular signaling.
SARS-CoV-2 can perturb almost every component of this network.
Longitudinal immunologic studies have demonstrated persistent abnormalities in T-cell populations, inflammatory signaling and adaptive immunity after SARS-CoV-2 infection.⁶
The resulting condition is not necessarily conventional immunodeficiency.
It may instead be immune misregulation.
The immune system may simultaneously be:
- excessively inflammatory;
- inefficient at clearing particular antigens;
- metabolically stressed;
- depleted of some functional lymphocyte populations;
- enriched for exhausted or activated cells;
- and redirected toward persistent viral or tissue antigens.
This creates a paradox:
an immune system that is highly activated can nevertheless be less effective at controlling latent viruses.
That paradox is characteristic of several chronic viral and inflammatory diseases.
III. EBV: The Most Studied Candidate
Of all latent viruses, EBV has generated the strongest long-COVID literature.
The biological rationale is compelling.
EBV resides within B cells, and SARS-CoV-2 profoundly alters B-cell biology and adaptive immune responses. If immune surveillance becomes dysfunctional, EBV-infected cells may enter a more active transcriptional state.
This does not necessarily mean explosive viral replication.
There may be a spectrum:
latency → early lytic transcription → protein expression → partial replication → productive replication.
Consequently, different laboratory tests detect different stages of the process.
A positive EBV antibody test does not necessarily mean active replication.
An increase in early-antigen antibodies may suggest recent activity.
EBV DNA in blood is stronger evidence.
Viral RNA transcripts are stronger still because they demonstrate viral gene expression.
The distinction is essential when interpreting long-COVID studies.
IV. What the Clinical Evidence Shows About EBV
Several early studies found increased serologic evidence of EBV reactivation among people with persistent symptoms after COVID-19.⁷–⁹
A particularly influential investigation associated evidence of EBV reactivation with fatigue and cognitive symptoms after SARS-CoV-2 infection.⁸
More recent work has strengthened the association.
A 2025 study of patients with persistent fatigue and post-exertional malaise found EBV DNA more frequently among patients with post-COVID symptoms than among recovered controls.⁹
A 2026 study additionally reported increased interferon-γ responses to EBV peptides among patients with long-term post-COVID symptoms and examined the relationship between EBV reactivation and circulating blood microaggregates.¹⁰
These observations are intriguing but must be interpreted cautiously.
An EBV-reactivation signal could mean:
- EBV contributes directly to symptoms;
- EBV amplifies an inflammatory process initiated by SARS-CoV-2;
- long COVID causes immune abnormalities that permit EBV reactivation;
- or both EBV and long COVID reflect another underlying immunologic disturbance.
Only prospective intervention can reliably distinguish these possibilities.
V. The 2026 Multi-Omic Breakthrough
The strongest recent evidence comes from a large longitudinal investigation of viral reactivation during and after COVID-19.
Researchers analyzed more than 1,100 hospitalized patients using an unusually broad collection of measurements, including viral RNA sequencing, host transcriptomics, immune-cell profiling, cytokines, proteomics and metabolomics.¹
This matters because earlier studies frequently relied upon indirect markers such as antibody titers.
The investigators instead asked:
Which viruses are transcriptionally active, when do they become active, where are they detected, and what happens to the host simultaneously?
They found evidence of reactivation involving multiple viral families, including:
- Herpesviridae;
- Anelloviridae;
- and other persistent or chronic viruses.
Different viruses followed different temporal trajectories.
EBV tended to appear earlier.
CMV and HSV-1 demonstrated different kinetics.
Anelloviridae showed another pattern, including persistent activity into convalescence.
This finding undermines the simplistic idea that “viral reactivation” represents one biological event.
It is better understood as a dynamic alteration of the host’s viral ecosystem.
VI. A New Concept: Virome Dysregulation
The most interesting implication of the multi-omic findings may be conceptual rather than diagnostic.
Suppose the human virome normally exists in equilibrium:
latent viruses ↔ immune surveillance
SARS-CoV-2 introduces a major perturbation:
SARS-CoV-2 → immune dysregulation → loss of viral equilibrium → secondary viral activity
The secondary viruses then produce additional antigens and stimulate the immune system.
The system becomes:
SARS-CoV-2 → immune disruption → viral reactivation → inflammation → further immune disruption.
This may constitute a positive feedback loop.
The original virus does not necessarily need to remain continuously replicative.
It may only need to initiate an immunologic state that permits another virus to become active.
VII. CMV: A Potential Vascular Amplifier
CMV deserves special consideration because of its relationship to immune and endothelial physiology.
CMV can affect:
- endothelial cells;
- monocytes;
- T cells;
- NK cells;
- vascular signaling;
- and inflammatory pathways.
The 2026 longitudinal study found CMV activity with distinct kinetics from EBV.¹
This raises the possibility that CMV reactivation could contribute to a different long-COVID phenotype from EBV.
One hypothetical pathway is:
SARS-CoV-2 endothelial injury
plus
CMV reactivation
leading to
persistent endothelial activation
followed by:
- impaired vasodilation;
- altered platelet-endothelial interactions;
- inflammatory signaling;
- microvascular dysfunction;
- and impaired tissue perfusion.
The hypothesis is attractive because vascular dysfunction is increasingly recognized as one component of long COVID.²,³
But it remains a hypothesis.
VIII. HSV and the Problem of Invisible Tissue Reactivation
HSV illustrates one of the fundamental limitations of blood testing.
HSV establishes latency inside sensory neurons.
If reactivation occurs within a ganglion or neural tissue, there may be little or no detectable virus in peripheral blood.
A patient could therefore theoretically have:
neural HSV reactivation → local inflammatory signaling → neuropathic or autonomic dysfunction
without:
detectable blood viremia.
This raises an important methodological principle:
A negative blood test does not necessarily exclude tissue-specific viral reactivation.
Future research will therefore need increasingly sophisticated approaches to determine whether particular tissues contain transcriptionally active virus.
IX. Anelloviruses: Signal or Pathogen?
The findings concerning Anelloviridae are particularly provocative.
Anelloviruses are widespread and persistent. Their presence is not, by itself, evidence of disease.
Yet persistent Anelloviridae transcription during convalescence was associated in the 2026 study with measures of physical impairment and fatigue.¹
There are at least three possible interpretations.
Interpretation 1: Anelloviruses are pathogenic
The viruses themselves may contribute to chronic disease.
Interpretation 2: They are immunologic biomarkers
Anellovirus activity may reflect inadequate immune surveillance.
Interpretation 3: They are ecological indicators
Changes in anellovirus activity may indicate that the entire virome has become destabilized.
The third possibility is particularly interesting.
Anelloviruses could eventually become biomarkers of host immune state, even if they are not themselves pathogenic.
X. The Host Response May Be More Important Than the Virus
A reactivated virus does not necessarily need to destroy tissue directly.
It may instead activate the host immune system.
The 2026 study found associations between viral reactivation and changes in:
- cytokines;
- T-cell populations;
- monocyte activity;
- transcriptional programs;
- metabolism;
- platelet pathways;
- and inflammatory signaling.¹
This suggests that secondary viral activity may function as an immune amplifier.
The mechanism might therefore be:
virus → immune activation → tissue dysfunction
rather than:
virus → direct cytotoxicity → tissue destruction.
This distinction could explain why conventional imaging may appear relatively normal in some severely symptomatic patients.
Functional disturbances can occur without extensive anatomical destruction.
XI. Cytokines and the Inflammatory Feed-Forward Loop
Viral reactivation can stimulate interferons, chemokines and inflammatory cytokines.
Among the pathways implicated in recent studies are:
- IL-6;
- IL-10;
- CXCL10;
- CXCL11;
- tumor-necrosis-factor signaling;
- and pathways involving T-cell activation.¹
A possible sequence is:
viral antigen
↓
pattern-recognition receptors
↓
interferon signaling
↓
chemokine production
↓
monocyte and lymphocyte recruitment
↓
endothelial activation
↓
tissue inflammation
↓
additional release of self-antigens
↓
autoimmune activation
The result is a system in which viral reactivation can indirectly contribute to pathology even if viral replication itself is modest.
XII. Viral Reactivation and Autoimmunity
The relationship between viral reactivation and autoimmunity is especially important.
EBV is already implicated in several autoimmune diseases, including multiple sclerosis and systemic lupus erythematosus.¹¹,¹²
Several mechanisms may explain this relationship.
Molecular mimicry
A viral antigen may resemble a host protein.
B-cell stimulation
EBV directly alters B-cell biology.
Bystander activation
Inflammation may activate autoreactive lymphocytes that were previously quiescent.
Epitope spreading
Tissue injury can expose new self-antigens.
Altered antigen presentation
Inflammation changes which peptides are presented to T cells.
Thus, viral reactivation could potentially transform an acute antiviral response into a chronic autoimmune state.
The sequence might become:
SARS-CoV-2 → EBV reactivation → B-cell activation → autoreactive clones → chronic inflammation.
This is one reason viral reactivation and autoimmunity should not be considered competing hypotheses.
They may represent successive stages of the same process.
XIII. Endothelial Dysfunction
The endothelium may be one of the major convergence points for SARS-CoV-2 and secondary viral infection.
Both SARS-CoV-2 and CMV can affect endothelial biology.
Endothelial activation can produce:
- reduced nitric-oxide availability;
- abnormal vasoconstriction;
- increased vascular permeability;
- platelet activation;
- leukocyte adhesion;
- complement activation;
- and altered microvascular flow.
A 2026 Scientific Reports study reported an association among EBV reactivation, circulating blood microaggregates and persistent post-COVID symptoms.¹⁰
The study is provocative, but it should not be interpreted as establishing that so-called “microclots” cause long COVID.
The more cautious conclusion is that viral reactivation and abnormal blood-cell or platelet behavior may coexist within a broader inflammatory vascular state.
That relationship deserves controlled mechanistic investigation.
XIV. The Nervous System
The nervous system is particularly vulnerable to disturbances in immune and vascular homeostasis.
Several mechanisms may converge:
- neuroinflammation;
- endothelial dysfunction;
- blood–brain-barrier impairment;
- altered cerebral perfusion;
- autonomic dysfunction;
- mitochondrial stress;
- autoimmunity;
- and possibly neurotropic viral reactivation.
The blood–brain barrier normally protects neural tissue from circulating immune mediators.
Long COVID has been associated with abnormalities in this barrier and with persistent systemic inflammation.¹³
If viral reactivation simultaneously increases systemic immune signaling, the nervous system may become particularly susceptible.
This provides a potential biological bridge to:
- brain fog;
- cognitive slowing;
- headaches;
- sensory symptoms;
- autonomic dysfunction;
- neuropathic pain;
- and post-exertional neurological worsening.
XV. Mitochondrial Physiology
Viruses are metabolically demanding organisms.
Replication requires:
- nucleotides;
- amino acids;
- lipids;
- ATP;
- and extensive host-cell biosynthetic activity.
Viral reactivation can therefore alter cellular metabolism.
At the same time, chronic inflammatory signaling can impair mitochondrial oxidative phosphorylation and increase oxidative stress.
The result could be a vicious cycle:
inflammation → mitochondrial dysfunction → metabolic stress → impaired immune regulation → greater susceptibility to viral reactivation.
This mechanism is particularly relevant to fatigue.
A patient may have apparently adequate resting energy production but inadequate metabolic reserve when physiologic demand increases.
That conceptual framework is consistent with the characteristic phenomenon of post-exertional malaise, although it does not establish that viral reactivation is its cause.
XVI. Genomics: Why Some People Reactivate Viruses
The existence of viral reactivation after SARS-CoV-2 raises a fundamental question:
Why one person and not another?
Host genetics almost certainly contributes.
EBV biology provides a useful model.
The ability to control EBV depends upon genetic determinants affecting:
- antigen presentation;
- CD8+ T-cell responses;
- NK-cell signaling;
- cytokine pathways;
- B-cell regulation;
- and immune tolerance.
Recent genomic research has identified substantial genetic and immunologic determinants of EBV-associated disease.¹⁴
HLA variation is particularly important because HLA molecules determine which viral peptides are presented to T cells.
Thus:
host genotype → antigen presentation → antiviral surveillance → latent-virus control
could influence susceptibility to post-COVID viral reactivation.
The same principle may apply to CMV, HSV and other persistent viruses.
XVII. Viral Genomics
The viruses themselves are genetically heterogeneous.
EBV strains contain substantial sequence variation, and recent sequencing studies have identified geographic and disease-associated genomic differences.¹⁵
Viral genomic variation may influence:
- cell tropism;
- immune evasion;
- transcriptional programs;
- latency;
- reactivation;
- and molecular mimicry.
For long COVID, this raises another unresolved question:
Could particular viral strains be more likely to reactivate after SARS-CoV-2 infection?
There is currently insufficient evidence to answer that question.
But it is increasingly technically feasible.
Future studies could sequence:
- the SARS-CoV-2 infecting strain;
- the patient’s EBV/CMV/HSV genome;
- host HLA genotype;
- immune-cell repertoire;
- and viral transcripts.
This would permit a genuinely mechanistic analysis of host–virus interactions.
XVIII. Persistent SARS-CoV-2 Versus Reactivated Viruses
An essential distinction must be made between two phenomena.
Persistent SARS-CoV-2
SARS-CoV-2 itself or its components may remain detectable in tissues after acute infection.
Secondary viral reactivation
A different persistent virus—such as EBV or CMV—becomes transcriptionally or replicatively active.
The two mechanisms may coexist.
Indeed, they may reinforce one another.
Persistent SARS-CoV-2 antigen could maintain immune activation.
That immune dysregulation could permit EBV or CMV reactivation.
The secondary virus could then add another layer of inflammatory signaling.
Thus some long-COVID patients may experience a multi-virus state rather than a single persistent infection.
XIX. The Clinical Course
The clinical course of viral-reactivation-associated long COVID can be conceptualized in four stages.
Stage 1: Acute SARS-CoV-2 infection
The initial infection produces:
- viral replication;
- innate immune activation;
- cytokine signaling;
- endothelial stress;
- lymphocyte redistribution.
Latent viruses may reactivate during this period.
In severe COVID-19, reactivation may simply reflect critical illness.
Stage 2: Early convalescence
SARS-CoV-2 declines.
The patient may nevertheless continue to experience:
- fatigue;
- cognitive dysfunction;
- dyspnea;
- palpitations;
- sleep abnormalities;
- sensory symptoms.
Some viral-reactivation signals decline.
Others persist.
Stage 3: Established PASC
A subset develops persistent disease.
At this stage, continued viral activity may potentially maintain:
- inflammation;
- immune dysregulation;
- endothelial dysfunction;
- metabolic abnormalities;
- and neurological symptoms.
Stage 4: Relapsing or resolving disease
Some patients gradually recover.
Others have a fluctuating course.
The possibility that intermittent viral reactivation contributes to relapses is biologically plausible but presently unproven.
XX. Why Relapses Are Scientifically Interesting
Many patients describe long COVID as a disease that is not static.
Symptoms can improve and then abruptly worsen.
Potential triggers include:
- physical exertion;
- psychological stress;
- sleep deprivation;
- another infection;
- vaccination;
- hormonal changes;
- metabolic stress.
One possible mechanism is transient reactivation of persistent viruses.
Another is autonomic dysregulation.
Another is immune activation unrelated to viral replication.
These mechanisms are not mutually exclusive.
Future longitudinal studies should therefore obtain viral, immune and physiologic measurements during both symptomatic exacerbation and recovery.
That design could answer a crucial question:
Does viral transcription increase before a relapse?
If the answer is yes, the causal hypothesis becomes substantially stronger.
XXI. What Would Prove Causality?
The field needs to move beyond association.
The strongest evidence would follow five steps.
1. Temporal precedence
Viral reactivation must precede the relevant symptoms.
2. Tissue localization
The virus must be demonstrated in a biologically relevant compartment.
3. Functional effect
The viral activity must alter cellular or physiologic function.
4. Phenotype specificity
The effect should correspond to a recognizable clinical phenotype.
5. Therapeutic reversal
Suppressing the virus should improve the phenotype.
The fifth criterion is decisive.
If EBV-positive patients with fatigue improve consistently after a well-designed EBV-directed intervention, causality becomes compelling.
If they do not, EBV may be a biomarker rather than a driver.
XXII. Why Antiviral Therapy Is Not Yet Established Treatment
The biological hypothesis is attractive enough that some clinicians and patients have experimented with antiviral therapy.
That does not constitute evidence of efficacy.
Antivirals such as acyclovir, valacyclovir, ganciclovir and valganciclovir have distinct targets, pharmacology and toxicities.
There is currently insufficient evidence to recommend routine treatment of long COVID on the basis of presumed EBV or CMV reactivation.
The 2026 international clinical guideline continues to emphasize evidence-based diagnosis and individualized management rather than routine antiviral treatment for presumed latent-virus reactivation.¹⁶
This is an area in which rigorous clinical trials are urgently needed.
XXIII. A Better Clinical-Trial Strategy
Future trials should not enroll all patients with long COVID as though they possess the same disease.
Instead, patients should be stratified biologically.
EBV-positive phenotype
Possible characteristics:
- EBV transcription;
- EBV DNA;
- compatible antibody profile;
- fatigue;
- cognitive symptoms.
CMV-positive phenotype
Possible characteristics:
- CMV transcription;
- vascular or inflammatory phenotype;
- immune-cell abnormalities.
HSV-positive neurological phenotype
Possible characteristics:
- evidence of HSV activity;
- neuropathic symptoms;
- autonomic dysfunction;
- compatible neurological physiology.
Anellovirus-associated phenotype
Possible characteristics:
- persistent Anelloviridae activity;
- fatigue;
- physical-function impairment.
Such trials could determine whether specific viruses are:
drivers, amplifiers, biomarkers, or innocent bystanders.
XXIV. The Importance of Tissue
The next generation of research must move beyond blood.
Blood is convenient but biologically incomplete.
A virus can persist in:
- B cells;
- lymph nodes;
- gut;
- endothelial compartments;
- respiratory tissues;
- sensory ganglia;
- or other anatomical reservoirs.
A negative blood PCR does not exclude localized viral activity.
The ideal future study would combine:
- blood;
- saliva;
- stool;
- respiratory specimens;
- immune-cell fractions;
- tissue when clinically justified;
- cell-free DNA/RNA;
- viral transcriptomics;
- proteomics;
- and longitudinal symptom measurement.
The objective is not merely to find a virus.
It is to establish the relationship:
virus → tissue → molecular effect → physiologic consequence → symptom.
XXV. Viral Reactivation as Part of a Larger Disease Network
The most plausible model is not that viral reactivation explains all long COVID.
Rather, it may interact with several other mechanisms.
Persistent SARS-CoV-2
Provides continuing antigenic stimulation.
Viral reactivation
Adds secondary antigenic stimulation.
Autoimmunity
Produces tissue-directed immune responses.
Endothelial dysfunction
Impedes microvascular physiology.
Autonomic dysfunction
Disturbs cardiovascular and gastrointestinal regulation.
Mitochondrial dysfunction
Reduces metabolic reserve.
Microbiome disturbance
Changes mucosal immunity and systemic signaling.
The resulting network might look like:
SARS-CoV-2
↓
immune dysregulation
↙ ↓ ↘
viral persistence — viral reactivation — autoimmunity
↓ ↓ ↓
endothelial dysfunction
↓
autonomic + metabolic + neurologic dysfunction
↓
persistent multisystem disease
This model accommodates the extraordinary heterogeneity of long COVID.
XXVI. The Most Important Scientific Caveat
It is tempting to say:
“EBV reactivation causes long COVID.”
The evidence does not justify that statement.
A more defensible conclusion is:
SARS-CoV-2 infection can be associated with reactivation of persistent viruses, and accumulating evidence suggests that persistent or recurrent viral activity may contribute to chronic symptoms in a biologically defined subset of patients.
That distinction matters enormously.
EBV infection is nearly ubiquitous.
Therefore, the mere presence of EBV cannot explain why only some people develop long COVID.
The critical variable may instead be:
what the virus is doing, where it is doing it, when it is doing it, and how the host responds.
XXVII. The Emerging Concept of a Post-SARS-CoV-2 Virome Disorder
Taken together, the evidence supports a broader hypothesis.
SARS-CoV-2 may not simply cause a persistent disease of one virus.
It may cause a disturbance of the host–virome relationship.
The normal state:
persistent viruses + immune surveillance = equilibrium
After SARS-CoV-2:
SARS-CoV-2 + altered immunity = virome instability
The consequences could include:
- EBV reactivation;
- CMV reactivation;
- HSV reactivation;
- altered anellovirus activity;
- persistent SARS-CoV-2 antigen;
- autoimmune responses;
- endothelial inflammation;
- metabolic stress.
The clinical syndrome would then represent the cumulative physiologic consequences of this altered ecosystem.
This hypothesis explains why long COVID can look so different from one patient to another.
XXVIII. Conclusion
The viral-reactivation hypothesis has progressed considerably since the earliest observations of EBV antibodies in people with persistent symptoms after COVID-19.
The most important advance has been methodological.
Researchers are increasingly able to distinguish viral presence from viral activity, and viral activity from its consequences.
Recent longitudinal multi-omic work demonstrates that SARS-CoV-2 infection can be accompanied by reactivation of several persistent viruses and that different viruses display distinct temporal trajectories.¹ The association of persistent viral activity during convalescence with measures of long-COVID disease strengthens the possibility that secondary viral activity may participate in chronic pathogenesis.
EBV remains the leading candidate because of its ubiquity, biology and relationship to immune dysregulation. CMV offers a plausible connection to endothelial and immune pathology. HSV provides a potential explanation for tissue-specific neurological phenomena that may not be visible in blood. Anelloviruses raise the possibility that changes in the broader human virome may themselves serve as biomarkers of immune dysfunction.
But the central scientific question remains unresolved.
Does viral reactivation cause long COVID, perpetuate it, amplify it, or simply mark the immune disturbance produced by SARS-CoV-2?
The answer will not come from another cross-sectional antibody study.
It will require longitudinal sampling, tissue-specific viral transcriptomics, host genomics, single-cell immunology, functional experiments and, above all, randomized trials in biologically defined patients.
If those studies demonstrate that suppressing a reactivated virus reverses a corresponding clinical phenotype, the consequences will be profound.
Long COVID would no longer be conceptualized solely as a lingering aftermath of SARS-CoV-2.
It would become, in some patients, a secondary postviral disorder of the human virome—a disease in which the initial infection destabilizes a lifelong relationship between host and persistent viruses, allowing one or more normally controlled viral populations to become active participants in chronic disease.
That possibility is now sufficiently supported to warrant intensive investigation, but not yet sufficiently established to justify treating every patient with long COVID for an assumed latent-virus reactivation.
The distinction between those two positions—between biological plausibility and demonstrated causality—is where the next decisive advances in long-COVID research will occur.
Numbered References
- Maguire C, et al. Virus reactivation in acute and long COVID-19. Nature. 2026. Longitudinal multi-omic analysis of viral transcriptomics, host responses and clinical outcomes.
- Oba S, Hosoya T, Iwai H, Yasuda S. Long COVID: mechanisms of disease, multisystem sequelae, and prospects for treatment. Immunol Med. 2026;49:35-58.
- Vlaming-van Eijk LE, Tang G, Bourgonje AR, et al. Post-COVID-19 condition: clinical phenotypes, pathophysiological mechanisms, pathology, and management strategies. J Pathol. 2025;266:369-389.
- Damania B, Kenney SC, Raab-Traub N. Epstein-Barr virus: biology and clinical disease. Cell. 2022;185:3652-3670.
- Guo R, Gewurz BE. Epigenetic regulation of the Epstein-Barr virus latent-lytic switch. J Med Virol. 2026.
- Phetsouphanh C, Darley DR, Wilson DB, et al. Immunological dysfunction persists for 8 months following initial mild-to-moderate SARS-CoV-2 infection. Nat Immunol. 2022;23:210-216.
- Gold JE, Okyay RA, Licht WE, Hurley DJ. Investigation of long COVID prevalence and its relationship to Epstein-Barr virus reactivation. Pathogens. 2021;10:763.
- Peluso MJ, Deeks SG, Mustapic M, et al. SARS-CoV-2 and Epstein-Barr virus coinfection in association with post-acute sequelae of COVID-19. J Clin Invest. 2023.
- Farid E, et al. Epstein-Barr virus reactivation in post-COVID-19 patients with persistent symptoms. Auris Nasus Larynx. 2025;52:442-446.
- Wick N, Hermann M, Lisch C, et al. Clinical relevance of circulating blood microaggregates and reactivation of Epstein-Barr virus in long-term post-COVID syndrome patients. Sci Rep. 2026;16:12559.
- Gondé A, Fournier B, Latour S. Human genetic and immunological determinants of Epstein-Barr virus-driven disease. Curr Opin Virol. 2026;76:101538.
- Lanz TV, Brewer RC, Ho PP, et al. Clonally expanded B cells in multiple sclerosis and Epstein-Barr virus biology. Nature. 2022;603:321-327.
- Greene C, et al. Blood-brain barrier disruption and sustained systemic inflammation in individuals with long COVID-associated cognitive impairment. Nat Neurosci. 2024.
- Population-scale sequencing resolves determinants of persistent Epstein-Barr virus DNA. Nature. 2026.
- EBV genome variations and association with diseases. J Med Virol. 2026.
- Cao B, Soriano JB, Wang Q, et al. Clinical practice guideline for long COVID prevention and treatment. Eur Respir J. 2026;68:2502611.
- Gáspár Z, Szabó BG, Ceglédi A, Lakatos B. Human herpesvirus reactivation and its potential role in the pathogenesis of post-acute sequelae of SARS-CoV-2 infection. Geroscience. 2025;47:167-187.
- Clinical and mechanistic literature on viral-induced autoimmunity, including molecular mimicry, bystander activation and epitope spreading. Hum Immunol. 2026.