Current Takeaway
A consistent pattern of gut microbial imbalance has emerged across ME/CFS and Long COVID cohorts: patients show reduced bacterial diversity, depletion of short-chain fatty acid (SCFA)-producing species, and—in a finding relatively new to the field—altered fungal communities whose composition shifts in an age-dependent manner. These bacterial and fungal changes correlate with fatigue severity and with markers of intestinal barrier disruption such as FABP2, suggesting that a permeable gut lining may allow inflammatory signals to enter the bloodstream and sustain systemic immune activation. A complementary line of evidence points to mucosal sites—the gut and the oral cavity—as reservoirs for herpesviruses and human endogenous retroviruses that remain undetectable in peripheral blood, offering a possible explanation for why blood-based viral searches in ME/CFS have yielded inconsistent results.
Two converging mechanistic proposals are shaping the current research direction. One framework situates the gut-brain axis centrally, with MAIT cells responding to microbial antigens and probiotics or dietary fiber being proposed—though not yet proven—as ways to restore homeostasis. A second and more specific proposal identifies microbiota-derived extracellular vesicles (GMEVs) as functional carriers that physically cross from a disrupted gut into systemic circulation, activate macrophages and brain-resident microglia, and thereby translate intestinal dysbiosis into neuroinflammation. Both proposals remain mechanistically plausible but have not been validated in large prospective human trials, and causal directionality—whether dysbiosis drives the disease or results from it—is not established.
Why This Matters
Dysbiosis and intestinal barrier failure provide a testable biological interface between the gut microenvironment and the systemic immune, metabolic, and neurological abnormalities documented in ME/CFS and Long COVID. Identifying specific microbial signatures and their vesicular byproducts creates potential targets for non-invasive diagnostics (stool sequencing, mycobiome profiling) and for interventions such as dietary fiber supplementation, prebiotics, probiotics, and GMEV-directed therapies. The mucosal viral reservoir hypothesis also redirects sampling strategy away from blood and toward saliva and stool, which could recover viral signals that blood testing has systematically missed.
State of Evidence
- Established: ME/CFS patients have measurably lower gut bacterial diversity and reduced levels of SCFA-producing genera (especially Rikenellaceae and Ruminococcaceae), with lower concentrations of acetate, butyrate, isobutyrate, and isovalerate correlating with fatigue severity scores across a matched cohort.
- Plausible but early: The gut mycobiome is disrupted in ME/CFS with age-dependent patterns—reduced fungal richness in younger patients but paradoxically increased richness in the elderly—and machine-learning models built on fungal taxa achieve high discriminatory accuracy in single-centre cohorts. Mucosal tissues harbour herpesvirus and HERV signals absent in paired blood samples, suggesting compartmentalised viral persistence. Microbiota-derived extracellular vesicles from Long COVID individuals can impair intestinal barriers and activate microglia in mouse and human iPSC models. Intestinal-damage biomarkers such as FABP2 are elevated in ME/CFS and may index barrier permeability. MAIT cell activation by gut microbial antigens is proposed as a bridge between dysbiosis and systemic neuroinflammation.
- Not established: A causal direction between gut dysbiosis and ME/CFS onset or progression. The clinical efficacy of probiotics, prebiotics, high-fibre dietary interventions, or GMEV-targeted therapies in reducing ME/CFS or Long COVID symptom burden. Whether mucosal viral persistence is a primary driver of symptoms or an opportunistic consequence of immune suppression secondary to the disease.
- Key limitations: Most bacterial and fungal cohort studies are cross-sectional, single-centre, and conducted in geographically restricted populations, limiting generalisability. The mycobiome diagnostic models showing near-perfect accuracy require independent external validation before clinical use. The GMEV mechanistic study is a preprint and has not completed peer review. Longitudinal studies tracking dysbiosis from illness onset through recovery phases are largely absent.
Timeline
2025-11-20 - Review proposes mucosal tissues as overlooked viral reservoirs in ME/CFS
A narrative review by Perera et al. synthesised evidence on viral detection across tissue compartments in ME/CFS, arguing that the documented inconsistency of blood-based viral findings reflects a methodological sampling bias rather than an absence of viral pathology. Saliva and gut biopsy data show herpesvirus (EBV, HHV-6) reactivation and human endogenous retrovirus (HERV) expression in mucosal tissue even when the same viruses are undetectable in paired peripheral blood, identifying the gastrointestinal and respiratory mucosae as probable long-term reservoirs maintaining chronic immune dysregulation. The review stops short of establishing a direct causal chain from mucosal viral persistence to specific symptoms, and the practical barrier of invasive gut biopsies remains a limiting factor in building larger evidence datasets—though the authors note that saliva and stool sampling offer less invasive alternatives.
Sources:
2025-12-12 - Case-control study maps bacterial dysbiosis and SCFA depletion to fatigue severity
A matched case-control study by Cheng et al. enrolled 40 ME/CFS patients and 40 healthy controls, using 16S rRNA sequencing to characterise gut bacterial composition and GC-MS to quantify six short-chain fatty acids in stool. Patients showed significantly lower alpha diversity and a consistent reduction in Rikenellaceae and Ruminococcaceae—families associated with SCFA production—alongside significantly depressed concentrations of acetate, butyrate, isobutyrate, and isovalerate; lower SCFA levels correlated directly with higher FSS and FS-14 fatigue scores and lower quality-of-life ratings. Participants consuming higher dietary fibre had measurably higher acetate and isovalerate, raising the hypothesis that fibre supplementation could partially restore the SCFA deficit. The cross-sectional design means the study cannot determine whether SCFA depletion initiates or amplifies disease, and interventional trials to test dietary or probiotic correction of this deficit have not yet been conducted.
Sources:
2025-12-30 - Age-stratified sequencing reveals distinct fungal dysbiosis patterns in ME/CFS
Gan et al. performed high-throughput ITS sequencing on stool from 59 ME/CFS patients and 59 matched controls, stratified into young, middle-aged, and elderly cohorts, making this one of the first systematic characterisations of the gut mycobiome in ME/CFS. Fungal richness was reduced in younger and middle-aged patients but unexpectedly elevated in the elderly cohort, indicating that pooling age groups masks biologically distinct subpopulations; key discriminatory taxa included Preussia, Aspergillus, and Chaetomium. Random Forest diagnostic models built on these age-specific fungal signatures achieved classification accuracies up to 100% within the study cohort, suggesting strong within-sample discriminatory signal. Replication in larger, geographically diverse cohorts is required before these models have any clinical diagnostic utility, and the functional contribution of these fungal taxa to host immune activation or metabolic disruption remains unexplored.
Sources:
2026-02-01 - Comprehensive review integrates gut-brain axis evidence and identifies FABP2 as a barrier biomarker
A broad clinical and mechanistic review by Xu et al. synthesised evidence through mid-2025 on ME/CFS aetiology, pathophysiology, diagnostics, and treatment, with dedicated sections on the gut-brain axis and its overlap with Long COVID. The review identifies FABP2 (intestinal fatty acid-binding protein 2) as a circulating biomarker of enterocyte damage that may index intestinal permeability in patients, and highlights MAIT cell activation by microbial antigens as a candidate immune bridge between gut dysbiosis and systemic neuroinflammation. It positions probiotics and other microbial preparations as biologically plausible interventions for restoring gut-immune-brain homeostasis, while acknowledging that large-scale clinical trial evidence for any microbiome-directed therapy in ME/CFS remains absent. The review’s broad scope means individual mechanistic claims are not independently replicated within the paper itself, and it does not establish causal ordering between gut, immune, and neurological abnormalities.
Sources:
2026-03-02 - Preprint demonstrates microbiota-derived extracellular vesicles activate microglia and impair gut barrier
Aranguren et al. conducted a multi-model mechanistic study, transplanting Long COVID microbiota into germ-free mice and isolating the gut microbiota-derived extracellular vesicles (GMEVs) from Long COVID patients for direct testing on human iPSC-derived microglia and macrophages. GMEVs alone were sufficient to impair intestinal barrier integrity, trigger pro-inflammatory responses in immune cells, and induce neuroimmune activation and systemic inflammation in mice following oral administration—establishing them as functional mediators of the gut-brain signalling axis rather than passive byproducts of dysbiosis. This positions GMEV composition and abundance as potential therapeutic targets and opens a rationale for interventions that reduce dysbiosis-driven vesicle production. The study is a preprint and has not completed peer review; the mouse oral-administration model may not replicate the route or dose of GMEV exposure occurring in human patients, and whether GMEV profiles differ between ME/CFS and Long COVID with ME/CFS phenotype has not been directly compared.
Sources:
2026-05-22 - Comparative study reveals overlapping gut microbiome alterations in ME/CFS and Long COVID
A cross-sectional case-control study of 163 participants (76 ME/CFS, 50 Long COVID, 37 healthy controls) in Bulgaria, published in Biomedicines by Donchev et al., demonstrated that both post-infectious cohorts share overlapping gut dysbiosis relative to healthy controls. Both patient groups showed enrichment of Sutterella and Lawsonibacter alongside depletion of Terrisporobacter and Lachnospiraceae. Predicted pathway analysis suggested shared decreases in anaerobic acetate/hydrogen carbon flow and inositol degradation capacity. Multivariate regression linked gut microbiome variations most strongly to fatigue and physiosomatic symptom severity rather than cognitive or affective symptoms. Because of the cross-sectional study design, it cannot establish causal relationships or temporal sequence. Furthermore, information on potential microbiome-modifying confounding factors (diet, prebiotic/probiotic use, supplements) was incomplete, and severely affected, bedbound patients were underrepresented.
Sources:
2026-06-26 - Study links gut microbial aromatic amino acid metabolites and AHR activation to neurocognitive symptoms in ME/CFS
A case-control study of 67 participants (32 ME/CFS, 35 controls) published in MicrobiologyOpen by Esteban et al. investigated the production of tryptophan metabolites and Aryl-Hydrocarbon Receptor (AHR) agonists by gut microbes. The study demonstrated gut dysbiosis (lower alpha diversity and lower Firmicutes:Bacteroidetes ratio) in ME/CFS, alongside changes in a tryptophan-metabolite-correlated subcommunity of Clostridia. Targeted metabolomics revealed nine stool metabolites significantly elevated in ME/CFS patients, including the AHR agonists indole, indoleacetate, and tryptophan, alongside tyrosine and phenylalanine metabolites. Although overall ex vivo stool AHR activation did not differ by disease status, elevated AHR agonist activity was strongly associated with neurocognitive symptoms across all participants, suggesting microbial aromatic amino acid metabolites acting on AHR may contribute to neurocognitive dysfunction (brain fog) in ME/CFS. As a cross-sectional study, it cannot establish causal relationships or monitor temporal fluctuations.
Sources:
Open Questions
- Does correcting SCFA depletion through dietary fibre, prebiotics, or direct SCFA supplementation produce measurable reductions in fatigue or barrier permeability in prospective ME/CFS trials?
- Are gut fungal dysbiosis patterns—particularly the paradoxical increase in richness among elderly patients—driven by immune suppression, antibiotic history, or primary metabolic factors, and do they influence bacterial dysbiosis or operate through independent pathways?
- Can GMEV cargo composition (protein, lipid, or nucleic acid content) distinguish ME/CFS from Long COVID and from other post-viral syndromes, and do specific vesicle species predict neuroinflammatory severity?
- Do mucosal viral reservoirs (EBV, HHV-6, HERVs) in gut or oral tissue persist independently of blood titres, and does antiviral treatment targeting mucosal compartments alter the dysbiosis or systemic symptom burden?
- Is intestinal barrier disruption (indexed by FABP2 or occludin loss) a cause, a consequence, or a co-occurring feature of the metabolic and immune abnormalities documented in ME/CFS?