Synthesis and Open Questions
Genetic and epigenetic research in ME/CFS has matured substantially over the past decade, progressing from underpowered candidate gene studies to genome-wide approaches, from speculation about epigenetic involvement to empirical demonstration of DNA methylation and histone modification changes, and from simple gene lists to integrated multi-omics analyses. This body of evidence establishes genetic predisposition as a substantial contributor to ME/CFS risk while highlighting the complex polygenic architecture and gene-environment interactions that determine disease manifestation.
1 Key Established Findings
Several conclusions now rest on firm empirical ground:
Moderate heritability Twin studies consistently demonstrate heritability estimates of \(h^2 = 0.3\)–$ 0.5$, indicating that 30–50% of disease liability reflects genetic factors. This moderate heritability implies substantial genetic contribution while confirming environmental factors’ essential role.
Polygenic architecture with neuronal convergence DecodeME (n\(>\) 15,000) identified 8 genome-wide significant loci with small effect sizes (OR 0.93–1.10) and SNP heritability of 9.5% (DecodeME Consortium, Ponting, et al. 2025). The most notable finding is convergence on brain-expressed genes involved in neuronal development and synaptic function (CA10, SHISA6, SOX6, LRRC7, DCC, UNC13C), confirmed by MAGMA tissue enrichment and independently supported by Stanford rare-variant analyses (Section Replication Status: Not Yet Replicated (By Design)). Genetic correlations are strongest with IBS (\(r_g = 0.75\)), sleep, and mood traits, but absent for classical autoimmune diseases (Section DecodeME Brain Tissue Enrichment).
Gene-environment interaction The observation that only a minority of individuals exposed to triggering infections develop ME/CFS, combined with familial aggregation patterns, validates gene-environment interaction as the central etiological framework. Genetic variants establish vulnerability; environmental triggers are necessary for disease expression.
Immune ambiguity and autophagy signals DecodeME immune-related hits (RABGAP1L, BTN2A2, TAOK3, HLA-DQA1*05:01) lie in gene-dense regions creating assignment ambiguity, and the absence of genetic correlation with autoimmune diseases constrains the autoimmune hypothesis. Unexpectedly, autophagy genes (FBXL4 for mitophagy, CCPG1 for ER-phagy) provide a genetic link to mitochondrial dysfunction (Chapter Energy Metabolism and Mitochondrial Function), suggesting constitutional vulnerability in organelle quality control.
Epigenetic reprogramming Demonstration of DNA methylation changes, altered histone modifications, and dysregulated microRNA expression establishes epigenetic reprogramming as a plausible mechanism for chronicity. These potentially reversible modifications provide therapeutic targets.
This synthesis achievement combines findings of differing replication strength. Moderate heritability (\(h^2 = 0.3\)–$ 0.5$) from twin studies is the most robust element, replicated across multiple independent twin registries. Polygenic architecture is confirmed by DecodeME but locus-level replication is pending (see Achievement DecodeME GWAS Findings). Gene-environment interaction is well-supported conceptually but the specific genetic variants mediating susceptibility remain largely unidentified. Epigenetic reprogramming evidence comes from small studies with limited replication—DNA methylation findings in particular show inconsistent patterns across studies, partly due to cell-type heterogeneity in whole-blood samples.
2 Integration with Broader ME/CFS Pathophysiology
The genetic and epigenetic findings detailed in this chapter provide the constitutional substrate upon which the dysregulated physiological systems described in other chapters develop.
The immune dysfunction documented in Chapter Immune System Dysfunction—NK cell cytotoxicity impairment, T cell exhaustion, cytokine dysregulation—reflects both genetic predisposition (HLA types, immune gene variants, TRPM3 polymorphisms) and epigenetic reprogramming (T cell chromatin states characteristic of exhaustion, methylation of immune genes). Genetic susceptibility determines baseline immune function capacity; epigenetic changes following infection establish chronic dysfunction states.
The metabolic dysfunction of Chapter Energy Metabolism and Mitochondrial Function—reduced oxidative phosphorylation, impaired ATP production, glycolytic shifts—similarly combines genetic vulnerability (mitochondrial gene variants, metabolic enzyme polymorphisms) with acquired epigenetic silencing of metabolic genes. The gene expression patterns show reduced expression of mitochondrial and metabolic pathway genes, potentially reflecting both genetic determinants of baseline expression and epigenetic downregulation following metabolic stress.
Neurological manifestations (Chapter Neurological and Neurocognitive Dysfunction) may reflect TRPM3 dysfunction (with genetic variants affecting baseline expression and function), neurotransmitter system genetic variants, and epigenetic changes affecting blood-brain barrier integrity and neuroinflammation. The cognitive and autonomic symptoms could arise from the intersection of genetic liability and acquired epigenetic modifications.
This integrated model suggests that ME/CFS arises when genetic predisposition across multiple systems (immune, metabolic, neurological, cardiovascular) encounters environmental triggers sufficient to induce epigenetic reprogramming. The specific symptom profile reflects which genetic vulnerabilities predominate and which epigenetic changes occur, explaining clinical heterogeneity. Chapter Integrative Models and Multi-System Pathophysiology builds upon this genetic-epigenetic foundation to develop comprehensive multi-system models of ME/CFS pathophysiology, examining how constitutional factors interact with acquired dysfunction to create stable pathological states (Section Multi-System Integration and Synthesis).
3 Unresolved Questions and Future Directions
Despite substantial progress, critical questions remain:
GWAS identifies associated genomic loci but typically does not pinpoint causal variants or affected genes. For DecodeME-identified loci, fine-mapping studies using dense genotyping and functional genomics are needed to identify specific causal variants, determine which genes they affect, and elucidate mechanisms by which they influence disease risk. Do ME/CFS risk variants affect transcription factor binding sites, alter splicing, modify protein sequence, or influence other molecular processes?
Observed epigenetic changes could represent disease-driving mechanisms or secondary consequences of chronic illness. Longitudinal studies examining epigenetic changes before, during, and after disease onset would address causality. Do epigenetic changes precede symptom development in at-risk individuals? Do they persist during remission or normalize with symptom improvement? Can experimentally reversing specific epigenetic modifications (using CRISPR-based epigenome editing or small molecule epigenetic drugs) alleviate cellular dysfunction in patient cells?
Does ME/CFS comprise genetically distinct subtypes with different molecular mechanisms? Cluster analyses based on genetic profiles, gene expression patterns, or epigenetic signatures might identify patient subgroups with different pathophysiological mechanisms, prognoses, and treatment responses. Such molecular subtyping could enable personalized treatment selection.
Given that epigenetic modifications are potentially reversible, can therapeutic interventions normalize methylation patterns, histone modifications, or microRNA expression? Would such normalization translate to clinical improvement? Trials of epigenetic-modifying drugs (HDAC inhibitors, methyltransferase inhibitors, demethylating agents) could test this hypothesis, though broad epigenetic drugs have significant toxicities. More targeted approaches using small molecules affecting specific epigenetic enzymes or dietary interventions affecting metabolite availability might offer safer therapeutic windows.
Can polygenic risk scores identify individuals at high genetic risk who might benefit from preventive interventions? Following the 2004 Bergen Giardia outbreak model, future post-infection cohorts could stratify by genetic risk and test whether early interventions (aggressive rest, anti-inflammatory treatments, metabolic support) prevent chronic illness development in high-risk individuals. Such prevention trials could validate genetic risk prediction and identify modifiable factors in the gene-environment interaction.
Genetic correlation analyses suggest shared genetic liability between ME/CFS, fibromyalgia, irritable bowel syndrome, and Long COVID. Do these conditions represent different manifestations of the same underlying genetic vulnerability, or do they have partially overlapping but distinct genetic architectures? Detailed comparison of GWAS findings across conditions would address this question and might reveal common therapeutic targets applicable across multiple chronic overlapping pain conditions.
The genetic and epigenetic foundations of ME/CFS, while increasingly well characterized, point toward a future of precision medicine approaches where genetic profiling informs diagnosis, prognostication, and treatment selection, and where therapeutic interventions target the specific molecular pathways dysregulated in individual patients.
4 Connective Tissue Genetics and ECM Regulation
The high comorbidity between ME/CFS and hypermobile Ehlers-Danlos syndrome (hEDS) suggests genetic overlap in connective tissue regulation pathways. The following speculations explore potential genetic mechanisms linking ME/CFS to connective tissue pathology.
Certainty: 0.40. LOX requires copper as cofactor + P5P (B6) for activation. Copper deficiency → impaired collagen crosslinking → hypermobility phenotype even without hEDS mutations. ME/CFS patients may develop functional LOX deficiency due to: (1) copper deficiency from malabsorption or diet; (2) P5P deficiency impairing LOX activation; (3) acquired LOX inhibition from oxidative stress. This could explain connective tissue manifestations in ME/CFS patients without hEDS mutations.
Mechanistic components:
- LOX catalyzes lysine-derived crosslinks in collagen and elastin
- Copper deficiency directly impairs LOX enzymatic activity
- P5P (B6) required for LOX cofactor synthesis
- Oxidative stress can inhibit LOX function
Clinical implications: Copper supplementation (2-4 mg/day), vitamin B6 (25-50 mg P5P), or vitamin C (LOX cofactor) could improve collagen crosslinking in ME/CFS patients with confirmed deficiency. Serum copper and ceruloplasmin testing would identify candidates.
Testable predictions:
- ME/CFS patients with hypermobility but no hEDS mutations will show LOX pathway abnormalities (low copper, low P5P, or elevated LOX inhibitors)
- Collagen crosslink biomarkers (pyridinoline, deoxypyridinoline) will correlate with LOX pathway status
- LOX pathway correction (copper/B6 supplementation) will improve hypermobility scores and reduce connective tissue symptoms
Limitations: Copper and B6 levels have not been systematically measured in ME/CFS. LOX activity assays are not clinically available. The contribution of LOX deficiency to ME/CFS connective tissue symptoms is untested.
Certainty: 0.40. miRNAs regulate collagen, elastin, MMP gene expression. ME/CFS miRNA dysregulation is established; may produce characteristic ECM-destabilizing miRNA signature. Specific miRNAs (miR-29, miR-196, miR-140) are known to regulate collagen synthesis and degradation; their dysregulation could produce connective tissue manifestations in ME/CFS.
Mechanistic components:
- miR-29 family suppresses collagen gene expression (COL1A1, COL3A1)
- miR-196 regulates ECM composition and remodeling
- miR-140 modulates cartilage ECM homeostasis
- ME/CFS-specific miRNA signature may tilt balance toward ECM degradation
Testable predictions:
- ME/CFS patients will show elevated miR-29, miR-196, miR-140 in blood and possibly affected tissues
- These miRNA elevations will correlate with collagen degradation markers
- Antagomir treatment (miRNA inhibitors) in cell culture will normalize collagen production in ME/CFS fibroblasts
Limitations: Comprehensive miRNA profiling in ME/CFS is limited. Tissue-specific miRNA expression patterns are unknown. Antagomir therapeutics are experimental and not clinically available.
Certainty: 0.30. Polymorphisms in LOX, P4H, P3H, transglutaminase may predispose to ME/CFS in hypermobile patients. These enzymes are critical for collagen and elastin crosslinking; genetic variants producing partial loss-of-function could create subclinical connective tissue vulnerability that, when combined with environmental triggers (infection, oxidative stress), produces ME/CFS. Genetic testing could identify high-risk individuals for personalized cofactor dosing.
Candidate enzymes and polymorphisms:
- LOX (lysyl oxidase): copper-dependent crosslinking enzyme
- P4H (prolyl 4-hydroxylase): collagen hydroxylation
- P3H (prolyl 3-hydroxylase): collagen stability
- Transglutaminase-2: ECM crosslinking
Therapeutic approach: Individuals with high-risk polymorphisms could receive prophylactic cofactor support (copper, vitamin C, vitamin B6, iron for P4H/P3H) to maximize residual enzyme activity, potentially preventing or mitigating ME/CFS onset.
Testable predictions:
- ME/CFS patients with hEDS features will show enrichment of LOX, P4H, P3H, or transglutaminase polymorphisms
- Functional assays will show reduced enzyme activity in polymorphism carriers
- Cofactor supplementation will improve enzyme activity and reduce connective tissue symptoms in polymorphism-positive ME/CFS patients
Limitations: No GWAS has identified these loci in ME/CFS. Functional significance of most polymorphisms is unknown. Cofactor dosing regimens are untested. Prevention trials would require large cohorts and long follow-up.
Certainty: 0.35. Polymorphisms in iron regulatory genes — hepcidin (HAMP), ferroportin (SLC40A1), and matriptase-2 (TMPRSS6) — may influence ME/CFS susceptibility and severity by altering the hepcidin-ferroportin setpoint (Speculation Iron Dysregulation: Hepcidin Setpoint Shift and the Hypoferremia-Ferroptosis Trap, Chapter Energy Metabolism and Mitochondrial Function). Common TMPRSS6 variants (rs855791, rs4820268) affect iron status in the general population; the A736V missense variant (rs855791) reduces matriptase-2 activity, increasing hepcidin and lowering serum iron and transferrin saturation by 5–10% per allele (Nai et al. 2011) (Benyamin et al. 2014). In ME/CFS, carrying the TMPRSS6 736V variant could amplify the IL-6-hepcidin axis (Speculation Hepcidin-Inflammation Axis as Endocrine-Immune Bridge, Chapter Endocrine and Metabolic Dysfunction), producing a lower iron setpoint and greater functional iron deficiency in response to the same inflammatory stimulus.
Mechanism. The hepcidin-ferroportin axis determines tissue iron distribution. HAMP polymorphisms that increase hepcidin transcription (promoter variants, G-582A) would exacerbate inflammation-driven hepcidin elevation, worsening the hypoferremia-ferroptosis trap. SLC40A1 variants that reduce ferroportin expression or increase its hepcidin-sensitivity (Q248H, N144H) would make enterocytes and macrophages more susceptible to hepcidin-mediated ferroportin degradation. TMPRSS6 (matriptase-2) downregulates hepcidin by cleaving hemojuvelin (HJV), a BMP co-receptor that drives HAMP transcription. TMPRSS6 loss-of-function variants (736V) reduce HJV cleavage, allowing higher hepcidin at any given IL-6 level. The net effect is additive genetic loading: each risk variant shifts the hepcidin-ferroportin setpoint toward iron retention in macrophages and iron deficiency in muscle, compounding the energy deficit.
Prediction. ME/CFS patients carrying TMPRSS6 rs855791 V/V (736VV) will show: (1) higher hepcidin at matched IL-6 levels compared to T/A carriers; (2) lower serum iron, transferrin saturation, and MCV; (3) greater fatigue severity and muscle weakness; (4) poorer response to iron supplementation (because hepcidin-mediated ferroportin blockade prevents oral iron absorption). SLC40A1 and HAMP variant carriers will show directionally similar but quantitatively distinct profiles.
Limitations. No study has tested iron regulatory gene variants in ME/CFS. Effect sizes from population studies are small (2–5% variance in iron parameters per allele); their clinical relevance in a disease with multiple iron-compounding factors is unknown. Gene-environment interactions (dietary iron intake, inflammation level, menstrual status) would need to be modelled to detect variant effects.
Certainty: 0.35. Fatty acid desaturase genes FADS1 and FADS2 encode Delta-5 and Delta-6 desaturases, the rate-limiting enzymes for long-chain polyunsaturated fatty acid (LC-PUFA) synthesis from dietary precursors. Common FADS cluster polymorphisms (rs174537, rs174546, rs174583) determine inter-individual variation in LC-PUFA levels, including arachidonic acid (AA), eicosapentaenoic acid (EPA), and docosahexaenoic acid (DHA) (Schaeffer et al. 2006) (Ameur et al. 2012). Since AA, EPA and DHA are the substrates for SPM biosynthesis (resolvins, maresins, protectins) (Serhan, Libreros, and Nshimiyimana 2022), FADS variants that reduce LC-PUFA levels may impair the resolution phase of inflammation in ME/CFS by limiting SPM precursor availability. This is consistent with the emerging picture of resolution failure in post-viral syndromes: Engert et al. (2026) demonstrated that sleep disturbance dysregulates SPM biosynthesis in Long COVID (Engert et al. 2026), and Rauf et al. (2026) characterized PASC as a disorder of impaired innate immune resolution (Rauf, Naveed, and Asghar 2026). This connects to the gut-microbiome lipid mediator axis (Speculation Gut-Microbiome Lipid Mediator Axis: SPM Precursor Conversion Deficiency, Chapter Gastrointestinal and Microbiome Dysfunction): patients with both low-genetic SPM-precursor capacity (unfavourable FADS variants) AND low bacterial SPM-precursor supply (dysbiosis) would face a compounded deficit.
Mechanism. The minor (low-activity) allele at rs174537 (G allele, frequency ~30% in Europeans) is associated with: (1) 15–25% lower AA and EPA in plasma and red blood cell membranes; (2) reduced conversion of linoleic acid (LA) to AA and alpha-linolenic acid (ALA) to EPA/DHA; (3) altered AA/EPA ratio affecting the balance of pro-inflammatory vs pro-resolving mediator production. In ME/CFS, low EPA and DHA availability would reduce resolvin E1, D1 and protectin D1 synthesis, impairing neutrophil clearance, macrophage efferocytosis, and inflammatory resolution. The resulting failure-to-resolve could shift the inflammatory trajectory from self-limited to chronic.
ME/CFS context. If FADS polymorphisms contribute to SPM insufficiency in ME/CFS, they would predict: (a) differential response to omega-3 supplementation (high-activity allele carriers will benefit more because they have the enzymatic capacity to convert EPA/DHA to SPMs); (b) interaction with gut microbiome composition (dysbiosis-reduced bacterial SPM-precursor supply compounds low host conversion capacity); (c) enrichment of low-activity FADS haplotypes in the ME/CFS population, particularly in patients with prominent non-resolving inflammatory features.
Falsifiable predictions. (1) ME/CFS patients carrying the low-activity FADS haplotype (rs174537 GG) will show lower plasma AA, EPA, and DHA, and correspondingly lower SPM levels (resolvin D1, resolvin E1) compared to high-activity (TT) carriers. (2) Omega-3 supplementation (EPA 2 g + DHA 1 g/day, 12 weeks) will increase plasma SPM levels and reduce inflammatory markers (hs-CRP, IL-6) preferentially in FADS high-activity carriers. (3) ME/CFS patients with both unfavourable FADS genotype AND low gut Faecalibacterium abundance will show the lowest SPM levels and highest PEM severity — a gene-microbiome interaction. (4) FADS genotype will not affect NK cytotoxicity or complement activation (SPM-independent pathways), providing a specificity control.
Limitations. FADS polymorphisms have not been studied in any ME/CFS genetic cohort. The SPM resolution model assumes that SPM availability is rate-limiting for inflammatory resolution in ME/CFS — this has not been demonstrated. Omega-3 supplementation has been tested in ME/CFS with mixed results (Chapter Supplements and Nutraceuticals), but never stratified by FADS genotype or with SPM endpoints. The predicted gene-microbiome interaction requires large sample sizes (n > 500) for adequate statistical power.