Skip to content

Kerrebijn I, et al. The genetic architecture of fibromyalgia across 2.5 million individuals. Nat Med. 2026;32:3060-3070. PMID 42521817

One-paragraph summary

A multi-ancestry genome-wide association study meta-analysis was conducted across 2,563,755 individuals from 11 cohorts — 54,629 fibromyalgia (FM) cases and 2,509,126 controls — identifying 26 risk loci, the first robustly established for the condition. The strongest association was with a coding variant in HTT, the causal gene for Huntington's disease. Gene prioritisation implicated the HTT regulator GPR52 together with a set of genes with neural roles: DCC, DRD2/NCAM1, MDGA2 and CELF4 (with CAMKV additionally named in the preprint version, PMID 41001472). Partitioned heritability was enriched exclusively within brain tissues and neural cell types. FM showed strong positive genetic correlations across chronic pain, psychiatric and somatic disorders, exceeding 0.7 with low back pain, post-traumatic stress disorder and irritable bowel syndrome. Despite the large sex difference in FM prevalence, the genetic architecture was nearly identical between males and females. The authors position this as the first robust genetic evidence defining fibromyalgia as a central nervous system disorder.

Key findings

  • 26 genome-wide significant risk loci from 54,629 cases — versus zero loci from the previous largest single-biobank FM GWAS (Moscati 2023, PMID 37334860), which explicitly concluded that larger samples were required.
  • Heritability enrichment confined to brain tissues and neural cell types. This is a partitioning result, not a narrative inference: no peripheral tissue class showed enrichment.
  • Top signal in HTT, a coding variant, with the HTT-regulating orphan GPCR GPR52 prioritised — a druggable target with no prior mechanistic account in pain biology.
  • Convergence with chronic-pain GWAS. DCC was previously the single gene shared between male and female multisite chronic pain associations (Johnston 2021, PMID 33830993); it reappears here.
  • Genetic correlations rg > 0.7 with low back pain, PTSD and IBS — quantitative genetic support for the chronic overlapping pain conditions framework, and for the view that these are not independent diagnoses.
  • Sex-invariant genetic architecture despite a several-fold prevalence difference between women and men.

Limitations

  • Case definition varies across 11 cohorts. FM ascertainment in biobanks generally relies on diagnostic codes or self-report rather than prospectively applied ACR criteria; this introduces misclassification in both directions and mixes criteria eras (1990 / 2010–2011 / 2016).
  • Control definition. With 2.5 million controls, an appreciable number are undiagnosed cases — FM has substantial diagnostic delay and under-recognition — which biases effect estimates toward the null and inflates required sample size rather than creating false positives.
  • Multi-ancestry but likely European-weighted. Locus discovery and any derived polygenic score are not automatically portable across ancestries.
  • Common variants only. Twenty-six loci explain a small fraction of FM liability; rare-variant and structural contributions are unaddressed.
  • Loci are not mechanisms. HTT and GPR52 are association signals with no demonstrated causal pathway to nociplastic pain; gene prioritisation is statistical, not experimental.
  • Tension with adjacent phenotypes. The brain-exclusive heritability enrichment sits awkwardly against skeletal-muscle tissue specificity reported for chronic widespread musculoskeletal pain (Rahman 2021, PMID 33926923). Whether this reflects genuinely different phenotypes or methodological difference is unresolved.

Why it matters

This is the paper that gave fibromyalgia a genome, and it changes three arguments at once.

First, it converts "FM is a central nervous system disorder" from a mechanistic inference drawn from neuroimaging and quantitative sensory testing into a genetic finding. Heritability enrichment confined to brain tissue is a strong, hard-to-explain-away constraint, and it does so without depending on the interpretation of any single imaging or psychophysical study. This is directly relevant to the long-running legitimacy dispute documented in history and nosology and patient experience and advocacy.

Second, it sharpens rather than resolves the sex-ratio problem. If the genetic architecture is essentially identical in men and women while prevalence is not, then the sex ratio is driven by something non-genetic — ascertainment, criteria performance, hormonal or environmental factors. That makes the criteria-artefact hypothesis in diagnostic criteria a live, testable explanation rather than a rhetorical one.

Third, it supplies the field's first genetically supported druggable target. FM drug development has been repurposing-driven throughout its history — antidepressants, anticonvulsants, a muscle relaxant — with no target nominated by FM biology itself. GPR52 is an orphan GPCR already of interest as a Huntington's disease target, which means tool compounds and medicinal chemistry exist. Whether that translates is entirely unknown, but the pipeline described in clinical trials landscape currently contains nothing derived from FM-specific biology, and this is the first candidate that could be.

Cited by wiki pages

  • omics and emerging science
  • clinical-trials-landscape