Skip to content

Pathophysiology — peripheral mechanisms

TL;DR — Since 2013, multiple laboratories have found objective small-fibre pathology (SFP) in a substantial minority-to-half of fibromyalgia (FM) patients: 41% of distal-leg skin biopsies diagnostic for small-fibre polyneuropathy versus 3% of controls in one series (Oaklander 2013, PMID 23748113), 63% with reduced intraepidermal nerve fibre density (IENFD) in a 117-woman cohort (Evdokimov 2019, PMID 31376174), and a pooled prevalence of 49% (95% CI 38–60%) across 8 studies and 222 participants — 45% by skin biopsy, 59% by corneal confocal microscopy (Grayston 2019, PMID 30314675). Microneurography shows abnormal, spontaneously active or mechanically sensitized silent C-nociceptors in 76.6% of FM patients (Serra 2014, PMID 24243538). What this means is unsettled: FM small-fibre changes differ morphologically from classical small-fibre neuropathy (SFN) (Doppler 2015, PMID 26164586), do not correlate with pain intensity in every series (Kosmidis 2014, PMID 25304055), and do not organize sensory phenotype in FM the way they do in idiopathic distal sensory polyneuropathy (Burgess 2024, PMID 39264538). Meanwhile, peripheral anaesthetic blockade demonstrably reduces both local and remote hyperalgesia (Staud 2009, PMID 19540671), so peripheral input is at minimum a maintaining factor. The historical muscle-pathology and muscle-ischaemia hypotheses were tested and largely abandoned (Simms 1996, PMID 8860798).

1. The small-fibre line of evidence

1.1 Founding studies (2013)

Study Design Key result
Üçeyler 2013, PMID 23474848 25 FM, 10 patients with monopolar depression without pain, matched healthy controls; QST, pain-related evoked potentials (PREP), skin punch biopsy at lower leg and upper thigh Normal neurological and standard neurophysiological examination (large-fibre neuropathy excluded). FM had increased cold and warm detection thresholds vs controls (p < 0.001) — not seen in the depression group. PREP: increased N1 latency at feet (p < 0.001), reduced amplitudes at face, hands and feet (p < 0.001). Total (p < 0.001) and regenerating (p < 0.01) intraepidermal fibres reduced at leg and thigh; dermal unmyelinated fibre bundles reduced; myelinated fibres spared
Oaklander 2013, PMID 23748113 27 patients meeting ACR 2010 criteria plus a physician's FM diagnosis; 30 matched controls; MNSI, UENS, distal-leg skin biopsy, autonomic function testing 41% of FM biopsies diagnostic for SFPN vs 3% of controls; MNSI and UENS scores higher (all p ≤ 0.001). Abnormal autonomic testing equally prevalent in both groups → FM-associated SFPN appears primarily somatic. Aetiology screen: all glucose tolerance tests normal; 8 subjects had dysimmune markers, 2 had hepatitis C serologies, 1 family showed apparent genetic causality

Üçeyler framed the results as "pointing towards a neuropathic nature of pain"; Oaklander framed them as evidence that some illnesses labelled FM are in fact unrecognized SFPN — a distinct, sometimes treatable disease. The two framings are not the same claim, and the difference persists in the field.

1.2 Replications, proportions and heterogeneity

Study n Proportion with reduced innervation Notes
Kosmidis 2014, PMID 25304055 46 FM (41 women), 34 controls 15/46 = 32.6% with reduced IENFD (mean 4.83 ± 2.5 fibers/mm in the reduced subset vs 7.35 ± 1.85 in controls, p < 0.0001) No correlation between Neuropathic Pain Symptom Inventory scores and IENFD. No difference in epidermal Langerhans cells or IL-6 staining. Controls were considerably younger (mean 31.7 vs 52.5 y) and differently sexed (16/34 male) — a real confound
Doppler 2015, PMID 26164586 32 FM, 12 SFN, 24 controls Confirmed reduced IENFD and abnormal QST Electron microscopy: mean unmyelinated axon diameter reduced in FM vs both SFN and controls (p < 0.05), i.e. FM small-fibre changes are morphologically distinguishable from SFN — "different pathomechanisms may lead to small fiber loss in the 2 disorders"
Evdokimov 2019, PMID 31376174 117 women with FM vs major depression with chronic widespread pain (MD-P) and healthy women 63% with reduced IENFD vs 10% (MD-P) and 18% (healthy), p < 0.001 Four distinct innervation patterns (normal; distally reduced; proximally reduced; both). Generalized IENFD reduction associated with higher pain intensity, greater impairment, more stabbing pain and paraesthesias, more anxiety (p < 0.05 each), plus lower corneal nerve fibre density (p < 0.01) and length (p < 0.05)
Malik 2019, PMID 30801480 22 FM, two-site skin biopsy 5/22 positive for IgM anti-TS-HDS antibodies; 4 of those 5 had non-length-dependent SFN (p = 0.0393) TS-HDS positivity rate did not differ from a control database (p = 0.41); trend only toward more non-length-dependent SFN in FM (p = 0.06). Small and exploratory

1.3 Meta-analytic estimate

The single pooled estimate available is Grayston 2019 (PMID 30314675): systematic search of six databases, 8 eligible full-text studies, 222 participants.

  • Overall pooled prevalence of small-fibre pathology in FM: 49% (95% CI 38–60%), I² = 68%
  • By skin biopsy: 45% (95% CI 32–59%), I² = 70%
  • By corneal confocal microscopy: 59% (95% CI 40–78%), I² = 51%

The authors conclude this is "compelling evidence of a distinct phenotype involving SFP in fibromyalgia". The counterweight is the size of the evidence base — 222 participants across 8 studies, moderate-to-high heterogeneity, no chronic-pain comparison arm in the pooled analysis.

1.4 Nociceptor function: microneurography

Morphology aside, single-fibre recordings show functional abnormality. In 30 female FM patients, 17 women with SFN and 9 female controls, 186 / 114 / 66 C-nociceptors were recorded. Mechanosensitive nociceptors in FM behaved normally, but silent (mechano-insensitive) nociceptors were abnormal in 76.6% of FM patients: spontaneous activity in 31% of silent nociceptors (SFN 34%, controls 2.2%) and mechanical sensitization in 24.2% (SFN 22.7%, controls 3.7%). Abnormally high activity-dependent slowing of conduction velocity at 0.25 Hz was more common in FM than in SFN and was proposed as a distinguishing feature (Serra 2014, PMID 24243538).

Large-diameter fibres may also be involved. Single-fibre recordings in people with FM found that a larger proportion of Aβ low-threshold mechanoreceptors responded to cold stimulation than in controls, although — contrary to the matching mouse experiments — those human fibres showed reduced responses to mechanical stimuli (Israel 2025, PMID 40898777). If confirmed, this widens the peripheral account beyond small fibres and would help explain the tactile allodynia, paraesthesia and dysaesthesia that the small-fibre account alone does not.

1.5 Corneal confocal microscopy (CCM)

CCM images the densest small-fibre innervation in the body non-invasively. In 17 women with FM vs 17 age-matched controls, corneal stromal nerve thickness was 5.0 ± 1.0 µm vs 6.1 ± 1.3 µm (p = 0.01) and sub-basal plexus nerve density 85 ± 29 vs 107 ± 26 /mm² (p = 0.02); stromal-nerve slenderness was associated with neuropathic pain descriptors, but only with patients and controls analysed together (Fisher's exact p = 0.007) (Ramírez 2015, PMID 26094164). Corneal nerve loss in FM is also documented within the larger Würzburg cohort (Evdokimov 2019, PMID 31376174), and CCM contributes the higher of the two pooled prevalence estimates (59%) (Grayston 2019, PMID 30314675). CCM has been proposed as a surrogate endpoint for small-fibre degeneration across neurological disease, with FM listed among the conditions where its diagnostic utility is comparable to IENFD (Petropoulos 2021, PMID 35295436); ocular involvement in FM more broadly (dry eye, reduced corneal sensitivity, thinner stromal nerves) is reviewed in Zdebik 2021 (PMID 32512032).

2. What SFN findings do — and do not — imply

Four non-exclusive interpretations are live.

(a) Cause. If degenerating nociceptors generate ectopic activity, that could drive widespread pain. Support: microneurographic spontaneous activity and mechanical sensitization (Serra 2014, PMID 24243538); association of generalized IENFD loss with greater pain and impairment in the largest cohort (Evdokimov 2019, PMID 31376174).

(b) Consequence or epiphenomenon. Support: the passive-transfer work shows loss of intraepidermal innervation is induced in mice by patient IgG, i.e. denervation can be downstream of a circulating factor rather than a primary event (Goebel 2021, PMID 34196305 — see autoimmunity-and-inflammation). IENFD did not correlate with neuropathic pain scores in one 46-patient series (Kosmidis 2014, PMID 25304055).

(c) Comorbidity / distinct disease. Support: FM small-fibre morphology differs from SFN (reduced axon diameter; Doppler 2015, PMID 26164586) and FM shows a different activity-dependent slowing signature (Serra 2014, PMID 24243538) — arguing FM-SFP is not simply "SFN plus an FM label".

(d) Misdiagnosis. Oaklander's explicit claim: some patients carrying an FM label have SFPN, which — unlike FM — can be tested for objectively and sometimes treated definitively once a cause (diabetes, dysimmunity, hepatitis C, genetic sodium-channel disease) is found (Oaklander 2013, PMID 23748113). A separate consecutive series from the same centre — all 39 adults with a physician's FM diagnosis plus ACR 2010 criteria who were tested for SFPN, 14 positive and 25 negative — found that pain severity did not differ between the two groups; the dysautonomia component subscore of the Small-fiber Symptom Survey did (10.42 ± 4.0 vs 7.16 ± 4.0, p = 0.019), as did paraesthesias, each with a receiver-operating-characteristic AUC of 0.729 — i.e. symptoms discriminate only fairly, and no secondary questionnaire discriminated at all (Lodahl 2017, PMID 29430562).

Three findings actively constrain interpretation (a):

  1. Sensory phenotype does not track SFP in FM. Comparing 25 FM, 23 idiopathic distal sensory polyneuropathy (IDSP) and 25 healthy volunteers, FM showed gain of function to heat and blunt pressure and higher wind-up ratio, IDSP showed loss of function; by corneal nerve fibre length, SFP was present in 13/25 FM and 22/23 IDSP. Crucially, sensory phenotypes were associated with the presence of SFP in IDSP but not in FM (Burgess 2024, PMID 39264538).
  2. Cortical pain-evoked responses do not follow the denervation. In 22 FM patients with proximal denervation, 18 with normal skin biopsy and 7 with proximal and distal IENFD reduction, laser-evoked potential amplitude and topography were not coherent with epidermal nerve fibre loss, and instead tracked pain expression and psychopathological factors (Vecchio 2022, PMID 35074721).
  3. Clinical discrimination remains hard. A practical review lists the features that lean toward SFN (stabbing pain, paraesthesias, dry eyes/mouth, sweating and skin-colour changes, thermal hypoaesthesia, length-dependent distribution, metabolic/Sjögren/sarcoid/HIV associations) versus FM (headache, temporomandibular disorder, post-traumatic stress, childhood abuse history), while conceding substantial overlap (Bailly 2021, PMID 34082128).

3. Muscle: what was tested and what survived

The "fibrositis" era assumed a muscle lesion. It did not survive controlled testing. A systematic review of the muscle literature concluded (Simms 1996, PMID 8860798):

  • Early morphological abnormalities came from studies with poor patient selection and inadequate controls; later studies showed only non-specific or mild changes.
  • MR-spectroscopy studies failed to confirm abnormal muscle metabolism, at tender and non-tender sites alike; earlier apparent abnormalities were confounded by deconditioning.
  • Muscle blood-flow abnormalities were likewise explicable by deconditioning; the microcirculatory-ischaemia/sympathetic-dysfunction hypothesis was not supported.
  • EMG showed no excessive muscle tension and no defective sympathetic function; strength differences were attributable to submaximal voluntary effort.
  • Conclusion: "muscle tenderness in fibromyalgia cannot be explained on the basis of primary muscle abnormalities" — structural or functional — and future pathophysiologic work should focus on central mechanisms.

That conclusion is the historical hinge from peripheral to central models. A later hypothesis relocated the peripheral lesion from muscle to fascia (intramuscular connective tissue), proposing fibroblast-derived IL-6 and a dysfunctional healing response as the nociceptive drive (Liptan 2010, PMID 20006283). This is an explicitly hypothetical paper published in a bodywork journal, not an evidence base; no confirmatory controlled data were located in this session.

4. Peripheral input maintains central sensitization: the local-anaesthetic experiments

Two Staud studies are the cleanest bridge between the peripheral and central accounts.

Study Design Result
Staud 2009, PMID 19540671 RCT, double-blind, placebo-controlled; 28 female FM, 22 female controls; 1% lidocaine vs placebo injected into trapezius tender points, with shoulder pain standardized by tonic mechanical stimulation Lidocaine (not placebo) raised trapezius pressure-pain thresholds in both groups (p < 0.001); in FM it also reduced remote heat hyperalgesia at the forearm (p = 0.02). No significant effect on clinical FM pain — attributed by the authors to the very low dose (50 mg)
Staud 2014, PMID 24193993 62 female FM randomized to 100 mg lidocaine, 200 mg lidocaine, or saline into both trapezius and gluteal muscles Primary mechanical hyperalgesia at shoulders and buttocks decreased more after lidocaine than saline (p = 0.004); secondary heat hyperalgesia at the arms likewise (p = 0.04). Clinical FM pain fell 38% but did not differ between lidocaine and saline; placebo-related factors (expectation of relief) accounted for 19.9% of the variance in post-injection clinical pain

Read together: blocking peripheral impulse input reliably reduces evoked hyperalgesia, locally and remotely, which is direct evidence that ongoing peripheral input sustains the sensitized state. It does not show that peripheral input is the source of spontaneous clinical pain — in both trials the effect on clinical pain was either absent or indistinguishable from placebo.

5. Is "fibromyalgia with SFN" a distinct subgroup?

Arguments for:

  • Split prevalence: pooled 49% of patients have measurable SFP (Grayston 2019, PMID 30314675), and 14/39 (36%) in a consecutive test-based series whose authors put the figure at ~40% (Lodahl 2017, PMID 29430562) — roughly half do not.
  • Within FM, generalized IENFD reduction marks a more severe phenotype (more pain, more impairment, more stabbing pain and paraesthesias, more anxiety, lower corneal nerve density) (Evdokimov 2019, PMID 31376174).
  • Discriminating symptoms exist (dysautonomia items, paraesthesias) with AUC ≈ 0.73 (Lodahl 2017, PMID 29430562).
  • Some SFP is associated with candidate autoantibodies (anti-TS-HDS), pointing to a potentially treatable dysimmune subtype (Malik 2019, PMID 30801480; Oaklander 2013, PMID 23748113).

Arguments against, or at least unresolved:

  • SFP status does not organize FM sensory phenotype the way it organizes IDSP phenotype (Burgess 2024, PMID 39264538).
  • SFP status does not track cortical pain-evoked responses (Vecchio 2022, PMID 35074721) and, in at least one series, does not track pain severity at all (Kosmidis 2014, PMID 25304055; Lodahl 2017, PMID 29430562).
  • FM small-fibre morphology differs from SFN morphology, so "FM with SFN" may be a category error — it may be FM with a different small-fibre change (Doppler 2015, PMID 26164586).
  • Denervation can be induced by patient IgG in mice, making it plausibly a marker of the disease process rather than a separate disease (Goebel 2021, PMID 34196305).
  • No prospective study located here shows that SFP status predicts treatment response — the practical test of a real subgroup.

The IASP nociplastic framework accommodates both poles: it lists autoreactive antibodies acting at the dorsal root ganglion alongside aberrant cerebral pain processing as FM mechanisms, and treats peripheral and central sensitization as typical but non-specific features (Kosek 2024, PMID 39560415). Reviews from the central camp now describe FM as a continuum from purely peripherally driven to purely centrally driven pain (Sluka 2016, PMID 27291641).

Open questions

  • What is the true prevalence of small-fibre pathology in FM, and against what comparator? The pooled 49% (95% CI 38–60%) rests on 8 studies and 222 participants with I² = 68% and no chronic-pain control arm (Grayston 2019, PMID 30314675); individual series range from 32.6% (Kosmidis 2014, PMID 25304055) to 63% (Evdokimov 2019, PMID 31376174).
  • Is small-fibre loss cause or consequence? Patient IgG produces loss of intraepidermal innervation in mice (Goebel 2021, PMID 34196305), which would make denervation downstream; but ectopic silent-nociceptor activity is present in 76.6% of patients (Serra 2014, PMID 24243538), which would make it upstream. No longitudinal human study resolving the order was located.
  • Does SFP status define a clinically actionable subgroup? It marks severity in one cohort (Evdokimov 2019, PMID 31376174) but does not organize sensory phenotype (Burgess 2024, PMID 39264538), cortical responses (Vecchio 2022, PMID 35074721) or, in some series, pain severity (Lodahl 2017, PMID 29430562).
  • How much of the "FM" population is actually misdiagnosed treatable SFPN? Oaklander found dysimmune markers in 8, hepatitis C serologies in 2, and apparent genetic causality in 1 family among FM patients with diagnostic biopsies (PMID 23748113); no systematic aetiological yield study in a large FM cohort was located.
  • Why does peripheral blockade reduce evoked hyperalgesia but not clinical pain more than placebo? Both lidocaine trials show this dissociation (Staud 2009, PMID 19540671; Staud 2014, PMID 24193993), and the 2014 trial attributes 19.9% of clinical-pain variance to expectancy.
  • Is the fascia hypothesis testable? It has been proposed (Liptan 2010, PMID 20006283) but no controlled confirmation was located; meanwhile the muscle hypotheses it descends from were rejected on MRS, EMG and blood-flow grounds (Simms 1996, PMID 8860798).

References

  1. Üçeyler N, Zeller D, Kahn AK, et al. Small fibre pathology in patients with fibromyalgia syndrome. Brain. 2013;136(Pt 6):1857-67. PMID 23474848
  2. Oaklander AL, Herzog ZD, Downs HM, Klein MM. Objective evidence that small-fiber polyneuropathy underlies some illnesses currently labeled as fibromyalgia. Pain. 2013;154:2310-2316. PMID 23748113
  3. Grayston R, Czanner G, Elhadd K, et al. A systematic review and meta-analysis of the prevalence of small fiber pathology in fibromyalgia: implications for a new paradigm in fibromyalgia etiopathogenesis. Semin Arthritis Rheum. 2019;48:933-940. PMID 30314675
  4. Kosmidis ML, Koutsogeorgopoulou L, Alexopoulos H, et al. Reduction of intraepidermal nerve fiber density (IENFD) in the skin biopsies of patients with fibromyalgia: a controlled study. J Neurol Sci. 2014;347:143-7. PMID 25304055
  5. Doppler K, Rittner HL, Deckart M, Sommer C. Reduced dermal nerve fiber diameter in skin biopsies of patients with fibromyalgia. Pain. 2015;156:2319-2325. PMID 26164586
  6. Evdokimov D, Frank J, Klitsch A, et al. Reduction of skin innervation is associated with a severe fibromyalgia phenotype. Ann Neurol. 2019;86:504-516. PMID 31376174
  7. Serra J, Collado A, Solà R, et al. Hyperexcitable C nociceptors in fibromyalgia. Ann Neurol. 2014;75:196-208. PMID 24243538
  8. Ramírez M, Martínez-Martínez LA, Hernández-Quintela E, Velazco-Casapía J, Vargas A, Martínez-Lavín M. Small fiber neuropathy in women with fibromyalgia. An in vivo assessment using corneal confocal bio-microscopy. Semin Arthritis Rheum. 2015;45:214-9. PMID 26094164
  9. Burgess J, Marshall A, Rapteas L, et al. Idiopathic distal sensory polyneuropathy and fibromyalgia syndrome: a comparative phenotyping study. Pain Ther. 2024;13:1541-1558. PMID 39264538
  10. Lodahl M, Treister R, Oaklander AL. Specific symptoms may discriminate between fibromyalgia patients with vs without objective test evidence of small-fiber polyneuropathy. Pain Rep. 2017;3:e633. PMID 29430562
  11. Vecchio E, Quitadamo SG, Ricci K, et al. Laser evoked potentials in fibromyalgia with peripheral small fiber involvement. Clin Neurophysiol. 2022;135:96-106. PMID 35074721
  12. Malik A, Lopate G, Hayat G, et al. Prevalence of axonal sensory neuropathy with IgM binding to trisulfated heparin disaccharide in patients with fibromyalgia. J Clin Neuromuscul Dis. 2019;20:103-110. PMID 30801480
  13. Bailly F. The challenge of differentiating fibromyalgia from small-fiber neuropathy in clinical practice. Joint Bone Spine. 2021;88:105232. PMID 34082128
  14. Staud R, Nagel S, Robinson ME, Price DD. Enhanced central pain processing of fibromyalgia patients is maintained by muscle afferent input: a randomized, double-blind, placebo-controlled study. Pain. 2009;145:96-104. PMID 19540671
  15. Staud R, Weyl EE, Bartley E, Price DD, Robinson ME. Analgesic and anti-hyperalgesic effects of muscle injections with lidocaine or saline in patients with fibromyalgia syndrome. Eur J Pain. 2014;18:803-12. PMID 24193993
  16. Simms RW. Is there muscle pathology in fibromyalgia syndrome? Rheum Dis Clin North Am. 1996;22:245-66. PMID 8860798
  17. Liptan GL. Fascia: a missing link in our understanding of the pathology of fibromyalgia. J Bodyw Mov Ther. 2010;14:3-12. PMID 20006283
  18. Goebel A, Krock E, Gentry C, et al. Passive transfer of fibromyalgia symptoms from patients to mice. J Clin Invest. 2021;131:e144201. PMID 34196305
  19. Petropoulos IN, Bitirgen G, Ferdousi M, et al. Corneal confocal microscopy to image small nerve fiber degeneration: ophthalmology meets neurology. Front Pain Res (Lausanne). 2021;2:725363. PMID 35295436
  20. Zdebik N, Zdebik A, Bogusławska J, Przeździecka-Dołyk J, Turno-Kręcicka A. Fibromyalgia syndrome and the eye — a review. Surv Ophthalmol. 2021;66:132-137. PMID 32512032
  21. Kosek E. The concept of nociplastic pain — where to from here? Pain. 2024;165(11S):S50-S57. PMID 39560415
  22. Sluka KA, Clauw DJ. Neurobiology of fibromyalgia and chronic widespread pain. Neuroscience. 2016;338:114-129. PMID 27291641
  23. Israel MR, Berwick R, Vastani N, et al. Aβ low threshold mechanoreceptors contribute to sensory abnormalities in fibromyalgia. Brain. 2025;148:4016-4029. PMID 40898777