🤖 AI Analysis This analysis was generated by Claude (Anthropic) using the Free Energy Principle (Friston) and Monoamine Systems frameworks (Robbins), synthesising 35 peer-reviewed open-access papers from PubMed Central (2014–2025). Not a substitute for medical advice.
2–4%
of population affected
6:1
female-to-male ratio
8 years
average diagnosis delay
30.7%
of COVID survivors develop FM criteria
25.6%
pain reduction at progesterone peak

What is Fibromyalgia?

Fibromyalgia (FM) is a chronic pain syndrome characterised by widespread musculoskeletal pain, fatigue, sleep disturbances, cognitive dysfunction ('fibro fog'), and heightened sensitivity to touch, light, and sound. It is classified in ICD-11 (code MG30.01) as 'chronic primary widespread pain' and as 'nociplastic pain' by the International Association for the Study of Pain (IASP, 2017). Unlike inflammatory arthritis or tissue-damage-driven pain, fibromyalgia arises not from injury to muscles or joints, but from a fundamental dysregulation of how the central nervous system processes pain signals.

The term 'nociplastic pain' — introduced by the IASP in 2017 — captures this distinction precisely. It refers to pain arising from altered nociception without clear evidence of tissue damage or nerve injury. In fibromyalgia, blood tests return normal, X-rays show no pathology, and nerve conduction studies are unremarkable. Yet the pain is entirely real: functional MRI studies consistently show abnormal activity in pain-processing brain regions, particularly the insula, anterior cingulate cortex, and prefrontal cortex. FM patients experience pain because their nervous system is generating it — not because something is broken in the periphery.

Fibromyalgia was formally recognised in modern medicine when the American College of Rheumatology (ACR) published its first diagnostic criteria in 1990, based on tender point examination. In 2010 and again in 2016, the ACR updated these criteria to remove the tender point requirement, replacing it with the Widespread Pain Index (WPI) and Symptom Severity Scale (SSS). This shift reflected a deeper scientific understanding: FM is not primarily a muscle disorder but a disorder of pain amplification in the central nervous system. Today, diagnosis is entirely clinical — based on symptom pattern and duration — and does not require laboratory tests or imaging.

One of the most significant challenges in fibromyalgia is the diagnostic delay. On average, patients wait 8 years from the onset of symptoms to receiving a confirmed diagnosis. This delay is not simply a failure of clinical recognition; it reflects the fact that fibromyalgia symptoms — fatigue, widespread pain, poor sleep, cognitive fog — are non-specific and overlap with dozens of other conditions. Patients are frequently dismissed, misdiagnosed with depression or hypochondria, or subjected to extensive workups that return normal results. This invalidation itself contributes to the psychological burden of the condition and can worsen central sensitisation by increasing stress, disrupting sleep, and activating the HPA axis.

Core Symptoms Explained

Widespread Pain

FM pain affects all four body quadrants for ≥3 months. It is typically deep, aching, burning, or throbbing and migrates between locations. Unlike joint pain, it is felt diffusely in muscles, tendons, and connective tissue. The mechanism is central sensitisation: the spinal cord and brain have become hyperexcitable, amplifying normal sensory signals into painful ones.

😴

Non-Restorative Sleep

FM patients wake unrefreshed regardless of sleep duration. Polysomnography reveals alpha-delta sleep intrusion: waking brainwaves disrupt deep restorative slow-wave sleep. Even 8–9 hours fails to restore energy. Crucially, sleep deprivation itself amplifies pain sensitivity in healthy people — creating a vicious cycle where pain disrupts sleep and sleep disruption worsens pain.

🧠

'Fibro Fog'

'Fibro fog' encompasses difficulties with working memory, word retrieval, concentration, and processing speed. It is a neurological — not psychiatric — symptom, linked to reduced norepinephrine activity in the prefrontal cortex. Brain imaging shows reduced grey matter density and abnormal connectivity. For many patients, cognitive disability is as debilitating as the pain.

😰

Fatigue

FM fatigue is a profound total exhaustion present regardless of activity level, not relieved by rest. It is driven by dopamine depletion in the mesolimbic reward system: dopamine normally provides the motivational push to initiate activity. When depleted, small tasks feel insurmountably effortful. This is why FM fatigue overlaps so deeply with ME/CFS and long COVID — all three involve the same dopaminergic pathways.

🌡️

Central Sensitisation

Central sensitisation is the defining pathophysiology of FM. The CNS becomes hypersensitive: light touch causes pain (allodynia), mild stimuli produce exaggerated pain (hyperalgesia). The sensitisation also affects auditory, visual, and olfactory processing — explaining light and sound sensitivity. Quantitative sensory testing (QST) consistently shows reduced pain thresholds across the entire body, confirming this is a global nervous system phenomenon, not localised pathology.

🦠

Associated Conditions

FM rarely exists alone. Most patients also experience IBS, migraines, restless legs, anxiety, and depression. These are mechanistically linked — not coincidental — all involving the same serotonin, norepinephrine, and dopamine dysregulation. The gut-brain axis is critical: 90–95% of the body's serotonin is produced in the gut, so gut microbiome dysbiosis directly impairs pain modulation.

Who Gets Fibromyalgia and Why?

Fibromyalgia affects an estimated 2–4% of the global population — roughly 200 million people worldwide. The condition shows a striking sex disparity: women are diagnosed approximately 6 times more often than men, though post-COVID research suggests this ratio may partly reflect different triggering mechanisms. Peak onset occurs between ages 30–55, overlapping precisely with perimenopause — not coincidentally, since estrogen and progesterone are powerful modulators of serotonin and GABA pain pathways.

Risk factors include female sex, middle age, a personal or family history of depression or anxiety, prior trauma, and prior infections. FM is frequently triggered by an identifiable event — viral illness, physical injury, extreme stress, or a hormonal transition. This 'trigger hypothesis' aligns with the central sensitisation model: a stressor activates the HPA axis, depletes monoamines, disrupts sleep, and primes the nervous system for amplified pain signalling. Once established, this state becomes self-sustaining even after the original trigger resolves.

The Science: Two Frameworks

Understanding fibromyalgia requires moving beyond the traditional biomedical model. Two complementary scientific frameworks provide the most coherent account: one from theoretical neuroscience (Karl J. Friston), one from pharmacology and behavioural neuroscience (Trevor W. Robbins). Together they explain FM at both the computational level (how the brain processes information) and the neurochemical level (which molecules go wrong and why).

🔮

Karl J. Friston

Free Energy Principle & Active Inference

Karl Friston's Free Energy Principle proposes that the brain's fundamental goal is to minimise the difference between what it predicts and what it actually receives from its senses. The brain is not a passive receiver — it is an active prediction machine that continuously generates a 'generative model' of the body and world. Sensory inputs are evaluated against prior predictions; the brain only updates its model when incoming signals cannot be explained away.

Pain, in this framework, is the brain's prediction that the body is being damaged — not a signal from injured tissue. In healthy people, this prediction is regularly corrected by sensory signals: the pain of innocuous stimuli resolves when sensory evidence confirms no actual damage occurred. The brain down-weights its pain predictions accordingly.

In fibromyalgia, this corrective mechanism has failed. The brain assigns excessive 'precision' — excessive confidence — to its pain predictions, refusing to update them in the face of contradictory sensory evidence. The result is a self-reinforcing loop: pain predictions persist and amplify, and every attempt by the nervous system to resolve them is suppressed. This is central sensitisation, reframed in computational terms.

The descending pain inhibitory pathway — running from the brain through the periaqueductal grey (PAG) into the spinal dorsal horn — is the biological mechanism the brain uses to suppress ascending pain signals it deems uninformative. In FM, this pathway is hypoactive. Serotonin and norepinephrine (its key neurotransmitters) are depleted, impairing the brain's ability to 'turn down the volume' on pain. Active inference predicts this: if the brain is overconfident in pain predictions, it won't engage descending inhibition, since doing so would mean accepting that those pain signals carry useful corrective information.

Sleep, in the Friston framework, is when the brain recalibrates its generative model — consolidating prediction errors and updating priors. Non-restorative sleep in FM means this recalibration fails nightly: pathological pain priors are reinforced rather than corrected. Stress (elevated cortisol, HPA axis dysrhythmia) further reduces the brain's capacity for flexible updating, locking the generative model into a state of chronic pain prediction.

  • Central sensitisation = elevated precision on pain signals (brain trusts predictions over sensory evidence)
  • Descending inhibition failure = brain suppresses correction signals rather than updating its model
  • Sleep disruption = failed nightly recalibration, pathological priors reinforced
  • Gut-brain axis = interoceptive signals cannot be resolved against brain priors
  • Treatment implication: therapies that update the generative model (exercise, CBT, sleep improvement) are mechanistically superior to symptom masking
⚗️

Trevor W. Robbins

Monoamine Systems & Descending Pain Control

Trevor Robbins' work on monoamine neurotransmitter systems provides the neurochemical complement to Friston's computational framework. While Friston explains the information-processing failure in FM, Robbins identifies the molecular machinery through which that failure occurs. These three monoamines — serotonin, norepinephrine, and dopamine — are not merely mood chemicals; they are the fundamental regulators of pain modulation, arousal, cognition, and motivation.

Serotonin (5-HT) is produced in the raphe nuclei of the brainstem. Descending serotonergic fibres project into the spinal dorsal horn, gating which pain messages reach consciousness. In FM, cerebrospinal fluid serotonin and its metabolite 5-HIAA are consistently reduced. Serum tryptophan (serotonin's precursor) is also lower than controls. The consequence: pain signals that would normally be suppressed reach the brain unfiltered.

Norepinephrine (NE) originates in the locus coeruleus — the brain's arousal and attention nucleus. NE plays two distinct roles in FM: (1) descending noradrenergic pathways synergise with serotonergic pathways in dorsal horn inhibition — when both are depleted, the inhibitory deficit compounds; (2) prefrontal NE regulates working memory and concentration via alpha-2 receptors. Prefrontal NE depletion produces exactly the 'fibro fog' picture: impaired working memory, poor concentration, slowed processing. This is why SNRIs — boosting both serotonin and NE simultaneously — are among the most effective FM pharmacological treatments.

Dopamine drives the fatigue and anhedonia of FM. Produced in the VTA and substantia nigra, dopamine provides the 'wanting' and motivational drive for effortful action. In FM, dopamine receptor binding is reduced in the caudate nucleus and putamen — critical for movement initiation and reward processing. This explains why FM fatigue is not simply tiredness but a fundamental impairment of motivational capacity: the brain lacks the dopaminergic signal that makes effort feel worthwhile.

Serotonin (5-HT)
Raphe nucleus → Dorsal horn → Cortex
FM: ↓ CSF 5-HT, ↓ 5-HIAA → pain gate fails, sleep disrupted, mood impaired
Norepinephrine (NE)
Locus coeruleus → Dorsal horn + Prefrontal cortex
FM: ↓ NE → descending inhibition fails + fibro fog (α2-receptor)
Dopamine (DA)
VTA / Substantia nigra → Nucleus accumbens + Basal ganglia
FM: ↓ DA receptor binding in caudate/putamen → fatigue, anhedonia, impaired motivation

The critical insight: these three deficits are not independent — they are deeply interrelated. Serotonin promotes dopamine release; norepinephrine modulates serotonin in the raphe nuclei; dopamine influences norepinephrine synthesis. A deficit in one cascades through the others. This is why FM is a multisystem condition — and why treatments targeting all three monoamines simultaneously (exercise, which acutely increases all three; or SNRIs combined with sleep optimisation) outperform single-target approaches.

Post-COVID / Long COVID and Fibromyalgia

30.7%

of COVID-19 survivors met ACR fibromyalgia criteria at 6 months post-infection (Ursini et al., 2021, N=616)

FibroCOVID: A New Clinical Entity

In 2021, Italian rheumatologist Giacomo Ursini and colleagues published a landmark study examining 616 COVID-19 survivors at 6 months post-infection. They found that 30.7% — nearly 1 in 3 — met ACR 2016 fibromyalgia criteria. They coined the term 'FibroCOVID' for this post-COVID fibromyalgia syndrome, characterising it as a distinct manifestation within the post-COVID-19 spectrum. This was not pre-existing FM being revealed: the criteria were specifically not met before infection.

One of the most scientifically important findings was the unexpected sex distribution. Classical FM affects women at 6:1 compared to men. But in the FibroCOVID cohort, 43% of cases were male — dramatically higher than classical FM. This confirms that COVID-19 disease severity — not female hormonal vulnerability — is the driving trigger for post-COVID FM. Male vulnerability here is explained by higher rates of severe COVID-19 in men (ACE2 expression differences, testosterone-driven immunosuppression, metabolic comorbidities), not inherent male FM susceptibility.

The strongest predictors of FibroCOVID were obesity (OR 82.82; 95% CI 32.19–213.08) and male sex (OR 9.95; 95% CI 6.02–16.43). BMI correlated with FM severity at R=0.763 (p<0.0001). The researchers proposed that obesity facilitates the COVID→FM transition via adipose-tissue-driven chronic low-grade inflammation, priming the nervous system for central sensitisation — consistent with obesity being a risk factor for classical FM via adipose-derived cytokines (leptin, adiponectin, TNF-α) that cross the blood-brain barrier and modulate neuroinflammatory tone.

Predictor Odds Ratio 95% CI
Male gender 9.95 6.02–16.43
Obesity 82.82 32.19–213.08
BMI ↔ FM severity R=0.763 p<0.0001

The Biological Mechanisms: How COVID Triggers FM

SARS-CoV-2 triggers FM through at least six converging pathological pathways, each attacking a different component of the pain regulation system. This explains why post-COVID FM persists long after viral clearance.

1. Direct CNS neuroinvasion: SARS-CoV-2 enters the CNS via ACE2 receptors on neurons and glial cells, directly infecting brainstem nuclei and triggering microglial activation. Activated microglia produce pro-inflammatory cytokines (IL-1β, IL-6, TNF-α) that lower pain thresholds and disrupt descending inhibitory pathways. Post-mortem studies confirm SARS-CoV-2 RNA in CNS tissue with persistent neuroinflammation months after acute infection.

2. Fascial myofibroblast activation (Plaut 2023): COVID-19's cytokine storm activates fascial myofibroblasts — contractile cells embedded in connective tissue throughout the body. This elevates intramuscular pressure from 12.23 mmHg to 33.48 mmHg — a threefold increase that directly stimulates nociceptors throughout the musculoskeletal system, producing the diffuse deep burning muscle pain characteristic of FM.

3. Mitochondrial dysfunction & oxidative stress: COVID-19 impairs mitochondrial function directly (viral proteins interfere with the electron transport chain) and indirectly (hypoxia). Mitochondrial dysfunction in sensory neurons impairs serotonin and norepinephrine production and reduces the neuronal capacity to maintain normal membrane potentials, further lowering pain thresholds. Elevated oxidative stress markers are consistently found in both COVID-19 and FM.

4. Gut microbiome dysbiosis: SARS-CoV-2 extensively infects the GI tract, disrupting the gut microbiome — the primary source of tryptophan (serotonin precursor) and short-chain fatty acids that modulate CNS function. Post-COVID dysbiosis depletes serotonin precursors, increases intestinal permeability, and allows bacterial endotoxins into systemic circulation, triggering low-grade inflammation that amplifies pain sensitisation. Gut microbiome abnormalities persist ≥6 months after COVID-19 recovery.

Shared Mechanisms: Long COVID = Fibromyalgia?

In 2023, Clauw & Calabrese published a landmark paper in Annals of the Rheumatic Diseases drawing explicit parallels between long COVID and fibromyalgia. Their central argument: after excluding patients with identifiable organ damage, the remaining long COVID patients present with a symptom profile indistinguishable from FM — and show the same neuroimaging abnormalities (increased insula and anterior cingulate cortex activity), the same autonomic dysfunction (heart rate variability abnormalities, orthostatic intolerance), and the same pharmacological response profile (benefit from SNRIs and gabapentinoids; no benefit from opioids or NSAIDs).

Goldenberg (2025), reviewing the latest evidence in Expert Review of Neurology, went further: proposing that long COVID, fibromyalgia, and ME/CFS are three clinical presentations of one underlying disorder — central sensitisation syndrome. This has direct treatment implications: the entire FM evidence base becomes applicable to long COVID, and decades of FM pathophysiology research provides a ready-made framework for treating the post-COVID epidemic of chronic fatigue and pain.

Feature Long COVID Fibromyalgia
Widespread pain
Post-exertional malaise
Brain fog / cognitive dysfunction
Non-restorative sleep
Neuroinflammation (neuroimaging)
Central sensitisation (QST)
Autonomic dysfunction (HRV)
Normal standard workup
SNRIs / TCAs effective
Opioids ineffective / contraindicated

The Friston View: Collapsed Inference After COVID

From Friston's Free Energy Principle perspective, COVID-19 is a catastrophic assault on the brain's generative model. During acute infection, the virus floods the body with unprecedented conflicting interoceptive signals — fever oscillations, hypoxia, severe fatigue, widespread pain, GI distress, and neuroinflammation — all simultaneously. The brain's prediction machinery is overwhelmed; it cannot generate a coherent prior. The generative model destabilises.

As the acute phase resolves, the brain rebuilds its generative model — but the template it has available is the one it just learned: 'my body is sick, painful, and exhausted.' This pathological prior becomes the new baseline. Even when the virus is cleared, the brain continues predicting pain and fatigue, because that is what recent experience taught it to expect. From a computational perspective, this is a hyperprior of illness: a deeply entrenched belief about the body's state that requires enormous updating evidence to revise. Without targeted therapeutic interventions — CBT to update catastrophic beliefs, graded exercise to provide corrective sensory evidence of function, sleep improvement for nightly recalibration — the hyperprior of illness persists indefinitely.

Menopause and Fibromyalgia

25.6%

pain reduction when progesterone peaks — lost entirely after menopause (Schertzinger et al., 2018)

The Hormonal Pain Connection

The epidemiology of FM presents a long-standing puzzle: why does it affect women 6× more often than men, and why does onset peak precisely in the 30–55 year range? The answer lies in the profound role that sex hormones — estrogen, progesterone, and testosterone — play in regulating pain thresholds, serotonin synthesis, and CNS excitability. These are not merely reproductive chemicals; they are powerful neuroactive compounds that modulate the very systems that fail in FM.

A landmark study by Carranza-Lira & Villalobos (2014) made this concrete: 29% of premenopausal women with climacteric symptoms (hot flashes, sleep disturbance, mood changes) also met FM criteria, compared to just 4% without climacteric symptoms — a sevenfold difference. The perimenopause represents the single most significant FM onset risk window in the female lifespan, because it is characterised by maximum hormonal instability — not simply declining levels.

Progesterone
Progesterone's painkilling action works via its metabolite allopregnanolone — a potent GABA-A receptor modulator that reduces neuronal excitability throughout the CNS. Schertzinger et al. (2018): 25.6% pain reduction at peak progesterone (mid-luteal phase, days 18–22). Entirely lost at menopause.
Lost at menopause → GABA inhibition collapses → pain amplification
Testosterone
Testosterone suppresses TNF-α (a pain-sensitising cytokine) and down-regulates TRPV1 ('capsaicin receptors' detecting noxious stimuli). Women lose ~50% of testosterone between ages 20–50, accelerating at menopause. Lower testosterone → higher FM symptom severity across multiple studies.
Progressive decline with age → neuroinflammation increases, TRPV1-mediated pain sensitivity rises
Estrogen
Estrogen's relationship with pain is paradoxical. Stable estrogen is protective — it upregulates serotonin synthesis (via tryptophan hydroxylase-2) and increases 5-HT2A receptor density. But rapidly fluctuating estrogen (perimenopause) is pro-nociceptive — cycling amplifies sensitisation via mast cell activation and substance P release. Hence the paradox: perimenopausal women (fluctuating) are at higher FM risk than postmenopausal women (stable low).
Maximum fluctuation (perimenopause) = maximum FM risk; stable low (postmenopause) = risk plateau

The Estrogen-Serotonin-Pain Axis: A Molecular Explanation

The estrogen-serotonin-pain axis is the central mechanistic pathway linking menopause to FM. Estrogen directly stimulates transcription of tryptophan hydroxylase-2 (TPH2) — the rate-limiting enzyme for brain serotonin synthesis. When estrogen declines at menopause: TPH2 activity falls → raphe nucleus serotonin production decreases → the descending serotonergic pain inhibitory pathway weakens → less inhibitory control over ascending pain signals → more pain reaches the brain.

Beyond synthesis, estrogen also regulates serotonin receptor density. Estrogen maintains high 5-HT2A receptor density in the prefrontal cortex and limbic system — critical for mood regulation and pain modulation. After menopause, 5-HT2A density falls, impairing serotonin signalling even at the same synaptic serotonin concentrations. Women undergoing surgical menopause (bilateral oophorectomy) show even more dramatic FM risk increases than natural menopause, because the abrupt estrogen withdrawal leaves no time for neural adaptation.

The Kynurenine Pathway: Tryptophan Hijacked

Tryptophan (serotonin's amino acid precursor) can be metabolised through two competing pathways. The serotonin pathway (preferred when estrogen is present) produces serotonin via TPH2. The kynurenine pathway (activated by inflammation and estrogen withdrawal) instead converts tryptophan into kynurenine → quinolinic acid. Quinolinic acid is an endogenous NMDA receptor agonist — the same receptors involved in central sensitisation and the induction of long-term potentiation in pain pathways.

At menopause, estrogen withdrawal simultaneously suppresses the serotonin pathway AND activates the kynurenine pathway — a double hit: serotonin availability falls (pain inhibition weakens) AND quinolinic acid accumulates (NMDA-driven central sensitisation is induced). Lang et al. (2025) demonstrated that menopausal women have elevated kynurenine-to-tryptophan ratios, correlating with reduced BDNF (brain-derived neurotrophic factor). Lower BDNF means reduced capacity for the nervous system to 'recover' from sensitisation — explaining why post-menopausal FM is often particularly persistent.

Friston: Menopause as Generative Model Disruption

Throughout reproductive life, the brain's generative model has been calibrated on the predictable rhythms of the menstrual cycle — regular fluctuations in progesterone, estrogen, and testosterone that the brain has learned to anticipate. Pain thresholds rise and fall predictably; mood and sleep follow established patterns. This cyclical predictability is a fundamental component of the body's interoceptive model.

At perimenopause, this predictability collapses. Hormone levels fluctuate unpredictably — estrogen can rise and fall tenfold within a single cycle; progesterone surges become irregular. The brain suddenly receives interoceptive signals — hot flashes, heart palpitations, mood shifts, fragmented sleep — that its existing generative model cannot predict. From Friston's perspective: massive, unresolvable prediction errors. These demand costly recalibration of the entire interoceptive model — and if this recalibration fails (when monoamines are already depleted, sleep is disrupted, stress is high), the nervous system defaults to elevated precision weighting on pain signals: central sensitisation, and FM onset.

GPER: Receptor-Level Dysfunction

A 2019 study by Koca et al. uncovered a finding that challenges the simple 'low estrogen → FM' model. FM patients had elevated GPER (G-protein coupled estrogen receptor) — 0.11 vs 0.059 ng/mL in healthy controls (p=0.037) — despite normal circulating estrogen. GPER is a membrane-bound receptor that mediates pro-nociceptive effects when overactivated: it promotes neuroinflammation, lowers pain thresholds, and amplifies sensory signalling. Elevated GPER in FM suggests the problem is not simply estrogen deficiency but estrogen receptor dysfunction — the estrogen signal is being misprocessed at the receptor level even when hormone levels are normal.

Hormone Therapy Evidence: What Does It Mean for FM?

Given the clear biological mechanisms linking estrogen withdrawal to FM pathophysiology, the question of whether hormone replacement therapy (HRT) benefits fibromyalgia is clinically important. The evidence is encouraging, with significant caveats.

Study Intervention FM Outcome
Dias et al. 2023 (N=32) 12-week transdermal estradiol + oral micronised progesterone 30% FIQ-R decline (p=0.0001); significant pain, fatigue, and sleep improvement
Athnaiel et al. 2023 (systematic review) Steady-state HRT (various formulations) ~50% pain perception reduction across studies
Schertzinger et al. 2018 Daily hormonal tracking across menstrual cycle in FM patients 25.6% pain reduction at progesterone peak; testosterone negatively correlated with daily pain

Key Insight: Stability Over Level

The key principle: it is not the absolute estrogen level that matters — it is stability of delivery. Transdermal estradiol (patches, gels, rings) delivers at a consistent rate, avoiding the sharp peaks and troughs of oral preparations. Oral estrogen undergoes first-pass hepatic metabolism, producing variable plasma concentrations — precisely the fluctuations that amplify FM pain. This explains why perimenopause (maximum fluctuation) is more dangerous for FM than postmenopause (low but stable), and why the route of HRT administration has direct clinical significance.

Treatment: What Works and Why

FM treatment is fundamentally different from nociceptive or inflammatory pain treatment. Because FM is a CNS disorder — not a peripheral tissue disorder — treatments targeting peripheral inflammation (NSAIDs, corticosteroids) or blocking nociceptive signals (opioids) are ineffective or harmful. Effective treatment must address the three underlying dysfunctions: the maladaptive generative model (Friston), the depleted monoamine systems (Robbins), and the disrupted sleep-pain cycle. The most effective approach is therefore multimodal — combining non-pharmacological interventions, pharmacological monoamine restoration, and when applicable, hormonal therapy.

Non-Pharmacological (First-Line)

Strong evidence

Graded Aerobic Exercise

Exercise is the single most evidence-based FM treatment — superior to any drug in the long term. Cochrane review: 13 RCTs, 839 participants, significant reductions in pain, fatigue, and mood; benefits last 24–208 weeks. Aquatic exercise is especially effective (buoyancy reduces nociceptive load while maintaining cardiovascular stimulus). Mechanism: acute aerobic exercise releases serotonin, norepinephrine, and dopamine simultaneously — directly restoring all three depleted monoamine systems. Sustained programmes also upregulate BDNF, increase pain thresholds, and recalibrate the generative model. Critical clinical note: must be graded (start very low, increase slowly) and paced — aggressive exercise causes post-exertional flares that worsen FM.

Strong evidence

Cognitive Behavioural Therapy (CBT)

CBT has the strongest evidence of any psychological FM intervention. In Friston's terms, CBT is the closest approximation to directly reprogramming the maladaptive generative model — targeting pain catastrophising (which elevates precision weighting on pain predictions), sleep hygiene, activity pacing, and unhelpful illness beliefs. The mechanism is neurobiological: catastrophising activates the amygdala, elevates cortisol, and depletes serotonin. Breaking the catastrophising cycle reduces amygdala hyperactivity, lowers cortisol, and allows monoamine partial recovery. Especially powerful when combined with exercise — together they address both the computational (CBT) and neurochemical (exercise) components of FM.

Moderate evidence

Tai Chi / Yoga / Qi Gong

Mind-body practices converge physical exercise (monoamine release) with cognitive training (mindfulness addresses attentional bias toward pain signals). Tai Chi meta-analysis (657 patients): significant improvements in pain, sleep, fatigue, and physical function. Yoga reduces pain catastrophising and perceived stress — both lower pain amplification. The group setting additionally provides social connection, which has analgesic properties via oxytocin release. Especially important for patients who cannot tolerate vigorous exercise.

Moderate evidence

Acupuncture + Massage

Network meta-analysis of CAM treatments for FM (41 studies, N=2,877): the combination of acupuncture + massage had a 99.2% probability of reducing the FIQ score — highest of any CAM combination. Acupuncture modulates pain via endogenous opioid release (enkephalins, endorphins) and normalises descending inhibitory pathways. Massage reduces substance P levels at tender points and promotes parasympathetic activity, lowering stress-mediated sensitisation. Neither alone matches the combination, suggesting additive or synergistic mechanisms.

Pharmacological

FDA Approved

Duloxetine (SNRI)

Duloxetine (60 mg/day) is the most widely used FDA-approved FM drug. As an SNRI, it simultaneously blocks reuptake of both serotonin and norepinephrine, directly addressing the two key monoamine deficits driving descending inhibition failure and fibro fog. Additional finding: one study showed duloxetine use was associated with significantly reduced COVID-19 incidence — via anti-inflammatory and possible antiviral properties — making it a particularly rational choice for post-COVID FM. Onset of benefit: 4–8 weeks. Side effects: nausea (usually transient), insomnia, dry mouth, sexual dysfunction.

FDA Approved

Milnacipran (SNRI)

Milnacipran (50–100 mg twice daily) is the second FDA-approved SNRI for FM. Its pharmacological profile differs importantly from duloxetine: milnacipran has ~3:1 norepinephrine:serotonin selectivity, making it particularly useful when dominant symptoms are fatigue and cognitive dysfunction (primarily NE-deficit driven) rather than mood. Twice-daily dosing provides more stable plasma concentrations than once-daily duloxetine. Approved in Europe (as Ixel) for depression as well as FM.

FDA Approved

Pregabalin (Gabapentinoid)

Pregabalin (150–450 mg/day) is the only FDA-approved FM drug targeting a non-monoamine mechanism. It binds to the alpha-2-delta subunit of voltage-gated calcium channels in the spinal dorsal horn, reducing calcium influx and decreasing release of excitatory neurotransmitters (glutamate, substance P, CGRP) at presynaptic terminals — directly reducing central sensitisation. NNT: 4–14 depending on outcome measure. Also improves sleep quality (secondary analgesic benefit via improved generative model recalibration). Shows positive effects in severe COVID-19 cases, suggesting utility for post-COVID FM. Side effects: dizziness, sedation, weight gain, peripheral oedema.

Most evidence of all FM drugs

Amitriptyline (TCA, low dose)

Despite being off-patent, amitriptyline (10–25 mg nightly) has the largest evidence base of any single FM medication — 36.5% global FM status improvement. Multiple mechanisms explain its broad utility: inhibits serotonin + NE reuptake (addresses both monoamine systems); blocks sodium channels (reduces neuronal hyperexcitability); antagonises H1 histamine receptors (promotes deep sleep, directly addressing non-restorative sleep and nightly recalibration failure); low-dose NMDA receptor blockade (reduces central sensitisation). The sedative properties at FM doses are a therapeutic benefit — improved sleep architecture directly reduces next-day pain.

For Menopause + FM

Clinical evidence

Hormone Replacement Therapy (HRT)

For perimenopausal or postmenopausal FM patients, HRT addresses the root hormonal cause. Dias et al. (2023): 30% FIQ-R reduction after 12 weeks of transdermal estradiol + oral micronised progesterone. Athnaiel et al. (2023) systematic review: ~50% pain perception reduction with steady-state HRT. Key principles: (1) Transdermal over oral — stable concentrations avoid pain-amplifying fluctuations; (2) Micronised progesterone over synthetic progestins — it metabolises to allopregnanolone (direct GABA-A analgesic benefit), which synthetic progestins do not; (3) Consistent dosing over cyclical — avoid the pain-amplifying fluctuations that characterise perimenopause. HRT also restores estrogen's anti-inflammatory effects on IL-6 and TNF-α, addressing the inflammaging component.

What Doesn't Work — and Why

Contraindicated

Opioids (pure mu-agonists)

Opioids are not merely unhelpful — they are actively harmful and contraindicated in all major FM guidelines. The mechanism: in FM, pain is generated by a maladaptive generative model with excessive precision on pain predictions. Opioids temporarily reduce pain sensitivity but do so by downregulating endogenous opioid receptors (tolerance) and impairing the brain's own endogenous pain modulation system. Long-term opioid use in FM causes opioid-induced hyperalgesia — a paradoxical pain sensitivity increase — directly worsening central sensitisation. Opioids also reduce slow-wave and REM sleep (worsening nightly recalibration failure), impair dopaminergic function (worsening fatigue), and disrupt the gut microbiome. Multiple studies show significantly worse FM outcomes on long-term opioids versus non-opioid regimens.

No evidence

NSAIDs / Corticosteroids

NSAIDs and corticosteroids target peripheral inflammation — and FM has no peripheral tissue inflammation. CRP, ESR, synovial fluid, and imaging are all normal in FM. NSAIDs reduce prostaglandin synthesis from peripheral tissues, contributing nothing to the central sensitisation mechanism. Multiple RCTs have failed to show NSAIDs outperforming placebo for FM pain. Worse, long-term NSAID use impairs gut microbiome (potentially worsening gut-brain axis dysfunction) and adds cardiovascular and GI risk with no compensating benefit.

Convergence: The Double Vulnerability

The most important clinical insight from integrating the post-COVID and menopause research is the concept of convergent vulnerability: these two conditions amplify FM risk through the same biological pathways. Understanding this convergence identifies the highest-risk populations and supports early, proactive, multimodal intervention before FM becomes entrenched.

Post-COVID

  • CNS neuroinvasion → microglial activation
  • Cytokine storm → neuroinflammation
  • HPA dysregulation → cortisol dysrhythmia
  • Serotonin, NE, DA depletion
  • Gut dysbiosis → tryptophan depletion
  • Sleep disruption → failed recalibration

Four Pillars of FM

1. Neuroinflammation
2. HPA Dysregulation
3. Monoamine Depletion
4. Sleep Disruption
FIBROMYALGIA

Menopause

  • Estrogen withdrawal → serotonin ↓ → descending inhibition fails
  • Progesterone loss → allopregnanolone ↓ → GABA ↓
  • Kynurenine pathway → quinolinic acid → NMDA sensitisation
  • Inflammaging (IL-1, IL-6, IL-17A, TNF-α)
  • HPA dysregulation → cortisol dysrhythmia
  • Sleep disruption via vasomotor symptoms

The Double Vulnerability: When Both Occur Together

Consider a perimenopausal woman — aged 45–55, already experiencing hormonal instability — who contracts moderate-to-severe COVID-19. In this individual, both pathological pathways are simultaneously active, and they are not merely additive; they are synergistic.

On the hormonal side: her estrogen is fluctuating maximally, serotonin synthesis is already compromised, progesterone is declining (GABA inhibitory tone falling), and the kynurenine pathway is already partially activated. Her nervous system is primed for central sensitisation — hovering just below the FM clinical threshold. Then COVID-19 hits. The viral cytokine storm triggers microglial activation, further depleting serotonin and norepinephrine, directly inducing neuroinflammation, disrupting the gut microbiome (further reducing tryptophan availability), dysregulating the HPA axis, and fragmenting sleep. Both pathways now reinforce all four FM pillars simultaneously.

In Friston's terms: the perimenopausal brain is already struggling with an unstable generative model (it cannot predict its own hormonal state). COVID-19 then catastrophically floods this already-destabilised model with overwhelming conflicting interoceptive signals. The model collapses into a pathological hyperprior of illness — establishing the conditions for persistent, treatment-resistant FM. Epidemiological data from the COVID-19 pandemic confirms this risk: higher rates of long COVID and post-COVID FM in perimenopausal women compared to both younger women and men of the same age.

The practical implication: perimenopausal women who contract COVID-19 should be proactively monitored for FM development at 3, 6, and 12 months post-infection. If post-COVID FM criteria are met, early multimodal therapy — transdermal HRT (where appropriate), SNRIs, graded exercise, CBT, and sleep optimisation — offers the best chance of preventing chronic entrenchment of the sensitised state. Waiting for 'spontaneous recovery' is likely to allow the pathological generative model to consolidate, making subsequent treatment progressively less effective.

Treatment Convergence: One Table, Three Conditions

Because FM, long COVID, and menopausal FM share the same four pathological pillars, their treatments converge to a remarkable degree. This convergence is clinically significant: 30 years of FM research directly informs the treatment of the long COVID epidemic.

Treatment FM Long COVID Menopausal FM Mechanism of shared benefit
SNRIs (duloxetine, milnacipran)✓✓✓✓✓✓✓✓Restore serotonin + NE descending inhibition; address fibro fog (prefrontal NE)
TCAs (amitriptyline 10–25 mg)✓✓✓✓✓✓✓Serotonin/NE reuptake + H1 blockade (deep sleep) + NMDA blockade (central sensitisation)
Pregabalin/Gabapentin✓✓✓✓✓✓✓α2-δ Ca²⁺ channel blockade → reduce excitatory neurotransmitter release → reduce central sensitisation
CBT✓✓✓✓✓✓✓✓✓Update maladaptive generative model; reduce pain catastrophising; reduce HPA activation
Graded aerobic exercise✓✓✓✓ (paced)✓✓✓Restore all three monoamines simultaneously; upregulate BDNF; provide corrective sensory evidence for the generative model
HRT (transdermal estradiol + micronised progesterone)✓✓✓Restore estrogen-serotonin-TPH2 axis; restore progesterone-allopregnanolone-GABA-A analgesia; reduce inflammaging; stabilise generative model
Sleep optimisation (TCA + sleep hygiene + CBT-I)✓✓✓✓✓✓✓✓✓Restore slow-wave sleep → enable nightly generative model recalibration → break pain-sleep vicious cycle

Sources & References

All documents from peer-reviewed open-access literature (PubMed Central / NIH).