The treatments that work in fibromyalgia are unglamorous, slow, and mostly not drugs — with one qualification added in August 2026, in section 7. That is a disappointing sentence to read and it is the accurate one — and the genetics have made it more defensible rather than less, because every effective treatment turns out to be aimed at the system the genetics implicate.
The UK guideline this chapter follows. Fibromyalgia falls within
NICE NG193, on chronic primary pain in over-16s, and the shape of that guideline is
worth knowing before the detail: exercise and psychological therapy are the first-line
recommendations, antidepressants are offered as an option, and a list of drug classes is specifically
not to be started — opioids, NSAIDs, paracetamol, benzodiazepines,
corticosteroids and antiepileptics including the gabapentinoids. That is unusual and
deliberate: a pain guideline whose central recommendations are not medicines.
The single most useful framing. The target is not the painful
area. It is the system that decides how loud the signal gets — which is why treatments
aimed at sleep, activity and nervous system regulation outperform anything aimed at the muscles that
hurt, and why scans and injections into painful sites so reliably disappoint.
1. Movement — first line in the guideline, and hardest in practice
- Graded exercise has the best evidence of anything here, across aerobic activity, strength work and aquatic exercise, with benefits on pain, function, fatigue and mood. NG193 recommends a supervised exercise programme as a core treatment — it is not the thing you do because the drugs did not work.
- And the commonest reason it fails is starting too high. A good day leads to doing too much, which causes a flare, which teaches that exercise makes things worse. Start below what feels achievable and increase slowly enough that it feels almost pointless — that is the technique, not a lack of ambition.
- Pacing rather than pushing. Consistent moderate activity across the week beats alternating between overdoing it and recovering.
- Warm water helps many people start, because it removes load while allowing movement.
- Expect weeks to months, not days. This is the intervention that most rewards persistence and most punishes impatience.
2. Sleep, treated as a target rather than a symptom
Unrefreshing sleep is not merely a consequence of the pain — the relationship runs both ways, and poor sleep measurably lowers pain thresholds in healthy volunteers. Treating sleep often improves pain more than treating pain improves sleep.
- Cognitive behavioural therapy for insomnia (CBT-I) is the first-line treatment, and is more effective and more durable than sleeping tablets. Available through the NHS and through digital programmes.
- Screen for obstructive sleep apnoea and restless legs, both of which are commoner here and both of which are treatable in their own right.
- The sleep chapter covers the general principles.
3. Psychological therapy — also first line, and the offer most often refused
This is the offer people most often refuse, for an understandable reason
- Being sent to psychology after years of being disbelieved reads as confirmation that nobody thinks the pain is real. That reaction is reasonable given the history, and clinicians who do not anticipate it will lose the patient.
- But it is not offered because the pain is imagined. It is offered because the nervous system's regulation of pain is modifiable by learning, attention and behaviour — which is a statement about neurophysiology, not about character.
- NG193 recommends CBT or acceptance and commitment therapy specifically, alongside the exercise programme, as core treatment rather than as an adjunct. Effects are modest and real, comparable to the drugs and with fewer costs.
- Pain reprocessing and similar newer approaches are promising and less established. Emerging
- And depression and anxiety are common alongside — partly as consequence, partly shared biology. Treating them is worth doing on its own terms and is not the same as treating the fibromyalgia.
4. Medicines
All of these are modest in effect. A treatment that helps a third of people meaningfully is a good treatment in this condition, and that is roughly the ceiling.
| Drug | Position |
|---|---|
| Amitriptyline, low dose | The class NG193 does offer — antidepressants are the one drug group the guideline supports considering, all of them off-label for this. Usually the first tried. Doses far below antidepressant doses, taken at night, often helping sleep and pain together. Anticholinergic effects — dry mouth, constipation, morning grogginess — are the usual limit. Used off-label for this |
| Duloxetine | An SNRI with reasonable evidence in fibromyalgia, useful where low mood coexists. Nausea early on is common and usually settles |
| Pregabalin and gabapentin the row where UK and US practice diverge |
Trial evidence exists, and pregabalin is licensed for fibromyalgia in the United States. It is not the UK position. NG193 advises against starting antiepileptics, including the gabapentinoids, for chronic primary pain, so a GP following the guideline will not usually initiate one — and that is a considered judgement about benefit against harm rather than an oversight. Sedation, weight gain and dizziness are common, both are Class C controlled drugs in the UK, and the dependence and withdrawal problems were underestimated for years. If you are already taking one, that is a different question. The guidance is to review whether it is helping and to discuss reducing or stopping if it is not — not to stop abruptly, which risks withdrawal. That conversation belongs with your prescriber |
| Opioids | Should not be used. They do not work well in nociplastic pain, they cause dependence, and they can produce opioid-induced hyperalgesia — making pain worse by the mechanism that is already the problem. Coming off them, done slowly and with support, frequently improves pain rather than worsening it |
| NSAIDs and paracetamol | Generally disappointing here, because there is no inflammation to treat. Useful for a coexisting nociceptive problem |
| Steroids | No role. Their ineffectiveness is itself informative about the mechanism |
| Cannabis-based products | Contested. Some people report benefit; trial evidence is weak and inconsistent, and the UK prescribing route is narrow |
5. What is being tested, and what is being sold
- Non-invasive brain stimulation and low-intensity focused ultrasound aimed at deep pain-processing regions are under investigation. Investigational, not approved, not available in clinic — but they are the same shift the genetics represent: treating fibromyalgia as a brain condition and targeting it directly. Worth knowing about; not worth pursuing privately yet.
- Supplements have not delivered. Vitamin D correction matters if you are deficient, as it does for anyone. Magnesium, CoQ10 and the rest have thin and inconsistent evidence. See supplements.
- Leaky gut, adrenal fatigue and detox protocols are not mechanisms this condition has. They are sold into the space the health system left empty, which is a real failing — but an expensive answer is not better than an honest one.
6. What exists elsewhere — the first new drug in fifteen years
Approved in the United States on 15 August 2025. Not available in the UK, in any
form. Tonmya (cyclobenzaprine hydrochloride) comes as 2.8 mg sublingual tablets — one
nightly for the first two weeks, then two, giving 5.6 mg. The maximum is
2.8 mg in older adults and in mild liver impairment. It is taken at bedtime and aimed at the non-restorative sleep that characterises
fibromyalgia, rather than at the pain directly. It is the first new FDA approval for
fibromyalgia in over fifteen years, and only the fourth ever — after pregabalin (2007),
duloxetine (2008) and milnacipran (2009). Checked 15 August 2026: no UK licence, no MHRA
authorisation, and NICE has not been asked to look at it. NG193 is unchanged.
What it is, stated accurately, because the marketing is doing work
- It is not a new molecule. Cyclobenzaprine is a muscle relaxant that has been in use in the United States since 1977. What is new is the formulation — a low dose held under the tongue, which is absorbed directly and bypasses the liver, producing less of the long-lived metabolite that makes the ordinary oral form sedating the next day.
- The evidence is two positive Phase 3 trials, RELIEF and RESILIENT, in about 1,000 people, both showing a statistically significant reduction in daily pain against placebo at 14 weeks. That is a real result. The effect sizes are modest, and in the same range as the existing options rather than beyond them.
- There was a third Phase 3 trial, and it missed. RALLY randomised 514 people to the same regimen and failed its primary endpoint — an interim analysis in July 2021 led the independent monitoring committee to recommend stopping it as unlikely to succeed. It appears in the label's efficacy section alongside the two that worked. Two positive trials out of three is the honest description, and this site names the null one for the same reason it names JELIS: a drug's evidence base is the trials that were run, not the trials that read well.
- "First-line treatment" is the manufacturer's phrase, not a guideline position, and this page will not adopt it. No guideline anywhere has placed it first line, and NG193's ordering — exercise and psychological therapy as core treatment — is untouched by a drug approval in another country.
- Cyclobenzaprine has never been marketed in the UK in any form, so there is no equivalent product and no off-label route. Milnacipran, the third of the four, was never UK-marketed either — which is worth knowing, because it means the American and British drug options for this condition have differed for fifteen years, not only since last August.
- And it is a tricyclic, with the cautions that implies. Contraindicated with monoamine oxidase inhibitors and within 14 days of stopping one, with a serotonin syndrome warning alongside other serotonergic drugs; and contraindicated in recent myocardial infarction, arrhythmias, heart block and heart failure. There is an embryofetal warning, so contraception advice applies. None of that is exotic for the class — it is the reason a fifty-year-old muscle relaxant is not simply prescribed off-label.
- What to do about it here: nothing, yet. It is worth knowing exists, and worth raising with a specialist if sleep is the dominant problem — because targeting sleep is a strategy available to you now, through CBT-I and the approaches in section 2, whatever happens to this particular drug.
And the qualification promised at the top of this chapter. The statement that the effective treatments are mostly not drugs remains true of what is available in the UK, and remains the shape of NG193. It is no longer true of the global landscape, and a page that said so without this section would be describing 2024.
7. Getting the system to work for you
- Ask for the diagnosis to be recorded. A named diagnosis on the record changes how the next clinician reads the notes, and its absence is why people are re-investigated for years.
- Ask specifically whether this has been considered as nociplastic pain, and whether there is a plan covering activity, sleep, psychological therapy and medication. That is the question that produces a plan rather than a test.
- New or different symptoms still need assessing. Having fibromyalgia does not protect you from anything else, and diagnostic overshadowing — attributing every new symptom to the existing label — is a documented and dangerous pattern.
- Flares are part of the pattern rather than evidence of failure. Having a written plan for them, made when well, is worth more than deciding what to do during one.
And the realistic expectation. The aim is meaningful improvement in function and pain, not cure. Most people improve with a combined approach and few become symptom-free, and the treatments that produce most of that improvement are the ones requiring the most patience. Anybody promising more than that is selling something.