Fibromyalgia: Exercise and Pacing

What Works

Cinematic anatomical illustration representing progressive loading and muscular adaptation

1. Aerobic exercise, the backbone High

2–3× per week · 25–40 min per session · 100+ min per week · start low, build over 6–12 weeks

Walking, swimming, pool work, cycling. Choose the one you will actually keep doing.

Pain reduced across 17 trials (MD −0.49, 95% CI −0.90 to −0.08). Backed by 167 trials covering 11,012 people, where every exercise type reduced pain except stretching.

2. Keep going for at least 13 weeks High

This is the parameter most often broken in practice. Trials that measured benefit ran this long. A six-week block stops before the window opens.

3. Use an adequate dose, not a minimal one High

Across 19 trials and 857 people, programmes meeting proper guideline targets beat under-dosed ones on pain, sleep, fatigue and overall condition. The only thing dose did not change was mood.

4. Resistance training, the long-term anchor Moderate

1–2 sets · 4–20 reps · 2× per week · 8–12 weeks · moderately hard weight

Ranked the single best option beyond three months in a network analysis of 51 trials. The researchers rate the improvements as clinically meaningful but the certainty of the evidence as low.

5. Water-based exercise, the entry point Moderate

Ranked the best option within the first three months. Useful when symptoms or fear of movement make land-based work too much to start with.

6. Supervision and a low starting intensity Moderate

Across 89 trials, having an expert such as a physical therapist supervising was the strongest single factor protecting against people dropping out. Starting at lower intensity was the next strongest.

More options (moderate and emerging evidence)

Combined aerobic and strength training Moderate
Produced the largest effect on low mood of any format.

Mind-body work: Pilates, qigong, tai chi Moderate
Best supported where fatigue rather than pain is the dominant complaint.

Virtual reality and exercise gaming Emerging
Had the lowest drop-out rate of any format tested. Reach for it when sticking with it, rather than the exercise itself, is the problem.

Activity pacing Emerging
An add-on, and only for the boom-and-bust pattern. See The Nuance below for where this genuinely helps and where it is oversold.

Pain education alongside exercise Emerging
Understanding why the pain is amplified makes movement less frightening. An addition to exercise, not a replacement.

What does not work

Stretching on its own as the whole plan. Across 167 trials it was the single exercise type that failed to reduce pain, and it lost head-to-head against both aerobic and strength work.
Pacing as a stand-alone, first-line treatment. The evidence behind it is far thinner than its popularity suggests.
Cutting how often you train when you flare. Frequency and duration did not predict who dropped out. Intensity did.
Six-week programmes. They finish before the point at which benefit was measured.
Building a programme around a vibration plate. The entire evidence base is four small trials from one country.

Exercise Prescription

Exercise	How to do it	Sets × reps	Frequency
Walking or swimming	Steady, easy pace where you could still hold a conversation	25–40 minutes	2–3× per week
Sit-to-stand	Stand up from a chair without using your hands, sit back down slowly	1–2 × 8–12	2× per week
Wall or counter push-up	Hands on a wall or worktop, lower your chest toward it, push back	1–2 × 8–12	2× per week
Supported step-up	Step up onto a low step holding a rail, step down slowly	1–2 × 8–12 per leg	2× per week
Band or dumbbell row	Pull a band or light weight toward your ribs, squeeze the shoulder blades	1–2 × 8–12	2× per week

Pick a weight that feels moderately hard but leaves you a few repetitions in reserve. This mirrors the dose that worked in the trials.

The flare rule

When symptoms flare, reduce the intensity. Do not reduce how often or how long you train. In the research, intensity predicted who dropped out. Frequency and duration predicted nothing. The instinct to train less often targets the two things that were not the problem.

Red Flags — Get Checked First

Widespread pain has look-alikes, and some of them are urgent. See a doctor before starting any programme if you have:

New pain and stiffness in both shoulders or both hips, if you are over 50. If that comes with headaches, scalp tenderness, jaw pain when chewing, or any change in your vision, treat it as a same-day emergency. It can threaten your sight permanently.

Unexplained weight loss, fevers, or night sweats
Swollen joints, or morning stiffness lasting hours
Numbness, tingling or weakness that is getting worse
Chest pain or fainting during exertion
Back pain that started before you were 45 and eases with movement rather than rest

Cinematic anatomical illustration representing systemic warning signs

This page is written for people who already have a fibromyalgia diagnosis. If your widespread pain has not been properly assessed yet, that assessment comes first. Nothing here replaces it.

Progress Checkpoints

Fibromyalgia is a long-term condition rather than an injury with a finish line, so these are progress markers, not a discharge test.

Completing your sessions for four weeks straight without abandoning during a flare
Over 100 minutes a week of aerobic work, held for four weeks
Strength work twice a week, held for four weeks
Able to step the intensity up without a multi-day crash
One objective measure improved from where you started: steps per day, fitness, or strength
You can state the flare rule correctly without looking it up
The programme has run at least 13 weeks before you judge whether it worked

How Confident Are We?

Overall: Moderate-High

The spread between claims here is unusually wide, so it is worth seeing them separately rather than as one number.

Exercise reduces pain	High
Exercise is tolerable, drop-out matches do-nothing groups	High
Stretching alone is the weakest option	Mod-High
Supervision and intensity drive whether people stick with it	Mod-High
Strength training gains are clinically meaningful	Moderate
Water first, strength later	Moderate
A precise optimal weekly amount exists	Low
Pacing on its own	Low
Diagnostic criteria and meaningful-change thresholds	Not retrieved

What would change our mind on pacing

A fibromyalgia-specific trial of at least 200 people that tests pacing on its own, separated from the therapy, mindfulness and ergonomics it is usually bundled with, against exercise alone and against both combined, split by whether people are boom-and-bust or activity-avoidant, followed for six months. A positive result in the boom-and-bust group would move pacing from an add-on to a co-primary treatment. A null result would justify dropping the word from routine advice.

What would change our mind on the ideal dose

A trial of at least 300 people randomising session length (25–40 minutes against 60–90 minutes) crossed with frequency (twice against three times a week) at matched weekly totals, supervised, running at least 16 weeks, with pain, overall impact and drop-out all measured. If longer and less frequent won at matched volume, the current two-to-three-times-a-week default would be wrong.

Go Deeper

Tired of being told to rest, stretch and pace yourself with nothing to back it up? The Verdict reviews the actual evidence behind one health claim every week, free.

Join The Verdict — free

The Full Picture — Anatomy, Diagnosis & Evidence

What's Actually Going On

Cinematic anatomical illustration of the nervous system and pain signalling

Fibromyalgia sits at the amplified-processing end of pain. The pain is produced by altered signal processing rather than by ongoing tissue injury. That single fact drives everything: there is no healing tissue to protect, so there is no biological reason to hold back load, and the processing is turned up, so the same load genuinely feels worse.

Two things are true at once, and most people only believe one of them. The loss of fitness is real and measurable. Across 99 studies and 9,853 people, those with fibromyalgia had lower aerobic fitness and lower muscle function than healthy controls. And the effort signal is turned up on top of it, with a heightened perception of effort and a reduced ability to switch on the muscle that is available.

Where people actually start is the part that gets missed. Only 37.7% meet the 150-minutes-a-week activity target. The average is 5,664 steps a day against a 6,000 reference, and 39 minutes a day less activity than matched people without the condition. So the prescription is almost never "do less". It is "build something that was never there in the first place".

How to Identify It

Cinematic anatomical illustration representing clinical assessment

There is no physical test that diagnoses fibromyalgia No validated test. This evidence review retrieved no diagnostic-accuracy data at all, so no test-accuracy figures appear anywhere on this page Data not retrieved. Anyone quoting them to you should be able to show you the source.

Diagnosis is clinical, and it is largely a process of ruling out the look-alikes. What points toward fibromyalgia:

Widespread, symmetrical, persistent pain, above and below the waist
Fatigue that is not proportional to activity, and unrefreshing sleep
Often low mood and difficulty concentrating
An absence of inflammatory, systemic or one-sided features

What points away from it: restricted joint range, joint swelling, morning stiffness lasting hours, raised inflammation on blood tests, or weight loss and fevers. Those belong to other conditions.

Cinematic anatomical illustration representing differential diagnosis

An honest gap. No verified clinical practice guideline was retrieved for this condition during this review. Five guideline-level references surfaced by the automated literature scan could not be traced to a real source and were removed rather than quoted. Everything on this page rests on pooled trial evidence instead. That is a limitation worth knowing about, and it is why no formal diagnostic criteria are reproduced here.

The Debate

Is exercise too hard on people with fibromyalgia?

Common clinical convention

Exercise is poorly tolerated, so the dose has to be kept minimal and built around avoiding flares.

vs

89 trials, 3,702 people, 2024

Drop-out was 19.2%, statistically the same as the comparison groups. What predicted dropping out was high intensity and the absence of expert supervision. How often and how long people trained predicted nothing.

The lever is intensity and support, not volume. Cut how hard, keep how often.

Is pacing a core treatment?

Standard advice

Pacing is one of the first things people with fibromyalgia are told to do.

vs

PACE trial secondary analysis, 2014

The largest trial ever to test pacing as its own therapy found it came out behind graded exercise on both muscle pain and joint pain. That trial was in chronic fatigue syndrome rather than fibromyalgia, and its methods are contested.

Not proof that pacing is useless. It is proof that the confidence behind the standard advice has not been earned. The fibromyalgia-specific evidence is a 31-person trial where pacing was bundled with four other treatments.

Is less actually more on dose?

A widely repeated 2025 claim

Longer sessions, fewer times a week, under 180 minutes a week total, beats standard advice.

vs

Two verified analyses

Programmes meeting proper guideline targets beat under-dosed ones, and the dose review lands on 2–3 times a week, 25–40 minutes, over 100 minutes a week.

The "less is more" source could not be verified against any primary record and was removed from this review. Until it can be, do not use it as a reason to under-prescribe.

Honest Limitations

Supervision is doing invisible work

What the research showed: continuous expert supervision was the single strongest factor protecting against drop-out, and the trials that produced these results used supervised sessions of 50 to 60 minutes.

The real-world gap: most people are handed a leaflet and a six-week review. The trial results were never produced under those conditions.

What to do about it: treat contact and structure as an active ingredient rather than a nicety. A group class, a training partner, remote check-ins, or exercise gaming can all carry some of what supervision provides.

The evidence is overwhelmingly from women

What the research showed: the dosing reviews, the network analysis and the activity data are female-only or almost entirely female by design, tracking who the condition affects most.

The real-world gap: every specific number on this page is an extrapolation when applied to men.

What to do about it: use the same structure and hold the same confidence in the direction, but less in the exact figures. Watch the response rather than assuming it transfers.

Thirteen weeks collides with how services are funded

What the research showed: the trials converge on at least 13 weeks, with intensity built over the first 6 to 12.

The real-world gap: programmes are routinely commissioned or attempted at four to six weeks.

What to do about it: set the three-month expectation at the very start. A six-week block does not fail because the exercise was wrong. It stops before the point at which anything was measured.

The Nuance

Cinematic anatomical illustration representing clinical decision-making

The simple version of this page is "exercise works, pacing does not". That is too neat, and worth correcting in two directions.

Pacing genuinely helps one specific group. If your pattern is doing far too much on a good day and paying for it for the next three, then evening activity out across the week matters. In the trial that looked at this properly, changes in pacing were what actually drove the improvement in function for that subgroup. What the evidence does not support is pacing as the default advice for everyone, given ahead of exercise, which is how it usually arrives.

And the drop-out finding is a group-level result. It does not mean nobody struggles. It means exercise is not the uniquely intolerable thing it gets described as, and the two things that made it tolerable in the trials were expert supervision and a gentle start. Both are things you and your clinician control.

The practical version of all of this is unglamorous. Start easier than your instinct says, keep the schedule steady through the bad weeks, cut intensity rather than frequency when you flare, get someone qualified alongside you if you possibly can, and hold your judgement until week twelve.

Sources

Kundakci et al., 2022, Pain (PMID 34813518) — 167 randomised trials, 11,012 people. All exercise types reduced pain except flexibility work.
Rodriguez-Dominguez et al., 2025, Brazilian Journal of Physical Therapy (PMID 40319533) — network meta-analysis, 51 trials, 2,873 people. Water-based best short-term, resistance best long-term.
Vancampfort group, 2024, Archives of Physical Medicine and Rehabilitation (PMID 37331421) — 89 trials, 3,702 people. Drop-out 19.2%, matching control groups.
Fernández-Rodríguez et al., 2025, European Journal of Pain (PMID 39805734) — 17 trials. Aerobic dose and progression.
Zhang et al., 2024, Musculoskeletal Care (PMID 39004771) — 19 trials, 857 people. Higher guideline compliance outperformed lower.
Da Silva Sarturi et al., 2022, Rheumatology International (PMID 34652480) — resistance training dosing.
Maestre-Cascales et al., 2024, European Journal of Pain (PMID 37458315) — 15 trials, formally graded. Meaningful improvements, low certainty.
Antunes / Marques group, 2021, European Journal of Pain (PMID 32976664) — 50 studies, 3,562 people. Direct comparison of exercise parameters.
Bourke, White et al., 2014, Psychological Medicine (PMID 23967878) — PACE trial secondary analysis, in chronic fatigue syndrome.
Ünal et al., 2026, American Journal of Occupational Therapy (PMID 41894618) — activity management trial, 31 people.
Vancampfort group, 2023, Musculoskeletal Care (PMID 37186016) — 22 studies, 5,997 people. Baseline activity levels.
2022, PLoS One (PMID 36264901) — 99 studies, 9,853 people. Fitness, muscle function and perception of effort.
Sosa-Reina et al., 2017, BioMed Research International (PMID 29291206) — pooled pain effect.
Busch et al., 2013 (PMID 24362925); Bidonde et al., 2014 (PMID 25350761); Bidonde et al., 2017 (PMID 28950401) — Cochrane reviews on resistance, water-based and vibration training. All now 9 to 13 years old.

Exercise is the best-proven treatment for fibromyalgia, but only if you start easier and stay longer than feels sensible.