The Cochrane review of acupuncture for fibromyalgia, published in 2013 by Deare and colleagues, found nine trials with 395 participants. Six of them, with 286 participants, compared acupuncture with sham, and on that comparison the review found moderate-quality evidence of no difference in pain, fatigue, sleep or overall well-being. Within those six, the two trials of electroacupuncture found less pain than sham for about a month. Three meta-analyses published since, drawing on more trials, found acupuncture better than sham for pain. The European League Against Rheumatism (EULAR) gave acupuncture a weak recommendation in 2017.
NICE classes fibromyalgia as chronic widespread pain, a type of chronic primary pain: pain that no underlying condition adequately accounts for. The trials below measure pain alongside fatigue, sleep, stiffness and overall well-being, and the results often differ between them. Sham acupuncture explains the controls that imitate treatment.
The Cochrane review
Deare and colleagues searched English and Chinese databases to January 2012. Three of the nine trials used electroacupuncture, in which a small current passes between needles, and six used manual acupuncture. All but one used a fixed formula of points rather than points chosen for each patient. They rated the quality of each finding with GRADE, the standard scale that runs from high to very low.
| Compared with | Trials (participants) | Finding | Quality |
|---|---|---|---|
| Sham, all acupuncture | 6 (286) | No difference in pain, fatigue, sleep or well-being; less stiffness | Moderate |
| Sham, electroacupuncture only | 2 (104) | Pain 13% lower; better well-being and sleep | Not rated separately |
| No treatment, electroacupuncture | 1 (13) | Pain 22 points lower on 0 to 100 | Low |
| Standard therapy of antidepressants and exercise, with acupuncture added | 1 (58) | Pain 3 points lower on 0 to 10, a 30% reduction | Moderate |
| Antidepressants | 1 (38) | Pain 17 points lower on 0 to 100 | Low |
Four trials that followed patients for six to seven months found no difference between acupuncture and its comparison by then. One low-quality trial found worse physical function after manual acupuncture than after sham. No serious adverse events were reported, though the review found too few adverse events of any kind to be certain of the risks. The authors concluded that electroacupuncture is probably better than manual acupuncture for pain, stiffness, well-being, sleep and fatigue, that manual acupuncture probably does not improve pain or function, and that the effect lasts up to a month after treatment.
Reviews since 2013
| Review | Trials | Pain | Other outcomes |
|---|---|---|---|
| Zhang et al. 2019, Journal of Pain Research | 12 trials, against sham or conventional drugs; 9 pooled against sham for pain | 1.04 points lower than sham on 0 to 10, moderate quality | Quality of life better than sham; advantage also present at long-term follow-up |
| Kim et al. 2019, Evidence-Based Complementary and Alternative Medicine | 10 trials (690 participants), 8 pooled | SMD −0.49 against sham | Sleep and general status better; no difference in fatigue |
| Zheng and Zhou 2022, Journal of Pain Research | 13 studies, 12 pooled, all against sham or simulated acupuncture | SMD −0.42 against sham | Well-being better; no evidence of effect on fatigue, sleep, physical function or stiffness. Evidence for pain moderate |
SMD is the standardised mean difference, the measure reviews use to combine trials that scored pain on different scales; a negative value favours acupuncture. Kim and Zheng disagree on sleep, and neither found an effect on fatigue. Zhang’s review includes the Spanish trial described below. A correction to Kim’s review, issued in 2026, states that against one type of sham, simulated acupuncture that does not pierce the skin at the relevant points, the advantage was not statistically significant when that subgroup was analysed alone.
The largest sham-controlled trial
A trial in primary care in southern Spain by Vas and colleagues, published in 2016, randomised 164 people with fibromyalgia to individualised acupuncture or sham, one 20-minute session a week alongside their usual medication. Pain fell by 41.0% with acupuncture and 27.1% with sham at 10 weeks, and the difference persisted at 12 months, at 19.9% against 6.2%. Side effects were mild and infrequent. Unlike most trials in the Cochrane review, this one chose points for each patient. Earlier trials by Vas and colleagues in other chronic pain conditions were outliers in the individual patient data analyses covered in Chronic pain.
What the guidelines say
EULAR revised its recommendations for managing fibromyalgia in 2017, by Macfarlane and colleagues. Its committee assessed eight reviews covering up to 16 trials and 1,081 participants and rated acupuncture “weak for”, with 93% agreement. It noted the Cochrane finding of a 30% improvement in pain when acupuncture was added to standard therapy, and that the evidence comparing real with sham acupuncture was less consistent. Exercise was the only therapy given a “strong for” recommendation. This page found no later revision.
NICE’s 2021 guideline on chronic pain, NG193, applies to fibromyalgia as a form of chronic primary pain. It says to consider a single course of acupuncture or dry needling, under limits on setting, cost and hours that Chronic pain sets out. Clinical guidelines sets both positions beside other bodies’.
What the reviews could not settle
The fibromyalgia trials are small: the Cochrane review’s largest comparison had 286 participants, and several of its findings rest on a single trial. The sources disagree on the sham comparison, the Cochrane review finding no difference for acupuncture overall and later reviews finding one, and they differ in which trials they admit and how they class shams. Whether electroacupuncture does better than manual acupuncture rests on two trials. Few trials followed patients beyond a few months: in the Cochrane review, four did, and any difference had gone by six to seven months, while the Spanish trial found one at 12 months. Pain, fatigue, sleep and stiffness respond differently, so a result on one does not carry to the others.
AcuiQ’s fibromyalgia page lists the protocols individual studies prescribed, each with its citation, and the diffuse musculoskeletal pain in fibromyalgia page does the same for the term some of those studies used.