Acupuncture

What it is

Acupuncture is a treatment in which needles are inserted at chosen points on the body. Trials compare real acupuncture with sham needling, which may use non-penetrating needles [1].

What the research says

For migraine, a Cochrane review of 22 trials (4,985 people) found that acupuncture cut headache frequency compared with no acupuncture [2]. Against sham needling, about 50% of people on acupuncture had their headaches at least halved, against 41% on sham [2]. A trial sequential meta-analysis of 20 studies also found fewer migraine episodes than sham [3].

For chronic low back pain, a Cochrane review of 33 trials found clearly more pain relief than no treatment, but a small gain over sham needling that fell below the amount usually considered meaningful [4]. A 2020 sham-controlled trial of 121 adults found no significant difference in pain, but a larger improvement in disability scores [5]. Compared with sham, the Cochrane review found no clear gain in back-specific function, while an overview of Cochrane reviews judged a small function gain over sham as probable [4][6].

For osteoarthritis pain, a pooled analysis of 39 trials with adequate allocation concealment found acupuncture beat sham by about 0.2 standard deviations, and the effect persisted at one year with a drop of about 15% [7]. Knee trials disagree: electroacupuncture beat sham at 8 weeks (60% vs 47% meaningful improvement) while manual acupuncture did not [8], and superficial needling was no better than sham at 4 weeks [1].

Doses used in studies

Migraine prevention trials ran for at least 8 weeks [2]. In a 2020 low back pain trial, electroacupuncture was given as 12 sessions, twice a week, over 6 weeks [5]. In a 2021 knee trial, intensive acupuncture was given 3 times a week for 8 weeks [8]; another knee trial gave 10 sessions over 4 weeks [1].

Safety and interactions

The most commonly reported side effects in back pain trials were needle-point pain, bruising, hematoma, bleeding and worsening of back pain [4]. In migraine trials, people on acupuncture were no more likely than people on sham to drop out because of side effects [2]. The abstracts read for this page did not report data on pregnancy, interactions with medicines or upper limits, and did not describe people with bleeding disorders. Anyone who takes blood thinners or has a bleeding problem should ask a doctor before starting.

Caveats and open questions

Trials comparing acupuncture with no treatment or with drugs could not blind participants, which raises the risk of bias [2]. Most back pain trials had high risk of bias, and the function findings are of low or very low certainty [4]. Only one migraine trial followed people for 12 months [2]. A recruiting trial of tai chi and acupuncture for migraine (NCT06967909) is still running and may add to the picture.