Period pain

Primary dysmenorrhoea is period pain with no underlying disease to account for it: cramping pain low in the abdomen that starts in the first hours of a period. The 2016 Cochrane review of acupuncture for it, by Caroline Smith and colleagues at Western Sydney University, puts it among the commonest causes of period pain, affecting up to three in four women at some point in their reproductive years. That review pooled 42 randomised trials with 4,640 women and rated the evidence for every comparison low or very low. Later reviews and the larger single trials report lower pain scores with acupuncture or acupressure than with no treatment or with anti-inflammatory painkillers, and much less agreement when the comparison is a sham. This page reports what each source found and how each measured pain.

The Cochrane review

The review searched to September 2015 for trials in women of reproductive age with moderate to severe primary dysmenorrhoea, comparing acupuncture, electroacupuncture or acupressure with a sham or placebo, with medication, or with no treatment. Its main outcome was pain. Many trials reported pain scores that were too skewed to average, so for several comparisons the authors could only say which way each trial pointed. The review uses GRADE, the standard system for rating how far a body of evidence can be trusted, from high to very low.

The authors concluded that the evidence was insufficient to show whether acupuncture or acupressure works for primary dysmenorrhoea. They named the reasons: risk of bias, poor reporting, inconsistency between trials and a risk of publication bias, meaning that trials with unfavourable results may not have reached print. Most comparisons had no data on adverse events at all. The review has not been updated since 2016.

Later reviews

A 2018 review by Woo and colleagues at Kyung Hee University searched ten databases to December 2017 and included 60 trials, 49 of them in its meta-analyses. Against no treatment, it found lower menstrual pain with manual acupuncture and with electroacupuncture. Against NSAIDs, it found lower pain with manual acupuncture and with warm needle acupuncture, in which moxa is burned on the needle handle. It rated most trials at low or unclear risk of bias and described the included studies as low in quality.

A 2024 network meta-analysis by Li and colleagues in Guangzhou compared eight non-drug treatments across 33 trials searched to October 2022. A network meta-analysis combines direct comparisons with indirect ones, so that two treatments never tested against each other can still be ranked through a shared comparator. In the network analysis, exercise, acupuncture and topical heat each came out about 3 points lower on a 0 to 10 pain scale than placebo or no treatment. The authors described the effects as probable, over the short term, and called for larger and better trials.

Acupressure

Acupressure is pressure on points with a finger or a device rather than a needle, and most trials of it for period pain press SP06 (Sanyinjiao). A 2016 review by Ukachukwu Abaraogu and colleagues found six trials of SP06 acupressure with 461 participants, generally of low quality and high risk of bias. Pressure given by trained staff lowered pain immediately and for up to three hours; acupressure that women gave themselves needed several monthly cycles before pain fell. A 2025 review by Yu and colleagues at Beijing University of Chinese Medicine included 23 trials, 20 of them in its meta-analysis, and found lower pain with acupressure than with placebo acupressure (1.58 points on a 0 to 10 scale), oral medication (1.11 points) or usual treatment (1.29 points). It rated the evidence low in quality, and only two of its trials reported adverse events, all of them mild. Acupressure describes the technique.

Trials that shaped the evidence

A 2008 German trial led by Claudia Witt at the Charité in Berlin compared acupuncture with no acupuncture in usual care. Of 649 women, 201 agreed to be randomised; those who received 15 sessions over three months reported average pain of 3.1 on a 0 to 10 scale after three months, against 5.4 in the control group. Both groups could use their usual medical care, so the trial measured what adding acupuncture changed, not what acupuncture does against a sham. Its economic analysis put acupuncture at 3,011 euros per quality-adjusted life year, within the thresholds health systems commonly use.

A 2011 Australian trial by Caroline Smith and colleagues compared nine sessions of acupuncture over three months with control treatment using a placebo needle, in 92 women aged 14 to 25. At three months pain was lower in the acupuncture group, but not by a statistically significant margin. At six months the acupuncture group had shorter periods of pain and used fewer extra painkillers; at 12 months neither difference remained.

A 2014 multicentre trial by Liu and colleagues at six hospitals in China randomised 501 women to electroacupuncture at SP06, at an unrelated point, GB39 (Xuanzhong), or at a site that is not an acupuncture point. Pain after the first session was 4 mm lower on a 100 mm scale at SP06 than in either comparison group. The authors judged that difference statistically significant but too small to matter clinically, and found no difference between the three groups in menstrual symptoms over three cycles. A 2017 trial by Mike Armour and colleagues tested whether timing or electrical stimulation changed the result in 74 women across four arms, and found pain fell in all four groups with no difference between them; the authors noted the trial may have been too small to detect one.

What guidance says

NICE’s Clinical Knowledge Summary for general practice in the UK, last revised in October 2023, recommends an NSAID for primary dysmenorrhoea, with paracetamol where an NSAID is unsuitable or not enough, or a trial of hormonal contraception. It suggests two non-drug measures alongside them: heat, such as a hot water bottle, and high-frequency transcutaneous electrical nerve stimulation (TENS). Acupuncture and acupressure are not among them. Clinical guidelines sets out what other guideline bodies say about acupuncture, condition by condition.

What the reviews could not settle

The open question is the sham comparison. Trials against no treatment and against NSAIDs point one way, but neither design can separate the effect of needling from the attention, expectation and ritual of a course of treatment, and the Cochrane review found the sham-controlled trials too few and too inconsistent to answer it. Sham acupuncture explains why no sham is inert. The reviews also could not compare styles of acupuncture, numbers of sessions or timing within the cycle with any confidence, and they had little long-term follow-up and little systematic reporting of harms. None of the sources above concerns secondary dysmenorrhoea, period pain caused by an underlying condition such as endometriosis.

AcuiQ’s primary dysmenorrhoea page lists the protocols individual studies prescribed, each with its citation, and the dysmenorrhoea and menstrual pain pages hold studies that did not specify the type. Reading a protocol explains each field of an entry.