Premenstrual syndrome (PMS) is a set of distressing physical, behavioural and psychological symptoms that, in the definition used by NICE’s Clinical Knowledge Summary, occur in the second half of the menstrual cycle. Premenstrual dysphoric disorder (PMDD) is its severe form: a 2025 review in American Family Physician describes it as symptoms that substantially interfere with work, school, social life or relationships, or cause significant distress. The 2018 Cochrane review of acupuncture and acupressure for both, by Mike Armour and colleagues at Western Sydney University, found five randomised trials with 277 women and rated the evidence low or very low. Other reviews, which mostly count the share of women who improved rather than measure symptom scores, report better results with acupuncture than with drugs or sham treatment, and describe the trials behind those results as small and at high or unclear risk of bias. This page reports what each source found and against what each compared acupuncture.
The Cochrane review
The review searched to October 2017 for trials that randomised women with PMS or PMDD to acupuncture or acupressure against a sham, usual care or a waiting list, or against a drug recommended by the International Society for Premenstrual Disorders. Treatment courses ran from 7 to 28 sessions. No included trial compared acupuncture or acupressure with an active treatment, so the review has nothing to say about how either compares with selective serotonin reuptake inhibitors (SSRIs), a class of antidepressant. Certainty is rated with GRADE, the standard scale for how far a body of evidence can be trusted, from high to very low. Symptom scores are on the Daily Record of Severity of Problems (DRSP), a daily symptom diary.
| Comparison and outcome | Trials (women) | What the review found | Certainty |
|---|---|---|---|
| Acupuncture against sham: mood and physical symptoms | 1 (67) | About 9 points lower on each DRSP subscale, against sham scores of 52 (mood) and 46 (physical) | Low |
| Acupuncture against sham: response rate | 2 (100) | No clear difference once the authors allowed for the large disagreement between the two trials | Very low |
| Acupuncture against sham: quality of life | 1 (67) | 2.85 points higher on the WHOQOL-BREF, the World Health Organization’s short quality-of-life questionnaire | Low |
| Acupuncture against sham: adverse events | 3 (167) | Too few data to tell whether rates differed | Very low |
| Acupuncture against no treatment | 1 (14) | Symptom score 13.6 points lower; no adverse events in either group; the review was uncertain of the result | Very low |
| Acupressure against sham acupressure | 1 (90) | Fewer women with moderate to severe PMS at the end of the trial (risk ratio 0.64), and higher physical and mental quality-of-life scores; no data on adverse events | Low |
The authors concluded that the limited evidence suggests acupuncture and acupressure may improve physical and psychological symptoms against a sham, and named small samples, detection bias (outcome assessors who knew the allocation) and selective reporting as the main limits. They asked for trials with validated PMS measures, adequate blinding and comparison groups that reflect current best practice. The review has not been updated since 2018.
Other systematic reviews
Five other reviews pooled or summarised trials of acupuncture for PMS. The two most recent measured success mostly as an “effective rate”: the share of women each trial judged improved, which Zhang and colleagues report most trials calculated from a fall in symptom scores.
| Review | Searched to | Trials (women) | What it found |
|---|---|---|---|
| Cho and Kim, 2010 | January 2009 | 9 | Four trials reported larger symptom reductions with acupuncture than with drugs. Only two reported how they randomised and concealed allocation, and only two reported on harms, with one small haematoma. |
| Kim and colleagues, 2011 (Kyung Hee University) | June 2009 | 10 | More women improved with acupuncture than with any control (8 trials, risk ratio 1.55), than with progestin or anti-anxiety drugs (4 trials, 1.49) and than with sham (2 trials, 5.99). Most trials were at high risk of bias in randomisation, allocation concealment and blinding. |
| Jang and colleagues, 2014 | Trials published 2002 to 2012 | 8 acupuncture trials, alongside 11 of herbal medicine | Symptoms fell by half or more from baseline, with no serious adverse events reported. |
| Zhang and colleagues, 2019 (Chengdu University of Traditional Chinese Medicine) | September 2018 | 15 (1,103) | Higher effective rate than drugs (10 trials, odds ratio 4.16) and than sham (3 trials, 23.02). Starting acupuncture at different points in the cycle made no measurable difference. |
| Xie and colleagues, 2026 (Changchun University of Chinese Medicine) | March 2025 | 21 (1,818) | A network meta-analysis of acupuncture, ear acupressure, moxibustion and related methods. Against drugs, acupuncture had a higher effective rate (3 trials, risk ratio 1.45); against placebo, one trial each for acupuncture and ear acupressure. Two of the 21 trials reported adverse events. |
A network meta-analysis combines direct comparisons with indirect ones, so that two methods never tested against each other can still be ranked.
Xie and colleagues ranked one acupuncture style, which they call Jianpi-Shugan, first for effective rate and ear acupressure first for symptom
scores, and rated most trials at unclear risk of bias. Zhang and colleagues described the methodological quality of their trials as poor and
their samples as small. The points the trials in Zhang’s review used most often were SP06 (Sanyinjiao),
LR03 (Taichong) and RN4 (Guanyuan).
Trials that shaped the evidence
A 2002 Croatian study by Habek and colleagues compared acupuncture with placebo acupuncture and reported success in 77.8% of the acupuncture group against 5.9% of the placebo group. One woman had a small haematoma under the skin at the needle site.
A 2013 Brazilian trial by Carvalho and colleagues assigned 30 women with PMDD alternately, rather than at random, to acupuncture or sham acupuncture, twice a week for two cycles. Scores on the Hamilton anxiety scale fell by 58.9% with acupuncture and 21.2% with sham; scores on the Hamilton depression scale fell by 52.0% and 19.6%.
A 2017 Iranian trial by Bazarganipour and colleagues randomised 97 university students with PMS to acupressure at
LR03, at LI04 (Hegu) or at a placebo point, for 20 minutes a day over the 14 days before each period, for
three cycles; after the first cycle the women pressed the points themselves. By the second and third cycles fewer women in the two real-point
groups had moderate or severe PMS, and their anxiety, depression and quality-of-life scores were better than in the placebo group. The two real
points did not differ.
What guidance says
The Royal College of Obstetricians and Gynaecologists’ Green-top Guideline No. 48, published in 2016, tabulates acupuncture among complementary therapies as showing some benefit across 10 studies with 235 participants, at high risk of bias, and asks for further data before any recommendation. It advises telling women that the evidence for some complementary medicines conflicts, and notes a placebo response of 36% to 43% in PMS trials. The RCOG’s guideline page says the committee reviewed it in May 2023 and that an update is under development.
NICE’s Clinical Knowledge Summary for UK general practice, last revised in September 2024, draws partly on the RCOG guideline and the Cochrane review. Among complementary treatments it advises telling women there is weak evidence that acupuncture may help physical and affective symptoms for some women. The American Family Physician review names acupuncture or acupressure among the options that may be used alongside cognitive behavioural therapy, exercise and agnus castus, with SSRIs as first-line treatment.
What the reviews could not settle
The Cochrane review’s sham-controlled results rest on single trials of 67 and 90 women, and its pooled response rate showed no clear difference. The larger reviews’ effects against drugs and sham come from trials that their own authors rated at high or unclear risk of bias, measured with effective rates. The Cochrane review found no trial against an SSRI, and the PMDD trial reported here is small, with 30 women. With a placebo response of up to 43% in PMS trials, a sham comparison carries most of the weight, and the Cochrane review found no more than three sham-controlled trials for any outcome. A 2026 protocol from Beijing University of Chinese Medicine describes a pilot trial of auricular acupressure against sham in 66 women with PMDD, measuring depression and anxiety over three cycles.
AcuiQ’s premenstrual syndrome page lists the protocols individual studies prescribed, each with its citation. Period pain and Breast pain report the evidence on two other conditions tied to the menstrual cycle, and the evidence hub lists every condition covered.