Acupuncture for dysmenorrhoea
Abstract
Background
Primary dysmenorrhoea is the most common form of period pain and affects up to three‐quarters of women at some stage of their reproductive life. Primary dysmenorrhoea is pain in the absence of any organic cause and is characterised by cramping pain in the lower abdomen, starting within the first eight to 72 hours of menstruation.This review examines the currently available evidence supporting the use of acupuncture (stimulation of points on the body using needles) and acupressure (stimulation of points on the body using pressure) to treat primary dysmenorrhoea.
Objectives
To determine the effectiveness and safety of acupuncture and acupressure in the treatment of primary dysmenorrhoea when compared with a placebo, no treatment, or conventional medical treatment.
Search methods
We searched the following databases: the Cochrane Menstrual Disorders and Subfertility Group Trials Register (to September 2015), Cochrane Central Register of Controlled Trials (CENTRAL) (the Cochrane Library), MEDLINE, EMBASE, PsycINFO, CINAHL and Chinese databases including Chinese Biomedical Literature Database (CBM), China National Knowledge Infrastructure (CNKI), VIP database and registers of ongoing trials.
Selection criteria
We included all published and unpublished randomised controlled trials (RCTs) comparing acupuncture with sham acupuncture or placebo control, usual care, pharmacological treatment or no treatment. We included the following modes of treatment: acupuncture, electro‐acupuncture, and acupressure. Participants were women of reproductive age with primary dysmenorrhoea during the majority of the menstrual cycles or for three consecutive menstrual cycles, and moderate to severe symptoms.
Data collection and analysis
We calculated odds ratios (ORs) for dichotomous outcomes and mean differences (MDs) or standardised mean differences (SMDs) for continuous outcomes, with 95% confidence intervals (CIs). We pooled the data where appropriate. Our primary outcomes was pain. Secondary outcomes included menstrual symptoms, quality of life, and adverse effects.
Main results
We included 42 RCTs (4640 women). Acupuncture or acupressure was compared with a sham/placebo group, medication, no treatment or other treatment. Many of the continuous data were not suitable for calculation of means, mainly due to evidence of skew.
1. Acupuncture studies
Acupuncture versus sham or placebo control (6 RCTs)
Findings were inconsistent and inconclusive. However, the only study in the review that was at low risk of bias in all domains found no evidence of a difference between the groups at three, six or 12 months. The overall quality of the evidence was low. No studies reported adverse events.
Acupuncture versus NSAIDs
Seven studies reported visual analogue scale (VAS) pain scores, but were unsuitable for pooling due to extreme heterogeneity (I² = 94%). In all studies the scores were lower in the acupuncture group, with the mean difference varying across studies from 0.64 to 4 points on a VAS 0 ‐ 10 scale (low‐quality evidence). Four RCTs reported rates of pain relief, and found a benefit for the acupuncture group (OR 4.99, 95% CI 2.82 to 8.82, 352 women, I² = 0%, low‐quality evidence). Adverse events were less common in the acupuncture group (OR 0.10, 95% CI 0.02 to 0.44, 4 RCTs, 239 women, 4 trials, I² = 15%, low‐quality evidence).
Acupuncture versus no treatment
Data were unsuitable for analysis, but pain scores were lower in the acupuncture group in all six studies reporting this outcome. The quality of the evidence was low. No studies reported adverse events.
2. Acupressure studies
No studies of acupressure reported adverse events.
Acupressure versus sham or placebo control
Data were unsuitable for pooling, but two studies reported a mean benefit of one to three points on a 0 ‐ 10 VAS pain scale. Another four studies reported data unsuitable for analysis: all found that pain scores were lower in the acupuncture group. No studies reported adverse events. The quality of the evidence was low.
Acupressure versus NSAIDs
One study reported this outcome, using a 0 ‐ 3 pain scale. The score was higher (indicating more pain) in the acupressure group (MD 0.39 points, 95% CI 0.21 to 0.57, 136 women, very low‐quality evidence).
Acupressure versus no treatment
There was no clear evidence of a difference between the groups on a VAS 0 ‐ 10 pain scale (MD ‐0.96 points, 95% CI ‐2.54 to 0.62, 2 trials, 140 women, I² = 83%, very low‐quality evidence).
Authors' conclusions
There is insufficient evidence to demonstrate whether or not acupuncture or acupressure are effective in treating primary dysmenorrhoea, and for most comparisons no data were available on adverse events. The quality of the evidence was low or very low for all comparisons. The main limitations were risk of bias, poor reporting, inconsistency and risk of publication bias.
PICOs
Plain language summary
Acupuncture for period pain
Review question
Cochrane researchers reviewed the evidence about the effect of acupuncture and acupressure versus a control to reduce menstrual pain and symptoms in all women of reproductive age with primary dysmenorrhoea. Primary dysmenorrhoea is defined as pain without any underlying cause.
Background
Dysmenorrhoea, also known as period pain, is commonly experienced by younger women. Symptoms may include cramping pain in the lower abdomen that may spread to the lower back or front thigh, nausea, vomiting, diarrhoea, headache, fatigue, anxiety, and dizziness. Use of conventional treatments is high, but 20% to 25% of women still have inadequate pain relief. We therefore need other approaches to effectively manage symptoms of period pain. Acupuncture (the stimulation of points on the body using needles) and acupressure (stimulation of points on the body using pressure) are used to treat period pain, although we do not know how effective it is in reducing such pain.
Study characteristics
We have included 42 trials of acupuncture and acupressure compared to a control (sham/placebo, medication, Chinese herbs, no treatment or usual care) in a total of 4640 women of reproductive age with period pain. Twenty‐two studies were undertaken in China. Eight studies were undertaken in Iran, four studies in Taiwan, two studies in Korea, and one each in Australia, Germany, Hong Kong, Thailand, Turkey, and the USA.The evidence is current to September 2015.
Key results
There was insufficient evidence to demonstrate whether or not acupuncture or acupressure is effective in treating primary dysmenorrhoea, and for most comparisons no information was available on adverse events.
Quality of the evidence
The quality of the evidence was low or very low for all comparisons. The main limitations were risk of bias, poor reporting, inconsistency and risk of publication bias.
Authors' conclusions
Summary of findings
| Acupuncture compared with placebo, no treatment or active controls for primary dysmenorrhoea | ||||||
| Population: women with primary dysmenorrhoea Settings: outpatient, community Intervention: acupuncture Comparison: placebo , no treatment or active controls | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Control | Acupuncture | |||||
| Pain: Acupuncture versus sham or placebo control Follow up: 1‐12 months | Findings were inconsistent and inconclusive:
| 477 | ⊕⊕⊝⊝ | Data unsuitable for calculation of means | ||
| Adverse events: Acupuncture versus sham or placebo control | No studies reported this outcome | |||||
| Pain: Acupuncture versus NSAIDs Follow‐up: 1 ‐ 4 months | Pain scores were lower in the acupuncture group in six of seven studies (n = 389/449) that reported VAS scores, with the mean difference varying across studies from 0.64 to 4 points on a VAS 0 ‐ 10 scale [In 4 studies (n = 352) reporting rates of pain relief, rates were higher in women having acupuncture than in women having NSAIDs] | [OR 4.99 (2.82 to 8.82)] | 611 (10 RCTs) | ⊕⊕⊝⊝ | Continuous data unsuitable for pooling, I² = 94%. Consistent direction of effect | |
| Adverse events: Acupuncture versus NSAIDs Follow‐up: 3 months | 83 per 1000 | 9 per 1000 (2 to 38) | OR 0.10 (0.02 to 0.44) | 239 (4 RCTs) | ⊕⊕⊝⊝ | |
| Pain: Acupuncture versus no treatment Follow‐up: 1 ‐ 6 months | Six studies reported this outcome, using a VAS scale. The findings were consistent but unsuitable for pooling, in most cases because data were skewed. | 384 (6 RCTs) | ⊕⊕⊝⊝ | Data unsuitable for calculation of means | ||
| Adverse events: Acupuncture versus no treatment | No studies reported this outcome | |||||
| *The basis for the assumed risk is the median control group risk across studies. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1Downgraded one level for serious risk of bias: all or most studies at high or unclear risk of bias in one or more domains assessed. | ||||||
| Acupressure compared with placebo, no treatment or active controls for primary dysmenorrhoea | ||||||
| Population: women with primary dysmenorrhoea Settings: outpatient, community Intervention: acupressure Comparison: placebo, no treatment or active controls | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Control | Acupressure | |||||
| Pain: Acupressure versus sham or placebo control Follow‐up: 2 days to 3 months | All studies reported a benefit in the acupressure group. Two studies reported a mean benefit of 1 ‐ 3 points on a 0 ‐ 10 VAS scale | 538 (5 RCTs) | ⊕⊕⊝⊝ | Data were unsuitable for pooling. I² = 98% when the 2 studies with data suitable for analysis were pooled | ||
| Pain: Acupressure versus NSAIDs Follow‐up: 3 months | 1 study used a 0 ‐ 3 pain scale and reported a higher score (indicating more pain) in the acupressure group (MD 0.39 points, 0.21 to 0.57 higher). | 136 (1 RCT) | ⊕⊝⊝⊝ | |||
| Pain score: Acupressure versus no treatment Follow‐up: 1 ‐ 3 months | There was no conclusive evidence of a difference between the groups on a VAS 0 ‐ 10 scale. MD 0.96 points lower in the acupressure group (‐2.54 lower to 0.62 higher) | 140 (2 RCTs) | ⊕⊝⊝⊝ | Lower scores indicate less pain, I² = 83% | ||
| Adverse effects: Acupressure versus placebo , no treatment or active controls | No studies reported this outcome | |||||
| *The basis for the assumed risk is the median control group risk across studies. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1Downgraded one level for serious risk of bias: all or most studies at high or unclear risk of bias in one or more domains assessed. | ||||||
Background
Description of the condition
Dysmenorrhoea (period pain) can be described as painful uterine cramps of menstrual origin (Proctor 2006) and is classified as primary or secondary dysmenorrhoea. The focus of this review is on primary dysmenorrhoea, the most common form of period pain in young women (French 2008), and the most frequently underreported cause of period pain (Jamieson 1996). Primary dysmenorrhoea is defined as pain in the absence of any organic cause and is most common in women under the age of 25, with pain starting within three years of menarche (Coco 1999). Primary dysmenorrhoea’s characteristic symptom is crampy, colicky spasms of pain in the suprapubic area, occurring within eight to 72 hours of menstruation and peaking within the first few days as menstrual flow increases (Proctor 2006). In addition to painful cramps, many women with primary dysmenorrhoea experience other menstrual‐related symptoms, including back and thigh pain, headaches, diarrhoea, nausea and vomiting (Proctor 2006). Studies describing the prevalence of primary dysmenorrhoea report wide ranges of rates. A recent review of high‐quality studies found prevalence rates to range from 16.8% to 81% (Latthe 2006). In a review of women under 20 years of age, higher prevalence rates of 43% to 91% were found (Zahradnik 2010).These wide‐ranging rates are most likely explained by the varying criteria used to assess dysmenorrhoea between surveys, and the descriptions used to classify the severity of pain (Jamieson 1996).
The largest contributing physiological factor in primary dysmenorrhoea is increased amounts of prostaglandins present in the menstrual fluid (Dawood 2006). These prostaglandins are released when the endometrial lining breaks down during menses. Prostaglandins, especially PGF₂α, stimulate myometrial contractions, reducing uterine blood flow and causing uterine hypoxia. This is responsible for the painful cramping that characterises primary dysmenorrhoea (Dawood 2006). Secondary symptoms associated with primary dysmenorrhoea, such as headache, nausea and vomiting, are due to prostaglandins and their metabolites entering general circulation (Howard 2000), and are similar to side effects seen in women who have induction of labour via prostaglandin gel (Dawood 2006).
Description of the intervention
Acupuncture is a component of traditional East Asian medicine, with various forms of acupuncture present historically in Japan, China and Korea (White 2004). Acupuncture is the insertion of fine needles into the body at specific acupuncture points, while acupressure is the stimulation of acupuncture points by means of pressure, usually applied by fingers or thumbs (Beal 1999).
Acupuncture, in common with many other non‐pharmacological therapies, is a complex intervention (MacPherson 2007). Inherent in the idea of complex interventions is the concept that a mixture of specific and non‐specific mechanisms contribute to the action of the therapy. Acupuncture can take place within a number of different theoretical frameworks, with the following being most common: Traditional Chinese, Five Element, Japanese or Biomedical (sometimes called “Western”) (Bovey 2005; Hopton 2012; Robinson 2012). These frameworks dictate many facets of the treatment, including diagnosis, point location, point selection, co‐interventions such as moxibustion, needling style, retention time for needles, and stimulation style.
A traditional Chinese medicine (TCM) acupuncture consultation typically includes but is not limited to: history taking, self‐care and lifestyle advice, TCM diagnosis and explanation, palpation of the pulse and other parts of the body, observation of the tongue, insertion and manipulation of needles, as well as common co‐modalities, such as moxibustion and cupping (Langevin 2011). Western or biomedical acupuncture more commonly uses a neurophysiological approach to point selection and often includes lifestyle advice based on biomedical considerations (Hopton 2012). Ear acupuncture, or auriculotherapy, is a relatively modern therapy and is based on the idea that there is a reflex connection between the ear and the rest of the body, allowing conditions to be treated by stimulation of certain parts of the ear (Andersson 2007). Serious adverse events from acupuncture are rare; minor side effects may include tiredness, pain at the site of needling, and headache (MacPerson 2004).
How the intervention might work
The exact mechanism of acupuncture to treat period pain is unclear. However, there has been extensive research into the mechanisms by which acupuncture may exert its pain‐relieving effects, with current evidence suggesting that endogenous opioids, modulation of neurotransmitters (such as dopamine), and anti‐inflammatory effects mediated via the hypothalamus–pituitary–adrenal (HPA) axis may all play a role (Napadow 2008). Because of the quick onset of the pain‐relieving effect of acupuncture observed in some studies, it is likely that the immediate analgesic effects observed are modulated, at least in part, by endogenous opioid release (Wang 2008).
Overall, current experimental data support three possible non‐exclusive mechanisms for the observed long‐term effects of acupuncture on primary dysmenorrhoea:
-
Descending pain modulation
-
An increase in uterine blood flow via ovarian sympathetic nerve reflex
-
A change in prostaglandin levels.
Smith 2010 proposes a descending pain modulation hypothesis, where pain relief in primary dysmenorrhoea occurs via stimulation of common acupuncture points, including REN4, SP6 and SP8, that transmit signals via afferent pathways to the midbrain. This information is integrated and changes women’s perception of pain. This alteration in descending pain modulation changes the perceived pain intensity. Animal models suggest that needling regulates neuro‐endocrine activities including levels of progesterone (Liu 2009) and needling point SP6 has been shown to increase ovarian blood flow via a reflex response in rats (Stener‐Victorin 2006). A similar reflex response increasing uterine blood flow via SP6 has been found in humans (Stener‐Victorin 1996) and appears to be related to some of the analgesic benefits of acupuncture in primary dysmenorrhoea. Lin 2008 found that after three months of acupuncture treatment there was a reduction in PGF₂α levels. A recent meta‐analysis, including Chinese‐language research, has shown that the effect of acupuncture on prostaglandins is still unclear due to poor trial methodology and small sample sizes (Xu 2014b). It is likely that acupressure exerts its benefits via mechanisms similar to acupuncture, with pressure rather than needle stimulation activating the pain‐relieving mechanism(s). The appropriateness of using sham acupuncture as a control is currently debated in the literature (Langevin 2011). Sham acupuncture devices deliver significant non‐specific effects, greater than that of placebo pills, so it is likely that these compromise at least some of the therapeutic benefit that acupuncture may deliver (Kaptchuk 2006).
Why it is important to do this review
Menstrual pain remains an important women's health issue and has medical, social, and economic consequences (Dawood 1990). Many young women do not seek help for their period pain or are undertreated (O'Connell 2006). In one study, 98% of adolescents used non‐pharmacological methods such as heat, rest, or distraction, with a perceived effectiveness of 40% or less (Campbell 1999). Conventional treatments have focused on the use of non‐steroidal anti‐inflammatory drugs (NSAIDs) and the oral contraceptive pill (OCP). NSAIDs work as prostaglandin synthetase inhibitors, as do OCPs, where the progestogen component of the OCP has a direct effect on uterine smooth muscle to reduce myometrial activity. The efficacy of these conventional treatments is high, although 20% to 25% of women have inadequate pain relief (Dawood 1985; Henzl 1985). NSAIDs appear to be a very effective treatment for dysmenorrhoea, although women using them need to be aware of the substantial risk of adverse effects (Marjoribanks 2015). There is some evidence in favour of oral contraceptives for pain management (Wong 2009). These two drug groups are now considered standard treatments for primary dysmenorrhoea (Dawood 1988; Dawood 1990). Other evidence‐based approaches to effectively managing symptoms of dysmenorrhoea are therefore needed. A number of clinical trials have been performed to study the efficacy of acupuncture for period pain, although it remains uncertain whether the existing evidence is rigorous enough to reach a definitive conclusion.
Objectives
To determine the effectiveness and safety of acupuncture and acupressure in the treatment of primary dysmenorrhoea when compared with a placebo, no treatment, or conventional medical treatment.
Methods
Criteria for considering studies for this review
Types of studies
All randomised controlled trials (RCTs) were eligible for inclusion. We included cross‐over trials if they had pre‐cross‐over data.
We excluded quasi‐randomised trials and trials where we were unable to ascertain if the trial was truly randomised.
Types of participants
Inclusion criteria
Women needed to meet the following criteria to be included in the review:
-
of reproductive age (15 to 49 years);
-
primary dysmenorrhoea, i.e. no identifiable pelvic pathology as indicated by pelvic examination, ultrasound scans, or laparoscopy;
-
primary dysmenorrhoea (self‐reported pain) during the majority of the menstrual cycles or for three consecutive menstrual cycles;
-
moderate to severe primary dysmenorrhoea (pain that does not respond well to analgesics, affects daily activities, or has a high baseline score on a validated pain scale).
Exclusion criteria
If participants in the trial met any of the following exclusion criteria, we excluded the trial from the review:
-
diagnosed secondary dysmenorrhoea (e.g. fibroids, endometriosis);
-
dysmenorrhoea resulting from use of an intra‐uterine device (IUD);
-
mild or infrequent dysmenorrhoea.
Types of interventions
Any RCT involving acupuncture (manual insertion of needles to points located on the body, and to the ear) and electro‐acupuncture as treatment for primary dysmenorrhoea. We made an amendment to the protocol to include trials of acupressure; this covers pressure applied using blunt studs or seeds. We also included application of minimal moxibustion on a small number of points, but we excluded trials of moxibustion alone, due to a different mode of application and action.
Interventions were compared to placebo control (including invasive and non‐invasive placebo controls), no treatment, pharmacological management, other types of control groups (for example, wait list controls, where the control group will receive the intervention after a waiting period), or other conventional treatments.
Types of outcome measures
We required reporting of at least one of the following primary outcomes for a trial to be included. We recorded data for each of the following outcomes, where available:
Primary outcomes
1. Pain score, as a continuous variable measured preferably by the visual analogue scale (VAS) or other validated scales. Pain relief, measured as dichotomous outcomes (i.e. pain relief: yes or no).
Secondary outcomes
1. Overall improvement in generic menstrual symptoms (e.g. nausea, tiredness) measured by changes in overall dysmenorrhoeic symptoms that were either self‐reported or investigator‐observed, or any other similar measures.
2. Reported use of additional medication, measured as the proportion of women requiring analgesics.
3. Restriction of daily life activities, measured as the proportion of women who reported activity restrictions.
4. Absence from work or school, measured as the proportion of women reporting absences from work or school, and also as hours and days of absence as a more selective measure.
5. Quality of life, measured by a validated scale, for example the Short Form (SF) 36.
6. Adverse effects from treatment, measured as incidence of side effects and types of side effects.
Search methods for identification of studies
We sought all published and unpublished RCTs of acupuncture and acupressure versus control, using the following search strategy, without language restrictions and in consultation with the Menstrual Disorders and Subfertility Group (MDSG) Trials Search Co‐ordinator.
Electronic searches
We searched the Cochrane Menstrual Disorders and Subfertility Group Trials Register (to September 2015), Cochrane Central Register of Controlled Trials (CENTRAL) (the Cochrane Library), MEDLINE, EMBASE, PsycINFO, CINAHL, and Chinese databases including Chinese Biomedical Literature Database (CBM), China National Knowledge Infrastructure (CNKI) and VIP database.
The search terms are presented in the appendices (Appendix 1; Appendix 2; Appendix 3; Appendix 4; Appendix 5; Appendix 6; Appendix 7; Appendix 8).
Searching other resources
We searched for additional studies in the reference lists of the relevant trials identified.
We also searched ProQuest Dissertations and Theses, Web of Knowledge, AMED (the Allied and Complementary Medicine Database), OpenGrey, LILACS, Acubase, Google Scholar and PubMed.
We searched the following clinical trial registries for ongoing trials: Australian and New Zealand Clinical Registry (www.anzctr.org.au/), Chinese Clinical Trial Register (www.chictr.org); Current Controlled Trials (controlled‐trials.com); Clinical trials.gov (clinicaltrials.gov); ISRCTN (www.isrctn.org/); and WHO International Clinical Trial Registration Platform search portal (www.who.int/trialsearch/).
Data collection and analysis
Selection of studies
Two review authors (CS and MA) screened the titles and abstracts of articles found in the search, and discarded trials that were clearly not eligible. One review author (XL) searched for and selected the trials from the Chinese databases. Two out of the three review authors (CS, MA, XL) undertook trial selection.
Three review authors (XZ, XL, ZL) translated Chinese papers. CS, MA, and XL independently assessed whether trials met the inclusion criteria, with disagreements resolved by discussion. If articles contained insufficient information to make a decision about eligibility, CS attempted to contact authors of the original reports to obtain further details. If details of randomisation were unclear in the reporting, we contacted all trial authors to ascertain if the study was truly randomised. We made a first contact and then sent a reminder. We translated letters and email into Chinese.
Data extraction and management
Following an assessment for inclusion, CS, MA, XZ, XL, ZL and JS independently extracted data. Author XZ and colleague JM reviewed the trial published by review author CS. We resolved discrepancies by discussion. For each included trial we extracted data regarding the location of the trial, the methods of the trial (as per assessment of risk of bias), the participants (age range, eligibility criteria), the nature of the interventions, and data relating to the outcomes specified above. We collected information on reported benefits and adverse effects. We extracted data and entered them onto a form (Appendix 9) sourced from the Cochrane Handbook for Systematic Reviews of Interventions (Cochrane Handbook; Higgins 2011). Two review authors (CS and MA) checked and entered data into Review Manager 5 (New Reference).
Assessment of risk of bias in included studies
Two review authors independently assessed risks of bias, using the criteria described in the Cochrane Handbook (Higgins 2011) ( Appendix 10). The tool consists of six items, with three potential responses: 'yes', 'no', and 'unclear'. In all cases a judgement of ‘yes’ indicates a low risk of bias and a judgement of ‘no’ indicates a high risk of bias. If insufficient detail was reported our judgement was usually ‘unclear’. We also made a judgement of 'unclear' if we knew what happened in the study but the risk of bias was unknown to us, or if an entry was not relevant to the study at hand (particularly for assessing blinding and incomplete outcome data, or when the outcome being assessed by the entry had not been measured in the study).
We assessed the following characteristics: sequence generation, allocation concealment, blinding (or masking), incomplete data assessment, selective outcome reporting, and other sources of bias. We resolved disagreements that arose at any stage by discussion between the review authors or with a third party, when necessary. We generated a 'Risk of bias' assessment table for each study. We assessed other aspects of trial quality including the extent of blinding (if appropriate), whether groups were comparable at baseline, the extent of losses to follow‐up, non‐compliance, whether the outcome assessment was standardised, and whether an intention‐to‐treat analysis was undertaken, according to the Cochrane Handbook (Higgins 2011). This information is presented in the Characteristics of included studies.
Measures of treatment effect
We performed statistical analysis in accordance with the Cochrane Handbook (Higgins 2011). We performed statistical analysis using Review Manager 5 software (RevMan 2014). For dichotomous data, we expressed results for each study as odds ratios (ORs) with corresponding 95% confidence intervals (CIs), using the Mantel‐Haenszel method. We expressed continuous data as mean differences (MDs) with 95% CIs, or as standardised mean differences (SMDs) if outcomes were conceptually the same but measured in different ways in the different trials. A standard rule of thumb for interpreting effect sizes is that 0.2 represents a small effect, 0.5 a moderate effect and 0.8 a large effect.
Unit of analysis issues
The primary analysis was per woman randomised. We included trials with multiple arms and describe them in the Characteristics of included studies, for example acupuncture compared with placebo acupuncture and with no acupuncture. If there were two acupuncture groups, we combined data from both treatment arms into one group. For studies with a placebo control and no treatment control group, we divided the shared intervention evenly between groups as described in the Cochrane Handbook (Higgins 2011). Where outcomes were repeated measures, we undertook analysis of outcomes at the end of the intervention.
Dealing with missing data
We analysed data on an intention‐to‐treat basis, as far as possible. We did not impute missing data but we did report the proportion lost to follow‐up and analysed only the available data.
Assessment of heterogeneity
We identified and measured heterogeneity by visually inspecting the overlaps of the CIs for the results of individual studies. If there was poor overlap, this was suggestive of statistical heterogeneity and we included a more formal Chi² test. A low P value (or a large Chi² statistic relative to its degrees of freedom) provided evidence of heterogeneity of intervention effects (variation in effect estimates beyond chance).
We measured inconsistency across trials in the meta‐analysis using the I² statistic. This describes the percentage of total variation across studies that is due to heterogeneity rather than chance (Higgins 2003; Higgins 2011). The interpretation of the I² statistic is as follows:
-
10% to 40% might not be important;
-
30% to 60% may represent moderate heterogeneity;
-
50% to 90% may represent substantial heterogeneity;
-
75% to 100% considerable heterogeneity.
Assessment of reporting biases
We planned to investigate potential biases of publication using the funnel plot or other analytical methods (Egger 1997). If there were 10 or more studies in the meta‐analysis we would have investigated reporting biases (such as publication bias) using funnel plots. If visual assessment had suggested asymmetry, we would have explored possible reasons.
Data synthesis
We combined the data using a fixed effect model in the following comparisons:
1. Acupuncture versus control.
2. Acupressure versus control.
Subgroup analysis and investigation of heterogeneity
We planned to undertake a subgroup analysis based on different types of acupuncture therapies: manual acupuncture versus electro‐acupuncture.
Sensitivity analysis
In the presence of substantial heterogeneity (an I² statistic greater than 50%) we aimed to examine the causes by prespecified subgroup using a sensitivity analysis. Where subgroup analysis failed to explain the heterogeneity, we analysed data using a random‐effects model. A priori, we had planned to perform sensitivity analyses on results to look at the possible contribution of: (1) differences in methodological quality, with trials of high quality (low risk of bias) compared to all trials; and (2) publication bias by country.
Overall quality of the body of evidence: Summary of findings table
We prepared 'Summary of findings' tables using GRADEPRO software. These tables evaluate the overall quality of the body of evidence for the main review outcomes (pain and adverse effects), using GRADE criteria (study limitations, consistency of effect, imprecision, indirectness and publication bias). We justified, documented and incorporated judgements about evidence quality (high, moderate, low) into the results for each outcome.
Results
Description of studies
See Characteristics of included studies, Characteristics of excluded studies, Characteristics of ongoing studies, and Characteristics of studies awaiting classification.
Results of the search
The updated search to September 2015 identified 73 potentially eligible studies of which we included 34: (Aghamiri 2005; Bazarganipour 2010; Bu 2011; Cao 2011; Charandabi 2011; Darabi 2010; Han 2012; Hu 2005; Kashefi 2010; Lee 2007; Li 2012b; Ma 2010; Ma 2013; Mirbagher‐Ajorpaz 2011; Peng 2012; Qiao 2013; Ruan 2011; Shi 2011; Song 2013; Sriprasert 2015; Sun 2011; Wang 2005b; Wang 2013b; Wang 2014a; Wong 2010; Xu 2013; Xu 2014; Yeh 2013a; Youn 2008; Yu 2014; Zafari 2011; Zhang 2012; Zhang 2013a; Zhang 2013b), and excluded 39 trials. Two RCTs previously included in the review (Jiang 2007; Wu 2007) we now exclude from this update. We now have 42 included and 65 excluded trials. see Figure 1.
Study flow diagram: 2016 update
Included studies
Study design and setting
The review includes 42 studies. Forty studies were parallel design and two studies were cluster‐randomised trials. Twenty‐eight trials had two study groups. Eight trials had three groups and six trials had four arms.
Study location and sources of women
Twenty‐two studies were undertaken in China. Eight studies were undertaken in Iran, four studies in Taiwan, two studies in Korea, and one each in Australia, Germany, Hong Kong, Thailand, Turkey, and the USA.
Participants
The studies included 4640 women with primary dysmenorrhoea. Twenty studies recruited participants from gynaecology hospital clinics. Fifteen trials recruited college or high school students, and five trials recruited from the community. Details were not reported in two trials. Women with secondary dysmenorrhoea were excluded, although this diagnosis was not clearly reported. One study used cancer or carbohydrate antigen 125 (CA 125) to eliminate a diagnosis of endometriosis (Wang 2009a). Evidence of a physical assessment confirming the diagnosis of primary dysmenorrhoea was reported in six trials, confirmed by ultrasound and physical examination. In all other 36 trials no clear, specific details were reported on how pelvic pathology was eliminated.
Interventions
Ten trials used acupressure and 32 trials provided stimulation using acupuncture (including manual and electro‐acupuncture). Three of these acupuncture trials also reported using moxa (warming needle). Acupuncture and acupressure varied in point selection, frequency of treatment, and number of treatments. A summary of the characteristics of the acupuncture and acupressure interventions is described in Appendix 11.
Comparative and control groups varied between studies. Nine trials used placebo controls. Placebo techniques varied between invasive and non‐invasive techniques. It is possible that the 11 control groups that used invasive pressure or needle stimulation may have had some physiological action. Comparisons with medication using NSAIDs and oral contraceptives were used in 15 trials. Eight trials used a no‐treatment control. One trial used Chinese herbal medicine. Two studies reported unspecified "usual care". One trial used rest. One trial used visits or social support. Three trials used a mixture of no treatment and active sham points. One trial used two groups comprising sham and medication.
Outcomes
All studies assessed the primary outcome of pain. VAS scales were widely used. Nine studies used a subjective assessment of pain and other menstrual symptoms. Six studies reported adverse events in a form that could be entered into meta‐analysis (Cao 2011; Peng 2012; Ruan 2011; Sriprasert 2015; Witt 2008; Zhang 2013a).
Excluded studies
We excluded 64 trials. Seventeen trials did not meet the inclusion criteria for the intervention; for example, the use of point injection acupuncture, use of moxibustion alone, or laser acupuncture. Nine trials did not meet the inclusion criteria for the control group; for example, comparisons between acupuncture versus acupuncture at another point. We exclude 35 trials due to insufficient reporting of randomisation; we were unable to ascertain the true randomisation status from the author, or trials were quasi‐randomised. No clinically relevant outcomes were reported in three trials. One trial did not meet the age criteria for participants. We present further background information on these trials in the table Characteristics of excluded studies.
Risk of bias in included studies
See Figure 2 and Figure 3 for a graphical summary of the 'Risk of bias' assessments of the included studies made by the review authors, based on the six 'Risk of bias' domains. Overall we judged one trial to be at a low risk of bias on all domains, 32 trials at high risk of bias in at least one domain and 12 trials rated as at an unclear risk of bias in one or more domains (but with no domains rated as at high risk).
Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Allocation
Using the Cochrane criteria, which rate the adequacy of randomisation allocation and concealment, we rated most trials (36) at a low risk of bias with adequate generation of randomisation sequence. In 11 trials the sequence was computer‐generated (Han 2012; Ma 2010; Ma 2013; Ruan 2011; Shi 2011; Smith 2010; Song 2013; Sriprasert 2015; Wang 2009a; Witt 2008; Zhi 2007). The sequence was drawn by lot in three trials (Chen 2004; Chen 2010; Zhang 2013a) and from random‐number tables in 20 trials (Bazarganipour 2010; Cao 2011; Helms 1987; Hu 2005; Kashefi 2010; Li 2008; Li 2012b; Lee 2007; Mirbagher‐Ajorpaz 2011; Qiao 2013; Sun 2011; Wang 2005b; Wang 2013b; Wang 2014a; Youn 2008; Xu 2013; Xu 2014; Yu 2014; Zhang 2012; Zhang 2013b), and coin toss in one trial (Yeh 2013a). The risk was unclear in six trials.
We rated allocation concealment at low risk of bias in 12 trials (27%). Central randomisation was undertaken in seven trials (Han 2012; Ma 2010; Ma 2013; Ruan 2011; Smith 2010; Song 2013; Witt 2008) and using sealed envelopes in five trials (Chen 2004; Chen 2010; Sun 2011; Sriprasert 2015; Wang 2009a). We assessed all other studies (30) to be at an unclear risk of bias due to insufficient reporting.
Blinding
We rated blinding at low risk of bias in three trials (Ma 2010; Smith 2010; Wang 2009a). All these trials were assessed as double‐blind with blinding of analyst and participants. Twenty‐eight trials were at high risk of bias, and the risk of bias was unclear in 11 trials.
Incomplete outcome data
We assessed incomplete outcome reporting to be at a low risk of bias in 30 trials (Bu 2011; Cao 2011; Charandabi 2011; Han 2012; Helms 1987; Li 2008; Li 2012b; Ma 2010; Mirbagher‐Ajorpaz 2011; Peng 2012; Qiao 2013; Ruan 2011; Shi 2011; Smith 2010; Song 2013; Sun 2011; Wang 2005b; Wang 2009a; Wang 2013b; Wang 2014a; Witt 2008; Wong 2010; Xu 2013; Yu 2014; Zafari 2011; Zhang 2012; Zhang 2013a; Zhang 2013b; Zhi 2007; Xu 2014). Nine trials were at high risk of bias, of which Chen 2004 had a 50% dropout rate (Aghamiri 2005; Bazarganipour 2010; Chen 2004; Chen 2010; Lee 2007; Ma 2013; Sriprasert 2015; Yeh 2013a; Youn 2008). We rated three trials as unclear, due to incomplete reporting (Darabi 2010; Hu 2005; Kashefi 2010).
Selective reporting
We rated the risk of selective reporting as low risk in one trial for which we were able to obtain the protocol published on a clinical trial registry (Smith 2010). We judged six other trials to be at low risk, due to receipt of entire theses. One study we rated at high risk because not all their secondary outcomes were reported. We assessed 34 trials to be unclear risk of bias.
Other potential sources of bias
We rated the risk of bias from other sources as low in five trials (Qiao 2013; Smith 2010; Song 2013; Wang 2013b; Witt 2008). We found one trial to be at high risk due to imbalances at baseline (Han 2012). We rated all other 36 trials as being at unclear risk of bias.
There were imbalances at randomisation in two trials. Smith 2010 reported imbalances in body mass index, smoking, and socio‐economic indices, and these were adjusted for in the primary analyses. Witt 2008 reported imbalances at randomisation for physical functioning, bodily pain, and physical component score of the SF36. For the majority of studies these data were not reported.
Five trials conducted and reported an intention‐to‐treat analysis (Ma 2013; Ruan 2011; Smith 2010; Wang 2014a; Witt 2008).
Four trials reported the source of funding. Three trials received government research funding (Chen 2004; Chen 2010; Smith 2010) and one trial received funding from health insurance organisations (Witt 2008).
Effects of interventions
See: Summary of findings for the main comparison Summary of findings: acupuncture for primary dysmenorrhoea; Summary of findings 2 Summary of findings: acupressure for primary dysmenorrhoea
We present results separately for acupuncture and acupressure, and by type of control group. Data were excluded from Charandabi 2011 due to data presented within subcategories of dysmenorrhoea. We could not obtain raw data for Helms 1987; Lee 2007; Ma 2010; Wang 2013b; Youn 2008, and present a narrative reports of their results.
1) Acupuncture versus sham/placebo control
Data were presented for pain outcomes reported after the intervention and at one to three months.
1.1 Primary outcome: Pain
1.1.1 Outcome: pain score
Five studies reported this outcome (Darabi 2010; Lee 2007; Smith 2010; Song 2013; Youn 2008). Data were unsuitable for pooling, in most cases because standard deviations (SDs) were not reported or data were skewed.
One study reported no evidence of a difference between the groups (Lee 2007) at one month.
Two studies reported evidence of a benefit in the intervention group at three months (Darabi 2010; Song 2013), but two (Smith 2010; Youn 2008) found no evidence of a benefit at this time point.
The single study with longer follow‐up reported no evidence of a difference between the groups at six or 12 months (Smith 2010).
Details of findings and rationale for not pooling are presented in Table 1.
| Study | Measure | Follow‐up | Acupuncture group Mean score | Control group Mean score | P < 0.05 | Comment |
| VAS (0 ‐ 10) | 3 months | 2.12, SD 1, n = 50 | 4.24, SD 0.91, n = 50 | Yes | This study reported implausibly low SDs. We assumed that these might be standard errors, but recalculation based on this assumption resulted in skewed data | |
| Measure of Menstrual Pain Questionnaire (scale unclear) | 1 month | 4.46 | 4.58 | No | Total n = 63. No SDs reported | |
| McGill Questionnaire Pain scale (0 ‐ 5) (Melzack 1975) | 3 months | 2, SD 2.5, n = 46 | 2.6, SD 2.7, n = 46 | No P = 0.32 | Data skewed | |
| 6 months | 2.2, SD 3 | 2.8, SD 2.7 | No P = 0.11 | Data skewed | ||
| 12 months | 2.6, SD 2.7 | 2.8, SD 2.7 | No P = 0.60 | Data skewed | ||
| VAS (0 ‐ 10) | 3 months | 4.2, SD 7.28, n = 25 | 8.77, SD 16.3, n = 96 | Yes | Data skewed | |
| Measure of Menstrual Pain Questionnaire (scale unclear) | 3 months | MD 0.54, 95% CI 0.12 to 1.19 | No | Total n = 80. No primary data reported | ||
SD: standard deviation
1.1.2 Outcome: pain relief
One study (Helms 1987) reported this outcome. After treatment for three cycles the rate of post‐treatment pain relief was higher in the acupuncture group (OR 17.5, 95% CI 1.6 to 191.9; 22 women).
Secondary outcomes
1.2 Outcome: overall improvement in menstrual symptoms
One trial (Smith 2010) compared acupuncture with a sham acupuncture control and found no evidence of a difference between the groups in the number of women reporting an improvement in menstrual symptoms at three months (OR 1.57, 95% 0.61 to 4.04, 92 women), six months (OR 1.47, 95% CI 0.62 to 3.48, 92 women) or at 12 months (OR 2.35, 95% CI 0.80 to 6.94) Analysis 1.2. At three months women in this study reported fewer mood changes in the acupuncture group (53%) compared with the control group (72%) (OR 0.72, 95% CI 0.53 to 1.00; P = 0.05).
Lee 2007 found a reduction in the severity of menstrual symptoms using the Menstrual Severity Symptom list, with the acupuncture scores reduced from 60.09 to 32.48, and the control scores declining from 62.40 to 39.65. Data were unsuitable for analysis as no standard deviations were reported.
1.3 Outcome: use of additional medication
Three trials reported on this outcome (Darabi 2010; Helms 1987; Smith 2010).
Smith 2010 found no evidence of a difference between the groups in the rate of additional medication use at three months (OR 0.91, 95% CI 0.40 to 2.09, 92 women) (Analysis 1.3) or at 12 months (OR 1.51, 95% CI 0.62 to 1.3.7). A reduction in medication use was found at six months for women in the acupuncture group (OR 0.34, 95% CI 0.14 to 0.79).
Helms 1987 reported that mean daily strength of analgesic medication (calculated using a study‐defined measure) dropped 41% in the intervention group after treatment, but increased among controls (48 women) (data not reported in meta‐analysis).
Darabi 2010 reported less use of medication in the acupuncture group compared with the control at three months (data not shown).
1.4 Outcome: restriction of daily living activities
There was no evidence of a difference between the groups at the end of the trial (OR 0.72, 95% CI 0.29 to 1.81, 92 women, 1 trial) (Analysis 1.4).
1.5 Outcome: absence from school or work
No trials reported on this outcome.
1.6 Outcome: quality of life
There was no evidence of a difference between groups on any quality of life domain, measured using the SF36 (92 women, 1 trial, high‐quality evidence) (Analysis 1.5, Analysis 1.6; Analysis 1.7; Analysis 1.8; Analysis 1.9; Analysis 1.10; Analysis 1.11).
1.7 Adverse events
No trial reported on this outcome.
2) Acupuncture versus medication
Data were presented for pain outcomes at one to three months.
2.1 Primary outcome: pain
2.1.1 Outcome: pain score
Acupuncture versus NSAIDS
Seven studies reported this outcome using a VAS 0 ‐ 10 scale, but data were unsuitable for pooling due to extreme heterogeneity (I² = 94%) for which there was no obvious explanation. In six of seven studies (389 women) acupuncture was associated with reduced pain scores compared to NSAIDs, with the mean differences varying across studies from 0.64 to 4 points (Analysis 2.1; Figure 4). The overall quality of the evidence was very low.
Forest plot of comparison: 2 Acupuncture versus medication, outcome: 2.1 Pain score.
One trial (Han 2012) used an unspecified scale and was not included in the analysis; this trial found acupuncture was associated with reduced pain scores (MD ‐3.70, 95% CI ‐4.68 to ‐2.80, 120 women).
Acupuncture versus combined oral contraceptives (COC)
Combined oral contraceptives were found to be more effective than acupuncture in reducing pain scores as measured by a 0 ‐ 10 numeric rating scale (NRS) in one small trial (MD 2.0, 95% CI 0.59 to 3.41, 52 women; Analysis 2.1).
2.1.2 Outcome: pain relief
Acupuncture versus NSAIDS
Women having acupuncture were more likely to report pain relief than women having NSAIDs (OR 4.99, 95% CI 2.82 to 8.82, 4 trials, 352 womenI² = 0%, low‐quality evidence (Analysis 2.2).
Acupuncture versus combined oral contraceptives
There was no evidence of a difference between acupuncture and combined oral contraceptives in the number of women reporting pain relief (OR 0.39, 95% CI 0.12 to 1.21, 42 women).
Secondary outcomes
2.2 Outcome: menstrual symptoms
2.2.1 Acupuncture versus NSAIDS
Women having acupuncture reported a lower score for menstrual symptoms than women having NSAIDs (MD ‐2.25 points, 95% CI ‐3.33 to ‐1.17, 6 trials, 440 women, I² = 89%, Analysis 2.3.1). Statistical heterogeneity was high but the direction of effect was consistent. These studies used the self‐report Clinical Principle of New Chinese Drug Clinical Research (scale unknown).
Three studies used other instruments and all reported reduced menstrual symptoms in the acupuncture group; however the scales of the measures used were unclear: Cao 2011 (MD ‐2.16, 95% CI ‐3.57 to ‐0.75, 53 women), Xu 2013 (MD ‐6.32, 95% CI ‐7.15 to ‐5.49, 45 women), and Zhi 2007 (MD ‐2.83, 95% CI ‐4.25 to ‐1.68, 64 women)
2.2.2 Acupuncture versus juyuansuan tamoxifen
Women having acupuncture reported a lower score for menstrual symptoms than women having juyuansuan tamoxifen (MD ‐4.22, 95% CI ‐5.12 to ‐3.32, one RCT, 80 women, Analysis 2.3.2).
2.3 Outcome: use of additional medication
2.3.1 Acupuncture versus combined oral contraceptives
One trial of acupuncture compared with COC found no evidence of a difference between the groups in the reduction from baseline in rescue medication used at three months (MD 0.56 tablets, 95% CI ‐1.03 to 2.15, 42 women, Analysis 2.4).
2.4 Outcome: quality of life
2.4.1 Acupuncture versus combined oral contraceptives
One trial of acupuncture compared with COC found no evidence of a difference between the groups in the quality of life at three months, measured using the SF‐36 (MD ‐9.14, 95% CI ‐19.23 to 0.95, one trial, 42 women; Analysis 2.5).
2.5 Outcome: adverse events
2.5.1 Acupuncture versus NSAIDs
The adverse event rate was lower in the acupuncture group than in the NSAIDs group (OR 0.10, 95% CI 0.02 to 0.44, 239 women, 4 trials, I² = 15%, low‐quality evidence) (Analysis 2.6.1).
2.5.1 Acupuncture versus combined oral contraceptives
There was no evidence of a difference in adverse events between COCs and acupuncture (OR 1.12, 95% 0.34 to 3.63, 1 trial, 52 women, Analysis 2.6).
Outcome: restriction of daily living activities
No trials reported on this outcome.
Outcome: absence from work or school
No trials reported on this outcome.
3) Acupuncture versus Chinese herbs
3.1 Primary outcome: Pain
3.1.1 Pain score
Acupuncture was associated with a lower pain score (on an unspecified scale) than Chinese herbs (MD ‐4.58, 95% CI ‐5.79 to ‐3.37, 1 RCT, 120 women, low‐quality evidence, Analysis 3.1).
Secondary outcomes
3.2 Outcome: menstrual symptom improvement
Women having acupuncture were more likely to report improvement in menstrual symptoms than those using Chinese herbs (OR 7.00, 95% CI 2.22 to 22.06, 120 women, 1 RCT, low‐quality evidence, Analysis 3.2).
3.3 Outcome: use of additional medication
No trials reported on this outcome.
3.4 Outcome: restriction of daily living activities
No trials reported on this outcome.
3.5 Outcome: absence from work or school
No trials reported on this outcome.
3.6 Outcome: quality of life
No trials reported on this outcome.
3.7 Outcome: adverse events
No trials reported on this outcome.
4) Acupuncture versus no treatment
4.1 Primary Outcome: Pain
4.1.1 Outcome: pain score
Six studies reported this outcome, using a VAS scale, after a single session of acupuncture (Shi 2011), three months (Song 2013; Xu 2014), four months (Sun 2011), and six months (Xu 2013; Zhang 2013b).
Data were unsuitable for pooling, in most cases because data were skewed. All studies reported a benefit in the acupuncture group. Details of findings at up to three months and rationale for not pooling are reported in Table 2.
| Study | Measure | Follow‐up | Acupuncture group Mean score | Control group Mean score | P < 0.05 | Comment |
| VAS | Post‐treatment | 30.05, SD 25.62 n = 20 | 48.2, SD 7.25, n = 10 | Yes | Data skewed | |
| VAS | 3 months | 4.1, SD 7.8, n = 25 | 25.23, SD 24.43, n = 48 | Yes | Data skewed | |
| VAS | 3 months | 42.62, SD 8.14, n = 28 | 62.77, SD 12.15, n = 28 | Yes | ||
| VAS | 3 months | 2.96, SD 0.31, n = 25 | 7.7, SD 0.3, n = 20 | Yes | This study reported implausibly low SDs | |
| VAS | 3 months | 3, SD 0.19, n = 48 | 8.13, SD 0.21, n = 48 | Yes | This study reported implausibly low SDs | |
| VAS | 3 months | 3.37, SD 1.64, n = 55 | 5.03, SD 2.37, n = 29 | Yes |
SD: standard deviation
Secondary outcomes
4.2 Outcome: menstrual symptoms
Five studies reported this outcome, using various measures, after a single session of acupuncture (Shi 2011), three months (Song 2013; Xu 2014), four months (Sun 2011), and six months (Xu 2013, Zhang 2013b).
Two of the studies reported data suitable for pooling, using the Retrospective Symptom Scale (RSS) (Sun 2011) or the CMSS (Bu 2011). When data were pooled using the SMD, the results suggested a moderate to large benefit in the acupuncture group (SMD ‐0.81, 95% CI ‐1.17 to ‐0.45, two RCTs, 132 women, I² = 0%).
The other three studies reported data that were unsuitable for analysis. All three studies reported a benefit in the acupuncture group. Details of the findings of these studies at up to three months and the rationale for not pooling are reported in Table 3
| Study | Measure | Follow‐up | Acupuncture group Mean score | Control group Mean score | P < 0.05 | Comment |
| CMSS | 3 months | 1.56, SD 0.87, n = 382 | 2.4, SD 0.81, n = 40 | Yes | This study reported implausibly low SDs | |
| RSS | 3 months | 11.56, SD 0.25, n = 48 | 36.76, SD 2.35, n = 48 | Yes | This study reported implausibly low SDs | |
| VAS | 3 months | 2.81, SD 0.16 n = 48 | 4.85, SD 0.19, n = 48 | Yes | This study reported implausibly low SDs. |
SD: standard deviation
4.3 Outcome: use of additional medication
No trials reported on this outcome.
4.4 Outcome: restriction of daily living activities
No trials reported on this outcome.
4.5 Outcome: absence from work or school
No trials reported on this outcome.
4.6 Outcome: quality of life
No trials reported on this outcome.
4.7 Outcome: adverse events
No trials reported on this outcome.
5) Acupuncture versus usual care
One study made this comparison (Witt 2008). The 'usual care' group was a wait list control, but women could use any additional conventional treatments they chose.
5.1 Primary Outcome: Pain
5.1.1 Outcome: pain score
Acupuncture was associated with a lower pain score than usual care, measured with a 0 ‐ 10 numerical rating scale (MD ‐2.09, 95% CI ‐2.99 to ‐1.19, 117 women, 1 RCT, high‐quality evidence) (Analysis 5.1).
Secondary outcomes
5.2 Outcome: menstrual symptoms
No trial reported on this outcome
5.3 Outcome: use of additional medication
No trial reported on this outcome.
5.4 Outcome: restriction of daily living activities
No trial reported on this outcome.
5.5 Outcome: absence from work or school
There was no evidence of a difference between groups in days of absence from school or work, measured by a self‐reported questionnaire (continuous) (MD 0.06, 95% CI ‐0.54 to 0.66, 117 women, 1 trial) (Analysis 5.2).
5.6 Outcome: quality of life
5.6.1 Outcome: quality of life: physical health
Acupuncture was associated with improved physical function compared with usual care, measured using the SF36 (MD 5.57, 95% CI 2.68 to 8.46; 117 women, 1 trial) (Analysis 5.3).
5.6.2.Outcome: quality of life: mental health
There was evidence that acupuncture improved mental health compared with usual care, measured using the SF36 (MD 10.49, 95% CI 3.63 to 17.35; 117 women, 1 trial) (Analysis 5.4).
5.6.3 Outcome: quality of life: bodily pain
There was evidence that acupuncture reduced bodily pain compared with usual care, measured using the SF36 (MD 20.10, 95% CI 9.90 to 30.30; 117 women, 1 trial) (Analysis 5.5).
5.6.4 Outcome: quality of life: general health
There was no evidence of a difference in general health compared with usual care, measured using the SF36 (MD 6.38, 95% CI ‐0.26 to 13.02, 117 women, 1 trial) (Analysis 5.6).
5.6.5 Outcome: quality of life: vitality
Acupuncture was associated with improved vitality compared with usual care, measured using the SF36 (MD 18.12, 95% CI 11.52 to 24.72; 117 women, 1 trial) (Analysis 5.7).
5.6.7 Outcome: quality of life: social function
There was evidence that acupuncture improved social function compared with usual care, measured using the SF36 (MD 20.27, 95% CI 11.52 to 29.02; 117 women, 1 trial) (Analysis 5.8).
5.6.8 Outcome: quality of life: emotional role
There was evidence that acupuncture improved emotional role compared with usual care, measured using the SF36 (MD 14.16, 95% CI 1.29 to 27.03; 117 women, 1 trial) (Analysis 5.9).
5.7 Outcome: adverse events
There was no evidence of a difference in self‐reported side effects in the acupuncture group compared with the usual care group (OR 0.27, 95% CI 0.05 to 1.34; 117 women, 1 trial) (Analysis 5.10).
6) Acupressure versus sham/placebo control
6.1 Primary outcome: pain
6.1.1 Outcome: pain score
Five studies reported this outcome at two days (Yeh 2013a), one month (Mirbagher‐Ajorpaz 2011), two months (Aghamiri 2005; Kashefi 2010) and three months (Bazarganipour 2010).
Two of the studies reported data suitable for analysis. We did not pool them because of high heterogeneity (I² = 98%). Both found that acupressure was associated with lower pain scores, measured on a VAS 0 ‐ 10 scale (MD ‐3.50 points, 95% CI ‐4.10 to ‐2.90, 1 trial, 100 women (Aghamiri 2005); MD ‐1.37 points, 95% CI ‐2.07 to ‐0.67, 1 trial, 81 women (Kashefi 2010); Analysis 6.1).
The other three studies reported data that were unsuitable for analysis. Two found a benefit in the acupressure group. Details of the findings of these studies are reported in Table 4. The third trial (Bazarganipour 2010) reported pain severity as categorical data; pain severity was lower in the acupressure group than in the control group (P < 0.001).
| Study | Measure | Follow‐up | Acupuncture group Mean score | Control group Mean score | P < 0.05 | Comment |
| VAS 0 ‐ 10 | 1 month | 1.66, SD 1.98, n = 15 | 4.9, SD 1.37, n = 40 | Yes | Skewed data | |
| VAS 0 ‐ 10 | 2 days | 2.03, SD 2.02, n = 50 | 3.36, SD 2.65, n = 50 | Yes | Skewed data |
SD: standard deviation
Secondary outcomes
6.2 Outcome: severity of menstrual symptoms
There was no evidence that acupressure improved menstrual symptoms using the Menstrual Distress Questionnaire compared with the placebo control (MD ‐1.84, 95% CI ‐3.71 to 0.03; 171 women, 2 RCTs, I² = 74%, low‐quality evidence, inconsistent direction of treatment effect) (Analysis 6.2).
6.3 Outcome: use of additional medication
No trial reported on this outcome.
6.4 Outcome: restriction of daily living activities
No trial reported on this outcome.
6.5 Outcome: absence from work or school
No trial reported on this outcome.
6.6 Outcome: quality of life
No trial reported on this outcome.
6.7 Outcome: adverse events
No trial reported on this outcome.
7) Acupressure versus medication
Primary outcome: pain
7.1.1 Outcome: pain score
Acupressure versus NSAIDs
One study reported this outcome (Zafari 2011).
One study reported a higher pain score at two months in women having acupressure than in those using ibuprofen, measured using a 0 ‐ 3 scale (MD 0.39, 95% CI 0.21 to 0.57, 136 women, 1 RCT) (Analysis 7.1).
Secondary outcomes
7.2 Outcome: severity of menstrual symptoms
No trial reported on this outcome.
7.3 Outcome: use of additional medication
No trial reported on this outcome.
7.4 Outcome: restriction of daily living activities
No trial reported on this outcome.
7.5 Outcome: absence from work or school
No trial reported on this outcome.
7.6 Outcome: quality of life
No trial reported on this outcome.
7.7 Outcome: adverse events
No trial reported on this outcome
8) Acupressure versus no treatment
8.1 Primary outcome: Pain
8.1.1 Outcome: Pain score
Two studies reported this outcome, at one month (Chen 2004) and three months (Chen 2010). There was no evidence of a difference between the acupressure and the no‐treatment groups in pain scores measured on a 0 ‐ 10 VAS scale (MD ‐0.96, 95% CI ‐2.54 to 0.62, 2 trials, 140 women, I² = 83%) (Analysis 8.1; Figure 5). Statistical heterogeneity was very high, with no obvious explanation, and findings should be regarded with caution.
Forest plot of comparison: 8 Acupressure versus no treatment, outcome: 8.1 Pain score (VAS 0 ‐ 10).
Secondary outcomes
8.2 Outcome: menstrual symptoms
Three studies reported this outcome (Chen 2004; Chen 2010; Wong 2010). There was no evidence of a difference between the groups in menstrual symptom scores, measured using the Menstrual Distress Questionnaire (MD ‐1.90, 95% CI ‐5.57 to 1.77, 160 women, 3 RCTs, I² = 81%). Statistical heterogeneity was very high and the direction of effect was inconsistent, so findings should be regarded with caution.
8.3 Outcome: use of additional medication
No trial reported on this outcome.
8.4 Outcome: restriction of daily living activities
No trial reported on this outcome.
8.5 Outcome: absence from work or school
No trial reported on this outcome.
8.6 Outcome: quality of life
No trial reported on this outcome.
8.7 Outcome: adverse events
No trial reported on this outcome.
Sensitivity analysis
We had planned, a priori, to undertake sensitivity analyses on the results to look at the possible contribution of differences in methodological quality, with trials of high quality (low risk of bias) compared to all trials. We did not pursue this, since only one was at an overall low risk of bias.
Subgroup analysis
We did not undertake any subgroup analysis, as there was little diversity by mode of stimulation of the included trials (five trials of manual acupuncture and one trial of electro‐acupuncture).
Reporting bias
There were insufficient studies to explore publication bias by visual scanning of a funnel plot.
Discussion
Summary of main results
We include 42 trials and data from 4640 women in this updated review.
Pain
For our primary outcome (pain), there was no consistent evidence of a difference between women receiving acupuncture and those receiving a sham control. The only study in the review that was at low risk of bias in all domains found no evidence of a difference between the groups at three, six or 12 months. However there was limited evidence to suggest a benefit for women receiving acupuncture compared with NSAIDs, juyuansuan tamoxifen, Chinese herbs, no treatment, or usual care. Comparisons of the effectiveness of acupuncture versus combined oral contraceptives were inconsistent. There was very limited evidence to suggest a benefit for acupressure compared with a sham control, but other findings for acupressure were inconsistent.
Secondary review outcomes
For our secondary review outcomes (which include menstrual symptoms, use of additional medication, restriction in activities, quality of life and adverse events), there was no consistent evidence of a benefit from acupuncture compared with a sham control. None of these studies reported adverse events.
Acupuncture versus other comparators offered very limited evidence to suggest that it was more effective than NSAIDs, juyuansuan tamoxifen or Chinese herbs for treating menstrual symptoms, was more effective than usual care for improving quality of life, and was less likely than NSAIDs to cause adverse events. There was no evidence of a difference between acupuncture and other comparators for other secondary review outcomes.
Acupressure versus other comparators found no evidence of a difference between the groups for any secondary review outcomes.
Overall completeness and applicability of evidence
Most of the trials did not report on outcomes other than pain, very few reported on adverse events or quality of life, and for many comparisons there were few data. There were also few trials with follow‐up of outcomes in the medium and long term, as most had follow‐up of three months or less.
There were variations in the eligibility criteria between trials, with the majority lacking detail in the reporting, or excluding some criteria for primary dysmenorrhoea. However, the majority of women in the community seeking acupuncture or acupuncture support are unlikely to have investigative procedures to exclude secondary dysmenorrhoea. Distinguishing between primary and secondary dysmenorrhoea was not well reported or undertaken in some trials. This may have impacted on the homogeneity of women included in the trials and the severity of symptoms, and could have potentially influenced the responsiveness to acupuncture. Trials recruited women from settings similar to those where women would access acupuncture or acupressure. Studies were conducted in different countries, and consequently reflects the different styles of acupuncture administered in the studies.
Twenty‐five of the trials were conducted in China and the applicability of their findings to women in other settings is unclear. The characteristics of acupuncture treatment, including variations in the duration, mode of stimulation, frequency, and selection of acupuncture points, may influence the quality of acupuncture and treatment effect. How acupuncture is practised varies significantly between countries, for example daily treatment is common in China while very rare in other countries. This is reflected in the differing treatment frequency of the included trials. It is possible that acupuncture may not have been therapeutically effective, and in some cases may not represent best clinical practice. In contemporary clinical practice, for example, the duration of treatment would generally be longer than one month or one menstrual cycle. Among acupuncturists trained in TCM, standardised prescriptions would not be used and treatment would be individualised based on the diagnosis, with acupuncture treatment varying between sessions. For those trials involving comparisons with medication, no data were reported on compliance, so it is unclear if these study groups could have been affected by suboptimal doses of medication. Placebo controls remain controversial, due to afferent activity arising from these applications.
Quality of the evidence
There remain almost no high‐quality trials of acupuncture and acupressure for the management of primary dysmenorrhoea, and only one study was assessed at a low risk of bias on all domains.
The quality of reporting was poor in 70% of the trials. Many trials were at a high risk of bias in relation to randomisation. Rates of follow‐up were high in the majority of trials, with only a small number of trials reporting a small loss of participants. For many studies blinding of participants and the practitioner was not possible, and reporting indicated that the outcomes could have been influenced by a lack of blinding, and consequently were rated at a high risk of bias. (Figure 2; Figure 3).
Continuous measures of effectiveness varied and were not always interpretable, and in many cases the data were skewed or unreliable, which precluded pooling them. Where pooling was possible, we found substantial unexplained heterogeneity for some outcomes. We were unable to explore the effect of study quality using a sensitivity analysis, due to the lack of studies at low risk of bias. We were also unable to undertake any subgroup analyses. However there was clear variation in study characteristics and delivery of the intervention that may contribute to this heterogeneity.
The quality of evidence using GRADE was low for the two outcomes of pain relief and menstrual symptoms (summary of findings Table for the main comparison). We downgraded the outcomes because of high statistical heterogeneity, imprecision and the risk of bias.
Potential biases in the review process
We attempted to minimise publication bias. Although our search was comprehensive and we included studies identified in languages other than English, we cannot rule out the possibility that we may have missed some studies.
Agreements and disagreements with other studies or reviews
There are three other reviews of acupuncture to treat primary dysmenorrhoea. Cho 2010 included nine trials out of 27 trials identified, Xu 2014b included two out of 20 trials identified and Yang 2008b identified 30 trials. Cho 2010, Yang 2008b and Xu 2014b included other modalities of TCM, which we excluded from our review, and included trials for which we were unable to ascertain the randomisation details or that we excluded because they did not meet the eligibility criteria. All three reviews found promising evidence for the use of acupuncture to treat primary dysmenorrhoea compared with pharmacological medicine or Chinese herbal medicine. The findings from the Yang 2008b review were more cautious, due to conflicting results, in part due to a greater number of trials included which added to the heterogeneity of the findings. The findings from all three reviews were influenced by the methodological flaws of the trials.
Study flow diagram: 2016 update
Methodological quality graph: review authors' judgements about each methodological quality item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Forest plot of comparison: 2 Acupuncture versus medication, outcome: 2.1 Pain score.
Forest plot of comparison: 8 Acupressure versus no treatment, outcome: 8.1 Pain score (VAS 0 ‐ 10).
Comparison 1 Acupuncture versus sham/placebo acupuncture, Outcome 1 Pain score.
Comparison 1 Acupuncture versus sham/placebo acupuncture, Outcome 2 Improvement in menstrual symptoms.
Comparison 1 Acupuncture versus sham/placebo acupuncture, Outcome 3 Use of additional medication.
Comparison 1 Acupuncture versus sham/placebo acupuncture, Outcome 4 Restriction of daily living activities.
Comparison 1 Acupuncture versus sham/placebo acupuncture, Outcome 5 Quality of life: physical health.
Comparison 1 Acupuncture versus sham/placebo acupuncture, Outcome 6 Quality of life: mental health.
Comparison 1 Acupuncture versus sham/placebo acupuncture, Outcome 7 Quality of life: bodily pain.
Comparison 1 Acupuncture versus sham/placebo acupuncture, Outcome 8 Quality of life: general health.
Comparison 1 Acupuncture versus sham/placebo acupuncture, Outcome 9 Quality of life: vitality.
Comparison 1 Acupuncture versus sham/placebo acupuncture, Outcome 10 Quality of life: social function.
Comparison 1 Acupuncture versus sham/placebo acupuncture, Outcome 11 Quality of life: role emotional.
Comparison 2 Acupuncture versus medication, Outcome 1 Pain score.
Comparison 2 Acupuncture versus medication, Outcome 2 Pain relief.
Comparison 2 Acupuncture versus medication, Outcome 3 Menstrual symptom score.
Comparison 2 Acupuncture versus medication, Outcome 4 Use of additional medication.
Comparison 2 Acupuncture versus medication, Outcome 5 Quality of life.
Comparison 2 Acupuncture versus medication, Outcome 6 Adverse events.
Comparison 3 Acupuncture versus Chinese herbs, Outcome 1 Pain score.
Comparison 3 Acupuncture versus Chinese herbs, Outcome 2 Menstrual symptom improvement.
Comparison 4 Acupuncture versus no treatment, Outcome 1 Pain score.
Comparison 4 Acupuncture versus no treatment, Outcome 2 Severity of menstrual symptoms.
Comparison 5 Acupuncture versus usual care, Outcome 1 Pain score.
Comparison 5 Acupuncture versus usual care, Outcome 2 Absence from work or school.
Comparison 5 Acupuncture versus usual care, Outcome 3 Quality of life physical health.
Comparison 5 Acupuncture versus usual care, Outcome 4 Quality of life: mental health.
Comparison 5 Acupuncture versus usual care, Outcome 5 Quality of life: bodily pain.
Comparison 5 Acupuncture versus usual care, Outcome 6 Quality of life: general health.
Comparison 5 Acupuncture versus usual care, Outcome 7 Quality of life: vitality.
Comparison 5 Acupuncture versus usual care, Outcome 8 Quality of life: social function.
Comparison 5 Acupuncture versus usual care, Outcome 9 Qualit of life: role emotional.
Comparison 5 Acupuncture versus usual care, Outcome 10 Adverse events.
Comparison 6 Acupressure versus sham/placebo, Outcome 1 Pain score (VAS 0 ‐ 10).
Comparison 6 Acupressure versus sham/placebo, Outcome 2 Severity of menstrual symptoms.
Comparison 7 Acupressure versus medication (NSAIDs), Outcome 1 Pain score.
Comparison 8 Acupressure versus no treatment, Outcome 1 Pain score (VAS 0 ‐ 10).
Comparison 8 Acupressure versus no treatment, Outcome 2 Menstrual symptoms (MDQ).
| Acupuncture compared with placebo, no treatment or active controls for primary dysmenorrhoea | ||||||
| Population: women with primary dysmenorrhoea Settings: outpatient, community Intervention: acupuncture Comparison: placebo , no treatment or active controls | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Control | Acupuncture | |||||
| Pain: Acupuncture versus sham or placebo control Follow up: 1‐12 months | Findings were inconsistent and inconclusive:
| 477 | ⊕⊕⊝⊝ | Data unsuitable for calculation of means | ||
| Adverse events: Acupuncture versus sham or placebo control | No studies reported this outcome | |||||
| Pain: Acupuncture versus NSAIDs Follow‐up: 1 ‐ 4 months | Pain scores were lower in the acupuncture group in six of seven studies (n = 389/449) that reported VAS scores, with the mean difference varying across studies from 0.64 to 4 points on a VAS 0 ‐ 10 scale [In 4 studies (n = 352) reporting rates of pain relief, rates were higher in women having acupuncture than in women having NSAIDs] | [OR 4.99 (2.82 to 8.82)] | 611 (10 RCTs) | ⊕⊕⊝⊝ | Continuous data unsuitable for pooling, I² = 94%. Consistent direction of effect | |
| Adverse events: Acupuncture versus NSAIDs Follow‐up: 3 months | 83 per 1000 | 9 per 1000 (2 to 38) | OR 0.10 (0.02 to 0.44) | 239 (4 RCTs) | ⊕⊕⊝⊝ | |
| Pain: Acupuncture versus no treatment Follow‐up: 1 ‐ 6 months | Six studies reported this outcome, using a VAS scale. The findings were consistent but unsuitable for pooling, in most cases because data were skewed. | 384 (6 RCTs) | ⊕⊕⊝⊝ | Data unsuitable for calculation of means | ||
| Adverse events: Acupuncture versus no treatment | No studies reported this outcome | |||||
| *The basis for the assumed risk is the median control group risk across studies. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1Downgraded one level for serious risk of bias: all or most studies at high or unclear risk of bias in one or more domains assessed. | ||||||
| Acupressure compared with placebo, no treatment or active controls for primary dysmenorrhoea | ||||||
| Population: women with primary dysmenorrhoea Settings: outpatient, community Intervention: acupressure Comparison: placebo, no treatment or active controls | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Control | Acupressure | |||||
| Pain: Acupressure versus sham or placebo control Follow‐up: 2 days to 3 months | All studies reported a benefit in the acupressure group. Two studies reported a mean benefit of 1 ‐ 3 points on a 0 ‐ 10 VAS scale | 538 (5 RCTs) | ⊕⊕⊝⊝ | Data were unsuitable for pooling. I² = 98% when the 2 studies with data suitable for analysis were pooled | ||
| Pain: Acupressure versus NSAIDs Follow‐up: 3 months | 1 study used a 0 ‐ 3 pain scale and reported a higher score (indicating more pain) in the acupressure group (MD 0.39 points, 0.21 to 0.57 higher). | 136 (1 RCT) | ⊕⊝⊝⊝ | |||
| Pain score: Acupressure versus no treatment Follow‐up: 1 ‐ 3 months | There was no conclusive evidence of a difference between the groups on a VAS 0 ‐ 10 scale. MD 0.96 points lower in the acupressure group (‐2.54 lower to 0.62 higher) | 140 (2 RCTs) | ⊕⊝⊝⊝ | Lower scores indicate less pain, I² = 83% | ||
| Adverse effects: Acupressure versus placebo , no treatment or active controls | No studies reported this outcome | |||||
| *The basis for the assumed risk is the median control group risk across studies. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1Downgraded one level for serious risk of bias: all or most studies at high or unclear risk of bias in one or more domains assessed. | ||||||
| Study | Measure | Follow‐up | Acupuncture group Mean score | Control group Mean score | P < 0.05 | Comment |
| VAS (0 ‐ 10) | 3 months | 2.12, SD 1, n = 50 | 4.24, SD 0.91, n = 50 | Yes | This study reported implausibly low SDs. We assumed that these might be standard errors, but recalculation based on this assumption resulted in skewed data | |
| Measure of Menstrual Pain Questionnaire (scale unclear) | 1 month | 4.46 | 4.58 | No | Total n = 63. No SDs reported | |
| McGill Questionnaire Pain scale (0 ‐ 5) (Melzack 1975) | 3 months | 2, SD 2.5, n = 46 | 2.6, SD 2.7, n = 46 | No P = 0.32 | Data skewed | |
| 6 months | 2.2, SD 3 | 2.8, SD 2.7 | No P = 0.11 | Data skewed | ||
| 12 months | 2.6, SD 2.7 | 2.8, SD 2.7 | No P = 0.60 | Data skewed | ||
| VAS (0 ‐ 10) | 3 months | 4.2, SD 7.28, n = 25 | 8.77, SD 16.3, n = 96 | Yes | Data skewed | |
| Measure of Menstrual Pain Questionnaire (scale unclear) | 3 months | MD 0.54, 95% CI 0.12 to 1.19 | No | Total n = 80. No primary data reported | ||
| SD: standard deviation | ||||||
| Study | Measure | Follow‐up | Acupuncture group Mean score | Control group Mean score | P < 0.05 | Comment |
| VAS | Post‐treatment | 30.05, SD 25.62 n = 20 | 48.2, SD 7.25, n = 10 | Yes | Data skewed | |
| VAS | 3 months | 4.1, SD 7.8, n = 25 | 25.23, SD 24.43, n = 48 | Yes | Data skewed | |
| VAS | 3 months | 42.62, SD 8.14, n = 28 | 62.77, SD 12.15, n = 28 | Yes | ||
| VAS | 3 months | 2.96, SD 0.31, n = 25 | 7.7, SD 0.3, n = 20 | Yes | This study reported implausibly low SDs | |
| VAS | 3 months | 3, SD 0.19, n = 48 | 8.13, SD 0.21, n = 48 | Yes | This study reported implausibly low SDs | |
| VAS | 3 months | 3.37, SD 1.64, n = 55 | 5.03, SD 2.37, n = 29 | Yes | ||
| SD: standard deviation | ||||||
| Study | Measure | Follow‐up | Acupuncture group Mean score | Control group Mean score | P < 0.05 | Comment |
| CMSS | 3 months | 1.56, SD 0.87, n = 382 | 2.4, SD 0.81, n = 40 | Yes | This study reported implausibly low SDs | |
| RSS | 3 months | 11.56, SD 0.25, n = 48 | 36.76, SD 2.35, n = 48 | Yes | This study reported implausibly low SDs | |
| VAS | 3 months | 2.81, SD 0.16 n = 48 | 4.85, SD 0.19, n = 48 | Yes | This study reported implausibly low SDs. | |
| SD: standard deviation | ||||||
| Study | Measure | Follow‐up | Acupuncture group Mean score | Control group Mean score | P < 0.05 | Comment |
| VAS 0 ‐ 10 | 1 month | 1.66, SD 1.98, n = 15 | 4.9, SD 1.37, n = 40 | Yes | Skewed data | |
| VAS 0 ‐ 10 | 2 days | 2.03, SD 2.02, n = 50 | 3.36, SD 2.65, n = 50 | Yes | Skewed data | |
| SD: standard deviation | ||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Pain score Show forest plot | 3 | Mean Difference (IV, Fixed, 95% CI) | Subtotals only | |
| 2 Improvement in menstrual symptoms Show forest plot | 1 | Odds Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 2.1 At three months | 1 | 92 | Odds Ratio (M‐H, Fixed, 95% CI) | 1.58 [0.61, 4.04] |
| 2.2 At six months | 1 | 92 | Odds Ratio (M‐H, Fixed, 95% CI) | 1.47 [0.62, 3.48] |
| 2.3 At 12 months | 1 | 92 | Odds Ratio (M‐H, Fixed, 95% CI) | 2.35 [0.80, 6.94] |
| 3 Use of additional medication Show forest plot | 1 | Odds Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 3.1 At three months | 1 | 92 | Odds Ratio (M‐H, Fixed, 95% CI) | 0.91 [0.40, 2.09] |
| 3.2 At six months | 1 | 92 | Odds Ratio (M‐H, Fixed, 95% CI) | 0.34 [0.14, 0.79] |
| 3.3 At 12 months | 1 | 92 | Odds Ratio (M‐H, Fixed, 95% CI) | 1.51 [0.62, 3.70] |
| 4 Restriction of daily living activities Show forest plot | 1 | 92 | Odds Ratio (M‐H, Fixed, 95% CI) | 0.72 [0.29, 1.81] |
| 5 Quality of life: physical health Show forest plot | 1 | 92 | Mean Difference (IV, Fixed, 95% CI) | ‐2.90 [‐6.33, 0.53] |
| 6 Quality of life: mental health Show forest plot | 1 | 92 | Mean Difference (IV, Fixed, 95% CI) | 4.40 [‐3.59, 12.39] |
| 7 Quality of life: bodily pain Show forest plot | 1 | 92 | Mean Difference (IV, Fixed, 95% CI) | ‐7.5 [‐16.71, 1.71] |
| 8 Quality of life: general health Show forest plot | 1 | 92 | Mean Difference (IV, Fixed, 95% CI) | 2.30 [‐6.98, 11.58] |
| 9 Quality of life: vitality Show forest plot | 1 | 92 | Mean Difference (IV, Fixed, 95% CI) | 2.20 [‐6.98, 11.38] |
| 10 Quality of life: social function Show forest plot | 1 | 92 | Mean Difference (IV, Fixed, 95% CI) | ‐0.5 [‐9.53, 8.53] |
| 11 Quality of life: role emotional Show forest plot | 1 | 92 | Mean Difference (IV, Fixed, 95% CI) | 0.0 [‐14.11, 14.11] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Pain score Show forest plot | 8 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 1.1 NSAIDs (VAS 0 ‐ 10 scale) | 7 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 1.2 Combined oral contraceptive (NRS 0 ‐ 10 scale) | 1 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 2 Pain relief Show forest plot | 5 | Odds Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 2.1 NSAIDs | 4 | 352 | Odds Ratio (M‐H, Fixed, 95% CI) | 4.99 [2.82, 8.82] |
| 2.2 COC | 1 | 52 | Odds Ratio (M‐H, Fixed, 95% CI) | 0.39 [0.12, 1.21] |
| 3 Menstrual symptom score Show forest plot | 7 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 3.1 NSAIDS | 6 | 440 | Mean Difference (IV, Random, 95% CI) | ‐2.25 [‐3.33, ‐1.17] |
| 3.2 Juyuansuan tamoxifen | 1 | 80 | Mean Difference (IV, Random, 95% CI) | ‐4.22 [‐5.12, ‐3.32] |
| 4 Use of additional medication Show forest plot | 1 | 42 | Mean Difference (IV, Fixed, 95% CI) | 0.56 [‐1.03, 2.15] |
| 5 Quality of life Show forest plot | 1 | 42 | Mean Difference (IV, Fixed, 95% CI) | ‐9.14 [‐19.23, 0.95] |
| 6 Adverse events Show forest plot | 5 | Odds Ratio (M‐H, Fixed, 95% CI) | Subtotals only | |
| 6.1 NSAIDs | 4 | 239 | Odds Ratio (M‐H, Fixed, 95% CI) | 0.10 [0.02, 0.44] |
| 6.2 COC | 1 | 52 | Odds Ratio (M‐H, Fixed, 95% CI) | 1.12 [0.34, 3.63] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Pain score Show forest plot | 1 | 120 | Mean Difference (IV, Fixed, 95% CI) | ‐4.58 [‐5.79, ‐3.37] |
| 2 Menstrual symptom improvement Show forest plot | 1 | 120 | Odds Ratio (M‐H, Fixed, 95% CI) | 7.00 [2.22, 22.06] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Pain score Show forest plot | 6 | Std. Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 2 Severity of menstrual symptoms Show forest plot | 2 | 132 | Std. Mean Difference (IV, Fixed, 95% CI) | ‐0.81 [‐1.17, ‐0.45] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Pain score Show forest plot | 1 | 117 | Mean Difference (IV, Fixed, 95% CI) | ‐2.09 [‐2.99, ‐1.19] |
| 2 Absence from work or school Show forest plot | 1 | 117 | Mean Difference (IV, Fixed, 95% CI) | 0.06 [‐0.54, 0.66] |
| 3 Quality of life physical health Show forest plot | 1 | 117 | Mean Difference (IV, Fixed, 95% CI) | 5.57 [2.68, 8.46] |
| 4 Quality of life: mental health Show forest plot | 1 | 117 | Mean Difference (IV, Fixed, 95% CI) | 10.49 [3.63, 17.35] |
| 5 Quality of life: bodily pain Show forest plot | 1 | 117 | Mean Difference (IV, Fixed, 95% CI) | 20.10 [9.90, 30.30] |
| 6 Quality of life: general health Show forest plot | 1 | 117 | Mean Difference (IV, Fixed, 95% CI) | 6.38 [‐0.26, 13.02] |
| 7 Quality of life: vitality Show forest plot | 1 | 117 | Mean Difference (IV, Fixed, 95% CI) | 18.12 [11.52, 24.72] |
| 8 Quality of life: social function Show forest plot | 1 | 117 | Mean Difference (IV, Fixed, 95% CI) | 20.27 [11.52, 29.02] |
| 9 Qualit of life: role emotional Show forest plot | 1 | 117 | Mean Difference (IV, Fixed, 95% CI) | 14.16 [1.29, 27.03] |
| 10 Adverse events Show forest plot | 1 | 117 | Odds Ratio (M‐H, Fixed, 95% CI) | 0.27 [0.05, 1.34] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Pain score (VAS 0 ‐ 10) Show forest plot | 2 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| 2 Severity of menstrual symptoms Show forest plot | 2 | 171 | Mean Difference (IV, Fixed, 95% CI) | ‐1.84 [‐3.71, 0.03] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Pain score Show forest plot | 1 | Mean Difference (IV, Random, 95% CI) | Subtotals only | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Pain score (VAS 0 ‐ 10) Show forest plot | 2 | 140 | Mean Difference (IV, Random, 95% CI) | ‐0.96 [‐2.54, 0.62] |
| 2 Menstrual symptoms (MDQ) Show forest plot | 3 | 160 | Mean Difference (IV, Random, 95% CI) | ‐1.90 [‐5.57, 1.77] |
