The efficacy and safety of acupuncture in women with primary dysmenorrhea

Supplemental Digital Content is available in the text


Introduction
Primary dysmenorrhea is defined as cramping pain during menstruation without any identifiable pelvic pathology, [1] and it affects most women throughout the menstrual years. [2] Many studies have reported that the prevalence of primary dysmenorrhea varied from approximately 50% to 90%, [3][4][5][6] and 13% to 51% had to limit daily activities, such as school or work absenteeism. [2] In the consensus guidelines of primary dysmenorrhea, [7] nonsteroidal anti-inflammatory drugs (NSAIDs) and oral contraceptives (OCs) are recommended as first-line treatments. However, some patients did not experience pain reduction with NSAIDs and did experience side effects such as nausea, dyspepsia, headache, or drowsiness. [8,9] In addition, OCs may not be suitable for patients attempting to become pregnant, and might cause adverse effects such as nausea, vomiting, weight gain, or vaginal bleeding. [10,11] Acupuncture, derived from China, is a therapeutic modality using the insertion of fine needles with the concepts of Yin and Yang and the circulation of qi. Acupuncture acts primarily by stimulating the nervous system, by local effects due to local antidromic axon reflexes, and by releasing opioid peptides and serotonin. Today, acupuncture is regarded as part of conventional medicine. It is no longer only "alternative medicine," and it is used in Western medicine. [12] In particular, acupuncture has been widely used to alleviate diverse pains [13] including menstrual pain.
Many clinical trials had been conducted to show efficacy of acupuncture on menstrual pain, and 6 systematic reviews (SRs) have been previously conducted to evaluate the efficacy of acupuncture on primary dysmenorrhea. [14][15][16][17][18][19] However, the previous SRs included acupressure, the stimulation of acupoints without skin penetration, [14][15][16][17][18] which made the evaluation of acupuncture difficult. Some studies analyzed all types of acupuncture together, [14][15][16][17] which increased the heterogeneity. One latest study [19] included all the types of acupuncture except acupressure and analyzed the results separately, but it did not include newly published studies in 2017. Thus, we found it necessary to conduct a study with rigorous criteria thatexcludedacupressureandincluded allothertypesofacupuncture that penetrate the skin, such as embedding therapy, and to synthesize the data according to the type of acupuncture to reduce heterogeneity. We conducted this study with these criteria to determine the efficacy and safety of acupuncture on primary dysmenorrhea.

Study registration
The protocol for this study was registered in PROSPERO: CRD42017069258.
2.2. Eligibility criteria 2.2.1. Types of studies. We included all randomized controlled trials (RCTs) that measured pain intensity and related outcomes to evaluate the efficacy of acupuncture in women with primary dysmenorrhea. Case studies, case series, noncontrolled trials, review articles, letters, conference papers, abstracts, and poster presentations were excluded. Studies not written in English, Chinese, or Korean were also excluded.

Types of participants.
We included female patients of reproductive age suffering from primary dysmenorrhea. The definition of primary dysmenorrhea was based on cyclic pelvic pain during menstruation without any gynecological pathology such as endometriosis, adenomyosis, or uterine myoma. Patients with secondary dysmenorrhea or serious medical conditions were excluded.

Types of interventions.
Manual acupuncture (MA), electroacupuncture (EA), auricular acupuncture (AA), and any other type of acupuncture using needle insertion were included in our study. Pharmacopuncture and acupressure were excluded. Other types of acupuncture that are rarely used in Korean clinical practice, such as eye acupuncture and floating acupuncture were also excluded. Types of control interventions included in our studies were no treatment, placebo acupuncture, and oral medications such as NSAIDs and OCs. Herbal medicines or other traditional medicine treatments used in the control group were excluded from our study.

Outcomes.
The primary outcome was pain intensity after the intervention period as measured by any validated scale, such as the visual analog scale (VAS) or numeric rating score (NRS). The secondary outcomes were pain relief measured by total effective rate (TER) or improvement rate; related symptoms measured by the seven-point verbal rating scale (VRS), Cox menstrual symptom scale (CMSS), Cox retrospective symptom scale (RSS), or menstrual symptom score (MSS); quality of life as measured by the 36-item Short Form health survey (SF-36); pain intensity after a follow-up period; and adverse events (AEs).

Study selection
All studies found based on the search results were saved into EndNote; duplicated studies were excluded. After deleting the duplicates, 3 reviewers, WHL, HSJ, and LHJ, selected the relevant studies independently by title and abstract, and finally selected the included studies using the full text. Any disagreements were resolved by discussion among the 3 reviewers and an arbiter, PKS.

Data extraction
Three authors, WHL, HSJ, and LHJ, extracted data from the included studies according to the predetermined data forms. The following items were extracted: baseline demographics (journal, author, and year of publication); participants (sample size, sex, and age); intervention (type of acupuncture, periods, and frequency of treatment, and follow-up period); control; and outcome.

Risk of bias assessment
WHL, HSJ, and LHJ independently assessed the risk of bias for each included study using the following criteria from the Cochrane Handbook for Systematic Reviews of Interventions [20] : random sequence generation; allocation concealment; blinding of participants and personnel; blinding of outcome assessment; incomplete outcome data; and selective reporting. We assessed these 6 criteria using "Low" ("L"), "Unclear" ("U"), and "High" ("H") as a key for judgements. "Low" indicated a low risk of bias, "Unclear" indicated that the risk of bias was uncertain, and "High" indicated a high risk of bias. Disagreements were resolved by discussion among the 3 reviewers and an arbiter, Park KS.

Data synthesis
In our review, for studies using the same type of acupuncture, comparator, and outcome measures, the meta-analysis was performed using Review Manager software (RevMan v. 5.3). To assess the effect of acupuncture on primary dysmenorrhea, dichotomous data were analyzed using a risk ratio (RR) with 95% confidence intervals (CIs), and continuous data were analyzed using mean differences (MD) and 95% CIs or standardized mean differences (SMD) with 95% CIs if different scales were used. The chi-square and I 2 tests were used to assess statistical heterogeneity. [20] If I 2 > 50% or P < .1, we considered that there was substantial heterogeneity among the trials, and if I 2 > 75%, we considered that there was serious heterogeneity. When serious heterogeneity was indicated, we found sources of heterogeneity by subgroup or sensitivity analysis. Subgroup analysis was conducted according to the treatment periods, and sensitivity analysis was done by excluding each heterogeneous trial. In case of substantial heterogeneity, a random effects model was used; otherwise, a fixed effects model was used to synthesize the data. However, if there were few studies for pooling, a fixed effects model was implemented because it is difficult to obtain a precise estimate of the between-studies variance. [21] If the number of the appropriate studies was only 1, or data were unsuitable for quantitative synthesis, descriptive synthesis of the findings was performed. If the number of studies for pooling was more than 10, publication bias was assessed using a funnel plot. [22] 3. Results

12-18
Once per day 3-5 days before menstruation started to the second day of menstruation started, 3 cycles After pain alleviated.

min
Oral administration of ibuprofen 1 T 3 times daily for 3 days before menstruation started.
VRS. Two studies [73,76] reported that there was no significant difference between the groups. Data were unsuitable for pooling www.md-journal.com because they reported the results only in graphs, which made it hard to extract raw data.
RSS. Three studies [47,73,76] reported this outcome. Two studies [73,76] showed there was no significant difference in RSS, but the other study [47] showed that EA was significantly more effective than no treatment in RSS-COX2 (P < .05). Data were unsuitable for pooling because they reported the results only as graphs, which made it hard to extract raw data.
VRS. Three studies [73,75,76] reported VRS, but only 2 [75,76] were included in the meta-analysis because the third did not provide SDs. As shown in Figure 5B, after treatment of 1 menstrual cycle, the VRS in the EA group was lower than the placebo group, but there was no significance (n = 347, MD = À 0.20, 95% CI [À0.43, 0.03], P = .10, I 2 = 61%). The other study [73] also reported a change of VRS in the EA group that was lower than the placebo group, but there was no significance (n = 322, reduction from 3.94 to 3.08 vs reduction 3.72 to 3.02).
AEs. Five studies [47,73,75,76,78] reported AEs. Three studies [47,76,78] reported there were no AEs, and one [73] of the other studies reported one case of dizziness after EA. The other study [75] reported one case of minimal bleeding in the EA group, and one case of minimal bleeding and one case of pain after insertion in the placebo group.
AEs. Two studies [54,65] reported AEs. One study [54] reported 5 cases of nausea, vomiting, and fever in the NSAIDs group, and the other study [65] reported there were no AEs.
AEs. One study [71] reported 6 cases of gastrointestinal discomforts, headache, dizziness, and insomnia in the NSAIDs group.

Summary of the main results
This systematic review was aimed to summarize and evaluate acupuncture treatment to reduce menstrual pain and its associated symptoms. As a result, we suggest that acupuncture might have beneficial effects for improvement of dysmenorrhea and remain efficacious after short-term follow-up.
We conducted comparisons separately according to the characteristics of interventions and controls. MA was significantly more effective than no treatment, and NSAIDs for reduction of menstrual pain and its associated symptoms, and remained effective after a short-term follow-up compared to no treatment and NSAIDs. The MA-induced analgesic effect could be explained by C-fiber involvement during the practitioners' manipulation for the de-qi response. [85] However, no significant difference was observed between MA and placebo acupuncture or between MA and OCs. It was difficult to determine the superior effect of OCs compared to MA because there was only one relevant study. [11] The results showed that EA was significantly more effective at reducing menstrual pain than no treatment, [47,73,76,78] placebo acupuncture, [40,48,[73][74][75][76]78,79] but not effective at improving its associated symptoms. [47,73,75,76] The results comparing with NSAIDs were insufficient to determine the efficacy of EA. The mechanism of EA-induced analgesia could be explained by inducing the release of endorphins [86,87] and the decrease of the pulsatility index in the uterine arteries, [88] which might be related to primary dysmenorrhea. [1] The reason that there was no difference between MA and placebo acupuncture and the relatively small difference between EA and placebo acupuncture was thought to be that placebo acupuncture also had positive effects. Several factors might explain the positive effects. First, some participants receiving placebo acupuncture may want pain relief, and it may affect the outcome psychologically. [89] Second, placebo acupuncture may stimulate cutaneous touch receptors and/or skin nociceptors and modulate the activity in the brain areas associated with pain management. [90] WA was significantly more effective at reducing menstrual pain than NSAIDs, but the efficacy for the associated symptoms was inconclusive due to the small sample size. The results showed WA with NSAIDs might also relieve menstrual pain compared to NSAIDs alone. WA increases the circulation of qi and blood through the needle body during thermal heating. It provides analgesic effects by stimulating nerve transfer and relaxing uterine muscle spasms. [91] CET might also be effective for primary dysmenorrhea. CET is a therapeutic modality based on acupuncture theory and continuous stimulation of acupoints with embedded thread, and its continuous stimulation prolongs the effects of acupuncture. In addition, the embedded thread gradually liquefies and is absorbed, and stimulates the points physically and chemically. [26] With this mechanism, CET might be considered to demonstrate analgesic effects and maintain the effects for short-term follow-up.
Severe AEs of acupuncture were not observed. Thirteen of the 60 studies reported AEs of acupuncture. Most of the reported AEs were regional pain or discomfort, hematoma, and dizziness. Those mentioned were mild, similar to previously known AEs. [92] The applicability of acupuncture to primary dysmenorrhea in other settings is unclear. Fifty-seven of the trials were conducted in Asian countries: 55 in China, 1 in Thailand, and 1 in South Korea. The acupuncture practitioners might have different treatment skills according to the nations in which they were trained, and the participants might have different preconceptions and familiarity with acupuncture according their cultures. [89] In addition, the variability of the details of interventions and controls could make applicability unclear.

Strengths and limitations of this review
Six SRs which evaluate the efficacy of acupuncture on primary dysmenorrhea have previously been conducted, [14][15][16][17][18][19] and 2 of them were published in 2016 [17] and 2017 [19] , respectively. However, there were some differences between these 2 SRs and www.md-journal.com our review. They may arise from the different search strategies, inclusion criteria, and analysis methods. In particular, the Cochrane review [17] analyzed 42 studies, just separating the treatment types into acupuncture and acupressure. Liu et al's [19] review analyzed 23 studies with similar strategies to our review, did not include 10 trials newly published in 2017, and did not include other modalities of acupuncture such as WA or CET, frequently used in clinical fields. Our review included all types of acupuncture that stimulate acupoints by penetrating the skin, including CET, and synthesized data separately according to the characteristics of the interventions and controls. Our study had some limitations, and those results mentioned above should be interpreted with caution. One was that most of the included trials achieved a low or unclear risk of bias. The unclear judgements appeared mostly in the domains of allocation concealment and blinding of participants/practitioners/outcome assessors, because the details were not described. The blinding of participants is critical for subjective outcomes such as pain, [93] but blinding of both participants and practitioners was difficult due to the characteristics of acupuncture intervention. The other limitation was that there was substantial heterogeneity among the pooled trials. We tried to reduce the heterogeneity by synthesizing the data separately depending on the characteristics of the interventions and controls, subgroup analysis, and sensitivity analysis, but the unresolved heterogeneity in some cases still existed. We considered this heterogeneity derived from the small sample sizes in some outcomes and the methodological variations among the included studies. The methods of interventions varied in the frequency, duration of each session, selection of acupoints, and de-qi methods. The variations of controls also appeared in different components of NSAIDs. These variations could influence the results of the trials, and were considered to cause unresolved heterogeneity.

Implications of this review for practice and research
To provide convincing evidence of the efficacy of acupuncture for primary dysmenorrhea, future RCTs should adhere to rigorous standards assessing the risk of bias, such as conducting randomization allocation concealment and trying to avoid performance bias. In addition, those trials should be reported as STRICTA guidelines [84] to clear the specific method of each intervention.

Conclusions
The results of this study suggest that acupuncture might reduce menstrual pain and associated symptoms more effectively compared with no treatment or NSAIDs, and the efficacy could be maintained during a short-term follow-up period. However, the efficacy of acupuncture compared to a placebo was not convincing. The safety of acupuncture appeared because a few mild AEs were reported. Our suggestions had limitations because the quality of the included RCTs was low, and methodological restriction existed in this study. More rigorously designed trials are required to confirm our findings.