Efficacy and safety of acupuncture-related therapies in symptomatic endometriosis: a systematic review and network meta-analysis

Abstract

Objective

To compare the effectiveness and safety of different acupuncture-related therapies combined with pharmacotherapies for treating symptomatic endometriosis.

Methods

A pre-defined search strategy was conducted across eight databases (Chinese Biomedical Literature Service System, China National Knowledge Infrastructure, Wanfang Data Knowledge Service Platform, China Science and Technology Journal Database, PubMed, Embase, Cochrane Library, and Web of Science) from inception to May 1, 2023. The included studies were evaluated for methodological quality using the Cochrane risk of bias assessment tool. The surface under the cumulative ranking (SUCRA) was applied to rank the acupuncture-related therapies for each outcome.

Results

Twenty-three randomized controlled trials (RCTs) involving one thousand five hundred forty-five patients were included in the network meta-analysis. Ear electroacupuncture (SUCRA = 83.0%), needle warming moxibustion with Modified Neiyi Zhitong Formula (SUCRA = 80.6%), and auricular needle-embedding (SUCRA = 79.6%) demonstrated a statistically significant reduction in comprehensive symptoms compared to the control group. Studies have shown that body electroacupuncture (OR = 4.33, 95% CI 1.20–15.61), acupoint catgut (OR = 4.32, 95% CI 1.08–17.25), and auricular needle-embedding (OR = 7.56, 95% CI 1.89–30.28) are statistically significantly more effective than conventional treatment.

Conclusion

The results of this analysis suggest that acupuncture-related therapies are effective in managing symptomatic endometriosis. Further high-quality randomized controlled trials are warranted to explore their efficacy and safety in greater depth.

Trial registration

Our study protocol was registered with the International Platform of Registered Systematic Review and Meta-analysis Protocols (INPLASY); registration number: INPLASY202380077.

Supplementary Information

The online version contains supplementary material available at 10.1007/s00404-025-07979-8.

What does this study add to the clinical work

The study provides summarized evidence on acupuncture-related therapies combined with pharmacotherapies for treating symptomatic endometriosis. The study also discovered that ear electro-acupuncture, when used as the sole or main intervention, has potential efficacy and safety for patients with symptomatic endometriosis.

Introduction

Endometriosis (EMs) is a significant public health issue, affecting approximately 10% to 15% of women of childbearing age [1]. This condition is a chronic inflammatory disorder characterized by the presence of endometrial tissue outside the uterus, manifesting in symptoms such as pelvic abdominal pain, dysmenorrhea, infertility, dyspareunia, pelvic nodules, and endometriomas. Among these, pain is the primary clinical indicator of infertility and disease severity in EMs. As a systemic disease marked by widespread inflammation and an unclear etiology, it is challenging to eliminate, prone to recurrence, and carries a potential risk of malignant transformation [2]. The pain associated with EMs severely impacts patients’ quality of life.

Western medical approaches to EMs include medication, surgical intervention, and expectant management; however, these therapies often result in adverse effects such as menstrual irregularities, gastrointestinal discomfort, or allergic reactions [3, 4]. Recently, a combination of pharmacological and non-pharmacological treatments has been recommended, including acupuncture techniques (auricular, electroacupuncture, body acupuncture, and needle warming moxibustion), moxibustion, Neiyi Zhitong Formula, Guixiang Wenjing Zhitong capsules, and Tong Bining for managing endometriosis-related pain [5, 6]. Research indicates that acupuncture and moxibustion can promote pelvic circulation, alleviate uterine smooth muscle spasms, regulate hormones, and relieve symptoms related to pelvic inflammatory disease and menopausal syndrome [7]. Despite these advancements, direct comparisons of the efficacy and safety of various acupuncture and moxibustion methods remain limited, making it difficult to comprehensively evaluate their clinical benefits.

The clinical efficacy rate (CER) for endometriosis assesses the extent of symptom relief, including improvements in abdominal pain, dysmenorrhea, abnormal menstruation, dyspareunia, and endometriomas. Furthermore, the Visual Analog Scale (VAS) scores derived from existing literature have been systematically ranked to offer evidence-based recommendations for both patients and clinicians.

Materials and methods

Registration

This systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Table S1) [8]. The study has been registered with INPLASY under the registration number INPLASY202380077.

Literature search

A comprehensive search was performed across eight electronic databases: Chinese Biomedical Literature Service System, China National Knowledge Infrastructure, Wanfang Data Knowledge Service Platform, China Science and Technology Journal Database, PubMed, Embase, Cochrane Library, and Web of Science. The search covered the period from the inception of each database to May 1, 2023. Clinical randomized controlled trials comparing various acupuncture and moxibustion techniques for symptomatic endometriosis were included. The detailed search strategy is presented in Table 1, with PubMed used as an example; the specific search terms and approach are provided in S1 Appendix.

Table 1.

Search strategy on PubMed

#1 ((((endometriosis[MeSH Terms])) OR (endometrioses[Title/Abstract])) OR (endometrioma[Title/Abstract])) OR (endometriomas[Title/Abstract])

#2 (((((Needles[MeSH Terms]) OR (Acupuncture[MeSH Terms])) OR (Acupuncture Therapy[MeSH Terms])) OR (Moxibustion[MeSH Terms])) OR (Electroacupuncture[MeSH Terms])) OR (Acupuncture Points[MeSH Terms])

#3 (((((((((((Pharmacopuncture[Title/Abstract]) OR (needle[Title/Abstract])) OR (Acupuncture Treatment[Title/Abstract])) OR (Acupuncture Treatments[Title/Abstract])) OR (Treatment, Acupuncture[Title/Abstract])) OR (Therapy, Acupuncture[Title/Abstract])) OR (Moxabustion[Title/Abstract])) OR (moxibustion therapy[Title/Abstract])) OR (abdominal acupuncture[Title/Abstract])) OR (fire needle[Title/Abstract])) OR (warming needle moxibustion[Title/Abstract])) OR (herb-partition moxibustion[Title/Abstract])

#4 (((((((Acupuncture Point[Title/Abstract]) OR (Point, Acupuncture[Title/Abstract])) OR (Points, Acupuncture[Title/Abstract])) OR (Acupoints[Title/Abstract])) OR (Acupoint[Title/Abstract])) OR (Shiatsu[Title/Abstract])) OR (Zhi Ya[Title/Abstract])) OR (Chih Ya[Title/Abstract])

#5 ((acup*?point*[Title/Abstract]) OR (acup*[Title/Abstract]))

#6 #2 OR #3 OR #4 OR #5

#7 ((((Random Allocation[MeSH Terms]) OR (Randomized Controlled Trial[MeSH Terms])) OR (Allocation, Random[MeSH Terms])) OR (Randomization[MeSH Terms])) OR (RCT)

#8 #1 AND #6 AND #7

Study selection

Participants: included patients met the diagnostic criteria for endometriosis [9, 10]. (2) Interventions: the treatment group underwent various acupuncture or moxibustion therapies. (3) Control interventions: the control groups received placebo, sham acupuncture, conventional treatment, traditional medicine, or no treatment. (4) Outcomes: the outcomes included CER, VAS scores, and adverse events. The studies focus on pain evaluation in EMs as the primary endpoint. Adverse events were defined as side effects unrelated to the intended purpose of acupuncture treatments. CER was assessed following the “Guiding Principle for Clinical Study of New Chinese Drugs” [11], categorized into four grades: recovery, evident, effective, and effectless. Recovery: abdominal pain, dysmenorrhea, abnormal menstruation, dyspareunia, pelvic masses, and related symptoms completely resolved (VAS score drops to 0, with no recurrence after ceasing treatment for one menstrual cycle). Evident: symptoms were significantly alleviated (VAS score reduced by more than 1/2), allowing patients to function without pain medication. Effective: symptoms showed improvement (VAS score reduced by 1/4 to 1/2), with no worsening within three months post-treatment. Effectless: symptoms did not substantially change or worsened (VAS score reduced by less than 1/4). Study types: only clinical randomized trials were included.

Exclusion criteria

(1) Studies categorized as reviews, case reports, systematic reviews, experience summaries, animal studies, or case observations. (2) Duplicate publications, repeated citations, or papers using similar methodologies and results recently published by the same author, with only the most authoritative selected. (3) Studies failing to evaluate the effects of acupuncture and moxibustion interventions on endometriosis. (4) Studies with incomplete or invalid data. (5) Studies with clear methodological errors, such as statistical flaws or improper randomization.

Literature management

Initial screening and exclusion of the literature were performed using NoteExpress software. Relevant data were recorded and summarized in Excel tables for further analysis. A third researcher was involved to resolve any discrepancies in the data. The remaining studies were reviewed independently by two researchers following the planned process. The inclusion results were cross-checked, and disagreements were resolved through discussion or, when necessary, a final decision by a third researcher.

Data extraction and quality assessment

The extracted data included the following information—(1) study details: title, author, publication year, methodology, study type, and sample size. (2) Participant characteristics: baseline data such as age, gender, disease duration, and follow-up period. (3) Interventions and indicators: specific interventions, randomization methods, diagnostic criteria, disease progression, treatment duration, follow-up time, and adverse reactions. This also included clinical outcome indicators and factors influencing the risk of bias. Two investigators (YS and HZ) independently screened each article, with both blinded to each other’s findings. Based on the pre-defined inclusion criteria, the reviewers rigorously screened and selected eligible articles. Data extraction from these studies was conducted using a standardized data collection form. Disagreements between reviewers were resolved through discussion until a consensus was reached. If unresolved, the third review author (XYZ) was consulted. The risk of bias was independently assessed by two researchers using the Cochrane Collaboration’s criteria for evaluating trial quality [12]. The following seven items were considered: (1) random sequence generation, (2) allocation concealment, (3) double-blinding of participants and personnel, (4) blinding in outcome assessment, (5) withdrawals and loss to follow-up, (6) selective reporting, (7) other sources of bias. Each item was classified as having a low, uncertain, or high risk of bias [13].

Statistical analysis

Odds ratios were used to represent categorical data and bicategorical variable outcomes, while mean differences were applied for continuous variables. Both effect sizes and their 95% confidence intervals were used to report odds ratios and mean differences [14]. Heterogeneity tests, network meta-evidence maps, and contribution maps were generated using Stata software (version 15.1), JAGS (version 4.3.0), and RStudio (version 4.3.1) for data analysis.

Network meta-analysis (NMA) was conducted using Stata software (version 15.1) with polymerization and analysis performed through Markov chain Monte Carlo simulation chains in a Bayesian framework, following the PRISMA NMA guidelines. According to the Stata software instructions, a P value greater than 0.05 indicates the consistency test is passed [14–16].

The validity and convergence of the model were assessed through convergence diagnostics. The fixed-effect consistency model was selected for correlation testing. The model achieved convergence prior to 50,000 iterations, demonstrating stability and eliminating the need for additional iterations or calculations. Convergence diagnostics are presented in Figs. 1 and 2.

Fig. 1.

Convergence diagnostics of CER

Fig. 2.

Convergence diagnostics of VAS

The intervention hierarchy was summarized and reported using a P score [17]. The P score ranges from 0 to 1, where a value of 1 represents the best treatment without uncertainty, while 0 represents the worst treatment without uncertainty. The P score or surface under the cumulative ranking curve (SUCRA) can also be interpreted as the percentage effectiveness or acceptability of an acupuncture intervention. However, such scores should be interpreted cautiously unless clinically significant differences between interventions are evident. To assess potential bias in smaller studies, network meta-analysis (NMA) funnel plots were generated to examine publication bias. These plots were visually evaluated using symmetry criteria [18].

Consistency and inconsistency tests were performed on P values of all direct and indirect comparisons between all studies, and most P values were greater than 0.05, indicating that the effect of consistency between included studies was acceptable. Details will be marked in S2 and S3 Tables.

The heterogeneity test revealed an I² value of 34% (< 50%) and a P value of 0.13 (> 0.1) in the Q test, indicating no significant heterogeneity among the included studies. This justified the use of a fixed-effects model for meta-analysis. To ensure the study’s accuracy and stability, a sensitivity analysis was conducted. Among the nine studies involving VAS data, Tian L (2022) [19], Jianfang N (2020) [20], and Qiaoyi P (2019) [21] were found to significantly influence heterogeneity. After excluding these studies, the heterogeneity test of the remaining six articles again showed no heterogeneity (I² = 37% < 50%, P = 0.16 > 0.1). A fixed-effects meta-analysis was subsequently performed only after these exclusions. The detailed results are presented in Figs. S1 and S2.

Results

Study description

A total of 8835 studies were retrieved from 8 databases and imported into NoteExpress (version 3.2.0) and Excel for literature management. After removing 302 duplicate studies, excluding 7830 non-compliant RCTs, and screening 680 titles and full texts deemed inappropriate, 23 studies involving 1776 patients met the inclusion criteria. The literature search process is illustrated in Fig. S3.

The details of the included studies are summarized in Table 2. Control group interventions comprised the following: Shu-Mu acupuncture (one study) [22], acupuncture (two studies) [23, 24], needle warming moxibustion (two studies) [25, 26], needle warming moxibustion combined with Modified Neiyi Zhitong Formula (one study) [25], auricular needle-embedding (one study) [27], ear electroacupuncture (one study)[28], body electroacupuncture (two studies) [28, 29], acupoint catgut embedding (three studies) [30–32], moxa stick moxibustion (two studies) [20, 29], and laser moxibustion (one study) [19]. In the included studies, acupuncture refers to the insertion of filiform needles into the skin at specific acupuncture points on one side of the patient’s body. The procedure involved an even reinforcing-reducing technique based on the patient’s sensations, such as aching, distension, or numbness. Ten studies reported CER as an outcome indicator, nine studies reported VAS as an outcome indicator, and four studies provided data on adverse events.

Table 2.

Basic information about the included studies

First author	Year	Total/experiment/control	Age/experiment/control	Intervention	Prescription	Outcome
Dongfang [33]	2011	T: 30 C: 28	T: 34.6 ± 5.9 C: 34.6 ± 4.9	T: Abdominal acupuncture C: CT (Chinese proprietary medicine: Tianqi Tongjing Capsules ) Length of Intervention:3 menstrual cycles Freq: 3 times a week Duration: 20 min	Zhongwan, Xiawan, Qihai, Guanyuan, Zhongji	CER; McGill; Serum CA125
Sun [22]	2006	T: Shu-Mu acupuncture, acupuncture: respectively 30 C: 30	T: Shu-Mu acupuncture: 34.70 ± 8.71 Acupuncture: 35.68 ± 7.12 C: 36.24 ± 6.78	T: Shu-Mu acupuncture, acupuncture C: CT (Western medicine) Length of intervention: 3 circles, 1 circle by 20d, interval 10d each circle Freq: 7 times a week Duration: 30 min	Shu-Mu: Ganshu, Pishu, Shenshu, Qimen, Zhangmen, Jingmen Acupuncture: Hegu, Zhongji, Guanyuan, Sanyinjiao	CER; B-scan pelvic mass; Serum CA125; Adverse events
Hu [34]	2020	T: 18 C: 18	T: – C: –	T: Herbs-partition moxibustion C: Blank Length of intervention:3 months Freq: T: 1 time per week Duration: T: 1 h	Hypogastrium: Shenshu, Guanyuan, Qugu Lumbosacral Region: Mingmen, Yaoshu	CER; Serum CA 125; SF-36; Adverse events
Rubi-Klein [23]	2010	T: 47 C: 54	T: – C: –	T: Acupuncture C: CT (non-specific acupuncture) Length of intervention: 5 weeks Freq: 2 times a week Duration: –	Acupuncture: Ciliao, Guilai, Zusanli, Zhongji, Sanyinjiao, Taichong, Ququan, Yinlingquan, Xuehai, Yingu CT: Jianzhen, Fengshi, Zhongfu, Touwei	VAS; PDI
Liang [24]	2018	T: 53 C: 53	T: – C: –	T: Acupuncture C: CT (non-acupoint acupuncture) Length of intervention: 3 menstrual cycles Freq: 3 times a week Duration: 30 min	Acupuncture: Sanyinjiao, Zhaohai, Taichong, Qichong, Guanyuan CT: less than 5 mm on each shoulder and upper arm surface	VAS; EHP-30; MPI; BDI; POMS; Adverse events
Xi [25]	2022	T: 80 C: 80	T: 31.78 ± 5.42 C: 32.36 ± 5.73	T: NWM + MNF C: NWM Length of intervention: 3 months Freq: 3 times a week Duration: 30 min	NWM: Guanyuan, Zigong, Qihai, Sanyinjiao Neiyi Zhitong Formula: Danshen, Ezhu, Xiangfu, etc	CER; Scores of TCM syndromes; VAS; Serum index
Dongfang [27]	2001	T: 37 C: 30 Normal: 8	T: – C: –	T: AN C: CT(traditional Chinese formula: Ezhu Danshen Decoction)/Blank Length of intervention: 3 menstrual cycles Freq: qod,4 times in total Duration: –	AN: Tingzhong, Pizhixia, Neifenmi, Jiaogan, Neishengzhiqi TCM: Danshen, Chishao, etc	CER; Scoring criteria for dysmenorrhea; Classification criteria of disease degree; Peripheral blood β-EP;
Huifang [30]	2018	T: 36 C: 36	T: 33.94 ± 5.98 C: 31.31 ± 5.86	T: Acupoint catgut C: Acupuncture Length of intervention:12 weeks Freq: T:1 time 2 weeks C: 3 times a week Duration: T: – C:30 min	Catgut: Sanyinjiao, Xuehai, Diji, Zigong, Guanyuan Acupuncture: Sanyinjiao, Xuehai, Diji, Zigong, Guanyuan	VAS; Dysmenorrhea; symptom score; Serum prostaglandin level
Tian [19]	2022	T: 38 C: 38	T: – C: –	T: Laser moxibustion C: CT (Sham laser moxibustion) Length of intervention: 4 weeks Freq: 3 times a week Duration: 30 min	Zigong	GBS scores; VAS
Jianfang [28]	2009	T: 42 C: 42	T: 34.93 ± 6.15 C: 33.83 ± 5.88	T: Ear electroacupuncture C: Body electroacupuncture Length of intervention: 3 treatments Freq: qod,10 times per treatment Duration: 30 min	Ear-EA: Zigong, Pizhixia Neifenmi, Jiaogan Body-EA: Qihai, Guanyuan, Sanyinjiao, Diji, Tianshu, Zigong	Degree of dysmenorrhea; Overall effectiveness; PGs
Liu [35]	2015	T: 43 C: 43	T: 33.7 ± 4.4 C: 32.5 ± 4.6	T: A-M C: CT (gestrinone) Length of intervention: T: 3menstrual cycles C: 3 months Freq: T: for a week C: once a day Duration: T: 30 min C: –	A-M: Guanyuan, Qihai, Taixi, Zigong, Dahe, Diji, Sanyinjiao + Guizhi Fuling Capsule gestrinone: 2.5 mg/day, 2 times a week	CER
Lu [29]	2014	T: 37 C: 32	T: – C: –	T: M C: CT (traditional Chinese formula: Ezhu Danshen Decoction) Length of intervention: T: 3 weeks C: 1 month Freq: T: 3 times a week C: – Duration: T: 30 min C: –	M: Zhongji, Guanyuan, Qihai CT: Ezhu, Danshen, Chishao, Xiangfu	CER
Jianfang [20]	2020	T: 35 C: 35	T: 32.29 ± 2.31 C: 31.82 ± 2.54	T: M C: CT (Dydrogesterone tablets) Length of intervention: 3 menstrual cycles Freq: T: 2 times per day C: 2–25d per menstrual cycle Duration: T: 20–30 min C: –	M: Guanyuan, Tianshu, Zhongji, Shenque Dydrogesterone Tablets: once a tablet, 2 times per day	VAS; CMSS; NMPP; DMPP; EHP-5
Qiuxia [36]	2018	T: 30 C: 30	T: 37.60 ± 4.64 C: 35.87 ± 4. 67	T: Body electroacupuncture C: Body electroacupuncture + sham acupoint Length of intervention: 3 menstrual cycles Freq: 1 time a day per menstrual cycle Duration: 20 min	Body-EA: Ciliao, Zhongji, Zusanli, Sanyinjiao, Diji Sham acupoint: sham	VAS; CER; SF-36 scores; EHP-5 scores; Safeties (routine blood test, urinalysis, hepatorenal function)
Lili [31]	2022	T: 38 C: 37	T: 35.85 ± 4.68 C: 35.26 ± 4.56	T: Acupoint catgut C: CT (GnRH-α) Length of intervention: 3 months Freq: T: - C: 4 pills, 3 times per day Duration: –	Catgut: Guanyuan, Tianshu, Zhongji, Zigong, Qihai, Sanyinjiao, Xuehai, Taichong, Zusanli GnRH-α: 150 μg, 1 time per day	CER; FSH; E2; LH; Hemorheology
Qiang [26]	2021	T: 36 C: 36	T: 37.23 ± 4.83 C: 36.89 ± 5.27	T: NWM C: CT (Chinese proprietary medicine: Guixiang Wenjing Zhitong Capsules) Length of intervention: 3 menstrual cycles Freq: T: once a week Duration: T: 20 min	Guiyang WenjingZhitong Capsules: 4 ~ 6 pills, 3 times per day NWM: Shangliao, Ciliao, Zhongliao, Xialiao	VAS; PGE2; TCM Symptom score; CER
Dongdong [37]	2020	T: 43 C: 43	T: 39.79 ± 4.55 C: 40.98 ± 5.00	T: Gestrinone + acupoint catgut C: CT(Gestrinone) Length of intervention: 3 months Freq: T: 1 time per month C: 2 times a week Duration: –	Catgut: Guanyuan, Tianshu, Zhongji, Zigong, Diji, Qihai, Ciliao, Taichong, Xuehai, Zusanli, Sanyinjiao	CER; VAS; Serum IL-1β; TNF -α
Meirong [38]	2021	T: 30 C: 30	T: 32.36 ± 4.65 C: 32.83 ± 5.03	T: Eelectroacupuncture + Tong Bining Granules C: CT (Sham acupuncture + Tong Bining Granules) Length of intervention: 3 months Freq: 3 times a week Duration: 20 min	EA: Qihai, Guanyuan, Taixi, Dahe, Zigong, Diji Sham acupuncture: 1cun lateral	CER; Dysmenorrhea symptom score; TCM symptom score; Peripheral blood CA125; IL-6; TNF-α; Adverse events
Qiaozhang [39]	2012	T: 32 C: 30	T: - C: -	T: Abdominal acupuncture + infrared C: Abdominal length of Intervention: 3 months Freq: 3 times or more per menstrual cycle Duration: 25 min	Abdominal acupuncture: Zhongwan, Xiawan, Qihai, Guanyuan, Zhongji Infrared: abdomen	CER; Dysmenorrhea symptom score; Serum CA125
Peng [40]	2017	T: 45 C: 45	T: 35.8 ± 7.8 C: 33.5 ± 6.8	T: Herbs-partition moxibustion C: CT (mifepristone tablets) Length of intervention: 3 months Freq: T: qod; C: once a day Duration: –	Herbs-partition moxibustion: Dazhui, Mingmen mifepristone tablets: 12.5 mg, once per day, for 3 months	CER; serum CA125; Blood-serum IL-8; Pelvic mass size
Guangxian [32]	2014	T: 28 C: 29	T: - C: -	T: Acupoint catgut C: NWM Length of intervention: 3 months Freq: 3 times or more per menstrual cycle Duration: 25 min	Acupoint catgut: Sanyinjiao, Shenshu, Ciliao, Xuehai, Zigong, Guanyuan NWM: Same	Dysmenorrhea Observation Scale; Serum CA125; SF-36
Qiaoyi [21]	2019	T: 35 C: 35	T: 35.13 ± 7.12 C: 34.28 ± 6.81	T: NWM C: CT (contraceptive pills) Length of intervention: T: 3 menstrual cycles C: 3 months Freq: T: 3 times per menstrual cycle C: 1 time per day Duration: T: 30 min C: –	NWM: Zusanli, Sanyinjiao, Tianshu, Qihai, Zhognwan, Siman, Shuidao Drospirenone:1 time a day, for 28 days	VAS; SF-36
Ning [41]	2022	T: 30 C: 31	T: 34.23 ± 5.95 C: 33.87 ± 6.75	T: A + Shaofu Zhuyu decoction C: Shaofu Zhuyu decoction Length of intervention: 3 menstrual cycles Freq: once per day Duration: 30 min	A: Shangliao, Ciliao, Zhongliao, Xialiao	CER; Serum CA125; VAS

T experimental group, C control group, CT conventional treatment (mainly pharmacotherapy and non-pharmacotherapy) included placebo (sham laser acupuncture, sham acupuncture, etc.), Chinese proprietary medicine (Gui Xiang Wenjing Zhitong capsules, Tong bining, etc.), and western medicine (dydrogesterone tablets, GnRH-α, gestrinone, mifepristone tablets, contraceptive pills, etc.), VAS Visual Analog Scale, CER clinical efficacy rate, A acupuncture, AN auricular needle-embedding, A-M acupuncture, and medicine, NWM needle warming moxibustion

Risk of bias

Thirteen [19–21, 23, 25, 28, 30, 31, 33–37] employed randomization methods such as random number tables, computer-generated stochastic methods, or random number cards, which were classified as low risk for selective bias. The remaining studies mentioned randomization but lacked specific details, leading to an uncertain risk classification. The blinding method in four studies [19, 28, 38, 39] was deemed low risk, with two employing double blinding and the other two using single blinding. Three studies [23, 39, 40] reported patient dropouts, which were classified as high risk for withdrawal bias. The remaining studies either did not report any dropouts or lacked sufficient information, resulting in a low or unclear risk classification. None of the studies published research protocols, leading to an unclear classification for reporting bias. Specific details are provided in Fig. 3.

Fig. 3.

Risk bias assessment table of the included studies. A Risk of bias assessment by individual trials; B overall risk of bias assessment using the Cochrane tool

Effect of interventions

Clinical efficacy evidence map: Ten studies reported on clinical efficacy rates, encompassing a total of eleven acupuncture types. The evidence network diagram is presented in Fig. S4. In the diagram, each of the 11 intervention nodes forms a closed loop. The size of the dots represents the sample size, while the thickness of the lines corresponds to the number of randomized controlled trials.

The CER for acupuncture treatments of EMs was higher compared to conventional treatments, and the difference was statistically significant (OR = 2.87, 95% CI (1.90, 4.82), P < 0.0001) (Fig. 4).

Fig. 4.

Probability diagram of CER. A Acupuncture, B body electroacupuncture, C ear electroacupuncture, D needle warming moxibustion, E needle warming moxibustion and Modified Neiyi Zhitong Formula, F Shu-Mu acupuncture, G acupoint catgut, H acupuncture and medicine, I auricular needle-embedding, J control group (conventional treatment), K moxa stick moxibustion

This forest plot shows the results of subgroup analysis for Chinese Patent Medicine and Traditional Chinese Medicine in the treatment of a particular condition. In general, neither the Chinese Patent Medicine nor the Traditional Chinese Medicine subgroup demonstrated significant treatment effects. The risk ratio for the Chinese Patent Medicine group was 1.06 (95% CI 0.93–1.21), while the risk ratio for the Traditional Chinese Medicine group was 1.07 (95% CI 0.86–1.33). Both groups showed risk ratios close to 1, and their 95% confidence intervals included 1, suggesting no statistically significant difference in treatment effects. In addition, heterogeneity within the Chinese Patent Medicine group and the Traditional Chinese Medicine group was I² = 0.0% (P = 0.959) and I² = 0.0% (P = 0.894), respectively, indicating consistency across studies. The overall heterogeneity between the groups was P = 0.931, suggesting no significant differences between the groups (Fig. S5).

We constructed separate funnel plots for the outcome indicator CER to examine possible publication bias. Visual examination of the funnel plot did not reveal any significant publication bias. The details are shown in Fig. S6.

The results from the network meta-analysis demonstrated that, compared to the control group (conventional treatment), ear electroacupuncture [MD = 9.43, 95% CI (1.35, 65.89)], needle warming moxibustion combined with Modified Neiyi Zhitong Formula [MD = 8.00, 95% CI (1.49, 42.76)], auricular needle-embedding [MD = 7.56, 95% CI = (1.89, 30.28)], body electroacupuncture [MD = 4.33, 95% CI (1.20, 15.60)], acupuncture and medicine [MD = 3.53, 95% CI (0.88, 14.09)], acupoint catgut [MD = 3.37, 95% CI (1.25, 9.10)], needle warming moxibustion [MD = 2.38, 95% CI (0.72, 7.86)], Shu-Mu acupuncture [MD = 1.38, 95% CI (0.28, 6.80)], and moxa stick moxibustion [MD = 1.00, 95% CI (0.23, 4.43)] all showed superior effectiveness in improving CER. The probability ranking of the various acupuncture and moxibustion interventions in terms of CER improvement was led by ear electroacupuncture, with a SUCRA of 83.0%, as shown in Fig. 5. A comparison of the different interventions is presented in Tables 3 and 4.

Fig. 5.

SUCRA plot for CER

Table 3.

League table on CER

Table 4.

League table on VAS

Ear-EA ear electroacupuncture, NWM + MNF needle warming moxibustion + Modified Neiyi Zhitong Formula, Body-EA body electroacupuncture, A-M acupuncture and medicine, Catgut acupoint catgut embedding, NWM needle warming moxibustion, Shu-Mu A Shu-Mu acupuncture, Moxa moxa stick moxibustion, AC acupuncture, CT conventional treatment

According to the SUCRA values (Fig. 5), C (ear electroacupuncture) (SUCRA = 83.0%) was the preferred option for improving CER, followed by E (needle warming moxibustion combined with Modified Neiyi Zhitong Formula) (SUCRA = 80.6%), I (auricular needle-embedding) (SUCRA = 79.6%), B (body electroacupuncture) (SUCRA = 63.4%), H (acupuncture and medicine) (SUCRA = 56.9%), G (acupoint catgut embedding) (SUCRA = 56.2%), D (needle warming moxibustion) (SUCRA = 43.3%), F (Shu-Mu acupuncture) (SUCRA = 28.4%), K (moxa stick moxibustion) (SUCRA = 25.2%), A (acupuncture) (SUCRA = 18.8%), and J (conventional treatment) (SUCRA = 14.6%).

A total of nine studies reported the VAS, involving nine acupuncture and moxibustion methods, and the evidence network diagram is shown in Fig. S7. The figure includes nine intervention nodes, which do not form any closed loop.

The VAS for acupuncture and moxibustion in the treatment of EMs was lower than that for conventional treatment, with a statistically significant difference (MD = −0.73, 95% CI (−0.93, −0.54), P < 0.0001) (Fig. S2). We created separate funnel plots for the outcome indicator CER to explore potential publication bias. The results indicated that the funnel plot was not perfectly symmetrical, suggesting possible publication bias or a small sample effect within the research network. The details are shown in Fig. S8.

According to the SUCRA values (Fig. 6), G (conventional treatment) (SUCRA = 99.4%) was the preferred option for statistically significant improvement in VAS, followed by D (needle warming moxibustion combined with Modified Neiyi Zhitong Formula) (SUCRA = 85.1%), C (needle warming moxibustion) (SUCRA = 66.3%), H (laser moxibustion) (SUCRA = 40%), E (acupoint catgut embedding) (SUCRA = 38.5%), B (body electroacupuncture) (SUCRA = 33%), A (acupuncture) (SUCRA = 32.9%), and F (auricular needle-embedding) (SUCRA = 4.9%).

Fig. 6.

SUCRA plot for VAS

Adverse events

Among the 23 studies included, 4 mentioned adverse reactions [22, 23, 32, 36]. Four studies reported “no adverse reactions” [32, 39–41], and 4 studies documented the symptoms and instances of adverse reactions in both the experimental and control groups [22, 23, 32, 36]. The most frequently mentioned issue was “dizziness,” which was typically addressed with symptomatic treatments. No specific treatment measures were provided. Details are shown in Table 5.

Table 5.

Adverse reactions or adverse events after different acupuncture and moxibustion methods

First author/year	Intervention	Control (cases)	Experiment	Treatment
Yuanzheng Sun [22], 2006	Acupuncture vs. Western medicine	Weight gain (14), acne (11), breast distending pain (5), nausea and vomiting (7), hot flush (10), vaginal bleeding (8), liver dysfunction (8)	Weight gain (3), acne (3), breast distending pain (1)	Unclear
Katharina Rubi-Klein [23], 2010	Acupuncture vs. non-specific acupuncture	Unmentioned	Dizzy (2)	Unclear
Dongdong Zuo [36], 2020	Acupoint catgut vs.. Western medicine	Dizzy (2), feeble (2), stomach upset (3), leg swelling (1)	Dizzy (1), stomach upset (1)	Unclear
Guangxian Chen [32], 2014	Acupoint catgut vs. needle warming moxibustion	Skin burn (3)	Unmentioned	Immediately local ice-packing and inunction of flower oils for trauma, then gauze covering, keep the wound clean

Discussion

This network meta-analysis examined the effects of acupuncture-related therapies on the management of EMs across 23 studies involving 1776 participants. All included studies reported randomization, but nine studies [22, 23, 27–29, 34, 38, 39, 41] did not specify the method of random sequence generation. Regarding allocation concealment, six studies [19, 21, 22, 24, 32, 33] provided details of the allocation concealment procedure. In terms of blinding, five studies [19, 24, 34, 38, 39] were deemed to have a low risk of bias. The acupuncture-related interventions assessed in the studies included acupuncture, ear electroacupuncture, auricular needle-embedding, body electroacupuncture, needle warming moxibustion, acupoint catgut embedding, Shu-Mu acupuncture, moxa stick moxibustion, and laser moxibustion.

The clinical effective rate is the proportion of patients who show improvement, excluding those for whom the treatment was ineffective. Ear electroacupuncture ranked first in terms of clinical efficacy rate.

The etiology of EMs is multifaceted. According to TCM, endometriosis is primarily caused by blood stasis in the uterus, with stagnation impacting the meridians and viscera. This condition is linked to psychological factors, cold coagulation with qi stagnation, and physical deficiency. Blood stasis leads to pain due to obstruction and eventually forms nodules and masses, making blood stasis a key pathological factor. Modern medicine, while recognizing the complexity of endometriosis, acknowledges that its exact etiology remains unclear. It is thought to involve multiple factors, including retrograde menstruation, coelomic metaplasia, genetic predisposition, immune dysfunction, inflammation, psychological factors, endocrine influences, and environmental exposures [42].

Acupuncture, a hallmark TCM therapy, is used to treat endometriosis through various techniques, including conventional acupuncture, specialized acupoint combinations, thunder-fire moxibustion, warm needle acupuncture, indirect moxibustion, abdominal acupuncture, auricular acupoint embedding, acupoint thread embedding, acupoint application, combined acupuncture–medication therapy, combined moxibustion–medication therapy, thread embedding with herbal medicine, heat-sensitive moxibustion, laser moxibustion, and acupoint application with moxibustion. Extensive prior research has explored the therapeutic mechanisms of acupuncture from both TCM and Western medical viewpoints [43]. From the TCM perspective, the core therapeutic principles involve regulating qi and blood circulation, promoting blood stasis removal, reducing inflammation, and alleviating pain through the stimulation of specific acupoints. This helps achieve meridian dredging, qi-blood harmonization, yin-yang balance, and the elimination of pathogenic factors. From the Western medical perspective, acupuncture is thought to reduce angiogenesis, modulate immune factors, regulate inflammatory mediators, lower estrogen levels, improve ovarian function, regulate hemorheology and hypercoagulable states, and influence the expression of apoptosis-related factors, ultimately improving the local microenvironment and inhibiting lesion progression [44].

In evaluating the efficacy and safety of acupuncture for EMs, variations in research findings and methodologies should be carefully considered. These differences may stem from patient population heterogeneity (age, disease duration, symptom severity), treatment protocol selection (acupoint choice, moxibustion type, treatment duration), and study design variations (sample size, randomization, blinding methodology). Thus, it is essential to take these factors into account when interpreting the effects of acupuncture interventions for EMs. Regarding efficacy, thunder-fire moxibustion and warm needle acupuncture may offer superior pain relief due to their enhanced thermal effects. Specialized acupoint combinations may provide more tailored treatment options for specific symptoms. Acupoint thread embedding and combined thread-herbal approaches could deliver more sustained therapeutic effects through prolonged acupoint stimulation and pharmacological action. In terms of safety, traditional interventions such as acupuncture and moxibustion are generally regarded as safe. Modern techniques, like laser moxibustion, may offer improved safety profiles by avoiding the thermal injury risks typically associated with traditional moxibustion. However, acupuncture point embedding may pose minor, reversible risks of foreign body reactions [45, 46].

This review also includes clinical studies of Chinese patent medicines and herbal formulations, such as the Modified Neiyi Zhitong Formula, Ezhu Danshen Decoction, Guixiang Wenjing Zhitong Capsules, Tong Bining Granule, and Shaofu Zhuyu Decoction. These formulations primarily feature herbs known for their blood-activating and stasis-removing properties, such as E Zhu, Dan Shen, Xiang Fu, Chi Shao, Dang Gui, and Yan Hu Suo. From a pharmacological perspective, Chinese herbal medicine demonstrates a multi-target, multi-pathway mode of action. Previous studies indicate that Chinese herbs may function through several mechanisms, including inhibiting inflammatory factors, enhancing phagocytosis, modulating cellular immunity, regulating metabolic homeostasis, modulating oxidative responses, exhibiting anti-tumor effects, and regulating signaling pathways. Therefore, oral Chinese herbal medicine may be considered for endometriosis patients who experience pain and menstrual irregularities [47–49]. The optimal treatment approach should consider multiple factors, including patient age, preferences, reproductive plans, pain intensity, and disease severity. This suggests that multi-pathway pharmaceutical interventions could have substantial implications for the development of precision medicine in managing EMs.

Strengths and limitations

The CER offers several advantages: (1) symptom relief can alleviate the pain and discomfort experienced by patients with endometriosis. (2) Symptom relief is also crucial for managing and controlling the disease. By addressing symptoms, the progression of the condition can be halted, thus reducing the risk of complications. (3) Symptom relief can improve patients’ overall life satisfaction and positively impact their mental and emotional well-being. However, due to the inherent variability in clinical practice, various factors must be considered, which presents challenges to the scientific rigor and reliability of research findings.

The VAS also has several benefits: (1) it is simple to apply, and patients can quickly understand and complete the scale. (2) The continuous nature of the scale allows for a more precise and sensitive measure of pain, making it suitable for patients of varying ages and cognitive abilities. However, aside from pain relief (VAS), other menstrual symptoms, such as pelvic girdle (PG) pain and serum CA125, have not been assessed in the 14 studies reviewed regarding the effects of acupuncture and moxibustion on women with endometriosis. Therefore, future RCTs should incorporate objective and quantitative assessments of other aspects of endometriosis to address the limitations of prior research. The details are shown in Figs. 4 and 7.

Fig. 7.

Probability diagram of VAS. A Acupuncture, B body electroacupuncture, C needle warming moxibustion, D needle warming moxibustion and Modified Neiyi Zhitong Formula, E acupoint catgut, F auricular needle-embedding, G control group (conventional treatment), H laser moxibustion, I moxa stick moxibustion

Several limitations should be noted: (1) the overall quality of the research included in the literature is relatively low, with considerable risk of bias; (2) only 23 studies were included, and just 9 studies mentioned the observation indicator—VAS. Therefore, the validity of the results remains uncertain; (3) although a network meta-analysis was conducted, a significant proportion of the studies had small sample sizes, which may reduce the statistical power of the findings; (4) discrepancies were observed in the rankings of the observation indicators. In addition, many studies did not report adverse events, limiting the ability to assess the safety of different acupuncture and moxibustion methods; (5) not all of the included studies provided details on syndrome differentiation and treatment. Moreover, there was variation in the selection of acupuncture points and treatment duration, which may have impacted the results. Consequently, the conclusions of this study should be interpreted with caution and considered carefully; (6) the lack of certain acupuncture and moxibustion therapies, due to the limited number of original studies, prevented a comprehensive comparison of the efficacy and safety of all available methods; (7) our search strategy only included studies published in Chinese and English, potentially introducing sampling bias. In addition, no multi-center studies involving women of different racial backgrounds were included, meaning the applicability of these therapies to non-Chinese populations remains uncertain.

Conclusion

For patients with endometriosis, acupuncture, ear electroacupuncture, body electroacupuncture, needle warming moxibustion, needle warming moxibustion combined with Neiyi Zhitong Formula, Shu-Mu acupuncture, acupoint catgut, acupuncture and medicine, auricular needle-embedding, conventional treatment, moxa stick moxibustion, and laser moxibustion can achieve significant therapeutic effects. The clinical effective rate and Visual Analog Scale (VAS) were ranked with a certain degree of probability. Addressing pain associated with endometriosis is of paramount importance for both clinical practice and future research. However, the reliability of these outcomes must be confirmed through multi-center, high-quality, large-sample randomized controlled trials that directly compare various Western medicines with proprietary Chinese medicines to improve the quality of evidence.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

YS and RJ wrote the manuscript and rectified the article (data analysis and examination); XZ and HZ collected data and extracted the data. Meanwhile, quality assessment was checked; YJ checked the article’s contents and various data by relevant software; CL, ZY, MZ, and SY made the data management; XT and JY checked the writing of the article and the authenticity of the content of the paper, funding acquisition, the reliability of the data, the credibility of the conclusion, whether it conformed to legal norms, academic norms, and moral norms.

Funding

This study was funded by the National Natural Science Foundation of China (No. 325019048 and No. 325021028).

Data availability

All data are available in the manuscript and they are shown in figures and tables.

Declarations

Conflict of interest

The authors report no conflicts of interest in this work.

Ethical approval

Not applicable.