Acupuncture monotherapy for endometriosis-related pain: A systematic review and meta-analysis

The results of this study suggest that acupuncture may effectively alleviate pain in patients with ERP. These findings support acupuncture’s potential role as a complementary therapy in managing ERP. However, these conclusions should be interpreted cautiously due to methodological limitations identified in the included studies. Future clinical trials should include more diverse populations and adopt standardized, high-quality protocols to improve the reliability and applicability of the results. Additionally, large-scale, rigorously designed RCTs are needed to confirm acupuncture’s efficacy and evaluate its long-term outcomes and safety.

Endometriosis is a common estrogen-dependent gynecological disorder characterized by the presence of endometrial-like tissue outside the uterine cavity. [ 1 – 3 ] It affects 10% to 15% of reproductive-aged women, with pelvic endometriosis being the most prevalent manifestation. [ 4 – 8 ] The condition is frequently associated with progressive symptoms, including secondary dysmenorrhea (71–87% of cases), chronic pelvic pain (up to 50% of patients), and infertility (30–50% of affected women). [ 4 – 8 ] Emerging evidence suggests these symptoms may arise from neuroplastic changes in both peripheral and central pain pathways, which could also contribute to chronic pain syndromes. [ 9 ] Given its substantial impact on quality of life, developing effective pain management strategies for endometriosis remains a critical research priority. Current therapeutic approaches include pharmacological agents, surgical excision, combined surgical-medical treatment, and assisted reproductive technologies. [ 10 – 12 ] However, these interventions often provide inadequate pain relief, with 20% to 40% of patients reporting persistent symptoms posttreatment. [ 10 – 12 ] Pharmacological options are particularly limited by adverse effects; for instance, gonadotropin-releasing hormone analogs induce hypoestrogenic states, leading to menopausal symptoms such as vasomotor disturbances and vaginal atrophy. [ 13 ] These limitations highlight the need for safer, more effective alternatives to manage endometriosis-related pain (ERP). Acupuncture is a safe and effective therapeutic modality for managing various disorders, including stroke, insomnia, and pain-related conditions. [ 14 – 17 ] It is reported to be emerged as a potential adjunctive therapy, with mechanistic studies suggesting it may modulate pain through neuroendocrine pathways (e.g., β-endorphin release) and anti-inflammatory effects. [ 18 – 21 ] Clinical trials report significant reductions in dysmenorrhea pain scores and improved quality of life metrics following acupuncture treatment. [ 18 , 20 , 21 ] Nevertheless, existing evidence remains constrained by methodological shortcomings, including small sample sizes and heterogeneous study designs. [ 22 – 24 ] While 2 prior meta-analyses demonstrated acupuncture’s efficacy for ERP, their conclusions were limited by the inclusion of non-randomized trials and combined acupuncture-moxibustion interventions. [ 22 – 24 ] To address these gaps, this systematic review and meta-analysis specifically evaluates the efficacy and safety of acupuncture monotherapy versus conventional treatments for ERP. Our study aims to provide higher-quality evidence to inform clinical decision-making for refractory ERP cases.

Conceptualization: Fan Yang, Long Wang, Ying-wei Wang, Liu-ci Chu. Data curation: Fan Yang, Long Wang, Liu-ci Chu. Investigation: Liu-ci Chu. Methodology: Fan Yang, Long Wang. Project administration: Liu-ci Chu. Resources: Fan Yang, Long Wang, Ying-wei Wang. Supervision: Liu-ci Chu. Validation: Fan Yang, Long Wang, Ying-wei Wang, Liu-ci Chu. Visualization: Fan Yang, Long Wang, Ying-wei Wang, Liu-ci Chu. Writing – original draft: Fan Yang, Long Wang, Ying-wei Wang, Liu-ci Chu. Writing – review & editing: Fan Yang, Long Wang, Ying-wei Wang, Liu-ci Chu.

This systematic review and meta-analysis was exempt from ethical approval as it utilized exclusively secondary data extracted from previously published studies, with no direct involvement of human participants or collection of identifiable personal data. We implemented a comprehensive search strategy across 6 major electronic databases (PUBMED, EMBASE, Cochrane Library, China National Knowledge Infrastructure, WANGFANG, and VIP Information Chinese Journal Service Platform) from their inception through March 1, 2025. The search was designed to identify all relevant randomized controlled trials evaluating the comparative efficacy of acupuncture versus non-acupuncture interventions for ERP. Our search methodology incorporated controlled vocabulary (MeSH terms) and free-text keywords including, but not limited to: “endometriosis,” “pelvic pain,” “acupuncture therapy,” and “randomized controlled trial (RCT).” The complete search strategy of all databases are presented in Table 1 , demonstrates the rigorous approach employed to ensure maximal retrieval of pertinent studies while maintaining methodological precision. Search strategies of included databases. Inclusion criteria were established as follows: participants with a confirmed diagnosis of ERP; study design limited to RCTs; intervention groups receiving acupuncture as a standalone therapy; and control groups receiving non-acupuncture treatments. No restrictions were applied regarding participants’ demographic characteristics (e.g., gender, ethnicity, or geographic location) to enhance the generalizability of findings. Exclusion criteria were implemented to ensure methodological rigor: studies with inadequate sample sizes (<10 participants per treatment arm); duplicate publications or redundant datasets; nonrelevant research designs (including animal studies, case reports, conference abstracts, review articles, noncontrolled trials, and studies combining multiple interventions); RCTs with incomplete outcome data or insufficient methodological details; and investigations involving concomitant therapies that could confound the assessment of acupuncture’s therapeutic effects. This selective approach was adopted to maintain analytical precision in evaluating acupuncture’s exclusive efficacy for ERP management. The primary outcome was pain intensity. It was measured by visual analogue scale (VAS; 10-mm scale from 0 [no pain] to 10 [worst pain]). [ 25 ] The secondary outcomes were serum CA125 level (U/mL), cure rate (CR, complete symptom resolution), and overall effective rate (OER; ≥50% symptom enhancement). CA125 was selected as a secondary outcome because of its established role in the diagnosis and monitoring of endometriosis. Its serum levels show a moderate correlation with disease severity, especially in pelvic endometriosis. [ 26 , 27 ] Elevated CA125 levels are attributed to secretion by ectopic endometrial tissue and activation of peritoneal mesothelial cells during inflammatory responses. [ 26 , 27 ] Serum CA125 also reflects peritoneal inflammation, as evidenced by its correlation with peritoneal fluid concentrations. [ 27 ] Clinically, it is used to evaluate treatment efficacy and predict the risk of pelvic adhesions, particularly when combined with inflammatory markers such as the platelet-to-lymphocyte ratio. [ 28 ] Study selection, data extraction, and quality assessment were conducted independently by 2 investigators to ensure methodological rigor. All identified studies underwent systematic screening using predefined eligibility criteria. A standardized extraction form captured: publication metadata (title, first author, publication year); participant demographics (age, sex distribution); methodological details (sample size, randomization technique, blinding protocol, allocation concealment); intervention characteristics (treatment modality, dosage regimen, session frequency, treatment duration); and primary and secondary outcomes. Discrepancies were resolved through consensus discussions with a third reviewer when required. Methodological quality was evaluated using the Cochrane Risk of Bias Tool 2.0 (RoB 2), which systematically assesses 7 critical domains: randomization process; allocation concealment; blinding of participants and personnel; blinding of outcome assessors; completeness of outcome data; selective reporting; and other potential biases. Each domain received categorical classification (low, high, or unclear risk of bias) to facilitate comprehensive study quality evaluation. All statistical analyses were performed using RevMan software (version 5.4; Cochrane Collaboration). Continuous outcome measures (e.g., pain scores, biomarker levels) were analyzed using mean differences (MD) with 95% confidence intervals (CIs), while dichotomous outcomes (e.g., treatment response rates) were expressed as odds ratios (OR) with 95% CI. Heterogeneity across studies was quantified using the I ² statistic; a value ≥50% indicating large heterogeneity. [ 29 ] If large heterogeneity was observed, a random-effects model was applied to incorporate between-study differences. Otherwise, a fixed-effects model was utilized. [ 29 ] All analyses were conducted according to intention-to-treat principles where possible.

Our systematic search initially identified 286 potential records (Fig. 1 ). Following duplicate removal and eligibility screening, we excluded 277 records, retaining 9 randomized controlled trials for final analysis. These trials collectively enrolled 535 participants meeting our inclusion criteria. The study populations exhibited an age range of 17.0 ± 2.1 to 36.24 ± 6.78 years, representing a clinically relevant spectrum of reproductive-aged women. Table 2 summarizes the key characteristics of included studies, detailing: participant demographics; intervention protocols (acupuncture techniques, treatment duration/frequency); comparator treatments; and primary outcome measures. This structured presentation enables direct comparison of methodological approaches across studies while highlighting potential sources of clinical heterogeneity. General characteristics of included studies. C = control group, CA125 = carbohydrate antigen 125, CR = cure rate, EA = electroacupuncture, OER = overall effective rate, T = treatment group, VAS = visual analog scale. Flow diagram of study selection. The methodological quality of all included trials was systematically evaluated using the Cochrane Risk of Bias tool (RoB 2.0), with results presented in Figure 2 . Among the 9 eligible studies, [ 30 – 38 ] all adequately reported on random sequence generation and addressed incomplete outcome data. Three trials [ 32 , 33 , 38 ] demonstrated proper allocation concealment procedures, while an equivalent number [ 32 , 33 , 37 ] implemented effective blinding of participants and study personnel. Two studies [ 32 , 33 ] specifically reported blinding of outcome assessors. The remaining bias domains exhibited variable reporting quality across the included trials, as visually summarized in Figure 2 . Risk of bias summary. Five RCTs involving 300 participants demonstrated a statistically significant reduction in VAS scores in the acupuncture group compared to controls (MD = −1.67, 95% CI [−2.85, −0.49]; I ²=88%, P  = .006; Fig. 3 ). [ 30 , 32 , 33 , 37 , 38 ] Subgroup analysis yielded a more consistent effect (MD = −1.56, 95% CI [−2.07, −1.05]; I ²=0%, P  < .01), indicating that differences in population characteristics may account for the variability in the main analysis. Funnel plot showed approximate symmetry, suggesting minimal publication bias (Fig. 4 ). Meta-analysis of VAS. VAS = visual analogue scale. Funnel plot of VAS. VAS = visual analogue scale. Three studies (n = 224) reported no significant differences in serum CA125 levels (MD = −1.46, 95% CI [−20.69, 17.76]; I ²=79%, P  = .88) between the intervention and control groups (Fig. 5 ). [ 33 , 36 , 38 ] However, subgroup analysis revealed a significant reduction in CA125 levels in the intervention group compared to the control group (MD = −12.09, 95% CI [−23.71, −0.46]; I ²=0%, P  = .04), suggesting a more homogeneous treatment effect in specific populations. Meta-analysis of CA125. Four trials (n = 235) reported significantly higher CR with acupuncture (OR = 2.61, 95% CI [1.38, 4.95]; I ²=0%; P  = .003; Fig. 6 ). [ 31 , 34 – 36 ] Meta-analysis of CR. CR = cure rate. The pooled analysis of 4 studies (n = 235) showed a marginally significant benefit for acupuncture (OR = 2.18, 95% CI [0.99, 4.80]; I ²=0%; P  = .05; Fig. 7 ). [ 31 , 34 – 36 ] Meta-analysis of OER. OER = overall effective rate.

In recent years, the incidence of endometriosis has significantly increased, rendering it one of the most common gynecological disorders encountered in clinical settings. Secondary dysmenorrhea is a typical symptom of endometriosis, and ERP is the most prominent clinical manifestation in a majority of affected patients. Oral administration of danazol capsules remains a conventional therapeutic approach; the drug functions by suppressing ovarian activity via hormonal modulation, inducing a pseudo-pregnant or pseudo-menopausal state. This hormonal environment inhibits the proliferation of ectopic endometrial tissue, leading to a reduction in lesion size and alleviation of dysmenorrhea symptoms. Nevertheless, the clinical use of danazol is often limited by adverse effects such as irregular vaginal bleeding, headaches, weight gain, hepatic dysfunction, and symptoms of estrogen deficiency. Moreover, the high rate of symptom recurrence upon drug discontinuation further limits its long-term utility. In contrast, acupuncture has been increasingly explored as a complementary or alternative treatment for ERP. Its therapeutic mechanism is believed to involve the suppression of prostaglandin secretion, enhancement of the pain threshold, and rapid alleviation of severe dysmenorrhea. These effects are achieved through improved pelvic blood flow, which promotes the resolution of adhesions and fibrotic changes associated with ectopic endometrial tissue. Acupuncture also appears to inhibit the pathological proliferation, secretion, and hemorrhage of ectopic endometrial cells, thereby alleviating clinical symptoms and improving patient outcomes. The association between CA125 levels and ERP likely reflects a combination of direct inflammatory mechanisms and underlying pathological processes. Elevated CA125 levels are associated with a proinflammatory environment marked by elevated cytokines, including interleukin-1β, interleukin-6, and tumor necrosis factor-α. [ 39 ] These cytokines can activate peripheral nociceptors and promote central sensitization by influencing neuronal activity in the dorsal root ganglia, thereby intensifying pain perception. [ 39 ] CA125 levels have also been correlated with lesion infiltration depth, suggesting a potential association with disease severity. [ 40 ] Higher CA125 levels have also been linked to the extent of pelvic adhesions, which may aggravate pain through mechanical distortion and reduced organ mobility. [ 40 , 41 ] However, our meta-analysis showed no significant change in serum CA125 levels after acupuncture (MD = −1.46; 95% CI: −20.69 to 17.76; P  = .88), suggesting that this biomarker may be insensitive to the therapeutic effects of acupuncture on ERP. Several previous studies have investigated the role of acupuncture in the management of ERP and endometriosis. [ 16 , 19 , 42 ] One study [ 16 ] examined the therapeutic effects of acupuncture or electroacupuncture alone and in combination with moxibustion, catgut implantation, herbal medicine, and acupoint therapy. The control group in this study received sham acupuncture, herbal remedies, or pharmaceutical agents. Findings suggested that acupuncture significantly reduced both pain scores and serum CA125 levels. Another study [ 19 ] focused on acupuncture and moxibustion, used either alone or alongside interventions such as acupoint catgut embedding, bloodletting, and acupoint application. Compared to herbal medicine or drug treatment alone, combination therapies that included acupuncture demonstrated superior outcomes. A third study [ 42 ] compared manual acupuncture and moxibustion with conventional therapies, sham acupuncture, and no treatment. Results showed that manual acupuncture was effective in improving pain and quality of life; however, the authors emphasized that current evidence remains insufficient to fully validate the efficacy of acupuncture and moxibustion for endometriosis. In the present meta-analysis, we sought to directly compare the effectiveness of acupuncture with herbal or pharmaceutical treatments for ERP. A total of 9 studies involving 535 patients were included. Our findings, which partially align with previous reports, [ 16 , 19 , 42 ] demonstrated statistically significant differences in treatment outcomes. Specifically, in terms of pain relief measured by the VAS, the calculated MD was −1.67, with a 95% CI of −2.85 to −0.49, and a P -value of .006, indicating a significant advantage for acupuncture. In addition, CR rates were also significantly higher in the acupuncture group, with an OR of 2.61 (95% CI: 1.38–4.95; P  = .003). While the OER showed a potential trend favoring acupuncture (OR = 2.18; 95% CI: 0.99–4.80), the P -value of .05 indicates only borderline statistical significance. Collectively, the current findings suggest that acupuncture may offer superior efficacy in alleviating ERP compared to herbal medicine or conventional drugs. Nonetheless, due to limited evidence regarding long-term outcomes and potential adverse events, further high-quality, large-scale randomized controlled trials are necessary to substantiate these preliminary findings and comprehensively evaluate the clinical value of acupuncture in the management of endometriosis. This study has several important limitations that should be considered. First, the risk-of-bias assessment revealed methodological weaknesses in a considerable number of the included RCTs. These involved performance bias (potential unintentional influence on outcomes), detection bias (variation in assessment methods), and attrition bias (dropouts affecting data completeness). Such biases may compromise the internal validity of individual trials and reduce the overall reliability of the meta-analytic findings. Second, the small sample sizes of the included trials (ranging from 14 to 106 participants) may have limited the statistical power to detect treatment effects and reduced the generalizability of the results. Third, considerable heterogeneity was observed in key outcomes such as VAS scores and serum CA125 levels. This heterogeneity likely reflects methodological inconsistencies, including variations in study design, acupuncture protocols, and participant characteristics. These inconsistencies complicate the interpretation of pooled outcomes and raise concerns about the consistency of treatment effects. Finally, although the meta-analysis focused specifically on acupuncture for endometriosis, considerable variation in treatment parameters was evident across studies. Differences in acupoint selection, needle insertion techniques, stimulation methods, treatment frequency, and session duration may have contributed to the observed heterogeneity and influenced clinical outcomes. Therefore, caution is warranted when extrapolating these findings to other acupuncture protocols or broader patient populations. Future research should prioritize large-scale, methodologically rigorous RCTs using standardized intervention protocols to confirm these findings and evaluate the long-term efficacy and safety of acupuncture in this context.