Clinical Effect and Contributing Factors of Acupuncture for Insulin Resistance in Diabetes Mellitus: A Systematic Review and Pairwise and Exploratory Network Meta-Analysis

Abstract

Objects:

The purpose of this study is to investigate whether (1) acupuncture is effective in improving insulin resistance (IR) in patients with diabetes mellitus (DM) and (2) the effect of acupuncture varies depending on the type and dosage of acupuncture.

Methods:

PubMed, Embase, Web of Science, Cochrane Library, OpenGrey, ClinicalTrials.gov, and the World Health Organization International Clinical Trials Registry were searched from their inception up to April 26, 2025. The risk of bias was assessed using the Revised Cochrane risk-of-bias tool for randomized trials. The Grading of Recommendations Assessment, Development and Evaluation approach was used to evaluate the credibility of findings from each outcome. Systematic review and pairwise and exploratory network meta-analysis (NMA) of randomized controlled trials (RCTs) were conducted to investigate the effectiveness and safety of acupuncture on IR in patients with DM.

Results:

We included 16 RCTs from 2328 citations with 1087 participants. When compared with usual care, acupuncture had a significant effect on the homeostatic model assessment of IR (HOMA-IR; standardized mean difference [SMD] = −1.13, 95% confidence interval [CI]: −1.61 to −0.64), fasting blood sugar (FBS; SMD = −0.90, 95% CI: −1.45 to −0.35), and glycated hemoglobin (HbA1c; SMD = −0.66, 95% CI: −1.11 to −0.20) but not on 2-h blood glucose (2hBG; SMD = −1.32, 95% CI −2.83 to 0.19). When compared with sham acupuncture, acupuncture had a significant effect on FBS (SMD = −0.71, 95% CI: −1.18 to −0.25) but not on HbA1c (SMD = −0.14, 95% CI: −0.48 to 0.19). Subgroup analysis revealed that high-dose acupuncture had a more beneficial effect on HOMA-IR and 2hBG. According to the NMA, electroacupuncture (EA) might be the most promising acupuncture type for improving IR. However, we failed to analyze safety outcomes due to the inadequate data across the included studies.

Conclusion:

The findings suggested that acupuncture could be an effective therapy to improve IR in patients with diabetes. EA and high-dose acupuncture are two potential contributing factors.

Keywords

diabetes mellitus insulin resistance acupuncture acupuncture type acupuncture dose

Introduction

Diabetes mellitus (DM) is emerging as a common metabolic disease, leading to a growing number of life-years lost.¹ DM is associated with a variety of serious complications including cardiovascular disease, retinopathy, etc., and increases the risk of cancer.^2–4 World Health Organization (WHO) reported that the prevalence of DM rose from 200 million in 1990 to 830 million in 2022. In 2021, DM was the direct cause of 1.6 million deaths.⁵ Insulin resistance (IR) is a major cause of DM and plays an important pathophysiological role in DM.⁶ It is commonly associated with visceral adiposity, glucose intolerance, hypertension, dyslipidemia, endothelial dysfunction, and elevated levels of inflammatory markers.⁷

Nonpharmacological approach is an important part of DM therapy.⁸ Pharmacological approaches may be included when lifestyle modification alone is not sufficient to achieve desired clinical outcomes.⁹ Recent research supports a holistic, integrative approach to manage DM, combining nonpharmacological and pharmacological treatments.¹⁰

Acupuncture has long been recognized as a potentially effective nonpharmacological therapy for the treatment of DM.^11–13 Several randomized controlled trials (RCTs) suggested that acupuncture showed benefits for IR, but there are also published trials reporting conflicting evidence regarding the effectiveness.^11–15 Moreover, current guidelines differ substantially in whether acupuncture should be recommended for patients with DM, posing a challenge to clinical decision-making.^16,17 Although some systematic reviews were conducted to evaluate the effect and safety of acupuncture to treat DM, few focused on IR.^18,19 These systematic reviews did not include several latest published trials or considered only one type of acupuncture or comparators.^18–20 In addition, one previous RCT suggested that electroacupuncture (EA) had a greater effect on IR than manual acupuncture (MA).²¹ Notably, the effect of acupuncture may be dose independent commonly based on four parameters: number of acupoints, De Qi response, frequency of treatment per week, and duration of treatment. The higher the scores for these four parameters, the higher the dose of acupuncture. Growing evidence showed that there was a positive correlation between higher-dose acupuncture and better treatment outcomes.^22,23 However, there is still a lack of comprehensive studies exploring whether these factors contribute to the therapeutic effects of acupuncture. It is important to summarize present research outcomes and provide convincing evidence for clinicians and researchers. Therefore, we conducted a systematic review and pairwise and exploratory network meta-analysis (NMA) to determine whether (1) acupuncture is effective to improve IR in patients with DM and (2) the effect of acupuncture differs according to acupuncture type and acupuncture dose.

Method

Design

We conducted a systematic review and pairwise and exploratory NMA. The report of this study was guided by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement.²⁴ We registered a protocol for this study in the PROSPERO (CRD420251046279).

Search methods

We searched PubMed, Embase, Web of Science, and Cochrane Library databases from the inception up to April 26, 2025. Moreover, OpenGrey, ClinicalTrials.gov, and the WHO International Clinical Trials Registry were searched for unpublished studies. The search strategy is described in Supplementary Table S1.

Study selection

Inclusion criteria were based on the PICOS (participants, interventions, comparators, outcomes, and study design) approach. To be eligible for inclusion, studies had to meet the following criteria:(1) population: adults diagnosed with DM; (2) intervention: MA or EA regardless of number of the sessions, needling manipulation, types of needles, stimulation methods, or stimulation areas; (3) comparison: usual care (UC) or sham acupuncture (SA); (4) outcome: homeostatic model assessment of IR (HOMA-IR), fasting blood sugar (FBS), 2-h blood glucose (2hBG), and glycated hemoglobin (HbA1c); and (5) study design: all studies that are described as RCTs, blinded, or open.

The exclusion criteria were (1) RCTs that reported only improvement rates; (2) vague description of interventions; and (3) data that were used in other included studies.

Pairs of independent investigators (C.M. and D.H.) searched databases for relevant articles and removed duplicate records, with discrepancies adjudicated by a third investigator (T.L.). After removing duplicate records, pairs of independent investigators (C.M. and D.H.) screened references and extracted study-level data, with discrepancies adjudicated by a third investigator (T.L.).

Risk of bias and quality assessment

Pairs of independent investigators (D.L. and D.H.) assessed the risk of bias of the included RCTs by the Revised Cochrane risk-of-bias tool for randomized trials (RoB2), classifying risk of bias as high, moderate, or low.²⁵ We incorporated results into the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach to assess the credibility of findings from each outcome. GRADE graded confidence in results of each treatment comparison as high, moderate, low, or very low.²⁶

Data extraction

Pairs of independent investigators (D.L. and N.L.) extracted data from the included studies, with discrepancies adjudicated by a third investigator (T.L.).

We extracted outcome data (expressed as mean and standard deviation) for changes from the initiation to the end of treatment. The primary outcome was HOMA-IR, calculated by fasting glucose (mmol/L) × fasting insulin (mIU/L)/22.5.²⁷ The secondary outcomes were FBS, 2hBG, and glycated HbA1c. The safety outcome was adverse events (AEs; participants experiencing any AE or serious AEs).

We also extracted study characteristics (name of the first author, country of origin, year of publication, and sample size), participant characteristics (mean age and gender trends), acupuncture characteristics (type, duration, and frequency), and acupoints.

Acupuncture dose

We assessed the acupuncture dose by using a semiquantitative scoring instrument published previously.²⁸ The scoring instrument includes the following four parameters: (1) number of acupoints; (2) de qi response; (3) frequency of treatment per week, and (4) duration of treatment. De qi refers to a series of sensations (soreness, numbness, fullness, heaviness, etc.) experienced by patients when acupuncture is performed at acupoints.²⁹ De qi response can be evaluated by de qi scale/questionnaire.³⁰ Based on the sum of scores, we defined three levels of doses in acupuncture treatment: high, moderate, and low.

Data synthesis and analysis

Due to the inherent differences in the effect size observed across studies, the true effect estimate was computed using a random-effects model. Because the units were different across studies, the standardized mean difference (SMD) was used as the summary statistic. Forest plots (with 95% confidence interval [CI]) were constructed for overall analysis. Funnel plots were visually assessed to determine the risk of publication bias for outcomes that were evaluated in more than five studies.^31–33 In addition, we compared the funnel plot and Egger’s test to assess the risk of publication bias under specific circumstances, with Egger’s test indicating publication bias when p < 0.05.

We assessed heterogeneity using I² and Q statistic, as described in section 9.5.2 of the Cochrane Handbook for Systematic Reviews of Interventions. Using this section, we used the following interpretation of the I² statistic: 0%–40%: might not be important; 30%–60%: may represent moderate heterogeneity; 50%–90%: may represent substantial heterogeneity; 75%–100%: considerable heterogeneity; as for Q statistic, the level of significance for heterogeneity was set at p < 0.05.³⁴

Subgroup analyses were performed to explain the heterogeneity. Moderator variables for subgroup analysis were related to acupuncture type and acupuncture dose. We assessed the sensitivity of our findings by one-study-removed analysis. To perform this type of sensitivity analysis, the overall results are assessed multiple times, each time excluding one of the included studies.³⁵ Moreover, all analysis above was performed with StataSE 16 and R version 3.2.2.

Exploratory NMA

We performed a NMA to compare the efficacy of EA and MA on HOMA-IR. Network graph was drawn, and the random-effects frequentist NMA was fitted to assume a common random effect in the network. Meanwhile, the league tables were created to display the relative degree of HOMA-IR among EA and MA. p-Scores were applied to rank gait training on the basis of balance outcomes. p-Scores ranged from 0 to 1, with higher p-score indicating greater effect. In addition, we used the comparison-adjusted funnel plot and Egger’s test to assess the risk of publication bias under specific circumstances, with Egger’s test suggesting publication bias when p < 0.05. We used global and local methods to test the inconsistency of the research results. For global inconsistency, we evaluated inconsistency statistically using the design-by-treatment test. We assessed local inconsistency by splitting network estimates into the contribution of direct and indirect evidence (node-splitting test). All analysis above was performed by StataSE 16 and R version 3.2.2.

Result

Literature search and characteristics of the included studies

In total, 2430 studies were identified by searching the databases in accordance with protocol. A total of 966 studies were excluded due to duplication. Following the screening of titles and abstracts, 1117 studies were excluded. According to 6 studies not retrieved, 311 studies were assessed for eligibility. Following the review of complete texts, 295 studies were excluded for the reasons listed in the flow chart. Finally, 16 studies were included in this study (Fig. 1), containing 1087 participants, 642 (59.1%) of whom received acupuncture and 445 (40.9%) control intervention. The mean age of participants ranged from 24.5 to 61 years, with approximately 35.51% being male. Supplementary Table S2 provides a summary of the baseline characteristics of RCTs included in this study.

FIG. 1.

Study flow diagram. WHO, World Health Organization.

Risk of bias

As evaluated by the RoB2, 1 (6.25%) study was assessed as low risk of bias, 12 (75%) as some concerns, and 3 (18.75%) as high risk of bias (Fig. 2 and Supplementary Table S3). For randomization process, 2 (12.5%) studies were assessed as some concerns and 14 (87.5%) as low risk of bias; for deviations from intended interventions, 13 (81.25%) studies were assessed as some concerns and 3 (18.75%) as low risk of bias; for missing outcome data, 16 (100%) studies were assessed as low risk of bias; for measurement of the outcome, 7 (43.75%) studies were assessed as low risk of bias, 6 (37.5%) as some concerns, and 3 (18.75%) as high risk of bias; for selection of the reported result, 7 (43.75%) studies were assessed as some concerns and 9 (56.25%) as low risk of bias.

FIG. 2.

Risk of bias graph for each included study.

Acupuncture dose

Acupuncture regimens and parameters differed in acupoints, de qi response, frequency, and duration of treatment. The number of acupoints varied from 2 to 32; 7 trials (43.75%) had 20–32 acupoints, whereas 7 trials (43.75%) had 2–6 acupoints. Eight trials clearly required de qi response, and eight trials failed to report relevant details. Frequency ranged from once weekly to twice per day, whereas most trials reported that frequencies were one, two, or three times weekly (12 trials, 75%). The duration ranged from 1 week to 12 weeks; however, most trials had durations of 3, 4, or 12 weeks (11 trials, 68.75%). Ultimately, the acupuncture total doses of included trials were classified as high dose (8 trials, 50%), moderate dose (7 trials, 43.75%), and low dose (1 trial, 6.25%).

Homeostatic model assessment of insulin resistance

HOMA-IR was available in 8 trials^{12,13,15,21,36,38,39,51} with 318 participants in the intervention groups and 259 in the control groups. As shown in Figure 3A, acupuncture showed significant effect on HOMA-IR when compared with UC (SMD = −1.13, 95% CI: −1.61 to −0.64); I² was 79.6% and p < 0.001. Subgroup analysis (shown in Table 1) revealed high-dose acupuncture showed significant effect on HOMA-IR, while medium-dose acupuncture showed no significant effect on HOA-IR. Subgroup analysis (shown in Table 1) also revealed both MA and EA showed significant effect on HOMA-IR. I² and p-value obviously decreased in the MA subgroup. The certainty of the evidence was moderate (shown in Supplementary Table S4). Finally, no obvious publication bias was found (shown in Supplementary Fig. S1A).

FIG. 3.

Forest plot at the end of treatment. 2hBG, 2-h blood glucose; FBS, fasting blood sugar; HbA1c, glycated hemoglobin; HOMA-IR, homeostasis model assessment of insulin resistance.

Table 1.

Subgroup Analysis

Outcome	Type of subgroup		Number of participants	Standardized mean difference (95% CI)	I² ( p-value)
Acupuncture vs. usual care
HOMA-IR	Acupuncture dose	High dose	411	−1.23 (−1.74 to −0.71)	80% ( p < 0.001)
	Acupuncture dose	Medium dose	30	−0.20 (−0.54 to 0.14)	Not applicable
	Acupuncture type	Manual acupuncture	402	−0.90 (−1.23 to −0.58)	54% ( p = 0.07)
	Acupuncture type	Electroacupuncture	40	−2.84 (−3.74 to −1.94)	Not applicable
FBS	Acupuncture dose	High dose	411	−0.73 (−1.44 to −0.02)	91% ( p < 0.001)
	Acupuncture dose	Medium dose	247	−1.13 (−2.14 to −0.12)	92% ( p < 0.001)
	Acupuncture type	Manual acupuncture	565	−0.78 (−1.38 to −0.18)	91% ( p < 0.001)
	Acupuncture type	Electroacupuncture	93	−1.25 (−2.24 to −0.26)	79% ( p = 0.03)
HbA1c	Acupuncture dose	High dose	68	−1.14 (−1.65 to −0.62)	Not applicable
HbA1c	Acupuncture dose	Medium dose	163	−0.42 (−0.73 to −0.11)	0% ( p = 0.83)
2hBG	Acupuncture dose	High dose	128	−0.50 (−0.86 to −0.15)	0% ( p = 0.87)
2hBG	Acupuncture dose	Medium dose	149	−1.35 (−3.92 to 1.23)	97% ( p < 0.001)
Acupuncture vs. sham acupuncture
FBS	Acupuncture dose	High dose	140	−0.80 (−1.56 to −0.04)	78% ( p = 0.03)
		Medium dose	103	−0.34 (−0.90 to 0.23)	41% ( p = 0.18)
		Low dose	50	−1.36 (−1.98 to −0.74)	Not applicable
	Acupuncture type	Manual acupuncture	224	−0.67 (−1.16 to −0.17)	67% ( p = 0.03)
	Acupuncture type	Electroacupuncture	69	−0.69 (−2.08 to 0.71)	85% ( p = 0.01)
HbA1c	Acupuncture dose	High dose	80	−0.45 (−0.89 to 0.00)	Not applicable
HbA1c	Acupuncture dose	Medium dose	144	−0.02 (−0.36 to 0.31)	4% ( p = 0.35)

2hBG, 2-h blood glucose; CI, confidence interval; FBS, fasting blood sugar; HbA1c, hemoglobin; HOMA-IR, homeostasis model assessment of insulin resistance.

Fasting blood sugar

FBS was available in 16 trials^{11–15,21,36–44,51} with 506 participants in the intervention groups and 445 in the control groups. As shown in Figure 3B and Figure 3C acupuncture showed significant effect on FBS when compared with UC (SMD = −0.90, 95% CI: −1.45 to −0.35); I² was 90.4% and p < 0.001. Meanwhile, acupuncture also showed significant effect on FBS when compared with SA (SMD = −0.71, 95% CI: −1.18 to −0.25); I² was 71.0% and p = 0.004. Significant effects were observed in most of the subgroups, whereas no significant effects were observed in medium-dose acupuncture and EA subgroups. The certainty of the evidence was low (shown in Table 1). I² and p-value obviously decreased in the MA and medium-dose acupuncture subgroups. The certainty of the evidence was low (shown in Supplementary Table S4). Finally, obvious publication bias was found in the acupuncture versus UC analysis, whereas no obvious publication bias was found in the acupuncture versus SA analysis (shown in Supplementary Fig. S1B and C).

2-h blood glucose

2hBG was available in five trials^{11,14,15,38,51} with 138 participants in the intervention groups and 139 in the control groups. As shown in Figure 3C, acupuncture did not show significant effect on 2hBG when compared with UC (SMD = −1.32, 95% CI: −2.83 to 0.19); I² was 96.3% and p < 0.001. Subgroup analysis (shown in Table 1) revealed that high-dose acupuncture showed significant effect on 2hBG, whereas medium-dose acupuncture showed no significant effect. I² and p-value obviously decreased in the high-dose acupuncture subgroup. The certainty of the evidence was low (shown in Supplementary Table S4). Moreover, we could not assess publication bias due to the scarcity of studies.

Glycated HbA1c

HbA1c was available in six trials^{11,37,38,41,43,44} with 229 participants in the intervention groups and 226 in the control groups. As shown in Figure 3D, acupuncture showed significant effect on HbA1c when compared with UC (SMD = −0.66, 95% CI: −1.11 to −0.20); I² was 64.5% and p = 0.60. As shown in Figure 3E and Figure 3F acupuncture did not show any significant effect on HbA1c when compared with SA (SMD = −0.14, 95% CI: −0.48 to 0.19); I² was 33.7% and p = 0.210. Subgroup analysis (shown in Table 1) revealed that both high-dose and medium-dose acupuncture showed significant effect on HbA1c when compared with UC, whereas both high-dose and medium-dose acupuncture showed no significant effect when compared with SA. The certainty of the evidence was low (shown in Supplementary Table S4). Moreover, we could not assess publication bias due to the scarcity of studies.

Sensitivity analysis

The sensitivity of our findings for all four meta-analysis outcomes was assessed by one-study-removed analysis. With regard to 2hBG (shown in Supplementary Fig. S2C), the effect of acupuncture changed significantly after removing Shen et al.¹¹ Otherwise, findings essentially remained the same in sensitivity analyses of HOMA-IR, FBS, and HbA1c (shown in Supplementary Fig. S2), indicating that the inclusion of these studies did not have any major influence on these results.

Exploratory NMA

As shown in Figure 4, the HOMA-IR was available in eight trials, comparing MA and EA (53 patients) with the UC (52 patients). Compared with the UC, both MA and EA demonstrated significant effect. Ranking on the basis of p-score identified EA as the best, MA as the second, and UC as the worst (shown in Fig. 5). However, we did not find any significant difference in HOMA-IR between MA and EA according to the league table (Table 2). Designby-treatment test (shown Supplementary Table S5) and node-splitting test (shown in Supplementary Table S6) showed significant global and local inconsistency. Moreover, we did not find publication bias in HOMA-IR NMA (shown in Supplementary Fig. S3).

FIG. 4.

Network plot for HOMA-IR.

FIG. 5.

Forest plot for HOMA-IR in network meta-analysis. CI, confidence interval; SMD, standardized mean difference.

Table 2.

League Table for Homeostasis Model Assessment of Insulin Resistance According to Their Relative Effects and 95% Credibility Intervals

Acupuncture	−0.23 (−0.92 to 0.46)	0.88 (0.44 to 1.32)
−0.59 (−1.20 to 0.02)	Electroacupuncture	2.80 (1.55 to 4.04)
1.03 (0.60 to 1.45)	1.62 (0.94 to 2.31)	Usual care

Dark red: intervention name, Light red: row vs column comparison, Blue: column vs row comparison.

Safety

Eleven trials did not assess safety outcomes, and the remaining five^{14,37,40–42}did not assess these adequately. Therefore, none of the included studies provided a detailed or systematic assessment of safety outcomes, and meta-analysis of AEs could not be conducted due to inadequate data. However, while no AEs were reported, the absence of reported AEs does not imply the absence of actual events. The failure to detect AEs likely reflects insufficient safety reporting rather than true safety.

Discussion

Principal findings

To our knowledge, this is the first systematic review and pairwise and exploratory NMA to comprehensively explore the effect of acupuncture on IR in patients with DM considering the type of comparisons, type of acupuncture, and acupuncture dose. Our findings showed that acupuncture had significant effect on HOMA-IR, FBS, and HbA1c but had no significant effect on 2hBG, when compared with UC. We also find that acupuncture significantly decreased FBS but not HbA1c, when compared with SA.

According to the subgroup analysis, we found both MA and EA had significant effect on HOMA-IR and FBS. Furthermore, an exploratory NMA was conducted to compare MA and EA. We found that EA was the most promising to improve IR, although there was no significant difference between MA and EA. For acupuncture dose, we found high-dose acupuncture had a significant and better effect on HOMA-IR and 2hBG when compared with UC, as well as a significant and better effect on HbA1c when compared with SA.

Strengths and limitations

This review has several key strengths. First, we assessed the effect of acupuncture with different types of comparisons including UC and SA, which makes our findings more comprehensive and reliable. Second, we assessed the acupuncture dose using a previously published scoring instrument. To our knowledge, this is the first meta-analysis to explore the impact of acupuncture dose on treatment effect in patients with DM through subgroup analysis. Third, MA and EA are two common types of acupuncture used to treat DM.^45,46 Exploratory NMA and subgroup analysis were conducted to compare the effects of these two types of acupuncture on IR and determine which is more promising.

Our review has some limitations. First, we failed to compare the effect of acupuncture on HOMA-IR and 2hBG with SA due to the scarcity of studies. Second, the included trials showed substantial heterogeneity, which reduced evidence grade to moderate or low. Part of the observed heterogeneity could be explained by acupuncture type and acupuncture dose. One-study-removed analysis revealed that Shen et al.¹¹ contribute significantly to the heterogeneity. The acupoint selection in this study differed from that in other studies, which might contribute to the heterogeneity. Third, global and local inconsistency was observed in NMA. Although random-effects model was applied to overcome the impact of inconsistency, the result of NMA should be interpreted in a conservative manner. Fourth, with <10 studies included in bias publication analysis (acupuncture vs. UC: 9 studies in FBS, 6 studies in HOMA-IR; acupuncture vs. SA in FBS), funnel plot and Egger’s test are challenging. The small number of studies reduces the reliability of these assessments, and findings regarding potential publication bias should be considered exploratory and interpreted with caution. Fifth, obvious publication bias were observed in the FBS (acupuncture vs. UC) analysis. One explanation might be that three trials were excluded because of insufficient data available, and the authors did not reply to our request for more details. Sixth, although five trials mentioned AEs, most lacked systematic methods for AE collection. The imprecise reporting of safety outcomes across the included studies might limit our meta-analysis to fully evaluate the risk of acupuncture. Greater emphasis on rigorous and transparent safety reporting in future trials is needed to support evidence-based decision-making.

Comparisons with previous studies

At the time of writing, there are three pairwise meta-analysis investigating the effect of acupuncture on patients with DM.^18–20 Two pairwise meta-analysis written by Wang¹⁹ et al and Chen et al.¹⁸ did not include HOMA-IR outcome, which is the most commonly used index to assess IR.⁴⁷ A recent pairwise meta-analysis²⁰ demonstrated that acupuncture is superior to nonacupuncture in reducing FBG and HOMA-IR. However, in this review, the definition of nonacupuncture group was not clear. This review did not provide enough granular evidence to guide clinicians on whether or not to choose acupuncture in comparison with other therapies. Our study supported that the effects of acupuncture differ depending on different controls, providing sufficient and comprehensive evidence for clinical practice.

Notably, few previous studies investigated the associations between acupuncture dose and the effect of acupuncture on DM. Our findings showed that acupuncture dose might be one of the contributing factors to its therapeutic effects. Besides, this is the first exploratory NMA to investigate the comparative effect of different acupuncture types, and the results showed EA was the most promising acupuncture in improving IR. Finally, four RCTs were published recently, and we included these recently published RCTs^32–44 and performed an up-to-date and comprehensive meta-analysis and exploratory NMA.

Implications for clinical practice and research

IR is a systemic disorder that has proven to be associated with various metabolic disorders.⁷ A number of prospective observational studies have indicated that IR is related to the risk of DM, obesity, and polycystic ovary syndrome.^6,48,49 Acupuncture has drawn the attention of the public and been proposed as an effective intervention to improve IR. However, the effect of acupuncture on IR in patients with DM remains inconclusive as some studies reported large gains,^11–15 whereas others reported limited or insignificant improvement. The discrepancies in acupuncture-induced gains might result from differences in acupuncture type and acupuncture dose. Compared with UC, our findings revealed that acupuncture had significant effect on HOMA-IR, FBS, and HbA1c but not 2hBG. Compared with SA, our findings revealed that acupuncture had significant effect on FBS but not 2hBG. According to NMA results, EA might be associated with better effect on IR improvement. This is not surprising as EA provides a stronger treatment dose to acupoints with electrical stimulation on the basis of needle acupuncture. However, the comparative effectiveness of EA versus MA should be interpreted with caution due to the fact that EA is not significantly superior to MA according to the league table. Therefore, head-to-head comparative trials are urgently needed to verify this finding.

In terms of acupuncture dose, high-dose acupuncture showed better effect on HOMA-IR and 2hBG. Considering challenges in determining optimal dose, there is an urgent need for study to find the proper acupuncture dose to obtain the optimal therapeutic outcomes.⁵⁰ Our findings explained why some clinical trials using inadequate acupuncture dose showed negative results and supported the adequate use of acupuncture dose in the treatment of DM. Future studies are needed to investigate the dose–effect relationship of acupuncture. Furthermore, detailed reporting of acupuncture dose in clinical trials is essential.

Conclusion

Our findings suggested that acupuncture was effective and safe in improving IR in patients with DM. Furthermore, we found two potential contributing factors—EA and high-dose acupuncture, which possibly indicate providing greater and more stable effects. With the limitations of this study in mind, these findings should be considered in a conservative manner. Furthermore, more rigorous and high-quality clinical trials are needed to validate our findings and provide more direct evidence.

Authors’ Contributions

T.L.: Conceptualization, project administration, software, and writing—original draft. D. Liu and D. Li: Methodology. D.H.: Data curation, resources, and writing—original draft. C.M.: Data curation and resources. N.L: Data curation and validation. Y.P. and Y.X.: Formal analysis and visualization. Y.Z.: Investigation, software, and supervision. N.Z.: Funding acquisition, supervision, and writing—review and editing. Y.W.: Conceptualization, funding acquisition, investigation, supervision, and writing—review and editing.

Footnotes

Author Disclosure Statement

The authors have no potential conflicts of interest to report.

Funding Information

No funding was received for this article.

Supplemental Material

References

Abel

, Gloyn

, Evans-Molina

, et al. Diabetes mellitus—Progress and opportunities in the evolving epidemic. Cell, 2024; 187(15):3789–3820; doi: 10.1016/j.cell.2024.06.029

Gal

, Burchell

. Diabetes mellitus and the kidneys. Vet Clin North Am Small Anim Pract, 2023; 53(3):565–580; doi: 10.1016/j.cvsm.2023.01.006

Wong

, Sattar

. Cardiovascular risk in diabetes mellitus: Epidemiology, assessment and prevention. Nat Rev Cardiol, 2023; 20(10):685–695; doi: 10.1038/s41569-023-00877-z

Harborg

, Kjærgaard

, Thomsen

, et al. New horizons: Epidemiology of obesity, diabetes mellitus, and cancer prognosis. J Clin Endocrinol Metab, 2024; 109(4):924–935; doi: 10.1210/clinem/dgad450

Murray

CJL

, GBD 2021 Collaborators. Findings from the global burden of disease study 2021. Lancet, 2024; 403(10440):2259–2262; doi: 10.1016/S0140-6736(24)00769-4

Młynarska

, Czarnik

, Dzieża

, et al. Type 2 diabetes mellitus: New pathogenetic mechanisms, treatment and the most important complications. Int J Mol Sci, 2025; 26(3):1094; doi: 10.3390/ijms26031094

, Chi

, Wang

, et al. Trends in insulin resistance: Insights into mechanisms and therapeutic strategy. Signal Transduct Target Ther, 2022; 7(1):216; doi: 10.1038/s41392-022-01073-0

Rahul

, Chintha

, Anish

, et al. Effectiveness of a non-pharmacological intervention to control diabetes mellitus in a primary care setting in Kerala: A cluster-randomized controlled trial. Front Public Health, 2021; 9:747065; doi: 10.3389/fpubh.2021.747065

Wang

, Kang

, Liu

, et al. Fighting diabetes mellitus: Pharmacological and non-pharmacological approaches. Curr Pharm Des, 2020; 26(39):4992–5001; doi: 10.2174/1381612826666200728144200

10.

Osman

AAM

, Seres-Bokor

, Ducza

. Diabetes mellitus therapy in the light of oxidative stress and cardiovascular complications. J Diabetes Complications, 2025; 39(2):108941; doi: 10.1016/j.jdiacomp.2024.108941

11.

Shen

, Kong

. Effects of acupuncture on mood and glucose metabolism in the patient of type 2 diabetes. Chin Acupunct Moxibust, 2007; 27(10):741–743.

12.

Cai

, Zhao

, et al. Effect of acupuncture on serum leptin level in patients with type II diabetes mellitus. Acupunct Res, 2011; 36(4):288–291; doi: 10.13702/j.1000-0607.2011.04.014

13.

Firouzjaei

, Li

, Wang

, et al. Comparative evaluation of the therapeutic effect of metformin monotherapy with metformin and acupuncture combined therapy on weight loss and insulin sensitivity in diabetic patients. Nutr Diabetes, 2016; 6(5):e209; doi: 10.1038/nutd.2016.16

14.

Wang

, Kao

, Chen

, et al. A single-blinded, randomized pilot study evaluating effects of electroacupuncture in diabetic patients with symptoms suggestive of gastroparesis. J Altern Complement Med, 2008; 14(7):833–839; doi: 10.1089/acm.2008.0107

15.

El-Shamy

, El-Kholy

, Labib

, et al. Ameliorative potential of acupressure on gestational diabetes mellitus: A randomized controlled trial. J Complement Integr Med, 2018; 16(1):10.1515/jcim; doi: 10.1515/jcim-2018-0011

16.

Tong

, Jia

, Wang

, et al. Guideline for diagnosis and treatment of type 2 diabetes mellitus with integrated Chinese and western medicine. Jilin J Chin Med, 2024; 44(10):1117–1127; doi: 10.13463/j.cnki.jlzyy.2024.10.001

17.

Van der Schueren

, Vrijlandt

, Thomson

, et al. New guideline of the European Medicines Agency (EMA) on the clinical investigation of medicinal products in the treatment and prevention of diabetes mellitus. Diabetologia, 2024; 67(7):1159–1162; doi: 10.1007/s00125-024-06162-z

18.

Chen

, Liu

, Sun

, et al. Acupuncture for type 2 diabetes mellitus: A systematic review and meta-analysis of randomized controlled trials. Complement Ther Clin Pract, 2019; 36:100–112; doi: 10.1016/j.ctcp.2019.04.004

19.

Wang

, Xu

, Wan

, et al. Acupuncture in treating obesity combined with type 2 diabetes mellitus: A systematic review and meta-analysis of randomized controlled clinical trials. Complement Ther Clin Pract, 2022; 49:101658; doi: 10.1016/j.ctcp.2022.101658

20.

, Chen

, Liu

, et al. Effect and safety of acupuncture for type 2 diabetes mellitus: A systematic review and meta-analysis of 21 randomised controlled trials. Chin J Integr Med, 2022; 28(5):463–471; doi: 10.1007/s11655-021-3450-2

21.

Ya-L

. Observation of the effect of electric acupuncture in treating type 2 diabetes with lung heat and fluid injury type. Chin J Tradit Chin Med Pharm, 2014; 29(6):1818–1821; doi: 10.13288/j.11-2166/r.2013.10.012

22.

Liu

, Duan

, Zhou

, et al. Clinical effect and contributing factors of acupuncture for knee osteoarthritis: A systematic review and pairwise and exploratory network meta-analysis. BMJ Evid Based Med, 2024; 29(6):374–384; doi: 10.1136/bmjebm-2023-112626

23.

, Li

, Yang

, et al. Correlation between acupuncture dose and pregnancy outcomes in women with polycystic ovary syndrome undergoing in vitro fertilization-embryo transfer: A systematic review. BMC Complement Med Ther, 2024; 24(1):407; doi: 10.1186/s12906-024-04695-9

24.

Page

, McKenzie

, Bossuyt

, et al. The PRISMA 2020 statement: An updated guideline for reporting systematic reviews. BMJ, 2021; 372:n71; doi: 10.1136/bmj.n71

25.

Sterne

JAC

, Savović

, Page

, et al. RoB 2: A revised tool for assessing risk of bias in randomised trials. BMJ, 2019; 366:l4898; doi: 10.1136/bmj.l4898

26.

Brozek

, Akl

, Alonso-Coello

, et al.; GRADE Working Group. Grading quality of evidence and strength of recommendations in clinical practice guidelines. Part 1 of 3. An overview of the GRADE approach and grading quality of evidence about interventions. Allergy, 2009; 64(5):669–677; doi: 10.1111/j.1398-9995.2009.01973.x

27.

Abbasi

, Okeke

, Reaven

. Evaluation of fasting plasma insulin concentration as an estimate of insulin action in nondiabetic individuals: Comparison with the homeostasis model assessment of insulin resistance (HOMA-IR). Acta Diabetol, 2014; 51(2):193–197; doi: 10.1007/s00592-013-0461-2

28.

Sun

, Tu

, Lin

, et al. Correlation between acupuncture dose and effectiveness in the treatment of knee osteoarthritis: A systematic review. Acupunct Med, 2019; 37(5):261–267; doi: 10.1136/acupmed-2017-011608

29.

Park

, Ryu

, Liu

, et al. A literature review of de qi in clinical studies. Acupunct Med, 2013; 31(2):132–142; doi: 10.1136/acupmed-2012-010279

30.

Yuan

, Wang

, Hu

, et al. A review of the methods used for subjective evaluation of De Qi. J Tradit Chin Med, 2018; 38(2):309–314.

31.

Jaruvongvanich

, Poonsombudlert

, Ungprasert

. Smoking and risk of microscopic colitis: A systematic review and meta-analysis. Inflamm Bowel Dis, 2019; 25(4):672–678; doi: 10.1093/ibd/izy296

32.

Zhang

, Li

, Zhao

, et al. The efficacy and safety of pilates exercise in patients with knee osteoarthritis: A systematic review with meta-analysis of randomized controlled trials. Ann Med, 2025; 57(1):2540616; doi: 10.1080/07853890.2025.2540616

33.

Zhang

, Luo

, Xue

, et al. The impact of exercise on prostate-specific antigen and testosterone in prostate cancer: A systematic review and meta-analysis. Aging Male, 2025; 28(1):2549292; doi: 10.1080/13685538.2025.2549292

34.

Higgins

JPT

, Green

., editors. Cochrane Handbook for Systematic Reviews of Interventions [Internet]. Version 5.1.0. London: The Cochrane Collaboration; 2011. Available from: https://training.cochrane.org/handbook

35.

Patsopoulos

, Evangelou

, Ioannidis

. Sensitivity of between-study heterogeneity in meta-analysis: Proposed metrics and empirical evaluation. Int J Epidemiol, 2008; 37(5):1148–1157; doi: 10.1093/ije/dyn065

36.

Wang

, Liu

, Xu

. A clinical observation on electro-acupuncture for 35 cases of type 2 diabetes mellitus with intense stomach heat pattern. J Tradit Chin Med, 2013; 54(10):852–854 + 857.

37.

, Huang

, Zhang

, et al. The short-term effects of acupuncture on patients with diabetic gastroparesis: A randomised crossover study. Acupunct Med, 2015; 33(3):204–209; doi: 10.1136/acupmed-2014-010670

38.

Yang

, Zhang

, Li

, et al. Effect of acupuncture for regulating spleen and stomach on aspirin resistance in patients with type 2 diabetes mellitus: A randomized controlled trial. Zhongguo Zhen Jiu, 2021; 41(11):1241–1247; doi: 10.13703/j.0255-2930.20201104-0002

39.

Wang

, Liu

, Xu

. Efficacy observation of acupuncture and tapping therapy in the treatment of type 2 diabetes of yin deficiency pattern combined with stasis in the patients. Chin Acupunct Moxibust, 2014; 34(8):731–735; doi: 10.13703/j.0255-2930.2014.08.004

40.

Simadibrata

, Budihardjo

, Srilestari

. The effect of electroacupuncture at the MA-IC 3 endocrine ear acupoint on fasting blood glucose levels in type 2 diabetes mellitus patients. J Phys: Conf Ser, 2017; 884:e012129; doi: 10.1088/1742-6596/884/1/012129

41.

Kazemi

, Wang

, et al. Therapeutic effects of acupuncture on blood glucose level among patients with type-2 diabetes mellitus: A randomized clinical trial. J Tradit Chin Med Sci, 2019; 6(1):101–107; doi: 10.1016/j.jtcms.2019.02.003

42.

Ismail

, Shaaban Abd El-Azeim

. Immediate fasting blood glucose response to electroacupuncture of ST36 versus CV 12 in patients with type 2 diabetes mellitus: Randomized controlled trial. Fam Med Prim Care Rev, 2021; 23(4):437–441; doi: 10.5114/fmpcr.2022.110370

43.

Najafi

, Ghorbani

, Kordi Yoosefinejad

, et al. The effect of acupressure on fasting blood glucose and glycosylated hemoglobin levels in diabetic patients: A randomized controlled trial. Int J Community Based Nurs Midwifery, 2021; 9(2):152–158; doi: 10.30476/ijcbnm.2021.86059.1318

44.

Salmani Mood

, Yavari

, Bahrami Taghanaki

, Mahmoudirad

. The effect of acupressure on fasting blood glucose, glycosylated hemoglobin and stress in patients with type 2 diabetes. Complement Ther Clin Pract, 2021; 43:101393; doi: 10.1016/j.ctcp.2021.101393

45.

Feng

, Fang

, Wang

, et al. Acupoint therapy on diabetes mellitus and its common chronic complications: A review of its mechanisms. Biomed Res Int, 2018; 2018:3128378; doi: 10.1155/2018/3128378

46.

Peplow

. Electroacupuncture treatment of insulin resistance in diabetes mellitus. Acupunct Med, 2015; 33(5):347–349; doi: 10.1136/acupmed-2015-010942

47.

Tahapary

, Pratisthita

, Fitri

, et al. Challenges in the diagnosis of insulin resistance: Focusing on the role of HOMA-IR and Triglyceride/Glucose Index. Diabetes Metab Syndr, 2022; 16(8):102581; doi: 10.1016/j.dsx.2022.102581

48.

Tong

, Xu

, Huang

, et al. Obesity and insulin resistance: Pathophysiology and treatment. Drug Discov Today, 2022; 27(3):822–830; doi: 10.1016/j.drudis.2021.11.001

49.

Shang

, Zhou

, Hu

, Feng

. Effect of diet on insulin resistance in polycystic ovary syndrome. J Clin Endocrinol Metab, 2020; 105(10):dgaa425; doi: 10.1210/clinem/dgaa425

50.

Fei

Y-T

, Cao

H-J

, Xia

R-Y

, et al. Methodological challenges in design and conduct of randomised controlled trials in acupuncture. BMJ, 2022; 376:e064345; doi: 10.1136/bmj-2021-064345

51.

Yang

, Liu

. BO’s abdominal acupuncture for obese type 2 diabetes mellitus. Chin Acupunct Moxibust, 2015; 35(4):330–334; doi: 10.13703/j.0255-2930.2015.04.005

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

44.69 MB

0.16 MB

0.17 MB

0.10 MB

0.22 MB

27.84 MB

31.92 MB