Acupuncture for glucose metabolism: An updated meta-analysis of randomized controlled trials

This meta-analysis represents the most up-to-date assessment of the efficacy of acupuncture for glucose metabolism, analyzing 14 randomized clinical trials that included 1,179 participants with various metabolic conditions. The study encompassed patients with type 2 diabetes, prediabetes, polycystic ovary syndrome, metabolic syndrome, obesity, and non-alcoholic fatty liver disease, reflecting the breadth of clinical application of acupuncture in metabolic disorders. The researchers conducted a comprehensive search of multiple databases, including PubMed, Web of Science, Embase, and Chinese databases, covering the period from January 2017 to October 2025. Stratified analysis by type of control revealed important findings about when and how acupuncture may be most effective.

The most consistent results were observed when acupuncture was used as an adjuvant therapy to conventional treatment, compared with conventional treatment alone. In this comparison, acupuncture demonstrated statistically significant reductions in HOMA-IR (mean difference -0.96), HbA1c (-0.70%), and fasting glucose (-0.83 mmol/L). Interestingly, when compared only with placebo/sham controls, acupuncture did not show statistically significant benefits for HOMA-IR, HbA1c, or fasting glucose, suggesting that specific effects beyond placebo may be limited or variable. Subgroup analyses by type of acupuncture revealed that manual acupuncture may have more consistent benefits than other modalities such as laser acupuncture or acupressure.

This potential superiority of manual acupuncture may be related to obtaining the 'Deqi' sensation — a specific sensory response considered crucial in traditional Chinese medicine to mediate therapeutic efficacy. The most frequently used points were Sanyinjiao (SP-6), Zhongwan (CV-12), and Zusanli (ST-36), located mainly on the spleen and stomach meridians — areas traditionally associated with digestive and metabolic regulation. Stimulation of these specific points is known to activate sensory nerve fibers that transmit signals to the central nervous system, modulating autonomic outflow by increasing vagal activity and reducing sympathetic tone. This neuroendocrine response may promote insulin secretion and improve hepatic and peripheral insulin sensitivity.

However, interpretation of these results must take significant limitations into account. Statistical heterogeneity was extremely high (I² > 90% for most primary outcomes), indicating substantial variability across studies that could not be explained by meta-regression analyses. This heterogeneity suggests that the benefits of acupuncture are highly context-dependent, including specific acupuncture protocols, patient characteristics, and study methodology. In addition, significant publication bias was detected for some outcomes, along with risks of performance and detection bias inherent to the challenges of blinding practitioners and participants in acupuncture studies.

Most studies were conducted in China, affecting the geographical generalizability of the findings, and few included long-term follow-up. Despite these limitations, the study advances the field by providing an updated synthesis and by attempting to distinguish between efficacy (versus sham control) and comparative effectiveness through stratified analysis. The results suggest that the most promising signal for acupuncture is as adjuvant therapy to conventional treatment, not necessarily as a stand-alone treatment superior to medications. For clinicians, this suggests that acupuncture may be considered a potential adjuvant for patients seeking complementary approaches, with manual acupuncture appearing to be the most promising.