Sustained efficacy of acupuncture in post-stroke dysphagia: a systematic review and meta-analysis

Objective To systematically evaluate the short-term and long-term (≥3 months post-treatment) efficacy and safety of acupuncture for post-stroke dysphagia (PSD), providing comprehensive evidence for clinical rehabilitation strategies. Methods Conducted in strict adherence to the PRISMA guidelines, a comprehensive search of Web of Science, PubMed, Embase, the Cochrane Library, VIP, Wanfang, CNKI, and CBM was performed from inception to March 18, 2026 for randomized controlled trials (RCTs) published in English or Chinese. Two independent reviewers performed study selection, data extraction, and risk of bias assessment. Meta-analysis was conducted utilizing Review Manager 5.4 and Stata 18.0. Results A total of 23 RCTs involving 2,224 patients were included. For short-term outcomes, acupuncture interventions significantly improved the Water Swallowing Test (WST) effective rate (RR = 1.50, 95% CI [1.31, 1.73], p < 0.001) and reduced Standardized Swallowing Assessment (SSA) scores (MD = −3.33, 95% CI [−4.46, −2.19], p < 0.001). These significant clinical benefits were robustly sustained for long-term outcomes (≥ 3 months follow-up), as evidenced by the WST effective rate (RR = 1.38, 95% CI [1.21, 1.58], p < 0.001), objective SSA scores (MD = −4.92, 95% CI [−6.54, −3.31], p < 0.001), and subjective SWAL-QOL scores (MD = 23.56, p = 0.010). Subgroup analyses indicated that initiating acupuncture during the subacute/recovery phase yielded pronounced WST improvements. Regarding stimulation modalities, while electroacupuncture showed a marginal potential advantage for the WST ( p = 0.08), no statistically significant superiority was established for the objective SSA scores ( p = 0.56). Acupuncture-related adverse events were confined to mild local reactions (RD = 0.11, p < 0.001). Conclusion Acupuncture is a safe and potentially effective approach for post-stroke dysphagia, improving both short-term swallowing function and quality of life. Notably, these benefits appear to be sustained for at least 3 months post-treatment. Nevertheless, the overall certainty of evidence for short-term efficacy ranges from low to moderate, and the certainty for long-term sustained efficacy ranges from low (WST and SSA) to very low (SWAL-QOL), primarily due to inherent performance bias, publication bias, and imprecision due to limited long-term sample sizes, rather than clinical inconsistency. Systematic review registration https://www.crd.york.ac.uk/PROSPERO/view/CRD420261345975 , identifier (CRD420261345975).