Combined noninvasive brain stimulation and acupuncture for post-stroke motor recovery: a systematic review and network meta-analysis.

OBJECTIVE: Stroke is highly prevalent among older adults, and age-related reductions in neuroplastic reserve may limit motor recovery. This systematic review and network meta-analysis aimed to compare and rank the effects of noninvasive brain stimulation combined with acupuncture on post-stroke motor recovery, spasticity, balance, and activities of daily living (ADL). METHODS: PubMed, Embase, the Cochrane Library, Web of Science, Scopus, CNKI, SinoMed, Wanfang, and CQVIP were searched from inception to July 1, 2026. Randomized controlled trials evaluating noninvasive brain stimulation, including TMS-based NIBS or transcranial direct current stimulation, combined with acupuncture or electroacupuncture were included. Pairwise and frequentist random-effects network meta-analyses were performed. Risk of bias was assessed using the Cochrane tool, and confidence in network estimates was evaluated using CINeMA. RESULTS: Forty-six randomized controlled trials involving 4,080 participants were included. Combined interventions generally showed greater effects than single-modality therapies or conventional rehabilitation, although heterogeneity was substantial. In the network meta-analysis, both TMS-based NIBS plus acupuncture and tDCS plus acupuncture showed favorable ranking probabilities across several outcomes. tDCS plus acupuncture ranked highly for FMA total, FMA-UE, and ADL, whereas TMS-based NIBS plus acupuncture ranked favorably for FMA-LE, spasticity, and balance. Differences between the two combined regimens were generally not statistically significant. CINeMA indicated moderate, low, or very low confidence across network estimates, with no high-confidence evidence. CONCLUSION: Combined noninvasive brain stimulation and acupuncture may provide additional benefits for post-stroke motor and functional recovery. However, current findings should be interpreted cautiously because of substantial heterogeneity, sparse evidence for some intervention nodes, risk-of-bias concerns, and limited confidence in several network estimates. SYSTEMATIC REVIEW REGISTRATION: https://www.crd.york.ac.uk/PROSPERO/view/CRD420251051398, Identifier: CRD420251051398.