Non-invasive neural stimulation improves upper limb function in subacute stroke: meta-analysisBrain stimulation may help stroke survivors regain arm function

IntroductionNon-invasive neural stimulation (NINS) has been increasingly adopted as an adjunctive strategy in stroke rehabilitation. However, its effectiveness in improving upper limb function during the subacute stage remains uncertain. This study aimed to evaluate the effects of different NINS modalities on upper limb function and activities of daily living in patients with subacute stroke.MethodsA systematic review and meta-analysis of randomized controlled trials (RCTs) was conducted. PubMed, Embase, Web of Science, Cochrane Library, and major Chinese databases (CNKI, Wanfang, VIP) were systematically searched from inception to October 2025. Eligible studies included adult patients with subacute stroke (7 d–6 m) receiving NINS. The primary outcomes were upper limb function and activities of daily living. Outcome data were extracted at the end of the intervention for quantitative synthesis. Standardized mean differences (SMDs) with 95% confidence intervals (CIs) were calculated using a random-effects model. Subgroup and meta-regression analyses prespecified in the study protocol were conducted based on recovery stage and stimulation modalities.ResultsA total of 24 RCTs (n = 1,091 patients) were included. Compared with the control group, statistically substantial between-group differences were observed for FMA-UE [SMD = 0.60, 95% CI (0.30, 0.89)], Barthel Index [SMD = 0.77, 95% CI (0.42, 1.13)], ARAT [SMD = 0.66, 95% CI (0.40, 0.93)], and WMFT [SMD = 0.75, 95% CI (0.22, 1.27)] (all p  0.05). Subgroup analyses showed substantial between-group differences in some modality-specific comparisons; however, no clear superiority was observed. In the transcranial direct current stimulation (tDCS) subgroup, a statistically substantial between-group difference in MAS scores was observed. Residual heterogeneity may reflect unmeasured confounders and limit the interpretability of pooled estimates.ConclusionNINS may be associated with modest, context-dependent improvements in upper limb function and activities of daily living in patients with subacute stroke. However, given the substantial heterogeneity and methodological variability, current evidence remains limited and should be interpreted cautiously. Further large-scale, high-quality trials are required to clarify optimal stimulation modalities and clinical applicability.Systematic review registrationPROSPERO, CRD420251242784.