Introduction: Prior meta-analyses have demonstrated positive summary effect of tDCS on Fugl-Meyer upper extremity scores (FM-UE) improvement as well as dose-response relationship. With new published manuscripts, we are not sure whether this positive effect still holds.
Hypothesis: tDCS leads to better improvements in post-stroke motor recovery when compared to sham, and in a dose-dependent manner.
Methods: We identified sham-controlled, randomized studies with ≥5 tDCS sessions from PubMed until July 2015. We derived effect size as standardized mean differences (Hedge’s g) of change scores, defined as FM-UE improvements. We calculated summary effect size by using random effect model, taking any data heterogeneity into account. We plotted current, charge (current х duration), total charge (charge х sessions) and their density counterparts (by dividing each with pad size). To assess the dose-response relationship, inverse-variance-based meta-regression was performed.
Results: We identified 9 studies with 234 subjects comparing active stimulation (anodal, cathodal or bihemispheric) with sham stimulation against FM-UE change scores. Stroke patients receiving tDCS improved significantly better than sham (summary Hedge’s g = 0.65, 95% CI = [0.18, 1.13], p = 0.03), with superior results in patients with chronic stroke (average time since stroke >2 years, summary Hedge’s g = 1.11, 95% CI = [0.43, 1.83], p = 0.001) than those with acute stroke (average time since stroke <2 months, summary Hedge’s g = 0.18, 95% CI = [-0.30, 0.66], p = 0.47). Dose-response relationship showed that FM-UE change scores correlated positively with charge density (R2 = 0.20, p = 0.036) and negatively with pad size (smaller the pad, better the effect; R2 = 0.58, p = 0.002).
Conclusion: This updated meta-analysis supports that tDCS leads to better improvement in UE motor outcome than with sham, especially in patients with chronic stroke.