Placebo effect of rTMS on post-stroke motor rehabilitation: a meta-analysis

Previous studies have shown that placebo repetitive transcranial magnetic stimulation (rTMS) was effective on post-stroke motor rehabilitation. However, the placebo effect has not been systematically assessed. Therefore, this meta-analysis was conducted to resolve this issue and explore potential influencing factors further. PubMed, Embase, web of science and the Cochrane Library were searched for published randomised controlled trials (RCTs) with placebo rTMS treatment of stroke recovery until May 2019. The placebo effect size (Hedges' g) was estimated using the motor outcome of pre- and post- placebo rTMS treatment. Meta-regression analysis was also performed to explore potential influencing factors for the placebo effect. Twenty-six placebo-controlled trials (including 381 patients in placebo group) were selected. Effect size results (Hedges'g = 0.466, 95% CI 0.207-0.726;P < 0.05) showed a medium and significant placebo rTMS effect on improving post-stroke motor recovery. The mean ratio of the effect size of sham to real stimulation was 56%. Meta-regression analysis did not find significant result except for the treatment sessions, which was significantly correlated with the placebo effect size (r = 0.465,p = 0.031). In the follow-up observations (1, 2 and 3 months), the sham rTMS groups manifested gradually increased motor improvement, which was similar to the real group, but the amplitude was lower, which was sustained for at least 3 months. Placebo effect of rTMS on post-stroke motor recovery was medium but significant. Regarding different sham styles, the number of stimulation sessions had an impact on the effect.