RESUMO
BACKGROUND: A review of the safety profile of exercise training in multiple sclerosis (MS) has not been conducted since 2013. OBJECTIVE: We undertook a systematic review and meta-analysis of randomised controlled trials (RCTs) of exercise training published since 2013 and quantified estimated population risks of clinical relapse, adverse events (AE) and serious adverse event (SAE). METHODS: Articles reporting safety outcomes from comparisons of exercise training with non-exercise among persons with MS were identified. The risk of bias was established from study's internal validity assessed using Physiotherapy Evidence Database (PEDro). Rates and estimated mean population relative risks (RRs; 95% confidence interval (CI)) of safety outcomes were calculated, and random-effects meta-analysis estimated the mean RR. RESULTS: Forty-six interventions from 40 RCTs (N = 1780) yielded 46, 40 and 39 effects for relapse, AE, adverse effects and SAE, respectively. The mean population RRs ((95% CI), p-value) for relapse, AE and SAE were 0.95 ((0.61, 1.48), p = 0.82), 1.40 ((0.90, 2.19), p = 0.14) and 1.05 ((0.62, 1.80), p = 0.85), respectively. No significant heterogeneity is observed for any outcome. CONCLUSION: In studies that reported safety outcomes, there was no higher risk of relapse, AE, adverse effects or SAE for exercise training than the comparator. Exercise training may be promoted as safe and beneficial to persons with MS.