Adverse outcomes in trials of graded exercise therapy for adult patients with chronic fatigue syndrome

Graded exercise therapy (GET) is an effective treatment for chronic fatigue syndrome (CFS), but concerns have been raised about its safety. Two randomised controlled trials have not supported these concerns. We further assessed safety outcomes in all ten published trials of GET for CFS. We undertook meta-analyses of three outcomes: Self-ratings of Clinical Global Impression (CGI) change scores of 6 or 7 (“much worse” or “very much worse”), numbers of participants withdrawing from treatments, and numbers of participants dropping out of trial follow up. We provide risk ratios (95% confidence intervals (CI)), comparing GET with control interventions. The 10 trials involved 1279 participants. CGI scores of 6 or 7 were reported by 14/333 (4%) participants after GET and 26/334 (8%) participants after control interventions (RR (CI): 0.62 (0.32, 1.17)). Withdrawals from treatment occurred in 64/535 (12%) participants after GET and 53/534 (10%) participants after control interventions (RR (CI):1.21 (0.86, 1.69)). Drop-outs from trial follow up occurred in 74/679 (11%) participants after GET and 41/600 (7%) participants after control interventions (RR (CI): 1.51 (1.03, 2.22)). The certainty of this evidence was rated low by GRADE, due to imprecision. There was no evidence of excess harm with graded exercise therapy by either self-rated deterioration or by withdrawing from GET, in comparison to control interventions. More GET participants dropped out of trial follow up in comparison to control interventions. Future research should ascertain the most effective and safest form of graded exercise therapy. •Participants were no worse after graded exercise therapy than control interventions.•No more participants withdrew from graded exercise therapy than control interventions.•More participants dropped out from trial follow up after graded exercise therapy.•Graded exercise therapy probably is safe when properly prescribed and supervised.