The Effects of Intradialytic Exercise on Key Indices of Sarcopenia in Patients With End-stage Renal Disease: A Systematic Review of Randomized Controlled Trials.

Eleven of the 14 trials were rated at high risk of bias, so the results rest on studies where the findings may be distorted. Allocation was not properly concealed in 7 trials, outcome assessors were not blinded in 11, and 9 did not analyse people in the group they were randomised to. Ten of the 14 trials had 50 people or fewer, which limits how precise the results are. There were no sham exercise comparators, so some of the improvement could come from attention or from simply being in a trial. The exercise programmes and the outcome measures varied so much between trials that the authors could not pool the numbers, so this review can only suggest what might work rather than settle it. Participants in some trials were functioning at a high level to begin with, so the results may not apply to frailer dialysis patients. There is no agreed clinically important difference for the quality-of-life measures used in this population, so it is unclear whether the quality-of-life improvements matter to patients. The evidence for aerobic exercise was rated low quality, and the review could not compare exercise directly against a sham condition.

Do not read the phrase 'strong evidence' as proof that this works. That label describes how consistent the trial results were, not how well the trials were run: 11 of the 14 were rated at high risk of bias, most had 50 people or fewer, and the authors could not pool the data at all. The review itself concludes there is still not enough evidence to say what exercise dose to prescribe.