Potential effects of intradialytic exercise therapy on physical performance in hemodialysis patients based on the intervention period: a systematic review and meta-analysis

Abstract Background Intradialytic exercise during hemodialysis (HD) improves physical function and shows high adherence. In Japan, the 2018 guideline by the Japanese Society of Renal Rehabilitation supports intradialytic exercise, but medical insurance only covers the first 90 days. As new studies have emerged since the guideline was published, we conducted a systematic review and meta-analysis to evaluate the impact and optimal duration of exercise interventions beyond 12 weeks. Methods We followed a preregistered protocol (PROSPERO CRD42025642273) and PRISMA-P guidelines. Randomized controlled trials of adult patients undergoing HD performing structured exercise were identified via MEDLINE database (PubMed, March 2017–November 2024) and ICHUSHI-Web database. Two independent reviewers screened studies, extracted data, and assessed risk of bias using Cochrane RoB 2.0. Outcomes included peak oxygen uptake (VO 2 peak), maximal oxygen uptake (VO 2 max), 6 min walk distance, timed up and go (TUG) test, handgrip strength, sit-to-stand (STS) performance, and biochemical measures (hemoglobin, standard weekly urea Kt/V [std Kt/V]). Pooled mean differences (MD) with 95% confidence intervals (CI) were calculated using a mixed-effects model (Hartung–Knapp–Sidik–Jonkman method). Results A total of 71 studies, including 43 newly identified studies, were included. Risk of bias was moderate-to-high. Interventions > 12 weeks significantly improved 6 min walking distance (MD 52.4 m [95% CI 35.4–69.3], p < 0.01), VO 2 peak (MD 3.5 mL/kg/min [95% CI 1.96–5.04], p < 0.01), VO 2 max (MD 5.29 mL/kg/min [95% CI 2.36–8.22], p < 0.01), hemoglobin (MD 0.94 g/dL [95% CI 0.09–1.78], p = 0.03), and std Kt/V (MD 0.16 [95% CI 0.12–0.20], p < 0.01). TUG (MD −1.68 s [95% CI −2.96 to −0.41], p = 0.02), handgrip strength (MD 4.08 kg [95% CI 1.95–6.22], p < 0.01), and STS performance (MD −2.75 s [95% CI −4.45 to −1.05], p < 0.01) significantly improved in the integrated results of undefined study periods, with nonsignificant trends in studies of > 12 weeks of exercise. Conclusions Our findings suggest that exercise interventions lasting more than 12 weeks can enhance physical performance. Nevertheless, the applicability of these results to older patients remains uncertain, as most evidence is derived from middle-aged cohorts. Trial registration Registered in the PROSPERO database (CRD42025642273).