Relative efficacy of exercise prehabilitation interventions and their components: a systematic review and network meta-analysis of randomised controlled trials

BACKGROUND: Exercise is the most common approach to prehabilitation; the optimal exercise interventions to improve postoperative outcomes are unknown. METHODS: This was a systematic review with network meta-analysis that followed a prespecified protocol (CRD420251034396). Peer-reviewed search strategies were applied from inception to 4 April 2025, to Ovid MEDLINE, Embase, CINAHL, PsycINFO, Web of Science, and Cochrane Central Register of Controlled Trials (CENTRAL), without language restrictions. We included all RCTs of prehabilitation that included at least one exercise intervention (aerobic, strength, pulmonary, or functional exercises) and that reported at least one critical outcome (postoperative complications, length of stay, health-related quality of life, or physical recovery). P-scores were used to rank interventions by their probability of being most beneficial. Best practices, including duplicate review, were followed. Confidence in network meta-analysis informed certainty ratings. RESULTS: We screened 7660 citations, reviewed 1480 full-text articles, and included 96 unique RCTs (N=7396). Combined aerobic, pulmonary, and functional exercise were most likely to reduce complications (P-score 0.94, odds ratio 0.16, 95% CI 0.05-0.60; very low certainty) and length of stay (P-score 0.97, mean difference -3.4 days, 95% CI -5.7 to -1.0; moderate certainty). Combined aerobic and strength exercise was most likely to improve health-related quality of life (P-score 0.93, standardised mean difference 0.67, 95% CI 0.07-1.27; low certainty). No exercise interventions significantly improved physical recovery. CONCLUSIONS: The efficacy of exercise prehabilitation interventions may be outcome-specific. Implementation of aerobic and pulmonary exercise prehabilitation may be warranted. Low-risk-of-bias trials are required to confirm or refute available data, which are mostly low certainty.