Nonpharmacological interventions for frailty and sarcopenia in older adults with gastrointestinal cancers: A Bayesian network meta-analysis

To compare the effects of nonpharmacological interventions on frailty, sarcopenia, and related physical-function outcomes in older adults with gastrointestinal cancers, we conducted a Bayesian random-effects network meta-analysis of 21 randomized controlled trials from 10 countries. Most interventions did not show clear benefits for frailty, sarcopenia, gait speed, Timed Up and Go performance, sit-to-stand performance, skeletal muscle index, or 6-min walk distance. Prehabilitation exercise showed a low-certainty benefit for handgrip strength, whereas enhanced nutritional support and intensive physiotherapy showed limited benefits for patient-reported physical function. However, these findings were based on small, sparsely connected networks, and the certainty of evidence was generally low or very low. No single intervention was consistently effective across outcomes. These results highlight the need for larger, cancer-specific trials using standardized outcome definitions and clearer reporting of intervention intensity, adherence, safety, and feasibility.