This systematic review of individual patient data investigated the effects of exercise on cancer-related fatigue and moderators of exercise interventions. Randomised controlled trials in the Predicting OptimaL cAncer Rehabilitation and Supportive care (POLARIS) database were included in the review if they reported fatigue outcomes. All principal investigators from the trials in the POLARIS database provided individual patient data under a data sharing agreement. The methodological quality of each trial was assessed using the Cochrane risk of bias tool. The main outcome was fatigue after completion of the exercise intervention measured using any scale. Potential moderators of treatment effect were based on previous trials and meta-regression analyses and included age, sex, marital status, education level, body mass index, cancer type, treatment type (surgery, chemotherapy, radiotherapy, hormone therapy), and presence of distant metastases. Characteristics of the exercise programs, including frequency, intensity, type, supervision, session time and volume, were also explored as moderators.

Thirty-one trials (n = 4,366 participants) were included. Of these, 2,437 participants were randomised to an exercise intervention group, and 1,929 to a control group. All trials were conducted in high-income countries, including the Netherlands, United States, Australia, Canada, Germany, United Kingdom and Norway.

Exercise reduced fatigue compared to control (effect size -0.17; 95% confidence interval (CI) -0.22 to -0.12). None of the demographic or clinical characteristics of individual participants moderated the effects of the intervention on fatigue. Compared to control, supervised exercise had larger effects on fatigue than unsupervised exercise interventions (effect size -0.18; 95% CI -0.28 to -0.08). Within supervised interventions, those with a duration of up to 12 weeks showed the larger effects (effect size -0.29; 95% CI -0.39 to -0.20) than those with a duration longer than 24 weeks (effect size -0.11; 95% CI -0.22 to 0.00). No other exercise-related characteristics were identified as moderators of supervised exercise interventions. Within unsupervised interventions, neither duration nor exercise-related characteristics moderated the effect of exercise interventions on fatigue.

Exercise interventions reduce fatigue across all subgroups of patients and types of cancer, supporting a role for exercise in clinical practice for people with cancer. The strongest effects on fatigue were noted in supervised exercise interventions with a duration of up to 12 weeks.

van Vulpen JK et al. Moderators of exercise effects on cancer-related fatigue: a meta-analysis of individual patient data. Med Sci Sports Exerc 2020;52(2):303-14

Read more on PEDro.