Effect and moderators of exercise on fatigue in patients with cancer: Meta-analysis of individual patient data.

Abstract

104 Background: Fatigue is a common and disabling complaint in patients with cancer and can be reduced by exercise. To further personalize exercise prescriptions, moderators of exercise effects on fatigue should be investigated. However, most randomized controlled trials (RCTs) are not adequately powered to identify heterogeneity in responses to exercise. Therefore, we conducted a meta-analysis using individual patient data (IPD) of exercise RCTs to investigate the effect and moderators of exercise on cancer-related fatigue. Methods: Within the Predicting OptimaL cAncer RehabIlitation and Supportive care (POLARIS) consortium, principal investigators of 34 exercise RCTs worldwide have shared their IPD, including in total 4366 cancer patients. A 1-step IPD meta-analysis, using a linear mixed-effect model with a random intercept on study was undertaken to investigate effect on fatigue. The result, a between-group difference in standardized z-scores, corresponds to a Cohen’s d effect size. An interaction term was included in the model to assess potential moderators including demographic (sex, age, marital status, education), clinical (body mass index, distant metastasis), intervention-related (timing, delivery mode, duration) and exercise-related (type, frequency, intensity, duration) characteristics. Results: Exercise significantly reduced fatigue (β = -0.17, 95% CI -0.22;-0.12). The effect was not moderated by demographic, clinical or exercise-related characteristics. Supervised exercise had significantly larger effects on fatigue than unsupervised exercise (βdifference= -0.18, 95%CI -0.28;-0.08). Compared to the control group, supervised exercise significantly improved fatigue (β = -0.23, 95%CI = -0.29;-0.17), while unsupervised exercise did not (β = -0.04, 95%CI = -0.13;0.04). Conclusions: Exercise significantly reduces cancer-related fatigue across subgroups formed on the basis of demographic and clinical characteristics. The effect of exercise is significantly larger when performed under supervision. Hence, exercise, and preferably supervised exercise, represents a viable intervention for the prevention and treatment of fatigue among patients with cancer.