Condition or cognition? Mechanism of change in fatigue in a randomized controlled trial of graded exercise therapy or cognitive behavior therapy for severe fatigue in patients with advanced cancer.

Abstract

OBJECTIVE\nFatigue remains one of the most common and distressing symptoms during treatment for advanced cancer. The TIRED trial demonstrated cognitive behavior therapy s (CBT) significant and clinically relevant effects to reduce fatigue among patients with advanced cancer, while graded exercise therapy (GET) did not prove beneficial. The present study aims to determine the mechanisms by which CBT and GET affect fatigue.\n\n\nMETHOD\nThe TIRED trial randomized 134 patients with advanced cancer to CBT (n = 46), GET (n = 42), or usual care (n = 46). At 14 weeks, 126 evaluable patients provided fatigue data and of those 117 received ≥1 CBT or GET session or usual care. We tested a prespecified multiple mediation model with four potential mediators (physical activity, exercise capacity, self-efficacy, and fatigue catastrophizing) assessed at baseline and at 14 weeks. Post-hoc analyses also included perceived physical activity and emotional functioning as potential mediators.\n\n\nRESULTS\nA total of 82 of 117 patients completed all required measures. CBT reduced fatigue indirectly through its effect on self-efficacy, ab = -3.292; 97.5% bootstrap CI [-6.518 to -0.598]. CBT participants experienced an increase in fatigue self-efficacy, with greater self-efficacy associated with decreased fatigue severity. There was no evidence that changes in physical activity, exercise capacity, perceived physical activity, fatigue catastrophizing, or emotional functioning mediated CBT s or GET s effects on fatigue.\n\n\nCONCLUSIONS\nThe effect of CBT was attributable to changes in cognition, that is, increased self-efficacy led to reduced fatigue severity. No significant mediators for GET were found. The findings inform further refinement of interventions for fatigue in this seriously ill population. (PsycInfo Database Record (c) 2021 APA, all rights reserved).