Abstract

Many clinical conditions are associated with a high incidence of chronic fatigue. While some physiological causes for chronic fatigue are established, e.g., processes connected to inflammation, psychological factors may also contribute. The metacognitive theory of dyshomeostasis proposes that a mismatch between cognitive predictions and sensory evidence for actions undermines self-efficacy and perception of control, contributing to chronic fatigue. We aimed to investigate alterations in prediction for physical exertion in participants with chronic fatigue using a new paradigm based on observation, therefore avoiding sensory feedback from the periphery. Participants watched randomised sets of videos with people exercising at different physical exertion levels. Participants had to predict the rate of physical exertion (RPE) of the individuals observed in the videos. Additionally, questionnaires for chronic fatigue, disability, mood, clinical history and body characteristics were assessed. 49 complete data sets from participants with chronic fatigue and 74 data sets from control subjects were analysed in this study. Compared with the control group, participants with chronic fatigue predicted a significantly higher RPE for the observed exercising individuals across all exertion levels. Multiple linear regression models revealed that in the control group, the variance of the bias in the prediction of exertion was significantly explained by the characteristics of the individuals observed in the videos. However, in the chronic fatigue group, the variance of bias in the prediction of exertion was strongly explained by the characteristics of the observer, i.e. fatigue levels and disability. Outcomes revealed that participants with chronic fatigue predicted higher exertion levels during observations and that levels were strongly influenced by their clinical symptoms, suggesting a prediction bias for exertion being present even without performing physical tasks. [Abstract copyright: Copyright © 2026 The Author(s). Published by Elsevier B.V. All rights reserved.]