Somatosensory processing in long COVID fatigue and its relations with physiological and psychological factors.

Abstract

We aimed to quantify somatosensory processing in long COVID and interactions between somatosensory processing, fatigability, fatigue, autonomic function, mood and illness beliefs. Eighty-eight participants (44 long COVID and 44 controls) were invited to complete two testing sessions, where fatigue was induced by either a cognitive or a physical task in a cross-over design; all participants completed at least one session. Baseline questionnaires assessed trait fatigue, autonomic symptoms, mood and illness beliefs. Pre- and post-task measures included somatosensory processing, state fatigue, perceived effort and heart rate variability (HRV). Group differences and task-related changes were analysed using multivariate and linear mixed models. There was no multivariate group effect on baseline somatosensory measures (P = 0.172), nor did they change following exertion or associate with post-exertion fatigue. Long COVID participants reported greater fatigue than controls, with 64% meeting criteria for severe fatigue. State fatigue was greater at baseline, throughout both exertion tasks (all P < 0.001), and increased more during exertion compared with controls. Despite this, cognitive and physical performance changed similarly across tasks, with no group differences in fatigability (P = 0.199-0.441). Long COVID participants had lower resting HRV, indicating autonomic dysfunction, but HRV was not associated with fatigue. Only group status (P < 0.001) and pre-task fatigue (P < 0.001) were associated with post-exertion fatigue. Within long COVID participants, greater depression (P < 0.001) and perceived illness threat (P = 0.033) were associated with greater trait fatigue. The absence of somatosensory abnormalities provides no support for the sensory attenuation model of fatigue in long COVID.