Multidimensional Characterization of Long COVID Fatigue

We performed a multidimensional analysis of mood, cognition, sleep and circadian rhythms in patients with post-acute sequelae of SARS-CoV-2 infection (PASC) with the objective of characterizing the phenotype of PASC fatigue.

We recruited adult patients from a Neuro-COVID-19 Clinic with persistence of disabling symptoms beyond 6 weeks from acute infection. Self-reported symptoms were assessed with Patient-Reported Outcomes Measurement Information System instruments. We evaluated cognitive performance using NIH Toolbox measures and assessed sleep and rest-activity rhythms by 7 days of wrist actigraphy. We performed level 2 polysomnography in a subset of 20 participants.

We studied 58 participants: 83% White, 59% female and 91% not hospitalized for COVID-19. Fatigue severity was significantly correlated with worse self-reported cognitive abilities but not with objectively measured cognitive performance and with greater depression symptoms, several rest-activity rhythm and light exposure disruption measures, and greater actigraphy measured sleep time and time in bed. A multivariable model found significant, independent associations between fatigue severity and subjective cognitive abilities, depression symptoms, and rest-activity rhythm disruption.

Long total sleep times, disruption of light exposure and circadian rest-activity patterns, depression and subjective cognitive impairment are associated with PASC fatigue. Behaviorally influenced sleep and circadian abnormalities may exacerbate fatigue and be targets for therapeutic interventions.