Segregation and integration of resting-state brain networks in a longitudinal long COVID cohort

Long COVID is characterized by debilitating fatigue, likely stemming from abnormal interactions among brain regions, but the neural mechanisms remain unclear. Here, we utilized a nested-spectral partition (NSP) approach to study the segregation and integration of resting-state brain functional networks in 34 patients with long COVID from acute to chronic phase post infection. Compared to healthy controls, patients with long COVID exhibited significantly higher fatigue scores and shifted the brain into a less segregated state at both 1 month and 3 months post infection. During the recovery of fatigue severity, there was no significant difference of segregation/integration. A positive correlation between network integration and fatigue was observed at 1 month, shifting to a negative correlation by 3 months. Gene Ontology analysis revealed that both acute and long-term effects of fatigue were associated with abnormal social behavior. Our findings reveal the brain network reconfiguration trajectories during post-viral fatigue progression that serve as functional biomarkers for tracking neurocognitive sequelae.