Brain dynamics of attentional, default-mode and limbic networks are disrupted at rest in Post-COVID-19 Syndrome

Abstract Background Post-COVID-19 Syndrome (PCS) is characterised by persistent fatigue, cognitive impairments, and affective symptoms, yet its underlying neural mechanisms remain poorly understood. While static neuroimaging studies have identified resting-state connectivity abnormalities in PCS, such approaches fail to capture the brain’s dynamic functional organisation. This represents a missed opportunity to understand how alterations at the dynamic macroscale interactions give rise to the complex and fluctuating symptom profile of PCS. Cognitive and emotional processes rely on the brain’s capacity to flexibly reconfigure large-scale networks over time; disruptions in the exploration and transition between brain states may therefore play a central role in PCS pathophysiology. Methods Resting-state fMRI data were acquired from 20 individuals with PCS (mean age = 41.8 years, SD = 9.4) and 20 age- and sex-matched healthy controls (mean age = 40.6 years, SD = 8.1) using a multi-echo sequence. Following denoising with multi-echo independent component analysis, we applied Leading Eigenvector Dynamics Analysis (LEiDA) to identify recurrent patterns of whole-brain phase synchrony. The optimal number of dynamic brain states was determined using the Dunn index. For each state, we quantified probability of occurrence, lifetime, and transition probabilities, and mapped spatial topographies onto canonical functional networks. Group differences were assessed using ANCOVAs controlling for age, sex, and handedness. Exploratory associations with clinical symptoms, cognitive performance, and inflammatory markers were examined using both frequentist and Bayesian approaches. Results Five recurrent dynamic brain states were identified. Compared with controls, PCS participants showed reduced probability of occurrence and shorter lifetime of a visual/dorsal attention state, alongside increased probability of a limbic/default mode network (DMN) state. PCS was also characterised by reduced transitions between visual/dorsal attention and frontoparietal–DMN states, and increased transitions from somatomotor/visual states toward the limbic-DMN configuration. Exploratory analyses revealed that greater expression of the limbic-DMN state was negatively associated with global cognitive performance (MoCA) and positively associated with serum IL-1β levels. Conclusions PCS is associated with a reorganisation of intrinsic brain dynamics, marked by a shift from externally oriented attentional states toward limbic-DMN configurations. This imbalance may reflect reduced cognitive flexibility and increased sensitivity to interoceptive or affective signals. The observed links between dynamic brain-state expression, cognitive impairment, and peripheral inflammation support a systems-level mechanism through which immune signalling may influence neural function in PCS. Dynamic functional connectivity offers a sensitive framework for capturing these alterations and may inform future mechanistic models of post-viral dysfunction and therapeutic targeting. Highlights Post-COVID-19 Syndrome (PCS) is linked to altered intrinsic brain dynamics at rest. Patients show reduced engagement of visual and attentional brain states. Increased recruitment of limbic-DMN states suggests a shift toward more internally, emotionally charged focussed brain activity. Brain-state dynamics are putatively associated with global cognition and IL-1β. Reduced brain-state flexibility may underlie cognitive and emotional symptoms in PCS.