Inflammation-associated brain functional network topological disruption in female nurses with SWSD: associations with...

Shift work sleep disorder (SWSD) is prevalent among female nurses and is associated with significant health morbidities. While inflammation is implicated in SWSD, how it relates to brain network alterations and clinical symptoms remains underexplored. This study aimed to investigate the associations among peripheral inflammation, brain functional network topological disruptions, clinical symptoms, and transcriptomic signatures in female nurses with SWSD.

Fifty female nurses with SWSD and 50 healthy daytime-working controls (HCs) comparable in age and education underwent clinical assessments, quantification of peripheral inflammatory markers, and resting-state functional magnetic resonance imaging (rs-fMRI). Graph theory was applied to rs-fMRI data to assess brain network topology. Mediation analyses were used to evaluate the pathways linking inflammation, network topology, and symptoms. Imaging transcriptomics, leveraging the Allen Human Brain Atlas, was used to identify gene expression patterns correlated with network alterations. Machine learning models were employed to assess the utility of these multimodal features in classifying SWSD.

Compared with HCs, nurses with SWSD exhibited immune dysregulation (elevated levels of interferon α (IFN-α), IFN-γ, interleukin 4 (IL-4), IL-5, IL-17A, and particularly IL-6). Graph analysis revealed altered global network topology (reduced global efficiency and small-worldness, increased local efficiency, clustering coefficient, and characteristic path length) alongside significant nodal changes, notably increased local efficiency and clustering coefficient in the left medial superior frontal gyrus (SFGmed.L). These topological alterations were significantly correlated with the severity of clinical symptoms. Mediation analyses indicated that global small-worldness mediated the relationship between IL-6 levels and poor sleep quality, whereas the local efficiency of SFGmed.L mediated the associations between IFN-γ levels and anxiety and cognitive performance. The support vector classifier model accurately differentiated nurses with SWSD from HCs (accuracy: 90%). Imaging transcriptomics identified spatial gene-expression patterns associated with altered nodal topology, particularly involving genes related to cytokine signaling and cellular regulation.

Our findings suggest that systemic inflammation is associated with characteristic brain functional network disruptions in female nurses with SWSD, and that these disruptions are associated with clinical symptoms. These inflammation-related neurobiological alterations, together with spatially associated transcriptomic signatures, provide novel insights into SWSD pathophysiology and may help identify potential biomarkers and therapeutic targets.