Altered Cortical Morphology and Morphometric Networks in Cushing’s Disease Linked to Multidimensional Clinical Measures

Cushing's disease (CD) involves chronic hypercortisolism and often presents with neuropsychiatric impairments. However, the underlying structural and network-level brain alterations, and their associations with multidimensional clinical measures remain unclear. In this study, we systematically compared brain morphology and network topology between 66 CD patients and 64 healthy controls. Using surface-based morphometry and individual morphological brain networks constructed from Jensen-Shannon divergence similarity, we analyzed cortical thickness, surface area, volume, as well as global and nodal topological properties. Compared with controls, CD patients exhibited widespread reductions in cortical thickness and surface area. Notably, cortical thickness in the right superior frontal gyrus and bilateral temporal regions showed significant negative correlations with 8 a.m. cortisol levels, Self-Rating Anxiety Scale (SAS) scores, and Self-Rating Depression Scale (SDS) scores, and positive correlations with Montreal Cognitive Assessment (MoCA) scores. Morphological networks preserved small-world architecture but showed reduced local efficiency and hierarchical organization. Nodal analysis showed impaired hub function in limbic and basal ganglia regions. The negative association between left temporal pole nodal efficiency and 24-hour urinary free cortisol remained robust after outlier and non-parametric validation, while other clinical correlations were exploratory. By integrating multiple cortical morphological features and individual-level network analysis, our findings identify morphological metrics in the prefrontal, limbic, and basal ganglia regions as potential neuroimaging biomarkers that may guide monitoring brain structural damage, disease progression, and psychiatric impairments in CD.