Comparison of whole-brain structural connectivity between bipolar disorder and major depressive disorder: A DTI-based connectome analysis.

Background: Bipolar disorder (BD) and major depressive disorder (MDD) are both characterized by disruptions in brain network connectivity, yet the specific patterns of structural connectivity that differentiate these disorders remain poorly understood.

Methods: This study utilized network-based modeling of whole-brain structural connectivity to investigate shared and disorder-specific patterns of white matter organization in these disorders. Diffusion tensor imaging (DTI) data were obtained from 37 BD patients, 54 MDD patients, and 62 healthy controls (HC). Connectivity matrices were constructed using 68 cortical and 14 subcortical nodes (based on the Desikan-Killiany atlas), with edges weighted by streamline counts obtained through probabilistic tractography. Network-based statistics were employed to examine group differences in subnetwork connectivity, and comparisons of global efficiency, node strength, and local efficiency were conducted across groups.

Results: Individuals with BD exhibited significantly higher subnetwork connectivity (pFWE < 0.001) and global efficiency (p < 0.001, Bonferroni corrected) than both HCs and MDD patients, with no significant differences observed between HC and MDD groups. Node-based analyses showed greater node strength and local efficiency mainly in several temporal lobe regions compared to the other two groups (p ≤ 0.001, Bonferroni corrected). Exploratory analyses suggested that local brain connectivity features may be associated with anhedonia in BD patients (p < 0.05, uncorrected). These results remained consistent after excluding medicated patients.

Conclusions: Our findings suggest that BD is characterized by enhanced brain network efficiency, offering new insights into the neurobiological distinctions between BD and MDD.