ALTERATIONS IN SHORT-RANGE STRUCTURAL CONNECTIVITY ACROSS THE PSYCHOSIS SPECTRUM: FINDINGS FROM THE B-SNIP STUDY

Abstract Background Schizophrenia (SZ) and bipolar disorder (BD) have been increasingly viewed as psychotic mood disorders along a shared spectrum. Long-range and short-range structural connectivity have been implicated in both disorders, conceptualising them as “disconnection syndromes”. There has been a rise in neuroimaging tools to understand the overlap and boundaries between the two disorders, which has shifted our focus towards appreciating traits in addition to diagnosis. Our recent pilot study examining short-range U-fibers found in superficial white matter (SWM) found shared and distinct traits among people with SZ and BD and we aimed to investigate SWM further using data from the Bipolar-Schizophrenia Network on Intermediate Phenotypes (B-SNIP) consortium. Methods Using diffusion weighted imaging (DWI), we performed whole brain tractography in 113 people with SZ, 69 people with SA disorder, 49 people with psychotic BD and 77 healthy controls using BrainVISA and Connectomist 2.0. Segmentation and labelling of SWM tracts were performed using a comprehensive U-fiber atlas. ComBat was applied to remove site effects and principle components analysis was performed to identity networks of bundles used for comparative analyses. Results Principle component analysis revealed a network comprised of 8 short tracts in frontal, parietal, and temporal regions that had decreased anatomical connectivity in patients, regardless of diagnosis, relative to healthy controls. This network overlaps, in part, regions that differed between patients (SZ and BD) and healthy controls in our recent pilot study. However, we were unable to detect differences between people with SZ, SA disorder and psychotic BD. Discussion We demonstrate that short U-fibers are likely vulnerable to pathological processes in psychotic illnesses, encouraging further understanding of their anatomy and function. Our lack of findings between patient groups may reflect a more homogeneous population (three subgroups of psychosis) and may suggest that abnormalities in SWM are less likely due to mood disturbances.