Neural White Matter Distinctions in Biologically Classified Psychosis

Psychotic disorders share common phenomenology and genetic risk factors, but the specific etiologies and therefore targeted pharmacological treatments for these disorders remain elusive. Schizophrenia (SZ), schizoaffective disorder (SZA), and bipolar disorder with psychosis (BPP) are all diagnosed based on clinical symptoms, which in their lack of diagnostic specificity, evidence homogeneity between disorders historically considered distinct. Separating probands based on neurobiology rather than clinical symptoms alone may yield more meaningful classifications which then offer better targets for study of disease mechanisms and treatments. The Bipolar and Schizophrenia Network for Intermediate Phenotypes (BSNIP) consortium has collected brain and behavioral data from probands with these three disorders and used k-means clustering and multivariate discriminant analysis to regroup probands into three ”Biotypes” based on electrophysiological, oculomotor, and cognitive measures; this gave two linear discriminant functions that were termed cognitive control and sensorimotor reactivity. The present study examines measures of neural white matter, which were not included in the original Biotypes analysis, in order to determine whether Biotypes provide better group separation than diagnostic classification. Diffusion-weighted imaging data were acquired from probands with SZ (n=30), SZA (n=39), and BPP (n=10), as well as healthy comparison (HC) subjects (n=76); fractional anisotropy (FA) values indexing white matter structure were calculated and compared between these three diagnostic groups, as well as between the three Biotypes previously derived. Between diagnoses, FA measures were lower for SZ probands than HC across several white matter tracts, but often not different between BPP probands and HC. In contrast, FA measures were lower across several tracts for Biotype 2, which includes probands from all three diagnostic categories, than for HC or the other Biotypes. Given that probands within Biotype 2 rate highly on sensorimotor reactivity, this decrease in white matter structure may be related to possibly compensatory, elevated neural activity. These results suggest that white matter differences between Biotypes may reflect true biological alterations in psychosis, which ultimately provide better classification and possible targets for etiological