Differentiating biomarker features and familial characteristics of B-SNIP psychosis Biotypes.

Abstract

Idiopathic psychosis shows considerable biological heterogeneity across cases. The Bipolar-Schizophrenia Network for Intermediate Phenotypes (B-SNIP) used psychosis-relevant biomarkers to identify psychosis Biotypes, which will aid etiological and targeted treatment investigations. Here, our previous approach (Clementz et al. 2022) is updated, which supports the development of an efficient psychosis Biotype diagnostic procedure called ADEPT. Psychosis probands (n = 1907), their first-degree biological relatives (n = 705), and healthy participants (n = 895) completed a biomarker battery composed of cognitive performance, saccades, and auditory EEG/ERP measurements. EEG and ERP quantifications were modified from previous Biotypes iterations. Multivariate integration using multiple approaches reduced biomarker outcomes to 11 "bio-factors." Twenty-four different approaches indicated bio-factor data among probands were best described by three subgroups. Numerical taxonomy with k-means clustering yielded psychosis Biotypes; Rand Indices evaluated individual-case consistency of Biotype assignments. Psychosis subgroups, their non-psychotic first-degree relatives, and healthy individuals were compared across bio-factors. The three psychosis Biotypes differed significantly on all 11 bio-factors, especially prominent for general cognition, antisaccades, ERP magnitude, and intrinsic neural activity. Rand Indices showed excellent individual-case consistency of Biotype membership when samples included more than 1000 subjects. Canonical discriminant analysis described composite bio-factors that simplified group comparisons: "Pattern-2" (high antisaccade errors, low BACS, high ongoing EEG) captured Biotype-2, "Pattern-1" (low ERP amplitudes, low intrinsic EEG) captured Biotype-1, and "Pattern-3" (low frontal P3 complex, accentuated S2 ERP, faster saccadic reaction times) captured Biotype-3. First-degree relatives had patterns like their proband for general cognition, antisaccades, ERP magnitudes, and intrinsic brain activity. These outcomes refine and extend operations for characterizing biologically distinct psychosis Biotypes. They also show that over 1000 observations are useful for achieving consistent individual-case diagnostic assignments. First-degree relative data implicate specific bio-factors as familial within idiopathic psychosis which may inform genetic studies.