Identifying Neurobiological Psychosis Biotypes Using Multi-Scale Functional Network Connectivity and its Latent Independent Subspace
Psychotic disorders present a great amount of biological heterogeneity. We explore this by leveraging multiscale functional network connectivity (msFNC) to identify distinct neurobiological psychosis biotypes. Resting-state fMRI data from 2103 B-SNIP 1&2 participants (1127 with psychosis, 350 relatives, 626 controls) were analyzed to obtain subject-specific multiscale intrinsic connectivity networks (msICNs) and compute msFNCs. A low-dimensional neurobiological subspace of multiscale connectivity features termed Latent Network Connectivity (LNC) was identified, consisting of three components (cognitive-related, typical, psychosis-related). Projections of psychosis participants' msFNC onto this subspace revealed three distinct biotypes through unsupervised learning. These biotypes, characterized by cognitive, clinical, and connectivity profiles, spanned all DSM diagnoses (Schizophrenia, Bipolar, Schizoaffective disorders), revealing their overlapping connectivity profiles. Biotype-1, the most cognitively impaired compared to other biotypes, showed Cerebellar-Subcortical and Visual-Sensorimotor hypoconnectivity, alongside Visual-Subcortical hyperconnectivity. Biotype-2 was most cognitively preserved and showed Visual-Subcortical, Subcortical-Sensorimotor, and Subcortical-Higher Cognition hypoconnectivity. Biotype-3 exhibited intermediate cognitive function, showing Cerebellar-Subcortical hypoconnectivity alongside Cerebellar-Sensorimotor and Subcortical-Sensorimotor hyperconnectivity. Notably, unaffected relatives displayed intermediate connectivity patterns, with 55% classified into the same biotype group as their affected family member--a significantly higher rate than random chance (p-valueRelatives-to-Biotype-1 < 0.001, p-valueRelatives-to-Biotype-2 < 0.05, p-valueRelatives-to-Biotype-3 < 0.001) compared to a non-significant 37% alignment using DSM-based diagnosis. Cognitive performance reliably aligns with distinct brain connectivity patterns, which are also evident in unaffected relatives, supporting their construct validity. These findings underscore the limitations of DSM-based classifications in capturing the biological complexity of psychotic disorders and highlight the potential of neurobiological biotypes to enhance our understanding of their spectrum.