Replication of biological subgroups using resting-state functional connectivity normative modeling in patients with PSD

Motivation: Current findings in resting-state functional connectivity in patients with psychosis spectrum disorders (PSD) are inconsistent. This inconsistency is likely due to the heterogeneous nature of PSD. Furthermore, there is an inadequacy in capturing this heterogeneity. Goal(s): Demonstrate that normative modeling is a technique better capable of capturing heterogeneity and creating biologically based, less-heterogeneous subgroups across multiple patient cohorts. Approach: Functional connectivity values from 3 unrelated cohorts were mapped to a normative range and classified into 1 of 4 subgroups. Results: Subgroups were found to have significantly higher deviation load overlap than the overall group. Impact: Biological subgroupings using normative modeling is a promising and robust method to define patient subgroups at a biological, rather than clinical, level.