Decoding Transdiagnostic Subtypes in Emotional Disorders

Emotional disorders (EDs), including anxiety disorders, major depression (MDD), and post-traumatic stress disorder (PTSD), show overlapping symptoms and neurobiological heterogeneity, limiting current diagnostic frameworks. Transdiagnostic biomarkers are critical for precision psychiatry.

Resting-state functional connectivity (FC) from 261 ED patients and 201 healthy controls (HCs) was decomposed via nonnegative matrix factorization (NMF). Differential FC features defined latent disease factors, classifying patients into subtypes with distinct network/structural profiles.

The analysis identified five representative functional networks (RFNs) underlying the heterogeneity of emotional disorders. RFN1 reflected connectivity between sensorimotor and dorsal attention networks, while RFN2 was defined by theory-of-mind-related connectivity. RFN3, characterized by thalamocortical connections, was specifically linked to somatic sensation processing. RFN4 involved connectivity within the default mode network, and RFN5 captured multi-network-subcortical integration. These RFNs delineated five subtypes with distinct clinical profiles: ST1 (mixed anxiety/depression) exhibited RFN1-driven somatic vigilance and comorbid depressive symptoms; ST2 (mild symptoms) demonstrated resilience associated with RFN2; ST3 (somatic anxiety-dominant) showed RFN3-related elevations in somatic anxiety and panic symptoms; ST4 (anxiety-centric) was marked by RFN4-associated self-referential anxiety; and ST5 represented a depression-pure subgroup mapped to RFN5.

This study proposes a neurobiologically informed stratification for EDs, revealing transdiagnostic subtypes with distinct network signatures and clinical implications. The findings bridge symptom heterogeneity to circuit-level dysfunction, offering potential biomarkers for personalized interventions and advocating for transdiagnostic approaches in psychiatric research.