Cortical excitability mapping stratifies neurobiological subtypes of schizophrenia with genetic and molecular signatures

Schizophrenia (SCZ) is marked by profound biological and clinical heterogeneity, presenting major challenges for accurate diagnosis and personalized treatment. Traditional classifications based solely on clinical presentation are limited by inter-individual variability, overlapping symptom profiles, and low stability across disease stages and treatment states. Dysregulation of the excitation–inhibition (E–I) balance within neural circuits is thought to underpin diverse positive and negative symptoms. Classification based on neural excitability may therefore provide critical insights into disentangling this heterogeneity. We applied a cortical excitability (CE) mapping approach to spatially characterize E–I dysregulations in 77 drug-naïve first-episode SCZ (FES) patients and 76 healthy controls (HCs). CE abnormalities were identified using voxel-wise comparisons, and patients were subsequently clustered into subtypes based on the spatial patterns of CE alterations. Longitudinal analyses assessed the subgroups’ clinical trajectories over 12 months of antipsychotic treatment. Furthermore, CE maps were integrated with transcriptomic and neuroreceptor datasets to analyse potential molecular mechanisms underlying the observed CE abnormalities. Relative to HCs, FES patients exhibited CE abnormalities primarily in the bilateral frontal lobes, sensorimotor cortex, and right cuneus. Two subtypes were identified, differing in both the spatial extent of CE abnormalities and their clinical profiles: FES1 showed more widespread CE reductions across frontal and association cortices, associated with greater affective and cognitive burden, whereas FES2 demonstrated a comparatively preserved CE profile and milder symptom expression. Subsequent transcriptomic and receptor analyses revealed distinct biological underpinnings: FES1 was associated with synaptic dysfunction and neurodevelopmental disruption, while FES2 reflected multisystem involvement potentially accompanied by compensatory processes. CE-based subtyping captures clinically and biologically meaningful heterogeneity in FES. Moreover, this approach provides a promising framework for bridging neural circuit dysfunction with molecular signatures and may advance precision psychiatry in SCZ.