Heterogeneity and classification of recent onset psychosis and depression: a multimodal machine learning approach

Paris Alexandros Lalousis, Stephen Wood, Lianne Schmaal, Katie Chisholm, Lowri Griffiths, Renate Reniers, Alessandro Bertolino, Stefan Borgwardt, Paolo Brambilla, Joseph Kambeitz, Rebekka Lencer, Christos Pantelis, Stephan Ruhrmann, Raimo K.R. Salokangas, Frauke Schultze-Lutter, C. Bonivento, Dominic Dwyer, Adele Ferro, Theresa Haidl, Marlene RosenAndré Schmidt, Eva Meisenzahl, Nikolaos Koutsouleris, Rachel Upthegrove, PRONIA Consortium

Research output: Contribution to journalArticlepeer-review

160 Downloads (Pure)


Diagnostic heterogeneity within and across psychotic and affective disorders challenges accurate treatment selection, particularly in early stages. Delineation of shared and distinct illness features at the phenotypic and brain levels may inform the development of more precise differential diagnostic tools. We aimed to identify prototypes of depression and psychosis to investigate their heterogeneity, with common, comorbid transdiagnostic symptoms. Analysing clinical/neurocognitive and grey matter volume (GMV) data from the PRONIA database, we generated prototypic models of recent-onset depression (ROD) vs. recent-onset psychosis (ROP) by training support-vector machines to separate patients with ROD from patients with ROP, who were selected for absent comorbid features (pure groups). Then, models were applied to patients with comorbidity, i.e., ROP with depressive symptoms (ROP+D) and ROD participants with sub-threshold psychosis-like features (ROD+P), to measure their positions within the affective-psychotic continuum. All models were independently validated in a replication sample. Comorbid patients were positioned between pure groups, with ROP+D patients being more frequently classified as ROD compared to pure ROP patients (clinical/neurocognitive model: χ2=14.874; p<0.001; GMV model: χ2=4.933; p=0.026). ROD+P patient classification did not differ from ROD (clinical/neurocognitive model: χ2=1.956; p=0.162; GMV model: χ2=0.005; p=0.943). Clinical/neurocognitive and neuroanatomical models demonstrated separability of prototypic depression from psychosis. The shift of comorbid patients towards the depression prototype, observed at the clinical and biological levels, suggests that psychosis with affective comorbidity aligns more strongly to depressive rather than psychotic disease processes. Future studies should assess how these quantitative measures of comorbidity predict outcomes and individual responses to stratified therapeutic interventions.
Original languageEnglish
Article numbersbaa185
Pages (from-to)1130-1140
Number of pages11
JournalSchizophrenia bulletin
Issue number4
Early online date5 Feb 2021
Publication statusE-pub ahead of print - 5 Feb 2021


  • MRI
  • comorbidity
  • depression
  • gray matter volume
  • mach ine learning
  • psychosis
  • transdiagnostic

ASJC Scopus subject areas

  • Psychiatry and Mental health


Dive into the research topics of 'Heterogeneity and classification of recent onset psychosis and depression: a multimodal machine learning approach'. Together they form a unique fingerprint.

Cite this