Transition to psychosis is among the most adverse outcomes of clinical high-risk (CHR) syndromes encompassing ultra-high risk (UHR) and basic symptom states. Clinical risk calculators may facilitate an early and individualized interception of psychosis, but their real-world implementation requires thorough validation across diverse risk populations, including young patients with depressive syndromes.
We validated the previously described NAPLS-2 (North American Prodrome Longitudinal Study 2) calculator in 334 patients (26 with transition to psychosis) with CHR or recent-onset depression (ROD) drawn from the multisite European PRONIA (Personalised Prognostic Tools for Early Psychosis Management) study. Patients were categorized into three risk enrichment levels, ranging from UHR, over CHR, to a broad-risk population comprising patients with CHR or ROD (CHR|ROD). We assessed how risk enrichment and different predictive algorithms influenced prognostic performance using reciprocal external validation.
After calibration, the NAPLS-2 model predicted psychosis with a balanced accuracy (BAC) (sensitivity, specificity) of 68% (73%, 63%) in the PRONIA-UHR cohort, 67% (74%, 60%) in the CHR cohort, and 70% (73%, 66%) in patients with CHR|ROD. Multiple model derivation in PRONIA–CHR|ROD and validation in NAPLS-2–UHR patients confirmed that broader risk definitions produced more accurate risk calculators (CHR|ROD-based vs. UHR-based performance: 67% [68%, 66%] vs. 58% [61%, 56%]). Support vector machines were superior in CHR|ROD (BAC = 71%), while ridge logistic regression and support vector machines performed similarly in CHR (BAC = 67%) and UHR cohorts (BAC = 65%). Attenuated psychotic symptoms predicted psychosis across risk levels, while younger age and reduced processing speed became increasingly relevant for broader risk cohorts.
Clinical-neurocognitive machine learning models operating in young patients with affective and CHR syndromes facilitate a more precise and generalizable prediction of psychosis. Future studies should investigate their therapeutic utility in large-scale clinical trials.
Bibliographical noteAcknowledgments and Disclosures
PRONIA is a collaboration project funded by the European Union under the 7th Framework Programme under grant agreement no. 602152. NAPLS-2 was supported by the National Institutes of Health (NIH) (Grant Nos. U01 MH081902 [to TDC], P50 MH066286 [to CEB], U01 MH081857 [to BAC], U01 MH82022 [to SWW], U01 MH066134 [to JA], U01 MH081944 [to KSC], R01 U01 MH066069 [to DOP], R01 MH076989 [to DHM], U01 MH081928 [to LS], and U01 MH081988 [to EFW]). The HARMONY collaboration was supported by the NIH administrative supplement 3U01MH081928-07S1 (to LS). The funders were not involved in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
NK, LK-I, MR, SR, JA, DOP, CEB, BAC, KSC, DHM, TM, LS, MT, EFW, SWW, EM, FS-L, JK, RKRS, PB, SB, SW, and TDC were responsible for study design. NK, MW, LK-I, RS, MR, SR, JA, DOP, CEB, BAC, KSC, DHM, TM, LS, MT, EFW, SWW, EM, FS-L, JK, RKRS, PF, RL, AB, PB, RU, SB, SW, and TDC were responsible for data collection. NK, MW, and DBD were responsible for data analysis. NK, MW, PF-P, KSC, FS-L, PM, REG, and TDC were responsible for data interpretation. NK, MW, and TDC wrote the manuscript. DBD, LK-I, RS, PF-P, MR, SR, AA, JA, DOP, CEB, BAC, KSC, DHM, TM, MT, EFW, SWW, EM, FS-L, JK, RKRS, PF, RL, AB, PB, RU, SB, SW, PM, REG, and TDC reviewed the manuscript. NK takes the final responsibility for the decision to submit this work for publication.
We would like to thank Sen Dong, M.Sc., for the implementation of the models described in this manuscript in the NeuroMiner Model Libary available at www.proniapredictors.eu.
NK and EM hold an issued patent US20160192889A1 (“Adaptive pattern recognition for psychosis risk modelling”). All other authors report no biomedical financial interests or potential conflicts of interest.
- Clinical high-risk states
- First-episode depression
- Machine learning
- Psychosis prediction
- Reciprocal external validation
- Risk calculators
ASJC Scopus subject areas
- Biological Psychiatry