2026 ECNP Congress

PS01-1208

The exposome score for schizophrenia in a first-episode psychosis sample: predictive performance and sample-specific recalibration

Ana Echevarria-Bruno — E. Guerra-Blacio, P. Andrés-Camazón, A. Sánchez, V. Cavone, A. Andreu-Bernabeu, J. Galváñ, M.A. Ayora, M.R. López Urbán, R. Rodríguez Jiménez, J.L. Ayuso Mateos, A. Muñoz-San José, M.F. Bravo Ortíz, K. McDowell, J.C. Leza, M. Díaz-Marsá, Á. Ibáñez, E. Baca, C. Arango, C.M. Díaz-Caneja · Institute of Psychiatry and Mental Health- Hospital General Universitario Gregorio Marañón, Department of Child and Adolescent Psychiatry- IiSGM- CIBERSAM- School of Medicine- Universidad Complutense- Madrid- Spain, Madrid, Spain

2026 ECNP Congress

Background: Evidence suggests that environmental factors play a dual role in psychotic disorders, influencing both the risk of developing psychosis and its subsequent prognosis and clinical outcomes [1, 2].The estimated exposome score for schizophrenia (ES-SCZ), originally developed as a compounded measure of cumulative environmental liability, has been proposed as a tool for gene-environment interaction analyses in population-based cohorts [3, 4]. The ES-SCZ accounts for a range of schizophrenia-associates exposures such as cannabis use, winter birth, hearing impairment, and childhood adversity [5] and it is a validated and robust polygenic and exposome score for schizophrenia.

Objectives: In the present study, the ES-SCZ was applied to characterize environmental risk within a first-episode psychosis sample, while evaluating whether its predictive validity and discriminative capacity are retained in this cohort relative to the original sample consisting of individuals diagnosed with schizophrenia. Additionally, we aimed to examine whether recalibrating the score by re-estimating the weights of individual environmental risk factors within our sample would yield a better-fitted model, potentially improving predictability relative to the original score.

Methods: Data from 753 participants (416 patients and 337 controls) were derived from a cross-sectional case-control study in patients with first-episode psychosis and an age- and gender-matched sample of controls participating in an observational longitudinal study conducted at seven University Hospitals in Madrid, Spain (AGES-CM). Information about demographic and ES-SCZ variables was collected through ad-hoc questionnaires and an adapted version of the Childhood Experience of Care and Abuse (CECA) questionnaire. χ2 tests and Student’s t were used to compare exposure to environmental variables. Binomial logistic regression was used to assess the association of individual ES-SCZ environmental risk factors with case-control status, and to derive sample-specific β coefficients, which were subsequently used as recalibrated weights to compute a refined composite score (Re-ES-SCZ). To determine the discriminative function of ES-SCZ when identifying individuals diagnosed with psychosis in our sample, receiver operating characteristic (ROC) analysis was performed.

Results: ROC analyses indicated that the ES-SCZ demonstrated acceptable discriminative capacity in distinguishing individuals in the first-episode psychosis group from controls, yielding an AUC of 0.67 (95% CI: 0.63; 0.71). Sample-specific recalibration of individual risk factor weights produced a marginally improved AUC of 0.72 (95% CI: 0.68; 0.76) for the Re-ES-SCZ; however, the difference between the two scores did not reach statistical significance (Z= -1.45, p= .148), suggesting that recalibration yielded modest but inconclusive gains in predictive performance.

Conclusions: The ES-SCZ demonstrated adequate predictive validity when applied to a first-episode psychosis sample, supporting its efficacy for environmental risk stratification and characterization in psychotic disorders beyond schizophrenia. Sample-specific recalibration yielded marginal but non-significant improvements in predictive performance. These findings suggest that population-tailored refinement of the score holds potential for improving risk stratification and advancing understanding of the complex environmental contributions to the aetiology of psychotic disorders.

References

  1. Guloksuz, S., van Os, J., Rutten, B.P., 2018. The exposome paradigm and the complexities of environmental research in psychiatry. JAMA Psychiatry. 75(10), 985-986.
  2. Belbasis, L., Köhler, C.A., Stefanis, N., et al., 2018. Risk factors and peripheral biomarkers for schizophrenia spectrum disorders: an umbrella review of meta-analyses. Acta Psychiatr. Scand. 137(2), 88-97.
  3. Erzin, G., Gülöksüz, S., 2021. The Exposome Paradigm to Understand the Environmental Origins of Mental Disorders. Alpha Psychiatry. 22(4), 171-176.
  4. Pries, L.K., Lage-Castellanos, A., Delespaul, P., et al., 2019. Estimating Exposome Score for Schizophrenia Using Predictive Modeling Approach in Two Independent Samples: The Results From the EUGEI Study. Schizophr. Bull. 45(5), 960-965.
  5. Erzin, G., et al., 2021. Examining the association between exposome score for schizophrenia and functioning in schizophrenia, siblings, and healthy controls: Results from the EUGEI study. Eur. Psychiatry. 64(1), e25.