Replicated evidence that endophenotypic expression of schizophrenia polygenic risk is greater in healthy siblings of patients compared to controls, suggesting gene-environment interaction. The EUGEI study

Jim van Os*, Lotta-Katrin Pries, Philippe Delespaul, Gunter Kenis, Jurjen J Luykx, Bochao D Lin, Alexander L Richards, Berna Akdede, Tolga Binbay, Vesile Altınyazar, Berna Yalınçetin, Güvem Gümüş-Akay, Burçin Cihan, Haldun Soygür, Halis Ulaş, Eylem Şahin Cankurtaran, Semra Ulusoy Kaymak, Marina M Mihaljevic, Sanja Andric Petrovic, Tijana MirjanicMiguel Bernardo, Bibiana Cabrera, Julio Bobes, Pilar A Saiz, María Paz García-Portilla, Julio Sanjuan, Eduardo J Aguilar, José Luis Santos, Estela Jiménez-López, Manuel Arrojo, Angel Carracedo, Gonzalo López, Javier González-Peñas, Mara Parellada, Nadja P Maric, Cem Atbaşoğlu, Alp Ucok, Köksal Alptekin, Meram Can Saka, Celso Arango, Michael O'Donovan, Bart P F Rutten, Sinan Guloksuz, Genetic Risk and Outcome of Psychosis (GROUP) Investigators

*Corresponding author for this work

Research output: Contribution to journalArticleAcademicpeer-review

22 Citations (Web of Science)


BACKGROUND: First-degree relatives of patients with psychotic disorder have higher levels of polygenic risk (PRS) for schizophrenia and higher levels of intermediate phenotypes.

METHODS: We conducted, using two different samples for discovery (n = 336 controls and 649 siblings of patients with psychotic disorder) and replication (n = 1208 controls and 1106 siblings), an analysis of association between PRS on the one hand and psychopathological and cognitive intermediate phenotypes of schizophrenia on the other in a sample at average genetic risk (healthy controls) and a sample at higher than average risk (healthy siblings of patients). Two subthreshold psychosis phenotypes, as well as a standardised measure of cognitive ability, based on a short version of the WAIS-III short form, were used. In addition, a measure of jumping to conclusion bias (replication sample only) was tested for association with PRS.

RESULTS: In both discovery and replication sample, evidence for an association between PRS and subthreshold psychosis phenotypes was observed in the relatives of patients, whereas in the controls no association was observed. Jumping to conclusion bias was similarly only associated with PRS in the sibling group. Cognitive ability was weakly negatively and non-significantly associated with PRS in both the sibling and the control group.

CONCLUSIONS: The degree of endophenotypic expression of schizophrenia polygenic risk depends on having a sibling with psychotic disorder, suggestive of underlying gene-environment interaction. Cognitive biases may better index genetic risk of disorder than traditional measures of neurocognition, which instead may reflect the population distribution of cognitive ability impacting the prognosis of psychotic disorder.

Original languageEnglish
Article number003329171900196
Pages (from-to)1884-1897
Number of pages14
JournalPsychological Medicine
Issue number11
Early online date15 Aug 2019
Publication statusPublished - Aug 2020


  • Cognition
  • genetics
  • schizophrenia
  • schizotypy

Cite this