S174. POLYENVIRONMENTAL AND POLYGENIC RISK SCORES AND THE EMERGENCE OF PSYCHOTIC EXPERIENCES IN CHILDREN AND ADOLESCENTS

Abstract

BackgroundPsychotic experiences (PE) include subliminal hallucinations and delusions without the characteristic functional impairment that constitutes a psychotic disorder. PE are prevalent during childhood and adolescence and studies show a clear link with higher risk to clinical psychosis and schizophrenia. The persistence and accumulation of psychosocial problems are also well established risk factors, but how they interplay with genetic risk is still unclear, especially during developmental stages. Polygenic risk score to schizophrenia (PRS-SZ) and the polyenviromic risk score (PERS) are two validated measures created to assess the contribution of each factor on the development of such psychopathology. Our aim was to verify if PRS and PERS jointly are able to predict psychotic experiences in a cohort of children and adolescents, considering two time-points.MethodsWe analyzed data from the High Risk Cohort (HRC) for Psychiatric Disorders, composed of 2511 children and adolescents from São Paulo and Porto Alegre. PRS-SZ was calculated using summary statistics from the PGC and corrected for the ten first principal components (PC) of the GWAS. In order to calculate the PERS, we used data corresponding to the nine variables that are consider on the score, being respectively, winter or spring birth, urbanicity, cannabis use, advanced paternal age, obstetric and perinatal complications, physical and sexual abuse, neglect and paternal death, therefore if the person is exposed to one or more enviromic factor the odds ratio corresponding to that factor are added up and divided by all factors considered on the calculation, generating the final score. PE was assessed through the Community Assessment of Psychotic Experiences (CAPE) and a latent variable was generated through confirmatory factor analysis producing a good model fit. The prediction model was performed using different linear regressions where the clinical outcome was the CAPE score and PRS and PERS as independent variables. We performed Spearman’s correlations in order to observe possible correlation between our variables.ResultsOur sample varied from 9 to 18 years old (Mean: 13.49, SD: 1.9, 53.9% male) and a total of 1704 individuals provided available CAPE scores, PRS and PERS. When Spearman’s correlations were performed, we observed a non-significant weak positive correlation between PERS x CAPE (R2 = 0.0118, p = 0.623) and between PRS x CAPE (R2 = 0.0292, p = 0.228) and a non-significant negative correlation between PERS x PRS (R2 = -0.03051, p = 0.207). Lastly, we perform a multiple linear regression and used in the model the ten first PC as covariables and, we observed that with an increase in one unit in the PRS, the model explain positively about 8% of the PE variance (R2 = 0.007986 (F(12;1691) = 2.143, p = 0.01225). When we used the PRS already adjusted by ten first PC in the model, this significance is lost (R2=0.0008381 F(2);1701, p=0.1804 (PC2 and PC8 explaining the most of variance).DiscussionPrevious studies have shown a lack of significant association between PRS-SZ and PE for youth samples. Our results are in line with such results, but also depict a trend direction for those variables. Although all correlations were non-significant statically, they show us their direction as discussed below. The higher PERS, higher the psychotic experience, suggesting known environment risk factors for psychosis play a role in the report of PE as well. The higher PRS, the higher psychotic experience also. On the other hand, we found a negative correlation between PERS and PRS.In addition, PERS and PRS jointly were not able to predict psychotic experience. Although non-significant, our results may shed light on knowledge of disease.