PT  - JOURNAL ARTICLE
AU  - A.R. Docherty
AU  - Andrey A. Shabalin
AU  - Daniel E. Adkins
AU  - Frank Mann
AU  - Robert F. Krueger
AU  - Archie Campbell
AU  - Caroline Hayward
AU  - David J. Porteous
AU  - Andrew M. McIntosh
AU  - Kenneth S. Kendler
TI  - Subthreshold psychosis symptoms associated with molecular genetic risk in a population-based cohort: Findings from Generation Scotland
AID  - 10.1101/821041
DP  - 2019 Jan 01
TA  - bioRxiv
PG  - 821041
4099  - http://biorxiv.org/content/early/2019/10/28/821041.short
4100  - http://biorxiv.org/content/early/2019/10/28/821041.full
AB  - Importance Subthreshold psychosis symptoms in the general population may be associated with genetic risk for schizophrenia. In this analysis, empirically-derived symptom factor scores led to a detection of significant and robust polygenic signal.Objective This study sought to optimize genetic association with data-driven symptom factor scores, accounting for cohort factor structure and sex differences.Design EFA-derived symptom factor scores were regressed onto PRS for schizophrenia in models accounting for age and genetic ancestry principal components. Follow-up examination of symptom factor score associations with other related genetic risks included ADHD, autism, bipolar disorder, major depression, and neuroticism.Participants This study examined the newly expanded symptom dataset from the Northern European ancestry cohort, Generation Scotland: Scottish Family Health Study (N = 9,105 individuals 18-65 years of age) comprising common variant and subthreshold psychosis symptom data. A total of 5,391 females and 3,713 males with age M[SD] = 45.2 [13] were included in the final analyses.Main Outcome and Measure Subthreshold psychosis symptoms were measured using the Schizotypal Personality Questionnaire-Brief (SPQ-B). Primary phenotypic factor scores and genome-wide polygenic risk scores (PRS) reflected weighted sum scores and were examined as continuous measures. Polygenic risk scores were calculated from genome-wide association summary statistics using 7,358,674 imputed common genetic variants passing quality control.Results In males, symptom factor scores were positively associated with polygenic risk for schizophrenia alone and implicated primarily interpersonal/negative symptoms. In females, symptom factor scores were positively associated with polygenic risks for ADHD and autism but not schizophrenia. Scores were robustly associated with genetic risk for neuroticism across both males and females.Conclusions and Relevance This study detected a significant association of subthreshold psychosis symptoms with genetic risk for schizophrenia and neuroticism in a population-based sample. Furthermore, important sex differences suggest a need for better understanding of schizophrenia risk assessment in females.Question What molecular genetic risks are associated with subthreshold psychosis symptoms in the general population?Findings In a large population-based cohort (N = 9,084), significant associations of polygenic risks with symptoms were observed. Symptoms were associated with genetic risk for schizophrenia in males, for ADHD and autism spectrum disorder in females, and for neuroticism across both males and females.Meaning Associations of genetic risk with symptoms in the general population are highly significant and suggest important sex differences.