Association of schizophrenia polygenic risk score with manic and depressive psychosis in bipolar disorder

Abstract

Bipolar disorder (BD) is highly heterogeneous in symptomatology. Narrowing the clinical phenotype may increase the power to identify risk genes that contribute to particular BD subtypes. This study was designed to test the hypothesis that genetic overlap between schizophrenia (SZ) and BD is higher for BD with a history of manic psychosis. Analyses were conducted using a Mayo Clinic Bipolar Biobank cohort of 957 bipolar cases (including 333 with history of psychosis during mania, 64 with history of psychosis only during depression, 547 with no history of psychosis, and 13 with unknown history of psychosis) and 778 controls. Polygenic risk score (PRS) analysis was performed by calculating a SZ-PRS for the BD cases and controls, and comparing the calculated SZ risk between different psychosis subgroups and bipolar types. The SZ-PRS was significantly higher for BD-I cases with manic psychosis than BD-I cases with depressive psychosis (Nagelkerke’s R2 = 0.021; p = 0.045), BD-I cases without psychosis (R2 = 0.015; p = 0.007), BD-II cases without psychosis (R2 = 0.014; p = 0.017), and controls (R2 = 0.065; p = 2 × 10−13). No other significant differences were found. Our results show that BD-I with manic psychosis is genetically more similar to SZ than any other tested BD subgroup. Further investigations on genetics of distinct clinical phenotypes composing major psychoses may help refine the current diagnostic classification system.

A principal component approach to improve association testing with polygenic risk scores

“…Both the PRS-PCA (K = 11) and the Opt-Perm approaches reproduced our previous finding that the PRS for SZ is higher in cases with a history of manic psychosis (N = 336) than those without a history of psychosis (N = 309) 11 . Both methods also showed cases with manic psychosis had higher genetic load for educational attainment.…”

“…This choice performed well in the simulations and, unlike the PRS-PCA approach with only K = 5, was able to reproduce the finding of Markota et al . 11 .…”

“…We generated genotypes by sampling without replacement from the Mayo Clinic Bipolar Disorder Biobank sample, followed by generating phenotypes conditional on (or independent of) the genotypes. The Mayo Clinic Bipolar Disorder Biobank collection, genotyping, and genetic data quality control has been described in previous publications 11,12 and is summarized in the supplement.…”

“…We explored the performance of the methods using samples sizes of N = 500 and 1500 with a balanced case-control design. Using GCTA 13 , we simulated the liability of a trait with realistic effect sizes across the genome by randomly choosing the effect size of each SNP from a normal distribution with mean equal to log and standard deviation of the corresponding SNP in the summary statistics from the Psychiatric Genomics Consortium (PGC) Schizophrenia (SZ) GWAS 14 which we previously showed was associated with psychosis during mania in bipolar disorder 11 . To avoid assigning “causal” effects to SNPs in LD, we first clumped the summary statistics using PLINKv1.90 (--clump-kb 250 -- clump-p=1 --clump-r2=0.1) to obtain 93 802 approximately independent SNPs.…”

“…We used the various PRS approaches to test for association of each PRS with the history of psychosis during mania in BD cases. We recently demonstrated that psychosis during mania is associated with polygenic risk of schizophrenia 11 . No large GWAS exists for this phenotype, thus, PRS approaches can be quite useful here to elucidate potential differences in genetic background between bipolar cases with and without psychosis, and the genetic overlap of this phenotype with other psychiatric traits in addition to SZ.…”