LAUSR

Background Cognitive impairment is a frequent and serious problem in patients with various forms of severe mental illnesses, including Schizophrenia (SZ) and Bipolar Disorder (BP). Cognitive performance in patients with bipolar disorder or schizophrenia has been found to be very similar in underlying structure, and recent research has suggested genomic linkage to several different cognitive phenotypes. In addition, recent evidence from the Cognitive Genomics Consortium (COGENT) study, focusing on a pooled population of more than 35,000 healthy European ancestry patients from 24 studies, reported an estimated SNP heritability of 21.5%. The current study examined the genomic correlates of cognitive and functional skills performance in 8,051 U.S. veterans with either SZ or BP. Methods Veterans Affairs (VA) Cooperative Studies Program (CSP) Study #572 evaluated cognitive and Functional Capacity (FC) measures among BPI and SZ patients, using a subset of tests from the MATRICS Consensus cognitive battery (MCCB) and the UCSD Performance based skills assessment brief version (UPSAB). In conjunction with the VA Million Veteran Program, and using a customized Affymetrix Axiom genotyping array for samples from 3,388 SZ and 4,663 BP CSP #572 patients (N=8,051 total), we conducted genome-wide association analyses for European Americans (3,952) and African Americans (2,598) separately. Results There are no SNPs with genome-wide statistically significant results for either EA or AA samples. We also applied MetaXcan to calculate gene-level test statistics across 44 tissues and did not identify any associated genes. However, based on the comparison between summary statistics from this study and those from published GWAS of 55 diseases/traits, there is statistically significant evidence of genetic correlation between cognitive function and schizophrenia (1.7E09), but not bipolar illness. Discussion No SNPs with strong evidence of association with cognitive performance were identified in this study, but genetic correlation analysis between this study and published GWAS on schizophrenia from other studies suggests shared factors between cognitive functions and schizophrenia. Later analyses will perform targeted examination of the linkage of previously identified genes and SNPs to overall performance and specific cognitive phenotypes.