Pairwise genetic meta-analyses between schizophrenia and substance dependence phenotypes reveals novel association signals

Abstract

Almost half of individuals diagnosed with schizophrenia also present with a substance use disorder, however, little is known about potential molecular mechanisms underlying this comorbidity. We used genetic analyses to enhance our understanding of the molecular overlap between these conditions. Our analyses revealed a positive genetic correlation between schizophrenia and the following dependence phenotypes: alcohol (rg = 0.3685, SE = 0.0768, P = 1.61 x 10-06), cannabis use disorder (rg = 0.309, SE = 0.0332, P = 1.19 x 10-20) and nicotine dependence (rg = 0.1177, SE = 0.0436, P = 7.0 x 10-03), as well as lifetime cannabis use (rg = 0.234, SE = 0.0298, P = 3.73 x 10-15) and drinks per week (rg = 0.0688, SE = 0.0217, P = 1.5 x 10-03). We further constructed latent causal variable (LCV) models to test for partial genetic causality and found evidence for a potential causal relationship between alcohol dependence and schizophrenia (GCP = 0.6, SE = 0.22, P = 1.6 x 10-03). This putative causal effect with schizophrenia was not seen using a continuous phenotype of drinks consumed per week, suggesting that distinct molecular mechanisms underlying dependence are involved in the relationship between alcohol and schizophrenia. To localise the specific genetic overlap between schizophrenia and substance use disorders, we conducted a gene-based and gene-set pairwise meta-analysis between schizophrenia and each of the four individual substance dependence phenotypes in up to 790,806 individuals. These bivariate meta-analyses identified 44 associations not observed in the individual GWAS, including five shared genes that play a key role in early central nervous system development. These genes may play an important role in substance dependence in schizophrenia, and, as a result, could represent important targets for future treatment or early intervention, as comorbid substance dependence is associated with poor treatment adherence, greater chronicity and increased mortality.