Pleiotropic genetic architecture linking schizophrenia and substance use disorders

Schizophrenia (SCZ) frequently co-occurs with substance use disorders (SUDs), yet the genetic basis of this comorbidity remains unclear. Using the latest European-ancestry genome-wide association studies (GWAS) for SCZ, cannabis use disorder (CanUD), opioid use disorder (OUD), problematic alcohol use (PAU), tobacco use disorder (TUD), and a general addiction factor (AF), together with two SCZ and one SUD case-control samples with individual-level genotype data, we applied multiple complementary genomic approaches to characterize their shared genetic architecture. Significant positive genome-wide genetic correlations were observed across all SCZ–SUD pairs. Local genetic correlation analyses identified multiple genomic regions contributing to this shared architecture, with both positive and negative correlations, and evidence of genomic regions shared across multiple SCZ–SUD pairs. Polygenic overlap analyses indicated substantial sharing (25-50%) of trait-associated variants between SCZ and SUDs. Genomic structural equation modelling supported a common latent factor underlying SCZ and all SUDs, accounting for approximately 23% of SCZ variance. Cross-trait polygenic risk score (PRS) analyses showed bidirectional associations between SCZ and SUD genetic liability. Mendelian randomization analyses provided evidence for a bidirectional causal relationship between SCZ and CanUD. Horizontal pleiotropy analyses identified numerous loci with concordant and discordant effects across traits, including loci shared among multiple SCZ–SUD pairs. Gene mapping and enrichment analyses indicated pathways related to neuroplasticity, synaptic transmission, immune system, metabolism and proteolysis, including both shared and SCZ–SUD-specific biological processes. Overall, these findings suggest that part of SCZ liability reflects genetic susceptibility to SUDs with potential implications for patient stratification and clinical management.