The International Cannabis Consortium: What Did We Learn About The Genetics Of Cannabis Use

Background Cannabis is the most frequently used and abused illicit drug worldwide and cannabis (ab)use is associated with social, physical, and psychological problems. Twin and family studies have shown that cannabis use and abuse are heritable traits. The International Cannabis Consortium was initiated with the aim of identifying genetic risk variants for cannabis use phenotypes by meta-analysing data from many contributing cohorts. The (partly preliminary) results of the International Cannabis Consortium will be presented: genome-wide association (GWA) meta-analyses on lifetime cannabis use and age at initiation of use. Additionally, findings from several follow-up studies will be presented, including the genetic association of cannabis use with use of other substances, schizophrenia, and conduct disorder. Methods GWA analyses of lifetime cannabis use were performed by each contributing group independently (13 groups, total N=32,330) and were subsequently meta-analysed. We tested for replication, and the SNP results were used to perform a gene-based test of association. We also estimated the total SNP-based heritability and the genetic correlation between lifetime cannabis use and cigarette use based on LD-score regression analysis. Secondly, we meta-analysed GWA results of age at initiation of cannabis use from 8 groups (N=24,222) using a survival analysis. Again SNP results are followed up by a gene-based test of association and an estimate of SNP-based heritability. In follow-up projects, LD-score regression analyses were used to determine the genetic correlation of cannabis use with nicotine, alcohol, and caffeine use, as well as schizophrenia and conduct disorder. We also created polygenic risk scores for cannabis use in an independent target sample and determined to what extent these polygenic scores predicted conduct symptoms. Results Although none of the SNPs were significantly associated with lifetime cannabis use, the gene-based analysis identified 4 significantly associated genes, including NCAM1, CADM2, SCOC and KCNT2. Interestingly, NCAM1 was previously reported to be associated with nicotine and other substance use. All SNPs combined explained 20% of the liability of lifetime cannabis use. For age at initiation of cannabis use, we identified five SNPs (in high linkage-disequilibrium) that were genome-wide significant. Results of the gene-based test and SNP-based heritability are not available yet. Follow-up studies show a significant genetic correlation of cannabis use with smoking initiation, alcohol use per week, as well as with schizophrenia. Furthermore, polygenic risk scores for cannabis use were significantly associated with symptoms of conduct disorder. Discussion The findings of the two largest meta-analyses of GWA studies of cannabis use phenotypes are presented. Several interesting genetic loci were identified, revealing important new candidate genes for cannabis use. Future functional studies should explore the impact of the identified genes on the biological mechanisms of cannabis use. We also show that genes underlying cannabis use are in part overlapping with genes underlying use of other substances and mental health phenotypes, including nicotine and alcohol use, schizophrenia, and symptoms of conduct disorder.