28. THE INTERNATIONAL CANNABIS CONSORTIUM: NEW FINDINGS FROM THE LARGEST GWASS ON EVER USE AND FREQUENCY OF USE

Cannabis is one of the most widely used psychoactive substances worldwide, and genetic factors play a significant role in individual differences in cannabis use. Twin studies report moderate heritability for both lifetime cannabis use and frequency of use, and somewhat higher estimates for cannabis use disorder. To date, large-scale genome-wide association studies (GWASs) on cannabis use have focused primarily on ever use or cannabis use disorder, with most studies limited to individuals of European ancestry. Here, we present the first well-powered GWAS of cannabis use frequency (N=269,160 cannabis users), along with an updated GWAS meta-analysis of lifetime cannabis use, more than doubling the sample size of previous efforts and adding the first trans-ancestry findings (N=467,889 European ancestry; N=122,224 non-European ancestry). GWASs were conducted on various cannabis use phenotypes—ever use, frequency of use (lifetime maximum and recent use)—using data from > 24 cohorts. Meta-analyses were performed with METAL and GENESIS, allowing for both ancestry-specific and cross-ancestry analyses. Using a set of functional annotation methods, we identified likely causal genes and implicated brain cell types. Genomic structural equation modeling (Genomic SEM) and Genetic Network Analysis (GNA) were used to examine shared liability across different cannabis use phenotypes. We estimated SNP-based heritability and calculated genetic correlations with a broad range of phenotypes, including cognitive function, socioeconomic status, psychiatric disorders, personality traits, and other substance use, using LD Score Regression. Out-of-sample polygenic scores were generated in the Swedish Twin Registry (N = 32,033) to examine associations with a range of outcome measures. Mendelian Randomization was applied to test potential causal links with key mental health traits. Preliminary results identified 3 independent genome-wide significant SNPs for cannabis use frequency (SNP-h² = 2.3%) and 32 loci for ever use in European ancestry samples (SNP-h² = 7.0%). Additional findings for non-European ancestry groups will be presented. CADM2 emerged as a high-confidence gene across multiple analytic methods. The different cannabis use traits clustered together, with frequency showing a stronger overlap with cannabis use disorder than ever use. Significant genetic correlations were observed between cannabis use (ever use and frequency) and a range of psychiatric, cognitive, personality, and risk-related traits, with some interesting opposing associations for ever use and frequency. MR analyses suggest putatively causal associations between cannabis use and mental health outcomes, with both ever-use and frequency showing causal effects across seven psychiatric conditions, and frequency of use demonstrating robust associations with ADHD, anxiety, and schizophrenia. This study provides new insights into the genetic and biological underpinnings of cannabis use, particularly cannabis use frequency, and expanding findings to hitherto underrepresented ancestry groups. Our findings highlight the value of large-scale genetic studies in uncovering the architecture of substance use and its relationships with broader psycho-behavioral traits, with implications for prevention and intervention strategies.