Genetic contribution to the comorbidity between attention-deficit/hyperactivity disorder and substance use disorders

Abstract

We characterized the genetic architecture of the attention-deficit hyperactivity disorder-substance use disorder(ADHD-SUD) relationship by investigating genetic correlation, causality, pleiotropy, and common polygenic risk.Summary statistics from genome-wide association studies (GWAS) were used to investigate ADHD (Neff =51,568), cannabis use disorder (CanUD, Neff = 161,053), opioid use disorder (OUD, Neff = 57,120), problematicalcohol use (PAU, Neff = 502,272), and problematic tobacco use (PTU, Neff = 97,836). ADHD, CanUD, and OUDGWAS meta-analyses included cohorts with case definitions based on different diagnostic criteria. PAU GWAScombined information related to alcohol use disorder, alcohol dependence, and the items related to alcohol problematic consequences assessed by the alcohol use disorders identification test. PTU GWAS was generated amulti-trait analysis including information regarding Fagerstr¨om Test for Nicotine Dependence and cigarettes perday. Linkage disequilibrium score regression analyses indicated positive genetic correlation with CanUD, OUD,PAU, and PTU. Genomic structural equation modeling showed that these genetic correlations were related to twolatent factors: one including ADHD, CanUD, and PTU and the other with OUD and PAU. The evidence of a causaleffect of PAU and PTU on ADHD was stronger than the reverse in the two-sample Mendelian randomizationanalysis. Conversely, similar strength of evidence was found between ADHD and CanUD. CADM2 rs62250713was a pleiotropic SNP between ADHD and all SUDs. We found seven, one, and twenty-eight pleiotropic variantsbetween ADHD and CanUD, PAU, and PTU, respectively. Finally, OUD, CanUD, and PAU PRS were associatedwith increased odds of ADHD. Our findings demonstrated the contribution of multiple pleiotropic mechanisms tothe comorbidity between ADHD and SUDs.