Table 1.

Summary of epidemiology, genetic epidemiology, and molecular genetic findings for substance use disorders

	Alcohol use disorder (AUD)	Nicotine use disorder (NicUD)	Cannabis use disorder (CanUD)	Opioid use disorder (OUD)	Cocaine use disorder (CocUD)a
SUD epidemiology and recent developments	◦ Lifetime prevalence rate = 29.1%41 ◦ Alcohol use and intoxication contributes to three million worldwide deaths annually122	◦ Lifetime prevalence rate = 27.9%41 ◦ Nicotine use and related-disease contributes to seven million worldwide deaths annually123	◦ Lifetime prevalence rate = 6.3%41 ◦ Recent legalization in Western countries is correlated with increased use, including among pregnant women; it remains to be seen whether this influences prevalence rates of CUD15	◦ Lifetime prevalence rate = 2.1%41 ◦ Despite lower prevalence relative to other SUDs, OUD poses large disease burden due to overdose deaths53	◦ Lifetime prevalence rate = 2.4%41 ◦ From 2012 to 2018, the rate of overdose deaths related to cocaine use increased from 1.4% to 4.5%53
SUD genetic epidemiology	◦ AUD heritability (h2) = 0.50–0.6451,61,117 ◦ SNP-heritability (h2SNP) = ~0.07–0.10128	◦ NicUD heritability (h2) = ~0.30–0.704,111 ◦ SNP-heritability (h2SNP) = ~0.0998	◦ CanUD heritability (h2) = 0.40–0.80118 ◦ SNP-heritability (h2SNP) = ~0.07–0.1261	◦ OUD heritability (h2) = ~0.508,67,114 ◦ SNP-heritability (h2SNP) = ~0.11127	◦ CocUD heritability (h2) = 0.40–0.8069 ◦ SNP-heritability (h2SNP) = ~0.27–0.3014,55
Notable GWAS risk-loci to date BOLD = loci have achieved p ≤ 5.0x10–9 with respective SUD in at least one study	◦ ADH1Bb,c (rs1229984) 17,32,36,78,84,105,121,124,128 ◦ ALDH2b,c (rs671)37,62,81,99 ◦ DRD2b (e.g. rs4936277)78,105,128 ◦ KLBb,c (e.g. rs13129401) 78,105,128 ◦ GCKRb,c (rs1260326)17,78,84,105,128 ◦ SLC39A8b,c (rs13107325)78,84,105	◦ CHRNA5b,c (rs16969968)31,45,46,84,98 ◦ CHRNA5-A3-B4b,c (multiple loci)31,45,46,84,98 ◦ CHRNA4b,c (rs151176846)45,46,84,98 ◦ DNMT3Bb (rs910083)46 ◦ DBHb,c (rs13284520)84,98	◦ FOXP2b (rs7783012)61 ◦ CHRNA2b (rs4732724)23,61 ◦ EPHX2b (rs4732724)61 ◦ CSMD1b (rs77378271)107 ◦ PDE4Bb (gene-wise)61	◦ OPRM1b (rs1799971)127 ◦ CNIH3b (rs10799590)92 ◦ KCNG2b (rs62103177)33 ◦ RGMAb (rs12442183)16 ◦ BEND4c (gene-wise)96	◦ FAM53Bb (rs2629540)34 ◦ HIST1H2BDb (gene-wise)14 ◦ C1QL2b (gene-wise)55 ◦ STK38b (gene-wise)55 ◦ KCTD20b (gene-wise)55
Notable genetic correlations (rg) with psychiatric/substance use traits	◦ Drinks per week (rg = +0.77)128 ◦ Ever smoked regularly (rg = +0.55)128 ◦ Lifetime cannabis use (rg = +0.39) 128 ◦ Major depression (rg = +0.39)128 ◦ Risk-taking (rg = +0.30)128	◦ Alcohol dependence (rg = +0.56)98 ◦ Cigarettes per day (rg = +0.95)98 ◦ Major depression (rg = +0.38)98 ◦ Schizophrenia (rg = +0.16)98 ◦ Smoking initiation (rg = +0.40)98	◦ Alcohol use disorder (rg = +0.55)61 ◦ Educational attainment (rg = −0.39)61 ◦ Lifetime cannabis use (rg = +0.50)61 ◦ Schizophrenia (rg = +0.31)61 ◦ Smoking initiation (rg = +0.66)61	◦ ADHD (rg = +0.36)127 ◦ Alcohol dependence (rg = +0.73)127 ◦ Drinks per week (rg = +0.38)127 ◦ Ever smoked regularly (rg = +0.51)127 ◦ Major depression (rg = +0.35)127	◦ ADHD (rg = +0.50)14 ◦ Ever smoked regularly (rg = +0.34)14 ◦ Major depression (rg = +0.40)14 ◦ Risk-taking (rg = +0.35)14 ◦ Schizophrenia (rg = +0.20)14
Notable CNV and exome/genome sequencing efforts	◦ Genome-wide meta-analysis of CNV associations in AUD cases112: ◦ identified nine CNV regions suggestively associated with AUD (e.g. 5q21.3 deletion)	◦ Exome-chip meta-analysis fine mapped rare coding variants for nicotine use outcomes10: ◦ identified 124 significant associations. ◦ 1.0–2.2% of phenotypic variance explained by rare variation	◦ Low-coverage WGS found two gene regions significantly associated with CanUD39: ◦ C1orf110 gene (protein-coding region) ◦ MEF2B gene (regulatory region)	◦ Three significantly associated CNVs82: ◦ a 18q12.3 deletion ◦ a Xq28 deletion ◦ a 18q12.3 deletion	◦ Several targeted sequencing44 and CNV studies (e.g. NSF gene)13 have been conducted, although no strong evidence has emerged. CocUD whole-genome, whole-exome, and CNV studies are needed
Notable efforts incorporating non-European populations	◦ Kranzler et al. (2019)78: AUD cases: EUR = 34 658; AFR = 17 267; LAT = 3449; EAA = 164; SAA = 44 ◦ Walters et al. (2018)121: AUD cases: EUR = 11 569; AFR = 3335	◦ Quach et al. (2020)98: N = 46 213 EUR smokers; N = 11 787 AFR smokers ◦ Hancock et al. (2018a)46: N = 28 677 EUR smokers; N = 9925 AFR smokers	◦ Johnson et al. (2020b)61: CanUD cases: EUR = 17 068; AFR = 3848 ◦ Sherva et al. (2016)107: CanUD cases: EUR = 2884; AFR = 1572	◦ Zhou et al. (2020a, b)127: OUD: EUR cases = 8259, EUR opioid-exposed controls  = 71 200; AFR cases = 4032, AFR opioid-exposed controls = 26 029	◦ Huggett and Stallings (2020a, b)55: CocUD: EUR cases = 3370; AFR cases = 2349 ◦ Gelernter et al. (2014a)34: provided CocUD case–control data for Huggett and Stallings (2020a, b) analyses

Abbreviations for notable samples incorporating non-European populations: European Ancestry (EUR), African Ancestry (AFR), Latino or Hispanic Ancestry (LAT); East Asian American (EAA); South Asian American (SAA).

Note: numeric superscripts correspond to numbered in-text citations (See Supplementary Material).

^aAt the time of review, CocUD sample sizes remain substantially smaller than other SUDs; thus, current CocUD findings and downstream analyses (e.g. h²SNP, r_g) should be interpreted with caution and require replication in well-powered samples. Efforts to extend CocUD sample sizes are underway.

^bDenotes findings with SUD or problematic use.

^cDenotes findings with substance consumption measure.