Skip to main content
NCBI home page

This is a preprint.

It has not yet been peer reviewed by a journal.

The National Library of Medicine is running a pilot to include preprints that result from research funded by NIH in PMC and PubMed.

medRxiv logoLink to medRxiv
[Preprint]. 2024 Nov 30:2024.11.26.24318011. [Version 1] doi: 10.1101/2024.11.26.24318011

Advancing Gene Discovery for Substance Use Disorders Using Additional Traits Related to Behavioral Disinhibition

Holly E Poore 1, Chris Chatzinakos 2, Travis T Mallard 3,4, Sandra Sanchez-Roige 5,6,7, Fazil Aliev 1, Alexander Hatoum 8; COGA Collaborators, Irwin D Waldman 9, Abraham A Palme 5,7, K Paige Harden 10,11, Peter B Barr 2,†, Danielle M Dick 1,†
PMCID: PMC11623735  PMID: 39649581

Abstract

Importance:

Substance use disorders (SUDs) frequently co-occur with each other and with other traits related to behavioral disinhibition, a spectrum of outcomes referred to as externalizing. Nevertheless, genome-wide association studies (GWAS) typically study individual SUDs separately. This single-disorder approach ignores genetic covariance between SUDs and other traits and may contribute to the relatively limited genetic discoveries to date.

Objective:

To identify the most effective model for capturing genetic relationships between SUDs and externalizing phenotypes, optimizing the detection of genetic influences on SUDs while maintaining specificity.

Design:

We used Genomic SEM to estimate SNP effects on a broad factor representing liability to externalizing and SUDs, on factors representing liability to behavioral disinhibition and SUDs separately, and on residualized SUDs. Subsequent gene-based, tissue expression, and polygenic score (PGS) analyses were used to compare the ability of these alternative approaches to identify genetic influences on SUDs.

Setting:

This study was carried out from May 2023 - September 2024.

Participants:

We used GWAS summary statistics based on samples of European ancestry from previous studies of externalizing and SUD phenotypes in the main multivariate GWAS (N > 2.2 million). We used two independent samples to estimate polygenic associations, a family-based sample enriched for substance use problems (COGA; N = 7,530) and a population-based sample representative of the United States, (All of Us; N = 77,442).

Exposures:

N/A

Main Outcomes and Measures:

Across the three factors (Externalizing; SUDs; Behavioral Disinhibition) and four residualized SUDs (alcohol, tobacco, opioid, and cannabis), we compared the number, putative function, previous associations of significant genomic risk loci and genes, and variance explained by polygenic scores in substance use outcomes.

Results:

We identified genomic risk loci and genes uniquely associated with Externalizing that are relevant to the neurobiology of substance use. Genes identified for residual SUDs were involved in substance-specific processes (e.g., metabolism). The Externalizing PGS accounted for the most variance in substance outcomes relative to the PGS for the other factors and residual PGS appeared to capture substance specific signals.

Conclusions and Relevance:

Our findings suggest that modeling both a broad genetic liability to externalizing behaviors and substance-specific liabilities enhances the detection of genetic effects related to SUDs and explains more variance in substance use outcomes.

SUDs frequently co-occur, and twin studies indicate that this overlap is due in large part to shared genetic influences. This shared genetic influence, which broadly impacts SUD risk, accounts for up to 74-80% of the genetic influences on alcohol use disorders and 62-74% of genetic influences on other substance use disorders1,2, with the remaining genetic risk being substance specific. Despite this, gene-identification efforts most commonly study SUDs individually, potentially hampering our ability to identify genes involved in SUDs. More recently, multivariate genome-wide association studies3,4 (GWAS) have been applied to SUDs, providing further evidence that most genomic risk is shared5.

SUDs also frequently co-occur with other psychiatric disorders and behavioral traits characterized by behavioral disinhibition, such as childhood conduct disorder, adult antisocial behavior, and personality traits related to impulsivity6,7. Here too, twin data suggest that shared genetic influences contribute to this overlap, with these traits and disorders loading together with SUDs on a common underlying shared genetic factor8,9. In twin studies, this underlying latent factor is highly heritable (~80%), moreso than any of the disorders or traits studied individually. In the psychological literature, this spectrum of behaviors and disorders, including SUDs, is typically referred to as Externalizing7,10.

In the current study, we leverage genetic correlations among SUDs and related behavioral disinhibition phenotypes to improve power to detect genetic effects on SUDs11. This builds our previous work12, in which we modeled the relationship between SUDs and behavioral disinhibition traits. The best fitting models were (1) a common factor in which all SUDs and traits related to behavioral disinhibition loaded onto a single factor, and (2) a two-factor model in which the factor representing shared SUD risk correlated .9 with a latent factor underlying other disorders and traits related to behavioral disinhibition. Here, we extend this work by conducting multivariate genomic analyses from data on > 2.2 million individuals of these three factors, representing broad externalizing, behavioral disinhibition, and SUDs.

The value of jointly analyzing genetically correlated traits to improve power is evidenced by previous multivariate genomic investigations, including from our group, in which we analyzed seven externalizing related traits and identified over 550 independent genetic loci that were significantly associated with broad externalizing risk. Nevertheless, use of an overly broad model may result in identification of genetic variants that are related to psychopathology more broadly, rather than specific to SUDs. Thus, in the current study, we carry forward both the common and two-factor models from our previous work and evaluate the ability of each to increase gene-identification for without sacrificing specificity of SUDs. To evaluate power, we compared the magnitude of genetic signal, number of significant genomic risk loci and genes, and variance accounted for by resulting polygenic scores. To evaluate specificity, we explored the functions and previous associations of identified risk loci and genes and differential tissue expression and polygenic prediction of specific phenotypes in two independent samples.

Methods

Results are reported using STREGA reporting guidelines13.

Multivariate GWAS

We used Genomic SEM11 to estimate SNP effects in two models (Figure 1), identified in our recent paper12, which were drawn from previous separate GWAS of externalizing14 and addiction risk3 with a total sample size of 2,219,357 individuals (Supplementary Table 1). In our two-factor model, the SUD factor included problematic alcohol use PAUD)15, problematic tobacco use (PTU)4, opioid use disorder (OUD)16, and cannabis use disorder (CUD)17. The remaining non-SUD indicators from the original externalizing model (attention deficit hyperactivity disorder [ADHD]18, risk taking [RISK]19, number of sexual partners [NSEX]19, age at first sex [FSEX]19, smoking initiation [SMOK]20, and cannabis initiation [CANN]21) loaded onto a behavioral disinhibition factor (boxes B and C in Figure 1). In our common factor model, all behavioral disinhibition phenotypes and SUDs loaded onto a common externalizing factor (Box A in Figure 1). We retained the same GWAS as were used in the respective original multivariate GWAS except in the case of OUD, for which a better powered GWAS was available.

Figure 1.

Figure 1.

Open in a new tab

Path diagrams of models used in the current analyses. Box A represents the broad Externalizing factor onto which all behavioral disinhibition and SUD phenotypes load. Boxes B-E represent residual SUD phenotypes. Boxes F and G represent narrower factors reflecting Behavioral Disinhibition and SUD phenotypes, respectively. Single headed arrows indicate factor loadings whereas the double headed arrow indicates a correlation between the two factors.

We performed multivariate genome-wide association analyses by estimating SNP effects on each of the latent factors (boxes A, F, and G in Figure 1) and the residual SUDs in the single factor model (boxes B-E in Figure 1). These residual GWAS capture genetic influences unique to that disorder after accounting for what it shared with other behavioral disinhibition phenotypes and SUDs. All GWAS included only individuals whose genomes were most similar to those from reference panels sampled from Europe (hereto referred to as “EUR”). Clumping was performed in Functional Mapping and Annotation of Genome-Wide Association Studies (FUMA) version 1.6.122 using an r2 threshold of ≥ 0.6 to define independent significant SNPs, a second threshold of r2 ≥ 0.1 to define lead SNPs, and a maximum distance between LD blocks of 250kb to merge into a locus.

To characterize differences in statistical power among the multivariate GWAS, we examine the mean χ2, λGC, and number of genome-wide significant risk loci for each factor and residual SUD. We evaluated the novelty of our findings by comparing the genomic risk loci identified in our analyses to 1) loci identified in the original externalizing and addiction risk GWAS and 2) loci previously identified for substance use phenotypes in the GWAS literature. This latter test was performed by comparing the genomic risk loci for our factors and correlated SNPs (r2 > 0.1) to those in the NHGRI-EBI GWAS Catalog23 (version e111_r2024-05-05). Finally, we assessed the relative performance of our two models by comparing the degree of heterogeneity of SNP effects. We did so by calculating QSNP heterogeneity statistics, which can be used to identify SNPs that have an effect on one or more indicator phenotypes that is better explained by pathways independent of the factor. If a factor is truly capturing the majority of genetic variance shared among its indicator phenotypes, there will be few QSNP loci relative to the number of factor loci.

Biological annotation

Gene-based methods.

We used three methods to identify genes associated with the three latent genomic factors. First, we used multi-marker analysis of genomic annotation (MAGMA; version 1.08)24, in which genome-wide SNPs were mapped to 18,235 protein-coding genes from Ensembl v102, and SNPs within each gene were jointly tested for association with each factor. We evaluated Bonferroni corrected significance adjusted for the number of genes (one sided p < 2.74x10−6). Next, we used MetaXcan25 to conduct a Transcriptome-Wide Association Study (TWAS) using genetically regulated expression models from GTEx v826. This analysis leveraged GWAS summary statistics to estimate gene-trait associations. Within-tissue Bonferroni correction was applied to identify statistically significant TWAS genes. Finally, we used summary-data-based Mendelian randomization (SMR)27 to 1) test the extent to which gene expression mediated the relationship between SNPs and the phenotype and 2) distinguish causality and pleiotropy models from the linkage model using the heterogeneity in dependent instruments (HEIDI) test. The latter test was used to identify genes that are more likely to be functionally relevant to the phenotype and should therefore be prioritized for follow up. We identified genes of interest as those that met SMR test Bonferroni correction significance threshold and had a HEIDI test p-value > .05.

To account for differences in gene-based mapping methods and ensure higher confidence in the genes associated with each factor, we further evaluated the intersection of genes that were identified in all three gene-based analyses. As our primary goal is to improve gene-identification for SUDs, we explored the specificity of genes associated with each factor as well as evidence that these genes are relevant to the neurobiological pathways implicated in SUDs. To do this, we quantified the number of genes that were unique to each factor (i.e., associated with one factor and not the other two factors) and explored the traits with which these genes have been previously associated. We also used MAGMA to identify genes associated with each residual SUD phenotype.

Tissue expression.

We also used MAGMA tissue expression analysis to test the relationship between expressed genes in brain, liver, and lung tissues and genetic associations across the three factors. We used weights from GTEx v8 for 15 brain, liver, and lung tissue types26. We report standardized beta coefficients of the associations to highlight any differences in direction or magnitude of associations across the three factors.

Polygenic scores

We calculated polygenic scores from GWAS of each latent genomic factor (Externalizing, Behavioral Disinhibition, and SUD) among EUR individuals from the Collaborative Study on the Genetics of Alcoholism (COGA; N =7,530), a multi-site family-based study28 and All of Us (Nmax = 77,442), a national cohort study29,30 (see Supplementary Material for addition sample description). We also calculated PGS for the residual SUD phenotypes (alcohol, opioid, tobacco, and cannabis use disorders) in COGA. These analyses allowed us to compare the total variance explained by each of the PGS and identify the specificity, or loss thereof, in polygenic prediction of substance use phenotypes. In COGA we included phenotypes related to substance initiation, consumption, and use disorders. In All of Us we analyzed diagnoses of alcohol use disorder, tobacco use disorder, other drug disorder (DUD; i.e., use disorder for all other drugs, including opioids or cannabis), major depressive disorder, bipolar disorder, and schizophrenia using phecodes in participants’ linked electronic health records (EHR). We included data from release 7 (May 6, 2018 to February 23, 2023) on those with available EHR and whole genome sequence data.

We used PRS-CS31 to adjust original GWAS beta weights for linkage disequilibrium and Plink232 to construct each PGS from these weights. We evaluated the incremental R2/pseudo-R2 (ΔR2) attained by adding the polygenic score to a regression with baseline covariates (e.g., age, sex, and ancestry PCs). We used least squares regression for continuous outcomes and logistic regression for categorical ones and adjusted the standard errors in COGA to account for the family structure. We estimated 95% CIs for ΔR2 using bootstrapping (1,000 iterations).

Results

Multivariate GWAS

Phenotype Mean χ2 λGC GWS Risk Loci
Externalizing 3.45 2.58 708
Behavioral Disinhibition 3.26 2.49 631
SUD 1.42 1.34 48
Residual AUD 1.21 1.18 14
Residual OUD 1.10 1.09 0
Residual PTU 1.25 1.18 23
Residual CUD 1.03 1.04 1
Open in a new tab

After merging across the ten sets of summary statistics, 5,963,905 SNPs were available for analysis. Table 1 shows the mean χ2, λGC, LD score intercept, and number of genome-wide significant risk loci for each factor and residual phenotype. Relative to the Behavioral Disinhibition and SUD factors, the Externalizing factor GWAS produced higher mean χ2 and λGC values, suggesting that it is better powered to detect genetic influences.

We identified 708, 631, and 48 genomic risk loci for the Externalizing, Behavioral Disinhibition, and SUD factors, respectively (Figure 2; Supplementary Table 2). Of the 708 Externalizing loci and their correlates within LD regions (r2 > .1), 187 (26%) were not identified in the previous Externalizing14 or Addiction Risk3 GWAS and 424 (60%) were not previously associated with a substance use trait. Although we identified the greatest number of risk loci using the broad Externalizing factor, we observed an increased number of hits for Behavioral Disinhibition and SUD relative to their original multivariate GWAS as well. Of the 631 Behavioral Disinhibition loci and their correlates, 94 (15%) were not identified in the previous Externalizing GWAS. Similarly, of the 48 SUD loci and their correlates, 33 (69%) were not identified in the previous Addiction Risk GWAS, and 6 (13%) were not previously associated with any substance use trait in the GWAS Catalog. Finally, we identified 14, 23, and 1 genomic risk loci for the residual AUD, PTU, and CUD phenotypes, respectively, whereas the residual OUD GWAS did not identify any genome-wide significant hits (Supplementary Table 3). Many of these loci were mapped to genes involved in metabolism or receptor activity of specific substances. For example, residual AUD loci were mapped to genes in the alcohol dehydrogenase family (ADH1B, ADH4, ADH5, ADH6), which is involved in metabolism of alcohol and other related substances, and loci for residual PTU were mapped to genes in the nicotinic acetylcholine receptor family of proteins (CHRNA3, CHRNA5, CHRNA6, CHRNB3, CHRNB4).

Figure 2.

Figure 2.

Open in a new tab

Manhattan plots of (top to bottom) Externalizing, Behavioral Disinhibition, and SUD. Brown points represent novel SUD loci (loci not previously associated with a substance use trait). Top loci are mapped to the nearest gene using ANNOVAR37 annotation.

We also performed SNP-level tests of heterogeneity (QSNP) to investigate the extent to which SNPs identified with the factors had consistent, pleiotropic effects on the constituent indicators of that factor. We identified 134 and 112 QSNP for the Externalizing and two-factor Behavioral Disinhibition and SUD model, respectively. Only one QSNP loci from the Externalizing model overlapped with any of the 708 genomic risk loci identified for that factor. This was rs4702, a noncoding transcript variant located in the gene FURIN that has previously associated with OUD33. Similarly, only three significant loci from the Behavioral Disinhibition and SUD model were QSNP loci. These loci were rs4702 and rs1229984, a missense variant in ADH1B associated with alcohol phenotypes34,35, and rs13135092, an intronic variant located in SLC39A8 also previously associated with AUD36. Of the 38 loci identified across all four residual SUDs, 14 overlapped with QSNP loci from the Externalizing and/or SUD factor, highlighting the pattern that QSNP loci represent substance-specific genetic effects.

Biological annotation

Identifying genes for latent genomic factors.

MAGMA identified 1325, 1125, and 69 genes associated with Externalizing, Behavioral Disinhibition, and SUD, respectively. Metaxcan identified 3463, 3078, and 207 genes. SMR and HEIDI tests prioritized 977, 868, and 20 genes for Externalizing, Behavioral Disinhibition, and SUD, respectively. The intersection of genes identified by these two methods resulted in 142, 127, and 5 high confidence genes significantly associated with Externalizing, Behavioral Disinhibition, and SUD, respectively (Figure 3a; Supplementary Table 4). We found 38 unique genes for Externalizing, 23 for Behavioral Disinhibition, and 3 for SUDs (Figure 3b). Of the 38 genes unique Externalizing genes, 34 (89%) have been previously associated with a substance use phenotype (e.g., initiation, consumption, or use disorder),23 including CDH12, which has been associated with nicotine dependence, and KLHL29, which has been associated with cannabis, nicotine, and alcohol use. Of the three unique SUD genes, PPP6C has been associated with CUD, AUD, and OUD. In addition, we identified 44, 26, and 3 genes that were not identified in their respective original GWAS.

Figure 3.

Figure 3.

Open in a new tab

(a) Venn diagrams showing the overlap of genes identified by the MetaXcan, SMR, and MAGMA analyses for Externalizing (blue), Behavioral Disinhibition (orange), and SUD (yellow). (b) upset plot showing intersecting sets of high confidence genes (those identified by all three gene-based methods) across Externalizing, Behavioral Disinhibition, and SUD.

Identifying genes for residual SUDs.

We also used MAGMA to identify genes associated with the four residual SUDs. We found 34, 38, and 1 genes associated with residual AUD, PTU, and OUD, respectively (Supplementary Table 5). We did not identify any significant genes associated with residual CUD. The genes associated with residual SUDs were unique to that phenotype, meaning that there were no overlapping genes identified across residual SUDs. We saw substance-specificity in the genes associated with the residual SUD phenotypes. The majority of genes identified for the residual SUDs were previously associated with a substance use phenotype and many were associated only with a phenotype for that specific substance. For example, of the 34 genes associated with residual AUD, 22 were previously associated with any substance use phenotype and 19 of those were previously associated only with an alcohol phenotype.

Tissue expression.

We found little evidence of differential associations across the three factors with gene expression in brain, liver, and lung tissues (Supplementary Figure 1). The direction of effects was the same across factors, but the magnitude of the associations for SUDs was significantly smaller than those for Behavioral Disinhibition and Externalizing, highlighting the differences in power, but not evidence of differences in magnitude of association.

Polygenic scores

Factor PGS.

Figure 4a and 4b compare the variance accounted for by each of the factor PGS in relevant substance use outcomes in COGA and All of Us (full results in Supplementary Tables 8 and 9). EXTPGS explained the most variance in all phenotypes except alcohol initiation in COGA and AUD diagnosis in All of Us, although the confidence intervals for the R2 estimates overlapped in many instances. BDPGS typically predicted the second most variance, except in the case of AUD symptoms and FTND scores in COGA, and AUD in All of Us. In COGA, EXTPGS explained between 1.6% (CUD) and 3.3% (AUD) of the variance and SUDPGS explained .39% (OUD) and 2.9% (FTND symptoms). In All of Us, EXTPGS explained between 3.1 (AUD) and 7.1% (PTU) of the variance in SUD diagnosis and SUDPGS accounted for 3.1% (AUD) and 4.4% (DUD).

Figure 4.

Figure 4.

Open in a new tab

Variance explained by the factor PGS in (a) COGA and (b) All of Us. Panel (c) shows the variance explained by residual SUD PGS in COGA. We report R2 estimates and 95% confidence intervals.

Residual SUD PGS.

We next compared the variance explained by the residual SUD PGS in substance initiation and use disorders in COGA (Figure 4c). Only the resPTUPGS accounted for a significant portion of the variance in any variable, explaining 3.2% of the variance in problematic tobacco use. Nevertheless, the R2 estimates for each use disorder showed substance specific prediction in use disorders, such that the residual PGS predicted the most variance in the use disorder phenotype corresponding to that substance (e.g., resAUDPGS predicted the most variance in AUD symptoms etc).

Discussion

Despite evidence that SUDs share phenotypic6 and genetic3,8,12 variance with each other and with other traits related to behavioral disinhibition, gene identification efforts typically study individual SUDs in isolation. This misses potentially important genetic variance untagged and may in part explain the slow progress in gene discovery for SUDs relative to other complex behavioral phenotypes38. Here, we drew from the twin literature on the nature of genetic influences on SUDs1,2,8 and capitalized on recent advances in multivariate statistical genetic methods11, to model the shared genetic architecture of SUDs and other externalizing phenotypes with the goal of improving gene discovery for SUDs.

Our results compellingly demonstrate that modeling genetic covariance of SUDs alongside related externalizing traits improves gene discovery for SUDs. In addition, we demonstrate that use of the externalizing factor as a gene-identification target did not reduce our ability to detect SUD-specific effects. Furthermore, our results indicate that use of a broad externalizing factor does not result in loss of specificity for SUD related signal. We base this argument on 1) the comparable number of QSNP loci between the two models, suggesting similar levels of heterogeneity; 2) the identification of substance use related genes in the Externalizing GWAS that were not identified in the GWAS for other factors; 3) lack of evidence for differential tissue expression or PGS associations across the three factors; and, in fact, 4) evidence that the Externalizing factor was best powered to capture SNP associations and that PGS derived from the externalizing factor account for variance in substance related outcomes. Finally, we demonstrate that substance-specific genetic associations (e.g., ADH1B, CHRNA5) are effectively captured by the residual SUD GWAS, indicating that the most insight into genetic influences on SUDs can be obtained by examining both the shared genetic liability to externalizing and the residual genetic variance in each SUD.

The finding that novel insights into the genetic influences on SUDs are best gained by studying broad genetic liability to externalizing in conjunction with residual SUD genetic variance is consistent with previous structure of psychopathology and twin literature. The importance of broad spectra and individual signs and symptoms is emphasized in recent structural models of psychopathology (e.g., HiTOP)10,39. Twin studies further support this, as externalizing factors capture a large proportion of etiological variance in their indicators, including SUDs, but individual disorders often retain a small, but statically significant, proportion of specific etiological variance8.

Limitations & Future Directions

These findings should be interpreted in light of a few limitations. First and foremost, our analyses include only data from participants of European ancestry, thereby limiting the generalizability of our results. This was a practical choice as we relied on previously published GWAS and sufficiently powered GWAS of non-European ancestry samples are not yet available for these outcomes. Our group is working to expand to more diverse genomic populations (https://osf.io/7pfgj/). Second, our analyses are limited to disorder-level phenotypes and may mask symptom specific heterogeneity. Recent phenotypic studies exploring the comorbidity between SUDs and among SUDs and other forms of psychopathology suggest symptom-specific effects, 40-43 which we are unable to model Finally, although SUDs manifest the strongest relationships with other externalizing phenotypes, there is evidence of an internalizing pathway to problematic substance use44,45 which is not modeled in the current study.

Conclusions

Modeling a common genetic liability to externalizing resulted in a greater number of genomic risk loci, identified novel genes, and produced a polygenic score that accounted for more variance in substance use outcomes. Our findings suggest that capitalizing on the shared genetic architecture among externalizing phenotypes can advance gene discovery for SUDs.

Supplementary Material

Supplement 1
media-1.docx (28.2KB, docx)
Supplement 2
media-2.xlsx (139.4KB, xlsx)

Key Points.

Question:

Can we advance gene discovery for substance use disorders (SUDs) by incorporating information about correlated outcomes related to behavioral disinhibition?

Findings:

Combining genetic effects common to SUDs and behavioral disinhibition in a multivariate genome-wide association study of > 2.2 million individuals identified 708 genomic risk loci and accounted for more variance in SUD phenotypes compared with modeling each set of phenotypes separately.

Meaning:

SUDs and behavioral dysregulation are influenced by a common set of common variants; modeling their joint contributions improves power for genetic discovery and polygenic prediction.

Acknowledgements

The authors thank The Externalizing Consortium. Principal Investigators: Danielle M. Dick, Philipp Koellinger, K. Paige Harden, Abraham A. Palmer. Lead Analysts: Richard Karlsson Linnér, Travis T. Mallard, Peter B. Barr, Sandra Sanchez-Roige. Significant Contributors: Irwin D. Waldman. The Externalizing Consortium has been supported by the National Institute on Alcohol Abuse and Alcoholism (R01AA015416 – administrative supplement to DMD), and the National Institute on Drug Abuse (R01DA050721 to DMD). Additional funding for investigator effort has been provided by K02AA018755, U10AA008401, P50AA022537 to DMD, as well as a European Research Council Consolidator Grant (647648 EdGe to Koellinger). The content is solely the responsibility of the authors and does not necessarily represent the official views of the above funding bodies. The Externalizing Consortium would like to thank the following groups for making the research possible: 23andMe, Add Health, Vanderbilt University Medical Center’s BioVU, Collaborative Study on the Genetics of Alcoholism (COGA), the Psychiatric Genomics Consortium’s (PGC) Substance Use Disorders working group, UK10K Consortium, UK Biobank, and Philadelphia Neurodevelopmental Cohort. All code necessary to replicate this study is available upon request.

The authors thank Million Veteran Program (MVP) staff, researchers, and volunteers, who have contributed to MVP, and especially participants who previously served their country in the military and now generously agreed to enroll in the study.

We gratefully acknowledge All of Us participants for their contributions, without whom this research would not have been possible. We also thank the National Institutes of Health’s All of Us Research Program for making available the participant data examined in this study.

Finally, we thank The Collaborative Study on the Genetics of Alcoholism (COGA), Principal Investigators B. Porjesz, V. Hesselbrock, A. Agrawal; Scientific Director, A. Agrawal; Translational Director, D. Dick, includes ten different centers: University of Connecticut (V. Hesselbrock); Indiana University (H.J. Edenberg, T. Foroud, Y. Liu, M.H. Plawecki); University of Iowa Carver College of Medicine (S. Kuperman, J. Kramer); SUNY Downstate Health Sciences University (B. Porjesz, J. Meyers, C. Kamarajan, A. Pandey); Washington University in St. Louis (L. Bierut, A. Agrawal, S. Hartz); University of California at San Diego (M. Schuckit); Rutgers University (J. Tischfield, D. Dick, R. Hart, J. Salvatore); The Children’s Hospital of Philadelphia, University of Pennsylvania (L. Almasy); Icahn School of Medicine at Mount Sinai (A. Goate, P. Slesinger); and Howard University (D. Scott). Other COGA collaborators include: C. Holzhauer, M. Hesselbrock (University of Connecticut); J. Nurnberger Jr., L. Wetherill, X., Xuei, D. Lai, S. O’Connor, (Indiana University); G. Chan (University of Iowa; University of Connecticut); D.B. Chorlian, J. Zhang, P. Barr, S. Kinreich, G. Pandey, Z. Neale (SUNY Downstate); N. Mullins (Icahn School of Medicine at Mount Sinai); A. Anokhin, K. Bucholz, F. Dong, A. Hatoum, E. Johnson, V. McCutcheon, J. Rice, S. Saccone (Washington University); F. Aliev, Z. Pang, S. Kuo, S. Brislin (Rutgers University); A. Merikangas (The Children’s Hospital of Philadelphia and University of Pennsylvania); H. Chin and A. Parsian are the NIAAA Staff Collaborators. We continue to be inspired by our memories of Henri Begleiter and Theodore Reich, founding PI and Co-PI of COGA, and also owe a debt of gratitude to other past organizers of COGA, including Ting- Kai Li, P. Michael Conneally, Raymond Crowe, and Wendy Reich, for their critical contributions. This national collaborative study is supported by NIH Grant U10AA008401 from the National Institute on Alcohol Abuse and Alcoholism (NIAAA) and the National Institute on Drug Abuse (NIDA).

COGA Collaborators: Bernice Porjesz, Victor Hesselbrock, Tatiana M. Foroud, Arpana Agrawal, Howard J. Edenberg, John I. Nurnberger Jr, Yunlong Liu, Samuel Kuperman, John Kramer, Jacquelyn L. Meyer, Chella Kamarajan, Ashwini K. Pandey, Laura Bierut, John Rice, Kathleen K. Bucholz, Marc A. Schuckit, Jay Tischfield, Andrew Brooks, Ronald P. Hart, Laura Almasy, Danielle M. Dick, Jessica E. Salvatore, Allison Goate, Manav Kapoor, Paul Slesinger, Denise M. Scott, Lance Bauer, Leah Wetherill, Xiaoling Xuei, Dongbing Lai, Sean J. O’Connor, Martin H. Plawecki, Spencer Lourens, Laura Acion, Grace Chan, David B. Chorlian, Jian Zhang, Sivan Kinreich, Gayathri Pandey, Michael J. Chao, Andrey P. Anokhin, Vivia V. McCutcheon, Scott Saccone, Fazil Aliev, Peter B. Barr, Hemin Chin & Abbas Parsian

Footnotes

*

A list of authors and their affiliations appears at the end of the paper

References

  • 1.Kendler KS, Jacobson KC, Prescott CA, Neale MC. Specificity of genetic and environmental risk factors for use and abuse/dependence of cannabis, cocaine, hallucinogens, sedatives, stimulants, and opiates in male twins. Am J Psychiatry. Apr 2003;160(4):687–95. doi: 10.1176/appi.ajp.160.4.687 [DOI] [PubMed] [Google Scholar]
  • 2.Kendler KS, Myers J, Prescott CA. Specificity of genetic and environmental risk factors for symptoms of cannabis, cocaine, alcohol, caffeine, and nicotine dependence. Archives of general psychiatry. 2007;64(11):1313–1320. [DOI] [PubMed] [Google Scholar]
  • 3.Hatoum AS, Colbert SMC, Johnson EC, et al. Multivariate genome-wide association meta-analysis of over 1 million subjects identifies loci underlying multiple substance use disorders. Nature Mental Health. 2023;1(3):210–223. doi: 10.1038/s44220-023-00034-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Hatoum AS, Johnson EC, Colbert SMC, et al. The addiction risk factor: A unitary genetic vulnerability characterizes substance use disorders and their associations with common correlates. Neuropsychopharmacol. 2021/November/08 2021;doi: 10.1038/s41386-021-01209-w [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Miller AP, Bogdan R, Agrawal A, Hatoum AS. Generalized genetic liability to substance use disorders. The Journal of Clinical Investigation. June/05/ 2024;134(11)doi: 10.1172/JCI172881 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Krueger Markon, Patrick Benning, Kramer. Linking Antisocial Behavior, Substance Use, and Personality: An Integrative Quantitative Model of the Adult Externalizing Spectrum. J Abnorm Psychol. 2007;116(4):645–666. doi: 10.1037/0021-843X.116.4.645 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Krueger RF, Hobbs KA, Conway CC, et al. Validity and utility of Hierarchical Taxonomy of Psychopathology (HiTOP): II. Externalizing superspectrum. https://doi.org/10.1002/wps.20844. World Psychiatry. 2021/June/01 2021;20(2):171–193. doi:https://doi.org/ 10.1002/wps.20844 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Krueger Hicks, Patrick Carlson, Iacono McGue. Etiological Connections among Substance Dependence, Antisocial Behavior, and Personality: Modeling the Externalizing Spectrum. J Abnorm Psychol. 2002;111(3):411–424. doi: 10.1037/0021-843X.111.3.411 [DOI] [PubMed] [Google Scholar]
  • 9.Kendler KS, Lönn SL, Maes HH, Lichtenstein P, Sundquist J, Sundquist K. A Swedish Population-Based Multivariate Twin Study of Externalizing Disorders. Behavior Genetics. 2016/March/01 2016;46(2):183–192. doi: 10.1007/s10519-015-9741-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Kotov R, Krueger RF, Watson D, et al. The Hierarchical Taxonomy of Psychopathology (HiTOP): A dimensional alternative to traditional nosologies. J Abnorm Psychol. May 2017;126(4):454–477. doi: 10.1037/abn0000258 [DOI] [PubMed] [Google Scholar]
  • 11.Grotzinger AD, Rhemtulla M, de Vlaming R, et al. Genomic structural equation modelling provides insights into the multivariate genetic architecture of complex traits. Nat Hum Behav. May 2019;3(5):513–525. doi: 10.1038/s41562-019-0566-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Poore HE, Hatoum A, Mallard TT, et al. A Multivariate Approach to Understanding the Genetic Overlap between Externalizing Phenotypes and Substance Use Disorders. Addiction Biology. 2023;28(9):e13319. doi: 10.1111/adb.13319 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Little J, Higgins JP, Ioannidis JP, et al. STrengthening the REporting of Genetic Association Studies (STREGA)--an extension of the STROBE statement. Genet Epidemiol. Nov 2009;33(7):581–98. doi: 10.1002/gepi.20410 [DOI] [PubMed] [Google Scholar]
  • 14.Karlsson Linnér R, Mallard TT, Barr PB, et al. Multivariate analysis of 1.5 million people identifies genetic associations with traits related to self-regulation and addiction. Nat Neurosci. 2021/October/01 2021;24(10):1367–1376. doi: 10.1038/s41593-021-00908-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Zhou H, Kember RL, Deak JD, et al. Multi-ancestry study of the genetics of problematic alcohol use in over 1 million individuals. Nature Medicine. 2023/December/01 2023;29(12):3184–3192. doi: 10.1038/s41591-023-02653-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Deak JD, Zhou H, Galimberti M, et al. Genome-wide association study in individuals of European and African ancestry and multi-trait analysis of opioid use disorder identifies 19 independent genome-wide significant risk loci. Molecular Psychiatry. 2022/October/01 2022;27(10):3970–3979. doi: 10.1038/s41380-022-01709-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Johnson EC, Demontis D, Thorgeirsson TE, et al. A large-scale genome-wide association study meta-analysis of cannabis use disorder. Lancet Psychiat. Dec 2020;7(12):1032–1045. doi: 10.1016/s2215-0366(20)30339-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Demontis D, Walters RK, Martin J, et al. Discovery of the first genome-wide significant risk loci for attention deficit/hyperactivity disorder. Nat Genet. Jan 2019;51(1):63–75. doi: 10.1038/s41588-018-0269-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Linnér R, Karlsson, Biroli P, Kong E, et al. Genome-wide association analyses of risk tolerance and risky behaviors in over 1 million individuals identify hundreds of loci and shared genetic influences. Nature Genetics. 2019/02/January 2019;51(2):245–257. doi: 10.1038/s41588-018-0309-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Liu M, Jiang Y, Wedow R, et al. Association studies of up to 1.2 million individuals yield new insights into the genetic etiology of tobacco and alcohol use. Nat Genet. 2019/February/01 2019;51(2):237–244. doi: 10.1038/s41588-018-0307-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Pasman JA, Verweij KJH, Gerring Z, et al. GWAS of lifetime cannabis use reveals new risk loci, genetic overlap with psychiatric traits, and a causal influence of schizophrenia. Nat Neurosci. Sep 2018;21(9):1161–1170. doi: 10.1038/s41593-018-0206-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Watanabe K, Taskesen E, van Bochoven A, Posthuma D. Functional mapping and annotation of genetic associations with FUMA. Nature Communications. 2017/November/28 2017;8(1):1826. doi: 10.1038/s41467-017-01261-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Sollis E, Mosaku A, Abid A, et al. The NHGRI-EBI GWAS Catalog: knowledgebase and deposition resource. Nucleic Acids Research. 2023;51(D1):D977–D985. doi: 10.1093/nar/gkac1010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.de Leeuw CA, Mooij JM, Heskes T, Posthuma D. MAGMA: Generalized Gene-Set Analysis of GWAS Data. PLOS Computational Biology. 2015;11(4):e1004219. doi: 10.1371/journal.pcbi.1004219 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Barbeira AN, Dickinson SP, Bonazzola R, et al. Exploring the phenotypic consequences of tissue specific gene expression variation inferred from GWAS summary statistics. Nature Communications. 2018/May/08 2018;9(1):1825. doi: 10.1038/s41467-018-03621-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Consortium G. The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science. Sep 11 2020;369(6509):1318–1330. doi: 10.1126/science.aaz1776 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Zhu Z, Zhang F, Hu H, et al. Integration of summary data from GWAS and eQTL studies predicts complex trait gene targets. Nature genetics. 2016;48(5):481–487. [DOI] [PubMed] [Google Scholar]
  • 28.Dick DM, Balcke E, McCutcheon V, et al. The collaborative study on the genetics of alcoholism: Sample and clinical data. Genes, Brain and Behavior. 2023;22(5):e12860. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Investigators AoURPG. Genomic data in the All of Us Research Program. Nature. Mar 2024;627(8003):340–346. doi: 10.1038/s41586-023-06957-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Investigators TAoURP. The “All of Us” Research Program. New England Journal of Medicine. 2019/August/15 2019;381(7):668–676. doi: 10.1056/NEJMsr1809937 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Ge T, Chen CY, Ni Y, Feng YA, Smoller JW. Polygenic prediction via Bayesian regression and continuous shrinkage priors. Nature Communications. Apr 16 2019;10(1):1776. doi: 10.1038/s41467-019-09718-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Purcell S, Neale B, Todd-Brown K, et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. The American journal of human genetics. 2007;81(3):559–575. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Kember RL, Vickers-Smith R, Xu H, et al. Cross-ancestry meta-analysis of opioid use disorder uncovers novel loci with predominant effects in brain regions associated with addiction. Nat Neurosci. Oct 2022;25(10):1279–1287. doi: 10.1038/s41593-022-01160-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.Gelernter J, Sun N, Polimanti R, et al. Genome-wide Association Study of Maximum Habitual Alcohol Intake in >140,000 U.S. European and African American Veterans Yields Novel Risk Loci. Biol Psychiatry. Sep 1 2019;86(5):365–376. doi: 10.1016/j.biopsych.2019.03.984 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Kember RL, Vickers-Smith R, Zhou H, et al. Genetic Underpinnings of the Transition From Alcohol Consumption to Alcohol Use Disorder: Shared and Unique Genetic Architectures in a Cross-Ancestry Sample. Am J Psychiatry. Aug 1 2023;180(8):584–593. doi: 10.1176/appi.ajp.21090892 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Zhou H, Sealock JM, Sanchez-Roige S, et al. Genome-wide meta-analysis of problematic alcohol use in 435,563 individuals yields insights into biology and relationships with other traits. Nat Neurosci. Jul 2020;23(7):809–818. doi: 10.1038/s41593-020-0643-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res. Sep 2010;38(16):e164. doi: 10.1093/nar/gkq603 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Deak JD, Johnson EC. Genetics of substance use disorders: a review. Psychol Med. 2021;51(13):2189–2200. doi: 10.1017/S0033291721000969 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Kotov R, Krueger RF, Watson D, et al. The Hierarchical Taxonomy of Psychopathology (HiTOP): A Quantitative Nosology Based on Consensus of Evidence. Annual Review of Clinical Psychology. 2021;17(Volume 17, 2021):83–108. doi: 10.1146/annurev-clinpsy-081219-093304 [DOI] [PubMed] [Google Scholar]
  • 40.Forbes MK, Watts AL, Twose M, et al. A Hierarchical Model of the Symptom-Level Structure of Psychopathology in Youth. Clinical Psychological Science. 2024:21677026241257852. doi: 10.1177/21677026241257852 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Savage JE, Barr PB, Phung T, et al. Genetic Heterogeneity Across Dimensions of Alcohol Use Behaviors. American Journal of Psychiatry. 2024:appi.ajp.20231055. doi: 10.1176/appi.ajp.20231055 [DOI] [PubMed] [Google Scholar]
  • 42.Mallard TT, Savage JE, Johnson EC, et al. Item-Level Genome-Wide Association Study of the Alcohol Use Disorders Identification Test in Three Population-Based Cohorts. Am J Psychiatry. Jan 2022;179(1):58–70. doi: 10.1176/appi.ajp.2020.20091390 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 43.Watts AL, Sher KJ, Heath AC, Steinley D, Brusco M. “General Addiction Liability” Revisited. Clinical Psychological Science. 2024:21677026241245070. doi: 10.1177/21677026241245070 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.King SM, Iacono WG, McGue M. Childhood externalizing and internalizing psychopathology in the prediction of early substance use. Addiction. 2004;99(12):1548–1559. [DOI] [PubMed] [Google Scholar]
  • 45.Hussong AM, Jones DJ, Stein GL, Baucom DH, Boeding S. An internalizing pathway to alcohol use and disorder. Psychol Addict Behav. Sep 2011;25(3):390–404. doi: 10.1037/a0024519 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplement 1
media-1.docx (28.2KB, docx)
Supplement 2
media-2.xlsx (139.4KB, xlsx)

Articles from medRxiv are provided here courtesy of Cold Spring Harbor Laboratory Preprints

RESOURCES