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It has not yet been peer reviewed by a journal.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]. 2025 Aug 24:2025.08.19.25332327. [Version 1] doi: 10.1101/2025.08.19.25332327

A Multivariate Genomic Investigation of the Externalizing Spectrum and Suicide Risk

Peter B Barr 1,2,3,4, Mallory Stephenson 5, Holly E Poore 6,7, Chris Chatzinakos 1,2,3, Xuejun Qin 8,9,10, Allison E Ashley-Koch 8,9,10, Elizabeth R Hauser 8,9,10, Jacquelyn L Meyers 1,2,3, Roseann E Peterson 1,2,3, Sandra Sanchez-Roige 11,12,13, Travis T Mallard 14,15, Danielle M Dick 6,7, K Paige Harden 16,17; COGA Collaborators*, Mihaela Aslan 18,19, Philip D Harvey 20,21, Jean C Beckham 8,9,10, Tim B Bigdeli 1,2,3,†, Nathan A Kimbrel 8,9,10,†, on behalf of the VA Million Veteran Program (MVP) and Cooperative Studies Program (CSP) #572
PMCID: PMC12393625  PMID: 40894127

Abstract

Suicidal thoughts and behaviors originate from heterogeneous mechanisms, including behavioral disinhibition characteristic of “externalizing” disorders (e.g., substance use disorders, antisocial personality disorder, etc.). Prior work has demonstrated strong genetic overlap between externalizing and suicide attempts. In the current analysis, we investigate the co-occurrence between a broader array of suicide phenotypes (i.e., suicide deaths, non-fatal attempts, suicidal ideation) and the externalizing spectrum using data from the Million Veteran Program (MVP) Cohort. We leverage the large-scale MVP database to (1) estimate a latent genomic factor for externalizing comprised of MVP data (MVP-EXT) using genomic structural equation modeling (GenomicSEM), (2) validate these results against prior externalizing models and other traits, (3) examine the genetic overlap between externalizing and suicide outcomes using multiple approaches (e.g., genetic correlations, polygenic scores, and post mortem brain tissue of suicide deaths), and (4) explore whether phenotypic externalizing is prospectively associated with death by suicide. We identify 155 loci in our meta-analysis of European-like (EUR-like, N = 310,498) and African-like (AFR-like, N = 99,949) MVP participants. MVP-EXT showed a strong genetic correlation with a prior, non-MVP externalizing factor (rG = 0.87, 95% CI = 0.83, 0.91) and suicide attempt in both EUR-like (rG = 0.67, 95% CI = 0.60, 0.74) and AFR-like (rG = 0.62, 95% CI = 0.42, 0.81) veterans. MVP-EXT polygenic scores were associated with suicidal ideation (OR = 1.09, 95% CI = 1.05, 1.13) and suicide attempts (OR = 1.20, 95% CI = 1.13, 1.27) in independent cohorts. MVP-EXT associated genes showed significant enrichment particularly within inhibitory neurons in suicide deaths compared to deaths from other causes. A phenotypic score for externalizing was prospectively associated with death by suicide in MVP (HR = 1.39, 95% CI = 1.33, 1.45). In total, our results reiterate that, while the relation between suicide with internalizing disorders has generally received more attention, externalizing is an important risk factor for suicide related behaviors. Greater attention should be paid to these problems as potential antecedents of suicide-related behaviors.

INTRODUCTION

Suicide remains a leading preventable cause of death, Suicide rates in the United States increased ~37% between 2000 and 2022. Deaths by suicide have contributed significantly to the decline in U.S. life expectancy, alongside other “deaths of despair”1,2. In addition to suicide deaths, the number of U.S. adults reporting a lifetime suicide attempt has increased in recent birth cohorts3. Suicide attempts are more prevalent in those with co-occurring psychiatric problems. For example, among those with an alcohol use disorder (AUD), the prevalence of lifetime suicide attempt is more than triple the prevalence of the general population (17.5%)4 and 40% of those seeking treatment for AUD report at least one suicide attempt5–8. Additionally, meta-analyses of attention-deficit/hyperactivity disorders (ADHD) and suicide related outcomes reveal strong associations with attempt, plans, and death by suicide9. While mood-related problems are important risk factors for suicide present in these comorbid psychiatric problems10, other mechanisms related to behavioral disinhibition and impulsivity are also important for suicide risk.

Disorders related to behavioral disinhibition (e.g., substance use disorders, ADHD, etc.), generally referred to as externalizing disorders11,12, share a common etiology13. Twin and family studies suggest that the liability for externalizing is highly heritable14,15. Recent multivariate genome wide association studies (GWAS) have found robust evidence for shared genomic factors for externalizing16–18. This genetic liability to externalizing also is genetically correlated with suicide attempt16,19, and externalizing polygenic scores (PGS) are associated with suicide-related outcomes20,21. Moreover, cases of death by suicide were enriched for genomic risk for a variety of traits related to externalizing, including AUD and other mental health problems22. However, we do not yet fully understand the nature of genetic overlap between externalizing and suicide-related outcomes.

In the current analysis, we explore the overlap between the externalizing spectrum and suicide-related outcomes in the Million Veteran Program (MVP) cohort. First, we validated a latent genomic factor for externalizing in MVP (MVP-EXT), comparing results to a previously published genome wide association study (GWAS)16. Next, compared the overlap of MVP-EXT and suicide-related outcomes using genetic correlations and polygenic scores in external cohorts to further explore these associations. Third, we used GWAS results to explore cell type enrichment in postmortem brain tissue of individuals who died by suicide compared to non-suicide deaths. Lastly, we performed a prospective analysis that evaluated the relationship between externalizing from electronic health records (EHR) and suicide mortality within the MVP cohort.

METHODS

The Million Veteran Program Cohort (MVP)

MVP links genomic laboratory testing, survey-based self-report data, and EHR data, with the goal of enhancing precision medicine initiatives23. Enrolled participants reflect the population that utilizes the VA, with over-representation of older and male individuals, as well as higher rates of multiple morbidities and chronic conditions related to externalizing compared to the general population24,25. Participants are active users of the VA healthcare system. Informed consent and authorization per the Health Insurance Portability and Accountability Act (HIPAA) were the only inclusion criteria. Once enrolled, participants’ EHR data are linked with their genetic data. The current analysis uses Release 4 of MVP data [February 2022] and was approved by the VA Central Institutional Review Board (IRB). All participants provided written informed consent.

MVP participants were genotyped on the MVP 1.0 Axiom array26. Genetic similarity of participants was classified using the HARE method27, which harmonizes the closest inferred ancestral population with self-reported race and ethnicity. Genotypic data were imputed to the Trans-Omics for Precision Medicine (TOPMed) reference panel28. We used data from the N = 467,101 veterans most similar to European reference panels (EUR-like) and the N = 124,717 veterans most similar to African reference panels (AFR-like) groups, as these had adequate statistical power for inclusion in the multivariate GWAS.

Genome Wide Association Study (GWAS) Phenotypes

We collated GWAS on externalizing-related traits within MVP that matched those in Externalizing Consortium GWAS (EXT1.0)16 as closely as possible when available. Our outcomes for the GWAS included in the multivariate models came from electronic health records (EHR) and the MVP Baseline Survey. EHR data were converted to phecodes, which are clusters of ICD-9/10-CM codes29,30. We defined a lifetime diagnosis for any given phecode as two or more occurrences of that phecode in their EHR, consistent with prior EHR analyses31,32. We examined all lifetime diagnoses without exclusion for overlap. We used phecodes for Substance addiction and disorders (Phecode 316, DUD), Alcohol-related disorders (Phecode 317, AUD), Tobacco use disorder (Phecode 318, TUD), and Attention deficit hyperactivity disorder (Phecode 313.1, ADHD). Questions for lifetime smoking (SMOK) and binge drinking (BINGE) came from the Baseline Survey. We performed all univariate GWASs using SAIGE33 to adjust for relatedness and included age, gender, and the first 20 genetic principal components as covariates (full details presented in the supplementary information). We filtered input GWAS to MAF > 1% with imputation scores (INFO) ≥ 0.80.

Multivariate GWAS and downstream analyses

We performed a confirmatory factor analysis using GenomicSEM34, which is robust to sample overlap and sample-size imbalances35. Within the EUR-like population we used the provided linkage disequilibrium (LD) scores36. For the AFR-like models, we used within-sample LD scores calculated using cov-LDSC37. We carried the best fitting models forward for multivariate GWAS. We ran the GWAS results from each population through FUMA (Functional Mapping and Annotation of GWAS)38 to identify independent loci (r2 < 0.1) among genome-wide significant SNPs. To further quantify the polygenic overlap between externalizing and suicide attempt, we applied bivariate MiXeR39 which uses a bivariate Gaussian mixture model to estimate the proportion of shared influential genetic variance between two traits. Finally, we ran all results through a standard series of post-GWAS pipelines (detailed in the supplementary information).

Single cell enrichment in post-mortem brain tissue of suicide deaths

To assess the collective expression of MVP-EXT related genes in those that died by suicide vs those who died by other causes, we applied the single-cell Disease-Relevance Scoring (scDRS)40 method. Each potential candidate gene was weighted by its MAGMA41 Z-score and adjusted for gene-specific technical noise in single-cell data obtained from an extensive postmortem dataset, consisting of 450K cells from the dorsolateral prefrontal cortex (DLPFC) of 40 human donors from the VA’s National PTSD Brain Bank (NPBB)42. This sample consisted of 16 “control” individuals (non-suicides related death) and 24 “cases” (confirmed suicide deaths, 12 with post-traumatic stress disorder and 12 with major depressive disorder), described in detail elsewhere43. The scDRS approach produced cell-specific raw disease scores. We generated 1,000 sets of cell-specific raw control scores using matched control gene sets, ensuring consistency in gene set size, mean expression, and expression variance with the candidate genes. Next, we normalized both the raw disease (e.g., suicide) and control scores for each cell, resulting in normalized disease and control scores. We evaluated cell type-level associations to identify broad cell types linked to the disease-status and to examine variability across individual cells within each cell type. All results were adjusted for multiple testing using a false discovery rate (FDR)44.

Independent validation cohorts

We included two independent cohorts to perform follow-up analyses. First, the Collaborative Study on the Genetics of Alcoholism (COGA) is a multi-site study of families densely affected with AUD and community-ascertained comparison families45–47. Participants completed a poly-diagnostic interview, the Semi-Structured Assessment for the Genetics of Alcoholism (SSAGA)48. The final analytic sample consisted of 10,986 COGA participants with genetic data49 (NEUR-like = 7,601; NAFR-like = 3,385). Second, the National Longitudinal Study of Adolescent to Adult Health (Add Health) is a nationally representative study of participants in the United States recruited as adolescents and followed into adulthood50. Our final analytic sample for Add Health consisted of 6,883 individuals (NEUR-like = 5,122; NAFR-like = 1,761). Full descriptions of each cohort are presented in the supplemental information.

Polygenic scores (PGS)

We estimated polygenic scores (PGS, aggregate measures of risk alleles weighted by GWAS effect sizes) using PRS-CSx51, which integrates GWAS summary statistics across multiple populations to improve the predictive power of PGS in the populations that typically lack well-powered GWAS results. Because of variation in allele frequencies and LD, PGS lose predictive accuracy when there is mismatch between the population of the discovery GWAS and target sample, even within relatively homogenous clusters52. PRS-CSx employs a Bayesian approach to correct GWAS summary statistics for the non-independence of SNPs in the genome. We converted PGS into Z-scores.

Prospective investigation of externalizing and suicide-related mortality in EHR

Lastly, to better contextualize externalizing in relation to prospective risk for death by suicide, we explored associations between externalizing and post-enrollment mortality. For each participant, we created a count of the past 12-month externalizing EHR codes (relative to MVP enrollment) and examined whether this count was associated with mortality via competing risk models53. We used cause of death (COD) from the ICD codes provided in the linked National Death Index (NDI) data up to December 2021 (v21, see Supplemental Table 1 for specific codes). The competing risk model allowed us to accurately estimate the hazards for specific causes of death (e.g., suicide death) in the context of competing causes (e.g., all other causes of death). We explored two competing risk models: 1) suicide-death in the context of risk for all other causes of death, and 2) “deaths of despair” (suicide, alcohol, and drug related deaths) in the context of risk for all other causes of death. We included age, gender, and race-ethnicity (non-Hispanic White, Black/African-American, Hispanic or Latino/a/x, American Indian/Alaskan Native, Asian, Native Hawaiian/Pacific Islander, multiracial, other race or ethnicity, or missing race or ethnicity) as covariates. Analyses were conducted using the survival package in R54,55.

It is important to note this paper includes language related to both race-ethnicity, which reflects socially-constructed categories, and genetic similarity, which uses empirical assignment based on available reference panels, because both are relevant for the current analyses. Prior work has established that racism, discrimination, and adverse social conditions — to which marginalized populations are disproportionally exposed — are relevant to suicide outcomes56–58. Additionally, we follow best practices for handling genetic data from diverse populations to limit bias from population stratification59. The inclusion of both concepts in no way endorses the notion that these reflect discrete biological categories.

RESULTS

Creating and validating a latent externalizing factor in MVP

Table 1 presents the individual GWAS results for the 6 indicators in the multivariate GWAS. Our final model for EUR-like veterans (Neff = 310,498) contained all six indicators and showed reasonably good fit to this model specification (Figure 1, Panel A; χ2= 104.10, df = 7, p = 1.53x10−19; CFI = 0.99, SRMR = 0.06). For the AFR-like veterans, the genetic correlation between SMOK and TUD was indistinguishable from one. We therefore retained only TUD, as it had greater statistical power. We also replaced ADHD with the attention problems subscale of the Medical Outcomes Survey as a proxy (MOS-ATTN, h2SNP [SE] = 0.033 [0.009], LDSC intercept = 1.003, Mean χ2 = 1.038, λGC = 1.035, rG with ADHD = 0.8) due to the low power of the ADHD indicator. The final model for AFR-like veterans (Neff = 99,949) therefore, largely mirrored that of the EUR-like veterans with one fewer indicator (Figure 1, Panel B; χ2 = 131.51, df = 8, p = 1.38x10−24; CFI = 0.96, SRMR = 0.09).

TABLE 1:

EXTERNALIZING PHENOTYPES INCLUDED IN THE MULTIVARIATE GWAS

Label Population Description Source N (eff) h2SNP (SE) LDSC int. Mean χ2 λGC
ADHD EUR-like Attention deficit hyperactivity disorder EHR 41,861 0.073 (0.013) 1.014 1.076 1.071
AUD EUR-like Alcohol use disorder EHR 249,476 0.101 (0.005) 1.024 1.529 1.414
DUD EUR-like Drug use disorder EHR 175,636 0.113 (0.005) 1.036 1.435 1.350
TUD EUR-like Tobacco use disorder EHR 365,610 0.091 (0.003) 1.065 1.749 1.570
SMOK EUR-like Ever smoker Survey 181,724 0.114 (0.005) 1.009 1.559 1.424
BINGE EUR-like Binge drinking frequency Survey 191,545 0.053 (0.004) 1.035 1.238 1.216
ADHD AFR-Like Attention deficit hyperactivity disorder EHR 3,753 0.096 (0.070) 0.999 1.011 1.008
AUD AFR-Like Alcohol use disorder EHR 94,280 0.050 (0.006) 1.008 1.154 1.139
DUD AFR-Like Drug use disorder EHR 87,879 0.049 (0.006) 1.016 1.150 1.140
TUD AFR-Like Tobacco use disorder EHR 102,346 0.049 (0.005) 1.010 1.172 1.155
SMOK AFR-Like Ever smoker Survey 45,638 0.050 (0.008) 0.998 1.071 1.061
BINGE AFR-Like Binge drinking frequency Survey 32,627 0.055 (0.011) 1.030 1.088 1.094
Open in a new tab

All statistics estimated using LD Score regression. N (eff) = effective sample size (4/[[1/N Cases] + [1/N Controls]], h2SNP = observed scale SNP-based heritability, LDSC int = LD Score regression intercept, Mean χ2 = mean χ2 statistic. λGC = genomic inflation factor (median χ2 statistic divided by the expected median of the χ2 distribution with 1 df).

Figure 1: Multivariate GWAS models of externalizing in MVP.

Figure 1:

Open in a new tab

Final models for EUR-like (Panel A) and AFR-like (Panel B) populations used in the subsequent multivariate GWASs. Df = degrees of freedom. AIC = Akaike’s information criterion, CFI = comparative fit index, SRMR = standardized root mean squared residual.

Next, we performed multivariate GWASs within the EUR-like and AFR-like veterans, separately, followed by an inverse-variance weighted fixed effects meta-analysis in METAL60. There were 155 independent genome wide significant loci in the meta-analysis, of which only 11 were significant Q-SNPs (p <.05/155 = 3.23 x10−4). Of the 138 of these loci that were present in EXT1.0, 94.2% (95% CI = 88.9%, 97.5%) were sign concordant and 116 (84.1%) of the loci were nominally significant (p < .05).

Finally, we took several steps to validate the MVP-EXT factor. First, we ran the genetic correlation between MVP-EXT and EXT1.0, which was strong (rG = 0.87, 95% CI = 0.83, 0.91). Second, we compared the genetic correlations of MVP-EXT and EXT1.0 with 93 external traits. The correlation of rG effect size estimates was very strong (r = 0.96, 95% CI = 0.93, 0.97). Third, we tested the association between the MVP-EXT PGS and a matched latent factor of externalizing in COGA and Add Health. The PGS was associated with externalizing in each cohort, as well as on overall effect (Beta META = 0.20, 95% CI = 0.18, 0.22). Taken together, these results suggest that the MVP-EXT factor largely recapitulated the model from EXT1.0 (full results from GWAS and post-GWAS biological characterization presented in the supplementary information and Supplemental Tables 2–8).

Exploring the genomic overlap between externalizing and suicide related outcomes

We next characterized the genetic overlap between MVP-EXT and suicide related phenotypes. First, we estimated the genetic correlations between MVP-EXT and both suicide attempt (SUI)61 and suicidal ideation (IDE)62. Figure 2 (Panel A) presents genetic correlation estimates for the EUR-like and AFR-like results side by side. Genetic correlations with SUI were strong for both the EUR-like (rG = 0.67, 95% CI = 0.60, 0.91) and AFR-like (rG = 0.74, 95% CI = 0.42, 0.81) participants. Genetic correlations with IDE were weak, but significant in the EUR-like (rG = 0.12, 95% CI = 0.08, 0.17) veterans and null for AFR-like (rG = −0.01, 95% CI = −0.20, 0.18) veterans. The results from MiXeR indicated that MVP-EXT shared ~77% of its influential variants with suicide attempt. Within this shared component, the variants that influence both MVP-EXT and suicide attempt had high sign concordance between SNPs (79%, SE = 0.03), supporting a model in which externalizing and suicide attempt were distinct traits with substantial polygenic overlap (full results in Supplemental Table 9).

Figure 2: Genetic correlations across populations.

Figure 2:

Open in a new tab

Genetic correlations (and 95% confidence intervals) between MVP-EXT and suicide attempt (SUI) and suicidal ideation (IDE) for both European-like and African-like results (Panel A). Genetic overlap between MVP-EXT and SUI using bivariate MiXeR in European-like results (Panel B). Venn diagram depicting the estimated number of influencing variants in thousands shared (grey) between and unique to suicide attempt (SUI, blue) and externalizing (MVP-EXT, orange). Conditional quantile–quantile (Q–Q) plots show observed versus expected −log10 p values in SUI and MVP-EXT as a function of the significance of association with the other trait. Log likelihood curves show goodness of model fit as a function of the count of shared influential variants.

Next, we examined the association between MVP-EXT PGS and suicide outcomes (full results in Supplemental Table 10). The MVP-EXT PGS was positively associated with lifetime suicidal ideation and suicide attempt in EUR-like participants from both COGA (Suicidal ideation: OR = 1.18, 95% CI = 1.11, 1.25; Suicide attempt: OR = 1.46, 95% CI = 1.32, 1.61) and Add Health (Suicidal ideation: OR = 1.07, 95% CI = 1.01, 1.13; Suicide attempt: OR = 1.13, 95% CI = 1.03, 1.24), but not the AFR-like participants in either COGA (Suicidal ideation: OR = 0.98, 95% CI = 0.89, 1.08; Suicide attempt: OR = 1.07, 95% CI = 0.94, 1.22) or Add Health (Suicidal ideation: OR = 1.04, 95% CI = 0.93, 1.16; Suicide attempt: OR = 0.97, 95% CI = 0.82, 1.15). For both suicidal ideation and suicide attempt, the meta-analyzed estimate was significant (Suicidal ideation: OR = 1.09, 95% CI = 1.05, 1.13; Suicide attempt: OR = 1.20, 95% CI = 1.13, 1.27). For context, Figure 3, Panel B presents the proportion of respondents in COGA reporting a lifetime suicide attempt across deciles of 1) the MVP-EXT PGS and 2) EXT-factor scores derived from the phenotypes used for replication. We see a similar pattern across predictor (PGS vs factor scores) and population, whereby the associations are stronger in EUR-like participants.

Figure 3: Polygenic associations between MVP-EXT PGS in COGA and Add Health.

Figure 3:

Open in a new tab

European-like, African-like, and meta-analysis associations between MVP-EXT PGS and suicide attempt (SUI) and suicidal ideation (IDE) in COGA and Add Health (Panel A). Proportion of individuals reporting lifetime suicide attempt across MVP-EXT PGS deciles (left) and phenotypic factor score deciles (right) in European-like and African-like COGA participants (Panel B).

Lastly, we examined whether the genetic architecture of externalizing converges to similar cell type enrichment across suicide deaths vs deaths from other causes (Figure 4A). Figure 4B demonstrates the cell-type clusters for comparison. We observed significant enrichment within inhibitory neurons, astroglia, and oligodendrocyte progenitor cells (OPCs) in the meta-analyzed results (Figure 4C); however, only the inhibitory neurons were significant in the AFR-like results, a finding that is likely attributable to lower statistical power.

Figure 4: Cell-type enrichment for MVP-EXT results in post-mortem brain tissue.

Figure 4:

Open in a new tab

Cell type enrichment across suicide deaths vs deaths from other causes (Panel A). Cell-type clusters (Panel B). Tests for cell-type enrichment across European-like, African-like, and meta-analysis results (Panel C).

Prospective phenotypic analyses in electronic health records

Finally, we created counts of past 12-month externalizing codes from the time of MVP enrollment (N = 840,865 as of 12/31/2021, overlapping with NDI time frames). Demographic information is presented in Supplemental Table 11. Of these participants, 13.8% (N = 116,365) were deceased by the end of 2021. Of the deceased, N = 1,268 were classified as suicides, while the more expansive definition for deaths of despair (which includes suicide) made up N = 4,410 of the deaths. The results from the competing risk models are presented in Table 2. Externalizing codes were associated with both death by suicide (HR = 1.39, 95% CI = 1.33, 1.45) and the broader deaths of despair category (HR = 1.96, 95% CI = 1.93, 2.00). In both models, the externalizing was associated with all other causes of mortality (HR = 1.42, 95% CI = 1.41, 1.43; HR = 1.38, 95% CI = 1.38, 1.39). For those with 4+ past year externalizing codes (compared to those with none), the 5-year cumulative incidence for suicide death was nearly 5x greater and the 5-year cumulative incidence for deaths of despair was almost 20x greater.

TABLE 2:

COMPETING RISK MODELS AND 5-YEAR CUMULATIVE INCIDENCE FOR COUNT OF EXTERNALIZING DIAGNOSES AND SUICIDE-RELATED MORTALITY

Outcome HR/Cumulative incidence 95% CI 95% CI P-value
Suicide death 1.39 1.33 1.45 6.18x10−51
5-year cumulative incidence:
0 externalizing codes 0.10% 10.77% 10.95% -
1-3 externalizing codes 0.20% 13.50% 13.80% -
4+ externalizing codes 0.48% 14.76% 15.83% -
All other deaths 1.42 1.41 1.43 2.23 x10−308
5-year cumulative incidence:
0 externalizing codes 10.86% 0.09% 0.11% -
1-3 externalizing codes 13.65% 0.18% 0.22% -
4+ externalizing codes 15.30% 0.38% 0.58% -
Death of despair 1.96 1.93 2.00 2.23 x10−308
5-year cumulative incidence:
0 externalizing codes 0.19% 0.18% 0.20% -
1-3 externalizing codes 0.91% 0.87% 0.95% -
4+ externalizing codes 3.65% 3.37% 3.93% -
All other deaths 1.38 1.38 1.39 2.23 x10−308
5-year cumulative incidence:
0 externalizing codes 10.78% 10.69% 10.86% -
1-3 externalizing codes 12.93% 12.78% 13.08% -
4+ externalizing codes 12.13% 11.64% 12.61% -
Open in a new tab

HR = hazard ratio. CI = confidence interval. Lists of ICD-10 codes for suicide death and deaths of despair listed in Supplemental Table 1. All competing risk models covaried for age (at enrollment), gender, and race-ethnicity.

DISCUSSION

Externalizing has been consistently linked to suicide related outcomes at the phenotypic20,63–65 and genetic level16,20,21,66,67, pointing to the importance of behavioral disinhibition and impulsivity in suicidality. In the current analysis, we sought to characterize how genetic liability for externalizing is related to suicide outcomes within the veteran population using the VA’s Million Veteran Program Cohort to further explore this overlap, using genetic data from European-like and African-like populations. We were able to recapitulate a latent externalizing factor that overlapped strongly with prior results (rg = .87) which contained no MVP data16. Additionally, we expanded prospective analyses for externalizing and mortality into the full MVP sample.

We found consistent evidence for genetic overlap between externalizing and suicide-related outcomes. Genetic correlations between MVP-EXT and suicide related outcomes across the EUR-like and AFR-like results had similar patterns of associations. MVP-EXT was strongly associated with suicide attempt in both populations, and results from MiXeR demonstrated that these are distinct phenotypes that have significant polygenic overlap. While MVP-EXT was only significantly associated with suicidal ideation in the EUR-like results, the confidence intervals overlapped with those in the AFR-like results, suggesting that this may be the result of the lower power in the AFR-like results. In the polygenic score analyses, the association between MVP-EXT PGS and suicide-related outcomes was null for the AFR-like participants. There are several possible explanations for this lack of association. First, it is possible we did not have sufficient power, as seen in the genetic correlation with ideation in the AFR-like participants. Second, our results are primarily derived from EHR. Marginalized racial and ethnic populations receive less engagement in EHR (e.g., fewer number of EHR actions performed within a patient’s EHR per hour) compared to those who are non-Hispanic White 68, leading to a potentially biased assessment of externalizing. Lastly, in line with the “social push” model 69 of gene-environment interaction, the null association could reflect the increased relevance of adverse social conditions amongst AFR-like participants (e.g., racism, discrimination)56–58, who are predominantly African-American. Importantly, these are not mutually exclusive explanations.

Beyond genetic analyses, which may capture more trait-like (e.g., aggerate lifetime risk) associations, current – as of time of enrollment – externalizing codes were associated with suicide related mortality, potentially capturing time-dependent risk. Our analyses demonstrate that both current and lifetime risk are important for the externalizing-suicide relationship. Importantly, the association with MVP-EXT and suicide-related mortality remains in the context of risk for other causes of death. Given the multitude of negative health correlates of EXT found in previous analyses of MVP (e.g., cirrhosis, chronic airway obstruction, viral hepatitis C, etc.)21 these results help frame the clinical importance of externalizing beyond the specific health consequence related to substance use. Development and implementation of externalizing risk screeners at intake may help to mitigate a variety of negative health outcomes, including suicidality.

This research has several important limitations. First, while we utilized the largest sample of non-European participants available in MVP, we still lacked sufficient power within the AFR-like participants to identify loci associated with externalizing. As future releases of MVP become available, it will be important to include larger samples and data from additional populations. Second, these analyses focused solely on the role of genetic risk. Suicidal behaviors are complex, and social factors (e.g., unemployment, experiences of homelessness, social support) also play an important role56. Future efforts should examine genomic risk factors in conjunction with well-established social and clinical determinants of suicide-related outcomes, as has been done with SUD70. Third, our sample was predominantly comprised of men (~85%) and there are large gender differences in suicidal behaviors, especially in regard to suicide attempt vs death. Whether the results from the current analysis generalize broadly remains an open question.

Suicide related behaviors remain a critical public health problem with a complex etiology. While the relation between suicide with internalizing and mood-related disorders has received greater attention, externalizing problems are an important factor in risk for suicide related behaviors. We recapitulated here finding from a previous GWAS of externalizing problems using data derived exclusively from MVP. Our analysis demonstrates that externalizing risk is broadly associated with suicide attempt, with variations across populations. Importantly, current diagnoses of externalizing problems are associated with future suicide-related mortality. Efforts at early intervention and suicide screening should consider addressing problems related to externalizing and behavioral disinhibition in addition to other common suicide-related risk factors.

Supplementary Material

Supplement 1
media-1.docx (1.2MB, docx)
Supplement 2
media-2.xlsx (81.5MB, xlsx)

ACKNOWLEDGEMENTS

This research is based on data from the Million Veteran Program, Office of Research and Development, Veterans Health Administration, and was supported by MVP000. This work was also funded by grant #1I01CX001729 from the Clinical Services Research & Development (CSRD) Service of VA ORD to Drs. Beckham and Kimbrel, by the MVP CHAMPION program, which is a collaboration between the VA and the Department of Energy (DoE), and by a CSRD Senior Research Scientist award (lK6BX003777) to Dr. Beckham. Dr. Kimbrel was also supported by a VA Research Career Scientist Award (#I01BX005881) from VA ORD. Drs. Barr, Bigdeli, Aslan, and Harvey are supported by VA Cooperative Studies Program (CSP) #572. Drs. Peterson, Bigdeli, and Meyers are supported by the National Institute of Mental Health (R01MH125938). Dr. Peterson is also supported by the National Institute on Alcohol Abuse and Alcoholism (P50AA022537) and the Brain Behavior Research Foundation NARSAD grant 28632 PS Fund. Drs. Barr and Dick are also supported by the National Institute of Drug Abuse (R01DA050721) and the National Institute of Alcohol Abuse and Alcoholism (R01AA015416). Dr. Sanchez-Roige was supported by funds from the California Tobacco-Related Disease Research Program (TRDRP; Grant Number T32IR5226) and the National Institute on Drug Abuse (DP1DA054394). Dr. Mallard is supported by funds from NIH T32HG010464. This publication does not represent the views of the Department of Veteran Affairs (VA), the National Institutes of Health, or the United States Government. Dr. Barr had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation of the manuscript; and decision to submit the manuscript for publication. The MVP Publication Committee reviewed and approved the manuscript.

We would also like to thank The Externalizing Consortium for sharing the GWAS summary statistics of externalizing. 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), and the National Institute on Drug Abuse (R01DA050721). Additional funding for investigator effort has been provided by K02AA018755, U10AA008401, P50AA022537, 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 Inc., Add Health, Vanderbilt University Medical Center’s BioVU, Collaborative Study on the Genetics of Alcoholism (COGA), the Psychiatric Genomics Consortium’s Substance Use Disorders working group, UK10K Consortium, UK Biobank, and Philadelphia Neurodevelopmental Cohort. We would like to thank the many studies that made these consortia possible, the researchers involved, and the participants in those studies, without whom this effort would not be possible. We would also like to thank the research participants and employees of 23andMe.

This research uses data from Add Health, funded by grant P01 HD31921 (Harris) from the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD), with cooperative funding from 23 other federal agencies and foundations. Add Health is currently directed by Robert A. Hummer and funded by the National Institute on Aging cooperative agreements U01 AG071448 (Hummer) and U01AG071450 (Aiello and Hummer) at the University of North Carolina at Chapel Hill. Add Health was designed by J. Richard Udry, Peter S. Bearman, and Kathleen Mullan Harris at the University of North Carolina at Chapel Hill.

Most importantly, we would like to thank the U.S. Veteran participants in MVP for their service, and for their time, samples, and continued participation in VA research. Without them, this work would not be possible.

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, A. Anderson); SUNY Downstate Health Sciences University (B. Porjesz, J. Meyers); Washington University in St. Louis (L. Bierut, A. Agrawal, S. Hartz); University of California at San Diego (M. Schuckit); Rutgers University (D. Dick, R. Hart, J. Salvatore, J. Tischfield); 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); D. Lai, J. Nurnberger Jr., L. Wetherill, X., Xuei, S. O’Connor, (Indiana University); J. Kramer (University of Iowa), G. Chan (University of Iowa; University of Connecticut); C. Kamarajan, A. Pandey, D.B. Chorlian, P. Barr, S. Kinreich, G. Pandey, Z. Neale, S., C. Chatzinakos, J. Zhang, Saenz deViteri, R. Christian, A. Bingly (SUNY Downstate); G. Pathak (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, J. Moore (Rutgers University); A. Merikangas (The Children’s Hospital of Philadelphia and University of Pennsylvania); M. Gitik, NIAAA Staff Collaborator. 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).

Footnotes

CONFLICTS OF INTEREST DISCLOSURES

Dr. Dick is the Chief Scientific Officer of Thrive Genetics, Inc. Dr. Dick is also on the Advisory Board for the Seek Women’s Health Company. She owns stock in both companies. She received royalties from authoring the book, The Child Code: Understanding Your Child’s Unique Nature for Happier, More Effective Parenting, published by Avery, an imprint of the Penguin group. Dr Harvey reported receiving personal fees from Boehringer Ingelheim, Bioexcel, Karuna Therapeutics, Minerva Neuroscience, Alkermes, Sunovion, and Roche; royalties from WCG Endpoint Solutions; and equity from i-Function outside the submitted work. These did not impact or influence the collection, development, analysis, or interpretation of the contents of this manuscript. No other disclosures were reported.

DATA AVAILABILITY

All data sources are described in the manuscript and supplemental information. Only data from existing studies or study cohorts were analyzed, some of which have restricted access to protect the privacy of the study participants. The full summary statistics for the discovery MVP GWAS are available via dbGaP (accession: phs001672.v5.p1). Add Health genetic data can be obtained through dbGaP (accession: phs001367.v1.p1). COGA genetic data are available through dbGaP (accession: phs000763.v1.p1).

REFERENCES

  • 1.Case A, Deaton A. Rising morbidity and mortality in midlife among white non-Hispanic Americans in the 21st century. Proc Natl Acad Sci U S A. 2015;112(49):15078–15083. doi: 10.1073/pnas.1518393112 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Tilstra AM, Simon DH, Masters RK. Trends in “Deaths of Despair” among Working-Aged White and Black Americans, 1990-2017. Am J Epidemiol. 2021;190(9):1751–1759. doi: 10.1093/aje/kwab088 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Olfson M, Blanco C, Wall M, et al. National Trends in Suicide Attempts Among Adults in the United States. JAMA Psychiatry. 2017;74(11):1095–1103. doi: 10.1001/jamapsychiatry.2017.2582 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Potash JB, Scott Kane MH, Chiu Y feng, et al. Attempted Suicide and Alcoholism in Bipolar Disorder: Clinical and Familial Relationships. Am J Psychiatry. 2000;157:12. [DOI] [PubMed] [Google Scholar]
  • 5.Sher L. Alcoholism and suicidal behavior: A clinical overview. Acta Psychiatr Scand. 2006;113(1):13–22. doi: 10.1111/j.1600-0447.2005.00643.x [DOI] [PubMed] [Google Scholar]
  • 6.Koller G, Preuß UW, Bottlender M, Wenzel K, Soyka M. Impulsivity and aggression as predictors of suicide attempts in alcoholics. Eur Arch Psychiatry Clin Neurosci. 2002;252(4):155–160. doi: 10.1007/s00406-002-0362-9 [DOI] [PubMed] [Google Scholar]
  • 7.Whiteford HA, Degenhardt L, Rehm J, et al. Global burden of disease attributable to mental and substance use disorders: findings from the Global Burden of Disease Study 2010. The Lancet. 2013;382(9904):1575–1586. doi: 10.1016/S0140-6736(13)61611-6 [DOI] [PubMed] [Google Scholar]
  • 8.Modesto-Lowe V, Brooks D, Ghani M. Alcohol dependence and suicidal behavior: From research to clinical challenges. Harv Rev Psychiatry. 2006;14(5):241–248. doi: 10.1080/10673220600975089 [DOI] [PubMed] [Google Scholar]
  • 9.Septier M, Stordeur C, Zhang J, Delorme R, Cortese S. Association between suicidal spectrum behaviors and Attention-Deficit/Hyperactivity Disorder: A systematic review and meta-analysis. Neurosci Biobehav Rev. 2019;103:109–118. doi: 10.1016/j.neubiorev.2019.05.022 [DOI] [PubMed] [Google Scholar]
  • 10.Pompili M, Gonda X, Serafini G, et al. Epidemiology of suicide in bipolar disorders: a systematic review of the literature. Bipolar Disord. 2013;15(5):457–490. doi: 10.1111/bdi.12087 [DOI] [PubMed] [Google Scholar]
  • 11.Achenbach TM. The classification of children’s psychiatric symptoms: a factor-analytic study. Psychol Monogr. 1966;80(7):1–37. doi: 10.1037/H0093906 [DOI] [PubMed] [Google Scholar]
  • 12.Kotov R, Cicero DC, Conway CC, et al. The Hierarchical Taxonomy of Psychopathology (HiTOP) in psychiatric practice and research. Psychol Med. 2022;52(9):1666–1678. doi: 10.1017/S0033291722001301 [DOI] [PubMed] [Google Scholar]
  • 13.Anttila V, Bulik-Sullivan B, Finucane HK, et al. Analysis of shared heritability in common disorders of the brain. Science (1979). 2018;360(6395):eaap8757. doi: 10.1126/science.aap8757 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Young SE, Stallings MC, Corley RP, Krauter KS, Hewitt JK. Genetic and environmental influences on behavioral disinhibition. Am J Med Genet. 2000;96(5):684–695. doi: [DOI] [PubMed] [Google Scholar]
  • 15.Hicks BM, Schalet BD, Malone SM, Iacono WG, McGue M. Psychometric and Genetic Architecture of Substance Use Disorder and Behavioral Disinhibition Measures for Gene Association Studies. Behav Genet. 2011;41(4):459–475. doi: 10.1007/s10519-010-9417-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.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. Published online 2021:1–10. doi: 10.1038/s41593-021-00908-3 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Baselmans B, Hammerschlag AR, Noordijk S, et al. The Genetic and Neural Substrates of Externalizing Behavior. Biological Psychiatry Global Open Science. 2022;2(4):389–399. doi: 10.1016/j.bpsgos.2021.09.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Clapp Sullivan ML, Schwaba T, Harden KP, Grotzinger AD, Nivard MG, Tucker-Drob EM. Beyond the factor indeterminacy problem using genome-wide association data. Nat Hum Behav. 2024;8(2):205–218. doi: 10.1038/s41562-023-01789-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Mullins N, Kang JE, Campos AI, et al. Dissecting the Shared Genetic Architecture of Suicide Attempt, Psychiatric Disorders, and Known Risk Factors. Biol Psychiatry. 2022;91(3):313–327. doi: 10.1016/j.biopsych.2021.05.029 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Bigdeli TB, Barr PB, Rajeevan N, et al. Correlates of suicidal behaviors and genetic risk among United States veterans with schizophrenia or bipolar I disorder. Mol Psychiatry. 2024;29(8):2399–2407. doi: 10.1038/s41380-024-02472-1 [DOI] [PubMed] [Google Scholar]
  • 21.Barr PB, Bigdeli TB, Meyers JL, et al. Correlates of Risk for Disinhibited Behaviors in the Million Veteran Program Cohort. JAMA Psychiatry. 2024;81(2):188–197. doi: 10.1001/jamapsychiatry.2023.4141 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Docherty AR, Shabalin AA, DiBlasi E, et al. Genome-Wide Association Study of Suicide Death and Polygenic Prediction of Clinical Antecedents. American Journal of Psychiatry. 2020;177(10):917–927. doi: 10.1176/appi.ajp.2020.19101025 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Gaziano JM, Concato J, Brophy M, et al. Million Veteran Program: A mega-biobank to study genetic influences on health and disease. J Clin Epidemiol. 2016;70:214–223. doi: 10.1016/j.jclinepi.2015.09.016 [DOI] [PubMed] [Google Scholar]
  • 24.Kramarow EA, Pastor PN. The health of male veterans and nonveterans aged 25-64: United States, 2007-2010. NCHS Data Brief. 2012;(101):1–8. [PubMed] [Google Scholar]
  • 25.Boersma P, Cohen RA, Zelaya CE, Moy E. Multiple Chronic Conditions Among Veterans and Nonveterans: United States, 2015-2018. Natl Health Stat Report. 2021;(153):1–13. Accessed March 28, 2022. https://europepmc.org/article/med/33663648 [PubMed] [Google Scholar]
  • 26.Hunter-Zinck H, Shi Y, Li M, et al. Genotyping Array Design and Data Quality Control in the Million Veteran Program. Am J Hum Genet. 2020;106(4):535–548. doi: 10.1016/j.ajhg.2020.03.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Fang H, Hui Q, Lynch J, et al. Harmonizing Genetic Ancestry and Self-identified Race/Ethnicity in Genome-wide Association Studies. Am J Hum Genet. 2019;105(4):763–772. doi: 10.1016/j.ajhg.2019.08.012 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Taliun D, Harris DN, Kessler MD, et al. Sequencing of 53,831 diverse genomes from the NHLBI TOPMed Program. Nature. 2021;590(7845):290–299. doi: 10.1038/s41586-021-03205-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Denny JC, Bastarache L, Ritchie MD, et al. Systematic comparison of phenome-wide association study of electronic medical record data and genome-wide association study data. Nat Biotechnol. 2013;31(12):1102–1110. doi: 10.1038/nbt.2749 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Wu P, Gifford A, Meng X, et al. Mapping ICD-10 and ICD-10-CM codes to phecodes: Workflow development and initial evaluation. J Med Internet Res. 2019;21(11):1–13. doi: 10.2196/14325 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Zheutlin AB, Dennis J, Linnér RK, et al. Penetrance and pleiotropy of polygenic risk scores for schizophrenia in 106,160 patients across four health care systems. American Journal of Psychiatry. 2019;176(10):846–855. doi: 10.1176/appi.ajp.2019.18091085 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Bigdeli TB, Voloudakis G, Barr PB, et al. Penetrance and Pleiotropy of Polygenic Risk Scores for Schizophrenia, Bipolar Disorder, and Depression among Adults in the US Veterans Affairs Health Care System. JAMA Psychiatry. 2022;79(11):1092–1101. doi: 10.1001/jamapsychiatry.2022.2742 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 33.Zhou W, Nielsen JB, Fritsche LG, et al. Efficiently controlling for case-control imbalance and sample relatedness in large-scale genetic association studies. Nat Genet. Published online 2018:212357. doi: 10.1038/s41588-018-0184-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 34.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. 2019;3(5):513–525. doi: 10.1038/s41562-019-0566-x [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Lee PH, Anttila V, Won H, et al. Genomic relationships, novel loci, and pleiotropic mechanisms across eight psychiatric disorders. Cell. 2019;179(7):1469–1482. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Waldman ID, Poore HE, Luningham JM, Yang J. Testing structural models of psychopathology at the genomic level. World Psychiatry. Published online 2020. doi: 10.1002/wps.20772 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Luo Y, Li X, Wang X, et al. Estimating heritability and its enrichment in tissue-specific gene sets in admixed populations. Hum Mol Genet. 2021;30(16):1521–1534. doi: 10.1093/hmg/ddab130 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Watanabe K, Taskesen E, van Bochoven A, Posthuma D. Functional mapping and annotation of genetic associations with FUMA. Nat Commun. 2017;8(1):1–11. doi: 10.1038/s41467-017-01261-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Frei O, Holland D, Smeland OB, et al. Bivariate causal mixture model quantifies polygenic overlap between complex traits beyond genetic correlation. Nat Commun. 2019;10(1):2417. doi: 10.1038/s41467-019-10310-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 40.Zhang MJ, Hou K, Dey KK, et al. Polygenic enrichment distinguishes disease associations of individual cells in single-cell RNA-seq data. Nat Genet. 2022;54(10):1572–1580. doi: 10.1038/s41588-022-01167-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.de Leeuw CA, Mooij JM, Heskes T, Posthuma D. MAGMA: Generalized Gene-Set Analysis of GWAS Data. PLoS Comput Biol. 2015;11(4):e1004219. doi: 10.1371/journal.pcbi.1004219 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Friedman MJ, Huber BR, Brady CB, et al. VA’s National PTSD Brain Bank: a National Resource for Research. Curr Psychiatry Rep. 2017;19(10):73. doi: 10.1007/s11920-017-0822-6 [DOI] [PubMed] [Google Scholar]
  • 43.Chatzinakos C, Pernia CD, Morrison FG, et al. Single-Nucleus Transcriptome Profiling of Dorsolateral Prefrontal Cortex: Mechanistic Roles for Neuronal Gene Expression, Including the 17q21.31 Locus, in PTSD Stress Response. American Journal of Psychiatry. 2023;180(10):739–754. doi: 10.1176/appi.ajp.20220478 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 44.Benjamini Y, Hochberg Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society Series B (Methodological). 1995;57(1):289–300. http://www.jstor.org/stable/2346101 [Google Scholar]
  • 45.Agrawal A, Brislin SJ, Bucholz KK, et al. The Collaborative Study on the Genetics of Alcoholism: Overview. Genes Brain Behav. 2023;22(5):e12864. doi: 10.1111/gbb.12864 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Begleiter H. The Collaborative Study on the Genetics of Alcoholism. Alcohol Health Res World. 1995;19(3):228–236. [PMC free article] [PubMed] [Google Scholar]
  • 47.Dick DM, Balcke E, McCutcheon V, et al. The collaborative study on the genetics of alcoholism: Sample and clinical data. Genes Brain Behav. 2023;22(5):e12860. doi: 10.1111/gbb.12860 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 48.Bucholz KK, Cadoret R, Cloninger CR, et al. A new, semi-structured psychiatric interview for use in genetic linkage studies: a report on the reliability of the SSAGA. J Stud Alcohol. 1994;55(2):149–158. [DOI] [PubMed] [Google Scholar]
  • 49.Johnson EC, Salvatore JE+, Lai D, et al. The collaborative study on the genetics of alcoholism: Genetics. Genes Brain Behav. 2023;22(5):e12856. doi: 10.1111/gbb.12856 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Harris KM, Halpern CT, Whitsel E, et al. The National Longitudinal Study of Adolescent to Adult Health: research design. Add Health: The National Longitudinal Study of Adolescent to Adult Health. Published online 2009. http://www.cpc.unc.edu/projects/addhealth/design
  • 51.Ruan Y, Lin YF, Feng YCA, et al. Improving polygenic prediction in ancestrally diverse populations. Nat Genet. 2022;54(5):573–580. doi: 10.1038/s41588-022-01054-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Ding Y, Hou K, Xu Z, et al. Polygenic scoring accuracy varies across the genetic ancestry continuum. Nature. 2023;618(7966):774–781. doi: 10.1038/s41586-023-06079-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 53.Putter H, Fiocco M, Geskus RB. Tutorial in biostatistics: competing risks and multi-state models. Stat Med. 2007;26(11):2389–2430. doi: 10.1002/sim.2712 [DOI] [PubMed] [Google Scholar]
  • 54.Therneau Terry M., Grambsch Patricia M. Modeling Survival Data: Extending the Cox Model. Springer; 2000. [Google Scholar]
  • 55.Therneau TM. A Package for Survival Analysis in R. Published online 2024. https://CRAN.R-project.org/package=survival
  • 56.Blosnich JR, Montgomery AE, Dichter ME, et al. Social Determinants and Military Veterans’ Suicide Ideation and Attempt: a Cross-sectional Analysis of Electronic Health Record Data. J Gen Intern Med. 2020;35(6):1759–1767. doi: 10.1007/s11606-019-05447-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 57.Coimbra BM, Hoeboer CM, Yik J, Mello AF, Mello MF, Olff M. Meta-analysis of the effect of racial discrimination on suicidality. SSM Popul Health. 2022;20:101283. doi: 10.1016/j.ssmph.2022.101283 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58.Paradies Y, Ben J, Denson N, et al. Racism as a Determinant of Health: A Systematic Review and Meta-Analysis. PLoS One. 2015;10(9):e0138511-. 10.1371/journal.pone.0138511 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 59.Peterson RE, Kuchenbaecker K, Walters RK, et al. Genome-wide Association Studies in Ancestrally Diverse Populations: Opportunities, Methods, Pitfalls, and Recommendations. Cell. 2019;179(3):589–603. doi: 10.1016/j.cell.2019.08.051 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60.Willer CJ, Li Y, Abecasis GR. METAL: fast and efficient meta-analysis of genomewide association scans. Bioinformatics. 2010;26(17):2190–2191. doi: 10.1093/bioinformatics/btq340 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 61.Docherty AR, Mullins N, Ashley-Koch AE, et al. GWAS Meta-Analysis of Suicide Attempt: Identification of 12 Genome-Wide Significant Loci and Implication of Genetic Risks for Specific Health Factors. American Journal of Psychiatry. 2023;180(10):723–738. doi: 10.1176/appi.ajp.21121266 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62.Ashley-Koch AE, Kimbrel NA, Qin XJ, et al. Genome-wide association study identifies four panancestry loci for suicidal ideation in the Million Veterans Program. PLoS Genet. 2023;19(3):e1010623. doi: 10.1371/journal.pgen.1010623 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 63.Edwards AC, Ohlsson H, Sundquist J, et al. The role of substance use disorders in the transition from suicide attempt to suicide death: a record linkage study of a Swedish cohort. Psychol Med. 2024;54(1):90–97. doi: 10.1017/S0033291722002240 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 64.Ferrari AJ, Norman RE, Freedman G, et al. The burden attributable to mental and substance use disorders as risk factors for suicide: Findings from the Global Burden of Disease Study 2010. PLoS One. 2014;9(4):e91936. doi: 10.1371/journal.pone.0091936 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 65.Lannoy S, Ohlsson H, Kendler KS, et al. Risk of suicidal behavior as a function of alcohol use disorder typologies: A Swedish population-based study. Addiction. 2024;119(2):281–290. doi: 10.1111/add.16351 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 66.Colbert SMC, Hatoum AS, Shabalin A, et al. Exploring the genetic overlap of suicide-related behaviors and substance use disorders. American Journal of Medical Genetics, Part B: Neuropsychiatric Genetics. 2021;186(8):445–455. doi: 10.1002/ajmg.b.32880 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 67.Colbert SMC, Mullins N, Chan G, et al. Polygenic Contributions to Suicidal Thoughts and Behaviors in a Sample Ascertained for Alcohol Use Disorders. Complex Psychiatry. 2023;9(1-4):11–23. doi: 10.1159/000529164 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 68.Yan C, Zhang X, Yang Y, et al. Differences in Health Professionals’ Engagement With Electronic Health Records Based on Inpatient Race and Ethnicity. JAMA Netw Open. 2023;6(10):e2336383–e2336383. doi: 10.1001/jamanetworkopen.2023.36383 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 69.Boardman JD, Daw J, Freese J. Defining the environment in gene-environment research: Lessons from social epidemiology. Am J Public Health. 2013;103(SUPPL.1):S64--S72. doi: 10.2105/AJPH.2013.301355 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 70.Barr PB, Driver MN, Kuo SIC, et al. Clinical, environmental, and genetic risk factors for substance use disorders: characterizing combined effects across multiple cohorts. Mol Psychiatry. 2022;27(11):4633–4641. doi: 10.1038/s41380-022-01801-6 [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 (1.2MB, docx)
Supplement 2
media-2.xlsx (81.5MB, xlsx)

Data Availability Statement

All data sources are described in the manuscript and supplemental information. Only data from existing studies or study cohorts were analyzed, some of which have restricted access to protect the privacy of the study participants. The full summary statistics for the discovery MVP GWAS are available via dbGaP (accession: phs001672.v5.p1). Add Health genetic data can be obtained through dbGaP (accession: phs001367.v1.p1). COGA genetic data are available through dbGaP (accession: phs000763.v1.p1).

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

RESOURCES