Associations between Suicidal Thoughts and Behaviors and Genetic Liability for Cognitive Performance, Depression, and Risk-Taking in a High-Risk Sample

Abstract

<b><i>Background:</i></b> Suicidal thoughts and behaviors (STBs) and nonsuicidal self-injury (NSSI) behaviors are moderately heritable and may reflect an underlying predisposition to depression, impulsivity, and cognitive vulnerabilities to varying degrees. <b><i>Objectives:</i></b> We aimed to estimate the degrees of association between genetic liability to depression, impulsivity, and cognitive performance and STBs and NSSI in a high-risk sample. <b><i>Methods:</i></b> We used data on 7,482 individuals of European ancestry and 3,359 individuals of African ancestry from the Collaborative Study on the Genetics of Alcoholism to examine the links between polygenic scores (PGSs) for depression, impulsivity/risk-taking, and cognitive performance with 3 self-reported indices of STBs (suicidal ideation, persistent suicidal ideation defined as ideation occurring on at least 7 consecutive days, and suicide attempt) and with NSSI. <b><i>Results:</i></b> The PGS for depression was significantly associated with all 4 primary self-harm measures, explaining 0.6–2.5% of the variance. The PGS for risk-taking behaviors was also associated with all 4 self-harm behaviors in baseline models, but was no longer associated after controlling for a lifetime measure of DSM-IV alcohol dependence and abuse symptom counts. Polygenic predisposition for cognitive performance was negatively associated with suicide attempts (<i>q</i> = 3.8e−4) but was not significantly associated with suicidal ideation nor NSSI. We did not find any significant associations in the African ancestry subset, likely due to smaller sample sizes. <b><i>Conclusions:</i></b> Our results encourage the study of STB as transdiagnostic outcomes that show genetic overlap with a range of risk factors.

Genetic differences between suicide deaths and deaths of undetermined intent

“…Even those adjusted FGRS estimates that differ significantly from the null hypothesis (i.e., OR = 1) are of small effect size, particularly relative to the effect sizes observed for several of the phenotypic registrations. Standard polygenic risk scores based on individual-level molecular data, when derived from and applied to psychiatric outcomes, vary in their estimated effects, but it is not unusual to see ORs that are higher than those in the current report (Campos et al, 2021;Fanelli et al, 2022;Johnson et al, 2021). However, the nature of our research question -i.e., whether the aggregate genetic risk can distinguish between members of two groups who might overlap considerably -is somewhat more nuanced than, for example, efforts to test whether polygenic scores are associated with more clearly delineated groups such as those with versus without schizophrenia.…”

Correlates of Risk for Disinhibited Behaviors in the Million Veteran Program Cohort

“…Recent multivariate genome wide association studies (GWAS) have found robust evidence for a latent genomic factor for externalizing disorders 11,12 , composed of input GWAS related to substance use disorders, risky sexual behaviors, personality characteristics, and neurodevelopmental disorders. Importantly, genetic liability for externalizing disorders overlaps with other phenotypes of public health relevance, such as suicidal thoughts and behaviors 11,[13][14][15] , SUDs 11,[16][17][18] , and a range of other medical conditions (e.g., ischemic heart disease, liver disease, viral hepatitis) 11 . The widespread impact of risk for externalizing disorders makes it a potential target for early intervention and prevention.…”

Polygenic contributions to suicidal thoughts and behaviors in a sample ascertained for alcohol use disorders

“…Non-suicidal self-injury and suicidal thoughts and behaviors are transdiagnostic, co-occurring with psychiatric disorders ranging from bipolar disorder [20] to eating disorders [21], but are most frequently studied in the context of depression. Polygenic risk scores (PRS; an estimate of one's genetic liability to a trait of interest) for major depressive disorder, depressive symptoms and/or broad depression are associated with self-reported suicidal thoughts and behaviors ranging from selfharm (in general) to suicidal ideation, suicide attempt and death by suicide [19,[22][23][24][25][26][27].…”

“…46) [17] and suicide attempt (r g =0.78) [18] and depression [29] is also not perfectly genetically correlated with death by suicide (r g =0.42) [30]. Some studies have found significant associations between observed suicidal thoughts and behaviors and PRS of selfharm [16,23] and suicide attempt [18,24,31]. However, with the exception of a recent study from Mullins et al [18] which identified a significant association between suicide attempt and a PRS constructed from a GWAS of suicide attempt conditioned on major depressive disorder [18], little research has explored whether genetic liability to non-suicidal self-injury and suicidal thoughts and behaviors that are independent from genetic liability to depression explain variance in suicidal thoughts and behaviors.…”

“…However, major depressive disorder [28] is only partially genetically correlated with suicidal ideation (r g =0.46) [17] and suicide attempt (r g =0.78) [18] and depression [29] is also not perfectly genetically correlated with death by suicide (r g =0.42) [30]. Some studies have found significant associations between observed suicidal thoughts and behaviors and PRS of self-harm [16,23] and suicide attempt [18,24,31]. However, with the exception of a recent study from Mullins et al .…”