Genetic relationships and biobehavioral pathways between suicidality and comorbid mental disorders: a comprehensive cross-phenotype analysis

Background The close relationships between mental disorders and suicidality are frequently seen in epidemiology. Shared genetic liabilities and brain structure variation may underlie these associations. Therefore, we aimed to investigate the phenotypic and polygenetic associations between multiple mental disorders and different levels of suicidality, as well as mediation by grey morphology and white matter tracts.

Methods Using raw data from the UK Biobank (UKB) European population, we first evaluated the phenotypic and polygenic relationships between 12 mental disorders and gradient scales of suicidality. We then accessed data from the All of Us (AoU) diverse cohort to replicate findings in European and African American populations. Demographic and social factors, including age, sex, BMI, education, deprivation, income, smoking, and chronic pain were included as covariates. Second, we used existing genome-wide association study (GWAS) summary statistics from 12 major mental conditions to estimate genetic correlations and identify pleiotropic genes using a combination of statistical genetics tools. Third, we further explored the potential mediation effects of brain structure on the relationship between mental disorders and suicidality through structural equation modeling and Mendelian randomization analyses.

Results In the UKB European population, 150,861 eligible individuals were retained after standard GWAS quality control. Nine out of 12 mental disorders showed both significant phenotypic and polygenic correlations with gradient suicidality (P_bonferroni <0.05). Using GWAS summary statistics, we also observed positive global and regional genetic correlations between the 12 mental disorders and suicidality (r_g ranging from 0.25 to 0.68, P_bonferroni <0.05). Across pairs of suicidality and other mental disorders, we identified 73 out of 136 pleiotropic functional genes (including 58 novel ones associated with suicidality) shared by two or more pairs. These genes were enriched in pathways including regulation of immune system process and DNA, nucleosome and chromatin organization, and phenotypes as common mental disorders and brain morphology. Finally, the association between mental disorders and suicidality was significantly mediated by several structural brain imaging features.

Conclusion This study underscores the urgent need to address the shared and distinct genetic architecture of suicidality and its related mental conditions. Combining longitudinal population-level biobanks with disease-ascertained GWAS data is warranted to further enhance our understanding of this complex phenomenon. Our research findings will guide future suicide prevention and precise treatment among individuals with or without major mental disorders.