Polygenic Risk for Suicidality in Youth With and at Clinical High-risk for Bipolar Disorder

Suicide is the second leading cause of death among youth. Bipolar disorder (BD), particularly early-onset BD, is associated with significantly elevated risk for suicide. Genetic factors play an important role in suicide risk in other psychiatric disorders, but studies have yet to examine genetic factors underlying suicide risk in youth BD. We therefore examined the polygenic risk associated with suicide-related phenotypes among youth with and at clinical high-risk (HR) for BD. Participants included 278 (n = 146 BD; n = 132 HR) White youth ages 13 to 20 years. HR participants primarily included youth who were clinically referred for BD assessment and/or treatment but did not meet the diagnostic criteria for BD. Lifetime suicidal ideation, nonsuicidal self-injury (NSSI), and suicide attempts (SA) were collected via standardized interviews. Deoxyribonucleic acid (DNA) was extracted from saliva samples genotyped on Infinium Global Screening Arrays. Polygenic risk scores (PRS) were based on summary statistics from published genome-wide association studies for BD, schizophrenia, MDD, SA, and risk taking. Logistic regression was used to analyze PRS with suicidal ideation (SI), self-harm (SA and/or NSSI), and SA with sex and first 2 population principal components as covariates. Lifetime rates of SI were 63%, 61% for self-harm, and 21% for SA specifically. MDD PRS was associated with SI (PRS continuous shrinkage [PRSCS]: β = .39; standard error [SE] = 0.16; p = .017); however, this did not remain significant after correction for testing multiple PRS and multiple suicide-related outcomes. None of the suicide-related phenotypes were significantly associated with PRS for BD, schizophrenia, SA, or risk taking in the current sample. We found a nominal association in our sample, whereby that higher PRS for MDD was associated with SI. A number of factors may explain the limited findings: 1) all of the PRSs were derived from adult studies, which may not extend to youth; 2) given their age, participants still face the risk of future suicidality; 3) PRS is limited to common single nucleotide polymorphisms (SNPs), and excludes many potential genetic markers of interest; and 4) the sample size is small. Despite these limitations and the tentative nature of our results, these findings provide encouragement for follow-up in future larger samples of youth with BD.