Genetic Vulnerability for Preadolescent Suicidal Thoughts and Behaviors

Suicidal thoughts and behaviors (STBs) are increasingly prevalent but understudied in preadolescent children (1). Rates of suicide attempts (SAs) and deaths among preadolescent children have been rising more sharply than in other age groups over the past 10 years in the United States, and suicide is now the second leading cause of death among 10- to 14-year-olds (2). Between 2000 and 2020, there was a 4.5-fold increase in SAs by ingestion in 10- to 12-year-olds (3). However, we understand little about the epidemiology, etiology, and biological underpinnings of STB in this age group.

The Adolescent Brain Cognitive Development (ABCD) Study provides an unparalleled opportunity to address gaps in our understanding of STB in preadolescent children. The ABCD Study is the largest longitudinal study of child health and brain development in the United States, which enrolled 11,880 U.S. children 9 to 10 years of age between 2016 and 2018. The study includes 10 years of serial multi-informant assessments of cognition, development, physical and mental health, social and emotional functioning, culture, and environment; it also includes neuroimaging and biospecimen analyses of genetics, epigenetics, hormones, and environmental exposures. In the ABCD cohort, lifetime prevalence of suicidal ideation (SI) was alarmingly high at 14.33%, with 1.26% reporting previous attempts (1).

In the current issue of Biological Psychiatry, Lee et al. (4) examine how genetic vulnerability to 8 common psychiatric disorders is associated with STBs in preadolescent children from the ABCD cohort (4). They used polygenic risk scores (PRSs)–combinations of genetic variants associated with genetic liability for a specific condition–for attention-deficit/hyperactivity disorder (ADHD), anorexia nervosa, anxiety disorder, autism spectrum disorder, bipolar disorder, major depression (MD), posttraumatic stress disorder, and schizophrenia. The authors measured the association between each of these 8 PRSs and SA, SI, and nonsuicidal self-injury at the baseline and year 1 study visits for 4344 subjects in the ABCD cohort of European ancestry. They found that PRSs for MD were associated with youth-reported SAs at baseline and year 1, and that PRSs for ADHD were associated with youth-reported SI at year 1, after controlling for age, sex, socioeconomic status, family history of suicide, and child psychopathology measured with the Child Behavior Checklist. The variance explained by PRSs for ADHD and MD–though modest–was higher than previous PRS studies for STBs in adults. Importantly, the authors found no significant associations between PRSs and parent-reported SA, SI, or nonsuicidal self-injury; significant associations were found only with youth-reported SA and SI. In addition, while their primary analysis suggested a separate effect for ADHD PRSs on SI and MD PRSs on SA, controlling for family history of suicide and youth psychopathology may be overly conservative and may obscure the considerable overlap between genetic vulnerability to each condition. In fact, their univariate analyses also found an association between ADHD and SAs (in addition to SI) at year 1, as well as an association between MD and SI (in addition to SAs) at year 1.

A complementary article by Joo et al. (5) examined the association between PRSs for psychiatric disorders and STBs in the ABCD sample. These authors found that PRSs for ADHD were significantly associated with SI and SA, corroborating Lee et al.’s results (4). Joo et al. (5) also found a significant association of PRSs for schizophrenia and “general happiness” with STBs in the whole sample, and associations of PRSs for MD, posttraumatic stress disorder, and autism spectrum disorder in addition to ADHD, schizophrenia, and general happiness with STBs in the subsample with European ancestry. The findings of Joo et al. (5) that diverge from those of Lee et al. (4) likely relate to a difference in predictors (PRSs measuring 24 psychiatric traits), outcomes (combined parent and child report of SI and SA), and sample (no exclusion of participants based on ancestry in the overall analysis and a larger European ancestry sample [N = 5718]).

Other studies of STBs with data from the ABCD cohort are beginning to provide further insights into preadolescent suicidality. For example, Lawrence et al. (1) identified ADHD and MD as diagnostic risk factors for SI (in addition to oppositional defiant disorder, conduct disorder, and generalized anxiety disorder) (1). In a study on neurocognition and STBs, lifetime parent-reported STBs were associated with poorer episodic memory, which is known to be associated with attentional deficits (6). Shoval et al. (7) reported an association between externalizing symptoms and STBs and found that ADHD medication use was protective against STBs, particularly for children with more externalizing symptoms.

Taken together, these results support an association of both ADHD and MD with STBs in preadolescent children. This is consistent with research showing that both MD and ADHD are heritable conditions and are associated with STBs throughout the lifespan, including in preadolescent children (1). ADHD is the earliest onset psychiatric disorder, often beginning in the preschool years, and was the most common psychiatric diagnosis among preadolescent suicide decedents in the United States from 2003 to 2012 (8). Lee et al. (4) found that PRSs for both MD and ADHD were associated with a broad range of externalizing psychopathology, highlighting the overlapping psychopathological impact of these two sets of PRSs in preadolescence. Thus, children with ADHD or other externalizing disorders may be a particularly compelling clinical population to target for early suicide prevention efforts. The study by Lee et al. (4) furthers our understanding of how ADHD relates to STBs in preadolescent children. Previously, some researchers found that comorbidities such as MD mediate the relationship between ADHD and SI (9), and others have suggested that impulsivity in ADHD contributes to the conversion of SI to behavior in the presence of comorbidity (8). On the contrary, Lee et al. (4) found a relationship of PRSs for ADHD with both SI and SAs. More research is needed to better understand the specific psychological, cognitive, and functional Impairments that contribute to the association between ADHD and STBs. Also compelling is that the associations between PRSs for ADHD/MD and STBs were found only when using the youth report data. Other researchers have reported low concordance between youth-reported and parent-reported STBs (1), and this study further highlights that STBs in this age group may often go undetected by parents. Thus, efforts to screen or evaluate for STBs in preadolescent children should include both child and parent reports.

There are a few limitations to this study. First, only children in the ABCD cohort of European ancestry were included (36% of the sample). This was methodologically sound because the genome-wide association study data used to identify PRSs are based on Individuals of European ancestry and avoided false positive associations related to mixing subpopulations with different allele frequencies. However, this subsample had significantly higher levels of parental education, household Income, and married parents, and It excluded Black youth, who have higher rates of STBs in this sample (10) and higher rates of death by suicide nationwide (8). These results may not translate to populations that are most impacted by preadolescent suicide in the United States. The authors appropriately note the importance of creating genome-wide association study datasets for populations underrepresented in genetics research. Finally, as noted above, controlling for family history of suicide and child psychopathology may be overly conservative. Genetic risk contributes to the association between child and parent STBs, and psychopathology likely partially mediates the effects of ADHD and MDD PRSs on STBs.

Despite these limitations, Lee et al. (4) contribute a valuable study using this large population-based sample of 9- and 10-year-olds in the United States, showing that genetic predisposition to ADHD and MD is significantly associated with SI and SAs in this sample of patients of European ancestry. Although addressing the lack of diversity should be a priority for genetics research, their results are important given how little is known to guide suicide preventive efforts for preadolescent children. More research Is needed to better understand additional environmental and clinical factors that contribute to STBs In children with ADHD and MD across racial and ethnic groups, and how best to intervene. One enticing lead from another ABCD study Is that in youth with externalizing and ADHD symptoms, treatment with a stimulant was protective against STBs (7). This supports the possibility that early identification and treatment of both ADHD and MD could reduce STBs in young children.