Prevalence and correlates of suicidal ideation and suicide attempts in preadolescent children: A US population-based study

Suicide is a leading cause of death among youth, yet little is known about predictors of suicide risk in preadolescent children. Existing data suggest suicide attempts and deaths in this age group have not decreased and may be rising. Prior nationally representative studies typically focused on adolescents, non-US samples, or only suicide deaths, leaving a gap in understanding suicidal ideation and attempts in US preadolescents. Early suicidal thoughts and behaviors are clinically important and may portend poorer outcomes. Limited, non-representative studies suggest associations with economic disadvantage and possibly female sex, with mixed findings and small samples; prior ABCD analyses found higher ideation in males but did not disentangle ideation from attempts or examine racial/ethnic categories separately. Diagnostically, preadolescent suicidality has been linked to psychiatric disorders, especially comorbidity; beyond depression, anxiety, disruptive behavior disorders, and ADHD may be relevant, but no nationally representative data have examined which diagnoses predict ideation versus attempts. Additionally, it is unknown what proportion of preadolescents with suicidality access treatment and which factors predict care. This study aims to provide nationally representative estimates of prevalence and to identify sociodemographic and diagnostic predictors of suicidal ideation and attempts in US preadolescents, as well as predictors of treatment utilization among suicidal youth.

Prior work indicates adolescent females often show higher suicidal ideation and attempt rates than males, but evidence is mixed in preadolescents. Some studies found associations between preadolescent suicidality and economic disadvantage and female sex, while others found no sex differences. Earlier ABCD analyses reported higher ideation (but not attempts) among males and no racial/ethnic differences, though racial/ethnic identities were collapsed and ideation conflated with attempts. Diagnostic risk in preadolescents includes depression and possibly anxiety and disruptive behavior disorders; comorbidity appears particularly potent. Population-based adolescent data show many suicidal youths do not receive mental health care, but analogous data for preadolescents are lacking. The current study builds on ABCD findings by separately modeling ideation and attempts, disaggregating sociodemographic categories, and examining both diagnostic predictors and treatment utilization.

Design and sample: Cross-sectional analysis of the ABCD Study baseline dataset (release 2.0.1), collected 2016–2018. The ABCD Study is a longitudinal, multi-site, multimethod, multi-informant cohort following US children for 10 years. The analytic sample included 11,875 children aged 9–10 at enrollment, recruited at 22 sites to approximate national sociodemographic variation (age, sex, race, SES, urbanicity). Weighting procedures were applied to yield nationally representative estimates. Measures: Suicidal ideation and behavior were assessed via computerized K-SADS-PL DSM-5 administered to both child and parent/guardian, covering passive/active ideation and attempts (current: past 2 weeks; lifetime). Following clinical convention, a child was coded positive for lifetime ideation or lifetime attempt if either informant endorsed relevant items. For analyses, two contrasts were defined: (1) pure suicidal ideation (lifetime ideation without any lifetime attempt) vs. no ideation/attempt; (2) among youth with lifetime ideation, attempt vs. no attempt. Sociodemographics: Parent/guardian reported child sex, race (collapsed to White, Black, Multiracial, Other), ethnicity (Hispanic vs non-Hispanic), family income (five brackets: < $25k, $25–49k, $50–74k, $75–99k, ≥ $100k), parental education (less than high school, high school/GED, some college, college graduate), and marital status (married vs not married). Youth reported sexual orientation (collapsed to gay or bisexual; not gay or bisexual; did not understand the question). Categories were collapsed to ensure adequate cell sizes. Psychiatric diagnoses: K-SADS-PL DSM-5 lifetime diagnoses were coded positive if criteria met by parent, child, or both, either currently or in the past. Diagnoses included MDD, separation anxiety, social anxiety, specific phobia, GAD, OCD, PTSD, conduct disorder, ODD, ADHD; eating disorders collapsed; any psychotic disorder collapsed. Treatment utilization: Parents reported whether the child had ever received any mental health or substance use services (yes/no; "not sure" excluded). Due to sparse data, specific treatment types were not modeled. Statistical analysis: Weighted cross-tabulations estimated lifetime and current prevalence of ideation, attempts, diagnoses, and treatment utilization overall and within subgroups. Bivariate logistic regressions tested associations of sociodemographics and diagnoses with outcomes. The Benjamini–Hochberg procedure controlled for multiple comparisons in bivariate analyses. Multivariate logistic regressions modeled: (a) all sociodemographics together; (b) any diagnosis vs none; (c) number of diagnoses (none, single, two or more); and (d) individual diagnoses simultaneously, each controlling for all sociodemographics. Separate models were run for outcomes: pure ideation (excluding attempters), attempt among ideators, and any treatment utilization among youth with pure ideation. Results reported as odds ratios (ORs) with 95% CIs. Analyses performed in SPSS 27 with survey weights to account for complex sampling.

Prevalence: - Lifetime suicidal ideation: 14.33% (SE 0.37); lifetime suicide attempt: 1.26% (SE 0.12). - Current suicidal ideation: 3.62% (SE 0.20); current attempt: 0.26% (SE 0.06). - Lifetime pure ideation (ideation without attempt): 13.15% (SE 0.35). - Among those with lifetime ideation, 8.70% (SE 0.82) reported a lifetime attempt. Sociodemographic predictors of pure ideation (multivariate): - Lower odds: female (OR 0.71, 95% CI 0.62–0.81); Hispanic (OR 0.73, 0.60–0.87); Black (OR 0.77, 0.62–0.95); parental education high school/GED (OR 0.78, 0.62–0.97). - Higher odds: sexual minority (OR 3.81, 2.49–5.83); multiracial (OR 1.39, 1.13–1.70); parents not married (OR 1.30, 1.10–1.53); family income $25–49k (OR 1.29, 1.03–1.63) and $50–74k (OR 1.39, 1.15–1.69) vs ≥$100k. Diagnostic predictors of pure ideation (multivariate): - Any psychiatric disorder (OR 2.67, 2.32–3.08); two or more disorders (OR 3.79, 3.24–4.43) vs none. - Individual disorders: MDD (OR 4.47, 3.54–5.64), GAD (OR 1.85, 1.41–2.44), conduct disorder (OR 1.78, 1.28–2.47), ODD (OR 2.00, 1.65–2.41), ADHD (OR 1.43, 1.20–1.69). Sociodemographic predictors of attempt among ideators (multivariate): - Higher odds: sexual minority (OR 2.55, 1.02–6.38); lowest income < $25k (OR 2.86, 1.31–6.26) and $25–49k (OR 3.08, 1.62–5.86) vs ≥$100k. Diagnostic predictors of attempt among ideators (multivariate): - Number of diagnoses: two or more (OR 2.17, 1.26–3.72) vs none; single disorder not significant. - Individual diagnoses: social anxiety (OR 2.39, 1.25–4.58) increased odds; separation anxiety associated with lower odds (OR 0.47, 0.24–0.90). MDD not significant for attempts among ideators. Treatment utilization: - Any lifetime treatment predicted by lifetime ideation (OR 3.48, 3.03–3.99) and attempts (OR 2.30, 1.53–3.45). - Among youth with pure ideation, 34.59% (SE 1.39) received treatment; among ideators with attempts, 54.82% (SE 4.91) received treatment. - Sociodemographic predictors of treatment among youth with pure ideation (multivariate): lower odds for females (OR 0.63, 0.48–0.83), Black (OR 0.57, 0.37–0.89), Hispanic (OR 0.65, 0.43–0.97), parental education high school/GED (OR 0.45, 0.27–0.75); higher odds for lowest income < $25k (OR 1.84, 1.11–3.05) and parents not married (OR 1.58, 1.13–2.21). - Diagnostic predictors of treatment among youth with pure ideation (multivariate): one diagnosis (OR 4.89, 2.98–8.02), two or more (OR 13.84, 8.73–21.93); specific diagnoses associated with higher odds included social anxiety (OR 1.91, 1.09–3.34), GAD (OR 3.08, 1.96–4.84), ODD (OR 2.33, 1.65–3.29), and ADHD (OR 2.11, 1.53–2.92).

This nationally representative study shows that suicidal ideation is common in US preadolescents, while attempts are less frequent but still clinically significant. Unlike patterns typical in adolescence, female sex was associated with lower odds of ideation, and sex did not predict attempts, underscoring the need to assess suicidality in young males. Sexual minority status and lower family income markedly increased risk for attempts among ideators, highlighting socio-structural vulnerabilities. Comorbidity emerged as a robust correlate of both ideation and attempts, with a clear dose–response effect; this aligns with theories positing that greater psychological pain and impairment heighten suicide risk. MDD strongly predicted ideation but did not distinguish ideators who attempted, suggesting depression relates more to the emergence of suicidal thoughts than to the transition to action in this age group. Externalizing disorders (conduct disorder, ODD, ADHD) were associated with ideation but not attempts, suggesting a potential developmental pathway where ideation manifests earlier with externalizing psychopathology and behavioral enactment may emerge later. Social anxiety disorder uniquely predicted attempts among ideators, whereas separation anxiety was protective. Treatment engagement was low relative to adolescent benchmarks, with notable disparities: females, Black, and Hispanic youth with ideation were less likely to receive care. Findings emphasize the need for early, targeted prevention, routine screening in pediatric and school settings, and efforts to improve equitable access to services, particularly for sexual minority, low-income, and racially/ethnically minoritized youth and those with comorbid conditions.

In a large, nationally representative sample of US 9–10-year-olds, lifetime suicidal ideation affected roughly one in seven, and about one in twelve ideators reported a lifetime attempt. Risk concentrated among sexual minority youth, multiracial youth (for ideation), and those from lower-income families (for attempts), with psychiatric comorbidity substantially elevating risk. MDD was the strongest predictor of ideation, whereas social anxiety and diagnostic comorbidity predicted attempts. Treatment utilization among suicidal preadolescents was suboptimal and marked by disparities for females, Black, and Hispanic youth. The study contributes critical baseline epidemiology and risk stratification for preadolescent suicidality and highlights targets for prevention and intervention. Future research should leverage longitudinal ABCD follow-ups to establish temporal relationships and developmental pathways, examine more granular sociodemographic identities, and evaluate the types, duration, and effectiveness of treatments to inform evidence-based care pathways.

Key limitations include: (1) cross-sectional design precluding causal inference and temporality; (2) necessity to collapse certain sociodemographic (e.g., race, sexual orientation) and diagnostic categories due to small cell sizes, limiting fine-grained subgroup analyses; (3) treatment utilization was examined as any lifetime service use without detail on modality, duration, timing, or effectiveness; and (4) although suicidality was assessed via structured interview with parent and child reports combined to maximize detection, potential reporting discrepancies remain inherent to multi-informant assessments.