Prevalence of and Factors Associated With Self-injurious Thoughts and Behaviors in a Nationally Representative Sample of Preadolescent Children in Great Britain

Introduction

Self-injurious thoughts and behaviors (STIBs) in childhood have received growing recognition as an important clinical concern. Supporting this view, SITBs in childhood are concurrently associated with psychiatric impairment, particularly externalizing disorders,¹ and prospectively predict negative outcomes, including subsequent SITBs¹ and psychiatric hospitalization.² Furthermore, the National Institute of Mental Health has identified childhood suicide as a priority, motivated by the observation that compared to SITBs studies in adolescents and adults, there is a paucity of research with children.³ Several fundamental aspects of childhood SITBs remain unknown, including their epidemiology, with prior studies featuring at-risk or clinical samples.^1,2 The current study characterized the prevalence and correlates SITBs in a nationally representative sample of preadolescents.

Methods

The 1999 and 2004 surveys of Mental Health of Children and Young People in Great Britain were conducted with youth in England, Scotland, and Wales, obtained with permission from the UK Data Archives. For this study, analyses were restricted to individuals under age 13 (unweighted n = 13,126). Participants were 49.35% female, 88.78% white, 5.00% South Asian, 3.34% black, and 2.77% other (mean_age = 8.53 years, SE = .02).

A structured diagnostic interview for epidemiological research to assess for psychiatric disorders, the Development and Well-Being Assessment (DAWBA),⁴ was conducted with children of ages 11 and 12, and with parents of children 5 to 12. The DAWBA included seven questions about SITBs. Queries about self-injurious thoughts covered the last month, and those about self-injurious behaviors the last month and lifetime. Parental psychiatric functioning was also assessed.⁵

Logistic regression was conducted for sociodemographic characteristics, and psychiatric disorders in children, and parental psychiatric functioning in relation to current thoughts of self-injury and lifetime self-injurious behavior, respectively. Sociodemographic characteristics (age, sex, race, and parents’ employment status) were assessed at the bivariate level. Those found to be significant predictors were included as covariates with all remaining predictors in the corresponding multivariate model. To avoid confounding self-injurious thoughts with behaviors (i.e., the possibility that an association with the former is better accounted for by the presence of children with the latter), children with a history of self-injurious behaviors were included in the reference group in analyses of self-injurious thoughts. All analyses were weighted to obtain nationally representative estimates.

Results

In total, 2.18% (SE = .13%) of children experienced past-month self-injurious thoughts without any history of behavior (i.e., “pure” ideators), .80% (SE = .08%) had past-month self-injurious behaviors, and 2.41% (SE = .14%) had lifetime self-injury. Among sociodemographic correlates, only minority race was unrelated to self-injurious thoughts (OR = .81, 95% CI = .54–1.21) or behaviors (OR= .67, 95% CI = .43–1.04).

Multivariate analyses are presented in Table 1. Age, male sex, both parents being unemployed, current parental psychiatric functioning, depression, anxiety, and externalizing disorders, except hyperkinesis, were associated with self-injurious thoughts. All predictors were significant for self-injurious behaviors, except depression and parental employment.

Table 1.

Multivariate analyses of correlates of self-injurious thoughts and behaviors among preadolescent children (unweighted n = 13,126)

Predictors	Current Thoughts of Self-Injurya		Lifetime Self-Injurious Behaviors
	OR (95% CI)	p	OR (95% CI)	p
Age	1.22 (1.15–1.30)	<.001	1.35 (1.27–1.44)	<.001
Sex (Male)	1.35 (1.05–1.75)	.02	1.32 (1.03–1.68)	.03
Parental Employment Status
Both Parents Unemployed	1.61 (1.17–2.21)	<.01	.99 (.71–1.38)	.93
One Parent Unemployed	1.21 (.86–1.69)	.27	1.06 (.77–1.45)	.74
Both Parents Employed	1.00	—	1.00	—
Poor Parental Psychiatric Functioning	1.14 (1.10–1.18)	<.001	1.09 (1.05–1.13)	<.001
Depression	7.87 (2.92–21.21)	<.001	1.92 (.75–4.91)	.17
Anxiety	4.32 (2.89–6.44)	<.001	2.50 (1.64–3.82)	<.001
Oppositional Defiant Disorder and Conduct Disorder	2.45 (1.61–3.72)	<.001	4.74 (3.31–6.80)	<.001
Hyperkinesis	1.15 (.61–2.17)	.66	1.98 (1.15–3.43)	.01

Note: CI = confidence interval; OR = odds ratio

^aChildren with a lifetime history of self-injurious behaviors were included in the reference group to allow for analyses of ideation unconfounded by behavior.

Discussion

This study provides the first analyses of the prevalence as well as sociodemographic and clinical correlates of SITBs in a nationally representative sample of preadolescents, demonstrating that although low base-rate phenomena, SITBs are associated with significant psychiatric comorbidity. Interestingly, SITBs rates were higher among boys in the current and prior studies,¹ contrasting with greater rates of suicidal ideation and attempts among females in adolescence.⁶ Insofar as males are more likely to die by suicide⁶ and earlier onset of psychiatric problems is associated with worse trajectories, whether preadolescents with SITBs are at particular risk of suicide deaths warrants future study. Additionally, whether SITBs in preadolescents in the general population are associated with negative later-life outcomes¹ or have changed in prevalence over time should be evaluated.