Adolescent boys with autism spectrum disorder (ASD) may be at higher risk for self-harm, according to results of a study, published in BMC Medicine.

Self-harm among adolescents is 1 of the strongest predictors for future suicide. Approximately 18% of adolescents report self-harming with gender, depression, anxiety, low self-esteem, and impulsivity contributing to risk. A recent study found that a majority of adults (66%) with newly diagnosed ASD endorsed contemplating suicide and 35% had planned to or had attempted suicide.

As self-harm in adolescence predicts future suicide and suicide appears to be more common among adults with ASD, this study investigated whether ASD in adolescence is associated with self-harm. Investigators at King’s College London in the United Kingdom evaluated data from the National Pupil Database of adolescents aged 11-17 who were enrolled in 4 state-maintained school and resident programs from their 11th birthday or April 2009 to presentation at an emergency department (ED) for self-harm, their 18th birthday, or March 2013. Demographic and clinical characteristics were evaluated for self-harm risks using matching and imputation approaches.

Continue Reading

The entire study cohort comprised 56,708 girls and 56,578 boys among whom 437 and 2026 had ASD, respectively. The study population was diverse, in which >60% were non-White, >25% reported English as their second language, 80% lived in deprived neighborhoods, and >25% were eligible for free school meals.

The majority of adolescents with ASD (>75%) attended mainstream schools and 12%-15% had severe or profound learning difficulties.

Adolescents with ASD had numerically higher rates of self-harm among both boys (0.5% vs 0.2%) and girls (<2.3% vs 1.1%) then the adolescents without ASD, respectively.

Girls (n=834) and boys (n=186) who presented with self-harm were aged mean 15.6 (SD, 1.4) and 15.9 (SD, 1.9) years at first presentation, 48.8% and 44.6% had a mental health history, 42.8% and 47.3% were White, 38.4% and 33.9% were in the most deprived national neighborhood deprivation quintile, and 74.0% and 51.1% presented with self-poisoning or overdose, respectively.

Among girls, self-harm was positively associated with hyperkinetic disorder (adjusted hazard ratio [aHR], 3.58; 95% CI, 2.03-6.29; P ≤.01); poor school attendance (aHR, 2.84; 95% CI, 2.70-3.51; P ≤.01); behavioral, emotional, and social special needs (aHR, 2.31; 95% CI, 1.84-2.88; P ≤.01); age (aHR, 1.28; 95% CI, 1.21-1.35; P ≤.01); and receiving free school meals (aHR, 1.22; 95% CI, 1.02-1.48; P ≤.05).

For boys, hyperkinetic disorder (aHR, 8.0; 95% CI, 5.0-12.8; P ≤.01); poor school attendance (aHR, 6.50; 95% CI, 4.24-9.92; P ≤.01); behavioral, emotional, and social special needs (aHR, 3.14; 95% CI, 2.19-4.70; P ≤.01); ASD (aHR, 2.73; 95% CI, 1.47-5.09; P ≤.01); and age (aHR, 1.38; 95% CI, 1.22-1.57; P ≤.01) were positively related with self-harm.

In a sensitivity analysis, ASD remained significantly associated with self-harm among boys.

The major limitation of this study was the small sample size of individuals who self-harmed and had ASD, limiting power.

This was the first longitudinal study to evaluate the intersection between self-harm in adolescents and ASD. The investigators observed that boys with ASD were at an almost 3-fold higher risk for self-harm than boys without ASD. No relationship between self-harm and ASD was observed among girls.

We spoke with Sophie Epstein, Phd; Emily Widnall, MsC, BA; and Johnny Downs, PhD from the Institute of Psychiatry, Psychology and Neuroscience, King’s College London about this study.

What was the motivation behind this study?

We were aware that there is some research in adults showing that those with Autism Spectrum Disorders (ASD) are at higher risk for self-harm than the general population. We also know that rates of self-harm are particularly high in the adolescent age group. But despite this, there has been very little research to examine whether adolescents with ASD are at higher risk for self-harm than the general population of adolescents. We know that adolescents with ASD are a vulnerable group in many ways, including having high rates of a range of mental health difficulties and so we wanted to understand whether this group is at higher risk for self-harm and would benefit from additional support or interventions. By linking routinely collected mental health data with education data, we had access to data representing the whole population of adolescents attending state schools in 4 boroughs in South East London including information on ASD and on presentations to Accident and Emergency (A&E) departments with self-harm. Therefore, we were able, for the first time, to examine the risk for self-harm in adolescents with ASD compared with this very large general population sample.

Were you surprised by the findings?

We knew that in almost all previous studies, rates of self-harm are higher amongst girls than boys. We also knew that a diagnosis of ASD is more common in boys than girls, therefore neither of these findings in our study was surprising. However, it was somewhat surprising to find an association between ASD and self-harm only in boys but not in girls. In the paper, we discuss that the absence of an association in girls could be explained by the fact that ASD is often under-recognized in girls which could have been the case in this sample. This would mean that some girls who did, in fact, have ASD, were counted as part of the group without ASD, resulting in an underestimation of the true effect of ASD on self-harm in girls.

What do the findings of this study mean for our understanding of the role of gender in ASD and/or self-harm?

As already discussed, most previous studies have found higher rates of self-harm in girls than boys, and this study was no exception. In our sample, approximately 1.5% of girls presented to A&E with self-harm and only 0.3% of boys. If we compare boys with ASD to girls with ASD in this study, the 2 groups had similar risks for self-harm. This tells us that both girls and boys with ASD are vulnerable and would be likely to benefit from additional support to prevent the incidence of self-harm or to manage self-harm in those already presenting with this behavior.

Do these results have any implications for potential interventions?

Our findings suggest that young people with ASD may be a particularly vulnerable group when it comes to self-harm. Self-harm prevention interventions could be targeted to groups who are at higher risk and these results suggest that young people with ASD may be 1 of these groups. Interventions could also target the reasons why young people with ASD might be at higher risk for self-harm. For example, it is sometimes the case that co-existing mental health problems are missed, identified late or not treated effectively in those with ASD. Interventions to identify problems earlier or improve treatments for this specific group could help to reduce future risk for self-harm.

What are the next steps in the research of self-harm in ASD?

This study looked at a whole population and compared the risk for self-harm in those with ASD to the risk in the general population. We do not yet know what increases the risk for self-harm for some young people with ASD more than others. We did not have high enough numbers of young people with ASD in this sample to explore this question, but this would be an important area for future research. Other interesting areas could be to look at self-harm more generally, not just that presenting to A&E. This is important because although self-harm presenting to A&E represents the most severe self-harm, overall, only a small proportion of young people who self-harm present to hospital, and many more self-harm, but do not come to the attention of medical services.

Do the results of this study say anything else about self-harm in adolescents beyond ASD?

As well as findings relating to ASD, our study also found other factors which seem to be related to self-harm. These include absence or exclusion from school, where those with less than 80% attendance at school had a 3 times higher risk of presenting to A&E with self-harm compared with those with more than 80% attendance. Although we do not yet fully understand the reasons for this finding, it does suggest that young people with poor attendance at school are another particularly vulnerable group. We also found that those with a diagnosis of ADHD were about 4 times more likely to self-harm than those without ADHD and there has been very little previous research examining self-harm in those with this diagnosis. Another important aspect of this study is the method and data used. By linking mental health and education data, we had access to a very large sample of whole-population data, collected routinely by health and education services. This type of study helps to make sure all groups, including those who might be less likely to respond to a survey, are represented.

Disclosure: An author declared affiliations with industry. Please refer to the original article for a full list of disclosures.


Widnall E, Epstein S, Polling C, et al. Autism spectrum disorders as a risk factor for adolescent self-harm: a retrospective cohort study of 113,286 young people in the UK. BMC Med. 2022;20(1):137. doi:10.1186/s12916-022-02329-w

This article originally appeared on Psychiatry Advisor