Links Between Self-injury and Suicidality in Autism

Background: Autistic individuals without intellectual disability are at heightened risk of self-injury, and appear to engage in it for similar reasons as non-autistic people. A wide divergence of autistic perspectives on self-injury, including those who frame it as a helpful coping mechanism, motivate investigating the link between self-injury, suicide ideation, and attempts which has been reported in typically developing individuals.

Method: One hundred three autistic participants completed the Non-Suicidal Self-Injury Assessment Tool (NSSI-AT), the Suicide Behaviors Questionnaire (SBQ-R), and the Interpersonal Social Evaluation List (ISEL-12) across two online studies. Logistic regression was conducted to predict self-harming status via responses to questions on suicidality, and to predict whether certain self-injurious behaviors, including cutting, were especially associated with suicide ideation and attempts. Non-parametric correlation analysis examined relationships between suicide ideation/attempts and other variables that might characterize self-harmers especially at risk of suicidality. These included perceived access to social support, purposes or reasons for self-injury, the number of different self-injurious behaviors engaged in, the duration and lifetime incidence of self-injury, and the individual's feelings about their self-injury.

Results: While self-injuring status was significantly predicted by responses to a question on suicide ideation and attempts, there was no relationship between suicide ideation/attempts and a participant's personal feelings about their self-injury. The method of cutting was also predicted by suicide ideation and attempts, though other methods common in autistic people were at borderline significance. Use of self-injury for the regulation of low-energy emotional states like depression, for self-punishment or deterrence from suicide, and for sensory stimulation, was associated with suicide ideation and attempts, as was the number of self-injurious behaviors engaged in. There was no significant relationship between suicide ideation/attempts and the duration and lifetime incidence of self-injury or social support.

Conclusions: These preliminary data suggest that while individuals might frame their self-injury as a positive or neutral thing, there remains a concerning relationship between self-injury and suicidality which exists regardless of individual feelings on self-injury. This is consistent with the theoretical perspective that self-injury can be a "gateway" through which individuals acquire capability for lethal suicidal behaviors. The data highlight that particular methods (cutting) and reasons for self-injury may be of significant concern, but this information, which might be of extreme value for clinicians, requires further investigation and validation.