Mechanisms Underlying Thoughts of Self-Harm and Depression in Autism Spectrum Disorder: Findings from a Nationally Representative Sample

Background: Across the lifespan people on the autism spectrum experience extremely high rates of co-morbid psychopathology, the most common being anxiety and depression (Bellini, 2006; Hofvander et al., 2009; Lever & Geurts, 2016; Uljarevic et al., 2017), and sleep disorders (Humphreys et al., 2013; Richdale & Schreck, 2009). The high rate of depression in the autistic community is particularly concerning as this factor alone is likely to place many individuals at heightened risk of suicide (Hedley, Uljarević, Wilmot, Richdale, & Dissanayake, 2017). Indeed, recent research suggests risk for suicide, including suicidal ideation and death, is elevated in this population (Cassidy et al., 2014; Hirvikoski et al., 2016). However, and although there is emerging and promising work in this area, more theoretical work is needed to identify the mechanisms that might lead to increased risk of poor mental health outcomes, including depression, suicidal ideation and behavior, in the autistic population.

Objectives: The aim of the present study was to examine the associations between loneliness and social support, ASD trait severity, and depression and thoughts of self-harm in an Australian nationally representative sample of people on the autism spectrum.

Methods: Participants were 185 individuals (92 female; M_age = 37.11, SD = 15.41 years) who were participating in one of two nationwide longitudinal studies (school leavers, adults) into autism. Participants self-reported a formal diagnosis of ASD and, for inclusion in the present study, were required to return a score > 65 on the AQ-Short. Participants also completed the University of California Los Angeles Loneliness Scale-Short Form (ULS-8) (Hays & DiMatteo, 1987), the Social Support Questionnaire-Shortened Version (SSQ-6) (Sarason, Sarason, Shearin, & Pierce, 1987), and the Patient Health Questionnaire-9 (PHQ-9) (Kroenke, Spitzer, & Williams, 2001), which includes a question concerning self-harm.

Results: Forty-nine percent of participants returned scores in the clinical range for depression and 36% reported recent suicidal ideation. Females, comprising almost 50% of the sample, returned higher depression scores than males, however no differences were identified between males and females in terms of self-harm. Regression analyses revealed that loneliness, satisfaction with social support and ASD traits predicted depression scores, and satisfaction with social support and depression predicted thoughts of self-harm. Path analysis showed that ASD trait severity was independently related to depression, the effect of number of social supports on depression was mediated by loneliness and satisfaction with social support, and that effects of loneliness and satisfaction with social support on self-harm were mediated by depression. The pattern of relationships was nearly identical for males and females.

Conclusions: This study identified high rates of depression and self-harm in a nationally representative Australian population sample. The finding that ASD traits independently predict depression suggests that autism may present a risk factor for the development and maintenance of psychiatric disturbance. Our model further suggests that loneliness and social support operate respectively as protective and risk factors for depression and self-harm in autism.