Autism Spectrum Conditions in People Who Died By Suicide in the UK.

Background: There is a significantly increased risk of suicidal ideation, suicidal behaviours (Cassidy et al. 2014) and death by suicide (Hirvikoski et al. 2015) in people with Autism Spectrum Conditions (ASC) compared to the general population. However, previous research has focused on prevalence of suicidality in clinic-based ASC samples. No research has yet explored suicide in general population samples, or compared the characteristics of suicide in those with and without ASC. This is key to understanding and thus preventing suicide among those with ASC.

Objectives: 1) To establish whether ASC diagnoses are over-represented among people who died by suicide in the UK; and 2) To compare the characteristics of suicide in those with and without ASC.

Methods: The first wave of data collection from a UK psychological autopsy study is presented. Coroners’ inquest records for the period 2014-2016, ruling a suicide or open verdict, were requested from two UK cities. These records were scrutinised for evidence of ASC diagnoses; Definite Diagnosis (where the individual had a formal diagnosis of an ASC); Strong Evidence (where two independent judges agreed that the individual met all the criteria for ASC but had no formal diagnosis); Probable Diagnosis (where the two independent judges agree there are many signs, but there are gaps in the evidence); and No Evidence (where the person had no clear signs of ASC, even if other diagnoses were present and which were noted). Characteristics of those who died by suicide were also noted (gender, age at death, previous suicide attempts, suicidal ideation, mental and physical health conditions, environmental stressors, recent access to services).

Results: 219 coroners inquest records were assessed, 153 of which were ruled a likely suicide according to ICD-10 criteria. Of these, 11.76% had evidence of ASC, significantly higher than the rate of ASC in the general population in the UK (1%); 0.7% were classed as Definite Diagnosis; 1.3% as Strong Evidence; and 9.8% as Probable Diagnosis. There were no significant differences between the groups in terms of age, gender, number of physical or mental health conditions, presence of environmental stressors or previous suicidal ideation/attempts.

Conclusions: A significant minority of cases of death by suicide in the UK general population show evidence of the person having previously undiagnosed ASC. Preliminary analysis from the first wave of data collection so far suggests that individuals with and without evidence of ASC diagnosis did not significantly differ in terms of their broad characteristics (age at death, gender, number of physical or mental health conditions, presence of environmental stressors, likelihood of experiencing suicidal ideation or attempts, and recent access to services). Considering that many adults remain undiagnosed with ASC (Lai and Baron-Cohen, 2015), it is important for clinicians to be aware of increased suicide risk in ASC, and where ASC diagnosis is suspected, consider screening for suicidal thoughts and behaviours. These results will inform training for medical and allied professionals in an effort to improve suicide prevention, and highlight ASC as a particularly vulnerable group in the population.