Suicidal tendencies hard to spot in some people with autism

To effectively screen for suicidality in people with autism, we need to learn how to ask questions that lead to real answers.

By Katherine Gotham
26 September 2017 | 7 min read
Illustration of people sitting at a party with looming shadows.
Illustration by Cinyee Chiu
Illustration by Cinyee Chiu
This article is more than five years old.
Neuroscience—and science in general—is constantly evolving, so older articles may contain information or theories that have been reevaluated since their original publication date.

Last year, I met a young man with autism who participated in a study at my lab. He did not meet the criteria for depression, and no one in my team would have guessed, based on our interactions with him, that he thought about ending his life. ­­But as part of our research protocol, we asked him directly whether he had thoughts of suicide. Only then did we learn that he had taken action on plans to end his life on two previous occasions.

Later in the interview he told us that if he did not meet a certain highly unrealistic career goal within a specific amount of time, he planned to kill himself. He repeated this often, sometimes matter-of-factly and other times almost defensively, as if challenging us to dissuade him.

Being well

Perspectives on quality of life in autism

After spending a day with him in the lab, I would characterize this man as angry, frustrated and dejected. But none of this was apparent during his first few hours with us, and I doubt that it would have been obvious during a brief or routine screen for suicidality.

These and other experiences have led me to conclude that to effectively screen for suicidality in people with autism, we need to learn how to ask questions that lead to real answers.

The statistics around suicide in people with autism are startling: Up to 50 percent of adults with the condition have considered ending their own lives, a rate two to three times that seen in the general population1.

Adults newly diagnosed with autism are particularly at risk, having spent decades feeling atypical without knowing why2. Rates of suicide attempts and death are also elevated among individuals on the spectrum3,4.

In the general population, we often think of suicidality as going hand in hand with depression. Although depression does emerge as the leading predictor of suicidality in people with autism, there is mounting evidence that a substantial proportion of people on the spectrum who contemplate suicide would not meet the criteria for depression3,5,6.

Pattern recognition:

Risk factors for suicide in the general population span sociodemographic and cognitive vulnerabilities, such as hopelessness or impulsivity; environmental stressors, such as financial or legal problems; and mental health conditions — most notably depression, bipolar disorder, schizophrenia and substance abuse.

The man who visited our clinic was young, unmarried, underemployed and disabled (if you consider autism a disability). All of these features are risk factors for suicidality in the general population. They also describe a large proportion of adults on the autism spectrum.

Our participant was not depressed, manic or psychotic. What’s more, he had clear life goals and enjoyed discussing them.

It was hard for us to measure his degree of hopelessness because it wavered with his black-and-white thinking: He sometimes spoke with certainty that he would reach his career goal and resolve all of his perceived problems; at other times, he doubted this outcome and seemed discouraged by life.

Anecdotal evidence suggests that some people with autism approach suicide practically, even dispassionately — as a valid plan when nothing has worked to help them fit into this world. Others struggle with rigid thinking and poor impulse control, making them susceptible to suicidal tendencies during difficult situations or negative moods. These two profiles are at opposite poles in terms of planning and passion, and yet this young man seemed to fit in both boxes. Still, we could not spot his elevated suicide risk before a lengthy mental health interview.

As a field, we have learned about emotional distress in people with autism through studies that fall under many topic headings: quality of life, emotion regulation, depression, anxiety, access to services and more. Communicating across these research areas is a crucial first step to learning how to recognize people with autism who are at high risk for suicide.

Several clinical research initiatives have led the call for this type of multidisciplinary conversation. Sarah Cassidy and Jacqueline Rodgers, respectively from Coventry and Newcastle universities in the United Kingdom, were instrumental in organizing an international summit on suicide in autism, hosted in May; the same team founded a special-interest group on suicide in autism that convened at the 2016 and 2017 annual meetings of the International Society for Autism Research.

These gatherings set specific goals: to develop tools for assessing suicide risk, identify risk factors and protective factors, and identify strategies for suicide intervention and prevention.

Start by asking:

No screening tool is perfect. Suicide screens designed for the general population tend to err on the side of caution, yielding a high rate of false positives — that is, some people who screen positive do not represent a threat to their own lives.

It is possible that existing tools will be even less specific among people with autism, but we don’t have the luxury of waiting for better ones. There are steps we can take now to assess people with autism who may be at risk.

When a person with autism mentions suicidal thoughts or tendencies, either verbally or on a questionnaire, we should always ask in a calm, unapologetic and non-reactionary manner if she has thoughts about hurting or killing herself. We must take care not to tacitly assume, pass judgment or make it easy to prevaricate. In other words, do not say, “You’ve probably never thought about suicide, right? I just have to ask this.”

At a minimum, we should assess the frequency and intensity of her suicidal thoughts, the presence of a plan to act on them, access to lethal means and history of attempts. We should also note any substance use, medication changes or side effects, and any risk factors related to autism, such as impulsivity, repetitive thoughts about suicide, bullying or social isolation. And we should ask about what supports she has and her reasons to live.

If we are still even moderately concerned about suicide after this assessment, we should directly enlist family support for her (with her consent, in the case of adults). We should work with her to create a safety plan that she can refer to during times in which she is more likely to think or act on suicidal thoughts. Safety plans commonly list coping strategies and contact information for supportive family members and friends, as well as mental health services.

A blog by Autism Speaks board member Sallie Bernard, titled “8 Critical Measures to Counter Suicide,” is a quick and potentially life-saving read. Another resource is a guide from the Suicide Prevention Resource Center that describes suicide screening considerations in the general population. 

We can make the educated guess that, for every person on the spectrum who contemplates ending his life, there are even greater numbers struggling with anger, dejection, hopelessness and other precursors to suicidal ideation. Until we establish tools to screen for suicidality and emotional health problems in this population, let’s use our training and common sense to flag and support individuals at risk in the autism community.

Katherine Gotham is assistant professor of psychiatry at Vanderbilt University in Nashville, Tennessee. 

  1. Segers M. and J. Rawana Autism Res. 7, 507-521 (2014) PubMed
  2. Cassidy S. et al. Lancet Psychiatry 1, 142-147 (2014) PubMed
  3. Cassidy S. and J. Rodgers Lancet Psychiatry 4, e11 (2017) PubMed
  4. Hirvikoski T. et al. Br. J. Psychiatry 208, 232-238 (2016) PubMed
  5. Horowitz L.M. et al. J. Autism Dev. Disord. Epub ahead of print (2017) PubMed
  6. Mayes S.D. et al. Res. in Autism Spect. Disord. 7, 109-119 (2013) Abstract