The Link Between Autism and Violence Isn’t Autism

After some very disturbing shootings involving individuals who reportedly had been diagnosed with an autism spectrum disorder, there has been increased attention to the possible link between autism and violence. As much as many people wanted to push back on this possible association immediately, there were some aspects of autism that at least made such a link theoretically feasible. After all, many individuals with autism do find it difficult to empathize emotionally with other people and some also suffer with aggressive outbursts. Missing in these discussions, however, has been good solid data on the issue, which is why this recent study published online in the Journal of the American Academy of Child and Adolescent Psychiatry, offers some much needed information.

The Stockholm Youth Cohort study examined medical and legal records on all children born in Stockholm county. This enabled them to analyze the official registries of nearly 300,000 individuals between the ages of 15 and 27. These registries record psychiatric diagnoses that are made of residents while convictions of violent crimes are listed on a different database, the Swedish National Crime Registry. Many other variables were also taken into account for this study including sex, age, the presence of other psychiatric disorders, crime history of parents, and family income, among others.

The pattern of results that they found nicely illustrate both 1) why people might think a link between autism and violence exists, and 2) why this conclusion is ultimately much more complicated. In terms of raw numbers, a total of 4.4% of individuals with autism had been convicted of a violent crime versus 2.6% of individuals without autism. Analyzing this in a more statistically comprehensive way, the authors of the study found that there continued to be about a 40% increased risk between autism and conviction for a violent crime event after mathematically accounting for things such as age, sex, and some parental factors. Furthermore, this increased risk appeared stronger among individuals who had autism but who did not meet criteria for an intellectual disability. However, when the researchers also took into account other psychiatric disorders the individuals had, in particular ADHD and Conduct Disorder, the association with autism faded away and became statistically nonsignificant. Indeed, the presence of autism among individuals diagnosed with ADHD or Conduct Disorder tended to reduce the risk of violent crime. (Yes I know some people don’t count autism as a psychiatric disorder but that is another story that you can read more about here.)

The authors concluded that what appears to be a link between autism and violent crime is actually explained by other psychiatric disorders, including ADHD and Conduct Disorder.

For those unfamiliar with the diagnosis, Conduct Disorder is a psychiatric disorder that describes people who tend to frequently break rules and violate the rights of others. In many ways, it is a diagnosis that explicitly describes criminal behavior. At the risk of getting a little technical, it is a bit statistically funky to include essentially criminal behavior as an independent predictor of other criminal behavior but the fact that co-occurring ADHD also diminishes the link between autism and violence lends some additional credence that it is not autism per se that is driving the association with violent crime. About 25% of the autism group also met criteria for ADHD while only about 4% met criteria for conduct disorder.

The bottom line for this study was that it does not appear that autism on its own is a risk factor for becoming violent against other people. This news will likely be highly welcomed by the autism advocacy community which has maintained this view all along and now has some good data to support it. However, the study could raise some concern among other mental health advocacy groups that are already concerned about the stigma associated with the perception that psychiatric disorders are associated with an increased propensity toward violence. Here, it might be useful to pull back from the fancy statistical analyses a bit and keep in mind that 96% of the individuals with an autism diagnosis (with many also meeting criteria for another diagnosis as well) were not convicted of any violent offense.

@copyright by David Rettew, MD

David Rettew is author of Child Temperament: New Thinking About the Boundary Between Traits and Illness and a child psychiatrist in the psychiatry and pediatrics departments at the University of Vermont College of Medicine.

Follow him at @PediPsych and like PediPsych on Facebook.