Autistic students have, in the past, broken their teacher’s arms, knocked out their teeth, and even given them concussions.  There have been incidents in which students have banged their own heads against walls, scratched their arms until they bled, and bitten their fingers.  There have also been incidents in which their classmates have been injured.  So schools are rightly concerned about autistic kids engaging in behaviors—self-injury, punching, biting, and kicking—that are potentially dangerous to themselves or others.

Nevertheless, schools cannot treat every autistic child as a time-bomb, ready to explode at any moment.  There are certainly some students on the autism spectrum who must be treated with great care, but there are also many who have outbursts only under extreme circumstances, and still others who pose no threat at all.  Unfortunately, the research on the prevalence of aggression in this population remains limited, and what exists has various weaknesses.  Nevertheless, it is worth reviewing, because it shows that “the violent autistic child” is not nearly as common as the general public, as well as many teachers and school administrators, assume.

Estimates of “self-injurious behavior” (SIB), for example, have been skewed by the populations sampled.  One group of researchers looked at 250 children and teens with autism who were enrolled in genetic studies at the Hospital for Sick Children in Toronto, Canada.  They found that 52.3% had engaged in SIB at some point in their life.[1]  This study was often cited in the years after its publication in 2012, and the idea that more than half of autistic kids injured themselves became widely accepted.  In 2016, however, a different group of scholars published the results of their research on more than 8,000 autistic children tracked by the Autism and Developmental Disabilities Monitoring Network in the United States.  They pointed out that the 2012 study, conducted in a hospital, had “over-sampled” kids with challenging behaviors and major impairments.  The 2016 study placed the percentage of autistic kids who self-injured at around 27.7%.  This is still a significant number, but it is only about half that of the earlier, widely-cited study.[2]

Research on aggression against other people has been complicated by disagreements about terminology (the authors of one study noted that other researchers were reluctant even to use the term “aggression”[3]) and weakened by failure to distinguish clearly between the prevalence and persistence of different forms of aggression.  One study, based on a fairly large sample of children, concluded that 68% had at one time or another demonstrated aggression against their care-givers, and 49% had at one time or another been aggressive towards non-caregivers.[4]  It should be noted, however, that these figures covered the children’s entire lifetime, including the period when they were toddlers (who generally tend to do a fair amount of hitting and kicking, even if they are neurotypical.)  When the researchers examined behavior at the time of the study, they found that 56% of the autistic children sampled were “currently” aggressive towards their caregivers, while 32% were aggressive towards non-caregivers.

The authors of this study focused on these general numbers, which they claimed showed that the prevalence of aggression among autistic children was “high.”  However, when they broke down their figures still further, to look at the prevalence of different kinds of violence, it turns out that a much smaller number (35.4% of all the kids in the study) were currently engaged in what the researchers called “definite aggression”—hitting, kicking, punching, etc.  The other children in the “aggressive” category (roughly 25% of the total) were currently practicing only “mild aggression,” defined as playing roughly, verbally threatening other people, or lashing out after being provoked.  Most importantly, 39.8% of the sample showed no aggressive behavior at all.   lt turns out, then, that of the autistic kids in this study, more were currently avoiding all aggressive behaviors than were involved in “definite aggression.”  If we combine the non-aggressive and mildly aggressive categories, it turns out that 65% of the sample studied actually seem pretty similar to “normal” kids.  However, in practice it is quite difficult to know how autistic aggression compares with neurotypical aggression, since studies on aggression in autism generally involve no control group of non-autistic children.[5]

A number of researchers have examined the “risk factors” for self-injurious and aggressive behaviors.  In terms of SIB, one study found that abnormal sensory processing was the most important predictor of self-injury[6]  Other researchers conclude that SIB is particularly common not only in those with abnormal sensory processing, but also those with regressive forms of autism, irritability, hyper-activity, mood issues, sleep issues, and severe communication limitations.[7]  There are some indications that SIB may decline as communication improves over time.[8]  Factors associated with aggression against others include youth (aggressive behavior declines with age among autistic as well as neurotypical children), social and communication problems, higher levels of “repetitive behaviors” (stimming), and—oddly enough—higher family income.[9]  A very high percentage of autistic children and adolescents (50-80%) suffer from sleep problems.  One recent study found a particularly significant correlation between lack of sleep and various problem behaviors, including hyperactivity, irritability, and physical aggression in autistic youth.[10]

The scientific evidence, then, suggests that a significant minority of young people with autism will engage in self-injurious behaviors (27.7%) and significant aggression against others (35.4%).  (A further area of concern is “meltdown” behavior, which I will address in another post.)  According to scientists, the individuals who engage in these behaviors tend to be younger children, those who have gone through early regression, those who are irritable and hyperactive due to poor sleep, those unable to communicate in other ways, and those with the kinds of sensory processing that make the world unpredictable and often painful.  Teachers and administrators would do well to consider and try to mitigate these factors before they condemn autistic children who “act out.”

[1] Emma Duerden, Hannah Oatley, Kathleen Mak-Fan, et al., “Risk Factors Associated with Self-Injurious Behaviors in Children and Adolescents with Autism Spectrum Disorders,” Journal of Autism and Developmental Disorders 42 (2012), 2460-70.

[2] Gnakub Soke, Steven Rosenberg, Richard Hamman, et al., “Brief Report:  Prevalence of Self-Injurious Behaviors among Children with Autism Spectrum Disorder:  A Population-Based Study,” Journal of Autism and Developmental Disorders 46 (2016), 3607-14.

[3] Cristan Farmer and Michael Aman, “Aggressive Behavior in a Sample of Children with Autism Spectrum Disorders,” Research in Autism Spectrum Disorders 5 (2011), 317-23.

[4] Stephen Kanne and Micah Mazurek, “Aggression in Children and Adolescents with ASD:  Prevalence and Risk Factors,” Journal of Autism and Developmental Disorders 41 (2011), 926-37.  The sample was made up of children enrolled in a multi-university research study on autism, which—like the hospital study mentioned above—probably “oversampled” those with challenging behaviors.

[5] There are many studies of aggressive behavior among children who have suffered trauma, who have been raised in poverty, etc.  I have found it difficult to find estimates for aggression among neurotypical children as a whole.  And in any case, different measures are used in studies on autistic and studies on non-autistic children, which makes comparisons virtually impossible.

[6] Emma Duerden, Hannah Oatley, Kathleen Mak-Fan, et al., “Risk Factors Associated with Self-Injurious Behaviors in Children and Adolescents with Autism Spectrum Disorders,” Journal of Autism and Developmental Disorders 42 (2012), 2460-70

[7] G. Soke, S. Rosenberg, R. Hamman, et al., “Factors Associated with Self-Injurious Behaviors in Children with Autism Spectrum Disorders:  Findings from Two Large National Samples,” Journal of Autism and Developmental Disorders 47 (2017), 285-96;

[8] Jeffrey Danforth, “Self-Injurious Behavior (SIB),” in Fred Volkmar, Encyclopedia of Autism Spectrum Disorders (New York:  Springer, 2013), 110-39.

[9]  Stephen Kanne and Micah Mazurek, “Aggression in Children and Adolescents with ASD:  Prevalence and Risk Factors,” Journal of Autism and Developmental Disorders 41 (2011), 926-37.  One might speculate that aggressive behaviors are attributed to factors other than autism in children with lower family incomes.

[10] Micah Mazurek and Kristin Sohl, “Sleep and Behavioral Problems in Children with Autism Spectrum Disorders,” Journal of Autism and Developmental Disorders 46 (2016), 1906-15.