Increases in autism prevalence: real or illusory?

This post was authored by one of our colleagues, Zeph Turner, who is a Clinician with Vermont Wellness, a collaborative of Vermont Therapists.

Throughout this article, I use identity-first language (“autistic person”) instead of person-first language (“person with autism”), because identity-first language reflects the preferences of a majority of autistic adults (Taboas et al., 2023).

Across the U.S., families and care providers have been concerned about an apparent rise in cases of autism spectrum disorder (ASD), sometimes framed as an “epidemic” of ASD. The U.S. Department of Education reports that in the 10 years from 1993 to 2003, the nationwide rate of autism diagnoses rose by 657% (Arkowitz & Lilienfeld, 2012). These numbers can seem concerning: is there something causing a rise in the number of autistic people in each generation? Does this really mean that autism is getting more common? Or is something else going on?

There are several reasons why the prevalence of a certain diagnosis, like autism, can seem to increase over time. Firstly, the condition could really become more common, with more people experiencing the symptoms than before. For example, rates of anxiety disorders increased during the onset of the COVID-19 pandemic (Mental Disorders, 2022), probably because people felt anxious about COVID. In other words, people really did get more anxious, and that was why the numbers of diagnosed anxiety disorders went up.

But that’s not the only reason why the prevalence of a particular diagnosis can increase. An increase in prevalence could also reflect a change in diagnostic criteria for that diagnosis, “lowering the bar” for a person to be diagnosed. As an example, consider the diagnosis of post-traumatic stress disorder (PTSD). Early conceptualizations of PTSD focused on traumatic events that occurred during wartime. The syndrome went through a number of names when it was first invented, most of which referred to combat: “shell shock” (referring to blast waves from exploding artillery shells), “war neuroses”, or “Combat Stress Reaction” (Friedman, 2024). As scientific understanding of PTSD increased, it became clear that PTSD symptoms also occur in people who experience traumatic events outside of combat, such as sexual assault. The diagnostic criteria for PTSD were updated to reflect this. When PTSD was considered a wartime disorder, only veterans or soldiers could be diagnosed with it. But when the diagnosis was updated to include other traumatic events, many people who had never been in combat became eligible for a diagnosis of PTSD, so the prevalence of the disorder went up–not because more people were being traumatized and developing PTSD symptoms, but because the definition of PTSD was broadened to include more people who had already been experiencing symptoms of trauma.

An increase in the prevalence of a disorder could also be due to an increase in awareness of the disorder among providers or clients. As more providers learn about a disorder, they may be more likely to recognize and diagnose it in their clients, when before the disorder may have been ignored or labeled with another diagnosis instead. And as clients and the general public become more aware of a disorder, they may be more likely to recognize it in themselves or their loved ones, and look for a provider who can diagnose the condition.

In the case of ASD, many of these factors have been at play. The diagnosis of autism was first applied in 1943 to a man named Donald Triplett, who passed away at age 89 in 2023 (Mandy, 2023). We know that diagnosis rates before 1943 must have been zero, so they have to have gone up at some point! Since that first diagnosis, it has taken time for clinicians to learn about the diagnosis of autism and be trained in recognizing clients who may be autistic. In recent decades, national-scale media campaigns have attempted to raise awareness of autism spectrum disorder. Increased awareness has likely spurred parents, care providers, and autistic people themselves to look into the possibility of an autism diagnosis, when historically they may have simply thought that their child, client, or they themselves were simply introverted, shy, or nerdy (Dillenburger et al., 2013). Some public health campaigns have focused on early identification of autistic children to enable early intervention (Wallis, 2021). These campaigns have promoted universal autism screening in primary care physicians’ offices, which can increase the number of autistic people who are recognized and diagnosed early in life. Alongside generally increased awareness, clinicians have increasingly recognized the way autism can present in women and people of color, who were historically overlooked and underdiagnosed. 

Additionally, the diagnostic criteria for ASD have evolved over time to recognize an increasingly broad range of presentations, like those that were formerly associated with the diagnosis of Asperger’s Syndrome (now considered a part of the autism spectrum umbrella in the DSM-5-TR) (Arkowitz & Lilienfeld, 2012). While the autism diagnosis once focused narrowly on people who had obvious neurodevelopmental differences and a very specific presentation, our broader understanding of ASD today includes many people who may have flown under the radar in past decades and would not have been identified as autistic. These changes can increase the number of people who are diagnosed with autism, even if the number of people with autistic traits remains the same over time.

How can we know whether the apparent increases in the prevalence of autism are due only to these factors, or also reflect a real increase in the number of autistic people in the population? One way is to get a large sample of people of all ages from the general population, and assess all of those people for autism, using a consistent set of diagnostic criteria. Testing a sample of people using consistent diagnostic criteria helps eliminate the effects of increasing autism awareness and changing diagnostic criteria over time. If the older people in the sample have a lower rate of autistic traits than the younger people, that might be a clue that the rate of autistic people is really increasing from generation to generation. However, if the number of autistic people is the same across all ages, that would suggest that the number of autistic people in the population is staying the same over time.

Terry Brugha, a professor of psychiatry at the University of Leicester, did a groundbreaking study of just this type in the UK (Brugha, 2016). Using a random sample of 618 adults from age 16 to 90, Brugha and his team of researchers tested everyone in the sample for autism using age-appropriate diagnostic assessments. His results? The rate of autism was the same for people of all ages in his sample, at about 1.1% using the criteria his team selected. In other words, he found evidence that the number of people born with autistic traits over the past century or so has been relatively steady–there has been no increase in the number of autistic people in the population over time, despite the fact that more people are being diagnosed.

Evidence suggesting that the prevalence of autism in the population has been steady over time can help push back on alarmist and stigmatizing claims that there is an “epidemic” of autism that must be controlled. Instead, we can understand autism as a condition that is now better-recognized and better-understood, but which has likely affected a small fraction of the population for generations, well before the diagnosis was first described by scientists. Although it’s difficult, if not impossible, to know how the prevalence of autism has changed throughout the course of human history, I believe that autistic people have probably always been among humans since prehistory, acting as valued members of society and accepted as part of the incredible diversity of the human species.

References

Arkowitz, H., & Lilienfeld, S. O. (2012, August 1). Is There Really an Autism Epidemic? Scientific American. https://www.scientificamerican.com/article/is-there-really-an-autism-epidemic/

Brugha, T. (2016). Autism and Ageing: Epidemiology and Demographics. In Autism Spectrum Disorder in Mid and Later Life (pp. 334–344). Jessica Kingsley Publishers.

Dillenburger, K., Jordan, J. A., McKerr, L., Devine, P., & Keenan, M. (2013). Awareness and knowledge of autism and autism interventions: A general population survey. Research in Autism Spectrum Disorders, 7(12), 1558–1567.

Friedman, M. J. (2024, May 9). History of PTSD in Veterans: Civil War to DSM-5. National Center for PTSD. https://www.ptsd.va.gov/understand/what/history_ptsd.asp

Mandy, W. (2023). The old and the new way of understanding autistic lives: Reflections on the life of Donald Triplett, the first person diagnosed as autistic. Autism, 27(7), 1853–1855. https://doi.org/10.1177/13623613231194476

Mental disorders. (2022, June 8). World Health Organization. https://www.who.int/News-Room/Fact-Sheets/Detail/Mental-Disorders

Taboas, A., Doepke, K., & Zimmerman, C. (2023). Preferences for identity-first versus person-first language in a US sample of autism stakeholders. Autism, 27(2), 565–570. https://doi.org/10.1177/13623613221130845

Wallis, K. E. (2021). The Roadmap to Early and Equitable Autism Identification. Pediatrics, 148(Supplement 1), 21–24. https://doi.org/10.1542/peds.2021-050693E