The DSM-5 consolidates many of the historical diagnostic understandings of autism, but it does so at the possible risk of excluding people with milder forms of the disorder. 

The Earliest Understanding of Autism

As the understanding of autism spectrum disorder has grown, so has the way autism has been defined and diagnosed. When the disorder was first categorized in 1943 by Leo Kanner, an Austrian-American psychiatrist, it was thought to be a type of childhood schizophrenia that developed from neglectful parenting.

Over the next few decades, autism was reclassified as a set of developmental disorders (still erroneously believed to be the result of bad parenting) before finally being seen as a spectrum condition with numerous presentations of impairment. 

This evolution is seen in how the Diagnostic and Statistical Manual of Mental Disorders (DSM), produced by the American Psychiatric Association (APA), has repeatedly revised its diagnostic definition of autism spectrum disorder. When the second edition of the DSM was published in 1952, it used Leo Kanner’s prognosis of autism as a deep form of emotional disturbance because of how he noticed that children with autism displayed “extreme aloneness” and an “anxiously obsessive desire for the maintenance of sameness.”

The DSM-II classified autism as a psychiatric condition, calling it a form of childhood schizophrenia with separation from reality. So entrenched was the theory that “pathologically aloof parents” were responsible for their child’s emotional withdrawal that psychologists publicly attributed autism to so-called “refrigerator mothers.”


By the 1970s, the theory of autism being the product of cold and detached parenting had fallen out of favor. More and more research pointed to the roots of the disorder lying in brain development, which suggested a biological factor.

In 1980, the third edition of the DSM was published. It created a separate diagnosis for autism and removed it from schizophrenia, calling autism a “pervasive developmental disorder.”

While DSM-II did not cover the diagnostic process, instead leaving it to a clinician’s own observation and interpretation, DSM-III provided three specific criteria for diagnosing autism:

  • A lack of interest in engaging with other people
  • Severe impairments in communicating
  • Unusual responses to the environment

Autism & PDD-NOS

DSM-III was revised in 1987. The revision significantly changed the criteria for diagnosing autism.
It added a diagnosis for pervasive developmental disordernot otherwise specified (PDD-NOS), which broadened the understanding of what could be considered autism by adding a relatively milder form of the disorder.

The revision also did away with a provision in the first edition of DSM-III, which stated that all the three criteria for diagnosing autism had to be present in the first 30 months of a child’s life. 

The 1987 revision reflected the emerging understanding that autism is not a single condition. Instead, it is a spectrum of conditions that can appear throughout a patient’s life, all without even using the word “spectrum.”

Asperger’s Syndrome

The DSM-III revision provided 16 criteria across the three that had been established in the previous edition. The APA required that eight of the criteria had to be met in order for a clinician to be able to diagnose autism.

The addition of PDD-NOS gave clinicians the freedom to include children who did not meet all the criteria for a diagnosis of autism but who still needed developmental or behavioral support. 

The idea of autism being a spectrum of conditions was not codified until the release of DSM-IV in 1994. In addition to both autism and PDD-NOS, DSM-IV introduced “Asperger’s disorder” (also known as Asperger syndrome), placing it at the milder end of the spectrum. It also introduced childhood disintegrative disorder, which is marked by severe reversals and regressions in development.

DSM-IV also included Rett syndrome, which affected movement and communication (in girls more so than boys). 

The Hunt for the Autism Gene

This diversification of autism in the DSM reflected the trend of research at the time that the development of autism could be traced back to clients’ genetics, and that each respective category of autism would, in time, be associated with specific problems and treatments specific to those problems. 

With this understanding of autism, the 1990s saw researchers working to identify the genes that they believed contributed to the development of autism in the womb. Many studies looked extensively at the human genome, finding hundreds of so-called “autism genes,” but failed to find any that could be exclusively linked to autism.

As time went on, it became evident that the hunt for a genetic driver for autism, and then devising corresponding treatments for the five conditions of autism laid out in DSM-IV, would not be successful. Going back to the drawing board, researchers and geneticists concluded that autism should be recharacterized as a diagnosis in and of itself, with a spectrum that ranged from mild to severe. 

While this was happening, there was scarce uniformity in how clinicians in different states (or even within the same state) would diagnose autism, Asperger syndrome, or PDD-NOS. Some researchers noted that parents lobbying for a diagnosis, sometimes based on the services available in their state and the eligibility criteria thereof, would occasionally lead to a spike in diagnoses. 

How the DSM-5 Diagnoses Autism

With these factors in mind, DSM-5, released in 2013, was the first to use the term “autism spectrum disorder,” characterized by “persistent impairment in reciprocal social communication and social interaction” and “restricted, repetitive patterns of behavior” in early childhood. Each diagnostic group includes specific behaviors that clinicians have to identify to return a positive diagnosis on a client. 

To some controversy, DSM-5 removed Asperger syndrome, PDD-NOS, childhood disintegrative disorder, Rett syndrome, and classic autism from its classifications. In their place, it introduced social communication disorder for children who have only language and social impairments but not the restricted, repetitive patterns of behavior. 

Diagnostic Controversies

The drastic reclassifications of certain presentations of autism surprised many advocates. Some fear that as diagnoses disappear from the DSM, they may lose access to their services or their health insurance.

For example, clients who identified themselves as having Asperger syndrome based on its inclusion in DSM-IV said seeing their symptoms in the manual helped them define their difficulties in ways they were unable to do before. Not only did this give them an explanation for their experiences, it gave them hope for therapy and management of their symptoms. The removal of the diagnosis from DSM-5 was akin to losing their identity. 

Furthermore, specialists debated whether the more stringent diagnostic criteria laid out in DSM-5 would make it harder for clients with milder symptoms to access healthcare that would help them, or whether the stringent criteria would slow the surge in rates of positive diagnoses. 

In the time since DSM-5 came out, some of the fears that people on the more subtle end of the autism spectrum would lose their healthcare have not borne out. However, there is significant research to suggest that the tighter definition of autism does exclude certain categories of people (specifically girls, older people, and patients with milder symptoms) from getting a diagnosis that might help them manage their symptoms of autism.