Share this article on:

This Way In: Are the proposed criteria for diagnosing autism spectrum disorders too restrictive?

Rukovets, Olga

doi: 10.1097/01.NNN.0000415684.84037.60
Departments: The Waiting Room

This Way In: Are the new criteria for diagnosing autism too restrictive?

A proposed change in the way autism spectrum disorders (ASDs) are diagnosed has some doctors, patients, and parents concerned that specialized services in schools and institutions may be compromised.

The American Psychiatric Association has proposed in its fifth edition of Diagnostic and Statistical Manual of Mental Disorders (DSM)—the reference manual for mental health disorders in the United States—that the subcategories of ASD be eliminated. As a result, the new manual would subsume Asperger's syndrome and pervasive developmental disorder-not otherwise specified (PDD-NOS) under the single category of autism spectrum disorder (ASD).

Ginny Kelly, who serves on the board of the Autism Society of Acadiana in Louisiana, worries that her 18-year-old son, Johnathan, who has Asperger's, would not be considered on the autism spectrum under the new criteria—which could affect his eligibility for services. “Why would we want to dismiss a tangible diagnosis when we're finally beginning to make strides in treatment?” Kelly asks. (The American Psychiatric Association will be inviting public comment during the spring of this year—go to for more information.)

Back to Top | Article Outline




Autism is a developmental disorder that appears in the first three years of life, affecting the brain's normal development of social and communication skills. Asperger's syndrome is often considered a high-functioning form of autism. PDD-NOS is an ASD that does not fully meet the criteria for autism or Asperger's syndrome.

The changes to the DSM are largely a response to concerns that the current criteria are too broad, resulting in overdiagnosis and misdiagnosis of children who either do not have ASD or have other disorders with similar symptoms, such as ADHD, obsessive-compulsive disorder, and social anxiety disorder. Misdiagnosis can be especially harmful when children are put on inappropriate medications.

But some experts believe the proposed criteria are too restrictive and would lead to the loss of services for higher-functioning individuals who have ASD, such as those with Asperger's syndrome. The new criteria do provide fewer ways to diagnose someone in terms of combinations of symptoms. (Visit to see a full comparison of the criteria, and go to for the American Academy of Neurology's guidelines on screening and diagnosing of autism.)

Fred R. Volkmar, M.D., Irving B. Harris Professor in the Child Study Center and Professor of Pediatrics, Psychiatry, and Psychology and chief of child psychiatry at Yale-New Haven Children's Hospital, says that changing the criteria could make it more difficult to interpret study results. “If you start changing the diagnostic criteria too much, understanding the results of clinical trials becomes more complicated because you're looking at different populations,” says Dr. Volkmar, who resigned from the DSM-5 Committee because he disagreed with the process.

Back to Top | Article Outline


According to Rebecca Landa, Ph.D., director of the Center for Autism & Related Disorders at Kennedy Krieger Institute in Baltimore, MD, the new criteria are mostly a reorganization of the current ones. “They provide more information about symptoms and enable health care professionals to rate ASD from mild to very severe. Importantly, the new criteria eliminate some of the confusion that results when, for example, one doctor diagnoses a child with PDD-NOS, and another diagnoses the same child with autism,” Dr. Landa says.

Max Wiznitzer, M.D., an associate professor of pediatric neurology at Case Western Reserve University in Cleveland, OH, and a Fellow of the American Academy of Neurology, agrees. He states that the DSM-5 eliminates subgroups solely for the sake of clarity, not to exclude anyone from diagnosis. “Although we know differences exist in ASD, where we draw the lines between these subgroups causes confusion,” Dr. Wiznitzer says. Because ASD is a disorder that often passes through many different hands—neurologists, psychiatrists, psychologists, school counselors—it is very important to have a detailed and uniform system of diagnosis, he says. Ultimately, though, the diagnostic criteria are only as good as the specialist using them, stresses Dr. Wiznitzer.

Back to Top | Article Outline


An analysis published online in the Journal of the American Academy of Child & Adolescent Psychiatry in November 2011 suggests that the proposed criteria are better at ensuring that young people without an ASD are not mistakenly diagnosed with one.

Using a large Internet-based registry, researchers compared caregiver-reported symptoms of 14,744 siblings (8,911 with ASD and 5,863 without ASD) from two questionnaires in order to test different statistical models for diagnosing ASD. The best model had two categories—people with autism and people without autism. “Furthermore, the ‘autism’ category had only two symptom dimensions: one was social communication and interaction, and the other was repetitive behavior. This is in keeping with the general structure of the proposed [DSM-5] diagnostic criteria,” says lead study author Thomas W. Frazier, Ph.D., of the Cleveland Clinic Children's Hospital Center for Autism and Center for Pediatric Behavioral Health in Cleveland, OH. (In contrast, the current criteria have three symptom dimensions: social communication, social interaction, and restriction/repetitive behavior.)

However, when Dr. Frazier and colleagues analyzed caregiver-reported symptoms associated specifically with high-functioning ASD, they found that the proposed criteria could exclude anywhere from 20 percent to 40 percent of people with very high-functioning ASD, he says. “If you have folks who no longer meet the diagnostic criteria for ASD, then the question becomes: Would they meet the criteria for something else? And would that something else be sufficient for them to get the services that they need?” Dr. Frazier asks.

According to Dr. Wiznitzer, the database that Dr. Frazier used cannot accurately measure potential exclusion because it contains caregiver-reported data without the validation of an ASD expert—and because the symptoms for high-functioning ASD used in the analysis were imprecise. Dr. Wiznitzer says the take-away message of the study is that the new criteria will lead to more accurate diagnosis.

For an extended conversation on the proposed revision, listen to a podcast with Dr. Thomas W. Frazier at

©2012 American Academy of Neurology