Commentary Takes Issue with Criticism of New Autism Definition

DSM-5 Experts Call Study Flawed

ARLINGTON, Va. – Several recent articles incorrectly state that new diagnostic criteria for autism spectrum disorder (ASD) in the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) could reduce the number of diagnosed individuals by approximately thirty percent. The latest paper is simply a meta-analysis of several previous papers with no new data being presented. Members of the DSM Neurodevelopmental Disorders Work Group have analyzed the studies on which this claim is based and found that there were serious flaws in the analysis of the data.

The Work Group members state: “We believe the archived data used in a number of these analyses have too many inherent limitations to assess the criteria proposed for the DSM-5, particularly in regard to sensitivity and specificity.” Those limitations stem from the study samples dating to 1994 and the restrictive way data from that sample were collected and evaluated. They make any legitimate review and comparison virtually impossible and do not justify claims that the number of children diagnosed with ASD under DSM-IV-TR criteria would not qualify for services under DSM-5 criteria.”

In DSM-5, Asperger‘s disorder, pervasive developmental disorder (not otherwise specified) and childhood disintegrative disorder were consolidated within the overarching category of ASD. The change signals how symptoms of these disorders represent a continuum from mild to severe, rather than being distinct disorders. The new category is expected to help clinicians more accurately diagnose people with relevant symptoms and behaviors by recognizing the differences from person to person, instead of providing general labels that tend not to be consistently applied across different clinics and centers.

Developing more useful diagnostic criteria for clinicians and individuals with ASD was the core objective of the Work Group‘s efforts. For example, the new measures indicate increased sensitivity in regard to age of onset. DSM-IV required functioning delays to be present prior to age 3; DSM-5 criteria extended this period until social demands exceed limited capacities, as long as symptoms were present in early childhood. Despite what some critics have suggested, the issue of containing autism rates was not considered by the Work Group, nor was it a factor in revising the criteria.

David Kupfer, M.D., chair of the DSM-5 Task Force, praised the Work Group for extraordinarily thorough, thoughtful and detailed work. “We remain open to any concerns the academic and advocacy communities might have, but we strongly support the decisions that these leading researchers and clinicians have made,” Kupfer said. “The proposed ASD criteria are backed by the scientific evidence.”

DSM-5, which is published by the American Psychiatric Association, is used by health care professionals in the United States and much of the world as the authoritative guide to the diagnosis of mental disorders. DSM-5 was released in May 2013, culminating a 14-year revision process.

The American Psychiatric Association is a national medical specialty society whose more than 35,000 physician members specialize in the diagnosis, treatment, prevention and research of mental illnesses, including substance use disorders. Visit the APA at




Media Contacts

Erin Connors, 202-609-7113

Press Line, 202-459-9732