A large, multicenter study suggests that the proposed new definition of autism would exclude relatively few children diagnosed under current criteria. The report appears today in the American Journal of Psychiatry.

Importantly, the investigators used the most recent proposed criteria for diagnosing autism spectrum disorder (ASD). Once finalized, the new criteria will become part of the 5th edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5). It is scheduled for publication next spring.

Healthcare providers currently use the DSM-IV to diagnose ASD. A panel of experts has been revising the manual’s ASD criteria with the goal of better identifying affected individuals.

Understandably, many worry that a too-narrow definition will exclude children and adults who need autism-related services. The current ASD criteria include the presence of three types of symptoms: social impairment, communication difficulties and repetitive behaviors.  The new system will focus on just two symptom categories: social communication impairments and repetitive behaviors. 

Psychologist Catherine Lord, Ph.D., of New York’s Weill Cornell Medical College, led the study. The investigators also included Autism Speaks Weatherstone Fellow Vanessa Hus, M.Sc.

Dr. Lord and her team reviewed the case records of 4,453 children previously diagnosed with ASD using the DSM-IV system. They also reviewed the records of 690 children with non-ASD conditions such as language disorder. Based on these records, they tried to gauge how well the proposed DSM-5 criteria identified children with ASD and excluded those with other disorders.

According to their analysis, the proposed new criteria produced fewer “false positives” than the old definition of ASD. That is, fewer children who did not have ASD were incorrectly diagnosed as being affected. The DSM-5 criteria also identified 91 percent of those diagnosed under the DSM-IV system.

The 9 percent of children who would have lost their diagnosis under the new criteria did so primarily because their social impairments were not severe enough to meet the DSM-5 criteria. The report did not include adults, and it remains unclear how the proposed changes will affect them. 

“In general, the results suggest that the proposed DMS-5 definition is doing a good job at identifying and distinguishing children with ASD,” says Autism Speaks Chief Science Officer Geraldine Dawson, Ph.D. “The major limitation is that they were based on analysis of case files, not the diagnosis of actual children by clinicians.  To understand the real-word effects of the DSM-5, we need the results of field trials. We also need studies of adults.”

Autism Speaks is currently funding such a field trial. Investigators at the Medical University of South Carolina are comparing the effects of DSM-IV versus DSM-5 criteria in diagnosing ASD in grade-schoolers. In addition, Autism Speaks is funding a large retrospective study comparing the effects of the two sets of criteria.

“Autism Speaks will continue to play a proactive role in monitoring the effect of the planned revisions,” Dr. Dawson adds.  “We are dedicated to ensuring that all who struggle with autism symptoms continue to receive the treatment and services they need.”