Toddlers With Autism: Is That Early Diagnosis Reliable?

F. Perry Wilson, MD, MSCE


May 01, 2019

Welcome to Impact Factor, your weekly dose of commentary on a new medical study. I'm Perry Wilson.

This week, I was struck by a study that may usher in a new era of autism therapy. And no, it's not a study of a drug, psychotherapy, or even a new test. It's a study[1] appearing in JAMA Pediatrics that shows the stability of early autism diagnoses.

Although that concept may not sound exciting, it should, because it opens a ton of doors. Researchers at the University of California San Diego (UCSD) reported on 1269 toddlers seen in their autism evaluation center.

About 75% of the kids were referred as part of an early screening program that occurred at well-child visits, and the remainder were referred because of parental or provider concerns.

In the United States, the typical child with autism spectrum disorder is diagnosed at age 4, but the amount of brain development that has occurred by that time is staggering. Can we diagnose autism earlier? And if so, how stable is that diagnosis over time? Can we really diagnose a child with autism at 12 months of age? On subsequent evaluation, how many turn out to be developing typically?

The answer surprised me.


Of 400 kids diagnosed with autism spectrum disorder at a median age of 17 months, just seven (2%) were found to have typical development on subsequent evaluations; 84% continued to have a diagnosis of autism spectrum disorder. A few transitioned into other categories, such as autism features or language delay, but not many.

Not all early diagnoses were so stable. The most unstable diagnosis was language delay, which had resolved in nearly 50% of children on subsequent evaluations, which should be reassuring to parents of young kids whose language is a bit behind.


But these findings should be equally encouraging to those of us who care deeply about the treatment of autism spectrum disorder. The remarkable stability of the diagnosis means that we can comfortably begin treatment earlier without worrying that we are directing scarce resources to kids who will end up being just fine without them. Earlier treatment, beginning as early as 12 months, may really change the landscape of the disorder.

There are, of course, some caveats here.

First of all, the UCSD autism evaluation center is staffed with incredibly highly trained experts. Their ability to diagnose autism in the very young may not be generalizable to the rest of the medical community.

Additionally, they were not blinded to the original diagnosis on subsequent evaluations. We can imagine that when you walk into a room to see a kid whom you have been told carries an autism diagnosis, you may be primed to make that diagnosis again. That said, initial diagnoses of developmental delay and language delay were not nearly as stable, so maybe the anchoring bias isn't quite as firm as intuition suggests.

In any case, I'm excited about this study. It's an example of what we can do. With good resources and good training, we may institute a new paradigm for autism treatment based upon early, accurate, and stable diagnoses.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.