The Autism Spectrum Diagnosis Might Not Last — Should DSM-6 Break it Apart?

If you've been following autism research for any length of time, you know the diagnosis has been anything but stable. In 2013, the DSM-5 made a massive change: it swept away separate diagnoses like Asperger's syndrome, autistic disorder, and PDD-NOS, replacing them all with a single umbrella — Autism Spectrum Disorder (ASD). That was supposed to simplify things, to recognize that autism is, well, a spectrum.

But a growing body of research — and a new SciShow deep dive — is asking a provocative question: Did we go too far? And should the next DSM break the spectrum back apart?

The Autism Spectrum Diagnosis Might Not Last

July 14, 2026 — The story of autism diagnosis is a story of science trying to catch up with reality. When Leo Kanner first described "early infantile autism" in 1943, and Hans Asperger described a similar but distinct condition the following year, the assumption was that these were two different disorders. That split held for decades. In the DSM-IV (1994), you had five different autism-related diagnoses: autistic disorder, Asperger's disorder, childhood disintegrative disorder, Rett syndrome, and PDD-NOS (pervasive developmental disorder not otherwise specified).

The Great Unification of 2013

The DSM-5 task force made a bold argument: these categories weren't as distinct as clinicians thought. The genetic and neurological evidence suggested overlapping mechanisms. A child diagnosed with Asperger's and a child diagnosed with "classic autism" often shared more similarities than differences — especially when you accounted for the wide range of support needs. So they collapsed all of it into Autism Spectrum Disorder, with three severity levels (Level 1: requiring support, Level 2: requiring substantial support, Level 3: requiring very substantial support).

Proponents argued this would reduce misdiagnosis, eliminate the confusing boundary between Asperger's and "high-functioning autism," and focus treatment on actual support needs rather than labels.

What's Wrong With the Spectrum

Here's the problem — and it's one that autism researchers and autistic self-advocates have been pointing out for years. The spectrum metaphor is misleading. Imagine a color spectrum: every point is a slightly different shade. Autism isn't like that. The "spectrum" of autism is more like a constellation — different clusters of traits, genetic markers, and neurological patterns that don't line up on a single axis from "mild" to "severe."

A 2019 study in Biological Psychiatry analyzed the genetic data of thousands of autistic individuals and found something striking: there are distinct genetic subtypes that correlate with different clinical presentations. Some subtypes are strongly linked to language delays. Others are associated with co-occurring intellectual disability. Still others map to the profile historically called Asperger's — strong language skills but profound social communication challenges.

The problem with the single ASD label is that it lumps all of these together. A clinical trial for a medication targeting one autism subtype might fail because the participants had a different subtype altogether. School interventions designed for one group may not work for another. And families searching for answers are told their child has "autism spectrum disorder" — which tells them almost nothing about what kind of support their child actually needs.

What the Research Now Shows

Since the DSM-5 was published, the evidence has only grown stronger that "autism" is not one condition but many. A 2023 Nature Genetics study identified 33 new autism-associated genes, bringing the total to over 200 — and they don't all point to the same biological pathway. Some affect synaptic pruning. Others affect mitochondrial function. Some affect immune regulation.

In other words, we're probably looking at multiple different neurodevelopmental conditions that happen to share some surface-level features. Calling them all "autism spectrum disorder" is like calling pneumonia, lung cancer, and a broken rib "chest pain syndrome" — technically true in terms of one symptom, but completely useless for treatment and prognosis.

The APA's Future DSM Strategic Committee, chaired by Dr. Maria Oquendo, published a series of roadmap papers in January 2026. While they haven't committed to any specific changes for DSM-6 (targeting a 2029 release), the papers explicitly call for incorporating genetic and biomarker data into diagnostic criteria. That's a clear signal: the next DSM may move beyond behavior-only diagnosis.

Should We Bring Back Asperger's?

This is the emotional heart of the debate. For many people, losing the Asperger's diagnosis in 2013 was painful — it erased a community identity. The term "Aspie" had become a source of pride for many self-advocates. On the other hand, Hans Asperger's collaboration with the Nazi euthanasia program has made the name itself deeply problematic.

The question isn't whether to bring back the old names. It's whether we need more precise diagnostic categories — whatever we call them. Some researchers have proposed "autisms" (plural) as a framework. Others suggest specifiers that capture the genetic subtype, developmental trajectory, and co-occurring conditions. Still others argue for a dimensional approach — not categories at all, but profiles across multiple domains: language, social communication, sensory processing, cognitive flexibility, and motor skills.

The autistic community itself is divided. Some argue that breaking the spectrum back apart could stigmatize certain subgroups. Others argue that the current system already fails the people with the highest support needs — who are invisible under the "spectrum" label, lumped in with self-advocates who need very different accommodations.

What This Means for Families

If you're a parent of a newly diagnosed child — or an adult seeking a diagnosis — this uncertainty can feel overwhelming. But here's the good news: the research is moving toward precision, not away from it. We're getting better at understanding what's actually going on in an individual's brain, not just assigning a label. Genetic testing, brain imaging, and biomarker panels are increasingly part of the diagnostic process.

Some advocacy organizations, including the Autism Science Foundation, have called for a "reconceptualization" of autism that separates the condition into biologically defined subtypes. The goal isn't to create new stigma — it's to create better interventions. A treatment that works for one autism subtype might be useless or even harmful for another. Knowing the subtype means knowing the right path forward.

The Bottom Line

The autism spectrum diagnosis was an improvement over the chaos of the DSM-IV era. But it may have been a transitional step — not the final destination. As genetic and neurological research advances, the DSM-6 is likely to look very different from the DSM-5. Whether that means separate categories, dimensional profiles, or something entirely new, one thing is clear: the era of treating autism as a single spectrum is probably coming to an end.

Science moves forward. And sometimes that means admitting that a perfectly reasonable solution from 2013 needs to be revisited in 2026. The autism diagnosis might not last — and that might actually be a good thing.

By Allan Ali, Health & Science Correspondent