If you’re on the autism spectrum, like me chances are your formal diagnosis  has changed over the years.

The shift from distinguishing  things like Aspergers to grouping everyone under the same  Autism Spectrum Disorder umbrella happened fairly recently. Maybe you adopted “Autism  Spectrum Disorder” with open arms.

Or maybe you continue to use the terms  you’ve been using from the beginning. Different people have different preferences, and there’s no one right way to talk about it. Even among the people coming up with these terms!

In the last few years, some researchers have started  pedaling back from the decision to smush all of autism into one big group. They suggest that we should once more  refer to subcategories of autism. But not the old ones.

This time, they’re proposing four  groups based on not just behavior, but on genes and age of diagnosis. And it might help people find  the right kind of support. Here’s a new way to think about autism. [intro music] The way one person describes their  autism diagnosis and themselves will be different from how someone else  with the same experience describes it.

In this video, we'll use the  terms "Autism Spectrum Disorder" and identity first language  like "Autistic People," but please know we respect whatever terms  you choose to use to describe yourself. And some people refer to  autistic symptoms as "traits," because "symptoms" might imply there's always something  wrong that needs to be fixed. In fact, many autistic people feel  they don't need to be "fixed," and neurodiversity is fine actually.

But we're using the word "symptoms" because these are things that can  be challenging for autistic people, and having appropriate support  can help them live happier lives. The papers we're talking about  today refer to them as "symptoms." So with that established, let’s  talk about this new research. In my lifetime, we’ve seen professionals  differentiate autistic people based on the extent of their social  communication and repetitive behaviors.

We’ve also seen a broader  acceptance of these conditions under one umbrella of Autism Spectrum Disorder. But not long after Autism Spectrum Disorder was written into the Diagnostic and  Statistical Manual of Mental Disorders, some researchers began saying we  should split it back up into subgroups. Now, they’re not suggesting  that we go back to the way it was before we unified under one roof.

Instead, they’re proposing new flavors of autism  based on what scientists believe today. After combing through diagnostic tests and behavioral reports from more than  5,000 people diagnosed with autism, a group of researchers in 2025  identified these categories: The largest portion of their  sample fell into a group they called SocialBehavioral. A little more than a third of the  individuals had this flavor of autism.

This group was united by their challenges with communication, attention, and anxiety, but not significant developmental delays. In addition to their autism, they often had ADHD, anxiety, or major depression. Those bonus diagnoses are called comorbidities.  And they were common among people  in the SocialBehavioral group, but not so much in the next group, who were said to experience Moderate challenges.

This was the second largest group, encompassing about one third  of the sample population. People in the Moderate autism group also didn’t have developmental delays. And compared to the SocialBehavioral group, they had far fewer difficulties with  the social and behavioral stuff.

Which you might have guessed from the name. To be clear, compared to someone who  hasn’t been diagnosed with autism, people in the Moderate group  still had more symptoms. But they weren’t experiencing  it all to the extremes found in other autism sub-categories.

Like the next category: Mixed autism spectrum disorder  with developmental delays. Yeah, they could have workshopped  this title a bit more, but what it lacks in brevity  it makes up for in clarity. The almost 20% of the sample  population that made up this group had a mixture of what the other groups  had, with added developmental delays.

This group often had intellectual  disabilities, language delays, and movement disorders. But did not usually have the  other comorbid diagnoses, like ADHD or anxiety, that can come along for the ride  in the SocialBehavioral group. Which brings us to the least  common of the four categories, with 10% of the sample population.

They  were categorized as Broadly affected. People in the Broadly affected  group experienced it all. And this time, when I say they experienced it all, I mean including the comorbidities.

So we have the whole spectrum broken down into SocialBehavioral, Moderate,  Mixed, and Broadly affected. Like every new concept, these categories started out as an  idea based on a limited set of data. But as the researchers sought  to test their categories, they found validation everywhere they looked.

Parent reports and clinicians  supported these classifications. And a totally different sample  population of autistic people also fell into the same categories. So it doesn’t just describe these individuals.

The evidence for these categories  isn't just behavioral, either. The researchers found genetic differences among the four groups that support the  establishment of the new categories. But before we get into that,  all science needs funding.

So please excuse this ad Thanks for watching SciShow! We’ve heard you loud and clear that  the SciShow audience loves rocks. That’s why we keep making new Rocks Box videos highlighting another  geological marvel every month.

And to go along with those videos, we send Rocks Box subscribers a  sample of whatever mineral or fossil we talked about in that month’s video, right to your door. you could get chalcopyrite or enhydro agate there's water in there, it's really cool or spoiler alert, rainbow moonstone The Rocks Box is so popular that we’ve had a wait list in the past  because we can only source so much of each rock. But right now, we have rocks to spare! So if you’ve been hoping to get your  very own Rocks Box, now’s the time!

Find the monthly Rocks Box  subscription, individual samples, and Rocks Box merch at Complexly.store/rocks. So please excuse this ad. Members of each of the four groups had  genetic profiles that resembled each other and were unlike that of  people in the other groups.

So if you only looked at everyone’s genes, you’d have a good chance of  putting them in the same categories as if you were basing it  on their diagnostic tests. For example, people in the SocialBehavioral  group generally had genetic variants that other researchers have already  linked to ADHD and depression. And since people in that category also  tend to rack up those additional diagnoses, the researchers see that as support for  grouping a portion of autistic people into the SocialBehavioral category.

Those in the Moderate group had some variation in what researchers refer  to as less essential genes. Remember, these are people who  tend to demonstrate mild symptoms. There’s still something going on there, but these genetic variations aren’t  associated with super severe symptoms.

People in the Mixed group had genetic variation that’s much less common if you looked  for it in any other population. … Which also lines up with this being  one of the less common categories. And in these cases, we’re not talking  one or two letters of DNA swapped. They had multiple genetic  variations going on at once.

And people in the Broadly affected group had the most of what are called  high-impact non-inherited variants, though they had a lot of inherited variants, too. This includes stuff like the genes  and proteins involved in fragile X, which is known for developmental  delays, intellectual disabilities, and feelings like hyperactivity and aggression. That’s pretty much exactly how the  Broadly affected group was described.

So the researchers point to the genetic evidence as further support for these four categories. It might also help explain why people who fell into different categories were  generally diagnosed at different ages. Now, the age that you’re diagnosed  depends on a lot of factors.

Too often, people are living  with symptoms for years before getting a diagnosis due to social,  cultural, financial, or other reasons. But symptoms can appear at  different ages for different people. And sometimes, there’s a genetic connection with when your symptoms become more noticeable.

Maybe you’ve heard the relatively new term AuDHD. It describes anyone with both autism and  attention-deficithyperactivity disorder, and a lot of those folks are diagnosed as adults. including me There’s a significant genetic correlation between later-diagnosed autism  and comorbidities like that. … Even if it doesn’t fully explain what’s going on. On the other side of the age range, people in the Broadly  affected and Mixed categories are likely to be diagnosed earlier in life, likely because of the developmental  delays characteristic of those categories.

If you’re not making eye  contact, have low muscle tone, or other physical displays of autism that  a parent might recognize in a toddler, they can bring it up with a  pediatrician when you’re young. And again, people in those Mixed  and Broadly affected categories often have genetic variations that  coincide with early diagnosis. A person’s genes, the way they navigate the world, and the age at which autism became apparent all provide some support  for this “four categories of autism” idea.

And with these new categories, people might even have an  easier time accessing therapies that work best for them. We know from a survey answered  by 200 autistic adults without intellectual disabilities that  finding the right support is crucial. Some of them shared what it was like to  work with mental health professionals who didn’t get what their  flavor of autism was all about.

Often the results were life-threatening. I’m about to talk about suicide. So please skip to the time code on screen if that discussion would be unhealthy for you.

These specific individuals  with autism were surveyed because they’re in a population that we know is at risk for contemplating and dying by suicide. So it’s not representative of  everyone holding an autism diagnosis. But it’s a valuable sample for  conversations around suicide.

Despite their risk, more than half of the people surveyed reported unmet mental health  or social support needs. And those who did get help said  it literally saved their lives. But getting the right kind  of support turns out to be as important as getting it at all.

Survey participants reported that getting therapy that turned  out to be inappropriate for them was disempowering and made them feel worse. If you’re in the SocialBehavioral group, but you work with someone whose area of expertise is more aligned with the Broadly affected group, it could lead to feelings of  hopelessness and incapacity. Nobody wants that.

And having these new categories as  shared language or reference points could help people avoid those mismatches. Now, I said “could” because it’s  definitely not an inevitability. Introducing new labels always comes with risks.

If your new classification feels less aligned with who you are than how you  previously described yourself, it could just reopen the same can  of worms we’ve been trying to close. Time will tell if this one  does more good than bad, or if it even catches on. There are other researchers who have  recently divided the autism spectrum into completely different groups.

And no one knows which classification system will stand the test of time. Plus, remember how I said those researchers found validation everywhere they looked? Well, they didn't look everywhere.

Like, you know, asking autistic folks  if those new labels seemed appropriate. Ideally a new classification system would have the stamp of approval  from folks it will directly impact. As the statistician George Box once said, “all models are wrong, but some are useful.” [ outro ]