Autism Is Not One Thing: Subtypes and How to Determine Your Child's Subtype

About this episode

In this episode, I'm taking you inside one of the most important shifts happening in autism science right now: subtyping. I walk you through the latest research on why autism is best understood as many different conditions under one umbrella, not a single biology with a single path. We look at a landmark 2025 study from Princeton, published in Nature Genetics, that identified four distinct subtypes of autism across more than 5,000 children, each with shared behavioral profiles and shared genetics. I also cover biological subtypes including mitochondrial dysfunction (which my team at Columbia was the first to identify directly in the brain) and maternal autoantibody-related autism (MAR autism), an immune-driven subtype studied extensively by Dr. Judy Van de Water at UC Davis.

This episode is for you if: you've been told your child "has autism" but no one has helped you understand what kind, your child's profile doesn't seem to match other autistic children you've met and you want to know why, you want to understand the genetics and biology behind your child's development, or you're looking for a science-backed framework that goes beyond one-size-fits-all treatment.

Throughout, I want you to remember: understanding your child's biology is not about finding something wrong. Every child, with autism or without, is their own unique subtype, and the journey we are on is to understand them more deeply so we can support their unique path to thriving.

Transcript

Click a time to play the video from that point. This transcript comes from the episode’s captions and may contain small errors.

Welcome to Magnificent Minds, where we explore the science of child development with both heart and reason, bringing clarity, compassion and scientific evidence to questions that matter deeply to families. Today's episode is about one of the most important shifts happening in autism science right now, and it's this autism is not one thing. It is not one biology. It is not one developmental path. It is not one prognosis. And it should not be approached as if every autistic child needs the same explanation, the same therapy, or the same future forecast.

For years, we've used the single label autism spectrum disorder to describe a very broad range of children and adults. That label has been useful diagnostically. It helps us identify shared features in social communication and repetitive behavior. But when we want to understand causes, biology, prognosis or treatment response that single label starts to break down. And that is exactly where the science is moving now toward subtyping. Today I want to walk you through the latest, most respected research on autism subtypes, including biological subtypes like genetic types of autism and maternal autoantibody related autism or Ma, autism, and developmental and behavioral subtypes that help us understand how different children grow over time.

Because the future of autism care is not one size fits all. It is precision. It is integration, and it is whole child care built around the actual child in front of us. Why does subtyping matter? Because autism is best understood as a recognizable pattern. What we can see in a child's behavior. But it is not a single biological condition. It's helpful to draw a comparison. For example, a fever is something we can recognize and we can identify. We can know when someone has a fever, but it can be caused by so many different things a virus, a bacterial infection, autoimmune conditions.

So we would never treat all fevers the same way without also asking why the fever is happening. Or if we were to think about how we approach something else like cancer. Imagine if we treated all types the same. We would never do that. We know breast cancer is not leukemia. Leukemia is not melanoma. Different biology, different risk factors, different treatments. And so we know from so many aspects of medicine that we need to look deeper. Autism has been treated too often as though it were one condition with one cause and one treatment pathway.

But that's increasingly at odds with the science. Some autistic children have rare genetic syndromes. Some have immune related differences in their biology. Some show a developmental regression. Some have marked motor delays from very early in life and may not walk until age two or older, whereas others are walking and even climbing or running very early. Some have a lot of gastrointestinal and feeding difficulties, whereas others don't have any. Some have mild features of autism, but very significant ADHD or anxiety, or later on, depression.

Others don't have any of those. Some children are diagnosed very early. Others later. And new evidence suggests those may not simply be the same autism discovered at different times. They may reflect partly different genetic profiles. So when families ask what is the best care for my child's autism, the most honest answer is often before we answer that, well, we need to understand which autism we are looking at. Soon, parents will be able to ask what kind of autism or what subtype of autism does my child have?

And as doctors we’ll be able to give them an answer. So today I want to walk you through the latest and most important research findings on autism subtypes. The first is a study published in 2025 by a Princeton led team of researchers. The study was published in an elite scientific journal called Nature Genetics. They analyzed data from more than 5000 autistic children who are part of one of the largest autism research initiatives in the world, called SPARK, which stands for Simons Foundation Powering Autism Research for knowledge.

What they identified are four different subtypes of autism, and children within each subtype share a common behavioral and developmental profiles, and they also share common genetics. So what are the four subtypes? You may be able to see which one your child falls within. The first subtype is called social and behavioral challenges. And a little over one third of autistic children fall into this group. They meet early developmental milestones like walking and talking on time for the most part.

The main differences show up in how they interact socially with others. They often also have anxiety, ADHD, or challenges related to managing their mood. So these children are children whose early development looks mostly on track. But if you look more closely and especially as they develop over time, more differences show up in social interactions and in things like their behavioral flexibility and emotional regulation. The next subtype is called moderate challenges and also affects about one third of autistic children.

Like the other group, they're early developmental milestones like when they start to walk and talk are also largely on time. They tend to be quite independent with their daily life skills. They have mild and subtle features of autism that tend to show up in specific situations, like understanding social cues and handling transitions. In my own mind, I think of this group as moderate social challenges without a lot of other challenges to go with it. The next subtype is called mixed ASD with developmental delay, and about 20% of autistic children fall into this group.

These children are later to talk, walk, or develop other skills. So they're often diagnosed first with developmental delays before they are diagnosed with autism. They have the difficulties with socialization and communication that are characteristic of autism, but they also have difficulties with other aspects of development like cognitive and motor development, for example. And the fourth subtype is called broadly affected. And about 10% of autistic children fall into this group compared to the other groups. These children have the most significant challenges across multiple domains, including sensory, motor, cognitive, language, social, and emotional challenges.

And they also have the highest likelihood of having a number of different medical co-occurring conditions. For example, one is epilepsy. These children often need the most support because many areas of development are affected, like communication, learning behavior, and daily living skills. This does not mean every child fits neatly into just one subtype, or just one subtype for their whole life, even. There actually is quite a bit of overlap between these different four subtypes. So don't worry if your child doesn't seem to fit one.

Not all children will. In terms of genetics, well, children in the moderate challenges and the social behavioral challenges, subtypes tend to have what we call a distributed genetic profile, which simply means there isn't one single genetic difference causing their autism. Instead, there are many small genetic differences, each contributing a little bit. So instead of one big switch being flipped, it's many tiny dials being turned slightly that shape how the brain is developing.

And in fact, those tiny turns of those dials are things a lot of people have. They're called common variants. So a lot of people without autism have them. These common genetic variants, can, kind of come together and have almost a combined effect that can affect brain development and might lead to autism. When you think about these little dials, one might slightly influence attention. One might influence how the brain processes sensory information. Another might influence social motivation. So it's kind of the the combination of all of these different dials that could influence the brain and could lead to autism.

The genetics for the other two subtypes, mixed autism with developmental delay and broadly affected, is more like one big switch that's turned on. These more often have genetic differences that are what we sometimes call high impact, and that alter how the brain develops. So genetic testing is important for all subtypes of autism. But what we expect to see from the testing is different for all. The information is very helpful to have and can influence the care we provide. Just knowing about these four subtypes and their behavioral and genetic profiles puts you in a better position to help your child going forward.

Let me explain a little bit more about how this looks for each of the subtypes. If your child fits best in the social and behavioral challenges subtype, they probably are showing strong language and cognitive skills. They are probably doing well with academic learning and pretty independent in their life skills, but where they may need more support is around social understanding. Things like managing anxiety, regulating their emotions, handling things like transitions and organizing and knowing all of this. You can provide the right supports and all these things often improve a great deal with those right supports in place.

If your child fits the moderate challenges profile, it's possible their challenges were missed or misunderstood for years. So you'll want to be more tuned in to the subtle signs that they're having difficulties. You'll want to look for stress that's emerging as demands increase, especially in school. If your child falls within the mixed ASD with developmental delay subtype or the broadly affected subtype, then you'll know the breadth and intensity of services they need across many service types. Working with many professionals is likely to be more so.

You'll want to have an integrated, multidisciplinary team as early as possible, and you'll want to be working with professionals who understand the full range of co-occurring medical conditions and knows how to test for and monitor and appropriately treat these. You'll want that high level of expert and coordinated care as early as possible. You'll want to get medical tests like genetic and metabolic testing done as early as you can, because the findings could really guide medical care. Next, I want to change directions a little bit. We have talked about genetics, but we haven't yet talked about some of the very exciting and important research on other ways to understand autism subtypes biologically.

The first is one that I researched with my team at Columbia University for many years, and we published in 2014 a research paper that was titled Mitochondrial Dysfunction as a neuro biological subtype of autism spectrum Disorder. This was the first research study to show that a subgroup of individuals with autism, and that group included children and adults, have evidence for mitochondrial dysfunction directly in the brain. We used a type of brain imaging called magnetic resonance spectroscopy and were able, through very advanced imaging technology and analysis, to detect parts of the brain with mitochondrial dysfunction in children and adults with autism.

And we found that about 1 in 8 had this particular finding. Now we looked directly at the brain, but other research studies have looked in the blood or other areas of the body and found an even higher percentage of signs of mitochondrial dysfunction. So this is an important area of biological subtyping for autism because first it's quite common. And second, there are good tests to identify it. And good treatments. The next biological subtype I want to tell you about is one called maternal Autoantibody related autism.

And it's an immune subtype. Maternal autoantibody related autism is sometimes called M-A-R autism or MAR autism. To stand for maternal Autoantibody related autism. Now, this is a subtype of autism in which the child's mother has a specific pattern of antibodies during pregnancy. So there are a number of different patterns that might show up, through an autoimmune process. The mother's body produces antibodies, which can cross through the placenta and then interact with the developing fetal brain.

Now, a number of research studies have been published on this over the past 20 years by the scientist and immunology researcher, Doctor Judy van de water and her team of researchers at the University of California, Davis. What are the important takeaways for you to know? Well, first, this subtype of autism, where the mothers, have these antibodies is more likely in the broadly affected group that we talked about earlier. Second, there is a test that can identify these antibodies in mothers who have a child with autism. And the test can be ordered by a doctor.

Third, there isn't yet a particular treatment for either the mothers or the children with this autism subtype. That's any different than what we would otherwise do, but it might still be helpful in some instances to have this information. And this is something your doctor can guide you with. And it's always, always so important to remember that mothers are never, ever to blame. No parent chooses their own immune biology. This is not about finding fault. It's about understanding as precisely as we can the biological mechanisms so that no stone is left unturned.

And so we can support every child and every family completely. Before we wrap up, I want to touch on another important research study that was published in 2025, in the leading scientific publication called nature. This research looked at autism subtypes through the lens of age at diagnosis. What the study shows us is that sometimes those who are diagnosed later have a different subtype than those who are diagnosed earlier. They truly have different genetics, different biology, and also different behavioral and mental health profiles. Sometimes later diagnosis like age six or above, means something was missed earlier on and we need to try to do better.

But sometimes it doesn't mean something was missed, but rather that a child has a different subtype of autism. Again, the message comes back to autism is not one thing. So what does our understanding of autism subtypes mean for treatment? Well, in plain language, if autism has meaningful subtypes, then treatment should not be organized around one standard package for everyone. That standard package today most often looks like high intensity ABA at 30 to 40 hours per week, but the best science tells us something far more individualized is needed.

Most children need targeted medical workup that may include genetics, EEG, metabolic testing, immune evaluation, sleep assessment, GI assessment. Some may need early intensive language and motor support because their type includes marked developmental delay. Some may need interventions designed around high anxiety, ADHD traits or sensory motor dysregulation, and some may need much more support than others in daily living and adaptive functioning.

This is also why a whole child approach matters so much. So where are we then? If we're taking into account all of the new research on autism subtypes. Well, we are past the point where it makes sense to talk about autism as if it were one biologically uniform condition. But we are not yet at the point where every child can receive a precise subtype label that perfectly predicts their future or their response to any type of therapy. We are in this transitional period and in the transition period. The best approach is a whole child approach.

Autism, we know, is not one thing. It is a broad behavioral umbrella that can arise from so many different biological and developmental pathways. So the goal is to understand which child, which biology, which developmental path, and which supports are more likely to help that child to thrive. That is the future. Not one size fits all autism care, but subtype informed whole child care. If there's one single message I want to leave you with today, it's this no matter which subtype your child falls within, they have ahead of them tremendous possibilities and opportunities for growth, learning, connection with other people, and joy.

And at the end of the day, I think every person really with, with autism or without is their own unique subtype. And the journey we're on is to understand it more and more deeply. And for each of us to become more and more of the person we're meant to be. The insights we are gaining into autism subtypes coming out of the latest research studies like the ones we reviewed today, are helping to make that possible. Thank you so much for being here. My hope is that these conversations help you feel more informed, less overwhelmed, and more connected to the science, to your child, and to a community that's learning and growing together.

If you found today's discussion helpful, please share it with someone, maybe a parent or a professional who you think might benefit. And as always, I appreciate you being part of this journey toward understanding and supporting every child's unique path to thriving. Until next time, take good care of yourself and your family and keep believing in what's possible.