The current language around autism is failing us

Update: This was written in reaction to what another person was saying. I have since learned that is not best way to write articles. I fear I may have added to the confusion rather than detracted from it.

This was also one of the first articles I wrote on a complicated topic for a general audience. I still think it’s still accurate and decent enough to leave up. I may try to write an updated article on the same topic in the future.

Growing up, I thought that autism was essentially synonymous with intellectual disability. Autism was the children with special needs. They were placed in different classrooms than the one I was in. The brief times I saw them—in the hallways, in the cafeteria, often accompanied by an adult—it was very clear that there was a difference between them and the other students. They needed the extra help they were getting.

It wasn’t until much later that I learned that intellectual disability doesn’t always go with autism. There are some individuals with autism that are high functioning. A stereotypical depiction is that of Sheldon Cooper in The Big Bang Theory. He’s smart, he’s a physicist, but he’s particular about certain things, and he’s extremely stubborn.

Is it strange that we use the same word to describe both kinds of people? Yes, it is.

What is autism?

The Diagnostic Statistical Manual of Mental Disorders V (DSM 5) is the standard reference manual used by clinicians to make mental health diagnoses. Because clinicians are the ones who make the diagnoses (they choose who to slap the label on to), the DSM 5’s criteria is the definition of autism. All the following criteria must be met for a diagnosis: 1) presence of social deficits, 2) presence of restricted and repetitive behaviors and interests, 3) symptoms must be present in early childhood, 4) symptoms must significantly impair the individual’s life, 5) and if intellectual disability is present then social communication must be below that expected for the “general developmental level”.

In addition to the traits listed in the DSM 5, individuals with autism are more likely than the general population to suffer from a broad range of other problems including cleft lips, congenital heart problems and other congenital abnormalities, gastrointestinal problems, schizophrenia, ADHD, seizures, social anxiety, mood disorders, and intellectual disability.

This diversity seen in autism can be explained by the fact that autism has multiple distinct—possibly overlapping—etiologies (or causes). One etiology results in autism+some comorbidities. In fact, we already know of some of these causes.

There are several genes that are known to strongly associate with autism. This includes several syndromes like Fragile X Syndrome (caused by a change to gene called FMR1 on the X chromosome,) Rett Syndrome (caused by a change to a gene called MECP2 also on the X chromosome), 16p11.2 deletion syndrome (caused by a deletion of several genes on the short arm of chromosome 16). All three of these syndromes result in autism, intellectual disability, and interestingly, in seizures, too. But each of these syndromes also has some unique features. For example, individuals with Fragile X Syndrome tend to have a larger head size but individuals with Rett Syndrome tend to have a smaller head size. 16p11.2 deletion syndrome also associates with childhood onset obesity, and Rett syndrome is known for progressive loss of certain motor functions.

There are many other syndromes such as Angelman syndrome and CHD8-related syndrome that have their own unique features. To note, there is still variability in how individuals with syndromic autism present (not everyone will have all of the features associated with the syndrome), but the presentation from individual to individual is still more consistent than non-syndromic autism, which means that parents of children with these disorders have a better idea of what to expect as their children grow older.

Other new research

One of the hopes over the past few decades of research was that all of these different genetic causes would converge onto a single biological pathway. However, I don’t think this is going to happen. A different hope of autism research was that if there were several biological pathways, we would be able to distinguish them and subtype autism. This still hasn’t happened, but there are clues.

Autism with intellectual disability (autism+ID) and autism without intellectual disability (autism+no ID) are linked to different kinds of genetic variants. Genetic variants are differences in the genome from person to person. These can be point mutations, deletions, or duplications. A de novo variant is a genetic variant that is a new mutation in the child (neither parent has it). When associated with disease, de novo variants tend to be more harmful than inherited variants. Cases of syndromic autism are often due to de novo variants. But in non-syndromic autism, too, individuals with autism+ID have a higher burden of damaging de novo variants than individuals with autism+no ID.

For a long time, it was also thought that common variants—variants common to the gene pool—also contribute to autism. Common variants are those that have been passed down from generation to generation for hundreds of generations, and so are not so harmful. It wasn’t until 2019 that a Danish study was the first to discover a handful of common variants that contribute to autism. The researchers took these variants that they found and asked the question: what other traits do these variants associate with? (A lot of variants in our genome associate with more than one trait.) What they found was quite striking: those variants associated with higher intelligence not with lower intelligence. (I also really like their Figure 2 in this paper and encourage people to look if they want to see what other traits correlated.)

These two pieces of evidence (the contribution of de novo vs. common variants) reflect the real life differences we see between those that need intensive support versus those that are the Sheldon Coopers of the world.

Why did we ever think it was a good idea to group them together?

The purpose of diagnostic categories is to aid several groups of people: clinicians, patients, patient’s support system, and researchers. Each group has different needs, and ideally, our diagnostic categories would serve them all.

The clinicians are interested in making a differential diagnosis and providing care: if two distinct diagnoses have a similar presentation and require the same care it’s ok to group them together from the clinician’s perspective. For example, when diagnosing for the common cold, it doesn’t matter which strain of the virus a patient has—the treatment is the same—rest and wait it out.

And it is a subset of clinicians who define what autism is—the American Psychiatric Association who publish the DSM 5. I am speculating a bit here—but bear with me—the American Psychiatric Association is comprised of physicians who don’t generally provide the behavioral support that individuals with autism need. So, from their perspective the care they provide for all of these different kinds of autism is the same—since there are no medical treatments. They make the diagnosis and send them on their way to the next person.

Perhaps this is why autism has only broadened in scope over the years. It’s pretty well known (in circles that are interested in autism) that the most recent version of the DSM 5 collapsed autistic disorder, Asperger’s disorder, and pervasive developmental disorder into a single disorder now called autism spectrum disorder. From their perspective, collapsing them into one category makes diagnosis easier and makes no difference in care. To add to the confusion, they’ve decided to pair the word autism with the word spectrum, which evokes images of the color spectrum, and, at least to me, makes it sound like there is only one axis of variability and that axis is on a continuous scale.

It is the psychologists, parents, teachers, and so many others that provide the actual care. And these groups, I think, would benefit from having clearer distinctions. I don’t have direct experience with care of individuals with special needs, but I can only imagine that care provided to individuals with autism+ID is very different than that provided to autism+no ID.

Another reason that there may be benefit to changing diagnostic categories is that individuals (and parents) with any disease often seek out others with the disease for mutual support. Finer resolution categories will help them find others that have closer experiences.

Yet another reason for changes to categorization is the growing frustration in advocacy communities. Groups with different needs are at odds. Should autism be considered a disorder at all? A change in terminology, again, will help here.

Finally, we can talk about researchers. Now, as I said earlier, one of the hopes over the past many decades was that we would discover a handful of converging biological pathways to explain autism in general. Over 1000 genes have been implicated in autism. Researchers looked for functional similarities between the genes and found many different processes are involved including gene regulation, synapse formation, synapse function, protein synthesis, metabolism.

Since a reasonable hypothesis is that any converging pathways would converge somewhere in the brain, and the same is true for intellectually disability then wouldn’t it make sense to study these two populations separately? This is often (but not always) done, and researchers may choose to divy up the population using different criteria—for example, they may take into consideration the presence of developmental delay. Changing the categories will add more consistency.

Closing remarks

What we consider to be autism and not autism is a matter of definitions. As humans, we enjoy sorting ourselves: Hogwarts houses, big five personality traits, or DSM 5 diagnostic categories. It’s a fun thing to do, and I certainly enjoy doing it myself.

Outside of strictly entertainment purposes, when creating categories it’s important to consider how useful these categories are and who these categories serve. Do they help us to predict other traits? Do they help us, or our support systems, with providing care? Do they help researchers research? If the categories don’t serve us then it’s time to rethink them.

I think that even if we do something as simple as calling it autism disorders (with an “s”), it can already be powerful in changing how people think about these conditions.