There are many stories about autism in the alternative medicine space. One of the most common one is the myth that vaccines cause autism, something that has been disproven many times. Vaccines don't cause autism, and anyone who tells you that they do is either remarkably ignorant or lying.

But what not everyone knows is that even the most basic aspects of this myth fails to hold water. A common claim that you'll find from pundits such as RFK Jr. and the American right-wing these days generally is that autism rates are skyrocketing, and that this must have been caused by something. They usually then go on to blame vaccines, because they can't think of any other reasonable explanations for the increase.

In fact, we know that autism rates have basically not increased in the last fifty or so years. Instead, we have simply redefined what the term "autism" means to capture a bigger group of people.

Let's look at the data.

The Beginning

It's important when we're discussing autism rates to start at the beginning. Autism was first coined as a descriptive term used to talk about schizophrenic patients in the early 1900s. It wasn't a diagnosis per se, but instead was a clinical sign that meant that a person was withdrawn and not interested in talking or interacting with others.

As the decades moved on, the term became more and more specific. By the 1930s, autism was mostly used to describe children who were entirely uninterested or incapable of communication, but it was still more descriptive than diagnostic. If a physician used the term autistic to describe a child in the 30s, it was usually as an addendum to another diagnosis.

During the 1940s, the term became more specific largely due to the work of Leo Kanner and Hans Asperger. These physicians used the term autism to refer to children with severe developmental delays, who displayed a group of similar symptoms such as repetitive behaviours and difficulty communicating. However, the term was still very specific — it was sometimes called infantile or childhood schizophrenia, and was exclusively diagnosed in children with very severe disabilities. It was considered a similar disease to schizophrenia, and based on the Freudian beliefs of the time was thought to be caused by some issue with mothering or attachment.

As the psychiatric world moved away from Freud towards a more scientific understanding of mental health, the diagnosis shifted considerably. By the 1970s we knew that autism was caused by differences in brain development rather than parenting, and the third edition of the American Psychiatric Association's Diagnostic and Statistical Manual third edition (DSM III) recognized this by completely rewriting the definition of autism. It was now a completely separate disorder, not related to schizophrenia, and had a substantially broadened diagnostic criteria.

The criteria for an autism diagnosis kept getting broader. In the early 90s, the DSM IV came out, recategorizing autism yet again. Now, you could have autism, Asperger syndrome, or infantile autism, expanding the number of children who fit into the diagnostic category even further.

Finally, in 2013, the DSM V was published. With decades of data and a huge push from the autistic community, the diagnosis was changed once more to our current category — autism spectrum disorder (ASD). This category is now very broad, which was an intentional choice to help provide services to the greatest number of people who might need them. Rather than the severe intellectual disability of the 1940s and 50s, ASD is a group of behaviours that can be any severity as long as they are persistent and impact people's daily functioning in a significant way.

ASD is also divided into three categories — people who may require support, people who need quite a lot of support, and people who need very substantial daily help sometimes for their entire lives. This is again aimed at trying to help people who may need a lot of services to assist in their health, but also has the impact of making the diagnostic criteria apply to more people.

The Epidemic

If you read back through that previous section, it may be obvious already what this article is about, but let's look at the claims here. The principle argument is that the rate of autism has exploded in recent years, and is now massively higher than it was in the past.

This is undeniably true, based on population statistics. Initial estimates from the 60s when prevalence studies were first conducted indicated that 4.5 in 10,000 children had autism. This data indicated that the syndrome was very uncommon, even rare.

In the intervening years, the rate of autism has increased enormously. The diagnosis rate had leapt to about 60 in 10,000 by the early 00s, was up to roughly 100 per 10,000 by 2010, and is now closer to 300 per 10,000 based on the most recent statistics. That's an increase of nearly two orders of magnitude in less than a century, which is huge by any measure. The rate is also continuing to increase, with the number doubling or even tripling every decade for the last 50 years.

So what's going on?

If you believe the uninformed, it's clearly vaccines. But the real explanation is much simpler and also very obvious — we changed the definition of autism. You can't directly compare the original rate of autism to the numbers of autistic children today, because these researchers defined autism as only the most severe, nonverbal children with severe disabilities.

This has been proven in a range of scientific studies. One paper from Denmark estimated that at least 60% of the increase in diagnoses of autism was due to reporting differences and changes in diagnosis. Another Australian study looked at the global rates of autism and found that there was no evidence that there was any true increase in the numbers once you accounted for diagnostic changes between 1990 and 2010. A study in California found that 26% of the total increase in autism could be explained by people who were previously diagnosed with mental retardation being diagnosed instead with autism when those two diagnoses were modified.

In other words, this isn't just theoretical — we can say with a great deal of certainty that this is the cause of the observed increase in autism. It's not that autism is becoming more common, it's that we have redefined the term to be far more inclusive. Something like 90% of the people who are currently diagnosed with autism would not be considered autistic if they had been born in the 40s, 50s, or 60s. Similarly, many people over the age of 50 who were never diagnosed probably would meet the diagnostic threshold if they were to be born today.

This is a very simple explanation, and it's extremely common in medicine. Broadening the diagnostic criteria for high blood pressure lead to more people getting diagnosed with high blood pressure. The same is true for diabetes. It happens with mental health diagnoses such as ADHD as well.

There are some other factors at play other than diagnostic criteria. If you look at the studies I cited above, they also discuss better understanding by clinicians of the diagnosis, better reporting to health authorities, and even just better recording once clinical records went online rather than being paper-based. But it's clear that the majority of the increase in autism is very simply due to the fact that we changed what the term autism means.

There is no autism epidemic. There is just a better recognition of a condition that we are even now working towards understanding better.