By Tara Haelle
January 5, 2015
“...most in the field believe that there is a ‘real’ increase that merits ongoing investigation – along with the continued search for a cause of any sort for autism, but the general consensus is that the ‘real’ increase is not nearly as large as the raw numbers suggest.”
Almost two thirds of the increase in autistic Danish children results from how autism is diagnosed and tracked, found a new study in JAMA Pediatrics, lending more support to the idea that the apparent rise in autism rates, or at least most of it, is unlikely to be “real.”
That is, the increase is likely more about previously-unidentified autistic individuals getting an autism diagnosis than more individuals actually developing autism.
Although the term “autism” has been around formore than a century, it didn’t start taking on a meaning we would recognize today until the 1940s, and it remained linked to schizophrenia, a completely unrelated psychiatric disorder, until the 1960s. The definition has continued to expand become more detailed, and thus autism prevalence has been adjusted accordingly over the years.
In Denmark in particular, the diagnostic criteria for autism expanded in 1994 to include a spectrum of disorders with a broader list of symptoms, thereby widening the definition of autism. Then in 1995, national data tracking began to include diagnoses made from outpatient patient visits rather than just diagnoses of those admitted to a healthcare facility.
Since every Danish resident has a complete health record maintained by the Danish government, researchers can use this national health registry to study an entire population with lower likelihood of bias from those included or excluded in a study.
By examining the health records of all children born in Denmark from 1980 through 1991 – nearly 668,000 children – the researchers determined that 60 percent of the increase in autism rates in Denmark could be attributed to those two changes in the way Danish autistic children have been counted since the mid-1990s.
The change in the diagnostic criteria accounted for a third of the increase in autism, and including outpatient diagnoses in the statistics account for 42 percent of the increase.
What makes this study strong is its size, with well over a half million children born over a decade, and the use of the national health registry for the data. “The authors have at hand a more exhaustive and thorough database than is available pretty much anywhere else, and their findings are echoed by other studies,” said Glen Elliott, chief psychiatrist and medical director of Children’s Health Council in Palo Alto, CA. But it’s less clear, Elliott told me, how much these findings can be generalized to the U.S.
Autism prevalence in Denmark, 54 of every 10,000 children, is still far lower than that of the U.S., with 147 of 10,000 children, though the latter, he said, uses a less precise tool* than direct access to patient records, as used in this study.
“Likely, the effect of diagnostic changes is similar between Denmark and the U.S., although we don’t really have side-by-side studies of which I’m aware to prove that,” Elliott said. “The effect of awareness and early detection of autism spectrum disorders could well be different between the two countries.”
The way autism is defined in the U.S. has changed dramatically since 1980, when it first appeared in the DSM-III as “Infantile Autism” and could only be diagnosed in children whose symptoms began before they were three years old. Autism spectrum disorders have expanded to include diagnosis without a specific age requirement beyond the “early developmental period” and without requiring significant language impairment in the recently revised DSM-5.
|Autism rates have increased over the past several decades,|
but evidence suggests better awareness and identification of
autistic individuals accounts for the majority of that increase.
The vast majority of people diagnosed with autism spectrum disorders today would never have qualified under the 1980 classification, and no formal classification separate from schizophrenia existed before then. So it’s not surprising that numbers have increased in the U.S. The CDC puts the current incidence of autism spectrum disorders at 1 in 68, or about 1.5 percent,* though this figure may also mislead those unfamiliar with the broad diversity – quite literally, the neurodiversity – that autism encompasses, since popular images of autism still unfortunately misrepresent the condition.
The 1 in 68 refers to individuals all over the autism spectrum, from those who need only some supports and accommodations to those who need very substantial support. It is the latter group, especially those who have little spontaneous language (verbal or nonverbal) who are the stereotypical face of autism, but they are the minority of autistics.
The idea that increases in autism rates reflects changes in diagnostic definitions rather than a real change in the developmental condition’s prevalence is not new. Other recent studies have found similarly that real rates of autism have not changed significantly over the past couple of decades, and a 2013 study found that identified autistic children “clustered” where resources for diagnosis and treatment were greater. In other words, where there’s more access to diagnostic services, the autism rate is higher.
But this reality is often denied or ignored by a significant proportion of autism advocates, primarily those who subscribe to pseudoscience like that peddled at Age of Autism. They may also claim that autism does not occur at the same rates in other countries as it does in the U.S. because they don’t recognize that the same diagnostic, awareness and identification issues exist just about everywhere.
For example, a 2011 study in South Korea found a prevalence higher than that of the U.S. (2.6 percent, or about 1 in 38), but two thirds of the cases were un-diagnosed.
However, none of this means there isn’t a real increase in those affected – it’s not clear yet. As Elliott told me regarding the population in this study, “it is true that their analyses suggest that ‘most’ of the increase is explained by changes in diagnostic criteria and outpatient contacts, but ‘most’ is still only 60 percent, leaving 40 percent of their increase unexplained.” (And that unexplained bit will also vary across countries.)
So, if there is a real increase – again, it’s not clear – do we have any idea what might account for it?
Well, we know for darn sure it isn’t vaccines. (Seriously, there IS NO LINK.) It’s possible a small amount of the increase may relate to the increase in the average age of parents. Both women and men are becoming parents later now than at any other time in history, and autism spectrum disorders appear linked to older parents, though the data on older fathers is more solid than that on older mothers. And then there is a long, long, ridiculous list of other things linked to autism (car exhaust? labor induction? really?) that have little to no solid evidence.
“No one has offered a convincing explanation for why an increase is occurring, if it is,” Elliott said. “I think most in the field believe that there is a ‘real’ increase that merits ongoing investigation – along with the continued search for a cause of any sort for autism, but the general consensus is that the ‘real’ increase is not nearly as large as the raw numbers would suggest.”
Elliott said the possibility of other factors can’t be ruled out, and researchers have continued looking for possible environmental factors.
But, “the take-away message,” he said, “is that at least a significant amount of the perceived increase of autism spectrum disorders is due to changes in criteria we use to make the diagnosis and the increased awareness about autism spectrum disorders that has led both to earlier diagnoses and greater likelihood of detection.”