In March the US Centers for Disease Control and Prevention (CDC) the newly measured autism prevalences for 8-year-olds in the United States, and headlines roared about a “1 in 88 autism epidemic.” The fear-mongering has led some enterprising folk to latch onto our nation’s growing chemophobia and link the rise in autism to “toxins” or other alleged insults, and some to sell their research, books, and “cures.” On the other hand, some researchers say that what we’re really seeing is likely the upshot of more awareness about autism and ever-shifting diagnostic categories and criteria.
Even though autism is now widely discussed in the media and society at large, the public and some experts alike are still stymied be a couple of the big, basic questions about the disorder: What is autism, and how do we identify—and count—it? A close look shows that the unknowns involved in both of these questions suffice to explain the reported autism boom. The disorder hasn’t actually become much more common—we’ve just developed better and more accurate ways of looking for it.
Leo Kanner first described autism almost 70 years ago, in 1944. Before that, autism didn’t exist as far as clinicians were concerned, and its official prevalence was, therefore, zero. There were, obviously, people with autism, but they were simply considered insane. Kanner himself noted in a 1965 paper that after he identified this entity, “almost overnight, the country seemed to be populated by a multitude of autistic children,” a trend that became noticeable in other countries, too, he said.
In 1951, Kanner wrote, the “great question” became whether or not to continue to roll autism into schizophrenia diagnoses, where it had been previously tucked away, or to consider it as a separate entity. But by 1953, one autism expert was warning about the “abuse of the diagnosis of autism” because it “threatens to become a fashion.” Sixty years later, plenty of people are still asserting that autism is just a popular diagnosis du jour (along with ADHD), that parents and doctors use to explain plain-old bad behavior.
Asperger’s syndrome, a form of autism sometimes known as “little professor syndrome,” is in the same we-didn’t-see-it-before-and-now-we-do situation. In 1981, noted autism researcher Lorna Wing translated and revivified Hans Asperger’s 1944 paper describing this syndrome as separate from Kanner’s autistic disorder, although Wing herself argued that the two were part of a borderless continuum. Thus, prior to 1981, Asperger’s wasn’t a diagnosis, in spite of having been identified almost 40 years earlier. Again, the official prevalence was zero before its adoption by the medical community.
And so, here we are today, with two diagnoses that didn’t exist 70 years ago (plus a third, even newer one: PDD-NOS) even though the people with the conditions did. The CDC’s new data say that in the United States, 1 in 88 eight-year-olds fits the criteria for one of these three, up from 1 in 110 for its 2006 estimate. Is that change the result of an increase in some dastardly environmental “toxin,” as some argue? Or is it because of diagnostic changes and reassignments, as happened when autism left the schizophrenia umbrella?
To most experts in autism and autism epidemiology, the biggest factors accounting for the boost in autism prevalence are the shifting definitions and increased awareness about the disorder. Several decades after the introduction of autism as a diagnosis, researchers have reported that professionals are still engaging in “diagnostic substitution”: moving people from one diagnostic category, such as “mental retardation” or “language impairment,” to the autism category. For instance, in one recent study, researchers at UCLA re-examined a population of 489 children who’d been living in Utah in the 1980s. Their first results, reported in 1990, identified 108 kids in the study population who received a classification of “challenged” (what we consider today to be “intellectually disabled”) but who were not diagnosed as autistic. When the investigators went back and applied today’s autism diagnostic criteria to the same 108 children, they found that 64 of them would have received an autism diagnosis today, along with their diagnosis of intellectual disability.
Further evidence of this shift comes from developmental neuropsychologist Dorothy Bishop and colleagues, who completed a study involving re-evaluation of adults who’d been identified in childhood as having a developmental language disorder rather than autism. Using two diagnostic tools to evaluate them today, Bishops’ group found that a fifth of these adults met the criteria for an autism spectrum diagnosis when they previously had not been recognized as autistic.
Another strong argument against the specter of an emergent autism epidemic is that prevalence of the disorder is notably similar from country to country and between generations. A 2011 UK study of a large adult population found a consistent prevalence of 1% among adults, “similar to that found in (UK) children” and about where the rates are now among US children. In other words, they found as many adults as there were children walking around with autism, suggesting stable rates across generations—at least, when people bother to look at adults. And back in 1996, Lorna Wing (the autism expert who’d translated Asperger’s seminal paper) tentatively estimated an autism spectrum disorder prevalence of 0.91% [PDF] based on studies of children born between 1956 and 1983, close to the 1% that keeps popping up in studies today.
One study in South Korea found a significantly higher rate of autism, but it used a different methodology and different study population. In fact, the part of the Korean study that was most comparable to other studies found an autism rate of .8%—about the same as in other countries.
Toting up these three known reasons for why autism prevalence is rising—consistent clinical prevalence rates across generations, many people who fit the criteria for autism going unidentified, and evidence of diagnostic substitution—we don’t need to dig much further to explain what looks superficially like an “autism epidemic” in the U.S. Because of greater awareness of autism and the flexibility of the diagnostic tools used, we’ve recently been diagnosing people with autism who previously would have received other diagnoses or gone unidentified.
And now, after the autism rate’s 70-year boom, it may soon take a sharp drop—but that will have nothing to do with environmental factors. The manual currently used to make mental-health diagnoses, the DSM-IV (the “bible of psychiatry”), is in the process of being updated. The proposed DSM-V criteria for diagnosing autism would, according to some studies, shift down the number diagnosed. In fact, Asperger’s and PDD-NOS would disappear altogether. If the number of people diagnosed as autistic decreases under the new criteria, as predicted, should it send us scurrying to look for environmental factors that are decreasing to explain the decline? Obviously not.