Autism is the poster child for the dramatic effects that changes in diagnostic criteria can have on apparent prevalence rates of disorders. CDC reports that the prevalence of autism was 0.05% in 1980, vs. 1.5% in 2015—a 30-fold increase. Most agree that diagnostic changes, especially DSM-4, drove this “epidemic.” Let’s look at the three most important benchmarks in autism diagnosis history.
DSM-3 DSM-3 was published in 1980, and the diagnosis of “infantile autism” required all of the following:
Onset before 30 months of age
Pervasive lack of responsiveness to other people
Gross deficits in language development
Peculiar speech patterns such as immediate and delayed echolalia
These days, such criteria would apply to only the most severe cases of autism.
DSM-4 With DSM-4’s publication in 1994, diagnosing autism became more complicated, and paradoxically, much more likely. The updated manual included an umbrella category called “pervasive developmental disorder,” and it encompassed five conditions related to autism: autistic disorder, Asperger’s disorder, Rett’s disorder, childhood disintegrative disorder, and for not qualifying for any other diagnosis, PDD not otherwise specified.
Focusing specifically on “autistic disorder,” DSM-4 introduced the well-known triad of symptoms: impairment in social interaction, impairment in communication, and restricted, repetitive behaviors or interests. Each of these symptoms included 4 items (for a total of 12 items), and a child met the autism threshold if 6 of the 12 were met. Gone were the DSM-3 requirements of gross language deficits, pervasive lack of responsiveness, or bizarre responses to the environment. Instead, DSM-4 proposed symptom domains that varied widely in severity. It was inevitable (and perhaps reasonable) that clinicians would bring many more children into the autism fold.
The most problematic aspect of DSM-4 was that the separate disorders were too similar to be reliably diagnosed.
DSM-5 DSM-5 was published in 2013, and the neurodevelopmental disorders work group decided to make some pretty drastic changes. First of all, since DSM-4’s publication, Rett’s syndrome was shown to be a genetic neurological disorder, so that diagnosis has been eliminated from DSM and is now under the purview of pediatricians and neurologists. More significantly, there was a scientific consensus that the other four DSM-4 autistic disorders were actually a single disorder that varied in severity in two core domains.
Therefore a new label was created, autism spectrum disorder, and the following DSM-4 disorders were lumped into this new category: autistic disorder, Asperger’s disorder, childhood disintegrative disorder, and PDD not otherwise specified. Don’t look for any of these in DSM-5—they’re gone.
A third major change was to truncate the symptom triad into a dyad. Since all communication is social, separating communication from social interaction didn’t make sense. So there are now two major symptoms categories: 1) impaired social communication and social interaction, and 2) restricted, repetitive patterns of behavior or interest.
A child who qualifies for both of these impairments now falls on the autism spectrum, and your next step is specifying the degree of severity, which is based on a judgment of how much treatment support the child will need (see table for details).
Finally, a new disorder was carved out called social communication disorder—reserved for people with impairment in social communication and interaction but without any restricted repetitive behaviors. The thought was that this might become a new diagnostic home for those patients formerly diagnosed with Asperger’s syndrome.
The DSM-5 changes were greeted with concern in some quarters because of studies indicating that the new criteria could substantially reduce autism diagnoses. The jury is still out, but one recent study found that the vast majority of patients with DSM-4 autism met criteria for DSM-5 ASD or for the new SCD (Kim YS et al, JAACAP;53(5):500–508).
CCPR Verdict: Autism spectrum disorder: A simpler and probably more useful label for diagnosing a complex problem.