Monday, November 12, 2012

Autism and the New DSM-5 Criteria: Who Will Be Left Behind?

Emily Willingham

When news broke that the autism spectrum categories of Asperger’s disorder and Pervasive Developmental Disorder-Not Otherwise Specified (PDD-NOS) would get subsumed into the wider maw of a general “autism disorder,” people worried. They worried about autistic people who are quite verbal or who have typical cognitive skills. What would happen to individuals whose autism doesn’t manifest in those terms as profound? The biggest concern was a new category for the Diagnostic and Statistical Manual of Mental Disorders (DSM)-5, social communication disorder. Would people like my son, diagnosed with Asperger’s and whose autism includes echolalia, anxiety, motor deficits, repetitive behaviors, learning differences, and other features well beyond the social, get rolled into what looks like a flimsy, catchall not-safety net of “social communication disorders”? And what other kind of communication is there if not social?

Based on early reports, the concerns were legit. One alarming presentation at a January 2012 conference in Iceland claimed that only 45% of people diagnosed with Asperger’s disorder or PDD-NOS would retain an autism spectrum diagnosis under the new guidelines. Those findings were preliminary and controversial, but further worrisome conference reports trickled in, all suggesting that the new autism diagnostic criteria proposed for the DSM-5 would exclude a good-sized percentage of people currently diagnosed on the autism spectrum.

Then a peer-reviewed version of the study presented in Iceland emerged in March 2012 suggesting that the new criteria would be pretty good at excluding appropriately people who aren’t on the autism spectrum but have limited ability to capture people who are. According to that study, the autistic people who would be overlooked under the new guidelines happen to be those without intellectual disability and who would fit the current criteria for Asperger’s disorder or PDD-NOS. People like my son.

Each of these studies had some limitation -- the choice of population, not originally designed to evaluate DSM-5 criteria -- that made their conclusions wiggly. Now there’s a new study out, one packed with data, focused on three clinical populations with different features, assessing how well the new criteria compare to the old in capturing autistic people and excluding the non-autistic. The headlines promise us that we will see “less impact” with the new definition or that the proposed guidelines “won’t miss autism cases” much more than the old ones. That, of course, begs the question of why we’d need to change the criteria at all.

But a review of the paper (abstract here) calls into question blithe assurances that people who might fit current Asperger’s and PDD-NOS diagnoses have nothing to worry about. What the authors really found was that the new criteria could end up excluding about 1 out of 10 children who meet the existing criteria for spectrum disorders. Given the oft-trumpeted 1 in 88 with autism in the United States, that rate of exclusion of people who should be included translates into a tens of thousands of individuals. And let’s not forget that the existing criteria, according to this study, correctly identify those who have autism at a higher rate than the proposed criteria do.

Where the old and new might distinctively part ways is in how well they exclude people who don’t have autism from receiving an autism diagnosis. The old criteria don’t do this as well, although how well depends on the specific population and diagnosis. From where I’m sitting, it looks like we’re making an untenable bargain: missing out on more people who have autism to avoid misdiagnosing people as having it when they don’t. The people we miss will live with an unrecognized disability that interventions and supports could help, and the people diagnosed with it who don’t have it will receive therapies that may or may not help them. Is no better option available? I think there is.

In their discussion, the authors make the following worrisome and strange observation:
Given concerns that the restricted and repetitive behavior requirement might lead to reduced identification of children previously diagnosed with ASDs under DSM-IV and possible misclassification as social communication disorder, we examined why some children with PDDs did not meet the DSM-5 criteria. In all three groups, most children who did not meet the criteria did so because they did not demonstrate the required impairments in social and communication functioning, and not because they did not meet the restricted and repetitive behavior criteria. In fact, few children did not meet the restricted and repetitive behavior requirement in DSM-5. These results suggest that few children with ASDs are likely to be misclassified as having social communication disorder and lend further support to the addition of the restricted and repetitive behavior criterion.
If I’m reading that reticulated verbiage correctly -- and I tried it before, during, and after coffee -- some children who had pervasive developmental disorders failed to meet the new DSM-5 criteria for autism because they did not have the “required impairments in social communication and functioning.” Yet they tended to meet the repetitive and restrictive behavior criteria, which was unexpected following a report finding the contrary. It’s been my experience that those behaviors themselves can impair social communication function, so it’s difficult for me to see how a clinician draws a clean line between, say, “repetitive questioning or extreme distress at small changes” (under repetitive/restrictive behavior) and “difficulties adjusting behavior to suit different social contexts” (social communication). And what happens to children–this study did not address autistic adults–who slip through a diagnostic crack because of this failure to meet the social communication criterion?

As it turns out, that particular category of symptoms continues to be the most problematic among the new criteria. The authors say that the results “revealed potential challenges in the interpretation of DSM-5 criterion A3.” If you read “criterion A3,” you can see why:
Deficits in developing and maintaining relationships, appropriate to developmental level (beyond those with caregivers); ranging from difficulties adjusting behavior to suit different social contexts through difficulties in sharing imaginative play and  in making friends  to an apparent absence of interest in people.
That’s so open to interpretation as to be irrelevant.

I talked to a clinical expert about it. Deborah Budding, neuropsychologist and supervising faculty at Harbor-UCLA Medical Center’s Neuropsychology training program, in addition to being in private practice, wrote to me that this finding also increases concerns that the criteria addressing language and communication will still exclude specific groups, such as the most superficially verbally adept on the spectrum, like my compulsively verbalizing son. The social communication section requires that all criteria be met, she said, and “this will essentially exclude people with superficially ‘good’ language -- large vocabularies and oromotor skills allowing intelligibility -- who nevertheless struggle to actually use language effectively to communicate.” What she references there is pragmatic language, the ability to use those big vocabularies as a tool to communicate with others. From our experience, the big vocabulary words get a lot of play because they’re fun to say, over and over again, not for the meaning they carry.

Girls are a particular concern. They may form a fairly large hidden corner of the autism spectrum, obscured by social communication functions normed to a male population but diminished in their functional capacities among other girls. Girls who are officially diagnosed as autistic tend to have more intense manifestations, possibly because those less intensely affected don’t ping the clinical radar. This latest study involved children who were on the radar, and the girls it assessed could well be a selected subpopulation of the most severely affected. This selection would have flattened any sex-based distinctions that might have emerged about how effective the new criteria would be at diagnosing females.

The authors themselves recognize some of the failings of the criteria, noting that “there is much room for improvement with respect to specificity,” the accurate exclusion of people who do not have autism. But, they say, the stated goal of the DSM-5 committee was to do a better job of separating autism spectrum disorders from non-spectrum disorders, which they list as “language disorders, intellectual disability, attention deficit hyperactivity disorder, and anxiety disorder.” Was a loss of more autistic people to an accurate diagnosis the marching companion of that goal?

The thing is, the newly proposed criteria do contain one requirement that, I think, should outshine and outscore all categorization or labels. Holding to this single requirement could resolve this endless dilution of the validity of psychiatric labels, which get no help from now-you-see-it-now-you-don’t diagnoses like Asperger’s. It’s the fourth item they list, and it reads as follows, simply: “Symptoms together limit and impair everyday functioning.”

And I close with this question: If we identify symptoms that are limiting to everyday functioning, why do we even have to slot them into a label? Labels for syndromes have current uses, such as obtaining services or insurance payments, but wouldn’t symptoms that negatively affect everyday function serve the same purpose? Among other things, addressing specific symptoms for mitigation instead of slapping on labels could reduce the stigma that some labels carry. We all have gaps that need mitigating. If we addressed specific gaps instead of wholesale labels, we’d all overlap with each other in some way, for anxiety, attention deficits, compulsive or obsessive or impulsive behaviors, depression, mania, intellectual disability or learning differences in specific areas, or addiction–and it would be a lot harder to stigmatize what we have in common, regardless of our overall neurobiology. Furthermore, we wouldn’t have to worry about applying lists of criteria that result in the wrong labels or therapies for the wrong people, or–worse in my mind–overlooking entirely the people who need help and support.


This essay was originally published at