Monday, February 20, 2006

On the differential validity of Asperger's disorder and Autism

“The validity of a diagnostic construct should be judged on its predictive power relative to variables of clinical or research interest,” is one of the last statements made by Klin, Pauls, Shultz and Volkmar (2005) in their article examining diagnostic approaches to Asperger Syndrome (AS). It sounds like it is a cry of desperation.

And arguably so. The diagnostic question of “Whether to Asperger or not?” causes clinicians and researchers a great deal of stress. Indeed, Klin, et al (2005). state that the reason that the Pervasive Developmental (PDD) Category was splintered into many different pieces was in order to stimulate logical debate and research into the area. Indeed, it has done so, they state, in the case of, for instance, Rhett Syndrome, where a host of bio-genetic and neurological studies have been conducted. In this instance, it appears as if the experiment worked out well.

In 12 years, research has led us to a better understanding on Rhett syndrome, while in the same amount of time (and perhaps more, if you count the time from Asperger’s original paper in 1944), we are none the wiser about AS. Witness confusion among various individuals on the NASP listserv when discussing the syndrome. Twelve years after the fact, and the leading researchers are still attempting to compile evidence to support the validity of the syndrome.

Although many discuss Wing’s paper (1981) in which she documents to this generation her encounters with an individual with Asperger’s as being seminal, I would also count the text '>Asperger’s Syndrome by Klin, Volkmar and Sparrow (2000) as equally important, as it represents a collection of research on the construct. And the team from the Yale Study Center continues to pose some interesting questions for us.

Before I continue, I need to let you know that as a practicing school psychologist, I was always wary of parents who told me that AS was a syndrome along the same continuum as Autism (AU). My belief at the time was that this was a ploy to secure educational as well as social services (Medicaid housing after a certain age) many of which were limited to individuals with a classification or diagnosis of autism. I thought was that these parents felt if their child has a disorder similar to or along the same continuum as AU, they would be eligible for the same services as other children.

However, my thoughts changed once I began to work with many of these children in a clinical setting. Clinically, I began to view these children as being on a spectrum or dimension, some children had more of the disorder and some had less of it; the nature of their difficulties was not necessarily related to their label or diagnosis (PDD-NOS vs. AU vs. AS), but related to intellectual capacity, language abilities and social issues. Also, in terms of interventions, they all appeared to react the same.

Research indicates that there are many Verbal-Performance differences in childrens with AS. In general, the findings suggest better verbal than non-verbal skills, but most of the research was conducted using outdated assessment instruments and outdated theories of cognitive ability.

Much of the research however, was not conducted with the CHC theory.Catell, Horn, Caroll, McGrew, and Flanagan, have all indicated that most IQ tests (specifically the Wechsler tests) do not appear to measure what they purport to measure. For instance, on the Wechsler tests, the Verbal IQ (which is "supposed" to measure verbal skills) is actually measuring a skill called crystallized knowledge or Gc, a sum of your cultural, verbal and linguistic knowledge.

This makes sense.

However, the Performance IQ (which is "supposed" to measure non-verbal reasoning) does not measure visual-spatial reasoning or Gv solely; it also taps fluid reasoning, or Gf – the ability to apply reason to deal with novel situations. Taken in this light, it makes perfect sense when you say that within an AS individual, Gc is greater than Gf.

To focus on VP differences, is in essence, very 20th century. Intellectual ability is not a simple construct, but incredibly complex and different depending on the child’s developmental level.

So, conceptually the idea of the groups having cognitive difference is still not demonstrated. A second area of investigation that Klin et al. highlight are the parental histories of both groups. Reviews of parental psychopathology revealed the presence of a Broader Autism Phenotype (BAP), a pattern of higher rates of social difficulties (as well as some other cub-clinical syndromes nearing anxiety and obsessional disorders) present in family members with AU, AS, and PDD-NOS. This concept of BAP is interesting, as it describes a familial, possibly genetic pathway to the disorder (which the authors note may be different for males and females). Although there were differences among the groups in terms of BAP presence in the families of children with AU, AS, and PDD-NOS, these differences did not approach significance.

A third area to investigate is the “onset of language” criterion; AS children tend to develop language as typically developing, AU children have language delays. The way that the DSM-IV-TR criteria are currently laid out, Woodbury-Smith, Klin and Volkmar (2005) indicate that the AS onset criteria is a poor criterion, as most of the data is often collected retrospectively. Thus, the parental reporting of “no problems in the development of early language” is a problem, as it is sensitive to reporter bias. Therefore, it is difficult, to ascertain truly if language development was normal or not. Oftentimes, we have the common occurrence of “regressed language” even though the DSM-IV-TR indicates that even in the uncommon world of psychopathology regressed language is a rare phenomenon (keep in mind, it is rare even within the domain of the rare).

So the onset of language is a poor discriminating criterion – what about the quality of language? In AU it is echolaic, idiosyncratic and repetitive, while in AS it is formal, pedantic and has an odd prosody. However, clinically I have bumped into some cases where developmentally a child began displaying echolalic, idiosyncratic and repetitive language only to later (after intense ABA intervention) develop language which with formal, pedantic and with an odd prosody.

My question then is fundamental is: If the difference in cognitive profiles and the onset of difficulties is not different, then is it worth distinguishing two constructs on the basis of one difference if they have two similarities? Scientifically, the concept of null difference needs to apply UNTIL we find a difference. Until there is a group of studies that demonstrate systematic and consistent differences, we must consider the two groups the same.

Then, the bigger picture question is this: If two disorders present slightly different phenotypes, but are treated equally and effectively with the same procedures, is it really necessary to discriminate between the two? Your thoughts....

Technorati tags: , , ,


Anonymous said...

"So the onset of language is a poor discriminating criterion – what about the quality of language? In AU it is echolaic, idiosyncratic and repetitive, while in AS it is formal, pedantic and has an odd prosody."

What you fail to mention, perhaps, is that the AS criteria for language is also echolalic, idiosyncratic and repetitive while it can be, at the same time, "scripted"--that is, repeating something that has been heard elsewhere and used in many different situations.

"If the difference in cognitive profiles and the onset of difficulties is not different, then is it worth distinguishing two constructs on the basis of one difference if they have two similarities?"

While my tendency is to question the validity of AU and AS being on the same spectrum, some of the same treatments do work for both disorders. Isn't that also true for other medical problems? Still, you wouldn't call the medical problems which respond to the same types of treatment by the same names, would you? If you would, why wouldn't AS be classified as epilepsy, or cancer, or high blood pressure? After all, some of the treatments for those problems also work for persons with AU and AS.

I think it is important to distinguish between the two disorders precisely because some treatments work for one disorder and not the other. Who wants to spend time on treatments that have been shown not to work for the specific malady encountered?

Andrew Livanis said...

I should mention that I performed a "double-take" and referred back to the DSM-IV-TR (2000). The only reference in the criteria for Asperger's Disorder (re: 299.80) is:

"There is no clinically significant delay in language (e.g., single words used by age two years, communicative phrases used by age 3 years)."

Although different authors have documented "scripted" speech, it is not yet a formal criterion. Nevertheless, I have seen some children with a diagnosis of AU "speak" like children with a diagnosis of AS.

As for the differential labelling of AS and AU, my thoughts, as always are much more complex.

The truth of the matter is that whether or not AU and AS are the same entity or belong on the same spectrum, there are very different from cancer; here is why:

Nobody ever died directly from begin diagnosed AU. Radiation and chemotherapy are not the preferred treatments for AS/AU. You can beat cancer, but AS/AU is with you for the long haul. You can also work to prevent cancer and prevent AU/AS, however, if your mother ingests a lot of mercury, you are not more likely to be born with cancer. As far as I know, there are no scientifically tested similar therapies for AS/AU and cancer.

Similarly, you cannot say that if you had a child with AU and a child with AS, with similar cognitive profiles that the same treatments would not work? I used social stories with all children on the spectrum and see good results.

(Please note, that if you present to me scientifically based, evidence-supported interventions for AU and AS, we can debate; please refrain from citing "fad" treatments.