Saturday, February 4, 2012

Info on the DSM V


Autism and Asperger’s in the DSM-V: Thoughts on clinical utility

Last week after writing about the DSM-V “Temper Dysregulation Disorder with Dysphoria,” I received several emails asking my opinion regarding the proposed merger of autism and Asperger’s disorder into a single ‘spectrum’ category.  This change has clearly generated some significant political debate in the media and the blogosphere, with some in favor of the change (see for example Dr. Roy Ginker’s NYT article), while others have expressed reservations about the potential impact that this change may have in the autism and Asperger’s community. So I wanted to keep my contribution to this discussion somewhat removed from the political/social issues associated with the change, and instead focus on the scientific/clinical basis for this specific move.  Thus, my aim with this post is not to take a position for or against the proposed DSM-V changes. Instead, I simply want to provide some background information about some of the research data and clinical issues that may have contributed to the DSM-V committee’s decision to propose the merger of all ASDs into a single category.

As simple background, according to the DSM-IV, the basic diagnostic distinction between autism and Asperger’s disorder is absence of clinically significant delays in language, cognitive development, and adaptive functioning in the Asperger’s group. The rest of the diagnostic criteria (impairments in social interactions, restricted repetitive and stereotype patterns of behaviors) between autism and Asperger’s is identical. This makes it difficult to differentiate children with Asperger’s from those with High Functioning Autism (HFA; i.e., those who meet the diagnosis of autism but perform in the average to above average rage in intellectual tests). Therefore, two teens with otherwise identical clinical profiles would be diagnosed differently if they differ on their history of language and cognitive delays. The child with a history of language/cognitive delays would be diagnosed with HFA and the child without a history of language/cognitive delays would be diagnosed with Aspeger’s. I mention this because any discussion about the science of the possible differences between these two categories is limited by the fact that both groups have been selected, by definition, to be different. Thus, the question is not whether these two groups are different – they are different because we have defined them differently. The question is whether these two groups actually represent two distinct typologies that go beyond the distinction of language/cognitive delay vs. no delay.

So what would drive the DSM-V to propose the merger between Asperger’s and Autism? In essence, the questions are 1) whether these two conditions represent two different disorders or are simply variations within a larger spectrum, and 2) whether having two categories, as defined today, is clinically useful. If Asperger’s and Autism are simply the same disorder separated by an arbitrary distinction (language/cognitive delays), having two categories would not help us in our understanding or treatment of the conditions, and keeping them as separate categories may be an obstacle for research because it encourages researchers to focus on a domain that may not be relevant or informative. However, if the language/cognitive delay distinction reflects differences between two truly distinct categories, the existence of two categories rather than one should help us make more effective interventions, inform our clinical decisions, or help us better understand the phenomenology of both conditions. Has this been the case?

Let me address the clinical impact of these two conditions from the perception of clinicians (Note: although I am basing these statements on my experience as a clinician interacting at academic/training settings, I admit that this may not represent the experience and practice of all clinicians). I interact weekly with graduate students who are learning how to conduct neuropsychological evaluations for children and adolescents. Often these students have already developed a schema, or prototype, of the child or adolescent with Asperger’s. They would describe such a child as someone who has intense and unusual interests, maybe superior skills in some area such as music or art, rigidity in behaviors and interests, and social and communication ‘deficits’ leading to difficulties interacting and relating to others. The problems begin when we start seeing actual assessment cases. For example, recently a doctoral intern and I sat in supervision to discuss a case of a teenage boy who could be described as having a “perfect” Asperger’s profile, fitting both the student’s schema and the DSM-IV criteria; except for one thing: the client had a documented history of language delays. There was no question about the diagnosis: If the teen had a history of “language delays’ the diagnosis is autism. My student then asked me, so if this is HFA, how does Asperger’s look like? I replied, just like this.

Therefore, in clinical settings, HFA and Aspeger’s disorder look mostly identical, assuming the clinician follows DSM guidelines. But the most important question is whether the current diagnostic difference is clinically useful.  When debating the Autism vs. Asperger’s diagnostic question, I have always asked my students and supervisors whether the diagnostic difference would change anything regarding our approach to the case. This is the most critical question: would our recommendations or conclusions change based on the final diagnosis that we provide (autism vs. Asperger’s)? The answer is usually, if not always, no. Given identical clinical profiles, the recommendation for treatment, school accommodations, parental interventions, and so forth, would be the same for two adolescents who only differ on the presence or absence of language delays in early childhood. The provision of a diagnosis of autism vs. Asperger’s may lead to different political/personal/social consequences, but clinically, the current DSM-IV distinction between these two conditions, and the research that has come out of this distinction, has not informed or improved our clinical practice (e.g., selection of treatment, assessment, prognosis, etc). This is likely one of the main reasons that led the DSM committee to suggest the merger of Asperger’s and Autism.

But why has the DSM-IV distinction failed to improve clinical services or lead to a greater understanding of these conditions? One possibility is that these two conditions are variations of a greater spectrum and that the language/cognitive delay difference is arbitrary (see for example Bennett et al., 2008 for a study showing identical clinical outcomes between HFA and Asperger’s). In such a case, the merger of the two conditions would better reflect the true nature of the conditions as a variations within a single spectrum. However, another possibility is that the DSM-IV criteria is simply wrong. Under that hypothesis, research has failed to find utility for this classification because of an erroneous diagnostic criteria which led to the incorrect classification of people. Some support for this later position was provided by the research team of Fred Volkmar at the Yale University Child Study Center (Klin et al., 2005). They proposed a new diagnostic criteria for Asperger’s disorder that was more inline Asperger’s original 1944 observation of his cases. Under this system HFA and Asperger’s would differ on 3 specific domains:

1. Nature of social impairments: HFA would be characterized by self-isolation and lack of interest while Aspeger’s would be characterized by interest in social relations and ‘seeking others’ (social motivation) but in a socially insensitive or atypical manner.

2. Nature of language impairment: HFA would be characterized by delayed, echolalic and stereotyped language while Asperger’s would be characterized by adequate or precocious language but with difficulties in the use of language (pragmatics).

3. In addition, the Asperger’s diagnosis would include one-sided verbosity and the presence of factual, circumscribed interest that interferes with the person’s functioning (e.g., education and social interactions).

Interestingly, some research has shown differences between HFA and Asperger’s when using the Klin criteria above (see for example Mazefsky and Oswald. 2006). Thus, it is possible that the lack of clinical utility of the current DSM-IV diagnostic distinction between HFA and Asperger’s is due to a lack of validity of the DSM-IV criteria rather than the lack of validity of the constructs of HFA and Asperger’s as two distinct syndromes. So why did the DSM-V committee recommend the merger of these two conditions rather than a redefinition of the Asperger’s criteria? It appears that their interpretation of the totality of the data is that there is no sufficient evidence to validate these two conditions as two separate syndromes regardless of diagnostic criteria used, and that the differences observed are better accounted for by differences in language, IQ, and severity, rather than features of the disorder.

From the DSM-V committee:

Differentiation of autism spectrum disorder from typical development and other “nonspectrum” disorders is done reliably and with validity; while distinctions among disorders have been found to be inconsistent over time, variable across sites and often associated with severity, language level or intelligence rather than features of the disorder.

Update: I just noticed that Dr. Mohammad Ghaziuddin, an accomplished autism and Asperger’s researcher and clinician working at the University of Michigan, just published an opinion piece on the Journal of Autism and Developmental Disorders arguing for a redefinition of Asperger’s rather than its merger with Autism. He argues that the current DSM-IV definition is incorrect and a new updated definition (following the Klin’s criteria outlined above) would be more accurate and clinically useful. He states:

…what is needed is a revision of its criteria taking into account, its quality of social impairment (active but oddrather than aloof and passive); idiosyncratic interests (oftensophisticated and intellectual); communication style (oftenpedantic and verbose); and age of onset/emergence of symptoms (often around 7–8 years). In addition, effortsshould continue to establish its validity not only from autism but also from other conditions.


Klin, A., Pauls, D., Schultz, R., & Volkmar, F. (2005). Three Diagnostic Approaches to Asperger Syndrome: Implications for Research Journal of Autism and Developmental Disorders, 35 (2), 221-234 DOI: 10.1007/s10803-004-2001-y

Bennett, T., Szatmari, P., Bryson, S., Volden, J., Zwaigenbaum, L., Vaccarella, L., et al. (2008). Differentiating Autism and Asperger Syndrome on the Basis of Language Delay or Impairment. Journal of Autism and Developmental Disorders, 38(4), 616-625. doi: 10.1007/s10803-007-0428-7

No comments: