Proposed DSM V rationale for eliminating Asperger Syndrome

http://www.dsm5.org/ProposedRevisions/Pages/proposedrevision.aspx?rid=97#

Please see the rationale section for Autistic Disorder (Autism Spectrum Disorder) for more details.

In making the recommendation to delete Asperger’s disorder, the following questions were considered:

Q.1. Have the DSM-IV diagnostic criteria for Asperger Disorder worked?

The ‘Asperger’ label has proved popular, ‘acceptable’, and has widened recognition of autism spectrum disorder (ASD) in combination with good language and intelligence. In addition, the introduction of this diagnostic entity has achieved the intended aim of prompting research into possible differences between this and other subgroups of PDD, with more than 500 published articles on Asperger syndrome.


1.1. Do the DSM-IV criteria work in clinical practice?

A number of published papers have argued that the DSM-IV Asperger disorder criteria do not work in the clinic (e.g., Mayes et al., 2001; Miller & Ozonoff, 2000; Leekam, Libby, Wing, Gould & Gillberg, 2000). Specifically, key problems exist in applying the current criteria:

·         Early language details are hard to establish in retrospect, especially for older children and adults; average age of first diagnosis is late (7 years according to Mandell et al. 2005;  11yrs, Howlin & Asgharian, 1999).

·          The trumping rule means most/all Asperger cases should strictly be diagnosed as having ‘Autistic disorder’ (Miller & Ozonoff, 2000; Bennett et al, 2008; Williams et al, 2008), although clinicians prefer to give the more specific term (Mahoney, et al.,1998)

o    Specifically, since language delay is not a necessary criterion for Autistic disorder, to meet criteria for Asperger disorder (without being trumped by Autistic disorder), a person would need to fail to meet Communication criteria for Autistic disorder. In practice, the Communication criterion (B.2.) of “marked impairment in the ability to initiate or sustain a conversation with others” is typically met by even very able individuals fitting the Asperger picture.

As a result, ‘Asperger syndrome’ is used loosely with little agreement: e.g. Williams et al (2008) survey of 466 professionals reporting on 348 relevant cases, showed 44% of children given Asperger, PDD-NOS, atypical autism, or ‘other ASD’ label actually fulfilled criteria for Autistic Disorder (overall agreement between clinician’s label and DSM-IV criteria; Kappa 0.31). 

 

1.2. Do the DSM-IV criteria delineate a meaningful subgroup for research or practice?

In part because of the difficulty in applying the criteria (as outlined in section 1.1.), different research groups often uses different criteria,  and quality of early language milestone information is variable (Eisenmajer et al., 1996; Klin et al., 2005; Woodbury-Smith, Klin, & Volkmar, 2005).  Different criteria lead to different samples being identified (see Klin et al, 2005 comparison of 3 diagnostic approaches; also Kopra et al., 2008; Woodbury-Smith et al., 2005).
 

Research suggests early language criteria do not demarcate a distinct subgroup with different:
Course/outcome: Children with autism who develop fluent language have very similar trajectories and later outcomes to children with Asperger disorder (Bennett et al., 2008; Howlin, 2003; Szatmari et al., 2000) and the two conditions are indistinguishable by school-age (Macintosh & Dissanayake, 2004), adolescence (Eisenmajer, Prior, Leekam, Wing, Ong, Gould & Welham 1998; Ozonoff, South and Miller 2000) and adulthood (Howlin, 2003).
Cause/aetiology: Autism and Asperger syndrome co-occur in the same families (Bolton et al., 1994; Chakrabarti & Fombonne, 2001; Lauritsen et al., 2005; Ghaziuddin, 2005; Volkmar et al., 1998). No clear evidence to date of distinct aetiology.
Neuro-Cognitive profile: mixed evidence, for example some authors have reported worse motor functioning in Asperger than HFA (Klin et al., 1995; Rinehart et al, 2006), while others have not found significant group differences (Jansiewicz et al., 2006; Manjiviona & Prior, 1995; Miller & Ozonoff, 2000; Thede & Coolidge, 2007). Evidence is similarly mixed for differentiation of Asperger group by lower performance than verbal IQ profile (for, Klin et al, 1995; against, Barnhill et al., 2000; Gilchrist et al., 2001; Ozonoff, South & Miller, 2000; Spek et al., 2008), better theory of mind (for, Ozonoff et al, 2000 ; against, Dahlgren & Trillingsgaard, 1996;            Spek et al, in press JADD; Barbaro & Dissanayake 2007) or executive function (for, Rinehart et al, 2006; reviewed by Klin, McPartland & Volkmar, 2005 ; against, Miller & Ozonoff, 2000; Thede & Coolidge, 2007; Verte et al., 2006) .  Note the risk of circularity for group differences relating to verbal ability, since early language development (grouping criterion) is generally predictive of later language abilities (Paul & Cohen, 1984; Rutter, Greenfield & Lockyer, 1967; Rutter, Mawhood & Howlin, 1992).
Treatment needs/response: no empirical studies demonstrating the need for different treatments or different responses to the same treatment, and in clinical practice the same interventions are typically offered. 
Q.2. Does the existing research literature allow us to suggest new criteria to diagnose Asperger Disorder, in contrast to Autistic Disorder/ASD?

The current clinical and research consensus appears to be that Asperger disorder is part of the autism spectrum, although with possible over-use of the term it is quite likely that other (non-ASD) types of individuals have received this label.

Research field currently reflects two views: 

That Asperger disorder is not substantially different from other forms of ‘high functioning’ autism (HFA); i.e. Asperger’s is the part of the autism spectrum with good formal language skills and good (at least Verbal) IQ. Note that ‘HFA’ is itself a vague term, with underspecification of the area of ‘high functioning’ (performance IQ, verbal IQ, adaptation, or symptom severity).
That Asperger disorder is distinct from other subgroups within the autism spectrum (see Matson & Wilkins, 2008, review): e.g. Klin, et al. (2005) suggest the lack of differentiating findings reflects the need for a more stringent approach, with a more nuanced view of onset patterns and early language (e.g. one-sided verbosity, unusual circumscribed interests).
2.1. What are the proposed differences? How strong is the evidence?

Several recent comprehensive reviews of the topic are available (Howlin, 2003; Macintosh & Dissanayake, 2004; Matson & Wilkins, 2008; Witwer & Lecavalier, 2008). Matson & Wilkins (2008) suggest current criteria could work if refined and supplemented. However, the research literature to date is not able to provide strong, replicated support for new or modified criteria likely to distinguish a meaningfully different group with Asperger disorder versus autism with good (current) language and IQ. Witwer and Lecavalier’s (2008) perhaps more comprehensive review concludes there is little evidence that Aspergers is distinct, and that current IQ is the main differentiating factor. Bennett et al’s (2008) follow-up study suggests that language impairment at 6-8years might have greater prognostic value than early language milestones, and Szatmari et al (2009) argue (on the basis of later developmental trajectory) for a distinction between ASD with (autism) versus without (Aspergers) structural language impairment at 6-8 years.

The draft criteria for ASD proposed by the Neurodevelopmental disorders workgroup would include dimensions of severity that include current language functioning and intellectual level/disability.

 


Q.3.  If Asperger disorder does not appear in DSM-V as a separate diagnostic category, how will  continuity and clarity be maintained for those with the diagnosis?

The aim of the draft criteria is that every person who has significant impairment in social-communication and RRBI should meet appropriate diagnostic criteria.  Language impairment/delay is not a necessary criterion for diagnosis of ASD, and thus anyone who shows the Asperger type pattern of good language and IQ but significantly impaired social-communication and repetitive/restricted behavior and interests, who might previously have been given the Asperger disorder diagnosis, should now meet criteria for ASD, and be described dimensionally. The workgroup aims to provide detailed symptom examples suitable for all ages and language levels, so that ASD will not be missed by clinicians in adults of average or superior IQ who are experiencing clinical levels of difficulty.

There may be some individuals with subclinical features of Asperger/ASD who seek out a diagnosis of ‘Asperger Disorder’ in order to understand themselves better (perhaps following an autism diagnosis in a relative), rather than because of clinical-level impairment in everyday life. While such a use of the term may be close to Hans Asperger’s reference to a personality type, it is outside the scope of DSM, which explicitly concerns clinically-significant and impairing disorders.  ‘Asperger-type’, like ‘Kanner-type’, may continue to be a useful shorthand for clinicians describing a constellation of features, or area of the multi-dimensional space defined by social/communication impairments, repetitive/restricted behaviour and interests, and IQ and language abilities.