Intended for healthcare professionals

Rapid response to:


New recommendations on autism spectrum disorder

BMJ 2011; 342 doi: (Published 09 May 2011) Cite this as: BMJ 2011;342:d2456

Rapid Response:

Autism is not a scientifically valid or clinically useful diagnosis

Szatmari (1) supports the DSM V working group proposal to eliminate
subtypes such as Aspergers syndrome and move to the more generic 'Autistic
Spectrum Disorders' (ASDs) as there is little evidence to support the
validity of various sub-types. This is true, however Szatmari failed to
tackle the lack of evidence to support the validity of ASDs per se. This
is the conclusion 2 ex-services users, both diagnosed with Aspergers, and
I came to following our extensive review of the literature in our recently
published book 'The Myth of Autism' (2). As Szatmari points out
"prospective studies have shown remarkably diverse results", "genetic
studies have also emphasised heterogeneity in risk factors", "response to
treatment (is) more closely related to cognitive and language abilities"
and overall "the quality of the evidence on which their conclusion is
based is relatively weak". In Szatmari's editorial, a scientific basis for
establishing any meaningful conclusion is remarkably absent with the
closest being the rather vague and scientifically and clinically unhelpful
"autistic spectrum disorder represents a 'family of dimensional

Autism and its spectrum expanded from a diagnosis said to potentially
affect 4.5 per 10,000 to one now claimed to affect160 per 10,000 - an over
3500% increase in prevalence in just 4 decades, without any new scientific
knowledge emerging about what this diagnosis corresponds to at the
biological level (i.e. a largely ideological based expansion).The high
ratio of males to females in ASD diagnoses poses a problem. Genetic
mechanisms need to account for this, such as autism being transmitted
through sex chromosomes, and thus far molecular genetic studies have
failed to find an X or Y-link. Despite a large number of molecular genetic
studies, including whole genome scans, evidence for only a very small
proportion of the assumed total genetic risk has been found and even this
may be related more to an association with general learning difficulties
than autism specifically. Neuroimaging studies have shown a lack of
consistently replicated findings. In reviewing the inconsistency in these
studies, sample heterogeneity (such differences reflecting IQ differences
rather than ASD specifically) is a regularly encountered problem. In our
review we were also unable to find any evidence of methodologically sound
and replicated research that demonstrates that particular interventions
(whether educational, psychological, social, or pharmacological)
specifically and differentially help those who have any form of autism.
Until specific treatments for ASD are adequately demonstrated through
replicated controlled trials, we cannot and should not assume that the
diagnosis has any clinical value.

We concluded that to genuinely open up a constructive approach to the
group who attract the 'autism' label, we must first remove the artificial
distraction created by such a limiting, imagination-less (dare I say
autistic?) approach to these people's lives. The task of good clinical
care is to take each individual and their family's narratives into account
(as Szatmari concedes) and find creative possibilities for change and for
more hopeful potential stories that can emerge instead of the stigmatising
'lifelong disability' script a diagnosis can instead lead to.

Szatmari P. New recommendations on autism spectrum disorder. BMJ

Timimi S, McCabe B, Gardner N. The Myth of Autism: Medicalising Mens'
and Boy's Social and Emotional Competence. Basingstoke: Palgrave-
MacMillan; 2010.

Competing interests: No competing interests

23 July 2011
Sami Timimi
Consultant Child and Adolescent Psychiatrist
Lincolnshire Partnership NHS Foundation Trust