Susan Greenfield and autistic spectrum disorder: was she misrepresented?

I have had many emails in response to my open letter to Baroness Greenfield. All but one have been approving. The one exception is an eminent Professor who has chided me for misrepresenting her views. I am reproducing here our unedited email correspondence. I have anonymised the name of the correspondent, as he has not given permission for it to be used, though I will happily break the anonymity if he wishes me to do so, so he can take credit for his arguments.

As a non-celebrity scientist, I would like to get on with my day job and do some data analysis, and so have decided to reproduce the debate here, so that others can pursue it. Please feel free to comment, though please note, I will delete any comments that are off-topic, i.e. those not pertaining to issues around the validity of Greenfield’s claims, and the extent to which they have been misrepresented.

Sent: 10 August 2011 13:27 
To: Dorothy Bishop
Re: Misrepresentation of Greenfield’s article

 Dear Professor Bishop,

In your blog of 28 September 2010 you flattered yourself with the aspiration of being a “Paragon”. However, your blog of 4 August 2011 betrays that aspiration and violates the principles of scientific debate. You are misrepresenting Greenfield’s article in New Scientist. To claim that she is blaming what you call “internet use” for the grievous condition of autism is a travesty. The word autism does not appear in that article; Greenfield specifically refers to “autistic spectrum disorders”. Nevertheless, you implore her to “stop talking about autism” and unpleasantly characterise her comments as “illogical garbage”. For clarity I shall repeat myself: autism is not the subject of that article.

It is imperative that scientists engage with all sectors of society and do so accurately, honourably and without intemperate, personal comments. Publishing an assertion which misrepresents the evidence is unacceptable. Furthermore, your blog ignores Greenfield’s explicit references to peer-reviewed papers which provide data consistent with aspects of her general hypothesis (which is not about autism). Perhaps I should remind you of one of the key sentences in Greenfield’s article: “it is not the technologies themselves that I'm criticising, but how they are used and the extent to which they are used”.

Your failure to live up to the aspiration you expressed in your blog of 28 September 2010 saddens me and many other members of our community. In that blog you stated: “Paragons write personal letters to authors”. However, given the public pronouncements which you have made, a public retraction of your misrepresentation is now required. Your earlier experiences as an journal editor will no doubt confirm this requirement.

From: Dorothy Bishop
Sent: 10 August 2011 16:31 
Re: Misrepresentation of Greenfield’s article

I have no intention of withdrawing what I have said. I am happy to defend it.
You seem to think there is a clear distinction between autism and 'autistic spectrum disorders'.
There is not; many people treat them as synonyms, and those who interpret them differently regard ASD as a milder form of the same condition. There is no justification for linking either the severe or the broader category with internet use. The argument I made about a cause needing to precede it effect applies just as much to ASD, broadly defined, as to core autism. ASD does not suddenly appear in middle childhood - the symptoms are evident from around 2 years of age, and so are not plausibly caused by internet use.
If the article is not 'about' ASD/autism, then why does Greenfield mention it at all? This really does upset parents of affected children.
And isn't she aware of the large literature debating reasons for the increasing prevalence of ASD/autism diagnosis? - if she is going to cite this to support her argument, then it behoves her to do her homework.
It is really not acceptable to use innuendo to imply associations, but then back off if challenged to produce evidence.

There is a more fundamental problem here. Susan Greenfield is listened to because she is a scientist. But unlike other scientists engaged in public communcation, she does not confine herself to explaining science to a broader audience. She uses the media to promote her own new theories. What she conspicuously does not do is to publish these ideas in the peer-reviewed scientific literature. This is a shame because it means she has become disconnected from the rest of the scientific community. I would have been happy to voice my criticism by the more conventional means of peer review, which would have been private, or as commentary on a scientific paper, but I am denied that opportunity because Susan Greenfield does not publish these ideas in the scientific literature. Since her views are widely distributed through magazines and newspapers, those of us who find them flawed have no alternative but to challenge them in the public domain. I am aware that a great many people have made 'intemperate personal comments' about Susan Greenfield, but I do not accept that I have done so; I criticised the ideas rather than the person.

I might add that yours is the first critical comment I've had. I have had numerous supportive emails and comments from scientists who have not only written to say they agree, but have thanked me for raising this.

Sent: 11 August 2011 09:49 
To: Dorothy Bishop
Re: Misrepresentation of Greenfield’s article

Dear Professor Bishop,

Thanks for your response.
You present yourself as sanguine about conflating Autism and Autistic Spectrum Disorders. I find this surprising and alarming.
Your case now rests on your conviction that all of the adolescents or adults who are currently being diagnosed with any Autistic Spectrum Disorder (at an increasing incidence) could have been diagnosed as such “from around 2 years of age”. Please direct me towards peer-reviewed prospective studies which support this claim.

From: Dorothy Bishop

Sent: 11 August 2011 10:49 
Re: Misrepresentation of Greenfield’s article

I will send you some peer-reviewed papers when I have some free time, but meanwhile, please see Criterion C in the DSM5 proposed revision, as well as the rationale section, which explains the terminology.
You might also find it useful to talk to Professor Sir Michael Rutter, who is the world's leading expert on autism.


Sent: 11 August 2011 12:00 
To: Dorothy Bishop
Re: Misrepresentation of Greenfield’s article

Criterion C in the link you have provided does not address the matter in question: namely, whether there is well-controlled evidence which supports your conviction that all of the adolescents or adults who are currently being diagnosed with any Autistic Spectrum Disorder (at an increasing incidence) could have been diagnosed as such “from around 2 years of age”.
Criterion C merely raises a circular argument, which would be susceptible to unreliable retrospection.
I will indeed raise these matters with Michael Rutter.
But, more importantly, I look forward to receiving from you peer-reviewed papers which substantiate your specific claims.
Dorothy Bishop
Sent: 11 August 2011 15:51 
Re: Greenfield’s article

Your initial complaint was that I had misrepresented Greenfield because I had failed to distinguish ASD and autism. I trust the DSM5 document has clarified the point for you and you now accept this was not misrepresentation.
You are now demanding that I provide peer reviewed evidence for my supposed "conviction" that "all of the adolescents or adults who are currently being diagnosed with any Autistic Spectrum Disorder (at an increasing incidence) could have been diagnosed as such “from around 2 years of age”.
I have sent you information pointing out that it is is part of the diagnostic criteria for ASD to have onset in early childhood.
This is not a circular argument. It is merely pointing out that ASD, as defined by gold standard diagnostic criteria, could not be caused by environmental influences that only start in later childhood.  I reiterate the last sentence from the DSM 5 rationale section: "Autism spectrum disorder is a neurodevelopmental disorder and must be present from infancy or early childhood, but may not be detected until later because of minimal social demands and support from parents or caregivers in early years."
Note that this does not mean that all children with ASD will be diagnosed in childhood, but it does mean that they have evidence of autism in early childhood.  This is typically identified by an interview instrument such as the Autism Diagnostic Interview.
To clarify my argument.
1. When asked for evidence that the internet is changing people's brains, Greenfield stated, among other things, "There is an increase in people with autistic spectrum disorders."
To most people this would imply that she is saying the internet is a causal factor in the increase in autistic spectrum disorders.
2. There has been an increase in autistic spectrum diagnoses over the years.
However, this evidence comes from epidemiological studies that do use standard diagnostic criteria including the onset criteria (see attached articles).
3. Since internet use cannot plausibly cause a disorder starting in a toddler, this is not a valid argument.

You now demand that I prove that "all of the adolescents or adults who are currently being diagnosed with any Autistic Spectrum Disorder (at an increasing incidence) could have been diagnosed as such “from around 2 years of age”. "
This is an attempt to move the goalposts. Of course diagnosis is not perfect. There may be misdiagnosed cases. The fact that you demand this evidence suggests that Greenfield's argument (as filtered by you) is now :

a) there are children who don't have autism in early childhood but who develop some kind of quasi-autism in middle childhood
b) this is caused by internet use
c) such cases account for the increase in ASD diagnoses, even though they don't meet DSM criteria for ASD
Do you have any evidence for any of these postulates ?
If that is not what you are saying, what exactly is the claim?

You have also not responded to the point I made about the appropriate place for a scientist to publish new scientific theories. Do you think it is appropriate to make statements about aetiology of a major neurodevelopmental disorder in a non peer-reviewed journal such as New Scientist, when there is no peer-reviewed work to back them up, even if the causal claims are by innuendo rather than direct statement?
If you would like your point of view have broader recognition, I would be happy to publish this correspondence on my blog, so that Greenfield's position and the supposed limitations of my arguments could be given wider publicity.

 pdfs of the following papers were attached:
Baird G, Simonoff E, Pickles A, Chandler S, Loucas T, Meldrum D, Charman T: Prevalence of disorders of the autism spectrum in a population cohort of children in South Thames: the Special Needs and Autism Project (SNAP). Lancet 2006, 368 (9531):210-215.
Baron-Cohen S, Scott FJ, Allison C, Williams J, Bolton P, Matthews FE, Brayne C: Prevalence of autism-spectrum conditions: UK school-based population study. British Journal of Psychiatry 2009, 194:500-509.
Brugha, T. S., McManus, S., Bankart, J., Scott, F., Purdon, S., Smith, J., et al. (2011). Epidemiology of Autism Spectrum Disorders in Adults in the Community in England. Arch Gen Psychiatry, 68(5), 459-465.
Fombonne, E. (2005). The changing epidemiology of autism. Journal of Applied Research in Intellectual Disabilities, 18, 281-294.
Kim, Y. S., Leventhal, B. L., Koh, Y.-J., Fombonne, E., Laska, E., Lim, E.-C., et al. (2011). Prevalence of autism spectrum disorders in a total population sample. American Journal of Psychiatry.
Rutter, M. (2005). Incidence of autism spectrum disorders: Changes over time and their meaning. Acta Paediatrica, 94, 2-15.
Taylor, B. (2006). Vaccines and the changing epidemiology of autism. Child: care, health and development, 32(5), 511-519.
Williams, J. G., Higgins, J. P. T., & Brayne, C. E. G. (2006). Systematic review of prevalence studies of autism spectrum disorders. Archives of Disease in Childhood, 91, 8-15.
Wing, L., & Potter, D. (2002). The epidemiology of autistic spectrum disorders: is the prevalence rising? Ment Retard Dev Disabil Res Rev, 8, 151-161.

P.S. 13.52 on 12th August 2011

A further response from xxx

Dear Professor Bishop,
I am astonished by your peremptory decision to publish our correspondence without permission. I ask you to add the response below, without any editing, as a matter of urgency.

Dear Professor Bishop,
In your first email you stated: “ASD does not suddenly appear in middle childhood - the symptoms are evident from around 2 years of age”. This non-ambiguous statement means that all people who are diagnosed with an Autistic Spectrum Disorder after early childhood will have been displaying its symptoms from around 2 years of age.
You now point out: “it is part of the diagnostic criteria for ASD to have onset in early childhood”. The difference from your initial statement is salient. Thus, it is the case that that unless those symptoms are present in early childhood, an Autistic Spectrum Disorder may not, by definition, be diagnosed.
In this context, you draw attention to “Criterion C in the DSM5 proposed revision”. As I am sure you realise, DSM5 will not supersede DSM-IV until 2013. The criteria you describe as “gold standard diagnostic criteria” are part of a proposed revision.
I shall consider just one matter arising:
Autistic Disorder and Asperger’s Disorder are addressed separately under DSM-IV. The current diagnostic criteria for Asperger’s Disorder (DSM-IV) include the following: “There is no clinically significant general delay in language (e.g. single words used by age 2 years, communicative phrases used by age 3 years). There is no clinically significant delay in cognitive development or in the development of age-appropriate self-help skills, adaptive behaviour (other than in social interaction), and curiosity about the environment in childhood”. Indeed, a delay in social interaction is the only age-related point mentioned; no critical age is given for its onset.
I recognise that the revisions for DSM5 under current consideration are being guided by the following:
”Asperger’s Disorder. The work group is proposing that this disorder be subsumed into an existing disorder:  Autistic Disorder (Autism Spectrum Disorder)”.
If this were to be enacted, diagnosis of Asperger’s Disorder would be precluded, unless its symptoms were present in early childhood (as specified by Criterion C). Again, I feel it is appropriate to ask for evidence which supports your original statement: “the symptoms are evident from around 2 years of age”. According to your gold standard DSM5, this must apply to Asperger’s Disorder. It is reasonable for me to ask whether this has been substantiated by prospective studies which are free from potentially unreliable parental retrospection. I may be in error, but I have found no such study among the papers you kindly sent me. I sincerely apologise if I have overlooked something relevant.
The immensely complex matters of aetiology and diagnosis are not given due consideration if proposed revisions (which are still subject to consultation) are presented as “gold standard”.
In my preceding email I wrote: “I look forward to receiving from you peer-reviewed papers which substantiate your specific claims”. I am saddened to note that you have chosen to misrepresent this polite request as “demanding”. It seems that our discourse will not be fruitful and that it should be closed.