New autism prevalence 1.5% in UK

31 May

A new study published (officially) tomorrow discusses ‘Prevalence of autism-spectrum conditions: UK school-based population study’.

Its an interesting study for quite a few reasons. Firstly, it offers a new autism prevalence of 1.5% (1 in 66). That’s the message that the press will no doubt focus on (and, as Kristina blogs, already have). And I’ve absolutely no doubt that our friends from JABS, Age of Autism and various other anti-vaccine fringe groups will be painting this as part of their ‘evidence’ that we’re in the throes of a massive autism ‘epidemic’.

However, the paper itself is very nuanced and is clear in its messages. However, to be absolutely sure I was correct in the conclusions I drew I had an email conversation with Professor Baron-Cohen before writing this entry.

Point 1: This study confirms the baseline rate of 1% as asserted by the Baird et al paper

Baird et al (2006) asserted that their findings would offer a baseline rate of autism prevalence, that prevalence being 1%. This figure was ascertained by looking at a SEN population of a South Thames cohort. Baron-Cohen et al (2009) confirmed that figure:

These authors took the decision to screen only the ‘at-risk’ population and assert that their estimate should be regarded as the minimum figure. Our results from screening the entire school-aged population support this assertion…

In other words Baron-Cohen et al also looked at the ‘at risk’ population and also found a prevalence of 1%.

Point 2: Baron-Cohen et al identify a further 0.5% to make a total prevalence of 1.5%

What is different about the Baron-Cohen paper is that as well as looking at the ‘at risk’ group they _also_ looked at mainstream schools. Using the CAST screening tool, this study identified a previously unknown prevalence of 0.5% within this mainstream environment.

Our results from screening the entire school-aged population…highlights the reality that there are children with autism- spectrum conditions, notably children with high-functioning autism, who remain undetected in primary schools. These children may use strategies to mask their social and communication difficulties such as going to the computer room at playtime. They may be quiet and cooperative at school and not difficult to manage and therefore teachers may not be aware that they have difficulties. Primary schools in the UK are typically small and foster a supportive and nurturing environment. It may not be until these children move to secondary school that their true differences are revealed.

Often I have heard people asking how it is possible that people with autism could possibly be missed. The Baron-Cohen et al paper gives a graphic answer to that question.

Point 3: Caution should be applied in assuming that results ascertained in Cambridgeshire could be applied across the rest of the country

The area is very affluent within the UK and has excellent autism resources for autistic children. It is a given that many families have moved into the area to try and exploit those services. This would have a positive effect on prevalence that is not consistent with the majority of the UK.

Our study does not report on migration of families but given the level of services for and awareness of autism-spectrum conditions in Cambridgeshire, this remains a distinct possibility. Caution should therefore be employed in assuming that the figures reported here can be applied nationwide.

Professor Baron-Cohen and I had the following exchange about the autism ‘epidemic’:

KL: What would you say to someone who says that your paper is strong evidence of an ‘autism epidemic’ (because you know they will)?

SBC: I think the term ‘epidemic’ of most value in relation to contagious diseases, which autism is not.

KL: Can I rephrase my question? Would you say your findings support the idea that there has been a true rise in prevalence? As oppose to the seven items you say have caused a seeming rise in autism earlier in your paper?

SBC: There has been a real rise in prevalence but what is at issue are the causes of this rise. In the paper we summarize the quite ordinary factors that might have driven the rise, such as better recognition, growth of services, and widening diagnostic criteria.

So next time someone who likes to bandy about the phrase ‘epidemic denier’ like he knows what he’s talking about when he claims that the ‘epidemic deniers’ say that autism is just better recognised these days, tell him there’s a lot more than just one reason:

Prevalence estimates for autism-spectrum conditions have shown a steady increase over the past four decades. In 1978, the consensus estimate for classic autism was 4 in 10 000; today autism-spectrum conditions (including classic autism) affect approximately 1% of the population. This massive increase is likely to reflect seven factors: improved recognition and detection; changes in study methodology; an increase in available diagnostic services; increased awareness among professionals and parents; growing acceptance that autism can coexist with a range of other conditions; and a widening of the diagnostic criteria.

18 Responses to “New autism prevalence 1.5% in UK”

  1. María Luján May 31, 2009 at 23:14 #

    This manuscript does not consider the importance of the number of home-schooled autistic children or autistic children that are not under a school method at home- but learning sign language or other AC methods. If the home schooled autistic children are going to be considered in the analysis, probably the numbers would be even higher.
    1-Are there statistics of UK about the number of home-schooled autistic children ?
    2-Are there statistics in Uk of autistic children receiveing some kind of education at home but not home schooled- formally- or included in public/private schools?

  2. Kev May 31, 2009 at 23:19 #

    Easy answer to both Maria – no 🙂

    I think that’s part of the point of his caution – Cambridgeshire may not be representative of the whole of the UK.

  3. Joseph May 31, 2009 at 23:31 #

    In the US, about 2% of children are home-schooled. I’m not sure what it is in the UK. In any case, it’s unlikely the numbers would change a lot.

    Even if home-schooled children had twice the prevalence of autism, the overall prevalence would be about 1.53 in 10,000.

  4. Joseph May 31, 2009 at 23:37 #

    That’s 1.53% in my last comment, not 1.53 in 10,000 🙂

  5. dr treg May 31, 2009 at 23:54 #

    It is interesting that autism (previously known as childhood schizophrenia) and schizophrenia share a prevalence of 1 in 100, are associated with immunological abnormalities, genetic abnormalities,dendritic spine abnormalities and clinical features.
    It seems that others have considered the possibility that autism and schizophrenia are two sides of the same coin.

  6. Matt June 1, 2009 at 03:31 #

    There is an interesting cross over (multiple complex developmental disorder (MCDD)) between autism and schizophrenia where the kids display autistic symptoms early in life and tend to develop schizophrenic behaviors later as they age. Their autistic symptoms tend to be less severe than in profoundly autistic kids. The interesting, to me, aspect of this is that classicly autistic kids, at least in literature I have seen, show elevated cortisol levels, whereas the MCDD kids show flat marginal cortisol responses to new stressors. A cite on the research is below.
    Multiple Complex Developmental Disorder (MCDD) represents a distinct group within the autistic spectrum based on symptomatology. Unlike autistic children, part of MCDD children develop schizophrenia in adult life. Despite the differences, patients of both disorders are mainly characterized by abnormal reactions to their social environment. At the biological level, we showed in a previous study that MCDD children have a reduced cortisol response to psychosocial stress. Given the fact that autistic children clinically show more social impairments, it was hypothesized that they may have even further decreased cortisol responses to psychosocial stress than MCDD patients. Therefore, 10 autistic children were compared to 10 MCDD children and 12 healthy control children in their response to a psychosocial stressor, consisting of a public speaking task. In order to test whether any impairments in the biological stress response are specific for psychosocial stress, the autistic children were compared with 11 MCDD children and 15 control children in their response to a physical stressor, consisting of 10 min of bicycle exercise. Heart rate and salivary cortisol levels were used as indicators of response to the stress tests. Autistic children showed a relatively elevated cortisol response to psychosocial stress, in contrast to MCDD children who showed a reduced cortisol response. No differences in heart rate or cortisol responses to the physical stress test were found. The specific difference between autistic and MCDD children in their cortisol response to psychosocial stress indicates that the disturbed reactions to the social environment observed in these disorders may have different biological backgrounds.

  7. passionlessDrone June 1, 2009 at 14:58 #

    Hi Matt –

    Very interesting. I wonder what impact age has on cortisol levels in a ‘normal’ population? If those values remained stable, perhaps this could be a mechanism to further get a handle on if there are more people with autism in the range under eighteen versus those that are in the thirty/forty/fifty range.

    I guess what this would really tell us is, are there the same number of people with elevated or decreased cortisol responses in different age cohorts. Even still, as part of a larger arsenal of biomarkers, it could be valuable information.

    It would also be very interesting to see what the cortisol response from some of these kids with ‘autism’ who don’t need a diagnosis or services is like.

    – pD

  8. cpu52362 June 1, 2009 at 21:34 #

    Faking ones way through life trying to survive is not the same as not needing a diagnosis pD.

    I would like to explain more, but I am not sure I have the words that would make it any clearer for you.

  9. passionlessDrone June 1, 2009 at 22:04 #

    Hi CPU52362 –

    Well, in this instance, I believe it was the researcher who said that many of these children don’t need a diagnosis. It just isn’t listed here; I’d read it in a different analysis of the story elsewhere.

    He said the undiagnosed cases are likely to be at the mild end of the spectrum, are coping well with their families and may not need a diagnosis.

    Prof Baron-Cohen said that for some families the autism label may ‘raise anxieties’ and be intrusive, rather than helpful.

    I’m not sure where this statement fits into your criteria.

    – pD

  10. laurentius June 1, 2009 at 22:11 #

    Do you know what the truth is because I do not think Cambridge do.

    That is that autism is a socially constructed diagnosis and disability, to wit the rate and proportion will continue to increase because whilst human evolution and diversity does not change, the social environment we find ourselves in does.

    We are increasingly marginalised and disabled by a society that advances in directions that do not allow us our full humanity, and those who marginalise us are effectively war criminals, because this is class war.

  11. rajensen088 June 2, 2009 at 13:22 #

    The APA working group for the PDD’s last month published its recommendations for changes in diagnostic criteria for the PDD’s that will be coded in the forthcoming DSM-V. If these recomendations are implemented, prevelance rates will climb to as high as 1/4 rendering the concept of autism even more meaningless than currently diagnostic schemes.

    Recommended changes would remove the term Pervasive Developmental Disorders and replace it with Autism Spectrum Disorders. Autistic Disorder, PDD/NOS and Asperger Syndrome would be removed and replaced with a severity scale (mild, moderate, severe). Two new categories would be added, Atypical ASD and normal variation (socially iolted or social ‘awkwardness’.

    Any person who is on a introverted side of a Bell Curve of normal variation of personality types would now be labelled as having a personality disorder (ASD). The prevelance of introverted personality types as been consistently calculated as 25% of the general population, hence the eventual prevelance rate o 1/4.

  12. Prometheus June 5, 2009 at 22:06 #

    When do we reach the point where autism – as it is variously diagnosed – becomes so common that the “diagnosis” is useless? If the criteria have become so broad as to include 1.5% of children, is there any point to making such a useless diagnosis?

    There are four reasons to make a diagnosis:

    [1] Treatment – most treatments in medicine are diagnosis-specific, so making an accurate diagnosis is the first step to applying the correct treatment.

    [2] Prognosis – most disorders have a consistent course or at least a known range of outcomes. Making a diagnosis allows the patient (or their parents) and practitioner to make appropriate plans for the future.

    [3] Epidemiology – accurate diagnosis allows epidemiologists to track a disorder to see if it is increasing, decreasing or stable in prevalence. Treatment research also needs accurate diagnosis in order to both select subjects and controls but also to see the effects of treatment. If the diagnosis becomes too broad or is constantly changing, it is impossible to do either accurate epidemiology or accurate treatment research.

    [4] Patients and their families (and their physicians) do better when they have a name for the disorder. It seems more settled to be able to say “I have terminal moraine.” than to say “I’m not sure what I have, but my gums itch all the time.” A name gives the illusion of control over the disorder, even if there is no treatment possible.

    We have reached the point where the only benefit to the diagnosis of “autism” is [4].

    Also, when a condition reaches a certain prevalence in a population, it starts looking more like a “trait” and less like a “disorder”. At some point – as “RAJensen” above alluded – the diagnosis of “autism” will be so broad as to convey no useful information at all. It will be of no more practical use than skin or hair color.


  13. dr treg June 5, 2009 at 22:50 #

    Re. Treatment: This assumption may be incorrect. Conventional treatments are not always “correct”. 46% of conventional treatments have unknown effectiveness.
    Re. Prognosis: Doctors focus on diagnosis rather than prognosis. Actuaries are specialists in prognosis.
    Re. Epidemiologists: Autism is not a notifiable disease and reporting is probably variable due to there being so many agencies involved.
    Re. Having a name for the disorder: Is there evidence that having a name for the disorder improves the outcome in autism?

    The word “autism” will probably be equated with a general symptom rather than a diagnosis which will lead onto investigating for a specific diagnosis in the future i.e. similar to “cough” leading onto different investigations for the cause of cough.

  14. rajensen088 June 16, 2009 at 12:51 #

    How vague, ambigous and subjective the concept of ‘autism’ has become can be seen in studies of previously healthy adults who are stroke patients.

    Studies over many decades have demonstrated that adult stroke patients, as a group, are seen as possessing significant social-communication impairments eg:

    The proposed changes that may be included in the DSM-V related to one of the designations (Less severe ASD) is defined as:

    “Clear impairments in social communication. Meets all diagnostic criteria including symptom severity greater than threshold”

    It would be interesting to apply one of the ‘Gold Standard’ diagnostic schemes (ADOS-G) to a representative group of previously healthy adult stroke patients to test the validity and specificity of ADOS-G

  15. Joseph June 16, 2009 at 14:14 #

    The ‘autism’ construct may be vague, subjective, and its mapping to anything biological unclear, but I have seen no evidence that this was any different with the narrowest, most strict criteria devised: that of Kanner’s.

    The outcome, treatment and epidemiology of Kanner autism was no more well defined than that of ASD as understood today.

  16. rajensen088 June 17, 2009 at 03:58 #

    Kanner’s definition was very highly specific. He described all the children in his paper as so unresponsive to people, including mothers and fathers that “If mother or father were absent for an afternoon or a month upon returning the children appeared to be unaware that the mother or father was ever absent”. Compare that to parents description of their ASD children on internet blogs.

    Kanner’s definition ‘A pervasive lack of responsiveness to other people’ was encoded in DSM-III. By 1994 (the start of the so-called autism epidemic) Kanners concept was completly removed from all diagnostic schemes including DSM-IV-TR, ICD-10 and every single one of the so-called ‘Gold Standard’ diagnostic tools such as ADOS, ARI, being replaced by the puzzling ‘Qualitative impairment in social reciprocity’.

    With respect to comparisons between ASD and previously healthy adult stroke patients you do realize that the best evidence based treatements, in many cases are the same, eg Physical Therapy (motor coordination), Occupational Therapy (fine motor coordination) and of course Speech and Language therapy.

    The major differnce between these two groups are. in ASD the brain damage is caused by a disruption in early brain development while in adult stroke patients the damage is done to a previously fully developed brain.

    Outcomes in both groups are variable with a percentage range of between 5% and 20% associated with a normal outcome.

  17. Joseph June 17, 2009 at 13:39 #

    Kanner’s definition was very highly specific. He described all the children in his paper as so unresponsive to people, including mothers and fathers that “If mother or father were absent for an afternoon or a month upon returning the children appeared to be unaware that the mother or father was ever absent”. Compare that to parents description of their ASD children on internet blogs.

    So? On what basis is a construct that includes “a pervasive lack of responsiveness to people” more valid than a construct that doesn’t include it? There are probably reasons why that kind of fell out of the picture. It might not have been useful to include in the criteria.

    Additionally, as Kanner’s patients grew up, this “pervasive lack of responsiveness” was not always there. Even as children, you see in the 1943 descriptions that they were not entirely unresponsive to people.

    Kanner also said autism involved “either mutism or the kind of language that does not seem intended to serve the purpose of interpersonal communication” (Kanner, 1951.) Yet, few of the children he saw could be described in this manner, apparently.

    The major differnce between these two groups are. in ASD the brain damage is caused by a disruption in early brain development while in adult stroke patients the damage is done to a previously fully developed brain.

    The major difference is that brain damage in stroke is evident, whereas in ASD it is not. There are other obvious differences. For example, a stroke can leave you paralyzed or even kill you, whereas ASD does not. Average survival after stroke is 7 years. Yes, I’d say there’s a bit of a difference in outcome.

  18. Erik October 14, 2009 at 19:35 #

    This front page Atlantic Magazine article is a great counter point to the new stats:

Leave a Reply to passionlessDrone Cancel reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: