Simon Baron-Cohen from Cambridge University has a long-standing interest in demonstrating that Autism Spectrum Conditions (ASC) are more prevalent in populations presenting with high systemizing skills. These are people who like predicting how things function, classifying observations in systems, etc. and who also show less empathy to others. He has focused his work around the issues of male/female brains and has presented autism as being equivalent to an excess maleness. I should clarify for our readers that in fact, this refers to High Functioning Autism (HFA) or Asperger Syndrome (AS) rather than ASC as a whole. In other words, HFA/AS have a lot more systemizing skills (i.e. some sort of scientific skills) together with a less empathic presentation, and this is essentially what the condition is. He came to this conclusion partly because of how AS/ASF people rate in his Empathy/ Systemizing questionnaire (tests available online, look for Systemizing Quotient and Empathy Quotient). The trouble is that empathy is essentially assessed within a social context and this will confound how AS/HFA people will respond to the questionnaire (and hence be rated). They do have impairment in socialization, in accordance with their diagnosis, but at no point did Simon Baron-Cohen tease out the empathy from the socialization aspect in his evaluations. AS/HFA people have low empathy quotients because they tend to dislike social settings, not primarily because of their low empathy, i.e. extra maleness according to Simon Baron-Cohen. This means, no accurate conclusion can in fact be drawn on the empathic skills (or excess maleness) of AS/HFA people based on this sort of work.
However, this went on for some years with studies after studies, digging deeper and deeper into this Autism High Functioning/ Excess maleness question, moving on to molecular /genetic issues and linking this excess maleness to excess testosterone. He went as far as proposing an antenatal screen based on the mother’s testosterone levels, something he later regretted after facing a number of criticisms. High testosterone levels are related to a number of NON-Autistic personality traits, and are not specific or even representative of autism as a whole, something I discussed here. Simon Baron-Cohen also proposed that autism is on the increase because people with high systemizing brains, IT people, scientists are more able to meet nowadays, thanks to the Internet, marry and have children and that as a consequence of these novel mating trends (assorting mating theory), we have the high rates of autism we see today across the western world.
So, it comes as no surprise to see his latest research in the Netherlands, looking at autism rates in three towns of comparable sizes, with one being presented as being the IT hub of the country. The study can be found here.
What are the main findings of this study?
“The prevalence estimates of ASC in Eindhoven [the IT Hub] was 229 per 10,000, signiﬁcantly higher than in Haarlem (84 per 10,000), and Utrecht (57 per 10,000), whilst the prevalence for the control conditions were similar in all regions.”
This is equivalent to 1 child in 44 in Eindhoven, 1 in 119 in Haarlem and 1 in 175 in Utrecht.
What was the methodology used?
“The schools were asked to provide a count of the total number of children in the school with any of these developmental conditions, speciﬁed per diagnostic subtype and by age and gender. The schools were instructed to only include formal diagnoses in their count (i.e. diagnoses made by a clinical professional, e.g. a clinical psychologist or psychiatrist).”
“Since both attention-deﬁcit hyperactivity disorder (ADHD) and dyspraxia are also developmental conditions and have a similar diagnostic process to ASC, the number of cases with a formal diagnosis of ADHD and dyspraxia were also examined as control conditions.”
“Of the 659 schools invited, 369 schools (56.0%) took part, providing diagnostic information on 62,505 children. Response in the Eindhoven region was higher (75.5%) than in the Haarlem (49.8%) and Utrecht regions (45.7%).”
“Negative binomial regression was used to investigate the multivariable effects of region and school type, with an additional model to investigate the differences between boys and girls.”
For clarification, we do not see actual numbers but only estimated numbers.
The conclusions reached by the authors are:
“The aim of this study was to test a prediction from the hyper-systemizing theory (Baron-Cohen 2006, 2008) that ASC are more common among children in areas where individuals who are talented systemizers are attracted to work and raise a family. Eindhoven is a candidate region of this kind being the hub for IT and technology in the Netherlands.”
“As predicted, this estimate of the prevalence of school-aged children with a formal ASC diagnosis was signiﬁcantly higher in the Eindhoven region, compared to the Haarlem and Utrecht region. This is consistent with the idea that strong systemizing in parents could be a risk factor for having a child with ASC, although there are other factors that could relate to the increased prevalence in the Eindhoven region.”
The authors have listed a few possible confounding factors: Possible higher awareness of ASC in Eindhoven (parents and professionals), over diagnosis services in Eindhoven or under diagnosis in Utrecht. Interestingly, it is stated in the discussion: “If the responding schools are representative of the comparable schools in the region, the difference in response should not have confounded our findings.” If, indeed (see below).
What do you think of this?
Let me tell you what I think of this study.
1-First of all, why have the authors not confirmed that parents are indeed in IT technology, if they want to link IT parents to ASD? It would be an obvious factor to capture in their evaluation, and one that is crucial to their main conclusion, I would have thought.
2-Secondly, there are only superficial explanations as to why there is such a high discrepancy in the schools response rates across these three cities. Are there more special needs schools than the responding ones from Eindhoven, or are the numbers the same? No mention of this. What are the actual ages of the kids covered by the survey in the three towns? Why is there so little information available to define these 3 sample groups? I would guess, it is because they have actually limited information on the children, because of the way the population was screened. We must stress that the main potential explanation to a difference in rates is that the populations have been captured differently- and this has to be fully addressed, with fuller transparency in order to proceed towards any kind of conclusion.
3-Thirdly, are dyspraxia and ADHD really a suitable validation to show that the populations have been captured in similar representative manners? – In my opinion, no they aren’t and the reason is that unlike what the authors claim, Dyspraxia (at least in the UK, and I doubt very much it is different elsewhere) is essentially NOT diagnosed even when the kids have clear motor planning issues. As for ADHD, we also have a huge number of children in schools, who present with hyperactivity that are simply seen as being difficult and challenging without receiving any proper diagnosis. I would estimate that the figures given are well under the actual values for Dyspraxia and ADHD. In other words they are not accurate estimations of numbers and therefore cannot be used as reference points.
4- Fourthly, what else is happening on Eindhoven? A quick look in Wikipedia gives us a good account of the actual high industrial development.
“Philips’ presence is probably the largest single contributing factor to the major growth of Eindhoven in the 20th century. It attracted and spun off many hi-tech companies, making Eindhoven a major technology and industrial hub. In 2005, a full third of the total amount of money spent on research in the Netherlands was spent in or around Eindhoven. A quarter of the jobs in the region are in technology and ICT, with companies such as FEI Company (once Philips Electron Optics), NXP Semiconductors (formerly Philips Semiconductors), ASML, Toolex, Simac, CIBER, Neways, Atos Origin and the aforementioned Philips and DAF.”
“Eindhoven has long been a centre of cooperation between research institutes and industry. This tradition started with Philips (the NatLab was a physical expression of this) and has since expanded to large cooperative networks.”
Do these industries relate in any other ways to the higher Autism rate? This possibility has not been explored in the discussion section. If a quarter of the total work force causes a near 4 fold increase in autism rate compared to the population of Utrecht, that would suggest a 16 fold increase in the likeliness to have children with autism in the IT profession. Surely something that would have been picked up before, other prevalence studies have not found any correlation between autism and social status.
5- Lastly, and I will stop it here, let’s look at what is happening in Utrecht, the town that was found to have the lowest rate of ASC. Again, a look at wikipedia gives a good account of the city development and this is what is found, as you will see there is hardly any industry at all. It is interesting to see that the university is the largest in the country, one would have thought, according to the systematizing brain theory, that it would have foster a large concentration of systemazing brains, apparently not.
The town looks so peaceful that I could not resist including a photo of the Dom Tower above.
“Utrecht University, the largest university of the Netherlands, as well as several other institutes for higher education.
“The economy of Utrecht depends for a large part on the several large institutions located in the city. Production industry has a relatively small influence in Utrecht. Rabobank, a large bank, has its headquarters in Utrecht.”
“Utrecht is the centre of the Dutch railroad network and the location of the head office of the Nederlandse Spoorwegen (Dutch Railways). NS’s former head office ‘De Inktpot’ in Utrecht is the largest brick building in the Netherlands (the “UFO” featured on its facade stems from an art program in 2000). The building is currently used by ProRail.”
Railway industry? Could it be another niche of high concentration of AS/HFA, known to like railways, trains, time schedules… well apparently not. (That was a joke BTW).
Can anyone tell me how the data presented in this paper substantiates a link between Parental IT Hub/ high systemazing brains and Autism?
I appreciate the authors observe some caution in the presentation of their data, none the less, wouldn’t it have been wiser to have more conclusive findings to present for a publication? What are those people living in Eindhoven supposed to make of the news reports that accompanied the publication? And what are all those IT/ scientists parents also supposed to think? Are they really at higher risk to have children with an autism spectrum conditions?
And aren’t the implications as far as the whole spectrum is concerned somehow insulting to all those parents who have reported another side of autism, the regressive autism, with the novel development of severe clinical issues alongside the behavioural changes that characterise autism? Presenting autism as a condition resulting from a specific brain-wiring difference trivialises the distress experienced by affected individuals and their family.
Baron-Cohen, S, & Wheelwright, S, (2004) The Empathy Quotient (EQ). An investigation of adults with Asperger Syndrome or High Functioning Autism, and normal sex differences. Journal of Autism and Developmental Disorders, 34, 163-175.
Baron-Cohen (2003). The Essential Difference: The Truth About the Male and Female Brain. Basic Books, Perseus Books Group.
ADHD is a condition characterized by three main core features; hyperactivity, impulsivity, and attention problems.
It is estimated that it affects between 3-5% of children of school age, though other estimates suggest this could be as high as 10% of the school population (Rowland et al. 2002). Fewer children of Hispanic American and African American backgrounds seem to have received a diagnosis of ADHD compared to White Americans, however it is very possibly due to increased poverty and reduced insurance coverage and possibly other cultural issues rather than a true difference with respect to ethnicity. The condition not only affects the US, but also most western countries (e.g. UK, NICE guidelines 2008, rest of Europe) andNorth Africa. ADHD rates have been found to dramatically increase over the last two decades and the condition is now the most commonly diagnosed in school age children (e.g Mandell et al. 2005; and more recent reports).
ADHD is often seen in Autism Spectrum Disorders (ASD). And like ADHD, ASD rates have also dramatically increased in the last 2 decades.
So what options do we have in response to ADHD, whether this is in association with Autism or not?
The most common answer is Ritalin-like medication. A very large proportion of children diagnosed with the condition receives medication. Children as young as 4 years old are being prescribed drugs. This causes major concerns both in the US and in the UK.
Simple interventions however do exist. Nutrition, particularly with good supplementation of Omega-3 (Richardson et al. 2006), but also dietary modification with exclusion of foods the child is sensitive to (Pellser et al. 2011) have been found to be effective.
It is worth highlighting this latest Lancet study.
The study used a number of ADHD rating scales based on parents, teachers (non-blinded) ratings or blinded paediatrician ratings. The strengths are multiple ratings, large sample size (n=50) with matched controls (n=50), overall heterogeneous population representative of the general population of children with ADHD.
What it is important to know is that the restricted diet the children are placed on is not designed based on any allergy test- it is a standard diet that consists of a few hypoallergenic foods, rice, meat, vegetables, pears and water, completed with fruits, potatoes and wheat. The challenge test conducted after 5 weeks on the respondent children consists of either 3 high IgG foods or 3 low IgG foods. Whether or not the foods were of low IgG or high IgG response, there was deterioration of ADHD symptoms upon challenge.
I have been in touch with the Lead author, Dr Lidy Pelsser who clarified that the low Ig G group means Zero IgG level- if this is the case, the group’s conclusions that IgG levels are irrelevant to ADHD symptoms is correct. In my experience of allergy tests though, at least with regard to IgE levels, a low level does not equal to zero or not detectable. A low level is a low detectable level and this can have biological implications.
I have also contacted the lab ImuPro in Australia to get some clarification with regard to their rating of low IgG levels. If Low IgG level are detectable low level, the conclusion on potential relevance of IgG testing would potentially be quite different, until it is proven that response to foods with No IgG at all (i.e. not detectable as opposed to low) also lead to return of ADHD symptoms.
These methodology details are however not very important in the light of the very good outcomes of the restriction diet: see figure 2 of the paper.
Figure 2: Distribution of behaviour scores at start and end of the first phase. Scores according to (A) masked paediatrician ratings and (B) unmasked teacher ratings. To facilitate comparison between the various measures, scores have been standardised as percentages of the maximum score per measure. Bars=maximum and minimum score. Shaded boxes=interquartile range. Horizontal bars within boxes=median. ADHD=attention-defi cit hyperactivity disorder. ARSall=ADHD rating scale total score (maximum score 54). ARSatt=ADHD rating scale inattention score (maximum score 27). ARShyp=ADHD rating scale hyperactivity and impulsivity score (maximum score 27). ODD=oppositional defi ant disorder (maximum score 8).
The authors conclude that a trial diet for a 5-week period should be proposed to every child with ADHD, followed by a challenge procedure to define which food the child would react to.