Simon Baron-Cohen: The Man Overturning Conventional Thinking on Autism

The Cambridge professor claims the syndrome may be related to differences between the male and female brain.

The homepage of the Cambridge Psychology website looks very innocent. The top features a black strip embedded with pictures of faces: a woman thinking, a man with a slight smile on his lips, a thoughtful child. “Welcome to the Graduate Parents Project,” declares the blue heading. “We are a research group, within the Departments of Psychology and Psychiatry at the University of Cambridge, specializing in online psychological research. This research project is testing if there is any link between what you as parents studied at university and how your child is developing.”

Behind the description lies an ambitious, and certainly provocative, theory. Among other things, the goal of the research is to reveal whether parents with autistic traits (even if they are not diagnosed as autistic per se) who choose certain professions (for example, engineering, physics or mathematics) and marry others in those fields with similar traits, will tend to give birth to children with a higher risk of autism.

This online study is the brainchild of Prof. Simon Baron-Cohen, the director of the Autism Research Center at the University of Cambridge. He’s a British Jew with warm ties to Israel, the cousin of entertainer Sacha Baron-Cohen, and one of the leading and most controversial researchers of autism in the world today.

At the end of November, he was scheduled to visit Israel to speak at the conference of the Israeli Association for Child Development and Rehabilitation, a visit he says he was eagerly anticipating. But Operation Pillar of Defense was launched and he was forced to give his speech by videoconference. In measured tones he spoke about his theory, which could explain autism and also perhaps teach us a great deal about the differences between the male and female brain, and about the nature of talent.

When Baron-Cohen embarked on his research career in the 1990s, he actually wanted to find an answer to one of the greatest mysteries in the field of autism: Why are there many more autistic males than females? When American psychiatrist Leo Kanner studied autism in the 1940s, he discovered that the ratio of boys to girls suffering from the syndrome was 4:1. In 1944, when Austrian pediatrician and child psychiatrist Hans Asperger began to describe the syndrome later named after him (a milder form of autism), he first diagnosed it only in boys. Asperger theorized that the symptoms he discovered were “a certain version of male intelligence.”

Today we know that there are nine times as many boys as girls among those diagnosed as suffering from Asperger’s. But for 50 years there was no explanation for the phenomenon. When Asperger’s own writings were translated into English, Baron-Cohen – who was beginning his career in autism research at the time – thought it would be interesting to examine the connection between the syndrome and gender. What he discovered caused him to call the autistic brain “an extreme form of the male brain.”

Of course, it is hard to describe the differences between the male and female brain without stepping into a mine field of gender politics. “There are basic differences, even if they were denied for decades,” Baron-Cohen claimed in his video lecture. For example, the male brain is, on average, larger than the female brain, but the female brain develops earlier and faster. The brain centers relating to language and processing emotions are on average larger in women.

“Hundreds of studies of various cultures have proven that, on average, boys play more than girls with constructional toys like Lego and toy cars, and girls play more with dolls,” says Baron-Cohen. For those who hasten to attribute the differences to purely social influences, he adds that there have been similar findings among certain species of monkeys, too. And, of course, men and women tend to choose different professions: engineering and mathematics, as compared to therapeutic professions, respectively.

In order to eliminate the social component, Baron-Cohen and his research collaborators showed newborn infants a face and a mechanical object and even the infants conformed to the gender stereotype: More girls looked at the face for a longer time, and more boys looked at the object longer.

Baron-Cohen describes the differences between the male and female brain on an axis ranging from “empathizing” to “systemizing.” Empathizing is being able to identify the thoughts and emotions of another person, and having an appropriate emotional response to these. Systemizing is the drive to understand and construct systems of all kinds, from software to a car engine. Usually, greater ability to empathize means less ability to systemize, and vice versa. There are few people who excel in both.

Women usually score higher than men in tests that examine empathizing, whereas men generally score higher in systemizing. In both cases, the autistic brain is off the charts of the “normal” male brain. In empathy tests, typical men score lower than women, and autistic people without any mental retardation with a relatively high IQ and normal language ability score even lower than men.

There are no surprises up to this point. But in questionnaires examining the level of systemizing manifested by an interest in complex systems, understanding of such systems and attention to detail typical males score higher than women, whereas autistic people have the highest score. People suffering from autism, says Baron-Cohen in a phone interview, spend a lot of time playing with objects, but tend not to look at human faces – the large amount of information that is expressed in faces is simply too confusing. On the other hand, they are capable of spending hours carefully examining a complex picture, recalling precisely how an engine is built or building magnificent towers from Lego.

Looking for the root

In his latest research, Baron-Cohen is trying to progress beyond circumstantial evidence and to reach the root of the phenomenon of autism to be more exact, the womb. The connection he found between the male brain and the autistic brain led him to wonder whether there could be a connection between the level of male hormones to which the male or female fetus is exposed in the mother’s uterus, and the chances of developing autism.

In a study he conducted in Cambridge, he examined the level of the male hormone testosterone in amniotic fluid taken from 235 pregnant women.

Eight years later, he returned to the same women and examined their children for the presence of autistic traits (that is, traits found not only in autistic people, but which in high concentration are liable to attest to autism: for example, a tendency toward seclusion, linguistic difficulties and a rigid personality). He discovered that the higher the level of testosterone in the fetal fluid, the larger the number of autistic traits in the children.

In a new study, on which he reported at the conference last month, Baron-Cohen was assisted by comprehensive research conducted in Denmark. Samples of fetal fluid from 1980 to the present were saved in that country; Baron-Cohen had over 100,000 samples at his disposal.

The Danes also meticulously record all the children diagnosed as autistic, which enabled him to cross-reference data and find a possible connection between the levels of testosterone in the fetal fluid of a specific individual, and the chances that he or she would be diagnosed as autistic later on. The results of that study, the researcher says, will be available in another two months.

What can we learn from the theory about women or girls who may have a talent for mathematics and engineering for instance, do they have a high level of fetal testosterone?

Baron-Cohen: “Possibly, but we have to be cautious because in our study the tests of systemizing were done at ages 6 to 9. It would be interesting to visit these children when they grow up and choose what to study at university, and to see if it is possible to predict their choice of degree – science versus humanities, for example – from the levels of fetal testosterone.”

If the Danish study proves what Baron-Cohen thinks it will, it will be possible to claim with certainty that autism has hormonal causes (too), which opens the door to a large number of possibilities. He warns, however, against the idea of making a prenatal diagnosis of the syndrome, although if his findings are proved correct, they may also be used for developments in this realm by other scientists.

What about developing medicines? Will it be possible to develop a way of treating the level of fetal testosterone? Baron-Cohen emphasizes that this is not his intention: “It’s too early to talk about that, for two reasons. First of all, there has been no study about the risks of hormone therapy for the mother or the child. Testosterone influences not only autistic traits, and if we manipulate hormones, we may change the way in which the child develops in multiple ways.

“For example, such a medication might reduce the child’s interest in mathematics or physics. The second question is ethical: What exactly are we treating? The idea of a cure for autism is itself controversial. Some people with autism say they don’t want to be cured, because autism gives them a different way of looking at the world.”

What Baron-Cohen suggests is in effect a new way of looking at the disability. “It is possible that by studying autism we’ll learn about the nature of talent. Supposedly there’s no connection between scientific talent and autism, but if we look closely, we find a very basic connection. If you have a child with autism, and he or she has good intelligence, with no delay in language learning maybe there is an advantage to autism as well? Maybe it gives them a better understanding of mathematics, or science? After all, the essence of science and the essence of autism is to notice patterns that others have not noticed.”

You refer only to high-functioning autism. Your optimistic assessments are liable to sound unrealistic to parents of autistic children without language skills or with a low IQ. For how many autistic people is it even relevant?

“One study, which was published in 2006, estimated that 50 percent of those diagnosed with autism have an average or above average IQ. You need a lot of time and money to do population studies, which include thousands of children, and to look at the percentages and that doesn’t happen often.”

According to the Centers for Disease Control and Prevention in the United States, 62 percent of American children diagnosed as autistic have an IQ of 70 and above.

For people with autism who conform to this profile, Baron-Cohen notes, fields such as theoretical physics, mathematics or engineering – which require a high level of systemizing but don’t require much communication with people – can be an excellent field of study. The same is true of computer-based learning, as opposed to learning that requires interaction with the teacher.

Baron-Cohen is now examining whether the tendency to choose partners who are similar to us which is called assortative mating also comes into play in the context of autism. Many studies have proven, for example, that people suffering from depression, bipolar disorder or personality disorders tend to marry people similar to them. Baron-Cohen seeks to examine the theory of whether two people with a tendency toward systemizing – a tendency that he identifies by their choice of profession – will give birth to children with a higher risk of autism.

Does this mean male and female engineers, or male mathematicians and female physicists, should be careful in choosing a partner?

“I’m not worried that people will take such things into consideration. The choice of a partner is usually not rational. People fall in love or feel that they suit one another. I do think it’s important to understand the genetics and if the genes involved in mathematical ability or in understanding systems really are related to genes for autism, that could change the way in which we think about autism. Not only as a disability, but also as a talent.”

The Dutch test case

In order to examine his theory, last year Baron-Cohen chose Eindhoven as a test case. The Dutch city has been the home of the Philips electronics company for the past 100 years and one of The Netherlands’ technology centers.

Baron-Cohen decided to study Eindhoven after receiving a report from parents there about a high percentage of children with autism. Indeed, he found that children there had a two to four times greater risk of being diagnosed as autistic when compared to children in two other Dutch cities of similar size. Ostensibly, there was proof here of the fact that parents with a high tendency toward systemizing (i.e., working in high-tech) will give birth to a higher percentage of autistic children.

However, the study was also criticized, mainly because Baron-Cohen did not mention the age and education of the parents data that have been found in other research to be related to the incidence of autism among children.

It was pointed out that a study done in California in 2010 did not, for example, find a concentration of children diagnosed as autistic around Silicon Valley, but rather in areas with highly educated parents, who may have tended to diagnose their children more.

Baron-Cohen reacts to criticism of his study by claiming that the aim was not to collect information about the parents, because the source of the data was documentation by schools.

The “Graduate Parents Project” project on the Internet, which seeks parents who will answer some questions and guarantees strict confidentiality, is looking at the respondents’ age and education.

As far as Silicon Valley is concerned, another study, done in 2009 in the San Francisco area, examining a connection between the parents’ profession and the chance that their children will be diagnosed as being on the autistic spectrum, found a slightly higher probability of the syndrome among the children of mothers engaged in technological professions. No similar connection to the father’s profession was found, however.

To what extent are Baron-Cohen’s theories accepted? According to Dr. Mitchell Schertz, director of Neurodevelopmental Services for Meuhedet Health Services in the Northern and Samaria District and chairman of the Israeli Association for Child Development and Rehabilitation, “Baron-Cohen offers a way of thinking that is not often examined. It’s not a proven fact, [but] this is a man who is very talented and organized in his thinking he is constructing a theory and trying to establish it. The theories are provocative, but he also has good data.”

Schertz adds that the problem is that autism in general has not been sufficiently studied. “Only in the past 20 years have we begun to study the field systematically. We don’t know much.”

Are Baron-Cohen’s studies relevant to us? More than we tend to think. According to his 2009 study – in which the mothers of 3,370 children filled out an autism-spectrum questionnaire – for every three children in the sample who had already been diagnosed as autistic, there were another two who had undiagnosed autism. These data conform to what some journalists call the “autism epidemic” the sharp increase in the percentage of people diagnosed with the syndrome in recent decades. In the United States, according to Centers for Disease Control and Prevention figures, the frequency of autism in 2008 was 1 out of 88 children, as compared to 1 out of 150 in 2000.

In Israel, according to data supplied by Schertz, the frequency among children up to the age of 12 has reached 0.48 percent: i.e., 1 out of 150 children. That is a substantial increase compared to the statistics from the 1980s, when 1 out of 1,000 children reportedly had autism.

Is this a real epidemic? Baron-Cohen refuses to overreact: “It’s almost certain that the increase in autism can be attributed to routine causes such as better identification, increasing awareness, the growth of clinical and educational services, and the expansion of the definition/diagnosis of autism to include Asperger’s syndrome and partial cases” a reference to the so-called autistic spectrum of disorders. And the relatively early age at which children are tested for the syndrome today (starting at the age of 2) is likely to cause an increase in the number of those diagnosed as suffering from it.

Schertz admits that the definition of autism has become much broader in recent years. “When Kanner described the first cases of autism in the 1940s, he was talking about children with significant damage to intelligence, with retardation. Only in recent years are there also descriptions of high-functioning children with Asperger’s. There’s a lot that we don’t know [there are] all kinds of syndromes that cause autistic behavior, but many children with autism without those syndromes. There is no specific and focused genetic cause here, as in Down Syndrome.”

But Schertz does not accept the theory of diagnosis being more efficient today and accounting for a higher rate of autism. “The moment an organization like the CDC coordinates reports from various centers in various places, and everyone reports the same thing, we can reasonably assume that the findings represent reality. I do see a change between 2002 and 2012. And I ask: Where were all the autistic children 20 years ago? Who examined them? There were far fewer special kindergartens to treat autism. What organizations did they attend?”

Maybe they fit into ordinary frameworks, like children with learning disabilities?

“A child with a learning disability functions fully in the community, gets married, starts a family, participates in society. A child with autism does not function normally on a daily basis. He needs all kinds of support. That’s why the probability that they all attended ordinary kindergartens seems less likely to me. I’m expressing myself cautiously, because I think there are lots of things that we don’t know.”

For his part, Baron-Cohen warns against overdiagnosis. “Here, too, as in every other medical situation, a diagnosis is required only when the person is suffering. If someone finds himself in a job where he or she doesn’t have to interact too much with people, and, more importantly, if the peer group responds to him or her positively, maybe they don’t need help. If they are very lonely and unhappy, or have signs of depression, that could lead to a diagnosis. Not [because of] the autism itself, but its consequences.”

Brian Harris
Susan Wilhite