A researcher who describes autism as a condition of the "extreme male brain" says fetuses exposed in the womb to high levels of the male hormone testosterone are more likely than others to develop autistic traits as children.
Simon Baron-Cohen, director of Cambridge University's Autism Research Centre, has shown in past research that men are more likely than women to score low on tests of empathy but high on tests of "systemizing" recognizing rules and patterns characteristics that, in the extreme, define autism. That's what led Baron-Cohen to regard the disorder which is about three to four times as prevalent in boys as in girls as one of the extreme male brain and to search for a link to male hormones. (See "The Year in Medicine: From A to Z.")
In his new paper, published on Monday in the British Journal of Psychology, Baron-Cohen studied 235 pairs of mothers and children over eight years, periodically giving the children questionnaires designed to measure autistic traits. None of the children in the study received an autism diagnosis, but Baron-Cohen found that those who had been exposed to higher testosterone levels in the womb measured via amniocentesis during pregnancy had a greater chance of displaying autism-associated traits such as poor social skills, imagination and empathy and high aptitude in certain memory-retention exercises.
"The study highlights for the first time the association between fetal testosterone and autistic traits and indicates that fetal testosterone not only masculinizes the body; it masculinizes the mind," he says.
Not all experts agree. Laurent Mottron, a professor of psychiatry at the University of Montreal who wrote an accompanying review of Baron-Cohen's study, told TIME that Baron-Cohen's study, including the questionnaires used to measure autistic traits, presented major "logical and factual flaws." Because the children in the study were normally developing rather than autistic, the study showed only that exposure to testosterone was associated with typically male cognition, not a disorder.
In an e-mail, Mottron wrote, "Baron-Cohen used a questionnaire which scores high in autistics. This questionnaire also scores higher in [nonautistic] men than in women. This only demonstrates that the autism questionnaire is a very weak and broad instrument, which is unable to differentiate autism and male characteristics. It does not demonstrate that autism is linked to testosterone."
Said Mottron: "The parallel between a male cognitive profile and an autistic cognitive profile is weak and is true only for a minority of tasks. The logical fuzziness and possible logical flaws of the Extreme Male Brain model result in the [scientific] community not following Baron-Cohen's work."
Such vehemently opposing views are typical in the fractured landscape of autism research. Like the thoughts of autism patients, the causes of the disorder remain inscrutable to researchers. Most scientists agree there is a genetic basis, but there is little question that environmental factors play a part too; there have been documented cases of autistic patients with a nonautistic identical twin.
Researchers who are more sympathetic to Baron-Cohen's work, like James B. Adams, a professor at Arizona State University's School of Materials, do not discount the theory that testosterone exposure is linked to autism but believe the association may be mediated by other potential causes. For his part, Adams believes autism is related to exposure to mercury a controversial charge that most research has failed to support and, Adams says, elevated testosterone levels are linked to the depletion of glutathione, a substance in the body that protects it from toxic metals. "So Baron-Cohen's work ties in with the mercury hypothesis," Adams says.
Baron-Cohen says his theory could also complement the genetic hypothesis, as "it is the baby's genes that determine how much testosterone the baby makes"; the mother's testosterone is believed not to pass through the placenta into the amniotic fluid.
In an interview with Britain's Guardian newspaper, Baron-Cohen made front-page headlines in his home country by expressing confidence that his research would soon allow doctors to screen for autism through amniocentesis which involves extraction of amniotic fluid with a needle the same procedure that allows parents to test for Down syndrome, and decide whether to terminate a pregnancy. Although a prenatal measure of testosterone is not a definitive test for autism, Baron-Cohen suggested that a debate was needed over whether such a test would be desirable.
The debate comes on the heels of the announcement of Britain's first fetus screened for the breast-cancer gene, BRCA-1, which sparked a wider debate over "designer babies" a discussion that even Baron-Cohen's supporters say is premature when it comes to autism. "We are a long way from unlocking the mysteries of autism in a way that could give us those abilities," says Adams. "Saying someone has autism is like saying someone has a fever; they have a set of symptoms, but we don't know yet what could be the cause or causes." If such a test could be made available, should it be allowed? "I have a daughter with autism," Adams says. "So that's a very difficult question."
Read TIME's "Heroes of the Environment, 2008."