What causes autism? The lack of answers or even particularly good leads is frustrating to say the least. Not surprisingly, people both within the scientific community and the general public are hungry for answers, and my general opinion is that the more information the better. However, I’m a little puzzled by a report from the British Association for the Advancement of Science’s annual Festival of Science.

According to Simon Baron-Cohen‘s ’extreme male brain’ theory, people with autism show extreme versions of behaviors that are normal in men. In general, men tend to empathize less and systematize more than women. These drives are taken to an extreme degree in people with autism, resulting in the behaviors associated with autism, including reduced eye contact and verbal skills and increased repetitive behavior and orderliness, according to Baron-Cohen.

Is the hormone that causes male behaviors involved in autism? At the Festival of Science, Baron-Cohen and Bonnie Auyeng reported that fetal testosterone contributed to ‘autistic traits’ in normal eight-year-old children. The researchers calculated an ‘autism spectrum quotient’ from questionnaires about children’s social behaviors and cognitive skills completed by their mothers. Fetal testosterone levels recorded eight years earlier accounted for more than 20% of the variability in this quotient.

According to the researchers, these data suggest that elevated testosterone levels in the womb may contribute to traits associated with autism. However, based on the researchers’ reasoning, wouldn’t an alternative explanation be that fetal testosterone correlates with male-typical behaviors? Animal studies have shown that testosterone produced in the fetal testes masculinizes the brain (allowing male-typical behavioral patterns). Perhaps the present study indicates that male-typical behaviors are graded, with high levels of fetal testosterone producing ‘super males’.

While interesting, it’s not clear to me that these data are directly relatable to autism. For that, we’ll need to see Baron-Cohen’s next study, involving clinical data and amniotic samples from 90,000 people with and without autism.