A compelling new study has found no link between testosterone in men and reduced cognitive empathy, a trait that is characteristically impaired in autism spectrum disorders.

Ever since the very first clinical account of autism, almost eighty years ago, far more male children and adults have been placed on the spectrum than females. In all that time, no one has been able to figure out why.

Today, one of the most popular and commonly cited hypotheses is also one of the most controversial. First proposed by psychologist Simon Baron-Cohen, the 'extreme male brain' (EMB) hypothesis suggests high levels of prenatal hormones in the womb can lead to hyper-masculine traits later on, such as a reduced ability to read emotions in others.

Evidence for this theory is highly disputed on many different levels. Because it's too dangerous to manipulate prenatal testosterone in human subjects, most previous studies on the role of testosterone have relied on correlations only.

Researchers have, for example, tested testosterone in adults and used digit ratios as a proxy for prenatal testosterone levels.

What's more, much of this research is small and inconsistently replicated. A study co-authored by Baron-Cohen in 2011, for instance, was only conducted among 16 young women. It found that when testosterone was administered to these participants, it reduced their performance on a test of reading others' emotions, called Reading the Mind in the Eyes Test (RMET).

The study also found a link between the ratio of the second and fourth finger, known as the 2D:4D ratio, and a woman's sensitivity to testosterone - an idea that has turned up mixed results and is highly controversial as proxy for the prenatal environment.

"People referred to that original study as if it was a solid stepping stone," Amos Nadler, a neuro-economist at the University of Toronto, told Science magazine.

"Pardon my French, but nobody was calling bullshit."

Delving deeper, Nadler and his colleagues conducted two randomised controlled studies with samples that were 15 and 25 times greater than the original 2011 study. They also decided to focus only on males, as males are roughly four times more likely to be diagnosed with autism.

Conducting one of the largest studies to date, the researchers administered testosterone or a placebo to 643 male participants. To avoid giving people the same test twice - a pitfall of previous research - the researchers split participants into two groups.

The first group, composed of 243 volunteers, had their testosterone levels measured for a baseline and then completed the first half of a cognitive empathy test, the same one that was used in the 2011 study. This group was then given a single dose of testosterone gel or a placebo on the shoulder and finished the rest of the test.

The second group, composed of 400 people, undertook the same routine but this time with a nasal gel. The fingers of each participant in each group were also measured.

"Our research shows that there is not a causal relationship between testosterone levels and the capacity to understand others' thoughts and feelings, which is a hallmark impairment of autism spectrum disorders," Nadler told ScienceAlert.

The behavioural economist says the new results cast serious doubt on previous studies, suggesting that past conclusions were "underpowered and therefore unreliable". As such, he says they make poor evidence for the extreme male brain theory, and he suggests that the role of testosterone in cognitive empathy is much more complex than we often think.

"We found that there is no evidence to support this effect of testosterone, but that doesn't rule out any possible effects," says Gideon Nave, a marketing researcher at the University of Pennsylvania.

"From what we know, though, it seems that if testosterone does have an influence, the effect is complex, not linear. Reality is typically not that simple."

The new findings are supported by other recent studies that have found no relationship between prenatal testosterone exposure and autistic traits.

Peter Bos, who studies how hormones influence our brains at Leiden University, finds the new study convincing, but he says we should be very careful with the implications we draw from it.

After all, the researchers did not directly examine autistic traits, so the findings are limited in what they can actually tell us about the role of testosterone.

What's more, tests like the RMET are often critiqued for requiring a broad range of skills beyond just emotion recognition, and they aren't usually clinically accepted.

"This work is definitely impressive in that it seems carefully executed and the sample is very large, far beyond what we are used to in this field," Bos told ScienceAlert.

"It convincingly shows that in males a single administration of testosterone gel has no effect on performance on the RMET."

But while he admires the authors' efforts, Bos said he was surprised that women were not included in such a large study. Testosterone is thought to be an important factor in shaping sex-differences, so he says we cannot assume that steroid effects are similar in both males and females.

For example, males, and especially males high in autistic traits, might show a poor performance on cognitive empathy tests that will not get worse after a dose of testosterone. They've already hit the floor. Whereas, females tend to have a stronger response to testosterone because they have a higher ceiling.

"I don't think this study challenges the evidence behind the EMB theory, I think it is too early for such a claim," Bos told ScienceAlert, "but it shows that story is more complex than we previously assumed.

He says that maybe if the results were replicated among females, he would be willing to agree with that claim.

The findings were published in the Proceedings of the Royal Society B.