Of the many proposed triggers for autism, one of the most controversial is the “extreme male brain” hypothesis. The idea posits that exposure to excess testosterone in the womb wires both men and women to have a hypermasculine view of the world, prioritizing stereotypically male behaviors like building machines over stereotypically female behaviors like empathizing with a friend. Now, a study is raising new doubts about this theory, finding no effect of testosterone on empathy in adult men.
The work does not directly address whether high levels of prenatal testosterone cause autism or lack of empathy. That would require directly sampling the hormone in utero, which can endanger a developing fetus. But the new study’s large size—more than 600 men—makes it more convincing than similar research in the past, which included no more than a few dozen participants, experts say.
The extreme male brain hypothesis was first proposed by psychologist Simon Baron-Cohen at the University of Cambridge in the United Kingdom. In 2011, he and colleagues found that women given a single hefty dose of testosterone fared significantly worse at the Reading the Mind in the Eyes test (RMET), which asked them to gauge the emotional states of others based on their facial expressions. The women’s performance seemed to track with a controversial metric called the 2D:4D ratio, the relative lengths of the second and fourth fingers. Men—and people with autism—tend to have a longer ring finger than index finger, and some researchers believe that is due to higher prenatal exposure to testosterone. (Others are skeptical.)
Taken together with the observation that men are at least 10 times more likely to develop autism, Baron-Cohen argued that the condition results from a “masculinization” of the brain in utero.
But as the theory caught on in psychology circles, few seemed to notice that the small follow-up studies that other researchers attempted failed to replicate the original findings, says Amos Nadler, a neuro-economist and visiting scholar at the University of Toronto in Canada. “People referred to that original study as if it was a solid stepping stone,” he says. “Pardon my French, but nobody was calling bullshit.”
Nadler and colleagues decided to run a much larger study. They also focused on men rather than women, partly because men are disproportionately affected by autism. After measuring the baseline testosterone levels of 243 volunteers, the team gave each man the first half of the RMET. The scientists then administered a single dose of testosterone gel applied to the shoulders, or a placebo. Then the participants took the second half of the RMET and once again, the team measured their performance and testosterone levels. In a separate experiment, the team followed a similar protocol with 400 additional men at a different location but administered the testosterone as a nasal gel.
There was no difference in performance on the empathy test between the placebo and testosterone groups in either study, the team reports today in the Proceedings of the Royal Society B. The researchers also measured participants’ 2D:4D ratios, and found no detectable pattern that could connect the purported testosterone exposure in the womb to empathy levels. “Even when we tried to be very forgiving about it, there was nothing there. Like, no effect,” Nadler says.
To Melissa Hines, a neuroscientist also at Cambridge, the team’s inability to replicate earlier findings is “unsurprising.” The small size of past studies and the unreliability of the 2D:4D measure make it unlikely that the results would replicate in women either, she says.
Still, Baron-Cohen finds it “strange” that Nadler’s team conducted the study exclusively in men “because to expect to see an effect of testosterone, one needs to study people who at baseline have relatively low” testosterone, he says—in other words, women.
But Nadler says Baron-Cohen’s objection is “incongruous” with the original paper’s hypothesis, which predicted that people with higher prenatal testosterone levels would be more sensitive to the empathy-blunting effects of testosterone. That means men should show “more pronounced effects [than women], not less,” he says.
Nadler cautions, however, that the study can’t disprove any theories about testosterone’s impact on the developing brain. “What we’re saying is that giving people testosterone as adults has no influence on their ability to understand people’s emotions.”
He hopes the study will dispel any misconceptions that by controlling testosterone—for example, using medications to block it in pregnant women—it might be possible to prevent autism. “If there’s no relationship, then we shouldn’t give people false hope.”
*Correction, 4 September, 11:35 a.m.: This story has been updated to reflect Amos Nadler's status, the date of the original study, and the experimental setup of the current one.