A small new study has once again found a link between elevated hormone levels in the womb and the likelihood of developing autism. But this time, instead of major androgens such as testosterone, researchers are pointing the finger at estrogen.

Currently, prenatal hormones such as androgens and estrogens are thought to influence the development of a fetus's brain. When these levels are out of whack, there's a possibility they could lead to autism later on in life.

This idea was first put forward roughly two decades ago; in that time, several studies have shown that prenatal hormones – including testosterone, cortisol, and progesterone – are higher in male fetuses who later develop autism.

These findings are often used to explain why so many men are diagnosed with autism compared to women. But while the majority of research has so far focused on androgens, the role of estrogen in autism has hardly been studied at all.

Researchers from the University of Cambridge and the State Serum Institute in Denmark have now provided the first evidence that prenatal estrogen levels may be more predictive of autism than even androgens.

"This new finding supports the idea that increased prenatal sex steroid hormones are one of the potential causes for the condition," says lead author Simon Baron-Cohen, director of the Autism Research Centre at Cambridge University.

"Genetics is well established as another, and these hormones likely interact with genetic factors to affect the developing foetal brain."

Testing 275 amniotic fluid samples included in the Danish Biobank, researchers measured the levels of four types of prenatal estrogens, including estradiol, estrone, estriol and progesterone.

Ninety-eight of these samples belonged to male individuals who later went on to develop autism. When their amniotic samples were compared to 177 fetuses who did not, there was a considerable difference in their hormone levels.

On average, all four estrogens were significantly elevated in those who would later develop autism, with estradiol levels having the strongest positive effect.

In fact, compared to their previous study from 2015, which focused on prenatal androgen levels in the Danish Biobank, the authors found an even stronger link between prenatal estrogens and the odds of developing autism.

The sample size used is admittedly small, and they appear to contradict some other studies that have shown estrogen can actually reduce the odds of autism in female humans, while limiting autistic behaviour in zebra fish and mice. So we need to take this study with a pinch of salt.

The authors explain, however, that their recent findings correspond to a slightly earlier time point in pregnancy, and that this could ultimately make all the difference in the way these hormones interact with the developing brain.

"These elevated hormones could be coming from the mother, the baby or the placenta," says co-author Alex Tsompanidis, who studies the role of hormones in autism at Cambridge.

"Our next step should be to study all these possible sources and how they interact during pregnancy."

The authors caution that their goal for further research is not to prevent autism or somehow screen it using these prenatal hormones - it's not a disease, after all. It's only to understand the condition better.

Over the years, it's become increasingly clear that estrogens and androgens cannot simply be split into 'masculinising' and 'feminising' hormones, and although this new study is small, it adds further weight to this idea.

As a result, the authors are now curious to see whether the same patterns are observed in autistic females.

"This is a terrific example of how a unique biobank set up 40 years ago is still reaping scientific fruit today in unimagined ways, through international collaboration," says co-author Arieh Cohen, a biochemist based at the State Serum Institute in Copenhagen.

The new research has been published in Molecular Psychiatry.