SALT LAKE CITY — University of Utah researchers have discovered a link between increased levels of a type of estrogen in babies in their second trimester and risk for autism, according to a new study.

The findings could eventually help doctors identify babies at risk of autism early in their mothers’ pregnancies and monitor them more closely, as well as provide early interventions to ensure the children’s well-being, said Dr. Deborah A. Bilder, the study’s first author.

Bilder serves as a U. associate professor of psychiatry and principal investigator of the Utah Registry of Autism and Developmental Disabilities.

The research, published recently in the Journal of Autism and Developmental Disorders, was a collaboration between the University of Utah and Intermountain Healthcare.

Obstetricians from both hospitals recruited more than 10,000 pregnant women throughout Utah from 1999 to 2002 and asked those who were willing to consent for researchers to use serum samples from their blood for research.

In the new study, researchers looked at samples from 72 of the women to determine what was in their serum during the beginning of their second trimesters “to see what in the serum could be associated with autism among the babies that are born,” Bilder explained.

Specifically, they were interested in steroids.

“Because we know that conditions which are associated with higher risk of autism are often related to inflammation and abnormal steroids. There’s a lot of work that’s currently being done looking at inflammation, as it should be, in regard to the role it might play in the baby’s development of autism,” Bilder said.

“But there is no work being done is looking at the steroids that accompany inflammation or accompany other sources of stress beyond inflammation that reflect how the mother, the placenta and the baby are responding to stress.”

Researchers wanted to learn whether there were differences in steroids between samples drawn from moms whose babies developed autism compared to moms whose babies didn’t. They looked at birth certificate data to identify autism cases among children of mothers who provided the samples.

Both the control group and the group of mothers whose children had autism were selected so that 50% of each group had an exposure to a condition such as gestational diabetes, hypertension and preeclampsia. Previous studies have shown links between those conditions and autism risk.

The researchers looked at several different steroids in the blood samples. Bilder said she expected to find increased levels of steroids that were known to be associated with the conditions, like testosterone. She found those steroids, but they did not reach a statistical significance, according to Bilder.

Bilder also expected progesterone and testosterone in the kids who developed autism to be elevated.

“But that’s not what I found. Instead, what I found, is estradiol being elevated in the children who developed autism,” she said.

Estradiol is a type of estrogen. Lower levels of estradiol in a baby usually indicate a concern with the baby, and high estrogen levels are not currently associated with abnormal conditions.

But as the researchers looked at the steroid hormones that were measured, “what we realized is that the higher estrogen levels being produced by the placenta actually may be stimulating the baby’s development of his or her stress response.”

Usually, a baby’s stress response takes time so that when the pregnancy reaches full term, the baby has developed its own stress response. But elevated levels of estrogen cause the baby’s stress response to develop early, Bilder said.

That prepares babies, when there is an issue, to survive outside the mom. It causes early growth of the lungs, gut and skin so that if the baby doesn’t make it all the way through pregnancy, it’s more likely to survive, she said.

The findings indicated that in the babies with autism, something set off their stress response early.

Studies have shown that children with autism have an abnormal stress response, according to Bilder. She believes the mechanism that triggers the early stress response during pregnancy may still be affecting children with autism past delivery.

Bilder doesn’t think doctors should target the higher estradiol levels or try to lower them. Instead, because it signals a “protective mechanism, that baby is surviving,” doctors should target something that doesn’t jeopardize the baby’s survival.

“By being able to have a way of looking at the baby’s well-being in that regard, I think that opens up the door to considering how can you reduce the stress on that baby?” Bilder explained.

She also doesn’t want the findings to be used to select pregnancies, even if it became possible to discover autism earlier. The researchers didn’t identify a threshold for normal or abnormal levels of estradiol. Instead, Bilder believes earlier identification could help doctors monitor the babies more closely, and provide early intervention if needed.

The findings also highlight a unique partnership between two separate health care systems.

“What really excites me so much about this, first of all, is I think it demonstrated what’s possible here in Utah, really, because of the willingness of two otherwise rival health care systems to work together for a common goal, which is to reduce the risk of autism and improve the well-being of individuals in their families with autism. And there’s a clear commitment by Intermountain and the University for that,” Bilder said.

The research was also supported by the Utah Department of Health, the Utah State Board of Education, the Utah Registry of Autism and Developmental Disabilities, and the Pedigree and Population Resource at Huntsman Cancer Institute.