Researchers examined the records of 320,000 children to determine if there’s a link between pregnant women with diabetes and a risk of autism in their children.

Expectant mothers concerned about the risk of autism in their unborn child have something new to worry about.

According to a new study of more than 320,000 children, intrauterine exposure to gestational diabetes may be associated with an increased risk of autism spectrum disorders.

The study was recently published in JAMA. According to Anny H. Xiang, Ph.D., of Kaiser Permanente Southern California in Pasadena, she and her colleagues analyzed data from a single healthcare system in order to study the link between maternal diabetes and the risk of children developing autism. The study group included woman who were known to be diabetic prior to pregnancy and those who were diagnosed during pregnancy.

Xiang was careful to note that the discovery of a link does not mean that gestational diabetes causes autism in children. What’s more, since there’s no magic pill to prevent autism, she suggests, “Women should see their doctor to make sure that blood sugar is normal when planning for pregnancy and throughout pregnancy.”

Xiang, an adjunct research associate professor at the University of Southern California’s Keck School of Medicine, said the study included 322,323 children born between 1995 and 2009 at Kaiser Permanente Southern California hospitals. The researchers adjusted for other factors, such as maternal age, household income, race/ethnicity, and the child’s gender.

Get the Facts: What Is Autism? »

The researchers found no link to a risk of autism when the mothers had pre-existing type 2 diabetes. The increased risk was independent of smoking, pre-pregnancy body mass index, and gestational weight gain.

The specifics of the study revealed that 6,496 children (2 percent) were exposed to pre-existing type 2 diabetes; 25,035 (8 percent) were exposed to gestational diabetes; and 290,792 (90 percent) were unexposed.

Following birth, and with a median age of 5.5 years, 3,388 children were diagnosed as being on the autism spectrum, with 115 exposed to pre-existing type 2 diabetes, 130 exposed to gestational diabetes at 26 weeks or less, and 180 exposed at more than 26 weeks. That left 2,963 unexposed.

The reason the numbers are reported as before or after 26 weeks’ gestation has to do with the design of the study, in which the researchers formed three groups of women to look at the relationship between gestation and the risk of autism in offspring.

In Xiang’s words: “We used the tertile distribution (appropriately 33 percent in each group) of gestational weeks to define the cut points. It turns out that the tertile cut points were 26 weeks and 30 weeks.”

The team found that results for the second group (26 to 30 weeks) and the third group (more than 30 weeks) were similar. So in reporting the results, “the latter two groups were combined.”

The Basics: What Is Gestational Diabetes? »

Scientists have long known that fetal exposure to maternal hyperglycemia may have long-lasting effects on organ development and function. Previous studies have revealed long-term risks of obesity and related metabolic disorders in offspring of women who had diabetes prior to pregnancy, as well as those whose hyperglycemia was first detected during pregnancy.

Whether such exposure can disrupt fetal brain development and heighten risk of neurobehavioral developmental disorders in offspring is less clear.

Dr. Xiang says more research is called for.

“We are looking for collaborations with scientists in the field of autism research to understand potential biological mechanisms,” she said. “There may be multiple pathways, such as hyperglycemia may cause hypoxia in the fetus, oxidative stress in cord blood and placental tissue, chronic inflammation, and epigenetics (external modifications to DNA that turn genes ‘on’ or ‘off’).”

This study was supported by Kaiser Permanente Southern California Direct Community Benefit Funds.

Break It Down: Autism Rates by State »