Blood test for mom could predict baby’s due date

May 5, 2021

A simple blood test might help figure out baby's arrival date. (Pexels/Amina Filkins)

Biomarkers in expectant women’s blood change throughout pregnancy, according to a new study that lays the groundwork for a test that could pinpoint the date of a baby’s arrival.

In a study published Wednesday in Science Translational Medicine, researchers analyzed blood samples taken from 63 women during their third trimester to look for clues that indicated how far along they were, a prediction that could help doctors better care for other pregnant patients.

“We wanted to see if, in the blood, we can find a signature that is predictive of the countdown to labor. This is important because, right now, clinicians don’t really have a tool to predict when labor will actually happen,” said study lead author Ina Stelzer, a postdoctoral researcher in anesthesiology, perioperative and pain medicine at Stanford University. “Especially for shorter or longer pregnancies — so, preterm or post-term pregnancies — it could help clinicians in their clinical management of pregnancies and counseling of pregnant women to inform them where they are in pregnancy and what they can expect.”

At the moment, a due date is based on the first day of a woman’s last menstrual period, as well as on an ultrasound, according to Stelzer. But because these calculations use a fixed gestation period of 40 weeks, which doesn’t take into account variation in individual pregnancies, they’re often inaccurate.

With the goal of developing a more precise way to forecast when a woman will give birth, Stelzer and her colleagues analyzed blood samples collected throughout the third trimester to look for compounds, also known as biomarkers, associated with time to birth. The team only included women who went into labor spontaneously, excluding others who were medically induced or had cesarean sections without first going into labor.

The team measured levels of all the proteins and chemical compounds, including hormones, in each of the blood samples. The researchers also looked at measures of immunity, including the abundance of different types of antibodies and immune cells present in the blood and the responses of immune cells to a challenge that mimics a bacterial infection.

All in all, the analysis characterized more than 7,000 different blood biomarkers. There was a distinct change in the composition of biomarkers around two to four weeks before labor, indicating a shift from maintaining pregnancy to a period that the team called prelabor, where the mother’s body prepares for birth.

To build a statistical model that predicted time-to-labor from this suite of markers, the researchers used data from 53 of the women to train a machine-learning algorithm. Then they tested the model’s accuracy with data from the remaining 10 women.

The model accurately predicted due date, give or take about two weeks, in both term and preterm pregnancies.

“The study wasn’t designed to make a distinction between healthy or preterm pregnancies, but we were really happy to see that the prediction also worked for preterm births,” Stelzer told The Academic Times. “We had five preterm births in the cohort, so this is just a first glimpse.”

The researchers found that about 45 biomarkers were most important for predicting when a woman went into labor. Some, such as the hormone cortisol and a protein expressed by the fetal membrane, steadily increased throughout the third trimester. In contrast, levels of some biomarkers, including one associated with inflammation, surged about 30 days before the due date.

In future studies, the researchers plan to examine more women to study both preterm and post-term pregnancies. By examining a larger group of mothers-to-be, the team also aims to predict the date of labor more precisely.

Down the road, the researchers hope to develop a test that clinicians could use in the routine care of pregnant women. According to Stelzer, such a tool could identify women at risk of a preterm birth, which can be managed with the use of treatments that prolong the pregnancy, for example. Likewise, a test could help doctors care for patients experiencing post-term pregnancies.

“Usually those pregnancies are induced,” Stelzer said. “But for gynecologists or obstetricians, it would be helpful to know if it’s actually necessary to induce, because that comes with its own complications. If they could see that the mom is already progressing towards labor anyway, there would be no need to induce.”

One limitation of the research was that it included patients from just one hospital, Stelzer said. Study participants were predominantly white or Asian, and very few were Black or Hispanic. In future studies, the researchers will include multiple hospitals and include a more diverse group of women.

According to Stelzer, this research is the first step toward more comprehensive monitoring of women's health during pregnancy. At the moment, doctors typically monitor only things such as blood glucose, blood pressure and weight in expectant mothers, she said.

“We do a lot for the fetus: there’s 3D ultrasound and prenatal screening tests for genetic diseases and all that,” she explained. “But there is huge potential to use maternal biology throughout pregnancy to monitor health.”

The study, “Integrated trajectories of the maternal metabolome, proteome, and immunome predict labor onset,” published May 5 in Science Translational Medicine, was authored by Ina A. Stelzer, Kazuo Ando, Julien J. Hédou, Dorien Feyaerts, Laura S. Peterson, Kristen K. Rumer, Eileen S. Tsai, Edward A. Ganio, Dyani K. Gaudillière, Amy S. Tsai, Benjamin Choisy, Lea P. Gaigne, Franck Verdonk, Danielle Jacobsen, Gavin M. Traber, Mathew Ellenberger, Natalie Stanley, Martin Becker, Anthony Culos, Ramin Fallahzadeh, Ronald J. Wong, Gary L. Darmstadt, Maurice L. Druzin, Virginia D. Winn, Ronald S. Gibbs, Xuefeng B. Ling, Karl Sylvester, Brendan Carvalho, Michael P. Snyder, Gary M. Shaw, David K. Stevenson, Kévin Contrepois, Martin S. Angst, Nima Aghaeepour and Brice Gaudillière, Stanford University School of Medicine; Mohammad S. Ghaemi, Stanford University School of Medicine and National Research Council Canada; Xiaoyuan Han, Stanford University School of Medicine and University of the Pacific; and Sonia Gavasso, Stanford University School of Medicine and Haukeland University Hospital.

We use cookies to improve your experience on our site and to show you relevant advertising.