GLP-1 Exposure in Pregnancy Linked to Contradictory Outcomes

Three studies attempting to tease out the potential effects of glucagon-like peptide 1 receptor agonists (GLP-1 RA) taken before or during pregnancy were presented at the annual meeting of the Society for Maternal-Fetal Medicine (SMFM) 2025 Annual Pregnancy Meeting.

Contradictory results from two of the cohort studies suggest the need for better understanding of GLP-1s during pregnancy before any consideration of changes to current recommendations, Olga Grechukhina, MD, an assistant professor of OB/GYN at Yale School of Medicine, New Haven, Connecticut, not involved in the studies, told Medscape Medical News.

“These discrepant results call for further translational and mechanistic studies that would explore the direct effects of this medication on placental function,” Grechukhina said. “In the absence of robust data to support the safety around the use of GLP-1 agonists in pregnancy and inconsistent but possible association with hypertensive disorders of pregnancy, this medication should continue to be halted during pregnancy.”

Preconception Use of GLP-1s

Rather than looking at the use of GLP-1s during pregnancy, the first of the studies looked at exposure prior to conception, though the study cohort included a proportion of patients who did not stop the medication until after conception.

The current evidence base has shown similar rates of miscarriage, live birth, preterm delivery, gestational diabetes, hypertension in pregnancy, maternal weight gain, and neonatal birth weight between those with and without preconception exposure to GLP-1s said Audrey Merriam, MD, MS, an associate professor of OB/GYN at Yale School of Medicine, when introducing her group’s research, led by Nishita Pondugula, MS, of Yale New Haven Hospital in New Haven, Connecticut.

So far, “small observational human studies showed no increased risk of congenital malformations or adverse pregnancy outcomes after antepartum GLP-1 RA exposure,” Merriam said.

Patients who stop taking semaglutide typically regain about two thirds of the weight they lost during treatment, Merriam noted, so she and her colleagues investigated whether differences in gestational weight gain occurred between those who did and did not take GLP-1s up to 1 year before conception.

In their retrospective cohort study, the researchers compared all pregnancies at their institution between 2014 and 2024 in which patients took GLP-1 drugs up to 1 year before conception, based on electronic medical records (EMRs). The 243 patients who met this inclusion criterion had an indication for GLP-1s due either to pregestational diabetes or to weight management. Each of these indication groups was compared with randomly selected control participants drawn from EMR data between 2021 and 2022.

The researchers compared the 103 patients taking a GLP-1 for diabetes to a cohort of 175 patients with pregestational diabetes who did not take a GLP-1.

The 140 patients taking a GLP-1 for obesity were compared with a cohort of 200 other patients not taking a GLP-1, including 100 with a body mass index (BMI) between 30 and 40 and 100 with a BMI ≥ 40. The matched cohorts were similar in age, proportion using public insurance, and proportion who were multiparous.

Most participants (92%) had taken only one GLP-1 drug, and most (63%) had taken semaglutide. Others took liraglutide, dulaglutide, tirzepatide, and exenatide.

Overall, the patients had taken their last dose of GLP-1 an average 14 days before and a median 30 days after conception.

The researchers then compared gestational weight gain between GLP-1-exposed and -unexposed patients, adjusting for baseline differences between the pregestational diabetes groups for race/ethnicity and chronic hypertension diagnosis. For those with obesity, the researchers adjusted for differences in parity, hypertensive disorders of pregnancy, and polycystic ovary syndrome diagnosis.

In patients with pregestational diabetes, there were no significant differences between those with and without preconception GLP-1 exposure in terms of how many gained more or less weight than recommended.

In those with obesity, however, patients with GLP-1 exposure were significantly less likely to be below gestational weight gain recommendations (adjusted odds ratio [aOR], 0.38; 95% CI, 0.18-0.80).

The researchers then stratified the groups to assess whether there was a difference between those with preconception and postconception GLP-1 exposure when it came to gaining less weight than recommended. Among those with obesity, only those who had taken GLP-1s after conception were significantly less likely to gain less weight than recommended (aOR, 0.29; 95% CI, 0.12-0.72).

The approximately two thirds reduction in odds of being below recommended weight gain for those with postconception GLP-1 exposure “may reflect nonpregnant findings showing rapid weight regain immediately after cessation,” Merriam said.

Reduced Risk for Hypertensive Disorders

A separate poster showed the same team’s analysis of hypertensive disorders of pregnancy risk in those who took GLP-1 medications during the year prior to conception. Pondugula and her colleagues reported that the use of GLP-1 medications prior to conception cut the odds in half of developing a hypertensive disorder of pregnancy.

For that analysis, the researchers adjusted for baseline differences between the pregestational diabetes groups in prepregnancy BMI, race/ethnicity, and chronic hypertension diagnosis. For those with obesity, the researchers adjusted for differences in parity, chronic hypertension, gestational diabetes, preterm birth, polycystic ovary syndrome diagnosis, and gestational weight gain.

Among those with obesity, 37.5% of unexposed participants developed a hypertensive disorder of pregnancy compared with 24% of those who had taken a GLP-1 (aOR, 0.5; P =.007).

There were no significant differences in either cohort between those with and without GLP-1 exposure for large for gestational age fetuses, fetal growth restriction, preterm birth, or fetal demise. There were no significant differences between the exposed and unexposed obesity groups for gestational diabetes, use of insulin, or use of metformin.

When the researchers compared those with only preconception exposure to GLP-1s and those who had been taking GLP-1s after conception, the aORs remained significant only for those exposed after conception.

Less Preterm Birth but Other Complications More Likely

A different group looked specifically at GLP-1 exposure during pregnancy and pregnancy complications. Semaglutide was linked to less risk for preterm birth but greater risk for several other pregnancy complications, Emily N. Adams, MD, of Johns Hopkins School of Medicine, Baltimore, and colleagues reported in a retrospective cohort study. Among 4,606,975 pregnant patients with type 2 diabetes, 0.2% of them (n = 9214) took semaglutide during pregnancy based on data from the Vizient Clinical Database.

Patients taking semaglutide had a significantly lower incidence of preterm birth than those receiving standard care (aOR, 0.74; P <.001) after adjustment for maternal age, gestational age at study entry, obesity, pregestational diabetes, and chronic hypertension. However, pregnancies exposed to semaglutide had a greater risk for preeclampsia (aOR, 1.16), HELLP syndrome (aOR, 1.88), and eclampsia (aOR, 2.74; all P <.001) compared with unexposed pregnancies.

“Our results underscore the need for careful monitoring of patients exposed to semaglutide in pregnancy,” the authors wrote.

Improved Glycemic Control

“I was not surprised to see that semaglutide exposure was associated with reduced risk of preterm birth,” Grechukhina told Medscape Medical News. “I suspect that might be related to improved glycemic control, which in turn implies that fewer patients would require iatrogenic preterm birth for poorly controlled blood sugar. Interestingly, the study by Pondugula, et al did not find this association, which may be related to the lower number of participants.” 

But Grechukhina was surprised by the other results of Adams and colleagues — the association between semaglutide exposure and increased risk for hypertensive disorders of pregnancy — because it contradicts the study by Pondugala and colleagues. She did note, though, that a limitation in the Adams and colleagues’ study is “it is unclear what counted as ‘exposure’ to semaglutide in relation to the onset of pregnancy.”

No external funding or disclosures were noted for any of the presented studies here, and Grechukhina had no disclosures.

Tara Haelle is a science/health journalist based in Dallas.