Study of 756,636 Pregnancies Ties GLP-1 Preterm Risk to Diabetes

Women who filled prescriptions for semaglutide (Ozempic, Wegovy) or liraglutide (Saxenda) around conception had higher odds of preterm birth only when the drugs were prescribed for diabetes, according to a Danish study of 756,636 pregnancies published in Human Reproduction Open. Among women using the same drugs for weight management, the researchers found no significant link to preterm birth or the other complications they studied.

The authors say diabetes itself, rather than the medication, may explain the extra risk. They also stress that the drugs have not been shown to be safe in pregnancy.

How the Danish team ran the study

Researchers at Copenhagen University Hospital Hvidovre used Denmark’s national health registries. They looked at 756,636 single-baby pregnancies among 480,231 women, all delivered between Oct. 1, 2009, and Dec. 31, 2023.

A pregnancy counted as possibly exposed if the woman filled a prescription for semaglutide or liraglutide in the eight weeks before or after the first day of her last menstrual period. That 16-week window covers the time before many women know they are pregnant and the early weeks when major organs start to form. Of all the pregnancies, 529 had prescription records showing exposure.

The team checked for preterm birth (before 37 weeks), pre-eclampsia, gestational diabetes, babies unusually large for their gestational age, stillbirth and placental problems. They adjusted for the mother’s age, body mass index, smoking, location, education and existing diabetes, plus the month and year of pregnancy.

What the numbers showed

At first, exposed women seemed to have several complications more often. Most of those differences stopped being statistically significant after the adjustments.

One remained. Compared with unexposed women, the adjusted odds of preterm delivery were 70% higher for women taking liraglutide for diabetes and 84% higher for those taking semaglutide for diabetes. The researchers’ announcement also gave estimated increases of about 11% for semaglutide and 9% for liraglutide among women with pre-existing diabetes. No such rise appeared when the drugs were prescribed for weight management.

“This suggests that the underlying condition of diabetes, rather than the medication, may be driving this association,” said first author Kathrine Hviid, a PhD student in the department.

Limits of the research

The study was observational. It can show associations but cannot prove that the drugs caused early births. The team also relied on prescription records, so it could confirm that a woman filled a prescription but not that she took the drug. The researchers believe most did, because the injections are expensive in Denmark: a 1mg injection of Wegovy costs about €180, and Ozempic about €114 with the diabetes subsidy.

Study leader Henriette Svarre Nielsen said the findings do not justify dropping current guidance to stop GLP-1 drugs eight weeks before a planned pregnancy, and they are not enough to set new counseling rules based on why the drugs were prescribed. She noted that this guidance rests on early studies in model organisms, not real-world evidence.

Why it matters

GLP-1 receptor agonists, first developed for type 2 diabetes, are now the medicines with the sharpest rise in prescriptions worldwide, Nielsen said. In Denmark, about 70% of people using weight loss drugs are women, so some will take them before realizing they are pregnant.

In an invited commentary, Yeyi Zhu and Monique Hedderson of Kaiser Permanente Northern California wrote that the results could support more balanced, individualized talks between doctors and patients exposed around conception. For women with diabetes, they said, the study is a reminder that diabetes is a risk factor for obstetric complications and that blood sugar control before and during pregnancy matters.

The authors say future studies, including randomized trials, need to separate women taking the drugs for diabetes from those taking them for weight loss.

The study is published in Human Reproduction Open (DOI 10.1093/hropen/hoag015), with the commentary at DOI 10.1093/hropen/hoag016.

Source: ScienceDaily