# Periconceptional GLP-1 receptor agonist exposure and obstetric outcomes: a Danish nationwide cohort study

## TL;DR

A Danish study found that using GLP-1 receptor agonists for diabetes around the time of conception is linked to a higher risk of preterm birth, but not when used for weight management.

## Contribution

The study provides new nationwide evidence linking GLP-1 receptor agonist use for diabetes to increased preterm birth risk, but not for weight management.

## Key findings

- Periconceptional GLP-1 receptor agonist use for diabetes was associated with elevated preterm birth risk (liraglutide aOR 1.70, semaglutide aOR 1.84).
- No increased preterm birth risk was observed for GLP-1 receptor agonist use in women without pre-existing diabetes.
- The findings suggest that diabetes-related factors may contribute more to preterm birth risk than the medication itself.

## Abstract

What is the association between periconceptional GLP-1 receptor agonist exposure and risk of obstetric complications?

Periconceptional GLP-1 receptor agonist exposure was associated with increased preterm birth risk when used for diabetes treatment (liraglutide aOR 1.70, 95% CI 1.17–2.48; semaglutide aOR 1.84, 95% CI 1.24–2.7) but not for weight management, suggesting the underlying diabetes rather than the medication may be the causal factor.

GLP-1 receptor agonists are rapidly expanding in use among reproductive-age women for diabetes and obesity treatment. While not approved for use in pregnancy, inadvertent periconceptional exposure occurs frequently. Limited safety data exist, with recent small studies suggesting no increased risk of major congenital malformations, but comprehensive obstetric outcome data remain lacking.

This nationwide observational cohort study used data from Danish health registries from October 2009 through December 2023. We analyzed 756 636 singleton pregnancies among 480 231 women, with 529 pregnancies having periconceptional GLP-1 receptor agonist exposure.

We identified periconceptional liraglutide or semaglutide exposure (prescription redemption within 8 weeks before/after last menstrual period) using the National Prescription Register. Exposed pregnancies were stratified by maternal pre-existing diabetes status and compared with propensity score-matched unexposed controls. Propensity score matching incorporated maternal age, BMI, smoking, geographic region, education, pre-existing diabetes, parity, and temporal factors.

Before adjustment, exposed women had higher rates of multiple obstetric complications. After propensity score matching, only the risk of preterm birth remained elevated for exposed women. This increased risk was confined to women using GLP-1 receptor agonists for diabetes treatment (liraglutide aOR 1.70, 95% CI 1.17–2.48; semaglutide aOR 1.84, 95% CI 1.24–2.71). Among women without pre-existing diabetes using these medications for weight management, no association with preterm birth was observed (liraglutide aOR 1.01, 95% CI 0.58–1.76; semaglutide aOR 0.71, 95% CI 0.30–1.70).

Study limitations include the absence of data regarding medication compliance post prescription redemption, potential misclassification due to parallel importation, and the inability to control for unmeasured confounding factors. The observational design cannot establish causality. Most semaglutide weight-loss prescriptions occurred late in the study period, limiting long-term follow-up data.

The results are compatible with the hypothesis that diabetes-related factors, rather than GLP-1 receptor agonist exposure itself, may contribute to the increased preterm birth risk. If so, there are important implications for preconception counselling and may inform future guidelines for GLP-1 receptor agonist use in reproductive-age women.

KVRH, KB, DW, and HSN acknowledge funding from the Novo Nordisk Foundation (NNF21OC0069257 and NNF220C0077221) and the AP Moller Foundation. LMH was supported in part by grants from NordForsk (id: 105545), the Novo Nordisk Foundation (NNF17OC0027594 and NNF17OC0027812), and the Villum Foundation (‘Nation-Scale Social Networks’). KSL was supported by the Independent Research Fund Denmark (8045-00047B), NordForsk (id: 156298), and Centre for Childhood Health (id: 72 2024_F_008 and 2024_I_001). SM: Advisory boards: AstraZeneca, Boehringer Ingelheim, Intarcia Therapeutics, Novo Nordisk, Sanofi, Abbott Lab, Bayer, Amgen; Lecture fees: AstraZeneca, Novo Nordisk, MSD; Research grant recipient: Novo Nordisk; Novo Nordisk foundation, Boehringer Ingelheim; Support for attending meetings and/or travel: Novo Nordisk, Boehringer-Ingelheim, Bayer. Grants were paid to the institution, Hvidovre Hospital, University of Copenhagen, with no personal fee. None of the grants has any relation to the work presented in the paper. SM is also a consultant for Netdoktor and has served as principal investigator in relation to the development of drugs for the treatment of type 2 diabetes and obesity in collaboration with Novo Nordisk and Bayer, with funds paid to the institution where he is employed, with no personal fee and with no relation to the work reported in this article. HSN: lecture fees on own research: Novo Nordisk A/S, Ferring Pharmaceuticals, Merck A/S, Astra Zeneca, Cook Medical, Gedeon Richter and Ibsa Nordic. K.B holds stock/share or stock/share options with Novonesis, Genmab, and Novo Nordisk (held in personal investment portfolio). The remaining authors have no other disclosures.

N/A