If You Conceived While Taking a Weight-Loss Injection: What a Large Danish Study Found

A nationwide study of more than 750,000 pregnancies looked at whether GLP-1 medicines like semaglutide and liraglutide lead to babies being born early

A large study from Denmark followed more than 750,000 pregnancies and asked whether women who were taking a GLP-1 weight-loss or diabetes injection around the time they conceived were more likely to give birth early. The answer turned out to depend on why the medicine had been prescribed. Women who were taking it for diabetes did have a higher chance of an early birth, but women taking it purely for weight management did not — which suggests that the diabetes, rather than the medicine itself, is what raises the risk.

Why this question came up in the first place

Medicines known as GLP-1 receptor agonists — the best-known are semaglutide and liraglutide, sold under several brand names — were originally developed to treat type 2 diabetes. They work by mimicking a natural gut hormone that helps the body handle blood sugar and that also reduces appetite. Over the past few years they have become extremely widely used for weight loss as well.

That success created an unexpected situation. Many of the people taking these medicines are women in their twenties, thirties and forties. Losing a significant amount of weight can also make it easier to become pregnant, because obesity interferes with ovulation. So a growing number of women are conceiving while still on the injection, or shortly after stopping it. The manufacturers advise stopping the medicine at least two months before trying to conceive, but that advice only helps if a pregnancy is planned — and many are not.

Until recently, nobody could give these women a clear answer about what the exposure meant for their baby. That gap is what this study set out to fill [1], and a companion article published in the same journal issue explains what the results should and should not be taken to mean [2].

What families and doctors faced before this study

For a long time, the advice given to women with obesity who wanted to become pregnant was essentially to lose weight first. Major professional organisations built their guidance around this idea — both the American College of Obstetricians and Gynecologists [3] and the International Federation of Gynecology and Obstetrics [4] describe weight management before pregnancy as one of the most useful things that can be done to improve the odds of a healthy pregnancy. But that guidance was written before there was a medicine that could reliably produce that much weight loss, and it never anticipated the question of what happens if a woman is still on the medicine when she conceives.

The scientific evidence that existed was patchy and, frankly, contradictory. One reassuring piece came from a study that pooled records from four Nordic countries, the United States, and Israel and found no increase in birth defects among babies exposed to these medicines early in pregnancy [5]. But when researchers looked at other outcomes — how long the pregnancy lasted, whether the mother developed high blood pressure or diabetes of pregnancy — the studies disagreed sharply. One American hospital system reported more problems, including early birth [6]. Another study using a large database reported fewer problems [7]. A 2026 review that gathered thirty-six studies together concluded that no consistent pattern of harm to babies had yet been demonstrated, but that the evidence was too thin to be confident either way [8]. Families asking a straightforward question were getting three different answers.

How the Danish researchers approached it

Denmark keeps national health records that can be linked using a personal identification number given to every resident, so researchers can follow what medicines people collected from pharmacies, what happened during their pregnancies, and how their babies were born — for an entire country, over many years, without having to recruit anyone.

The research team, based at Copenhagen University Hospital Hvidovre and the University of Copenhagen, looked at every single-baby pregnancy in Denmark that ended in a birth between October 2009 and December 2023. That came to 756,636 pregnancies among 480,231 women. They counted a woman as "exposed" if she had collected a prescription for liraglutide or semaglutide within eight weeks before or eight weeks after her last menstrual period — a sixteen-week window chosen because it covers the time when a woman may not yet know she is pregnant. That produced 529 exposed pregnancies: 295 involving liraglutide and 234 involving semaglutide [1].

Here is the difficulty the researchers had to solve. The women taking these medicines were not like the other women in the study. Their average body mass index was around 33 or 34, compared with about 23 or 24 for everyone else. And a third to nearly half of them already had diabetes before pregnancy, compared with roughly one and a half in every hundred of the other women. Simply comparing the two groups would tell you about obesity and diabetes, not about the medicine.

To get around this, the researchers used a statistical technique called propensity score matching. In plain terms, they built comparison groups of unexposed women who resembled the exposed women as closely as possible on age, weight, smoking, education, region, number of previous births, whether they had diabetes, and the year of the pregnancy. Then they compared like with like. They did this six different ways, including deliberately leaving weight and diabetes out of the matching to see how much difference those two factors alone were making.

What they found

When the exposed women were compared with the general population — without any adjustment — almost everything looked worse. Early birth (before 37 weeks) occurred in 12 out of every 100 liraglutide pregnancies and 14 out of every 100 semaglutide pregnancies, compared with 5 out of every 100 unexposed pregnancies. Pre-eclampsia, a serious blood pressure condition of pregnancy, occurred roughly three times as often. Babies were more likely to be large for their gestational age.

But once the women were properly matched with similar women, nearly all of these differences disappeared. There was no longer any increase in pre-eclampsia, in diabetes developing during pregnancy, in birth weight, in babies being unusually large or unusually small, or in the weight of the placenta. There were also no differences in stillbirth or in the two placental complications the researchers examined.

One difference remained: babies were still being born slightly earlier. On average, pregnancies exposed to liraglutide were about two days shorter, and those exposed to semaglutide about six days shorter, than closely matched comparison pregnancies [1].

Then came the finding that changes the whole picture. When the researchers separated women who were taking the medicine for diabetes from women taking it for weight loss, the extra risk of early birth appeared only in the diabetes group. Among women with diabetes, the odds of early birth were roughly 70 to 84 per cent higher. Among women without diabetes taking the same medicines for weight management, there was no increase at all [2].

This matters because a medicine does not know why it was prescribed. If the injection itself were shortening pregnancies, it should have done so in both groups. The fact that it only appeared in women with diabetes points strongly to the diabetes as the real cause. Diabetes present before pregnancy has been known for twenty-five years to raise the chance of an early birth [9], largely because it can damage blood vessels and the placenta in ways that sometimes make doctors decide to deliver a baby early for safety [10]. A few weeks of weight loss just before conception cannot undo years of that process.

What this means for you and your baby

If you were taking one of these medicines for weight management and you did not have diabetes, this study offers real reassurance. In this large national dataset, women in your situation were no more likely to give birth early, no more likely to develop pre-eclampsia or pregnancy diabetes, and their babies were no more likely to be unusually small or unusually large.

If you have diabetes and were taking one of these medicines, the higher chance of an early birth is real — but the most useful thing to know is that it is very likely coming from the diabetes rather than the injection. That reframes what you and your team should focus on: managing your blood sugar carefully and keeping to your monitoring appointments will do more for your pregnancy than worrying about an exposure that has already happened.

It is worth keeping the size of the effect in perspective. A shift of two to six days in the length of pregnancy is small at the level of any one baby. It matters mostly at the level of populations, because babies born even a few weeks early — the "late preterm" group — are somewhat more likely to need help with breathing, blood sugar, feeding, or jaundice than babies born at full term [11], and preterm birth remains the single biggest contributor to newborn illness worldwide, affecting an estimated 13.4 million babies in 2020 [12]. If you would like to read more about how a mother's diabetes can affect a child over the longer term, our related article on maternal diabetes and child neurodevelopment covers that ground.

What this study could not answer

No single study settles a question like this, and the researchers were open about the gaps. Their records show which prescriptions were collected from the pharmacy, but not whether a woman actually used the medicine, at what dose, or when she stopped. The records also contain nothing about how long a woman had had diabetes or how well controlled it was — so some of the extra risk seen in the diabetes group could reflect more severe diabetes rather than anything else. And 529 exposed pregnancies is not a large number once divided between two medicines and two reasons for taking them, so small effects could still be hiding in the data. The study was also funded in part by a foundation connected to the company that makes both medicines, which is worth knowing even though the results come from independent national records.

What researchers are working on next

The next generation of studies needs information Danish registers do not hold: how long each woman had diabetes, what her blood sugar control looked like, when she stopped the injection, and what dose she was on. Researchers also want pregnancy registries that women can join early, so that pregnancies ending in miscarriage — which this study could not include, because it counted only pregnancies ending in a birth — are captured too. And a newer family of medicines acting on two gut hormones at once is now reaching women of childbearing age, with even less pregnancy data behind it [8]. For now, the message from Denmark is a useful one: for women without diabetes, an accidental exposure around conception does not appear to have shortened pregnancy or harmed the baby's growth.

References

  1. Hviid KVR, Banasik K, Mortensen LH, Madsbad S, Strandberg-Larsen K, Geiker NRW, Westergaard D, Nielsen HS. Periconceptional GLP-1 receptor agonist exposure and obstetric outcomes: a Danish nationwide cohort study. Human Reproduction Open. 2026;2026(2):hoag015. doi:10.1093/hropen/hoag015 ↩
  2. Zhu Y, Hedderson MM. When drugs meet disease: disentangling diabetes, obesity, and periconceptional GLP-1 receptor agonist safety. Human Reproduction Open. 2026;2026(2):hoag016. doi:10.1093/hropen/hoag016 ↩
  3. Sagi-Dain L. Obesity in pregnancy: ACOG Practice Bulletin, Number 230. Obstetrics & Gynecology. 2021;138:489. doi:10.1097/AOG.0000000000004527 ↩
  4. Hanson MA, Bardsley A, De-Regil LM, et al. The International Federation of Gynecology and Obstetrics (FIGO) recommendations on adolescent, preconception, and maternal nutrition: "Think Nutrition First". International Journal of Gynaecology and Obstetrics. 2015;131:S213–S253. doi:10.1016/S0020-7292(15)30023-0 ↩
  5. Cesta CE, Rotem R, Bateman BT, et al. Safety of GLP-1 receptor agonists and other second-line antidiabetics in early pregnancy. JAMA Internal Medicine. 2024;184:144–152. doi:10.1001/jamainternmed.2023.6663 ↩
  6. Maya J, Pant D, Fu Y, et al. Gestational weight gain and pregnancy outcomes after GLP-1 receptor agonist discontinuation. JAMA. 2025;334:2186–2196. doi:10.1001/jama.2025.20951 ↩
  7. Imbroane MR, LeMoine F, Nau CT. Preconception glucagon-like peptide-1 receptor agonist use associated with decreased risk of adverse obstetrical outcomes. American Journal of Obstetrics and Gynecology. 2025;233:116.e1–116.e7. doi:10.1016/j.ajog.2025.01.019 ↩
  8. Ozbek L, et al. Safety of GLP-1 and dual GLP-1/GIP receptor agonists in preconception, pregnancy, and lactation: a systematic review of maternal, fetal, and neonatal outcomes. Diabetes, Obesity and Metabolism. 2026;28(6):4503–4528. doi:10.1111/dom.70699 ↩
  9. Sibai BM, Caritis SN, Hauth JC, et al. Preterm delivery in women with pregestational diabetes mellitus or chronic hypertension relative to women with uncomplicated pregnancies. American Journal of Obstetrics and Gynecology. 2000;183:1520–1524. doi:10.1067/mob.2000.107621 ↩
  10. Domingueti CP, Dusse LMSA, das Graças Carvalho M, de Sousa LP, Gomes KB, Fernandes AP. Diabetes mellitus: the linkage between oxidative stress, inflammation, hypercoagulability and vascular complications. Journal of Diabetes and its Complications. 2016;30:738–745. doi:10.1016/j.jdiacomp.2015.12.018 ↩
  11. Teune MJ, Bakhuizen S, Gyamfi Bannerman C, et al. A systematic review of severe morbidity in infants born late preterm. American Journal of Obstetrics and Gynecology. 2011;205:374.e1–374.e9. doi:10.1016/j.ajog.2011.07.015 ↩
  12. Ohuma EO, Moller A-B, Bradley E, et al. National, regional, and global estimates of preterm birth in 2020, with trends from 2010: a systematic analysis. The Lancet. 2023;402(10409):1261–1271. doi:10.1016/S0140-6736(23)00878-400878-4) ↩