By Dr. Asnat Raziel, bariatric surgeon specializing in obesity treatment, metabolic health and healthy longevity; Director of the Multidisciplinary Center for Obesity Treatment, Assia Medical, Assuta Ramat HaChayal, Tel Aviv
GLP-1 in diabetes and pregnancy: a new Lancet review sums up what we know
GLP-1 drugs have become part of routine diabetes care, including for women of reproductive age. So what happens to GLP-1 treatment in diabetes and pregnancy, when a woman conceives while taking it? Many women also ask whether they can go back to the drug after giving birth. A new systematic review in The Lancet Obstetrics, Gynaecology & Women’s Health offers answers. It analyzed data on more than 40,000 women with diabetes. The main conclusion is reassuring: exposure to these drugs around the start of pregnancy did not raise the rate of birth defects. At the same time, the researchers stress that a lot of data is still missing. That is why they also published practical recommendations from an international expert panel.
Why does this question matter so much?
Diabetes in pregnancy is becoming more common. This is true for type 1 diabetes, type 2 diabetes and gestational diabetes. According to the review, about one in six pregnancies worldwide now involves high blood sugar. This raises the risk of birth defects and pregnancy loss. It also raises the risk of preeclampsia, a very large baby and cesarean delivery. For this reason, getting blood sugar and weight under control before pregnancy is a key step toward a healthy pregnancy.
This is where GLP-1 drugs come in. Examples include semaglutide (Ozempic and Wegovy), tirzepatide (Mounjaro) and liraglutide (Saxenda). They control blood sugar and lead to significant weight loss. On the other hand, regulators have not approved them for use during pregnancy. The drug labels recommend stopping them before conceiving and avoiding them while breastfeeding. Until now, this advice has relied mainly on animal studies. In real life, however, many women become pregnant while on treatment. I have written about this before, in an article on incretins, fertility and pregnancy.
What did the study examine?
An international team led by researchers from the University of Leicester and King’s College London carried out the work. They searched four medical databases up to December 19, 2025. Out of 1,392 studies, 23 met the criteria. Of these, 7 went into the meta-analysis, a combined statistical analysis. Notably, 19 of the 23 studies on GLP-1 drugs in diabetes and pregnancy appeared in 2023 or later, which shows how young this field is. Most of the women in the studies had type 2 diabetes. The studies also included women with type 1 diabetes and women with previous gestational diabetes.
Beyond the data analysis, the researchers invited 42 experts from five continents to take part in a Delphi process. In this method, experts reach agreement through rounds of questionnaires. Of the experts, 33 took part in the first round and 32 in the second. Each recommendation required at least 70% agreement. What’s more, women living with diabetes shared their own experience to help shape the questions and comment on the final document.
Birth defects: the reassuring finding on GLP-1 in diabetes and pregnancy
This is the biggest fear of any woman who finds out she is pregnant during treatment. Four studies, with 42,144 women exposed to the drugs, gave a clear answer. In the exposed group, 1,440 babies had a birth defect, or about 3.4 in 100. The comparison group included 137,091 women. There, the rate was the same: 4,705 cases, again about 3.4 in 100. The risk ratio was 1.02 (95% confidence interval: 0.96 to 1.08). In other words, there was no difference at all.
It is important to understand the timing of exposure. In most studies, women took the drug from about 90 days before their last menstrual period through the first trimester. A similar finding emerged from 21 pregnancies described in case reports. The researchers saw no pattern of defects that pointed to the drug. Likewise, data from drug company clinical trials pointed in the same direction. Of 112 women who received a GLP-1 drug, 2 (1.8%) had a baby with a birth defect. By comparison, the placebo group had 2 cases among 38 women (5.3%).
Miscarriage: a finding to interpret with caution
Here the picture is less simple. Two studies, with 41,046 exposed women, showed a high rate of early pregnancy loss. The exposed group had 4,031 cases, or about 9.8 in 100. In the comparison group, the rate was about 7.5 in 100. The risk ratio was 1.31 (1.26 to 1.35). However, the largest study contributed 99.5% of the weight of the analysis. That study combined miscarriages and elective abortions into a single outcome. So the researchers state plainly that the finding is hard to interpret.
Other studies support this caution. One study looked only at elective abortions. There, the rate was much higher in the GLP-1 group (adjusted risk ratio 3.89). In contrast, a case-control study looked only at miscarriages in women with type 2 diabetes. It found no increase in risk at all (odds ratio 1.05, confidence interval 0.75 to 1.48). Much of the gap may therefore reflect unplanned pregnancies that women chose to end. That is not necessarily direct harm to the fetus. Still, this remains an open question that needs more research.
Preterm birth, preeclampsia and birth weight
For the other outcomes, the combined analysis found no differences. Preterm birth: risk ratio 0.99 (0.81 to 1.21) across six studies. High blood pressure in pregnancy and preeclampsia: 1.00 (0.81 to 1.24). Cesarean delivery: 0.98 (0.85 to 1.14). Stillbirth or newborn death: 1.18 (0.87 to 1.61), not statistically significant. Small for gestational age: 1.11 (0.66 to 1.86). Interestingly, a baby that was large for gestational age tended to be less common: 0.78 (0.60 to 1.01).
It is worth mentioning a nationwide Danish study that the meta-analysis did not include. That study did find more preterm births in women with pre-existing diabetes. The adjusted odds ratio was 1.70 for liraglutide and 1.84 for semaglutide. In addition, an American study looked at women with diabetes who stopped the drug. They gained an average of 12.41 kg during pregnancy, compared with 9.58 kg in the comparison group. The difference was 2.82 kg. The researchers attribute this to weight regain after stopping treatment. I covered this in more detail in an article on GLP-1 drugs and pregnancy weight gain.
After birth and breastfeeding: what we know about GLP-1 drugs
Data on the period after birth are especially scarce. The only controlled study included 72 women after gestational diabetes who were not breastfeeding. Liraglutide plus metformin led to a 7.2% loss of body weight over 84 weeks. Metformin alone led to a loss of only 3.1%. Insulin sensitivity, blood pressure and the lipid profile also improved.
As for breastfeeding, two pharmacokinetic studies tested whether the drug passes into breast milk. In a study of semaglutide in eight breastfeeding women, testing found no drug in any milk sample. A preliminary study of tirzepatide in five women showed a similar result. Even so, the researchers emphasize that there are still no clinical safety data on breastfed babies. This is especially true for the new pills, which contain absorption enhancers.
Expert panel recommendations on GLP-1 in diabetes and pregnancy
The experts make clear that these recommendations reflect expert opinion in areas where information is still lacking. They are not evidence-based guidelines. Nevertheless, this is the first international consensus on the subject.
GLP-1 for type 2 diabetes before pregnancy
GLP-1 drugs are likely to help control blood sugar and weight before pregnancy. The experts recommend 12 to 18 months of treatment for the greatest benefit. Six months is the shortest course that may help. The first target is an HbA1c (glycated hemoglobin, a measure of average blood sugar over the past three months) below 6.5%. The second target is a BMI below 25, or a loss of at least 5% to 10% of body weight. There is no preference between semaglutide and tirzepatide. In addition, women must use contraception during treatment and afterward. The waiting period is one month after stopping tirzepatide and two months after stopping semaglutide. Finally, it is important to plan ahead for how to prevent weight regain after stopping.
If you find out you are pregnant during treatment
The recommendation is clear: stop the drug as soon as you learn you are pregnant. You can find a detailed guide in my article on stopping weight loss drugs for pregnancy.
After birth
Experts strongly recommend breastfeeding for women with diabetes, for the benefit of both mother and baby. However, there are currently not enough safety data to support GLP-1 drugs during breastfeeding. Some women choose them anyway, for example because of uncontrolled blood sugar or weight gain on insulin. In that case, the experts recommend the lowest dose that works. It is also important to keep eating a complete diet. If weight loss is too fast, treatment should stop. I describe the next steps in my guide to losing weight after giving birth.
Type 1 diabetes and gestational diabetes
For type 1 diabetes, there are almost no data. Therefore, the decision is personal and depends on the balance of benefit and risk. If a woman does start treatment, her insulin dose needs adjusting. Her insulin pump settings also need a monthly review while the dose is going up. This helps prevent hypoglycemia and diabetic ketoacidosis. After gestational diabetes, the experts recommend considering at least 12 months of treatment before the next pregnancy. The risk is low, and the benefit may be high. This is especially true for women with prediabetes.
What should we remember about the study’s limitations?
The researchers rated the quality of the evidence as very low. The reason is that most studies were observational, and many did not adjust for weight and blood sugar control before treatment. Most of the data on GLP-1 drugs in diabetes and pregnancy relate to type 2 diabetes, and most of the women were of European ancestry. Moreover, there is almost no information on exposure in the second and third trimesters. As a result, the researchers describe the findings as hypothesis-generating only. They call for large prospective studies. You should also know that some of the authors reported ties to drug companies.
Bottom line: GLP-1 in diabetes and pregnancy
The largest review on the subject to date gave a reassuring answer. Exposure to GLP-1 drugs around the start of pregnancy did not raise the risk of birth defects, preterm birth or preeclampsia. On the other hand, the finding on pregnancy loss needs further study. For now, it does not count as proven harm. The practical message is simple. For women with type 2 diabetes who are planning a pregnancy, treatment may greatly improve their starting point. The condition is advance planning, contraception and a planned stop before conceiving. So it is worth talking with your treating doctor and your gynecologist before making any decision.
The information in this article is general and does not constitute personal medical advice.
References
Subcutaneous Semaglutide during Breastfeeding: Infant Safety Regarding Drug Transfer into Human Milk
Pregnancy effects from discontinued GLP-1 use
Funding: the UK National Institute for Health and Care Research (NIHR) and academic research funds. Some of the authors reported consulting fees, speaker fees and research grants from drug companies, including Novo Nordisk and Eli Lilly.
