1. Introduction and Background
Semaglutide, a GLP1 agonist, is used for glycemic control in Type 2 diabetes and weight management [1]. In animals (mice, rats and rabbits), GLP1 agonists is associated with embryofetal mortality, structural abnormalities and growth disturbances concerning delayed ossification and skeletal variants [2]. A recent cohort study on the potential risk of malformations included GLP‐1 receptor agonists but did not provide specific estimates for semaglutide [3]. Except for case reports, there are no studies on other pregnancy outcomes [1, 4, 5].
Unintentional use of semaglutide in pregnancy is rapidly increasing. Women who experience such exposure are difficult to advise due to the lack of human data. We report pregnancy and neonatal outcomes of pregnancies exposed to semaglutide based on data from a cohort of over 100 000 pregnancies.
2. Materials and Methods
The study population included all singleton pregnancies surviving the first trimester in the Central Denmark Region, which accounts for one fifth of Denmark's population. This encompassed pregnancies from a gestational age (GA) of 12 + 0/7 weeks onwards, with conception dates ranging from 1 June 2016 to 22 June 2023, regardless of foetal survival. In Denmark, there is universal access to healthcare free of charge.
Data were collected from the electronic charts from all hospitals in the region and linked on the individual level, including exact mother–child linkage, via a unique identifier from the Danish Civil Registration System.
Exposure to semaglutide (ATC A10BJ06) and/or insulin (ATC A10B) in first trimester (from the estimated conceptiondate to 70 days after, corresponding to GA 12 + 0/7), pregnancy and neonatal outcomes and informations regarding diabetes were collected from the electronic patient record.
Major malformations were classified according to EUROCAT classification including both prenatal (ultrasound data) and postnatal registered malformations as previously reported, with a deviation with reverse registration of congenital malformations (DQ) diagnoses found later in life (until June 2024) [6]. The neonatal outcomes of respiratory distress syndrome (RDS), hypoglycemia and jaundice were identified using ICD10 codes (RDS DP220, 228, 229/hypoglycemia DP703, 704/jaundice DP550, 551, 558‐560, 569, 570, 578‐585, 588‐593, 598, and 599). The outcomes of semaglutide‐exposed pregnancies were compared to pregnancies exposed to either insulin or no diabetes medication.
Statistical analyses were performed with Stata (Stata Corp, College Station, TX, United States).
Wilcoxon ranksum test was used for numerical data and Fishers exact test for dichotomous data.
Access to the data was approved according to Danish legislation (Sundhedsloven §46, permission [1‐45‐70‐88‐22, 5 April 2024]), and data were stored according to the EU GDPR rules.
The study was conducted in accordance with the Basic & Clinical Pharmacology & Toxicology policy for experimental and clinical studies [7].
3. Results
The population included 104 422 singleton pregnancies. Of these, 32 were exposed to semaglutide during the first trimester: 27 used Ozempic (22 of these in combination with insulin), and five used Wegovy. Seventeen discontinued semaglutide in the first trimester, and 14 used semaglutide throughout all trimesters. The insulin comparison group included 547 pregnancies with no simultaneous use of semaglutide. The semaglutide‐exposed group was older and had a higher prepregnancy BMI than the two comparison groups (Table 1).
TABLE 1.
Characteristics of singleton pregnancies exposed to semaglutide, insulin or nonexposure in first trimester.
| Semaglutide (± insulin) | Insulin (−semaglutide) | No registred semaglutide or insulin | |
|---|---|---|---|
| (n = 32) | (n = 547) | (n = 103 843) | |
|
Maternal age at birth or abortion, years mean, (SD) (Missing 0/0/212) |
33.7 (5.5) | 31.1 (5.0) | 30.4 (4.6) |
|
Weight prepregnant kg mean, (SD) (Missing 2/34/3895) |
98.8 (23.2) | 78.6 (18.4) | 70.3 (15.9) |
|
BMI prepregnant mean, (SD) (Missing 2/34/3990) |
34.7 (7.1) | 27.9 (6.1) | 24.9 (5.3) |
| Diabetes diagnoses (any), in pregnancies going further than 22 weeks, n (%) | 24/30 (80.0) | 519/528 (98.3) | 4681/100 887 (4.6) |
| Diabetes diagnoses (any) distribution: | 24 | 519 | 4681 |
| Type 1 diabetes, n (%) | < 10 a | 371 (71.5) | 36 (0.8) |
| Type 2 diabetes, n (%) | 15 (62.5) | 134 (25.8) | 96 (2.1) |
| Gestational diabetes, n (%) | < 3 a | 14 (2.7) | 4549 (97.2) |
| Livebirth, n (%) | 29 (90.6) | 517 (94.5) | 100 448 (96.7) |
|
GA birth, weeks + days (SD days) GA ≥ 22 + 0 (Missing 2/20/3055) |
38 + 0 (21.3) | 37 + 4 (16.4) | 39 + 6 (12.9) |
Masked numbers.
There was no increased risk of major malformations among semaglutide‐exposed pregnancies compared to insulin‐exposed or unexposed pregnancies (Table 2). Semaglutide‐exposed pregnancies had a higher risk of preterm birth, large for gestational age (LGA), neonatal hypoglycemia and jaundice compared to unexposed pregnancies, but the risks were similar to those in insulin‐exposed pregnancies (Table 2).
TABLE 2.
Pregnancy and neonatal outcomes of singleton pregnancies exposed to semaglutide compared to insulin or nonexposure in first trimester.
| Wilcoxon ranksum test, numerical data | Semaglutide (± insulin) | Insulin (− semaglutide) | Nonexposure of semaglutide or insulin | Semaglutide vs insulin | Semaglutide vs. nonexposure |
|---|---|---|---|---|---|
| (n = 30) | (n = 527) | (n = 100 788) | p value | p value | |
| (n = 29) | (n = 518) | (n = 99 833) | |||
|
GA birth, days GA ≥ 22 + 0 (n = 30/527/100788) |
265 | 263 | 279 | 0.04 | < 0.001 |
|
Weight at birth, g (n = 29/518/99833) |
3623 | 3706 | 3546 | 0.31 | 0.31 |
| Fishers exact test, dichotomous data | Cases/exposed, (%) | Cases/exposed, (%) | Cases/unexposed, (%) |
2‐sided Fisher's exact p value OR (95% CI) |
2‐sided Fisher's exact p value OR (95% CI) |
|---|---|---|---|---|---|
| Major malformations | 3/32 (9.4) | 76/547 (13.9) | 7985/103 843 (7.7) |
p = 0.784 0.64 (0.12–2.15) |
p = 0.485 1.24 (0.24–4.01) |
|
Preterm birth GA ≥ 22 + 0 − <GA 37 + 0 |
6/30 (20.0) | 122/527 (23.1) | 4846/100 788 (4.8) |
p = 1.000 0.83 (0.27–2.15) |
p = 0.002 4.95 (1.65–12.43) |
| SGA − 15% (10 p) of registered livebirths | 0/29 (0) | 21/514 (4.1) | 9447/99 522 (9.5) | — | — |
| LGA + 22% (3 p) of registered livebirths | 6/29 (20.7) | 218/514 (42.4) | 4184/99 522 (4.2) |
p = 0.021 0.35 (0.12–0.92) |
p = 0.001 5.94 (1.98–15.02) |
| RDS of registered livebirths | 3/29 (10.3) | 46/517 (8.9) | 2408/100 448 (2.4) |
p = 0.722 1.18 (0.22–4.08) |
p = 0.024 4.70 (0.91–15.34) |
| Hypoglycemia of registered livebirths | 3/29 (10.3) | 88/517 (17.0) | 1448/100 448 (1.4) |
p = 0.782 0.62 (0.12–2.10) |
p = 0.006 7.89 (1.53–25.78) |
| Jaundice of registered livebirths | 5/29 (17.2) | 107/517 (20.7) | 5947/100 448 (5.9) |
p = 1.000 0.80 (0.23–2.20) |
p = 0.017 3.31 (0.99–8.86) |
Abbreviations: GA, gestational age; LGA, large for gestational age; RDS, respiratory distress syndrome; SGA, small for gestational age.
4. Discussion
Semaglutide exposure was not associated with an increased risk of major malformations, but the analyses were based on only 32 exposed pregnancies with limited power. Compared to unexposed, semaglutide was associated with increased risks of preterm birth, LGA, neonatal hypoglycemia and jaundice. However, semaglutide was not associated with these outcomes when compared to pregnancies exposed to insulin, possibly reflecting that most semaglutide exposed also used insulin, were older and had a higher BMI.
The main limitation was that only 32 were exposed to semaglutide. Due to the small number of exposed individuals, it was not possible to adjust for potential confounders.
It is likely that the use of semaglutide among women without diabetes will increase. In our data, the first 27 exposed were Ozempic (2019–2023) and the five recent were women without diabetes with Wegovy exposure (2023). We speculate that unintended use of semaglutide in early pregnancy may increase proportionally to the increased use of semaglutide for weight loss, which may increase fertility among women with overweight. Further, maternal weight gain after discontinuation of semaglutide may per se cause obstetric complications.
In conclusion, 32 pregnancies exposed to semaglutide and insulin showed comparable risks of malformations, preterm birth, LGA, neonatal hypoglycemia and jaundice compared to pregnancies exposed solely to insulin. Given the current knowledge gaps and the limited availability of both animal and human studies, it remains unsafe to revise the recommendation against use during pregnancy. Unintended use of semaglutide during pregnancy is likely to increase, and larger studies are needed.
Conflicts of Interest
The authors declare no conflicts of interest.
Acknowledgements
LHP is supported by a Borregaard Clinical Ascending Investigator Grant from the Novo Nordisk Foundation (NNF18OC0054457). The Novo Nordisk Foundation had no role in the design and conduct of the study; the collection, analysis and interpretation of the data; or the preparation, review or approval of the manuscript.
Funding: This work was supported by the Borregaard Clinical Ascending Investigator Grant (NNF18OC0054457).
Contributor Information
Line Kolding, Email: lkolding@clin.au.dk.
Lars Henning Pedersen, Email: lhp@clin.au.dk.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.
References
- 1. Chao A. M., Tronieri J. S., Amaro A., and Wadden T. A., “Clinical Insight on Semaglutide for Chronic Weight Management in Adults: Patient Selection and Special Considerations,” Drug Design, Development and Therapy 16 (2022): 4449–4461, 10.2147/DDDT.S365416. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2. Muller D. R. P., Stenvers D. J., Malekzadeh A., Holleman F., Painter R. C., and Siegelaar S. E., “Effects of GLP‐1 Agonists and SGLT2 Inhibitors During Pregnancy and Lactation on Offspring Outcomes: A Systematic Review of the Evidence,” Frontiers in Endocrinology (Lausanne) 14 (2023): 1215356, 10.3389/fendo.2023.1215356. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3. Cesta C. E., Rotem R., Bateman B. T., et al., “Safety of GLP‐1 Receptor Agonists and Other Second‐Line Antidiabetics in Early Pregnancy,” JAMA Internal Medicine 184, no. 2 (2024): 144–152, 10.1001/jamainternmed.2023.6663. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Nuako A., Tu L., Reyes K. J. C., Chhabria S. M., and Stanford F. C., “Pharmacologic Treatment of Obesity in Reproductive Aged Women,” Current Obstetrics and Gynecology Reports 12, no. 2 (2023): 138–146, 10.1007/s13669-023-00350-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5. Drummond R. F., Seif K. E., and Reece E. A., “Glucagon‐Like Peptide‐1 Receptor Agonist use in Pregnancy: A Review,” American Journal of Obstetrics and Gynecology 232 (2024): 17–25, 10.1016/j.ajog.2024.08.024. [DOI] [PubMed] [Google Scholar]
- 6. Liu X., Kolding L., Momen N., Gasse C., and Pedersen L. H., “Maternal Antipsychotic use During Pregnancy and Congenital Malformations,” American Journal of Obstetrics and Gynecology MFM 5, no. 6 (2023): 100950, 10.1016/j.ajogmf.2023.100950. [DOI] [PubMed] [Google Scholar]
- 7. Tveden‐Nyborg P., Bergmann T. K., Jessen N., Simonsen U., and Lykkesfeldt J., “BCPT 2023 Policy for Experimental and Clinical Studies,” Basic & Clinical Pharmacology & Toxicology 133, no. 4 (2023): 391–396, 10.1111/bcpt.13944. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.