Should You Take a GLP-1 Before Trying to Conceive With PCOS?

What did the new study actually find?

Fewer adverse pregnancy outcomes in the treated group, across a composite of almost everything that can go wrong.

The researchers ran what they call a target trial emulation using TriNetX, a federated network of health records. Women aged 18 and over with PCOS who received liraglutide, semaglutide or tirzepatide before conception were matched one to one against women receiving metformin, with matching on demographics, anthropometrics, comorbidities and medications.¹

Outcome over two years	Hazard ratio	95% CI
Composite of adverse pregnancy outcomes	0.38	0.34 to 0.41
Type 2 diabetes	0.31	0.30 to 0.33
Major adverse cardiovascular events	0.88	0.81 to 0.95

The composite covered hypertensive, glycaemic, placental, delivery, labour, foetal growth, preterm and early pregnancy complications.¹ A hazard ratio of 0.38 means roughly 62% fewer of those events in the treated group.

The paper's own title says the therapy "reduces" these outcomes. That is a causal verb on an observational design, and the design cannot support it. The abstract's conclusion is more careful, describing results "consistent with potential benefits of pre-conception metabolic optimisation."¹

Why is a hazard ratio of 0.38 hard to take at face value?

Because of its size. A 62% reduction across eight categories of complication is larger than almost any drug achieves in any population.

The two groups were not assembled the same way. Metformin is first line for PCOS, cheap, and prescribed in primary care to almost anyone with the diagnosis, since insulin resistance is part of the condition for most women who have it. A GLP-1 before a planned pregnancy usually means specialist involvement, a weight-management or fertility pathway, and someone being monitored closely. Those women are not a random half of the PCOS population.

So the comparison may be partly measuring the care around the drug rather than the drug. Propensity matching on demographics, anthropometrics, comorbidities and medications does not capture whether someone was under specialist supervision, how much they wanted the pregnancy, or whether anyone was watching their blood pressure monthly. That is the standard criticism of this design, and the size of the effect gives it force here.

One line in the abstract is easy to read past: "Associations varied across some sensitivity analyses."¹ In an observational study, sensitivity analyses are where a real effect proves durable or an artefact falls apart. The abstract does not say which ones varied, or by how much.

The composite is loose enough to hide what drove it. Bundling gestational diabetes with preterm birth and placental complications produces one number, and that number can be driven almost entirely by one component. Improved glycaemic outcomes are the most biologically plausible part of this result, and a composite hides how much of the 62% is just that.

One more thing about the comparator. Metformin can be continued into pregnancy in some circumstances, and the PCOS guideline says it is not routinely recommended in pregnancy but could be considered where there is a risk of preterm birth.⁷ A GLP-1 is a drug you are told to stop before conceiving. So the two arms differ in their relationship to the pregnancy and not only in which drug they took, and the metformin group may include some women exposed into pregnancy while the GLP-1 group by definition were not.

None of this means the finding is wrong. It means the result needs a randomised trial before anyone treats it as settled.

What did the study say about pregnancy itself?

Nothing. The authors say so twice, in two separate sentences.

The abstract states: "This study cannot evaluate in-pregnancy safety or efficacy."¹ It then states: "These findings do not support GLP-1/GIP RA use during gestation."¹

Those two sentences are the ones the coverage will drop. A result framed as "GLP-1s cut pregnancy complications by 62%" reads as reassurance about the drugs around pregnancy, and the study was built so that it could not say that. Everyone in the exposed group had stopped before conceiving.

Has anyone studied pregnancies that were already exposed?

Yes, repeatedly. None of the three analyses found a malformation signal, and the new study is not the first large look at exposure around conception.

This is the question the new study says it cannot answer, and it is the one a frightened reader is actually asking. Three analyses have addressed it.

Analysis	What it looked at	Headline result
JAMA Internal Medicine, 202412	938 pregnancies with periconceptional GLP-1 exposure in women with pre-existing type 2 diabetes, drawn from 3,514,865 pregnancies across the Nordic countries, the United States and Israel, compared against 5,078 on insulin	Major congenital malformations in 8.3% of the GLP-1 group against 7.8% on insulin, adjusted relative risk 0.95, 95% CI 0.72 to 1.26
Endocrine Connections, July 202613	Meta-analysis of six studies, 286,599 women, of whom 43,577 were exposed to a GLP-1 drug within three months before conception or during pregnancy	No significant difference in major congenital malformations, relative risk 1.02, 95% CI 0.96 to 1.08
Diabetes, Obesity and Metabolism, March 202614	Systematic review of 36 studies covering preconception, pregnancy and lactation exposure	Periconceptional or early-pregnancy exposure not consistently associated with increased major congenital malformations, foetal growth restriction, stillbirth or neonatal mortality in adjusted analyses

Three analyses, none showing a malformation signal, and all three sets of authors declining to call it proof. The meta-analysis says the findings "should not be taken as proof of safety" and calls for prospective cohorts and randomised trials before pregnancy.¹³ The review says data on continued use throughout gestation "remain limited".¹⁴ The JAMA authors note that confounding by obesity and cardiovascular conditions "would preferentially affect GLP-1 receptor agonists", and that filled prescriptions may not reflect what was taken during the weeks when organs form.¹²

So the accurate summary is narrower than either "we know nothing" or "it is safe". Accidental exposure in the weeks around conception has been looked at in tens of thousands of pregnancies without a malformation signal appearing. Choosing to stay on one through a pregnancy is a different question, and that is the one with almost nothing behind it.

One finding on the next stage, since it comes up immediately. The same review reports a single pharmacokinetic study that found no detectable semaglutide transfer into human milk.¹⁴ One study is not a conclusion, and our breastfeeding guide covers what the labels say.

What does the PCOS guideline say?

It supports these drugs for weight, and it rules them out for exactly the purpose this study is about.

The 2023 International Evidence-Based Guideline for the Assessment and Management of PCOS is the document clinicians in this field actually work from, and it has four separate items on GLP-1 drugs.⁷

• On weight, conditional support. Anti-obesity medicines including GLP-1 receptor agonists "could be considered, in addition to active lifestyle intervention" at higher weight, following general population guidelines.⁷
• On reproductive outcomes, a different answer. Anti-obesity agents should be used for reproductive outcomes "only in research settings."⁷ That is the specific purpose the new study speaks to, and the guideline's position predates the study and has not been revised.
• On contraception, the thing the semaglutide label leaves out. Clinicians "should ensure concurrent effective contraception when pregnancy is possible" for women taking GLP-1 receptor agonists, "as pregnancy safety data are lacking."⁷
• On stopping, a warning about what follows. Shared decision making should weigh the possible need for long-term use, given the "high risk for weight regain after discontinuation."⁷

The Wegovy label says nothing about contraception. The PCOS guideline covers it, and the guideline is the document more likely to be in front of whoever wrote the prescription.

Two more points from the same guideline put the study in context. Metformin "should be considered in adults with PCOS and a BMI ≥ 25 kg/m2", so the study's comparator is the guideline-recommended drug and not a placebo or a weaker option.⁷ And letrozole, not any weight drug, "should be the first-line pharmacological treatment for ovulation induction" in anovulatory women with PCOS.⁷ A reader deciding whether to ask for a GLP-1 to help conceive should know that the first-line fertility treatment is something else entirely.

Does stopping undo the benefit?

No study has looked. The PCOS guideline names the mechanism that would undo it: weight returning once the drug stops.

The logic of the new study is that metabolic improvement before conception carries forward into the pregnancy. The practical problem is the gap in between. Semaglutide's label asks for at least two months between the last dose and a planned pregnancy, and conception does not always happen on schedule, so the real interval can run much longer.²

The guideline's own practice point names what tends to happen in that window: a high risk of weight regain after discontinuation.⁷ If the preconception benefit comes from the weight and the metabolic change rather than from the drug itself, then regaining weight between stopping and conceiving is the obvious way to lose it, and no study has looked at whether it does.

That is not an argument against preconception treatment. It is the question to put to a prescriber, alongside how long the wait is likely to be.

Whether to restart after stopping is a separate question again, and it changes once there is a baby to feed. Our guide to semaglutide and breastfeeding covers what the labels say about that stage.

How long before pregnancy should you stop?

It depends on which drug you are on. Both semaglutide labels give two months. Neither tirzepatide label gives anything at all.

The semaglutide figure is a label instruction and not an estimate. The Wegovy prescribing information instructs clinicians to "discontinue WEGOVY in patients at least 2 months before they plan to become pregnant", citing the potential for fetal harm and semaglutide's long half-life.² Two months, from the manufacturer, in the label.

Both semaglutide products say the same thing. The Ozempic label, revised 5/2026, instructs clinicians to "discontinue OZEMPIC in women at least 2 months before a planned pregnancy due to the long washout period for semaglutide."⁸ That is the same interval as Wegovy, in the label for the product many women with PCOS are actually prescribed, and it is the label that uses the word washout.

Product	Molecule	Elimination half-life	Preconception interval in the label	Contraception or birth control advice
Wegovy	Semaglutide	About 1 week, in circulation about 5 to 7 weeks after the last dose2	At least 2 months before a planned pregnancy2	None
Ozempic	Semaglutide	About 1 week, in circulation about 5 weeks after the last dose8	At least 2 months before a planned pregnancy8	None in section 8.38
Zepbound	Tirzepatide	Not quoted in the retrieved label	None given3	Non-oral or added barrier method for 4 weeks after starting and after each dose increase3
Mounjaro	Tirzepatide	About 5 days9	None given9	Non-oral or added barrier method for 4 weeks after starting and after each dose increase9
Saxenda	Liraglutide	About 13 hours6	None given6	None in the pregnancy section6

The usual explanation is pharmacokinetics, and it only half works. Liraglutide clears in roughly half a day, so a two-month rule would mean nothing for it and its label gives none.⁶ But tirzepatide's elimination half-life is about five days against semaglutide's one week.⁹² Five days and seven days are not far apart, and five half-lives of each is roughly 25 days against 35. The labels are much further apart than the drugs are. One molecule gets a stated interval and the other gets none, and the pharmacokinetics do not account for the gap.

Zepbound replaces the interval with a different instruction. Its label has no preconception discontinuation interval at all. It says to discontinue when a pregnancy is recognised. That answers a different question from when to stop.³

The two molecules are mirror images on birth control. Both tirzepatide labels warn that the drug may reduce the effectiveness of oral hormonal contraceptives through delayed gastric emptying, and advise switching to a non-oral method or adding a barrier method for four weeks after starting and for four weeks after each dose increase.³⁹ Neither semaglutide label gives contraception advice in its pregnancy or reproductive-potential sections.²⁸

So one molecule tells you when to stop and says nothing about preventing pregnancy, and the other tells you how to prevent pregnancy and says nothing about when to stop. Reading only one of them leaves a gap, and which gap you get depends on which drug you were prescribed.

If you are on something that is not one of these four, there is no label to read. Retatrutide, and everything sold research-labelled, has no pregnancy section because it has no approval.

Does it matter which brand you were prescribed?

For stopping before pregnancy, no. For what happens if you are already pregnant, the brand changes the instruction even when the molecule is identical.

Semaglutide is sold as Wegovy for weight and as Ozempic for type 2 diabetes. Tirzepatide is sold as Zepbound for weight and as Mounjaro for diabetes. Four products, two molecules, four separate labels. Many women with PCOS are on the diabetes product, either for co-existing diabetes or because that is what was prescribed, so the label in front of their prescriber may not be the one quoted in the coverage.

Product	If a patient is or becomes pregnant
Wegovy	"Discontinue WEGOVY in pregnant patients who are using it for weight reduction", on the basis that weight loss offers no benefit in pregnancy and may cause fetal harm2
Ozempic	"OZEMPIC should be used during pregnancy only if the potential benefit justifies the potential risk to the fetus"8
Zepbound	Discontinue when a pregnancy is recognised3
Mounjaro	"MOUNJARO should be used during pregnancy only if the potential benefit justifies the potential risk to the fetus"9

The logic is visible once the two are side by side. Weight loss has no benefit to offer a pregnancy, so the weight-management labels say stop. Glycaemic control does have something to offer, so the diabetes labels leave room for a judgement. The molecule is the same in each pair.

None of that is permission to stay on one. It means the instruction a prescriber is working from depends on which product was dispensed, and a page or a news story that quotes only the Wegovy label is describing one of four documents. If you are on a GLP-1 and pregnant or trying to be, the label to read is the one for your product, and the conversation is with whoever wrote the prescription.

What if the pregnancy was not planned?

PCOS makes this situation more likely, and it is the one the labels cover least well.

These drugs restore ovulation in PCOS. In one study of 96 women, 80 had absent or anovulatory cycles before treatment, and after six months on semaglutide 42 of those 80, 52.5%, had normal ovulatory cycles.⁴ Our guide on GLP-1s and periods covers that effect and its consequences in more detail.

Years of being told that conception is unlikely is what sets this up. Someone in that position may not be using contraception. A GLP-1 can make them fertile again within months. The semaglutide label's instruction is framed around a planned pregnancy, "before they plan to become pregnant", and offers nothing to someone whose fertility returned before any plan existed.

If a pregnancy happens while you are taking one, the label instruction is to stop when the pregnancy is recognised, and that is a call to make with an obstetric provider rather than alone.²³ Early exposure before a positive test is a common situation rather than a catastrophe, and it is a different thing from knowingly continuing. Our fertility and pregnancy guide covers what that conversation involves, including asking about a pregnancy exposure registry.

Who is actually collecting the pregnancy data?

Two registered studies. Both are small and neither has reported.

The reason the in-pregnancy question stays open is not that nobody is working on it. It is that prospective work is registry work. That depends on pregnancies happening and then being followed for a year afterwards.

• The Wegovy Pregnancy Registry. A Novo Nordisk prospective cohort comparing maternal, foetal and infant outcomes in pregnancies exposed to Wegovy against an unexposed reference group, across sites in the United States, the United Kingdom and Spain. Target 728 pregnant women. Data are collected at enrolment, at week 26 and at delivery, and infant data at four and twelve months after birth. Collection is planned to run about ten years, or until the target is reached.¹⁰
• A European cohort on first-trimester exposure. Registered as EUPAS50643, a comparative cohort whose stated primary objective is to "prospectively evaluate the risk of major birth defects" after first-trimester exposure to a GLP-1 agonist, against two reference groups, and which also looks at spontaneous pregnancy losses and terminations. Estimated 200 subjects. The record was first published in January 2023 and its status at the last update was planned.¹¹

Those are outcomes under study, not results. Neither has reported.

What that means for a decision now is simple arithmetic. A ten-year collection window on 728 pregnancies, and a 200-subject birth-defect cohort still listed as planned, is what would have to mature before anyone could answer the in-pregnancy question prospectively. The observational evidence already exists and is reassuring on malformations. What the registries would add is the part observational data cannot give: follow-up of the infants. The preconception question is the one this page is about, and it is the one a reader can actually act on. The new study is the largest single analysis of preconception treatment yet published, though not the first evidence on exposure around conception.

If you are pregnant and have taken one of these drugs, asking your obstetric provider about enrolling in a registry is how a case stops being anecdote. It is also the only route by which the next version of this page gets better numbers.

What is still unknown?

Five things. The first is the one the study was designed around.

• Whether the drug caused the benefit. An observational comparison against metformin, however large and however well matched, cannot separate the drug from the care that comes with it. A randomised preconception trial would, and none has been run.
• How long before conception is actually enough. Two months comes from semaglutide's half-life and is stated in a label. It is not a tested threshold, and no trial has compared stopping at one month against three.
• Which part of the composite moved. Eight categories of complication were counted together. The individual endpoints were analysed as exploratory secondary outcomes and the abstract does not give them.¹
• Anything about the compounds with no label. The study tested liraglutide, semaglutide and tirzepatide. It says nothing about retatrutide or about any research-labelled vial.
• What continued use through a pregnancy does. The reassuring findings all concern exposure around conception, in practice a few weeks before anyone knew. The systematic review states that data on continued use throughout gestation remain limited, and that is the exposure nobody has studied properly in either direction.¹⁴

When does this need a doctor rather than a search?

Every version of this question does. Four labels for two molecules say different things, the PCOS guideline adds an instruction none of them carries, and none of it can be read off a page without knowing which product you were given.

Three situations where the timing matters more than the research: you are on a GLP-1 and actively trying to conceive; you are on one and your cycles have returned after a long absence; or you are pregnant and have taken one in the past few months.

A hazard ratio answers none of those. They need someone who can see your records, knows which drug you are on and can order a test.

If you are in the third situation, stopping and contacting an obstetric provider is the label instruction. Ask about enrolling in a pregnancy exposure registry, because those registries are how the in-pregnancy evidence gets built, and there is one open for semaglutide.

Does this change anything about research-grade vials?

No. For this subject the gap is wider than usual.

The study tested three approved drugs with labels, prescribed and recorded in health systems. Semaglutide is also sold research-labelled: in our October 11 capture it appeared in 180 listings across 84 sellers at a median of $7.10 per mg.⁵ Across our GLP-1 compound library we track 25 compounds, of which 11 are approved somewhere and 14 are not.⁵

For those with no approval there is no pregnancy section, no reproductive-potential section, no discontinuation interval and no exposure registry. The two-month figure that exists for semaglutide exists because a manufacturer ran the pharmacokinetics and a regulator reviewed them. There is no equivalent for a vial. A preconception washout needs a known half-life and verified contents, and a vial offers neither.

Common questions

Does taking a GLP-1 before pregnancy improve outcomes if I have PCOS?

How long before trying to conceive should I stop semaglutide?

How long before pregnancy should I stop tirzepatide?

Can you take a GLP-1 while pregnant?

Do GLP-1 drugs affect birth control?

Can a GLP-1 make me fertile again if I have PCOS?

Does the PCOS guideline recommend GLP-1 drugs before pregnancy?

What is the first-line fertility treatment for PCOS?

Will I regain the weight between stopping and getting pregnant?

Is a hazard ratio of 0.38 reliable?

What did the study measure as an adverse pregnancy outcome?

Was the study randomised?

Is the Ozempic label different from Wegovy on pregnancy?

Does Mounjaro have the same pregnancy advice as Zepbound?

Do GLP-1 drugs cause birth defects?

I got pregnant while taking a GLP-1. Should I panic?

Is anyone collecting data on pregnancies exposed to semaglutide?

Does any of this apply to research-grade peptides?

Sources

1. Henney AE, Holmes PJ, Riley D, Heague M, Hapangama DK, Alam U, Cuthbertson DJ. Preconception incretin therapy reduces pregnancy and cardiometabolic outcomes in polycystic ovary syndrome. The Journal of Clinical Endocrinology & Metabolism, dgag423, advance article published 10 October 2026. doi:10.1210/clinem/dgag423. Abstract read at source; the full methods, results and discussion are paywalled and were not read, so no figure beyond the abstract is used here and the authors' full limitations section is not quoted. Target trial emulation using the TriNetX global federated health research network. Women aged 18 and over with PCOS or PMOS receiving pre-conception GLP-1 or dual GLP-1/GIP receptor agonist therapy, specified as liraglutide, semaglutide or tirzepatide, compared with metformin. One-to-one propensity score matching accounting for demographics, anthropometrics, comorbidities and medications; 48,126 patients per group after matching. Primary outcome a composite of adverse pregnancy outcomes over two years, covering hypertensive, glycaemic, placental, delivery, labour, foetal growth, preterm and early pregnancy complications: HR 0.38, 95% CI 0.34 to 0.41. Exploratory secondary outcomes individual pregnancy endpoints, MACE at HR 0.88, 95% CI 0.81 to 0.95, and type 2 diabetes at HR 0.31, 95% CI 0.30 to 0.33. States "Associations varied across some sensitivity analyses", "This study cannot evaluate in-pregnancy safety or efficacy", and "These findings do not support GLP-1/GIP RA use during gestation". Conclusion describes results as consistent with potential benefits of pre-conception metabolic optimisation.
2. Wegovy (semaglutide injection) prescribing information, revised 06/2026. Novo Nordisk. novo-pi.com/wegovy.pdf. Read at source. Section 8.3 Females and Males of Reproductive Potential instructs clinicians to "discontinue WEGOVY in patients at least 2 months before they plan to become pregnant", citing the potential for fetal harm and semaglutide's long half-life. Section 8.1 Pregnancy states "Discontinue WEGOVY in pregnant patients who are using it for weight reduction" and that weight loss offers no benefit to a pregnant patient and may cause fetal harm; for the MASH indication with advanced fibrosis it states the drug should be used in pregnancy only if the potential benefit justifies the potential risk to the fetus. Neither section contains contraception advice.
3. Zepbound (tirzepatide injection) prescribing information. Eli Lilly. drugs.com/pro/zepbound. Read at source via a drug-information database reproducing the label. Section 8.1 Pregnancy contains no instruction to discontinue before a planned pregnancy and no preconception interval; it instructs discontinuation when a pregnancy is recognised. Section 8.3 states that use of Zepbound may reduce the efficacy of oral hormonal contraceptives due to delayed gastric emptying, and advises patients to switch to a non-oral contraceptive method or add a barrier method for four weeks after initiation and for four weeks after each dose escalation.
4. Evidence That Semaglutide Represents an Important Tool for Treatment of Irregular Menses and Chronic Anovulation in Women with Polyendocrine Metabolic Ovarian Syndrome. Journal of Clinical Medicine 2026;15:5165. doi:10.3390/jcm15135165. Of 96 patients, 80 presented with oligomenorrhoea and anovulatory cycles before treatment. After six months of semaglutide, 42 of those 80, 52.5%, achieved normalised menstrual cycles and developed ovulatory cycles, accompanying weight loss greater than 10%. Previously cited in our guide on GLP-1 drugs and periods.
5. Peptide Decoding compound library and vendor price capture, October 11, 2026. peptidedecoding.com/prices. Our own data, counting one priced product at one seller as a listing. Semaglutide appears in 180 listings across 84 sellers, median $7.10 per mg among in-stock single vials. 25 compounds in the GLP-1 library are tracked, of which 11 are approved by a regulator somewhere and 14 are not.
6. Saxenda (liraglutide injection) prescribing information. Novo Nordisk. novo-pi.com/saxenda.pdf. Read at source. Section 12.3 gives "an elimination half-life of approximately 13 hours". Section 8.1 Pregnancy gives no preconception discontinuation interval and no instruction to stop before attempting pregnancy; it states "When a pregnancy is recognized, advise the pregnant patient of the risk to a fetus, and discontinue SAXENDA". No contraception advice appears in section 8.1. Section 8.3 was not present in the retrieved portion of the document and is not relied on here.
7. International Evidence-Based Guideline for the Assessment and Management of Polycystic Ovary Syndrome 2023. Monash University, guideline summary. monash.edu. Guideline summary read at source; the full guideline document was not retrieved, so recommendation grades beyond the labels given here are not quoted. Recommendation 4.5.1, consensus: anti-obesity medicines including GLP-1 receptor agonists "could be considered, in addition to active lifestyle intervention" at higher weight, following general population guidelines. Practice point 4.5.2: clinicians "should ensure concurrent effective contraception when pregnancy is possible" for women taking GLP-1 receptor agonists, "as pregnancy safety data are lacking". Practice point 4.5.3: gradual dose escalation is advised to reduce gastrointestinal side effects. Practice point 4.5.4: shared decision making should weigh the possible need for long-term use given the "high risk for weight regain after discontinuation". Consensus recommendation 5.9.1: anti-obesity agents should be used for reproductive outcomes "only in research settings". Evidence-based recommendation 4.3.1: metformin alone "should be considered in adults with PCOS and a BMI ≥ 25 kg/m2" for metabolic and anthropometric outcomes. The summary states "Metformin is not routinely recommended for use in pregnant women with PCOS", with section 4.11.2 noting it could be considered in some circumstances such as risk of preterm birth. Recommendation 5.3.1: "Letrozole should be the first-line pharmacological treatment for ovulation induction" in infertile anovulatory women with PCOS.
8. Ozempic (semaglutide injection) prescribing information, revised 5/2026. Novo Nordisk, via DailyMed. dailymed.nlm.nih.gov. Read at source. Section 8.3 Females and Males of Reproductive Potential: "Discontinue OZEMPIC in women at least 2 months before a planned pregnancy due to the long washout period for semaglutide." No contraception advice appears in section 8.3. Section 8.1 Pregnancy: "OZEMPIC should be used during pregnancy only if the potential benefit justifies the potential risk to the fetus." Elimination half-life of approximately 1 week, with semaglutide present in the circulation for about 5 weeks after the last dose. The Wegovy label gives 5 to 7 weeks for the same molecule at the higher weight-management doses.
9. Mounjaro (tirzepatide injection) prescribing information, revised 01/2026. Eli Lilly, via FDA Drugs@FDA. accessdata.fda.gov. Read at source. Section 8.3 gives no instruction to discontinue any stated interval before a planned pregnancy. It states that use "may reduce the efficacy of oral hormonal contraceptives due to delayed gastric emptying" and advises switching to a non-oral method or adding "a barrier method of contraception for 4 weeks after initiation and for 4 weeks after each dose escalation". Section 8.1 Pregnancy: "MOUNJARO should be used during pregnancy only if the potential benefit justifies the potential risk to the fetus." Section 12.3 gives "an elimination half-life of approximately 5 days".
10. Wegovy (semaglutide 2.4 mg) Pregnancy Registry Study: a prospective cohort study of safety outcomes in pregnancies exposed to Wegovy. Sponsor Novo Nordisk. UK Health Research Authority research summary. hra.nhs.uk. Read at source. Non-interventional post-authorisation safety study with a prospective cohort design, comparing pregnant women with obesity or overweight and a weight-related condition exposed to Wegovy during pregnancy against an unexposed reference group, described as a study "to compare maternal, foetal and infant outcomes". Target 728 pregnant women across sites in the United States, United Kingdom and Spain. Pregnancy data collected at enrolment, at week 26 and at delivery; infant data at 4 and 12 months after delivery. Data collection planned for approximately 10 years or until the target sample size is met. REC opinion dated 12 August 2025; IRAS ID 347411, REC reference 25/WS/0106. No NCT or EU PAS number is given on this page. Two ClinicalTrials.gov records titled as studies of exposure to Wegovy during pregnancy appeared in search results but their records could not be retrieved from here and nothing is quoted from them.
11. Birth defects after maternal exposure to GLP1 agonists in early pregnancy: a comparative ENTIS cohort study. EU PAS number EUPAS50643. European Medicines Agency catalogue of real-world data studies. catalogues.ema.europa.eu. Read at source. Cohort study in pregnant women aged 18 to under 46, drug class GLP-1 analogues, ATC code A10BJ. Primary objective stated as to "prospectively evaluate the risk of major birth defects", with spontaneous pregnancy losses including abortions and stillbirths, and pregnancy terminations, assessed after first-trimester exposure to a GLP-1 agonist against two reference groups. Estimated 200 subjects. Record first published 20 January 2023, last updated 19 April 2024, status planned. No results are reported in the record and none are quoted here.
12. Cesta CE, Rotem R, Bateman BT, Chodick G, Cohen JM, Furu K, Gissler M, Huybrechts KF, Kjerpeseth LJ, Leinonen MK, Pazzagli L, Zoega H, Seely EW, Patorno E, Hernández-Díaz S. Safety of GLP-1 receptor agonists and other second-line antidiabetics in early pregnancy. JAMA Internal Medicine 2024;184(2):144-152. doi:10.1001/jamainternmed.2023.6663. Published online 11 December 2023. Figures read from the published abstract via an institutional repository copy and cross-checked against a full-text copy; the JAMA Network page itself was behind a browser check and could not be opened from here. Cohort of 3,514,865 pregnancies across Nordic registers, a United States claims database and an Israeli health fund; 51,826 pregnancies in women with pregestational type 2 diabetes, of whom 15,148 were treated periconceptionally; 938 GLP-1 receptor agonist users and 5,078 insulin users. Standardised prevalence of major congenital malformations 8.3% for GLP-1 receptor agonists against 7.8% for insulin, with an adjusted relative risk of 0.95, 95% CI 0.72 to 1.26. Comparators in the same analysis: sulfonylureas 1.18 (0.94 to 1.48), DPP-4 inhibitors 0.83 (0.64 to 1.06), SGLT2 inhibitors 0.98 (0.65 to 1.46). Stated limitations include residual bias from channelling, that "confounding by obesity and cardiovascular conditions would preferentially affect GLP-1 receptor agonists", that filled prescriptions may not reflect exposure during embryogenesis, and that conditioning on live birth could introduce selection bias.
13. Liu X, Xiong B, Yang R, Wang H, Liu Y. Periconceptional use of GLP-1 receptor agonists and the risk of major congenital malformations: a systematic review and meta-analysis. Endocrine Connections 2026;15(7). doi:10.1530/ec-26-0280. Peking University Third Hospital. Gold open access. The publisher page and PubMed were both unreachable from here, so the abstract was obtained through the OpenAlex index. That index reconstructs abstract text rather than reproducing it as printed. Figures should be confirmed against the published abstract before this page goes live. As indexed: PubMed, Embase and Web of Science searched through December 2025 for trials and cohort studies of exposure within three months before conception or during pregnancy; six studies included, one prospective and five retrospective, 286,599 women, of whom 43,577 exposed and 243,022 unexposed. Primary outcome major congenital malformations, relative risk 1.02, 95% CI 0.96 to 1.08. Secondary outcomes as indexed: preterm birth 1.09 (0.80 to 1.49), large for gestational age 2.31 (0.56 to 9.44), small for gestational age 0.71 (0.39 to 1.27), stillbirth 1.16 (0.24 to 5.69); spontaneous abortion comparable in one study. Authors conclude no association with increased malformation risk but state the findings should not be taken as proof of safety given observational designs and a limited number of exposed pregnancies, and call for prospective cohorts and randomised trials. Note the large for gestational age point estimate of 2.31 sits inside a confidence interval running from 0.56 to 9.44, so "not significantly elevated" rests on imprecision rather than on a null result.
14. Özbek L, Shah EA, Al-Shiab R, İnal A, Guldan M, Afşar B, Covic A, Kanbay M. 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. Koç University. Published 26 March 2026. PubMed and the publisher were unreachable from here, so the abstract was obtained through the OpenAlex index. That index reconstructs abstract text rather than reproducing it as printed, and the figures should be confirmed before this page goes live. As indexed: PubMed/MEDLINE, Web of Science, Scopus and the Cochrane Library searched from inception to 23 September 2025; human studies of preconception, pregnancy or lactation exposure; two independent reviewers, risk of bias assessed with validated tools, narrative synthesis under PRISMA 2020; 36 studies included. Periconceptional or early-pregnancy exposure not consistently associated with increased major congenital malformations in adjusted analyses, nor with foetal growth restriction, stillbirth or neonatal mortality, against insulin-treated or disease-matched controls. Maternal outcomes including gestational diabetes, hypertensive disorders, preterm birth and gestational weight gain described as heterogeneous without a reproducible safety signal. Lactation data sparse, with one pharmacokinetic study reporting no detectable semaglutide transfer into human milk. Conclusion states data on continued use throughout gestation remain limited.

Citing this page. Peptide Decoding. Should You Take a GLP-1 Before Trying to Conceive With PCOS? Study figures from the Journal of Clinical Endocrinology & Metabolism of 10 October 2026; exposure evidence from JAMA Internal Medicine 2024, Endocrine Connections 2026 and Diabetes, Obesity and Metabolism 2026; discontinuation and birth control instructions from the Wegovy, Ozempic, Zepbound and Mounjaro labels; registry details from the UK Health Research Authority and the European Medicines Agency study catalogue; listing data from the October 11, 2026 capture. https://peptidedecoding.com/guides/glp1-preconception-pcos