GLP-1 and Pregnancy Safety: What the Evidence Actually Says

Here is the conversation almost nobody planned for: a wave of people lose 20, 40, 60 pounds on a GLP-1 drug, their cycles regularize for the first time in years, and a pregnancy test turns positive while a pen of semaglutide is still in the fridge. Clinicians even coined a nickname for it: “Ozempic babies.” So what actually happens to a pregnancy exposed to these medications, and what does the evidence say you should do?

Are GLP-1 drugs safe to take during pregnancy?

No. Semaglutide and tirzepatide are not recommended in pregnancy. Their U.S. labels advise against use because animal studies showed fetal harm, and human data remain limited. Guidance is to stop GLP-1 drugs at least two months before trying to conceive. Early reassuring human data exist for accidental exposure, but these drugs are not approved for use while pregnant.

Hold two ideas at once as you read the rest of this. If you took one of these drugs before you knew you were pregnant, the accidental-exposure data so far are more reassuring than frightening. And the recommendation to avoid them during a planned pregnancy still stands, because “not enough evidence to clear them” is different from “shown to be safe.” Both statements are true, and the rest of this page walks through why.

Why are GLP-1 medications not recommended in pregnancy?

The hesitation is not a reflex. It comes from a real gap in evidence plus a clear animal-data signal. In pregnant rats given semaglutide during organogenesis, researchers documented embryofetal mortality, structural abnormalities, and growth changes at maternal exposures below the maximum recommended human dose (Ozempic prescribing information, Novo Nordisk). That is the kind of finding that makes regulators pump the brakes.

On the human side, the honest answer is that we are still early. The FDA classifies these drugs as ones where a risk to a developing pregnancy cannot be ruled out, which is regulatory language for “we do not have enough data to clear them.” There is also a mechanistic worry that is easy to overlook: these medications work partly by suppressing appetite and slowing weight gain. Pregnancy is one of the few times in adult life when steady weight gain and robust nutrition are the goal, not the enemy. A drug engineered to do the opposite is, on its face, working against the biology of the moment.

Put those two concerns side by side and the caution makes sense. One is a direct signal from animal studies at relevant doses. The other is a logical mismatch between how the drug works and what a pregnancy needs. Neither proves harm in humans, but together they are more than enough reason for regulators and clinicians to say “not during pregnancy” until the human evidence is far larger and clearer than it is today.

What does the human data actually show about birth defects?

This is where the story gets more nuanced than the scary headlines suggest. The accidental-exposure data so far are cautiously reassuring, not alarming.

A multicenter prospective cohort study run across six European Teratology Information Services followed 168 pregnancies exposed to GLP-1 receptor agonists in the first trimester, most commonly liraglutide and semaglutide, with about 70% of users taking the drug for weight loss. The rate of major birth defects was 2.6% (3 of 117 pregnancies with known outcomes), statistically indistinguishable from the diabetes comparison group at 2.3%. The authors explicitly offered “reassurance in cases of inadvertent exposure” (Dao et al., 2024, PMC).

Broader reviews echo this. A 2025 systematic review in the European Journal of Obstetrics and Gynecology pooled studies covering more than a thousand semaglutide-exposed pregnancies and found no clear association with birth defects, while flagging that the evidence is limited by small samples, variability, and bias risk (EJOG, 2025). A 2025 first-trimester case series reached a similar verdict (Morton and He, 2025).

Two things can be true at once. If you took a GLP-1 before you knew you were pregnant, the data do not point to a high risk of malformation. That should lower the panic. At the same time, “no clear signal in a few hundred to a thousand pregnancies” is not the same as “proven safe across populations.” Rare risks need much larger numbers to detect.

How to read these studies without over-reading them

It helps to understand why researchers keep hedging even as the numbers look calm. A background rate of major birth defects sits around 3 to 5% in the general population, so a 2.6% figure in a small exposed group is genuinely reassuring on its face. But a few hundred pregnancies simply cannot rule out a rare defect that shows up in, say, one in a few thousand exposures. That is not a knock on the studies, it is the nature of statistics: you need very large numbers to detect very rare events. The honest interpretation is the one the authors themselves take, that inadvertent exposure looks reassuring while planned use stays off the table until the evidence base is much bigger. Anyone claiming these drugs are “proven safe” in pregnancy is reading past what the data can support.

Are there risks beyond birth defects?

Yes, and this is the part that gets lost. Birth defects are only one outcome. Newer work points to other concerns. A 2026 analysis reported that mothers with overweight or obesity exposed to semaglutide before and during pregnancy had higher risks of excessive gestational weight gain, gestational diabetes, excessive fetal growth, and cesarean delivery compared with nonusers (Medscape, 2026).

Some of this likely reflects a rebound effect. Stop a powerful appetite suppressant and weight, appetite, and blood sugar can swing back, sometimes sharply, at the exact moment a pregnancy begins. One reported pattern linked semaglutide discontinuation to fetal macrosomia (an unusually large baby) and neonatal low blood sugar. Miscarriage rates in the exposure cohorts have generally tracked with what you would expect in diabetes and obesity populations rather than spiking, with one cohort reporting roughly a 23% loss rate (Dao et al., 2024). The takeaway: the conversation should not be just “did the drug cause a defect” but “how is the whole metabolic picture managed across the transition off the drug.”

What happens metabolically when you stop a GLP-1 for pregnancy?

This is the underappreciated half of the story. A GLP-1 is not just a weight tool, it is holding down appetite, blood sugar, and often insulin resistance all at once. Take that hand off the controls and several things can move together. Appetite tends to rebound, sometimes strongly. Weight can start climbing again. And blood sugar can drift upward in someone whose glucose was only well behaved because the drug was doing the work. Landing all of that at the start of a pregnancy, when the body is already shifting its metabolism, is exactly why some of the adverse outcomes in the data cluster around discontinuation rather than around the drug itself.

The practical implication is that stopping a GLP-1 for pregnancy is a managed transition, not a light switch. This is where knowing your baseline numbers pays off: blood sugar and A1c, thyroid function, and the metabolic markers that shape a healthy pregnancy. If you have those measured before you come off the drug and again as you transition, you and your clinician can catch a rebound early instead of discovering it at a routine prenatal visit weeks later. Going into a pregnancy with your metabolic picture actually mapped, rather than guessed at, is one of the most useful things a person coming off one of these drugs can do.

A realistic scenario

Picture someone with PCOS who lost a significant amount of weight on semaglutide over a year. Her periods, absent for ages, became regular, and she and her partner decided to try for a baby. The tempting move is to stop the injection that week and start trying. The better move, and the one that avoids the rebound trap, looks different. She talks to her prescriber first, gets her blood sugar, A1c, and thyroid checked while still on the drug, then stops roughly two months ahead so it clears her system. During those two months she and her clinician watch for appetite and glucose creeping back and put a plan in place before conception, not after. When she does conceive, she is starting the pregnancy with a known metabolic baseline and a strategy already running, rather than scrambling to react to numbers that quietly drifted the moment the drug came off. Same goal, far less turbulence, because the transition was sequenced instead of abrupt.

The through-line in every version of this story is lead time. Almost everything that goes wrong in the surprise-pregnancy accounts traces back to no runway: the drug still on board, no baseline numbers, no bridge plan, and a rebound arriving unannounced. None of that is exotic to prevent. It just requires deciding in advance rather than at the moment a test turns positive.

When should you stop a GLP-1 before trying to conceive?

The label answer is clean: stop at least two months before a planned pregnancy. Both the Ozempic prescribing information and the MotherToBaby fact sheet recommend halting semaglutide about two months (eight weeks) before trying to conceive, because these molecules have long half-lives and clear the body slowly (MotherToBaby). Semaglutide has a half-life of roughly a week; tirzepatide is in a similar range. It takes multiple half-lives to fully wash out, which is why “stop the day you get a positive test” is not the strategy clinicians want.

The practical move if you are on a GLP-1 and pregnancy is on the table: bring it up with your prescriber before you start trying, not after. That buys time to taper, plan, and put a bridge in place for blood sugar or weight management. If you are reading this because you already conceived while on the drug, the message from the data is to stop now, loop in your clinician promptly, and resist the urge to spiral. Inadvertent early exposure has not shown a strong malformation signal in the studies above.

How do you plan a pregnancy if you are on a GLP-1?

A little sequencing turns a stressful situation into a manageable one.

1. Talk to your prescriber before you start trying, not after a positive test. This is the single most important step, because everything else depends on lead time.
2. Get your baseline numbers on record while still on the drug: blood sugar and A1c, thyroid, and the broader metabolic markers. You want a “before” picture to compare against as you come off.
3. Stop the GLP-1 about two months before trying, per label and MotherToBaby guidance, so it clears your system before conception.
4. Put a bridge plan in place for whatever the drug was managing. For glucose that may mean options with long pregnancy track records like insulin or, in many cases, metformin; for weight it means shifting the goal from loss to appropriate, monitored gestational weight gain.
5. Use reliable contraception until you are intentionally trying, because these drugs can restore fertility faster than people expect, which is the whole reason surprise pregnancies happen in the first place.

Done in that order, coming off a GLP-1 to conceive becomes a planned handoff rather than an abrupt stop with a rebound waiting on the other side.

What can you use instead during pregnancy?

For people who need glucose control, established options with long pregnancy track records exist, including insulin and, in many cases, metformin, alongside nutrition and activity changes. For weight, the strategy shifts entirely: pregnancy is not the window for active weight loss, it is the window for appropriate, monitored gestational weight gain. The goal moves from “lose” to “support a healthy pregnancy,” which is a different playbook than the one a GLP-1 was prescribed for.

None of these substitutions are do-it-yourself decisions. Which glucose plan fits you depends on your history, and the switch should be mapped out with your clinician before you conceive so there is no gap in control. If you want the broader mechanism behind how these peptide-based drugs work in the body, our guide to peptides covers the biology in plain language.

What about GLP-1 drugs while breastfeeding?

The data here are thin but lean cautiously optimistic. Semaglutide is expected to pass into breast milk only in small amounts, and one study of eight nursing women did not detect it in breast milk after weekly injections of up to 1.0 mg (MotherToBaby). The oral tablet form (Rybelsus) carries a more cautious label due to lack of published data. As with pregnancy, this is a per-person decision to make with your clinician, weighing your health needs against limited evidence.

One nuance worth raising with your clinician is timing. The postpartum period, especially while nursing, has its own nutritional demands, and restarting a strong appetite suppressant sits uneasily against those demands for some people. That does not make it automatically wrong, it makes it a conversation, and one that should weigh your specific health needs against the thin evidence base rather than defaulting either way.

Frequently asked questions

Can a GLP-1 drug cause miscarriage?

Current cohort data do not show a clear spike in miscarriage from GLP-1 exposure. One multicenter study reported about a 23% pregnancy loss rate, in line with what is expected in diabetes and obesity populations rather than above it (Dao et al., 2024). Evidence is still limited, so this is not a settled question.

I got pregnant while on Ozempic. Should I panic?

No. Stop the medication and contact your clinician promptly. Studies of first-trimester accidental exposure have not found a strong increase in major birth defects, which is genuinely reassuring, though longer-term and larger data are still being collected.

How long before pregnancy should I stop tirzepatide or semaglutide?

At least two months before trying to conceive, per the drug labels and MotherToBaby guidance, because these drugs clear the body slowly. Plan this with your prescriber in advance so you have a bridge plan for blood sugar or weight.

Are “Ozempic babies” a documented phenomenon?

The term describes unplanned pregnancies in people on GLP-1 drugs, often after weight loss restores ovulation in those with conditions like PCOS. It is a real pattern clinicians report, and it is a key reason contraception counseling matters when starting these medications.

Is metformin safer than a GLP-1 in pregnancy?

Metformin has a far longer pregnancy track record and is used in pregnancy in many cases, whereas GLP-1 drugs are not recommended. Any switch should be decided with your clinician based on your specific situation.

Will stopping the drug make my blood sugar or weight rebound during pregnancy?

It can, and that is one of the main reasons to plan the transition rather than stop abruptly. Appetite, weight, and blood sugar can swing back once the drug is gone, and some of the adverse outcomes in the data cluster around discontinuation. A bridge plan and baseline numbers agreed with your clinician are how you stay ahead of a rebound.

Does the two-month rule apply to tirzepatide too?

The clearest labeled guidance is for semaglutide, but tirzepatide has a similar long half-life and clears the body slowly as well, so the same “stop well ahead of conception” logic applies. Confirm the exact timing for your specific drug with your prescriber rather than assuming.

This article is for general information and is not medical advice. GLP-1 medications and pregnancy decisions are individual and high-stakes. Always consult a qualified clinician before starting, stopping, or changing any medication while pregnant or planning a pregnancy.