GLP-1 and Pregnancy in 2026: Washout Windows, the Pill Trap, and What OB-GYNs Now Say

A 32-year-old on Wegovy walks into her OB-GYN in April 2026 with a half-finished pen in her purse and one question. When can we start trying? Three weeks later, a 28-year-old on Mounjaro shows up at the same clinic six weeks pregnant, holding a positive test in one hand and the strip of birth-control pills she had taken faithfully for two years in the other. Different stories, same gap. Neither prescriber walked through the part that mattered most.

The headline guidance from Novo Nordisk, Eli Lilly, the FDA, EMA, Korea's MFDS, Japan's PMDA, China's NMPA, and the Gulf authorities lines up neatly in 2026: GLP-1 medications are contraindicated in pregnancy. Stop before conception. Use effective contraception. What the label does not spell out, and what most primary-care visits skip, is how long "before" actually means, why your pill might quietly fail in month one of titration, and what to do if you find out you are pregnant on a refill from last week.

Below is the version of that conversation a careful OB-GYN would have with you. As of April 2026, every regulator position, label change, and registry note here is current.

Why pregnancy plus GLP-1 became the 2025 conversation

For most of the GLP-1 era, roughly 2017 to 2023, pregnancy was a niche question. Early semaglutide and liraglutide trials enrolled mostly older patients with type 2 diabetes. Tirzepatide shifted the demographic. Wegovy shifted it again. By 2024 a meaningful share of GLP-1 users were women in their 20s and 30s, many with PCOS, many on the pill, many not aggressively planning a pregnancy and not aggressively preventing one either.

Two things hit at once. The press started writing about "Ozempic babies," the wave of unexpected pregnancies in women who had been told they could not conceive. Reuters, the New York Times, Time, and JAMA commentary all ran versions of the story through 2024 and 2025. At the same time, Eli Lilly rewrote the tirzepatide label in 2023 to warn that the drug can reduce the absorption of oral contraceptives. The recommendation: switch to a non-oral method or add a barrier for the first 4 weeks after starting tirzepatide and for 4 weeks after each dose escalation. That single label change is the largest piece of women-specific GLP-1 safety news to date. Most patients still have not heard about it.

Where things sit in April 2026: the regulator guidance is unanimous and conservative. The clinical reality is messier. The inadvertent-exposure data is reassuring but underpowered. And the practical questions, washout, contraception, postpartum restart, all have answers, but they are answers you almost certainly have to ask for.

Washout windows by molecule, with the half-life math underneath

OB-GYNs lean on a shorthand called "five half-lives." It is the period needed for a drug's blood concentration to drop to roughly 3% of peak, close enough to functional zero. Half-lives across the GLP-1 class differ by an order of magnitude, which is why the discontinue-before-conception window is not the same molecule to molecule.

Semaglutide is the long-tail molecule of the class. A weekly injection on Saturday still has clinically meaningful drug onboard the next Saturday, and the next, and the next. Two months is the manufacturer guidance, and most US OB-GYNs round up to a clean 60 days from the last dose to the start of trying. Tirzepatide clears faster, so the window is shorter. Four weeks is the modal counsel, with some practitioners stretching to 6 weeks for an extra cushion. Liraglutide clears in days, but the precautionary window stays at about 4 weeks because the manufacturer wants conservative real-world behavior.

Orforglipron is the new variable. Foundayo, Lilly's oral GLP-1, cleared the FDA on April 1, 2026 with no human pregnancy data and an animal-study signal at clinically relevant exposures. "Avoid in pregnancy" is the only honest answer right now. Until registry data accumulates over the next few years, Foundayo gets the most conservative handling of the four.

The half-life math also explains why "I missed one dose" is not a washout. Skipping a single weekly semaglutide injection drops the blood level by maybe 10%. Stopping for a month leaves you at about 1% of peak, close to but not at the manufacturer-recommended baseline. The two-month rule is built around the math, not around overcaution.

What the regulators say in 2026, in plain language

The labeling language is consistent across major agencies, with small differences in tone.

There is no major market in 2026 where a regulator endorses GLP-1 use during pregnancy. The unanimity is not because the risk is proven. It is because the risk has not been ruled out, and the obvious test, a placebo-controlled trial in pregnant women, is one nobody is going to run. The default is conservative, and it should be.

Labels say "contraindicated." The clinic conversation rarely says "two months." If your prescriber handed you a Wegovy starter pack last spring without naming the discontinue-before-pregnancy window, that is the gap to close at your next visit, not at the appointment when you tell them you are trying.

The oral-contraceptive failure that almost nobody warns you about

This is the underreported piece. If there is one section in this post worth forwarding to a friend on a GLP-1, it is this one.

GLP-1s slow gastric emptying. That is a feature, not a bug. Slower emptying is part of what produces satiety and the "food noise silence" so many users describe. The side effect is that anything you swallow at the same time, including an oral contraceptive pill, hits your bloodstream more slowly and at a lower peak concentration. For most oral medications that does not matter. For estrogen-progestin combination pills, where the contraceptive effect depends on hitting and holding a steady hormonal level, it matters a lot.

Eli Lilly took the lead on this in 2023. Mounjaro and Zepbound labels were updated to recommend that patients on combined oral contraceptives switch to a non-oral method (the patch, the ring, the IUD, the implant, the shot) or add a barrier method, for the first 4 weeks after starting tirzepatide and for 4 weeks after each dose escalation. The dose-escalation step is the part most often missed. If you were stable on 5 mg and your prescriber bumped you to 7.5 mg in March, your pill needs the same 4 weeks of backup all over again starting on the day of that bump. Same for the next step. Same for the one after that.

Novo Nordisk's labels for semaglutide are softer. The Wegovy and Ozempic prescribing information notes the potential for delayed absorption of co-administered oral medications and recommends caution. Some clinicians apply the same 4-week barrier-method rule from the tirzepatide label to semaglutide patients, especially during initiation and titration. Others do not. The data on semaglutide-specific contraceptive failure is thinner than for tirzepatide, but the underlying mechanism of delayed gastric emptying is identical. A meaningful share of the "Ozempic baby" stories trace back to women who had been on the pill for years without any issue.

Liraglutide and orforglipron sit in the same category. GI effects come on more slowly with liraglutide and the manufacturer guidance is less prescriptive, but the physiology is the same. If you are on an oral contraceptive and starting any GLP-1, the safe default is simple: assume your pill is no longer as protective as it was, and add a barrier or switch to a non-oral method until your prescriber tells you otherwise.

A few practical notes:

• The interaction is with oral contraceptives specifically. The patch, the ring, the depo-provera shot, the hormonal IUD, the copper IUD, and the implant are not affected by GLP-1 gastric emptying.
• Emergency contraception (Plan B, ella) is also oral. The same delayed-absorption mechanism may apply. If you need EC while on a GLP-1, talk to a pharmacist about whether the copper IUD as EC is a better choice for you.
• The 4-week window resets at every dose escalation on tirzepatide. Most titration ladders run 4 to 6 escalation steps from start to maintenance. That can mean 4 to 6 separate 4-week windows of reduced pill protection over the course of titration.

The "Ozempic baby" phenomenon, calibrated

Two mechanisms drive the fertility-uptick reports. They affect different populations, but the press tends to blur them together.

The first is the PCOS pathway. Polycystic ovary syndrome affects 6% to 12% of reproductive-age women globally, per the WHO. PCOS is associated with anovulation, insulin resistance, and weight gain. The OB-GYN literature has shown for decades that weight loss of 5% to 10% can restore ovulation in many PCOS patients. GLP-1s produce that magnitude of weight loss reliably. Wegovy in the STEP 1 trial showed average body-weight loss of around 14.9% at 68 weeks. Tirzepatide in SURMOUNT-1 showed roughly 20.9% at the highest dose at 72 weeks. For a PCOS patient who had not ovulated regularly in years, a 15% reduction is enough to restart cycles in a meaningful share of cases. Restored ovulation plus an inattentive contraceptive plan equals an unexpected pregnancy.

The second is the contraceptive-absorption pathway above. For women without PCOS, on the pill, in their late 20s and early 30s, who took their pill on time every day for years, the only thing that changed was starting the GLP-1. The pill failed because the absorption changed. The pregnancy was not a sudden fertility surge. It was a pharmacokinetic interaction.

As of April 2026 there is no published large registry that quantifies the conception-rate increase. Press accounts are anecdotal and the case series in the literature are small. What the OB-GYN community broadly agrees on: both mechanisms are real, both are underreported in patient counseling, and the fertility uptick is something to plan around rather than be surprised by.

If you have PCOS and you start a GLP-1, the working assumption should be that your fertility may improve as your weight comes down. If you are not actively trying, that improvement needs an active contraceptive plan, not the lower-stakes plan you had before.

What the inadvertent-exposure data shows so far

The honest answer: not enough yet, but what we have is reassuring.

Several case series published in 2024 and 2025, most with sample sizes in the hundreds rather than thousands, looked at pregnancy outcomes in women exposed to semaglutide or tirzepatide in early pregnancy, usually in the first trimester before they knew they were pregnant. Headline findings have been consistent across the series: no clear teratogenic signal, miscarriage rates roughly in line with population baselines, no excess of major congenital anomalies in the small samples reported.

That is good news. It is not enough news. The case series are underpowered to rule out small-effect harm. A drug could double the rate of a rare malformation and the case series so far would not see it. This is why every regulator and both manufacturers continue to recommend discontinuation before pregnancy and have not softened the contraindication language.

Two registries are worth knowing about if you find yourself in this situation:

• Lilly maintains a tirzepatide pregnancy registry, open since 2023. US patients can self-report exposure. The data is being aggregated to inform future labeling.
• Novo Nordisk maintains a semaglutide pregnancy registry on the same model.

Beyond the registries, free phone-counseling services exist for inadvertent-exposure questions. In the US, the CDC-funded MotherToBaby service offers free counseling on medication exposures during pregnancy and lactation. In the UK, the BUMPS service (UK Teratology Information Service for the public) offers a similar service. Both are staffed by clinicians who specialize in this question and who track the literature in real time.

If you took a GLP-1 dose in the first trimester before you knew you were pregnant, the data so far does not support panicking. It does support stopping the drug now, telling your OB-GYN at your next visit, and self-reporting to the manufacturer registry if you are in the US. The 2024 to 2025 case series are reassuring enough that an emergency call to your prescriber on a Saturday night is not warranted.

Postpartum and breastfeeding, the part nobody plans for

The conversation usually ends at "stop before pregnancy." It should not. Most women who go through pregnancy on a GLP-1 hiatus then face the question of when, or whether, to restart.

Labels are unanimous on lactation: not recommended during breastfeeding for any of the four molecules above. Human data on transfer to breast milk is insufficient to set a quantitative safety threshold, and the same conservative-default logic that applies to pregnancy applies here. If you are exclusively breastfeeding, the working assumption is that you are not back on the GLP-1 yet.

In practice, OB-GYNs and endocrinologists vary in how they handle the postpartum restart for women who are not breastfeeding or who have stopped. A common cautious approach is to wait at least 6 to 8 weeks postpartum before restarting, partly to let the body settle through immediate postpartum recovery, partly to make sure the patient is not in an early second pregnancy. Some practitioners wait longer if the patient is contemplating another pregnancy in the next 12 months. Others restart sooner if the metabolic indication is strong (severe insulin resistance, rapid postpartum weight gain in a patient with a history of class III obesity).

The bigger conversation, and the one most patients are not having, is what happens to weight regain during the GLP-1 hiatus. STEP 4 data published several years ago showed that women who stopped semaglutide regained roughly two-thirds of their lost weight within a year. The pregnancy hiatus is, on average, going to run 12 to 18 months from washout to a sensible postpartum restart. Plan for it as a real intervention period: strength training, structured nutrition support, lactation-aware caloric planning if you are nursing. Not a passive pause.

How this looks in your market

The regulator list above is uniform. Patient access patterns are not.

In the US, most OB-GYNs in 2026 are now actively asking GLP-1 patients about pregnancy intent at every visit. Pre-conception clinics increasingly screen for GLP-1 use and offer a washout-and-contraception package as routine. Insurance coverage is starting to flex around pregnancy intent. Some commercial plans pause Wegovy or Zepbound refills if pregnancy is disclosed, partly out of conservatism and partly because the manufacturer label says discontinue. If your refill gets denied in this scenario, the cleaner workaround than fighting the PA is to align with your OB-GYN on the washout window and let the refill lapse intentionally.

In the EU and UK, the contraindication is enforced more uniformly through the EMA-aligned SmPC. NHS-channel Wegovy in the UK, and EU national-system access in France, Germany, and the Netherlands, includes patient-information leaflets in local languages that name the pregnancy contraindication explicitly. The private slimming-clinic channel is where the counseling gaps tend to concentrate. If you got your pen from a private clinic in London, Berlin, or Madrid, the depth of pregnancy counseling at the prescribing visit is the variable to ask about directly.

In Korea, GLP-1s remain prescription-only at a 약국, with overseas direct purchase (직구) banned since October 2024. The MFDS contraindication is on the label in Korean. The system-level catch is that the average Korean prescriber visit is short, and the pregnancy conversation often does not happen unless the patient initiates it. If you are on Wegovy or Mounjaro in Korea and contemplating a pregnancy in the next 18 months, raise it at the next visit and ask for the washout timeline in writing.

In Japan, the cosmetic-clinic semaglutide channel is the watch-list segment. PMDA-registered endocrinology and obesity clinics handle the pregnancy conversation seriously. The cosmetic-import pathway often does not. If your semaglutide source is a Tokyo or Osaka cosmetic clinic, the pregnancy counseling baseline you should expect is lower, and you should treat the planning conversation as your job, not the clinic's.

In China mainland, 诺和盈 (Wegovy) launched in 2025 with the pregnancy contraindication printed on the box in Chinese. Hospital pharmacy is the safest channel. Community pharmacy varies by city in how thoroughly the pharmacist will walk through the contraindication at dispense.

In Saudi Arabia and the UAE, the SFDA and MOHAP follow the manufacturer SmPC. The licensed-pharmacy channel handles the contraindication consistently. The cash-pay private-clinic segment, especially in cosmetic weight-loss positioning, is where the counseling variability concentrates.

In Hong Kong and Taiwan, regulator alignment with EMA-style SmPC is solid. Drug Office and TFDA channels enforce the contraindication. As elsewhere, the failure mode is the unlicensed beauty-centre or cosmetic-clinic dispense path, which should not be your source for any GLP-1 in any market.

What to verify before your next prescription or refill

A short, practical checklist for any woman of reproductive age starting or continuing a GLP-1 in 2026:

1. Confirm the molecule on your pen and the half-life that goes with it. Semaglutide: 7 days. Tirzepatide: 5 days. Liraglutide: 13 hours. Orforglipron: 24 to 30 hours.
2. Map the discontinue-before-conception window to a calendar date. Semaglutide: 60 days from the last dose. Tirzepatide: 30 days. Liraglutide: 28 days. Orforglipron: avoid altogether.
3. If you are on an oral contraceptive, identify whether you started the GLP-1 in the past 4 weeks or had a dose escalation in the past 4 weeks. If yes, you need a barrier method or a non-oral alternative right now.
4. If you are not actively trying to conceive but also not aggressively preventing, re-assess. Both pathways above push the unplanned-pregnancy probability higher than the version of you a year ago might assume.
5. Confirm your prescriber knows your pregnancy intent. If they have not asked, tell them. If you have a planning horizon of 12 to 18 months, ask for the washout calendar in writing.
6. Save the registry contact. Lilly pregnancy registry for tirzepatide, Novo registry for semaglutide. If you need exposure counseling in the US, MotherToBaby. In the UK, BUMPS.
7. Know your postpartum plan before you stop, not after delivery. Will you breastfeed? For how long? Do you want to restart the GLP-1 at 6 weeks, 12 weeks, 6 months, or after weaning?

Questions to bring to your OB-GYN or endocrinologist

If you have a single 20-minute appointment to use, these are the questions that change the conversation.

• "Given the molecule I am on and my pregnancy planning horizon, what is the calendar washout date you would want me to hit before we start trying?" Asking for a date, not a duration, forces specificity.
• "Am I on an oral contraceptive that could be affected by gastric-emptying delay, and if so, do you want me on a barrier method or a non-oral method during initiation and after each dose escalation?" Most prescribers know the tirzepatide labeling. Many have not extended the same precaution to semaglutide patients on the pill.
• "If I find out I am pregnant on a refill I just took, what should I do in the first 24 hours?" The answer should be specific. Stop the drug, schedule an early OB visit, self-report to the manufacturer registry if appropriate, contact MotherToBaby or the local equivalent.
• "What is your postpartum restart plan if I am breastfeeding versus not breastfeeding?" Different prescribers handle this differently. Get yours on the record.
• "Given the STEP 4 weight-regain data, how do you want to manage the hiatus period of 12 to 18 months when I am off the drug for pregnancy and lactation?" This question surfaces whether your prescriber has actually thought about the hiatus as a metabolic intervention period or as a passive pause.
• "Do you have a written patient handout for women on GLP-1s who are planning pregnancies?" Some practices do. If yours does not, a printed summary of the answers above is worth requesting in writing.

What the next two years are likely to shift

Three things look set to move between now and 2028.

Inadvertent-exposure registry data will mature. The 2024 to 2025 case series will be supplanted by larger cohort analyses. If the reassuring early signal holds, labeling tone may soften slightly, though "contraindicated" is unlikely to become "permitted." The likelier change is clearer, less alarming language for women who discover an inadvertent first-trimester exposure.

Contraceptive-interaction guidance will probably standardize across the GLP-1 class. Right now the Lilly tirzepatide label is the most prescriptive and Novo's semaglutide labels are vaguer. As post-marketing data on semaglutide and oral contraceptives comes in, the FDA is the agency to watch for harmonization. If you are on Wegovy or Ozempic and on the pill, watch for label updates over the next 18 to 24 months.

Pre-conception counseling is heading toward becoming a standard part of the GLP-1 prescribing visit, not an add-on. Pre-conception clinics, OB-GYN groups, and the major endocrinology societies are all moving in this direction. By 2027 or 2028, being handed a GLP-1 starter pack without any pregnancy counseling should be the exception, not the rule.

For now, in April 2026, the gap is real, the questions above are the ones to ask, and most of the planning is on the patient. If you are on Wegovy and trying for a baby in 2027, your washout window starts the day you decide. If you are on Mounjaro and on the pill, your barrier method starts today. If you are on Foundayo and pregnancy is anywhere in the picture, the conversation with your prescriber is overdue.

For context on how the broader GLP-1 access landscape is evolving for women, our piece on who qualifies for GLP-1 in 2026 walks the eligibility criteria. For the verification mechanics that matter when supply tightens, and tightening supply is when grey-market substitution accelerates, the counterfeit-Wegovy global alert breakdown is the companion piece. And if you are weighing the Foundayo question specifically as the new oral GLP-1 with no pregnancy data, the Foundayo orforglipron guide covers the molecule in depth.

The next visit is the right time. Bring the questions. Bring the calendar. The decisions on the table are yours, but the inputs are too easy to leave on the prescriber's side of the desk by default.

Start managing your GLP-1 with Blueshot

AI coaching, injection scheduling, and weight tracking in one app