Tirzepatide Has a Contraception Instruction Semaglutide Doesn't — and the Timing Makes It Almost Continuous

The sentence most women never hear

Buried in Zepbound's prescribing information, in a section about drug interactions rather than anywhere a patient would think to look, is an instruction that has no equivalent on any semaglutide label.

It tells patients using oral hormonal contraceptives to switch to a non-oral method, or add a barrier method of contraception, for four weeks after starting the drug — and for four weeks after each dose escalation.

Read that second clause against the titration schedule and the arithmetic does something uncomfortable. Zepbound's dose steps are spaced four weeks apart. A woman who titrates on schedule finishes one four-week window and begins the next one on the same day.

Which means that for most of the months it takes to reach a maintenance dose, a woman on the pill is being told — by the label, if not by anyone who spoke to her — that her contraception needs backing up. Not once. Continuously.

Hormonal contraceptives that are not taken by mouth are explicitly not affected. The implant, the injection, the hormonal IUD, the patch, the ring: the label says these should not be affected, and that is the cleanest way out of the problem for a woman who does not want to think about it every month.

Why this is a Zepbound problem and not an Ozempic one

The two drugs are not interchangeable here, and the difference is mechanical.

Tirzepatide delays gastric emptying enough, particularly after a dose increase, that the absorption of a pill swallowed at the same time can be affected. Semaglutide's labels carry no comparable contraceptive instruction.

So a woman moving from Wegovy to Zepbound — a common enough switch — inherits a precaution she did not have before, in a document she has probably already read once and will not read again.

The story this page actually exists for

There is a phenomenon that the internet named "Ozempic babies" before the medical literature caught up with it, and a 2025 review in the American Journal of Obstetrics and Gynecology describes it without the nickname: because of the weight loss these drugs produce, many women previously diagnosed with oligomenorrhea — infrequent or absent periods — and unable to conceive have experienced unplanned pregnancies while taking the medications¹.

That is worth restating plainly, because it inverts how most women arrive at this question. The risk is not principally that a woman decides to get pregnant and wonders about timing. It is that fertility returns to someone who had been told, sometimes for years, that it would not — while she is taking a drug whose label says to stop it the moment pregnancy is recognized.

Combine that with a contraception instruction filed under drug interactions, and the gap is not really a medical one. It is a communication one.

What is actually known about exposure

Here the review is unusually careful, and compressing it would misrepresent it.

The animal data are not reassuring. In studies of small animals exposed to GLP-1 receptor agonists in pregnancy, there has been evidence of adverse outcomes in offspring including decreased fetal growth, skeletal and visceral anomalies, and embryonic death¹.

The human data are more reassuring, and thinner than they look. There are no prospective studies in humans. Case reports, cohort studies and population-based studies have not shown a pattern of congenital anomalies in infants¹.

The largest study has a hole in the middle of it. A large observational population-based cohort examined 938 pregnancies affected by type 2 diabetes, comparing outcomes after periconceptional exposure to GLP-1 receptor agonists against insulin, and concluded there was no significantly increased risk of major congenital malformations. But it carried no information on maternal glycemic control. As the reviewers point out, diabetic embryopathy is directly related to the degree of maternal hyperglycemia rather than to the diagnosis of diabetes itself — so that conclusion cannot properly be drawn without knowing it¹.

The review's own bottom line is the one to carry: patients should be counseled that there is not enough evidence to predict any adverse effects, or the lack thereof¹.

That sentence refuses to alarm and refuses to reassure, and it is the most accurate thing anyone can currently tell you.

What the label instructs, in order

Zepbound's position is unambiguous even where the evidence is not.

Weight loss offers no benefit to a pregnant patient and may cause fetal harm. Weight loss is not recommended during pregnancy, and the instruction is to discontinue the drug when a pregnancy is recognized.

Available data are insufficient. The label states directly that data in pregnant patients are insufficient to evaluate a drug-related risk of major birth defects, miscarriage, or other adverse maternal or fetal outcomes.

And the registry has not reported. The label's wording here is in the future tense, and it is the clearest single statement of where this field stands: "There will be a pregnancy exposure registry that monitors pregnancy outcomes in women exposed to ZEPBOUND." Patients and providers are encouraged to contact Eli Lilly at 1-800-LillyRx.

A registry described in the future tense is not a data source. It is an acknowledgment that the data does not exist yet — and, if you are exposed, an invitation to become part of the answer for whoever asks this question in five years.

Planning a pregnancy deliberately

For a woman who wants to conceive rather than one worrying about an accident, the practical question is how long to leave between stopping and trying.

A 2025 review of preconception use of GLP-1 and GLP-1/GIP receptor agonists for obesity treatment covers this ground², and the honest summary is that recommended washout intervals are derived from how long the drug takes to clear rather than from outcome data showing a particular gap is safe. There is no trial establishing an interval. Your prescriber's number will be a reasoned one, not a measured one, and it is worth asking which.

The other half of that conversation gets skipped more often than it should: coming off the drug before conceiving means the appetite effect ends before the pregnancy begins, and the weight-regain pattern is well documented. That is a genuinely difficult trade and it deserves discussing as one rather than being presented as a simple instruction.

The part nobody schedules time for

A qualitative study in JAMA Network Open interviewed 30 people across 15 US states — 19 of them women, mean age 54 — about their experience on these drugs. Among its findings, participants reported that information and clinical support were essential but highly variable³.

That is the finding this whole page sits on top of. The contraception instruction exists. It is specific, it is actionable, and it is printed in the document that comes with the drug. Whether a particular woman hears it appears to depend on who prescribed it to her and how long the appointment was.

If nobody has raised it with you, that is not unusual, and it is not a reason to assume it does not apply.