glp1ledger

Ozempic and Birth Control: Why the Label Carries No Warning, and What the Two Studies Behind It Tested

The Ozempic label has no contraception instruction; the tirzepatide labels do. Both semaglutide pill studies were run in postmenopausal women at steady state.

Ronald R · Edited by Caroline S · Published 2026-11-18

Illustration: Two plain white pills in a small ceramic dish on a warm wooden bathroom vanity with a folded towel.
Illustration

The Ozempic label carries no birth control warning. The tirzepatide labels, Mounjaro and Zepbound, carry one. That difference is real and it is in the documents, but the reason for it is not quite the one usually given. It does not show that semaglutide was tested in the same conditions as tirzepatide and passed. It shows that the two manufacturers measured the pill at different moments in treatment.

Every label figure below was read on 18 November 2026 from DailyMed: the Ozempic injection label (version 20, effective 1 June 2026), the Wegovy label (version 19, effective 18 June 2026) and the Zepbound label (version 40, effective 28 August 2026). The study figures come from the published papers, cited by PubMed number.

What the semaglutide labels say

The Ozempic label's interaction section reports a set of studies run "at semaglutide 1 mg steady-state exposure", in which a combined oral contraceptive, ethinylestradiol with levonorgestrel, was among the medicines tested. Its conclusion is that no clinically relevant drug-drug interaction was observed and that no dose adjustment is needed. There is no contraception paragraph anywhere in the document.

The Wegovy label says the same in one sentence: no clinically significant differences in the pharmacokinetics of ethinyl estradiol or levonorgestrel were observed with semaglutide. That summary sentence does not state the semaglutide dose at which the measurement was made.

The tirzepatide labels say something else entirely, and say it three times: in the interaction section, in the section on people who could become pregnant, and in the patient counselling text. Tirzepatide "may reduce the efficacy of oral hormonal contraceptives due to delayed gastric emptying", the delay "is largest after the first dose and diminishes over time", and patients on a pill are to be advised to switch to a non-oral method or add a barrier method for four weeks after starting and four weeks after each dose increase. The full side-by-side of every GLP-1 label's interaction lines is on GLP-1 drug interactions.

The two studies behind the semaglutide line

Study Who took part Semaglutide When the pill was measured Result (exposure ratio, with vs without)
Kapitza 2015, PMID 25475122 43 postmenopausal women with type 2 diabetes Weekly injection, escalated 0.25 mg × 4 weeks, 0.5 mg × 4 weeks, 1.0 mg × 5 weeks At 1.0 mg steady state Ethinylestradiol 1.11 (90% CI 1.06–1.15); levonorgestrel 1.20 (1.15–1.26); peak levels within limits
Jordy 2021, PMID 33782832 25 healthy postmenopausal women Oral semaglutide (the tablet) During daily dosing Ethinylestradiol 1.06 (1.01–1.10); levonorgestrel 1.06 (0.97–1.17); peak levels unaffected
For comparison: tirzepatide, Zepbound label 12.3 Not stated in the label A single 5 mg dose After the first dose Peak ethinyl estradiol −59%, total exposure −20%; peak delayed 2.5–4.5 hours

Read as written, the semaglutide results are reassuring in a narrow sense: neither hormone's level fell, and in the injection study levonorgestrel exposure actually rose by a fifth. The 1.0 mg study also recorded the expected effects of the drug in the same women — HbA1c fell by 1.1 points and weight by 4.3 kg — so semaglutide was clearly active while the pill was being measured.

What those studies did not test

Four limits follow directly from the designs, and none of them appears in the label sentence that summarises the result.

The first weeks were not measured. The injection study measured the pill at steady state on 1.0 mg, after thirteen weeks of treatment. The tirzepatide study measured it after a single first dose. The Zepbound label itself explains why that matters: the stomach-slowing effect is largest after the first dose and fades. The Ozempic label makes the matching point about semaglutide in its own pharmacology section — the drug causes "a delay of early postprandial gastric emptying", and in a separate study "no apparent effect on the rate of gastric emptying was observed with semaglutide 2.4 mg". A study that measures only at steady state is measuring after most of the effect that worries people has gone.

Each dose step was not measured either. Semaglutide is started at 0.25 mg and stepped up at intervals of at least four weeks. The tirzepatide warning attaches to every dose increase because each step restarts some of the early slowing. Nothing in either semaglutide study measured the pill around a step.

The higher doses are not described. The Ozempic label says its interaction studies were run at 1 mg. Ozempic is now prescribed up to 2 mg and Wegovy at 2.4 mg, and the label summaries do not describe a pill study at either dose.

Nobody's ovulation was measured. Both semaglutide studies enrolled postmenopausal women, the standard design for this kind of test, which measures hormone levels in the blood rather than whether the pill still prevents ovulation. Hormone levels are a reasonable stand-in, and tirzepatide was judged on the same kind of measure, but neither drug has a study showing what happens to ovulation or pregnancy rates in women of reproductive age.

None of this shows that Ozempic weakens the pill. It shows that the label's reassurance is a statement about steady-state hormone levels in postmenopausal women, and that the period of treatment where the tirzepatide data found a problem is the period the semaglutide data did not look at.

Vomiting is a separate route, and it has no evidence either way

The interaction studies are about absorption. A second, simpler route is the side effects themselves. In the Ozempic label's type 2 diabetes trials, vomiting was reported by 5% of patients on 0.5 mg and 9.2% on 1 mg, against 2.3% on placebo; diarrhoea by 8.5% and 8.8% against 1.9%.

CDC's U.S. Selected Practice Recommendations for Contraceptive Use, 2024 (PMC11340200) has a section on exactly this situation for combined pills, with a set of steps depending on how long the vomiting or diarrhoea lasts. Its own evidence summary is blunt: the effectiveness of the pill "might be decreased" theoretically, the recommendations are borrowed from the rules for missed pills, and "no evidence was found" on the effects of vomiting or diarrhoea on pregnancy, follicle development, hormone levels or cervical mucus. That applies to vomiting from any cause, semaglutide included.

What regulators have added

The UK's medicines regulator, the MHRA, issued a reminder on 5 June 2025 aimed at people using GLP-1 medicines for weight loss and diabetes. It said effective contraception is needed while on these medicines, that effective contraception includes the pill as well as the implant, coil and condoms, and that the extra step — adding a non-oral method — applied only to Mounjaro, especially in the four weeks after starting and after any dose increase. Its wording on Mounjaro referred specifically to people who are overweight and using the pill. The MHRA did not extend the advice to semaglutide.

In the US, the FDA's position is the label text: an instruction for tirzepatide, none for semaglutide.

How many people this concerns

A Danish nationwide register study published in July 2026 (Bager and colleagues, JAMA Network Open, PMID 42485041) compared 22,694 women aged 12 to 49 who started semaglutide with 229,640 who did not. Every pattern of hormonal contraception use was associated with later starting semaglutide; for women who had used only the combined pill, the adjusted hazard ratio was 1.42 (95% CI 1.34–1.51). The study was about who starts the drug, not about interactions, but it shows the overlap between pill users and semaglutide users is large and not incidental.

The adverse event reports

FDA's adverse event reporting database, read through openFDA on 18 November 2026 (data updated 28 October 2026), holds:

Molecule Reports naming it "Pregnancy on contraceptive" "Unintended pregnancy"
Semaglutide 78,023 7 6
Tirzepatide 181,068 4 16
Liraglutide 51,303 1 3

These numbers are small and they cannot answer the question. A report is a voluntary submission that a drug was being taken when something happened; it does not establish cause, nobody knows how many people took both a GLP-1 and a pill, and a single report can carry both terms. What the table does show is that neither molecule has produced a visible cluster of contraceptive-failure reports, and that the reports which exist are not confined to the drug that carries the warning.

What the record leaves open

The firm part is short. The Ozempic and Wegovy labels carry no contraception instruction; the Mounjaro and Zepbound labels do; and the studies behind each label were run at different points in treatment. The open part is what happens to the pill in the first weeks on semaglutide and after each dose step, at the 2 mg and 2.4 mg doses, and in women of reproductive age, because no published study has measured it. Vomiting adds a separate risk that guidance recognises and no study has quantified.

What the labels say about pregnancy itself — including the two-month gap the semaglutide labels give before a planned pregnancy — is on GLP-1s and pregnancy. The same contraception question arises with particular force in polycystic ovary syndrome, where the pill is a first-line treatment; that is covered on GLP-1s for PCOS. The rest of Ozempic's label record is on Ozempic side effects.

This page reports what the labels, studies and regulators say and is not medical advice. Which contraceptive method suits a particular person while on Ozempic is a decision for that person and the prescriber, as our editorial standards set out.

Frequently asked questions

Does Ozempic make birth control pills less effective?

The Ozempic label says no clinically relevant interaction was found and carries no contraception instruction. That conclusion rests on a study in 43 postmenopausal women with type 2 diabetes, measured at the 1 mg dose after 13 weeks of escalation, in which ethinylestradiol and levonorgestrel levels did not fall (exposure ratios 1.11 and 1.20). The first weeks of treatment and each dose increase, when the stomach slows most, were not what that study measured. The question of which method suits a particular person belongs with the prescriber.

Why do Mounjaro and Zepbound have a birth control warning but Ozempic does not?

Because their studies measured different moments. Lilly's tirzepatide study gave a combined pill alongside a single first 5 mg dose and found peak ethinyl estradiol 59% lower and total exposure 20% lower. Novo Nordisk's semaglutide studies measured the pill at steady state, after weeks of treatment, and found no reduction. Both labels describe their own data accurately; they simply answer different questions.

Can vomiting on Ozempic stop the pill from working?

It is a recognised theoretical risk, separate from any drug interaction. Vomiting was reported by 9.2% of patients on Ozempic 1 mg in the label's diabetes trials, against 2.3% on placebo. CDC's 2024 contraception guidance sets out steps to follow after vomiting or severe diarrhoea on a combined pill, but states that its advice is borrowed from the missed-pill rules because no evidence was found on what vomiting does to the pill's effectiveness.

Does Ozempic affect the coil, implant or injection?

Neither semaglutide label reports any interaction with non-oral contraception, and the mechanism discussed for the pill, slower stomach emptying, concerns medicines absorbed through the gut. The tirzepatide labels state explicitly that hormonal contraceptives not taken by mouth should not be affected. The UK regulator listed the implant, coil and condoms as non-oral methods in its 2025 reminder.

Are there reports of pregnancies on Ozempic despite birth control?

A few. FDA's adverse event database, read on 18 November 2026, holds 7 semaglutide reports coded 'pregnancy on contraceptive' and 6 coded 'unintended pregnancy' among 78,023 semaglutide reports. A report does not establish that the drug caused anything, there is no count of how many people used both, and one report can carry more than one term, so the numbers cannot be turned into a failure rate.

How long before trying to conceive should Ozempic be stopped?

The Ozempic label's patient information gives at least two months before a planned pregnancy, because semaglutide is cleared slowly. What the labels say about pregnancy, and how the semaglutide and tirzepatide labels differ on it, is set out on our page on GLP-1s and pregnancy. The timing for any individual is a question for the prescriber.