Do GLP-1 medications affect birth control, and how long before pregnancy should they be stopped?

The short answer

Tirzepatide and semaglutide carry opposite warnings, and almost every article on this subject blurs them together.

Tirzepatide can reduce the absorption of oral contraceptives, and its label requires switching to a non-oral method or adding a barrier method for four weeks after starting and four weeks after every dose increase. Semaglutide carries no contraceptive warning at all — the interaction was studied and found not clinically relevant.

Reverse that for pregnancy planning. Semaglutide labels advise stopping at least two months before a planned pregnancy. Tirzepatide labels specify no washout period.

The MHRA put the distinction in one sentence: "This only applies to those taking Mounjaro."

Evidence tier: Tier 1 — every label quotation below was read directly from the FDA-approved prescribing information. Educational content, not medical advice.

The key points:

• Tirzepatide: backup contraception for 4 weeks after start and after each dose increase.
• Semaglutide: no contraceptive warning. The interaction was tested and was negative.
• Semaglutide: stop 2 months before a planned pregnancy. Tirzepatide gives no figure.
• Non-oral methods are unaffected — the label says so explicitly.
• "Ozempic babies" is more likely restored fertility than contraceptive failure.

The two drugs point in opposite directions

Evidence tier: 1 — labels read directly.

That is not an oversight in either direction. Both were studied, and they produced different results.

What tirzepatide's label actually says

Evidence tier: 1 — Mounjaro §7.2, §8.3 and §12.3.

Section 8.3, verbatim:

"Use of MOUNJARO may reduce the efficacy of oral hormonal contraceptives due to delayed gastric emptying. This delay is largest after the first dose and diminishes over time. Advise patients using oral hormonal contraceptives to switch to a non-oral contraceptive method, or add a barrier method of contraception for 4 weeks after initiation and for 4 weeks after each dose escalation."

And section 7.2 answers the obvious follow-up directly: "Hormonal contraceptives that are not administered orally should not be affected."

The pharmacokinetics explain why the warning is time-limited rather than permanent, and this is the detail that gets lost. With a single 5 mg dose alongside a combined oral contraceptive:

Time to peak was delayed 2.5 to 4.5 hours.

Read those two columns against each other. The peak is cut by more than half; total exposure falls by around a fifth. This is not a drug being destroyed — it is a drug arriving late and flatter, because the stomach is emptying more slowly. Since contraceptive efficacy depends partly on reaching threshold concentrations, a blunted peak matters. And because the gastric-emptying effect is largest after the first dose and fades with repeat dosing, the risk window is the titration period rather than the whole course. Hence four weeks, repeated at each step up.

Zepbound's label carries identical wording and identical figures.

Semaglutide was tested, and the result was negative

Evidence tier: 1 — Ozempic §7.2 and §12.3, Wegovy §12.3, Rybelsus §7.2.

This is the part most coverage gets wrong, usually by assuming the drugs behave alike.

Ozempic's section 7.2 is short enough to quote in full:

"OZEMPIC causes a delay of gastric emptying, and thereby has the potential to impact the absorption of concomitantly administered oral medications. In clinical pharmacology trials, semaglutide did not affect the absorption of orally administered medications to any clinically relevant degree. Nonetheless, caution should be exercised when oral medications are concomitantly administered with OZEMPIC."

No mention of contraceptives. And crucially, this is not an untested gap. The Ozempic drug-interaction figure states that a combined oral contraceptive — ethinyl estradiol with levonorgestrel — was assessed at steady state. Wegovy's section 12.3 lists the result plainly: no clinically significant pharmacokinetic differences were observed for a list that explicitly includes ethinyl estradiol and levonorgestrel.

Rybelsus, the oral formulation, does not differ. Its interaction section flags levothyroxine, with exposure increased 33%, and drugs with a narrow therapeutic index. Contraceptives are absent.

So the honest framing is: semaglutide's contraceptive interaction was looked for and not found. That is a stronger statement than "no warning exists," and it is the one the labels support.

Other GLP-1s, briefly. Liraglutide and dulaglutide carry no advice to change contraception. Exenatide is the interesting exception — it does not tell you to switch methods, but it does instruct that oral medications dependent on threshold concentrations, naming contraceptives, be taken at least an hour before injecting. Its own data show why: taken 30 minutes after exenatide, ethinyl estradiol peak fell 45%; taken an hour before, it fell only 15%.

The pregnancy rule runs the other way

Evidence tier: 1 — §8.3 of each label.

Every semaglutide label carries a washout instruction. Wegovy:

"Because of the potential for fetal harm, discontinue WEGOVY in patients at least 2 months before they plan to become pregnant to account for the long half-life of semaglutide."

Ozempic and Rybelsus say the same thing in near-identical words, citing "the long washout period for semaglutide."

Neither Mounjaro nor Zepbound specifies any washout period before conception. Their reproductive-potential sections contain only the contraceptive advice.

On pregnancy itself, both molecules converge: available human data are insufficient to evaluate drug-related risk, animal reproduction studies suggest potential fetal risk, and for the weight-management indications the labels are direct — weight loss offers no benefit in pregnancy and may cause fetal harm, so the drug should be stopped when a pregnancy is recognised.

Registries exist, but only for the obesity products. Wegovy and Zepbound name pregnancy exposure registries; Ozempic, Rybelsus and Mounjaro do not. That is a real asymmetry given the same molecules are involved. Zepbound's label as of its September 2025 revision still says a registry "will be" established, suggesting it was not yet enrolling. No outcomes from either registry have been published.

So what is actually behind "Ozempic babies"?

Evidence tier: 1–2 — labels, regulator statement and meta-analyses.

Three different mechanisms get collapsed into one story. They are not equally supported.

Reduced contraceptive absorption. Established for tirzepatide, quantified above. Not established for semaglutide, where it was tested and found not clinically relevant. Separately — and applying to any drug — severe vomiting or diarrhoea can impair pill absorption. That is ordinary sick-day contraceptive advice, not a pharmacokinetic property of GLP-1s, and it should not be presented as one.

Restored ovulation after weight loss. This is the likeliest explanation for most cases, and the evidence is real but weaker than headlines suggest. A 2023 meta-analysis of 11 randomised trials in 840 women with PCOS found improved natural pregnancy rates (RR 1.72, 95% CI 1.22–2.43) and menstrual regularity — though heterogeneity on the menstrual outcome was 95.6%, which is very high. A more recent GRADE-assessed review reached a more cautious conclusion: modest short-term weight loss, but evidence "insufficient to draw conclusion regarding glucose, insulin, hirsutism, and menstrual regularity."

No randomised trial has ovulation or live birth as a primary endpoint. Both readings should be held at once: a plausible, partly-supported fertility effect, with the reproductive endpoints rated uncertain by the more rigorous assessment.

Anecdote and amplification. The phrase itself came from social media and press coverage, not from a safety signal. The one quantified regulatory figure we found was roughly 40 pregnancy reports across all GLP-1s — a spontaneous-report count, not an incidence, and with no comparison against the baseline pregnancy rate in reproductive-age women losing substantial weight.

The MHRA drew the line explicitly in June 2025, advising that people taking Mounjaro who use oral contraception should add a non-oral method, and then stating: "This only applies to those taking Mounjaro" — especially in the four weeks after starting and after any dose increase.

The PCOS guideline gives a different reason for the same advice

Evidence tier: 1 — 2023 international guideline.

The 2023 International Evidence-Based Guideline for PCOS supports considering GLP-1 receptor agonists alongside lifestyle intervention for weight management. It then adds a practice point worth reading carefully:

"Healthcare professionals should ensure concurrent effective contraception when pregnancy is possible for women who take GLP-1 receptor agonists, as pregnancy safety data are lacking."

Note the reasoning. That recommendation is grounded in absent pregnancy safety data — not in a drug interaction. It applies to the whole class, including semaglutide, and for a completely different reason than tirzepatide's absorption warning.

Two pieces of advice that look identical therefore rest on separate foundations. Conflating them is how people end up believing semaglutide interferes with the pill.

The guideline also flags gradual dose escalation to limit gastrointestinal effects, and the need to weigh long-term use against the high risk of weight regain after stopping.

What this means in practice

• On tirzepatide, using oral contraception: the label asks for a non-oral method or an added barrier method for four weeks from starting, and again for four weeks after every dose increase. The second half is the part people miss — each escalation restarts the window.
• On tirzepatide, using an IUD, implant, injection, patch or ring: the label says non-oral methods should not be affected.
• On semaglutide: no contraceptive change is indicated by the label. If you are also vomiting heavily, ordinary sick-day rules apply.
• Planning a pregnancy on semaglutide: the labels say stop at least two months beforehand.
• Planning a pregnancy on tirzepatide: no washout figure is given, which is a gap rather than a reassurance. This is a prescriber conversation.
• Either drug, pregnancy recognised: the weight-management labels say stop.

Limitations

This is educational content, not medical advice.

• Labels change. Every quotation here is from the revision current in September 2026; check the current label before relying on any of it.
• The tirzepatide PK figures come from a single-dose interaction study with one combined oral contraceptive. They are not a general statement about every formulation.
• We did not open the Saxenda or Trulicity PDFs — those two entries come from secondary summaries and are directional.
• The Byetta text we read carried an FDA notice that it may not be the latest approved version.
• No pregnancy registry outcomes have been published for either molecule; the human data available are pharmacovigilance and observational.
• Marko Maal, MSc Pharmacy reviewed this article. Reviewer attribution does not constitute a doctor-patient relationship.

The bottom line

The two most-used GLP-1s carry mirror-image warnings, and treating them as interchangeable gets both wrong. Tirzepatide has a specific, quantified, label-mandated contraceptive interaction and no stated pre-pregnancy washout. Semaglutide has a tested-and-negative contraceptive result and a two-month washout. Advice written for one is wrong for the other in both directions.

The "Ozempic babies" story is most plausibly about fertility returning to people who believed they were subfertile — a real and under-discussed consequence of substantial weight loss, particularly in PCOS — rather than about contraception failing. The exception is tirzepatide during titration, where the mechanism is documented and the label is specific.

And the contraception advice in the PCOS guideline exists for a third reason again: not interaction, not fertility, but the simple fact that pregnancy safety data for this class are lacking. Three different rationales, one superficially similar instruction, and they are worth keeping apart.

Related on this site

• GLP-1s for women: PCOS, fertility and cycle effects
• Peptides to avoid in pregnancy and breastfeeding
• Peptides and the menstrual cycle
• Peptides in perimenopause and menopause
• GLP-1 side effects and how to manage them
• GLP-1 titration schedules
• Stopping GLP-1s: rebound and regain

References

• Mounjaro (tirzepatide) prescribing information, §7.2, §8.1, §8.3, §12.3 — oral contraceptive advice, 4-week windows, and the Cmax/AUC figures. FDA label
• Zepbound (tirzepatide) prescribing information — identical contraceptive wording and PK; pregnancy exposure registry. FDA label
• Ozempic (semaglutide) prescribing information, §7.2, §8.3, §12.3 — no contraceptive warning; combined oral contraceptive assessed at steady state; 2-month pre-pregnancy discontinuation. FDA label
• Wegovy (semaglutide) prescribing information, §7.2, §8.1, §8.3, §12.3 — ethinyl estradiol and levonorgestrel among drugs with no clinically significant interaction; 2-month discontinuation; pregnancy registry. FDA label
• Rybelsus / Ozempic tablets prescribing information, §7.2 — levothyroxine flagged, contraceptives absent. FDA label
• MHRA, Women on "skinny jabs" must use effective contraception, 5 June 2025 — "This only applies to those taking Mounjaro." Press release · Companion guidance
• 2023 International Evidence-Based Guideline for the Assessment and Management of Polycystic Ovary Syndrome, §4.5 — recommendation 4.5.1 and practice points 4.5.2–4.5.4. Guideline
• Ma R, et al. Effects of GLP-1 receptor agonists on women with PCOS: a meta-analysis. BMC Endocrine Disorders 2023 — 11 RCTs, 840 women; natural pregnancy rate RR 1.72 (1.22–2.43). doi:10.1186/s12902-023-01500-5
• Forslund M, et al. European Journal of Endocrinology 2026;194(3):S25–S39 — GRADE-assessed; evidence insufficient on menstrual regularity and metabolic outcomes. doi:10.1093/ejendo/lvag033
• Byetta (exenatide) prescribing information, §7.1 and §12.3 — one-hour dosing separation for threshold-dependent oral drugs including contraceptives. FDA label (label carries an FDA currency notice)