Evidence review · from the 250-topic plan
Pregnancy, fertility, and the runway: stopping GLP-1s before conceiving
These medications and pregnancy don't mix — animal studies showed fetal harm, adequate human data doesn't exist, and the labels are unambiguous: discontinue when pregnancy is recognized. For planned pregnancy, the runway rule: semaglutide's label specifies stopping at least two months before conceiving — the number falls straight out of the half-life math (a week-long half-life needs ~five weeks to functionally clear, padded for safety); tirzepatide's faster clearance implies a shorter runway, with your prescriber setting the date. The hard part isn't the stopping — it's what stopping means: contraception until the runway completes, then trying-to-conceive months spent defending weight without the medication, with the regain data read honestly in advance. The protocol, the discovered-pregnancy scenario, breastfeeding, and the partner question — the whole file, plainly.
Why the rule exists
Three stacked reasons. Animal signal: reproductive studies in both molecules' files showed fetal harm at relevant exposures — the finding that writes warnings. Human data gap: pregnant patients are excluded from trials, so the reassurance that would soften the animal signal simply doesn't exist yet (registries tracking inadvertent exposures are accumulating it slowly). Biological plausibility: pregnancy is the one state where appetite suppression and reduced intake are the opposite of the assignment. Uncertainty plus plausibility plus signal equals the label's clarity — this is one of the class's few true bright lines.
The runway math
Semaglutide's two-month pre-conception rule is printed label text, and it's just pharmacokinetics wearing a calendar: ~seven-day half-life × four-to-five half-lives ≈ five weeks to functional clearance, rounded up because embryonic development starts before most tests turn positive. Tirzepatide's ~five-day half-life clears in roughly three-to-four weeks — labels frame discontinuation-before-planned-pregnancy with your prescriber owning the exact date; the conservative household translation many clinicians use is “about a month, and two never hurts.” The operational rule that makes both work: effective contraception until the runway is complete — which for tirzepatide users also means the four-week backup windows were part of this story all along.
Pregnancy discovered on therapy
It happens — the fertility-restoration pattern guarantees it — and the protocol is calm and specific: stop the medication now, call your prescriber and an obstetric clinician promptly, and don't panic. The exposure in most discovered cases is early and brief; registry data from inadvertent exposures, while limited, has not painted the catastrophic picture panic assumes; and the drug's own appetite effects fade over the weeks-long tail — expect hunger's return and feed it like the assignment it now is. What the moment is not: a reason for shame spirals or forum-diagnosis. It's a Tuesday-morning phone call with a clear first step already taken.
The trying-to-conceive gap — the part nobody plans for
Here's the file's honest center: conception can take months, and every one of them is spent off the medication, with the withdrawal data predicting pressure toward regain. Pretending otherwise helps no one; planning does. The pre-runway conversation with your clinician: a maintenance strategy for the gap — the protein-and-training floor doing its heaviest lifting, structured eating patterns installed while still on therapy (habits built in the quiet transfer better than habits built in the noise), realistic weight-band expectations agreed in advance, and — because obstetric medicine cares about trajectory more than a number — the reframe that a partial regain into a healthy pregnancy beats white-knuckle perfection into a stressed one. Timing strategy belongs here too: some patients and clinicians sequence major loss first, then the runway, then trying — using the therapy to reach the healthiest achievable pre-pregnancy baseline before the clock starts.
Breastfeeding, and the partner question
Breastfeeding: whether these peptides pass into human milk in meaningful amounts isn't established; labels counsel caution, and the standard clinical posture is to delay restarting until nursing concludes — another prescriber conversation with the same uncertainty-shaped honesty. The partner question: no established restriction exists for male partners' use around conception — no label instruction, no demonstrated paternal-mediated concern — though anyone worried should simply ask; “no known issue” and “don't discuss it” are different sentences. And the program test this file writes: intake that asked about pregnancy plans and contraception before shipping is practicing medicine; checkout that didn't left this entire article as your homework. Intake that plans ahead ↗
FAQ
How long before pregnancy should I stop semaglutide or tirzepatide?
Semaglutide's label specifies at least two months before a planned pregnancy; tirzepatide's shorter half-life implies a shorter runway with your prescriber setting the date — contraception stays on until the runway completes.
What if I get pregnant while on a GLP-1?
Stop the medication, call your prescriber and an obstetric clinician promptly, and don't panic — most discovered exposures are early and brief, and registries haven't shown the catastrophe panic assumes.
How do I keep weight off while trying to conceive?
Plan the gap before it starts: habits installed while still on therapy, the protein-and-training floor, agreed weight-band expectations — and the reframe that a healthy pregnancy trajectory beats perfect numbers.
Can I take GLP-1s while breastfeeding?
Milk transfer isn't established; labels counsel caution and standard practice delays restarting until nursing concludes — a prescriber decision.
Sources
- Wegovy and Zepbound prescribing information — pregnancy discontinuation and pre-conception language.
- Reproductive animal-study findings; pregnancy-exposure registry reports.
- Half-life clearance math — companion analysis; withdrawal-trial regain data.