The "Ozempic baby" phenomenon — a wave of unexpected pregnancies in women who had previously experienced infertility or irregular cycles — generated a lot of headlines, most of them framed as a curiosity or a surprise. It isn't either. It's the predictable result of a medication correcting the metabolic dysfunction that was causing hormonal dysregulation in the first place. Understanding why that happens explains both the opportunity these medications represent for women with PCOS and the risks that require careful planning.
This post covers three interconnected topics: what GLP-1 medications do to the reproductive hormonal axis in women with PCOS, what we know about pregnancy safety and the timing of discontinuation, and the oral contraceptive interaction that every woman on these medications needs to know about.
Why PCOS Disrupts the Reproductive Cycle — and What GLP-1 Medications Do About It
Polycystic ovary syndrome (PCOS) is the most common endocrine disorder in women of reproductive age, affecting roughly 8–13% of women globally. Despite its name, the central problem in most PCOS is not the ovarian cysts — those are a consequence, not a cause. The central problem, in the majority of cases, is insulin resistance and compensatory hyperinsulinemia (elevated circulating insulin levels that the body produces to overcome its own resistance to insulin's effects).
Here's why that matters for reproduction: elevated insulin directly stimulates the ovary's theca cells to produce excess androgens (male sex hormones — primarily testosterone). Elevated insulin also suppresses the liver's production of sex hormone-binding globulin (SHBG), a protein that binds testosterone and renders it inactive. The net result is a surge in free, biologically active testosterone that disrupts the hormonal feedback signals controlling ovulation.
The Insulin Resistance → Androgen Excess → Anovulation Cycle in PCOS
GLP-1 medications attack this cycle at its root. By improving insulin sensitivity and reducing insulin secretion relative to need, they lower the hyperinsulinemia driving androgen excess. Less insulin means less ovarian androgen stimulation, higher SHBG (allowing more testosterone to be bound and inactivated), and restoration of the hormonal feedback loop that allows regular ovulation.
Weight loss — which GLP-1 medications produce in most patients — has an additive effect through a separate mechanism: adipose tissue (fat) independently produces androgens and inflammatory cytokines (signaling molecules that promote inflammation) that further disrupt ovulation. Losing visceral fat reduces this secondary androgen source.
A 2026 meta-analysis of 7 randomized controlled trials evaluated liraglutide specifically in women with PCOS. Liraglutide significantly increased menstrual frequency (Hedges' g = 1.76 — a large effect size), reduced BMI, improved insulin resistance markers, lowered LH (luteinizing hormone, which drives androgen production) and free androgen index, and increased SHBG. A separate meta-analysis found GLP-1 receptor agonist use in women with PCOS was associated with improved menstrual regularity (SMD 1.72) and a 72% higher rate of spontaneous pregnancy compared to metformin alone. These are meaningful reproductive outcomes — not marginal improvements.
The Contraception Complication: What Every Woman on GLP-1 Therapy Needs to Know
GLP-1 medications slow gastric emptying — the rate at which food and medications transit from the stomach to the small intestine, where absorption primarily occurs. For most medications, this has limited clinical significance. For oral contraceptives, it can matter considerably.
If a birth control pill passes through the stomach more slowly, it reaches the absorptive surface of the small intestine later and potentially less completely. The resulting reduction in circulating estrogen and progestin concentrations may be sufficient to impair the pill's effectiveness — particularly during periods of dose escalation when gastric slowing is most pronounced.
The 2026 ADA Standards of Care and the 2023 International PCOS Guideline both specifically address this interaction. The ADA recommendation for tirzepatide is explicit: patients who take oral contraceptives should use a backup method of contraception — condoms, barrier methods, or a non-oral hormonal method — for 4 weeks after each dose increase until the maintenance dose is established. GLP-1 receptor agonists (semaglutide, liraglutide) have a similar but generally lesser effect on gastric emptying. The 2023 PCOS guideline, which recommends GLP-1 medications as an option for weight management in PCOS, explicitly states that effective contraception must be used concurrently unless pregnancy is desired — precisely because these medications may restore ovulatory function in women who had been inadvertently relying on anovulation as contraception.
The practical message is this: if you are a woman with PCOS who has had irregular or absent periods, and you begin a GLP-1 medication, you should not assume you remain infertile. The drug may restore ovulation before your periods even normalize. If pregnancy is not desired, ensure you have reliable contraception in place before starting — and if you're on the pill, discuss a backup method with your prescribing physician.
Pregnancy Safety: What We Know and What We Don't
This is the section I'm most careful about, because the clinical picture here is genuinely incomplete — and incomplete evidence in the context of pregnancy decisions requires particular honesty about uncertainty.
The good news: human observational data collected to date has not identified a clear pattern of congenital anomalies or pregnancy complications specifically attributable to GLP-1 medications. The concerning news: preclinical (animal) data raises questions that haven't been definitively answered in humans.
The FDA semaglutide label notes that in rat studies, semaglutide caused increased estrus cycle length and reduced corpora lutea (structures formed after ovulation) at all doses tested — findings attributed to reduced food intake and weight loss rather than direct hormonal toxicity. More significantly, animal studies showed reduced fetal weight, delayed bone formation (ossification), and skeletal variants in exposed offspring, coinciding with marked maternal weight loss. A 2025 Endocrine Society and European Society of Endocrinology joint guideline on preexisting diabetes and pregnancy recommends discontinuing semaglutide at least 2 months prior to attempting conception, based on this preclinical data and the absence of adequate human safety data.
The challenge here is genuine: the animal studies showing fetal effects used exposures that also produced significant maternal weight loss. It's not clear whether the fetal effects were caused by the drug directly or by the dramatic nutritional change associated with maternal weight loss. Human registry data collected from pregnancies inadvertently exposed to GLP-1 medications has not shown a clear teratogenic signal — but the numbers remain too small to be reassuring about rare outcomes.
There's a real tension embedded in the discontinuation recommendation that I want to be explicit about. Stopping GLP-1 therapy before pregnancy carries its own risks: most patients will regain significant weight and may experience return of hyperglycemia during the first trimester — the period of highest fetal sensitivity to metabolic disturbance. For women with type 2 diabetes, this is a meaningful clinical concern. For women with PCOS who achieved menstrual restoration on GLP-1 therapy, stopping the medication may cause return of anovulation, potentially making conception harder. These are tradeoffs that require individualized discussion with your physician, not a one-size-fits-all recommendation. The current guideline represents the most cautious reasonable position given incomplete data — but that position should be revisited as more human pregnancy exposure data accumulates.
The "Ozempic Baby" Phenomenon: What's Actually Happening
The reports of unexpected pregnancies in women on GLP-1 medications — primarily women who had previously experienced infertility or irregular cycles, many with PCOS — are a direct consequence of the reproductive mechanisms described above. When insulin resistance improves and androgens fall, ovulation can resume. It often does so before cycles normalize, meaning a woman may ovulate and conceive before she perceives any change in her period pattern.
This is not a medical mystery. It is the expected result of treating the metabolic condition that was impairing reproduction. What it requires is that prescribing clinicians explicitly counsel every woman of reproductive age starting GLP-1 therapy about the possibility of restored fertility — and that contraception be planned proactively, not reactively.
For women with PCOS who want to conceive, GLP-1 medications represent a genuinely meaningful addition to our therapeutic toolkit — not as fertility treatments in the formal sense, but as metabolic interventions that address a root cause of the reproductive dysfunction. The evidence for improved menstrual regularity, reduced androgen excess, and higher spontaneous pregnancy rates is real and clinically significant. My approach is to treat GLP-1 therapy in reproductive-age women with PCOS as a conversation that needs to happen in three directions simultaneously: what are your goals for weight and metabolic health, what are your reproductive plans, and what contraception are you using right now? Getting those three answers before the first injection prevents situations where the drug did exactly what it was supposed to do — and nobody told the patient it might.
A Practical Timeline for Women Planning Pregnancy
Before starting a GLP-1 medication
Discuss reproductive plans explicitly. If pregnancy is desired in the near term (<12 months), weigh the benefit of metabolic optimization before conception against the need to discontinue the drug prior to conception. Ensure contraception is in place if pregnancy is not immediately desired.
While on GLP-1 therapy — not planning pregnancy
Use reliable contraception. If you take oral contraceptives, use a backup method for 4 weeks after each dose increase (tirzepatide; GLP-1 agents similar caution applies). Do not assume prior infertility or irregular cycles protect you from conception.
Planning to attempt conception
Discuss timing of discontinuation with your physician. Current guidelines recommend stopping semaglutide at least 2 months before attempting conception. The specific timeline for other agents should be individualized. If you have type 2 diabetes, ensure alternative glucose management is in place before discontinuing.
Unexpectedly pregnant while on a GLP-1 medication
Contact your physician promptly. Discontinue the medication. This is not a reason for alarm based on current human data — the evidence to date has not shown a clear teratogenic pattern — but your care team should be informed and monitoring should be appropriate for your clinical situation. Report the exposure to the manufacturer's pregnancy registry if one is available for your medication.
After delivery, while breastfeeding
The picture here is more nuanced than the standard "not recommended" label implies — and worth understanding carefully if you are a mother who wants to breastfeed and also needs or wants to resume GLP-1 therapy.
Breastfeeding and GLP-1 Medications: A More Nuanced Picture
The conventional advice has been to avoid GLP-1 medications while breastfeeding due to limited data. That's still the cautious default — but the data that does exist is more reassuring than the label language suggests, and I think mothers deserve to understand it rather than simply receive a blanket restriction.
For tirzepatide, a dedicated lactation study measured breast milk concentrations following a single 5 mg dose in 11 breastfeeding women. Tirzepatide was essentially undetectable: concentrations fell below the limit of detection in 96% of samples (164 of 171). In the small number of samples where any drug was measurable, the cumulative amount detected over a 28-day window represented less than 0.02% of the maternal dose, with the last measurable concentration occurring just 5 days after the dose. The FDA tirzepatide label reflects this, advising physicians to weigh the developmental benefits of breastfeeding against the mother's clinical need — without a specific contraindication.
For semaglutide, concentrations in breast milk were below the lower limit of quantification in a lactation study of the oral tablet formulation. No human breast milk data are currently available for subcutaneous semaglutide specifically, though a systematic review confirmed no detectable semaglutide transfer into human milk in the one pharmacokinetic study that assessed this. This is consistent with the pharmacological properties of GLP-1 receptor agonists as large peptide molecules — their size makes breast milk transfer inherently unlikely, and even if transferred, gastrointestinal absorption by the infant is unlikely given peptide breakdown in the gut.
One important exception: oral semaglutide tablets (Wegovy oral formulation) contain an absorption enhancer called SNAC. SNAC and its metabolites are present in human milk, and because enzyme activity involved in SNAC clearance may be lower in newborns, the FDA specifically advises against breastfeeding during treatment with oral semaglutide tablets. This warning does not apply to subcutaneous semaglutide, which does not contain SNAC.
The available data — particularly for tirzepatide — is more reassuring than most patients are told. Breast milk transfer appears to be negligible, and the peptide structure of these molecules makes infant absorption through the gut unlikely even in the small amounts that might be present. For a mother who needs GLP-1 therapy for diabetes management or significant obesity-related health conditions, and who wants to breastfeed, I think a careful, individualized conversation with her physician is warranted rather than an automatic stop. That conversation should include close monitoring of both mother and infant, and should exclude oral semaglutide specifically given the SNAC concern. The data is not yet sufficient to make a blanket recommendation to resume — but it is sufficient to say the risks, if any, appear very small, and that the decision should be made thoughtfully rather than reflexively.
A Note on the Broader Evidence in This Space
I want to close by acknowledging something honestly: the reproductive medicine data for GLP-1 medications is less mature than the cardiovascular, metabolic, and weight loss data I've described in other posts in this series. Most of what we know comes from trials of liraglutide in PCOS populations — a shorter-acting, lower-efficacy drug than the current standard of care. The semaglutide and tirzepatide data in reproductive-age women is accumulating, and the picture will be meaningfully clearer in 3–5 years.
What we know now is enough to make thoughtful clinical decisions. What we know now is also not enough to be overconfident. The frame I use with patients is this: the reproductive benefits for women with PCOS appear real and meaningful; the pregnancy safety data is incomplete and warrants caution; and the planning conversations that prevent inadvertent exposures are entirely within our control if we have them proactively.
Sources
- Lu YT, et al. Efficacy of Liraglutide on Metabolic and Reproductive Outcomes in Women With Polycystic Ovary Syndrome: A Systematic Review and Meta-Analysis. Diabetes Obes Metab. 2026.
- Moiz A, et al. The Expanding Role of GLP-1 Receptor Agonists: A Narrative Review. EClinicalMedicine. 2025.
- Teede HJ, et al. Recommendations From the 2023 International Evidence-Based Guideline for the Assessment and Management of Polycystic Ovary Syndrome. J Clin Endocrinol Metab. 2023.
- American Diabetes Association. Pharmacologic Approaches to Glycemic Treatment: Standards of Care in Diabetes — 2026. Diabetes Care. 2026.
- Wyckoff JA, et al. Preexisting Diabetes and Pregnancy: An Endocrine Society and European Society of Endocrinology Joint Clinical Practice Guideline. J Clin Endocrinol Metab. 2025.
- Drummond RF, et al. Glucagon-Like Peptide-1 Receptor Agonist Use in Pregnancy: A Review. Am J Obstet Gynecol. 2025.
- FDA. OZEMPIC Drug Label. Updated January 2026.
- Walter MT, et al. GLP-1RA Partially Alleviates Obesity-Induced Reproductive Dysfunction via the SIRT-Associated Pathway. Eur J Pharmacol. 2025.
- Apovian CM, Aronne L, Barenbaum SR. Clinical Management of Obesity — Third Edition. The Obesity Society. 2025.
- FDA. ZEPBOUND (tirzepatide) Drug Label. Updated February 2026.
- FDA. WEGOVY (semaglutide) Drug Label. Updated February 2026.
- Ozbek L, et al. Safety of GLP-1 and Dual GLP-1/GIP Receptor Agonists in Preconception, Pregnancy, and Lactation: A Systematic Review. Diabetes Obes Metab. 2026.
- Spencer JP, Thomas S, Pawlowski RH. Medication Safety in Breastfeeding. Am Fam Physician. 2022.
- Muller DRP, et al. Effects of GLP-1 Agonists and SGLT2 Inhibitors During Pregnancy and Lactation on Offspring Outcomes. Front Endocrinol. 2023.