GLP-1 Drugs and PCOS Fertility: A New Review Finds a Natural-Pregnancy Signal, Not an IVF Answer

A woman starts a GLP-1 drug to lose weight. Her cycles become more regular. Then she finds herself asking a question the weight-loss headlines rarely address: could the treatment change her chances of getting pregnant? A new review of randomized trials offers a cautiously interesting answer. It found a signal for natural pregnancy in women with polycystic ovary syndrome (PCOS), but no clear improvement in overall pregnancy or IVF pregnancy. That distinction is the story. The result is not evidence that GLP-1 medicines are fertility drugs, and it is not a reason to take one while trying to conceive. It is a small, early clinical signal that deserves a careful look.

The result: one outcome moved, two did not

Published online in October 2026 in the European Journal of Obstetrics & Gynecology and Reproductive Biology, the systematic review and meta-analysis pooled four reports representing three randomized trials and 252 participants. The researchers compared GLP-1 receptor agonist-based treatment, sometimes combined with metformin, against metformin alone in women with PCOS. The database search ran through February 2026. For natural pregnancy, the pooled odds ratio was 3.10 (95% confidence interval 1.74–5.49). The authors rated certainty for this outcome as moderate and reported that trial sequential analysis supported the signal. For overall pregnancy, the odds ratio was 1.41 (0.81–2.45): not statistically significant, with evidence considered underpowered. For IVF pregnancy, the estimate was 1.01 (0.36–2.85), also not significant. The confidence intervals are wide, especially for the latter outcomes. “No significant difference” here does not establish that treatments are equivalent; it means this small evidence base could not resolve the question.

Why natural conception and IVF are separate questions

Natural conception depends on a person ovulating and having intercourse during a fertile window, among other factors. IVF adds ovarian stimulation, egg retrieval, fertilization in a laboratory, embryo development and transfer. A treatment that shifts weight, insulin resistance or cycle regularity might influence the first set of steps without changing IVF success after an embryo transfer. Those are not interchangeable endpoints. Nor does a higher natural-pregnancy odds ratio tell us how many additional people will have a live birth: the review’s primary outcome was pregnancy, not a pooled live-birth result. The authors explicitly called for larger trials with standardized outcomes, including live birth.

The odds ratio also should not be read as “three times as many pregnancies” in every clinic or patient group. It is a relative measure from pooled trials, not an individual forecast. Absolute chances depend on the people studied, comparison treatment, follow-up period and clinical circumstances. The studies were small and not all used identical drugs or protocols. Their results cannot be converted into a personal probability from the abstract alone.

What the component trials can—and cannot—tell us

The newer synthesis is important because it focuses on pregnancy endpoints, but it does not turn the underlying trials into a large definitive program. One of the primary studies, a randomized open-label trial published in 2025, assigned 100 overweight or obese Chinese women with PCOS to metformin alone or metformin plus weekly semaglutide for 16 weeks; 80 completed the intervention. Afterward, both groups received metformin during follow-up. The combination group lost more weight during treatment (about 6.09 kg versus 2.25 kg) and had a higher reported natural-pregnancy rate during follow-up (35% versus 15%). The trial was small, open-label, and restricted to a specific population. It does not show that semaglutide itself caused conception, that the result applies to people without overweight, or that the same outcome occurs with every GLP-1 medicine.

Older randomized research includes exenatide comparisons with metformin and a small study of liraglutide plus metformin before IVF. These trials are useful context, not proof of a class-wide fertility effect. The 2026 meta-analysis combines reports from only three trial populations (252 people in total), and its estimates should be read as preliminary. It does not establish which specific drug, dose, treatment duration or patient profile is responsible for the natural-pregnancy signal. It also does not identify a direct ovarian mechanism. Improvements in weight or insulin sensitivity may contribute, but the review cannot prove that pathway.

PCOS is more than weight—and the evidence has limits

PCOS is a heterogeneous endocrine and metabolic condition. Not everyone with PCOS has the same symptoms, body size, insulin resistance or fertility goals. The included trials largely concern women with overweight or obesity, so results should not be generalized automatically to lean PCOS or to all reproductive-age patients. Small studies can produce unstable estimates, and differences in follow-up and fertility care complicate comparisons. This is why the moderate-certainty rating for natural pregnancy should not be inflated into a practice-changing recommendation. The evidence for overall pregnancy was low certainty; the IVF result also remains uncertain.

There is a further practical distinction: medicines may affect the chance of ovulation, while the timing of exposure matters for a developing pregnancy. GLP-1 drugs are not established pregnancy treatments. The 2023 international evidence-based PCOS guideline says clinicians should ensure effective contraception when pregnancy is possible for women taking GLP-1 receptor agonists because pregnancy-safety data are lacking. It also says anti-obesity medicines may be considered for higher weight in adults with PCOS according to general population guidance, alongside active lifestyle intervention. Neither point means a GLP-1 drug should be used as a fertility prescription.

Conception planning is a separate safety conversation

For semaglutide, current U.S. prescribing information says to discontinue at least two months before a planned pregnancy because of its long washout period. That interval is product-specific; do not assume it applies identically to every GLP-1 medicine. If someone becomes pregnant while taking a GLP-1 drug, they should contact their prescribing clinician promptly for individualized advice rather than making medication changes from a blog post. A clinician can also plan how to manage the condition being treated while the medicine is stopped.

Stopping a medicine can bring its own trade-offs, including return of appetite, weight or glucose-related problems for some patients. A preconception plan should therefore be coordinated with the clinician managing PCOS, obesity or diabetes and, where relevant, a fertility specialist. The study results do not replace standard infertility evaluation, established ovulation-induction options, or individualized IVF counseling. They also do not support buying unapproved products or changing a prescribed dose to try to increase fertility.

What a useful next study would look like

The next step is not another headline built around a small percentage. Researchers need adequately powered randomized trials that define the population clearly, separate natural conception from assisted reproduction, and follow participants through pregnancy and live birth. Studies should report miscarriage and maternal and neonatal outcomes, explain when treatment stopped relative to conception, and compare like-for-like medicines and care pathways. They should also track what happens after withdrawal, since the period between stopping treatment and trying to conceive is clinically important.

Until those data exist, this review is best treated as a signal for researchers and a conversation point for clinicians—not a new fertility protocol. It supports asking better questions: whether restoring metabolic health changes spontaneous conception for some people with PCOS; which patients might benefit; how durable the effect is after treatment stops; and what the safest transition looks like. It does not answer those questions yet.

The Oria take

The finding is specific and worth attention: in three small randomized-trial populations, GLP-1-based treatment was associated with higher odds of natural pregnancy, while pooled evidence did not show a significant difference in overall or IVF pregnancy. The headline should preserve both halves. This is not proof that GLP-1 medicines improve fertility for everyone with PCOS, and it is not evidence they are safe to continue during pregnancy. People considering pregnancy should discuss contraception, medication timing and alternatives with their healthcare team; semaglutide’s two-month preconception discontinuation instruction comes from its own U.S. label.

Evidence grade: C. A recent RCT-only meta-analysis reports a moderate-certainty natural-pregnancy signal, but it pools only three trials and 252 participants. Overall pregnancy remains underpowered; IVF pregnancy shows no significant difference; live-birth evidence is not established. This is promising, limited evidence—not a fertility indication.

Sources: Kar E, et al. “The effect of glucagon-like peptide-1 receptor agonists versus metformin on pregnancy rates in women with polycystic ovary syndrome: a systematic review and meta-analysis of randomized trials.” European Journal of Obstetrics & Gynecology and Reproductive Biology, published online 15 October 2026. PubMed: https://pubmed.ncbi.nlm.nih.gov/42893679/ | DOI: https://doi.org/10.1016/j.ejogrb.2026.115499

Chen H, et al. “Effects of combined metformin and semaglutide therapy…in women with PCOS.” Reproductive Biology and Endocrinology (2025). https://doi.org/10.1186/s12958-025-01447-3 | 2023 International Evidence-based PCOS Guideline, ASRM: https://asrm.org/practice-guidance/practice-committee-documents/recommendations-from-the-2023-international-evidence-based-guideline-for-the-assessment-and-management-of-polycystic-ovary-syndrome | U.S. DailyMed Ozempic prescribing information: https://dailymed.nlm.nih.gov/dailymed/drugInfo.cfm?matchtype=e&setid=adec4fd2-6858-4c99-91d4-531f5f2a2d79

Medical disclaimer: Oria provides educational information, not medical advice, diagnosis, or treatment. This article is not a recommendation to start, stop, or change medication. GLP-1 medicines are not established fertility treatments and should not be used during pregnancy unless specifically directed by a qualified clinician. Discuss PCOS, fertility goals, contraception, pregnancy planning, and medication timing with your licensed healthcare professional.