Is Metformin Safe in Pregnancy? What the PCOS Trials Found for Mother and Child
Metformin has not been clearly linked to birth defects, but in PCOS pregnancies it did not prevent gestational diabetes, and trial children exposed in the womb had higher BMI and more allergies and eczema. The 2023 guideline does not recommend it routinely in pregnancy.
By Kymata Health Editorial Published Last checked
This guide
PCOS & Cycle- Question
- Is it safe to keep taking metformin during pregnancy if I have PCOS?
- Sources
- 7 cited
- Quick answers
- 3 questions
- Clinician review
- None; sources cited inline
Key takeaways
4 points- In the PregMet2 trial (487 pregnant women with PCOS), late miscarriage or preterm birth occurred in 5% on metformin vs 10% on placebo, a difference that was not statistically significant; gestational diabetes was 25% vs 24%2.
- Children exposed to metformin in the womb in the earlier PregMet trial had a higher BMI z-score at 5 to 10 years3, and a 2025 follow-up analysis found more allergies and eczema at 8 years4.
- The FDA label says published studies have not reported a clear link between metformin and major birth defects or miscarriage, but those studies cannot rule out risk5.
- The 2023 international guideline says metformin is not routinely recommended for pregnant women with PCOS, though it could be considered in some circumstances, such as a risk of preterm birth1.
Why this question comes up
Many women with PCOS take metformin, and its label notes it may restore ovulation in some women who were not ovulating, so an unplanned pregnancy is possible5. In its IVF section, the 2023 guideline suggests stopping metformin at the time of the pregnancy test or period unless it is otherwise indicated1.
Unlike insulin, metformin crosses the placenta, and fetal levels are comparable to the mother's6. That is why the question is not only whether it causes birth defects, but what it does to the pregnancy and to the child later.
Birth defects and miscarriage
The FDA label says limited data are not enough to determine a drug-related risk of major birth defects or miscarriage, and published studies have not reported a clear association5. It adds that those studies cannot definitely rule out risk because of small samples and inconsistent comparison groups5. For context, the label puts the US background risk at 2% to 4% for major birth defects and 15% to 20% for miscarriage in recognized pregnancies5.
What the PCOS pregnancy trials found
PregMet2 randomized 487 pregnant women with PCOS at 14 hospitals in Norway, Sweden and Iceland to metformin (up to 1,000 mg twice daily) or placebo from the late first trimester until delivery2. Late miscarriage or preterm birth occurred in 12 of 238 women on metformin (5%) and 23 of 240 on placebo (10%), an odds ratio of 0.50 that was not statistically significant (p=0.08)2. Gestational diabetes was 25% vs 24%, and serious adverse events did not differ substantially2.
When the authors pooled PregMet2 with two earlier trials in a post hoc analysis, the combined outcome was 5% vs 10%, which was statistically significant2. The trial was funded by the Research Council of Norway, the Novo Nordisk Foundation, St Olav's University Hospital and the Norwegian University of Science and Technology2.
The guideline summarizes that metformin in pregnant women with PCOS has not been shown to prevent gestational diabetes, late miscarriage, high blood pressure in pregnancy, preeclampsia or large babies1. It could be considered in some circumstances, such as a risk of preterm birth, to reduce preterm delivery and limit excess weight gain, and its side effects are mostly mild stomach symptoms that are not worse in pregnancy1.
Effects on children later
In a 5 to 10 year follow-up of the earlier PregMet trial, 141 of 255 invited children took part (55%); those exposed to metformin had a higher BMI z-score than those exposed to placebo (difference 0.41)3. A 2025 post hoc analysis of the PregMet trials (634 women, 145 children) found fewer infections in mothers on metformin but more allergies (odds ratio 4.83) and eczema (odds ratio 2.42) in exposed children at 8 years4. These were secondary analyses; the PedMet follow-up had the same public and foundation funders as PregMet223.
A 2025 meta-analysis of 18 studies (7,975 exposed children, any maternal reason for metformin) found no overall difference in BMI at an average age of 4.4 years, but in the PCOS subgroup exposed children were heavier, with low-certainty evidence7. Two of its authors reported ties to industry, including in-kind metformin and placebo from Apotex for another trial and honoraria from Novo Nordisk7. The guideline asks clinicians to tell women that long-term effects on children remain unclear, with a suggestion of higher childhood weight, though causality is not certain1.
Breastfeeding
Metformin passes into breast milk: lactation studies on the label estimate the infant dose at about 0.11% to 1% of the mother's weight-adjusted dose5. The label says there is not enough information on effects in breastfed infants, so the benefits of breastfeeding should be weighed with the mother's need for the drug5.
Safety notes and when to see a clinician
Do not stop a prescribed medicine on your own after a positive test; call the prescriber. The decision is different if you have type 2 or gestational diabetes, because poorly controlled diabetes itself raises risks for mother and baby5. For PCOS fertility options, see letrozole vs Clomid and our prenatal vitamin guide.
- Who should not take metformin: people with severe kidney impairment (eGFR below 30) or a known allergy to it5.
- Get urgent care for possible lactic acidosis, which the label says often starts with vague symptoms such as feeling generally unwell, muscle pain, trouble breathing, sleepiness or stomach pain5.
- Tell your prescriber if you are pregnant or planning to be, so the plan for metformin is decided before or early in pregnancy.
Quick answers
Questions
3 answersShould I stop metformin when I get a positive pregnancy test?
Ask your prescriber rather than stopping on your own. The 2023 guideline says metformin is not routinely recommended for pregnant women with PCOS but could be considered in some circumstances, and the answer is different if you have diabetes15.
Does metformin prevent miscarriage in PCOS?
The guideline says it has not been shown to prevent late miscarriage1. In PregMet2, the combined rate of late miscarriage and preterm birth was lower on metformin, but not significantly2.
Can I take metformin while breastfeeding?
It passes into breast milk at low levels, and the label says effects on breastfed infants are not well studied, so weigh it with your clinician5.
More in PCOS & Cycle
Keep reading
6 pages- PCOS & Cycle hubEvery guide, regulatory notes and sources
- How PCOS (PMOS) Is Diagnosed: Cycle Length, Androgens and the AMH Test
- Birth Control Pill vs Metformin for PCOS: What the 2023 Guideline Recommends
- Inositol for PCOS: What the Trials Show for Cycles, Weight and Fertility
- Letrozole vs Clomid for PCOS: Which Works Better for Getting Pregnant?
- Berberine vs Metformin for PCOS: What the Trials Show, and Why the Evidence Is Shaky
Keep reading: Supplements and vitamins
Supplements and vitamins: every pageSupplement & Nutrient EncyclopediaBrowse all topics
- Previous in this topicBerberine vs Metformin for PCOS: What the Trials Show, and Why the Evidence Is ShakySmall trials reported berberine matching or beating metformin on waist size and cholesterol in PCOS.
- Next in this topicPCOS (Now PMOS) and Irregular Cycles: Diagnosis, Treatment and Supplements, by the EvidencePCOS was renamed polyendocrine metabolic ovarian syndrome (PMOS) in 2026, but how it is diagnosed has not changed.
- Women's Hair Loss: Minoxidil, Spironolactone, Laser Caps and Supplements, by the EvidenceTopical minoxidil has the best evidence for female pattern hair loss, and low-dose oral minoxidil matched it in a pooled analysis of trials.
- SpearmintSpearmint (Mentha spicata) is a mint used as a tea for unwanted hair growth in PCOS and as a rosmarinic-acid extract for memory.
- InositolInositol is a sugar-like compound the body makes and also gets from food; myo-inositol is the form sold most for PCOS and pregnancy.
References
Every study and document cited in this guide, numbered as in the text.
Sources
7 sources- 1Recommendations From the 2023 International Evidence-based Guideline for the Assessment and Management of Polycystic Ovary Syndrome. J Clin Endocrinol Metab, 2023. PMID 37580314
- 2Use of metformin to treat pregnant women with polycystic ovary syndrome (PregMet2): a randomised, double-blind, placebo-controlled trial. Lancet Diabetes Endocrinol, 2019. PMID 30792154
- 3Intrauterine metformin exposure and offspring cardiometabolic risk factors (PedMet study): a 5-10 year follow-up of the PregMet randomised controlled trial. Lancet Child Adolesc Health, 2019. PMID 30704873
- 4Metformin Treatment in PCOS Pregnancies Reduces Maternal Infections and Increases the Risk of Allergies and Eczema in the Offspring: Post Hoc Analyses of Two Randomised Controlled Trials and One Follow-Up Study. BJOG, 2025. PMID 40785326
- 5Glucophage and Glucophage XR (metformin hydrochloride) tablets: prescribing information. U.S. Food and Drug Administration (Drugs@FDA), 2018.
- 6Metformin use in pregnancy: What about long-term effects in offspring?. Acta Obstet Gynecol Scand, 2024. PMID 38757307
- 7Intrauterine metformin exposure and adiposity outcomes in children: a systematic review and meta-analysis. BMJ Open, 2025. PMID 39779258
Dietary supplements are not intended to diagnose, treat, cure or prevent any disease, and these statements have not been evaluated by the FDA. Prescription options need a licensed clinician. Talk to your clinician or pharmacist before starting anything new, especially if you take prescription medicines, are pregnant or breastfeeding.