Semaglutide and GLP-1s for PCOS: what the evidence actually shows

Polycystic ovary syndrome is, at its metabolic core, a disorder of insulin resistance. Most women with PCOS carry more insulin resistance than body weight alone would predict, and that hyperinsulinemia drives the ovaries to overproduce androgens — the thread connecting the irregular cycles, the acne and hirsutism, and the trouble conceiving. So a drug class that reliably improves insulin sensitivity and drives real weight loss is an obvious thing to point at PCOS. That is the rationale for using semaglutide and other GLP-1 receptor agonists here. The question is what the human evidence actually shows — and where it stops.

The rationale: insulin resistance is the lever

Metformin has been the reflexive insulin-sensitizer in PCOS for decades, but its effect is genuinely variable and often modest.^[1] GLP-1 receptor agonists improve insulin sensitivity by a different route and add something metformin barely does: substantial weight loss. In people without PCOS, semaglutide produced roughly 15% mean weight loss in its pivotal obesity program — the magnitudes are laid out in our review of the semaglutide weight-loss trials. Because excess weight worsens insulin resistance, which worsens androgen excess, a drug that pulls on both the weight lever and the insulin lever should, in theory, unwind several PCOS features at once. The theory is sound. Now the data.

What the human evidence shows

A 2026 systematic review and meta-analysis pooled 18 randomized controlled trials of GLP-1 RAs in women with PCOS, comparing them against metformin, standard therapy or placebo. It found significant reductions in BMI, in HOMA-IR (a measure of insulin resistance), and in total testosterone.^[2] An earlier meta-analysis restricted to head-to-head trials against metformin reached a similar direction: across eight RCTs in 462 women, GLP-1 RAs edged out metformin on insulin sensitivity and BMI, with the benefit most pronounced in obese, insulin- resistant patients — though the authors were explicit that the underlying evidence was of moderate- to-low quality and therefore inconclusive.^[3] How the two drugs stack up more broadly is the subject of our metformin versus Ozempic comparison.

Cycles, ovulation and fertility

The reproductive signal is the most clinically interesting — and the thinnest. A 2026 meta-analysis of seven liraglutide RCTs in 330 overweight or obese women with PCOS found that liraglutide significantly increased menstrual frequency, lowered BMI and insulin resistance, reduced luteinizing hormone and the free androgen index, and modestly raised sex-hormone-binding globulin. Notably, the authors could not pool hard reproductive endpoints — ovulation and pregnancy rates — because too few trials reported them.^[4] For exenatide the picture is a little richer: a meta-analysis of nine RCTs in 785 women reported higher pregnancy and ovulation rates with exenatide than with metformin, alongside the expected BMI and insulin-resistance improvements — but again flagged the evidence as inconclusive and in need of larger trials.^[5] The honest summary: menstrual regularity improves fairly consistently; ovulation and live-birth data are still sparse.

The fertility and contraception caveat

This is where the story turns from “promising” to “handle carefully.” If a GLP-1 restores ovulation in a woman who had assumed she was effectively infertile, she can become pregnant unexpectedly — and GLP-1 receptor agonists are not established as safe in pregnancy and are advised to be stopped well before conception.^[6] That is the collision at the heart of the so-called “Ozempic babies” reports, which we cover in fertility and birth control on GLP-1s. There is also a specific pharmacology trap: tirzepatide can meaningfully reduce the absorption of oral contraceptives via delayed gastric emptying, so backup or non-oral contraception is recommended around dose initiation and escalation.^[7] Tirzepatide and semaglutide differ on this and other points, as our tirzepatide versus semaglutide comparison lays out.

The honest bottom line

GLP-1 receptor agonists do the two things PCOS most needs done: they lower weight and improve insulin sensitivity, and downstream of that they nudge androgens down and cycles toward regularity. The trial evidence supports those effects — modestly, over short horizons, in small studies. What the evidence does not yet support is calling any GLP-1 a PCOS treatment: none is approved for it, the reproductive-outcome data are immature, and the fertility-and-contraception considerations are real and non-trivial. This is not a cure, and it is not a self-directed experiment. It is a reasonable, off-label option to discuss with a clinician who knows your full picture — pregnancy plans included.