Semaglutide and GLP-1s for PCOS: The Evidence

The short version

Polycystic ovary syndrome is, for many women, a weight-and-insulin problem as much as a reproductive one — and GLP-1 medications act directly on that machinery. The randomized evidence shows they produce real but modest weight loss in women with PCOS. A 2026 systematic review and meta-analysis screened 9,654 studies and included 11 randomized controlled trials; GLP-1 receptor agonists added on top of usual care reduced BMI by a mean difference of 1.38 kg/m² (95% CI 0.38 to 2.39), rated low certainty².

What the same evidence does not show is an effect on the things women most want fixed. The same review found no difference in LDL cholesterol or triglycerides, and rated the evidence insufficient to draw a conclusion on glucose, insulin, hirsutism and menstrual regularity — and found no studies at all assessing quality of life, mental health or cost-effectiveness². And no GLP-1 is FDA-approved for PCOS: the Wegovy label's indications cover cardiovascular risk reduction, weight reduction and MASH, with no mention of polycystic ovary syndrome⁷. Every use here is off-label.

Why GLP-1s are in the PCOS conversation at all

The 2023 international evidence-based PCOS guideline positions lifestyle and, where weight is a driver, weight reduction as foundational, with medications considered alongside them rather than instead of them, because losing weight can improve ovulation, insulin resistance and metabolic markers¹. GLP-1 and dual GIP/GLP-1 drugs entered the picture because they are the most effective pharmacological weight-loss tools available: in STEP 1, 1,961 adults lost a mean 14.9% of body weight over 68 weeks on semaglutide 2.4 mg versus 2.4% on placebo⁵, and in SURMOUNT-1, 2,539 adults lost a mean 20.9% on tirzepatide 15 mg over 72 weeks versus 3.1%⁶. Both enrolled by BMI in general obesity populations. Neither enrolled women because they had PCOS.

What the PCOS-specific evidence actually shows

Three things about that table deserve saying out loud.

Chen 2025 is a genuine randomized trial, and it is small and open-label. One hundred women with PCOS by Rotterdam criteria were randomized to metformin 1000 mg twice daily, or the same metformin plus semaglutide 1 mg weekly, for 16 weeks. Eighty completed — an 80% completion rate. The combination arm lost meaningfully more weight and showed greater improvements in testosterone and CRP³. The pregnancy figures are the ones to handle carefully: they come from weeks 16 to 40, after semaglutide had been stopped and both groups were on metformin alone, and 35% versus 15% is read on the 80 women who completed, not the 100 who were randomized³. That is an encouraging signal in fewer than a hundred women, not a fertility result.

Ling 2025 is not a trial. It is a systematic review and meta-analysis of 19 randomized trials in 1,657 women, comparing liraglutide plus metformin against metformin alone⁴. It reports improvements across glycemic, hormonal and lipid measures — and heterogeneity (I²) between 79% and 98% on nearly every outcome, with the authors themselves rating overall certainty as low because of risk of bias and heterogeneity⁴. Pooled numbers that unstable are a direction, not a magnitude.

The two meta-analyses agree on the shape. Weight moves. Everything downstream of weight is uncertain, and both review teams say so in their own conclusions^2,4.

The caveats every PCOS patient should hear

These drugs are off-label for PCOS; approval is for obesity, cardiovascular risk reduction or type 2 diabetes, not PCOS itself⁷. GLP-1 medications are not used in pregnancy, and improving ovulation can increase the chance of conceiving — the Wegovy label directs patients using it for weight reduction to stop at least 2 months before a planned pregnancy because of semaglutide's long half-life⁷. That is a longer runway than most people plan for, and it makes contraception and pregnancy intentions a first-appointment conversation rather than a later one.

And the trials measured group averages over months. Individual response varies, gastrointestinal side effects are common, and nothing in this literature followed women long enough to tell you what happens to PCOS after the drug stops.

What good PCOS-aware care looks like

Look for a clinician who takes a real history, checks glucose and metabolic labs, raises contraception and pregnancy intentions before the first dose, and treats the medication as one lever alongside the lifestyle base the guideline emphasizes¹ — not a program that ships a vial without asking whether you are trying to conceive.

Where this sits on the evidence scale

Weight loss in PCOS: moderate, and modest in size. Pooled randomized evidence, consistently positive, but graded low certainty by the review authors themselves and amounting to roughly 1.4 BMI points.

Testosterone and inflammatory markers: low. Single small randomized trials pointing the same way.

Cycles, ovulation and fertility: insufficient. That is the meta-analysts' own word for menstrual regularity, and the pregnancy data come from completer denominators in trials of fewer than a hundred women.

Quality of life, mental health, cost-effectiveness: no studies exist. The 2026 review went looking and found none.

What would change this: an adequately powered randomized trial in PCOS with reproductive endpoints pre-specified and analyzed on everyone randomized. Until one exists, treat any PCOS marketing built on the fertility angle as running well ahead of its data. If you're weighing the two molecules, our semaglutide vs tirzepatide for mothers guide compares them directly.