Zepbound and PCOS: What I Found When I Went Looking for the Evidence

My sister sent me a photo of the box

Two weeks ago my sister texted me a picture of a Zepbound carton sitting on her kitchen counter, with one line underneath it: does this actually do anything for the PCOS, or is it just a weight drug?

She has had PCOS since she was nineteen, and has been handed metformin, the pill, and a printout about lifestyle change roughly once a year ever since. I am the one in the family who reads the studies, so I said give me a few days.

Here is what I came back with, and the honest version is thinner than either of us wanted.

PCOS is not on the label. Not anywhere on it.

I started with the prescribing information, because that is the one document that is legally obliged to be accurate.

Zepbound is approved for two things: reducing excess body weight and maintaining that reduction long term in adults with obesity, or with overweight plus at least one weight-related condition — and treating moderate to severe obstructive sleep apnea in adults with obesity¹. That is the entire list.

The word polycystic does not appear in that document. I searched it — not as an indication, not as a studied subgroup, not in a footnote. So if PCOS is why it was prescribed to you, that is your prescriber's clinical judgement, not something the label backs up.

What it is proven to do

Weight. And the numbers are not small.

In the 72-week trial the approval rests on, adults were randomised to 5 mg, 10 mg, 15 mg or placebo. At week 72 the 630 adults on 15 mg had lost 20.9% of their body weight. The 643 on placebo had lost 3.1%¹ — the placebo arm was not nothing, because everybody in both groups also got a roughly 500-calorie-a-day deficit and exercise counselling.

The responder rates, with their denominators: 90.9% of the 630 on 15 mg lost at least 5% of their body weight, against 34.5% of the 643 on placebo¹. That is adults only. Safety and effectiveness have not been established in anyone under 18¹.

Then I went looking for the PCOS evidence

A review published in Drugs in May 2026 mapped the incretin drugs against PCOS one by one, and its verdict on tirzepatide — the molecule in Zepbound — was a sentence I read twice: it has no PCOS-specific evidence, and cannot be recommended beyond extrapolation from the obesity and diabetes trials².

I did not want to just repeat that, so I ran the search myself. One human trial exists, and it landed three weeks after that review went out.

Sixty overweight or obese women with PCOS were randomised for 16 weeks to metformin alone, or metformin plus 5 mg of tirzepatide once a week. The combination group lost 10.4 kg; the metformin-only group lost 1.7 kg. Menstrual cycle recovery and total pregnancy rate were both higher in the combination group³.

Genuinely encouraging, and I still made my sister read the caveats with me. No placebo arm — the comparison was metformin, an active drug. Sixteen weeks. Open-label, so everyone knew who was getting what. Sixty women is sixty women. And the dose was 5 mg alongside metformin, not Zepbound titrated on its own.

The drug with the actual PCOS trials is not the one she was handed

This is the part that surprised me most. Sort the PCOS literature by how much randomised evidence exists per drug and the answer is liraglutide — the older daily injection nobody talks about anymore.

A 2026 systematic review pooled seven randomised trials covering 330 women with PCOS. Liraglutide increased menstrual frequency, lowered BMI, improved insulin resistance and lowered the free androgen index⁴. The authors are careful about it: heterogeneity for the menstrual finding was high, and ovulation and pregnancy could not be pooled at all because too few of the trials reported them.

So the honest ranking is liraglutide first, semaglutide second, tirzepatide last with one small open-label trial. Zepbound is not the best-studied PCOS drug. It is the best-studied weight drug, being taken by women who have PCOS.

The thing I made her promise before she started

This is the part I would tape to a fridge.

Zepbound delays gastric emptying, and its label tells women on oral hormonal contraceptives to switch to a non-oral method, or add a barrier method, for four weeks after starting and for four weeks after every single dose increase¹. Contraception you do not swallow is not affected.

Half the women I know with PCOS are on the pill for the cycle or the acne, and have never once thought of it as their birth control. Meanwhile the weight loss that improves PCOS is the same thing that can bring ovulation back. My sister was exactly that woman. One phone call to her prescriber sorted it.

When to stop reading and call somebody

Before you start, ask whether it is reasonable at all given your history, what the contraception plan is, and whether your metformin dose changes. Once you are on it, call for severe abdominal pain that goes through to your back, vomiting you cannot stop, or a positive pregnancy test — the label says weight loss offers no benefit in pregnancy and to stop the drug once a pregnancy is recognised¹.

I am a mom who takes this medication and reads the research, not a clinician. This is education, not medical advice.

What she said back

I told her to take it if her prescriber and her wallet agree, and to stop expecting a PCOS drug. Expect a weight drug. If the cycle follows, that is the weight loss reaching her hormones — a real thing that happens, which nobody has proven this particular molecule does.

She read all of that, thought about it for a minute, and texted back: so I'm the study.

I have not had a good answer to that yet.