How to Lose Weight With PCOS: What Works, in Order of Evidence

The short answer

Start with a calorie deficit you can keep and regular activity, because the 2023 international PCOS guideline says no single diet beats another¹. If your BMI is 25 or above, metformin is the guideline's first add-on¹. GLP-1 medicines produce the largest losses in PCOS trials. Inositol is the weakest rung: low harm, little effect on weight⁴.

Elena Voss is a mother writing for mothers, not a treating clinician, and holds no medical license. Every clinical figure here is cited inline to its primary source, the trial or the FDA label, so you can check it. This is background reading, not medical advice.

Why PCOS makes weight harder

PCOS is common. The guideline puts it at 10% to 13% of reproductive-aged women worldwide¹. For many of them, weight and the condition feed each other. Insulin resistance pushes the ovaries to make more androgens, higher androgens favor fat around the middle, and that central fat worsens insulin resistance. The loop is the reason a diet that works for a friend can seem to do half as much for you, and it is also the reason the drugs that help most here are the ones aimed at insulin and appetite.

None of this is a willpower story. The guideline goes out of its way to say that many women with PCOS meet weight stigma in health care and that clinicians should be aware of it¹. If an appointment has left you feeling blamed, that is a known failure of care, not a verdict on you.

The evidence ladder

Here is how the main options compare, using the best pooled evidence for each. The numbers come from different reviews with different comparison groups, so read across a row, not down the table as a league.

### Lifestyle comes first, and its evidence is thin

A Cochrane review pooled 15 randomized trials of diet, exercise and behavioral programs in 498 women with PCOS. Lifestyle treatment reduced weight by about 1.7 kg and BMI by about 0.3 points compared with minimal or no treatment, and it lowered the free androgen index². The authors rated the evidence low quality, and none of the trials measured live birth or regular cycles².

That modest number is not a reason to skip the base. It is the platform every other option is tested on top of. The guideline asks for at least 250 minutes a week of moderate activity, or 150 minutes of vigorous activity, plus muscle strengthening on two days, when the goal is modest weight loss and preventing regain¹. For a mother with small children, that can be three brisk 30-minute walks plus a few short sessions with the stroller or a kettlebell, spread through the week. On food, the guideline recommends a tailored energy deficit and healthy eating that suits your preferences rather than any branded diet¹.

Metformin: the guideline's first medicine

A meta-analysis written to inform the 2023 guideline compared metformin with placebo across randomized trials. With moderate certainty, metformin lowered BMI by about 0.53 points and improved insulin resistance and fasting glucose. It also raised the odds of mild gastrointestinal side effects about sevenfold³. The guideline turned that into a clear line: metformin alone should be considered in adults with PCOS and a BMI of 25 or more for weight and metabolic outcomes¹.

In plain terms, metformin is a small push on the scale and a meaningful one on blood sugar. It is cheap and generic, and starting low and going slow usually tames the stomach side effects.

Inositol vs metformin

Inositol is the supplement most women with PCOS hear about first. The review commissioned for the guideline pooled 30 trials with 2,230 participants and called the evidence for inositol "limited and inconclusive." Metformin may beat inositol on waist-hip ratio and hirsutism, the evidence on BMI was very uncertain, and myo-inositol caused fewer stomach side effects than metformin⁴.

The guideline's wording follows that closely. Inositol in any form could be considered based on your preferences, with limited harm but limited benefit for ovulation, hirsutism or weight, and metformin should be considered over inositol for hirsutism and central fat¹. It also notes that no specific type, dose or combination can be recommended, and that supplement quality can vary¹. A 2026 network meta-analysis of 16 trials agreed: myo-inositol with folic acid trended toward weight loss but did not reach statistical significance⁵.

If you like the idea of inositol, it is a reasonable add-on, not a replacement for the rungs above it. Tell your clinician you are taking it.

GLP-1s, and taking metformin with a GLP-1

This is where the larger numbers sit. The 2026 network meta-analysis set every option against placebo. GLP-1 medicines alone cut weight by about 5.2 kg, and GLP-1 plus metformin by about 5.6 kg, the largest effect in the network. No option clearly improved the insulin-resistance score, and the authors note they did not study cycles or fertility⁵. A separate 2026 meta-analysis of 11 trials found GLP-1s lowered BMI by about 1.4 points compared with control, rated low certainty, with too little data to judge cycles or hirsutism⁶. Our GLP-1 evidence guide for PCOS walks through those trials one by one.

Taking metformin with a GLP-1 is common and is the combination most of these trials tested. The guideline lists liraglutide and semaglutide among anti-obesity medicines that could be considered alongside active lifestyle change, with gradual dose increases to limit nausea¹. It also flags two cautions to discuss up front: the likely need for long-term use, because weight tends to return after stopping, and the lack of long-term safety data¹.

Zepbound or Ozempic for PCOS?

Neither drug is approved for PCOS, and no trial has compared them head to head in women with PCOS. Most PCOS trials used semaglutide or liraglutide; our Ozempic for PCOS page covers the semaglutide side and the approval question.

The tirzepatide (Zepbound) data are newer. In a 16-week randomized trial of 60 women with PCOS, low-dose tirzepatide (5 mg weekly) plus metformin cut weight by about 10.4 kg, against 1.7 kg on metformin alone. Cycles returned more often and pregnancy rates were higher in the combination group⁷. The trial was small, short and open-label. A UK cohort of 54,114 women prescribed tirzepatide found those reporting PCOS lost about 19% of their weight at 10 months, no different from women without PCOS. That figure comes from only 40 women with PCOS still in follow-up at that point, in an observational study⁸. For how the two molecules compare in general, see semaglutide vs tirzepatide for mothers.

When cycles come back: the birth-control catch

As weight falls, ovulation can return, sometimes before your period is regular. That is welcome news if you want a baby and a surprise if you don't. The guideline asks clinicians to make sure effective contraception is in place for women on GLP-1s when pregnancy is possible, because pregnancy safety data are lacking¹. Tirzepatide adds a second issue: it can make the pill less reliable for weeks after starting and after each dose increase. Our birth control guide has the four-week rule, and trying to conceive on a GLP-1 covers when to stop before trying. If you are breastfeeding, the injectable GLP-1s have no human milk data, so the decision means weighing the benefits of breastfeeding against your own need for the drug with your clinician.

Where this sits on the evidence scale

Lifestyle change rests on low-quality trials but is the base of every recommendation. Metformin has moderate-certainty evidence for a small weight effect. Inositol's evidence is low to very low and shows little weight benefit. GLP-1s show the largest losses, but the PCOS trials are small and short, and the tirzepatide data are one small randomized trial and one cohort. What would change this ranking: larger, longer randomized trials of semaglutide and tirzepatide in PCOS that measure cycles, fertility and weight regain after stopping.