GLP-1 by Subgroup: PCOS, Prediabetes, NAFLD/MASH, and Cardiovascular Disease (2026)

Last updated: June 2026

Medical disclaimer: This article is for informational purposes only and is not medical advice. Several uses below are off-label and not FDA-approved for that condition. Do not start, stop, or change treatment based on this. Consult your doctor.

GLP-1 drugs started as diabetes and weight-loss medicines. But the same biology that drives weight loss touches the heart, the liver, and the hormones behind PCOS. So the evidence now spreads across many conditions, and it is far stronger for some than others.

This guide walks through four populations one at a time. For each, we give you the real trial numbers and an honest label for how solid that evidence is. Cardiovascular disease has a landmark trial behind it. PCOS does not.

A word on why this matters. "GLP-1s help with X" gets repeated everywhere online, often without saying whether "help" means a 17,000-person outcomes trial or a 30-person pilot. Those are not the same thing. We grade each use so you can tell them apart.

How well do GLP-1s work for PCOS?

The honest answer: GLP-1s help with weight and insulin resistance in PCOS, but no GLP-1 is FDA-approved for PCOS. Every use here is off-label.

PCOS (polycystic ovary syndrome) tangles together weight gain, insulin resistance, irregular periods, and trouble ovulating. GLP-1 drugs target the metabolic side hard. That can ripple into the reproductive side.

The catch is the size of the evidence. Most studies are small, short, and single-center. We do not have the giant phase 3 trials that exist for diabetes or heart disease.

One 2025 randomized trial of 100 women with PCOS compared metformin alone to metformin plus semaglutide over 16 weeks. The combination group lost more weight, had lower fasting insulin, and reported more regular periods (Reproductive Biology and Endocrinology, 2025).

Smaller observational work backs this up. In one cohort of obese women with PCOS who had failed lifestyle changes, low-dose semaglutide led to meaningful weight loss and normalized menstrual cycles in roughly 80% of patients (PMC, 2023).

Why might it work beyond weight loss? Insulin resistance sits at the core of PCOS for many women. High insulin can push the ovaries to make more androgens, which disrupts cycles. A drug that lowers insulin resistance may break part of that loop.

Some researchers report hormonal improvements larger than weight loss alone would predict. That hints at a direct effect on the ovary, but it is a hypothesis, not a settled fact. We need bigger trials to know.

A few important cautions. These trials are tiny. None were designed to prove GLP-1s improve live birth rates. And several PCOS patients take these drugs to get pregnant, which raises a real safety problem we cover below.

There are active trials trying to fill the gap. Studies like NCT06222437 and NCT05646199 are testing semaglutide against metformin and measuring ovulation, cycle regularity, and androgen levels directly (ClinicalTrials.gov, NCT05646199). Until they read out, treat PCOS use as experimental.

Trial / source	Drug	N	Outcome	Evidence strength
Repro Biol Endocrinol 2025	Semaglutide + metformin	100	More weight loss, lower insulin, more regular cycles vs metformin alone	Weak (small RCT, 16 wk)
PMC cohort 2023	Low-dose semaglutide	~30	~80% normalized menstrual cycles after lifestyle failure	Very weak (observational)
NCT06222437, NCT05646199	Semaglutide	ongoing	Ovulation, menstrual regularity, androgens (not yet read out)	Pending

Bottom line for PCOS: promising signals on weight, insulin, and periods. Far from proof. Off-label, full stop.

Can GLP-1s prevent or reverse prediabetes?

The honest answer: yes, the evidence here is strong. Semaglutide reversed prediabetes in most people who took it in trials. This is one of the better-studied uses.

Prediabetes means your blood sugar is high but not yet in the diabetes range. Left alone, many people with prediabetes progress to type 2 diabetes. GLP-1 drugs appear to slow or reverse that slide.

The STEP program (the phase 3 trials behind Wegovy) tracked people with prediabetes at the start. In a pooled analysis, about 80% of those on semaglutide reverted to normal blood sugar by the end, versus 37% on placebo (Diabetes Care, 2022).

STEP 10 was built specifically to test this. After 52 weeks of semaglutide 2.4 mg, 81% of participants saw their prediabetes resolve, versus 14% on lifestyle changes alone (Lancet Diabetes & Endocrinology, 2024).

There is a real-world catch worth flagging. STEP 10 also followed people for 28 weeks after stopping the drug. Normal blood sugar held in only 44% of the semaglutide group off treatment, versus 18% of the lifestyle group (Lancet Diabetes & Endocrinology, 2024).

So the effect is large but partly tied to staying on the drug. Stop it, and some of the benefit fades, along with the weight loss.

There is a bigger-picture point too. Type 2 diabetes is progressive and hard to reverse once it sets in. Catching people at the prediabetes stage and pulling them back to normal blood sugar is exactly when an intervention can do the most good.

The semaglutide signal here is consistent across multiple STEP trials, not a one-off. In STEP 1, 3, and 4, few participants with prediabetes who took semaglutide had progressed to type 2 diabetes by the end, and most had returned to normal blood sugar. That consistency is part of why we grade it strong.

Trial	Drug	Population	Result	Evidence strength
STEP program pooled (2022)	Semaglutide 2.4 mg	Prediabetes at baseline	~80% reverted to normal glucose vs 37% placebo	Strong
STEP 10 (2024)	Semaglutide 2.4 mg	Obesity + prediabetes	81% remission at 52 wk vs 14% lifestyle	Strong (dedicated RCT)
STEP 10 off-treatment (2024)	Semaglutide 2.4 mg	Same	44% stayed normal off drug vs 18%	Strong (durability caution)

Note: Wegovy is FDA-approved for weight management, not for "treating prediabetes" as a stand-alone label. The prediabetes benefit is well documented but rides along with the weight-loss indication.

What do GLP-1s do for NAFLD/MASH (fatty liver)?

The honest answer: the evidence is strong and growing. Semaglutide now has positive phase 3 liver-histology data, but it is not yet FDA-approved for MASH. The only approved MASH drug is a different medicine entirely.

NAFLD (now often called MASLD) is fat buildup in the liver. MASH (formerly NASH) is the more serious form, with inflammation and scarring (fibrosis). It can lead to cirrhosis.

For contrast, set the bar first. In March 2024, the FDA approved resmetirom (Rezdiffra) as the first drug for MASH with moderate-to-advanced fibrosis. It is a thyroid-hormone-receptor drug, not a GLP-1 (FDA, 2024).

Now the GLP-1 data. The ESSENCE phase 3 trial tested semaglutide 2.4 mg in 800 people with MASH and F2-F3 fibrosis over 72 weeks. MASH resolved in 62.9% on semaglutide versus 34.3% on placebo (NEJM, 2025).

ESSENCE also hit its fibrosis endpoint. Liver scarring improved in 36.8% on semaglutide versus 22.4% on placebo. This was the first GLP-1 trial to clear both histology bars (NEJM, 2025).

Tirzepatide showed similar promise earlier. In the phase 2 SYNERGY-NASH trial, MASH resolved in up to 73.3% at the 15 mg dose versus 13.2% on placebo over 52 weeks (NEJM, 2024).

Retatrutide, the triple agonist still in development, posted the most dramatic liver-fat numbers. In a phase 2a substudy, liver fat fell about 82% on the 12 mg dose, and roughly 90% of participants cleared their fatty liver (Nature Medicine, 2024). But that measures fat, not the harder histology endpoints.

Why the standout liver-fat numbers? Retatrutide hits a third target, the glucagon receptor, which tells the liver to burn its own fat stores. Pure GLP-1 drugs do not engage that pathway directly. A dedicated phase 3 program for MASH is now underway to test whether the fat drop turns into real fibrosis improvement.

One thing to keep straight across all these drugs: liver fat, MASH resolution, and fibrosis improvement are three different things. Cutting fat is the easiest. Resolving inflammation is harder. Reversing scarring is hardest. ESSENCE matters because semaglutide moved the two hard endpoints, not just fat.

Drug	Trial (phase)	Key result	Approved for MASH?	Evidence strength
Resmetirom (not a GLP-1)	MAESTRO-NASH (3)	First FDA-approved MASH drug, F2-F3	Yes (Mar 2024)	Strong (approved)
Semaglutide	ESSENCE (3)	MASH resolution 62.9% vs 34.3%; fibrosis 36.8% vs 22.4%	No (off-label)	Strong (positive phase 3)
Tirzepatide	SYNERGY-NASH (2)	MASH resolution up to 73.3% vs 13.2%	No (off-label)	Moderate (phase 2)
Retatrutide	Phase 2a substudy	~82% liver-fat drop; ~90% cleared fatty liver	No (investigational)	Early (fat, not histology)

So GLP-1s look like real liver drugs now. But if you want an approved MASH medicine today, that is resmetirom, not semaglutide.

How much do GLP-1s reduce cardiovascular risk?

The honest answer: this is the strongest evidence of all. Semaglutide cut major heart events 20% in a landmark 17,604-person trial, and it earned an FDA heart indication. No off-label asterisk here.

The SELECT trial enrolled 17,604 adults with overweight or obesity and established cardiovascular disease, but no diabetes. They got semaglutide 2.4 mg or placebo on top of standard care.

The result was a 20% reduction in MACE (cardiovascular death, heart attack, or stroke). The composite event hit 6.5% on semaglutide versus 8.0% on placebo (NEJM, 2023).

That trial changed the label. In March 2024, the FDA approved Wegovy to reduce the risk of cardiovascular death, heart attack, and stroke in adults with heart disease who also have overweight or obesity (FDA, 2024). It was the first weight drug to carry a heart benefit on its label.

Tirzepatide brought its own heart data in a specific group. The SUMMIT trial tested it in 731 people with HFpEF (heart failure with preserved ejection fraction) and obesity. It cut worsening heart-failure events by 46% and improved quality of life (NEJM, 2024).

One nuance from SELECT worth knowing. The heart benefit showed up early and did not depend much on how much weight a person lost. That hints the drug protects the heart beyond just shrinking body weight.

Breaking the 20% down helps. Non-fatal heart attack risk dropped about 28%, cardiovascular death about 15%, and non-fatal stroke about 7%. The heart-attack signal carried the most weight in the composite.

The SUMMIT result deserves a second look too. HFpEF, the stiff-heart form of heart failure, has been notoriously hard to treat. For years almost nothing moved its outcomes. A 46% cut in worsening heart-failure events, plus better exercise tolerance and lower inflammation, is a genuinely big deal for that group.

The two trials together tell a story. Semaglutide protects the heart broadly in people with known cardiovascular disease. Tirzepatide carved out a specific, hard-to-treat slice and helped there too.

Trial	Drug	Population	Result	Evidence strength
SELECT (2023)	Semaglutide 2.4 mg	CVD + overweight/obesity, no diabetes	20% MACE reduction (6.5% vs 8.0%)	Very strong (landmark, FDA label)
SUMMIT (2024)	Tirzepatide	HFpEF + obesity	46% fewer worsening HF events	Strong (dedicated RCT)

If you only trust one use on this whole list, it is this one.

Which drug has the best evidence for which condition?

This master table is the centerpiece. It maps each drug to each condition and grades the evidence honestly. "FDA-approved" means approved for that exact use. Everything else is off-label or investigational.

Condition	Semaglutide (Wegovy/Ozempic)	Tirzepatide (Mounjaro/Zepbound)	Retatrutide (investigational)	Best-evidence pick
Cardiovascular (CVD)	FDA-approved; SELECT 20% MACE cut	Strong in HFpEF (SUMMIT, 46%)	No outcomes data yet	Semaglutide (approved + landmark)
Prediabetes	Strong; ~80% reverted (STEP)	Likely class effect; less direct data	No dedicated data	Semaglutide (STEP 10)
NAFLD/MASH	Strong phase 3 (ESSENCE), off-label	Strong phase 2 (SYNERGY-NASH)	Huge liver-fat drop, early phase	Semaglutide (positive phase 3)
PCOS	Weak, off-label (small RCTs)	Very thin off-label data	None	Semaglutide, but evidence is weak
Type 2 diabetes	FDA-approved (Ozempic)	FDA-approved (Mounjaro)	Phase 3 ongoing	Tirzepatide (higher A1C/weight effect)
Obesity/weight	FDA-approved (Wegovy)	FDA-approved (Zepbound)	Largest losses in phase 2	Tirzepatide today; retatrutide later

Reading the table: semaglutide owns the FDA-approved heart and weight uses and has the deepest off-label liver and prediabetes data. Tirzepatide edges it on raw diabetes and weight numbers and owns the HFpEF niche. Retatrutide looks strongest on paper for liver fat but has no late-stage outcome data yet.

A practical note on the grades. "FDA-approved" is the highest bar because it means a regulator reviewed the full safety and efficacy package for that exact use. "Strong" off-label means good trial data exists but the label has not caught up. "Weak" means the studies are small or observational. Match your expectations to the grade.

One more honest caveat. Class effects are real but not guaranteed. Just because semaglutide helped prediabetes does not prove tirzepatide does it identically. Where we wrote "likely class effect," that is an informed guess, not proven equivalence.

For more on retatrutide and the drugs behind it, see our next-gen GLP-1 pipeline guide.

What are the contraindications and cautions across these groups?

Each population carries its own risks. Here are the ones that matter most.

PCOS patients trying to conceive. This is the big one. GLP-1 labels advise stopping the drug before pregnancy because of possible harm to a developing fetus.

The exact washout window depends on the drug, since they clear at different speeds. Semaglutide is usually stopped about two months before trying to conceive. We cover the timing in our GLP-1 pregnancy washout guide.

The irony is sharp. A GLP-1 may restore ovulation in PCOS, then must be stopped right when fertility returns. Plan this with a doctor, not on your own.

Personal or family history of medullary thyroid cancer or MEN 2. Semaglutide and tirzepatide carry a boxed warning here, based on rodent data. These drugs are contraindicated in those patients (FDA Wegovy/SELECT, 2024).

History of pancreatitis. GLP-1 drugs have been linked to pancreatitis cases. Anyone with prior pancreatitis should discuss the risk carefully before starting.

MASH with advanced cirrhosis. The liver trials studied F2-F3 fibrosis, not full cirrhosis. Evidence does not yet extend to decompensated cirrhosis, so do not assume the ESSENCE numbers apply there.

Heart failure patients. SUMMIT was encouraging for HFpEF with obesity. It does not mean every heart-failure patient should take a GLP-1. The data is specific to that group.

Across all of these, side effects skew gastrointestinal: nausea, vomiting, diarrhea, constipation. Most are mild and fade after the dose-up phase, but they drive a lot of people to quit.

What the evidence does and does not support by indication

A plain-language scorecard to close out.

Cardiovascular: supported, and approved. SELECT is a true landmark, and the FDA label backs it. This is the most defensible use outside diabetes and weight.

Prediabetes: supported, with a durability catch. Reversal rates are high on the drug. Some benefit fades when you stop, so it is not a one-and-done cure.

NAFLD/MASH: supported by phase 3 data, but off-label. ESSENCE is real and positive. Still, the only FDA-approved MASH drug today is resmetirom, not a GLP-1.

PCOS: not yet supported by strong evidence. Small trials hint at weight, insulin, and cycle benefits. There are no large trials and no FDA approval. Treat any PCOS use as experimental.

What the evidence does NOT support: using these drugs as proven fertility treatments, as cures that work after you stop, or as approved MASH or PCOS therapies. Match the drug to the strength of its data, and talk to your doctor. You can also browse our medications directory for the approved-use details on each drug.

Frequently Asked Questions

Is semaglutide FDA-approved for PCOS?

No. There is no FDA-approved GLP-1 for PCOS. Any use for PCOS is off-label. Small trials suggest it can help with weight, insulin resistance, and menstrual regularity, but the evidence is far thinner than for heart disease or diabetes. Talk to a doctor who knows your full history.

Can a GLP-1 reverse my prediabetes?

In trials, yes, for most people while they take it. In STEP 10, 81% of people with prediabetes returned to normal blood sugar after 52 weeks on semaglutide, versus 14% on lifestyle alone. But some of that benefit faded after stopping the drug, so it is not a permanent fix on its own.

Which GLP-1 has the best evidence for fatty liver (MASH)?

Semaglutide has the strongest GLP-1 evidence, with positive phase 3 ESSENCE data showing MASH resolution in 62.9% of patients. Tirzepatide also looks strong in phase 2. But neither is FDA-approved for MASH. The only approved MASH drug is resmetirom (Rezdiffra), which is not a GLP-1.

How much does semaglutide reduce heart attack and stroke risk?

In the SELECT trial of 17,604 people with heart disease and excess weight, semaglutide cut major cardiac events by 20% versus placebo. That includes cardiovascular death, heart attack, and stroke. The FDA approved Wegovy for this use in March 2024, making it the strongest evidence on this list.

Should I stop my GLP-1 if I want to get pregnant?

Yes, this is standard advice. GLP-1 labels warn against use in pregnancy and recommend stopping before trying to conceive, often about two months ahead for semaglutide. This matters most for PCOS patients, since the drug may restore ovulation right as it needs to be stopped. Plan the timing with your doctor.

Related Reading

• Retatrutide: complete evidence review
• Next-gen GLP-1 pipeline 2026
• Orforglipron: complete evidence review

-- The GLP-1 Daily Team