PMOS stands for polyendocrine metabolic ovarian syndrome. Until May 2026 it was called PCOS, polycystic ovary syndrome, and most people still call it that. Your diagnosis letter says PCOS. Your doctor's computer says PCOS. Nothing about the condition changed. The name did.

The ovary is where it shows up most visibly. The cause sits elsewhere, mostly in how your body handles insulin, which is why the rest of this page spends more time on your liver and your pancreas than on your ovaries.

A doctor diagnoses it when two of three things are true: your periods are irregular or absent, your body makes more androgen than expected, and your ovaries hold an unusually high number of small follicles on an ultrasound. A blood test for AMH can stand in for the scan.² Those criteria have not changed and are not changing.

Where the old name came from

In 1935 two Chicago doctors, Irving Stein and Michael Leventhal, wrote up seven women whose periods had stopped and whose ovaries were enlarged and studded with small round structures. They called them cysts. The paper was titled Amenorrhea associated with bilateral polycystic ovaries, and the name held for ninety years.³

They were wrong about the cysts. Those structures are follicles: ordinary egg sacs, two to nine millimetres across, each one still holding an egg. They stalled before release. A cyst is a fluid-filled sac with nothing inside it.

The name failed in a second way. Plenty of women with the condition do not have those follicles at all. Naming a condition after a feature many patients lack sends doctors looking for the wrong thing, and it sent women home undiagnosed for decades.

Why it changed, and why that matters

An expert panel at the US National Institutes of Health called for a new name in 2012. Fourteen years passed before one arrived.⁴

What was different in 2026 was who got a vote. The process ran through 56 patient and professional organisations across six continents. In the first round, 9,358 patients voted alongside 3,656 doctors and researchers. Patients outnumbered clinicians more than two to one. The name they picked took 66% of first-place votes.⁴

Medicine has a long habit of naming women's conditions after what a doctor saw through an instrument, and an even longer habit of settling what those women should be called without asking any of them. This time the people living with it were most of the room.

The gap this is meant to close is measurable. In a survey of 1,385 women across 32 countries, a third waited more than two years for a diagnosis. Nearly half saw three or more doctors before anyone named it. Fewer than one in four were satisfied with what they were told about treatment, and more than half received nothing at all about long-term complications.⁶

If you read one section on this page, read this one. Almost everything below it follows from here.

Insulin is the hormone that moves sugar out of your blood and into your cells. In PMOS, muscle, fat and liver cells stop answering it properly. Your pancreas sees sugar sitting in the blood and does the obvious thing: it makes more insulin. Blood sugar stays normal. Insulin does not.

Now the part that explains the rest. Your ovary never stopped listening. Insulin has a second job besides handling sugar, which is telling cells to make hormones, and the ovary stayed fully responsive to that one. A body flooded with insulin is a body telling the ovary, hour after hour, to make more testosterone.

This is measured, not guessed. The most accurate test of insulin sensitivity is a clamp study, where researchers hold insulin and glucose at set levels and watch what the body does with them. A review pooling 28 clamp studies across 1,224 women with PMOS found insulin sensitivity 27% lower than in women without it. Carrying more weight lowered it roughly 15% further. The 27% gap was still there in women with a BMI around 22.⁷

That last detail is the one that should end the standard advice. Thin women with PMOS have the same insulin problem. Weight makes it worse. Weight did not cause it.

The six-week proof

In 1997, nineteen lean women with PMOS took metformin, a drug that lowers insulin, for four to six weeks. Their free testosterone fell by around 70%. The enzyme their ovaries use to make androgen quietened down. SHBG, a protein we will come back to, rose from 84 to 134. Nothing moved in the placebo group, and nobody lost weight.⁸

Lower the insulin and the ovary makes less testosterone. Six weeks, no weight change. Nothing else on this page has a demonstration that clean.

Genetic work points the same way. By using inherited gene variants as a natural experiment, which sidesteps the usual problem of not knowing which thing came first, researchers found that higher fasting insulin causes PMOS rather than the reverse. Low SHBG and higher body weight came out as causes too.⁹

Androgens get called male hormones, which is a poor label. Every woman makes them and needs them. In PMOS there is too much.

Androgen is a category, not one hormone. Five of them circulate. Ranked by how much sits in your blood: DHEAS, DHEA, androstenedione, testosterone, DHT. Only the last two bind the androgen receptor and do anything. The three above them are raw material, so the ranking by amount runs close to the reverse of the ranking by power.²⁷

Testosterone is the one your doctor measures. DHT is the other active one, and it is made out in the tissue rather than in a gland, converted from testosterone in skin and hair follicles.²⁷ Which is why two women with the same testosterone reading can have very different skin.

Most of the excess comes from the ovary. The adrenal glands, which sit on top of your kidneys, supply the rest: around 20 to 30% of women with PMOS have raised adrenal androgens as well.¹⁰

The ovary in PMOS is not a normal ovary being pushed hard. Take the cells that make androgen from a woman with PMOS, grow them in a dish with no insulin anywhere near them, and they still make more androgen than cells from a woman without it. The tendency is built in. Insulin turns up the volume on something already loud.

This is what produces the symptoms you can see in a mirror. Acne that shrugs off normal treatment. Hair arriving where you would rather it did not, and thinning where you would rather it did not. Periods that turn up whenever they feel like it.

A great many women with PMOS have a testosterone test land inside the normal range while every symptom says otherwise. There is a reason, and it is not that the lab got it wrong.

Testosterone travels through your blood stuck to a protein called SHBG, which your liver makes. About 65% of your testosterone is attached to it at any moment. Only one or two per cent floats free, and free testosterone is the only kind that can act on a hair follicle or an oil gland. The rest is in transit and cannot do anything.

A standard blood test measures the total. If your SHBG is low, your total can read perfectly normal while the free portion has doubled.

So what lowers SHBG. Insulin does. High insulin drives your liver to make fat, and that fat interferes with the specific switch the liver uses to produce SHBG. Less SHBG, more free testosterone, and no change at all in the number on your results letter.¹¹

This is the same mechanism running backwards in that 1997 trial, where lowering insulin pushed SHBG from 84 up to 134 while free testosterone dropped.⁸

The liver shows up again, more seriously. Across 36 studies covering 5,021 women, 43% of women with PMOS had fatty liver disease, against roughly a quarter of the general population.¹² Almost nobody gets told this at diagnosis.

In a typical month a batch of follicles starts growing, one pulls ahead of the others, and that one releases an egg. In PMOS the batch grows and nobody pulls ahead.

Two things block the selection.

AMH, which is not only a marker

Follicles produce AMH, and women with PMOS have two to four times more of it in their blood. That is not simply because there are more follicles. Each follicle makes more. Follicle cells taken from women with PMOS who were not ovulating produced 75 times more AMH in culture than cells from women who were.¹³

AMH stops the follicle responding to FSH, the hormone that tells it to grow, and switches off the enzyme that converts testosterone into oestrogen. A follicle that cannot make oestrogen cannot be chosen. It sits there at two to nine millimetres and waits.

Twenty or more of them in one ovary meets the diagnostic threshold. A typical ovary has around eight.²

The pulse that runs too fast

Your brain releases the signal controlling LH and FSH in pulses, and the speed sets the mix. Fast pulses make LH. Slow pulses make FSH. In PMOS the pulses run fast and stay fast, so you get more LH pushing the ovary to make androgen, and less FSH helping follicles grow.

Progesterone is what normally slows those pulses. Progesterone only appears after ovulation. No ovulation, no progesterone, no brake.¹⁴

Worth flagging: the strongest evidence here comes from sheep exposed to excess androgen before birth, which reproduce the fast-pulse pattern almost exactly and stop responding to progesterone. The human data agrees but is thinner.¹⁴

Put the last three sections together and you have a loop.

High insulin pushes the ovary to make androgen. Androgen and low FSH stall the follicles. Stalled follicles mean no ovulation. No ovulation means no progesterone. No progesterone means nothing slows your brain's pulses, so LH stays high, so the ovary keeps making androgen, so next month's follicles stall too.

No point on that circle is the beginning. That is why it does not sort itself out, and why the useful question is not what started it but where you can interrupt it. Insulin is the point with the most evidence behind it and the most room to move.

Weight is part of this and the usual conversation about it is the wrong one. What the research keeps landing on is not how much fat but how the fat behaves.

Compared with women of the same BMI who do not have PMOS, fat cells in PMOS are larger. Androgens make it harder for the body to create new fat cells, so the ones that exist overfill. Overfilled fat cells leak fatty acids and stop answering insulin, and the surplus fat ends up in the liver and muscle, where it causes more damage than it would have done in the first place.

Of everything measured in one review, fat cell size had the strongest negative relationship with insulin sensitivity, and it held after adjusting for both BMI and androgen levels.¹⁵

Which routes straight back to the liver, and to insulin.

PMOS runs in families. A Dutch study comparing 1,332 identical twins with 1,873 non-identical twins and sisters put heritability at about 70%.¹⁶

A 2025 study across Chinese and European populations found 94 regions of the genome involved, 73 of which nobody had reported before. It also pointed at granulosa cells, the follicle cells from the AMH section, as where those genes do their work.¹⁷

Those 94 regions account for only a small slice of the 70%. Most of the inheritance has not been located yet, which is worth saying because anyone selling you a genetic explanation for your own case is ahead of the science.

High heritability does not mean fixed. The same genetic methods showing PMOS is inherited also show that insulin, body weight and SHBG cause it, and all three move. Your genes set how easily the loop starts. They do not decide whether it keeps running.

Everything above is settled enough to sit in a textbook. This section is not. These are ideas with real research behind them that the field has not agreed on, and most of them get sold with far more confidence than the evidence carries.

Inflammation as a cause

CRP, a marker of inflammation, is raised in women with PMOS, and raised regardless of weight. A review of 63 studies covering 7,206 women found it clearly elevated, and more elevated in non-obese women than in obese ones.¹⁸

What that same review did not find was any significant effect on TNF-α, the inflammatory messenger with the clearest mechanical link to insulin resistance. CRP is a downstream marker rather than a mechanism, and a genetic pathway analysis found inflammation is not a major contributor.

Inflammation is real in PMOS and it probably makes insulin resistance worse. Calling it the root cause goes past the evidence. You will see it called the root cause constantly.

The gut microbiome

A 2025 review screened 4,377 studies and included 29. Its finding was inconsistent diversity across studies, with some reporting an increase, some a decrease and some no change at all. It graded the certainty of the evidence as low.¹⁹

Diet, body weight and metformin all reshape the gut microbiome, and a large share of women in these studies take metformin. There may well be something here. Nobody can currently say what it is.

Stress, cortisol and "adrenal PCOS"

You will see PMOS sorted into types, one of them blamed on stress and high cortisol. Cortisol control is normal in PMOS. ACTH, the hormone that tells the adrenal gland what to do, sits at normal levels and responds normally.¹⁰

What is real is that some women's adrenal glands are more sensitive to a normal signal, and specifically on the androgen side rather than the cortisol side. No recognised classification of PMOS contains a stress type.

Stress still matters. Genetic analysis suggests depression raises the risk of PMOS.⁹ Not by the route you have been told.

Whether AMH helps cause it

AMH is a good enough marker that the 2023 guideline accepts it instead of an ultrasound. Whether it helps cause the condition is open. The 2025 genome study flagged the AMH gene as causal.¹⁷ The Mendelian randomisation analysis found genetically predicted AMH is not causal.⁹ Both are good studies and they disagree.

One related claim worth killing: high AMH does not mean you are insulin resistant. Across 22 studies the relationship between the two comes out at 0.089, which is close to none.

Creatine

Creatine is having a moment in women's health. It is among the most studied supplements in existence, it costs very little, and its safety record in healthy people is good. Several well-known researchers have spent the past two years arguing that women were left out of creatine research for decades. They are right about that.

None of that work is about PMOS.

We went looking for trials of creatine in women with PMOS. There are none published. There are none registered either: as of September 2026, neither ClinicalTrials.gov nor the EU Clinical Trials Register lists a single study giving creatine to women with this condition.²⁰ Every recommendation you will read is borrowed from a different population. Here is what is being borrowed, and how well it travels.

Take blood sugar first. One trial gave 5 g of creatine a day to 25 people with type 2 diabetes for twelve weeks. Their HbA1c fell from 7.4 to 6.4 while the placebo group went from 7.5 to 7.6.²¹ That is a real result, with two catches. Both groups were also doing supervised exercise, so it shows creatine adding to training rather than doing the work itself. And it has never been repeated. In healthy people, several trials found no effect on glucose or insulin sensitivity at all.

Then there is the creatine kinase argument. You will sometimes read that women with PMOS have disturbed creatine metabolism, which would make topping it up sensible. The only study to measure it looked at 660 women in Qatar and found creatine kinase did not differ between women with PMOS and women without. It rose with body weight in both groups equally.²² That is the most PMOS-specific piece of creatine evidence in existence, and it points away from the idea.

Androgens are the part that matters most here, because PMOS is defined partly by having too many of them and because hair thinning is one of the symptoms women mind most. A 2009 trial in 20 male rugby players found creatine raised DHT, a potent androgen, by 56% during loading.²³ That study is where creatine's hair-loss reputation comes from. A 2025 trial in 45 resistance-trained men measured hormones and actual hair density across twelve weeks and found nothing.²⁴

So the scare did not replicate. Now notice the populations. Every androgen measurement in the creatine literature was taken in men. We could not find one study measuring testosterone or DHT in women taking creatine, in any group, let alone in women whose androgens are already high. For a condition where that is the central problem, the reassurance is entirely secondhand.

On mood, creatine added to an antidepressant produced better remission than the antidepressant alone in 52 women with major depression.²⁵ Worth knowing, given how often depression accompanies PMOS, though it is a trial in diagnosed depression rather than in this condition. On thinking and memory, EFSA reviewed 23 human trials in 2024 and concluded that no cause and effect has been established.²⁶

Which leaves a plausible idea with a decent safety record, no evidence in this condition, and one unexamined question about what it does to androgens in women. Creatine may well turn out to be useful here. Saying so today is a guess.

Whether PMOS is one condition at all

Andrea Dunaif's objection to the new name is also a scientific position. Her argument is that the genetics describe several distinct conditions that happen to share symptoms, and that studying them as one holds the field back. If she turns out to be right, some of the research on this page is an average of things that should have been measured separately.

This is the most interesting unanswered question in the field, and it will take another decade to settle.