PMOS Diagnosis Criteria: What Actually Gets Checked

The name changed in 2026. The criteria did not.

In May 2026 a global consensus process involving 56 academic, clinical and patient organisations renamed polycystic ovary syndrome polyendocrine metabolic ovarian syndrome, or PMOS, published in The Lancet. The consortium’s reasoning was that the old name was inaccurate — it implies pathological ovarian cysts that were never cysts, and it hides the endocrine and metabolic half of a condition that affects one in eight women.

What did not change: the biology, the treatment, or the diagnostic criteria. The tests your doctor orders and the thresholds they apply are the same ones as before the rename. If you want the background on the change itself, that story is here.

The criteria in current use come from the 2023 International Evidence-based Guideline, which refined the 2003 Rotterdam consensus rather than replacing it. That guideline still uses the word “PCOS” throughout — it predates the rename by three years.

The rule for adults: two of three

An adult meets the criteria when two of the following three features are present and other causes have been ruled out.

Criterion	What counts	Threshold
Ovulatory dysfunction	Irregular menstrual cycles, or confirmed anovulation via a serum progesterone level	Depends on years since your first period — see Table 2
Hyperandrogenism	Clinical signs (hirsutism, acne, female pattern hair loss) or a raised androgen level	Modified Ferriman–Gallwey score of 4–6 for hirsutism, varying by ethnicity; total and free testosterone above the lab’s validated range
Polycystic ovarian morphology	Ultrasound or anti-Müllerian hormone — one, not both	≥20 follicles in at least one ovary; or ovarian volume ≥10 mL; or an AMH above an assay-specific cut-off

The most useful thing in that table is what it lets you skip. If you have irregular cycles and hyperandrogenism, the guideline states plainly that an ovarian ultrasound is not necessary, and neither is an AMH. You already have two of three. A scan at that point adds cost and waiting time without changing the answer.

What counts as an irregular cycle

This is the criterion most often waved through with “they seem a bit irregular.” The guideline defines it by how long ago your periods started, not by age.

Time since first period	Counts as irregular
First year	Nothing — irregular cycles are normal here, part of the pubertal transition
1 to under 3 years	Under 21 days or over 45 days
3 years to perimenopause	Under 21 days, over 35 days, or fewer than 8 cycles a year
Over 1 year, any point	Any single cycle longer than 90 days
No period yet	None by age 15, or none more than 3 years after breasts began developing

Two practical notes. Regular cycles do not rule out ovulatory dysfunction — you can bleed on schedule without ovulating, and a mid-luteal serum progesterone is how that gets confirmed. And the combined pill produces a withdrawal bleed regardless, so a cycle history from years on the pill tells a clinician very little.

The androgen tests worth asking for by name

The guideline asks for total and free testosterone, with free testosterone either calculated as a free androgen index or measured by equilibrium dialysis. It specifically says laboratories should use tandem mass spectrometry (LC-MS/MS) rather than direct immunoassays, which have poor sensitivity and precision for detecting hyperandrogenism.

If testosterone comes back normal, androstenedione and DHEAS can be added, though both are less specific.

Two things distort the result. The combined pill raises sex hormone-binding globulin and suppresses androgen production, so a level taken on it is close to uninterpretable — the guideline suggests a minimum three-month washout if the test is genuinely necessary. And androgen levels well above the reference range point away from PMOS, toward causes such as congenital adrenal hyperplasia, Cushing’s syndrome or an androgen-secreting tumour.

Ultrasound or AMH — you should not have both

The 2023 guideline’s headline change was allowing an AMH blood test to replace the scan for defining polycystic ovarian morphology in adults. Either test may be used; the guideline says both should not be, specifically to limit over-diagnosis.

On ultrasound, the adult threshold is 20 or more follicles in at least one ovary, counted transvaginally. Where older equipment or a transabdominal approach makes accurate follicle counting unreliable, an ovarian volume of 10 mL or more is used instead.

AMH is also age-dependent, generally peaking between 20 and 25, and it runs lower at higher BMI and can be suppressed by recent combined pill use. A single number without that context is not a diagnosis.

What has to be ruled out first

Two of three is necessary, not sufficient. The diagnostic algorithm requires exclusion of other causes that produce the same picture — thyroid stimulating hormone, prolactin, 17-hydroxyprogesterone and FSH as standard, plus targeted testing for Cushing’s syndrome or an adrenal tumour where the clinical picture suggests it. Hypogonadotrophic hypogonadism, usually from low body fat or heavy training, is excluded clinically alongside LH and FSH.

Adolescents: both criteria, and no imaging

For anyone still within the adolescent window, the bar is higher rather than lower. Both hyperandrogenism and ovulatory dysfunction are required — two of three does not apply, because the third leg is not used at all. Ultrasound is not recommended, since there are no definitive criteria for polycystic ovarian morphology at this life stage, and AMH is not recommended either.

The evidence for that is in the numbers. In adolescents, pooled AMH sensitivity fell to 0.66 (95% CI 0.58–0.73) and specificity to 0.78 (95% CI 0.71–0.83) — well below the adult figures, and not good enough to diagnose on.

Where features are present but the criteria are not met, the guideline suggests recording an “increased risk” status and reassessing at or before eight years post-menarche. That is a holding position, not a dismissal, and it is worth asking for in writing.

Which two of three you meet decides your phenotype

The two-of-three structure allows four combinations, and those four combinations are the four phenotypes. They are not different diseases — they are different presentations, and they differ in how much metabolic and how much reproductive weight they carry.

That is the practical reason to know exactly which criteria you met rather than just the label. The four types are broken down here, and the rest of the diagnosis section assumes you know yours.

If you are still waiting

Delay is the norm, not your bad luck. In a cross-sectional survey of 1,385 women with a reported diagnosis, 33.6% said more than two years passed before they were diagnosed, and 47.1% saw three or more health professionals along the way.

Your next step

Before your next appointment, write down the length of your last six cycles — first day of bleeding to first day of the next. That single list is the piece of evidence most likely to be missing from your notes, and it maps directly onto Table 2.

Then ask for the specific tests rather than “hormone bloods”: total and free testosterone by LC-MS/MS, plus TSH, prolactin and 17-hydroxyprogesterone to exclude the alternatives. If you want a script for that conversation, including how to raise the name change without derailing the appointment, start here.