PCOS Has a New Name. Your Testing Is Still Running on 2003 Logic.

In May this year, the syndrome I've met the criteria for since my twenties got a new name. Nothing about how I would have been tested for it changed at all.

Then in May 2026, an international group of more than fifty patient and professional organisations, coordinated by a team led by Helena Teede at Monash University, formally renamed PCOS to PMOS: Polyendocrine Metabolic Ovarian Syndrome. A different expansion of the same three letters, because the old name was too narrow for what the condition actually is.

The reason why this happened? Most women carrying this diagnosis don't actually have cysts on their ovaries. What they have is multiple small follicles stuck in an arrested state. It feels like a more honest name for what's actually happening in the body.

I want to be precise about what did and didn't happen here, because the distinction matters more than the headline suggests.

A New Name, Eleven Years in the Making

The renaming process took eleven years and input from roughly 22,000 people worldwide, doctors, researchers, patients, and advocacy groups, and was endorsed by more than fifty organisations including the Endocrine Society and the American Society for Reproductive Medicine. The reasoning, as ASRM put it, was that "polycystic ovary syndrome" reduced a complex, whole-body endocrine and metabolic condition to a description of what the ovaries look like on an ultrasound, a description that's usually inaccurate at that. The new name puts the metabolic and endocrine features first, and drops the reference to cysts that mostly aren't there. Some of the researchers involved have also pointed to evidence that the same underlying metabolic and endocrine drivers show up in men, though that piece of the picture is still an active area of research rather than settled fact.

This is a real shift in how the condition is understood, and I don't want to undersell it. But the same reporting that announced the change was careful to note what it doesn't do: it doesn't immediately change diagnostic criteria, treatment guidelines, or clinical terminology in most settings. Patients should expect to keep seeing PCOS on insurance forms, referral letters, and prescriptions for a long while yet. The label has moved. The tests underneath it haven't.

The Rotterdam Problem

Whatever you call the condition, it's still diagnosed the same way it has been since 2003, when a consensus meeting in Rotterdam settled on a two-out-of-three rule: irregular or absent ovulation, clinical or biochemical signs of high androgens, and polycystic ovarian morphology on ultrasound. Meet two of the three and you qualify.

That framework was carried into a 2018 international evidence-based guideline, and then reissued again in 2023, when the same network of professional and patient organisations reconvened to update its recommendations for assessing and managing PCOS. It's worth being precise about what that update actually changed, because it's less than the date suggests. The guideline reaffirmed the same two-out-of-three structure drawn up in 2003. The one substantive change was allowing anti-Müllerian hormone, a blood test, as an alternative to an ultrasound for the ovarian morphology criterion. Where irregular cycles and clinical hyperandrogenism are both already present, the guideline says you don't need imaging or AMH at all.

So the most recent formal review of how this condition gets diagnosed, arriving the same year the metabolic and endocrine understanding of it hit international headlines, kept the twenty-year-old architecture intact and added one extra way to satisfy one of its three boxes. That's a defensible, evidence-led decision on its own terms. It's also a long way from what "endocrine and metabolic condition" implies a diagnostic pathway should look like. If PMOS is fundamentally about insulin signalling, androgen production, and whole-system hormonal regulation, a framework built almost entirely around ovarian appearance and menstrual regularity was already a strange fit before the rename. It's a stranger fit now.

This is the gap I actually want to spend this piece on. Not the name. The testing underneath it, because that's the part you have some influence over, regardless of what your file says at the top.

One Diagnosis, Several Different Drivers

Even before the rename, PMOS was never really one condition. Under the Rotterdam rule's own two-out-of-three logic, there are four different ways to qualify for it, formally described as phenotypes A through D depending on which combination of anovulation, high androgens, and polycystic ovarian morphology someone meets. Phenotypes A and B, both anovulatory and hyperandrogenic, carry the highest average metabolic risk. Phenotype D, ovulatory dysfunction and ovarian morphology without measurable hyperandrogenism, carries the lowest. Four meaningfully different metabolic pictures, filed under the same diagnostic label.

Clinically, the picture fractures further than even that. Integrative and functional practitioners increasingly describe presentations by their dominant physiological driver rather than by which Rotterdam boxes get ticked: elevated insulin driving ovarian androgen production in the most commonly seen presentation, chronic low-grade inflammation disrupting hormonal signalling in another, a temporary post-contraceptive rebound in a third, and adrenal-driven androgen excess in a fourth. This four-driver framework isn't part of the formal Rotterdam classification. It's a working clinical model rather than a diagnostic category, but the adrenal presentation specifically has real support underneath it in the peer-reviewed literature.

This is my own presentation. My androgen excess reads as adrenal rather than ovarian, DHEA-S elevated, testosterone unremarkable, and rather than the insulin resistance that gets assumed by default with androgen-driven PMOS, I run toward the opposite end of that spectrum: high insulin sensitivity. That's not a contradiction, and it's not rare. Elevated DHEA-S has been shown to correlate inversely with insulin resistance in PCOS cohorts, with adrenal androgen excess associated with a more favourable insulin and lipid profile than presentations with normal adrenal androgens. Elevated DHEA-S also shows up more often in the non-classic B and C phenotypes than in phenotype A, the presentations least likely to carry the textbook insulin-resistant picture.

For anyone with an androgen-driven presentation, prolactin is worth testing specifically alongside this. Elevated prolactin can suppress ovulation and raise androgen levels through an entirely separate mechanism, mimicking this picture rather than confirming it, and standard diagnostic guidance requires ruling it out before an androgen-driven diagnosis is treated as settled. It's also one of the markers most likely to be left off a rushed initial panel.

The point isn't to replace one label with four better ones. It's that a single testing protocol, applied uniformly, will systematically miss whichever driver isn't the one it was built around. Insulin testing alone won't explain an adrenal presentation. A panel built around ovarian markers alone won't catch an inflammatory one. Knowing which driver you're actually looking at changes what's worth testing for, and what the results, once you have them, actually mean.

Gap One: Ranges Built to Catch Disease, Not Dysfunction

Something I say often, because it comes up in nearly every set of bloodwork I review: standard reference ranges are built through statistical correlation with diagnosable disease. A marker doesn't get flagged until it's moved far enough to be reliably associated with a named condition. That's a defensible way to build a range if the goal is ruling disease in or out. It's a poor way to build a range if the goal is catching a system under strain before it tips into disease, which is most of what a proper hormone investigation is actually trying to do.

Androgen testing sits squarely inside this problem. Total testosterone reference ranges are wide enough to comfortably contain a woman with clinically significant hyperandrogenism, because the range was built to separate "clearly pathological" from "everything else," not to describe optimal or even typical female physiology. Free testosterone and SHBG, the two markers that actually tell you how much androgen is biologically available to tissue, are ordered far less often than total testosterone alone, largely because total testosterone is cheaper and technically permitted as a first-line test. AMH carries a version of the same problem from a different angle: different assay platforms produce meaningfully different values for the same blood sample, and there's still no consensus cutoff for what counts as a "polycystic" AMH result across laboratories. A result that reads as unambiguous polycystic morphology on one platform can sit comfortably in range on another.

None of this means the tests are useless. It means a single result, read against a population-derived range, is a much blunter instrument than it's usually presented as being.

Gap Two: The Right Test, the Wrong Day

Hormones are not static. LH, FSH, testosterone, and AMH all move meaningfully across the cycle, and some of them move meaningfully across the day too. The most useful baseline for LH, FSH, and androgens is drawn early in the follicular phase, cycle day two to five, before the rest of the cycle's hormonal cascade gets underway. The elevated LH-to-FSH ratio long associated with PCOS, generally a ratio above two to three, is specifically an early follicular phase finding. Draw it later in the cycle and the ratio can look completely different, not because anything about the underlying physiology has changed, but because you're reading a different part of the same story.

Testosterone adds a layer to this. Serum testosterone in women shows measurable diurnal variation, generally higher in the morning than the evening, on top of the cycle-day effect. A test drawn at 4pm on cycle day eighteen is measuring a different physiological moment to one drawn at 8am on cycle day three, and both can come back inside a lab's technically normal range while telling two very different stories about what's actually happening.

For women with the irregular or absent cycles that are themselves part of the diagnostic picture, this gets harder still. There's no reliable "day three" to aim for. In practice, this often means testing happens whenever the appointment lands, with no attempt to establish where in a cycle, or in the absence of one, that sample actually sits. The test is frequently the right one. The day it's taken on is rarely interrogated, and it changes what the result means.

Gap Three: One Point Cannot Describe a Pattern

The deeper issue underneath both of the gaps above is that most hormone testing happens once. One draw, one moment, one number to compare against a range.

Androgens and gonadotropins aren't secreted in a smooth, steady stream. LH is released in pulses, superimposed on top of the cycle-day and diurnal patterns already described, which means two blood draws taken an hour apart from the same woman on the same day can differ substantially. Progesterone is the clearest example of why this matters clinically. Progesterone is released episodically from the corpus luteum, correlating closely with LH pulses, and a single mid-luteal progesterone value is an unreliable way to assess whether the corpus luteum is functioning adequately, even when it's timed correctly. A low reading might mean anovulation. It might mean you caught a trough between pulses on an otherwise adequate cycle. One number can't tell you which.

This is the same principle I come back to constantly when I'm reading a full blood panel with a client: a single marker outside range is a data point. A pattern, repeated across a marker family or across cycles, is information. Ovulatory status confirmed once means very little on its own. Ovulatory status tracked across two or three cycles, alongside basal body temperature and cycle-timed bloodwork, starts to describe an actual system rather than a snapshot of it.

What a Proper Investigation Actually Looks Like

None of this is an argument against testing. It's an argument for testing with enough precision that the result means something.

A cycle-timed panel for PMOS generally includes early follicular (day two to five) LH, FSH, total and free testosterone, SHBG, DHEA-S, and prolactin, alongside a mid-luteal progesterone drawn roughly seven days after confirmed ovulation rather than on a fixed calendar day. Since the new name puts metabolic function on equal footing with the endocrine and ovarian picture, fasting insulin and glucose belong in the same panel, not as an afterthought. Thyroid function is worth including as a matter of course, both because thyroid dysfunction can produce a similar symptom picture and because the two conditions overlap more often than chance alone would predict. Where AMH is used, it's worth knowing which assay your lab runs and treating the result as one data point among several rather than a definitive read on ovarian morphology.

Where cycles are irregular enough that timing a draw is hard to pin down, basal body temperature charting over a full cycle or two adds something a single blood test structurally cannot: a continuous record of whether and when ovulation is actually happening, rather than an inference from one moment in time. Ovulation predictor kits are a reasonable alternative, and easy enough to pick up from a chemist, but they're trickier to use well. You need a rough sense of when you're likely to ovulate before you start testing, otherwise you can burn through a lot of strips without ever catching the LH surge.

I've written a full guide that walks through exactly this, how to time each hormone to the right point in your cycle, what a functional or optimal range looks like rather than just a technically normal one, and how to read the patterns across a full panel rather than one marker in isolation. It also covers the rest of a standard blood panel, so the sex hormones section isn't read apart from everything else your body is telling you.

Get the guide, How to Interpret Your Blood Test Results, here →

The Name Was Never the Hard Part

Renaming a condition that affects more than 170 million people worldwide took eleven years and input from 22,000 people. That's not a small undertaking, and the people who did that work deserve real credit. But naming a condition accurately and testing for it accurately are two different projects, and only one of them has actually happened yet.

Until testing practice catches up with what the new name already acknowledges, that this is a whole-system endocrine and metabolic condition and not simply an ovarian one, most of the responsibility for getting an accurate picture sits with the person doing the asking. That means requesting the fuller panel, asking what day a test was drawn on and why, and treating any single result as one data point in a pattern rather than a verdict.

I wrote previously about the gap between what PMOS genetics actually tell us and what "it's genetic" usually implies, which covers some of the same territory from a different angle: predisposition is a starting point for investigation, not an endpoint. If you're working through a PMOS diagnosis, old name or new, and want a proper look at the full picture, testing, timing, and what's actually driving your particular presentation, I work with women one-on-one to put this together properly.