PCOS — a metabolic condition, not just an ovarian one

Polycystic ovary syndrome is the most common endocrine disorder in people of reproductive age, and its name is close to a category error. The ovaries are where the condition is most visible, but the underlying disturbance involves insulin signaling, androgen production, and the hormonal axis that governs ovulation. Treating it as an ovarian problem leads to a narrow conversation about periods and fertility, and leaves the metabolic half unaddressed for years. We approach it as what it is: a lifelong endocrine and metabolic condition with reproductive consequences.

Why the name gets in the way

The polycystic ovary is not full of cysts. What ultrasound shows is a larger than usual number of small antral follicles, each a normal structure arrested early because the signal that would select one of them to mature is disordered. Calling them cysts suggests something that should be removed or drained, which is why people occasionally arrive expecting surgery. The follicles are a consequence of the endocrine picture, not the disease itself.

The disturbance underneath involves several systems at once. Insulin resistance is present in a large share of people with PCOS, in lean bodies as well as larger ones, and high insulin levels drive the ovary to produce more androgen while lowering the protein that binds testosterone in the blood, so more of it is biologically active. Disordered pulses of luteinizing hormone add to that androgen production. The result is irregular or absent ovulation, the skin and hair changes of excess androgen, and a metabolic profile that raises long-term cardiovascular and diabetes risk. The 2023 international evidence-based guideline recognized that the name is a poor description and a source of confusion; no replacement has reached consensus, so the label remains in use while the understanding has moved on.

How the diagnosis is actually made

Diagnosis in adults uses the Rotterdam criteria, endorsed by the 2023 international guideline and by ACOG: any two of three features, after other causes have been excluded. Two of three means there is no single test and no single appearance, and it means two people with the same diagnosis can look quite different from each other, which is part of why the condition is both overdiagnosed and underdiagnosed depending on where you go.

• Ovulatory dysfunction: cycles persistently longer than 35 days or shorter than 21, fewer than eight or nine periods a year, or absent periods, reflecting infrequent or absent ovulation
• Hyperandrogenism, either clinical (hirsutism in an androgen-dependent pattern, persistent acne, scalp hair thinning) or biochemical (elevated free or total testosterone on a reliable assay)
• Polycystic ovarian morphology on ultrasound, or an elevated anti-Mullerian hormone level, which the 2023 guideline accepts as an alternative to imaging in adults

Two points about the third criterion matter. In an adult who already has both irregular cycles and hyperandrogenism, ultrasound adds nothing to the diagnosis; it may be worth doing for other reasons, but not to confirm PCOS. In adolescents it should not be used at all, because multifollicular ovaries are normal in the years after menarche and imaging simply generates false positives. Some cycle irregularity is also normal in those years, so the adolescent diagnosis rests on hyperandrogenism plus persistent irregularity, and it is often correct to call someone at risk and reassess later rather than to label them at fifteen.

What has to be ruled out before the label is applied

PCOS is a diagnosis of exclusion, and skipping this step is the most common way to get it wrong. Thyroid dysfunction and elevated prolactin both cause irregular cycles and are checked with simple blood tests. Non-classical congenital adrenal hyperplasia mimics PCOS closely, with hirsutism and irregular periods, and is screened for with an early morning 17-hydroxyprogesterone level drawn in the follicular phase; it matters because the treatment differs and because it carries implications for a future pregnancy.

Where the picture suggests it, we look further. Cushing syndrome should be considered when there is rapid central weight gain, easy bruising, proximal muscle weakness, or purple striae. Markedly high testosterone, or virilization that came on quickly with voice change or clitoromegaly, prompts evaluation for an androgen-secreting tumor of the ovary or adrenal gland. None of this is routine for everyone, but the history and examination should be specific enough to know whether it applies to you.

Treatment organized by what you want, not by the diagnosis

There is no single treatment for PCOS, and a plan built around the diagnosis rather than your goals tends to satisfy no one. What you want changes over time: contraception in one decade, conception in the next, metabolic protection throughout. We set the plan by goal and revisit it when the goal shifts.

Cycle regulation and protecting the endometrium

When ovulation is infrequent, the uterine lining is exposed to estrogen without the progesterone that normally follows ovulation and stabilizes it. Over years that unopposed exposure raises the risk of endometrial hyperplasia and endometrial cancer, and this is the single most underexplained risk in PCOS. Combined hormonal contraception is the usual first-line answer, and it addresses cycle control and androgenic symptoms together. Where estrogen is unsuitable, cyclic progestin or a levonorgestrel intrauterine device protects the lining well. A reasonable working standard is a withdrawal bleed at least every three to four months, or continuous progestin exposure in its place.

Hirsutism, acne, and hair thinning

Combined hormonal contraception is first line here too, and it works by lowering ovarian androgen production and raising sex hormone binding globulin. It is slow: hair growth cycles mean six months is the earliest fair assessment, and expecting change at six weeks leads people to abandon treatment that was working. If the response is inadequate after about six months, an antiandrogen such as spironolactone is commonly added, which requires reliable contraception because of the risk to a male fetus. Mechanical and laser hair reduction complement medical treatment rather than competing with it, and for many people the combination is what actually delivers the result.

The ovary is where PCOS shows itself. The pancreas, the liver, and the blood vessels are where it does most of its long-term work.

Insulin resistance and metabolic risk

This is the part that tends to get deferred, and it should not be. Metformin improves insulin sensitivity and often modestly improves cycle regularity, and it is a reasonable addition for metabolic features or when contraception is not wanted. Sustained changes in eating pattern, resistance training, and sleep have genuine effects on insulin sensitivity independent of weight change, which is worth saying because weight-focused advice delivered without support is the most common complaint we hear about previous care. GLP-1 receptor agonists are increasingly used where obesity and metabolic risk coexist with PCOS; the general evidence for weight and cardiometabolic outcomes is strong, the PCOS-specific evidence is still developing, and they are not compatible with pregnancy, so timing relative to fertility plans has to be explicit.

Fertility

Most people with PCOS who want to conceive can. Anovulation is the main obstacle and it is usually treatable. Letrozole is first-line for ovulation induction and outperforms clomiphene citrate for live birth in PCOS; clomiphene, metformin as an adjunct, gonadotropins, and in vitro fertilization follow in selected situations. Ovulation can also return unpredictably, so if pregnancy is not the goal, irregular cycles are not contraception. Screening for glucose intolerance before pregnancy is worthwhile, since PCOS raises the risk of gestational diabetes and of hypertensive disorders of pregnancy.

The long view, and why we start early

PCOS carries a raised lifetime risk of impaired glucose tolerance and type 2 diabetes, dyslipidemia, hypertension, obstructive sleep apnea, and metabolic dysfunction-associated fatty liver disease, alongside the endometrial risk described above. Rates of depression and anxiety are also higher, and the guideline recommends screening for both rather than waiting for you to raise them. None of this is a forecast of what will happen to you, and the risks are modifiable to a meaningful degree, but they accumulate quietly over decades. That is why the metabolic conversation belongs at the first visit rather than at the one where a fasting glucose has already gone wrong.

In practice that means a baseline assessment of glucose tolerance, ideally an oral glucose tolerance test rather than a fasting glucose alone, plus lipids and blood pressure, repeated every one to three years depending on your results and risk factors. It also means naming the condition properly, because people who understand they have a metabolic condition tend to make different decisions over twenty years than people who have been told they have cysts on their ovaries.