If you have cycles longer than 35 days that never settled down, adult acne or chin hair that won't quit, or you've watched the scale climb without a clear reason — this is the entry that explains why those signs travel together. The thing to know: PCOS roughly doubles your odds of type 2 diabetes by your forties and meaningfully raises heart-disease risk over decades, but it is treatable, and recognising it early is one of the highest-leverage moves a woman can make in her own health.
Three things travel together in PCOS and feed each other: cycles that don't ovulate on time, more male-pattern hormones (androgens) than your body needs, and a body that doesn't respond properly to insulin. None of these alone is PCOS — but in combination they form a self-sustaining loop. Your body makes too much insulin trying to compensate; the ovary picks up that signal and overproduces androgens; the androgens disrupt egg maturation so cycles stall; and the ovary fills with small follicles that never finish growing. That last bit is where the name comes from — those aren't cysts in the dangerous sense, they're stalled follicles, frozen partway through development.
The reason it matters that insulin sits at the centre: insulin resistance in PCOS is direct, measurable, and only partly explained by weight. Studies using the gold-standard test of insulin sensitivity — clamping insulin levels while measuring how much glucose the body absorbs — find women with PCOS are about 27% less insulin-sensitive than women of the same weight without it Cassar et al. 2016. Carrying more weight makes the resistance worse, but losing the weight doesn't fully reverse it. This is why slim PCOS exists, and why it gets missed.
How the diagnosis works now
The 2023 International Evidence-based Guideline is the field's current standard, written by an international consortium covering 71 countries and replacing the 2018 version Teede et al. 2023. The diagnosis still rests on the same three-feature framework that has been in use since the early 2000s, but two pieces have shifted in ways that matter for whether you actually get diagnosed.
The framework: two of three features, after ruling out other conditions that look like PCOS. The features are irregular or absent ovulation (cycles longer than 35 days, or fewer than eight per year, persisting past adolescence), signs of too much androgen (either visible on the body — hirsutism, acne, scalp-hair thinning — or measurable in blood as elevated testosterone), and the polycystic ovary look. The 2023 update made two changes. First, ultrasound thresholds got stricter, so the "polycystic" criterion now requires more follicles than before — this trims the false positives. Second, and bigger: in adult women, a blood test for anti-Müllerian hormone (AMH) is now an accepted alternative to ultrasound. The reason that matters is logistical — most diagnosis used to be gated by gynaecology appointments and transvaginal scans. AMH is a tube of blood you can draw in primary care.
You don't fit one mould. Four phenotypes are recognised: the full classic version with all three features (the most common, around half to two-thirds of cases, and the most metabolically severe); a version with cycle and androgen problems but normal-looking ovaries; a version with hirsutism and polycystic ovaries but regular cycles (this is the one most often missed); and a non-androgenic version with cycle irregularity and polycystic ovaries but no excess testosterone (the mildest metabolically) Lizneva et al. 2016.
What most people get wrong about it
The name is misleading. "Polycystic" sounds like dangerous cysts that need treating; what's actually on the ultrasound is many small follicles that never finished maturing. They don't need surgical removal. The historical operation — cutting wedges out of the ovary — is gone for a reason.
It is not a disease of overweight women. The metabolic biology is there in lean women too, just lower in absolute terms — and slim PCOS is the version most often missed for years, because clinicians anchor on the weight. If your cycles never settled, your skin and hair give the androgenic story, and you're slim, you can absolutely still have it.
It does not resolve at menopause. The cycle problems become irrelevant when periods stop, but the metabolic and cardiovascular trajectory carries on, and may even pick up speed when the protective effect of oestrogen falls away. The framing of PCOS as a "fertility-years condition" is the single most damaging old idea the field is still trying to bury.
And irregular cycles in your teens are not automatic PCOS. Cycles take three years or so post-menarche to settle on their own. The 2023 guideline raised the bar specifically for teenagers, requiring longer cycle documentation and dropping the ultrasound criterion altogether for that age group — too many adolescents were getting the label prematurely and carrying it forward Teede et al. 2023.
How it actually shows up
For most women the signs accumulate slowly through the teens and twenties without ever cohering into a diagnosis. The standard pattern: periods that never settled into a predictable rhythm — cycles drifting out to 40, 50, sometimes 60 days, with the gap between them widening rather than narrowing as adulthood arrives. This is exactly the pattern that menstrual cycle tracking is built to surface — cycle length over 35 days is a core diagnostic clue, and a dated log turns a vague "my periods are all over the place" into something a clinician can act on rather than wave away as "just irregular". Adult acne that doesn't behave like teenage acne, concentrated along the jawline, chin, and back, often premenstrually flaring. Hair growing where it hasn't before — chin, upper lip, sideburns, lower abdomen below the navel, around the nipples, inner thighs. Scalp hair thinning at the crown or front, not the receding hairline pattern men get. Weight that climbs around the middle and resists ordinary diet-and-exercise effort more than it should. Difficulty conceiving, when that becomes the question — which is often when the diagnosis finally happens, sometimes a decade after the first signs.
The presentation looks different across ethnic backgrounds, and this matters for whether you get caught. South Asian women tend toward the metabolic end — higher insulin resistance and earlier diabetes at lower weights. Middle Eastern and Mediterranean women have the strongest hirsutism signal. East Asian women show the weakest dermatological signs, so the diagnosis there leans more on cycle pattern and bloodwork; using a Western hirsutism threshold misses real disease in those populations Lizneva et al. 2016.
What happens if it goes unmanaged
The unmanaged trajectory has rungs, and the rungs land at roughly predictable ages.
In your twenties the cost is mostly social and psychological. The constellation — cycles that don't run on time, skin and hair that don't read as healthy in a world that reads health through them, weight that creeps up on a diet that worked for friends — gets internalised as personal failure rather than recognised as a condition. The mental-health load is real and well-documented: pooled prevalence of depressive symptoms in women with PCOS runs around 36–42%, anxiety around 37–45%, and moderate-to-severe symptoms are four to five times the background rate for women without it Cooney et al. 2017. A lot of women carry years of self-blame for what is actually one diagnosable thing.
In your thirties, fertility brings the first real medical contact for many — and this is often when the diagnosis finally lands, sometimes after a decade of disconnected appointments. Conception is harder but treatable; the bigger story is what shows up in the obstetric paperwork. PCOS pregnancies carry meaningfully elevated odds of gestational diabetes, gestational hypertension, pre-eclampsia, miscarriage, preterm birth, and caesarean section. A meta-analysis of 63 studies pegs the odds ratio for gestational hypertension at 2.58 and pre-eclampsia at 1.87, with most of the excess persisting after matching for weight Bahri Khomami et al. 2024.
In your forties, the metabolic story becomes the story.
In your fifties and beyond, the elevated metabolic risk starts cashing out as cardiovascular events. Umbrella meta-analyses put coronary heart disease risk at about 1.4 times background and stroke at about 1.4 times background in PCOS Wekker et al. 2020. The American Heart Association now treats PCOS as a "risk-enhancing factor" for atherosclerotic heart disease — the same category as chronic inflammatory disease and early menopause — which in practice means clinicians are supposed to lower their threshold for statins and blood-pressure treatment in women with it Carmina and Lobo 2022.
Two more travelling companions that almost never get screened for, and should: obstructive sleep apnoea, which runs at 35–40% in adult women with PCOS versus around 6% in comparable women without it Kahal et al. 2020; and non-alcoholic fatty liver disease (now called MASLD), at 40 to 50 per cent prevalence in PCOS populations Rocha et al. 2017. Both compound the metabolic story. And one cancer risk that does deserve the attention: women who go years with very few periods build up unopposed oestrogen exposure on the lining of the uterus, which raises endometrial cancer risk to roughly three times background Barry et al. 2014. Maintaining at least a few withdrawal bleeds a year — whether through ovulation, combined contraception, or scheduled progestin — protects the endometrium and is one of the cheapest preventions in medicine.
What management actually looks like
The 2023 guideline frames management as lifelong rather than fertility-window. The structure is consistent across most women, with the dose adjusted to phenotype, weight, and life stage Teede et al. 2023.
The thing the guideline cannot do for you: get you into a clinician who actually applies it. The 2023 changes — AMH-based diagnosis without ultrasound, lifelong cardiometabolic framing — are uneven in primary-care adoption. If you suspect PCOS and your initial visit ended with "your cycles are just irregular, take the pill", a second opinion with an endocrinologist or a reproductive endocrinologist is worth the effort.
Where the management commonly falls apart
The most common failure mode in PCOS care is not that no one treats it — it's that the treatment narrows to the fertility window. A woman gets diagnosed when she struggles to conceive in her early thirties, is put on combined contraception or metformin to manage cycles and skin, has her baby (or doesn't), and then drifts out of endocrine follow-up. The long-term cardiometabolic surveillance — the annual glucose, the lipids, the blood pressure, the sleep-apnoea screen, the fatty-liver check, the mental-health screen — never gets scheduled, or peters out after a year. The metabolic phenotype carries on doing its work in the background, and the diabetes shows up at fifty as a surprise.
Second failure mode: the diagnosis is missed in the wrong-shape patient. Slim women without the obvious metabolic markers; East Asian women without the hirsutism signal; women whose primary symptom is mood or anxiety. Years of disconnected care follow — a dermatologist for the acne, a GP for the irregular cycles, a psychiatrist for the depression — without anyone joining the dots.
Third: anti-androgens prescribed without contraception. Spironolactone, finasteride, and cyproterone are teratogenic, and the combination of "trying to fix hirsutism" with "trying to conceive" is one of the more common high-stakes mismatches.
Fourth: metformin or a GLP-1 agonist sold to the patient as a slimming drug, then dropped when the scale doesn't move fast enough. Their job in PCOS is broader — glucose regulation, ovulation support, lowering the long-term diabetes trajectory — and dropping them on a weight-loss timeline misreads what they're doing.
What changes when you actually engage with it
The fastest thing that shifts is the framing. A woman in her late twenties or thirties walks into the diagnostic appointment carrying ten years of separate stories — the cycles, the skin, the weight that wouldn't move, the period of low mood she couldn't explain — and walks out with one. That reframe is not nothing. The qualitative literature on diagnosis-as-intervention isn't enormous, but the same finding shows up in clinic after clinic: people report relief that what they had been treating as a series of personal failures was a single biological pattern.
Within two to three months on the first-line combined contraceptive, cycles become predictable for the first time in a decade. Within six months, jawline acne fades and new hair growth slows. Hair already grown takes longer to retreat — that runs on a year or two timeline with sustained anti-androgen therapy plus removal methods. Scalp-hair density is the slowest to recover and the least likely to fully reverse, which is worth being honest about up front.
The metabolic gains are less visible and more important. Sustained weight reduction of even a few per cent, combined with treatment of the insulin resistance, bends the diabetes trajectory measurably — the same Nordic cohort that found doubled diabetes risk also found the risk is heavily weight-mediated, which means weight is a high-leverage handle here in a way it is not for everyone Glintborg et al. 2024. Restoring ovulation, where that matters, runs at roughly 28% per cycle live birth on letrozole — high enough that conception is usually a question of cycles, not feasibility Legro et al. 2014.
The long-game payoff is the one no one tells you about because it's measured against a counterfactual. The woman who recognises this in her twenties, gets her cycles protected, gets the sleep apnoea caught when it arrives, keeps the fatty liver from progressing, screens for diabetes annually, and treats the depression rather than living through it — she ends up in her fifties looking like a woman without PCOS at all. The intervention is partly medical and partly attention. Both work.
Related entries
Several conditions and interventions PCOS overlaps with deserve their own attention. Obstructive sleep apnoea sits next to PCOS so often it should be screened for as part of the routine workup. Metabolic syndrome and insulin resistance share most of the same biology and most of the same management. Non-alcoholic fatty liver disease (MASLD) is the silent partner most women with PCOS carry without knowing. Endometrial cancer screening becomes meaningful where chronic anovulation has gone unprotected. And on the treatment side: combined oral contraceptives, metformin, GLP-1 receptor agonists (semaglutide, tirzepatide), and inositol each warrant their own treatment when chosen specifically for the PCOS context.
- — PCOS roughly doubles your odds of type 2 diabetes by your forties — the metabolic risk outlasts fertility.
- — For the weight and insulin side of PCOS, GLP-1 drugs are increasingly used — they hit the metabolic core, not just the scale.
- — For cycles, acne, and excess hair, the combined pill is a standard treatment when pregnancy isn't the goal.
- — Myo-inositol is the best-evidenced supplement for PCOS — it nudges cycles back and calms insulin.
- — Berberine nudges insulin sensitivity — a modest, cheap option some use for the metabolic side of PCOS alongside myo-inositol.
- — NAC is one supplement with real PCOS evidence, improving ovulation, especially when standard treatment stalls.
- — Cycles longer than 35 days are a core diagnostic clue — tracking them is how the pattern surfaces.
- — PCOS quietly raises lifetime heart-disease risk — the part that outlasts fertility, and a reason to watch the female cardiac picture.
- — The same insulin resistance behind PCOS drives fatty liver — worth checking even if you feel fine.
- — PCOS strongly raises sleep-apnea risk, and apnea worsens the insulin problem — a loop worth screening.
- — Recurring painful boils in the armpits or groin alongside PCOS aren't a coincidence — that's hidradenitis suppurativa, and it needs its own care.
- — Metformin is a common PCOS treatment, and years on it quietly drain B12 — worth checking the level periodically.
1. Substance + claimed effects
Polycystic ovary syndrome (PCOS) is the most common endocrinopathy of reproductive-age women — a heterogeneous syndrome defined by some combination of ovulatory dysfunction, clinical or biochemical hyperandrogenism, and polycystic ovarian morphology, with insulin resistance as a near-universal but not formally diagnostic substrate Teede et al. 2023. Pooled global prevalence by Rotterdam criteria is roughly 10–13% of reproductive-age women, varying by region and ethnicity Bozdag et al. 2016. The entry covers PCOS as a condition adult women carry from menarche to menopause: how it is now diagnosed under the 2023 International Evidence-based Guideline (the field's current standard, replacing the 2018 update of the Rotterdam consensus), what it does to cycle regularity and fertility, the androgen-mediated dermatological signs (hirsutism, acne, androgenic alopecia), the metabolic-syndrome cluster it predisposes to (type 2 diabetes, NAFLD/MASLD, sleep apnoea, dyslipidaemia), the elevated cardiovascular risk that compounds over decades, the strong overlap with depression and anxiety, and the obstetric risk profile during pregnancy. The entry's editorial frame is recognition + literacy: PCOS is under-diagnosed (estimated diagnostic delay of years, with multiple physicians consulted before label is applied), under-managed past the fertility window, and the long-term cardiometabolic stakes are routinely understated to patients. The "what to do" question splits into clinician-directed work (OCPs, metformin, letrozole, GLP-1 RAs, anti-androgens) that the entry summarises but does not prescribe.
2. Evidence by addressing question
Mechanism
Science. PCOS is a self-sustaining loop of three intertwined endocrine derangements: hyperandrogenism, ovulatory dysfunction (driven by altered hypothalamic-pituitary GnRH/LH pulsatility), and insulin resistance with compensatory hyperinsulinaemia. Meta-analysis of euglycaemic-hyperinsulinaemic clamp studies — the gold-standard direct measure of insulin sensitivity — finds women with PCOS are ~27% less insulin-sensitive than BMI-matched controls, and this deficit is partially independent of adiposity though strongly amplified by it Cassar et al. 2016. The seminal Endocrine Reviews synthesis frames PCOS as a disorder of selective insulin resistance: the metabolic insulin-signalling pathway (PI3K) is impaired in muscle and adipose tissue, while the mitogenic/steroidogenic pathway (MAPK) remains intact in ovarian theca cells, so hyperinsulinaemia continues to drive theca-cell androgen synthesis even as systemic glucose handling deteriorates Diamanti-Kandarakis and Dunaif 2012.
Mechanism (loop detail). Hyperinsulinaemia stimulates CYP17α activity in ovarian theca cells, increasing androgen output. Insulin simultaneously suppresses hepatic sex hormone-binding globulin (SHBG) production, raising the free testosterone fraction. Elevated LH pulsatility (a partly genetic, partly insulin-mediated trait) amplifies theca-cell androgenesis further. Granulosa-cell aromatase activity is relatively reduced, so the androgen-to-oestrogen conversion is impaired, and follicular development arrests at the small antral stage — the "polycystic" morphology is many arrested follicles, not pathological cysts. Without an LH surge there is no ovulation, so cycles lengthen or stop. The hyperandrogenic phenotype (Rotterdam A, B, C) is metabolically driven; the non-hyperandrogenic phenotype (D) appears more neuroendocrine.
Evidence (does the diagnosis correspond to a real disorder, and how strong is the long-term-risk literature)
Science. The 2023 International Evidence-based Guideline — developed by an international collaboration of 39 societies across 71 countries, methodologically GRADE-rated, replacing the 2018 update — endorses modified Rotterdam diagnostic criteria as the standard: two of three of (i) ovulatory dysfunction, (ii) clinical or biochemical hyperandrogenism, (iii) polycystic ovarian morphology, after excluding mimics (thyroid disease, hyperprolactinaemia, non-classic congenital adrenal hyperplasia, Cushing's, androgen-secreting tumours) Teede et al. 2023. Key 2023 refinements over Rotterdam: ultrasound thresholds tightened (≥20 follicles per ovary on transvaginal high-frequency probe, or ovarian volume ≥10 mL), and crucially, serum anti-Müllerian hormone (AMH) is now accepted as an alternative to ultrasound for the morphological criterion in adults (not adolescents), enabling diagnosis in primary care without imaging. AMH meta-analysis pooled sensitivity 0.79 and specificity 0.87 for PCOS in adults.
Four phenotypes (A–D) capture the heterogeneity. Phenotype A (all three features) is the most prevalent (~50–67%) and the most metabolically severe; B (ovulatory dysfunction + hyperandrogenism, no PCOM) carries similar metabolic risk; C (hyperandrogenism + PCOM, regular cycles) is intermediate; D (PCOM + ovulatory dysfunction, no hyperandrogenism) is the mildest metabolic phenotype but still associated with elevated cardiometabolic risk above background Lizneva et al. 2016.
Long-term cardiometabolic outcomes. Multiple meta-analyses converge: PCOS roughly doubles type 2 diabetes risk and raises cardiovascular event risk by 30–70% over the lifecourse. The 2024 Nordic cohort (Glintborg et al.) of 99,892 women with PCOS and 446,055 controls across Denmark/Finland/Sweden found a crude hazard ratio for incident T2D of ~4.6, adjusted HR 2.9, with BMI ≥30 conferring 7.6–11.3× risk vs. BMI <25 in the PCOS population — i.e., obesity multiplies what is already a PCOS-elevated baseline Glintborg et al. 2024. The Wekker et al. 2020 umbrella meta-analysis of long-term cardiometabolic disease found pooled odds ratios of 2.87 for type 2 diabetes, 1.44 for coronary heart disease, and 1.36 for stroke in PCOS vs. controls Wekker et al. 2020. The American Heart Association now classifies PCOS as a "risk-enhancing factor" for atherosclerotic cardiovascular disease — the same category as chronic inflammatory disease and premature menopause — meaning it warrants more aggressive primary prevention than the calculated 10-year risk score alone would suggest Carmina and Lobo 2022.
Pregnancy complications meta-analysis (63 studies) finds increased odds of gestational diabetes, gestational hypertension (OR 2.58), pre-eclampsia (OR 1.87), miscarriage, preterm birth, and caesarean section in PCOS pregnancies, with effects partially but not fully attenuated by BMI matching Bahri Khomami et al. 2024. Endometrial cancer risk is elevated 2.8–3-fold by meta-analysis, driven by chronic anovulation producing unopposed oestrogen exposure on the endometrium — this is the strongest cancer signal in PCOS, with breast and ovarian cancer associations weak or absent after BMI adjustment Barry et al. 2014. Obstructive sleep apnoea prevalence pooled at 35–40% in adult PCOS vs. ~6% in matched controls, with a roughly 9-fold odds ratio Kahal et al. 2020. Non-alcoholic fatty liver disease (now MASLD) prevalence is pooled at ~40–50% in PCOS populations, driven by the shared insulin-resistance substrate plus hyperandrogenism as an independent contributor Rocha et al. 2017 Macut et al. 2017.
Contraindications / when not to act on this
The diagnosis itself has no contraindications; the management decisions do. Combined oral contraceptives — the guideline first-line for menstrual irregularity and hyperandrogenism — carry the usual contraindications (smokers over 35, history of VTE, migraine with aura, uncontrolled hypertension, oestrogen-sensitive cancers). Metformin contraindicated in advanced renal disease. Letrozole and clomiphene are ovulation-induction drugs prescribed in defined cycles, not chronic therapy. Spironolactone is teratogenic and must be paired with contraception. GLP-1 receptor agonists are not approved specifically for PCOS, are contraindicated in pregnancy and personal/family history of medullary thyroid carcinoma or MEN2, and the long-term reproductive-age data are still accumulating Teede et al. 2023.
Misconceptions
(i) "Polycystic" is a misnomer — the follicles are not cysts, they are arrested antral follicles, and the morphological criterion is not what makes PCOS a disorder. Women can have polycystic ovaries on ultrasound without PCOS (~20% of reproductive-age women), and women can have PCOS without polycystic morphology (Phenotype B). (ii) PCOS is not a disease of obese women — slim PCOS exists (Phenotype B/D more often), shares the underlying insulin resistance, and is under-diagnosed because clinicians anchor on weight. (iii) Irregular cycles in adolescence are not automatic PCOS; physiological anovulation is common in the first 3 years post-menarche and the 2023 guideline explicitly raises the diagnostic bar for adolescents to avoid over-labelling. (iv) PCOS does not resolve at menopause; cycle dysfunction becomes moot but the metabolic and cardiovascular risk trajectory persists and may accelerate. (v) "Cysts" do not need surgical removal — historical ovarian wedge resection and modern laparoscopic ovarian drilling are second-line fertility interventions, not curative.
Audience (who needs to recognise this in themselves)
The diagnostic triad presents as: cycles longer than 35 days or fewer than 8 per year that persisted past adolescence, unexplained hirsutism (Ferriman-Gallwey ≥4–6 depending on ethnic norms), persistent adult acne especially jawline/back, androgenic scalp thinning, difficulty conceiving, or a metabolic-syndrome workup that surfaces hyperinsulinaemia. Ethnic background modulates phenotype — South Asian women show higher metabolic severity at lower BMI; Mediterranean and Middle Eastern populations have higher hirsutism prevalence; East Asian women show lower hirsutism rates so dermatologic signs underweight diagnosis there Lizneva et al. 2016.
Alternatives (other diagnoses that mimic and must be excluded)
Before PCOS is settled the workup must rule out: thyroid disease (TSH), hyperprolactinaemia (prolactin), non-classic congenital adrenal hyperplasia (17-hydroxyprogesterone), hypothalamic amenorrhoea (FSH/LH pattern, low BMI/over-exercise history), Cushing's syndrome (clinical screen, dexamethasone test if suspected), and androgen-secreting tumour (rapid virilisation, very high testosterone >200 ng/dL, DHEAS) Teede et al. 2023.
Failure modes (where management goes wrong in practice)
Most common failure: the diagnosis is made at fertility presentation, cycle/skin is managed with OCP, and the long-term cardiometabolic surveillance (annual glucose tolerance test or HbA1c, lipid panel, blood pressure, weight, mental health screen, MASLD screen if obese) lapses once pregnancy is achieved or not pursued. Second most common: the diagnosis is missed in lean women, in non-hirsute East Asian women, or in women whose primary symptom is mood/anxiety, leading to years of disconnected symptomatic treatment. Third: anti-androgens prescribed without contraception cause teratogenic risk. Fourth: metformin prescribed as a slimming agent rather than for its glucoregulatory + ovulation-supporting effects, then discontinued when weight doesn't budge.
Practicalities (what management actually looks like)
The 2023 guideline structures management as a lifelong framework: (a) lifestyle (diet + activity, no single specific regimen dominant, weight maintenance/reduction proportional to baseline) is foundational; (b) combined oral contraceptives are first-line pharmacotherapy for cycle regulation and hyperandrogenism, with preference for low-dose ethinylestradiol and progestins with anti-androgenic or neutral profile; (c) metformin for metabolic features and as add-on for cycle/ovulation, particularly in BMI ≥25; (d) inositol (myo-inositol ± D-chiro-inositol) is now an option but with weaker evidence than metformin; (e) anti-androgens (spironolactone, finasteride, cyproterone) for hirsutism/acne where OCPs alone are insufficient, with mandatory contraception; (f) letrozole first-line for ovulation induction, superior to clomiphene (RR for live birth 1.54) Legro et al. 2014; (g) bariatric surgery for class III obesity with PCOS; (h) GLP-1 RAs (semaglutide, tirzepatide) increasingly used off-label for the obesity + insulin-resistance + cycle dysfunction triad — early trials show restoration of regular cycles in a meaningful fraction of obese PCOS women, but FDA labelling does not yet include PCOS as an indication and reproductive-age long-term data are immature Teede et al. 2023.
History
Described by Stein and Leventhal in 1935 as a triad of amenorrhoea, hirsutism, and bilaterally enlarged polycystic ovaries treated by wedge resection. NIH 1990 criteria required hyperandrogenism + ovulatory dysfunction (and excluded the lean ovulatory phenotype). The Rotterdam ESHRE/ASRM 2003 consensus added polycystic morphology as a third criterion and the two-of-three structure; the Androgen Excess Society 2006 position required hyperandrogenism. The 2018 international guideline endorsed Rotterdam. The 2023 guideline refined ultrasound thresholds and admitted AMH — the largest diagnostic update in two decades Teede et al. 2023.
Stakes (cost of leaving it under-managed across the lifecourse)
Science. The lifetime arc is: late teens/twenties — irregular cycles, acne, hirsutism, possible weight gain, often a long lag before label. Thirties — fertility difficulty surfaces, often the first contact with definitive diagnosis. Forties — insulin resistance progresses; type 2 diabetes incidence accelerates (HR ~3–5× background) Glintborg et al. 2024. Fifties+ — cardiovascular events begin to translate the elevated risk into real coronary disease and stroke incidence Wekker et al. 2020. Mental health load is constant: depressive symptoms in ~36–42% of women with PCOS pooled, anxiety in ~37–45%, with moderate-to-severe symptoms at 4–5× the background rate Cooney et al. 2017. Untreated chronic anovulation drives a 2.8-fold endometrial cancer risk Barry et al. 2014. Sleep apnoea is grossly under-screened despite a 35–40% prevalence and its independent contribution to insulin resistance, hypertension, and fatigue Kahal et al. 2020.
Payoff (what changes when it is recognised and managed)
Recognition itself reframes a constellation of symptoms a woman has often experienced as personal failure (weight, skin, fertility, mood) as a unified, biologically tractable condition — there is a small but real mental-health gain from the diagnosis alone, though the data on this are mostly qualitative. Treatment-mediated gains: combined OCP normalises cycles within 1–3 months and reduces androgenic symptoms over 6–12 months; metformin restores ovulation in 30–50% of anovulatory women; letrozole achieves live birth in ~28% per cycle in the PPCOS II trial Legro et al. 2014; lifestyle intervention with even 5% weight loss restores spontaneous ovulation in a meaningful subset and reduces type 2 diabetes incidence; GLP-1 RAs in trials produce 9–15% weight loss with concurrent menstrual regularisation. Cardiovascular benefit is inferred from general primary prevention, not yet demonstrated in PCOS-specific RCTs.
Out-of-scope (adjacent conditions / interventions to know about)
Hidden adjacent topics this entry should signpost but not cover: combined oral contraceptives as a substance; metformin; GLP-1 receptor agonists; inositol; obstructive sleep apnoea; metabolic syndrome / insulin resistance; depression and anxiety as standalone entries; endometrial cancer screening; hidradenitis suppurativa (associated with PCOS, separate condition); congenital adrenal hyperplasia (the principal differential).
3. The credibility range
Optimist case
The 2023 guideline is methodologically the strongest endocrine guideline in the field — international consortium, GRADE-evaluated, 254 specific recommendations and practice points, replacing a fragmented diagnostic landscape with a unified algorithm including AMH-based diagnosis that finally moves PCOS out of imaging gatekeeping. The cardiometabolic risk signal is now corroborated by multiple large national cohorts (Nordic 99,892-women cohort, UK CPRD analyses, Danish registries) — this is no longer disputable. Management has matured: letrozole has displaced clomiphene as first-line for ovulation; GLP-1 RAs are providing the metabolic lever PCOS treatment lacked for forty years; combined OCPs remain a clean first-line for cycle/androgen. The reframing of PCOS as a lifelong cardiometabolic condition rather than a fertility nuisance is the single most important clinical shift of the last decade.
Skeptic case
PCOS is a syndrome, not a disease — it is a pattern label over a heterogeneous biology that may turn out to be several distinct conditions (the A-vs-D phenotype split is suggestive). Diagnostic prevalence ranges from 8% to 21% depending on criteria, which means we are labelling somewhere between 80 million and 200 million women globally — much of the upper range may be normal-variant ovarian morphology and cycle variability. The AMH cutoff has not converged across labs and ethnicities, threatening reproducibility. Cardiovascular event risk in PCOS appears largely mediated by BMI and metabolic syndrome — i.e., it may be that obesity-in-the-context-of-PCOS rather than PCOS itself is the driver, and current cardiovascular RCT evidence in PCOS specifically is thin. GLP-1 RA enthusiasm in PCOS is outrunning the data: the trials are short, the long-term reproductive-age safety unclear, and prescribing is exploding (2.4% of PCOS women in 2021 to 17.6% in 2025). Mental-health prevalence numbers are inflated by clinic-sampled cohorts; community-sampled rates are lower. The diagnostic label can also harm: a woman in her early 20s with mild oligomenorrhoea and normal androgens carrying a "PCOS" tag for life may face insurance, fertility-narrative, and self-stigma consequences disproportionate to her actual disease severity.
Author's call
The diagnosis is real and meaningful, the long-term cardiometabolic risk is well-supported (high evidence on the disease side), and the field is moving in the right direction. Under-diagnosis is a bigger problem than over-diagnosis at the population level — diagnostic delay averages multiple years and multiple clinicians, and women living undiagnosed pay in fertility, metabolic disease, and mental-health load. But the management landscape is mid-flight: OCP/metformin/letrozole are settled, GLP-1 RAs and inositol are still maturing, and the cardiovascular-prevention question (do we treat PCOS specifically more aggressively, or just track the resulting metabolic-syndrome features?) is genuinely open. Controversy score is modest (2) — the field broadly agrees on diagnosis and first-line management; the disagreements are over edges (AMH cutoffs, adolescent thresholds, GLP-1 RA scope). Evidence score is high (4) — multiple Cochrane/umbrella meta-analyses, methodologically rigorous 2023 guideline, large national cohorts; not a 5 because the cardiovascular-prevention RCT evidence is observational/inferential rather than direct.
4. Stakeholder + incentive map
- Endocrine and reproductive societies (ESE, ASRM, ESHRE, AE-PCOS Society, Monash University CREPCOS) are the primary guideline drivers, methodologically rigorous, broadly aligned in 2023.
- Reproductive endocrinology / fertility industry has a long-standing fertility-frame on PCOS that historically narrowed it to anovulation; the metabolic and lifelong-condition framing is corrective.
- Pharma — combined OCP manufacturers (legacy market), Novo Nordisk / Eli Lilly (GLP-1 RAs surging into the PCOS off-label market), inositol supplement manufacturers (large consumer market on weaker evidence), spironolactone/finasteride generics (commoditised). GLP-1 makers have material incentive to expand into PCOS but have not yet sought a formal indication.
- Patient communities and online influencers are unusually active — PCOS has a strong patient-advocacy presence (Verity, PCOS Challenge, large Reddit and TikTok communities) which has been a force for diagnostic visibility and a vehicle for low-evidence supplement promotion (inositol, berberine, spearmint tea, low-carb absolutism).
- Naturopathic / functional medicine claims of "reversing PCOS" through diet — overstated, but the lifestyle-first message they amplify is concordant with the guideline.
- Primary care remains the diagnostic bottleneck — most women's symptoms present here, but PCOS literacy and the new AMH-based pathway are unevenly implemented; the diagnostic delay is largely a primary-care problem.
- Insurance / payer systems in some jurisdictions require infertility diagnosis to fund ovulation induction; this distorts who gets diagnosed when.
5. Population variability
- Ethnicity. South Asian women show higher metabolic severity at lower BMI and earlier T2D onset; Middle Eastern/Mediterranean populations have the highest hirsutism prevalence; East Asian women have the lowest hirsutism prevalence and a higher proportion of non-hyperandrogenic phenotype D, so dermatological criteria undercount disease there. Ferriman-Gallwey thresholds should be ethnicity-adjusted (≥4–6 instead of ≥8 in low-baseline populations) Lizneva et al. 2016.
- BMI. Obesity multiplies cardiometabolic risk substantially within PCOS (T2D HR 7.6–11.3× at BMI ≥30 vs. <25 within PCOS) but lean PCOS exists and is under-diagnosed. Insulin resistance is present in lean PCOS, just less severe in absolute terms Glintborg et al. 2024.
- Age band. Adolescent diagnosis carries higher false-positive rates; the 2023 guideline raises the bar for adolescents (no PCOM criterion, longer documentation of cycle irregularity required). Perimenopausal cycles regularise as ovarian reserve falls, but the metabolic phenotype persists into and through menopause.
- Phenotype. A and B (hyperandrogenic anovulatory) carry the heaviest metabolic load; D (non-hyperandrogenic) carries less but still elevated risk. Treatment intensity should track phenotype + BMI, not the diagnostic label alone.
- Reproductive intent. A woman not seeking pregnancy needs cycle-protection (endometrial protection via OCP or progestin), cardiometabolic surveillance, and mental health support — not fertility work. A woman seeking pregnancy needs ovulation induction (letrozole), pre-conception metabolic optimisation, and antenatal surveillance for GDM/preeclampsia.
6. Knowledge gaps
- Whether intensive PCOS-specific cardiovascular primary prevention (statin, BP, aspirin thresholds adjusted for PCOS as a risk-enhancer) reduces CV events in PCOS — no PCOS-specific CV-outcome RCT exists yet; current AHA "risk-enhancing factor" framing is inferential Carmina and Lobo 2022.
- Whether GLP-1 RAs should be first-line in obese PCOS over OCP+metformin — early trials promising but uncontrolled for long-term reproductive safety, and the conception-rebound effect when therapy is discontinued is poorly characterised.
- Whether AMH-based diagnosis without ultrasound performs equivalently across ethnicities and across assay platforms — assay standardisation is unfinished.
- The biological basis of the phenotype split (A/B vs. D): is PCOS one disease or two? Genetic studies suggest partly overlapping but distinct architecture.
- The mental-health component: is the depression/anxiety load a consequence of androgen excess and inflammation, or of the chronic-illness experience, or both? Causal direction is mixed in the data Cooney et al. 2017.
- Long-term outcomes in adolescent-diagnosed cohorts now reaching their forties — the natural history of "modern-criteria PCOS" is still being read.
Brief vs. coverage. The brief named diagnostic criteria, cycle regularity, fertility, metabolic risk, androgenic symptoms, and long-term cardiometabolic health — all five are covered. The 2023 guideline is the central diagnostic anchor; the AMH-based diagnostic option in adults is flagged as the structurally meaningful change. Metabolic risk is covered through three converging lenses: type 2 diabetes (Glintborg Nordic cohort), cardiovascular events (Wekker umbrella, Carmina AHA reframe), and the under-screened companions (OSA, MASLD, endometrial cancer). Mental health was added as a major consequence even though the brief did not name it — the prevalence data (Cooney et al.) is too strong to omit and the mood score reflects it.
Action choice. Settled on know over decide or test. PCOS is a condition the reader either has or doesn't; the primary work the entry does is recognition + literacy, and the downstream choices (OCP, metformin, letrozole, GLP-1, anti-androgens) all involve clinician input and were explicitly framed as out-of-scope-for-this-entry. A future entry titled "PCOS diagnostic workup" with action: test could plausibly split out the lab-and-imaging pathway, but it would duplicate this entry's evidence section.
Rating difficulties.
longevity: 3rather than 4: the CV risk signal is real and replicated, but the magnitude (HR 1.4 for CHD/stroke) is modest compared to the catalogue's 4-tier interventions, and no PCOS-specific CV-prevention RCT yet anchors the upper end.beauty_direct: 3rather than 4: hirsutism and acne treatment is clearly visible over months, but treatment effect is gradual rather than dramatic, and scalp-hair recovery is genuinely poor — calling 4 would oversell.focus: 1: there is no direct cognitive effect of the condition; the indirect lift via sleep/depression is real but already counted in the sleep/mood scores, so scoring it more than 1 would double-count.controversy: 2: the field is broadly aligned post-2023. Edge debates (AMH cutoffs across ethnicities, adolescent diagnostic thresholds, GLP-1 RA positioning) are genuine but not foundational; scoring 3+ would misread the state of the field.
Hard scoping calls.
- Did not cover adolescent diagnosis in depth, only flagged that the 2023 thresholds differ. A separate PCOS in adolescents entry warrants existence — the diagnostic, mental-health, and treatment-decision tradeoffs are different enough to justify it.
- Did not cover specific drug entries (OCP, metformin, letrozole, GLP-1 RAs, inositol, spironolactone). Each is referenced in the protocol section and listed in
related, but the entry deliberately stays at the recognition/literacy altitude — going deeper on each drug would turn this into a pharmacology textbook. - Did not cover Stein-Leventhal historical detail beyond the briefest gesture in the mechanism section. The history is interesting but did not earn a section; the editorial decision was to spend that paragraph on what the reader notices, not the chronology of nomenclature.
Future-link candidates. Entries that should cross-link to this one once they exist: combined-oral-contraceptive, metformin, glp-1-receptor-agonists, obstructive-sleep-apnea, metabolic-syndrome, endometrial-cancer-screening, inositol, spironolactone, letrozole-for-ovulation-induction, hirsutism-management. The related field on meta lists the first six; the remaining four are not yet wired.
Separate-entry candidates surfaced during writing.
- PCOS in pregnancy / preconception optimisation — the obstetric risk picture is substantial enough to warrant its own entry, and a pregnancy-specific reader has different decision points.
- GLP-1 receptor agonists for metabolic conditions — this is a broader entry that would naturally cross to PCOS but cover semaglutide/tirzepatide as substances. The PCOS-specific evidence is still maturing; deserves its own treatment when it is.
- The reframe / diagnostic-relief phenomenon — the qualitative finding that diagnostic recognition is itself an intervention applies beyond PCOS (endometriosis, ADHD, autism in adults). Out of scope here, but worth noting as a cross-cutting editorial pattern.
Audience scoping. Adult women, ages 18-39 and 40-59. The condition continues into 60+, but the entry's editorial centre of gravity is recognition + decision-making during reproductive years and the transition; the 60+ slice (post-menopausal metabolic/CV monitoring of pre-existing PCOS) is referenced in the stakes section but does not justify the age band, since the recognition framing wouldn't be the primary value to a 65-year-old reader.
Polycystic Ovary Syndrome (PCOS)
2023 International Evidence-based Guideline is GRADE-methodologically the strongest endocrine guideline in the field, with international consortium endorsement. Multiple Cochrane/umbrella meta-analyses on key outcomes (T2D, CV events, sleep apnoea, NAFLD, pregnancy complications, depression). Not a 5 because cardiovascular-prevention RCT evidence in PCOS specifically remains inferential rather than direct (Teede et al. 2023; Wekker et al. 2020).
Hirsutism is the principal hyperandrogenic dermatological sign in PCOS, prevalent in 65–75% of cases; acne and androgenic alopecia common. Combined oral contraceptive plus spironolactone produces clearly visible improvement over 3–6 months; topical eflornithine and hair removal act faster (Teede et al. 2023).
Long-term aesthetic trajectory is meaningfully altered: chronic hyperandrogenism drives progressive hirsutism and androgenic alopecia, and the obesity + metabolic-syndrome cluster contributes to skin and body-composition aging. Sustained management changes the aging trajectory rather than producing a quick visible win (Teede et al. 2023; Wekker et al. 2020).
Within months of starting first-line therapy (combined OCP, lifestyle, metformin or GLP-1 RA), cycles regularise, androgenic symptoms attenuate, and insulin resistance markers move. Mood and energy lift in a meaningful subset from diagnosis-reframe alone (Teede et al. 2023; Cooney et al. 2017).
PCOS roughly doubles type 2 diabetes incidence (adjusted HR 2.9 in the Nordic 99,892-women cohort), raises coronary heart disease risk 1.44× and stroke 1.36×, and is now classified by AHA as a cardiovascular risk-enhancing factor; effective management reduces but does not eliminate this excess (Glintborg et al. 2024; Wekker et al. 2020; Carmina and Lobo 2022).
Obstructive sleep apnoea is roughly 6× more common in PCOS than in age- and BMI-matched controls (pooled prevalence 35–40%), is independently associated with insulin resistance and cardiovascular risk, and is routinely under-screened; recognising and treating it produces clear sleep improvement (Kahal et al. 2020).
Pooled prevalence of depressive symptoms ~36–42% and anxiety symptoms ~37–45% in PCOS, with moderate-to-severe symptoms 4–5× background rate; the diagnostic reframe (recognising a unified condition behind weight, skin, fertility, and mood symptoms) is itself meaningful, and addressing the underlying drivers improves inner wellbeing (Cooney et al. 2017).
Lifetime cost is substantial: diagnostic workup (labs, ultrasound or AMH), ongoing OCP or metformin (commoditised but lifelong), letrozole cycles for fertility, possibly GLP-1 RAs (multi-thousand annual cost where uninsured), specialist visits, ongoing surveillance (annual glucose, lipids, BP, mental-health screen). Insurance variability adds noise (Teede et al. 2023).
Sustained lifestyle attention (weight maintenance, activity), daily medication, regular surveillance appointments, and active mental-health management together represent sustained ongoing demand — not full life reorganisation, but materially more than the background female health load (Teede et al. 2023).
Fatigue in PCOS is multifactorial — sleep apnoea (35–40% prevalence vs ~6% baseline), depression, anaemia from heavy menstrual bleeding, insulin resistance — and addressing these recovers a real but modest daily-energy gain (Kahal et al. 2020; Cooney et al. 2017).