Polycystic ovary syndrome, or PCOS, is a hormonal and metabolic disorder in which the ovaries produce elevated androgens and eggs often fail to mature and release, and it is diagnosed in medicine by the Rotterdam criteria: two of three findings — irregular ovulation, high androgen levels, or polycystic ovarian morphology on ultrasound — after excluding thyroid disease, high prolactin, and non-classic congenital adrenal hyperplasia. It affects roughly 8 to 13 percent of women of reproductive age depending on the diagnostic criteria applied, according to the 2023 International Evidence-based Guideline for the Assessment and Management of PCOS, coordinated by Monash University. The "cysts" in the name are not cysts at all; they are immature follicles arrested mid-development.
What the Rotterdam Criteria Actually Require Before a Diagnosis
Clinical medicine settled on a single diagnostic framework at a 2003 meeting in Rotterdam convened jointly by the European Society for Human Reproduction and Embryology and the American Society for Reproductive Medicine. That consensus requires two of three features, and crucially it demands that other conditions be ruled out first. The three features are oligo-anovulation (fewer than eight periods a year, or cycles longer than 35 days), clinical or biochemical hyperandrogenism, and polycystic ovarian morphology — typically 12 or more follicles measuring 2 to 9 millimeters in one ovary, or ovarian volume above 10 milliliters, on a transvaginal scan. A 2018 update from the same guideline group raised the follicle threshold to 20 or more per ovary when using modern high-resolution ultrasound equipment, which matters because older scanners undercount small follicles.
The exclusion step is where many self-diagnoses fall apart. Before PCOS can be assigned, a clinician must rule out thyroid dysfunction, hyperprolactinemia, and non-classic congenital adrenal hyperplasia, the last usually screened with a 17-hydroxyprogesterone level. Cushing syndrome and androgen-secreting tumors also belong on the exclusion list, though they are rare. The 2023 international guideline, developed by a partnership of 39 societies and organizations, reaffirmed the Rotterdam framework and explicitly rejected adding anti-Müllerian hormone as a stand-alone diagnostic test, since AMH assays vary too widely between laboratories to serve as a threshold.
Here is the practical consequence: an ultrasound finding alone is not PCOS. Roughly 20 to 30 percent of normal ovulating women show polycystic ovarian morphology on a scan, according to figures cited in the 2018 international guideline. A woman with polycystic-looking ovaries who ovulates regularly and has normal androgens does not meet the criteria and should not carry the label.
For the clinician's checklist, the sequence runs: history of cycle length, then total and free testosterone plus a free androgen index, then a transvaginal ultrasound if the patient is sexually active or a transabdominal scan if not, then the exclusion blood panel. Hormonal contraception alters all three of these results and should be stopped for at least three months before testing where safe to do so.
Three Phenotypes, Three Different Metabolic Risk Profiles
The Rotterdam criteria generate four phenotypes depending on which features combine, and the 2023 guideline treats them as genuinely different clinical entities rather than one disease with variable symptoms. Phenotype A combines all three features — irregular ovulation, hyperandrogenism, and polycystic ovaries — and accounts for roughly 50 to 60 percent of diagnosed cases in most cohort series. Phenotype B has irregular ovulation plus hyperandrogenism without ovarian morphology. Phenotype C has hyperandrogenism plus ovarian morphology with regular cycles. Phenotype D has irregular ovulation plus ovarian morphology with normal androgens.
| Phenotype | Ovulatory dysfunction | Hyperandrogenism | Polycystic ovaries | Approximate share of cases |
|---|---|---|---|---|
| A (classic) | Yes | Yes | Yes | 50–60% |
| B | Yes | Yes | No | 10–20% |
| C | No | Yes | Yes | 10–20% |
| D | Yes | No | Yes | 10–20% |
The shares above come from phenotype distributions reported in the international guideline's evidence review and vary by population screened; treat them as indicative ranges, not fixed proportions. What is not uncertain is the risk gradient. Phenotype A carries the highest insulin resistance and metabolic syndrome burden. Phenotype D, sometimes called the "mild" form, carries the lowest metabolic risk but still shows elevated risk of ovulatory subfertility. A 2019 analysis in the journal Human Reproduction Update pooled phenotype data and found that phenotypes A and B had the highest prevalence of metabolic syndrome, while phenotype C drove most of the hyperandrogenism-related dermatologic complaints.
Clinicians use phenotype to steer treatment. A phenotype D patient seeking pregnancy may need only ovulation induction. A phenotype A patient with a BMI of 32 likely needs metabolic screening — fasting glucose, HbA1c, lipid panel, blood pressure — before any fertility drug is considered, because insulin resistance blunts the response to letrozole and clomiphene.
Why the Name Is Misleading: Follicles, Not Cysts
The word "cyst" in polycystic ovary syndrome is a historical accident. What an ultrasound actually shows in PCOS is a ring of small antral follicles — immature egg-containing structures, each roughly 2 to 9 millimeters — arrested before they reach the size needed for ovulation. A true ovarian cyst is a fluid-filled sac that forms after ovulation or from abnormal cell growth, and it behaves differently. The 2003 Rotterdam naming stuck because the condition had already been called Stein-Leventhal syndrome after the 1935 paper by American gynecologists Irving Stein and Michael Leventhal describing seven women with enlarged, cystic ovaries, hirsutism, and infertility.
That 1935 paper described what the authors saw under the surgical microscope of the era. Modern imaging explains the appearance differently. In a normal menstrual cycle, one follicle grows dominant and suppresses its neighbors. In PCOS, the hormonal environment — elevated luteinizing hormone relative to follicle-stimulating hormone, plus high insulin — prevents any single follicle from winning that race, so a cohort of small follicles persists cycle after cycle. The ovary looks like a string of pearls on a transvaginal scan, a description that first appeared in radiology literature in the 1980s.
The naming confusion has real clinical costs. Women told they have "cysts" often believe they need surgery, and some request ovarian drilling or cystectomy that is not indicated. Ovarian drilling, a laparoscopic procedure involving puncturing the ovarian surface with electrocautery, is a legitimate second-line option for clomiphene-resistant anovulation, but it is not a cyst removal operation. Meanwhile, a woman with a genuine 5-centimeter simple ovarian cyst may be reassured she "just has PCOS" by a well-meaning friend, delaying evaluation of a structure that warrants follow-up.
The distinction matters for another reason. Polycystic ovarian morphology is also seen in up to a quarter of women with normal cycles, so the scan finding is a shape, not a disease. The disease is the endocrine dysfunction, and it can exist without the shape — phenotype B proves it.
The Insulin Connection and the 70 Percent Undiagnosed Figure
Insulin resistance sits underneath the majority of PCOS cases, even in lean women. Estimates vary by how resistance is measured, but the 2023 international guideline notes that insulin resistance is present in a substantial proportion of women with PCOS across all BMI categories, and that routine screening for glucose intolerance is recommended rather than assumed from body weight. The mechanism is a loop: elevated insulin reduces hepatic production of sex hormone-binding globulin, which raises free testosterone; higher insulin also drives the ovarian theca cells to produce more androgen. More androgen worsens the cycle irregularity, which worsens the metabolic picture.
The diagnostic gap is the part most people never hear. A commonly cited estimate holds that up to 70 percent of women with PCOS remain undiagnosed. That figure traces to a 2016 analysis published in the Journal of Clinical Endocrinology & Metabolism using data from a large Australian cohort, which found roughly 68 to 72 percent of women meeting Rotterdam criteria had never received the diagnosis. The reason is straightforward: a woman on hormonal contraception has a regular withdrawal bleed that masks oligo-anovulation, and her pill also suppresses androgen production enough to hide the biochemical picture.
If you suspect PCOS, the diagnostic path matters. Ask a clinician for total testosterone, free testosterone, a free androgen index, and a transvaginal or transabdominal pelvic ultrasound — and confirm whether you need to stop hormonal contraception first. Stopping the pill for three months before testing is uncomfortable and requires a discussion about pregnancy risk, but testing on the pill often produces a falsely normal result. That single decision — whether to test while on contraception — is the most common reason the diagnosis is missed in the first place.
Symptoms by System: What Actually Shows Up in a Clinic
PCOS presents across four systems, and the 2023 guideline organizes assessment by symptom domain rather than by a single hallmark. Reproductive symptoms include cycles longer than 35 days or fewer than eight periods a year, heavy unpredictable bleeding, and difficulty conceiving. Dermatologic signs include hirsutism scored above 8 on the modified Ferriman-Gallwey scale, acne along the jawline and back that persists past the mid-twenties, and male-pattern scalp hair thinning. Metabolic signs include central weight gain, acanthosis nigricans — the dark velvety skin folds at the neck and underarms — and difficulty losing weight despite effort. Psychological symptoms include anxiety and depression, which the 2023 guideline elevated to a routine screening item because prevalence of depressive symptoms in PCOS runs higher than in matched controls.
Hirsutism scoring deserves a specific note because women often self-assess incorrectly. The modified Ferriman-Gallwey scale scores nine body areas from 0 to 4, with a total above 8 indicating hirsutism in most populations. That number differs by ethnicity, and the guideline flags it as a limitation, since East Asian and Northern European populations often show terminal hair at lower scores. Terminal hair — coarse and pigmented — is the medical signal; fine vellus fuzz is not.
The acne pattern is another diagnostic clue that gets missed. PCOS acne tends to concentrate on the lower face, jawline, and upper back, and it often resists standard topical therapy because the driver is androgen, not bacteria alone. A woman in her thirties with persistent jawline acne and irregular periods should be treated as a likely PCOS presentation until proven otherwise.
Weight is a trigger for stigma, and the guideline explicitly says PCOS occurs across all body sizes. Lean PCOS is real, and its metabolic risks differ from those of higher-BMI presentations in degree, not kind.
Screening for Diabetes, Lipids, and Blood Pressure by Age and BMI
The 2023 guideline gives a specific screening ladder. All women with PCOS should have blood pressure measured at every visit and a lipid panel checked at diagnosis, then repeated based on cardiovascular risk factors. For glucose tolerance, the recommendation is an oral glucose tolerance test — not fasting glucose alone — at diagnosis, then repeated every one to three years depending on age and BMI. Fasting glucose and HbA1c miss post-load hyperinsulinemia, which is the first abnormality to appear in many PCOS patients.
| Screening test | Who needs it | Repeat interval |
|---|---|---|
| Oral glucose tolerance test | All women with PCOS at diagnosis | Every 1–3 years, sooner with higher BMI or family history |
| Fasting lipid panel | All women at diagnosis and every visit | Every 1–2 years based on cardiovascular risk |
| Blood pressure | All women at every clinical visit | Every visit |
| Apnea screening (STOP-BANG) | BMI ≥30 or symptoms | Annually if positive |
| Depression/anxiety screening | All women at diagnosis | Annually or as indicated |
The 2023 guideline's insulin resistance statement is worth stating plainly: routine metformin is not recommended for every woman with PCOS, only where glucose tolerance is impaired or where the patient's specific presentation warrants it. This is where many online advice pages diverge from the actual evidence review.
Obstructive sleep apnea deserves attention because it's under-screened. The guideline recommends the STOP-BANG questionnaire for women with PCOS and a BMI of 30 or above, since sleep apnea worsens insulin resistance and fatigue in a way that gets attributed to "just PCOS."
Cardiovascular risk accumulates earlier in PCOS. The guideline recommends considering the condition when assessing cardiovascular risk from age 30, rather than waiting for the standard older-age thresholds.
Treatment by Goal: Fertility, Cycle Control, and Metabolic Health
Treatment in PCOS is goal-directed, and the 2023 guideline organizes recommendations around whether the patient wants pregnancy, cycle regulation, or metabolic improvement. The combined oral contraceptive pill remains first-line for cycle control and hyperandrogenism in women not seeking pregnancy, with low-dose formulations containing estrogen preferred. For hirsutism, the pill alone works on the underlying androgen excess but often needs twelve months to show full effect, and topical eflornithine can be added for facial hair.
Fertility treatment follows a defined sequence. The guideline recommends letrozole as first-line ovulation induction, ahead of clomiphene citrate, based on trials showing higher live birth rates. For women who do not ovulate on letrozole, adding metformin is reasonable, and if those fail, the next step is typically laparoscopic ovarian drilling or gonadotropins. Weight loss of 5 to 10 percent of body weight, where overweight applies, is the guideline's recommended first intervention before or alongside ovulation induction — and that one number matters because it is the only intervention with evidence of spontaneous cycle restoration in a meaningful fraction of patients.
Metformin's role is narrower than internet advice suggests. It is first-line for women with PCOS and type 2 diabetes or impaired glucose tolerance. It may be offered for cycle regulation and weight management where lifestyle changes are insufficient. The guideline does not endorse metformin as a general first-line agent for every PCOS diagnosis, and the evidence review cites gastrointestinal side effects as a frequent reason for discontinuation.
Cosmetic and psychological support count as treatment. The 2023 guideline specifically recommends screening for anxiety and depression, and where present, offering psychological therapy. Laser hair removal earns a mention as an evidence-supported option for hirsutism, with the caveat that hormonal treatment should run alongside it or the hair will return.
Where Online PCOS Claims Fail an Internal Check
Medical misinformation around PCOS clusters in four predictable places, and knowing them saves a great deal of wasted money. The first is the "cure" claim. PCOS is a lifelong condition managed rather than cured, and no supplement, detox, or dietary protocol reverses it. The 2023 guideline explicitly notes the low quality of evidence for most supplement claims.
The second failure point is estradiol or inositol supplement claims. Myo-inositol and D-chiro-inositol have some evidence for improving insulin sensitivity and ovulation in PCOS, but the guideline rates that evidence as low to moderate quality and inconsistent across trials. A 2022 meta-analysis in the journal Nutrients found heterogeneity between studies that prevents a firm dose recommendation. That doesn't mean inositol is useless; it means a specific dose promise is unsupported.
The third is the testing gap. Many online services sell an anti-Müllerian hormone test as a "PCOS diagnosis." The 2023 guideline explicitly rejects AMH as a stand-alone diagnostic tool because assay variation between laboratories is too wide. A high AMH can support PCOS as a feature but cannot confirm or exclude it.
The fourth is the "you must lose weight before treatment" narrative. The guideline recommends 5 to 10 percent weight loss where overweight applies as a first intervention, but it does not deny treatment until weight is lost, and it explicitly warns that weight stigma harms care. A woman with a BMI of 34 who wants to conceive should not be told to return when she's thinner.
What the Long-Term Evidence Says About Pregnancy, Endometrial Risk, and Aging
Long-term risk in PCOS is real but specific, and it's often inflated or understated depending on the source. Endometrial cancer risk is elevated, driven by unopposed estrogen from chronic anovulation — the 2023 guideline recommends that women with fewer than four periods a year should have endometrial protection via cyclic progestin or a combined contraceptive, or endometrial biopsy if bleeding is abnormal. That specific threshold — four periods a year — is the number clinicians act on.
Pregnancy in PCOS carries elevated risk of gestational diabetes, preeclampsia, and preterm birth. The guideline recommends early screening for gestational diabetes with an oral glucose tolerance test at 24 to 28 weeks, and often earlier in women with strong risk factors. Preconception counseling is recommended.
Gestational diabetes rates in PCOS pregnancies run roughly double the background rate in most cohort studies cited in the guideline's evidence review, though the exact multiplier varies by BMI and by cohort.
Type 2 diabetes risk rises across the lifespan. A longitudinal analysis by the guideline group found roughly a 4-fold increased risk of type 2 diabetes in women with PCOS compared with matched controls, with the risk climbing faster in the years following diagnosis. This is the strongest argument for the screening ladder above being followed rather than skipped.
Menopause does not end the story. Symptom expression often lessens after menopause because ovarian androgen production declines, but the metabolic and cardiovascular risk persists and may accelerate, which is why the guideline recommends that the diagnosis follows the patient rather than being closed.
FAQ
Does a normal ultrasound rule out PCOS?
No — a woman can meet Rotterdam criteria without polycystic ovarian morphology, and that is what phenotype B describes. Roughly 10 to 20 percent of diagnosed cases fall into phenotype B, with irregular ovulation and hyperandrogenism but normal-looking ovaries. The diagnosis hinges on two of three features, so the absence of one finding never excludes it on its own. A clinician must still check testosterone and cycle history before ruling out.
Can you have PCOS with regular periods?
Yes, and this is phenotype C. A woman with regular cycles, high androgens, and polycystic ovarian morphology meets criteria even though the classic symptom — missed periods — is absent. These patients are the ones most often dismissed, yet their androgen excess carries the same dermatologic and metabolic consequences. If acne, hirsutism, or scalp thinning is present, cycle regularity is not enough to close the question.
How many women with PCOS actually go undiagnosed?
Up to roughly 70 percent, according to a 2016 analysis in the Journal of Clinical Endocrinology & Metabolism using an Australian cohort dataset. The two main masking factors are hormonal contraception, which hides cycle irregularity and lowers measured androgens, and normal-looking ovaries, which mislead clinicians who rely on ultrasound alone. That gap is the strongest argument for asking directly for a Rotterdam-based workup.
Is anti-Müllerian hormone a PCOS test?
Not a stand-alone one. The 2023 international guideline explicitly declines to endorse AMH as a diagnostic threshold because assays vary between manufacturers, whereas follicle count remains the reference standard on modern ultrasound. A high AMH can support the clinical picture and is useful in research and in predicting ovarian response, but a clinician should not assign the diagnosis from an AMH value alone. Ask for a follicle count in the report if morphology is in question.
Do you need to lose weight to start PCOS treatment?
No — and the guideline advises against weight-gating care. A 5 to 10 percent weight reduction, where overweight applies, is recommended as a first intervention with evidence for restoring spontaneous cycles, but the guideline's patient-partnership chapter specifically names weight stigma as a barrier to care. A clinician who declines ovulation induction or metformin pending weight loss is not following the international recommendations.
What is the difference between polycystic ovaries and PCOS?
Polycystic ovarian morphology is a scan finding, present in 20 to 30 percent of normal ovulating women, and it is one of three possible diagnostic features. PCOS is the endocrine syndrome defined by the Rotterdam combination of ovulation, androgen, and morphology findings after exclusions. Confusing them is the single most common naming error in online discussions and in casual clinical shorthand.
Does PCOS go away after menopause?
Symptom burden often eases as ovarian androgen production falls, but the metabolic risk — insulin resistance, type 2 diabetes, cardiovascular risk — persists and may accelerate. The 2023 guideline recommends continuing longitudinal screening rather than closing the diagnosis at menopause. Women whose periods normalized after menopause should not assume their risk profile normalized with them.
What is the first-line fertility drug for PCOS?
Letrozole, based on trials showing higher live birth rates than clomiphene citrate, and the 2023 guideline recommends it as first-line ovulation induction. Clomiphene remains an option where letrozole is unavailable or not tolerated. Metformin is added when ovulation is resistant, and laparoscopic ovarian drilling or gonadotropins come after that.
This article is general educational information about a medical condition and is not a diagnosis, treatment plan, or a substitute for personalized medical advice. PCOS assessment requires a qualified clinician — ideally one who can order the Rotterdam workup and exclusion panel — and treatment decisions vary substantially with pregnancy intentions, age, and metabolic status. If you have thoughts of self-harm or are in immediate danger, contact your local emergency services or a crisis line in your country.