PCOS: A Practical Guide to a Misunderstood Diagnosis

Polycystic ovary syndrome is one of the most common hormonal conditions in women, affecting between 8 and 13 in every 100 of reproductive age. It is also the one women are most often told about badly. In a two-minute conversation at the end of a scan, with the words "you have cysts" and little else.

Almost everything difficult about PCOS comes from the name. It suggests a problem with the ovaries, that the problem is cysts, and that it is one disease. None of those is quite true.

What is actually happening#

PCOS is a hormonal and metabolic condition in which the monthly conversation between the brain and the ovaries goes out of step.

In a typical cycle, the pituitary releases follicle-stimulating hormone (FSH) and luteinising hormone (LH) in a particular rhythm. FSH recruits a group of follicles. Small fluid-filled sacs, each holding an immature egg. One becomes dominant and LH surges to release it. The emptied follicle makes progesterone, which stabilizes the womb lining and, if there is no pregnancy, withdraws to trigger a period.

In PCOS that sequence stalls. LH runs high relative to FSH, and the ovary makes more androgens. Testosterone and related hormones all women produce in small amounts. Higher androgens inside the ovary interfere with follicle maturation, so instead of one follicle taking the lead, a dozen or more sit part-grown. Those stalled follicles are the "cysts" on the scan. They are not cysts as most people picture them. They are ordinary egg sacs that have paused, and they are not dangerous, painful, or something to be removed.

Because no follicle is released there is no progesterone surge, so the period does not come. And because the follicles remain, androgens stay up and the cycle stays stuck. It is a loop.

Insulin sits on top of this. Many women with PCOS have insulin resistance: tissues respond poorly to insulin, so the pancreas makes more of it. High circulating insulin does two things. It pushes the ovary to make more androgens, and it lowers a liver protein called sex hormone-binding globulin (SHBG), which normally mops up free testosterone. Less SHBG means more active testosterone even when the total on the blood test looks unremarkable. Insulin is not a side story in PCOS. For many women it is the engine.

How PCOS is actually diagnosed#

The international standard is the Rotterdam criteria. You need two out of three:

1. Irregular or absent ovulation. Cycles consistently longer than 35 days, fewer than about eight bleeds a year, or long unpredictable gaps.
2. Signs of high androgens. Clinical (excess coarse hair in a male pattern, persistent adult acne, male-pattern hair thinning) or biochemical (raised free or total testosterone, or free androgen index).
3. Polycystic ovarian morphology on ultrasound. A high follicle count or enlarged ovarian volume, on current thresholds. In adults, a raised anti-Müllerian hormone (AMH) level can be used in place of the scan.

And crucially: other explanations must be excluded first. Thyroid disease, raised prolactin, non-classic congenital adrenal hyperplasia, Cushing's syndrome and androgen-producing tumors all imitate PCOS. A diagnosis made from a scan alone is not a diagnosis.

Two points get missed. Ultrasound should not be used to diagnose PCOS within roughly eight years of a first period, because polycystic-looking ovaries are normal in adolescence. And you need only two criteria, so a woman with irregular cycles and clear hirsutism has PCOS on a normal scan.

The four patterns#

Meeting two of three criteria can happen in four ways, and they behave differently. This is the most useful thing to understand about PCOS.

Two women can both be correctly told they have PCOS and need almost nothing in common in their care. Pattern C with normal glucose and no plans to conceive needs a very different plan from pattern A with prediabetes and three periods a year.

The metabolic side almost nobody mentions#

This is what gets left out of the two-minute conversation, and what determines your health at 55.

Insulin resistance is present in a majority of women with PCOS, estimates commonly range from 65 to 80 per cent, and it occurs independently of weight. Excess weight worsens it, but slim women with PCOS show it at higher rates than slim women without. This is the biggest reason "lean PCOS" gets under-investigated.

What that raises the risk of, over years:

• Impaired glucose tolerance and type 2 diabetes, well above background risk and at younger ages.
• Gestational diabetes, which is why PCOS should be flagged early in pregnancy care.
• Fatty liver disease, often picked up as mildly abnormal liver enzymes.
• Dyslipidaemia, higher triglycerides, lower HDL, and higher blood pressure.
• Obstructive sleep apnoea, independently of weight, and frequently missed in women.
• Endometrial hyperplasia, from a lining exposed to estrogen for long periods without the progesterone that normally sheds it.

A 75 g oral glucose tolerance test detects abnormal glucose handling in PCOS better than fasting glucose or HbA1c alone, because the earliest problem is often a high two-hour value with a normal fasting number. Guidance suggests repeating metabolic screening every one to three years. Ask about it specifically, because a busy clinic defaults to a single fasting glucose.

Fertility: the honest version#

PCOS is the most common cause of infertility due to not ovulating. That is not the same as being infertile.

The problem is the supply of ovulation, not the quality of eggs. Women with PCOS tend to have a larger remaining egg pool and often reach menopause slightly later. If ovulation can be made to happen reliably, the rest of the machinery generally works.

What that means practically:

• If your cycles are highly irregular, ovulation predictor kits and app "fertile windows" are unreliable, because they assume a pattern you do not have. Many women spend months timing intercourse to a window that never opened.
• Modest weight reduction where there is excess weight, around 5 to 10 per cent, restores ovulation in a meaningful proportion of women. Real, but not universal, and never a reason to delay referral.
• Ovulation induction is the standard next step. Guidelines now favor an aromatase inhibitor over the older estrogen-receptor drug as first-line, on live birth rates. These are prescription decisions needing your own doctor, because the main risk is multiple pregnancy.
• If that does not work: gonadotrophins, ovarian drilling, IVF. Women with PCOS respond strongly to IVF stimulation, so hyperstimulation risk must be managed carefully.

One thing worth saying plainly: irregular cycles are not contraception. Ovulation in PCOS is unpredictable, not absent.

Hair, skin and how long change actually takes#

For many women this is the part that affects daily life most, and the part most likely to be brushed aside as cosmetic. It is not cosmetic. It is a visible sign of an internal hormonal state, and the distress it causes is measurable.

Hirsutism is coarse, dark hair in an androgen-dependent pattern: upper lip, chin, jawline, chest, around the nipples, lower abdomen, inner thighs, lower back. It is graded by a formal scoring system whose threshold varies with ethnicity. The same amount of hair means different things in a South Asian, Mediterranean, East Asian or Northern European woman. A doctor who ignores that over-diagnoses some women and dismisses others.

Acne in PCOS is persistent and adult-onset or adult-continuing, clustering along the jaw, chin and neck. Androgenic hair thinning affects the crown and part line, usually sparing the frontal hairline, and is often noticed first by a hairdresser.

Acanthosis nigricans, velvety dark thickening at the back of the neck, armpits or groin, is not a hygiene problem and does not scrub off. It is a visible marker of high insulin, and finding it should always prompt metabolic testing.

What helps, broadly: combined hormonal contraception raises SHBG and lowers free androgens; anti-androgens block the receptor and are used with reliable contraception because of effects on a male fetus; insulin-sensitising treatment helps some women; topical treatments exist for facial hair. For lasting hair reduction, laser and electrolysis do more than any tablet, best started alongside hormonal treatment rather than instead of it.

The timeline is the thing to hold on to. Hair follicles turn over on a three to six month cycle, so no androgen-directed treatment can show its effect faster than that. Acne responds sooner. Hair thinning is slowest of all. Most women who conclude a treatment failed stopped it at eight weeks.

What I actually see in clinic#

The consultation I have most often is not about ovaries. It is a woman in her late twenties or thirties who has been told she has PCOS, been told to lose weight, tried, and arrived believing the whole thing is her fault.

Two things are usually true. Her metabolic risk has never been properly assessed, often a single fasting glucose years ago, no lipids, no blood pressure record, no discussion of sleep. And nobody has explained the loop: that insulin drives androgens, androgens make ovulation harder, and the metabolic pattern of PCOS itself makes weight easier to gain and harder to lose. When people understand a mechanism, the self-blame drops away, and that is usually when they start doing the things that help.

The other consultation is the woman with completely regular periods, told for a decade that her jawline acne and facial hair are just how she is. She has pattern C, nobody ever measured an androgen level, and she has spent years feeling vain for minding.

One more thing: the psychological load of PCOS is underestimated everywhere. Depression, anxiety and disordered eating are all more common in PCOS, and international guidance recommends actively screening for them. If your appointments have never touched on how you are coping, that is a gap.

What genuinely helps, and what is oversold#

Things with real evidence behind them:

• Resistance training and regular aerobic activity. These improve insulin sensitivity partly independently of weight change, which matters when the scale refuses to move. One of the few interventions that works on the mechanism directly.
• Sleep. Short and disrupted sleep worsens insulin resistance within days. Treating sleep apnoea, where present, changes metabolic numbers.
• A dietary pattern you can sustain. Trials comparing low-carbohydrate, Mediterranean and low-glycaemic-index approaches in PCOS show broadly similar results. Adherence beats composition. There is no PCOS diet.
• Cycle regulation and endometrial protection where cycles are very infrequent. A conversation with your own doctor.
• Stopping smoking and moderating alcohol, on cardiovascular grounds.

Oversold: detox and "hormone balancing" protocols, expensive food-intolerance panels, supplement stacks marketed for PCOS with no trial data, and the claim that any diet cures PCOS. Inositol is the honourable exception, with genuine if modest trial evidence, and is worth raising with your doctor rather than buying off an advert.

At work#

Two things matter if you work shifts or in an industrial setting.

Night and rotating shift work independently worsens insulin resistance and disturbs the circadian signals that already run irregularly in PCOS. In occupational health I see this as a compounding effect: the same woman's glucose and cycle regularity are measurably worse on nights than on a fixed day pattern. If your metabolic screening was done during a run of day shifts, it may look better than your real baseline.

Second, unpredictable heavy bleeding after a long gap is a practical workplace problem, not only a medical one, particularly with restricted toilet access, long fixed shifts, PPE or hot environments. It is legitimate to raise with occupational health, and usually more fixable than people expect.

Getting the right assessment#

If you were diagnosed loosely, or suspect PCOS and have never been assessed, reasonable things to ask about:

Rapidly progressive androgen symptoms are the exception. Coarse hair growth, voice deepening or clitoral enlargement developing over months rather than years is not typical PCOS and needs prompt assessment for an androgen-producing tumor. It is rare, and it is the one thing here that should not wait for a routine appointment.

The bottom line#

PCOS is not a cyst problem and not one disease. It is a spectrum of four patterns that share a name and little else. The diagnosis needs two of three criteria and the exclusion of imitators, so a scan alone is never enough. For most women the long game is metabolic rather than gynecological, so glucose, lipids, blood pressure and sleep deserve as much attention as cycles, whatever your body size. And if fertility is the worry, the picture is better than the internet suggests: this is usually a problem of unreliable ovulation, which is one of the more treatable things in medicine.