Lean PCOS: when slim women do not get the diagnosis

You are sitting in the examination chair and saying the same sentence for the third time in three years. My period is not coming. Or it comes whenever it feels like it.

Then comes the answer you already know. But you are slim. That is surely not PCOS.

Many women know this pattern. It is rarely meant unkindly. It is a mental shortcut, and it is understandable, because in textbooks and image searches polycystic ovary syndrome is almost always pictured together with excess weight.

Except that excess weight appears in not a single diagnostic criterion. Not in the Rotterdam criteria from 2004, and not in the international guideline from 2023.

This article follows the question of what is different in women of normal weight with PCOS and what stays the same. It takes the evidence as it is: surprisingly clear in some places, thin and contradictory in many others. I make exactly those differences visible instead of smoothing them over.

Slim and still PCOS: why the diagnosis so often comes later

There is a moment that many women with an irregular menstrual cycle describe. They sit in the waiting room and silently rehearse how to say it so that it gets taken seriously.

That is not imagination. And it is not an accusation against gynaecology either. It is what happens when a condition is coupled to a body image in the collective memory.

The numbers on diagnostic delay exist. They apply to polycystic ovary syndrome as a whole, not specifically to slim women. That needs saying before the number arrives.

That slim women in particular wait longer is not proven by this. It is my clinical observation, and I mark it here expressly as that.

What I observe clinically: when a woman with a cycle disturbance and a BMI of 21 comes to my consultation, she has often already heard several explanations. Stress. Sport. Too much coffee. She had a hormone panel in her hands considerably less often than a woman with the same set of complaints and a higher weight.

Why this happens has little to do with individual people and a lot to do with structure. In routine care, the time budget per appointment is small. Whoever has little time works with patterns, and patterns come from what one sees often. Excess weight and PCOS frequently occur together, so the mental link is an obvious one. It is simply not a rule.

I deliberately do not explain the Rotterdam criteria in detail again here. That is set out at length in the foundational article on polycystic ovary syndrome, causes and symptoms. If you have not been there yet, that is the better entry point. This text builds on it.

For me the 15.6 percent is the genuinely interesting number. Not the waiting time, but what is missing afterwards. A diagnosis without an explanation is a label. And that is exactly why I am writing this article at such length.

And now you know why the first hurdle in lean PCOS is often not the laboratory, but the reflex that comes before it.

What lean PCOS means, and why the term wobbles

You probably read the term in English first. It rarely turns up in German practices, but constantly in forums and on Instagram.

There is a simple reason for that. Lean PCOS is not a diagnosis of its own. It is a description.

What is meant are women who meet the diagnostic criteria and are of normal weight. There is nothing more behind it. In no guideline anywhere does this term appear as an entity in its own right.

And this is exactly where the problem starts that gets overlooked across the whole internet: if you read two studies on lean PCOS, they often do not mean the same group of women.

How many women with PCOS are actually of normal weight?

The honest answer is: there is no clean prevalence study on this. Precisely because of the definition question.

What does exist are study populations. And the largest one that was analysed separately by BMI is a genetic meta-analysis.

Together with the specialist literature, which describes at most about half of those affected in Europe as obese, this gives the range you read everywhere: depending on the study and the cut, roughly one in five to one in two women with PCOS is of normal weight.

I deliberately name a range and not a single number. Anyone claiming one single percentage here is pretending there is a measurement that does not exist.

Which phenotypes predominate in this group

Because two out of three criteria are enough for the diagnosis, there are four possible combinations. They are usually labelled A to D. A is the complete form with all three features. D is the form without raised androgens, so a cycle disturbance plus the ultrasound appearance.

In women of normal weight, the milder combinations are proportionally more common. That is hardly surprising, because the metabolic component is on average weaker here. But it has a consequence that becomes important in the next section: it is precisely the form without hyperandrogenaemia that is hardest to tell apart from a different cause.

I put this sentence early in the text on purpose. It is the frame for everything that follows. And now you know why a term that everybody uses still wobbles.

The central question: do slim women with PCOS have insulin resistance?

This is the question I am asked most often. And it is the question where the most fudging happens online.

You will find two camps. One says: insulin resistance is the core of PCOS, in lean women too, always. The other says: insulin resistance is an excess weight problem, it is not an issue for you.

Both are partly right and both simplify. So I sort the data in order, from the best method downwards.

The gold standard: what the clamp studies show

The most precise way to measure insulin sensitivity is called the euglycaemic hyperinsulinaemic clamp. Put simply: blood sugar is held constant over hours and it is measured how much sugar has to be infused to do that while insulin is running. Whoever needs little responds poorly to insulin.

This procedure is elaborate and is therefore rarely done. But there is a meta-analysis that pooled all the available work.

That is the strongest finding for the statement that slim women can be affected too. It does not say, however, that every one of them is. An average is not an individual fate.

The famous 75 percent figure, and what stands next to it

When you read on English language advice sites that 75 percent of women with lean PCOS are insulin resistant, that number almost always comes from a single paper. Those sites usually do not name the source. I do it the other way round: here is the source, and here is the number that stands next to it and changes everything.

How does that fit together? Rather well, if you let both statements stand side by side instead of deleting one of them.

Compared with equally lean women without PCOS, insulin sensitivity is on average lower. Compared with obese women with PCOS, it is on average better. Both at the same time.

How insulin reaches into the hormone balance at all I explain here in only three sentences. In the liver, insulin can damp the production of SHBG, the transport protein for sex hormones. Less SHBG can mean more free testosterone in the blood, even if total testosterone stays the same. In addition, insulin can stimulate androgen production in the ovaries. This pathway is considered mechanistically well studied, which says nothing about the extent to which it comes into play in an individual woman. You will find the whole connection with all its intermediate steps in the article on insulin resistance and hormones in women.

There is one more finding that matters to me, because it explains why the standard piece of advice so often misses the mark in slim women.

And now you know why the right answer to this question is neither yes nor no, but: smaller, yet real, and to be checked individually in every woman.

If the BMI is normal, where does it sit then? Body composition, honestly

There is one sentence that appears on almost every page about lean PCOS. It runs roughly like this: even at a normal weight you have hidden belly fat around the organs, and that is the explanation.

The sentence is catchy. It answers the question without anything having to change about the picture of polycystic ovary syndrome. Except that the evidence carries it less well than it looks.

I will show you why, because this is one of the few places where the measurement method tips the result.

It goes even further. There is a small paper that found exactly the opposite.

What differs instead

So if it is not the amount of fat tissue, then what is it? The most interesting trail leads to the function of fat tissue.

Fat tissue is not a storeroom. It is a hormonally active organ. Among other things it releases leptin, which reports satiety and energy availability to the brain, and adiponectin, which supports insulin sensitivity in the tissue. The ratio of these two to each other says more than fat mass alone.

One more detail that fits the same picture well. A position paper from three US professional societies (Goodman NF et al., Endocr Pract 2015, PMID 26642102) describes that young, lean women affected showed smaller HDL particles, more VLDL particles and more LDL particles compared with controls matched for BMI and age. The cholesterol value in a normal laboratory panel can look unremarkable in the process. What differs is the composition, not necessarily the sum.

That is exactly the pattern of this whole section. On the coarse level everything looks normal. On the finer level something shows up. And the coarse level is precisely the one measured in routine care.

And now you know why, in a slim woman with PCOS, I do not go looking for hidden belly fat but for signals from the tissue.

The hormonal profile: LH, AMH and the adrenal gland

When you have your laboratory report in front of you, you see a row of abbreviations. LH, FSH, AMH, DHEAS, androstenedione. For most women that is a wall of letters.

In lean PCOS it is worth looking exactly here. Because something shifts here, and it does so systematically.

A short orientation. LH and FSH are the two control signals from the pituitary gland. They tell the ovary what to do. Put simply: FSH lets follicles mature, LH triggers ovulation and stimulates hormone production in certain cells. If LH calls too loudly relative to FSH over a longer period, an imbalance can arise that stimulates androgen production in the ovary and can bring maturation to a standstill.

DHEAS is the value where you are most likely to see this in the laboratory. It comes almost entirely from the adrenal cortex and barely fluctuates across the menstrual cycle. That is exactly why it is a usable marker when the question is where the androgens come from.

The large meta-analysis from 2025 confirms the neuroendocrine part of this picture.

AMH nevertheless remains an important value. It comes from the small follicles and reflects how many of them are currently present. The 2023 guideline now allows AMH in adults as an alternative to ultrasound when the question of polycystic ovarian morphology is on the table.

The authors of the large genetics analysis summarise the same thing from their perspective. They describe lean women affected as typically having higher gonadotropin ratios, less insulin resistance, higher androgens including the adrenal androgens and more favourable lipid profiles.

That also explains why a standard hormone panel in a slim woman can sometimes look unremarkable even though something is going on. If only total testosterone and TSH are measured, exactly those values that would be most likely to stand out in her are missing.

What belongs in a sensible panel and when it is taken is set out at length in the article testing hormones in women. I am not repeating it here.

And now you know why, in your case, the right values may simply not have been on the form yet.

The most important mix-up: functional hypothalamic amenorrhoea

If you read only one section of this article, please make it this one.

There is a second condition that is very common in slim women and that can look confusingly similar to polycystic ovary syndrome on ultrasound. It is called functional hypothalamic amenorrhoea.

Put simply: the brain turns reproduction down because it rates the energy situation in the body as insecure. The pacemaker in the hypothalamus pulses more slowly, LH and FSH fall, estradiol drops, the period stays away. This is not a defect. It is a protective circuit.

It is typically triggered by too little available energy, by a high training load, by strong psychological stress or by a combination of these. Often the woman herself does not notice that a gap exists, because she eats enough, just not enough for her level of activity.

Now comes the point that appears practically nowhere in the German language internet.

So how are the two told apart? There are numbers for this, from a paper that made exactly this comparison.

And now you know why, in a slim woman whose period is absent, I first ask about the course of events and only then look at the ultrasound image.

What has been studied when losing weight is not a lever

Now comes the section you were probably looking for first. And I have to disappoint you a little right at the start.

The lifestyle evidence in polycystic ovary syndrome has largely been collected in mixed groups in which excess weight predominated, and the best documented endpoint is frequently a change in weight. For women of normal weight, exactly that endpoint falls away. What then remains is thinner than the advice landscape suggests.

I sort it anyway, and honestly by level of evidence.

Exercise

On strength training you will find many clear statements online. A solid meta-analysis specifically for lean women affected does not exist. That is the honest information.

Mechanistically plausible, human studies for this subgroup thin: muscle is the body's largest consumer of glucose. More muscle mass means more storage space and more insulin independent uptake during exertion. That this changes the hormonal situation in a slim woman with PCOS is a reasonable assumption. It has not been tested in studies for this group. Strength training can still make sense for other good reasons, for bone, for strength in everyday life and for sleep.

Nutrition

For the lean subgroup there are no randomised nutrition studies. What exists is a review arguing for phenotype specific strategies without having solid intervention data in hand. I name it exactly like that, because such reviews are often quoted online as if they were study results.

What remains is the general level: a way of eating that keeps blood sugar calmer across the day is broadly plausible in PCOS and has not been tested specifically in lean women affected. How to approach that in practice without drowning in rules is set out in the article on blood sugar spikes and how to avoid them and in the article on cycle based nutrition. Figures on energy amounts are deliberately absent there, and in this text too.

Inositol

Sleep

The stress axis

How chronic stress reaches into the menstrual cycle is set out at length in the article cortisol, stress and female hormones. Here it is enough to note that the stress axis and the reproductive axis share the same control region in the brain.

Medication

Metformin, antiandrogens and the combined pill are used in PCOS depending on the pattern of complaints. All three are prescription only. They appear here for orientation, without dosing and without any judgement on the individual case.

In Germany metformin is licensed only for type 2 diabetes. Its use in PCOS is therefore off-label. It has to be discussed with you and documented by the prescribing practice, and statutory health insurance usually does not pay for it.

Each of these medicines carries risks that belong in the conversation. Metformin can cause gastrointestinal complaints and is not suitable when kidney function is impaired, because in rare cases lactic acidosis can occur. It is usually paused before operations and before examinations using iodine containing contrast media. Antiandrogens can harm an unborn child, which is why reliable contraception is mandatory alongside them, and for cyproterone acetate there is an official warning about the risk of meningioma. The combined pill can raise the risk of thrombosis, especially in smokers, in migraine with aura and after a previous thrombosis. I deliberately name no doses here.

Three sentences on this that I consider important. First: none of these medicines is started, changed or stopped on your own initiative, and certainly not because of a blog article. Second: if you take the pill for contraception, every change also affects contraceptive protection, and that belongs in the conversation with your gynaecologist. Third: if you are trying to conceive, time is a real factor, and a reproductive medicine work-up should not be postponed because you want to try something else first.

And now you know why an article that promises you a finished programme here informs you less well than one that names the gap.

What I also look at: inflammation, endocrine disruptors, exposure

If you leave an appointment with me holding a laboratory form with things on it you have never seen before, this section is the explanation for that.

Two clarifications first, so that the frame is right.

First: the gynaecological work-up remains the foundation. Ultrasound, hormone panel, exclusion of other causes. Everything I describe here comes on top of that, not in its place.

Second: environmental factors are not part of the recommended diagnostic pathway in the international guideline from 2023. That is not negligence on the part of the guideline authors, it is the logic of their procedure. They assess by GRADE and prefer endpoints from intervention studies. For endocrine disruptors and heavy metals in PCOS, essentially cross-sectional and case control studies exist. From associations without demonstrated benefit of a treatment, no recommendation can be derived. That is consistent and right.

The guideline is therefore not saying that environmental factors are irrelevant. It is saying that the data are not sufficient for a recommendation. That is a difference, and I make it visible here instead of concealing one of the two sides.

The finding that convinces me most

This is not an alternative medicine add-on. This is a finding from the standard literature, published in a regular specialist journal. It calls for an explanation that does not hang on weight. On the link between silent inflammation and metabolism we have a separate article: silent inflammation and weight.

Endocrine disruptors in everyday life

Xenoestrogens are substances that can dock onto oestrogen receptors in the body even though they are not the body's own hormones. They sit in plastics, coatings, cosmetics and packaging. How they occur in everyday life and where something can be changed with little effort is set out at length in the article xenoestrogens, the hidden endocrine disruptors in everyday life.

How does the specialist world classify something like this? There is an extensive scientific statement from the Endocrine Society for that, known as EDC-2. It describes non-monotonic dose response curves, effects in the low dose range and a particular vulnerability during developmental phases. The authors consider causal relationships in animal models to be established and the epidemiological human data to be consistent. In the same breath they expressly urge caution about inferring causality in humans.

This double statement is the whole point. The mechanism is to be taken seriously. Proof in humans is missing. Both at the same time.

Mould and mycotoxins

Zearalenone is a metabolic product of certain moulds and is described as a mycoestrogen, meaning a substance with oestrogen like properties. A systematic review pooled 104 papers from cell culture and animal models on this and found consistent effects on processes and tissues of female reproduction. The same paper expressly records that it could not include a single epidemiological study in humans meeting its criteria. Important for placing it: this oestrogen like action therefore comes from cell culture and animal experiments. There are no intervention studies in humans showing an effect on the menstrual cycle, and a link with polycystic ovary syndrome has not been demonstrated in humans. Everything further on this is in the article zearalenone, the mycoestrogen and what it means for hormones, and if you are wondering whether mould in your home could play any role at all, you will find the symptom overview under mould symptoms by organ system.

Heavy metals

Whether a measurement makes sense in an individual case, and which material is even suitable for it, is a separate question. Blood, urine and hair each say something different, and hair analyses are frequently over-interpreted. That is set out at length under measuring heavy metals: blood, urine or hair.

And now you know why an inflammatory marker sometimes appears on the form in my practice. What such a value means for treatment is expressly not answered by that. To this day there is no evidence of benefit for treating environmental exposures in PCOS. An additional value can open a question. It promises you nothing.

The diagnosis: which values, which timing, and what the pill changes

You have read this far. Now comes the practical part, and it is deliberately written so that you can take it into a consultation instead of diagnosing yourself.

Because that is the boundary I do not shift here: PCOS is a medical diagnosis. It is made through a work-up, because other causes have to be ruled out. A thyroid disorder, a raised prolactin, a late onset adrenogenital syndrome and rarer causes look similar from the outside and are treated differently.

The timing

Hormones are not static quantities. They are a daily course and a cycle course. That is why the time of the blood draw is part of what a value means.

The early follicular phase is usual, roughly cycle day two to five, counted from the first proper day of bleeding. LH, FSH, estradiol and prolactin are typically measured there. Androgens such as total testosterone, SHBG with the free androgen index calculated from it, androstenedione and DHEAS are frequently drawn at the same time. The thyroid is routinely added.

If your period does not come at all, this counting falls away. Measurement is then done independently of the day, after discussion, and the interpretation happens through the overall picture rather than through a reference range for a particular cycle day.

Progesterone belongs in the second half of the cycle, roughly seven days before the expected bleeding. It answers a single but important question: has ovulation taken place. A single low value on the wrong day answers nothing at all.

What many laboratory forms for slim women do not contain

From what you read in the previous sections, a practical consequence follows. If the findings in lean women affected lie more on the neuroendocrine and adrenal side, then those are exactly the values missing from a minimal panel.

Concretely: whoever measures only total testosterone and TSH does not see the LH to FSH ratio, does not see androstenedione and does not see DHEAS. Without SHBG the free androgen index cannot be calculated, and that is often more informative than total testosterone alone.

This is not a criticism of anyone. A basic panel is a sensible first step. It is simply a first step.

Fasting glucose or tolerance test

The same principle applies to metabolism. Fasting blood glucose is the most convenient value and the least sensitive one. In the Korean paper with 194 women affected, the two hour value in the oral glucose tolerance test was the best single screening parameter. Among the lean women affected, the rate of glucose disorders there was 5.9 percent, and thus 9.8-fold higher than in women of the same age in the general population. (Lee H et al., Endocrine 2009, PMID 19688613)

The same position paper from three US professional societies (Goodman NF et al., Endocr Pract 2015, PMID 26642102) also expressly records that clamp investigations have shown a degree of insulin resistance in both obese and lean women with PCOS. Whether a tolerance test, a fasting insulin or none of it makes sense for you belongs in the medical consultation. That is a weighing up and not a rule.

And if I am on the pill?

Then the informative value of a hormone panel is limited, and that is not a detail.

Combined preparations can lower LH and FSH and raise SHBG in the liver. A higher SHBG binds more testosterone, so the free androgen index can come out lower, quite independently of what the situation would look like without the pill. An unremarkable androgen profile on the pill therefore proves little.

What does not follow from this: that you should stop the pill in order to finally be able to measure. Whether, when and within what framework a break before diagnostics makes sense is decided by the practice treating you, and the question of contraception is a necessary part of that. What happens in the body after stopping and why the first months say little is set out in the article stopping the pill, what happens in the body.

And now you know why a normal BMI does not end the work-up but only shifts its direction a little.

Frequently asked questions about lean PCOS

These are the questions that come up most often in my consultations and in my inbox. They do not answer an individual situation, they give the general classification.

Where this topic connects to the rest of your body

Lean PCOS does not stand on its own. It hangs on blood sugar, on the stress axis, on the thyroid, on the skin and on how a woman deals with an unclear finding. Here are the articles that take things further in each case.