Is PCOS Hereditary? What Twin Studies and Family Risk Data Show

Does PCOS run in families, or is that a coincidence?

Sisters of a woman with PCOS are affected at roughly eight times the rate seen in the general population, and mothers at about six times, in the largest study built specifically to answer this question.

That study evaluated 78 mothers and 50 sisters of 93 women already diagnosed with PCOS. Using the diagnostic definition of its era — irregular cycles plus clinical or biochemical androgen excess, with look-alike conditions ruled out — it found PCOS in 19 of the mothers (24%) and 16 of the sisters (32%), against a roughly 4% rate in the general population the researchers used for comparison. When the count was narrowed to relatives who were premenopausal and not on hormonal treatment that could mask symptoms, the rates rose further, to 35% of mothers and 40% of sisters.

That is a single-center U.S. study from 2001, run before the current Rotterdam-based diagnostic criteria existed, and it is small by modern genetics standards. It is still the clearest family-clustering evidence available, and later work has not overturned the basic pattern: PCOS is several times more common among the first-degree relatives of an affected woman than in the population at large.

What twin studies show about the genetic contribution

Identical twins resembled each other on PCOS-related traits almost twice as strongly as non-identical twins and sisters did, in a 2006 Dutch study of more than 3,200 twins — the kind of gap that only appears when a trait has a real genetic basis, because both twin types typically share a similar upbringing.

The study drew on 1,332 monozygotic (identical) twins and 1,873 dizygotic (non-identical) twins and singleton sisters registered with the Netherlands Twin Register, using cycles under nine per year plus acne or hirsutism as a self-reported proxy for PCOS rather than a full clinical work-up — worth flagging, since it is less precise than an in-person diagnosis. The correlation between identical twin pairs was 0.71; between non-identical pairs it was 0.38. Identical twins share essentially all their genes, non-identical twins share about half, and both grow up in comparably similar environments — so the extra resemblance in identical pairs points specifically to genetics, not to a shared household.

PCOS genetics: what the gene-mapping studies have actually found

Researchers have now identified six regions of the genome that reach genome-wide statistical significance for PCOS, in a 2015 study of more than 5,000 women with the condition and almost 83,000 controls, with a second stage adding roughly 2,000 more clinically confirmed cases.

The six regions sit in or near the genes ERBB4, YAP1, THADA, FSHB, RAD50 and KRR1. Three of the four human epidermal growth factor receptor genes turned up among the strongest signals, which the study’s authors flagged as a genuinely new lead rather than confirmation of an existing theory. Separately, the same analysis used Mendelian randomization — a method that can support a causal direction rather than a simple correlation — to link higher BMI, higher insulin resistance and lower sex-hormone-binding globulin to PCOS risk itself, not just to its downstream symptoms.

One limit worth stating plainly: the study’s primary cases were “self-reported cases of White European ancestry.” PCOS genetics research has not yet been built out equally across populations, so how well these specific six regions generalize to other ancestries is not yet settled. The same analysis also found that PCOS-risk gene variants track with higher AMH levels in girls before any diagnosis is possible — a thread picked up in the next section.

Will my daughter get PCOS? What the evidence on daughters actually shows

Daughters of women with PCOS had an average AMH level of 24.0 pmol/L, against 14.4 pmol/L in daughters of unaffected mothers — a difference that showed up before puberty was even complete, in a study of 28 daughters of PCOS mothers and 33 daughters of controls, aged 8 to 16.

The same daughters of PCOS mothers also had higher free androgen index, higher testosterone, higher two-hour insulin after a glucose load, and larger average ovarian volume than the daughters of unaffected mothers. That is a real biological signal, detectable years before a diagnosis could be made. It is not the same thing as a predictive test: the study is small, it measured group averages rather than individual risk, and — critically — the 2023 guideline’s adolescent-specific rules explicitly recommend against using AMH to diagnose PCOS at this age, precisely because puberty itself produces a wide, overlapping range of values.

What a family history does, and does not, tell you

Most people diagnosed with PCOS cannot point to another affected relative, because population rates of roughly 6–10% mean plenty of cases are the first in their family — a family history raises the odds, but its absence tells you almost nothing.

The reverse also holds: in the family study above, the majority of mothers (76%) and sisters (68%) of an affected woman were not themselves affected, even with a first-degree relative who was. A family history moves the probability; it does not fix the outcome. What sits between the genetic risk and the diagnosis is a set of modifiable factors — insulin sensitivity and weight among them — which is why two sisters carrying a similar genetic background can end up with different symptom severity, and why this is described as a metabolic condition shaped by biology and circumstance together, not a verdict. It is also why relatives who share the same genetic risk can land in different places on the four PCOS phenotypes — one androgen-dominant, another lean and ovulatory — rather than a single shared presentation.

There is also no blood test or genetic panel that diagnoses PCOS from DNA alone. Whatever your family history, the diagnosis itself is still made the way it is made for everyone: against the criteria on what actually gets tested, drawn at the correct point in the cycle, not from ancestry.

Your next step, this week

Write down which relatives had irregular periods, fertility difficulty, adult acne or excess hair growth, and roughly when it started. Bring that history — not a request for genetic testing — to your first PCOS-related appointment; it helps a clinician interpret borderline results faster than starting from nothing. If the question is really about a daughter rather than yourself, the age-specific criteria and testing timeline is the more useful next page, because family history there changes what to watch for, not what to test for immediately. How the rest of that work-up actually unfolds covers what happens after that conversation.

One more note on terminology: in May 2026, a global consensus process renamed the condition polyendocrine metabolic ovarian syndrome, or PMOS. Nothing about the genetics or the family-risk data above changed — only the label did. This article uses PCOS because that is still what most people search.