An international team has analysed genetic data from more than 544,000 women and identified 29 regions linked to polycystic ovary syndrome (PCOS) – 13 of them previously unknown. The results connect genetic susceptibility to PCOS with later menopause and inherited patterns related to type 2 diabetes and cardiovascular disease. Nevertheless, despite the well-known effects of PCOS on ovulation, the study found no evidence that this inherited susceptibility increases the likelihood of remaining childless.
When Irene was diagnosed with PCOS as a young woman, she was told that irregular periods could make getting pregnant difficult. Twenty years later, she still menstruates, although many of her friends have gone through menopause. She has also developed type 2 diabetes and heart disease.
Irene is fictional, but stories like hers capture the puzzle at the heart of PCOS: how can impaired ovulation, a longer reproductive window and metabolic disease be part of the same condition?
A new international study led by researchers at the University of Copenhagen offers part of the answer. By nearly doubling the number of known genetic regions linked to PCOS, the team reveals how its reproductive and metabolic sides may share some of the same biological roots.
The research, published in Nature Genetics, was coordinated by Tugce Karaderi, Associate Professor at the Center for Health Data Science at the University of Copenhagen (Denmark), in collaboration with colleagues from the University of Cambridge (United Kingdom) and the University of Utah (United States).
“We already knew that PCOS can be accompanied by other health problems, but the underlying biology and the processes in the body that lead to them are still only partly understood. Our findings shed more light on the possible mechanisms,” says Tugce Karaderi.
A search through half a million genomes
Tugce Karaderi and colleagues wanted to move beyond the patterns doctors already recognise among women with PCOS and identify inherited differences that might explain them.
They combined genetic data from biobanks and health registries worldwide, including 20,818 women with PCOS and 523,695 without the diagnosis. They then searched for small DNA variations that appeared more often among women with PCOS.
The principle is straightforward: if a DNA variation appears more often among women with PCOS than among those without it, the surrounding region becomes a clue. It does not identify the responsible gene by itself but shows where to look for biological mechanisms.
“We compared genetic data from a large number of women with PCOS with genetic data from women without PCOS. The differences we found formed the basis for identifying genetic factors that affect the risk of PCOS,” says Tugce Karaderi.
Different analyses converge on hormones and metabolism
The team then asked whether the newly mapped regions were connected with hormone levels, body weight, blood sugar, ovarian function and age at menopause. This revealed a pattern involving both the hormonal control of ovulation and the metabolic pathways that regulate weight, insulin and blood fat.
This type of large-scale comparison is known as a genome-wide association study. It can identify regions linked to a condition but cannot by itself reveal the responsible gene or prove that a particular variant directly causes disease.
The team therefore tested the findings in several other ways. They examined related traits, looked for blood proteins connected with ovarian dysfunction and tested genetic risk scores in the UK Biobank and Copenhagen Hospital Biobank. The analyses measured different aspects of PCOS but together supported roles for reproductive hormones and metabolic regulation.
Because PCOS is diagnosed and recorded differently across countries, the team checked whether the results depended on how cases were defined. The results were broadly consistent across definitions, but most participants were of European ancestry, limiting how confidently the findings can be applied to other populations.
“We had to harmonise the way we selected patients and controls for the analyses. Because the study was international, we had to take into account how different health systems and data sources define who is a patient,” explains Tugce Karaderi.
Overlapping genetic pathways link ovulation, menopause and metabolic risk
Several variants were linked both to later menopause and a larger pool of egg-containing follicles. Women are born with a finite supply of these follicles. One possibility is that hormonal changes in PCOS slow depletion of this pool. Irregular or absent ovulation may therefore make pregnancy more difficult earlier in life, even though more follicles remain later.
Higher genetic susceptibility to PCOS also corresponded to 13% higher odds of type 2 diabetes and 9% higher odds of coronary artery disease among women for each standard-deviation increase in the genetic risk score.
The differences became smaller after body-mass index was considered, indicating that part of the overlap reflects shared genetic influences on body weight and metabolism rather than PCOS directly causing these diseases. Some remained, especially among women.
Greater exposure to active androgens may be one explanation. Sex hormone–binding globulin normally binds part of the testosterone in the blood; when its level is lower, more testosterone remains available to act on tissues. The findings support this pathway but cannot establish the full chain from genetic variation to disease.
Harder to conceive does not necessarily mean fewer children
PCOS remains a common cause of infertility because ovulation may be infrequent or absent. Yet in both the data from the United Kingdom and Denmark, genetic susceptibility did not correspond to a greater likelihood of remaining childless or having fewer children overall.
The distinction is important: difficulty becoming pregnant does not necessarily reduce the number of children a woman has over her lifetime. For some, problems earlier in life may be partly balanced by a longer or later reproductive window.
The team also calculated a polygenic risk score. PCOS is not caused by one decisive gene. Instead, the score combines the small effects of many variants into an estimate of inherited susceptibility.
In the UK Biobank, women in the highest fifth of the score had more than twice the odds of having PCOS as women in the lowest fifth. Higher scores were also linked to diabetes, coronary artery disease and a less favourable metabolic profile. The score therefore identifies inherited differences across populations but cannot yet predict how PCOS will develop for an individual woman.
“In the long term, a genetic risk score could become an additional layer in the doctor’s assessment of the patient. It could help determine who should receive early attention regarding fertility and who should also be monitored for signs of metabolic disease,” says Tugce Karaderi.
Genetic risk scores are not yet ready for the clinic
Before such a tool could be used clinically, studies would need to show that it improves prediction beyond symptoms, hormone measurements, family history and other information already available to doctors.
Some specialists have therefore proposed the name polyendocrine metabolic ovarian syndrome (PMOS). Whereas PCOS focuses attention on the ovaries, PMOS emphasises the interaction between hormones, metabolism and ovarian function. The term is not established in clinical practice but captures the wider biological picture revealed by the study.
“We are trying to put together the pieces of an incomplete puzzle. We have been able to add a number of new pieces, but much remains to be learned about PMOS and the conditions that accompany it,” concludes Tugce Karaderi.
For women like Irene, the study cannot predict what will happen. But it changes the picture of PCOS: difficulty becoming pregnant, later menopause and metabolic disease are not separate chapters but parts of the same lifelong condition – although its course differs from woman to woman.
