What Causes PCOS? The Hormones, Insulin, and Genetics Behind Polycystic Ovary Syndrome

Short Answer

PCOS does not come from one trigger you can point to. The most accurate short version is this: it is a complex hormonal–metabolic condition with an unknown exact cause, shaped by inherited risk and environment. In many people, one loop gets louder when cells respond less efficiently to insulin. The body makes more insulin to move glucose into cells, and higher insulin can push the ovaries toward more androgen production. Those androgens can interfere with follicle development and ovulation, which is why PCOS may show up as irregular periods, acne, or excess facial or body hair. (NICHD/NIH)

But insulin is not the whole story. PCOS can also involve changes in the brain–ovary signaling system — especially faster GnRH/LH signaling — which can favor ovarian androgen production and make normal follicle maturation harder. That is one reason PCOS looks different from person to person: one body may be more insulin-driven, another more reproductive-hormone-driven, and many sit somewhere in between. Genetics, inflammation, body weight, and environmental factors can shape how strongly these loops show up; they do not make PCOS a personal failing. In short, PCOS is better understood as several interacting drivers — insulin resistance, excess androgens, neuroendocrine signaling, and inherited susceptibility — rather than one single cause. (Blank et al., *Human Reproduction Update*, PMC3453528)

What causes PCOS — at a glance

PCOS does not have one proven cause. The most accurate way to think about it is: your baseline biology may make you more susceptible, then insulin signaling, ovarian androgen production, brain–ovary hormone rhythms, inflammation, weight distribution, and environment can push the syndrome into different patterns. That is why two people can both have PCOS and look very different on the outside. The exact cause is still considered unknown, but genetic and environmental factors, insulin resistance, and higher androgen activity are consistently linked to PCOS biology. (NICHD/NIH)

Mechanism vs. contributor vs. predisposition — how the pieces fit

PCOS is easier to understand when you stop looking for one villain. The exact cause is still unknown, and the best evidence points to a mix of genetic susceptibility, hormone signaling, insulin biology, and environmental or body-state factors that can push the system harder or softer. That means PCOS is not caused by something you “did wrong.” It is a syndrome — a pattern your body can arrive at through more than one route. (NICHD/NIH)

Predisposition is the part you may inherit. PCOS tends to run in families, and NIH describes genetic and environmental factors as contributors while still emphasizing that the exact cause remains unknown. Recent research has identified multiple gene variants linked with PCOS risk, which may help explain why one person’s PCOS looks mostly metabolic, another person’s looks mostly reproductive, and another person has mixed features. (NICHD/NIH)

Core mechanisms are the internal loops that make PCOS feel so stubborn. Insulin resistance is recognized as a key feature of PCOS, although routine clinical measures of insulin resistance are not always accurate enough to diagnose or track it by themselves. When insulin runs high, it can support higher androgen activity; and when androgen activity is high, follicles may not mature and ovulate in the usual rhythm. LH and GnRH signaling can also be altered, adding a neuroendocrine route into the same pattern of irregular ovulation and androgen-related symptoms. (Barber et al., *Clinical Medicine*, PMC5922706)

Modifiers are the dimmer switches. Weight, adiposity, inflammation, sleep, diet, activity, medications, stress physiology, and environmental exposures can influence how loudly PCOS shows up — especially through insulin sensitivity, inflammation, and hormone signaling. But “modifier” does not mean “blame.” PCOS-related insulin resistance can be at least partly independent of obesity, and the 2023 international guideline specifically emphasizes healthy lifestyle support alongside awareness of weight stigma. (2023 International PCOS Guideline, PMC10477934)

So the honest causal language is: genetics may predispose, insulin resistance and androgen excess may drive or maintain, neuroendocrine signaling may contribute, and lifestyle or body-state factors may modify. None of that means sugar, stress, weight, or birth control “caused” PCOS. It means your endocrine system is a network — and PCOS is what can happen when several parts of that network get pulled into the same reinforcing pattern.

There is no single cause — PCOS is multifactorial

Decades of research have not found one neat origin point for PCOS. The best current frame is that PCOS is a syndrome — a pattern that can emerge when several body systems push on each other at once: genes, insulin signaling, ovarian hormone production, brain–ovary signaling, inflammation, and environmental exposures. NICHD puts it plainly: genetic and environmental factors contribute, but the exact cause is still unknown. (NICHD/NIH)

As one 2026 review in Frontiers in Endocrinology summarizes it: “The etiology of PCOS involves a complex interplay of genetic, metabolic, hormonal, immunological and environmental factors, though its precise mechanisms remain incompletely understood.” (Chen et al., *Frontiers in Endocrinology*, 2026, PMID 41669247)

It is also common. One review describes PCOS as “a multifactorial endocrine disorder affecting about 10% of reproductive-age women” (Patel, *Frontiers in Endocrinology*, PMC12957162) — and newer WHO estimates place PCOS at about 10–13% of reproductive-aged women globally. That scale matters because “what caused this?” is not a niche question. It is a question many people ask while trying to make sense of irregular cycles, acne, excess hair growth, fertility concerns, weight changes, blood-sugar issues, or symptoms that do not fit anyone else’s version of PCOS. (WHO)

In plain terms: PCOS is not one switch that flips. It is more like a loop. Insulin can push the ovaries to make more androgens. Higher androgens can interfere with ovulation. Irregular ovulation can change cycle patterns. Genetics can make those loops easier to start or harder to quiet down. For some people, the metabolic side is loud; for others, the reproductive-hormone side is more obvious. That is why two people can both have PCOS and look very different on paper, in the mirror, and in the doctor’s office. NICHD notes that researchers have identified PCOS subtypes with different reproductive, metabolic, inflammatory, and symptom patterns, and Rotterdam-based research also describes multiple PCOS phenotypes rather than one single presentation. (NICHD/NIH)

Insulin resistance and high insulin — a central metabolic driver

Insulin resistance is one of the clearest ways PCOS becomes both a hormone condition and a metabolic one. Insulin’s everyday job is to help move glucose from your blood into your cells for energy. When muscle, fat, and liver cells respond less well to insulin, the pancreas often compensates by making more of it. In PCOS, that extra insulin can do more than affect blood sugar: it can push the ovaries toward making more androgens, especially when it acts alongside luteinizing hormone, and it can lower sex hormone-binding globulin from the liver, leaving more testosterone “free” and biologically active. That is why a metabolic signal can show up as very reproductive or skin-related symptoms — irregular ovulation, acne, oily skin, or more facial and body hair. (NICHD/NIH)

The same review frames PCOS this way: it is “defined by insulin resistance, androgen excess, and chronic inflammation, which drive both reproductive and metabolic complications.” (Patel, *Frontiers in Endocrinology*, PMC12957162)

A 2026 phenotype analysis put the mechanism directly: in the hyperandrogenic form of PCOS, insulin resistance tracked with androgen excess even after accounting for body size. As the authors state, “we confirm that insulin resistance drives the hyperandrogenic phenotype independently of obesity.” (Wang et al., *Frontiers in Endocrinology*, 2026, PMC12913134)

That last phrase matters. Insulin resistance is linked to PCOS whether or not someone carries excess weight — so PCOS is not simply “a weight problem,” and lean people can have it too. Older clamp studies and later lean-PCOS research support the same anti-stigma point: PCOS can involve impaired insulin action independent of obesity, while weight gain can still worsen insulin resistance in people who are already susceptible. (Dunaif et al., *Diabetes*, 1989, PMID 2670645)

This is also the reason clinicians pay attention to metabolic health in PCOS. Estimates vary by phenotype and by how insulin resistance is measured, but reviews commonly cite that roughly 50% to 90% of women with PCOS show insulin resistance. That does not mean insulin resistance single-handedly causes PCOS in everyone. It means it is a central, associated driver in many people — strong enough that guidelines emphasize metabolic risk, while also noting that routine insulin blood tests are not always clinically useful. (Barber et al., *Clinical Medicine*, PMC5922706)

Excess androgens (hyperandrogenism)

Androgens are hormones such as testosterone. They are often described as “male” hormones, but that can be misleading: everyone makes them, including in the ovaries and adrenal glands. In PCOS, androgen signaling is often higher than the body can comfortably handle. That excess can show up on the skin and hair follicles as acne, oilier skin, coarse facial or body hair, or scalp hair thinning — and it can show up inside the ovary as disrupted ovulation. Clinical or biochemical hyperandrogenism is one of the major diagnostic features used in current PCOS guidelines. (Cleveland Clinic)

One review describes it as “a multifactorial endocrine disorder characterized by hyperandrogenism, inflammation, and ovarian dysfunction.” (*American Journal of Physiology. Endocrinology and Metabolism*, 2026, PMC12990670)

The ovary is not just “overproducing testosterone” in isolation. In many people with PCOS, insulin is part of the push. When the body becomes less responsive to insulin, the pancreas may release more of it. Higher insulin can act on ovarian theca cells and work together with luteinizing hormone (LH), nudging the ovary toward more androgen production. It can also lower sex hormone-binding globulin (SHBG), a carrier protein that normally keeps some testosterone bound and less biologically active — so more testosterone may be available to tissues even when total levels do not tell the whole story. (Barber et al., *Clinical Medicine*, PMC5922706)

This is why androgen excess and insulin resistance are often described as a loop rather than a one-way cause. More insulin can amplify androgen effects; androgen excess can worsen metabolic signaling in some tissues; together they can interfere with the steady maturation of ovarian follicles. Instead of one follicle becoming dominant and ovulating on schedule, follicles may stall earlier in development, which helps explain irregular or absent ovulation in PCOS. (PMC9832677)

That loop is powerful, but it is not the only route into PCOS. Some people have a more insulin-driven pattern. Others have a stronger ovarian, adrenal, or neuroendocrine component. The important point is that hyperandrogenism is not a character flaw, a hygiene issue, or a simple “too much sugar” problem. It is a hormone-and-tissue signaling pattern — visible on the outside for many people, but rooted in ovarian, metabolic, and brain-hormone communication.

Neuroendocrine signaling — the brain–ovary axis and LH

PCOS is not only an ovarian story. Your ovaries respond to signals that start higher up, in the hypothalamus and pituitary — the brain centers that help time the menstrual cycle. The hypothalamus releases GnRH in pulses; the pituitary reads that rhythm and releases luteinizing hormone (LH) and follicle-stimulating hormone (FSH). When that rhythm becomes too fast or too LH-heavy, the ovary can receive a stronger androgen-making signal and a weaker, less coordinated follicle-maturing signal. That can make ovulation less predictable or stop it from happening regularly. Reviews of PCOS neuroendocrine biology describe increased GnRH pulsatility as driving preferential LH secretion, with downstream effects on ovarian androgen production and oligo/anovulation. (Garg et al., *Clinical Endocrinology*, 2022, PMID 35262967)

In plain language: the “brain–ovary conversation” can get stuck on a frequency that keeps pushing LH. LH is not bad — you need a well-timed LH surge to ovulate. The problem is different: in many people with PCOS, LH signaling is persistently altered, and that altered pattern can feed the same androgen-and-ovulation loop that shows up as irregular cycles, acne, excess hair growth, or difficulty predicting ovulation. Not everyone with PCOS has the same LH pattern, which is one reason LH or the LH/FSH ratio is not used as a stand-alone diagnostic test. (*Molecular and Cellular Endocrinology*, 2019, PMID 31461666)

A 2026 phenotype analysis reported that in its cohort, “menstrual cycle prolongation correlated with LH levels rather than metabolic markers, suggesting a predominant neuroendocrine aetiology.” (Wang et al., *Frontiers in Endocrinology*, 2026, PMC12913134)

The same study found that not every presentation is metabolic: “The Non-HA phenotype appears driven primarily by neuroendocrine dysregulation.” (Wang et al., *Frontiers in Endocrinology*, 2026, PMC12913134)

The practical takeaway: PCOS has more than one route. In some people, insulin resistance and metabolic signaling dominate the picture. In others, the neuroendocrine route — the timing and intensity of signals from brain to pituitary to ovary — may be more prominent. NICHD also describes PCOS as heterogeneous, with research pointing to reproductive and metabolic subtypes rather than one single cause. That is why one-size-fits-all explanations fall short: two people can both meet PCOS criteria, but the biology driving their cycles may not be identical. (NICHD/NIH)

Genetics and family predisposition

PCOS tends to run in families, but not in a clean “one gene equals one diagnosis” way. NIH describes the exact cause as still unknown, with genetic and environmental factors both contributing, and MedlinePlus Genetics notes that PCOS has no clear inheritance pattern even though affected people may have a close family member with the condition. In practical terms, this means your family history can raise your baseline likelihood — it does not write the outcome in advance. MedlinePlus estimates that 20% to 40% of women with PCOS have an affected mother or sister. (MedlinePlus Genetics)

Twin and family studies are why researchers take that family pattern seriously. In a Dutch twin-family study, PCOS-like features were more similar in identical twin sisters than in non-identical twin/sister pairs: the reported correlations were 0.71 for monozygotic twins and 0.38 for dizygotic twins and sisters. Later genetic reviews commonly interpret that kind of evidence as a substantial heritable component — roughly in the range of two-thirds to four-fifths depending on the definition and model used — but that is not the same as saying PCOS is genetically predetermined. (Vink et al., *J. Clin. Endocrinol. Metab.*, 2006, PMID 16219714)

As above, genetics is one strand of the “complex interplay of genetic, metabolic, hormonal, immunological and environmental factors.” (Chen et al., *Frontiers in Endocrinology*, 2026, PMID 41669247)

That “interplay” matters because PCOS genes seem to point toward several body systems at once: ovarian androgen production, follicle development, brain–ovary hormone signaling, insulin and glucose regulation, inflammation, and fat metabolism. Genome-wide studies have repeatedly flagged regions near or involving genes such as DENND1A, THADA, FSHR, LHCGR, INSR, RAB5B, YAP1, TOX3, HMGA2, ZNF217, and others. A large European-ancestry meta-analysis of 10,074 PCOS cases and 103,164 controls identified three newer loci near PLGRKT, ZBTB16, and MAPRE1 and replicated many earlier signals; more recent multi-ancestry work has reported a broader set of independent loci, underscoring that the map is still expanding. (McAllister et al., *Trends Endocrinol. Metab.*, PMC4346470; Day et al., *PLOS Genetics*, PMC6300389)

Specific inflammatory-gene variants have been implicated as one example of that genetic layer: one case-control study reports that “SNPs in IL-1 and IL-6 genes may influence susceptibility to PCOS.” (*Indian Journal of Medical Research*, 2025, PMC12883111)

Note the careful wording — may influence susceptibility, not cause. Most PCOS-associated variants are common, small-effect risk signals, not single “PCOS mutations.” Even GWAS signals usually mark a region of the genome that may contain the biologically relevant variant; they are rarely proof that one exact SNP caused a person’s symptoms. So the honest summary is this: you can inherit a higher baseline likelihood of PCOS, including a tendency toward higher androgens, irregular ovulation, insulin resistance, or inflammatory patterns. Genes load the dice. Your body’s hormones, metabolism, development, environment, and life stage help decide how — or whether — that susceptibility shows up. (Dapas & Dunaif, *Endocrine Reviews*, PMC9695127)

Weight, adiposity, and inflammation — modifiers, not blame

Weight is one of the most misunderstood parts of the PCOS story. Some people with PCOS have excess adiposity, and extra adipose tissue can make insulin resistance harder on the body. But that does not mean weight is the root cause of PCOS. PCOS can also show up in people who are not overweight, and research using insulin-clamp methods has found reduced insulin sensitivity in PCOS even when obesity or BMI are accounted for. In other words: weight can turn up the volume on the metabolic loop, but it is not the switch that created the syndrome. (Dunaif et al., *Diabetes*, 1989, PMID 2670645)

One 2026 paper puts the weight relationship this way: “Obesity, present in a significant proportion of PCOS patients, exacerbates insulin resistance (IR) and worsens reproductive outcomes.” (*Frontiers in Endocrinology*, 2026, doi:10.3389/fendo.2026.1758805)

Here’s the body-level reason this matters. When cells respond less well to insulin, the pancreas may release more insulin to move glucose out of the blood. Higher insulin can push the ovaries toward making more androgens and can lower sex hormone-binding globulin, leaving more active androgens in circulation. More androgens can worsen acne, excess hair growth, and ovulation problems. If excess adiposity is also present, it may add more insulin resistance and inflammatory signaling to that same loop — insulin resistance ↔ weight ↔ androgens — making symptoms more intense for some people. (NICHD/NIH)

Chronic low-grade inflammation is another recurring thread linked to the metabolic side of PCOS. Inflammation is not a moral statement about food, discipline, or body size; it is immune and metabolic signaling. A systematic review and meta-analysis found higher circulating C-reactive protein in PCOS, including analyses suggesting this pattern is not explained only by obesity. (Aboeldalyl et al., *Int. J. Mol. Sci.*, 2021, PMID 33800490)

Newer work points in the same direction: “growing evidence highlights mitochondrial dysfunction and chronic low-grade inflammation as central drivers of cardiovascular pathology in PCOS.” (*BioMed Research International*, 2026, PMC12835629)

So the clean framing is this: PCOS is not caused by lifestyle choices. Its exact cause is still not fully known, and authoritative medical sources describe it as involving genetic, hormonal, insulin-related, and environmental contributors rather than one personal behavior. Weight, adiposity, and inflammation belong inside the biology of PCOS as modifiers — things that can worsen or amplify the loop for some people — not as blame. The “independently of obesity” evidence is the anchor: insulin resistance can be part of PCOS even without excess weight, while obesity, when present, can add another metabolic burden. (NICHD/NIH)

What a wearable like Welltory can — and can't — tell you about PCOS

PCOS is diagnosed by a clinician, not by an app. Your doctor usually starts with your symptoms and menstrual history, checks for signs such as acne, excess hair growth, or insulin resistance, and may use blood tests to measure hormones and rule out other conditions; ultrasound can also be used to look at the ovaries and uterine lining. There is no single wearable signal that can detect PCOS, diagnose it, or explain why you developed it. (Mayo Clinic)

What consumer data can do is give you context. PCOS often travels with metabolic strain, sleep disruption, stress load, and changes in how your body regulates energy and recovery. Heart rate variability — HRV — is one way researchers study the autonomic nervous system, the body system that shifts you between “fight-or-flight” and “rest-and-digest.” In clinical studies, women with PCOS have shown differences in HRV and cardiac autonomic modulation compared with controls, and a 2024 systematic review found that this area is being studied as part of the broader cardiovascular and metabolic picture of PCOS. That does not mean your watch can diagnose autonomic dysfunction or prove PCOS; it means HRV can be a useful body-signal to track alongside symptoms. (de Sá et al., *Gynecological Endocrinology*, 2011, PMID 20645891; Mirzohreh et al., *Systematic Reviews*, 2024, PMC11271026)

This is where a wearable like Welltory fits: as a pattern notebook, not a medical test. If your cycle becomes more irregular, your sleep gets worse, your resting heart rate trends up, your HRV drops during high-stress weeks, or your recovery feels consistently poor, those patterns may help you describe what is happening more clearly at an appointment. They can also help you notice whether symptoms cluster around missed periods, poor sleep, illness, intense training, travel, or emotional stress.

Keep the boundary firm: wearable data is associated, trackable context — never a cause, never proof, never a diagnosis. Welltory can help you record your body’s patterns and start a better conversation with your doctor, but it does not diagnose or detect PCOS.

Related questions people ask

• Does stress cause PCOS? No — stress is not considered a root cause of PCOS. PCOS is usually explained as a multifactorial condition: genetic risk, androgen signaling, insulin resistance, and other biologic factors can all contribute, while the exact cause remains unknown. Stress can still matter because the stress-response system talks to the same brain–ovary hormone network that helps time ovulation. In real life, that means stress may aggravate symptoms, make periods less predictable, or make PCOS harder to live with — but it should be framed as context, not the thing that “gave you” PCOS. (NICHD/NIH)

• Does birth control cause PCOS? No. Hormonal birth control does not cause PCOS; it is commonly used to manage PCOS symptoms. Combined oral contraceptives can make bleeding more regular, lower androgen activity, and help with acne or excess hair growth in people with PCOS who are not trying to get pregnant. The confusing part is timing: while you’re taking hormonal contraception, it can create regular withdrawal bleeding or reduce bleeding, so irregular cycles that were already there may become obvious only after you stop. That can feel like “the pill caused it,” but medically it is better understood as symptoms becoming visible. (NICHD/NIH)

• Why does inositol help PCOS? Inositol comes up in PCOS because insulin resistance is common in PCOS, and inositol is studied as a supplement that may influence insulin sensitivity. That connects it to the “why” of PCOS for some people — especially those whose symptoms are strongly tied to insulin signaling. But this causes article should not turn that into a treatment claim: the evidence for inositol in PCOS is still described as limited and uncertain in a systematic review used for the 2023 international guideline update. Specific benefits, who might try it, safety, and dosing belong on the PCOS treatment page and should be discussed with a clinician. (Fitz et al., *J. Clin. Endocrinol. Metab.*, 2024, PMC11099481)

• Is PCOS caused by too much sugar / a bad diet? No single food, sugar, or “bad diet” causes PCOS. That framing is too simple and often unfair. A more accurate way to think about it is this: PCOS can involve an insulin-resistant system, and food patterns can push that system harder or make symptoms easier to manage. Diet is a modifier — something that can influence glucose, insulin demand, weight, energy, and symptom burden — not the original cause of the syndrome. (NICHD/NIH)

Comparison blocks (for quick extraction)

Cause vs. contributor. A cause would mean one clear origin point. PCOS does not have that. The better picture is a loop: insulin resistance can push the body to make more insulin; higher insulin can amplify ovarian androgen activity; excess androgens can interfere with follicle development and ovulation; brain–ovary signaling can add another layer. Weight, inflammation, diet, and environmental exposures can change how strongly that loop shows up, but they are not the single “reason” you have PCOS. Current NIH sources describe PCOS as a condition with genetic and environmental contributors whose exact cause remains unknown. (NICHD/NIH)

Insulin-driven vs. neuroendocrine-driven PCOS. PCOS can reach a similar symptom pattern through different routes. For some people, the metabolic route is louder: insulin resistance and higher insulin levels feed androgen excess, which then disrupts ovulation. For others, especially non-hyperandrogenic presentations, the stronger signal may come from the hypothalamus–pituitary–ovary axis — the brain and ovaries mistiming the hormone conversation that helps an egg mature and release. That is why PCOS is described in phenotypes, not as one identical profile in every body. (PMID 21263450)

PCOS vs. “just irregular periods.” Irregular cycles are a clue, not a diagnosis. Your period can shift because of thyroid dysfunction, high prolactin, stress, perimenopause, pregnancy, medications, or other endocrine conditions. PCOS is more specific: clinicians look for patterns such as irregular or absent ovulation, signs or lab evidence of excess androgens, and sometimes polycystic ovarian morphology — while also ruling out look-alike conditions. A tracker can help you show your pattern; it cannot diagnose PCOS for you. (NICHD/NIH)

Genes vs. destiny. Inherited susceptibility can raise your baseline risk, especially if PCOS-like symptoms run in your family, but genes are not a verdict. They shape the terrain your hormones and metabolism have to work on; they do not decide every outcome by themselves. Modifiable factors — sleep, movement, nutrition, stress load, weight changes, and treatment choices made with a clinician — can still influence insulin demand, ovulation patterns, and symptom severity. (NICHD/NIH)

Who needs extra caution / when to see a clinician

See a clinician if your periods become newly irregular or stay irregular, if you stop getting periods, or if you notice unusual hair growth on the face/body, stubborn acne, trouble getting pregnant, or weight changes you can’t explain. Those patterns can fit PCOS, but they are not PCOS-specific: thyroid disease, high prolactin, and adrenal conditions can look similar, so a proper diagnosis means checking for PCOS and ruling out mimics instead of guessing from symptoms alone. (NICHD/NIH)

Be extra careful about self-managing if you’re pregnant, trying to conceive, or already have diabetes, thyroid disease, high blood pressure, high cholesterol, or cardiovascular disease. PCOS care often needs to cover more than cycles and skin — guidelines emphasize metabolic risk, cardiovascular risk, fertility planning, and pregnancy monitoring, because these can change what testing and treatment are safest for you. A wearable can help you bring cleaner context to the visit — cycle timing, sleep, resting heart rate, stress load, and symptom patterns — but it cannot diagnose PCOS, tell you the cause, or replace labs and a clinician’s evaluation. (2023 International PCOS Guideline, PMC10477934)

How we made it

Made with AI tools, then edited, fact-checked, and medically reviewed by the Welltory team.

Discounts for blog readers: up to 36% off

See what affects your energy, stress, sleep, and daily state with Welltory

Explore your data