PCOS Open Controversies: What Experts Are Still Debating (And Why It Matters for You)

Why PCOS Is Still a Contested Diagnosis

PCOS is one of the most common hormonal conditions in women, yet its definition, cause, and optimal treatment remain genuinely unsettled. That is not a communication failure. It is a scientific one.

Prevalence estimates range from 6% to 13% of women of reproductive age globally, depending entirely on which diagnostic criteria a clinician applies. A woman evaluated at one clinic may receive a PCOS diagnosis; the same woman at a different clinic, using different criteria, may not. This diagnostic inconsistency shapes everything downstream: treatment access, fertility counseling, metabolic monitoring, and long-term cardiovascular risk management.

The disputes are not trivial academic disagreements. They affect whether you get diagnosed at all.

Three Sets of Criteria, Three Different Patient Populations

The NIH 1990 criteria require both clinical or biochemical hyperandrogenism and chronic anovulation. No ovarian morphology required. This definition is the most restrictive and captures a more metabolically at-risk group.

The Rotterdam 2003 consensus broadened the definition to require only two of three features: hyperandrogenism, oligo-anovulation, or polycystic ovarian morphology on ultrasound. This added two phenotypes that had no androgen excess at all, which remains controversial.

The Androgen Excess and PCOS Society (AES) 2006 criteria sit in between: androgen excess is required, but anovulation or polycystic ovaries only needs to fulfill one of the two remaining criteria.

A 2016 systematic review estimated that Rotterdam criteria diagnose significantly more women than NIH criteria, particularly younger women with regular cycles but polycystic ovarian morphology. Whether those additional women carry the same metabolic risk is the crux of the argument.

The Polycystic Ovary Morphology Problem

Polycystic ovarian morphology (PCOM) on ultrasound is a feature many clinicians rely on, but the threshold for calling it positive has shifted. The original 12-follicle threshold was set using older transabdominal equipment. With modern high-resolution transvaginal ultrasound, a much higher antral follicle count or ovarian volume threshold is appropriate, and the 2023 International Evidence-Based Guideline for PCOS now recommends a follicle count of 20 or more per ovary or an ovarian volume of 10 mL or more as the threshold. Many labs and clinicians have not updated their reporting accordingly. A woman told she has "polycystic ovaries" on a report from five years ago may have been assessed against an outdated standard.

The Four Phenotypes: Same Label, Very Different Biology

Rotterdam criteria define four PCOS phenotypes, and experts disagree about whether they should all carry the same diagnosis.

• Phenotype A (classic): Hyperandrogenism plus anovulation plus PCOM. Highest metabolic risk.
• Phenotype B: Hyperandrogenism plus anovulation, without PCOM. Also high metabolic risk.
• Phenotype C: Hyperandrogenism plus PCOM, with ovulatory cycles. Intermediate risk.
• Phenotype D (ovulatory PCOM): Anovulation plus PCOM, without hyperandrogenism. Lowest metabolic risk, and the most contested.

Phenotype D women do not have androgen excess by any measure. A growing number of reproductive endocrinologists argue they should not be grouped with women who have frank hyperandrogenism, because the metabolic implications, treatment needs, and long-term cardiovascular risk differ substantially. The counterargument is that phenotype D women still have menstrual irregularity requiring clinical attention and that splitting the diagnosis creates its own confusion.

This matters to you directly because a phenotype D diagnosis, if your clinician uses Rotterdam criteria, may lead to the same treatment algorithm applied to a woman with phenotype A, even though your biology and risk profile are different.

What Actually Causes PCOS? The Etiology Debate

Insulin Resistance: Central Cause or Downstream Effect?

Most PCOS experts agree that insulin resistance is present in approximately 65-80% of women with PCOS, regardless of body weight. The disagreement is about causality. Does insulin resistance drive androgen excess, or does androgen excess impair insulin signaling? Or does a third upstream factor, possibly genetic, trigger both?

The insulin hypothesis holds that hyperinsulinemia stimulates ovarian theca cells to overproduce androgens and suppresses hepatic sex-hormone-binding globulin (SHBG) production, amplifying free androgen levels. Metformin's partial effectiveness in restoring ovulation supports this model. But metformin does not work for every woman, and some lean women with PCOS have normal insulin sensitivity yet still have high androgens, which the pure insulin hypothesis cannot explain cleanly.

Neuroendocrine Dysregulation

A competing and not mutually exclusive framework centers on the hypothalamic-pituitary axis. In many women with PCOS, GnRH pulse frequency is elevated, favoring LH secretion over FSH, which drives androgen production in the ovary and impairs follicle maturation. Whether this is a primary defect or a feedback consequence of low progesterone from anovulation is still unresolved.

Genetics vs. In Utero Androgen Exposure

PCOS clusters in families, with heritability estimates around 70% in twin studies. Genome-wide association studies have identified loci near genes involved in gonadotropin signaling, insulin receptor function, and androgen biosynthesis, but no single PCOS gene has been identified. A separate hypothesis, supported by animal data, suggests that prenatal androgen exposure programs the hypothalamic-pituitary-ovarian axis toward the PCOS phenotype. Studies in daughters of women with PCOS show higher AMH levels and altered LH pulsatility from early life, suggesting fetal programming plays a role alongside genetics.

The Gut Microbiome Hypothesis

Emerging, less mature research suggests women with PCOS have altered gut microbiome composition compared to controls, with potential implications for inflammation, insulin signaling, and androgen metabolism. A 2019 meta-analysis found differences in bacterial diversity, but the field has not yet established whether microbiome changes are a cause, a consequence, or an epiphenomenon. This is an area where the evidence is genuinely preliminary, and you should be skeptical of supplement products claiming to "fix" PCOS through gut health at this stage.

Metabolic Risk: How Serious, and for Whom?

Women with PCOS have higher rates of type 2 diabetes, gestational diabetes, dyslipidemia, and hypertension than age-matched women without PCOS. But the size of that risk, and whether it persists after adjusting for BMI, is contested.

A 2019 NEJM review described the lifetime metabolic burden of PCOS as substantial but noted that many studies do not adequately control for adiposity. The landmark Rotterdam follow-up data suggested that lean women with PCOS have meaningfully lower metabolic risk than overweight women with PCOS, yet still elevated compared to lean women without PCOS.

This distinction matters in clinical practice. The 2023 International PCOS Guideline recommends a minimum of annual screening for impaired glucose tolerance and type 2 diabetes in all women with PCOS, regardless of BMI, using an oral glucose tolerance test rather than fasting glucose alone, because fasting glucose misses postload dysglycemia that OGTT captures. Many women with PCOS are not being offered this screening.

Cardiovascular Risk: A Real Signal or Confounded by Obesity?

Whether PCOS independently increases cardiovascular event risk is one of the most debated questions in the field. Large observational studies show higher rates of cardiovascular risk factors. A 2020 meta-analysis in the European Journal of Endocrinology found a significantly elevated risk of major adverse cardiovascular events in women with PCOS, but the authors acknowledged that most included studies did not fully control for obesity or metabolic syndrome. A definitive, adequately powered, BMI-adjusted prospective cardiovascular outcomes trial in women with PCOS has not been done.

Adolescent PCOS: When Is It Appropriate to Diagnose?

Diagnosing PCOS in adolescent girls is one of the most contested areas in the field, and the controversy has real consequences for young women.

Irregular periods, mild acne, and multicystic ovarian appearance on ultrasound are all normal features of early puberty. Applying Rotterdam criteria to a 14-year-old risks over-diagnosis and the psychological burden of a chronic condition label. Yet delaying diagnosis in a 17-year-old with frank hirsutism, oligomenorrhea, and high testosterone means years without appropriate metabolic monitoring or fertility counseling.

The 2023 International PCOS Guideline recommends that in adolescents, diagnosis requires both hyperandrogenism and ovulatory dysfunction (irregular cycles persisting beyond 2 years post-menarche), and explicitly states that polycystic ovarian morphology alone should not be used to diagnose PCOS in this age group. PCOM criteria developed for adults are not validated in adolescents. This is one of the few areas where international experts now have reasonable consensus, though adoption into clinical practice remains inconsistent.

PCOS Across Life Stages: What Changes and When

Reproductive Years

During your 20s and 30s, PCOS typically presents most overtly: irregular periods, hirsutism, acne, and fertility challenges. Anovulatory cycles make conception harder, but spontaneous ovulation occurs and natural pregnancy is possible. PCOS accounts for approximately 80% of anovulatory infertility cases.

Oral contraceptives remain the most commonly prescribed treatment for cycle regulation and hyperandrogenism, but they suppress endogenous hormone production and are not appropriate for women who are trying to conceive. When fertility is the goal, letrozole 2.5-7.5 mg on cycle days 3-7 is now the first-line ovulation induction agent per ACOG Practice Bulletin 194 and the 2023 PCOS Guideline, having displaced clomiphene based on the PPCOS II trial data.

Trying to Conceive and Pregnancy

Women with PCOS face a distinct set of decisions when trying to conceive that are not well captured in most general PCOS content. The framework below organizes the key contested and established points by decision stage.

Pre-conception: Metformin is sometimes continued through conception, but a 2017 Cochrane review found insufficient evidence that metformin reduces miscarriage risk in PCOS pregnancies compared to placebo, despite widespread use for this purpose. The ASRM and the 2023 PCOS Guideline do not recommend routine metformin use in pregnancy specifically to reduce miscarriage, and the evidence for this practice is genuinely weak. You should discuss stopping or continuing metformin with your prescriber based on individual metabolic factors.

During pregnancy: Women with PCOS have higher rates of gestational diabetes, preeclampsia, and preterm birth. A large Swedish registry study found a relative risk of 2.0 for gestational diabetes and 1.45 for preeclampsia in women with PCOS compared to controls. Early glucose screening at 16-18 weeks (rather than the standard 24-28 weeks) is recommended by some but not uniformly adopted.

Lactation: There is no evidence that PCOS itself impairs lactation. Some women with PCOS may have lower prolactin response or reduced milk supply, and this appears to be more common in women with higher androgen levels, though the data are limited. Metformin is excreted in breast milk in small amounts; available data suggest infant exposure is low and most guidelines consider it compatible with breastfeeding, though this should be individualized.

Perimenopause

The menopause transition in PCOS is under-studied and genuinely uncertain. Menstrual cycle regularity may improve as women approach perimenopause and ovarian follicle counts decline naturally. Some women report that their PCOS symptoms become less severe in their 40s, while others do not notice a change.

A 2011 study in Menopause found that women with a history of PCOS reached menopause 2 years later on average than controls, possibly because higher antral follicle counts early in life translate to a larger ovarian reserve that depletes more slowly. Whether the PCOS phenotype persists into post-menopause, and what that means for ongoing cardiovascular and metabolic monitoring, is not established. The Menopause Society does not yet have a dedicated PCOS and menopause guideline.

Post-Menopause

Androgen levels decline after menopause, and some markers of PCOS (hirsutism, irregular cycles) naturally resolve. But the metabolic legacy persists. Women who had PCOS in their reproductive years carry higher baseline metabolic risk into post-menopause, and this history should inform cardiovascular risk stratification, diabetes screening, and statin consideration. Clinicians who do not take a careful reproductive history may miss this context entirely.

The Androgen Measurement Problem

One of the most practical controversies in PCOS care is measurement quality. Testosterone is the most commonly ordered androgen, but total testosterone assays used in most general labs have poor sensitivity at the low concentrations found in women. Immunoassay-based testosterone measurements in women are frequently unreliable.

The gold standard for androgen measurement in women is liquid chromatography-tandem mass spectrometry (LC-MS/MS), but this is not universally available. The 2023 International PCOS Guideline specifies that clinicians should use a high-quality testosterone assay and, where possible, calculate free androgen index or free testosterone using SHBG, because total testosterone alone may miss biochemical hyperandrogenism in women who are not producing severely elevated levels.

A woman told her testosterone is "normal" may have had it measured on a low-sensitivity immunoassay. Free androgen index, calculated as (total testosterone / SHBG) x 100, is a more reliable marker of androgen excess in women and is what should be reviewed alongside the raw number.

Treatment Controversies: What Works and What Is Debated

Metformin vs. Lifestyle vs. Combined

Lifestyle modification (diet quality, physical activity, behavioral support) is recommended as first-line for women with PCOS and excess weight. The evidence for metformin is stronger for metabolic outcomes than for androgen or cycle outcomes. A 2012 Cochrane review found that in anovulatory women with PCOS, the combination of metformin plus clomiphene improved ovulation rates compared to clomiphene alone, but subsequent data from the PPCOS II trial showed letrozole alone outperformed both for live birth rates. Treatment hierarchy has shifted, but prescribing practices have not always kept up.

Inositols

Myo-inositol and D-chiro-inositol are widely used and sold as supplements for PCOS, particularly for insulin sensitization and ovarian function. A 2018 meta-analysis found improvements in hormonal markers and insulin resistance with myo-inositol, but study quality was variable and no large, well-powered RCT has established clinical endpoints (live birth, diabetes prevention). The 2023 PCOS Guideline notes inositols as potentially useful but does not give a formal recommendation due to evidence gaps. Patients are often using these without knowing the evidence base is genuinely uncertain.

GLP-1 Receptor Agonists

GLP-1 receptor agonists (semaglutide, liraglutide) are generating significant interest in PCOS, particularly for women with BMI >27 and insulin resistance. Small trials have shown reductions in weight, androgen levels, and improvements in cycle regularity. A 2022 RCT found that liraglutide 1.8 mg daily for 12 weeks improved menstrual frequency and reduced free androgen index compared to placebo in women with PCOS. These agents are not yet approved specifically for PCOS, the long-term data in PCOS are limited, and they are contraindicated in pregnancy. Women of reproductive age using GLP-1 agents require reliable contraception, because the teratogenic risk in humans is not established and animal studies raised concern. If you become pregnant while taking a GLP-1 agent, stop it immediately and contact your provider.

Who Gets Missed: The Evidence Gap by Race, Ethnicity, and Body Size

Most foundational PCOS research was conducted in predominantly white, European cohorts. PCOS prevalence appears similar across ethnic groups, but phenotypic expression differs. South Asian women with PCOS tend to have higher insulin resistance at lower BMI thresholds. East Asian women with PCOS may present with less hirsutism but similar metabolic risk, because androgen-driven hair follicle sensitivity differs across populations.

The diagnostic criteria were not validated in diverse populations. A clinician using a hirsutism scoring tool calibrated in white European women may underestimate androgen-driven hair changes in South Asian or Middle Eastern women, and may miss the diagnosis in East Asian women where body hair response to androgens is less pronounced.

"The current diagnostic criteria were built on data from a narrow slice of the global population of women with PCOS. We are still working out how to apply them equitably," as stated in the 2023 International Evidence-Based Guideline.

Lean women with PCOS (BMI <25) are another consistently under-served group. Clinicians may be less likely to consider the diagnosis, metabolic screening may be skipped because of the assumption that lean means metabolically healthy, and treatment discussions are often weighted toward weight loss advice that is irrelevant to their situation.

What the 2023 International Guideline Changed (and What It Left Open)

The 2023 International Evidence-Based Guideline for the Assessment and Management of PCOS, produced jointly by the Monash University team and endorsed by ACOG, the European Society of Endocrinology, and others, is the most current authoritative document. Key changes from prior guidance include:

• Updated PCOM threshold to 20 follicles per ovary or ovarian volume 10 mL or more, on modern ultrasound equipment only.
• Letrozole confirmed as first-line ovulation induction over clomiphene.
• Routine AMH measurement endorsed as a diagnostic tool (though not yet validated as a sole criterion).
• Explicit acknowledgment that psychological health (depression, anxiety, disordered eating, body image) requires screening and management alongside physical symptoms.
• Spironolactone endorsed for hirsutism management with contraceptive co-prescription given teratogenic risk (feminization of male fetuses). Women taking spironolactone must use reliable contraception. Spironolactone is classified as pregnancy Category C/D and should be stopped before any attempt to conceive.

What the guideline did not resolve: the phenotype D controversy, the optimal long-term cardiovascular monitoring protocol, whether PCOS diagnosis should be revisited at menopause, and the appropriate role of newer agents like GLP-1 receptor agonists.