Why phenotyping matters
A 22-year-old with irregular periods, elevated androgens, and polycystic ovaries on ultrasound; a 28-year-old with regular cycles, elevated androgens, and polycystic ovaries; a 32-year-old with irregular cycles, no elevated androgens, and polycystic ovaries — all three meet the Rotterdam diagnostic criteria for PCOS. All three may have received the same diagnosis. The clinical management that is appropriate for each of them is substantially different.
PCOS is defined by the presence of at least two of three features from the Rotterdam criteria:
- Oligo-ovulation or anovulation (infrequent or absent ovulation, usually presenting as irregular or absent periods)
- Clinical or biochemical signs of hyperandrogenism (elevated testosterone, DHEAS, or clinical signs such as hirsutism and acne)
- Polycystic ovarian morphology (PCOM) on ultrasound — defined as 12 or more follicles 2–9 mm in diameter in either ovary, or ovarian volume above 10 mL
When two or three of these features are present, the diagnosis is PCOS. But which two features are present determines which phenotype a patient has.
Phenotype A: The classic phenotype
Features: Hyperandrogenism + oligo-ovulation + polycystic ovarian morphology (all three criteria)
This is the most common phenotype in the Indian population and the most metabolically severe. Phenotype A patients have the highest rates of insulin resistance, dyslipidaemia, and metabolic syndrome. They are at the greatest long-term risk for type 2 diabetes and cardiovascular disease.
The combination of hyperandrogenism (which drives insulin resistance through androgen-insulin cross-talk) and anovulation (which means no progesterone exposure from ovulation) creates a hormonal environment that promotes endometrial proliferation and metabolic dysfunction simultaneously.
Management of Phenotype A typically prioritises insulin sensitisation and androgen management alongside cycle regularisation.
Phenotype B: Hyperandrogenic anovulatory without PCOM
Features: Hyperandrogenism + oligo-ovulation (without polycystic ovarian morphology)
These patients have the hormonal features of PCOS — elevated androgens and anovulation — but a normal-appearing ovarian structure on ultrasound. This phenotype is metabolically similar to Phenotype A: elevated insulin resistance risk and metabolic consequence.
Ultrasound-normal results can mislead patients and occasionally clinicians into questioning the diagnosis. The Rotterdam criteria explicitly allow this phenotype — PCOM is neither necessary nor sufficient for the diagnosis.
Phenotype C: Hyperandrogenic ovulatory PCOS
Features: Hyperandrogenism + polycystic ovarian morphology (with regular ovulation)
Phenotype C patients ovulate regularly — they have predictable cycles — but have elevated androgens and polycystic-appearing ovaries. This was identified as a distinct phenotype because the clinical presentation differs meaningfully from anovulatory PCOS.
These patients are metabolically intermediate: less severe insulin resistance than Phenotypes A and B, but not metabolically normal. They have a somewhat lower long-term risk of progression to diabetes than the anovulatory phenotypes.
Fertility is usually preserved in Phenotype C patients, which is relevant for clinical counselling.
Phenotype D: Non-hyperandrogenic PCOS
Features: Oligo-ovulation + polycystic ovarian morphology (without hyperandrogenism)
This is the most metabolically benign phenotype. Phenotype D patients have irregular cycles and polycystic-appearing ovaries but no elevated androgens. Some researchers and societies debate whether Phenotype D truly belongs within the PCOS spectrum.
Clinically, these patients may have fewer skin and hair manifestations (no hirsutism, acne, or scalp hair loss driven by androgens) and are less likely to have significant insulin resistance. However, cycle irregularity — and its consequences for endometrial health and fertility — is still a clinical priority.
The Indian prevalence picture
The 2024 JAMA Network Open study by Ganie et al. — the largest epidemiological study of PCOS in India, involving 9,824 participants — found a prevalence of 7.2% using NIH 1990 criteria and 19.6% using Rotterdam criteria. The difference reflects how many more patients the broader Rotterdam criteria capture, particularly Phenotypes C and D.
The most prevalent phenotype in the Indian cohort was Phenotype C, followed by Phenotype A. This has implications for how Indian clinicians should approach PCOS assessment: regular cycles do not rule out PCOS when androgens are elevated and PCOM is present.
What phenotyping changes about management
Phenotype determines:
- Metabolic monitoring intensity: Phenotypes A and B require closer monitoring for insulin resistance, dyslipidaemia, and diabetes risk. Phenotype D may require less aggressive metabolic surveillance.
- Androgen management: Phenotypes C and D without hyperandrogenism do not need anti-androgen therapy.
- Fertility approach: Phenotype D patients with regular ovulation require a different fertility evaluation than anovulatory phenotypes.
- Long-term risk counselling: The cardiovascular and metabolic risk discussion should be proportionate to phenotype severity.
A diagnosis of "PCOS" without phenotype classification is an incomplete diagnosis. A doctor who identifies which phenotype applies can offer a management plan that is specific to the patient's biological reality — not a generic protocol built for the average PCOS patient, who may not resemble any individual patient at all.