What are the diagnostic criteria for Polycystic Ovary Syndrome (PCOS)?

PCOS Diagnostic Criteria

PCOS is diagnosed when at least two of the following three Rotterdam criteria are present: (1) oligo- or anovulation, (2) clinical and/or biochemical hyperandrogenism, and (3) polycystic ovarian morphology on ultrasound—after excluding other causes of hyperandrogenism and menstrual irregularity. 1, 2

Core Diagnostic Framework

The Rotterdam criteria require meeting 2 out of 3 of the following features 1, 3:

1. Ovulatory Dysfunction

• Menstrual cycle length >35 days indicates chronic anovulation 1
• Cycles between 32-35 days or slightly irregular patterns (32-36 days) require further assessment for ovulatory dysfunction 3
• Persistent oligomenorrhea 2-3 years beyond menarche predicts ongoing menstrual irregularities 3

2. Hyperandrogenism (Clinical or Biochemical)

Clinical signs to assess:

• Hirsutism (gradual onset, intensifies with weight gain) 3
• Acne (severe or resistant to standard treatments including isotretinoin) 3
• Androgenic alopecia (vertex, crown, diffuse pattern, or bitemporal with frontal hairline loss in severe cases) 3
• Clitoromegaly (suggests rapid-onset virilizing neoplasm rather than PCOS) 3

Biochemical testing hierarchy:

• Total testosterone (TT): 74% sensitivity, 86% specificity—best initial marker 1, 4
• Calculated free testosterone (cFT): 89% sensitivity, 83% specificity—highest sensitivity when calculated using Vermeulen equation from high-quality TT and SHBG 1, 4
• Free androgen index (FAI): 78% sensitivity, 85% specificity—use cautiously when SHBG <30 nmol/L 1, 4
• Androstenedione (A4): 75% sensitivity, 71% specificity—useful when SHBG is low 1, 4
• DHEAS: 75% sensitivity, 67% specificity—most reliable for adrenal androgen production, particularly in women <30 years 1, 4

Critical testing methodology:

• Liquid chromatography-tandem mass spectrometry (LC-MS/MS) is mandatory for accurate androgen measurement, with superior specificity (92% vs 78%) and sensitivity (71% vs 74%) compared to direct immunoassays 1, 4

3. Polycystic Ovarian Morphology (PCOM)

Gold standard ultrasound criteria:

• Follicle number per ovary (FNPO) ≥20 follicles: 87.64% sensitivity, 93.74% specificity 5, 1
• Alternative markers when accurate follicle counting impossible: ovarian volume (OV) >10 mL or follicle number per single cross-section (FNPS) 5, 1
• Technical requirements: transvaginal ultrasound with ≥8 MHz transducer frequency in adults 5, 1

Age-Specific Diagnostic Modifications

Adults (18-50 years)

• Full Rotterdam criteria apply 5
• Transvaginal ultrasound is appropriate and recommended 5, 1

Adolescents (<20 years, ≥1 year post-menarche)

• Ultrasound should NOT be used as primary diagnostic criterion in adolescents <8 years post-menarche due to high false-positive rates from normal multi-follicular ovaries 1, 4, 3
• Diagnosis relies on persistent oligomenorrhea 2-3 years beyond menarche PLUS biochemical hyperandrogenism (elevated free testosterone and FAI) 4, 3
• Menstrual irregularities and anovulatory cycles are common in the first 2-3 years post-menarche due to hypothalamic-pituitary-ovarian axis immaturity 3

Essential Exclusion Criteria

Must rule out the following conditions before confirming PCOS diagnosis 1:

• Cushing's syndrome: buffalo hump, moon facies, hypertension, abdominal striae 1
• Androgen-secreting tumors (ovary or adrenal): rapid onset, severe hyperandrogenism, clitoromegaly 1, 3
• Non-classic congenital adrenal hyperplasia: measure 17-hydroxyprogesterone 1, 3
• Thyroid disease and prolactin disorders 1
• Primary hypothalamic amenorrhea and primary ovarian failure 1
• Acromegaly and genetic defects in insulin action 1

Mandatory Metabolic Screening

All women diagnosed with PCOS require screening for:

• Type 2 diabetes and glucose intolerance (fasting glucose; OGTT if BMI >30 kg/m² with normal fasting glucose) 1, 6
• Dyslipidemia with fasting lipoprotein profile 1
• Cardiovascular risk factors including BMI, waist-hip ratio, blood pressure 1, 6

Common Diagnostic Pitfalls

• Avoid relying on direct immunoassays for testosterone measurement—they produce inaccurate diagnoses 4
• Do not use ultrasound in adolescents <17 years as first-line investigation due to poor specificity 1, 3
• Do not diagnose based on hirsutism using Ferriman-Gallwey score alone—requires biochemical confirmation 6
• Screening for elevated plasma LH is no longer necessary; GnRH testing serves no purpose 6