Laboratory Evaluation for PCOS
For PCOS evaluation, draw a fasting morning sample (ideally days 3-6 of the menstrual cycle or any time if amenorrheic) that includes: total and free testosterone, LH, FSH, TSH, prolactin, and 17-hydroxyprogesterone to exclude mimicking conditions. 1, 2
Core Hormone Panel
The essential laboratory workup requires measuring specific hormones to both diagnose PCOS and exclude other conditions that present similarly:
Primary Androgen Assessment
- Total testosterone - Most laboratories should measure this, though assay quality varies significantly 1
- Free testosterone - When measured by gold-standard assays (LC-MS/MS or equilibrium dialysis), this is the most useful indicator of PCOS 1, 3
- SHBG - Allows calculation of Free Androgen Index (FAI = [Total testosterone/SHBG] × 100) if direct free testosterone unavailable 1, 3
Critical caveat: Immunoassays for testosterone are notoriously unreliable at the low concentrations seen in women. If your initial testosterone is normal but clinical suspicion remains high, recheck using a specialty laboratory with LC-MS/MS methodology 1, 4. The 2025 meta-analysis confirms that free testosterone measured by LC-MS/MS has superior diagnostic accuracy (AUC 0.90) compared to direct immunoassays (AUC 0.82) 3.
Gonadotropins
Mandatory Exclusion Tests
These three tests must be performed in all women being evaluated for PCOS 1, 2:
- TSH (and free T4 if TSH abnormal) - Excludes thyroid dysfunction causing menstrual irregularity 1, 7, 2
- Prolactin - Excludes hyperprolactinemia, which can present identically with amenorrhea and hirsutism 1, 7, 2
- 17-hydroxyprogesterone (8 AM fasting) - Screens for nonclassic congenital adrenal hyperplasia, found in 1.5-6.8% of hyperandrogenic women 1, 2
Additional Androgens (Selective Use)
DHEA-S - Consider measuring to assess adrenal androgen contribution 1, 4
Androstenedione - Has lower diagnostic accuracy (AUC 0.80) than testosterone or FAI 3, so not essential for routine diagnosis
Metabolic Screening
While not required for diagnosis, metabolic assessment is critical given PCOS-associated risks 7, 2:
- Fasting glucose and 2-hour oral glucose tolerance test (75g load) - ACOG recommends this for all PCOS patients due to increased diabetes risk 7
- Fasting lipid panel - Screen for dyslipidemia (total cholesterol, LDL, HDL, triglycerides) 7
- Fasting insulin - Some use glucose/insulin ratio <4.5 as insulin resistance indicator, though not standardized 8
Optional/Adjunctive Tests
- AMH (anti-Müllerian hormone) - Often elevated in PCOS and can be helpful adjunct 1, but not required for diagnosis and has limited specificity in adolescents 9
- Mid-luteal progesterone - If ovulatory status unclear; <6 nmol/L indicates anovulation 5
Timing and Practical Considerations
Optimal timing: Draw blood in early follicular phase (days 3-6 of cycle) in regularly cycling women, as reference ranges are standardized to this timeframe 4. In amenorrheic women, draw at any time but document this 1.
Morning collection: Testosterone and 17-OHP must be drawn in early morning (8 AM) due to diurnal variation 1, 4.
Assay familiarity: Know your local laboratory's reference ranges and assay methods - they vary significantly between facilities 1. If using immunoassays and results don't match clinical picture, send to specialty lab with mass spectrometry 1, 4.
Common Pitfalls to Avoid
- Don't rely solely on total testosterone from standard immunoassays - they lack sensitivity at female ranges 1, 3
- Don't skip the exclusion tests (TSH, prolactin, 17-OHP) - these conditions are treatable and have different management 2
- Don't check LH/FSH during mid-cycle - you may catch the ovulatory surge (high FSH/LH with estradiol >150 pg/mL) and misinterpret results 1
- Don't measure androgens post-ictally in epilepsy patients - prolactin will be falsely elevated 5