Hormonal Tests for PCOS
Measure total testosterone, androstenedione, and LH as the core hormonal panel for diagnosing PCOS in reproductive-age women, with total testosterone being the single most important test. 1
Essential Hormonal Tests
First-Line Tests (Highest Diagnostic Yield)
Total Testosterone: This is your most valuable single marker, with 70% sensitivity for PCOS 1. When combined with clinical features (oligomenorrhea and/or hirsutism) plus ultrasound findings, testosterone provides the best diagnostic accuracy.
Free Androgen Index (FAI): If available, FAI demonstrates the highest diagnostic value with 91.7% sensitivity and 87.0% specificity at a cutoff of 4.27 2. Calculate this using total testosterone and SHBG.
Androstenedione: This adds incremental value with 53% sensitivity 1. When you combine testosterone, androstenedione, or LH (either alone or together), you'll detect 86% of PCOS cases 1.
LH and FSH: Measure both to calculate the LH/FSH ratio, though this has limited standalone value (only 41-44% sensitivity) 1. The ratio is elevated in PCOS but should not be used as a sole diagnostic criterion 1.
Additional Tests to Consider
Anti-Müllerian Hormone (AMH): Excellent diagnostic marker across all PCOS phenotypes with 87% sensitivity and 97% specificity at a cutoff of 5.28 ng/mL 3. AMH cutoffs vary from 4.4 to 5.6 ng/mL depending on phenotype, with sensitivities ranging 86-97% 3.
DHEA-S: Helps exclude adrenal causes of hyperandrogenism, though it shows no significant difference between PCOS and controls in most studies 1. Recent evidence shows DHEA-S elevation doesn't consistently align with adrenal-origin androgen dominance 2.
SHBG: Useful for calculating FAI but doesn't differ significantly between PCOS and controls when measured alone 1. It correlates negatively with BMI in PCOS patients 1.
Tests with Limited Value
Abandon the LH/FSH ratio as your primary diagnostic criterion due to its low sensitivity (35-44%) 1. While mean LH levels are significantly higher in PCOS, only 35% of patients show abnormal values 1.
Prolactin and estradiol (E2) show no statistically significant differences between PCOS and control groups 4 and are primarily useful for excluding other diagnoses rather than confirming PCOS.
Critical Pitfalls to Avoid
Assay Variability: LH and FSH values differ significantly between assay methods even when using the same reference preparation 1. Establish your laboratory's normal range using regularly ovulating women in the early follicular phase 1.
Timing Matters: Draw all hormones in the early follicular phase (days 2-5) for women with any menstrual cycles. For those with amenorrhea, timing is less critical but document it.
Don't Over-Rely on Single Tests: The combination approach (testosterone + androstenedione + LH) captures 86% of cases 1, while FAI alone detects 86.5%, increasing to 97.3% when combined with total testosterone and androstenedione 2.
Additional Screening Beyond Androgens
The Endocrine Society guideline emphasizes screening for metabolic complications 5:
- Oral glucose tolerance test (preconceptual assessment)
- Blood pressure
- BMI assessment
These address the increased risk of gestational diabetes, pre-eclampsia, and preterm delivery in PCOS pregnancies 5.
Special Considerations for Adolescents
In adolescent populations, avoid using polycystic ovarian morphology and AMH levels as diagnostic criteria due to limited specificity during normal puberty 6, 7. Focus on persistent menstrual irregularities and clear hyperandrogenism only 7.