PCOS Hormone Ratio and Androgen Levels
For diagnosing PCOS, measure total testosterone (TT) and free testosterone (FT) using mass spectrometry as first-line tests; the LH:FSH ratio is NOT a diagnostic criterion and should not be used to diagnose PCOS, though it may provide supplementary information about disease phenotype. 1
First-Line Androgen Testing
The 2025 International PCOS Guidelines establish clear priorities for biochemical hyperandrogenism assessment:
Primary Tests (in order of preference):
- Total Testosterone (TT): Sensitivity 74%, specificity 86%, AUC 0.87 1
- Free Testosterone (FT): Sensitivity 89%, specificity 83%, AUC 0.85 1
- Ideally measured by equilibrium dialysis or ammonium sulfate precipitation
- Can be calculated as free testosterone (cFT) using the Vermeulen equation from high-quality TT and SHBG measurements 1
Alternative When Mass Spectrometry Unavailable:
- Free Androgen Index (FAI): Sensitivity 78%, specificity 85%, AUC 0.87 1
- Calculated from TT/SHBG ratio
- Caveat: Becomes inaccurate when SHBG <30 nmol/L 1
Critical Technical Point:
Liquid chromatography-tandem mass spectrometry (LC-MS/MS) is superior to immunoassays for all androgen measurements. 1 Direct immunoassays are prone to interference and reduced accuracy, particularly at the low testosterone concentrations typical in women 1.
Second-Line Androgen Tests
If TT or FT are not elevated but clinical suspicion remains high:
- Androstenedione (A4): Sensitivity 75%, specificity 71%, AUC 0.80 1
- DHEAS: Sensitivity 75%, specificity 67%, AUC 0.77 1
Important limitation: Both A4 and DHEAS have poorer specificity than TT/FT, increasing risk of overdiagnosis 1. Use only when first-line tests are negative but clinical features strongly suggest PCOS.
The LH:FSH Ratio: Not Diagnostic
The LH:FSH ratio is NOT part of the Rotterdam diagnostic criteria for PCOS and should not be used as a diagnostic test. 2
What the Research Shows:
- PCOS patients typically have elevated LH:FSH ratios (mean 1.25 vs 0.71 in controls) 2
- An LH:FSH ratio >2 occurs in only 18% of PCOS patients 3
- The ratio shows marked overlap between PCOS patients and controls 4
- Key finding: LH:FSH ratio correlates with ovarian morphology (antral follicle count) and anovulation, but NOT with androgen levels or hirsutism 2, 3
Clinical Utility of LH:FSH:
While not diagnostic, an elevated LH:FSH ratio may indicate:
- A neuroendocrine-dominant PCOS phenotype 3, 5
- Higher AMH levels and more prominent polycystic ovarian morphology 3
- Greater likelihood of anovulation 6, 3
- The normoandrogenic phenotype (Phenotype D) shows the highest LH:FSH ratios 3
Bottom line: The LH:FSH ratio provides phenotypic information but has no role in establishing the diagnosis of PCOS 2.
Clinical Hyperandrogenism
When biochemical testing is unavailable or equivocal, clinical hyperandrogenism can establish the diagnosis:
- Hirsutism: Present in 65-75% of PCOS patients (lower in Asian populations) 7
- Assess using modified Ferriman-Gallwey score
- Most reliable clinical marker of hyperandrogenism 7
- Acne: Present in 14-25% of PCOS patients 7
- Androgenic alopecia: Less frequent, presents later 7
Important: Acne and alopecia are poor markers for hyperandrogenism compared to hirsutism 7.
Practical Algorithm
- First: Measure TT and FT by LC-MS/MS
- If mass spectrometry unavailable: Use FAI (ensure SHBG >30 nmol/L)
- If TT/FT negative but high clinical suspicion: Consider A4 and DHEAS
- If biochemical testing unavailable: Use clinical hyperandrogenism (preferably hirsutism with Ferriman-Gallwey score)
- LH:FSH ratio: Optional for phenotyping, NOT for diagnosis
Common Pitfalls
- Don't use direct immunoassays for testosterone measurement—they lack accuracy in the female range 1
- Don't rely on LH:FSH ratio for diagnosis—it's not part of Rotterdam criteria 2
- Don't use FAI when SHBG <30 nmol/L—results become unreliable 1
- Don't assume all PCOS patients have elevated androgens—the normoandrogenic phenotype exists and is diagnosed by ovulatory dysfunction plus polycystic ovaries 6