Accuracy of Calculated Free Testosterone (cFT) Equations Based on SHBG Levels
Calculated free testosterone (cFT) equations become inaccurate when SHBG levels are below 30 nmol/L, particularly when combined with albumin levels ≤3.5 g/dL, at which point direct measurement by equilibrium dialysis is recommended for better accuracy. 1, 2, 3
SHBG Threshold for cFT Accuracy
- cFT calculations (most commonly using the Vermeulen equation) are reliable for most clinical evaluations but have specific limitations based on SHBG levels 1, 4
- The accuracy of cFT decreases significantly when SHBG concentrations fall below 30 nmol/L 1, 2
- When SHBG is <30 nmol/L, the Free Androgen Index (FAI) also becomes inaccurate 1
- The combination of low SHBG (≤30 nmol/L) with low albumin (≤3.5 g/dL) creates the greatest variance in cFT calculations, occurring in approximately 1.0-1.2% of samples 2, 3
Factors Affecting cFT Calculation Accuracy
- The Vermeulen equation appears to be the most robust approximation for cFT, with its accuracy being largely independent of SHBG, albumin, and total testosterone levels, though it tends to overestimate free testosterone 4
- Using a fixed albumin value of 4.3 g/dL (instead of measured albumin) produces no significant variance for most evaluations of cFT 2, 3
- Mass spectrometry methods (LC-MS/MS) for measuring total testosterone provide superior sensitivity and specificity compared to direct immunoassays when calculating cFT 1
- cFT calculated from high-quality total testosterone and SHBG measurements has shown high diagnostic accuracy with a pooled sensitivity of 0.89 and specificity of 0.83 1
Clinical Recommendations
- When SHBG levels are ≤30 nmol/L, especially with albumin ≤3.5 g/dL, free testosterone measurement by equilibrium dialysis is warranted for better accuracy 2, 3
- Equilibrium dialysis is considered the gold standard for free testosterone measurement but is resource-intensive and less practical for routine clinical use 1, 5
- For most clinical evaluations with normal SHBG levels (>30 nmol/L), calculated free testosterone using the Vermeulen equation provides reliable results 4, 6
- When using cFT calculations, prioritize total testosterone measurements by LC-MS/MS over immunoassay methods for improved accuracy 1
Common Pitfalls and Caveats
- Free Androgen Index (FAI) should not be used in men when SHBG concentrations are at extremes, and in women when SHBG concentrations are low 5
- Different cFT calculation equations can yield significantly different results, with mean biases ranging from 5.8% to 56.0% between methods 6
- Assay harmonization issues affect both testosterone and SHBG measurements, which can profoundly impact the accuracy of calculated free testosterone 5
- Reference intervals for cFT vary based on the equation used, making it important to use method-specific reference ranges 6