The free androgen index is inaccurate in women when the SHBG concentration is low.

OBJECTIVE/CONTEXT: The free androgen index (FAI) is known to give erroneous results in men, but it is still a commonly used test for the investigation of hyperandrogenism in women. This study aimed to compare the results of the FAI with the gold standard equilibrium dialysis method for free testosterone in women. DESIGN/PATIENTS: Free serum testosterone T (ED-T) and total serum T (T) were measured by equilibrium dialysis and LC-MS/MS in patients with polycystic ovarian syndrome (n = 130), normal female controls (n = 53) and normal males (n = 120). Calculated free T (cFT) and free androgen index (FAI) were also measured in these patients. In addition, cFT was retrospectively calculated in 4223 female patients with a normal T (<1.6 nmol/L) routinely investigated for hyperandrogenism. RESULTS: The cFT showed good agreement with measured ED-T, and the ratio cFT/ED-T was stable across all SHBG concentrations. In contrast, the FAI/ED-T ratio and the FAI/cFT ratio increased when the concentration of SHBG fell below 30 nmol/L. CONCLUSIONS: The FAI is not a reliable indicator of free T when the SHBG concentration is low and would give misleading information in a large number of women being investigated for hyperandrogenism.