Biological variation of total testosterone, free androgen index and bioavailable testosterone in polycystic ovarian syndrome: implications for identifying hyperandrogenaemia

Objective Hyperandrogenaemia is one of the three Rotterdam consensus diagnostic criteria for polycystic ovarian syndrome (PCOS) and may be measured by estimation of total testosterone, free androgen index (FAI) or bioavailable testosterone (BioT). The aim of this study was to compare the biological variability of total testosterone with that of the biological variability of both the FAI and BioT, to determine the least variable measurement for clinical practice. Design Comparative study. Patients Blood samples were collected after an overnight fast at 4-day intervals on 10 consecutive occasions from 12 PCOS patients and 11 weight- and age-matched control women. Measurements Duplicate samples of stored serum were analysed for total testosterone, SHBG and BioT in a single batch. Results The PCOS group had a significantly higher median BioT, FAI and total testosterone than controls. In both the PCOS and control groups, the intraindividual variance was small and similar for BioT and FAI. There was no significant difference between the within-subject biological coefficient of variation (CVI) for BioT, FAI and total testosterone. The maximum and minimum critical differences were +58% and -37% for BioT and +70% and -40% for FAI, respectively. Conclusion FAI appears to be the better diagnostic marker to distinguish hyperandrogenism in patients with PCOS, but once diagnosis has been made, all three methods should be equally good in monitoring further changes in the androgen status.