Endocrine Abstracts

Background: Free androgen index (FAI) is used to estimate free testosterone concentrations in patients undergoing investigation for polycystic ovarian syndrome (PCOS). However, it’s not widely used for diagnostic purpose due to lack of consensus in the literature. We present an appraisal of biochemical data from patients referred to a hyperandrogenaemia clinic with a view to establish an in-house cut-off for the diagnosis of hyperandrogenaemia in PCOS.

Method: Clinical and biochemical data were extracted from patient records (n = 220 from 2012-2020). Patient were categorised as (i) non-PCOS (Group 1), (ii) PCOS (hyperandrogenaemia) (Group 2) or (iii) PCOS (no hyperandrogenaemia) (Group 3) based on the Rotterdam criteria (2003). Hyperandrogenaemia was defined by a patient having ≥1 of the following: testosterone >1.9nmol/l , androstenedione >8.5nmol/l (both measured by LC-MS/MS) or FAI >4.5%.

Results: Patients were categorised as: Group 1 (n = 38), Group 2 (n = 132) and Group 3 (n = 46). Mean testosterone, androstenedione and FAI were 1.0 nmol/l , 3.8 nmol/l and 1.9% (Group 1); 2.2 nmol/l , 8.0 nmol/l and 6.8% (Group 2) and, 1.1 nmol/l , 4.6 nmol/l and 2.5% (Group 3). Testosterone and androstenedione were significantly different when Group 2 was compared to Groups 1 (P = <0.0001) and 3 (P = <0.0001). No significant difference was observed between Group 2 and Group 3 (P = 0.05). A significant difference was observed in FAI between all 3 groups (P = <0.02, all comparisons).

Conclusions: FAI was significantly different between all clinical groups and much lower than the frequently used cut-off of >4.5% currently used to identify hyperandrogenaemia. We propose that a lower FAI cut-off of 2.5% is adopted to help identify hyperandrogenaemia in patients undergoing investigation for PCOS.