Determing the cutoff value of free androgen index for the diagnosis of hyperandrogenemia

Objective To determine the normative cutoff value of free androgen index (FAI) among reproductive-age women in China. Methods A multistage systematic cluster random sampling was conducted by which a total sample of 3009 women aged 20-45 years were randomly selected from 16 communities in two urban and two rural regions, and about half of them (n=1526) were selected by random sampling. Among these women, 444 subjects without known factors affecting androgen levels were selected as the reference group to undergo clinical and laboratory tests. Clinical data of subjects were collected and the FAI was calculated according to the following equation. The normative cutoff value of the FAI was computed based on the 5th-95th percentiles and K-means cluster analysis (K=2). Results A reference range of 0.7-6.4 of FAI was recommended by the 5th and 95th percentiles for the reference group. K-means cluster analysis identifed 6.1 as the cutoff value of FAI for the total population. The means of FAI decreased with age increasing among the total population, either evaluated by K-means or percentile analysis. The FAI value was negatively correlated with age (r=-0.185, P<0.000 1), and positively correlated with BMI (r=0.250, P<0.000 1), after the adjustment of age. Conclusion A reference range of FAI was 0.7-6.4, and the cutoff value of FAI for HA was 6.1. According to this cutoff value, the hyperandrogenic women present with higher risks for metabolic disorders. Key words: Hyperandrogenemia; Free androgen index; Total testosterone; Sex hormone binding globulin; K-means cluster analysis