Abstract
To determine the predictive value of a raised androgen level with an elevated anti-Müllerian hormone (AMH) for the diagnosis or exclusion of polycystic ovary syndrome (PCOS), a prospective cross-sectional study of 170 women (105 with PCOS type A and 65 normal) was undertaken. AMH was combined with one of, total serum testosterone (T); calculated free androgen index; salivary testosterone (salT); serum androstenedione (A); salivary androstenedione (salA). The diagnostic sensitivity and specificity of AMH (>35 pmol/l) alone for PCOS were 55% and 79% respectively. The diagnostic sensitivity and specificity of AMH (>35 pmol/l) with either an elevated T or raised FAI level for PCOS showed 100% specificity and a 100% positive predictive value. Conversely, diagnostic exclusion of PCOS was shown by an AMH <35 pmol/l with a normal T or FAI salivary testosterone giving 100% specificity and 100% positive predictive value. AMH with an elevated A or elevated salA level gave specificities of 87% and 94%, and positive predictive values 80% and 94%, respectively. Therefore, the combination of an AMH with a cut off of 35 pmol/l combined with a raised T and/or a FAI will confirm PCOS whilst a normal AMH with a normal T and/or FAI will exclude PCOS, thus addressing diagnostic uncertainty.