Assessment of 5 alpha-reductase activity in hirsute women: comparison of serum androstanediol glucuronide with urinary androsterone and aetiocholanolone excretion.

ABSTRACT

OBJECTIVE: Recent evidence suggests that androstanediol glucuronide (AG), a metabolite of dihydrotestosterone (DHT) formed in skin, is frequently elevated in hirsute women, presumably reflecting enhanced 5 alpha-reductase activity. An alternative method of demonstrating 5 alpha-reductase activity is the androsterone (A)/aetiocholanolone (E) ratio in urine. A and E are the 5 alpha- and 5 beta-reduced metabolites, respectively, of androstenedione, which is the principal metabolite of dehydroepiandrosterone (D). Although serum AG and the urinary A/E ratio have both been considered valid methods for assessing 5 alpha-reductase activity, the two have not been previously compared in hirsute women. The present study was undertaken to assess 5 alpha-reductase activity in hirsute patients as determined by these two different methods. PATIENTS AND MEASUREMENTS We surveyed 47 untreated women (ages 17-33) with various degrees of hirsutism. Serum testosterone, bioavailable testosterone, dehydroepiandrosterone sulphate, and AG were determined. Additionally, A, E and D were measured in 24-hour collections of urine. RESULTS: For the 47 women, 37 had elevated blood levels of AG (17.4 +/- 2.2, mean +/- SEM; normal < 8 nmol/l), but only 18 of these had an increased urinary A/E ratio (> 1.5). All but one of the remainder had elevated urinary and/or serum androgen levels. Overall, no significant correlation between AG and A/E was observed. There was a highly significant correlation between AG in serum and A in urine (r = 0.82, P < 0.001). AG was also positively related to dehydroepiandrosterone sulphate (r = 0.64; P < 0.005), bioavailable testosterone (r = 0.6; P < 0.001), aetiocholanolone (r = 0.58; P < 0.001) and total testosterone (r = 0.52; P < 0.01). In contrast, A/E was not significantly related to androgen production. CONCLUSIONS: There is a poor correlation between AG and the A/E ratio in hirsute women. Although AG may be raised by increased 5 alpha-reductase activity, it is probably also affected by the presence of elevated androgens regardless of 5 alpha-reductase activity.