Objective: Acne vulgaris is androgen dependent but the hormonal mechanisms are unclear. Although there have been many studies of serum hormones in women with acne there are few studies in men and the results are conflicting. We have therefore carried out a further study in men.

Design and patients: Fifty men with acne vulgaris were age-matched against 50 normal men.

Measurements: Serum levels of dehydroepiandrosterone sulphate (DHEAS), 17 alpha-hydroxyprogesterone, 11-deoxycortisol, androstenedione and testosterone were measured by radioimmunoassay, sex hormone binding globulin (SHBG) by immunoradiometric assay and LH, FSH and oestradiol by automated ELISA.

Results: The acne patients had higher levels of androstenedione, median 7.35 nmol/l, (interquartile range 2-7) vs 6.05 (2.3), P = 0.004; testosterone, 21.7 nmol/l (7.5) vs 17.55 (7.7), P = 0.04; and free androgen index (FAI) 78.26 (40) vs 65.06 (20), P = 0.007, but also had higher levels of 11-deoxycortisol, 13.65 nmol/l (4.3) vs 12.0 (4.3), P = 0.022. The LH, FSH, 17 alpha-hydroxyprogesterone, DHEAS, oestradiol and SHBG levels were not significantly different. Examination of the Spearman rank correlation coefficient matrices for the serum levels of 17 alpha-hydroxyprogesterone, androstenedione and 11-deoxycortisol showed that the strongest correlation was between androstenedione and 11-deoxycortisol.

Conclusion: Although there was overlap between the results of the acne patients and controls the acne patients tended to have higher levels of androstenedione, testosterone, free androgen index and 11-deoxycortisol. The higher levels of 11-deoxycortisol are suggestive of 11 beta-hydroyxlase dysfunction which could be due to a primary adrenal defect or a consequence of raised androgens. Also, a pathway between androstenedione and 11-deoxycortisol has been described in sheep and, although unsubstantiated in man, requires consideration.