Correlation of clinical and biochemical hyperandrogenism in Thai women with polycystic ovary syndrome

Aim The aim of this study was to determine the correlation of clinical hyperandrogenism and biochemical hyperandrogenism (hyperandrogenemia) in Thai women with polycystic ovary syndrome (PCOS). Methods This cross‐sectional study was conducted at Siriraj Hospital, Thailand. Subjects were 145 women with PCOS who were diagnosed in accordance with the revised Rotterdam 2003 criteria and registered during January to July 2008. Clinical hyperandrogenism was assessed using the modified Ferriman–Gallwey score for hirsutism, the American Academy of Dermatology criteria for severity of acne, and the Ludwig scale for androgenic alopecia and virilization. Biochemical hyperandrogenism was determined from serum concentration of total testosterone (TT), dehydroepiandrosterone sulfate and free testosterone (FT). Results The participants had a mean age of 25.5 ± 6.5 years and a body mass index of 26.2 ± 6.9 kg/m2. The most common presenting symptom was oligomenorrhea or amenorrhea. The most common expression of clinical hyperandrogenism was acne (56.6%). Most of the participants (84.8%) had high serum‐FT. There was a statistically significant correlation between clinical and biochemical hyperandrogenism in the following pairs: hirsutism and FT (r = 0.3, P < 0.001); hirsutism and TT (r = 0.26, P < 0.001); and acne and TT (r = 0.26, P = 0.002). The others had little or no correlations. Conclusion Clinical hyperandrogenism is not a good predictor for biochemical hyperandrogenism in Thai women with PCOS. A modified Ferriman–Gallwey score cut‐off point of 8 has low sensitivity but high specificity for hyperandrogenemia; therefore, it is useful for the diagnosis but not useful for the exclusion of hyperandrogenemia in Thai women with PCOS.