Clinical and laboratory features in women with different phenotypes of polycystic ovary syndrome.

Background: Misdiagnosis is still an obstacle in the workup of polycystic ovary syndrome. We compared patients' characteristics among phenotypes of the syndrome and assessed the association of the characteristics with oligo-anovulation, hyperandrogenism, and polycystic ovary.

Methods: In the infertility department of a teaching hospital, we performed a cross-sectional study. Based on the Rotterdam Criteria, we included 136 patients with polycystic ovary syndrome and 46 healthy controls.

Results: The most common phenotype was A (N.=45;33%). Overall, A and C had larger Body Mass Index (P=0.019 and 0.030, respectively) and waist circumference (P=0.005 and 0.003) than control. Also, A and D had higher serum anti-mullerian hormone than control (both P<0.001) or phenotype C (P<0.001 and 0.01). Phenotypes A and C had higher insulin than control (both P=0.004). The highest level of estradiol was for control and the lowest for B. The lowest high-density lipoprotein was for C; and A to C phenotypes had higher triglyceride than control (P=0.002, <0.001, and 0.041). Larger body mass was associated with hyperandrogenism (adjusted Odds Ratio 95% CI=1.11 [1.01, 1.23], P=0.034); higher anti-mullerian hormone and high-density lipoprotein were related to oligo-anovulation (1.2 [1.07, 1.37], P=0.004; 1.1 [1.00, 1.13], P=0.041). Higher insulin and lower high-density lipoprotein were associated with polycystic ovary (1.11 [1.03, 1.21], P=0.013; 0.9 [0.82, 0.97], P=0.014).

Conclusions: There are potentials in Body Mass Index, waist circumference, serum anti-mullerian hormone, insulin, estradiol, high-density lipoprotein, and triglyceride concentration for differentiating the phenotypes of polycystic ovary syndrome and for more precise diagnosis of hyperandrogenism, oligo-anovulation, and polycystic ovary morphology.