[Exploration of the classification of polycystic ovarian syndrome].

OBJECTIVE: To investigate the clinical presentation, hormonal profile and metabolic abnormalities in subgroups of women with PCOS and explore a reasonable classification for PCOS.

METHODS: A cross-sectional study of 192 women with PCOS (14 - 38 years of age) was performed. The patients were divided into 3 groups of A, B and C according to the revised 2003 consensus on diagnostic criteria and also divided into 2 groups according to body mass index (BMI): group A (n = 110), long term anovulation, clinical and biochemical evidence of high androgen level, ovary enlargement with its size larger than 10 ml or number of small follicles of 2 - 9 mm >or= 12 under ultrasound with exclusion of other diseases caused by high androgen; group B (n = 46), long term anovulation, clinical and biochemical evidence of high androgen level; group C (n = 36), long term anovulation, ovary enlargement with its size larger than 10 ml or number of small follicles of 2 - 9 mm >or= 12 under ultrasound with exclusion of other disease caused by high androgen; obesity PCOS group (OB-PCOS, n = 70), BMI >or= 25 (kg/m(2)); no obesity PCOS group (NOB-PCOS, n = 122), BMI < 25 (kg/m(2)). One hundred and four women with bilateral tubal block factor caused infertility served as control group. Anthropometric measurements, Ferriman Gallwey hirsutism scoring, presence of acne and acanthosis nigricans were noted. Hormonal profile was assessed by measuring follicle-stimulating hormone (FSH), luteinizing hormone (LH), free testosterone (FT), prolactin (PRL), sex hormone binding globulin (SHBG). The metabolic profile was investigated by measurements of oral glucose tolerance test (OGTT), serum lipid levels, including total cholesterol (Chol), triglycerides (TG), high-density lipoprotein (HDL), and low-density lipoprotein (LDL). Hyperinsulinemia was estimated by measurement of fasting insulin (FINS) and insulin area under the curve (IAUC). The extent of insulin resistance (IR) and hyperandrogenism was estimated by homeostasis model assessment (HOMA) and free androgen index (FAI) respectively.

RESULTS: (1) Clinical phenotypes: the presence of obesity was 36.4% (70/192), among which 80.0% (56/70) were central obesity. Higher rates of acanthosis nigricans were observed in OB-PCOS group (35.7%, 25/70) compared with NOB-PCOS group (7.4%, 9/122; P < 0.01). Waist to hip ratio (WHR) was lower in group C than those in groups A and B (P < 0.05). (2) Endocrinology: FAI level was higher in OB-PCOS group than in NOB-PCOS group (P < 0.01), whereas LH/FSH ratio was lower in OB-PCOS group compared with NOB-PCOS group (P < 0.01). FAI level was higher in groups A and B than in group C (P < 0.01). SHBG, LH/FSH ratio did not differ between groups A, B, and C. (3) Metabolism: the prevalence of IR was 43.2% (83/192). A higher prevalence was observed in group OB-PCOS (82.8%, 58/70) compared with group NOB-PCOS (20.5%, 25/122; P < 0.01). FINS, HOMA-IR, glucose area under the curve (GAUC), IAUC and TG were higher in group OB-PCOS than in group NOB-PCOS (P < 0.01), whereas HOMA-IR, lipid profile did not differ between groups A, B, and C.

CONCLUSION: The classification according to the revised 2003 consensus on diagnosis reflects the basic characteristics of PCOS; while the classification based on obesity shows the severity of hyperandrogenism and degree of IR, and thus has substantial significance for evaluation of metabolic complications.