A cross-sectional comparison of clinical and endocrine parameters among phenotypes of polycystic ovarian syndrome in iranian population

Abstract

Background and objective: Polycystic ovary syndrome (PCOS), as a major hormonal issue among reproductive age women, is categorized into four phenotypes regarding hyperandrogenism, oligo-anovulation and polycystic ovary morphology. This study aims to compare and predict the phenotypes based on metabolic and hormonal parameters. Methods: This cross-sectional study was conducted using clinical, hormonal, and ultrasound information of PCOS women referred to a major referral infertility center in Tehran between January 2014 and April 2017. Based on the Rotterdam Consensus of 2003 criterion, the phenotypes were defined as A: presence of olig-anovulation, hyperandrogenism, polycystic ovary morphology; B: presence of olig-anovulation and hyperandrogenism; C: presence of hyperandrogenism and polycystic ovary morphology; and D: presence of olig-anovulation and polycystic ovary morphology. Results: Among of 635 patients, 372 (58.6%) were A, 28 (4.4%) were B, 34 (5.4%) were C, and 201 (31.7%) were D phenotypes of PCOS. Higher prolactin increases the odds of phenotype C than A, and higher anti-mullerian hormone increases the odds of phenotype A than B (OR = 1.22) and C (OR = 1.19). Older patients are more prone to be in phenotype D than A. Patients with higher mean serum glutamate-pyruvate transaminase has 0.03 more odds of phenotype B than A. History of abortion was associated with more likelihood of phenotype A than D. The receiver operating characteristic curve found a cut-off point of 7.15 for anti-mullerian hormone regarding morphology (area under curve = 0.66). Conclusion: This study showed that PRL, AMH, Age, SGPT, and history of abortion are affecting factors for distinguishing the phenotypes of PCOS. Keywords: Polycystic ovary syndrome, Phenotype, Hyperandrogenism, Anovulation, Morphology, Logistic regression