Higher metabolic risk with National Institutes of Health versus Rotterdam diagnostic criteria for polycystic ovarian syndrome in Turkish women.

Background: Polycystic ovary syndrome (PCOS) is a heterogeneous disease with well-established metabolic abnormalities in women of reproductive age. There are various diagnostic criteria to define and establish PCOS. However, data conflict regarding the optimal diagnostic criteria for PCOS and its metabolic consequences. We have evaluated the clinical, endocrine, and metabolic features between main PCOS phenotypes according to different diagnostic criteria.

Methods: In this prospective, case-control study, 175 consecutive women with PCOS, 41 ovulatory women with idiopathic hirsutism, and 109 healthy, nonhirsute, ovulatory controls were enrolled. The following diagnostic criteria were assessed: Hirsutism; ovulatory function; ovarian sonography; gonadotropin, testosterone, dehydroepiandrosterone (DHEA), 17-hydroxyprogesterone, fasting insulin, and lipid levels; oral glucose tolerance test; homeostasis model assessment for insulin resistance (HOMA-IR); body mass index (BMI); waist circumference; and the presence of metabolic syndrome.

Results: Of the 175 women meeting the Rotterdam criteria for PCOS, 121 (69%) had both androgen excess and ovulatory dysfunction, thus also meeting the National Institutes of Health (NIH) criteria. The other 54 (31%) had either androgen excess or ovulatory dysfunction. Women meeting the NIH criteria had the largest mean BMI, waist circumference, and HOMA-IR score; the highest mean testosterone, DHEA, triglyceride, and fasting insulin levels; the lowest mean progesterone level; and the greatest prevalence of insulin resistance (HOMA-IR score >3) among the groups.

Conclusions: The NIH criteria identify women at high risk for insulin resistance and metabolic syndrome. The Rotterdam criteria include women who have less severe metabolic implications.