Metabolically healthy polycystic ovary syndrome (MH-PCOS) and metabolically unhealthy polycystic ovary syndrome (MU-PCOS): a comparative analysis of four simple methods useful for metabolic assessment.

STUDY QUESTION Is it possible to distinguish metabolically healthy polycystic ovary syndrome (MH-PCOS) from metabolically unhealthy PCOS (MU-PCOS) by simple diagnostic tools such as body mass index (BMI), waist/hip ratio (WHR), at-risk category suggested by Androgen Excess Society (AES) and visceral adiposity index (VAI)? SUMMARY ANSWER VAI could be an easy and useful tool in clinical practice and in population studies for assessment of MU-PCOS. WHAT IS KNOWN ALREADY VAI is a good indicator of insulin sensitivity and cardiometabolic risk in oligo-ovulatory women with PCOS. STUDY DESIGN, SIZE, DURATION We conducted a cross-sectional study of 232 women with PCOS in a university hospital setting. PARTICIPANTS/MATERIALS, SETTING, METHODS Anthropometric, hormonal and metabolic parameters were evaluated. An oral glucose tolerance test measured areas under the curve (AUC) for insulin (AUC 2h insulin) and for glucose (AUC 2h glucose). Homeostasis model assessment of insulin resistance (HOMA2-IR), the Matsuda index of insulin sensitivity (ISI), the oral dispositional index (DIo) and VAI were determined. MAIN RESULTS AND THE ROLE OF CHANCE The prevalence of MU-PCOS according to the different criteria was: BMI, 56.0%; WHR, 18.1%; at-risk criteria of AES, 72.0% and VAI, 34.5%. The likelihood that a woman would exhibit MU-PCOS (except when diagnosed by the WHR criterion) showed a significant positive association with high HOMA2-IR [BMI criterion: (odds ratio (OR): 1.86; 95% confidence interval (CI): 1.43-2.41); risk criteria of AES (OR: 1.86; 95% CI: 1.36-2.56); VAI criterion (OR: 1.45; 95% CI: 1.17-1.80)] and a significant negative association with low ISI Matsuda [BMI criterion: (OR: 0.81; 95% CI: 0.72-0.91); risk criteria of AES (OR: 0.78; 95% CI: 0.69-0.89); VAI criterion (OR: 0.82; 95% CI: 0.71-0.94)]. Only MU-PCOS according to the VAI criterion showed a significant association with low DIo (OR: 0.85; 95% CI: 0.75-0.96); these women also showed a significant association with low luteal progesterone levels (OR: 0.97; 95% CI: 0.95-0.99). LIMITATIONS, REASONS FOR CAUTION The analysis is limited by the lack of a gold standard definition of metabolic health that would have allowed the execution of a receiver operator characteristic analysis of the four proposed criteria. WIDER IMPLICATIONS OF THE FINDINGS The results will facilitate the early recognition of cardiometabolic risk in women with PCOS before they develop overt metabolic syndrome.