Background and aims: The main objective was to evaluate the prevalence of the metabolic syndrome in Caucasian women with PCOS, using either of the currently proposed definitions (NCEP/ATPIII, IDF and AHA/NHLBI) and, therefore, to estimate the concordance between these three classifications. Secondary objectives were to evaluate: i) which individual criterion of the metabolic syndrome is most strongly associated with PCOS; and ii) whether the severity of hyperandrogenemia, hyperinsulinemia and insulin resistance may influence the presence of the metabolic syndrome in PCOS women.
Methods and results: The metabolic syndrome was assessed in 200 Caucasian women with PCOS and in 200 Caucasian controls, matched for age and BMI, considering the NCEP/ATPIII, IDF and AHA/NHLBI definitions. PCOS women had an increased prevalence of the metabolic syndrome compared with controls: 32 versus 23% with the NCEP/ATPIII, 39 versus 25% with the IDF and 37 versus 24% with the AHA/NHLBI, respectively (Cohen's Kappa index between the three classifications, P < 0.001). Multivariate logistic regressions revealed that among the individual criteria of the metabolic syndrome, only low HDL-cholesterol levels were significantly associated with PCOS (P < 0.001) which, in turn, are related to insulin(AUC) (P = 0.029) but not to androgens.
Conclusion: This case-control study indicates a high prevalence of the metabolic syndrome in Caucasian PCOS women that is independent of the diagnostic classification used. Furthermore, it shows that low HDL-cholesterol is the criterion which best explains the high prevalence of the metabolic syndrome in PCOS subjects which, in turn, is influenced by hyperinsulinemia, rather than by hyperandrogenemia.