To assess the agreement of the NCEP ATP-III and the IDF definitions of metabolic syndrome and to determine their predictive values for the diagnosis of insulin resistance. For this purpose, we recruited 150 subjects (94 women and 56 men) and determined the presence of metabolic syndrome using the NCEP-ATP III and IDF definitions. We evaluated their insulin sensitivity S(I) using Caumo's oral minimal model after a standardized hyperglucidic breakfast test. Subjects whose S(I) was in the lowest

more »

... artile were considered as insulin resistant. We then calculated sensitivity, specificity, positive and negative predictive values of both definitions for the diagnosis of insulin resistance. The prevalence of metabolic syndrome was 37.4% (NCEP-ATP III) and 40% (IDF). Agreement between the two definitions was 96%. Using NCEP-ATP III and IDF criteria for the identification of insulin resistant subjects, sensitivity was 55.3% and 63%, specificity was 68.8% and 67.8%, positive predictive value was 37.5% and 40%, negative predictive value was 81.9% and 84.5%, respectively. Positive predictive value increased with the number of criteria for both definitions. Whatever the definition, the scoring of metabolic syndrome is not a reliable tool for the individual diagnosis of insulin resistance, and is more useful for excluding this diagnosis.