Background: This study examines the characteristics and agreement between different definitions of metabolic syndrome (MetS) and insulin resistance (IR).
Methods: A total of 347 non-diabetic individuals who were ≥ 20 years of age were selected from the Tehran Lipid and Glucose Study (TLGS). Subjects were categorized as having MetS by the Adult Treatment Panel III (ATP III) and the Joint Interim Statement (JIS). IR was estimated by using the homeostasis model assessment (HOMA-IR).
Results: According to ATP III and JIS criteria 38.9% and 38.2% of subjects had MetS respectively. The sensitivity of ATP III was 52.3% and specificity was 65%; for JIS the sensitivity was 52.3%, with a specificity of 66.5%. Kappa between ATP III or JIS and HOMA-IR was 0.14 and 0.16, respectively. Based on receiver operating characteristic (ROC) analysis, the use of waist circumference (WC) and fasting plasma glucose (FPG) for the diagnosis of IR in women showed a diagnostic accuracy equal to or instead of counting MetS components using modified ATP III or JIS. WC optimal cut points for prediction of IR were 93.5 cm for men and 92.5 cm for women.
Conclusions: ATP III and JIS definitions have low sensitivities and specificities for detecting IR. There is poor agreement between these criteria and IR.