Objective: To analyze the prevalence of the metabolic syndrome (MetS) defined by three sets of Adult Treatment Panel III (ATPIII)-derived criteria, and the ability of each definition to identify insulin-resistance (IR) in a wide cohort of outpatient children.
Subjects and methods: Seven hundred and twenty-four children consecutively observed in the Outpatient Pediatric Clinic of Pozzuoli Hospital during the period 2004-2009 were included in the study. Diagnosis of the MetS was made using three definitions: Cook, Jolliffe (which adopt age- and gender-specific cut-points) and de Ferranti. Insulin sensitivity was evaluated by homeostasis model assessment of insulin resistance (HOMA-IR). IR was defined by the 90th percentile of HOMA-IR in healthy non-obese Italian children grouped by gender and Tanner stage. The ability of each definition to identify IR was evaluated in terms of sensitivity and specificity.
Results: The prevalence of the MetS in the overall cohort was 11, 12 and 24% using Cook, Jolliffe and de Ferranti criteria, respectively. Sensitivity and specificity in relation to IR were 19 and 94% with Cook criteria, 21 and 92% with Jolliffe criteria, and 39 and 84% with de Ferranti criteria.
Conclusions: The prevalence of the MetS in children increases with increasing body weight. Among the three definitions analyzed, de Ferranti identifies a larger number of children with the MetS. The prediction of IR is weak with all definitions; on the contrary, the absence of MetS identifies fairly well children with low degree of IR.