Instability of different adolescent metabolic syndrome definitions tracked into early adulthood metabolic syndrome: Tehran Lipid and Glucose Study ( TLGS )

Background and objective There are substantial controversies about the clinical utility of adolescent metabolic syndrome ( MetS ). The current study examined the stability of adolescent MetS by assessing the agreement and discriminative abilities of four different definitions of adolescent MetS and the adult MetS definition during a 10.4‐yr follow up. Subjects and Methods For this study, 1424 adolescents (55.2% female), who participated in the framework of the Tehran Lipid and Glucose Study were included. Kappa was calculated for agreement between adolescent MetS definitions [Cook, de Ferranti, pediatric National Cholesterol Education Program ( NCEP ) and pediatric International Diabetes Federation ( IDF )] and the adulthood MetS definition defined by the joint interim statement ( JIS ) criteria. MetS persistence, instability, and incidence were assessed, and for each of the four adolescent definitions, sensitivity, specificity, and area under receiver operating curve ( AUC ) for the counting of categorical adulthood MetS components was evaluated. Results The agreement between the four adolescent MetS definitions and JIS was poor (κ = 0.094–0.255). All definitions showed low sensitivity and high specificity, except for de Ferranti's, which contrary to other definitions, had higher sensitivity and lower specificity. All four adolescent definitions revealed generally low AUCs (0.601–0.647). Compared with the pubertal group (11–14 yr), the predictive power was slightly higher in the late‐pubertal group (15–18 yr). Cook's and de Ferranti's definitions showed fairly better predictive powers (0.647 and 0.644, respectively). Across all definitions, instability ranged between 5.4 and 19.6%. Conclusion The adolescent definitions show considerable amount of instability defined as poor agreement and low discriminative abilities tracked into early adulthood.