RESUMEN
This study investigated the association of six different anthropometric markers with metabolic syndrome to find the most suited to predict children at risk. Sample comprises 1324 Portuguese youth (701 girls, 623 boys), aged 10-17 years. Six anthropometric markers were included: body mass index (BMI), BMI z-score, tri-ponderal index (TPI), waist circumference (WC), WC/height ratio (WC/H), and WC/H adjusted ratio (WC/Hadj ). A standardized metabolic risk score (zMR) was computed by summing of standardized values for fasting glucose, triglycerides, high-density lipoprotein cholesterol, and mean arterial blood pressure. The associations between zMR and anthropometric markers were assessed using univariate and multivariate analyses. Receiver operating characteristic (ROC) curve analysis was used to identify the optimal values that best predict metabolic risk of each anthropometric marker. Among the studied predictors, BMI z-score, followed by BMI and WC, was most highly associated with zMR, while WC/Hadj was the weakest predictor. ROC analyses showed significant AUCs for all markers, yet the discrimination was poor (AUCs from 0.60 to 0.68), with sensitivity ranging from 45.5% to 67.5% and specificity from 72.6% to 81.9%. The optimal cut-off values to predict metabolic risk were 1.62, 23.1 kg/m2 , 71.0 cm, 18.0 kg/m3 , 0.47, and 0.50, for BMI z-score, BMI, WC, TPI, WC/H, and WC/Hadj , respectively. BMI z-score, followed by BMI and WC, were the most relevant anthropometric markers to predict metabolic risk in youth, while WC/Hadj was the worst predictor. Results suggest that anthropometric markers should continue to be used as clinical tools to identify youth at risk.