Anthropometric Indicators of Body Composition Associated With Lipid and Glycemic Profiles in Overweight Brazilian Children and Adolescents From 2008 to 2020

BackgroundAnthropometric indicators have been used to predict health problems. The objective was to determine which indicators present better correlation with dyslipidemia, hyperglycemia and peripheral insulin resistance, as well as the cutoff points capable of predicting lipid and glycemic alterations in Brazilian children and adolescents. MethodsA cross-sectional study conducted with 568 overweight individuals, aged between 5 and 18 years, living in Southeast and South Brazilian regions, submitted to anthropometric and body composition evaluation by bioimpedance, in addition to fasting laboratory tests [total cholesterol (TC), triglycerides (TG), low-density lipoprotein cholesterol (LDL-c), high-density lipoprotein cholesterol (HDL-c), fasting glycemia, and homeostasis model assessment-insulin resistance (HOMA-IR)]. Pearson's correlation was used to evaluate the association between anthropometric indicators and serum biomarkers. The ROC curve with Youden's J index was used to suggest anthropometric cutoff points with better ability to predict or rule out lipid and glycemic changes. ResultsCutoff points obtained for the z-score of body mass index (BMI), waist circumference (WC), and waist circumference for height (WC/H) showed high specificity (52 to 87%) and low sensitivity (23 to 59%), indicating greater ability to exclude changes in HDL-c, TG, and HOMA-IR levels. Cutoff points suggested for BMI ranged from +1.86 to +2.20 z-score. WC cutoff points ranged from +1.29 to +1.72, and, for the WC/H index, from +1.21 to +1.25. It was suggested the use of the following cutoff points to rule out changes in HDL-c, TG, and HOMA-IR values in clinical practice: BMI < z-score +2 and WC/H < z-score +1.29. In body fat percentage (BFP) analyses, the cutoff point < of 34% may be able to rule out changes in HDL-c (specificity of 70%), while the cutoff point > 36.6% may be able to predict changes in the HOMA-IR index (sensitivity of 76%). ConclusionIt is not yet possible to state which anthropometric parameter has the best correlation with lipid and glycemic alterations in overweight children and adolescents. We suggest considering BMI, WC, and WC/H cutoff points together to rule out changes in HDL-c, TG, and HOMA-IR, and use the BFP cutoff point to predict changes in HOMA-IR.