• Adriano Ruy Matsuo, Wendell Arthur Lopes, João Carlos Locatelli, Caroline Ferraz Simões, Gustavo Henrique de Oliveira, and Nelson Nardo.
• Department of Physical Education, Center of Health Sciences, State University of Maringa-UEM, Maringa, Brazil; Multidisciplinary Centre of Obesity Studies-NEMO, State University of Maringa, Maringa, Brazil.
• Nutrition. 2020 Jun 1; 74: 110744.

ObjectiveThe aims of this study were to assess the capacity of tri-ponderal mass index (TMI) to screen and predict insulin resistance (IR) in overweight Brazilian adolescents, comparing it with body mass index (BMI) predictive capacity; and to verify the IR predictive capacity of other assessment tools such as waist circumference (WC), and waist-to-height ratio (WHR).MethodsA cross-sectional study was carried out with 217 overweight adolescents, from both sexes, between 12 and 18 y of age. The participants were classified as having IR according to a previous established cutoff point of ≤3.16. IR was determined by the homeostasis model assessment of insulin resistance. The other assessment tools were measured with standard protocols.ResultsThere was no difference between TMI and BMI to explain the presence of IR. In girls, BMI presented a slightly better predictive capacity to explain IR than TMI. Moreover, WC was reported to be the most effective IR screening methods for girls. Receiver operating characteristic curves showed that TMI and BMI presented similar values of sensibility and specificity for boys. Nevertheless, BMI had a better sensibility and TMI had a better specificity for girls. Interestingly, WC demonstrated a strong sensibility for both sexes.ConclusionsTMI did not present a superior predictive accuracy for IR screening in overweight Brazilian adolescents when compared with BMI. TMI and BMI presented similar values of sensitivity and specificity for boys and WC a slightly superior IR predictive capacity for girls.Copyright © 2020 Elsevier Inc. All rights reserved.

17 keywords...

hide…