Accuracy of different cutoff points of body mass index to identify overweight according to body fat values estimated by DEXA

J Pediatr (Rio J). 2017 Jan-Feb;93(1):58-63. doi: 10.1016/j.jped.2016.04.010. Epub 2016 Aug 17.

Abstract

Objective: To evaluate the sensitivity and specificity of different cutoff points of body mass index for predicting overweight/obesity according to body fat values estimated by DEXA among Brazilian adolescents.

Methods: Cross-sectional study including 229 male adolescents aged 10-15 years, in which body adiposity and anthropometric measures were assessed. Nutritional status was classified by BMI according to cutoff points described in scientific literature.

Results: Moderate agreements were observed between body fat estimated by DEXA and cutoffs proposed by Cole et al. (K=0.61), Conde and Monteiro (K=0.65), Must et al. (K=0.61) and WHO (K=0.63). The BMI in continuous form showed good agreement with the Dexa (ICC=0.72). The highest sensitivity was observed for cutoff by Conde and Monteiro (0.74 [0.62, 0.84]) and the highest specificity by Cole et al. (0.98 [0.94, 0.99]). For the areas under the ROC curve of cutoff points analyzed, significant difference comparing the cutoff points by Cole et al. and Conde and Monteiro (0.0449 [0.00294, 0.0927]) was observed.

Conclusions: The cutoff proposed by Conde and Monteiro was more sensitive in identifying overweight and obesity when compared to the reference method, and the cutoff proposed by Cole et al. presented the highest specificity for such outcomes.

Keywords: Adolescent health; Body composition; Composição corporal; Obesidade; Obesity; Overweight; Saúde do adolescente; Sensibilidade e especificidade; Sensitivity and specificity; Sobrepeso.

MeSH terms

  • Adolescent
  • Anthropometry / methods
  • Body Composition
  • Body Mass Index*
  • Brazil
  • Child
  • Cross-Sectional Studies
  • Humans
  • Male
  • Overweight / classification
  • Overweight / diagnosis*
  • Pediatric Obesity / diagnosis*
  • Prognosis
  • ROC Curve
  • Sensitivity and Specificity