Diagnostic performance of body mass index in detection of obesity using different cutoff points for excess body fat

Nutrition. 2020 Nov-Dec:79-80:110950. doi: 10.1016/j.nut.2020.110950. Epub 2020 Jul 22.

Abstract

Objectives: The aim of this study was to assess the diagnostic sensitivity of body mass index (BMI) in detecting obesity according to different cutoff points in order to classify a high body fat percentage (%BF) in adolescents and young adults.

Methods: This was a cross-sectional study conducted with 2447 adolescents 18 and 19 y of age residing in São Luís, Brazil and 951 young adults 21 to 23 y of age residing in Ribeirão Preto, Brazil. Three references were used to define a high %BF (i.e., those of Williams et al., Ramírez-Vélez et al., and Macias et al.). The area under the receiver operating characteristic area under the curve (AUC) was used to assess the performance of BMI.

Results: Sensitivity ranged from 38.3% to 54.1% among boys and from 12.7% to 72.7% among girls. Among young adults, it ranged from 52.3% to 67.1% in men and from 33.7% to 86.6% in women. The AUC ranged from 0.69 to 0.76 among boys, from 0.56 from 0.85 among girls, from 0.75 to 0.80 among men, and from 0.67 to 0.88 among women. The best cutoff points for the BMI were 24.79 to 25.10 kg/m2 for boys, 21.89 to 27.04 kg/m2 for girls, 26.43 to 28.22 kg/m2 for men, and 23.34 to 29.28 kg/m2 for women.

Conclusion: The use of different references for the classification of a high %BF implied a difference in the diagnostic sensitivity of the BMI. Higher cutoff points resulted in greater sensitivity and ability to differentiate individuals with and without obesity.

Keywords: Body fat; Body mass index; Obesity; ROC curve; Sensitivity.

MeSH terms

  • Adipose Tissue*
  • Adolescent
  • Body Composition
  • Body Mass Index
  • Brazil
  • Cross-Sectional Studies
  • Female
  • Humans
  • Male
  • Obesity* / diagnosis
  • Obesity* / epidemiology
  • ROC Curve
  • Young Adult