Reference percentiles for tri‐ponderal mass index and its association with general and abdominal obesity among Iranian children and adolescents: A report from the SHED LIGHT study

Abstract Background The body mass index (BMI) has some limitations in identifying obesity, particularly among children and adolescents. Objectives We sought to determine the reference percentiles of tri‐ponderal mass index (TMI) for Iranian population, and its relation to BMI, waist circumference (WC), and waist circumference‐to‐height ratio (WHtR). Methods Baseline characteristics for 14 641 children and adolescents were applied. The area under the characteristics (AUC) curve and optimal cut‐off points for TMI were used to define the accuracy of TMI for discriminating general and central obesity compared to other anthropometric measurements. Results TMI levels remained stable from 6 to 18 years with a small range of difference. The highest AUCs for discriminating obesity according to BMI and WHtR ≥0.5 from non‐obese subjects were related to TMI (0.980) and TMI (0.912), respectively, and these findings were similar for both sexes. Moreover, the AUC for TMI to discriminate central obesity by WHtR was higher than that for BMI (0.912 vs. 0.833, p < .001), and this difference was similar among sex and age groups. The optimal value of TMI for discriminating general obesity, obesity by WC, and obesity by WHtR were 15.2 (sensitivity 96.2 and specificity 89.8), 15 (sensitivity 86.9 and specificity 79.9), and 13.9 (sensitivity 85.1 and specificity 81.5), respectively. Conclusions TMI discriminated central obesity by WHtR more accurately than obesity by BMI among Iranian children and adolescents irrespective of sex. Besides it remained relatively constant across age‐ and sex‐specific groups providing a single optimal cut‐off point for screening obesity during childhood and adolescence.