Use of Tri-Ponderal Mass Index in Predicting Late Adolescent Overweight and Obesity in Children Aged 7–18

BackgroundCurrent reference systems using body mass index (BMI) or BMI z-scores to estimate overweight and obesity risk in adolescents are complex to use. An easy and effective measure and cutoffs such as the tri-ponderal mass index (TMI) are in need for parents and grassroots health workers.ObjectiveThe aim of this study was to test whether cohort-derived TMI could be efficient for obesity prediction and to find out whether simplified TMI cutoffs could be used in the prediction.MethodsData were obtained from a 12-year retrospective growth cohort generated in Guangdong, China. A total of 17,815 children (53.9% were boys) with 151,879 follow-ups conducted annually between 2005 and 2016 were involved. Late adolescent overweight and obesity were defined based on the BMI z-score (WHO 2007 growth reference) of the last measurement, which happened at the mean age of 17.2 (SD: 0.7) for both sexes. Analysis of the area under the curve (AUC) of the receiver operating characteristic curves was used to find the most appropriate cutoff.ResultsIn total, 9,604 boys and 8,211 girls were included in the final analysis. TMI cutoffs performed better than WHO BMI cutoffs in the prediction of late adolescent overweight and obesity, with all corresponding AUCs <0.7. The simplified TMI cutoffs used to predict late adolescent overweight and obesity were 13.1 and 14.1 kg/m3 for children aged 7 to 15 years, respectively, with the corresponding AUCs ranging from 0.7315 (standard error, SE: 0.0132) to 0.9367 (SE: 0.0052). The cutoffs for predicting late adolescent overweight and obesity for children aged 16 to 18 years were 14.0 and 15.8 kg/m3, respectively, with the corresponding AUCs ranging from 0.9189 (SE: 0.0048) to 0.9841 (95% CI: 0.0027).ConclusionTri-ponderal mass index with the ease of administration in practice could be a promising alternative screening tool to BMI for the prediction of late adolescent overweight and obesity.

Abstract licence: CC BY

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.