Role of tri-ponderal mass index in cardio-metabolic risk assessment in children and adolescents: compared with body mass index.

OBJECTIVE: To investigate the performance of weight/heightn in discriminating obesity-related cardio-metabolic risks, and compare their performance with BMI in Chinese and American children. METHODS: 57,201 Chinese children aged 7-18 and 10,441 American children aged 12-18 with complete record of sex, age, height, weight, and waist circumference were included. Analyses and comparisons of BMI, weight/height2.5, and weight/height3 were predominantly discussed, while BMI z score, converted by BMI based on 2007 WHO growth standard, was set as the reference. Log-binomial regression models and areas under receiver-operating characteristic curves were used to examine their abilities on identifying cardio-metabolic risks, including elevated blood pressure, impaired fasting glucose, and dyslipidemia. Misclassification rates of each index were calculated. RESULTS: Weight/height3 is relatively stable during childhood in both populations. Odds ratio of weight/height3 in discriminating cardio-metabolic risks ranged from 1.09 (95% CI: 1.04, 1.14) to 1.23 (95% CI: 1.22, 1.25) and 1.06 (95% CI: 1.04, 1.08,) to 1.17 (95% CI: 1.15, 1.20) in Chinese and American participants, respectively. When 85th and 95th percentiles were set as thresholds for each sex, weight/height3 showed similar accuracy to BMI percentiles, and were more precise than BMI z scores. Misclassification rates of weight/height3 ranged from 19.1% (95% CI: 18.8%, 19.5%) to 34.7% (95% CI: 34.0%, 35.4%) compared to BMI z score, which ranged from 26.3% (95% CI: 26.0%, 26.7%) to 36.8% (95% CI: 36.0%, 37.5%) in Chinese participants. Results were similar in American participants. Combined use of weight/height3 and waist-to-height ratio did not change the classification accuracy. CONCLUSIONS AND RELEVANCE: Tri-ponderal mass index (TMI) performed superior to BMI z scores and similar to BMI percentiles in Chinese and American participants. TMI is stable in adolescents, and could be a more efficient indicator for screening obesity-related cardio-metabolic risks in routine health screening compared with BMI.