Xijie Wang1, Bin Dong2, Jun Ma3, Yi Song1, Zhiyong Zou1, Luke Arnold4. 1. Institute of Child and Adolescent Health, School of Public Health, Peking University, Beijing, PR China. 2. Institute of Child and Adolescent Health, School of Public Health, Peking University, Beijing, PR China. bindong@bjmu.edu.cn. 3. Institute of Child and Adolescent Health, School of Public Health, Peking University, Beijing, PR China. majunt@bjmu.edu.cn. 4. Department of Commissioning, South Western Sydney Primary Health Network, Campbelltown, Australia.
Abstract
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.
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.
Authors: Rossana Gómez-Campos; Rubén Vidal-Espinoza; Anderson Marques de Moraes; Evandro Lázari; Cynthia Lee Andruske; Luis Castelli Correia de Campos; Luis Urzua-Alul; Wilbert Cossio-Bolaños; Marco A Cossio-Bolanõs Journal: Front Nutr Date: 2021-12-24