BACKGROUND: We aimed to compare the accuracy of the ponderal index (PI) vs. BMI-for-age z-scores transformed (BMIz) in estimating body fat levels and classifying obesity in children and adolescents from a Brazilian urban population. METHODS: This is a cross-sectional study with 1149 participants (53.2% male), aged 6 to 18 years. Body fat percent (BFP) was obtained by multi-frequency bioelectrical impedance. Non-linear regression analysis provided the accuracy of both BMIz and PI in estimating BFP. False positive rate was obtained from the proportion of individuals placed at or above the 95th percentile for BMIz or PI, whereas their BFP was discordantly below the 95th percentile. RESULTS: PI and BMIz appeared with similar stability from childhood to adolescence for both boys and girls. The portion of the variability in BFP explained by BMIz (R2 = 0.74 and R2 = 0.75) was close to the variability in BFP explained by PI (R2 = 0.73 and R2 = 0.75) for boys and girls, respectively. False positive rate was higher for BMIz compared with PI among boys (21.8% vs. 3.9%) and girls (28.5% vs. 17.5%). CONCLUSIONS: PI is a promising index for replacing BMIz in children and adolescents due to its potential to reduce false diagnosis of obesity.
BACKGROUND: We aimed to compare the accuracy of the ponderal index (PI) vs. BMI-for-age z-scores transformed (BMIz) in estimating body fat levels and classifying obesity in children and adolescents from a Brazilian urban population. METHODS: This is a cross-sectional study with 1149 participants (53.2% male), aged 6 to 18 years. Body fat percent (BFP) was obtained by multi-frequency bioelectrical impedance. Non-linear regression analysis provided the accuracy of both BMIz and PI in estimating BFP. False positive rate was obtained from the proportion of individuals placed at or above the 95th percentile for BMIz or PI, whereas their BFP was discordantly below the 95th percentile. RESULTS: PI and BMIz appeared with similar stability from childhood to adolescence for both boys and girls. The portion of the variability in BFP explained by BMIz (R2 = 0.74 and R2 = 0.75) was close to the variability in BFP explained by PI (R2 = 0.73 and R2 = 0.75) for boys and girls, respectively. False positive rate was higher for BMIz compared with PI among boys (21.8% vs. 3.9%) and girls (28.5% vs. 17.5%). CONCLUSIONS: PI is a promising index for replacing BMIz in children and adolescents due to its potential to reduce false diagnosis of obesity.
Authors: Dusko Bjelica; Jovan Gardasevic; Zoran Milosevic; Predrag R Bozic; Bojan Masanovic Journal: Int J Environ Res Public Health Date: 2021-05-20 Impact factor: 3.390