Anthropol. Anz. 72/3 (2015), pp. 257–262 J. Biol. Clinic. Anthropol. published online 3 February 2015; published in print September 2015
Article
Body mass index (BMI) predicts percent body fat better than body adiposity index (BAI) in school children Dapeng Zhao, and Yunzhao Zhang Tianjin Key Laboratory of Animal and Plant Resistance, College of Life Sciences, Tianjin Normal University, Tianjin, China [email protected] [email protected] With 1 figure and 1 table
Summary: Background/Aims: Child obesity is associated with increased risk of adult obesity, and is considered as one important health risk factor. Appropriate indicators are required to identify potential risks of child adiposity. This study for the first time compares body mass index (BMI) and body adiposity index (BAI) for predicting percent body fat (PBF) in children. Methods: We measured statures, weights, and hip circumferences of 527 children of Han ethnicity and calculated BMI and BAI. PBF was obtained by bioelectrical impedance analysis. We adopted Pearson correlation analysis, linear regression analysis, and receiver operating characteristic (ROC) analysis. Results: For each sex, we found that: 1) BMI and BAI were significantly correlated with PBF; 2) the correlation coefficient between BMI and PBF was higher than that between BAI and PBF; 3) BMI better predicted PBF in the linear regression analysis; 4) the discriminatory capacity of the BMI is higher than the one of BAI in ROC analysis. Conclusion: Taken together, BMI is a more reliable PBF indicator predicting adiposity in children. This finding may aid future obesity monitoring and intervention in children. Key words: adiposity indicator, children, fat mass, obesity monitoring.
Introduction Obesity becomes one global challenge of human biology and now is prevalent in both adults and non-adults. Child obesity is associated with increased risk of adult obesity, and is considered as one important health risk factor for many diseases (Lobstein et al. 2004, Lissau 2005, Brownell et al. 2009, Puder et al. 2011). Therefore, appropriate indicators are urgently required to identify potential risks of children adiposity (Taylor et al. 2008, Horan et al. 2014). Body mass index (BMI) is routinely used as a common indicator to quantify adiposity for all ages. However, BMI is not able to differentiate between fat and fat-free mass (Garn et al. 1986, Nevill et al. 2006). Body adiposity index (BAI) was recently proposed as an alternative to BMI and it is measured without weight data, which is more convenient in field research (Bergman et al. 2011).
쏘 2015 E. Schweizerbart’sche Verlagsbuchhandlung, Stuttgart, Germany DOI: 10.1127/anthranz/2015/0499 eschweizerbart_XXX
www.schweizerbart.de 0003-5548/15/0499 $ 1.50
258
Dapeng Zhao, and Yunzhao Zhang
To date some researchers found that BMI performed similarly or better than BAI for predicting PBF (e.g. Freedman et al. 2012, Zhao et al. 2013) whereas others hold converse findings (e.g. Johnson et al. 2012, Sun et al. 2013). However, it is still unknown whether BAI performs better than BMI in children. This study aims to for the first time compare the accuracy of BMI and BAI for predicting PBF in a sample of children population with the same ethnicity.
Methods Study population Totally we randomly investigated 527 school children of Han ethnicity (6–12 years old, 281 boys and 246 girls) recruited from Shaogongzhuang primary school in Tianjin city of China in autumn 2013. Tianjin city is one municipality in China. Han ethnicity originates from the Han dynasty and now is widespread across the world, but mainly in China. The people number of Han ethnicity is about 1.22 billion in 2010 based on the 6th national census of Chinese population (Zhao et al. 2012, Zhao et al. 2015). The research was approved by local institutional review board. Permission was obtained from participants before the onset of the study. Our study accords with the Helsinki Declaration.
Anthropometry We measured the stature to the nearest 0.01 meter with a Martin stadiometer and the weight to the nearest 0.1 kilogram with a digital scale for participants who have light indoor clothing without shoes and overcoats. BMI was calculated based on the formula: BMI = body weight (kg)/stature (m2) Hip circumference was measured with a tape measure to the nearest 0.001 meter and calculated the BAI following the formula: BAI = (hip circumference (cm)/stature (m)1.5) – 18 PBF was measured by tetrapolar bioelectrical impedance analysis instruments (model type: BC-601, Tanita, Japan). We defined PBF of 욷 25 % in boys and 욷 30 % in girls belonging to obese samples (Freedman & Sherry 2009).
Data analysis Sex differences were examined by t-tests for each measure. We adopt Pearson correlation coefficients (r) to test correlations among BMI, BAI and PBF. The linear regression was applied to assess the better predictor of PBF. The coefficient of determination (r2) and standard error of the estimate (SEE) were calculated. Receiver operating characteristic curves (ROC curves) were also used in the present study. Cutoff values showed graphically the tradeoff between sensitivity and specificity, and were calculated based on the point on the ROC curve with the lowest value for the formula: (1-sensitivity)2 + (1-specificity)2. The area under the ROC curve was used as one good measure to display the discriminatory capacity of one predictor. All analyses were conducted by SPSS 21.0 and the significance level was set as p 울 0.05.
Results Boys had significantly higher scores than girls on BMI and BAI (Table 1). BMI and BAI showed a significant correlation for each sex (boys: r = 0.416, p < 0.01; girls: r = 0.208, p < 0.01). The correlation coefficient between BMI and PBF (boys:
eschweizerbart_XXX
BMI and BAI prediction on adolescent adiposity
259
Table 1. Sex differences on anthropometric dimensions of school children. Boys (N = 281) Range
Girls (N = 246)
Mean ± SD
Range
Mean ± SD
Age (years) Weight (kg) Stature (m) Hip circumference (cm) BMI (kg/m2)
6–12 8.48 ± 1.88 18–72 34.92 ± 11.38 1.08–1.72 1.34 ± 0.13 63–105 78.44 ± 9.22
6–12 8.50 ± 1.83 18–79 33.95 ± 11.98 1.10–1.66 1.35 ± 0.13 60–108 77.26 ± 10.69
13.42–32.35 18.89 ± 3.43
11.96–30.67 18.23 ± 3.48
BAI (cm/m1.5)
23.59–46.34 32.53 ± 4.15
13.01–42.58 31.59 ± 4.54
PBF (%)
6.7–59.4
23.27 ± 10.50
6.3–51.0
23.54 ± 8.80
Sex difference p-value p = 0.341 p = 0.817 p = 0.176 p = 0.028 < 0.05 p = 0.013 < 0.05 p = 0.752
r = 0.957, p < 0.01; girls: r = 0.918, p < 0.01) was higher than that between BAI and PBF for both sexes (boys: r = 0.489, p < 0.01; girls: r = 0.204, p < 0.01). Correlation coefficients in BMI-PBF, BAI-PBF and BMI-BAI correlations were higher in boys than those in girls. For each sex, we found that there was a significant positive correlation between age and PBF (boys: r = 0.244, p < 0.01; girls: r = 0.382, p < 0.01) as well as between age and BMI (boys: r = 0.338, p < 0.01; girls: r = 0.397, p < 0.01). The linear regression analysis showed that BMI was the better predictor of PBF in both boys (BMI: r2 = 0.915, SEE = 3.06, p < 0.01; BAI: r2 = 0.239, SEE = 9.18, p < 0.01) and girls (BMI: r2 = 0.842, SEE = 3.50, p < 0.01; BAI: r2 = 0.042, SEE = 8.63, p < 0.01). The variation (SEE) around the regression lines for each sex was greater for BAI comparisons than for BMI comparisons with PBF. The area under the ROC for BMI was higher than that for BAI for each sex (Fig. 1). ROC curves showed that the cutoff point of BMI in boys is 19.28 (95 % CI: 96.8–99.3 %) and provided a sensitivity of 93.8 % and specificity of 94.1 %. In girls the cutoff point of BMI is 19.63 (95 % CI: 96.5–99.6 %) and provided a sensitivity of 94.0 % and specificity of 93.4 %. The cutoff point of BAI in boys is 33.65 (95 % CI: 67.7–80.3 %) and provided a sensitivity of 65.6 % and specificity of 75.7 %. In girls, cutoff point of BAI is 32.01 (95 % CI: 66.4–81.8 %) and provided a sensitivity of 78.0 % and specificity of 64.3 %.
Discussion In the present study, participants were chosen with the same ethnicity and analysis was conducted for boys and girls respectively in order to remove the potential impact of race and gender. Multiple comparative analyses consistently show that BMI is the better PBF predictor although BAI could also be effectively adopted to predict PBF in children. These findings are consistent with related findings in adult Asian populations (e.g. Zhao et al. 2013) and other populations (e.g. Geliebter et al. 2013). Furthermore, we found that significant positive correlations between age and PBF/BMI exist for each sex, which accord with previous findings (Roche, 1992, Maynard et al. 2001). Our results should be treated with caution due to limited sample size, which lower statistic power to some extent.
eschweizerbart_XXX
260
Dapeng Zhao, and Yunzhao Zhang
Fig. 1. ROC analysis for BMI/BAI: the ROC curve is represented by the solid line; the diagonal reference line is represented by the dotted line; the area under the ROC for BMI was higher than that for BAI for boys and girls respectively. a. boys: BMI and PBF; b. boys: BAI and PBF; c. girls: BMI and PBF; d. girls: BAI and PBF.
The diagnostic criteria on children obesity vary and become one hot topic of debate and research. Various indicators have been used in predicting adolescent obesity, such as BMI, BAIp, waist circumference, waist circumference-to-stature ratio (Taylor et al. 2008, Freedman & Sherry 2009, Wilson et al. 2011, El Aarbaoui et al. 2013). Each adiposity indicator has its own limitation on accuracy. For instance, it is found that BMI is a good indicator of excess adiposity among relatively fat children whereas differences in BMIs of relatively thin children can be largely due to fat-free mass (Freedman & Sherry 2009). Furthermore, various PBF technical tools have been adopted such as dual energy X-ray absorptiometry, bioelectrical impedance analysis, skinfold anthropometry (Ellis et al. 1999, Freedman et al. 2013, Horan et al. 2014). Differences technical methods influence the relationship between adiposity indicators and corresponding PBF (e.g. BMI and BAI: Geliebter et al. 2013). In addi-
eschweizerbart_XXX
BMI and BAI prediction on adolescent adiposity
261
tion, body composition and obesity levels vary among different ethnic groups (ElSayed et al. 2011). Therefore, one should consider ethnic-specific characteristics, advantage of adiposity indicators, and PBF technical factors simultaneously in future study.
Acknowledgements This work was funded by Talent Introduction Fund of Tianjin Normal University (No. 5RL115) and Undergraduate Innovation Training Program Fund of Tianjin Normal University (No. 201 3082). We appreciate Hong Wang and Xiaofeng Zi (Shaogongzhuang primary school, China) for their friendly help during data collection period. The authors have no conflict of interest.
References Bergman, R.N., Stefanovski, D., Buchanan, T.A., Sumner, A.E., Reynolds, J.C., Sebring, N.G., Xiang, A.H. & Watanabe, R.M. (2011): A better index of body adiposity. – Obesity 19: 1083–1089. Brownell, K.D., Schwartz, M.B., Puhl, R.M., Henderson, K.E. & Harris, J.L. (2009): The need for bold action to prevent adolescent obesity. – J. Adolesc. Health 45: S8–S17. El Aarbaoui, T., Samouda, H., Zitouni, D., di Pompeo, C., de Beaufort, C., Trincaretto, F., Mormentyn, A., Hubert, H., Lemdani, M. & Guinhouya, B.C. (2013): Does the body adiposity index (BAI) apply to paediatric populations? – Ann. Hum. Biol. 40: 451–458. Ellis, K.J., Abrams, S.A. & Wong, W.W. (1999): Monitoring childhood obesity: assessment of the weight/height2 index. – Am. J. Epidemiol. 150: 939–946. El-Sayed, A., Scarborough, P. & Galea, S. (2011): Ethnic inequalities in obesity among children and adults in the UK: a systematic review of the literature. – Obes. Rev. 12: e516– e534. Freedman, D.S. & Sherry, B. (2009): The validity of BMI as an indicator of body fatness and risk among children. – Pediatrics 124: S23–S34. Freedman, D.S., Thornton, J.C., Pi-Sunyer, F.X., Heymsfield, S.B., Wang, J., Pierson R.N., Blanck, H.M. & Gallagher, D. (2012): The body adiposity index (hip circumference÷ height1. 5) is not a more accurate measure of adiposity than is BMI, waist circumference, or hip circumference. – Obesity 20: 2438–2444. Freedman, D.S., Ogden, C.L., Blanck, H.M., Borrud, L.G. & Dietz, W. H. (2013): The abilities of body mass index and skinfold thicknesses to identify children with low or elevated levels of dual-energy X-Ray absorptiometry-determined body fatness. – J. Pediatr. 163: 160–166. Garn, S.M., Leonard, W.R. & Hawthorne, V.M. (1986): Three limitations of the body mass index. – Am. J. Clin. Nutr. 44: 996–997. Geliebter, A., Atalayer, D., Flancbaum, L. & Gibson, C.D. (2013): Comparison of body adiposity index (BAI) and BMI with estimations of % body fat in clinically severe obese women. – Obesity 21: 493–498. Horan, M., Gibney, E., Molloy, E. & McAuliffe, F. (2014): Methodologies to assess paediatric adiposity. – Irish. J. Med. Sci. in press. doi: 10.1007/s11 845-014-1124-1. Johnson, W., Chumlea, W.C., Czerwinski, S.A. & Demerath, E.W. (2012): Concordance of the recently published body adiposity index with measured body fat percent in EuropeanAmerican adults. – Obesity 20: 900–903. Lissau, I. (2005): Action, prevention and epidemiology of paediatric obesity. – Acta Paediatr. 94: 30–37 Lobstein, T., Baur, L. & Uauy, R. (2004): Obesity in children and young people: a crisis in public health. – Obes. Rev. 5: 4–85.
eschweizerbart_XXX
262
Dapeng Zhao, and Yunzhao Zhang
Maynard, L.M., Wisemandle, W., Roche, A.F., Chumlea, W.C., Guo, S. S. & Siervogel, R.M. (2001): Childhood body composition in relation to body mass index. – Pediatrics 107: 344–350. Nevill, A.M., Stewart, A.D., Olds, T. & Holder, R. (2006): Relationship between adiposity and body size reveals limitations of BMI. – Am. J. Phys. Anthropol. 129: 151–156. Puder, J.J., Schindler, C., Zahner, L. & Kriemler, S. (2011): Adiposity, fitness and metabolic risk in children: a cross-sectional and longitudinal study. – Int. J. Pediatr. Obes. 6: e297– e306. Roche, A.F. (1992): Growth, maturation, and body composition: the Fels longitudinal study 1929–1991. – Cambridge University Press, Cambridge. Sun, G., Cahill, F., Gulliver, W., Yi, Y., Xie, Y., Bridger, T., Pace, D. & Zhang, H. (2013): Concordance of BAI and BMI with DXA in the Newfoundland population. – Obesity 21: 499–503. Taylor, R.W., Williams, S.M., Grant, A.M., Ferguson, E., Taylor, B.J. & Goulding, A. (2008): Waist circumference as a measure of trunk fat mass in children aged 3 to 5 years. – Int. J. Pediatr. Obes. 3: 226–233. Wilson, H.J., Dickinson, F., Griffiths, P.L., Azcorra, H., Bogin, B. & Varela-Silva, M.I. (2011): How useful is BMI in predicting adiposity indicators in a sample of Maya children and women with high levels of stunting? – Am. J. Hum. Biol. 23: 780–789. Zhao, D.P., Li, B.G., Yu, K. & Zheng, L. (2012): Digit ratio (2D: 4D) and handgrip strength in subjects of Han ethnicity: impact of sex and age. – Am. J. Phys. Anthropol. 149: 266– 271. Zhao, D.P., Li, Y., Zheng, L. & Yu, K. (2013): Body mass index, body adiposity index, and percent body fat in Asians. – Am. J. Phys. Anthropol. 152: 294–299. Zhao, D.P., Zheng, L., Li, Y., Yu, K. & Zhang, X. (2015): Area-level socioeconomic disparities impact adult overweight and obesity risks of Han ethnicity in China. – Am. J. Hum. Biol. 27: 129–132. Submitted: 24 June 2014; accepted: 13 November 2014. Address for correspondence: Dapeng Zhao, Tanjin Key Laboratory of Animal and Plant Resistance, Tianjin Normal University, College of Life Sciences, Binshui West Road 393, Xiqing District, Tianjin 300 387, China [email protected] [email protected]
eschweizerbart_XXX
Article
Body mass index (BMI) predicts percent body fat better than body adiposity index (BAI) in school children Dapeng Zhao, and Yunzhao Zhang Tianjin Key Laboratory of Animal and Plant Resistance, College of Life Sciences, Tianjin Normal University, Tianjin, China [email protected] [email protected] With 1 figure and 1 table
Summary: Background/Aims: Child obesity is associated with increased risk of adult obesity, and is considered as one important health risk factor. Appropriate indicators are required to identify potential risks of child adiposity. This study for the first time compares body mass index (BMI) and body adiposity index (BAI) for predicting percent body fat (PBF) in children. Methods: We measured statures, weights, and hip circumferences of 527 children of Han ethnicity and calculated BMI and BAI. PBF was obtained by bioelectrical impedance analysis. We adopted Pearson correlation analysis, linear regression analysis, and receiver operating characteristic (ROC) analysis. Results: For each sex, we found that: 1) BMI and BAI were significantly correlated with PBF; 2) the correlation coefficient between BMI and PBF was higher than that between BAI and PBF; 3) BMI better predicted PBF in the linear regression analysis; 4) the discriminatory capacity of the BMI is higher than the one of BAI in ROC analysis. Conclusion: Taken together, BMI is a more reliable PBF indicator predicting adiposity in children. This finding may aid future obesity monitoring and intervention in children. Key words: adiposity indicator, children, fat mass, obesity monitoring.
Introduction Obesity becomes one global challenge of human biology and now is prevalent in both adults and non-adults. Child obesity is associated with increased risk of adult obesity, and is considered as one important health risk factor for many diseases (Lobstein et al. 2004, Lissau 2005, Brownell et al. 2009, Puder et al. 2011). Therefore, appropriate indicators are urgently required to identify potential risks of children adiposity (Taylor et al. 2008, Horan et al. 2014). Body mass index (BMI) is routinely used as a common indicator to quantify adiposity for all ages. However, BMI is not able to differentiate between fat and fat-free mass (Garn et al. 1986, Nevill et al. 2006). Body adiposity index (BAI) was recently proposed as an alternative to BMI and it is measured without weight data, which is more convenient in field research (Bergman et al. 2011).
쏘 2015 E. Schweizerbart’sche Verlagsbuchhandlung, Stuttgart, Germany DOI: 10.1127/anthranz/2015/0499 eschweizerbart_XXX
www.schweizerbart.de 0003-5548/15/0499 $ 1.50
258
Dapeng Zhao, and Yunzhao Zhang
To date some researchers found that BMI performed similarly or better than BAI for predicting PBF (e.g. Freedman et al. 2012, Zhao et al. 2013) whereas others hold converse findings (e.g. Johnson et al. 2012, Sun et al. 2013). However, it is still unknown whether BAI performs better than BMI in children. This study aims to for the first time compare the accuracy of BMI and BAI for predicting PBF in a sample of children population with the same ethnicity.
Methods Study population Totally we randomly investigated 527 school children of Han ethnicity (6–12 years old, 281 boys and 246 girls) recruited from Shaogongzhuang primary school in Tianjin city of China in autumn 2013. Tianjin city is one municipality in China. Han ethnicity originates from the Han dynasty and now is widespread across the world, but mainly in China. The people number of Han ethnicity is about 1.22 billion in 2010 based on the 6th national census of Chinese population (Zhao et al. 2012, Zhao et al. 2015). The research was approved by local institutional review board. Permission was obtained from participants before the onset of the study. Our study accords with the Helsinki Declaration.
Anthropometry We measured the stature to the nearest 0.01 meter with a Martin stadiometer and the weight to the nearest 0.1 kilogram with a digital scale for participants who have light indoor clothing without shoes and overcoats. BMI was calculated based on the formula: BMI = body weight (kg)/stature (m2) Hip circumference was measured with a tape measure to the nearest 0.001 meter and calculated the BAI following the formula: BAI = (hip circumference (cm)/stature (m)1.5) – 18 PBF was measured by tetrapolar bioelectrical impedance analysis instruments (model type: BC-601, Tanita, Japan). We defined PBF of 욷 25 % in boys and 욷 30 % in girls belonging to obese samples (Freedman & Sherry 2009).
Data analysis Sex differences were examined by t-tests for each measure. We adopt Pearson correlation coefficients (r) to test correlations among BMI, BAI and PBF. The linear regression was applied to assess the better predictor of PBF. The coefficient of determination (r2) and standard error of the estimate (SEE) were calculated. Receiver operating characteristic curves (ROC curves) were also used in the present study. Cutoff values showed graphically the tradeoff between sensitivity and specificity, and were calculated based on the point on the ROC curve with the lowest value for the formula: (1-sensitivity)2 + (1-specificity)2. The area under the ROC curve was used as one good measure to display the discriminatory capacity of one predictor. All analyses were conducted by SPSS 21.0 and the significance level was set as p 울 0.05.
Results Boys had significantly higher scores than girls on BMI and BAI (Table 1). BMI and BAI showed a significant correlation for each sex (boys: r = 0.416, p < 0.01; girls: r = 0.208, p < 0.01). The correlation coefficient between BMI and PBF (boys:
eschweizerbart_XXX
BMI and BAI prediction on adolescent adiposity
259
Table 1. Sex differences on anthropometric dimensions of school children. Boys (N = 281) Range
Girls (N = 246)
Mean ± SD
Range
Mean ± SD
Age (years) Weight (kg) Stature (m) Hip circumference (cm) BMI (kg/m2)
6–12 8.48 ± 1.88 18–72 34.92 ± 11.38 1.08–1.72 1.34 ± 0.13 63–105 78.44 ± 9.22
6–12 8.50 ± 1.83 18–79 33.95 ± 11.98 1.10–1.66 1.35 ± 0.13 60–108 77.26 ± 10.69
13.42–32.35 18.89 ± 3.43
11.96–30.67 18.23 ± 3.48
BAI (cm/m1.5)
23.59–46.34 32.53 ± 4.15
13.01–42.58 31.59 ± 4.54
PBF (%)
6.7–59.4
23.27 ± 10.50
6.3–51.0
23.54 ± 8.80
Sex difference p-value p = 0.341 p = 0.817 p = 0.176 p = 0.028 < 0.05 p = 0.013 < 0.05 p = 0.752
r = 0.957, p < 0.01; girls: r = 0.918, p < 0.01) was higher than that between BAI and PBF for both sexes (boys: r = 0.489, p < 0.01; girls: r = 0.204, p < 0.01). Correlation coefficients in BMI-PBF, BAI-PBF and BMI-BAI correlations were higher in boys than those in girls. For each sex, we found that there was a significant positive correlation between age and PBF (boys: r = 0.244, p < 0.01; girls: r = 0.382, p < 0.01) as well as between age and BMI (boys: r = 0.338, p < 0.01; girls: r = 0.397, p < 0.01). The linear regression analysis showed that BMI was the better predictor of PBF in both boys (BMI: r2 = 0.915, SEE = 3.06, p < 0.01; BAI: r2 = 0.239, SEE = 9.18, p < 0.01) and girls (BMI: r2 = 0.842, SEE = 3.50, p < 0.01; BAI: r2 = 0.042, SEE = 8.63, p < 0.01). The variation (SEE) around the regression lines for each sex was greater for BAI comparisons than for BMI comparisons with PBF. The area under the ROC for BMI was higher than that for BAI for each sex (Fig. 1). ROC curves showed that the cutoff point of BMI in boys is 19.28 (95 % CI: 96.8–99.3 %) and provided a sensitivity of 93.8 % and specificity of 94.1 %. In girls the cutoff point of BMI is 19.63 (95 % CI: 96.5–99.6 %) and provided a sensitivity of 94.0 % and specificity of 93.4 %. The cutoff point of BAI in boys is 33.65 (95 % CI: 67.7–80.3 %) and provided a sensitivity of 65.6 % and specificity of 75.7 %. In girls, cutoff point of BAI is 32.01 (95 % CI: 66.4–81.8 %) and provided a sensitivity of 78.0 % and specificity of 64.3 %.
Discussion In the present study, participants were chosen with the same ethnicity and analysis was conducted for boys and girls respectively in order to remove the potential impact of race and gender. Multiple comparative analyses consistently show that BMI is the better PBF predictor although BAI could also be effectively adopted to predict PBF in children. These findings are consistent with related findings in adult Asian populations (e.g. Zhao et al. 2013) and other populations (e.g. Geliebter et al. 2013). Furthermore, we found that significant positive correlations between age and PBF/BMI exist for each sex, which accord with previous findings (Roche, 1992, Maynard et al. 2001). Our results should be treated with caution due to limited sample size, which lower statistic power to some extent.
eschweizerbart_XXX
260
Dapeng Zhao, and Yunzhao Zhang
Fig. 1. ROC analysis for BMI/BAI: the ROC curve is represented by the solid line; the diagonal reference line is represented by the dotted line; the area under the ROC for BMI was higher than that for BAI for boys and girls respectively. a. boys: BMI and PBF; b. boys: BAI and PBF; c. girls: BMI and PBF; d. girls: BAI and PBF.
The diagnostic criteria on children obesity vary and become one hot topic of debate and research. Various indicators have been used in predicting adolescent obesity, such as BMI, BAIp, waist circumference, waist circumference-to-stature ratio (Taylor et al. 2008, Freedman & Sherry 2009, Wilson et al. 2011, El Aarbaoui et al. 2013). Each adiposity indicator has its own limitation on accuracy. For instance, it is found that BMI is a good indicator of excess adiposity among relatively fat children whereas differences in BMIs of relatively thin children can be largely due to fat-free mass (Freedman & Sherry 2009). Furthermore, various PBF technical tools have been adopted such as dual energy X-ray absorptiometry, bioelectrical impedance analysis, skinfold anthropometry (Ellis et al. 1999, Freedman et al. 2013, Horan et al. 2014). Differences technical methods influence the relationship between adiposity indicators and corresponding PBF (e.g. BMI and BAI: Geliebter et al. 2013). In addi-
eschweizerbart_XXX
BMI and BAI prediction on adolescent adiposity
261
tion, body composition and obesity levels vary among different ethnic groups (ElSayed et al. 2011). Therefore, one should consider ethnic-specific characteristics, advantage of adiposity indicators, and PBF technical factors simultaneously in future study.
Acknowledgements This work was funded by Talent Introduction Fund of Tianjin Normal University (No. 5RL115) and Undergraduate Innovation Training Program Fund of Tianjin Normal University (No. 201 3082). We appreciate Hong Wang and Xiaofeng Zi (Shaogongzhuang primary school, China) for their friendly help during data collection period. The authors have no conflict of interest.
References Bergman, R.N., Stefanovski, D., Buchanan, T.A., Sumner, A.E., Reynolds, J.C., Sebring, N.G., Xiang, A.H. & Watanabe, R.M. (2011): A better index of body adiposity. – Obesity 19: 1083–1089. Brownell, K.D., Schwartz, M.B., Puhl, R.M., Henderson, K.E. & Harris, J.L. (2009): The need for bold action to prevent adolescent obesity. – J. Adolesc. Health 45: S8–S17. El Aarbaoui, T., Samouda, H., Zitouni, D., di Pompeo, C., de Beaufort, C., Trincaretto, F., Mormentyn, A., Hubert, H., Lemdani, M. & Guinhouya, B.C. (2013): Does the body adiposity index (BAI) apply to paediatric populations? – Ann. Hum. Biol. 40: 451–458. Ellis, K.J., Abrams, S.A. & Wong, W.W. (1999): Monitoring childhood obesity: assessment of the weight/height2 index. – Am. J. Epidemiol. 150: 939–946. El-Sayed, A., Scarborough, P. & Galea, S. (2011): Ethnic inequalities in obesity among children and adults in the UK: a systematic review of the literature. – Obes. Rev. 12: e516– e534. Freedman, D.S. & Sherry, B. (2009): The validity of BMI as an indicator of body fatness and risk among children. – Pediatrics 124: S23–S34. Freedman, D.S., Thornton, J.C., Pi-Sunyer, F.X., Heymsfield, S.B., Wang, J., Pierson R.N., Blanck, H.M. & Gallagher, D. (2012): The body adiposity index (hip circumference÷ height1. 5) is not a more accurate measure of adiposity than is BMI, waist circumference, or hip circumference. – Obesity 20: 2438–2444. Freedman, D.S., Ogden, C.L., Blanck, H.M., Borrud, L.G. & Dietz, W. H. (2013): The abilities of body mass index and skinfold thicknesses to identify children with low or elevated levels of dual-energy X-Ray absorptiometry-determined body fatness. – J. Pediatr. 163: 160–166. Garn, S.M., Leonard, W.R. & Hawthorne, V.M. (1986): Three limitations of the body mass index. – Am. J. Clin. Nutr. 44: 996–997. Geliebter, A., Atalayer, D., Flancbaum, L. & Gibson, C.D. (2013): Comparison of body adiposity index (BAI) and BMI with estimations of % body fat in clinically severe obese women. – Obesity 21: 493–498. Horan, M., Gibney, E., Molloy, E. & McAuliffe, F. (2014): Methodologies to assess paediatric adiposity. – Irish. J. Med. Sci. in press. doi: 10.1007/s11 845-014-1124-1. Johnson, W., Chumlea, W.C., Czerwinski, S.A. & Demerath, E.W. (2012): Concordance of the recently published body adiposity index with measured body fat percent in EuropeanAmerican adults. – Obesity 20: 900–903. Lissau, I. (2005): Action, prevention and epidemiology of paediatric obesity. – Acta Paediatr. 94: 30–37 Lobstein, T., Baur, L. & Uauy, R. (2004): Obesity in children and young people: a crisis in public health. – Obes. Rev. 5: 4–85.
eschweizerbart_XXX
262
Dapeng Zhao, and Yunzhao Zhang
Maynard, L.M., Wisemandle, W., Roche, A.F., Chumlea, W.C., Guo, S. S. & Siervogel, R.M. (2001): Childhood body composition in relation to body mass index. – Pediatrics 107: 344–350. Nevill, A.M., Stewart, A.D., Olds, T. & Holder, R. (2006): Relationship between adiposity and body size reveals limitations of BMI. – Am. J. Phys. Anthropol. 129: 151–156. Puder, J.J., Schindler, C., Zahner, L. & Kriemler, S. (2011): Adiposity, fitness and metabolic risk in children: a cross-sectional and longitudinal study. – Int. J. Pediatr. Obes. 6: e297– e306. Roche, A.F. (1992): Growth, maturation, and body composition: the Fels longitudinal study 1929–1991. – Cambridge University Press, Cambridge. Sun, G., Cahill, F., Gulliver, W., Yi, Y., Xie, Y., Bridger, T., Pace, D. & Zhang, H. (2013): Concordance of BAI and BMI with DXA in the Newfoundland population. – Obesity 21: 499–503. Taylor, R.W., Williams, S.M., Grant, A.M., Ferguson, E., Taylor, B.J. & Goulding, A. (2008): Waist circumference as a measure of trunk fat mass in children aged 3 to 5 years. – Int. J. Pediatr. Obes. 3: 226–233. Wilson, H.J., Dickinson, F., Griffiths, P.L., Azcorra, H., Bogin, B. & Varela-Silva, M.I. (2011): How useful is BMI in predicting adiposity indicators in a sample of Maya children and women with high levels of stunting? – Am. J. Hum. Biol. 23: 780–789. Zhao, D.P., Li, B.G., Yu, K. & Zheng, L. (2012): Digit ratio (2D: 4D) and handgrip strength in subjects of Han ethnicity: impact of sex and age. – Am. J. Phys. Anthropol. 149: 266– 271. Zhao, D.P., Li, Y., Zheng, L. & Yu, K. (2013): Body mass index, body adiposity index, and percent body fat in Asians. – Am. J. Phys. Anthropol. 152: 294–299. Zhao, D.P., Zheng, L., Li, Y., Yu, K. & Zhang, X. (2015): Area-level socioeconomic disparities impact adult overweight and obesity risks of Han ethnicity in China. – Am. J. Hum. Biol. 27: 129–132. Submitted: 24 June 2014; accepted: 13 November 2014. Address for correspondence: Dapeng Zhao, Tanjin Key Laboratory of Animal and Plant Resistance, Tianjin Normal University, College of Life Sciences, Binshui West Road 393, Xiqing District, Tianjin 300 387, China [email protected] [email protected]
eschweizerbart_XXX
