Weight and body mass index (BMI)

http://informahealthcare.com/ahb ISSN: 0301-4460 (print), 1464-5033 (electronic) Ann Hum Biol, Early Online: 1–11 ! 2014 Informa UK Ltd. DOI: 10.3109/03014460.2014.907444

RESEARCH PAPER

Weight and body mass index (BMI): current data for Austrian boys and girls aged 4 to under 19 years Michael Mayer1,2, Andreas Gleiss3, Gabriele Haüsler4, Martin Borkenstein5, Klaus Kapelari6, Gerhard Ko¨stl7, Michael Lassi8, Michael Schemper3, Klaus Schmitt2, and Peter Blu¨mel9

Ann Hum Biol Downloaded from informahealthcare.com by 194.107.184.200 on 05/13/14 For personal use only.

1

Department of Paediatrics, General Hospital ‘Barmherzige Schwestern’, Ried im Innkreis, Austria, 2Department of Paediatrics, Children’s und Maternity Hospital, Linz, Austria, 3Center for Medical Statistics, Informatics and Intelligent Systems and 4Department of Paediatrics and Adolescent Medicine, Medical University of Vienna, Austria, 5Department of Paediatrics, Medical University of Graz, Austria, 6Department of Pediatrics, University Children’s Hospital, Medical University of Innsbruck, Austria, 7Department of Paediatrics, LKH, Leoben, Austria, 8Department of Paediatrics, Landesklinikum Mo¨dling, Austria, and 9Department of Paediatrics, Gottfied von Preyer’sches Kinderspital, Vienna, Austria Abstract

Keywords

Background: BMI reference charts are widely used to diagnose overweight, obesity and underweight in children and adolescents. Aim: To provide up-to-date national reference values for Austria. Methods: A cross-sectional sample of over 14 500 children and adolescents (4–19 years) stratified by provinces according to age- and sex-specific population proportions was drawn via schooling institutions (kindergartens, schools and vocational colleges). The generalized additive models for location, scale and shape were used for a flexible estimation of percentile curves. Results: Austrian boys and girls have higher average weight compared with previous prevalence data. BMI centiles matching BMI values at age 18 years, which are used for defining thinness, overweight and obesity in adults, were calculated. In Austria, using reference values as thresholds, 18% of boys and 12% of girls are overweight (with thresholds passing through BMI 25.00–29.99 kg/m2 in adults) and 5% of boys and 3% of girls are obese (with thresholds passing through BMI 30.00 kg/m2 in adults). Conclusion: Overweight and obesity are common in Austria and their prevalence is increasing (using the same IOTF reference for international comparison). Up-to-date national BMI reference values are provided to classify children and adolescents according to the proposed overweight and obesity thresholds.

Austria, BMI, GAMLSS, obesity, overweight, reference

Introduction What is already known on this topic Overweight and obesity is increasing in many countries worldwide. The WHO classifies adults and children as underweight, overweight and obese according to their BMI. The International Obesity Taskforce (IOTF) favours reference data where centiles pass through adult cut-points, which are linked with mortality rates. What this study adds Provides national BMI reference values based on a representative sample of more than 14 500 children and adolescents.

Correspondence: Michael Mayer, Department of Paediatrics, General hospital ‘Barmherzige Schwestern’, Schlossberg 1, 4910 Ried im Innkreis, Austria. Tel: +43-7752-602-84266. Fax: +437752+602+95161. E-mail: [email protected]

History Received 22 December 2013 Revised 27 February 2014 Accepted 5 March 2014 Published online 23 April 2014

Provides age- and sex-dependent thresholds calculated according to the IOTF recommendations as a continuum to those used in adults avoiding more arbitrary thresholds in childhood and previous changes in diagnosis, e.g. overweight/obesity at age 18 years. Application of our new thresholds shows that overweight and obesity is frequent in Austria, although their prevalence is much higher in some other countries. Reference data for weight measurements traditionally have a role in clinical practice for monitoring the somatic development of a child and in decision-making for diagnostic workup for overweight. Children’s average weight is increasing and the percentage of overweight and obese children has risen during the past decades in many countries (Kelishadi, 2007; Li et al., 2006; McCarthy et al., 2005; Moreno et al., 2001), including Austria (Freisling & Elmadfa, 2007), although there are hints that a plateau has been reached in some industrialized countries (Ku¨hnis & Erne, 2012; O’Dea et al., 2011; Ogden et al., 2012). The World Health Organization (WHO) classifies adults (WHO, 2000) and children (WHO, 2006) as underweight,


2

M. Mayer et al.

overweight and obese according to their body mass index (BMI). The BMI for age is significantly associated with cardiovascular risk factors in both adults and children, although, like skin-fold thickness, its association with adverse risk-factor levels is not as strong as that of waist:height ratios (Freedman et al., 2009). Overweight children [Centers for Disease Control and Prevention (CDC) 85–95th percentile] are more likely than thinner children to have adverse levels of cardiovascular disease risk factors and to become obese adults (Freedman et al., 2009). Furthermore, childhood obesity is associated with premature death (555 years of age) (Franks et al., 2010). The BMI is accepted as a valid indirect measure of adiposity in children because it is associated with body composition and risk factors and is based on widely available measurements (Rolland-Cachera, 2011). It has a curvilinear relation with the percentage of body fat among children and it is one easy to measure value (Federico et al., 2011). Obesity and overweight can be defined on the basis of a variety of reference values, resulting in different cut-off points in the lower and upper range. German reference values, which were created by pooling data from 1985–1999 and are widely used in Austria, recommend the 90th and 97th percentile of their study for defining overweight and obesity (KromeyerHauschild et al., 2001). The CDC reference values, which were created by pooling historical data [1977 NCHS Growth Charts including data of NHES II (1963–1965), aged 6–11 years, NHES III (1966–1970), aged 12–17 years, NHANES I (1971–1974), aged 1–19 years; NHANES II (1976–1980), aged 6 months–19 years; NHANES III (1988–1994), aged 2–72 months], recommend the 85th and 95th percentile as cut-off values for overweight and obesity (Kuczmarski et al., 2002). In adults, a BMI 518.50 kg/m2 defines underweight, of 25.00–29.99 kg/m2 overweight and of 30 kg/m2 obesity. Data from Cole et al. (2000, 2007) showed that BMI centiles matching these BMI values at the age of 18 years are useful for defining thinness, overweight and obesity in children. This approach is favoured by the International Obesity Taskforce (IOTF) and has been accepted for international comparisons (Cole et al., 2000, 2007). The centiles are less arbitrary than other cut-off values because they pass through adult cut-off points which are linked with mortality rates (RollandCachera, 2011). Former Austrian BMI reference values are available which were calculated using data of 5145 children and adolescents aged 5–18 years from the Austrian Study on Nutritional Status 1991–1993 (Zarfl & Elmadfa, 1995). It has been shown that cut-off values differ if using national Austrian (Freisling & Elmadfa, 2007), German (Kromeyer-Hauschild et al., 2001) or international reference values (Cole et al., 2000) for overweight and obesity classification in Austrian children and adolescents as described by Freisling & Elmadfa (2007). As former studies were based on a small sample size and because overweight and obesity prevalence is still rising in many countries worldwide, this re-evaluation of prevalence in Austrian children and adolescents had to be done on a representative bigger sample. Furthermore, no study has used state-of-the-art statistical methods to calculate Austrian national BMI reference values.

Ann Hum Biol, Early Online: 1–11

Methods Study design Sampling aimed for a representative data basis for reference curve estimation and was, therefore, done from the general population via schooling institutions. This approach was also expected to guarantee a high degree of participation. A sufficiently large and broad data basis was required to ensure the newly established curves for diagnostics in Austria would be applicable nationwide for the next decades. Therefore, a sample of at least 12 000 children and adolescents stratified by provinces according to age- and sexspecific population proportions (Statistik Austria, 2009) was targeted. This sample size was also considered to allow the application of flexible state-of-the-art statistical modelling techniques. Based on the sample size target, a total number of 117 kindergartens, primary, secondary and vocational schools were contacted by an external healthcare system consulting company after written consent had been obtained from the Austrian Federal Ministry of Education, Arts and Culture. Whole classes were sampled in each of these institutions which had two effects: The final sample size exceeded the one targeted and children outside of the planned age range, i.e. below 4 and above 19 years of age, were included in our sample. A total of 15 301children from 15 368 originally recruited between November 2009 and June 2011 were eligible to be measured. The remaining 67 (0.44%) could not be measured due to plaster casts, Downs syndrome or spasticity. In total, 14 544 (7415 boys and 7129 girls) were aged within the range of 4 to under 19 years, the remaining 757 children and adolescents were older or younger, but of those data from 442 children and adolescents whose age was only half a year outside the target age range showed sufficient density to be used in the estimation process (see the statistical analysis section). The data for height are complete and only one observation for all other measures is missing. Anthropometric measurements Trained members of staff measured the participants following standardized procedures; 96.7% of all measurements were taken by one and the same experienced endocrinologist. Height was measured while children were sitting and standing using a portable Harpenden Stadiometer (Holtain Ltd., Crymych, UK) with a reading accuracy of 0.1 cm. Weight was measured using an electronic medical weighing device (Seca 899, Hamburg, Germany) with a reading accuracy of 0.1 kg. An (age-dependent) estimated jeans’ weight was subtracted from the total weight of participants whose weight measurement included jeans. Further details of the measurement process are given elsewhere (Gleiss et al., 2013). Statistical analysis We used the generalized additive models for location, scale and shape (GAMLSS) as described by Rigby & Stasinopoulos (2004) to estimate age-dependent percentile curves. The corresponding R-package GAMLSS (Stasinopoulos & Rigby, 2007) was used for computation. Separate GAMLSS models were used for each sex and the outcome variables, weight

Weight and BMI in Austrian boys and girls

DOI: 10.3109/03014460.2014.907444


and BMI, were estimated as described before by Gleiss et al. (2013). The estimates of the four Box-Cox power exponential (BCPE) parameters, m, , and , are given for each sex and half-year steps of age (Table 1). The first parameter, m, is the mean parameter, while the remaining three represent measures of variability, skewness and kurtosis. A BCPE distribution with ¼ 1 and ¼ 2 gives a normal distribution. Values of 51 and 41 represent right- and left-skewed distributions, respectively. For an observed value X of weight or BMI, a standard deviation score (SDS) is calculated by inserting these parameters at a given child’s age into the following formula (Rigby & Stasinopoulos, 2004): 1 SDS ¼ 1 1 þ signðzÞG1=t jz=cj =2 2 where denotes the standard normal distribution function, G1/ the standard Gamma distribution function with paramqffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi eter 1/ and c ¼ 22= ð1= Þð ð3= ÞÞ1 , where is the Gamma function. Z is given as: 8 h i < 1 X 1 if 6¼ 0 Z¼ : 1 log X if ¼ 0 For further details on statistical analysis see Gleiss et al. (2013). BMI centiles passing through adult cut-off points have been defined by Cole et al. (2000, 2007) for LMS models as follows: The sex-specific centile is determined at the age of 18 years according to the estimated LMS model for a given BMI value, e.g. 25.00 kg/m2, and then the respective curve represented by this centile is suggested as a threshold across the whole age range. In this study the GAMLSS model was used instead of the LMS model. The BMI centiles passing through adult cut-off points were called equicurves, short for equivalent curves, as they represent the equivalent of the adult cut-off point.

Results Percentile values for BMI (P3, P50, P97) and equicurves for BMI (16.00, 17.00, 18.50, 25.00, 30.00 and 35.00 kg/m2) are given for girls and boys aged 4 to under 18 years (Table 1). We used the already mentioned WHO definitions of thinness as a BMI518.5 kg/m2, overweight as a BMI 25.0 kg/m2 and obesity as a BMI 30.0 kg/m2 in adolescents aged 18 years and older. Figure 1 shows overweight and obesity rates of the children and adolescents from this study using the different reference thresholds for definition of overweight and obesity described in other studies (Cole et al., 2000; de Onis et al., 2007; Ogden et al., 2012; Rosario et al., 2010; WHO, 2006; Zarfl & Elmadfa, 1995). Figure 2 compares our sample using IOTF reference values for international comparison published by Cole et al. (2000) with other samples from all over the world also using IOTF reference values (Bac et al., 2012; Christoforidis et al., 2011; Duran et al., 2013; Freisling & Elmadfa, 2007; Kuczmarski et al., 2002; Murer et al., 2014;

3

Rosario et al., 2010; Telford et al., 2008; Tuan & Nicklas, 2009). Note that the overweight and obesity rate is higher in our study population than the Austrian reference values from the 2000s as published by Freisling & Elmadfa (2007). Figures 3 and 4 give the new BMI equicurves for Austrian girls and boys. Figures 5 and 6 compare the new Austrian BMI equicurves with the WHO standards for children aged 0–5 years (WHO, 2006). WHO centiles matching our BMI equicurves at the age of 4 years are +1 SDS in boys and girls for our BMI 25 kg/m2 equicurve and +2 SDS for the BMI 30.0 kg/m2 equicurve. Figures 5 and 6 also show our equicurves compared with international IOTF reference values published by Cole et al. (2000, 2007). Weight percentiles for boys and girls are given in Table 2. These percentiles only draw a picture of the real weight distribution in Austrian children and adolescents and should not be used as reference values for defining overweight, obesity or thinness. As already shown, BMI curves for centiles matching BMI values at the age of 18 years are more useful for defining thinness (BMI518.50 kg/m2), overweight (BMI 25.00– 29.99 kg/m2) and obesity (BMI 30.00 kg/m2) (see Cole et al., 2000, 2007).

Discussion The new Austrian overweight and obesity reference values were calculated using state-of-the-art statistical methods, in contrast to other reference values where cut-off points were calculated using historical data (de Onis et al., 2007; Kromeyer-Hauschild et al., 2001; Kuczmarski et al., 2002). Furthermore, the data is up-to-date and was obtained from a cross-section of the Austrian population of children and adolescents, of whom more than 96% were measured by one and the same experienced endocrinologist travelling throughout the country. Thresholds were calculated as advocated by the IOTF, such that thresholds in childhood are a continuum to those used in adults. When using other cut-off points (e.g. those of Kromeyer-Hauschild et al. (2001), which are used in daily routine in Austria at the moment), a girl aged 17 years and 11 months with a BMI of 29.00 kg/m2 might, merely because of the different thresholds used, be classified as obese because she is above the threshold (e.g. BMI 27.76 kg/m2 in girls) before her 18th birthday but as overweight immediately after. The threshold for obesity in adults is defined as a BMI of 30.00 kg/m2. Accordingly, when using equicurves, as described in this study, there would be no change in diagnoses following the 18th birthday. The same reclassification in adulthood would be necessary if we used BMI P85, BMI P90, BMI P95 (data not shown) or BMI P97 (see Table 1) derived from our study populations for classification of obesity or overweight in Austrian boys and girls. In Austria, overweight and obesity often used to be interpreted by weight percentiles (P90 and P97) based on a historical Swiss study population (Prader et al., 1989) which had only a minor obesity problem and had not included overweight or obese children. If body weight were used to classify overweight or obesity, many overweight and obese children might be missed as the weight P97 is 82 kg in 16-year-old girls and more than 100 kg in 16-year-old boys. Therefore, the body weight centiles are not adequate for

P3

13.476 13.393 13.206 13.067 13.004 12.992 13.042 13.143 13.262 13.381 13.498 13.617 13.741 13.881 14.041 14.233 14.477 14.767 15.086 15.417 15.749 16.070 16.367 16.632 16.860 17.059 17.237 17.398 17.546

Age (years)

4.00 4.50 5.00 5.50 6.00 6.50 7.00 7.50 8.00 8.50 9.00 9.50 10.00 10.50 11.00 11.50 12.00 12.50 13.00 13.50 14.00 14.50 15.00 15.50 16.00 16.50 17.00 17.50 18.00

19.037 19.297 19.450 19.740 20.203 20.758 21.381 22.071 22.825 23.625 24.430 25.202 25.909 26.525 27.033 27.448 27.798 28.098 28.352 28.568 28.758 28.931 29.085 29.221 29.339 29.444 29.534 29.616 29.690

P97

0.084 0.089 0.095 0.101 0.108 0.115 0.122 0.129 0.136 0.143 0.150 0.155 0.160 0.163 0.165 0.165 0.164 0.161 0.158 0.154 0.150 0.146 0.142 0.139 0.136 0.134 0.132 0.130 0.128

1.649 1.694 1.735 1.774 1.829 1.916 2.031 2.157 2.273 2.364 2.428 2.465 2.479 2.474 2.450 2.411 2.362 2.308 2.254 2.203 2.154 2.106 2.060 2.018 1.982 1.952 1.926 1.904 1.882

2.654 2.511 2.379 2.257 2.141 2.031 1.929 1.835 1.751 1.676 1.612 1.560 1.520 1.493 1.479 1.477 1.486 1.502 1.526 1.557 1.593 1.634 1.677 1.722 1.767 1.811 1.853 1.894 1.933

BCPE parameters

12.635 12.528 12.322 12.156 12.065 12.033 12.073 12.163 12.266 12.361 12.448 12.533 12.622 12.728 12.854 13.014 13.226 13.486 13.777 14.085 14.397 14.699 14.981 15.234 15.453 15.645 15.819 15.977 16.123

16.00 P0.225 13.179 13.087 12.892 12.742 12.668 12.647 12.691 12.786 12.897 13.006 13.111 13.216 13.327 13.454 13.601 13.782 14.015 14.295 14.605 14.929 15.255 15.570 15.862 16.124 16.350 16.547 16.724 16.885 17.033

17.00 P1.35 14.009 13.946 13.778 13.662 13.625 13.636 13.707 13.826 13.968 14.113 14.259 14.408 14.561 14.727 14.910 15.123 15.385 15.691 16.022 16.364 16.703 17.029 17.330 17.597 17.827 18.026 18.202 18.363 18.500

18.50 P9.35 16.996 17.134 17.159 17.281 17.540 17.890 18.328 18.836 19.387 19.951 20.505 21.026 21.500 21.917 22.271 22.578 22.868 23.151 23.421 23.674 23.909 24.124 24.315 24.478 24.615 24.732 24.834 24.926 25.000

25.00 P85.52

BMI equicurves

19.171 19.439 19.599 19.900 20.376 20.943 21.576 22.273 23.038 23.850 24.670 25.457 26.179 26.808 27.327 27.750 28.106 28.408 28.663 28.879 29.067 29.238 29.393 29.528 29.647 29.753 29.844 29.928 30.000

30.00 P97.27 21.293 21.666 21.945 22.420 23.096 23.830 24.567 25.346 26.226 27.205 28.230 29.241 30.189 31.026 31.726 32.297 32.758 33.124 33.407 33.627 33.815 33.994 34.169 34.338 34.503 34.657 34.793 34.921 35.040

35.00 P99.27 13.802 13.629 13.525 13.447 13.411 13.420 13.469 13.531 13.602 13.705 13.831 13.964 14.094 14.222 14.357 14.504 14.677 14.882 15.124 15.403 15.710 16.029 16.345 16.648 16.938 17.211 17.470 17.718 17.958

P3 15.793 15.593 15.514 15.508 15.591 15.756 15.979 16.213 16.455 16.729 17.031 17.340 17.641 17.927 18.198 18.460 18.730 19.016 19.325 19.659 20.010 20.364 20.708 21.034 21.341 21.626 21.892 22.143 22.379

P50 ¼ m

Percentiles

18.999 18.950 19.135 19.522 20.153 21.012 21.994 22.918 23.733 24.484 25.150 25.703 26.149 26.504 26.791 27.064 27.365 27.712 28.113 28.567 29.058 29.562 30.052 30.508 30.917 31.279 31.594 31.859 32.078

P97 0.079 0.081 0.085 0.090 0.098 0.106 0.115 0.124 0.131 0.138 0.144 0.148 0.153 0.156 0.157 0.158 0.159 0.158 0.158 0.156 0.155 0.153 0.151 0.150 0.148 0.146 0.144 0.142 0.140

2.038 2.313 2.520 2.651 2.713 2.716 2.670 2.581 2.462 2.331 2.187 2.034 1.880 1.736 1.614 1.520 1.455 1.417 1.406 1.417 1.446 1.487 1.532 1.577 1.619 1.658 1.693 1.725 1.754

1.454 1.561 1.666 1.768 1.877 2.001 2.140 2.277 2.404 2.520 2.613 2.670 2.690 2.678 2.641 2.587 2.523 2.451 2.374 2.293 2.214 2.139 2.069 2.006 1.948 1.897 1.850 1.809 1.772

BCPE parameters

12.619 12.553 12.514 12.471 12.449 12.462 12.509 12.564 12.622 12.705 12.801 12.892 12.967 13.034 13.102 13.184 13.293 13.435 13.616 13.835 14.086 14.353 14.620 14.877 15.123 15.356 15.578 15.792 16.001

16.00 P0.18

Boys

13.249 13.121 13.043 12.978 12.945 12.952 12.997 13.052 13.115 13.206 13.316 13.426 13.529 13.627 13.729 13.845 13.988 14.163 14.376 14.627 14.908 15.203 15.497 15.779 16.049 16.304 16.546 16.780 17.006

17.00 P0.824

14.099 13.906 13.792 13.710 13.676 13.691 13.746 13.814 13.894 14.006 14.143 14.290 14.436 14.583 14.736 14.901 15.091 15.311 15.568 15.861 16.181 16.512 16.839 17.153 17.452 17.734 18.001 18.256 18.500

18.50 P5.45

16.704 16.555 16.553 16.653 16.880 17.228 17.656 18.091 18.519 18.969 19.429 19.870 20.274 20.634 20.949 21.235 21.518 21.811 22.121 22.451 22.795 23.140 23.472 23.783 24.070 24.333 24.575 24.797 25.000

25.00 P77.2

BMI equicurves

18.361 18.283 18.411 18.709 19.214 19.913 20.725 21.506 22.221 22.906 23.542 24.096 24.562 24.945 25.259 25.545 25.844 26.173 26.539 26.943 27.373 27.809 28.230 28.621 28.972 29.285 29.559 29.796 30.000

30.00 P94.85

19.834 19.830 20.100 20.620 21.443 22.546 23.784 24.912 25.857 26.672 27.341 27.848 28.227 28.511 28.741 28.982 29.280 29.654 30.109 30.642 31.234 31.852 32.462 33.035 33.551 34.006 34.399 34.723 34.983

35.00 P98.46

M. Mayer et al.

15.418 15.441 15.354 15.336 15.419 15.565 15.780 16.052 16.352 16.659 16.963 17.255 17.533 17.799 18.053 18.310 18.594 18.905 19.227 19.546 19.856 20.148 20.413 20.645 20.841 21.009 21.157 21.291 21.411

P50 ¼ m

Percentiles

Girls

Table 1. BMI for girls and boys (4–18 years)—BCPE parameters (m, , , ), percentiles (P3, P50 and P97) and BMI equicurves corresponding to BMI 16.00 kg/m2, 17.00 kg/m2, 18.5 kg/m2, 25.00 kg/m2, 30.00 kg/m2 and 35.00 kg/m2 at the age of 18 years; equicurve’s percentile value as indicated for each BMI equicurves.


4 Ann Hum Biol, Early Online: 1–11


5


DOI: 10.3109/03014460.2014.907444

Figure 1. Overweight and obesity rates of Austrian boys and girls (this study population, aged 5–18 years) when using different reference values. We use equicurves for BMI 25.00 and BMI 30.00 at the age of 18 years to define overweight and obesity; Zarfl & Elmadfa (1995) used the 85th and 97th percentiles of their study as cut-off points for overweight and obesity; Rosario et al. (2010) used their BMI 25 and BMI 30 equicurves corresponding to BMI 25 and BMI 30 at the age of 18 years; Kromeyer-Hauschild et al. (2001) used their 90th and 97th percentile as cut-off points; CDC percentiles by Kuczmarski et al. (2002) used their 85th and 95th percentiles as cut-off points; WHO (2006) and de Onis et al. (2007) used their +1 SDS and +2 SDS percentiles; IOTF reference values by Cole et al. (2000) used BMI 25 and BMI 30 equicurves of their survey. The overweight rate (white bar) is always without the obesity rate (black bar). (a) Boys, (b) Girls.

weight classification, but only describe the real distribution of weight in Austrian boys and girls. They could be important for interpreting children’s weight gain or loss if crossing percentiles, but should not be used for interpreting overweight, obesity or thinness. The WHO recommends using the WHO BMI standards (WHO, 2006) for children up to the age of 5 years. We recommend the new national reference for Austrian boys and girls from the age of 4 years because they match WHO BMI standards quite well.

Our population showed higher BMI thresholds for children aged 7–12 years, but a less pronounced increase thereafter. The Austrian overweight and obesity rates are very similar to those measured in other European countries for girls, whereas the rates vary from country to country for boys (Figure 2). The Childhood Obesity Surveillance Initiative (COSI) found that, across Europe, an average of 24% of children aged 6–9 years are overweight or obese (Wijnhoven et al., 2013). In our cohort of 6–9 year olds, 26–29% were overweight or obese as defined by the WHO and 20–21% as

M. Mayer et al.



6

Figure 2. Overweight and obesity rates of boys and girls in the literature (all using the same IOTF reference for international comparison by Cole et al. (2000)): Freisling & Elmadfa (2007), Rosario et al. (2010), Murer et al. (2014), Bac et al. (2012), Christoforidis et al. (2011), Kuczmarski et al. (2002), Duran et al. (2013), Telford et al. (2008), Tuan & Nicklas (2009). Age ranges are different depending on the original study population (a ¼ 4–18 years; b ¼ 3–15 years; c ¼ 0–17 years; d ¼ 6–12 years; e ¼ 0–20 years; f ¼ 0–19 years, g ¼ 9–12 years; h ¼ 2–18 years). Rates for China, Indonesia and Vietnam include the obesity rate in the overweight rate, in all others the obesity rate is not included in the overweight rate. he white bar is the overweight rate (BMI 25.00–29.99 kg/m2 equicurve in Cole et al. (2000)), the grey bar is the overweight rate including the obesity rate (25.00 kg/m2 equicurve in Cole et al. (2000)), the black bar is the obesity rate (BMI 30.00 kg/m2 equicurve in Cole et al. (2000)). (a) Boys’ rates, (b) Girls’ rates.

defined by the IOTF criteria. Our study population is, therefore, within the range for prevalence of overweight and obesity observed in Europe. The prevalence of overweight including obesity in 9–12 year olds as defined by the WHO or IOTF reaches a peak of up to 37%, depending on sex and the reference used (Table 3), thereafter the prevalence declines in boys and girls. Therefore, children in Austria seem to show the greatest weight gain between the ages of 6–12 years. The less pronounced increase thereafter might on the other hand reflect that the overweight and obesity epidemic has not reached children aged more than

12 years that much and there might be an increase in this age range in the next years. Using the IOTF criteria to compare our overweight and obesity results from all age groups with results from our neighbouring countries (Ku¨hnis & Erne, 2012; Murer et al., 2014; Rosario et al., 2010), the prevalence of obesity in Austria is within the range of that in other central European countries (Figure 2). Table 4 shows rates and absolute numbers of misclassification if IOTF references for overweight or obesity by Cole et al. (2000) were applied to the Austrian population instead


DOI: 10.3109/03014460.2014.907444


7

Figure 3. BMI equicurves for Austrian girls corresponding to BMI cut-off points for adults at the age of 18 years. Equicurve means the centile passing through the corresponding adult WHO cut-off point for thinness/overweight/obesity (i.e. BMI 30.00 kg/m2 for obesity at 18 years means P97.27; see Table 1). After the age of 18 years the WHO classification of body composition is used instead of the centile (WHO, 2000).

Figure 4. BMI equicurves for Austrian boys corresponding to BMI cut-off points for adults at the age of 18 years. Equicurve means the centile passing through the corresponding adult WHO cut-off point for thinness/overweight/obesity (i.e. BMI 30.00 kg/m2 for obesity at 18 years means P94.85; see Table 1). After the age of 18 years the WHO classification of body composition is used instead of the centile (WHO, 2000).


8

M. Mayer et al.


Figure 5. Austrian BMI percentiles for girls (P3, P50, P97) in light dotted lines, Austrian BMI equicurves for girls (full black lines) aged 4–19 years passing through the corresponding adult WHO cut-off point for thinness, overweight or obesity at the age of 18 years compared with WHO growth standards for girls aged 0–5 years (+3 SDS; +2 SDS; +1 SDS; 0 SDS; 1 SDS; 2 SDS; 3 SDS) in black dashed lines (WHO, 2006) and the IOTF reference by Cole et al. (2000, 2007) for thinness, overweight and obesity in grey dashed lines for girls aged 2–18 years.

Figure 6. Austrian BMI percentiles for boys (P3, P50, P97) in light dotted lines, Austrian BMI equicurves for boys (full black lines) aged 4–19 years passing through the corresponding adult WHO cut-off point for thinness, overweight or obesity at the age of 18 years compared with WHO growth standards for girls aged 0–5 years (+3 SDS; +2 SDS; +1 SDS; 0 SDS; 1 SDS; 2 SDS; 3 SDS) in black dashed lines (WHO, 2006) and the IOTF reference by Cole et al. (2000, 2007) for thinness, overweight and obesity in grey dashed lines for boys aged 2–18 years.


DOI: 10.3109/03014460.2014.907444

9

Table 2. Weight in Austrian girls and boys aged 4–19 years (percentiles). These percentiles only draw a picture of the real weight distribution in Austrian children and adolescents should not be used as reference values for defining overweight or obesity or thinness. Girls


Age (years) 4.00 4.50 5.00 5.50 6.00 6.50 7.00 7.50 8.00 8.50 9.00 9.50 10.00 10.50 11.00 11.50 12.00 12.50 13.00 13.50 14.00 14.50 15.00 15.50 16.00 16.50 17.00 17.50 18.00 18.50 19.00

Boys

P3

P10

P25

P50

P75

P90

P97

P3

P10

P25

P50

P75

P90

P97

13.24 14.20 14.68 15.25 16.05 16.86 17.81 18.96 20.15 21.28 22.38 23.51 24.71 26.00 27.40 28.96 30.80 32.88 35.06 37.19 39.16 40.88 42.32 43.49 44.39 45.09 45.66 46.13 46.52 46.83 47.11

14.15 15.20 15.76 16.45 17.38 18.28 19.32 20.54 21.81 23.06 24.30 25.61 27.02 28.57 30.24 32.06 34.11 36.36 38.64 40.80 42.75 44.43 45.81 46.93 47.78 48.45 48.99 49.45 49.83 50.14 50.40

15.16 16.32 16.98 17.80 18.90 19.96 21.13 22.49 23.91 25.34 26.79 28.35 30.05 31.90 33.87 35.96 38.23 40.62 42.96 45.11 47.00 48.60 49.91 50.95 51.75 52.38 52.90 53.35 53.72 54.01 54.27

16.41 17.70 18.47 19.46 20.83 22.16 23.61 25.24 26.97 28.74 30.57 32.52 34.64 36.91 39.24 41.61 44.06 46.52 48.83 50.88 52.63 54.10 55.29 56.22 56.95 57.53 58.03 58.47 58.83 59.11 59.36

17.91 19.36 20.26 21.49 23.21 24.94 26.77 28.79 30.95 33.23 35.59 38.05 40.69 43.46 46.21 48.88 51.49 53.97 56.19 58.07 59.65 60.95 62.01 62.83 63.49 64.04 64.52 64.96 65.32 65.59 65.82

19.67 21.30 22.39 23.93 26.10 28.28 30.49 32.86 35.44 38.24 41.14 44.14 47.32 50.61 53.82 56.87 59.74 62.31 64.48 66.25 67.70 68.91 69.90 70.69 71.32 71.88 72.37 72.82 73.19 73.46 73.69

21.97 23.84 25.20 27.22 30.02 32.79 35.40 38.09 41.10 44.47 47.99 51.57 55.34 59.21 62.99 66.59 69.87 72.64 74.83 76.55 77.96 79.20 80.28 81.16 81.90 82.56 83.12 83.63 84.03 84.34 84.59

13.93 14.56 15.27 15.95 16.75 17.70 18.75 19.73 20.62 21.61 22.80 24.05 25.32 26.66 28.02 29.22 30.45 31.89 33.63 35.82 38.49 41.46 44.32 46.81 48.89 50.52 51.95 53.37 54.67 55.74 56.76

14.79 15.50 16.29 17.08 17.99 19.07 20.26 21.37 22.37 23.48 24.82 26.24 27.67 29.19 30.73 32.13 33.60 35.33 37.40 39.93 42.93 46.17 49.24 51.87 54.02 55.70 57.13 58.49 59.65 60.54 61.40

15.77 16.57 17.47 18.38 19.44 20.71 22.10 23.39 24.56 25.86 27.41 29.05 30.71 32.46 34.21 35.83 37.56 39.61 42.03 44.92 48.21 51.67 54.88 57.58 59.75 61.46 62.90 64.18 65.19 65.91 66.60

17.01 17.94 18.98 20.06 21.36 22.92 24.64 26.24 27.71 29.34 31.26 33.27 35.30 37.36 39.37 41.24 43.29 45.70 48.51 51.73 55.24 58.79 61.99 64.65 66.74 68.42 69.86 71.08 71.93 72.48 72.99

18.52 19.60 20.84 22.15 23.76 25.73 27.92 29.96 31.85 33.94 36.36 38.87 41.37 43.82 46.10 48.24 50.64 53.51 56.77 60.35 64.07 67.64 70.75 73.28 75.26 76.91 78.36 79.50 80.17 80.49 80.79

20.24 21.52 23.00 24.60 26.59 29.05 31.79 34.34 36.67 39.25 42.20 45.21 48.16 50.98 53.51 55.94 58.77 62.23 66.14 70.28 74.35 78.06 81.20 83.70 85.63 87.35 88.88 89.91 90.28 90.25 90.20

22.44 24.00 25.80 27.80 30.32 33.47 36.96 40.15 43.02 46.14 49.67 53.20 56.62 59.79 62.53 65.30 68.72 73.00 77.87 82.90 87.61 91.66 94.98 97.60 99.61 101.55 103.32 104.22 104.09 103.46 102.85

Table 3. Prevalence of overweight and obesity in our cohort defined by WHO and IOTF for different age groups. Boys Age (years) WHO 4–6 years 6–9 years 9–12 years 12–15 years 15–18 years IOTF 4–6 years 6–9 years 9–12 years 12–15 years 15–18 years

Girls

n

ow

ob

ow+ob

n

ow

ob

ow+ob

775 1397 1367 1603 1950

10.1% 11.0% 13.3% 14.5% 12.2%

10.2% 18.3% 23.9% 15.4% 13.2%

20.3% 29.3% 37.2% 29.9% 25.4%

787 1337 1345 1544 1846

10.5% 12.5% 13.8% 12.9% 9.6%

7.9% 13.2% 16.7% 9.2% 6.7%

18.4% 25.7% 30.5% 22.1% 16.3%

775 1397 1367 1603 1950

8.8% 13.0% 21.1% 19.1% 17.0%

2.7% 7.2% 6.4% 5.0% 6.9%

11.5% 20.3% 27.4% 24.1% 23.8%

787 1337 1345 1544 1846

11.9% 15.1% 19.3% 14.6% 12.2%

3.2% 6.1% 5.1% 3.6% 3.4%

15.1% 21.2% 24.3% 18.3% 15.7%

n, numbers; ow, overweight rate; ob, obesity rate; ow + ob, overweight including obesity rate.

of the reference values proposed in this study. In relation to the total number of children and adolescents living in Austria in 2012 (Statistik Austria, 2012) the calculated rate of false positives is 2.82%. This means that 36 335 children and adolescents would be classified as overweight or obese when using the IOTF reference values, while they would be classified as normal weight based on our national reference values. On the other hand, 16 969 children and adolescents in Austria (i.e. 1.32% of all Austrian boys and girls or 7.04% of all Austrian overweight or obese) would be classified as

normal weight using IOTF reference values (false negatives). These children and adolescents would be lost for adequate risk evaluation and treatment or prevention. The total direct medical expenditures attributed to the false positives account for $170 million in Austrian children based on an estimated $4685 per case for US children (Trasande, 2010). Only half of this amount is balanced by the total costs for risk evaluation and treatment of the false negatives. When comparing different countries, the possibility that different socioeconomic statuses might influence the results needs to be taken into account. Unfortunately, we were not able to obtain information on socioeconomic status in our study population. As sampling was done from the general population via schooling institutions, where a representative sample could be drawn and a high degree of participation was reached, we believe that the children measured represent the average mixture of the socioeconomic statuses in Austria, which is very similar to that in Germany. Account also has to be taken of the different years in which the prevalence data of other countries were collected, as the trend over time might influence the results. The differences between studies in age groups analysed is another important factor because they might hinder interpretation of the prevalence in all age groups. Our cross-sectional study found that the prevalence of overweight and obese girls is increasing in Austria in different age groups, with an overweight prevalence of 14.7% and obesity prevalence of 4.3% when using IOTF reference data

10

M. Mayer et al.


Table 4. Relative and absolute total numbers of children and adolescents with overweight or obesity in Austria defined by our new reference values and IOTF reference values by Cole et al. (2000) for different age groups based on the average population by age and sex in the year 2012 for Austria (Statistik Austria, 2012). This study


Age (years) Boys 4–6 years 6–9 years 9–12 years 12–15 years 15–18 years Girls 4–6 years 6–9 years 9–12 years 12–15years 15–18years

IOTF

n (AT)

ow + ob ow + ob (%) (#)

ow + ob (%)

ow + ob (n)

1 22 409 1 25 472 1 26 889 1 36 396 1 48 182

22.80% 22.80% 22.80% 22.80% 22.80%

27 909 28 608 28 931 31 098 33 785

11.50% 20.30% 27.40% 24.10% 23.80%

14 077 13 832 25 471 3137 34 768 5837 32 871 1773 35 267 1482

1 17 046 1 19 200 1 20 461 1 29 923 1 40 565

14.48% 14.48% 14.48% 14.48% 14.48%

16 948 17 260 17 443 18 813 20 354

15.10% 21.20% 24.30% 18.30% 15.70% MC 1 total MC 2 total

17 674 25 270 29 272 23 776 22 069 2.82% 1.32%

MC

726 8010 11 829 4963 1715 36 335 16 969

n (AT), number of all children in Austria in the year 2012 for each age group; ow + ob, overweight including obesity (relative rate (%) and absolute numbers (#)). Misclassification (MC): positive numbers means MC 1, negative ones means MC 2. Misclassification 1 (false positive) means that these children would be classified as overweight or obese on the basis of IOTF, but not on our national reference values. Misclassification 2 (false negative) means that these children would be classified as normal on the basis of IOTF but as overweight or obese if our national reference values were used.

for international comparison, compared with a prevalence of between 10.0% and 13.7% for overweight and between 0.2% and 3.1% for obesity in Austrian studies conducted in the early 2000s (Freisling & Elmadfa, 2007) (Figure 2). It was almost the same in the different age groups of boys, for whom overweight prevalence has risen from 12.4% and 14.5% in the early 2000s to 16.6% and the obesity prevalence from 3.0% and 5.1% to 6.0%. The overweight and obesity rates in boys and girls vary considerably depending on the definitions used in studies. The obesity rate in our study was lowest in both boys and girls when using our definition, but the overweight rate was in the upper range of the studies. If our study population were classified using recent German reference values by Rosario et al. (2010), the overweight and obesity rates would be higher than our results using the new Austrian thresholds, but if the reference values that have been used until now by KromeyerHauschild et al. (2001) were used the overweight rate would be lower but the obesity rate would be higher. Therefore, up-to-date national reference values are important to define overweight and obesity, but also thinness. Compared with CDC reference values (pooled data from 1963–1980 in Cole et al., 2000), the prevalence of overweight and obesity is higher in Austria nowadays than it was in the US up until the 1980s. The current prevalence of overweight and obesity in Austrian boys and girls is, however, much lower than it is now in the US (Duran et al., 2013). The prevalence in rural Texas is 50% higher in boys and 90% higher in girls than in our study population when using IOTF data as the common cut-off point for overweight and obesity.

Overweight and obesity rates were also calculated in our study population using the BMI data of this study and different thresholds (Kromeyer-Hauschild et al., 2001; Rosario et al., 2010; Zarfl & Elmadfa, 1995), CDC-percentiles by Kuczmarski et al. (2002), WHO-percentiles (de Onis et al., 2007; WHO, 2006), IOTF reference values as defined by Cole et al. (2000, 2007) (Figure 1). As an example, when our data was analysed using former Austrian reference values (Zarfl & Elmadfa, 1995), the prevalence of obesity was lower in boys and girls, but the prevalence of overweight was higher in boys but about the same in girls than when using our new thresholds. Therefore, it is important to know which reference values are used for interpretation of a study population. One accepted reference method (i.e. IOTF reference values by Cole (2000, 2007)) should be used for comparisons between different countries. Kromeyer-Hauschild et al.’s (2001) reference values, which are used routinely in Austria, under-estimate the overweight rate and over-estimate the obesity rate compared with our new thresholds in both boys and girls, but more pronouncedly in boys (Figure 1). Among the strengths of this study, we want to emphasize that nearly all children (496%) were measured by one and the same experienced person and that there are practically no missing values. No less important is the large number of children and adolescents included in the study (14 544), with an additional 442 participants only half a year outside of the age range investigated. Data from these additional participants ensured that the representativeness was not lost in the lower and upper age groups and could be used for the estimation process. Another strength of this study is that few children were excluded from measurements (only 67 out of 15 368) and, therefore, the sample investigated is highly representative for the Austrian boys and girls. The present study used a sampling scheme via schooling institutions in contrast to Rosario et al.’s (2010) study, where sampling was based on a stratified random selection of communities from which children were contacted at random for participation. The advantage of our approach is the high response rate from all eligible children and adolescents invited to participate in the study. The disadvantage is that we did not contact children and adolescents outside the regular schooling system. The proportion of school children our sampling system failed to capture was negligible because only 1.2% of pupils in Austria attended schools for the handicapped in 2011 and special school types like private schools account for only 2.7% of school-aged children, as reported by the OECD (2011). Unfortunately, although socioeconomic factors are known to influence the overweight and obesity rate, we were not allowed to ask for socioeconomic data in this investigation.

Conclusion As in many other countries in Europe and worldwide, overweight and obesity are common in Austrian children and adolescents, although the prevalence rates are higher in some other countries than in the Austrian population when IOTF references are used for common cut-off points by Cole et al. (2000). The prevalence of overweight and obesity is increasing in Austria compared with the findings of a study conducted in the early 2000s by Freisling & Elmadfa (2007), which used the

DOI: 10.3109/03014460.2014.907444

same IOTF common cut-off points. Up-to-date national BMI reference values for children aged 4 to under 19 years are now available to classify these children and adolescents according to BMI thresholds in adults (BMI ¼ 16.00; 18.50; 25.00; 30.00 kg/m2). The WHO growth standards (WHO, 2006: +2 SDS, +1 SDS, 1 SDS and 2 SDS) are recommended for children aged 4 years and younger.


Acknowledgements All authors planned the study and decided on the measurements to be done, in addition GH corresponded with the ministry, ML and MM did the measurements, AG and MS did the statistical analysis. The manuscript was written by MM, AG and PB and all authors reviewed the manuscript for scientific content and approved the final submitted and published versions. The authors thank Daniel Weghuber, Salzburg, Austria and Maria Fritsch, Vienna, Austria for scientific review and support. The authors thank Elise Langdon-Neuner for language editing of the manuscript. The sampling of the data was supported by grants of Merck-Serono Austria, data belong to the Austrian working group on ¨ ), part of the Austrian pediatric endocrinology and diabetes (APEDO ¨ GKJ). pediatric society (O

Declaration of interest The sampling of the data by ML was supported by grants from MerckSerono Austria, AG and MS received grants for the statistical analysis from Merck-Serono Austria. The authors report no further conflicts of interest.

References Bac A, Wozniacka R, Matusik S. 2012. Prevalence of overweight and obesity in children aged 6–13 – alarming increase in obesity in Cracow, Poland. Eur J Pediatr 171:245–251. Christoforidis A, Dimitriadou M, Papadopolou E, Stilpnopoulou D, Katzos G, Athanassiou-Metaxa M. 2011. Defining overweight and obesitiy among Greek children living in Thessaloniki: International versus local reference standards. Hippokratia 15:141–146. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. 2000. Establishing a standard definition for child overweight and obesity worldwide: International survey. BMJ 320:1240–1243. Cole TJ, Flegal KM, Nicholls D, Jackson AA. 2007. Body mass index cut offs to define thinness in children and adolescents: International survey. BMJ 335:194. de Onis M, Onyango AW, Borghi E, Siyam A, Nishida C, Siekmann J. 2007. Development of a WHO growth reference for school-aged children and adolescents. Bull WHO 85:660–667. Duran M, Gillespie J, Malina RM, Little BB. 2013. Growth and weight status of rural Texas school youth. Am J Hum Biol 25:71–77. Federico B, D’Aliesio F, Pane F, Capelli G, Rodio A. 2011. Body mass index has curvilinear relationship with the percentage of body fat among children. BMC Res Notes 4:301. Franks PW, Hanson RL, Knowler WC, Sievers ML, Bennett PH, Looker HC. 2010. Childhood obesity, other cardiovascular risk factors, and premature death. NEJM 362:485–493. Freedman DS, Dietz WH, Srinivasan SR, Berenson GS. 2009. Risk factors and adult body mass index among overweight children: The Bogalusa Heart Study. Pediatrics 123:750–757. Freisling H, Elmadfa I. 2007. Prevalence of overweight and obesity among Austrian children (3–15y): using three different sets of reference values for body mass index (BMI). Nutrition 31:6–11. Gleiss A, Lassi M, Blu¨mel P, Borkenstein M, Kapelari K, Mayer M, Schemper M, Haüsler G. 2013. Austrian height and body proportion references for children aged 4 to under 19 years. Ann Hum Biol 40: 324–332. Kelishadi R. 2007. Childhood overweight, obesity, and the metabolic syndrome in developing countries. Epidemiol Rev 29:62–76. Kromeyer-Hauschild K, Wabitsch M, Kunze D, Geller F, Geiß HC, Hesse V, von Hippel A, et al. 2001. Perzentilen fu¨r den Body-MassIndex fu¨r das Kindes- und Jugendalter unter Heranziehung verschiedener deutscher Stichproben. Monatsschr Kinderheilkd 149:807–818.


11

Kuczmarski RJ, Ogden CL, Guo SS, Grunner-Stawn LM, Flegal KM, Mei Z, Wei R, et al. 2002. 2000 CDC growth charts for the United States: methods and development. National Center for Health Statistics. Vital Health Stat 11:1–190. Ku¨hnis J, Erne S. 2012. Stabilisation in the prevalence of childhood overweight in Liechtenstein between 2004 and 2010. Schweizerische Zeitschrift fu¨r Sportmedizin und Traumatologie 60:4–7. Li C, Ford ES, Mokdad AH, Cook S. 2006. Recent trends in waist circumference and waist-height ratio among US children and adolescents. Pediatrics 118:e1390–e1398. McCarthy HD, Jarrett KV, Emmett PM, Rogers I. 2005. Trends in waist circumferences in young British children: a comparative study. Int J Obes (Lond) 29:157–162. Moreno LA, Fleta J, Sarrıá A, Rodrı´guez G, Gil C, Bueno M. 2001. Secular changes in body fat patterning in children and adolescents of Zaragoza (Spain), 1980–1995. Int J Obes Relat Metab Disord 25: 1656–1660. Murer SB, Saarsalu S, Zimmermann MB, Aeberli I. 2014. Pediatric adiposity stabilized in Switzerland between 1999 and 2012. Eur J Nutr 53:865–875. O’Dea JA, Nguyen Hoang TD, Dibley MJ. 2011. Plateau in obesity and overweight in a cross sectional study of low, middle and high socioeconomic status schoolchildren between 2004 and 2009. Int J Public Health 56:663–667. OECD. 2011. Education at a glance 2011: OECD indicators. Paris, France: OECD Publishing. Ogden CL, Carroll MD, Kit BK, Flegal KM. 2012. Prevalence of obesity and trends in body mass index among US children and adolescents, 1999–2010. JAMA 307:483–490. Prader A, Largo RH, Molinari L, Issler C. 1989. Physical growth of Swiss children from birth to 20 years of age. First Zurich longitudinal study of growth and development. Helv Paediatr Acta Suppl 52:1–125. Rigby RA, Stasinopoulos DM. 2004. Smooth centile curves for skew and kurtotic data modelled using the Box-Cox power exponential distribution. Stat Med 23:3053–3076. Rolland-Cachera MF. 2011. Childhood obesity: current definitions and recommendations for their use. Int J Pediatr Obes 6:325–331. Rosario AS, Kurth BM, Stolzenberg H, Ellert U, Neuhauser H. 2010. Body mass index percentiles for children and adolescents in Germany based on a nationally representative sample (KiGGS 2003–2006). Eur J Clin Nutr 64:341–349. Stasinopoulos DM, Rigby RA. 2007. Generalized additive models for location scale and shape (GAMLSS) in R. J Stat Software 23:1–46. Statistik Austria. 2009. Statistic on population demographics at 1.1.2008 (‘‘Statistik des Bevo¨lkerungsstandes am 1.1.2008’’, structured by age, gender and province). Vienna, Austria: Statistik Austria. Available online at: www.statistik.at (accessed March 6, 2009). Statistik Austria. 2012. Average population by age and sex in the year 2012 (‘‘Jahresdurchschnittsbevo¨lkerung 2012 nach Alter und Bundesland - Ma¨nner und Frauen’’). Vienna, Austria: Statistik Austria. Available online at: www.statitik.at (accessed February 11, 2014). Telford RD, Cunningham RB, Daly RM, Reynolds GJ, Lafferty ARA, Gravenmaker KJ, Budge MM, et al. 2008. Discordance of international adiposity classifications in Australian boys and girls – the LOOK study. Ann Hum Biol 35:334–341. Trasande L. 2010. How much should we invest in preventing childhood obesity? Health Affairs 29:372–378. Tuan NT, Nicklas TA. 2009. Age, sex and ethnic differences in the prevalence of underweight and overweight, defined by using the CDC and IOTF cut points in Asian children. Eur J Clin Nutr 63:1305–1312. WHO. 2000. Obesity: preventing and managing the global epidemic. Report of a WHO Consultation. WHO Technical Report Series 894. Geneva: World Health Organization. WHO Multicentre Growth Reference Study Group. 2006. Child growth standards: methods and development: length/height-for-age, weightfor-age, weight-for-length, weight-for-height and body mass index-forage. Geneva: World Health Organization. Wijnhoven TM, van Raaij JM, Spinelli A, Rito AI, Hovengen R, Kunesova M, Starc G, et al. 2013. WHO European Childhood Obesity Surveillance Initiative 2008: weight, height and body mass index in 6– 9-year-old children. Pediatr Obes 8:79–97. Zarfl B, Elmadfa I. 1995. Body Mass Index (BMI) als Indikator fu¨r das ¨ bergewicht bei Kindern und Jugendlichen – Ergebnisse der ASNS. U Akt Erna¨hr Med 20:201–206.