Institute of Social Paediatrics and Adolescent Medicine of Ludwig-Maximilians-University, Munich, Germany

Correspondence to: R von Kries, Department of Paediatric Epidemiology, Institute of Social Paediatrics and Adolescent Medicine of Ludwig-Maximilians-University, Heiglhofstr. 63, 81377 Munich, Germany. E-mail: ag.epi@lrz.uni-muenchen.de

OBJECTIVE: To examine the prevalence of overweight and obesity in the entire population of 5- and 6-y-old children entering school in Germany, Bavaria, and to assess time trends over the last 15 y and the impact of ethnicity.

DESIGN: Cross-sectional studies were based on the obligatory school entry health examinations: all health districts of Bavaria in 1997 (n=127 735); three health districts every 5 y from 1982 to 1997 (n=16 281).

MEASUREMENTS: Body mass index (BMI; kg/m²) was calculated and the prevalence of overweight and obesity was defined based on national and international agreed cut-off points. Ethnicity was measured as German and non-German nationality.

RESULTS: The prevalence of overweight and obese children as defined by international reference values was 9.4 and 3.1% for 5-y-old boys, 10.0 and 2.9% for 6-y-old boys, 12.2 and 3.3% for 5-y-old girls and 12.4 and 3.3% for 6-y-old girls. The whole BMI distribution in non-German children compared to German children was shifted to the right with median values in non-German children 0.3-0.5 kg/m² higher. In these the prevalence of overweight/obesity was 1.9/2.4 times higher for boys and 1.5/1.9 times higher for girls. The time trend between 1982 and 1997 shows an increase of the BMI distribution in the upper percentiles, whereas the lower percentiles did not change substantially. The increased prevalences of overweight/obesity for both sexes as defined by international references increased from 8.5/1.8% in 1982 to 12.3/2.8% in 1997.

CONCLUSION: This large study on all children entering school in Bavaria in 1997 shows patterns of overweight and obesity which are comparable with other European data but are lower than US and Australian data. Increasing prevalences since 1982 indicate that overweight and obesity in children are of increasing public health importance in Bavaria. The upwards shift of the BMI distribution in non-German children needs further investigation.

International Journal of Obesity (2002) 26, 1211-1217. doi:10.1038/sj.ijo.0802013

overweight; obesity; child; body mass index; prevalence; secular trend; ethnic groups; Germany

Introduction

Rising prevalences for childhood obesity are becoming a major public health concern in Europe, America and Australia.^{1,2,3,4,5,6,7} Overweight in childhood leads to severe psychological and health problems and predisposes to overweight in adulthood with all its known adverse health outcomes.^{8,9,10, 11}

To allow for sensible international comparisons and to assess time trends within a country valid national reference data are needed. New German reference data have recently been published.¹² The data source of these references were data of different studies carried out from 1985 to 1999 and the maximum number of cases per gender and age group did not exceed 2083 children. We had the opportunity to assess body mass index (BMI) values in all 5- and 6-y-old children (n=127 735) seen during the 1997 obligatory school entry health examinations in a German state, Bavaria (12 million inhabitants), and to assess the time trend since 1982 in three representative health districts in Bavaria.

The objective of this study was to compare these data to international and national figures and to assess the impact of time trend and ethnicity as defined by non-German citizenship.

Methods

Study population and data sources

In this cross-sectional study routine data on height and weight collected during the health examinations at the time of school entry in 5- and 6-y-old children in Bavaria, Germany, were analysed. Bavaria is the largest (70 552 km²) and the second most populated state of Germany (12 million inhabitants). The data were collected by health districts (n=75), where 1-19 doctors were involved in the investigations. School entry health examinations have been obligatory since 1985. This study was based on data from the 1997 obligatory school entry health examinations in all 75 health districts in Bavaria allowing analysis of data for the whole population.

In order to assess time trend we tried to identify representative public health offices with complete data available for at least 1982, 1987, 1992, 1997. These health districts were identified by the following procedure: out of all 75 health districts only 20 had data of their school health examinations accessible and sorted by year although not in electronic format. Therefore manual data collection was required. Because of limited resources eight health districts were chosen to represent big cities, small cities and rural areas. These health districts were visited by trained personal to abstract the required information from the school health examination data. Unfortunately, the data quality in these health districts was not consistent and not sufficient. We therefore had to constrain our further analysis to three districts, where the data met the following criteria: children had been examined by trained medical personal and the maximum difference between the number of available data sets and the children starting school did not exceed 30% in each study year. Some differences between the number of available data sets and the children starting school may result from mobility between districts, because the child's health card is sent to the new district. The finally selected health districts are one densely populated area (München Land: 421 inhabitants/km²) and two rural areas (Lichtenfels: 136 inhabitants/km²; Mühldorf a. I.: 132 inhabitants/km²). In the 1997 investigation the distribution of the BMI by age and gender in these three selected health districts was similar compared to the average of all Bavarian districts (data not shown), suggesting that these health districts are representative for Bavaria.

Description of variables

The school entry health examinations were either performed in the public health office or in the kindergarten. In the first case height and weight were measured with fixed stadiometers and balances and, in case of measurements outside the health offices, portable stadiometers and scales were used.

BMI was calculated as kg/m². Values for BMI under 10 kg/m² and above 30 kg/m² were excluded as implausible. The estimation of the prevalence of overweight and obesity was based on cut-off points derived from international data as recommended by the Childhood Obesity Working Group of the International Obesity Task Force.¹³ These cut-off points are linked to the widely accepted adult cut-off points of a BMI of 25 kg/m² (overweight) and 30 kg/m² (obesity). Data for 1997 are additionally compared to the recently published BMI percentile curves for German children.¹² In these, being overweight/obese was defined as the percentage of children whose BMI is equal or above the 90th (overweight)/97th (obesity) age- and sex-specific percentile.

The routine health examinations include a query on German and non-German nationality but no further questions regarding ethnicity, income or parental education.

Data analysis

The distribution of the BMI was described by mean, standard deviation and percentile values, and the proportion of overweight and obese children with their confidence limits. Differences between gender and nationality were tested using ²-test for categorical data. Trends in the prevalence of overweight and obesity were analysed with the ²-trend test by Cochran and Armitage.¹⁴ All analyses were carried out using SAS 6.12 statistical package.

Results

In 1997 a total of 129 881 children aged 5 and 6 y participated in the school entry examinations in Bavaria. Some 1.6% had missing information on height or weight and 0.1% were excluded because of implausible data on BMI, leading to a study population of 127 735 children (98.3%). For 902 boys and 751 girls data on nationality are lacking, so that the population is reduced to 126 083 children when data on nationality were also considered.

The number of 5- and 6-y-old children with available data for BMI in 1982 to 1997 ranged from 2773 to 4911 as a result of increasing birth rates in the 1980s in Bavaria and a postponement of school entry health examinations to after school entry in the year 1982 in one district (resulting in a higher number of 7-y-old children, who were not considered in this investigation).

Table 1 shows the characteristics of the BMI distribution for all Bavarian children investigated in the school health examination in 1997 by gender, age and nationality. Non-German nationality was reported in 8.4% of all children. In non-German children there was a shift in the entire BMI distributions resulting in median values which are 0.3-0.5 kg/m² higher than in German children and even larger differences regarding the upper percentiles (eg 95th percentile: non-German children 1.0-1.8 kg/m² higher BMI values than German children).

In order to allow for comparison of prevalence estimates with other studies we used recently published national and international reference values to define overweight and obesity (Table 2).

With regard to international reference values, 9.4% 5-y-old and 10.0% 6-y-old boys are overweight as compared to 12.2% 5-y-old and 12.4% 6-y-old girls (P<0.001). The proportion of obese children is 3.1% for 5-y-old and 2.9% for 6-y-old boys compared to 3.3% for 5- and 6-y-old girls each (P=NS).

Comparing German and non-German children, the shift in the upper percentiles results in a considerably higher proportion of overweight and obese non-German children in all age and gender categories (P<0.001). The frequency of overweight in non-German boys was on average 1.9 times higher (16.5 vs 8.8% for 5-y-old and 16.5 vs 9.3% for 6-y-old boys); the respective figures for obesity accounted for a factor of 2.4 on average in boys (6.8 vs 2.7% for 5-y-old and 5.9 vs 2.6% for 6-y-old boys). The figures for non-German girls accounted for a factor of 1.5 regarding overweight (16.5 vs 11.9% for 5-y-old and 18.5 vs 11.8% for 6-y-old girls) and for a factor of 1.9 regarding obesity (5.6 vs 3.1% for 5-y-old and 5.8 vs 3.0% for 6-y-old girls).

The proportion of obese children is almost identical if compared to the latest German or international reference values, whereas the proportion of overweight is lower with the German reference values compared to the international reference values.

Trends in body mass index 1982-1997

Table 3 shows the time trends in the prevalence of overweight and obesity between 1982 and 1997 in the three selected Bavarian health districts. The overall prevalence of overweight and obesity increased for both sexes between 1982 and 1997. Using international reference data, in 1982, 8.5% of all children (7.6% boys; 9.5% girls) were overweight and 1.8% (1.5% boys; 2.1% girls) were obese; in contrast 12.3% (11.0% boys; 13.6% girls) of all children were overweight and 2.8% (2.8% boys; 2.8% girls) were obese in 1997.

Figures 1 and 2 show the 10th, 25th, 50th, 75th, 90th and 97th percentile values for BMI over time for boys and girls. While the lower percentile values did not increase in girls and increased only slightly in boys, the upper percentile values increased more markedly for both sexes. The marked increase related to the upper percentiles indicates that the observed differences over time are not caused by a shift in the mean but by a shift in the upper tail of the distributions.

Discussion

In this study up-to-date data on overweight and obesity in German children are shown in sufficient detail to allow for comparisons with other populations. The strength of our data is that they are based on the entire population of 5- and 6-y-old pre-school children in a state with the second highest population of all German states. Such a huge number of children implies that data quality cannot be as good as in smaller studies on which other published data in Europe are based.^2,4,15,,,18 Misclassification on weight and height in our study, however, may be assumed to be random and therefore should not result in bias.

Data for the assessment for time trends had to be confined to three public health offices because of limited resources and because not all offices had complete and valid documentation over time. Since data for 1997 on these health districts are similar to those of all Bavaria, we assume that the trend in these districts reflects the trend in Bavaria. The presumptive completeness of the documentation of these offices was assessed by comparison of the number of available data sets and the corresponding number of children starting school in the respective year and region. The maximum differences in the three health districts enrolled were in the range of 5.2-12.0% in 1982, 5.3-20.9% in 1987, 4.9-25.5% in 1992 and 1.0-4.0% in 1997, suggesting that ascertainment bias should be negligible.

Comparisons between other countries are difficult, because percentile values are rarely reported and different reference values for the definition of overweight and obesity have been used.^{17,18,19,20,21} Only a few studies used the reference values by Cole et al.¹³ US data,^6,13 for example, show a higher prevalence of overweight children (boys: 18.1%; girls: 16.5%) compared to our data. In Australian children higher prevalences were found as well: 13.4% of boys and 19.5% of girls were overweight.⁷

In Europe, a representative study using the reference values of Cole et al¹³ in the UK² found lower prevalences of overweight in 4-6-y-old children, with a higher prevalence of overweight in Scottish compared to English children: the prevalence of overweight in Scottish boys/girls was 7.6/11.9% compared to 9.7/12.3% in our study. Higher mean values for some south European countries, however, have been reported, indicating that the whole distribution of BMI in south European children may be shifted to the right: in a large population based study on school children in Spain 1995, Moreno et al³ reported a mean BMI of 16.96 kg/m² for 6-y-old boys and 16.91 kg/m² for 6-y-old girls, which is higher than the respective BMI of 15.4 kg/m² in both boys and girls in our study.

In Germany, publications on the burden of overweight and obesity in pre-school children exist but are based on data with small sample size and mostly represent big cities.^{4,22,23,24,25,26} To our knowledge only the study of Zabransky et al²⁷ was representative for a German state (Saarland, 1 million inhabitants) and the study of Hebebrand et al²⁸ for west German states, but in the latter weight and height were asked by self-reported questionnaires. In a recently published study Kromeyer-Hauschild et al¹² summarized data from 17 studies on childhood obesity to generate new reference values for German children. Although this is a useful approach in the absence of national representative data for Germany, there are some limitations with respect to the heterogeneity of these data: big cities are over-represented, age-group-specific figures are based on varying numbers of children since the data are derived from different studies which each cover different age-groups, the studies were performed between 1985 and 1999 accounting for cohort effects, and the numbers of children per age and gender group varied from 80 to 2083 children (5-y-old boys/girls: 1301/1260; 6-y-old boys/girls: 2045/2083).

Although in our data on all 5- and 6-y-old children entering school in Bavaria lower proportions of overweight and higher proportions of obese children were found than in the study of Kromeyer-Hauschild et al¹² (by definition: proportions of overweight/obesity, 10/3%), these differences were small in size (maximum difference 2%) and all confidence intervals for the estimates do overlap in both studies except one (6-y-old boys).

In order to identify a possible secular trend in the prevalence of overweight and obesity in German children we analysed data from the last 15 y in three representative Bavarian health districts. The data show an increase in the proportion of children in the upper tail of the distribution over time. Similar findings regarding rising prevalences of overweight and obesity in childhood have been reported in repeated cross-sectional studies from other countries. In US children, nationally representative samples over a 30 y period showed that the prevalence of overweight and obesity has increased dramatically, especially in the last 10 y. While mean BMI increased from 17.6 to 19.5 kg/m² in 5-14-y-old US children between 1973 and 1994, even more marked differences were found in the upper percentiles, so that prevalences of children above the 95th percentile reference value doubled.^1,6 Although different definitions for overweight and obesity have been used in Europe, most studies on secular trends in BMI reported an overall increase of overweight and obesity (ie in the upper percentiles of the BMI distribution) with constant or only slightly increasing BMI values for the lower quartiles of the BMI distributions.^3,4 In British 4-6-y-old children Chinn et al² found little change in the prevalence of overweight and obesity defined by Cole et al¹³ between 1997-1984, whereas between 1984 and 1994 the prevalence of overweight (obesity) increased from 5.4% (1.4%) to 9.0% (1.7%) in English boys and from 9.3% (1.5%) to 13.5 (2.6%) in English girls. The same pattern was found in children of an East German city, with a slight increase of the prevalence of overweight (obesity) as defined by Rolland-Cachera¹⁸ from 10.0% (5.3%) to 11.8% (6.1%) in boys and 11.7% (4.7%) to 13.0% (5.3%) in girls between 1975 and 1985. Between 1985 and 1995 the prevalence of overweight increased to 16.3% in boys and 20.7% in girls and the prevalence of obesity to 8.2% in boys and 9.9% in girls.⁴ A similar trend has already been observed in a small study for 1968-1987 in a city in the north of Germany.²³

The predominant increase in the upper tail in the BMI distribution in Bavaria and other European countries over time contrasts with the findings in the non-German children in Bavaria: in non-German children the BMI of the entire population appears to be shifted to the right (not only the upper tail). To our knowledge, this is the first German data showing considerably higher BMI values in ethnic minorities. In Bavaria, 84% of all non-German children have south-east European origin (half of them Turkish).²⁹ While ethnic groups are generally considered in US studies, only a few European studies on BMI examined different ethnic groups within a country. In France, children with Maghrebian origin were found to have a higher prevalence of obesity.³⁰ In Britain, children from the Indian subcontinent were thinner than Afro-Caribbean and white children.³¹ Apart from genetics of the ethnic group, socio-economic status and lifestyle factors must be considered as potential explanations for these differences. In Bavaria, ethnic minorities from south-east European origin have a lower income and school education.²⁹ compared to families with German origin. Since we had no further data for characterisation of non-German children, the causes of the increased BMI values in non-German children remain speculative.

Conclusion

This large study on all 5- and 6-y-old children entering school in Bavaria 1997 shows patterns of overweight and obesity which are comparable with other European data. Increasing prevalences since 1982 indicate that overweight and obesity in children are an increasing Public Health problem in Bavaria. The upwards shift of the BMI distribution in non-German children needs further investigation.

Acknowledgements

This research was supported by the Bayerisches Staatsministerium für Arbeit und Sozialordnung, Familie, Frauen und Gesundheit, Democh Mauermaier Stiftung and Child Health Foundation Munich, Germany.

1 Freedman DS, Srinivasan SR, Valdez RA, Williamson DF, Berenson GS. Secular increases in relative weight and adiposity among children over two decades: the Bogalusa Heart Study. Pediatrics 1997; 99: 420-426. MEDLINE

2 Chinn S, Rona RJ. Prevalence and trends in overweight and obesity in three cross sectional studies of British Children, 1974-94. Br Med J 2001; 322: 24-26.

3 Moreno LA, Sarria A, Fleta J, Rodriguez G, Bueno M. Trends in body mass index and overweight prevalence among children and adolescents in the region of Aragon (Spain) from 1985 to 1995. Int J Obes Relat Metab Disord 2000; 24: 925-931. MEDLINE

4 Kromeyer-Hauschild K, Zellner K, Jaeger U, Hoyer H. Prevalence of overweight and obesity among school children in Jena (Germany). Int J Obes Relat Metab Disord 1999; 23: 1143-1150. MEDLINE

5 O'Loughlin J, Paradis G, Meshefedjian G, Gray-Donald K. A five-year trend of increasing obesity among elementary schoolchildren in multiethnic, low-income, inner-city neighborhoods in Montreal, Canada. Int J Obes Relat Metab Disord 2000; 24: 1176-1182. MEDLINE

6 Troiano RP, Flegal KM, Kuczmarski RJ, Campbell SM, Johnson CL. Overweight prevalence and trends for children and adolescents. The National Health and Nutrition Examination Surveys, 1963 to 1991. Arch Pediatr Adolesc Med 1995; 149: 1085-1091. MEDLINE

7 Magarey AM, Daniels LA, Boulton TJ. Prevalence of overweight and obesity in Australian children and adolescents: reassessment of 1985 and 1995 data against new standard international definitions. Med J Aust 2001; 174: 561-564. MEDLINE

8 Gortmaker SL, Must A, Perrin JM, Sobol AM, Dietz WH. Social and economic consequences of overweight in adolescence and young adulthood. New Engl J Med 1993; 329: 1008-1012. MEDLINE

9 Must A, Jacques PF, Dallal GE, Bajema CJ, Dietz WH. Long-term morbidity and mortality of overweight adolescents. A follow-up of the Harvard Growth Study of 1922 to 1935. New Engl J Med 1992; 327: 1350-1355. MEDLINE

10 Nieto FJ, Szklo M, Comstock GW. Childhood weight and growth rate as predictors of adult mortality. Am J Epidemiol 1992; 136: 201-213. MEDLINE

11 Serdula MK, Ivery D, Coates RJ, Freedman DS, Williamson DF, Byers T. Do obese children become obese adults? A review of the literature. Prev Med 1993; 22: 167-177. Article MEDLINE

12 Kromeyer-Hauschild K, Wabitsch M, Kunze D, Geller F, Gei HC, Hesse V et al. Perzentile für den Body-mass-Index für das Kinder- und Jugendalter unter Heranziehung verschiedener deutscher Stichproben. Monatsschr Kinderheilkd 2001; 149: 807-818.

13 Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. Br Med J 2000; 320: 1240-1243.

14 Armitage P, Berry G. Statistical methods in medical research 3rd edn Blackwell: Oxford, 1994.

15 Maffeis C, Schutz Y, Piccoli R, Gonfiantini E, Pinelli L. Prevalence of obesity in children in north-east Italy. Int J Obes Relat Metab Disord 1993; 17: 287-294. MEDLINE

16 Nuutinen EM, Turtinen J, Pokka T, Kuusela V, Dahlstrom S, Viikari J et al. Obesity in children, adolescents and young adults. Ann Med 1991; 23: 41-46. MEDLINE

17 Lindgren G, Strandell A, Cole T, Healy M, Tanner J. Swedish population reference standards for height, weight and body mass index attained at 6 to 16 years (girls) or 19 years (boys). Acta Paediatr 1995; 84: 1019-1028. MEDLINE

18 Rolland-Cachera MF, Sempe M, Guilloud-Bataille M, Patois E, Pequignot-Guggenbuhl F, Fautrad V. Adiposity indices in children. Am J Clin Nutr 1982; 36: 178-184. MEDLINE

19 Hammer LD, Kraemer HC, Wilson DM, Ritter PL, Dornbusch SM. Standardized percentile curves of body-mass index for children and adolescents. Am J Dis Child 1991; 145: 259-263. MEDLINE

20 White EM, Wilson AC, Greene SA, McCowan C, Thomas GE, Cairns AY et al. Body mass index centile charts to assess fatness of British children. Arch Dis Child 1995; 72: 38-41. MEDLINE

21 Leung SS, Lau JT, Tse LY, Oppenheimer SJ. Weight-for-age and weight-for-height references for Hong Kong children from birth to 18 years. J Paediatr Child Health 1996; 32: 103-109. MEDLINE

22 Hesse V, Bartezky R, Jaeger U, Kromeyer-Hauschild K, Zellner K, Vogel H et al. Körper-Masse-Index (KMI)¾Perzentilen deutscher Kinder im Alter von 0-18 Jahren. 1999.

23 Danker-Hopfe H, Roczen K. Secular trends in height, weight and body mass index of 6-year-old children in Bremerhaven. Ann Hum Biol 2000; 27: 263-270. MEDLINE

24 Schaefer F, Georgi M, Wühl E, Schärer K. Body mass index and percentage fat mass in healthy German schoolchildren and adolescents. Int J Obes Relat Metab Disord 1998; 22: 461-469. MEDLINE

25 Schwandt P, Geiss HC, Ritter MM, Ublacker C, Parhofer KG, Otto C et al. The prevention education program (PEP). A prospective study of the efficacy of family-oriented life style modification in the reduction of cardiovascular risk and disease: design and baseline data. J Clin Epidemiol 1999; 52: 791-800. MEDLINE

26 Mast M, Körtzinger I, König E, Müller MJ. Gender differences in fat mass of 5-7-year old children. Int J Obes Relat Metab Disord 1998; 22: 878-884. MEDLINE

27 Zabransky S, Weinand C, Schmidgen A, Schafmeister C, Muller S, Hollinger-Philipp R et al. [Saarland Growth Study: percentile charts for height, weight, and body mass index (BMI) for boys and girls, 4-18 years old, in Saarland.]. Wien Med Wochenschr 2000; 150: 145-152. MEDLINE

28 Hebebrand J, Heseker H, Himmelmann W, Schäfer H, Remschmidt H. Altersperzentilen für den Body-Mass-Index aus Daten der nationalen Verzehrstudie einschlielich einer Übersicht zu relevanten Einflufaktoren. Akt Ernähr Med 1994; 19: 259-265.

29 Bayerisches Landesamt für Statistik und Datenverarbeitung. Statistisches Jahrbuch für Bayern 1997 Bayerisches Landesamt für Statistik und Datenverarbeitung: München, 1997.

30 Roville-Sausse F. [Increase during the last 20 years of body mass of children 0 to 4 years of age born to Maghrebian immigrants.]. Rev Epidemiol Sante Publique 1999; 47: 37-44. MEDLINE

31 Chinn S, Hughes JM, Rona RJ. Trends in growth and obesity in ethnic groups in Britain. Arch Dis Child 1998; 78: 513-517. MEDLINE

Figure 1 Body mass index percentile values (kg/m²) by year in pre-school boys of three health districts in Bavaria (Germany), 1982-1997, using interpolation between the points. Dashed lines represent the baseline value of 1982.

Figure 2 Body mass index percentile values (kg/m²) by year in pre-school girls of three health districts in Bavaria (Germany), 1982-1997, using interpolation between the points. Dashed lines represent the baseline value of 1982.

Table 1 Body mass index characteristics (kg/m²) in pre-school children by gender, age and nationality in Germany (Bavaria), 1997

Table 2 Prevalence of overweight and obesity in pre-school children by gender, age and nationality in Germany (Bavaria), 1997, defined by German¹² and international¹³ cut-off points for overweight and obesity

Table 3 Prevalence and trends of prevalence in overweight and obesity in pre-school children of three health districts in Germany (Bavaria), 1982-1997