In 2002, a cross-sectional study was conducted at 12 primary schools in Bielefeld, a town of some 320.000 inhabitants in North-West Germany. The study population comprised of all children registered at these schools for school entry in 2002. Through the city public health office these children and their parents were invited to attend the compulsory medical examination taking place at the local school. The 12 schools (out of a total of some fifty primary schools in the town) were pragmatically selected based on information available for the preceding year. The aim was to include schools with a relatively balanced mix of migrant and German children and fair variation of social status among the population in the respective school districts.
Social and demographic information relevant to the children were derived from a 36 item-questionnaire filled out by parent or relative who accompanied the children [21
]. When two adults accompanied the child we collected interview data from only one parent or relative, usually the mother. The questionnaire was available in 4 languages (German, Turkish, Russian, Polish). Parents or relatives were interviewed on the day of the school health examination with the assistance of trained interviewers.
The medical examination of the children was carried out by medical staff from the public health office in Bielefeld using a standardized model of medical school health examination [22
]. During the physical and psychological examination, height and weight of children were recorded. All anthropometric measurements of the children were performed in the morning. Body weight was determined to the nearest 500 g using electronic or mechanic non -calibrated scales. During the measurements the children were dressed in light indoor clothes without shoes. No adjustments were made for clothing. Height was measured to the nearest 10 mm using a mobile scale.
657 children aged 6 and 7 appeared in the schools for pre-school examination. Interview data were collected from 565 adult companions (response rate of 86%). The sample size was further reduced to 537 data sets due to merging errors and missing medical data for some children. We also excluded all 12 children who had non-parental companions at the examination.
The statistical analyses were performed on merged data from 525 children (274 boys and 251 girls) and their parents, corresponding to an overall response rate of 79.9% (525/657). 267 (50.9%) parents were of German descent and 258 (49.1%) of Non-German descent. There were 240 females and 27 males among the German parents and 191 females and 67 males among the Non-German parents. Migrant status of the children was based on the migrant status of the interviewed parent. Parents who themselves or whose parents were born in Germany were classified as German, all others were defined as migrants.
The study population consisted of three main migrant groups: 36% of migrants were of Turkish origin, 25% originated from Russia and 15% from Poland. A total of 26 different countries of origin was noted.
A social class index consisting of three social classes (low, medium, high) was constructed using four indicators (primary qualification, professional education, employment status and size of household). We divided these variables into 3 levels (variable professional education in 2 levels) and scored 1–3 points. The sum of the points was than divided in 3 social classes (high social class 10–11 points, medium social class 7–9 points, low social class 4–6 points) [23
BMI was calculated as body weight divided by body height squared (kg/m2
). Overweight and obesity were defined according to the recommendations of the International Obesity Task Force IOTF, using international reference values based on data from six countries [24
]. These age and sex specific BMI cut-off points for overweight and obesity in children (between 2 and 18 years) were constructed using dataset specific centiles corresponding to the widely accepted adult cut-points of a BMI of 25 kg/m2
(overweight) and 30 kg/m2
For the calculation of overweight and obesity the exact age of the children was rounded to the nearest half year. Children with BMI at or above cut-off value corresponding to a BMI of 25 kg/m2
in late adolescence (e.g. BMI of 17.55/17.34 for 6 year old boys/girls) were classified as overweight (or obese when the respective cut-points were reached). For comparative purposes, we also used national BMI reference curves for Germany [25
] in some instances. Here, the age-and sex-specific 90th
BMI percentile are used as cut-point for overweight and obesity, respectively.
The analysis was performed on the data set containing merged data from both the parental survey and childrens' medical examination.
We used descriptive statistical methods to compare obesity and overweight between migrant and German children. Data were stratified according to migrant status, social status and duration of stay in Germany. Differences in proportions were tested by Fisher's exact or Chi-square test, trends in the prevalence of overweight were analysed with chi-square trend test.
Logistic regression models were used to simultaneously analyse the association between overweight and gender, social status, ethnic background and duration of stay in Germany among migrants. Adjusted odds ratios and 95% confidence intervals were calculated. The statistical analysis was performed with the software package SPSS (version 11.5; SPSS Inc, Chicago, IL).