This cross-sectional study was carried out from November 2001 until May 2002. The study population consisted of primary schoolchildren aged 12 and 13 years old, living in Istanbul. Out of totally 650 6th grade primary school children registered in the selected schools, 510 pupils (78.5%) were finally studied (257 males, 253 females). Inclusion of subjects was on a voluntary basis; prior to acceptance, children's parents or guardians were fully informed about the objectives and methods of the study and signed a consent form.
The study population was selected from one private and two public schools using the multi-stage sampling method. All children in the same class were invited to participate in the study to avoid ethical problems.
Approval to conduct this survey was granted by the Ethical Committee of Marmara University and the Turkish Ministry of Education.
Baseline data was collected during face-to-face interviews with children by a team of trained personnel. The data collected from the children consisted of physiological indices, such as biochemical, anthropometrical and cardiorespiratory fitness measurement and behavioral indices, such as dietary habits, estimation of energy and nutrient intake and physical activity assessment. These data are presented in more detail below.
From the data provided by the Ministry of National Education and the National Statistical Centre of Turkey three districts within the metropolitan area of Istanbul were identified. These three districts are inhabited by citizens with distinguished SES and consequently this is reflected on the SES of the pupils attending the schools within the respected school zones. From the high SES district a private school was selected randomly among all the private schools located in that zone, while for the middle and low SES districts one public school, from each one of these areas, among the public schools located in these two zones respectively was selected.
Hence for the needs of the present study the SES grouping of the pupils was based on the SES of the district their schools were located, while middle and high SES group children were grouped as one group.
Body weight was measured using a digital scale (Seca) with an accuracy of ± 100 g. Subjects were weighed without shoes, in the minimum clothing possible, i.e. underwear. Standing height was measured without shoes to the nearest 0.5 cm with the use of a commercial stadiometer, with the shoulders in relaxed position and arms hanging freely . Body Mass Index (BMI) was calculated by dividing weight (kg) by height squared (m2). Left triceps, biceps, subscapular and suprailiac skinfold thickness, were measured with a Lange skinfold calliper. The sum of these four skinfolds was then estimated.
Definition of underweight and overweight
Based on the US NHANES I data, in the current study the 5th percentile for BMI was used as a cut-off point for defining childhood underweight . This method is recommended by the Centre of Diseases Control (CDC), as well as for international use for adolescents aged 10–19 years by a WHO expert committee .
Regarding classification of the under study population in overweight and obese, the age- and sex-specific BMI cut-off points proposed by the Childhood Obesity Working Group of the International Obesity Task Force (IOTF) have been adopted in the current study . These cut-off points are based on the reference values of Cole et al (2000) and they are widely used in many studies with children and adolescents [21–23]. Due to the relatively low prevalence of obesity in the current population, for further examination of health indices associated with obesity per se, apart from the initial estimation of prevalence rate, in all other data analysis both overweight and obese subjects were grouped together as overweight .
Early morning venous blood samples were obtained from each child for biochemical screening tests, following a 12-h overnight fast. Professional staff performed venipuncture, using vacutainers to obtain 10 ml of whole blood. When blood collection was completed all samples were stored at 3–4°C and sent to Marmara University, Faculty of Health Education where the actual biochemical analysis took place.
One of these aliquots was used for the determination of several biochemical parameters. Total cholesterol (TC) was determined using Allain's method , while Fossati's method was used for triglycerides (TG) determination . High Density Lipoprotein-cholesterol (HDL-C) was measured by the heparin-manganese precipitation method. Low Density Lipoprotein-cholesterol (LDL-C) was calculated as follows: LDL-C = TC - (HDL-C + TG/5) .
Food consumption of children was assessed by the 24-hour recall technique on three consecutive days. Dietary data was collected from children during a face-to-face interview with a trained dietician. Dieticians were trained as a group to minimize inter-observer variation. During the interview, food models and photos of common Turkish dishes of various portions, as well as household cups and measures were used to define amounts, in order to obtain as accurate information as possible, regarding the type and amount of food and beverages consumed during the previous day. Macronutrient and micronutrient intake were calculated using the food database available in Marmara University, Faculty of Health Education. This database contains Turkish food composition tables for all food, including cooked Turkish dishes. Information on processed foods was obtained from food companies and national as well as international fast food chains.
Physical activity assessment
Physical activity during school hours and leisure time was assessed using a standardized physical activity diary completed by the children for two consecutive weekdays and one weekend day. A member of the research team crosschecked diary information during daily interviews. The diary was constructed for searching various physical activities ranging from mild to vigorous , while the time spent for each type of activity was recorded in hours.
Activities were classified into two groups, namely Sedentary and Light Activities (< 4 METs) and Moderate to Vigorous Physical Activities (MVPA) (> 4 METs). Typical activities in the Sedentary and Light category were watching TV, board and computer games, studying and extra curricular classes (e.g. music, language), etc. The MVPA category included activities such as walking, bicycling, rhythmic-gymnastics, dancing, basketball, soccer, athletics, tennis, swimming, running up and down, jumping rope and general participation in active outdoors games. Given the age group, MVPA was defined as continuous vigorous activity causing sweating and heavy breathing for periods longer than 15 minutes, but with occasional breaks in intensity, rather than the strict aerobic definition of 20 continuous minutes appropriate for adults.
Cardiorespiratory fitness assessment
Cardiorespiratory fitness was estimated indirectly according to children's performance on the Endurance 20 m shuttle Run Test (ERT). The ERT is a field test included in the European battery of physical fitness tests and is recommended by the Committee of Experts on Sports Research for the assessment of cardiorespiratory fitness in school children . During this test, subjects start running at a speed of 8.5 Km/h and speed is increased at various stages. The subjects move between two lines at a distance of 20 m apart, reversing direction and continuing backwards and forwards in accordance with a pace dictated by a sound signal on an audio tape, which gets progressively faster (0,5 Km/h every minute). Each stage of the test is made up of several shuttle runs, but the actual score of the subject is the last half stage fully completed before they drop out (the stages are 0.0, 0.5, 1.0, 1.5, 2.0 etc). In our study, the number of shuttle runs that each child completed was estimated and referred as ERT score. The higher the ERT score the better the cardiovascular function. The ERT is recommended for large groups of children since it is reliable, valid, non-invasive and requires limited facilities .
Descriptive statistics of continuous variables are expressed as the mean ± Standard Deviation (SD). At first, univariate linear regression analysis was performed in order to determine variables related to BMI. Only variables with significant association were used to subsequent analysis. All variables followed a normal distribution with the exception of BMI, HDL-C, LDL-C, TG, energy and fat intake, which were transformed using the natural logarithm before further analysis.
The chi-square test was used to compare prevalence and percentages between SES groups for both sexes. Differences between low and middle/ high SES children in relation to physiological and behavioral indices were determined using the Independent Sample T-test. All analyses were conducted using the SPSS 10.0 statistical software package for Windows. In all analyses a 5% significance level was used.