School-based interventions to prevent overweight and obesity in ...

Acta Pædiatrica ISSN 0803–5253

REGULAR ARTICLE

School-based interventions to prevent overweight and obesity in prepubertal children: process and 4-years outcome evaluation of the Kiel Obesity Prevention Study (KOPS) Sandra Danielzik, Svenja Pust, Manfred J. Muller ¨ ([email protected]) Institut fur ¨ Humanernährung und Lebensmittelkunde, Christian-Albrechts-Universität, Kiel, Germany

Keywords Children, Feasibility, Overweight, School-based intervention Correspondence Prof. Dr. Med., Manfred James Muller, ¨ Institut fur ¨ Humanernährung und Lebensmittelkunde, Agrarund Ernährungswissenschaftliche Fakultät, Christian-Albrechts-Universität zu Kiel, Dusternbrooker ¨ Weg 17-19, D-24105 Kiel, Germany. Tel: 0431 8805670 | Fax: 0431 8805679 | Email: [email protected] Received 3 January 2006; revised 14 September 2006; accepted 5 December 2006. DOI:10.1111/j.1651-2227.2007.00165.x

Abstract Aim: To evaluate the feasibility and 4-year outcome of school-based health promotion on overweight among 6–10-year-old children. Methods: Four-year follow-up data of 344 children participating in health promotion (I) as part of the Kiel Obesity Prevention Study (KOPS) at age 6 years, compared with 4-year changes in 1420 non-intervention children (NI). Nutritional knowledge was assessed before and 3 months after intervention in 1996 and 2004. Outcome was characterized compared to reference values for (i) BMI, (ii) triceps skinfold (TSF) and (iii) waist circumference (WC). Results: Process evaluation showed an increase in knowledge after intervention. However, the prevalence of children with good nutritional knowledge before intervention doubled from 1996 to 2004 but similar intervention-induced increases in knowledge (+50%) were observed. When compared with NI I increased remission of overweight with no significant effect on incidence. The effect was most pronounced in girls. The effect was affected by definition of overweight: when compared with parameters of fat mass (TSF and WC), BMI was showing a stronger effect (remission in girls ( I – NI): BMI: +13.4%, TSF: +18.7%, WC: +20.7%). Conclusions: School-based health promotion has sustainable effects on nutritional knowledge and remission of overweight being most pronounced in girls. The effect of intervention was most pronounced using TSF and WC as criteria of overweight.

INTRODUCTION The prevalence of overweight and obesity in children and adolescence has increased rapidly over the past 30 years all over the world (1,2). Since treatment of childhood obesity has limited success preventive measures are necessary but its effectiveness is not well documented (for review see 3). School health programs have the potential to influence health of nearly all children within existing institutional structures. In the U.S. it has been suggested that one third of the health objectives for the nation can be significantly influenced by school health programmes (4). However, feasibility of school-based obesity prevention programmes has not been assessed systematically and most school-based trials addressing childhood overweight produced some improvements in behavior but only minor or no effects on mean BMI (for reviews see 3,5). The latter finding addresses the definition of overweight. Most experts use body mass index (BMI) as a measure of fat mass and international reference values were published in 2000 (6). However, BMI is a suboptimal marker for total body fat. In addition, BMI does not reflect body fat distribution (5,7). Alternatively, skinfold thickness as well as waist circumferences have been recommended to identify individuals with increased risk of overweight-associated diseases (7). However, the possible value of fat mass and fat distribu-

tion as suitable outcome variables in prevention studies on children has not been investigated. The Kiel Obesity Prevention Study (KOPS) was started in 1996 as a cross-sectional as well as longitudinal 8-year follow-up study. The intention of KOPS was to investigate the determinants of childhood overweight (8). In addition, the effects of school-based and family-based interventions on overweight and obesity were/are followed. Our preliminary 1-year follow-up study data of school-based intervention suggested positive effects on nutritional status as measured by TSF but not by BMI (9). We now describe data on feasibility and 4-year outcome of school-based health promotion in children. In addition the influence of a parameter used to define overweight on the effect of intervention will be discussed. METHODS Study population The design of the KOPS has been well described previously (8–10). Briefly, KOPS investigated three cohorts of 6- (T0), 10- (T1) and 14- (T2) year old children. A subsection of each group was re-examined as part of two or three cohorts (i.e. they were followed longitudinally). The T0–T1 longitudinal data (i.e. the 6- to 10-year follow-up data) will be described in this paper.

C 2007 The Author(s)/Journal Compilation C 2007 Foundation Acta Pædiatrica/Acta Pædiatrica 2007 96, pp. 19–25

19

Danielzik et al.

Prevention of childhood overweight

Between 1996 and 2005 we investigated 4997 or 41% (cohort T0) and 4487 or 37% (cohort T1) of all 6- (first graders) and 10-year-old children (fourth graders) in Kiel (North Germany), respectively. There were no eligibility criteria except willingness to participate. All parents gave their informed written consent. The study was approved by the local ethical committee. Recruiting was carried out through health examinations by the school physicians. The KOPS cohorts T0 and T1 were found to be representative for all 6- and 10-year-old children in Kiel (10). One thousand seven hundred and sixtfour of the T0-children (35%) could be re-investigated at the age of 10 years. When compared with the total T0 population this subgroup of children did not differ in mean BMI, TSF, fat mass (according to BIA measurements) and waist circumference but the prevalence of children from high SES families was lower (35.7% vs. 43.9%). School-based intervention was performed between 1996 and 2001 in 18 ‘intervention schools’ in Kiel (three schools per year). These schools were randomly chosen from the 32 primary schools in Kiel. Seven hundred and eighty first graders (of the KOPS cohort T0), i.e. 6–7 years, underwent a 6 h curriculum of nutrition and activity. After 4 years 344 of the ‘intervention children’ (I, 44%) were re-investigated. The data of these children were compared with ‘non-intervention children’ (NI, n = 1420). To analyze the effectiveness of the intervention the T1–T0-changes in variables were considered. Nutritional status was the primary outcome. All investigations were blinded, i.e. examiners couldn’t identify ‘intervention children’. Intervention The intervention program was based on the assumption that attitudes towards behavior and other factors (e.g. familiar disposition, low SES) are predictive of weight change. Within KOPS the same behavioral and educational messages were given to all children, parents and teachers. These were (i) eat fruit and vegetables every day, (ii) reduce intake of high fat foods, (iii) keep active for at least 1 h a day and (iv) decrease TV consumption to less than 1 h per day. These messages were delivered to children within their first year at school by a 6 h course of nutrition education followed by 20 min of so-called ‘active breaks’. The intervention was offered by a skilled nutritionist together with a teacher. The messages were also given to parents on the occasion of a parental school meeting as well as to teachers who were trained regularly in a half day structured nutrition education program as part of a continuation class. Outcome parameters Nutritional status was assessed by anthropometric measurements. Methodological details for measurement of weight, height and skinfolds have been described previously (8,11). Waist circumference (WC) was measured midway between the lowest rib and the top of the iliac crest at the end of gentle expiration. BMI, triceps skinfold (TSF) and WC were used to characterize nutritional status independently. (1) Actual German BMI percentiles (12) (= ≥90th percentile), (2) German reference values for triceps skinfold from 1980

20

(13) (= ≥90th percentile) and (3) Dutch reference values for WC (14) (SDS > 1.3) have been used for classification of overweight to calculate prevalence, 4-year cumulative incidence and remission of overweight. Time between T0 and T1 was 3.8 ± 0.4 years. T1 data were corrected to 4 years. Process evaluation Feasibility of the intervention was tested in three different populations of children taking into account the possible effect of time as well as the school setting. Changes in nutritional knowledge were measured by a questionnaire in Kiel on two different occasions (in 1996 and 2004) in 189 and 175 children before as well as 6 weeks and 3 months after the intervention. Additionally the KOPS-project was conducted in a second German city, Osnabrueck (165 000 inhabitants, North Germany) in 2004 to assess feasibility of the program in another setting (the KOPS conducted in Kiel is called KOPS 1, whereas the program conducted in Osnabrueck is called KOPS 2). Because the outcome analysis of KOPS 2 was a 1-year follow-up, only the process evaluation of KOPS 2 is described. KOPS 2 was performed in 195 6-year-old children before and 3 month after intervention and assessed using the same questionnaire as in KOPS 1. In addition, 11 and 70 teachers in KOPS 1 and KOPS 2 were asked if they thought the intervention was feasible as part of a school curriculum. Statistics Statistical analysis was performed with SPSS 13.0 for Windows (SPSS Inc., Chicago, IL, USA). Results were presented as median and interquartile range (IQR). The nonparametric Mann–Whitney U-test was used to determine between-group differences. Pearson correlation coefficient was calculated to assess the relation between BMI, TSF and WC. The chi-square test compared prevalence, incidence and remission of overweight. Logistic regression models were used to compute Odds Ratios (OR). As dependent variable we considered prevalence of overweight versus prevalence of normal weight at T1. Incidence was calculated from 4-year development of children who were normal weight at 6 years. Remission was calculated as number of overweight children at age 6 years who became normal weight at age 10 years. Separate regression analyses were performed for boys and girls. The dichotomous grouping variable (intervention yes–no) was the independent variable. ORs were adjusted for age, baseline BMI, TSF or WC of children as well as SES and weight status of mothers because these variables were found to be the major determinants of childhood overweight (11). An OR > 1 indicated that intervention is related to overweight; an OR < 1 implied that intervention has a protective effect. By contrast, an OR > 1 reflected increased remission. Level of significance was set at p < 0.05 (two-sided). RESULTS Study population Fourty-nine and 50% of the children in NI and I were boys, respectively. The distribution of SES did not differ between


Danielzik et al.


Table 1 Parameters of nutritional status of the non-intervention and intervention children before (T0) and after the 4-years observation (T1) period (data are given as median [interquartile range]) Non-intervention group (NI) (n = 1420)

Age (years) Height (m) Weight (kg) BMI (kg/m2 ) TSF2 (mm) WC3 (cm)

Intervention group (I) (n = 344)

T0

T1

(T1–T0)

T0

T1

(T1-T0)

6.3 (6.0–6.5) 1.20 (1.17–1.24) 22.0 (20.4–24.5) 15.4∗ (14.6–16.4) 10.3 (9.0–13.0) 54.0 (52.0–57.6)

10.2 (10.0–10.5) 1.44 (1.39–1.49) 35.7 (31.7–42.0) 17.2 (15.8–19.6) 14.3 (10.1–19.4) 62.7 (58.7–68.3)

4.0 (0) 0.24 (0.22–0.26) 13.2 (10.7–17.5) 1.8 (0.9–3.3) 3.6 (0.0–7.3) 8.1 (5.0–12.3)

6.3 (5.9–6.5) 1.20 (1.16–1.23) 22.5 (20.5–24.5) 15.6∗ (14.8–16.7) 10.7 (9.0–13.9) 55.0 (52.0–58.0)

10.3 (9.9–10.5) 1.44 (1.40–1.48) 36.1 (31.7–41.2) 17.5 (16.0–19.1) 13.7 (10.2–18.5) 63.1 (59.5–67.8)

4.0 (0) 0.24 (0.22–0.26) 13.5 (10.7–17.1) 1.7 (0.7–3.1) 2.9 (−0.5–7.2) 8.3 (4.7–12.4)

∗

NI and I: 26.6% and 26.7% of children from NI and I were from low SES families, 30.2% and 26.5% from middle SES families and 43.2% and 46.8% were from high SES families, respectively. Table 1 shows parameters of nutritional status of NI and I as well as the 4-year changes in these parameters. There were no group differences nor at T0 or T1 with the exception of mean BMI at T0 which was lower in NI. Process evaluation Figure 1 shows the prevalence of children with good nutritional knowledge before and after intervention in KOPS 1 in 1996 and 2004 as well as in KOPS 2 in 2004. Before intervention the prevalence of children with good knowledge was the same in KOPS 1 2004 and KOPS 2 2004 but it was lower in KOPS 1 in 1996. The increase in nutritional knowledge after invention was similar in KOPS 1 in both years (+46% in 1996 and +49% in 2004) but it was lower in KOPS 2 (+20%). 100% and 97% of the teachers in KOPS 1 in 1996 and KOPS 2 in 2004 believed that the intervention program was feasible as part of a school curriculum. Outcome evaluation Comparison of parameters to define overweight The coefficients of correlation at T0 and T1 were 0.614 and 0.753 between BMI and TSF, 0.769 and 0.852 for BMI and WC and 0.523 and 0.673 for TSF and WC, respectively (p < 0.001 for all). The prevalence of overweight was different depending on its definition: it was lowest for BMI and highest for TSF (Table 2). The agreement in classification of children in over and normal weight according to all three parameters was 76.4% at T0 and 68.0% at T1. The agreement of two parameters at T0 and T1 were 84.0 and 73.1% for BMI and TSF, 87.7 and 86.6% for BMI and WC and 81.2 and 76.2% for TSF and WC, respectively.

good nutritional knowledge [%]

Significant difference between non-intervention and intervention group; Mann–Whitney test (p < 0.05); SES, socioeconomic status.; TSF, triceps skinfiold; WC, waist circumference.

100 90 80 70 60 50 40 30 20 10 0

before after

KOPS 1 (1996)

KOPS 1 (2004)

KOPS 2 (2004)

Figure 1 Prevalence of children with good nutritional knowledge before and 3 months after intervention in two occasions in KOPS 1 in 1996 and 2004 as well as in KOPS 2 in 2004

4-year changes in prevalence, incidence and remission The 4-year increase in prevalence of overweight was higher in NI when compared with I (ns). This effect was independent of the parameters used to define overweight. Girls were more often overweight when compared with boys (particularly with definition according to WC). The effect of intervention on the 4-year increase of overweight was stronger in girls, also particularly with definition according to WC. Cumulative 4-year incidence was lower in I when compared with NI (Table 3, ns). The difference in incidence between NI and I was similar according to TSF and WC but higher when compared with that of BMI. The effect of intervention was stronger in girls for definition according to BMI and TSF. Cumulative 4-year remission was higher in I when compared with NI and reached significance for TSF (Table 3). In girls the effect of intervention was higher and additional to the definition according to TSF, that of WC was significant.


21

Danielzik et al.


Table 2 Prevalence of overweight + obesity at T0 and T1 in non-intervention and intervention children with crude and adjusted Odds Ratio for prevalence at T1 (OR, 95% CI) (OR of NI was taken as 1) T0 Prevalence of overweight at T1 All

BMI

T1 – T0

T1

NI (%)

I (%)

NI (%)

I (%)

NI

I

9.1

10.8

16.3

15.4

+7.2

+4.6

20.2

24.4

42.2

37.8

+22.0

+13.4

18.5

19.0

27.5

23.9

+9.0

+4.9

7.5

10.4

15.4

16.2

+7.9

+5.8

19.3

23.1

41.1

38.7

+21.8

+15.6

16.7

15.6

25.9

23.7

+9.2

+8.1

10.6

11.1

17.1

14.6

+6.5

+3.5

21.0

25.7

43.2

36.8

+22.2

+11.1

20.2

22.4

29.1

24.1

+8.9

+1.7

†

TSF‡

WC‡

Boys

BMI†

TSF‡

WC‡

Girls

BMI†

TSF‡

WC‡

Crude OR (95% CI), p-value

Adjusted OR (95% CI), p-value

0.94 (0.68–1.30) p = 0.6968 0.83 (0.65–1.06) p = 0.1380 0.83 (0.63–1.09) p = 0.1776 1.06 (0.67–1.67) p = 0.8091 0.90 (0.64–1.27) p = 0.5659 0.89 (0.60–1.31) p = 0.5583 0.83 (0.52–1.33) p = 0.4422 0.77 (0.54–1.08) p = 0.1308 0.78 (0.53–1.14) p = 0.1975

0.70 (0.46–1.08) p = 0.1051 0.79 (0.80–1.03) p = 0.0824 0.73 (0.52–1.01) p = 0.0588 0.65 (0.35–1.20) p = 0.01669 0.86 (0.58–1.26) p = 0.4303 0.72 (0.45–1.14) p = 0.1613 0.72 (0.39–1.34) p = 0.3052 0.70 (0.48–1.04) p = 0.0744 0.74 (0.46–1.17) p = 0.1994

Definition of overweight according to † (12), ‡ (13) and ‡ (14). OR were adjusted for age, SES, BMI of the mother and baseline BMI, TSF or WC of the children.

DISCUSSION Process evaluation In the year 2004 before intervention about 50% of the children had a good nutritional knowledge, whereas only 20% of the children reached this result 8 years earlier (see Fig. 1). The ‘spontaneous’ increase in nutritional knowledge in KOPS 1 was most likely explained by ongoing KOPS 1 activities, which were conducted within these years and were accompanied by several reports on KOPS and its key messages in local newspapers. However, the 2004 data of KOPS 2 also showed a similarly high rate of children with good nutritional knowledge before intervention (50%). This data may indicate that the increase in good nutritional knowledge was a result of changes in our society. In 2004, children are more competent in nutrition than 8 years earlier (and it is likely that the parents improved their nutritional knowledge too). Although children increased knowledge, this does obviously not tackle the obesity endemic (the prevalence of overweight and obesity further increased in NI). With respect to the intervention itself, similar increases in knowledge were observed in 1996 and 2004. The effect of intervention (i.e. an increase of about +45%) was independent of initial nutritional knowledge. However, when compared with KOPS 1 the increase in knowledge was lower in KOPS

22

2. This difference is most likely explained by the higher rate of non-German children in the intervention schools in KOPS 2 (19.5% in KOPS 2 vs. 0.4% in KOPS 1l). In fact, subgroup analysis revealed that in KOPS 2 the German children had the same benefit from the intervention as the children followed as part of KOPS 1. Outcome evaluation The two main findings of our study were: (i) school-based health promotion had long-term effects on overweight, but these effects were seen in remission of overweight for girls only. By contrast, no effects on prevalence and incidence were seen. (ii) The effect of intervention was also influenced by the parameter used to define overweight. Ad (i): The majority of studies on school-based health promotion reported statistically significant effects on health knowledge, attitudes and behavior (3). By contrast only modest changes in nutritional status were observed within ‘real world’ school settings (15). It can be questioned why school-based interventions had only minor effects on prevalence of overweight: perhaps the interventions were not intensive enough or they were not focussed on certain and simple key messages (e.g. eliminate soft drink consumption). However high intensive interventions did


Danielzik et al.


Table 3 Cumulative 4-year incidence and remission of overweight + obesity with crude and adjusted odds ratios (OR, 95% CI) (OR of NI was taken as 1) Incidence All

NI (%)

I (%)

I – NI (%)

10.0

8.8

−1.2

33.3

30.4

−2.9

18.3

15.5

−2.8

10.6

9.7

−0.9

32.6

31.6

−1.0

18.1

14.4

−3.7

9.4

7.9

−1.5

34.0

29.1

−4.9

18.5

16.7

−1.8

20.9

29.7

+8.8

22.6

39.3

+16.7

31.9

40.0

+8.1

25.0

27.8

+2.8

23.1

37.5

+12.5

35.3

25.9

−9.4

18.2

31.6

+13.4

22.2

40.9

+18.7

29.3

50.0

+20.7

†

BMI

TSF‡

WC‡

Boys

BMI†

TSF‡

WC‡

Girls

BMI†

TSF‡

WC‡

Remission All

BMI†

TSF‡

WC‡

Boys

BMI†

TSF‡

WC‡

Girls

BMI†

TSF‡

WC‡

Crude OR (95% CI), p-value

Adjusted OR (95% CI), p-value

0.87 (0.56–1.34) p = 0.5255 0.88 (0.65–1.17) p = 0.3707 0.82 (0.57–1.17) p = 0.2734 0.90 (0.50–1.63) p = 0.7337 0.96 (0.64–1.44) p = 0.8283 0.76 (0.46–1.27) p = 0.2967 0.83 (0.43–1.58) p = 0.5660 0.80 (0.53-1.22) p = 0.2950 0.88 (0.53–1.46) p = 0.6261

0.76 (0.47–1.23) p = 0.2722 0.92 (0.68–1.25) p = 0.5957 0.76 (0.52–1.11) p = 0.1561 0.63 (0.32–1.25) p = 0.1793 1.00 (0.65–1.54) p = 0.9899 0.61 (0.35–1.04) p = 0.0705 0.88 (0.44–1.76) p = 0.7183 0.81 (0.52-1.27) p = 0.3596 1.01 (0.58–1.75) p = 0.9739

1.60 (0.70–3.64) p = 0.2640 2.21∗ (1.32–3.71) p = 0.0027 1.42 (0.81–2.49) p = 0.2190 1.15 (0.35–3.86) p = 0.8163 1.99 (0.94–4.24) p = 0.0735 0.64 (0.25–1.64) p = 0.3531 2.08 (0.67–6.41) p = 0.2039 2.42∗ (1.19–4.94) p = 0.0148 2.42∗ (1.17–5.01) p = 0.0175

1.84 (0.77–4.38) p = 0.1686 2.09∗ (1.20–3.62) p = 0.0087 1.54 (0.83–2.87) p = 0.1707 1.34 (0.34–5.23) p = 0.6750 1.89 (0.84–4.28) p = 0.1266 0.78 (0.26–2.36) p = 0.6584 2.46 (0.76–8.08) p = 0.1350 2.27∗ (1.06–4.85) p = 0.0339 2.45∗ (1.13–5.32) p = 0.0235

Definition of overweight according to † (12), ‡ (13) and ‡ (14). OR were adjusted for age, SES, BMI of the mother and baseline BMI, TSF or WC of the children.

not affect overweight; e.g. a 3-year school-based multicomponent intervention (with a classroom curriculum, food service, physical education together with a family intervention, strong support from tribal, educational and community

authorities) for reducing percentage body fat in American Indian elementary schoolchildren (i.e. Pathways) could not reduce the prevalence of overweight (16,17). In addition concentrating on simple key messages did not show unique


23

Danielzik et al.


successful outcomes, e.g. a cluster randomized controlled trial which aimed to reduce consumption of carbonated drinks in 7–11-year-old children were able to reduce daily consumption of carbonated drinks and additionally decreased prevalence of overweight after 12 months (18). In this trial all three sessions of the intervention focussed on the key message (carbonated drinks). By contrast a Norwegian study to increase fruit and vegetable consumption in children which also concentrated on the key message could not influence the consumption of fruit and vegetables after a 1-year intervention at school (19). Altogether, when compared with KOPS, this data suggests that more intensive or more focussed interventions have no better effectiveness. Lifestyle interventions implemented were not capable of stopping the obesity epidemic. Thus additional measures focussing on public health interventions are necessary (20). The public-health crisis demands increased funding for research into new dietary, physical activity, behavioral and environmental approaches for prevention of obesity (2). Ad (ii): Measurements of overweight in children take into account different methods. A primary purpose for defining overweight is to predict health risk. In field studies, anthropometric measurements (like BMI, skinfold thickness and waist circumference) have been most widely used (5). However, BMI fails to distinguish between muscle and fat mass but is a good marker of subcutaneous fat mass. By contrast WC is a good predictor of abdominal adiposity (21). Thus different groups of children were classified as overweight when BMI, TSF or WC were used to define nutritional status. Although 76% of the 6 year old children were uniformly classified as normal- or overweight according to all three parameters. At the moment the impact of individual parameter on health risk is not characterized in children. Because central fat mass is a risk factor for later metabolic and heart disease also in children (14) WC may be a more suitable method to define overweight when compared with BMI. A second argument against the use of BMI is the fact that when comparing BMI and WC, the secular increase in WC greatly exceeded that of BMI suggesting that the central accumulation of body fat has risen more steeply than total body mass derived from height and weight (22,23). Our data showed that using different parameters to characterize overweight affected the prevalence of overweight as well as the effect of intervention. The strongest effect was seen using TSF and WC (Table 3). However trends in overweight were independent of the method used but there were differences in the effect sizes. Our data suggests the use of BMI, skinfold thickness and WC to fully characterize the outcome of preventive studies.

ACKNOWLEDGEMENT Supported by: Deutsche Forschungsgemeinschaft (DFG Mu¨ 5.1, 5.2, 5.3 und 5.5), WCRF, London, Wirtschaftliche Vereinigung Zucker, Bonn, Danone Stiftung, Munchen, DAK, ¨ Hamburg. Contributors’ list: Dr. S. Danielzik: data acquisition, data analyses, concept, discussing the data and writing the

24

manuscript. S. Pust: data acquisition, processing data and process evaluation. Prof. Dr. M.J. Muller: study design, ¨ supervision of the study, discussing data and writing the manuscript.

References 1. Kumanyika S, Jeffery RW, Morabia A, Ritenbaugh C, Antipatis VJ. Obesity prevention: the case for action. Int J Obes 2002; 26: 425–36. 2. Ebbeling CB, Pawlak DB, Ludwig DS. Childhood obesity: public health crisis, common sense cure. Lancet 2002; 360: 473–82. 3. Summerbell CD, Waters E, Edmunds LD, Kelly S, Brown T, Campbell KJ. Interventions for preventing obesity in children (Review). The Cochrane Database Syst Rev 2005; 3: CD001871. 4. US Department of Health and Human Services. Healthy People 2010. Understanding and Improving Health. Objectives for Improving Health. Washington, DC: US Government Printing Office; 2000. 5. Report of the World Health Organisation. Obesity in children and young people: A crisis in public health (T Lobstein, L Baur, R Uauy for the IASO IOTF, Eds.) obesity reviews 2004; 5(suppl. 1): 1–104. 6. Cole TJ, Bellizzi MC, Flegal KM, Dietz WH. Establishing a standard definition for child overweight and obesity worldwide: international survey. BMJ 2000; 320: 1–6. 7. Seidell JC. Obesity: a growing problem. Acta Paediatr 1999; 88(suppl.): 46–50. 8. Muller MJ, Asbeck I, Mast M, et al. Prevention of obesity-more ¨ than an intention. Concept and first results of the Kiel Obesity Prevention study (KOPS). Int J Obes 2001; 25(Suppl.1): S66–S74. 9. Muller MJ, Mast M, Asbeck I, Langnase ¨ ¨ K, Grund A. Prevention of obesity – is it possible? Obesity reviews 2001; 2: 15–28. 10. Danielzik S, Pust S, Landsberg B, Muller MJ. First lessons of ¨ the Kiel Obesity Prevention Study (KOPS). Int J Obes 2005, 29(suppl. 2): S78–S83. 11. Danielzik S, Czerwinski-Mast M, Langnase ¨ K, et al. Parental overweight, socioeconomic status and high birth weight are major determinants of overweight and obesity in 5-7 yr old children. Baseline data of the Kiel Obesity Prevention Study (KOPS). Int J Obes 2004; 28: 1494–1502. 12. Kromeyer-Hauschild K, Wabitsch M, Kunze D, Geller F, Geiß HC, Hesse V, von Hippel A, Jaeger U, Johnson D, Korte W, Menner K, Muller G, Muller MJ, Niemann-Pilatus A, Remer T, ¨ ¨ Schaefer F, Wittchen H-U, Zabransky S, Zellner K, Ziegler A, Hebebrand J. Perzentile fur ¨ den Body Mass Index fur ¨ das Kindes- und Jugendalter unter Heranziehung verschiedener deutscher Stichproben. Monatsschrift Kinderheilkunde 2001; 149: 807–18. 13. Reinken L, Stolley H, Droese W, van Oost G. Longitudinale Korperentwicklung gesunder Kinder. II. Große, Gewicht, ¨ ¨ Hautfalten von Kindern im Alter von 1,5 bis 16 Jahren. Klein Padiat 1980; 192: 25–33. ¨ 14. Fredriks AM, van Buuren S, Fekkes M, Verloove-Vanhorick SP, Wit JM. Are age references for waist circumference, hip circumference and waist-hip ratio in Dutch children useful in clinical practice? Eu J Pediatr 2005; 164: 216–22. 15. Haymann LL, Williams C, Daniels SR, et al. Cardiovascular Health Promotion in the schools. A statement for health education professionals and child health advocates from the Committee on Atherosclerosis, Hypertension, and Obesity in


Danielzik et al.

16.

17.

18.

19.

Youth (AHOY) and the Council on Cardiovascular Disease in the Young, American Heart Association. Circulation 2004; 110: 2266–75. Caballero B, Davis S, Davis CE, et al. Pathways: A school-based program for the primary prevention of obesity in American Indian children. J Nutr Biochem 1998; 9: 535–43. Caballero B, Clay T, Davis SM, et al. Pathways: a school-based, randomized controlled trial for the prevention of obesity in American Indian schoolchildren. Am J Clin Nutr 2003; 78: 1030–8. James J, Thomas P, Cavan D, Kerr D. Preventing childhood obesity by reducing consumption of carbonated drinks: cluster randomised controlled trial. BMJ 2004; 328: 1237–42. Bere E, Veierod MB, Bjelland M, Klepp K-I. Outcome and process evaluation of a Norwegian school-randomized fruit


20. 21.

22.

23.

and vegetable intervention: Fruits and Vegetables Make the Marks (FVMM). Health Educ Res 2005 Oct 11 Jain A. Treating obesity in individuals and populations. BMJ 2005; 331: 1387–90. Brambilla P, Bedogni G, Moreno LA, Goran MI, Gutin B, Fox KR, Peters DM, Barbeau P, De Simone M, Pietrobelli A. Crossvalidation of anthropometry against magnetic resonance imaging for the assessment of visceral and subcutaneous adipose tissue in children. Int J Obes 2006; 30: 23–30. McCarthy HD, Ellis SM, Cole TJ. Central overweight and obesity in British youth aged 11-16 years: cross sectional surveys of waist circumference. BMJ 2003; 326: 624. Moreno LA, Pineda I, Rodriguez G, Fleta J, Sarria A, Bueno M. Waist circumference for the screening of the metabolic syndrome in children. Acta Pediatr 2002; 91: 1307–12.


25