Hostname: page-component-7c8c6479df-r7xzm Total loading time: 0 Render date: 2024-03-29T05:26:14.882Z Has data issue: false hasContentIssue false

Positive impact of a pre-school-based nutritional intervention on children's fruit and vegetable intake: results of a cluster-randomized trial

Published online by Cambridge University Press:  23 August 2011

Freia De Bock*
Affiliation:
Competence Center for Social Medicine and Occupational Health Promotion, Mannheim Institute of Public Health, Social and Preventive Medicine, University Medicine Mannheim, Heidelberg University, Ludolf-Krehl-Strasse 7–11, D-68167 Mannheim, Germany Children's Hospital, University Medicine Mannheim, Heidelberg University, Mannheim, Germany
Luise Breitenstein
Affiliation:
Competence Center for Social Medicine and Occupational Health Promotion, Mannheim Institute of Public Health, Social and Preventive Medicine, University Medicine Mannheim, Heidelberg University, Ludolf-Krehl-Strasse 7–11, D-68167 Mannheim, Germany
Joachim E Fischer
Affiliation:
Competence Center for Social Medicine and Occupational Health Promotion, Mannheim Institute of Public Health, Social and Preventive Medicine, University Medicine Mannheim, Heidelberg University, Ludolf-Krehl-Strasse 7–11, D-68167 Mannheim, Germany
*
*Corresponding author: Email freia.debock@medma.uni-heidelberg.de
Rights & Permissions [Opens in a new window]

Abstract

Objective

To assess the short-term impact of a nutritional intervention aimed at reducing childhood overweight in German pre-school children.

Design

Using a cluster-randomized study design with waiting-list controls, we tested a 6-month intervention administered once weekly by a nutrition expert consisting of joint meal preparation and activities for children and parents such as tasting and preparing nutritious, fresh foods. At baseline, 6 and 12 months, a parent-completed questionnaire assessed fruit and vegetable intakes (primary outcomes) and water and sugared drinks consumption (secondary outcomes). Direct measurement assessed BMI, skinfold thickness and waist-to-height-ratio. An intention-to-treat analysis used random-effects panel regression models to assess the intervention effect, adjusted for each child's age, gender, immigrant background and maternal education.

Setting

Eighteen pre-schools from three south German regions.

Subjects

Healthy children aged 3–6 years.

Results

Three hundred and seventy-seven (80 %) eligible pre-school children participated in the study. Of these, 348 provided sufficient data for analysis. The sample mean age was 4·26 (sd 0·78) years; the majority (53·2 %) were boys. Children's fruit and vegetable intakes increased significantly (P < 0·001 and P < 0·05, respectively); no significant changes in the consumption of water, sugared drinks or anthropometric measurements were noted.

Conclusions

Nutritional interventions in pre-schools have the potential to change eating behaviours in young children, which in the long term might reduce risk for developing overweight.

Type
Research paper
Copyright
Copyright © The Authors 2011

The prevalence of overweight in children has emerged as a public health issue of increasing importance in both developed and developing countries(Reference Wilkinson1). In Europe, for example, the prevalence of overweight in children currently is estimated to be 20 %, one-third of whom are obese(Reference Troiano and Flegal2, Reference Branca, Nikogosian and Lobstein3). Childhood overweight is defined as BMI equal to or greater than the 90th reference percentile in Europe (International Obesity Taskforce criteria(Reference Dietz and Bellizzi4)) or the 95th percentile in the USA (http://www.cdc.gov). Given the high prevalence and the known long-term effects of childhood overweight such as dyslipidaemia, hyperinsulinaemia, hypertension(Reference Reilly, Methven and McDowell5Reference Freedman, Dietz and Srinivasan7), persistent overweight in adulthood(Reference Friel, Kelleher and Campbell8Reference Must11) as well as its social and economic consequences(Reference Gortmaker, Must and Perrin12Reference Sargent and Blanchflower15), an urgent need for effective interventions early in childhood has become evident.

Factors that influence a child's weight are partly genetic and partly environmental(Reference Birch16, Reference Wardle and Cooke17). In industrialized societies, environmental factors such as high energy intake paired with physical inactivity can lead to an energy surplus and increased body fat(Reference Anderson and Butcher18). Nutritional surveys from several European countries show consumption trends among 3-year-old children exceeding recommendations for high-energy foods and beverages and failure to meet recommendations for fruits and vegetables(Reference Mensink, Kleiser and Richter19, Reference Yngve, Wolf and Poortvliet20). Changing children's eating behaviour is therefore potentially useful for preventing childhood overweight.

Previous work suggests that eating behaviour in early childhood might be changed more easily while these behavioural patterns are still developing(Reference Birch16, Reference Zwiauer21, Reference Oerter22). The pre-school setting may be especially appropriate for behavioural interventions(Reference Rummel23), as it is among the first in which young children encounter behavioural norms different from those in the home. Given high pre-school attendance rates in several industrialized countries(24, 25), nutritional interventions conducted in these settings could positively impact the health of many children(Reference Story, Kaphingst and French26).

Although few published reports of interventions in this area exist, those that do suggest the potential for successfully promoting healthy nutrition in pre-school settings. A questionnaire-based study of 4–6-year-old pre-school children in China, for example, observed a significant decrease in unhealthy dietary behaviours(Reference Hu, Ye and Li27), while a Canadian pre-school programme reported significant increases in the frequency of fruit and vegetable intake(Reference Cason28). Other studies conducted in primary and secondary schools, involving older children, provide further evidence that eating behaviour can be effectively modified through school-based nutritional interventions(Reference Friel, Kelleher and Campbell8, Reference Muth, Chatterjee and Williams29, Reference Anderson, Porteous and Foster30).

Most of these interventions primarily used educational strategies aimed at rising awareness in children(Reference Muth, Chatterjee and Williams29), parents and teachers. Based on theoretical frameworks of modelling(Reference Bandura31) and the general system theory(Reference Bertalanffy32), however, involvement of parents, teachers and peers in a participatory way – reaching beyond mere education – might be most effective.

While use of anthropometric outcomes has been infrequently reported in previous evaluations, a single study observed no significant differences in body weight and height-for-age(Reference Hu, Ye and Li27). Previous work, however, suggests that higher BMI in children is associated with physiological risk factors and even premature death(Reference Franks, Hanson and Knowler33). Therefore, it appears important that nutritional intervention studies incorporate both intermediate behavioural and anthropometric outcomes measures. The aim of the present study was thus to assess the effects of a pre-school-based nutritional intervention on both behavioural outcomes, like children's fruit, vegetable and water consumption, and anthropometric measures.

Methods

Setting and participants

The study was set in Baden-Württemberg, a federal state (population nearly 11 million) in south-west Germany with an extensive network of urban and rural pre-schools for 3–6–year-old children, cared for by certified teachers.

Pre-schools were eligible to participate in the study if they were located in one of three predefined regions and had applied to participate in the nutritional intervention module of a state-sponsored health promotion programme ‘Komm mit in das gesunde Boot’ (‘Come aboard the health boat’), with at least fifteen children participating. This programme comprised of two distinct arms was initiated in 2006 to encourage (i) healthy eating behaviour and (ii) physical activity(Reference De Bock, Fischer and Hoffmann34) among pre-school children, with the long-term goal of reducing childhood overweight. The current paper focuses on the nutritional intervention. Analyses of the physical activity module's effectiveness are conducted separately with interactions between both modules not assessed.

Twenty-three pre-schools (mean size: forty-eight children) met the above-mentioned criteria with eighteen (78 %) agreeing to participate (see Fig. 1), either with all children (in small pre-schools) or the older children (in larger schools), due to a maximum group size of twenty students per expert. The vast majority of parents (> 95 %) consented to their child participating in the programme if the pre-schools had scheduled participation. Children between 3 and 6 years of age attending one of the participating pre-schools and participating in the programme (n 473) were considered eligible for our study. A total of 377 pre-school children (80 %) were recruited in Autumn 2008 following written informed consent provided by each child's parents; at least one outcome measurement was obtained on 348 (93 %). The study was approved by the ethical committee of Heidelberg University, Mannheim Medical Faculty.

Fig. 1 (colour online) Enrolment, stratification, randomized allocation of pre-schools and resulting number of children in each arm: cluster-randomized trial of a pre-school-based nutritional intervention on children's fruit and vegetable intakes, south Germany

Study design

Our study used a cluster-randomized trial design, with natural pre-school-bound clusters of children randomized to either the intervention or the waiting-list control arm(Reference De Bock, Fischer and Hoffmann34). Pre-schools in the waiting-list control arm (n pre-schools 8, n children 183) received the same intervention 6 months later than the intervention arm (n pre-schools 10, n children 194). To account for seasonal differences in fruit and vegetable intake, the study was implemented with half of the pre-schools starting in Autumn 2008, the other half in Spring 2009.

As previous research in children has shown differences in eating behaviour(Reference Kolip35) and BMI by socio-economic status (SES)(Reference Kleiser, Schaffrath Rosario and Mensink36), we stratified the recruited pre-schools before randomization to balance aggregate pre-school social background and immigrant proportion (see De Bock et al.(Reference De Bock, Fischer and Hoffmann34)).

Intervention – theoretical framework and delivery

Nutritional interventions in older children have chosen various theoretical frameworks, yet many are based upon the assumptions of Bandura's(Reference Bandura31) social learning theory(Reference Lowe, Horne and Tapper37Reference Horne, Tapper and Lowe39) or Zajonc's(Reference Zajonc40) exposure effect(Reference Perry, Mullis and Maile41Reference Sullivan and Birch43). The theory of social learning suggests that children's eating behaviour evolves from imitating the behaviour they observe in significant others, like parents and peers(Reference Wardle and Cooke17). Thus, for an intervention to be most effective, it may be useful to include parents to modify children's behaviour(Reference Lindsay, Sussner and Kim44). Zajonc's exposure effect(Reference Zajonc40) suggests that even very young children are more likely to eat specific foods when these are repeatedly and consistently offered to them(Reference Wardle and Cooke17, Reference Birch, McPhee and Shoba45, Reference Cooke46). Our nutritional intervention incorporated elements from both theoretical frameworks.

The intervention was delivered by external nutrition experts who participated in intensive 4 d training sessions covering specific nutritional guidelines for children and basic communication skills. The experts delivered the intervention during fifteen standardized 2 h nutrition sessions, mostly during pre-school hours, conducted once weekly over a 6-month period with five of these sessions actively involving parents by targeting them alone (discussions on parents’ modelling role and nutritional needs of children) or together with their children (Table 1). Intervention activities consisted of familiarizing with different food types and preparation methods as well as cooking and eating meals together in groups of children, teachers and parents. One session additionally focused on healthy drinking behaviours. Pre-school group teachers assisted the external nutrition expert during each session to enable them to sustain intervention-related activities after the study end.

Table 1 The nutritional intervention, consisting of fifteen 2 h sessions once weekly over a period of 6 months. Ten modules only targeted children, another five parents and children or parents exclusively

Models for healthy eating within the intervention included: (i) use of nutrition experts; (ii) play acting with ‘pirate dolls’ used as props enjoying fruit and vegetables; (iii) active parental involvement; and (iv) involvement of other pre-school peers. The exposure effect was taken into account by repeatedly offering healthy snacks like fruit and vegetables and water to the children every week.

Process evaluation

Evaluating an intervention's effectiveness and processes at the same time may aid in understanding why an intervention was effective. Our process evaluation was based upon telephone interviews with the nutrition experts directly involved in intervention delivery. Our assessment used elements(Reference Glasgow, Edwards and Whitesides47) of the RE-AIM framework, specifically focusing on reach (proportion of target population participating in intervention), effectiveness and implementation (e.g. fidelity to or adaptation of the intervention schedule and its components and the effort required to organize intervention-related activities within each pre-school setting). Reach was reflected by the proportion of participating children and parents compared with the total number of targeted, eligible and recruited children and parents. Implementation ratings adapted from Glasgow et al.(Reference Glasgow, Edwards and Whitesides47) on fidelity to planned intervention delivery, total intervention content covered by the experts, effort to get activities organized and the level of parental commitment were given on an ordinal scale ranging from 0 = not at all/none to 5 = completely/a great deal, separately for parent and children's modules, based upon ratings by the nutrition experts.

Measurements

Quantitative outcome measurements were obtained before start (baseline) and at 6 and 12 months (sustainability) after baseline. Given a cross-over study design, pre-schools assigned to the waiting-list control arm underwent one additional measurement at 6 months before intervention start. Our choice of outcomes combines both intermediate behavioural and objective health parameters to grasp possible translation of behaviour change into anthropometric changes.

Parental questionnaire

The study questionnaire, completed by the child's parents, assessed multiple domains of behaviour including children's eating behaviour and physical activity. Wherever possible, we used items from existing surveys; given limited previous work in this area, however, it was necessary to adapt or create new items to assess specific outcomes. Given the potential inaccuracy of extensive food inventories completed by young children or by parents unable to observe their child during the course of the day, we relied instead on parental report of consumption of specific foods or beverages that reflect general eating behaviour. For the assessment of our primary outcomes, fruit and vegetable intakes, we used items developed by Bayer et al.(Reference Bayer, von Kries and Strauss48): ‘How many portions (size of a child's hand) of fruit does your child eat at average per day?’ and ‘How many portions (size of a child's hand) of vegetables does your child eat at average per day?’ Responses were given on a 6-point ordinal scale ranging from ‘none’ to ‘more than three portions per day’.

Secondary outcomes included the consumption of water (as assessed by an adaptation of an FFQ item: ‘How many [glasses] does your child drink per day (1 glass = 200 ml)?’) and sugared drinks (using the same item stem with sugared drinks defined as soft drinks, beverages containing juice concentrates, 100 % fruit juice and sugared tea/iced tea). Responses on a 6-point ordinal scale for both items ranged from ‘seldom or never’ to ‘four or more glasses per day’. After identifying consumption in each category, we created a single composite score of sugar-sweetened beverage consumption for each child.

Anthropometry and body composition

Trained study personnel performed all anthropometric measurements. Height (to the nearest 0·1 cm; model Seca 201, Seca Deutschland, Hamburg, Germany) and weight (to the nearest 0·1 kg; model Glaswaage Nr. 63685, Soehnle pharo, Nassau, Germany) were measured following a standardized protocol. BMI was calculated using the standard formula of weight (kg) divided by the square of height (m2). Overweight was defined by age- and gender-specific 90th percentile cut-offs of BMI taken from a representative sample of German pre-school children(Reference Kromeyer-Hauschild, Wabitsch and Kunze49).

Waist circumference was measured directly on the skin half-way between the top of the iliac crest and the lower rib(Reference Stolzenberg, Kahl and Bergmann50) with an accuracy of 0·5 cm using a non-elastic measuring tape (Seca Messband 201; Seca Deutschland) with waist-to-height-ratio (WTHR) calculated as a measure of central body fat(Reference Roswall, Bergman and Almqvist-Tangen51Reference Nambiar, Truby and Abbott53). Total body fat was determined by at least two averaged measurements at the triceps, biceps, subscapular, vertical hip and calf skinfolds using Lange Skin Fold Calipers (Beta Technology, Santa Cruz, CA, USA). These mean measures of all skinfolds (triceps to calf) were summed for each child to yield the outcome ‘skinfold sum’(Reference Freedman, Wang and Ogden54, Reference Eisenmann, Heelan and Welk55).

The quality of anthropometric measurements was maintained through biannual refresher trainings for study personnel and assessment of inter-rater reliability through duplicate measurement. The mean inter-rater correlation coefficient over all measures taken across all measurement time points was 0·87, suggesting a high level of reliability among study personnel, who were also blinded to group assignment.

Measurement of covariates and confounders

The analysis was adjusted for confounding factors at the child's level including age, gender(Reference Anderson, Economos and Must56), SES(Reference Muller, Koertzinger and Mast57) and immigrant background(Reference Kuepper-Nybelen, Lamerz and Bruning58). Sociodemographic information was collected through the parent-completed questionnaire. Immigrant background of each child was determined by parental response to three items: (i) child is a non-German national; (ii) child's parents are native speakers of a language other than German; and (iii) child primarily speaks a language other than German at home(Reference Schenk, Bau and Borde59). Responses (yes = 1; no = 0) were added to create a final value. The child's SES was estimated using his/her mother's self-report of the highest level of educational attainment as a proxy(Reference Ball, Cleland and Timperio60, Reference Semmler, Ashcroft and van Jaarsveld61). According to the international Standard Classification of Education (ISCED)(62), we defined three categories of maternal education: ‘high’ = ISCED level 5 or 6, ‘middle’ = ISCED level 4 and ‘low’ = ISCED level 1, 2 or 3.

Statistical analysis

The data set comprised up to three measurement points per child. Ordinal scale points included as outcome variables were treated as continuous. Descriptive statistical analysis was performed on all outcome measures and bivariate linear regression analyses reflected the intervention effect.

With 81 % of the children contributing more than one measurement to the data set, the effects of the intervention on the whole sample could be captured using a panel regression analysis. This technique can be useful when data consist of repeated observations in one or more cross-sectional units like individual children(Reference Allison63, Reference Halaby64). It allows analysing individual dynamics based upon within- and between-person differences in time, controlling for unobserved heterogeneity.

As our data stemmed from natural pre-school-bound clusters of children, we first determined the extent of clustering. Intraclass correlation coefficients (ICC) on the level of pre-schools were 0·016 and 0·014 for the primary outcomes of fruit intake and vegetable intake, respectively. With an average cluster size of 19·5 children per pre-school, the design effect () did not exceed 2, allowing us to ignore the issue of clustering in our analyses(65Reference Muthen and Satorra68).

Next, the panel regression models were built with the intervention effect expressed as change in ordinal scale points. In models for outcomes other than water intake, we were able to use random-effects panel regression since our estimators were unbiased, based on the Hausman(Reference Hausman69) specification test (χ 2 < 4·5; P > 0·1). Random-effects estimators generally are more efficient than fixed-effects estimators. Yet, for the analysis of water intake we used a fixed-effects panel regression as the random-effects estimators were biased (χ 2 = 6·60; P = 0·036). In our initial model, we used panel-robust estimators and adjusted the intervention effects to the time-variant factor age. A second model additionally adjusted for potential confounders including gender, immigrant background and maternal education. A supplemental analysis for the anthropometric outcome variables was performed separately in the subgroup of overweight children.

The study was powered on the primary outcome vegetable intake, for which we expected the smallest clinically relevant effect size, assumed to be 0·05 in the panel regression model. f 2 = 0·05 is traditionally considered a small to medium effect size(Reference Cohen70). With a power of 0·8, α = 0·05/3 = 0·0167 (adjusted for three measurement time points) and accounting for five independent predictors, we determined a minimal sample size of n 332. All statistical analyses were performed with the STATA 10 statistical software package (StataCorp LP, College Station, TX, USA).

Results

Of the 348 pre-school children (Table 2) providing data, 29·6 % (n 103) participated in measurements at all three time points, 51·4 % (n 179) children participated twice and 19·0 % (n 66) at one measurement point, with 58·0 % of the children (n 202) providing both pre- and post-intervention measurements. The majority of the sample (53·2 %) was boys. Nearly a third (32·4 %, n 113) of all children came from an immigrant background; fifty-seven (16·3 %) and seventy-four (21·2 %) had mothers with low and higher educational levels, respectively. Mean age was 4·26 (sd 0·78) years. The proportion of overweight children was 8·4 %(Reference Kromeyer-Hauschild, Wabitsch and Kunze49), a figure comparable to German national statistics (9 %)(Reference Kurth and Schaffrath Rosario71). Gender distribution, immigrant background and maternal educational levels were similar among participants across measurement time points. With the exception of mean age, there were no significant differences in children's characteristics between those providing both pre- and post-intervention measurements compared with the full sample (Table 2).

Table 2 Characteristics of the study participants (in percentage and absolute numbers): children participating in a cluster-randomized trial of a pre-school-based nutritional intervention on fruit and vegetable intakes, south Germany

†Participants with both pre- and post-intervention measurements had a higher mean age (P ≤ 0·01); all other differences were non-significant (P > 0·05). Significance levels quoted are two-sided.

‡Maternal education level was defined according to the International Standard Classification of Education (ISCED): ‘high’ = ISCED level 5 or 6; ‘middle’ = ISCED level 4; ‘low’ = ISCED level 1, 2 or 3.

At baseline, 60·0 % and 34·6 % of the pre-school children achieved the recommended amounts of at least two portions (200 g) of fruit and vegetables daily(Reference Alexy, Klausen and Kersting72), respectively. When comparing only pre-intervention and post-intervention measurements, there was a significant change from baseline in fruit and vegetable intakes (Fig. 2). Specifically, we found a mean increase from baseline of 0·17 points on the 6-point ordinal scale (P ≤ 0·05) for fruit consumption frequency and 0·22 points (P ≤ 0·01) for vegetable consumption, unadjusted for other characteristics. A change of 1 on the 6-point ordinal scale approximates one portion difference (size of a child's hand).

Fig. 2 Children's fruit (a) and vegetable (b) intakes before () and after () the pre-school-based nutritional intervention, indicated as the percentage of children distributed across intake categories measured as portions (a child's hand full) per day. Data presented include only children with both pre- and post-intervention measurements (n 202, see Table 2)

In the random-effects panel analysis including all measurements (Table 3) and controlling for age, we found an increase of 0·22 points on the ordinal scale of fruit consumption compared with consumption before intervention (‘within-child comparison’) and compared with children not receiving the intervention (‘between-child comparison’; P = 0·027). Participation in the intervention was also associated with a change in vegetable consumption (P = 0·027): we found an increase of 0·15 points on the ordinal scale in vegetable intake (also based upon ‘within-child comparison’ and ‘between-child comparison’). After adjusting for gender, immigrant background and maternal education, regression coefficients for the intervention effect were slightly higher and remained statistically significant for fruit (0·23; P = 0·0 0 1) and vegetable (0·15; P = 0·027) consumption. These estimators refer to the general intervention effect. Six-month and 12-month measurement points were not differentiated because we only had 12-month data for 18·9 % of pre-schools.

Table 3 The impact of the intervention on fruit and vegetable intakes† using random-effects panel analysis, adjusted for age (model 1a/2a) as well as gender, immigrant background and maternal educational level (model 1b/2b): pre-school children, south Germany

n, number of children contributing measurements; N, number of measurements included in the analysis.

Level of two-sided significance: *P ≤ 0·05, **P ≤ 0·01, ***P ≤ 0·001.

†Daily fruit and vegetable intakes as measured in child ‘hand’ portions on a 6-point ordinal scale, ranging from ‘none’ to ‘more than three’.

‡Reference category: girls.

§Reference category: children without immigrant background.

∥Maternal education level was defined according to the international Standard Classification of Education (ISCED): ‘high’ = ISCED level 5 or 6; ‘middle’ = ISCED level 4; ‘low’ = ISCED level 1, 2 or 3. Reference category: low maternal education.

The intervention did not have significant effects on daily water intake or the consumption of high-energy drinks (data not presented). We also found no significant intervention effect on BMI, WTHR or skinfold sum. In the subgroups of overweight and obese children (n 18 with pre- and post-term measurements), bivariate analyses also showed no significant effect of the intervention on BMI, WTHR and skinfold sum. A multivariate analysis adjusted for age was not possible due to the small size of these subgroups.

Process evaluation

On average, 23·1 (sd 12·1) children participated regularly in the lessons, with 16·5 (sd 9·5) parents present at the parents’ only and parent and children's sessions. In sum, 99 % and 69 % of the recruited children and parents participated, representing 47·8 % and 68·3 % reach into the target population (recruited and non-recruited) for children and parents, respectively (http://www.re-aim.org/tools/calculations/use-the-reach-calculator.aspx). For both parent and children's modules and all pre-schools, the implementation rate was high with all modules delivered completely (5·0/5); no session was cancelled. Intervention fidelity was high with the majority of interventions delivered as planned (ratings 4·4 (sd 0·5) and 4·2 (sd 0·3) on 5-point ordinal scale for parent and children's modules, respectively). Those responsible for administering the intervention reported expected seasonal adaptations of the activities like a delay in seed sowing related to the weather. Parental and teacher commitment was rated high (3·8 (sd 0·8) and 4·2 (sd 1·5), respectively). Some experts also related that parents sought more counselling after the sessions ended. The teachers were mostly regarded as very helpful and committed by the nutrition experts, with the amount of effort required to facilitate programme implementation rated as ‘low’ (2·0 (sd 1·6) out of 5 for both children's and parent modules).

Discussion

The pre-school-based nutritional intervention arm of the programme significantly increased fruit and vegetable intakes in our sample by around 0·23 and 0·15 portions daily (portion = size of a child's hand representing about 100 g), respectively. Although this change appears to be rather small on an individual scale, it is consistent with an increase of 6 % for both fruit and vegetable consumption in the whole sample and therefore might represent a relevant change on population level. According to a Norwegian study reported by Bere et al.(Reference Bere, Veierod and Skare73), a small increase of 2·5 g of fruit and vegetables daily over lifetime makes e.g. school fruit programmes cost-effective.

Our short-term findings are generally consistent with previous studies in older children: evaluations of school-based nutrition programmes in older children showed increases in fruit and vegetable intake(Reference Friel, Kelleher and Campbell8, Reference Cason28Reference Anderson, Porteous and Foster30, Reference Reinaerts, Crutzen and Candel74). Middle- to long-term increases are mostly reported in portions per day, ranging from about 0·09 and 0·07 more portions in fruit and vegetables, respectively(Reference Reinaerts, Crutzen and Candel74), to 0·28 portions of fruit and vegetables(Reference Hoffman, Thompson and Franko75), or 10 % daily increases in consumption of fruit(Reference Siega-Riz, El Ghormli and Mobley76). The 0·23 and 0·15 portion increases in our pre-school study thus are comparable with previous evidence on intervention effects in older children. Some previous intervention studies could also show significant increases in non-primary target foods(Reference Cason28, Reference Horne, Greenhalgh and Erjavec77). In our pre-school sample, we could not detect any significant change in water or sugar-sweetened drinks consumption. This could be due to the intervention design, which focused more on fruit and vegetable exposure and on testing new foods than on the consumption of drinks. Furthermore, sugar-sweetened drinks consumption in our sample was already at a very low level before the intervention commenced.

A change in eating behaviour might be viewed as an intervention success, even without an observable change in anthropometric parameters. Increased fruit and vegetable intake generally indicates a nutritional pattern characterized by less refined foods and a more balanced nutrition(Reference Alexy, Clausen and Kersting78). Such healthy nutritional patterns have been shown to lead to a definite yet delayed decrease of BMI and body fat in adults and adolescents, especially if maintained over a longer period(Reference Tohill, Seymour and Serdula79, Reference Buijsse, Feskens and Schulze80). This suggests that our intervention might still be effective in reducing the prevalence of childhood overweight over the long term, even if it cannot be demonstrated in the current study with a relatively short period of follow-up.

By our process evaluation, we observed that 99 % and 69 % of the recruited children and parents, respectively, were reached by the intervention. This points to the potential importance of this setting for interventions, especially because a vast majority of children attend pre-schools in developed Western countries(24, 25). Social learning theory also underlines the value of engaging individuals within and outside the family with potential influence on the development of behaviours and behavioural norms.

Interventions targeting multiple settings and combining healthy nutrition and physical activity promotion have successfully changed anthropometric parameters in older children(Reference Taylor, McAuley and Barbezat81, Reference Müller, Asbeck and Mast82). In contrast, nutritional interventions in pre-school children could not improve body composition measures, with only very few studies including anthropometric data as outcome variables(Reference Hu, Ye and Li27) at all. Even if our short-term findings are consistent with these previous results, our analysis should be repeated including the sustainability measurements from all pre-schools to come to a final conclusion. By the time of reporting, sustainability measurement data from only 18·9 % of pre-schools were available.

Similar to previous work conducted in controlled laboratory settings involving older children and adults(Reference Lowe, Horne and Tapper37, Reference Horne, Tapper and Lowe39, Reference Birch, McPhee and Shoba45, Reference Pliner, Pelchat and Grabski83Reference Forestell and Mennella86), the exposure effect reflected in our programme suggests that pre-school interventions can have a positive impact on eating behaviour of younger children. Other studies in school-aged children have shown that children's knowledge about nutrition may be improved and their preferences concerning food also changed by nutritional interventions(Reference Friel, Kelleher and Campbell8, Reference Cason28Reference Anderson, Porteous and Foster30). Although we did not collect comparable data given concerns about the reliability of self-reports in pre-school children, we observed parental report of an increase in fruit and vegetable consumption after the intervention. It is unclear whether this finding reflects a direct effect on children's nutritional behaviour or an indirect, beneficial effect on parental knowledge and practice.

The present study has a number of strengths. First, on an international scale, to our knowledge it is one of very few studies evaluating the impact of nutritional interventions in pre-schools(Reference Jebb and Lambert87, Reference Campbell and Hesketh88). This is an important research gap, as especially the pre-school period should be viewed as a window of opportunity for establishing healthy nutritional behaviours(Reference Reilly, Methven and McDowell5, Reference Rummel23, Reference Story, Kaphingst and French26, Reference Birch, McPhee and Shoba45). Second, the current study used objective anthropometric outcomes in addition to self-reported behavioural outcomes. This allows differentiation between behavioural and body compositional changes in the children, yet with the anthropometric parameters left unaffected by the intervention. A third strength concerns the theoretical underpinnings of the intervention: with parents playing a substantial role in influencing health behaviours of young children, their inclusion into the intervention may have been an essential precondition for an intervention success(Reference Lindsay, Sussner and Kim44). In our intervention, a third of the intervention units were delivered to parents only or parents and children instead of children only, and thereby a majority of the parental population could be reached. By this approach, the programme applied a general system theory-based approach(Reference Bertalanffy32) linking children, pre-school teachers and parents to change pre-school curricula.

Despite these strengths, the study also has several limitations. First, children were sampled from pre-schools that applied for a nutritional intervention, possibly resulting in limited generalizability. Second, in the current analysis, we did not have sustainability measurements from all participating pre-schools. Third, despite the calculation of a minimal sample size of 332, we only had 202 children providing both pre- and post-intervention measurements. Our analysis thus may have been underpowered for detecting changes in anthropometric outcomes or water and sugar-sweetened drinks consumption. Fourth, we evaluated only a limited number of behavioural outcomes. The intervention thus could have changed other eating behaviours, without the study design being able to detect these changes. Last, we did not measure nutritional intake directly, but focused instead on dietary patterns. Although dietary patterns are known to correlate quite well with nutrient intake(Reference Roman-Vinas, Ribas Barba and Ngo89), a food diary or daily weighing approach might have resulted in more accurate data, but was not considered feasible in this sample or study setting.

Conclusions

The present study demonstrates a significant increase in children's fruit and vegetable intake following a 6-month nutritional intervention in south German pre-schools. Fruit and vegetables are important components of a balanced nutrition(Reference Alexy, Clausen and Kersting78). Although we were unable to demonstrate an effect of our intervention on anthropometric measures after 6 months, we nevertheless anticipate that increased fruit and vegetable intake over a longer time period might eventually lead to a change in anthropometric features. Future studies evaluating interventions might wish to include long-term effectiveness and sustainability measurements, investigating whether behavioural change through interventions eventually leads to health-relevant changes in anthropometric data.

Acknowledgements

This work was supported by a grant from the Baden-Württemberg Stiftung. F.D.B. is supported by the European Social Fund and by the Ministry of Science, Research and the Arts Baden-Württemberg. The authors confirm that they do not have any conflict of interest. Author contributions were as follows. F.D.B.: conception and design of the study, interpretation of data, drafting and revising the manuscript; L.B.: analysis of data, drafting the manuscript; J.F.: conception and design of the study, grant writing. The authors thank the parents, children, families and pre-schools for their cooperation. They also sincerely thank David Litaker, MD, PhD (Departments of Medicine, Epidemiology and Biostatistics, Case Western Reserve University, Cleveland, OH, USA) for his very detailed and thorough review of the manuscript and Professor Dr Josef Brüderl (Department of Sociology, Mannheim University, Mannheim, Germany) for his support with panel analysis.

References

1.Wilkinson, KM (2008) Increasing obesity in children and adolescents: an alarming epidemic. JAAPA 21, 3136.CrossRefGoogle ScholarPubMed
2.Troiano, RP & Flegal, KM (1998) Overweight children and adolescents: description, epidemiology, and demographics. Pediatrics 101, 497504.Google Scholar
3.Branca, F, Nikogosian, H & Lobstein, T (2006) The Challenge of Obesity in the WHO European Region and the Strategies for Response. Copenhagen: WHO.Google Scholar
4.Dietz, WH & Bellizzi, MC (1999) Introduction: the use of body mass index to assess obesity in children. Am J Clin Nutr 70, 123125.Google Scholar
5.Reilly, JJ, Methven, E, McDowell, ZC et al. (2003) Health consequences of obesity. Arch Dis Child 88, 748752.CrossRefGoogle ScholarPubMed
6.Wirth, A (2008) Adipositas. Ätiologie, Folgekrankheiten, Diagnostik, Therapie. Heidelberg: Springer.Google Scholar
7.Freedman, DS, Dietz, WH, Srinivasan, SR et al. (1999) The relation of overweight to cardiovascular risk factors among children and adolescents: the Bogalusa Heart Study. Pediatrics 103, 11751182.CrossRefGoogle ScholarPubMed
8.Friel, S, Kelleher, C, Campbell, P et al. (1999) Evaluation of the Nutrition Education at Primary School (NEAPS) programme. Public Health Nutr 2, 549555.Google Scholar
9.Singh, AS, Mulder, C, Twisk, JW et al. (2008) Tracking of childhood overweight into adulthood: a systematic review of the literature. Obes Rev 9, 474488.Google Scholar
10.Nader, PR, O'Brien, M, Houts, R et al. (2006) Identifying risk for obesity in early childhood. Pediatrics 118, 594601.Google Scholar
11.Must, A (2003) Does overweight in childhood have an impact on adult health? Nutr Rev 61, 139142.Google Scholar
12.Gortmaker, SL, Must, A, Perrin, JM et al. (1993) Social and economic consequences of overweight in adolescence and young adulthood. N Engl J Med 329, 10081012.Google Scholar
13.Wang, SS, Brownell, KD & Wadden, TA (2004) The influence of the stigma of obesity on overweight individuals. Int J Obes Relat Metab Disord 28, 13331337.Google Scholar
14.Strauss, RS & Pollack, HA (2003) Social marginalization of overweight children. Arch Pediatr Adolesc Med 157, 746752.Google Scholar
15.Sargent, JD & Blanchflower, DG (1994) Obesity and stature in adolescence and earnings in young adulthood. Analysis of a British birth cohort. Arch Pediatr Adolesc Med 148, 681687.Google Scholar
16.Birch, LL (1998) Development of food acceptance patterns in the first years of life. Proc Nutr Soc 57, 617624.CrossRefGoogle ScholarPubMed
17.Wardle, J & Cooke, L (2008) Genetic and environmental determinants of children's food preferences. Br J Nutr 99, Suppl. 1, 1521.Google Scholar
18.Anderson, PM & Butcher, KE (2006) Childhood obesity: trends and potential causes. Future Child 16, 1945.Google Scholar
19.Mensink, G, Kleiser, C & Richter, A (2007) Lebensmittelverzehr bei Kindern und Jugendlichen in Deutschland. Ergebnisse des Kinder- und Jugendsurveys (KIGGS). Bundesgesundheitsbl Gesundheitsforsch Gesundheitsschutz 50, 609623.Google Scholar
20.Yngve, A, Wolf, A, Poortvliet, E et al. (2005) Fruit and vegetable intake in a sample of 11-year-old children in 9 European countries: The Pro Children Cross-sectional Survey. Ann Nutr Metab 49, 236245.Google Scholar
21.Zwiauer, KF (2000) Prevention and treatment of overweight and obesity in children and adolescents. Eur J Pediatr 159, Suppl. 1, 5668.CrossRefGoogle ScholarPubMed
22.Oerter, R (2002) Kindheit. In Entwicklungspsychologie. Ein Lehrbuch, pp. 209–258 [R Oerter and L Montada, editors]. Weinheim: Beltz.Google Scholar
23.Rummel, C (2001) Projektbeispiele zur Ernährungserziehung im Bereich Kindergarten. In Ernährungsziele unserer Gesellschaft [Bundesforschungsanstalt für Ernährung, editor]. Karlsruhe: Bundesforschungsanstalt für Ernährung.Google Scholar
24.National Institute for Early Education (2004) Preschool attendance reaches new highs, but some are still left behind. NIEER Preschool Policy Matters 2, issue 4; available at http://nieer.org/psm/index.php?article=74Google Scholar
25.European Primary Schools Association (1995) Preschool Education in the EU: Current Thinking and Provision. Luxembourg: Office for Official Publications of the European Communities.Google Scholar
26.Story, M, Kaphingst, KM & French, S (2006) The role of child care settings in obesity prevention. Future Child 16, 143168.Google Scholar
27.Hu, C, Ye, D, Li, Y et al. (2009) Evaluation of a kindergarten-based nutrition education intervention for pre-school children in China. Public Health Nutr 13, 253260.Google Scholar
28.Cason, KL (2001) Evaluation of a preschool nutrition education program based on the theory of multiple intelligences. J Nutr Educ 33, 161164.Google Scholar
29.Muth, ND, Chatterjee, A, Williams, D et al. (2008) Making an IMPACT: effect of a school-based pilot intervention. N C Med J 69, 432440.Google Scholar
30.Anderson, AS, Porteous, LE, Foster, E et al. (2005) The impact of a school-based nutrition education intervention on dietary intake and cognitive and attitudinal variables relating to fruits and vegetables. Public Health Nutr 8, 650656.Google Scholar
31.Bandura, A (editor) (1974) Die analyse von Modellierungsprozessen. In Lernen am Modell, pp. 9–67. Stuttgart:Klett.Google Scholar
32.Bertalanffy, L (1968) General System Theory: Foundations, Developments, Applications. New York: George Braziller.Google Scholar
33.Franks, PW, Hanson, RL, Knowler, WC et al. (2010) Childhood obesity, other cardiovascular risk factors, and premature death. N Engl J Med 362, 485493.CrossRefGoogle ScholarPubMed
34.De Bock, F, Fischer, JE, Hoffmann, K et al. (2010) A participatory parent-focused intervention promoting physical activity in preschools: design of a cluster-randomized trial. BMC Public Health 10, 49.Google Scholar
35.Kolip, P (2004) Influence of gender and social inequality on nutrition and overweight in children and adolescents. Bundesgesundheitsbl Gesundheitsforsch Gesundheitsschutz 47, 235239.Google Scholar
36.Kleiser, C, Schaffrath Rosario, A, Mensink, GB et al. (2009) Potential determinants of obesity among children and adolescents in Germany: results from the cross-sectional KiGGS Study. BMC Public Health 9, 46.Google Scholar
37.Lowe, CF, Horne, PJ, Tapper, K et al. (2004) Effects of a peer modelling and rewards-based intervention to increase fruit and vegetable consumption in children. Eur J Clin Nutr 58, 510522.Google Scholar
38.Rinderknecht, K & Smith, C (2004) Social cognitive theory in an after-school nutrition intervention for urban Native American youth. J Nutr Educ Behav 36, 298304.Google Scholar
39.Horne, PJ, Tapper, K, Lowe, CF et al. (2004) Increasing children's fruit and vegetable consumption: a peer-modelling and rewards-based intervention. Eur J Clin Nutr 58, 16491660.CrossRefGoogle ScholarPubMed
40.Zajonc, RB (1968) Attitudinal effects of mere exposures. J Pers Soc Psychol 9, 127.CrossRefGoogle Scholar
41.Perry, CL, Mullis, RM & Maile, MC (1985) Modifying the eating behavior of young children. J Sch Health 55, 399402.Google Scholar
42.Loewen, R & Pliner, P (1999) Effects of prior exposure to palatable and unpalatable novel foods on children's willingness to taste other novel foods. Appetite 32, 351366.Google Scholar
43.Sullivan, S & Birch, L (1990) Pass the sugar, pass the salt: experience dictates preference. Dev Psychol 26, 546551.Google Scholar
44.Lindsay, AC, Sussner, KM, Kim, J et al. (2006) The role of parents in preventing childhood obesity. Future Child 16, 169186.CrossRefGoogle ScholarPubMed
45.Birch, LL, McPhee, L, Shoba, BC et al. (1987) What kind of exposure reduces children's food neophobia? Looking vs. tasting. Appetite 9, 171178.Google Scholar
46.Cooke, L (2007) The importance of exposure for healthy eating in childhood: a review. J Hum Nutr Diet 20, 294301.Google Scholar
47.Glasgow, RE, Edwards, LL, Whitesides, H et al. (2009) Reach and effectiveness of DVD and in-person diabetes self-management education. Chronic Illn 5, 243249.Google Scholar
48.Bayer, O, von Kries, R, Strauss, A et al. (2009) Short- and mid-term effects of a setting based prevention program to reduce obesity risk factors in children: a cluster-randomized trial. Clin Nutr 28, 122128.Google Scholar
49.Kromeyer-Hauschild, K, Wabitsch, M, Kunze, D et al. (2001) Perzentile für den Body-mass-Index für das Kindes- und Jugendalter unter Heranziehung verschiedener deutscher Stichproben. Monatsschr Kinderheilkd 149, 807812.CrossRefGoogle Scholar
50.Stolzenberg, H, Kahl, H & Bergmann, KE (2007) Körpermaße bei Kindern und Jugendlichen in Deutschland. Ergebnisse des Kinder- und Jugendgesundheitssurveys (KiGGS). Bundesgesundheitsbl Gesundheitsforsch Gesundheitsschutz 50, 659669.Google Scholar
51.Roswall, J, Bergman, S, Almqvist-Tangen, G et al. (2009) Population-based waist circumference and waist-to-height ratio reference values in preschool children. Acta Paediatr 98, 16321636.CrossRefGoogle ScholarPubMed
52.Ashwell, M (2009) Obesity risk: importance of the waist-to-height ratio. Nurs Stand 23, 4954.Google Scholar
53.Nambiar, S, Truby, H, Abbott, RA et al. (2009) Validating the waist-height ratio and developing centiles for use amongst children and adolescents. Acta Paediatr 98, 148152.Google Scholar
54.Freedman, DS, Wang, J, Ogden, CL et al. (2007) The prediction of body fatness by BMI and skinfold thicknesses among children and adolescents. Ann Hum Biol 34, 183194.Google Scholar
55.Eisenmann, JC, Heelan, KA & Welk, GJ (2004) Assessing body composition among 3- to 8-year-old children: anthropometry, BIA, and DXA. Obes Res 12, 16331640.CrossRefGoogle ScholarPubMed
56.Anderson, SE, Economos, CD & Must, A (2008) Active play and screen time in US children aged 4 to 11 years in relation to sociodemographic and weight status characteristics: a nationally representative cross-sectional analysis. BMC Public Health 8, 366.Google Scholar
57.Muller, MJ, Koertzinger, I, Mast, M et al. (1999) Physical activity and diet in 5 to 7 years old children. Public Health Nutr 2, 443444.Google Scholar
58.Kuepper-Nybelen, J, Lamerz, A, Bruning, N et al. (2005) Major differences in prevalence of overweight according to nationality in preschool children living in Germany: determinants and public health implications. Arch Dis Child 90, 359363.Google Scholar
59.Schenk, L, Bau, AM, Borde, T et al. (2006) A basic set of indicators for mapping migrant status. Recommendations for epidemiological practice. Bundesgesundheitsbl Gesundheitsforsch Gesundheitsschutz 49, 853860.Google Scholar
60.Ball, K, Cleland, VJ, Timperio, AF et al. (2009) Socioeconomic position and children's physical activity and sedentary behaviors: longitudinal findings from the CLAN study. J Phys Act Health 6, 289298.Google Scholar
61.Semmler, C, Ashcroft, J, van Jaarsveld, CH et al. (2009) Development of overweight in children in relation to parental weight and socioeconomic status. Obesity (Silver Spring) 17, 814820.Google Scholar
62.United Nations Educational, Scientific and Cultural Organization (1997) International Standard Classification of Education, ISCED 1997. http://www.unesco.org/education/information/nfsunesco/doc/isced_1997.htm (accessed September 2010).Google Scholar
63.Allison, PD (1994) Using panel data to estimate the effects of events. Sociol Method Res 23, 174199.CrossRefGoogle Scholar
64.Halaby, CN (2004) Panel models in sociological research: theory into practice. Annu Rev Sociol 30, 507544.CrossRefGoogle Scholar
65.University of California at Los Angeles, Academic Tchnology Services (2010) Analyzing correlated (clustered) data. http://www.ats.ucla.edu/stat/stata/Library/cpsu.htm (accessed September 2010).Google Scholar
66.Hox, JJ (1998) Multilevel modeling: when and why. In Classification, Data Analysis and Data Highways, pp. 147154 [I Balderjahn, R Mathar and M Schader, editors]. New York: Springer.Google Scholar
67.Moss, S (2008) When to apply multi-level modelling. http://www.psych-it.com.au/Psychlopedia/article.asp?id=225 (accessed September 2010).Google Scholar
68.Muthen, BO & Satorra, A (1995) Complex sample data in structural equation modelling. Sociol Methodol 25, 267316.Google Scholar
69.Hausman, JA (1978) Specification tests in econometrics. Econometrica 46, 12511271.Google Scholar
70.Cohen, J (1988) Statistical Power Analysis for the Behavioral Sciences. Hillsdale, NJ: Lawrence Erlbaum Associates.Google Scholar
71.Kurth, B & Schaffrath Rosario, A (2007) The prevalence of overweight and obese children and adolescents living in Germany. Results of the German Health Interview and Examination Survey for Children and Adolescents (KiGGS). Bundesgesundheitsbl Gesundheitsforsch Gesundheitsschutz 50, 736743.CrossRefGoogle ScholarPubMed
72.Alexy, U, Klausen, K & Kersting, M (2008) Die Ernährung gesunder Kinder und Jugendlicher nach dem Konzept der optimierten Mischkost. Ernährungsumschau 55, 168177.Google Scholar
73.Bere, E, Veierod, MB, Skare, O et al. (2007) Free school fruit – sustained effect three years later. Int J Behav Nutr Phys Act 4, 5.CrossRefGoogle ScholarPubMed
74.Reinaerts, E, Crutzen, R, Candel, M et al. (2008) Increasing fruit and vegetable intake among children: comparing long-term effects of a free distribution and a multicomponent program. Health Educ Res 23, 987996.Google Scholar
75.Hoffman, JA, Thompson, DR, Franko, DL et al. (2011) Decaying behavioral effects in a randomized, multi-year fruit and vegetable intake intervention. Prev Med 52, 370375.Google Scholar
76.Siega-Riz, AM, El Ghormli, L, Mobley, C et al. (2011) The effects of the HEALTHY study intervention on middle school student dietary intakes. Int J Behav Nutr Phys Act 8, 7.CrossRefGoogle ScholarPubMed
77.Horne, PJ, Greenhalgh, J, Erjavec, M et al. (2011) Increasing pre-school children's consumption of fruit and vegetables. A modelling and rewards intervention. Appetite 56, 375385.CrossRefGoogle ScholarPubMed
78.Alexy, U, Clausen, K & Kersting, M (2008) Die Ernährung gesunder Kinder und Jugendlicher nach dem Konzept der Optimierten Mischkost. Ernährungsumschau 3, 168175.Google Scholar
79.Tohill, BC, Seymour, J, Serdula, M et al. (2004) What epidemiologic studies tell us about the relationship between fruit and vegetable consumption and body weight. Nutr Rev 62, 365374.CrossRefGoogle ScholarPubMed
80.Buijsse, B, Feskens, EJ, Schulze, MB et al. (2009) Fruit and vegetable intakes and subsequent changes in body weight in European populations: results from the project on Diet, Obesity, and Genes (DiOGenes). Am J Clin Nutr 90, 202209.CrossRefGoogle Scholar
81.Taylor, RW, McAuley, KA, Barbezat, W et al. (2008) Two-year follow-up of an obesity prevention initiative in children: the APPLE project. Am J Clin Nutr 88, 13711377.Google Scholar
82.Müller, MJ, Asbeck, I, Mast, M et al. (2001) Prevention of obesity – more than an intention. Concept and first results of the Kiel Obesity Prevention Study (KOPS). Int J Obes Relat Metab Disord 25, Suppl. 1, S66S74.CrossRefGoogle ScholarPubMed
83.Pliner, P, Pelchat, M & Grabski, M (1993) Reduction of neophobia in humans by exposure to novel foods. Appetite 20, 111123.Google Scholar
84.Hobden, K & Pliner, P (1995) Effects of a model on food neophobia in humans. Appetite 25, 101113.CrossRefGoogle Scholar
85.Hendy, HM (2002) Effectiveness of trained peer models to encourage food acceptance in preschool children. Appetite 39, 217225.Google Scholar
86.Forestell, CA & Mennella, JA (2007) Early determinants of fruit and vegetable acceptance. Pediatrics 120, 12471254.CrossRefGoogle ScholarPubMed
87.Jebb, SA & Lambert, J (2000) Overweight and obesity in European children and adolescents. Eur J Pediatr 159, Suppl. 1, S2S4.Google Scholar
88.Campbell, KJ & Hesketh, KD (2007) Strategies which aim to positively impact on weight, physical activity, diet and sedentary behaviours in children from zero to five years. A systematic review of the literature. Obes Rev 8, 327338.Google Scholar
89.Roman-Vinas, B, Ribas Barba, L, Ngo, J et al. (2009) Validity of dietary patterns to assess nutrient intake adequacy. Br J Nutr 101, Suppl. 2, S12S20.Google Scholar
Figure 0

Fig. 1 (colour online) Enrolment, stratification, randomized allocation of pre-schools and resulting number of children in each arm: cluster-randomized trial of a pre-school-based nutritional intervention on children's fruit and vegetable intakes, south Germany

Figure 1

Table 1 The nutritional intervention, consisting of fifteen 2 h sessions once weekly over a period of 6 months. Ten modules only targeted children, another five parents and children or parents exclusively

Figure 2

Table 2 Characteristics of the study participants (in percentage and absolute numbers): children participating in a cluster-randomized trial of a pre-school-based nutritional intervention on fruit and vegetable intakes, south Germany

Figure 3

Fig. 2 Children's fruit (a) and vegetable (b) intakes before () and after () the pre-school-based nutritional intervention, indicated as the percentage of children distributed across intake categories measured as portions (a child's hand full) per day. Data presented include only children with both pre- and post-intervention measurements (n 202, see Table 2)

Figure 4

Table 3 The impact of the intervention on fruit and vegetable intakes† using random-effects panel analysis, adjusted for age (model 1a/2a) as well as gender, immigrant background and maternal educational level (model 1b/2b): pre-school children, south Germany