Introduction

Childhood obesity is one of the biggest health challenges of the 21^st century. The problem is global and is progressively affecting many countries, where the prevalence of overweight and obesity among children has increased significantly in recent years. This is a concern shared by governments and institutions at international level, which has materialized in various prevention plans and strategies aimed at encouraging children and young people to adopt healthy lifestyles, mainly through healthy eating and regular physical activity.

In Spain, according to the 2019 ALADINO study, childhood obesity and overweight have reached figures of 17.3% and 23.3% respectively, i.e. 4 out of 10 (40.6%) of schoolchildren aged between 6 and 9 years are overweight.¹ The 2015 ALADINO-Euskadi report also showed a prevalence of childhood obesity and overweight of 34.2%.²

The pattern of food consumption in children is characterized by the following aspects: 4% of children and adolescents in the Basque Autonomous Community (BAC) do not eat breakfast; the consumption of dairy products and cereals consist of about 50% of estimated daily consumption; girls consume more fruit than boys, while boys have higher intakes of cereals, meat and juices. In general, daily food intake increases with age; however, the consumption of milk and its derivatives is higher among younger people and a higher consumption of milk, fish and fruit is observed in the high-socioeconomic status, while meat and dairy products are more consumed by socioeconomically disadvantaged people.³

In addition to the fact that the Basque Government's Directorate of Public Health and Addictions in 2019 presented “SANO”, the Comprehensive Strategy for the Prevention of Childhood Obesity in the BAC,² which aims to reduce the prevalence of childhood obesity and overweight in the BAC, the Basque Health Plan 2013-2020 also includes among its provisions measures to curb this situation. In its objective 4.1.5. (Child Health Area) establishes as a priority to monitor the risks of child malnutrition and develop prevention and approach mechanisms in coordination with social and educational services. To this end, the Basque Government Departments of Health, Education and Agriculture have promoted the “Initiatives for Healthy Eating in the Basque Country” Plan with the aim of promoting healthy eating and thus limiting the increase in chronic diseases such as cardiovascular diseases, diabetes and some types of cancer.⁴ This Plan contemplates 18 initiatives (included in 5 actions lines) to be developed until 2020, ranging from the reduction of salt and sugar consumption to the promotion of seasonal fruits and vegetables, or the promotion of healthy eating in school canteens, among many others. The present study corresponds to the last-mentioned initiative which requires the coordinated intervention of different actors (central kitchens, schools, food industry, distribution and authorities) working synergistically to achieve the improvement of healthy eating. These agents take care of the appearance of the dishes, improve their format and make them more attractive and appetising in order to facilitate healthier choices. One of the first result of this collaborative work was a Guide of recommendations for healthy eating at school age published in 2018, which have been adopted in BAC school canteens.⁵

Strategies to address the problem of obesity from an early age must be considered with a multidimensional approach, in which unhealthy eating habits together with sedentary lifestyles are the main aspects to reduce. In this comprehensive approach, school canteens are a key tool for proper child and youth nutrition. Schools, overall, provide a key platform for multiple community health benefits for children and young people.⁶ Recent research in countries such as Sweden or Brazil,^7,8 shows that school lunches are contributing positively to school children's dietary intake and are determinants of their long-term health, as they positively affect the consumption of healthy foods. In the case of Spain, there is not enough bibliography to support it.

Considering the different educational centres models currently existing in Spain and in the BAC more than two thirds of schools belong to the public network. Regarding the management of school canteens, in the same area, the model of catering (food prepared in a central kitchen and transported to the school for direct consumption), is predominant compared to the model that consist of food prepared on-site (school canteens that include a kitchen).^9,10

More than 300,000 schoolchildren eat daily in 905 school canteens in the BAC, so that school meals represent a significant proportion of children's food in the BAC. School menus must comply with the technical specifications established by the Department of Education,¹¹ which include the proposals within the framework of the Plan of Initiatives for Healthy Eating in the BAC,⁴ so they are carefully designed, but their usefulness depends on their actual consumption.

Thus, the aim of this study is to measure and determine whether the effort to improve the quality of the menus made by public institutions, central kitchens and education centres, have been effective, in terms of tolerance, among school population because if the acceptance to the modified menus is void, all the effort made will have proved fruitless.

To this end, this study wants to evaluate and measure the consumption of 25 types of food present in the school menu by analysing the fraction not consumed. In addition, the study tries to determine whether the type of school management (public vs. private with subsidy) and the food origin (processed on site vs. catering) influences consumption.

Method

Results

The influence of the type of management and food origin

We only found statistically significant differences (p<0.05) associated with the type of management or food origin (catering vs. in situ) for five of the food items: legumes with vegetables, oily fish, lean fish, meat and precooked foods. However, for these five foods, the influence of these fixed factors is always very small (explained variance: R²_m=0.86-2.34%). For the other 20 food items, neither the type of management (public vs. private with subsidy school), nor the origin (catering vs. in situ) seems to significantly influence their acceptability. Only for oily fish (Fig. 1), the effect of the origin of food depends on the type of management. In the case of legumes with vegetables, lean fish, meat and precooked foods, the only significant effect is associated with the food origin.

Figure 1. Effect of food origin and management type significantly associated with the consumption of oily fish. 

Variability explained by territory, school and individual

The random factors territory and school, together with the residual variability due to the individuals (or diners) themselves, quantify in the form of explained variance the effect associated with them (Fig. 2). In the case of territory, the estimated variances of the corresponding normal distributions (which are always indexed by zero mean), lie between 0-10. In the case of schools, between 5-15. And in the case of individuals, between 15-35. Except in one case (canned fruit), therefore, the greatest variability is always due to the individuals; the second most influential factor is the school itself, and the last (in quantitative terms) is the historical territory. In other words, the territory where meals are served is hardly associated with food wastage, while the school has a moderate influence and the individual himself has a strong influence.

Figure 2. Variability (estimated variance) in food wastage attributed to nested random factors territory, school and individuals. The largest source of variability is the individual, followed by the school and, finally, the territory. 

Discussion

Foods offered in school canteens from the BAC are generally well accepted. However, certain foods which are an essential part of healthy diets, such as cooked vegetables dishes, legumes, dishes with oily or lean fish, and fruit, have average residues of around 20% or more. A study in Bizkaia published in 2015 that evaluated the offer of school menus according to NAOS indicators pointed out that one of the stumbling blocks reported by central kitchens was the significantly poorer acceptance of vegetables, fish and fruit.¹⁸ Another study from Bizkaia also observed a lower success rate in vegetable intake compared to other foods as a whole.⁶ The lower overall acceptance of fruit and vegetables by the school population has also been demonstrated in other studies in Spain and the United States.^19,20 Furthermore, the lower acceptance of fish dishes, which is related to family habits, is also known in other countries, such as Italy, which has led to the search for recipes and presentations that improve their acceptance.²¹ Nevertheless, legumes, fish and fruit have obtained better acceptance in our environment than that observed in previous national studies.^19,22 Legumes are also better accepted in our area (BAC) than the two options of rice dishes, a food considered (together with meat and pasta) to be preferred by the Spanish child population.¹⁹

Management model and food origin do not seem to affect the acceptance of 80% of the food items (20 of 25 food items). However, the acceptance of five food items does seem to be influenced, although only very moderately (between 0.86% and 2.34% of the acceptance), by the type of management, the origin, or both.

The management model does not seem to affect, in general, the acceptance of food in schools. Only in the case of oily fish, however, was its acceptance positively affected (lower average wastage) in private schools when it came from central kitchens, although this does not seem to be the case in public schools, where the origin of the fish is irrelevant to its level of consumption. After inspecting the data underlying these results, we found that in the private schools that bring their food from central kitchens, during the study period only tuna/bonito in tomato sauce was served, while at all other levels, trout, salmon, cobbler and mackerel were also served, and prepared in a variety of ways (baked or battered). Thus, we believe that it is this particular preparation (bonito/tuna with tomato) that explains the optimal consumption of oily fish observed in the private school, as the tomato sauce softens the taste of the fish and makes it juicier.

It is considered that the fact that there are no differences in food consumption between Basque private and public schools can be at least partially explained by the fact that is the same central kitchens that supply both public and private schools. Among the studies reviewed that provide data on food acceptance in schools,^23-25 none of them compare differences in consumption between private and public schools. Even so, discrepancies have been found between public and private school students in terms of knowledge about healthy eating and a healthy lifestyle.^26,27

The food origin (in situ vs. catering) negatively affected, although also very moderately, the consumption of legumes with vegetables, oily fish, lean fish, meat and pre-cooked foods. For all these foods, intakes were lower (i.e. wastage was higher) when the food came from central kitchens than when it was prepared in on-site kitchens. This may be explained by the fact that processing food products in advance can generate organoleptic alterations and loss of quality due to excessive regeneration or simply because there is a long time between processing and consumption.²⁸ Long periods between processing and consumption can lead to hardening and browning unsalted meats,²⁹ wilting or oxidation of vegetables or fruits,³⁰ surface drying of products such as fish, pasta and eggs, overcooked products in legumes and loss of crispiness in frying,^31,32 especially in precooked foods. In contrast, on-site cooking facilitates healthier cooking techniques such as grilling or steaming and reduces the need for breading or sauces,³³ which serve to help maintain temperatures, while adding fat content to foods.⁵ The on-site origin also helps to increase the supply of salads or to keep leftover fruit for a few days until it is at optimum ripeness, for example.²⁴

The greatest source of variability in food consumption has been the individual, followed by the school (the magnitude of whose associated influence is quantified as approximately half of that associated with the individual) and, lastly, in an almost negligible way for most of the food, the territory. For this reason, we believe that further educational campaigns should be directed, in the first instance, to intervention at the individual (or family) level and, in the second instance, to schools. We have not found any findings to compare these results with similar studies in this field.

This study has three limitations. One of them is that we had to modify the initial sample size, due to difficulties in collaborating with school canteen monitors because of the pandemic period. Furthermore, it should be noted that although there is a technical specification that indicates the grammage that should be offered for each food in each age group,¹¹ it is difficult to control the amount offered to students and therefore there may have been small variations in the result of Comstock wastage. Finally, although the school canteen monitors responsible for visual assessment have been trained with audio-visual training,¹⁴ it has not been possible to examine the degree of accuracy of each monitor on the Comstock scale.

Conclusions

The overall acceptance of the 25 foods evaluated is acceptable, with consumption between 90% and 60%, although there is room for improvement in the case of vegetables, fish and fruit. We believe that the promotion of on-site kitchens would improve the acceptability of legumes with vegetables, fish, meat and pre-cooked foods. We recommend that future interventions to improve intake be designed to be more personalised.

Transparency declaration

The corresponding author on behalf of the other authors guarantee the accuracy, transparency and honesty of the data and information contained in the study, that no relevant information has been omitted and that all discrepancies between authors have been adequately resolved and described.