Measuring the obesogenic food environment in New Zealand primary schools

Transcription

1 HEALTH PROMOTION INTERNATIONAL Vol. 19. No. 1 Oxford University Press doi: /heapro/dah103 All rights reserved Printed in Great Britain Measuring the obesogenic food environment in New Zealand primary schools MARY-ANN CARTER 1 and BOYD SWINBURN 2 Department of Community Health, University of Auckland, New Zealand and National Heart Foundation of New Zealand, Auckland, New Zealand Present addresses: 1 Public Health Unit, Hawkes Bay District Health Board, PO Box 447, Napier, New Zealand and 2 Deakin University, Melbourne, Australia SUMMARY Childhood obesity is an increasing health problem in New Zealand and many other countries. Information is needed to guide interventions that reduce the obesogenic (obesitypromoting) elements of school environments. The aim of this study was to identify and measure the obesogenic elements of the school environment and the canteen sales of energy-dense foods and drinks. A self-completion questionnaire was developed for assessing each school s nutrition environment and mailed to a stratified random sample of New Zealand schools. The responses from primary schools (n=200, response rate 61%) were analysed. Only 15.5% of schools had purpose-built canteen facilities and over half ran a food service for profit (31% profit to the school, 24.5% profit for the contractors). Only 16.5% of schools had a food policy, although 91% of those rated the policy as effective or very effective. The most commonly available foods for sale were pies (79%), juice (57%) and sausage rolls (54.5%). Filled rolls were the most expensive item (mean $1.79) and fruit the least expensive (mean $0.47). The ratio of less healthy to more healthy main choices was 5.6:1, for snacks it was 9.3:1 and for drinks it was 1.4:1. In contrast, ~60% of respondents said that nutrition was a priority for the school. Only 50% felt there was management support for healthy food choices and only 39% agreed that mainly nutritious food was offered by the food service. Less healthy choices dominated food sales by more than 2:1, with pies being the top selling item ( per week). We found that the food environment was not conducive to healthy food choices for the children at New Zealand schools and that this was reflected in the high sales of relatively unhealthy foods from the school food services. Programmes that improve school food through policies, availability, prices and school ethos are urgently needed. Key words: nutrition; obesity; schools INTRODUCTION Childhood overweight and obesity is high among New Zealand children, with estimates varying between 20 and 30% depending on age, ethnic group and income (Ministry of Health, 1997). Childhood obesity has immediate effects on physical health and well-being (such as orthopaedic problems and lowered fitness), future disease risk (such as increased blood pressure and adverse blood lipid levels) and mental health. Obesity, once it develops, is difficult to treat, and prevention programmes aimed at children and adolescents are considered a high priority as there is a high risk of persistence of obesity into adulthood (Gill, 1997). Schools can reach almost all children and adolescents during their first two decades in life and are ideal settings for influencing health (Glanz et al., 1995). Schools provide opportunities to practice health-promoting behaviours, and are a critical part of the social environment that shape young people s behaviours (Parcel et al., 1989). No other public 15

2 16 M.-A. Carter and B. Swinburn institution has as much continuous and intensive contact with children (Reniskow, 1993; Cale, 1997; Centres for Disease Control, 1997). School health education takes place both inside and outside the classroom. The policies that a school adopts, the physical and social environment it provides, the curriculum it chooses, and the quality and methods of instruction all have the potential to significantly affect the health of students and others at school. The school environment can reinforce or hinder messages delivered through the health education curriculum and can promote or create barriers for healthy behaviour (Shilton and Corti, 1992). The context in which students learn about healthy eating are key factors in determining how receptive they are to nutrition education (Crockett and Sims, 1995). It is not known how obesogenic (obesitypromoting) the school environments are in New Zealand and there are no published survey instruments for assessing this important setting. We have used the ANGELO framework (Analysis Grid for Environments Linked to Obesity) to develop a self-completion questionnaire for schools to assess the obesogenicity of their environments (Swinburn et al., 1999). This framework divides the environment into microand macroenvironments (the school setting is a microenvironment), and environment type into physical (what is available), economic (what are the financial factors), policy (what are the rules) and socio-cultural (what are the attitudes, perceptions, beliefs and values). The purpose of the framework is to identify and try to quantify the potential impact of environmental factors on the promotion of unhealthy weight gain. METHODS A school environments questionnaire was developed using information obtained from semistructured interviews with 11 primary and secondary school health and physical education teachers. Schools in the Auckland region were purposely selected to give a range of school type, ethnic mix and socio-economic status of the school population. Questions were developed to assess the key elements of the physical, economic, policy and socio-cultural environments in schools for nutrition (reported here) and physical activity (not reported). School food sales were also included as an index of foods eaten. Foods and drinks sold through the school food service were classified by the researcher as main choice (pies, filled rolls, sandwiches and hot chips), snacks (other foods) and drinks. They were grouped further into more healthy (higher nutrient density, lower energy density) or less healthy foods (lower nutrient density, higher energy density). The questionnaire was reviewed by academic staff within the Department of Community Health, University of Auckland, to ensure that the questions were easily understood, free from bias, and appropriately structured. It was then pre-tested by three primary and three secondary school health and physical education teachers, where each question was reviewed to see whether teachers had interpreted the questions as the researcher had intended. The questionnaire was then modified accordingly and mailed to the 500 selected schools nationally in September 1999, and one mailed reminder was sent ~2 weeks later. The population for the national survey comprised all 2737 schools on the New Zealand Ministry of Education database for July 1999 (excluding special schools and the correspondence school) (Ministry of Education, 1999). A sample size of 337 was estimated on the basis of a desired precision of 5% on an expected 50% variable prevalence. The response rate was expected to be ~60%. Therefore, a total sample of 500 schools was selected by systematic sampling (alphabetical listings by local region) and stratified by type of school to give 70 secondary and composite schools, and 430 primary (ages 5 10 years) and intermediate schools (ages years). School demographic data such as location, school size and decile rating for socio-economic status (SES) were obtained from the Ministry of Education. Data were entered on an Excel spread sheet and exported to SAS Insight for statistical analysis. The sales of the food items were estimated for all New Zealand primary/ intermediate schools by simple extrapolation of the sample data to all the schools. RESULTS The response rate for primary and intermediate schools was 60.6% (n = 200), and for secondary schools it was 32.5% (n = 26). Due to the low response rate of secondary schools, only primary/intermediate schools were used in analyses. There was an even distribution of SES

3 Obesogenic elements in New Zealand primary schools 17 rankings in the responding schools, with between 7 and 12% of the schools in each decile group. The majority of schools (61.5%) had 200 pupils and 23% had 300 pupils. The physical environment (what was available?) One hundred and sixty-seven schools had a food service. Of these schools, 69% had no physical canteen facilities, 15.5% had inadequate facilities and only 15.5% had purpose-built canteens. The available foods for sale in the food service are shown in Table 1. The food items most frequently for sale were pies (79%) and sausage rolls (54.5%), and the most common drink for sale was juice (57.5%). Fruit was the least frequent item offered for sale (28%). Very few schools operated a vending machine (3.5%). Where vending machines were operated, soft drinks, fruit juice and milk were the most commonly sold items. The economic environment (what were the financial factors?) Of the schools that provided a food service (n = 167), 37% ran it for profit for the school, 29% contracted it out to private businesses and 34% ran it as a not-for-profit service for the students. The cost of food items sold in schools is shown in Table 1. Filled rolls were the most Table 1: Percentage of schools selling food items and the mean unit price (NZ$) with 95% confidence intervals (95% CI) Food items Available (%) Mean price 95% CI (NZ$) Main choice Pies Hot chips Sandwiches Filled rolls Snacks Potato chips Cakes/biscuits Sausage rolls Doughnuts Scones/muffins Fruit Yoghurt Drinks Soft drinks Juice Milk expensive main choice, and cost 30 cents more expensive than the next most expensive main choice, pies. In the foods grouped as snacks, there was very little price difference between choices, except that fruit was about half the price of the other choices. Beverages were all of a relatively similar price. Almost three-quarters of schools used food products for fundraising. The most popular food products for fundraising were pizzas (51%) and pies (44%). The policy environment (what were the rules?) Schools were asked whether they had a food policy, and if so, how effective that policy had been in promoting healthy eating in the school. Only 33 schools (16.5%) had a food policy. Of these schools, almost all (91%) rated the policy as moderately effective (n = 22) or very effective (n = 8) in promoting healthy eating. Twenty schools provided a copy of their food policy and these were analysed to identify which aspects of the school environment they addressed (Table 2). Most schools followed a similar format and included a rationale followed by some key policy statements. About one-third of the policies had a rationale that included a statement about the importance of the food service supporting the curriculum and classroom learning about food. For a quarter of schools the rationale included a statement about the role of nutrition in improving the health and development of children. The rationale for the rest of the policies was diverse, Table 2: Statements included in school food policies (n = 20) Statements included in the policies Number of schools, n (%) Types of food provided in the 18 (90) food service Classroom teaching programme 8 (40) related to food and nutrition Encouraging children to adopt 6 (30) healthy eating behaviours Operating the school food service 6 (30) for profit Providing information to parents 5 (25) about healthy food Types of foods not permitted 3 (15) in the school The use of food for fundraising 2 (10) Food served at school events 2 (10) Staff as role models 1 (5)

4 18 M.-A. Carter and B. Swinburn Table 3: Ratings of statements about nutrition within schools (%) Statement No. Strongly Agree Neutral Disagree Strongly responding agree disagree Nutrition is high on our list of priorities Management is supportive of healthy food provision Food service provides mainly foods with high nutritional value and included broad statements about food as part of the school environment or about the importance of making a profit from the food service. The socio-cultural environment (what were the attitudes and perceptions?) Schools were asked to rate a range of statements to indicate how they applied to their school (Table 3). About 60% of schools felt nutrition was high on their list of priorities, with about onethird being neutral to this statement. Only half of the schools thought that the management team of their school supported the provision of healthy foods through the school canteen or lunch order service. Only 39% of schools felt the school food service provided mainly foods with high nutritional value. Just under a third of schools were neutral in response to this statement. Food sales Menu items were divided into groups (mains, snacks and beverages), and within each group items were split into more healthy or less healthy choices. Filled rolls, sandwiches, muffins, scones, yoghurt, fruit, juice and milk were designated more healthy choices. Pies, hot chips, sausage rolls, cakes, potato chips, doughnuts and soft drinks were designated less healthy choices. In the mains group, 48 schools sold pies or chips but not sandwiches and rolls, while only one school sold sandwiches and not pies and chips. In the snack group, 40 schools sold no healthy choices while only seven schools offered no unhealthy choices. When beverages were compared, nine schools offered soft drinks but no juice or milk, while 48 schools sold juice or milk but not soft drinks. The ratio of less healthy to more healthy main choices was 5.6:1, for snacks it was 9.3:1 and for drinks it was 1.4:1. The quantity of each food item sold per week in New Zealand primary/intermediate schools was Table 4: Estimated total quantity of food items sold each week in New Zealand primary and intermediate schools (years 1 8) Food item extrapolated from the total quantity sold by the sample schools and is shown in Table 4. The less healthy food items clearly dominate the estimated unit sales by more than 2:1, with the leading item being the pies sold each week. Socio-economic status and school size There were no consistent, significant effects of either of SES decile rating or school size on the environmental and canteen sales data. DISCUSSION Units sold each week Mains Pies Chips Sandwiches Filled rolls Snacks Potato chips Cakes and biscuits Sausage rolls Doughnuts Muffins and scones Fruit Yoghurt Drinks Soft drinks Milk Juice In this survey of a representative sample of New Zealand primary/intermediate schools, we found that the food environment was not conducive to healthy food choices for the children at those schools and that this was reflected in the high

5 Obesogenic elements in New Zealand primary schools 19 sales of relatively unhealthy foods from the school food services. High-fat items (pies, sausage rolls) were the most available foods, filled rolls were the most expensive food item, few schools had food policies, and there was a high neutral response to questions about the level of support for healthy food choices in the schools. This neutrality could reflect the lack of interest in nutrition and may be a major barrier to improving school food environments. The results presented here highlight the urgent need to promote healthy food policies and food choices in schools. The positive impact of the National Heart Foundation School Food Programme on food sales has been well documented (Carter and Swinburn, 1999). There is substantial evidence that school food programmes that take a strong environmental, yet comprehensive approach can influence eating patterns of children (Hider, 2001). There are a number of specific implications from this survey for improving school food. The lack of facilities for canteens restricts the ability to provide freshly prepared food, and favours prepackaged products and foods like pies and sausage rolls that can be frozen and reheated. Two-thirds of the schools operated their food service for profit and this would have further placed the priority on profitability rather than healthiness. New Zealand schools have always been exempt from food safety legislation and therefore have no regulatory requirement to provide safe food preparation facilities. This allows schools to sell food regardless of whether appropriate equipment or facilities are available. The potential impact of food safety legislation is unknown. One possibility of food safety legislation could be improved food service facilities able to provide a wide range of healthier foods. But equally likely is that stricter rules may further limit fresh food options. The apparent lack of school management support for the provision of healthy foods in the food service is of concern. Less healthy food choices are generally more convenient for the food service providers. Therefore the lack of school involvement in promoting healthy food choices means that the less healthy options dominate. The rating by 40% of respondents that the school provided foods with high nutritional value is at odds with the food sales data and is likely to be an overestimation. While schools had positive attitudes to nutrition as an issue, they did not necessarily see the school environment as playing a part in nutritional outcomes. This may be because teachers believe they have little influence on altering some of the aspects of the environment (i.e. the food service), and focus on addressing their concerns through the curriculum and the classroom. Successful school food programmes create links between the curriculum, the food service, school policy and parents (Centres for Disease Control, 1997). This is the first systematic attempt to measure indicators of the school food environment at a national level and it is appropriate to reflect on the performance of the chosen indicators. Measuring the types and quantities of food sold provided a clear measure of the school environment. However respondent burden was increased through obtaining food sales data from food service staff. Prospective data on sales may provide a higher degree of accuracy, but the sales data from this survey were surprisingly consistent across the country, with very few instances of higher sales of more healthy items compared with less healthy items. Measuring the price differential between more healthy and less healthy foods was a robust indicator, but needed to be interpreted in light of the sales data. For example, in the snacks group where prices were relatively similar for most items, the less healthy foods sold many more units than the more healthy items. In the mains group, pies and sandwiches were about the same price yet three times more pies were sold than sandwiches. More research is required to determine how price influences food sales in schools and how this impacts on profitability. As an indicator of the policy environment, the presence or absence of a food policy was an important but crude measure. The assessment of the level of effectiveness of the policy was very subjective because no definition of what constituted effectiveness was given. For those schools with policies, the strong focus was on the school food service. A more comprehensive policy to address the wider food environment may achieve a greater impact, although to measure this with any certainty, a more comprehensive audit of policy implementation will be needed. Measuring attitudes to nutrition by asking one person s interpretation of the overall school attitudes provided insightful information. Although the responses were generally positive, the praise was relatively faint for school support for healthy

6 20 M.-A. Carter and B. Swinburn food and nutrition, and this tallied with the generally unhealthy food choices offered by and bought from the food service. Improving the robustness of this measure may require a small group within the school to respond collectively rather than relying solely on the judgement of one person. In summary, the food environment in New Zealand primary schools appears to be quite obesogenic. Increased efforts to improve the school food environment are urgently needed and these should be broad based, incorporating all the elements of the school food environment including policies, canteen facilities and school community support. This study has shown the value of surveys of the school food environment, but further research is also needed to improve the validity and robustness of the measures used in such surveys. Address for correspondence: Boyd Swinburn School of Health Sciences Deakin University 221 Burwood Highway Victoria 3125 Australia swinburn@deakin.edu.au REFERENCES Cale, L. (1997) Promoting physical activity through the school. British Journal of Physical Education, Spring, Carter, M. A. and Swinburn, B. (1999) Measuring the impact of a school food programme on food sales in New Zealand. Health Promotion International, 14, Centres for Disease Control (1997) Guidelines for school health programs to promote lifelong healthy eating, Journal of School Health, 67, Crockett, S. and Sims, L. (1995) Environmental influences on children s eating. Journal of Nutrition Education, 27, Gill, T. (1997) Key issues in the prevention of obesity. British Medical Bulletin, 53, Glanz, K., Lankenau, B., Foerster, S., Temple, S., Mullis, R. and Schmid, T. (1995) Environmental and policy approaches to cardiovascular disease prevention through nutrition: opportunities for state and local action. Health Education Quarterly, 22, Hider, P. (2001) Environmental interventions to reduce energy intake or density. A critical appraisal of the literature [report]. New Zealand Health Technology Assessment, 4, 2. Ministry of Education (1999) Directory of New Zealand Schools and Tertiary Institutions. Data Analysis and Management Division, Ministry of Education, Wellington. Ministry of Health (1997) Food and Nutrition Guidelines for Healthy Adolescents: A Background Paper. Ministry of Health, Wellington. Parcel, G., Simons-Morton, B., O Hara, N., Baranowski, T. and Wilson, B. (1989) School promotion of healthful diet and physical activity: impact on learning outcomes and self reported behaviour. Health Education Quarterly, 16, Reniscow, K. (1993) School-based obesity prevention. Population versus high-risk interventions. Annals of the New York Academy of Science, 699, Shilton, T. and Corti, B. (1992) School heart health promotion. Hygie, 11, Swinburn, B. A., Egger, G. J. and Raza, F. (1999) Dissecting obesogenic environments: the development and application of a framework for identifying and prioritising environmental interventions for obesity. Preventative Medicine, 29,