An evaluation of the impact of school food standards in England on children's diets

Abstract

Many children do not eat a healthy diet. In 2005, the nutritional content of school lunches in England received wide criticism. In 2006, a major policy change led to legislation specifying what food and drink could, and could not, be served in schools. This thesis considers the impact of the implementation of food and nutrient-based standards on children’s dietary intake at lunchtime and in their total diet, if the impact was equitable across the socio-economic spectrum, and if school lunch take-up changed. Methods Data collected pre and post-policy implementation in children aged 4-7y and 11- 12y were analysed. In the 4-7y olds, dietary data were collected on four consecutive days using an observational method in 12 primary schools, in Newcastle upon Tyne, UK (n=385 in 2003-4; n=632 in 2008-9). In 11-12y olds, dietary data were collected from two consecutive 3-day food diaries followed by a researcher-led interview in six middle schools, in Northumberland, UK (n=298 in 1999-2000; n=215 in 2009-10). Linear mixed effect models were used to analyse the effects of year (pre and post-policy), lunch type (school or homepacked lunch), level of socio-economic deprivation, and the interaction(s) between these factors on children’s total dietary intake. Logistic regression was used to examine the change in school lunch take-up by year and level of deprivation. Results At lunchtime, children who ate a school lunch post-policy implementation consumed a lower per cent energy from fat, saturated fat and absolute amounts of sodium. In the 4-7y olds, mean calcium (mg), vitamin C (mg) and iron (mg) intakes increased; in 11-12y olds, non-starch polysaccharides (g) and iron (mg) decreased. A child’s lunch type was associated with change in the total dietary intake in 4-7y olds; post-policy implementation children eating a school lunch had a healthier total diet compared with children eating a home-packed lunch. In 11-12y olds, there was limited evidence found that lunch type was associated with change in total diet. In both age groups children’s total dietary intake from ii | per cent energy saturated fat and non-milk extrinsic sugars remained above the recommended guidelines. There was some evidence that post-policy implementation, lunch type and level of deprivation were associated with differences in per cent energy from non-milk extrinsic sugars and vitamin C (mg) intake in the total diet of 4-7y olds; there was no such evidence found in 11-12y olds. Post-policy implementation, school lunch take-up decreased in both age groups. Conclusions The implementation of school food and nutrient-based standards in England has been associated with positive changes in children’s dietary intake at lunchtime. These changes were reflected in the total diets of the 4-7y olds but evidence was more limited in 11-12y olds. A key strength of this study is the unique evaluation of national policy enabled by the availability of preimplementation data. A key limitation is the use of repeat cross-sectional surveys; this limits the extent to which change in children’s diets can be attributed to the policy. Future regulation of school lunches should be evaluated prospectively. To improve children’s diets in all their complexity, future interventions also need to consider the social, environmental and behavioural contexts in which food choices are made or directed, both in and outside of the school environment.