Intended for healthcare professionals

Primary Care

School dinners and markers of cardiovascular health and type 2 diabetes in 13-16 year olds: cross sectional study

BMJ 2005; 331 doi: (Published 03 November 2005) Cite this as: BMJ 2005;331:1060
  1. P H Whincup (pwhincup{at}, professor of cardiovascular epidemiology1,
  2. C G Owen, senior research fellow in epidemiology1,
  3. N Sattar, professor of metabolic medicine2,
  4. D G Cook, professor of epidemiology1
  1. 1 Division of Community Health Sciences, St George's, University of London, London SW17 0RE
  2. 2 Department of Vascular Biochemistry, Division of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow G31 2ER
  1. Correspondence to: P H Whincup
  • Accepted 8 September 2005


Recent concern about the diets of British children and adolescents has focused on the nutritional content of school dinners.1 However, whether markers of nutrition, cardiovascular health, and type 2 diabetes differ between school pupils who eat school dinners and those whose school day meal is provided from home has been little studied. We have examined this question in a survey of state secondary school pupils.

Participants, methods, and results

The ten towns heart health study (third phase) took place in 72 secondary schools across England and Wales in 1998-2000.2 3 Parents gave written consent (response rate 66%). We assessed height, weight, waist and hip circumference, skinfold thicknesses, bioimpedance measurement (percentage body fat), and pubertal status.24 We measured blood pressure, and participants provided blood samples for measurements of plasma glucose, serum insulin, and blood lipids after an overnight fast.2 3 We measured serum leptin by radioimmunoassay, plasma vitamin C by high performance liquid chromatography, and serum folate by microbiological assay. Participants indicated whether they usually ate a school dinner, had a meal from home, or made other eating arrangements. Parental occupation was provided by the parent (75%) or the participant (25%) to determine household social class. We used standard linear modelling procedures to determine adjusted means, log transforming variables when necessary. We fitted town as a fixed effect and school as a random effect to allow for clustering at school level.

The table compares the characteristics of 1112 pupils (53% boys) who usually ate a school dinner or a meal from home; results are standardised for town, school, age, and sex. Among pupils who ate school dinners, anthropometric markers of adiposity were slightly but not significantly lower; mean levels of leptin, systolic blood pressure, ratio of total cholesterol to high density lipoprotein cholesterol, glucose, insulin, and folate were significantly lower in this group. Although pupils whose parents were in unskilled occupations or unemployed were more likely to eat school dinners than those from other social groups (66% v 38%, P < 0.0001), the differences (apart from those in systolic blood pressure) remained statistically significant after adjustment for social class. Additional adjustment for pubertal status and physical activity level had no appreciable effect. Restricting the analysis of the school dinner group to pupils who were eligible for free school meals did not materially affect the results.

Comparison of nutritional markers and risk factors among pupils eating school dinners and meals from home

View this table:


The differences in risk factor profile and nutritional status between pupils eating school dinners and those eating home meals were modest. Their long term importance remains uncertain, although if the differences in blood lipids, insulin, and leptin persist with increasing age these could be of public health importance. As the provision of school dinners has changed little in the past five years,1 the differences could be of continuing relevance. However, the extent to which the differences reflect dissimilarities in the composition of school dinners and home provided school meals or other aspects of the dietary patterns and health behaviours of the two groups remains uncertain.

Despite these uncertainties, we can draw two general conclusions. Firstly, the lower mean folate concentration seen among pupils eating school dinners suggests that new initiatives likely to increase the folate content of school dinners would be appropriate.5 Secondly, the other differences seen suggest that the average health status of pupils eating school dinners is no worse—and may actually be better—than that of pupils eating meals provided from home. This suggests that efforts to improve the diet and nutrition of British children and adolescents will need to extend beyond school dinners to tackle overall dietary patterns and their societal determinants if they are to be successful.

What is already known on this topic

The content of school dinners has been a cause for concern

Little information is available on the health of school pupils who do and do not eat school dinners

What this study adds

Several markers associated with chronic disease risk were potentially more favourable in pupils who ate school dinners, although serum folate concentrations were also lower

This article was posted on on 6 October 2005:

We are grateful to the research team members and to the participating schools and pupils. We thank P M Ueland and H Refsum (department of pharmacology, University of Bergen) for the measurement of serum folate.


  • Contributors PHW developed the idea for this report, conducted the study with assistance from DGC, and drafted the paper. CGO did the statistical analysis. NS provided analyses of leptin. All authors contributed to the final version of the paper. PHW is the guarantor.

  • Funding This study was supported by a Wellcome Trust project grant (No 051187/Z/97/A). CGO is supported by the British Heart Foundation (project grant No PG/04/072).

  • Competing interests None declared.

  • Ethical approval Ethical approval was obtained from all relevant local research ethics committees.



View Abstract