Background: Obesity and overweight are increasing problems to the health of Western populations, both in adults and children. Overweight or obesity in childhood are predictors of obesity and various chronic diseases in adulthood. Constituents of the diet, in particular dietary energy intake, are at least part of the reason for this, so the measurement of childhood diet has considerable public health importance. However, the assessment of children's diets is usually costly and often inaccurate. Aims: This PhD involved the design, development and testing of a novel website method of assessment of dietary energy and macronutrient intake for school-age children, with the aim of maximising participation and completion rates, while keeping costs lower than those associated with paper-and-ink survey techniques, such as food diaries. The results from this new method were tested for their agreement with the results obtained from a 5-day food diary used in the same subjects. Methods: Participants were 164 children aged 9-10 years from a range of schools in the North East of England. The website's content was based on recent evidence of important sources of energy in this age group, and its design was informed by research into children's memory and reporting of foods and drinks. Colour photographs of a range of foods were provided on a website, selected from keyage-specific references and refined during a pilot study, along with time-cued questions regarding their usual diet. A computer database automatically compiled and coded each child's responses, in order to produce an estimate of mean daily energy and macronutrient intakes. The food diary was adapted from an existing tool which has been used previously in similar participants, and mean daily nutrient intake calculations were made using the same database as the website. The two methods were compared in their calculated mean daily energy and macronutrient intakes, and their relative cost of use. Results: 154 children completed both the website and the diary in full. Significant correlations were found between the two methods' ranking of mean daily energy and macro nutrients intakes (r=0.193 p<0.05, to r=0.230 p<0.01). However, this did not reach the levels of correlation identified as adequate to make the new tool useful in public health research, i.e. correlation above r=0.5, more than 50% agreement on tertiles of intake, and weighted kappa greater than k=O.4. The mean daily energy intakes reported to the website and the diary were 14.46MJ and 6.04MJ respectively. Bland-Altman (difference) plots and limits of agreement modelling revealed systematic under-reporting to the website at lower levels of intake of all nutrients, and over-reporting at high levels, when compared with the diary. Analysis of sources of energy showed closer correlation between the two techniques (r=0.245 to 0.351, all p<O.O 1), but again this fell short of the specified threshold. A simple economic analysis showed substantial cost savings of the website over the diary technique. Discussion: The analysis of the novel website method of dietary assessment showed some strengths, such as high participation rate in the study and low but significant correlation with the more established technique. However, important limitations including a lack of agreement between the website and the diary, with the appearance of over-reporting in the website method mean this tool cannot be recommended to measure energy and macronutrient intake in children in its current form. Possible extensions and modifications to the website are suggested, including some which would aim to address the over-reporting, and hence improve its usefulness as a tool in public health research.
| Identifer | oai:union.ndltd.org:bl.uk/oai:ethos.bl.uk:517480 |
| Date | January 2008 |
| Creators | Taylor, Mark Steven |
| Publisher | Teesside University |
| Source Sets | Ethos UK |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
Page generated in 0.0021 seconds