A UK on-line 24h dietary recall tool for population studies: development, validation and practical application

Research that aims to discover whether what we eat affects our health has been hampered by lack of accurate tools to collect information on diet. Such a measurement tool would need to collect and analyse data in a reasonable time frame, with enough detail, flexible enough for different ages and ethnic groups and at minimal cost. At the moment there is no such dietary assessment tool available in the UK. This study will make use of the internet to develop a new UK focussed dietary assessment tool based on the Automated Multiple Pass Method (AMPM) developed in the USA. The AMPM uses 5 cycles of questioning to obtain a detailed account of food eaten in the past 24 hours. A search engine allows participants to select from several thousand food items in an extensive food database.

This project has three main components. 1) Development of a new UK on-line dietary assessment tool. Existing online tools from the US and France (ASA24, DietDay, Nutrinet Sante) will be compared using focus groups. 30 potential users will be divided into 3 groups according to age: adolescent; adult; elderly (post-retirement). Following this the new UK tool will be developed based on the American ASA24 tool as template. 2) The second part of the study is a validation study, to see how the new tool compares to existing ?gold standard? approaches. We will compare food and nutrient intakes assessed by the on-line tool with foods and nutrients obtained using an interviewer administered dietary recall (the AMPM) both in comparison to independent biomarkers of nutrients: specifically urinary N (for protein); potassium; sucrose and fructose; and total energy expenditure. 200 participants will provide repeated measurements over 6 months. We will look at differences in agreement between the biomarkers and nutrient intakes as assessed by the average of measurements from both the on-line recall and the AMPM. 3) The final part will be a pilot study to see how well the new tool will work in practise with two different epidemiological cohorts. A subgroup of 200 older women in the UK Women?s Cohort plus their adult children and adolescent grandchildren and 100 participants from the Clinical Biobank at Imperial, a clinic based cohort with mixed population characteristics, will complete 3 on-line recalls over 6 months.

Dissemination of the tool is key and open access vital. We intend to provide the software free of charge to other UK researchers.

Reliable assessment of the association between diet and health requires accurate and repeated measurements of diet in large prospective cohort studies. Incorporation of an internet based 24hr recall into such studies may substantially advance our understanding of the nutritional determinants of disease. Existing tools are limited in scope, and do not lend themselves to easy adaptation for different population groups due to their reliance on previously defined lists of foods. Presently, there is no on-line 24hr dietary recall tool available in the UK based on the principles of the Automated Multiple Pass Method, which could be suitable for use by groups from different ethnicities and of different socio-economic classes. We aim to develop, validate and apply this tool in practice.

This project has three main components: 1) Development of a UK on-line 24hr dietary recall. Existing online tools from the US and France (ASA24, DietDay, Nutrinet Sante) will be compared using focus groups, 30 potential users will be divided into 3 groups according to age: adolescent; adult; elderly (post-retirement). Following this the new UK tool will be developed based on the ASA24 as template. 2) Validation of foods and nutrients measured by this new on-line tool by comparing it with foods and nutrients obtained using an interviewer administered multiple pass recall method(AMPM) both in comparison to independent biomarkers of nutrients: specifically urinary N (for protein); potassium; sucrose and fructose; and total energy expenditure from Actiheart. 200 participants will provide repeated measurements over 6 months. Primary analysis will assess differences in agreement between the urinary N biomarker and protein intakes as assessed by the average of the measurements from both the on-line recall and the AMPM. 3) Pilot study for practical application of the on-line 24hr dietary recall in two different epidemiological cohorts. A subgroup of 200 older women in the UK Women?s Cohort plus their adult children and adolescent grandchildren will be asked to complete 3 on-line recalls over a 6 month period. 100 participants from the Clinical Biobank at Imperial, a clinic based cohort with mixed population characteristics will be asked to do the same. Dissemination of the tool is key and open access vital. We intend to provide the software free of charge to other UK researchers.