MICA: Partnership for Improvement and Innovation in Dietary Assessment Technology (PIIDAT)

Studies of diet and disease relationships are difficult to carry out and often generate conflicting results. Problems occur because measuring diet in large populations is challenging. Measurement error happens in dietary studies for lots of reasons including lack of expertise, limited food choices on assessment tools, inaccurate portion sizes, incomplete food databases and cost. Health practitioners also need suitable tools for assessment of diet in patients for diagnostic purposes and monitoring treatments.

The main aim of this proposal is to improve the design, conduct, analysis and interpretation of diet studies reporting food & nutrient intakes to reduce measurement error and bias. In addition, the partnership will empower researchers and clinicians through provision of valid tools to measure food and nutrient intakes. This will be of value for a range of stakeholder groups, with the potential to make important public health impacts.

Currently, unless researchers or practitioners are experts in the area of nutrition it is extremely difficult for them to include measurement of dietary behaviour in their studies. We aim to provide a resource to meet the challenges of measuring diet accurately and efficiently that incorporates new developments in technology. This will be achieved by creating a unique website containing valid, evidence based dietary assessment tools along with guidance on use linked to an analysis system generating results of foods and nutrients consumed. This resource will provide access to the best assessment tools for use in a standardised way; increasing the quality of research which is undertaken in this area.

Tools for inclusion on the website will be chosen by a group of experts to provide trustworthy selection of the best dietary assessment tools. These could be questionnaires; use an interview format; or on-line completion. We expect initially 10-20 tools to be included on the website. Existing paper based questionnaires included will be converted to an on-line format with downloadable pdfs. We will establish new standards for collecting, analysing and reporting studies of diet and health. The web-site hosting the validated dietary assessment tools (DAT-eLibrary) will be interactive and provide guidance for choice of the most appropriate tool for use; printable questionnaires and forms; on-line data entry and analysis functions.

At the same time, we will expand and improve existing food composition tables for the UK. The growing diversity of the UK diet with around 70,000 food items in stores has not been matched by an expansion of the standard food composition tables, with only ~7,000 items. We plan to link existing national public & food industry branded food tables to expand the choice of foods available. We will use back of pack nutrient information from food producers, fast food retailers & supermarkets to expand the food tables. National public and food industry branded food tables will be linked to support analysis of tools in the DAT-eLibrary. International food tables can also be incorporated.

The final element of the project will be to create an on-line interface between the dietary assessment tools and the food composition tables. The food tables will be linked to the tools in the DAT-eLibrary. The interface will allow users to access the tools and undertake analysis. Access will either be free or minimal cost to support sustainability of the resource keeping the database up-to-date.

This partnership will be a unique resource for nutrition researchers, scientists and health practitioners of value to both experts and non-experts. No other website combines in one place the resources proposed here, making this proposal the most comprehensive place to find support for research in nutrition & dietary assessment. Having a central on-line resource will be a cost effective approach to improving the quality of research in the diet and health area.

Objective 1: Establish new collection, analysis & reporting standards
An Expert Group (co-apps) will develop requirements for researchers undertaking & reporting dietary studies. Guidance for choice of tool will be established; reporting requirements will be published as a reference paper following CONSORT/STROBE. The group will be informed by a systematic literature review of validation studies assessing dietary intake &/or meal pattern. This will identify potential dietary assessment methods for inclusion in the Dietary Assessment Tool-eLibrary.

Objective 2: Creation of website hosting dietary assessment tools (DAT-eLibrary)
An open access web-based library of valid dietary assessment tools will be created. The website will include guidance for tool selection; the tools; online data entry; food/nutrient analysis & reporting. The number of tools included will depend on Expert Group selection, but ~10-20 tools will initally be available. The website will include an e-algorithm for users to aid decisions regarding selection of the best tool.

Objective 3: Expansion of existing UK food composition tables
Existing national food tables will be linked to food industry branded data. This will expand the tables to ~70,000 foods from 7,000. Date stamping will allow for future reformulation of products. We will use & validate back of pack nutrient information to expand the food tables. These tables will be linked to the tools in the DAT-eLibrary.

Objective 4: Creation of on-line interface between the dietary assessment tools and the food composition tables
This interface will enable users to access the tools, code dietary data and generate nutrient/food intake results. Providers of tools will be supported to prepare datasets of foods & nutrients along with associated analysis algorithms. These will be linked to the expanded food tables & a front-end web-based data entry system. The website will contain sections for both researchers and respondents.

Who will benefit?
Nutrition researchers, other scientists, healthcare practitioners & policy makers could all benefit. The list of beneficiaries is exemplified through attached supporting letters. These organisations include those delivering policy & practice - WHO, NICE, DoH, NOO; research funders and NGOs - Research Councils, NIHR, WCRF; research organisations - IARC, Cochrane, MRC Units, Universities. The PIIDAT will have international relevance and is supported by IARC (Slimani), WHO (Branca) & US NCI (Subar). Potential users of different elements of the resource include academia, industry, public sector funding bodies and regulators, as well as clinicians, dietitians, students and the general public. Brandbank & M&S are industrial partners with potential benefits supporting use of food/nutrient information; improving the quality of food labelling & standards of reporting of food research.

How will they benefit?
The PIIDAT will have a major impact on understanding the contribution of diet to risk of chronic diseases associated with dietary intake such as cancer, CHD and diabetes, & therefore to future preventive strategies. It will enhance aetiological research, supporting behaviour change interventions through monitoring diet. For example, ~40% of the population attributable fraction of major cancer types could be prevented by choice of appropriate foods, nutrition or physical activity (AICR, 2010).

New data sources will be created: 1) expanded & enhanced food composition tables, 2) DAT-eLibrary - a repository of valid dietary assessment tools, 3) interactive website containing resources 1 + 2 above plus Expert Group guidance and algorithms supporting choice of the most appropriate tool. The project will support researchers to generate higher quality outcomes from studies exploring diet and health; & clinicians enabling monitoring of diet in clinical groups. The interactive website will guide users to the most appropriate tool; providing the resources needed to generate food and nutrient results from data collected. A User Contract completed prior to use of the tools will support appropriate tool use and check on ethical approval. Expert guidance will support data collection and reporting of results to strengthen the evidence base linking diet and health and to facilitate rigorous pooling of data.

The Partnership will have high utility for systematic reviewers, evaluating the quality of methods when assessing risk of bias in studies. Developing a library of formally validated dietary assessment tools will expedite reviews, enabling reviewers to access an objective assessment of the instrument taking into account the history, validation & prior uses of the tool. Further, use of the DAT-eLibrary by researchers will potentially increase the consistency of methods and reporting of dietary data in primary studies, facilitating meta-analyses in systematic reviews, with more studies using the same measures. There has been a call for more consistent methods across most areas of Cochrane reviews. A longer term aim of PIIDAT is to see if better support for researchers & review authors, in method selection & quality assessments of instruments, results in quicker systematic reviews; containing studies with more consistent methods enabling stronger findings to be made. If successful, this approach could be extended to other areas such as measurement of physical activity.

The food composition table will be compliant with food tables in the EuroFIR project; facilitating use by researchers in 30+ countries. EuroFIR provides an interface to search multiple food composition databases. International tables can be linked to the resource supporting analysis of assessment tools from around the world.

Future developments will allow the general public to have access to the food database so that they can check food & nutrient information of products; and provision of weight loss & dietary behaviour modification resources.