Case study: Integrating platforms on food consumption and food composition

To ensure its recommendations are actionable, EuroDISH will perform two case studies on research infrastructures. The first case study focuses on the integration of two major existing platforms on food consumption and food composition (nutrient and other food components), relevant to enhancing comparable dietary assessment and monitoring throughout Europe.

IFR (Institute of Food Research, UK) and EuroFIR AISBL (European Food Information Resource, BE) have been working together with IARC (International Agency for Research on Cancer, World Health Organization, FR) on developing and testing an interface between the e-SMP (e-Standardized Methodological Platform, for standardised dietary intake data) and the EuroFIR Food Data Composition Platform (for harmonized nutrient and other food component data). E-SMP, which is still under development, is a platform intended to support the maintenance and dissemination of a highly standardised 24-hour dietary recall programme (EPIC-Soft) already implemented in seven countries (NL, BE, DE, FR, CH, AT, MT). These integrated platforms will facilitate future international studies by standardising dietary methodologies and allowing comparison of results among countries and across projects over time. Further aims are to improve cost-effectiveness in countries with limited local technical resources and skills in dietary assessment, to promote independence and flexibility, and to simplify communication and data synchronisation and (or) exchange.

As part of the development and testing phases, teams from JSI (Jozef Stefan Institute, SI), THL (National Institute for Health and Welfare, FI) and EU national food composition database compiler and (or) food consumption survey organisations (e.g. RIVM (National Institute for Public Health and the Environment) NL, MRI (Max Rubner-Institut) DE, CRA-NUT (Research Center on Food and Nutrition) IT and IFR UK) are mapping consumption data with composition information. If we are to understand what nutrients, bioactives and contaminants are being consumed and their effects on the health of European citizens, it is vital these datasets can be matched, despite different terminology and other technical barriers. Whilst this can be done manually, realistically it is more cost-effective to develop an algorithm for computer-assisted matching. However, the role of those experienced in dietary assessment will continue where terms cannot be matched automatically or need to be checked for the correct assignment. Preliminary results indicate that the quality of the algorithm matching is largely dependent on the completeness and quality of the composition databases, particularly in terms of indexing to LanguaL ('language of food', an automated method for describing, capturing and retrieving data about food) and the availability of food names in English. The initial results will provide feedback to improve the algorithm, and will help to identify gaps in the current composition databases, which, if addressed, could further improve the accuracy of matching.

The final task is to design and assess the potential for an interface between these two pre-existing e-platforms, potentially as part of larger food and health research infrastructure(s) including the other DISH pillars. Beneficiaries are discussing the potential and limitations as well as identifying barriers to implementation in the medium to long term (e.g. governance, access to data, technical requirements). Many of the national food consumption survey organisations in Europe are also national food composition database compiler organisations, meaning that a joint 'sustainability framework' could be developed.