In 1972 a working party of the Group of European Nutritionists met in Zurich (Switzerland) to consider the principles that should be used in preparing national tables of food composition. A small book based on a working paper for this conference and describing guidelines for the preparation of such tables was subsequently published (Southgate, 1974).
During those discussions it became clear that in the future more tables providing international coverage (e.g. for all of Europe) would be needed. Since then, widespread advances in computer techniques have made the creation of such international databases technically feasible; their development is impeded, however, by the variable analytical quality, the incompatibilities, and even the unknown provenance of existing compositional data. Furthermore, large areas of the world remain where little information on food composition is available.
In 1983 a conference was held at Bellagio (Italy), under the auspices of the United Nations University, to identify the tasks that needed to be carried out in order for internationally valid, consistent and usable food composition data to become available. During the discussions the creation of an International Network of Food Data Systems (INFOODS) was proposed (Rand and Young, 1983).
One of the first tasks for INFOODS was to revise and extend the earlier Southgate (1974) guidelines, which addressed issues relevant to the central problem of data quality and compatibility. Accordingly, one of us (HG) spent four months as an INFOODS Fellow working with the original guidelines' author (DATS) at the Food Research Institute in Norwich (United Kingdom). This initial work, continued and completed by correspondence, drew information from production and management of food composition data in the United Kingdom and United States and from Australian experience of producing data. In January 1985, a partially completed version was reviewed by a working group in Washington, DC (United States). A revised version, prepared on the basis of this review, was reviewed again by a number of international authorities; their comments were used in the version prepared in 1986.
After reviews by experts in the computer field, and considerable inputs from participants in the FLAIR Concerted Action No. 12 EUROFOODS-Enfant Project, the final revised version was prepared by correspondence and meetings between the authors while one of us (HG) was a Visiting Scientist at the International Agency for Research on Cancer (IARC), Lyon (France) in connection with the Nutrition and Cancer Programme.
In preparing a document of this kind, personal feelings and prejudices inevitably emerge; they are the responsibilities of the authors alone, who nonetheless beg their readers to remember that these idiosyncrasies developed during lengthy consideration of nutritional compositional data, their production and use.