Results from a comparative dietary assessment in Europe: II. Feasibility of pooling individual-based dietary data between countries.

Dietary investigations in four central European survey populations carried out in the German Democratic Republic, Poland and Denmark between 1982 and 1984 using different methodologies were analysed in order to assess the possibilities of characterizing the dietary habits of individual survey participants in a comparable fashion. This was done with the view of assessing the feasibility of a pooled cancer cohort study. For this purpose a method has been devised to combine dietary information derived by food frequency questionnaires and quantitative recording methods into a quantitative characterization of individuals' habits. A comparable characterization between different cultural settings could be demonstrated for a selected list of food items. The selection was determined by the food items considered in common in the different food frequency questionnaires and yielding sufficient and comparable variation as well as absolute amounts of intake. This was more clearly found for food items such as 'fruit' which experience a concise role in dietary habits. However, the observed discrepancies of the different dietary methods within the countries, and, most importantly, between the countries, result in distributions of average daily consumption values which are not deemed to be comparable.