The need for a strict methodology in dietary surveys: the experience of the WHO Osteoporosis Project.

There are three basic dietary data collection tools: direct weighing, diet diaries and diet recalls. Each technique has inherent strengths and weaknesses. Direct weighing requires great cooperation from the participants, is expensive and time consuming; diary methods tend to distort usual intake patterns, while retrospective data collection methods are subject to respondent memory bias. Therefore, there are differences in study outcomes depending on the methodologies used; comparisons of data collected in different studies would only be possible if a standardized method were utilized. The Project on Osteoporosis of the World Health Organization is a cross-national study, designed to explore the differences in the incidence of hip fracture and to ascertain risk and protective factors for bone loss and fractures in the older populations in different countries. The objective of the diet assessment is to evaluate the role of diet on osteoporotic hip fracture, and on loss of bone mass. Major effort during the preparation of the proposal was devoted to the development of strategies for assessing the validity and reliability of the instruments selected and the comparability of the results obtained in the participating countries. Such strategies include re-interviews of the participants, proxy interviews and comparisons of data obtained by the food frequency method and those recorded in diaries.