Structure of the standardized computerized 24-h diet recall interview used as reference method in the 22 centers participating in the EPIC project
Introduction
The European Prospective Investigation into Cancer and Nutrition (EPIC) is a large multicenter prospective cohort study involving ∼460 000 middle-aged men and women from nine European countries (Spain, Italy, France, Greece, Germany, The Netherlands, UK, Sweden and Denmark) [1]. Within each country, a dietary questionnaire was developed and validated to estimate the usual individual food intakes of the study subjects [2], [3], [4], [5], [6], [7], [8], [9], [10]. An internal calibration method was adopted, based on a single 24-h diet recall interview collected from a representative sample of ∼1200–5000 individuals per cohort. The rationale for this approach was to use the 24-h diet recall interview as a common reference method to adjust, at the group level, for systematic over- or underestimation of the mean food (or nutrient) intakes measured by the country-specific baseline questionnaires [11], [12], [13]. At the same time, this calibration method will be used to correct for attenuation bias when measuring the relative risk between diet and disease. The statistical power, to detect a real relationship between the dietary exposure and cancer or other chronic diseases, could thereby be increased.
One of the main statistical conditions for using the calibration approach in the EPIC project was that 24-h diet recall is highly standardized within and between countries. Despite some recent interest in the calibration concept in studies on diet [14], there is still little experience with the methodological and practical problems of standardizing dietary assessment methods within the context of large international multi-center studies. A software program named ‘EPIC-SOFT’ was therefore developed at the International Agency for Research on Cancer (IARC), in collaboration with the EPIC centers, to ensure the highest possible level of standardization of the 24-h diet recall interviews collected in the 22 EPIC centers.
The development of EPIC-SOFT was conducted in three main steps. The first stage was to identify the potential sources of errors associated with the 24-h diet recall measurements by carrying out a literature review [15], [16], [17], [18] and by analyzing a large pool of 24-h diet recalls collected during the EPIC pilot phases. The second stage was to develop the approaches to prevent or minimize each identified potential source of error associated with the 24-h diet recall in the international context of the EPIC study. The third stage was to develop the EPIC-SOFT program and country-specific versions prepared for each of the nine participating countries.
In this paper, we present the EPIC-SOFT program and its common structure and interview interface adopted to optimize the standardization of the dietary interview procedure within and between the EPIC centers. The rationale and methodological concepts developed for standardizing the computerized 24-h diet recall interviews are presented in another paper [19].
Section snippets
Main objectives and functions of the software
The main functions and logical structures implemented in the EPIC-SOFT program were designed to conduct interactive dietary interviews following a strictly standardized procedure within and between the EPIC centers:
- 1.
Information on all of the foods and beverages consumed during the recall day needed to be collected, entered, and coded automatically according to common rules.
- 2.
The software had to be user-friendly and convenient for use in large populations of different linguistic, socio–cultural,
General conceptual structure of EPIC-SOFT
The general structure of EPIC-SOFT (Fig. 1) was designed to pre-define and standardize the pathways to be followed during the 24-h diet recalls across all countries. This overall structure relies first on a common definition of ‘foods’, or ‘mixed recipes’ that might be reported by the subjects: a ‘food’ is a single item which is perceived and quantified as such (e.g. fruits, vegetables, fish, cakes, bread, sweets). A ‘mixed recipe’ is a complex dish where all the ingredients cannot be
Presentation of the EPIC-SOFT interface
The EPIC-SOFT interface was developed using the same basic design common to all country versions. This interface was conceived to equally guide all of the interviewers during the whole 24-h diet recall interview and minimize potential errors which might be due to varying degrees of knowledge of nutrition or computing. All common functions, in a given screen, are summarized in a pull down menu and can always be activated through the mouse, hot keys or using the keyboard (Fig. 2).
The 24-h diet
Maintenance/update of the EPIC-SOFT databases
Like any open-ended method, the computerized 24-h diet recall method needs regular update of its databases in order to add new foods, recipes, or other information reported by the study subjects. To maintain a high level of control and standardization of EPIC-SOFT databases and to facilitate updating, it was decided that only one version will be available per country and that any modifications to the EPIC-SOFT files will be centralized at IARC. In addition, a four-step hierarchical structure
EPIC-SOFT software and hardware requirements
EPIC-SOFT is written in Clipper. The system is PC IBM-compatible and needs ∼2 Mb of RAM memory and at least 5 Mb of hard disk space. MS-DOS version 3.00 or later is required.
Conclusion
In this paper we presented the general structure and principal functions of a new software (EPIC-SOFT) developed to obtain standardized 24-h diet recall interviews between the nine European countries involved in the EPIC project. It was decided to develop a computerized 24-h diet recall interview as it was felt that this was the most appropriate methodological approach to standardize the interview procedure [21], [22], and to minimize the potential sources of error associated with the 24-h diet
Summary
A computerized 24-h diet recall interview program (EPIC-SOFT) was developed for use as a common reference calibration method between nine European countries involved in a large prospective cohort study, EPIC. The EPIC-SOFT program, which was adapted for each participating country and translated into 9 languages (Danish, Dutch, English, French, German, Greek, Italian, Sweden and Spanish), was developed to perform highly standardized interactive interviews in 22 centers involved in the study. The
References (26)
- et al.
Computerized collection and analysis of dietary intake information
Comput. Methods Programs Biomed.
(1989) - et al.
Urine nitrogen as an independent validatory measure of dietary intake
Am. J. Clin. Nutr.
(1985) - et al.
Simultaneous measurement of seven carotenoids, retinol and alpha-tocopherol in serum by high performance liquid chromatography
J. Chromatogr.
(1997) - et al.
The EPIC project: Rationale and study design
Int. J. Epidemiol.
(1997) - et al.
Development of a semi-quantitative food frequency questionnaire to assess food, energy and nutrient intake in Denmark
Int. J. Epidemiol.
(1991) - et al.
Comparison of dietary assessment methods in nutritional epidemiology: weighted records versus 24 h recalls, food-frequency questionnaires and estimated-diet records
Br. J. Nutr.
(1994) - et al.
The Dutch EPIC food frequency questionnaire. I. Description of the questionnaire and validity and reproducibility for food groups
Int. J. Epidemiol.
(1997) - et al.
The reproducibility and relative validity of a food frequency questionnaire newly developed for the German part of the EPIC project. I. Food group intake
Int. J. Epidemiol.
(1997) - The EPIC group of Spain, Validity and reproducibility of a diet history questionnaire in Spain. I. Foods, Int. J....
- et al.
Reproducibility and validity of an extensive semi-quantitative food frequency questionnaire using dietary records and biochemical markers among Greek school teachers
Int. J. Epidemiol.
(1997)