ObjectiveTo validate an integrated food-frequency questionnaire (FFQ) developed to assess habitual food intake of Flemish and Italian-native subjects in Belgium as part of the European Collaborative Dietary Habit Profile in European Communities With Different Risk of Myocardial Infarction: the Impact of Migration as a Model of Gene/Environment Interaction (IMMIDIET Project).Methods
The semiquantitative FFQ contained 322 items on food and food preparation. FFQs filled by a sample (n = 70) of the Flemish-Flemish and Flemish-Italian IMMIDIET subpopulations were randomly selected. Five 24-h recalls, administered over a period of 1 y by the same sample, served for validation. Energy and macronutrients were calculated using the Dutch NEVO and the Belgian NUBEL food composition tables. Intakes of energy and macronutrients estimated by the FFQ and repeated 24-h recall, respectively, were compared by means of correlation coefficients, classification into quartiles, and Bland-Altman plotting.
Results
The FFQ overestimated intake of energy and most macronutrients by 40–70 % . This overestimation largely disappeared when values were expressed as energy percentage. Correlations ranked from 0.40 to 0.60 for energy and most macronutrients (median 0.53); correlations were lower (null to 0.41) for fat and higher (up to 0.90) for alcohol. Classification in quartiles of intake showed good agreement: 83 % were classified in the same or adjacent quartile of energy, and 66–90 % for macronutrients. Correlations and classification of macronutrient intake into quartiles remained similar when macronutrients were expressed as energy percentage. Stratification according to ethnic subgroup, age, body mass index, or social status showed no differences.
Conclusion
The IMMIDIET FFQ is a valuable tool for studies of the role of energy and macronutrients in disease etiology or outcome, but less suitable for estimating absolute intake levels.
