The reproducibility of dietary data from a self-administered questionnaire. The Tromsø Study

The reproducibility of a self-administered questionnaire about dietary habits was assessed by comparing answers from 201 men and women to identical questions posed in two surveys separated by approximately one year. In spite of possible changes in the diet in the time period between the two surveys, the concordance between the information was high, both concerning type of food item most commonly eaten and the frequency with which the food items were consumed. The highest reproducibility was found for food items consumed habitually (e.g. alcohol) or often (e.g. coffee). The results are in accordance with most other similar studies, and support the use of self-administered questionnaires in nutritional epidemiology.