Methodology of a two-step quantified nutritional questionnaire and its effect on results

The effects of methodology on the results of epidemiologic studies that involve collection of nutritional data, especially those involving long-term-onset illnesses such as cancer, have not yet been carefully evaluated. We present methodological features of a quantitative dietary history and physical activity questionnaire and discuss their contribution to the final results. The results of our analysis are as follows: 1) Forty-seven percent of the population consumed > 100 food items yearly. 2) The mean number of calories contributed by items eaten less than once a week exceeded 200 kcal/day in 50% of interviewees. 3) Seventy-six percent of the patients had undergone dietary changes during the course of adult life. Of the 379 individuals who reported no changes when asked general questions, 61.8% reported changes when asked specifically about each food item in the questionnaire. 4) Physical activity was significantly correlated with mean daily energy intake (r = 0.208, p < 0.001). We conclude that certain methodological features of questionnaires, such as extensive listing of food items, precise documentation of food quantity, inquiries about former eating habits, and inclusion of questions about physical activity, increase accuracy in evaluations of dietary habits. Our analysis provides practical information for future planning of nutritional questionnaires.