Alternatives in the collection and analysis of food frequency interview data

In addition to frequency of consumption, a food frequency questionnaire may assess amount consumed, often by using food models, and the stability of diet. A food frequency interview directed at preformed vitamin A and beta-carotene was administered to 130 cases and 309 controls in an ongoing population-based case-control study of lung cancer in New Mexico. The questionnaire measured frequency, amount, and stability of consumption for 55 food items. Different combinations of responses to these three types of questions were used to calculate indices of total vitamin A consumption. The index based on frequency alone had the lowest value and increased significantly when amount was combined with frequency. Consideration of past consumption had relatively little effect on absolute and relative estimates of intake. Spearman rank order correlations between index pairs were high. These results suggest that the use of frequency alone is appropriate when the objective of data collection is to establish subjects' relative intake of specific nutrients.