The validity of a short food frequency questionnaire and its ability to measure changes in food intake: a longitudinal study.

BACKGROUND In studies of health behaviour exploring factors associated with differences and changes in eating patterns of populations, diet is often measured with short food frequency questionnaires (FFQ). This study examines the validity of a short FFQ by comparing frequencies of food intake from the FFQ to information on food intake obtained by a diet history interview. METHODS Food intake was measured at two separate occasions in the same 329 individuals, first in 1987-1988 and 6 years later in 1993-1994. RESULTS In 1987-1988 the Spearman correlation coefficients were around r = 0.50 for most foods, with white and dark ryebread and light bread as extremes on the one hand (r = 0.10, r = 0.23 and r = 0.27, respectively) and coarse bread, fruit and cakes as extremes on the other (r = 0.61, r = 0.60, r = 0.60, respectively). In general, the correlations were higher at the second data collection in 1993-1994. At both data collections, the mean food intake from the diet history interview increased with increasing frequency category, indicating that the questionnaire was able to identify levels of food intake correctly. In general, when individual changes in food intake were assessed during the study period, those who reported a less frequent intake by the FFQ in 1993-1994 compared with 1987-1988 also had a lower mean daily intake according to the diet history information. CONCLUSION The short FFQ can quantify food intakes and, is also responsive to changes in food intake over time. Thus the short FFQ can be used to monitor changes in food patterns at a group level.