Short-term stability of dietary patterns defined a priori or a posterior

ObjectivesThe aim of this work was to evaluate whether the short-term stability (repeatability) of dietary patterns derived using a priori and a posterior approaches is affected by food items or food groups selected for the analysis. A priori methods seek to determine a person's adherence to a pre-specified (usually ‘healthy’) diet, whereas a posterior methods are used to record a person's (or more generally a population's) actual diet.Design and main outcome measuresTo test the a priori approach, the MedDietScore was selected as an index of adherence to the Mediterranean diet. A sample of 500 participants (aged 18–82 years, 38% male) completed a 76-item food frequency questionnaire (FFQ), while another 250 participants (aged 17–76 years, 34% male) completed a 36-item FFQ and the MedDietScore, twice within a 15-day interval. The MedDietScore was additionally indirectly calculated from participants’ 76-item FFQ and 36-item FFQ answers. To test the a posterior approach, principal-components analysis (PCA) was used to derive dietary patterns. PCA was applied to data derived from the 76-item FFQ, using as independent variables first the 50 food items recorded on the FFQ and second 12 food groups (i.e., aggregated data) derived from the FFQ items.ResultsBoth the indirectly calculated (i.e., through the two FFQs) and the directly recorded MedDietScore mean values were similar between the two administrations (p-values > 0.05). In addition, PCA using the 50 food items revealed three major dietary patterns that were similar in both administrations, explaining 23% and 25% of the dietary intake variability, respectively, while PCA using 12 food groups also revealed three similar patterns between the two administrations (explaining 43% and 46% of the dietary intake variability, respectively).ConclusionsThe results indicate the short-term stability of dietary patterns derived using either a priori or a posterior methods. For the a posterior approach the use of food groups, instead of food items, seems to explain more variation in dietary intake.