Validity and Reproducibility of a Semi-Quantitative Food Frequency Questionnaire Adapted to an Israeli Population

Abstract

Assessment of individual representative dietary practices in epidemiologic research has long been a challenge. Given the differences in eating habits and availability of foods between populations, a study was set to test the reproducibility and validity of an Israeli-adapted semi-quantitative food frequency questionnaire (SFFQ) used in case-control studies of cancer etiology. Two hundred volunteers between the ages 22-60 were randomly assigned to either complete an SFFQ independently or by nutritionist interview. Participants were then asked to fill out a 3-day food record (FR). SSFQ was repeated 3 months later. Reliability of the repeated SSFQ was examined using Pearson correlations; paired t-tests and Mann-Whitney Rank Test were used to measure reliability and validity of the SFFQ. Overall, there was >85% agreement ± 1 category between the two encounters in over 75% of the food items. Percent of calories from proteins, fats and carbohydrates were similar for both self-administered and interviewed groups. The majority of correlations between the 1st and 2nd SSFQs for the micronutrients were above 0.75. No differences were noted in percent energy from proteins, fats or carbohydrates between self-administered and interviewer-assisted SSFQ and FR. The Israeli FFQ was found in our study to be highly reproducible and valid compared to a dietary record.