How did EATS validate the DHQ and short screeners?

At the beginning of the study, participants filled out a brief questionnaire that asked about body weight, smoking history, and physical activity. Then, during the course of a year, the participants completed four telephone-administered 24-hour dietary recalls, with one recall per season. Following these 24-hour recalls, participants were randomized into two groups, each of which filled out two mail-administered FFQs, one month apart. One group completed the DHQ and the 1995 NCI-Block Health Habits and History Questionnaire; the other group completed the DHQ and the Willett FFQ (purple version). These versions of these FFQs were among the most widely used FFQs in epidemiologic research at the time. Two subgroups of participants also filled out one of the two short screeners.

Study investigators compared the performance of the DHQ to that of the Block and Willett FFQs. To further compare the performance of the three FFQs, study investigators determined correlations between each of the FFQs and "true" intakes, which were estimated using a measurement error model based on the four 24-hour recalls. The short screeners were evaluated by comparing their performance against the DHQ and the 24-hour recalls.

Analyses of the results showed that the DHQ performed as well as or better than the other two FFQs (see Comparative Validation of the Block, Willett, and NCI Food Frequency Questionnaires). Both short screeners were useful in estimating median intakes of fruits and vegetables (see Performance of Two New Cognitively Enhanced Fruit and Vegetable Short Assessment Forms).