We examined the relative validity of food intake for each meal type (breakfast, lunch, dinner, and snacks) and overall food intake obtained through the Meal-based Diet History Questionnaire (MDHQ). In total, 222 Japanese adults (111 for each sex) aged 30-76 years completed the web version of the MDHQ and then the 4-non-consecutive-day weighed dietary record (DR). The number of major food groups (n = 24 in total) for which no statistically significant difference was observed between median intakes estimated using the DR and MDHQ ranged from 8 (snacks) to 12 (dinner) among women, and from 8 (breakfast) to 13 (lunch) among men. The median values of the Spearman's correlation coefficients between the MDHQ and DR estimates ranged from 0.28 (dinner) to 0.54 (breakfast) among women, and from 0.24 (dinner) and 0.60 (breakfast) among men. Bland-Altman analyses generally showed wide limits of agreement and proportional bias. Similar results were obtained using the paper version of the MDHQ, which was completed after conducting the DR. In conclusion, the MDHQ has a satisfactory ability to estimate median intake and rank individuals according to consumption for many food groups, despite a limited ability to estimate food group intakes on an individual level.
Keywords: Japan; diet; dietary record; food; meal; questionnaire; snack; validity.