The purpose of the present study was aimed to identify dish items applicable in developing dish?based food frequency questionnaire (DFFQ) for a hypertension study of Koreans. The 24-hour recall data of 4,401 subjects aged 20~65 years from the 2001 Korean National Health and Nutrition Examination Survey were used for the analysis. Logistic regression model was used to identify the nutrient related with hypertension. Energy, fiber, sodium, calcium, carotene, vitamin B1 and vitamin C were associated with hypertension. Selection the top 30 dish items for these seven nutrients was performed based on their degree of contribution in supplying nutrients in terms of the cumulative percent contribution (cPC), as well as on their degree of explanation for between?person variation, in terms of the cumulative regression coefficient (cMRC). Rice supplied 43% of total energy consumption. Korean cabbage and radish kimchi were two major sources of sodium and it also covered the 27% of between person variation of sodium intake. Soybean paste soup and single item of orange?colored fruits supplied 43% of total vitamin C intake for Koreans and it covered the almost 79% of between person variation among the Korean population. Korean cabbage kimchi was the major source of fiber, calcium, sodium, carotene, and vitamin C for Koreans. In summary, the top 30 selected dish items supplying the 78% to 89% of the 7 nutrients. Those items also covered the 79% to 94% of between person variation of the 7 nutrients consumption. Therefore, the selected 30 dish items in each categories of nutrient could be applicable in developing dish based food frequency questionnaire for hypertension study.
This study examined within- and between-person variation in nutrient intakes in order to estimate the degree of precision in dietary assessment among 37 males and 46 females whose mean age was 70.4 years old. To collect dietary data, each subject was interviewed 5 or 6 times using a 24-hour recall method during a 3 month period. Within-person variation ranged from 23.5% to 101.2%. Lower values of within-person variation were observed in energy, carbohydrated, protein and phosphorous while higher value was observed in vitamin A. Between person variations of nutrient intakes ranged 12.6-23.5% in most nutrients. With 1 day dietary data, observed nutrient intakes were estimated to within 6-25% of the group's usual9true) intakes and 52-198% of the individuals' usual intakes. The values of these maximum percentage deviations became smaller when the number of dietary recalls increased. The results of this study suggest that the Korean elderly subjects appear to have a slightly less diverse diet compared to young Korean women. Within persons, intakes of nutrients largely from animal sources were more variable for the Korean elderly than for their counterparts in Western countries. This study also implies that commonly used 1 day dietary study may be appropriate for assessing group means of nutrient intakes, but clearly not appropriate for assessing individual's nutrient intakes.