Abstract
A food frequency questionnaire (FFQ), which can be used in studies investigating relationship between diet and chronic diseases in a rural area of Korea, was developed and validated. Food items were selected from two sources ; (1) preliminary survey in a rural area by 24-hour recalls and (2) National Nutritional Survey of 1991. Sixty-five food items were finally selected based on the frequency of consumption and contributions to major nutrient intakes. Portion size of each food item was determined considering the mean and median values of the amounts consumed by subjects in the preliminary survey. Frequency of consumption was asked in nine categories ranging from 'more than three times a day' to 'almost never'. The newly developed FFQ was administered in 24-hour recalls conducted in different seasons in the same area after the administration of FFQ. Sixty-one subjects completed both FFQ and all three repeated 24-hour recalls. The results of the two different survey methods showed that mean daily intake levels of energy, carbohydrate, vitamin A and vitamin C were significantly higher in FFQ compared to 24-hour recalls(p<0.05). Intake levels of energy, protein, fat, carbohydrate and iron were significantly correlated by Pearson's correlation coefficients(p0.05). Ranking on nutrient intake of the subjects by two method were significantly correlated(Spearman's correlation coefficients) in all above nutrients plus Ca and vitamin A. Percentage of subjects in the lowest or in the highest quintile by 24-hour recalls who belong to the nearest two categories by FFQ ranged from 46% to 83%, while the percentage falling into the opposite category were below 10% in most of the nutrients. The proportion of subjects classfied into the same quintiles by the two methods were between 23% and 33%, with an average of 28%. From the results, FFQ developed in this study seems to be useful in evaluating nutrient intake pattern of 1 year in adults living in rural area of Kyonggi province.