Development of a food frequency and quantity method for assessing dietary habits of Japanese individuals--comparison with results from 24hr recall dietary survey.

In a lifestyle modification program for the prevention of cardiovascular disease, quantitative assessment of the diet is essential. We developed a new food frequency and quantity method using portion-sized food models to apply in a lifestyle modification program. We conducted a dietary survey using this method of Japanese middle-aged men and women (n=76) who had undergone four 24hr dietary recalls one year before, and compared the results from the two surveys. Total energy intake and gram intake of macronutrients determined by our method were significantly lower than those determined by 24hr recalls. Correlation coefficients between nutrient intake obtained by the two methods were significant: r=0.532 (p<0.001) for carbohydrate, r=0.327 (p=0.004) for protein, r=0.334 (p=0.003) for total fat, r=0.419 (p<0.001) for saturated fatty acid, and r=0.230 (p=0.046) for polyunsaturated fatty acid. Correlations of nutrient intake were significantly increased with energy adjustment (p<0.05). Correlation coefficients between major food group intake were also significant: r=0.682 (p<0.001) for cereals, r=0.261 (p=0.023) for meat, and r=0.333 (p=0.003) for fish and shellfish. This method was thought to be appropriate for quantitative assessment of the intake of major nutrients and food groups of individuals in Japan.