Dietary Sources of Nutrient Consumption in a Rural Japanese Population

We determined the sources of nutrient intake of 59 men and 60 women in two rural towns in the Miyagi Prefecture, a northeastern part of Japan. Four 3-day food records were collected in four seasons within a year. The total dishes and recipes were classified into 197 items. Their percent contributions to the total population consumption of energy and 14 nutrients were calculated as the sum of the nutrient intake contributed by a given dish or recipe divided by the total nutrient intake from all the items. Rice was the largest contributor for energy (29.8%), protein (13.0%) and carbohydrates (45.3%). Miso soup, as a dish, was a leading contributor (7.1%) for fat. The largest contributor for sodium, calcium, carotene, vitamin C were miso soup (17.1%), milk (16.6%), spinach (23.6%), green tea (13.6%), respectively. The result suggests that the examination of nutrient sources based on dishes and recipes, rather than on food materials, may be useful in characterizing the dietary patterns of populations.


INTRODUCTION
Public health workers often give nutritional education in terms of nutrient contents of dishes and recipes. Therefore examining the contribution of individual dishes and recipes, rather than food materials, is important because the dishes and recipes are the unit of eating.
Several studies have reported dietary sources of nutrient consumption among populations in the US 1-4), Italy 5.6), UK 7), Germany 8) and Japan 9-2). Some studies examined or compared dietary sources of different groups of general population [13][14][15][16][17][18]. However, most studies examined the contribution of foods or food materials, rather than dishes and recipes.
In this study, based on diet records collected from 119 subjects in rural Japan, we classified all the dishes and recipes into 197 items to study the percent contribution for the total popula-tion intake of 15 nutrient variables. We also compared our results with studies conducted in western and the Japanese populations.

Study design and subjects
The subjects were residents in two rural towns in Miyagi prefecture in the northern part of the mainland Japan, who were a subsample of participants in the two cohort studies that were started in 1990 19) and 1994 20). Fifty-nine men (mean age=62.5 8.4SD) and 60 women (mean age=61.2 8.5SD) were selected on the voluntary basis. Written informed consent was obtained from all the subjects.

Dietary Sources of Nutrient Consumption
Dietary records Diet records were collected on three consecutive days four times in a year during November , 1996 through November, 1997. The records were collected in a way to cover both weekdays and weekends by avoiding the same days of the week, and also to cover four different seasons of the year (November, 1996;February, 1997;May, 1996;August, 1997). In total, 113 participants provided all the 12-day food records , 9-day records from 4 participants , 6-day records from 1 participant and 3-day records from 1 participant. The participants were instructed to record all foods and beverages consumed in a standardized booklet. They were asked to provide detailed descriptions of each food (open-ended) including the weights prepared and proportions consumed. The participants recorded their diets by themselves and research dietitians checked their records in a standardized way after the completion by the participants. The daily consumption of 15 nutrient variables was calculated from these records using Standard Tables of Food Composition published by the Science and Technology Agency of Japan 21).

Dish/recipe lists
Research nutritionists classified all the dish and recipes into 197 categories by considering the similarities in terms of nutrient contents and perceptions of subjects. First, we numbered all the foods by using a published book 22) (which classifies all the dishes) as a sample. Cooking oils and soy sauce with specific dishes or recipes were included into those dish/recipe categories. For example, soy sauce could be included in "roast fish" and "sashimi", or cooking-oil was included in "fries with vegetables", "fried eggs" and "tempra" etc. Miso soup was used as a recipe category, which included miso (bean paste) , vegetables and several protein foods such as eggs. We assigned a single category for all types of miso soup.

Calculations of contribution rates of dish/recipe items for the total population intake of nutrients
The percent contribution of each dish or recipe to the total population intake of energy and 14 nutrient variables was computed according to the procedures described by Block et al 23).
Fourteen nutrient variables presented were as follows: protein, fat, carbohydrates, calcium, sodium, potassium, phosphorus, iron, retinol, carotene, vitamin C, vitamin B1, vitamin B2, niacin. The 197 dishes/recipes were ranked according to their percent contributions to the total consumption of a nutrient. There were no marked differences between men and women regarding the percentage contributions of dish/recipe items to the total nutrient intakes. Therefore we presented the results for men and women combined.

RESULTS
The participants were aged 45-77 y in men and 47-76 in women. Their major occupations were farmers, self-employed and house-wives. The percentage of current smokers in men was 49.6% (28 subjects). No women currently smoked. Mean daily intake of both men and women were presented elsewhere 24). Table 1 presents the percentage contributions of dish/recipe items to the total intake of energy and macronutrients (protein, fat, carbohydrates). Rice was the largest source of energy, contributing 29.8% of total intake, followed by miso soup (4.8%). The largest contributor for protein was rice (13.0%), followed by miso soup (8.7%). Rice was the largest contributor (45.3%) for carbohydrates, followed by miso soup (3.4%). Miso soup was a leading contributor (7.1%) for fat, followed by milk (6.7%). In comparison with large contribution of the highest ranked item in energy (rice, 29.8%) and carbohydrates (rice, 45.3%), there was no remarkable single contributor for fat. When rice and miso soup, which were consumed frequently among a rural Japanese (78.2 per 100 person-meals for rice and 53.5 for miso soup), were combined, they were the most important contributors of macronutrients; 34.6% of total for energy, 21.7% for protein, 48.7% for carbohydrates and 11.2% for fat. Table 2a and 2b present the percentage contribution of dish/recipe items to the total intake of minerals (sodium , potassium, calcium, phosphorus and iron). Miso soup was a major contributor for minerals; ranked first for sodium (17 .1%) and potassium (10.6%) and second for calcium (16.6%). Milk was the largest contributor (16.6%) of calcium. Table 3a and 3b present the percentage contribution of dish/recipe items to the total intake of vitamins (carotene , retinol, vitamin C, vitamin B1, vitamin B2, and niacin) . The largest contributor for carotene was spinach (23.6%) , followed by miso soup (10.4%). Pork liver and chicken liver were major contributors, constituting of 23.4 % of total retinol intake although they were seldom consumed (0.26 for pork 0 .19 for chicken per 100 person-meals). In case of vitamin C , green tea (consumed 48.3 per 100 person-meals) was the largest contributor (13.6%), followed by spinach (9.0%) and miso soup (8.6%).

DISCUSSION
We presented dish/recipe based dietary sources to the total nutrient intakes. Several cautions should be required in interpreting the results. For instance, miso soup was used as a category of a recipe in this study, so that it included not only bean paste but also several vegetables and protein-rich foods such as eggs and tofu. There were not cooking oils in the list of contributors for fat intake because they were not dishes and included in other categories of dishes or recipes. Table 4 summarizes the studies conducted in western countries and Japan regarding the dietary sources for the nutrient intakes. We compared our result with those of other studies Table 1. Percentage contribution of dish/recipe items to the total intake of energy and macronutrients (protein, carbohydrates and fat).     conducted in general (adult) population. It gives us the information about the characteristics of habitual diet of their population although dietary sources depend on the method of classification. There are some differences between Japanese and western diets. Rice was the largest contributor for energy, protein and carbohydrates in Japanese populations. In western populations, breads and beef were major contributors for most of macronutrients. Milk was the largest contributor for calcium in all the populations.
There are some differences among the studies in Japanese populations. For example, in the present study, miso soup, as a recipe, was the largest contributor for fat. In comparison, Tokudome 10) reported chicken egg and Tsubono 9) reported vegetable oils as the largest contributor for fat. This is partly because we used miso soup as a recipe category while the other two studies considered bean paste as a single food material. In case of carotene, miso soup was the second largest con-tributor in our study. Carrot was ranked as the largest contributor in other studies for Japanese populations, but not ranked in our study because carrot was classified into several dishes, not as a food. The very high contribution of green tea to vitamin C intake is characteristics of the present study. The subjects of our study consumed green tea 48.3 per 100 person-meals. On the other hand, it was consumed only 16.7 per 100 personmeals in another study 9). One possible reason about the large contribution of green tea may be that the mean age of our subjects were relatively high (men: 62.5 ± 8.4 SD; women: 61.2 ± 8.5 SD) in comparison with other studies (men: 50.2 ± 5.2 SD ; women 46.9 ± 4.4 SD s) and aged 40-49 years 10)) and that green tea consumption is more common in elder than younger populations 25). Because major occupations of our subjects were farmers, self-employed and house-wives, it could lead to high intakes of green tea.
In conclusion, in this study of 12-day diet records collected from 119 men and women in rural northern Japan, we found rice and miso soup as important contributors of energy and macronutrients, and several characteristic dishes as the largest contributors of important nutrients. Our results also suggest that the examination of nutrient sources based on dishes and recipes, rather than on food materials, may be useful in characterizing the dietary patterns of populations. We used these data to develop a food frequency questionnaire (FFQ) for a rural Japanese population. Validation of the questionnaire is currently underway.