Trends in Dietary Intake of Folate, Vitamins B6, and B12 among Japanese Adults in Two Rural Communities from 1974 through 2001

BACKGROUND: The 5th edition of the Japanese food composition table enables us to evaluate intakes of folate and vitamins B6 and B12, which are associated with risk of chronic diseases. METHODS: We investigated long-term trends in dietary intake of those nutrients in two rural communities; Ikawa from 1974 through 2000, and Kyowa from 1982 through 2001. The 24-hour recall method was adopted. Intake of green tea interviewed from 1994 was used to examine food sources for these nutrients in the latest period, but not to evaluate long-term trends. Reduced intakes of nutrients due to cooking were not taken into account. RESULTS: Age-adjusted mean folate intake increased by 30% in Ikawa between the 1970’s and 1980’s, and then leveled off to the latest survey, while that in Kyowa did not change throughout the survey periods. The increased folate intake was primarily due to green/yellow vegetables. Mean vitamin B6 intake did not change except that it increased for Ikawa women in the 1980’s and decreased for Kyowa men in the latest period. No secular trend was found for mean vitamin B12 intake. The largest source for folate intake was total vegetables (38-58% of total intake) and the second largest source was alcohol/beverages including green tea (11-24%). Fish/shellfish was the largest source for vitamins B6 (16-23%) and B12 (77-84%). CONCLUSIONS: Dietary intakes of folate, vitamins B6 and B12 showed no notable long-term trend, except for an increased folate intake between the 1970’s and 1980’s due to an increased intake of green/yellow vegetables.

for the dieticians to follow the manual. In the interviews, actualsized food models, pictures of food materials and dishes, 30 and/or real foods and dishes were shown so that the subjects was easily able to recall what they had eaten. The same basic food models and the interview forms were used throughout the surveys. The intake of green tea was interviewed from 1994. As for rice and miso-soup, we asked the subjects to put the usual amount into a bowl and then we measured that quantity. We also investigated the frequencies of 18 major foods and food groups per week in order to confirm that foods in the 24-hour dietary recall were not so different from the usual foods taken. Milk and alcohol were asked as beverage in the food frequency questionnaire. Persons who had had special events such as a festival or a celebration were excluded from the surveys. The interview took approximately 30 minutes per subject.
Intakes of nutrients were estimated based on the Standard Tables of Food Composition in Japan (5th revised edition). 27 In the surveys before the 5th edition was issued in 2000, the data were coded based on the 4th edition of Food Composition Table. 31 We translated the data based on the 4th edition into those based on the 5th edition, using our original translation table. As for foods newly-appeared in the 5th edition, we checked the data in the latest surveys and confirmed that the newly-appeared foods were rarely taken among our subjects. It is possible that the amounts of folate and vitamins B6 and B12 contained in the same foods may change from the 1970's to the present, but we have no data to support this possibility. Therefore, we used the data in the 5th edition throughout the surveys.
The 5th edition provides the amounts of nutrients after cooking only for selected foods. 32,33 Thus, we evaluated all data as in conditions before cooking in order to investigate long-term trends, although there may be systematic overestimation of nutrient intakes.
We categorized 17 food groups (green/yellow and other vegetables, rice/cereal, fruits, fish/shellfish, meats, eggs, seaweed, potatoes, beans, milk/dairy products, alcohol/beverages, and others) based on the National Nutrition Survey in Japan. 34 For primary trend analyses, we did not include dietary intake of green tea because this intake had not been interviewed in the surveys before 1994. However, for secondary analyses we estimated folate and vitamins B6 and B12 from green tea in the latest survey period, in order to examine the proportion of these nutrient intakes.
Sex-and age (10 years)-specific mean values and standard deviations of folate and vitamins B6 and B12 were calculated for each survey period as described above. Sex-specific and ageadjusted mean values and standard errors were calculated by analysis of covariance. We also evaluated intakes of major food groups that contributed to dietary intakes of these nutrients. Differences in mean values from the earliest survey period were determined using the Student's t-test or analysis of covariance. SAS ® version 8.02 software (SAS Institute Inc., Cary, USA) was used for statistical analysis. P values less than 0.05 were regarded Trends in Folate, B6, and B12 among Japanese pregnant women, 1.6 mg/day for vitamin B6 in men and 1.2 mg/day in women, and 2.4 µg/day for vitamin B12. 26 In 2000, the Ministry of Science and Education revised the  Japanese standard food composition table for the first time in 18 years, 27 in order to reflect the diversification of Japanese dietary habits. Folate and vitamins B6 and B12 were included in the table, so that trends for these intakes could be evaluated, using the existing database of dietary surveys. Examination of these nutrient trends would be valuable for the formulation of public health recommendations.
The present study investigated the long-term trends in the dietary intake of folate and vitamins B6 and B12, according to population-based surveys in two Japanese communities between 1974 and 2001.
The subjects in this study are men and women aged 40 to 69 years living in the town of Ikawa, Akita prefecture, and the town of Kyowa, Ibaraki prefecture. Ikawa is located near the Sea of Japan in the north, where nearly a half of the town is forest and rice crop agriculture is the main industry. The town of Kyowa is on a plain and is located in the center of Japan, which has mainly horticultural and rice crop agriculture with other light industries.
The nutrition surveys were carried out in approximately 10% systematic samples of the participants aged 40 to 69 in the annual cardiovascular risk surveys. The participants aged over 70 were excluded from the nutrition survey because the accuracy of data based on the 24-hour dietary recall may decline with aging. For each participant in cardiovascular risk surveys, the recruitment for the nutrition surveys was made every 4 to 5 years. The subjects were not pre-informed of the recruitment for the nutrition study.
The surveys were from 1974 through 2000 for Ikawa and from 1982 through 2001 for Kyowa. All were conducted in autumns in Kyowa. In Ikawa, most of them were in springs but some data were taken in autumns from 1975 through 1981. To examine the season variations, 28 we compared the data taken in springs and autumns and confirmed that the seasonal differences were not large enough to affect the long-term trends. The study during was stratified into seven time periods for Ikawa: 1974Ikawa: -1977Ikawa: , 1978Ikawa: -1981Ikawa: , 1982Ikawa: -1985Ikawa: , 1986Ikawa: -1989Ikawa: , 1990Ikawa: -1993Ikawa: , 1994Ikawa: -1997Ikawa: , and 1998Ikawa: -2000, and into four periods for Kyowa: 1982Kyowa: -1986Kyowa: , 1990Kyowa: -1993Kyowa: , 1994Kyowa: -1997Kyowa: , and 1998Kyowa: -2001. When there were persons who undertook the nutrition survey more than once during one survey period, we used the data in the earliest year in each survey period. We adopted the 24-hour dietary recall method to collect the dietary data. 29 The subjects were interviewed on what they had eaten during 24 hours before the examination. Throughout the surveys, trained dieticians carried out the interviews based on our dietary-recall manual. We continued to hire the survey dieticians for a long term in order to avoid fluctuations in interviewing technique overtime. The training sessions were conducted when we employed new dieticians. Before each survey, a meeting was held When we took green tea into account in the latest survey, the proportion of folate intake from alcohol/beverages was 14% for Ikawa men, 11% for Ikawa women, 24% for Kyowa men, and 21% for Kyowa women, while those excluding green tea were 3%, 0.3%, 4%, and 0.3%, respectively. The proportions of folate intake by food group in the latest survey period are shown in Figure 1. Mean folate intake from alcohol/beverages including green tea in the latest survey period was 60.3 µg/day for men and 50.0 µg/day for women in Ikawa, and 101.6 µg/day for men and 91.5 µg/day for women in Kyowa.
In Ikawa, mean folate intake for both sexes from green/yellow vegetables doubled from 1974-1977 through 1982-1985, and then leveled off. Spinach mainly contributed to this increase, which comprised 55-82 % of total green/yellow vegetables in men and 50-75 % in women. Folate intake from spinach was 68 µg/day for men and 57 µg/day for women in 1974-1977, which increased three-fold to 186 and 146 µg/day for men and women, respectively, in 1982-1985. In Kyowa, folate intake from green/yellow vegetables decreased between 1982-1986 and 1990-1993. This was mainly due to the decreased intake from spinach from 71.7 to 29.0 µg/day for men and from 79.5 to 37.5 µg/day for women. The folate intake from broccoli slightly increased (2.9 µg/day in men and 7.8 µg/day in women). In 1998-2001, there was a higher folate intake from spinach in Ikawa (142 µg/day for men and 136 µg/day for women) than in Kyowa (47 µg/day for men and 38 µg/day for women). Mean folate intake from eggs, milk/dairy products, and beans increased in a similar manner, whereas that from rice/cereal decreased for both sexes. Mean folate intake from other vegetables declined for men but increased for women. In Kyowa, the mean folate intake from beans and milk/dairy products increased, whereas that from rice/cereal decreased for both sexes.
The proportions of vitamin B6 intake from major food sources were 16-23% from fish/shellfish, 12-17% from rice/cereal, 7-15% from green/yellow vegetables, 7-13% from other vegetables, 7-10% from meats, 5-8% from potatoes, 4-6% from beans, and 1-2% from eggs and milk/dairy products, for both sexes in each community ( Figure 1). There was a large sex-difference in the percent contribution from alcohol/beverages and fruits in both communities. The proportion of vitamin B6 intake from alcohol/beverages was 9-15% for men and 0.1-0.9% for women, and that from fruits was 3-6% for men and 8-13% from women. Ninety percent of the total vitamin B6 intake was from the above 11 food groups. In the vitamin B6 intake from alcohol/beverage, beer and sake were predominant sources. Ninety-nine percent of that intake from alcohol/beverage was from beer and sake (0.15-0.20mg for men in the latest survey period). Mean vitamin B6 intake from rice/cereal gradually decreased for both sexes in each community, whereas intake from vegetables, meats, beans, and milk/daily products increased except for that in Kyowa men.
The number of the survey participants was between 251 and 616 for men and between 158 and 339 for women in Ikawa, and between 381 and 690 for men and between 453 and 616 for women in Kyowa. Age-adjusted mean value (standard error) of body mass index in the latest survey period was 23.9 (0.2) for men and 24.4 (0.2) for women in Ikawa, and 23.7 (0.2) and 23.5 (0.2), respectively, in Kyowa. The percentage of subjects usually taking alcohol was 88% for men and 14% for women in Ikawa, and 76% for men and 11% for women in Kyowa.
In Ikawa, age-adjusted mean intake of folate in both sexes increased approximately 30% between 1974-1977 and 1982-1985, and then leveled off (Table 1). This trend was similar for each age group. For both men and women, age-adjusted mean folate intake in 1998-2000 was 393 µg/day, which was estimated in raw food conditions and by excluding green tea. There was no secular trend in folate intake for either men or women in Kyowa. Age-adjusted mean folate intake in 1998-2000 was 332-338 µg/day for both sexes. When age was taken into account, however, mean folate intake for both sexes declined at ages 40-59, but tended to increase at ages 60-69. Thus, there was little difference overall in the mean folate intake among age groups.
The age-adjusted mean intake of vitamin B6 increased approximately 10% from 1.13 mg/day in 1978-1981 to 1.27 mg/day in 1998-2000 for women in Ikawa. This trend was observed at ages 50-69, but not at ages 40-49. There was no secular trend in the mean vitamin B6 intake for men in Ikawa, and the age-adjusted mean value in 1998-2000 was 1.60 mg/day. Age-adjusted mean intake of vitamin B6 decreased 10% for men in Kyowa in the latest survey period. Age-adjusted vitamin B6 intake in 1998-2001 was 1.41 mg/day for men and 1.20 mg/day for women in Kyowa.
intake was approximately 20% higher in Ikawa than in Kyowa for both sexes during 1998-2001. This community difference was due mostly to a higher intake of fish/shellfish. The higher intake of seafood in Ikawa is explained by its geographical location near the Sea of Japan. Although the mean vitamin B6 intake did not differ between the communities, the mean intake was approximately 20% higher for men than for women in each community. This sex difference was largely due to higher intakes of seafood and beer and sake.
Although there was no significant secular trend in mean folate intake from meats, there were large variations for men in both communities. This large variation was due to liver, which contains a large amount of folate (810-1300 µg per 100 g or 255µg per portion size of 55g) compared with green/yellow vegetables (approximately 100-200 µg per 100 g). When we excluded liver in calculating mean folate intake from meats, the age-adjusted mean value was smaller, ranging between 2.7 and 4.2 µg/day for men in Ikawa and between 2.0 and 3.2 µg/day for men in Kyowa. Although liver caused the large variations in folate intake from meats, its influence on the total intake was small because liver was rarely eaten.
The declines in mean vitamin B6 intake for Kyowa men and in mean folate intake for Kyowa men and women aged 40-59 years could potentially cause nutritional problems. The decline in vitamin B6 intake was due to decreased intakes of both seafood and rice/cereals. The decline in folate intake among young age groups was due to decreased intake of vegetables and rice/cereals.
In our study, nutrient intakes were evaluated in raw food conditions because there was no systematic database about nutrient loss by cooking. Estimated changes of nutrients after cooking are 79 % for folate, 75 % for vitamin B6, and 97 % for vitamin B12. 33 Thus, dietary intakes of folate and vitamin B6 may be overestimeated systematically in the present study.
In summary, we investigated long-term trends in dietary intakes of folate and vitamins B6, and B12 among Japanese adults in two rural communities, and found that there was an increase in the dietary intake of folate for men and women between the 1970's and the 1980's, along with the increased intake of green/yellow vegetables. Folate intake was determined mainly by vegetables, especially spinach, while intakes of vitamins B6 and B12 were by fish/shellfish.