Abstract
Objective
To examine the reproducibility of 24-hour dietary recall for estimating dietary vitamin intakes by middle-aged Japanese men and women.
Methods and Results
The subjects were 132 men and 130 women aged 40–69 years, selected from participants in cardiovascular risk surveys conducted in 4 communities. The reproducibility of the 24-hour dietary recall was tested by comparing nutrient and food intake for two recalls conducted on the same season 1 year apart, designated as recalls 1 and 2. Differences in mean values between two recalls were tested using analysis of variance, and Spearman rank correlation coefficients for the two recalls were calculated for nutrient and food intakes. Mean values of energy and vitamins for both sexes were generally similar for the two recalls. The reproducibility of recall by men was high for vitamin B2, folate, pantothenic acid, and vitamin C and by women for vitamin B2, moderate by men for vitamins A, E, K, B1, B6 and niacin, and by women for vitamins A, E, K, B1, B6 and niacin, folate, pantothenic acid and vitamin C, and low by both men and women for vitamins D and B12. The reproducibility during 1985-1999 was generally lower than that of 1973–1984, but that for folate, pantothenic acid and vitamin C remained to be moderate in 1984–1999.
Conclusions
Although the reproducibility of 24-hour dietary recall varies among vitamins, moderate and sustained reproducibility was observed for folate, vitamin C and pantothenic acid.
Key words: Reproducibility, 24-hour dietary recall, vitamins, nutritional survey
Introduction
Dietary assessment is important for examining potential dietary effects on biomarkers and the disease risk. The 24-hour dietary recall has been one of the standard methods to assess nutrient intakes in population-based studies (1). This method is completely open ended, it allows an unlimited level of specificity for detail related to describing food and amounts. This higher level of specificity may not be obtainable using a limited number of food items in a structured questionnaire. A single 24-hour dietary recall per individual can be used to determine the average nutrient intake in defined subgroups of a population (1). Since vitamin content varies greatly for different foods, even those in the same food group, the 24-hour dietary recall may be sensitive for estimating group means of vitamin intakes for prospective study if its reproducibility is confirmed.
The main purpose of the study presented here was to examine the reproducibility of single 24-hour dietary recall for the estimation of vitamin intake by a middle-aged general population. Japanese consume a substantial amount of rice, approximately 29.0 % of total energy, as the principal diet component, together with various ingredients such as soybean products, fish, meat, poultry, milk and dairy products, while; fruit, vegetable, mushroom, seaweed, etc. constitute, the secondary diet components (2). Such a dietary pattern is likely to yield stable intakes of major nutrients, minerals and vitamins. In recent years, however, day-to-day variations in diets have increased because of the growing importation of various foods, the development of the food service industry and transportation, and the spread of convenience stores since around 1985 (3, 4, 5). There is a strong likelihood that the reproducibility of 24-hour dietary recall may have been affected by changes in dietary habits. Therefore, we also examined the reproducibility of 24-hour dietary recall during two survey periods, 1973-1984 and 1985-1999.
Methods
Populations
The subjects were residents of Ikawa, Akita Prefecture (a northeastern rural community), Kyowa, Ibaraki Prefecture (a central rural community), Yao, Osaka Prefecture (a suburban community), and Noichi, Kochi Prefecture (a southwestern rural community) participated in cardiovascular risk surveys conducted between 1973 and 1994 in Ikawa, between 1982 and 1999 in Kyowa, between 1975 and 1990 in Yao, and between 1975 and 1999 in Noichi. We adopted the 24-hour dietary recall method to collect the dietary data, we previously reported on the relationships between dietary intake and risk of cardiovascular disease (6, 7). We selected the subject aged 40-69 years old systematically, approximately 10 % of the participants responded to a 24-hour dietary recall at examination. Most of participants took part in 24-hour dietary recall for every 4 to 5 years, for this study, we selected the subjects who repeated single 24-hour dietary recalls 1 year apart, 3.3 % of the participants in 24-hour dietary recall.
The total number of participants in the present study was 262, 132 men and 130 women.
Assessment of Food and Nutrient Intakes
The 24-hour dietary recall was conducted by trained dietitians. Previously, Yoshino et al. described details of interview, food coding and assessment of nutrient intakes as used in the present study (8). The participants responded to two single 24-hour dietary recalls on the same season 1 year apart. Food items were grouped into 30 categories based on the National Nutrition Survey in Japan (2). Nutrient intakes were estimated based on the Standard Tables of Food Composition in Japan (5th revised edition) (9) and adjusted for total energy intake by using the residual method as well as log transformation (10).
The nutrients examined in this study were related to energy and vitamins, the latter being vitamin A (also retinol and carotene), vitamin D, vitamin E, vitamin K, vitamin B1, vitamin B2, Niacin, vitamin B6, vitamin B12, folate, vitamin C and pantothenic acid. We also examined major diet components for vitamins.
Statistical Analysis
The reproducibility of 24-hour dietary recall data was tested by comparing food and nutrient intakes, stratified by sex, adjusted for total energy intake, for the two recalls conducted 1 year apart, i.e. recalls 1 and 2. Differences in mean values between recalls 1 and 2 were tested using analysis of variance. Spearman rank correlation coefficients for nutrient and food intakes were calculated for the two recalls. Estimates of within (w2) and between-person variance (b2) were calculated by analysis of components of variance, and the ratio of within- to between-person variance (11) was presented. SAS version 9.1.3 software (SAS Institute Inc., Cary, NC, USA) was used for statistical analysis. All statistical tests were two-tailed and p<0.05 was regarded as significant.
The reproducibility of 24-hour dietary recalls was examined by sex and the two survey periods, 1973-1984 and 1985-1999.
Results
Table 1 shows Spearman rank correlation coefficients for recalls 1 and 2 and the mean values of energy and vitamin intakes throughout entire period, 1973-1999. There was no significant difference in mean nutrient intake between the two recalls for either sex, except for vitamin E and B2 for men. The correlation was moderate for vitamin B2 (0.43 for men and 0.44 for women), folate (0.45 for men), pantothenic acid (0.49 for men), vitamin C (0.52 for men), and weak for vitamin D (0.15 for men, 0.13 for women) and vitamin B12 (0.01 for men, 0.22 for women). The correlations for other vitamins ranged from 0.25 to 0.39. These correlations were similar or became somewhat weaker when adjusted for energy.
Table 1.
Spearman rank correlation coefficients, means and standard deviations for energy and vitamins intakes estimated from two recalls conducted one year apart throughout entire period, 1973-1999
| Correlation Crude | coefficients Energy-adjusted | Recall 1 Mean ± SD | Recall 2 Mean ± SD | |
|---|---|---|---|---|
| Men(n=132) | ||||
| Age | - | - | 51.4 ±7.5 | 52.3 ± 7.7 |
| Energy (kcal) | 0.54* | - | 2196 ± 643 | 2255 ± 642 |
| Vitamin A (µgRE) | 0.26* | 0.24 * | 803 ± 611 | 1061 ± 1534 |
| Retinol (µg) | 0.26* | 0.16 | 112 ± 87 | 300 ± 1391 |
| Carotene (µg) | 0.32* | 0.30 * | 4104 ± 3602 | 4492 ± 3433 |
| Vitamin D (µg) | 0.15 | 0.13 | 10.4 ± 11.0 | 11.3 ± 10.5 |
| Vitamin E (mgα-TE) | 0.35 * | 0.36 * | 7.7 ± 3.4 | 9.0 ± 5.3** |
| Vitamin K (µg) | 0.39 * | 0.36 * | 308 ± 248 | 327 ± 264 |
| Vitamin B1(mg) | 0.33 * | 0.26 * | 0.9 ± 0.4 | 1.0 ±0.4 |
| Vitamin B2 (mg) | 0.43 * | 0.36 * | 1.0 ±0.4 | 1.2 ±0.7** |
| Niacin (mg) | 0.29 * | 0.19 * | 16.9 ± 7.9 | 17.7 ±8.2 |
| Vitamin B6 (mg) | 0.31 * | 0.09 | 1.5 ± 0.6 | 1.6 ±0.6 |
| Vitamin B 12 (ug) | 0.01 | -0.04 | 9.6 ± 10.9 | 9.8 ± 9.9 |
| Folate (ug) | 0.45 * | 0.40 * | 370 ± 185 | 404 ± 228 |
| Vitamin C (mg) | 0.52 * | 0.48 * | 128 ±107 | 141 ± 111 |
| Pantothenic acid (mg) | 0.49 * | 0.27 * | 5.8 ± 2.0 | 6.2 ± 2.4 |
| Women (n=130) | ||||
| Age | - | - | 54.3 ± 8.1 | 55.5 ± 8.2 |
| Energy (kcal) | 0.42 * | - | 1696 ± 496 | 1667 ± 508 |
| Vitamin A (ugRE) | 0.28 * | 0.24 * | 944±1104 | 1036 ± 877 |
| Retinol (ug) | 0.27 * | 0.31 * | 189 ±666 | 193 ± 702 |
| Carotene (ug) | 0.25 * | 0.19 * | 4504 ± 5186 | 5039 ± 2974 |
| Vitamin D (ug) | 0.13 | 0.09 | 8.0 ± 7.9 | 8.3 ± 8.3 |
| Vitamin E (mga-TE) | 0.33 * | 0.36 * | 9.0 ± 21.9 | 7.6 ± 3.6 |
| Vitamin K (ug) | 0.25 * | 0.24 * | 272 ± 476 | 270 ± 169 |
| Vitamin B 1(mg) | 0.28 * | 0.18 * | 0.8 ± 0.3 | 0.8 ± 0.3 |
| Vitamin B 2 (mg) | 0.44 * | 0.48 * | 1.0 ±0.7 | 1.0 ±0.4 |
| Niacin (mg) | 0.33 * | 0.23 * | 13.7 ±6.8 | 13.1 ±6.4 |
| Vitamin B 6 (mg) | 0.31 * | 0.31 * | 1.2 ±0.5 | 1.2 ±0.4 |
| Vitamin B 12 (ug) | 0.22 * | 0.17 * | 8.0 ± 10.4 | 6.7 ± 6.9 |
| Folate (ug) | 0.32 * | 0.34 * | 371 ± 458 | 355 ± 154 |
| Vitamin C (mg) | 0.26 * | 0.26 * | 148 ±129 | 150 ± 87 |
| Pantothenic acid (mg) | 0.39 * | 0.34* | 5.2 ± 2.1 | 5.2 ± 1.5 |
p < 0.05 (two-sided p value);
Significantly different : p < 0.05 (for ANOVA).
Table 2 shows Spearman rank correlation coefficients for energy and vitamin intakes, their mean values, and corresponding ratios of within- to between-person variance stratified by the two survey periods, 1973-1984 and 1985-1999. In 1973-1984, there was no significant difference in mean vitamins intake between two recalls except for vitamin A and folate for women. In 1985-1999, Most vitamin intakes were tended to increase compared to the first survey period both men and women, however, total energy intake increase for men, decreased for women. In 1985-1999, there was no significant difference between two recalls except for vitamin E for men.
Table 2.
Spearman rank correlation coefficients, means and standard deviations, and ratios of within- to between-person variance* for energy and vitamins intakes estimated from two recalls conducted one year apart and stratified survey periods 1973-1984 and 1985-1999
| Correlation coefficients | 1973-1984 | 1985-1999 | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| 1973-1984 | 1985-1999 | Crude intakes | Crude intakes | |||||||
| Crude | Energy-adjusted | Crude | Energy-adjusted | Recall 1 Mean ± SD | Recall 2 Mean ± SD | w2/ b2* | Recall 1 Mean ± SD | Recall 2 Mean ± SD | w2/ b2* | |
| Men(1973-1984;n=92, 1985-1999;n=40) | ||||||||||
| Energy (kcal) | 0.53 ** | - | 0.61 ** | - | 2163 ± 662 | 2243 ± 669 | 0.7 | 2271 ± 598 | 2282 ± 583 | 0.6 |
| Vitamin A (µgRE) | 0.31 ** | 0.30 ** | 0.18 | 0.10 | 810 ± 622 | 1095 ±1788 | 4.0 | 787 ± 590 | 983 ± 662 | 7.2 |
| Retinol (µg) | 0.28 ** | 0.13 | 0.18 | 0.11 | 103 ± 75 | 339 ±1659 | - | 134±108 | 211 ± 246 | 23.5 |
| Carotene (µg) | 0.36** | 0.37 ** | 0.23 | 0.17 | 4140 ± 3704 | 4418 ± 3394 | 1.4 | 4021 ± 3400 | 4663 ± 3561 | 2.4 |
| Vitamin D (µg) | 0.17 | 0.13 | 0.12 | 0.15 | 10.4 ±11.3 | 10.7 ± 10.0 | 5.4 | 10.1 ± 10.6 | 12.6 ± 11.5 | 4.9 |
| Vitamin E (mgα-TE) | 0.30 ** | 0.39 ** | 0.42 ** | 0.23 | 7.5 ± 3.5 | 8.5 ± 5.6 | 3.4 | 8.2 ± 3.3 | 10.1 ± 4.4 *** | 1.8 |
| Vitamin K (µg) | 0.48 ** | 0.44 ** | 0.15 | 0.16 | 310 ±262 | 330 ± 269 | 1.4 | 303 ± 214 | 318 ±256 | 4.0 |
| Vitamin B 1(mg) | 0.32 ** | 0.31 ** | 0.33 ** | 0.17 | 0.9 ± 0.3 | 1.0 ±0.4 | 1.9 | 1.0 ± 0.5 | 1.0 ±0.4 | 2.5 |
| Vitamin B 2 (mg) | 0.45 ** | 0.40 ** | 0.27 | 0.02 | 0.9 ± 0.4 | 1.0 ±0.7 | 2.5 | 1.2 ± 0.5 | 1.5 ±0.8 | 3.8 |
| Niacin (mg) | 0.31 ** | 0.25 ** | 0.21 | 0.01 | 16.5 ± 8.0 | 16.7 ± 8.4 | 3.2 | 18.0 ±6.6 | 19.8 ± 7.3 | 6.6 |
| Vitamin B 6 (mg) | 0.28 ** | 0.03 | 0.37 ** | 0.16 | 1.5 ±0.6 | 1.6 ±0.6 | 3.2 | 1.6 ± 0.5 | 1.7 ±0.5 | 1.8 |
| Vitamin B 12 (µg) | -0.09 | -0.12 | 0.22 | 0.15 | 9.1 ± 10.8 | 8.9 ± 9.2 | -*** | 10.7 ± 11.2 | 11.8 ± 11.3 | 4.4 |
| Folate (µg) | 0.39 ** | 0.35 ** | 0.54 ** | 0.50 ** | 352 ± 186 | 382 ± 234 | 1.4 | 413 ± 179 | 455 ± 209 | 1.3 |
| Vitamin C (mg) | 0.50 ** | 0.45 ** | 0.55 ** | 0.56 ** | 123 ± 110 | 132 ± 93 | 1.3 | 138 ±99 | 163 ± 142 | 1.0 |
| Pantothenic acid (mg) | 0.47 ** | 0.26 ** | 0.50 ** | 0.27 | 5.6 ± 1.9 | 6.0 ± 2.5 | 1.0 | 6.4 ± 2.1 | 6.7 ± 2.3 | 1.5 |
| Women(1973-1984; n=84,1985-1999;n=46) | ||||||||||
| Energy (kcal) | 0.44 ** | - | 0.41 ** | - | 1712± 472 | 1716± 553 | 1.1 | 1666 ± 542 | 1579 ± 406 | 3.9 |
| Vitamin A (µgRE) | 0.24 ** | 0.15 | 0.40 ** | 0.40 ** | 766 ± 544 | 1083 ± 1001 *** | 38.4 | 1270 ± 1667 | 951 ± 587 | - |
| Retinol (µg) | 0.36 ** | 0.39 ** | 0.00 | 0.04 | 102 ± 86 | 200 ± 850 | 33.5 | 347±1104 | 178 ± 280 | - |
| Carotene (µg) | 0.22 ** | 0.11 | 0.38 ** | 0.39 ** | 3946 ± 3170 | 5264 ± 2999*** | 7.4 | 5522 ± 7545 | 4628 ± 2913 | 38.8 |
| Vitamin D (µg) | 0.17 | 0.11 | 0.10 | 0.13 | 7.7 ± 8.3 | 9.1 ± 9.1 | 8.7 | 8.7 ± 7.2 | 6.9 ± 6.4 | 8.6 |
| Vitamin E (mga-TE) | 0.38 ** | 0.50 ** | 0.20 | 0.03 | 6.7 ± 3.4 | 7.4 ± 3.4 | 1.4 | 13.2 ± 36.4 | 8.0 ± 3.9 | 49.4 |
| Vitamin K (µg) | 0.26 ** | 0.22 | 0.27 | 0.22 | 217 ± 160 | 264 ± 168 | 2.7 | 371 ± 766 | 281 ± 172 | - |
| Vitamin B 1(mg) | 0.27 ** | 0.16 | 0.30 ** | 0.20 | 0.8 ± 0.3 | 0.8 ± 0.3 | 3.9 | 0.8 ± 0.4 | 0.8 ± 0.3 | 9.4 |
| Vitamin B 2 (mg) | 0.47 ** | 0.48 ** | 0.24 | 0.21 | 0.8 ± 0.3 | 0.9 ± 0.4 | 1.3 | 1.3 ± 1.0 | 1.1 ±0.4 | 18.2 |
| Niacin (mg) | 0.39 ** | 0.30 ** | 0.17 | 0.02 | 12.8 ± 5.9 | 12.9 ± 7.1 | 2.2 | 15.3 ± 8.0 | 13.3 ± 5.2 | 12.8 |
| Vitamin B 6 (mg) | 0.38 ** | 0.31 ** | 0.16 | 0.21 | 1.2 ±0.4 | 1.2 ±0.4 | 1.7 | 1.3 ± 0.6 | 1.2 ±0.3 | 11.2 |
| Vitamin B 12 (µg) | 0.29 ** | 0.23 ** | 0.13 | 0.18 | 7.2 ± 7.1 | 7.2 ± 6.8 | 2.6 | 9.6 ± 14.5 | 5.6 ± 7.1 | - |
| Folate (µg) | 0.29 ** | 0.23 ** | 0.28 | 0.26 | 297 ± 124 | 344 ± 162*** | 3.1 | 505 ± 738 | 375 ± 137 | 35.1 |
| Vitamin C (mg) | 0.16 | 0.14 | 0.40 ** | 0.32 ** | 131 ± 91 | 137 ± 79 | 19.6 | 181 ± 175 | 172 ± 96 | 5.0 |
| Pantothenic acid (mg) | 0.38 ** | 0.22 ** | 0.40 ** | 0.31 ** | 4.9 ± 1.6 | 5.1 ± 1.6 | 1.5 | 5.7 ± 2.9 | 5.3 ± 1.5 | 5.4 |
Ratios of within- to between-person variance;
p < 0.05 (two-sided p value);
Significantly different: : p < 0.05 (for ANOVA).
The spearman correlation coefficient between recalls 1 and 2 indicates reproducibility of intakes vitamins among population.
The correlations for most vitamin intakes declined from 1973-1984 to 1985-1999, but some of vitamins showed different behavior by sex. The correlation for vitamin A declined for men (0.31 to 0.18), increased for women (0.24 to 0.40), for vitamin E, increased for men (0.30 to 0.42), decreased for women (0.38 to 0.20), for folate, increased for men (0.39 to 0.54), sustained for women (0.29 to 0.28), for vitamin C, moderate for men (0.50 to 0.55), increased for women (0.16 to 0.40). For pantothenic acid, remained moderate both sexes.
The ratio of within-to-between person variance explains the detail of vitamin intakes among population. In the case subjects that vitamin intakes varied greatly among recalls 1 and 2 observed and the distribution was skewed, then the ratio had a large unit, if skewness was greater, the ratio could not be calculated. The ratios of within- to between-person variance for each vitamin intake became greater in 1985-1999 than in 1973- 1984, except for vitamins D, E, B6, folate, pantothenic acid and vitamin C in men, and vitamins D and C in women. Vitamins which both of the correlation coefficient and the ratio maintained were vitamins E, B6, folate, pantothenic acid, and vitamin C in men and vitamin C in women.
Table 3 shows Spearman rank correlation coefficients for recalls 1 and 2 and mean intake of 11 out of 30 food groups that are major source for vitamins, in 1973-1984 and 1985-1999. There was no significant difference in mean intakes of food groups between the two recalls for either sex, except for green vegetables intakes for women in 1973-1984. Compared 19851999 with 1973-1984, mean intakes of rice and rice products declined. On the other hand, mean intakes of flour and flour products, meat and meat products and milk and dairy products increased for both sexes. The correlation for rice and rice products remained high; that for fresh fruits increased both sexes (0.34 to 0.64 for men, 0.15 to 0.45 for women); and that for green vegetables declined for men (0.42 to 0.12) and increased for women (0.25 to 0.42). On the other hand, the correlations for flour and flour products, fresh fish, fish products, egg and milk and dairy products declined from 1973-1984 to 1985-1999. The correlations for other food groups were low during both survey periods. The ratios within- to between-person variance for each food group became greater in 1985-1999 than in 1973-1984, except for fresh fruit, meat and meat product in both men and women, soy and soy products in men, and green vegetable, fish product, milk and milk product in women.
Table 3.
Spearman rank correlation coefficients, means and standard deviations, and ratios of within- to between-person variance* for intakes of food groups estimated from two recalls conducted one year apart and stratified by the survey periods 1973-1984 and 1985-1999
| 1973-1984 | 1985-1999 | |||||||
|---|---|---|---|---|---|---|---|---|
| Crude Correlation | Crude intakes (g/day) | Crude intakes (g/day) | ||||||
| coefficients | Recall 1 | Recall 2 | w2/ b2* | Recall 1 | Recall 2 | |||
| 1973-1984 | 1985-1999 | Mean ± SD | Mean ± SD | Mean ± SD | Mean ± SD | |||
| Men(1973-1984;n=92, 1985-1999;n=40) | ||||||||
| Rice and rice products | 0.64** | 0.49** | 299 ±142 | 276 ±126 | 0.5 | 269 ± 128 | 244 ± 118 | 0.7 |
| Flour and flour products | 0.29** | 0.16 | 59 ± 117 | 74 ± 119 | 4.3 | 85 ± 108 | 94 ± 119 | 9.1 |
| Soy and soy products | 0.26** | 0.36** | 70 ±96 | 84 ±95 | 3.2 | 81 ± 89 | 64 ± 85 | 1.1 |
| Green vegetables | 0.42** | 0.12 | 81 ± 88 | 90 ± 103 | 1.2 | 66 ± 79 | 102 ± 96 | 22.7 |
| Other vegetables | 0.20 | 0.10 | 94 ± 72 | 99 ± 107 | 9.6 | 136±110 | 102 ± 87 | 16.4 |
| Fresh fruits | 0.34** | 0.63** | 142 ± 204 | 144±157 | 2.5 | 128 ± 174 | 140 ± 200 | 0.6 |
| Fresh fish | 0.23** | 0.15 | 80 ± 90 | 78 ± 89 | 2.7 | 78 ± 72 | 79 ± 81 | 3.2 |
| Fish products | 0.14 | 0.05 | 32 ±50 | 26 ± 42 | 7.7 | 24 ± 40 | 31 ± 35 | 7.9 |
| Meat and meat products | 0.12 | 0.37** | 32 ±42 | 37 ± 44 | 6.3 | 56 ± 55 | 42 ± 43 | 3.2 |
| 0.17 | 0.13 | 25 ±33 | 24 ± 34 | 9.5 | 32 ± 44 | 32 ± 36 | 5.1 | |
| Milk and dairy products | 0.56** | 0.14 | 44 ± 82 | 56 ± 102 | 0.9 | 70 ± 102 | 91 ± 140 | - |
| Women(1973-1984;n=84, 1985-1999;n=46) | ||||||||
| Rice and rice products | 0.65** | 0.57** | 218 ± 113 | 208 ± 106 | 0.6 | 147 ± 71 | 137 ±56 | 1.0 |
| Flour and flour products | 0.31** | 0.16 | 51 ± 77 | 52 ± 86 | 3.7 | 81 ± 96 | 69 ± 81 | 8.0 |
| Soy and soy products | 0.07 | 0.00 | 50 ± 74 | 45 ± 55 | - | 56 ± 79 | 48 ± 49 | 17.8 |
| Green vegetables | 0.25** | 0.42** | 68 ± 69 | 100 ±70*** | 4.7 | 94 ± 107 | 84 ± 78 | 1.6 |
| Other vegetables | 0.21 | 0.02 | 96 ± 77 | 102 ± 82 | 4.6 | 126 ± 84 | 142 ± 96 | 51.4 |
| Fresh fruits | 0.15 | 0.45** | 211 ± 177 | 200 ± 158 | 7.8 | 157 ± 143 | 172 ± 143 | 1.6 |
| Fresh fish | 0.45** | 0.02 | 58 ± 57 | 53 ± 52 | 2.0 | 61 ± 71 | 47 ± 57 | 30.1 |
| Fish products | 0.26** | 0.00 | 28 ± 42 | 23 ± 33 | 11.3 | 33 ± 51 | 21 ± 32 | 6.9 |
| Meat and meat products | 0.12 | 0.14 | 30 ± 38 | 29 ± 35 | 4.6 | 33 ± 36 | 41 ± 44 | 3.7 |
| 0.12 | -0.06 | 5 ± 36 | 29 ± 35 | 2.9 | 28 ± 36 | 30 ± 30 | 5.8 | |
| Milk and dairy products | 0.40** | 0.26 | 57 ±94 | 62 ± 102 | 1.6 | 93 ± 109 | 103 ± 112 | 1.5 |
Ratios of within- to between-person variance;
p < 0.05 (two-sided p value);
Significantly different : p < 0.05 (for ANOVA).
Discussion
The Spearman rank correlation coefficients were moderate for vitamin B2 in men and women, folate, vitamin C and pantothenic acid in men, and low for vitamins D and B12 in both men and women in the entire periods. After stratification by the two survey periods, the correlation for vitamin B2 declined in both sexes. However, folate, vitamin C and pantothenic acid maintained moderate correlations between 1973-1984 and 1985-1999 in both sexes. In general, the ratios of within- to between-person variance for each vitamin intake became greater when Spearman correlation coefficients were declined from 1973-1984 to 1985-1999.
Declined reproducibility for vitamin B2 in both sexes corresponded to the decline of reproducibility for dietary intakes of milk and dairy products and egg, major sources (approximately 25 %) of vitamin B2 (2, 12).
The reproducibility for folate remained moderate in both men and women corresponding to the sustained reproducibility for green vegetables in women, and fresh fruits in both sexes, which were the main sources (approximately 40%) of folate (2, 8). The reproducibility for vitamin C also remained moderate in both men and women corresponding to the sustained reproducibility for fresh fruits in both sexes and green vegetables in women, which were the main sources (approximately 50 %) of that vitamin (2, 13).
The moderate and sustained reproducibility for pantothenic acid in both men and women was probably due to the distribution of this nutrient over a wide variety of food groups (9).
The low reproducibility for vitamins D and B12 in both men and women corresponded to the low reproducibility for fresh fish and fish products in both sexes, which were the main sources (approximately 80 %) for these vitamins (8, 13).
In conclusion, we found that the reproducibility of 24-hour dietary recall for estimating vitamin intakes varies depending on the vitamin concerned. Our result showed the reproducibility for most vitamins in recent years declined, but moderate and sustained reproducibility for folate, vitamin C and pantothenic acid were proved for both sexes.
Financial disclosure
None of the authors had any financial interest or support for this paper.
References
- 1.Willet W.C. Nutritional Epidemiology. 2nd ed. Oxford University Press; New York, NY: 1998. [Google Scholar]
- 2.Ministry of Health, Labor and Welfare. Nutrition Status Based on the National Nutrition Survey in Japan. 2001, Daiichi Shuppan. (in Japanese).
- 3.White paper on economics. 1987, Economic Planning Agency of Japan. (in Japanese).
- 4.White paper on the National Lifestyle. 1987: Economic Planning Agency of Japan. (in Japanese).
- 5.White paper on the National Lifestyle. 2005, Cabinet Office, Government of Japan. (in Japanese).
- 6.Iso H., Terao A., Kitamura A., Sato S., Naito Y., Kiyama M., Tanigaki M., Iida M., Konishi M., Shimamoto T., komachi Y. Calcium intake and blood pressure in seven Japanese populations. Am J Epidemiol. 1991;133:776–783. doi: 10.1093/oxfordjournals.aje.a115956. PubMed PMID: 2021144. [DOI] [PubMed] [Google Scholar]
- 7.Iso H., Sato S., Kitamura A., Naito Y., Shimamoto T., Komachi Y. Fat and protein intakes and risk of intraparenchymal hemorrhage among middle-aged Japanese. Am J Epidemiol. 2003;157:32–39. doi: 10.1093/aje/kwf166. 10.1093/aje/kwf166 PubMed PMID: 12505888. [DOI] [PubMed] [Google Scholar]
- 8.Yoshino K., Inagawa M., Oshima M., Yokota K., Umesawa M., Embo M., Yamagishi K., Tanigawa K., Sato S., Shimamoto T., Iso H. Trends in dietary intake of folate, vitamins B6, and B12 among Japanese adults in two rural communities from 1974 through 2001. J Epidemiol. 2005;15:29–37. doi: 10.2188/jea.15.29. 10.2188/jea.15.29 PubMed PMID: 15762091. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Japan Science and Technology Agency. Standard Tables of Food Composition in Japan. The fifth revised edition: National Printing Breau, 2000. (in Japanese).
- 10.Willet W.C., Stampfer M.J. Total energy intake: implications for epidemiologic analysis. Am J Epidemiol. 1986;124:17–27. doi: 10.1093/oxfordjournals.aje.a114366. [DOI] [PubMed] [Google Scholar]
- 11.Sempos C.T., Johnson N.E., Smith E.L., Gilligan C. Effects of intraindividual and interindividual variation in repeated dietary records. Am J Epidemiol. 1985;121:120–130. doi: 10.1093/oxfordjournals.aje.a113974. PubMed PMID: 3964987. [DOI] [PubMed] [Google Scholar]
- 12.Ogawa K., Tsubono Y., Nishino Y., Watanabe Y., Ohkubo T., Watanabe T., Nakatsuka H., Takahashi N., Kawamura M., Tsuji I., Hisamichi S. Dietary sources of nutrient consumption in a rural Japanese population. J Epidemiol. 2002;12:1–8. doi: 10.2188/jea.12.1. PubMed PMID: 11848179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Imaeda N., Tokudome Y., Ikeda M., Kitagawa I., Fujiwara N., Tokudome S. Foods contributing to absolute intake and variance in intake of selected vitamins, minerals and dietary fiber in middle-aged Japanese. J Nutr Sci Vitaminol. 1999;45:519–532. doi: 10.3177/jnsv.45.519. PubMed PMID: 10683805. [DOI] [PubMed] [Google Scholar]