Abstract
Background
The disparity of overall diet quality by personal educational attainment has been a public issue. However, it remains unknown which food groups contribute to the disparity. This cross-sectional study assesses which food groups explain associations between education and overall diet quality in Japanese women.
Methods
A total of 3,788 middle-aged (mean age, 47.7 years) and 2,188 older women (mean age, 74.4 years), who lived in 47 prefectures in Japan, provided data on their education (low, middle, and high) and dietary intakes from a diet history questionnaire. A diet quality score (possible score 0–70) was calculated based on seven food components. Mean diet quality scores, with adjustment for lifestyle and neighborhood variables, were estimated by education using a general linear model, and Dunnett’s multiple comparison was conducted. Additionally, mean scores of each food component were estimated by education and compared using the same manner.
Results
After adjustment for lifestyle and neighborhood variables, mean diet quality score of high or middle education was higher than low education for both generations. Middle-aged women with high and middle education had higher scores of ‘milk’, ‘snacks, confection, and beverages’, ‘fruits’, and ‘vegetable dishes’ than those with low education. Older women with high and middle education had higher scores of ‘sodium from seasonings’ and ‘fruits’ than those with low education.
Conclusions
This study suggests that positive associations between education and diet quality are explained by different food groups in middle-aged and older Japanese women, which are independent of lifestyle and neighborhood variables.
Key words: education, diet quality, Japanese
INTRODUCTION
The disparity of dietary intake by personal socioeconomic status (SES) has been a public issue.1–4 Especially, people with low education have consistently shown low level of overall diet quality, which may be partly due to lack of nutrition knowledge, cooking skills, or ability to use prevention messages.1–4 Considering that women influence their family’s diet, and their food preparation behaviors differ more greatly than men according to their educational level,5–7 it is important to understand mechanisms for associations between education and diet quality in women from a nutritional perspective.
Nevertheless, no study has investigated which food groups explain the associations between education and diet quality, although fruits and vegetables showed the most consistent association with education.1–3,8,9 Moreover, generation-specific associations in middle-aged and older women remain unknown. In previous studies, there is generational difference in time spent cooking6 and in food preference.10,11 An Australian study reported that education was more strongly associated with diet quality in adults than in the older people.4
In addition to individual lifestyle, such as smoking and physical activity, neighborhood contexts are potential mediators of associations between education and diet quality.3,4,12–18 Previous studies suggested that people with higher education tend to live in high SES areas12,13 with good access to supermarkets,14 or urban.3,15,16 Moreover, high SES areas mostly related to good diet quality.3,4,12,15–18 Therefore, adjustment for mediating effects by lifestyle and neighborhood variables are needed to investigate direct associations between education and diet quality.
The objective of this cross-sectional study is to assess which food groups explain associations between education and overall diet quality in middle-aged and older Japanese women, with adjustment for lifestyle and neighborhood variables.
METHODS
Survey design
The Three-Generation Study of Women on Diets and Health was conducted on dietetic students, their mothers, and grandmothers in northern and western Japan in 2011 and in eastern Japan in 2012.19,20 The present analysis used data from the mother’s and grandmother’s generation. Briefly, the survey was conducted in a total of 85 academic institutions with a nutrition department. The 7,016 students were requested to invite their mothers and grandmothers to join this study, and to distribute two questionnaires on dietary habits and lifestyle variables to those who had agreed to participate. In total, 4,044 mothers (response rate: 57.6%) and 2,332 grandmothers (response rate: 33.2%) answered both questionnaires. Mothers and grandmothers were considered middle-aged and older, respectively.
To analyze middle-aged women, participants aged 34–60 years were selected (n = 4,011). We excluded those living in eastern Japan who participated in the 2011 survey because their dietary habits and lifestyle would have been influenced by the Great East Japan Earthquake (n = 63). We also excluded those from the institution with the low response rate (n = 2), and those with missing information on the variables of interest (n = 158). To analyze older women, participants aged 61–94 years were selected (n = 2,320). We excluded those living in eastern Japan who participated in the 2011 survey (n = 47), and those from the institution with the low response rate (n = 1). We also excluded those with missing information on the variables of interest (n = 84). Consequently, the final sample sizes were 3,788 and 2,188 for middle-aged and older women, respectively.
This study was conducted according to the Declaration of Helsinki and all study procedure were approved by the Ethics Committee of the University of Tokyo Faculty of Medicine (no 3249). Written informed consent was obtained from each participant.
Calculation of the diet quality score
Dietary habits during the preceding month were assessed using a validated comprehensive diet history questionnaire (DHQ)21–23 for middle-aged women and a validated brief-type diet history questionnaire (BDHQ)21,22 for older women. Briefly, the DHQ and the BDHQ cover the consumption frequency (and portion size in the DHQ) of selected foods commonly consumed in Japan, as well as general dietary behavior and usual cooking methods.21–23 Estimates of the intake of food (151 items in the DHQ and 58 items in the BDHQ) and energy were calculated using an ad hoc computer algorithm, which was based on the Standard Tables of Food Composition in Japan.24 A relative validity of the DHQ and BDHQ has been previously investigated among ninety-two women aged 31–69 years using a 16-d dietary record as reference, in terms of energy-adjusted estimates of food groups and nutrients.21,22 Using the information from the DHQ and the BDHQ, overall diet quality was estimated using a previously developed food-based diet quality score (eTable 1).19,25 This score is based on six components recommended in the Japanese Food Guide Spinning Top and sodium from seasonings (seven components in total). When intake was within the recommended range, a score of 10 was assigned to that component. Energy-adjusted values of dietary intake were calculated using the density method (amount per 7,531 kJ) to compare with the recommended values. For a participant who fell short of or exceeded the recommended value, the score was calculated proportionately between 0 and 10. The seven scores were then summed to provide the diet quality score, which ranged from 0 to 70. In the present population, a higher diet quality score was associated with favorable nutrient intake patterns, such as higher intakes of dietary fiber and micronutrients and lower intakes of saturated fat and sodium.19
Education and lifestyle variables
All the variables were based on the participants’ self-reported information, except for the education of middle-aged women, which was obtained from their daughter’s or son’s questionnaires. Education was categorized as low, middle, and high. For middle-aged women, they corresponded to ≤12 years (ie, junior high school or high school), 13–15 years (ie, junior college or professional college), and ≥16 years (ie, university). For older women, they corresponded to ≤9 years (ie, junior high school), 10–12 years (ie, high school), and ≥13 years (ie, junior college, professional college, or university). Age at the time of the survey was calculated based on the birth date. Body mass index (BMI) was calculated as body weight (kg) divided by the square of body height (m). Living status (living alone or living with others), current marital status (married or widowed/divorced/never married), smoking status (current smoker, former smoker, or non-smoker), and prescription medicine use (yes or no) were also obtained. Physical activity was defined as the average total metabolic equivalent-hours score per day based on the frequency and duration of seven activities.26 Energy-adjusted diet cost (Japanese yen/4,184 KJ) was calculated in accordance with previous studies,20 by multiplying the consumption of each food item by food price using the 2012 Retail Price survey27 (141 items for DHQ, 57 items for BDHQ), supermarket websites (9 items for DHQ, 1 item for BDHQ), and fast-food restaurants (1 item for DHQ), and then summing these products. The employment status (housewife [not employed], part-time worker, or full-time worker) was obtained only for middle-aged women.
Neighborhood variables
Neighborhood contexts were assessed using urban-rural classification, percentage of workers in the primary sector industry, areal deprivation index (ADI), number of food retailers, and region. In this study, neighborhood was defined as a municipality where a participant lived during the past month. A total of 1,147 municipalities from all 47 prefectures were covered (ie, 60.3% of the total municipalities in Japan). Urban-rural classification was defined in the 2010 population census,28 as central cities (Tokyo 23 districts, ordinance-designated cities, and other major cities), the suburbs (municipalities that not only surrounded a central city but also have a 1.5% or higher population commuting to the city), and rural areas (the other municipalities). The percentage of workers in the primary sector industry (agriculture, fishing, and forestry) was obtained from the 2010 population census.28 ADI, as a measure of neighborhood SES reported by Nakaya et al,29,30 consisted of weighted sums of deprivation-related census variables (proportion of old couple households, old single households, single-mother households, rent houses, sales and service workers, agricultural workers, blue collar workers, and unemployment rate) using the 2010 population census.28 The number of food retailers in a municipality was obtained from the economic census for business activity in 2012.31 Municipalities were categorized according to the region.
Statistical analysis
All statistical analyses were performed for middle-aged and older women separately, using SAS statistical software package version 9.4 (SAS Institute, Cary, NC, USA). All reported P values are two-tailed, and P < 0.05 is considered statistically significant. First, basic characteristics of participants were described. Second, to investigate associations between education and the diet quality score, crude mean (model 1) and adjusted mean (model 2 and model 3) and their 95% confidence interval of diet quality score by education were estimated using a general linear model. Additionally, Dunnett’s multiple comparison was conducted with the use of low education as a reference category. In model 2, age and individual lifestyle variables were adjusted to investigate the associations between education and the diet quality score. Individual lifestyle variables were BMI, living status (only for older women because 99.2% of middle-aged women lived with others), current marital status, employment status (only for middle-aged women), smoking status, physical activity, prescription medicine use, and diet cost. In model 3, neighborhood variables (urban-rural classification, percentage of workers in the primary sector industry, ADI, number of food retailers, and region) were also adjusted for. In this study, the multilevel model was not used because of the small number of participants in a municipality (average: 3.6 for middle-aged women and 2.6 for older women) and because of the small municipality level variance in overall variance of the diet quality score (intraclass correlation coefficients for the null model: 0.008 for middle-aged women and 0.004 for older women). Third, the associations between education and the seven components of the diet quality score were estimated using the same manner.
RESULTS
Characteristics of participants
The characteristics of participants are shown in Table 1. The mean age of middle-aged and older women was 47.7 years and 74.4 years, respectively. In comparison to older women, middle-aged women had a lower mean BMI and a higher proportion of living with others, married, high education, current smoker, and non-user of prescription medicine. Additionally, middle-aged women had a higher score of total metabolic equivalent-hours, and spent less on diet than older women. The mean values of diet quality score and scores of ‘vegetable dishes’, ‘fish and meat dishes’, ‘fruits’, ‘snacks, confection, and beverages’, and ‘sodium from seasonings’ were higher in older women than middle-aged women. The intake of energy and total fat were on average higher in middle-aged women, while the intake of protein and carbohydrate were higher in older women. A higher proportion of middle-aged women lived in central cities than older women. Municipalities where middle-aged women lived had a lower percentage of workers in the primary sector industry and lower ADI, while no difference in the number of food retailers was found compared to older women.
Table 1. Characteristics of middle-aged and older Japanese womena.
| Middle-aged (n = 3,788) | Older (n = 2,188) | Pe | |||
| Individual lifestyle variables | |||||
| Age, years | 47.7 | 3.9 | 74.4 | 5.2 | <0.0001 |
| Body mass index, kg/m2 | 22.0 | 3.1 | 22.8 | 3.2 | <0.0001 |
| Living status | |||||
| Alone | 31 | (0.8) | 341 | (15.6) | <0.0001 |
| Live with others | 3,757 | (99.2) | 1,847 | (84.4) | |
| Marital status | |||||
| Married | 3,468 | (91.6) | 1,332 | (60.9) | <0.0001 |
| Widowed/divorced/never married | 320 | (8.4) | 856 | (39.1) | |
| Educationb | |||||
| Low | 1,818 | (48.0) | 981 | (44.8) | <0.0001 |
| Middle | 1,420 | (37.5) | 994 | (45.4) | |
| High | 550 | (14.5) | 213 | (9.7) | |
| Employment status | |||||
| Housewife | 742 | (19.6) | — | — | — |
| Part-time worker | 1,680 | (44.4) | — | — | |
| Full-time worker | 1,366 | (36.1) | — | — | |
| Smoking status | |||||
| Current smoker | 287 | (7.6) | 59 | (2.7) | <0.0001 |
| Former smoker | 315 | (8.3) | 86 | (3.9) | |
| Non-smoker | 3,186 | (84.1) | 2,043 | (93.4) | |
| Physical activity, total metabolic equivalent-hours score per day | 40.7 | 5.6 | 39.0 | 6.6 | <0.0001 |
| Prescription medicine use | |||||
| Yes | 958 | (25.3) | 1,684 | (77.0) | <0.0001 |
| No | 2,830 | (74.7) | 504 | (23.0) | |
| Diet cost, Japanese yen/4,184 kJ | 568 | 108 | 643 | 124 | <0.0001 |
| Diet quality scorec | 42.9 | 8.1 | 50.5 | 8.0 | <0.0001 |
| Components of diet quality scored | |||||
| Grain dishes | |||||
| score | 8.2 | 1.8 | 8.2 | 1.9 | 0.18 |
| g/4,184 kJ | 208.7 | 61.0 | 216.6 | 66.9 | <0.0001 |
| Vegetable dishes | |||||
| score | 7.0 | 2.4 | 9.0 | 1.7 | <0.0001 |
| g/4,184 kJ | 151.4 | 75.6 | 235.6 | 99.8 | <0.0001 |
| Fish and meat dishes | |||||
| score | 9.8 | 0.8 | 9.9 | 0.5 | <0.0001 |
| g/4,184 kJ | 113.8 | 40.1 | 153.3 | 50.4 | <0.0001 |
| Milk | |||||
| score | 6.1 | 3.6 | 6.0 | 3.7 | 0.26 |
| g/4,184 kJ | 69.3 | 64.3 | 69.0 | 55.9 | 0.85 |
| Fruits | |||||
| score | 3.2 | 2.7 | 5.3 | 3.1 | <0.0001 |
| g/4,184 kJ | 36.9 | 35.4 | 62.8 | 43.4 | <0.0001 |
| Snacks, confection, and beverages | |||||
| score | 4.9 | 4.0 | 6.0 | 4.1 | <0.0001 |
| kJ/4,184 kJ | 745.9 | 353.5 | 641.3 | 380.3 | <0.0001 |
| Sodium from seasonings | |||||
| score | 3.7 | 3.6 | 6.1 | 2.9 | <0.0001 |
| mg/4,184 kJ | 1,218.4 | 520.4 | 1,077.7 | 290.5 | <0.0001 |
| Energy, kJ/day | 7,711.4 | 2,224.9 | 7,369.8 | 2,271.5 | <0.0001 |
| Nutrient intake | |||||
| Protein, % of energy | 13.7 | 2.0 | 16.9 | 3.2 | <0.0001 |
| Total fat, % of energy | 29.1 | 5.9 | 25.7 | 5.1 | <0.0001 |
| Carbohydrate, % of energy | 54.3 | 7.1 | 55.9 | 7.4 | <0.0001 |
| Neighborhood variables | |||||
| Urban-rural classification | |||||
| Central cities | 854 | (22.5) | 418 | (19.1) | <0.0001 |
| Suburbs | 1,451 | (38.3) | 787 | (36.0) | |
| Rural areas | 1,483 | (39.1) | 983 | (44.9) | |
| Percentage of workers in primary sector industry | 4.5 | 5.8 | 5.1 | 6.2 | 0.0003 |
| Areal deprivation index score | 6.76 | 0.6 | 6.82 | 0.6 | 0.0004 |
| Number of food retailers | 325 | 281 | 335 | 294 | 0.18 |
| Region | |||||
| Hokkaido and Tohoku | 381 | (10.1) | 206 | (9.4) | 0.01 |
| Kanto | 1,074 | (28.4) | 544 | (24.9) | |
| Hokuriku and Tokai | 839 | (22.1) | 533 | (24.4) | |
| Kinki | 491 | (13.0) | 269 | (12.3) | |
| Chugoku and Shikoku | 514 | (13.6) | 353 | (16.1) | |
| Kyushu | 489 | (12.9) | 283 | (12.9) | |
aData are shown as mean and standard deviation for continuous variables and number (percentage) of participants for categorical variables.
bFor middle-aged women, low education is ≤12 years (junior high school or high school). Middle education is 13–15 years (junior college or professional college). High education is ≥16 years (university). For older women, low education is ≤9 years (junior high school). Middle education is 10–12 years (high school). High education is ≥13 years (junior college, professional college or university).
cPossible score ranging from 0 to 70.
dPossible score ranging from 0 to 10.
eStudent’s t-test for continuous variables and pearson’s chi-square test for categorical variables.
Associations of education with diet quality
As shown in Table 2 (for middle-aged women) and Table 3 (for older women), women with high and middle education had a higher total diet quality score than those with low education (model 1), even after adjustment for individual lifestyle variables (model 2), as well as individual lifestyle and neighborhood variables (model 3). However, the associations seemed to be due to different components of the diet quality score by generation. After adjustments for individual lifestyle and neighborhood variables (model 3), middle-aged women with high and middle education had higher scores of ‘milk’, ‘snacks, confection, and beverages’, ‘fruits’, and ‘vegetable dishes’ than those with low education. On the other hand, older women with high and middle education had higher scores of ‘sodium from seasonings’ and ‘fruits’ than those with low education. Although some lifestyle variables were associated with the diet quality score, no significant associations were found between neighborhood variables and diet quality except urban-rural classification (eTable 2).
Table 2. Associations of education with components of diet quality score in middle-aged women (n = 3,788)a.
| Model 1b | Model 2c | Model 3d | |||||||||||||
| Low | Middle | High | Low | Middle | High | Low | Middle | High | |||||||
| (n = 1,818) | (n = 1,420) | (n = 550) | (n = 1,818) | (n = 1,420) | (n = 550) | (n = 1,818) | (n = 1,420) | (n = 550) | |||||||
| mean (95% CIe) |
mean (95% CIe) |
Pf | mean (95% CIe) |
Pf | mean (95% CIe) |
mean (95% CIe) |
Pf | mean (95% CIe) |
Pf | mean (95% CIe) |
mean (95% CIe) |
Pf | mean (95% CIe) |
Pf | |
| Diet quality scoreg | 41.7 (41.3–42.1) |
43.4 (43.0–43.8) |
<0.001 | 45.4 (44.7–46.0) |
<0.001 | 42.0 (41.6–42.3) |
43.3 (42.9–43.7) |
<0.001 | 44.8 (44.1–45.5) |
<0.001 | 42.0 (41.6–42.3) |
43.3 (42.9–43.7) |
<0.001 | 44.8 (44.1–45.5) |
<0.001 |
| Components of diet quality scoreh | |||||||||||||||
| Grain dishes | 8.2 (8.1–8.3) |
8.2 (8.1–8.3) |
0.97 | 8.1 (7.9–8.2) |
0.27 | 8.1 (8.1–8.2) |
8.2 (8.1–8.3) |
0.22 | 8.2 (8.1–8.3) |
0.48 | 8.1 (8.0–8.2) |
8.2 (8.1–8.3) |
0.11 | 8.2 (8.1–8.4) |
0.22 |
| Vegetable dishes | 6.7 (6.6–6.8) |
7.3 (7.2–7.4) |
<0.001 | 7.5 (7.3–7.7) |
<0.001 | 6.9 (6.8–7.0) |
7.2 (7.1–7.3) |
<0.001 | 7.2 (7.0–7.3) |
0.01 | 6.9 (6.8–7.0) |
7.2 (7.1–7.3) |
<0.001 | 7.2 (7.0–7.4) |
0.01 |
| Fish and meat dishes |
9.8 (9.8–9.8) |
9.8 (9.8–9.9) |
0.17 | 9.9 (9.9–10.0) |
<0.001 | 9.8 (9.8–9.8) |
9.8 (9.8–9.9) |
0.75 | 9.9 (9.8–9.9) |
0.07 | 9.8 (9.8–9.8) |
9.8 (9.8–9.9) |
0.84 | 9.9 (9.8–9.9) |
0.12 |
| Milk | 5.6 (5.5–5.8) |
6.3 (6.1–6.4) |
<0.001 | 7.2 (6.9–7.5) |
<0.001 | 5.7 (5.5–5.9) |
6.2 (6.0–6.4) |
<0.001 | 7.0 (6.7–7.3) |
<0.001 | 5.7 (5.6–5.9) |
6.2 (6.0–6.4) |
<0.001 | 6.9 (6.6–7.2) |
<0.001 |
| Fruits | 2.9 (2.8–3.0) |
3.3 (3.1–3.4) |
<0.001 | 3.8 (3.6–4.1) |
<0.001 | 3.0 (2.9–3.2) |
3.2 (3.1–3.4) |
0.12 | 3.5 (3.3–3.7) |
<0.001 | 3.0 (2.9–3.2) |
3.2 (3.1–3.4) |
0.09 | 3.5 (3.3–3.8) |
<0.001 |
| Snacks, confection, and beverages |
4.6 (4.5–4.8) |
5.0 (4.8–5.2) |
0.01 | 5.4 (5.1–5.7) |
<0.001 | 4.7 (4.5–4.9) |
5.0 (4.8–5.2) |
0.04 | 5.3 (5.0–5.6) |
0.004 | 4.7 (4.5–4.8) |
5.0 (4.8–5.2) |
0.02 | 5.4 (5.0–5.7) |
0.001 |
| Sodium from seasonings |
3.9 (3.7–4.0) |
3.6 (3.4–3.7) |
0.04 | 3.5 (3.2–3.8) |
0.04 | 3.7 (3.6–3.9) |
3.6 (3.4–3.8) |
0.53 | 3.7 (3.4–4.0) |
0.99 | 3.7 (3.6–3.9) |
3.6 (3.4–3.8) |
0.55 | 3.7 (3.4–4.0) |
0.98 |
aLow education is ≤12 years (junior high school or high school). Middle education is 13–15 years (junior college or professional college). High education is ≥16 years (university).
bModel 1: crude model.
cModel 2: adjusted for age, body mass index, current marital status, employment status, smoking status, physical activity, prescription medicine use, and diet cost.
dModel 3: adjusted for variables in model 2, urban-rural classification, percentage of workers in primary sector industry, areal deprivation index score, number of food retailers, and region.
eConfidence interval.
fDunnett’s multiple comparison was conducted (reference: low).
gPossible score ranging from 0 to 70.
hPossible score ranging from 0 to 10.
Table 3. Associations of education with components of diet quality score in older women (n = 2,188)a.
| Model 1b | Model 2c | Model 3d | |||||||||||||
| Low | Middle | High | Low | Middle | High | Low | Middle | High | |||||||
| (n = 981) | (n = 994) | (n = 213) | (n = 981) | (n = 994) | (n = 213) | (n = 981) | (n = 994) | (n = 213) | |||||||
| mean (95% CIe) |
mean (95% CIe) |
Pf | mean (95% CIe) |
Pf | mean (95% CIe) |
mean (95% CIe) |
Pf | mean (95% CIe) |
Pf | mean (95% CIe) |
mean (95% CIe) |
Pf | mean (95% CIe) |
Pf | |
| Diet quality scoreg | 49.6 (49.1–50.1) |
50.9 (50.4–51.4) |
<0.001 | 52.4 (51.3–53.5) |
<0.001 | 49.7 (49.2–50.2) |
50.9 (50.4–51.3) |
0.004 | 52.2 (51.2–53.3) |
<0.001 | 49.8 (49.3–50.3) |
50.8 (50.3–51.3) |
0.013 | 52.2 (51.1–53.2) |
<0.001 |
| Components of diet quality scoreh | |||||||||||||||
| Grain dishes | 8.4 (8.3–8.5) |
8.1 (8.0–8.3) |
0.003 | 7.8 (7.6–8.1) |
<0.001 | 8.3 (8.2–8.4) |
8.2 (8.1–8.3) |
0.75 | 8.2 (8.0–8.4) |
0.86 | 8.3 (8.2–8.3) |
8.2 (8.1–8.3) |
0.72 | 8.2 (8.0–8.4) |
0.96 |
| Vegetable dishes | 8.9 (8.8–9.0) |
9.1 (9.0–9.2) |
0.08 | 9.4 (9.2–9.7) |
<0.001 | 9.0 (8.9–9.1) |
9.0 (8.9–9.1) |
0.98 | 9.2 (9.0–9.4) |
0.20 | 9.0 (8.9–9.1) |
9.0 (8.9–9.1) |
0.98 | 9.2 (9.0–9.4) |
0.17 |
| Fish and meat dishes |
9.9 (9.9–10.0) |
9.9 (9.9–10.0) |
0.55 | 10.0 (9.9–10.1) |
0.48 | 10.0 (9.9–10.0) |
9.9 (9.9–9.9) |
0.24 | 10.0 (9.9–10.0) |
0.98 | 10.0 (9.9–10.0) |
9.9 (9.9–9.9) |
0.19 | 10.0 (9.9–10.0) |
0.99 |
| Milk | 5.6 (5.4–5.8) |
6.3 (6.0–6.5) |
<0.001 | 6.4 (5.9–6.8) |
0.02 | 5.7 (5.5–5.9) |
6.2 (6.0–6.4) |
0.01 | 6.2 (5.7–6.7) |
0.17 | 5.7 (5.5–5.9) |
6.2 (6.0–6.4) |
0.01 | 6.1 (5.6–6.6) |
0.24 |
| Fruits | 4.8 (4.6–5.0) |
5.5 (5.3–5.7) |
<0.001 | 6.4 (6.0–6.8) |
<0.001 | 4.9 (4.8–5.1) |
5.4 (5.2–5.6) |
<0.001 | 6.0 (5.6–6.4) |
<0.001 | 5.0 (4.8–5.2) |
5.4 (5.2–5.6) |
0.004 | 6.0 (5.6–6.4) |
<0.001 |
| Snacks, confection, and beverages |
6.0 (5.8–6.3) |
6.0 (5.7–6.3) |
0.98 | 5.7 (5.1–6.2) |
0.43 | 6.1 (5.8–6.3) |
6.0 (5.7–6.2) |
0.77 | 5.6 (5.1–6.2) |
0.28 | 6.1 (5.8–6.4) |
5.9 (5.7–6.2) |
0.61 | 5.7 (5.1–6.2) |
0.29 |
| Sodium from seasonings |
5.9 (5.8–6.1) |
6.1 (5.9–6.3) |
0.42 | 6.7 (6.4–7.1) |
<0.001 | 5.8 (5.6–6.0) |
6.2 (6.0–6.3) |
0.01 | 7.1 (6.7–7.4) |
<0.001 | 5.8 (5.7–6.0) |
6.1 (6.0–6.3) |
0.02 | 7.0 (6.7–7.4) |
<0.001 |
aLow education is ≤9 years (junior high school). Middle education is 10–12 years (high school). High education is ≥13 years (junior college, professional college or university).
bModel 1: crude model.
cModel 2: adjusted for age, body mass index, living status, current marital status, smoking status, physical activity, prescription medicine use, and diet cost.
dModel 3: adjusted for variables in model 2, urban-rural classification, percentage of workers in primary sector industry, areal deprivation index score, number of food retailers, and region.
eConfidence interval.
fDunnett’s multiple comparison was conducted (reference: low).
gPossible score ranging from 0 to 70.
hPossible score ranging from 0 to 10.
DISCUSSION
To our knowledge, this is the first study to investigate which food groups explain the associations between education and overall diet quality by generation. This study found that positive associations between education and diet quality were explained by different food groups in middle-aged and older Japanese women, namely by ‘milk’ and ‘sodium from seasonings’, which are independent of lifestyle and neighborhood variables.
In previous studies, education was positively associated with overall diet quality in both middle-aged and older women.2–4,9 This is consistent with this study. Interestingly, however, this study showed that associations between education and food components of the diet quality score differed by generation. The score difference between high and low education for middle-aged women was the largest for ‘milk’, followed by ‘snacks, confection, and beverages’, ‘fruits’, and ‘vegetable dishes’, while in older women, it was the largest for ‘sodium from seasonings’, followed by ‘fruits’. Although the exact reason is unknown, generation difference in time spent cooking6 and food preference10,11 may be related to the associations. In previous studies, old age was associated with longer time spent cooking,6 putting more value on nutrition and less value on cost10,11 and convenience.10 Therefore, the present study suggested that middle-aged women with high education tended to improve their diet quality by eating food that did not require long cooking, such as increasing their intake of ‘milk’, ‘fruits’, and ‘vegetable dishes’ (such as salad), and decreasing their intake of ‘snacks, confection, and beverages’. This is partly consistent with a study in France, which shows that middle-aged women with high education spent less time in meal preparation but prepared food from scratch (eg, use of raw or fresh ingredients) more frequently than those with low education.7 Conversely, older women with high education may improve their diet quality by cooking in a healthy manner, such as reducing their intake of ‘sodium from seasonings’, in addition to increasing their intake of ‘fruits’. Since the main source of salt intake in Japan was discretionary salt (57.1%),32 older women with high education may use seasoning more carefully than those with low education.
Moreover, the present study shows that associations between education and diet quality were independent of neighborhood variables, suggesting that the associations were not mediated by neighborhood contexts in this study. This may be partly because of limited association of food intakes with neighborhood SES13 or store availability in Japan.18 Actually, no significant associations were found in this study between neighborhood variables and diet quality except urban-rural classification.
The present findings might be useful for developing a public health policy to reduce disparity of diet quality by education in Japan. There may be a greater need to address the issue of ‘milk’, ‘snacks, confection, and beverages’, ‘fruits’, and ‘vegetable dishes’ for middle-aged women, while ‘sodium from seasonings’ and ‘fruits’ for older women.
The present study has several limitations. First, the participants were mothers and grandmothers of dietetic students, and not a random sample of Japanese middle-aged and older women. Nevertheless, the participants were comparable to the general middle-aged and older women in terms of education (low, high: 54%, 16% for the former, and 40%, 11% for the latter)28 and the diet quality score (mean 42.0; standard deviation, 8.9 for the former, and mean 47.1; standard deviation, 9.6 for the latter),33 but not in terms of living status (those living alone: 9%, 20%, respectively),28 or current marital status (those who are married: 72%, 54%, respectively).28 Second, neighborhood variables were measured using municipality level data because it was the smallest spatial unit available in this study. However, municipality may be too large to be used as a neighborhood influencing diet (average area size: 204.4 km2). Moreover, the data was not representative of each municipality and only included a small number of participants in each municipality (average: 3.6 for middle-aged women and 2.6 for older women). In addition, duration of residence was not taken into account for analysis because of a lack of information. Third, a self-report dietary assessment was conducted which is subject to measurement error particularly caused by the misreporting of food intake. To minimize measurement error, we assessed the dietary habits using a well-established assessment questionnaire (ie. DHQ and BDHQ) with reasonable validity21–23 as well as energy-adjusted dietary values for calculating the diet quality score. Additionally, exclusion of energy intake misreporters, evaluated by the ratio of energy intake to basal metabolic rate (the Goldberg cut-off),19,20,34 did not change most results (data not shown), which may support the robustness of the present finding. Fourth, because different dietary assessment questionnaires were used for middle-aged (DHQ) and older (BDHQ) women, it is not possible to directly compare differences in the diet quality score by education between generations. Nevertheless, the diet quality score was similarly associated with nutrient intakes in both generations.19 Fifth, we lacked data on personal or household income as a residual confounder. People with low income tended to choose low-cost foods, which are generally energy-dense and nutrient-poor.1 Since the association between income and diet quality may be partly mediated by diet cost,1 the dietary cost was adjusted in this study to minimize potential influence of income on the association of education with diet quality.
In conclusion, this study suggests that positive associations between education and diet quality are explained by different food groups in middle-aged and older Japanese women, which are independent of lifestyle and neighborhood variables. The present finding might be useful for developing a public health policy to reduce disparity of diet quality by education in Japan.
ACKNOWLEDGEMENTS
The authors thank the members of the Three-generation Study of Women on Diets and Health Study Group. The members are listed in: Kobayashi S, Asakura K, Suga H, Sasaki S & Three-generation Study of Women on Diets and Health Study Group. High protein intake is associated with low prevalence of frailty among old Japanese women: a multicenter cross-sectional study. Nutr J 2013;12:164.
This work was supported by the Grants-in-Aid for Scientific Research (A) [grant number 22240077] from the Japan Society for the Promotion of Science. The Japan Society for the Promotion of Science had no role in the design, analysis or writing of this article.
Conflicts of interest: None declared.
Authorship: A.H. conceptualized the study question, conducted the statistical analysis, interpreted the data, and prepared the first draft of the manuscript. K.M. contributed to the concept and design of the survey, conceptualized the study question, interpreted the data, and provided critical input into the final draft of the manuscript. S.K., H.S. contributed to the concept and design of the survey, coordination of the fieldwork, data collection and management. S.S. designed and supervised the data collection and provided critical input into the final draft of the manuscript. All authors read and approved the final manuscript.
APPENDIX A. SUPPLEMENTARY DATA
The following is the supplementary data related to this article:
eTable 1. Scoring system of the diet quality score
eTable 2. Associations of individual lifestyle variables and neighborhood variables with diet quality score
REFERENCES
- 1.Darmon N, Drewnowski A. Does social class predict diet quality? Am J Clin Nutr. 2008;87(5):1107–1117. 10.1093/ajcn/87.5.1107 [DOI] [PubMed] [Google Scholar]
- 2.Rehm CD, Peñalvo JL, Afshin A, Mozaffarian D. Dietary intake among US adults, 1999–2012. JAMA. 2016;315(23):2542–2553. 10.1001/jama.2016.7491 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Livingstone KM, Olstad DL, Leech RM, et al. Socioeconomic inequities in diet quality and nutrient intakes among Australian adults: findings from a nationally representative cross-sectional study. Nutrients. 2017;9(10):E1092. 10.3390/nu9101092 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Backholer K, Spencer E, Gearon E, et al. The association between socio-economic position and diet quality in Australian adults. Public Health Nutr. 2016;19(3):477–485. 10.1017/S1368980015001470 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Shrivastava A, Murrin C, Sweeney MR, Heavey P, Kelleher CC; Lifeways Cohort Study Steering Group . Familial intergenerational and maternal aggregation patterns in nutrient intakes in the Lifeways Cross-Generation Cohort Study. Public Health Nutr. 2013;16(8):1476–1486. 10.1017/S1368980012003667 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Adams J, White M. Prevalence and socio-demographic correlates of time spent cooking by adults in the 2005 UK Time Use Survey. Cross-sectional analysis. Appetite. 2015;92:185–191. 10.1016/j.appet.2015.05.022 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Méjean C, Si Hassen W, Gojard S, et al. Social disparities in food preparation behaviours: a DEDIPAC study. Nutr J. 2017;16(1):62. 10.1186/s12937-017-0281-2 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Murakami K, Miyake Y, Sasaki S, Tanaka K, Ohya Y, Hirota Y; Osaka Maternal and Child Health Study Group . Education, but not occupation or household income, is positively related to favorable dietary intake patterns in pregnant Japanese women: the Osaka Maternal and Child Health Study. Nutr Res. 2009;29(3):164–172. 10.1016/j.nutres.2009.02.002 [DOI] [PubMed] [Google Scholar]
- 9.Schoufour JD, de Jonge EAL, Kiefte-de Jong JC, et al. Socio-economic indicators and diet quality in an older population. Maturitas. 2018;107:71–77. 10.1016/j.maturitas.2017.10.010 [DOI] [PubMed] [Google Scholar]
- 10.Glanz K, Basil M, Maibach E, Goldberg J, Snyder D. Why Americans eat what they do: taste, nutrition, cost, convenience, and weight control concerns as influences on food consumption. J Am Diet Assoc. 1998;98(10):1118–1126. 10.1016/S0002-8223(98)00260-0 [DOI] [PubMed] [Google Scholar]
- 11.Aggarwal A, Rehm CD, Monsivais P, Drewnowski A. Importance of taste, nutrition, cost and convenience in relation to diet quality: evidence of nutrition resilience among US adults using National Health and Nutrition Examination Survey (NHANES) 2007–2010. Prev Med. 2016;90:184–192. 10.1016/j.ypmed.2016.06.030 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Thornton LE, Crawford DA, Ball K. Neighbourhood-socioeconomic variation in women’s diet: the role of nutrition environments. Eur J Clin Nutr. 2010;64(12):1423–1432. 10.1038/ejcn.2010.174 [DOI] [PubMed] [Google Scholar]
- 13.Murakami K, Sasaki S, Okubo H, Takahashi Y; Freshmen in Dietetic Courses Study II Group . Neighborhood socioeconomic status in relation to dietary intake and body mass index in female Japanese dietetic students. Nutrition. 2009;25(7–8):745–752. 10.1016/j.nut.2009.01.010 [DOI] [PubMed] [Google Scholar]
- 14.McInerney M, Csizmadi I, Friedenreich CM, et al. Associations between the neighbourhood food environment, neighbourhood socioeconomic status, and diet quality: an observational study. BMC Public Health. 2016;16:984. 10.1186/s12889-016-3631-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Dolman RC, Wentzel-Viljoen E, Jerling JC, Feskens EJ, Kruger A, Pieters M. The use of predefined diet quality scores in the context of CVD risk during urbanization in the South African Prospective Urban and Rural Epidemiological (PURE) study. Public Health Nutr. 2014;17(8):1706–1716. 10.1017/S1368980013002206 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Trivedi T, Liu J, Probst J, Merchant A, Jhones S, Martin AB. Obesity and obesity-related behaviors among rural and urban adults in the USA. Rural Remote Health. 2015;15(4):3267. [PubMed] [Google Scholar]
- 17.Black C, Moon G, Baird J. Dietary inequalities: what is the evidence for the effect of the neighbourhood food environment? Health Place. 2014;27:229–242. 10.1016/j.healthplace.2013.09.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Murakami K, Sasaki S, Takahashi Y, Uenishi K; Japan Dietetic Students’ Study for Nutrition and Biomarkers Group . Neighborhood food store availability in relation to food intake in young Japanese women. Nutrition. 2009;25(6):640–646. 10.1016/j.nut.2009.01.002 [DOI] [PubMed] [Google Scholar]
- 19.Kuriyama N, Murakami K, Livingstone MBE, et al. ; Three-generation Study of Women on Diets and Health Study Group . Development of a food-based diet quality score for Japanese: associations of the score with nutrient intakes in young, middle-aged and older Japanese women. J Nutr Sci. 2016;5:e41. 10.1017/jns.2016.36 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Shiraki K, Murakami K, Okubo H, et al. ; Three-Generation Study of Women on Diets and Health Study Group . Nutritional correlates of monetary diet cost in young, middle-aged and older Japanese women. J Nutr Sci. 2017;6:e22. 10.1017/jns.2017.18 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Kobayashi S, Murakami K, Sasaki S, et al. Comparison of relative validity of food group intakes estimated by comprehensive and brief-type self-administered diet history questionnaires against 16 d dietary records in Japanese adults. Public Health Nutr. 2011;14(7):1200–1211. 10.1017/S1368980011000504 [DOI] [PubMed] [Google Scholar]
- 22.Kobayashi S, Honda S, Murakami K, et al. Both comprehensive and brief self-administered diet history questionnaires satisfactorily rank nutrient intakes in Japanese Adults. J Epidemiol. 2012;22(2):151–159. 10.2188/jea.JE20110075 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Sasaki S, Yanagibori R, Amano K. Self-administered diet history questionnaire developed for health education: a relative validation of the test-version by comparison with 3-day diet record in women. J Epidemiol. 1998;8(4):203–215. 10.2188/jea.8.203 [DOI] [PubMed] [Google Scholar]
- 24.Science and Technology Agency. Standard Tables of Food Composition in Japan. Tokyo: Official Gazette Co-operation of Japan; 2010 (in Japanese).
- 25.Sakai H, Murakami K, Kobayashi S, Suga H, Sasaki S; Three-generation Study of Women on Diets and Health Study Group . Food-based diet quality score in relation to depressive symptoms in young and middle-aged Japanese women. Br J Nutr. 2017;117(12):1674–1681. 10.1017/S0007114517001581 [DOI] [PubMed] [Google Scholar]
- 26.Ainsworth BE, Haskell WL, Herrmann SD, et al. 2011 Compendium of Physical Activities: a second update of codes and MET values. Med Sci Sports Exerc. 2011;43(8):1575–1581. 10.1249/MSS.0b013e31821ece12 [DOI] [PubMed] [Google Scholar]
- 27.Statistics Bureau, Ministry of Internal Affairs and Communications. The 2012 Retail Price survey (in Japanese): https://www.stat.go.jp/data/kouri/doukou/index.html; 2012 Accessed 01.10.18.
- 28.Statistics Bureau, Ministry of Internal Affairs and Communications. The 2010 Population Census (in Japanese): https://www.stat.go.jp/data/kokusei/2010/; 2010 Accessed 01.10.18.
- 29.Nakaya T. Evaluating socio-economic inequalities in cancer mortality by using areal statistics in Japan: a note on the relation between municipal cancer mortality and areal deprivation index. Proc Inst Stat Math. 2011;59(2):239–265 (in Japanese). [Google Scholar]
- 30.Nakaya T, Honjo K, Hanibuchi T, et al. ; Japan Public Health Center-based Prospective Study Group . Associations of all-cause mortality with census-based neighbourhood deprivation and population density in Japan: a multilevel survival analysis. PLoS One. 2014;9(6):e97802. 10.1371/journal.pone.0097802 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Statistics Bureau, Ministry of Internal Affairs and Communications. The 2012 Economic Census for Business Activity (in Japanese): https://www.stat.go.jp/data/e-census/2016/gaiyo.html; 2016 Accessed 01.10.18.
- 32.Asakura K, Uechi K, Masayasu S, Sasaki S. Sodium sources in the Japanese diet: difference between generations and sexes. Public Health Nutr. 2016;19(11):2011–2023. 10.1017/S1368980015003249 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Murakami K, Livingstone MBE, Sasaki S. Diet quality scores in relation to metabolic risk factors in Japanese adults: a cross-sectional analysis from the 2012 National Health and Nutrition Survey, Japan. Eur J Nutr. 2019;58(5):2037–2050. 10.1007/s00394-018-1762-6 [DOI] [PubMed] [Google Scholar]
- 34.Black AE. Critical evaluation of energy intake using the Goldberg cut-off for energy intake: basal metabolic rate. A practical guide to its calculation, use and limitations. Int J Obes Relat Metab Disord. 2000;24(9):1119–1130. 10.1038/sj.ijo.0801376 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.