Abstract
A meta-analysis of cohort studies found a positive association between white rice consumption and chronic disease risk, particularly in women. However, the association between rice intake and obesity remains inconsistent across populations. We aimed to examine the relationship between rice intake and obesity stratified by sex and age. This cross-sectional study used nationwide registry data from Japanese type 2 diabetes outpatients (2014–2019). Obesity was defined as BMI ≥25 kg/m2. The study included 1,565 outpatients aged 30–89 years (mean age: 62.3 ± 11.6 years), with 63.1% being male. Rice intake was associated with a diet low in energy from protein, fiber density, and dairy products. In adjusted multivariate analysis, older women in the highest tertile of rice intake had a higher prevalence of obesity (95% CI = 1.104–4.260, p trend = 0.042); however, this association lost significance after adjusting for fiber density (95% CI = 0.864–3.558, p trend = 0.080). In younger women, an inverse association with obesity emerged after fiber density adjustment in the supplementary quartile analysis. No significant associations were found in men. These results suggest that the association between rice intake and obesity is influenced by overall dietary quality rather than rice consumption alone. Promoting greater dietary diversity while maintaining traditional staples like rice may be a practical strategy to improve diet quality in Japan. Prospective studies in Japanese and other populations are needed to confirm these associations.
Introduction
In the Asia Pacific region, as well as parts of the Caribbean and Latin America, rice serves as a primary source of both calories and nutrition. In Southeast Asia, where consumption is the highest, rice can make up to 76% of caloric intake, and an upward trajectory in rice consumption has been observed in Africa [1]. Even though brown rice contains more nutrients, bran, and other biological compounds than white rice, consumers prefer polished white rice [2]. The nutrient content of rice varies, depending on factors such as rice variety, soil quality, milling degree and cooking methods [3]. Moreover, fortification of white rice with iron and folate is mandatory in countries such as the US [4].
One meta-analysis of 12 cohort studies revealed a positive association between white rice consumption and risk of overall chronic diseases in women, but not in men [5]. The authors highlighted that chronic disease events were more frequent in women, and although women tend to use more health-care services and have a longer lifespan, they live with greater disabilities. Studies have shown that women tend to have higher levels of protective HDL [6]; therefore, higher glycemic index diets, which reduce protective HDL, could have a greater impact on women with increased white rice intake [7]. Lifestyle differences, such as limited physical activity and unhealthy dietary patterns, which can vary by country, underscore the need for studies in different societies to clarify these findings. A positive association between white rice intake and the incidence of type 2 diabetes, especially in women, was reported in a meta-analysis of six prospective studies [8]. The authors suggested that the greater muscle mass in men could counteract the effects of white rice consumption on the development of the disease due to the higher metabolism of glucose.
Unfortunately, studies on the association between rice and obesity are conflicting: in Iranian adults, no association was found between rice intake and obesity [9]; however, rice-based dietary patterns were associated with greater BMI and waist circumference in a population of Hispanic elders from the US [10]. Whereas a cohort in China reported an inverse association between rice intake and obesity [11]. Different dietary patterns associated with rice intake, variations in age or sex demographics and other socioeconomic factors are all possible explanations for the conflicting findings.
Considering obesity as a key modifiable factor for type 2 diabetes and other chronic diseases, and the prominence of rice as an important energy source in various populations, this study aims to investigate the relationship between rice intake and obesity, while stratifying by sex and age, along with an exploration of dietary interactions in Japanese outpatients with type 2 diabetes.
Subjects and Methods
Study population and design
This cross-sectional investigation included outpatients diagnosed with type 2 diabetes and receiving treatment within Japanese clinics affiliated with the Japan Diabetes Clinical Data Management Study Group (JDDM) between December 2014 and December 2019. Comprehensive details outlining the JDDM and data collection software have been previously reported [12]. The cohort comprised 1,565 Japanese outpatients aged 30–89 years (mean age: 62.3 ± 11.6 years) with type 2 diabetes, of whom 987 (63.1%) were male.
Anthropometric, dietary, and lifestyle assessments
Participating clinics administered a lifestyle questionnaire, developed by JDDM, to willing outpatients. Participants self-reported their height and weight, while dietary habits were assessed using a validated self-administered Food Frequency Questionnaire [13] with the calculation of nutrient and food intakes being performed using standardized software (Eiyo-kun; Kenpakusha Co., Ltd., Tokyo, Japan) [14]. Participants with energy intake of ≤600 or ≥4,000 kcal per day were excluded from the analysis due to concerns about potential underreporting or overreporting. Supplement intake was not recorded. Physical activity was determined using a self-administered short version of the Japanese International Physical Activity Questionnaire [15, 16].
Statistical analysis
Categorical variables were compared using the chi-square test and expressed as numbers with percentages. Continuous variables were compared using the Jonckheere test and expressed as means with standard deviations. For metabolic equivalent of task (MET), the median was reported due to its highly skewed distribution. Binary logistic regression analysis was used to assess tertiles of rice intake and obesity, which was defined as body mass index (BMI) ≥25 kg/m2 according to the Japan Society for the Study of Obesity [17]. Demographic, lifestyle, and nutritional factors were analyzed as potential confounders. Participants were divided into four groups for the age- and sex-stratification analysis: younger men, older men, younger women, and older women. Younger groups were ≤65 years of age and older groups were >65 years of age. Supplementary analyses using quartiles of rice intake and MET stratification were also performed. Statistical analyses were performed using SPSS software (version 27.0; IBM Corp., Armonk, NY, USA). P < 0.05 was considered statistically significant. Results are presented with 95% confidence intervals.
Results
Characteristics of all study participants and those in four age- and sex-stratified groups
Table 1 shows the differences in lifestyle factors and intake of nutrient and food groups in younger and older men or women. Younger men had the highest intake of rice, processed meat/meat, and eggs, with the highest daily saturated fat intake. Younger women had the highest intake of sweets. Older men had the highest percentage of energy intake from carbohydrates. Older women had the highest dietary fiber density, percentage of energy intake from protein, total vegetables, fruits, soybeans/soy products, fish/seafood and dairy products. Regarding lifestyle habits, alcohol consumption, smoking rates, and physical activity were higher in men compared to women.
Table 1 Basic characteristics of all study participants in four age- and sex- stratification groups
|
All participants
n = 1,565 |
Younger women (≤65 y)
n = 301 |
Older women (>65 y)
n = 277 |
p |
Younger men (≤65 y)
n = 620 |
Older men (>65 y)
n = 367 |
p |
| Sex (men, %) |
987 (63.1) |
|
|
N/A |
|
|
N/A |
| Age (years) |
62.3 (11.6) |
55.3 (8.5) |
73.7 (5.4) |
<0.001 |
54.3 (7.7) |
72.9 (5.5) |
<0.001 |
| Body mass index (kg/m2) |
25.8 (4.6) |
27.1 (5.8) |
24.6 (4.0) |
<0.001 |
26.7 (4.5) |
24.2 (3.0) |
<0.001 |
| Diabetes duration (years) |
11.8 (7.8) |
9.7 (6.7) |
14.0 (8.1) |
<0.001 |
9.9 (6.6) |
14.9 (8.6) |
<0.001 |
| Current smoker (%) |
302 (19.3) |
31 (10.3) |
14 (5.1) |
<0.001 |
201 (32.4) |
56 (15.3) |
<0.001 |
| Current alcohol consumption (%) |
732 (46.8) |
94 (31.2) |
52 (18.8) |
<0.001 |
368 (59.4) |
218 (59.4) |
0.522 |
| Current insulin treatment (%) |
451 (28.8) |
87 (28.9) |
99 (35.7) |
0.048 |
148 (23.9) |
117 (31.9) |
0.004 |
| Current OHA and/or GLP treatment (%) |
1,277 (81.6) |
246 (81.7) |
227 (81.9) |
0.516 |
517 (83.4) |
287 (78.2) |
0.027 |
| MET (hour/week) |
15.4 (5.5–35.1) |
13.2 (5.1–30.2) |
12.3 (4.9–26.6) |
0.513 |
13.2 (4.5–35.5) |
23.1 (8.9–46.8) |
<0.001 |
| Energy (kcal/day) |
1,765.7 (437.1) |
1,687.6 (395.1) |
1,650.7 (350.8) |
0.150 |
1,818.7 (469.7) |
1,827.1 (446.3) |
0.533 |
| Carbohydrate/energy ratio (%) |
55.7 (6.5) |
54.2 (6.4) |
55.4 (4.9) |
0.006 |
55.6 (7.2) |
57.2 (6.1) |
<0.001 |
| Protein/energy ratio (%) |
14.7 (2.2) |
14.7 (2.0) |
15.6 (1.9) |
<0.001 |
14.4 (2.3) |
14.8 (2.1) |
0.002 |
| Fat/energy ratio (%) |
29.4 (5.3) |
31.0 (5.4) |
28.9 (4.0) |
<0.001 |
29.9 (5.7) |
27.9 (5.0) |
<0.001 |
| Na/K ratioa |
1.5 (0.4) |
1.4 (0.4) |
1.5 (0.4) |
0.018 |
1.6 (0.5) |
1.6 (0.4) |
0.636 |
Fiber density
(g/1,000 kcal/day) |
6.8 (1.9) |
7.3 (1.7) |
8.0 (1.8) |
<0.001 |
6.0 (1.5) |
7.1 (1.9) |
<0.001 |
| Saturated fat (g/day) |
17.9 (6.6) |
18.3 (6.8) |
16.5 (5.6) |
0.002 |
18.5 (7.0) |
17.6 (6.4) |
0.054 |
| Rice (g/day) |
254.8 (118.7) |
232.1 (100.9) |
236.1 (104.0) |
0.818 |
275.3 (131.5) |
252.9 (114.3) |
0.012 |
| Total vegetables (g/day) |
230.6 (118.2) |
247.8 (113.0) |
266.0 (119.4) |
0.089 |
196.1 (107.5) |
248.1 (123.9) |
<0.001 |
| Fruits (g/day) |
84.2 (71.3) |
85.1 (68.2) |
109.9 (71.2) |
<0.001 |
59.4 (62.7) |
105.7 (73.8) |
<0.001 |
| Soybeans/soy products (g/day) |
60.4 (42.1) |
59.6 (40.1) |
67.1 (40.8) |
0.009 |
54.9 (42.0) |
65.3 (43.6) |
<0.001 |
| Fish/seafood (g/day) |
72.9 (45.3) |
60.2 (35.6) |
83.4 (41.1) |
<0.001 |
68.5 (47.8) |
82.6 (47.5) |
<0.001 |
Processed meat/meat
(g/day) |
76.0 (49.5) |
79.6 (49.0) |
57.1 (35.0) |
<0.001 |
89.0 (54.5) |
65.1 (43.4) |
<0.001 |
| Eggs (g/day) |
27.3 (19.0) |
25.2 (16.3) |
24.5 (16.6) |
0.483 |
29.9 (21.4) |
26.9 (18.0) |
0.067 |
| Dairy products (g/day) |
131.5 (101.6) |
125.6 (90.6) |
148.6 (98.7) |
0.003 |
117.6 (103.2) |
147.0 (106.0) |
<0.001 |
| Sweets (g/day) |
51.1 (42.8) |
58.3 (41.5) |
47.1 (41.1) |
<0.001 |
51.6 (43.7) |
47.5 (42.8) |
0.053 |
Significant differences are highlighted in bold (p < 0.05).
MET, metabolic equivalent of task; OHA, oral antihyperglycemic agents; GLP, glucagon-like peptide.
Data are presented as mean (±standard deviation) and n (%). Physical activity (metabolic equivalent of task [MET] hour/week) as median (interquartile range).
a Na/K ratio was calculated as sodium intake (mg/day) divided by potassium intake (mg/day).
Table 2 presents the binary logistic regression analysis of associations between tertiles of rice intake in all participants and with age- and sex- stratifications. In the model adjusted solely for age (Model 1), no associations were found between tertiles of rice intake and obesity in all participants or subgroups. After the inclusion of energy intake adjustment (Model 2), a significant association emerged for the highest tertile of rice intake and obesity in the older women subgroup (95% CI: 1.106–4.036, p trend = 0.028). This association persisted after adjustments for lifestyle factors, diabetes duration, and diabetes treatment status (Model 3), and remained significant (95% CI: 1.104–4.260, p trend = 0.042). However, after additional adjustment for fiber density (Model 4), the association was no longer significant (95% CI: 0.864–3.558, p trend = 0.080).
Table 2 Binary logistic regression analysis of tertiles of rice intake and obesity in all participants and groups stratified by sex and age
|
Rice intake (mean; SD) |
Model 1 OR (CI) |
Model 2 OR (CI) |
Model 3 OR (CI) |
Model 4 OR (CI) |
All participants
(n = 1,565) |
T1 (141.4; 58.5) |
Reference |
Reference |
Reference |
Reference |
| T2 (259.3; 21.2) |
1.097 (0.852–1.413) |
1.070 (0.828–1.381) |
1.062 (0.820–1.375) |
0.924 (0.709–1.206) |
| T3 (381.3; 87.6) |
1.203 (0.939–1.540) |
1.136 (0.876–1.473) |
1.070 (0.822–1.393) |
0.893 (0.678–1.175) |
| p trend |
0.342 |
0.629 |
0.853 |
0.706 |
Younger Women
(n = 301) |
T1 (145.1; 55.2) |
Reference |
Reference |
Reference |
Reference |
| T2 (258.5; 21.6) |
1.175 (0.650–2.124) |
1.096 (0.602–1.996) |
1.169 (0.631–2.164) |
1.100 (0.590–2.052) |
| T3 (355.6; 46.9) |
0.808 (0.464–1.409) |
0.716 (0.403–1.272) |
0.686 (0.379–1.242) |
0.597 (0.319–1.119) |
| p trend |
0.520 |
0.383 |
0.273 |
0.166 |
Older Women
(n = 277) |
T1 (144.2; 55.8) |
Reference |
Reference |
Reference |
Reference |
| T2 (258.5; 20.8) |
0.849 (0.485–1.487) |
0.906 (0.513–1.599) |
0.962 (0.533–1.735) |
0.790 (0.423–1.472) |
| T3 (372.1; 74.9) |
1.825 (0.987–3.375) |
2.113 (1.106–4.036) |
2.169 (1.104–4.260) |
1.753 (0.864–3.558) |
| p trend |
0.056 |
0.028 |
0.042 |
0.080 |
Younger Men
(n = 620) |
T1 (140.3; 61.5) |
Reference |
Reference |
Reference |
Reference |
| T2 (260.0; 21.3) |
1.387 (0.904–2.130) |
1.366 (0.886–2.106) |
1.287 (0.828–2.001) |
1.164 (0.741–1.828) |
| T3 (396.7; 100.5) |
1.313 (0.889–1.941) |
1.261 (0.828–1.920) |
1.166 (0.756–1.798) |
1.020 (0.651–1.597) |
| p trend |
0.249 |
0.335 |
0.526 |
0.771 |
Older men
(n = 367) |
T1 (136.6; 59.8) |
Reference |
Reference |
Reference |
Reference |
| T2 (259.6; 21.2) |
0.910 (0.543–1.525) |
0.847 (0.501–1.433) |
0.855 (0.499–1.466) |
0.708 (0.406–1.232) |
| T3 (372.3; 81.7) |
1.056 (0.634–1.758) |
0.937 (0.549–1.596) |
0.880 (0.503–1.543) |
0.685 (0.381–1.234) |
| p trend |
0.855 |
0.824 |
0.836 |
0.362 |
Significant differences and trends are highlighted in bold (p < 0.05).
OR, odds ratio; SD, standard deviation; CI, confidence interval.
Model 1: Adjusted for sex (except in the stratified analysis by sex and age) and age.
Model 2: Model 1 plus energy intake.
Model 3: Model 2 plus diabetes duration, smoking status, alcohol consumption, insulin treatment, OHA or GLP use, physical activity (metabolic equivalent of task).
Model 4: Model 3 plus fiber density
Table 3 shows the differences in dietary intakes and lifestyle habits according to tertiles of rice intake in younger and older men and women. In all groups, rice intake was negatively associated with fiber density. In older women, processed meat/meat intake significantly increased with higher rice intake, while dairy product intake decreased. In younger women, higher rice intake was linked to significantly greater soybeans/soy products and fish/seafood consumption, and a non-significant decrease in sweets. In younger men, higher rice intake was associated with significantly greater total vegetables, soybeans/soy products, fish/seafood, and processed meat/meat consumption. In older men, higher rice intake was linked to significantly greater fish/seafood intake.
Table 3 Dietary and lifestyle factors according to tertiles of rice intake stratified by age and sex
| Groups |
Variables |
T1 |
T2 |
T3 |
p |
| Younger Women (n = 301) |
Body mass index (kg/m2) |
26.7 (5.2) |
27.7 (6.0) |
27.2 (6.8) |
0.694 |
| Energy (kcal/day) |
1,603.7 (409.1) |
1,716.6 (348.7) |
1,803.3 (380.9) |
<0.001 |
| Na/K ratioa |
1.5 (0.4) |
1.4 (0.3) |
1.4 (0.4) |
0.037 |
| Fiber density (g/1,000 kcal/day) |
7.8 (1.8) |
7.2 (1.5) |
6.6 (1.4) |
<0.001 |
| Saturated fat (g/day) |
18.2 (6.3) |
18.1 (6.2) |
18.5 (8.1) |
0.957 |
| Carbohydrate/energy ratio (%) |
52.2 (5.8) |
55.1 (5.9) |
56.8 (6.7) |
<0.001 |
| Protein/energy ratio (%) |
14.9 (2.3) |
14.4 (1.6) |
14.4 (2.0) |
0.347 |
| Fat/energy ratio (%) |
32.7 (4.8) |
30.3 (5.3) |
28.6 (5.7) |
<0.001 |
| Total vegetables (g/day) |
245.5 (120.4) |
248.7 (99.6) |
250.9 (112.6) |
0.532 |
| Fruits (g/day) |
87.1 (70.4) |
88.4 (67.9) |
78.9 (65.0) |
0.512 |
| Soybeans/soy products (g/day) |
53.7 (38.7) |
63.7 (38.9) |
66.0 (42.5) |
0.010 |
| Fish/seafood (g/day) |
56.2 (36.1) |
58.1 (28.3) |
68.9 (39.2) |
0.011 |
| Processed meat/meat (g/day) |
72.9 (40.6) |
80.3 (48.1) |
90.4 (60.1) |
0.054 |
| Dairy products (g/day) |
133.8 (92.5) |
118.0 (81.7) |
118.4 (94.9) |
0.128 |
| Sweets (g/day) |
62.6 (44.2) |
56.6 (39.4) |
52.7 (38.3) |
0.135 |
| MET (hour/week) |
16.5 (6.6–33.4) |
10.9 (4.9–29.7) |
11.0 (4.0–23.1) |
0.071 |
| Alcohol (kcal/day) |
18.1 (39.7) |
14.1 (43.0) |
20.5 (44.1) |
0.866 |
| Current smoker (%) |
17 (12) |
7 (9.3) |
7 (8.3) |
0.877 |
| Older Women (n = 277) |
Body mass index (kg/m2) |
24.3 (4.1) |
24.4 (4.1) |
25.4 (3.5) |
0.057 |
| Energy (kcal/day) |
1,556.6 (372.0) |
1,669.7 (316.5) |
1,795.9 (307.0) |
<0.001 |
| Na/K ratioa |
1.5 (0.4) |
1.5 (0.3) |
1.5 (0.4) |
0.887 |
| Fiber density (g/1,000 kcal/day) |
8.7 (1.7) |
7.5 (1.7) |
7.3 (1.8) |
<0.001 |
| Saturated fat (g/day) |
16.8 (6.0) |
16.4 (5.4) |
16.1 (5.0) |
0.413 |
| Carbohydrate/energy ratio (%) |
53.2 (4.7) |
56.1 (4.4) |
58.4 (4.2) |
<0.001 |
| Protein/energy ratio (%) |
16.1 (2.0) |
15.5 (1.7) |
14.8 (1.7) |
<0.001 |
| Fat/energy ratio (%) |
30.6 (3.9) |
28.3 (3.5) |
26.7 (3.4) |
<0.001 |
| Total vegetables (g/day) |
269.8 (114.5) |
251.5 (114.3) |
280.5 (134.6) |
0.793 |
| Fruits (g/day) |
111.2 (73.9) |
107.8 (70.0) |
110.8 (69.0) |
0.972 |
| Soybeans/soy products (g/day) |
67.1 (47.8) |
67.1 (37.4) |
67.3 (31.3) |
0.226 |
| Fish/seafood (g/day) |
78.3 (36.3) |
85.6 (43.4) |
89.7 (45.2) |
0.119 |
| Processed meat/meat (g/day) |
52.7 (30.8) |
55.5 (32.3) |
67.4 (43.8) |
0.031 |
| Dairy products (g/day) |
161.1 (98.2) |
151.6 (104.8) |
121.3 (85.4) |
0.009 |
| Sweets (g/day) |
47.5 (46.3) |
46.5 (36.5) |
47.2 (37.8) |
0.561 |
| MET (hour/week) |
12.7 (4.9–25.1) |
12.6 (4.4–29.0) |
11.7 (6.0–28.3) |
0.983 |
| Alcohol (kcal/day) |
10.7 (29.6) |
11.1 (34.0) |
11.2 (36.4) |
0.594 |
| Current smoker (%) |
7 (5.9) |
5 (5.3) |
2 (3.1) |
0.720 |
| Younger Men (n = 620) |
Body mass index (kg/m2) |
26.3 (4.1) |
26.4 (4.1) |
27.3 (5.1) |
0.075 |
| Energy (kcal/day) |
1,613.8 (449.0) |
1,772.5 (395.9) |
2,018.9 (453.4) |
<0.001 |
| Na/K ratioa |
1.6 (0.5) |
1.6 (0.4) |
1.5 (0.4) |
0.018 |
| Fiber density (g/1,000 kcal/day) |
6.5 (1.7) |
5.8 (1.4) |
5.6 (1.2) |
<0.001 |
| Saturated fat (g/day) |
17.7 (7.3) |
18.4 (6.7) |
19.3 (6.9) |
0.007 |
| Carbohydrate/energy ratio (g/day) |
52.6 (8.1) |
56.2 (5.9) |
57.7 (6.3) |
<0.001 |
| Protein/energy ratio (%) |
14.9 (2.7) |
14.1 (1.9) |
14.1 (2.0) |
0.002 |
| Fat/energy ratio (%) |
32.4 (6.2) |
29.6 (4.8) |
28.0 (5.1) |
<0.001 |
| Total vegetables (g/day) |
183.6 (109.5) |
181.2 (104.4) |
216.5 (105.2) |
<0.001 |
| Fruits (g/day) |
60.6 (67.6) |
55.6 (59.6) |
60.9 (60.7) |
0.614 |
| Soybeans/soy products (g/day) |
52.1 (41.1) |
50.1 (39.4) |
60.6 (43.9) |
0.012 |
| Fish/seafood (g/day) |
59.6 (46.0) |
62.4 (39.5) |
80.1 (52.1) |
<0.001 |
| Processed meat/meat (g/day) |
82.1 (53.5) |
88.9 (52.6) |
94.8 (56.1) |
0.008 |
| Dairy products (g/day) |
110.0 (83.2) |
122.7 (118.9) |
120.3 (106.4) |
0.880 |
| Sweets (g/day) |
54.7 (51.4) |
48.2 (34.4) |
51.3 (42.4) |
0.570 |
| MET (hour/week) |
16.0 (4.9–39.6) |
10.8 (3.3–24.0) |
10.8 (3.3–24.0) |
0.714 |
| Alcohol (kcal/day) |
89.4 (102.8) |
78.5 (107.9) |
67.4 (93.9) |
0.011 |
| Current smoker (%) |
69 (34.0) |
60 (35.3) |
72 (29.1) |
0.529 |
| Older Men (n = 367) |
Body mass index (kg/m2) |
24.4 (2.9) |
24.1 (3.1) |
24.1 (3.1) |
0.613 |
| Energy (kcal/day) |
1,665.8 (426.7) |
1,850.9 (388.4) |
1,978.5 (464.8) |
<0.001 |
| Na/K ratioa |
1.6 (0.4) |
1.7 (0.4) |
1.5 (0.4) |
0.157 |
| Fiber density (g/1,000 kcal/day) |
7.8 (2.2) |
6.8 (1.5) |
6.7 (1.8) |
<0.001 |
| Saturated fat (g/day) |
17.5 (6.4) |
17.7 (6.1) |
17.5 (6.8) |
0.873 |
| Carbohydrate/energy ratio (%) |
54.2 (6.3) |
57.7 (5.0) |
59.8 (5.6) |
<0.001 |
| Protein/energy ratio (%) |
15.4 (2.4) |
14.6 (1.9) |
14.3 (1.7) |
0.002 |
| Fat/energy ratio (%) |
30.3 (5.0) |
27.5 (4.1) |
25.8 (4.7) |
<0.001 |
| Total vegetables (g/day) |
239.9 (116.6) |
240.7 (127.5) |
264.3 (127.5) |
0.150 |
| Fruits (g/day) |
107.7 (68.7) |
102.8 (69.2) |
106.6 (83.5) |
0.489 |
| Soybeans/soy products (g/day) |
61.2 (45.0) |
65.2 (38.9) |
69.8 (46.3) |
0.065 |
| Fish/seafood (g/day) |
75.2 (47.8) |
84.9 (48.2) |
88.3 (45.8) |
0.017 |
| Processed meat/meat (g/day) |
62.3 (41.7) |
62.8 (43.2) |
70.5 (45.3) |
0.155 |
| Dairy products (g/day) |
156.5 (108.9) |
150.2 (100.4) |
133.5 (107.7) |
0.065 |
| Sweets (g/day) |
45.7 (42.0) |
49.4 (41.0) |
47.8 (45.6) |
0.778 |
| MET (hour/week) |
23.1 (11.5–39.8) |
28.4 (8.2–57.7) |
19.8 (8.2–46.5) |
0.686 |
| Alcohol (kcal/day) |
89.5 (96.3) |
80.1 (97.6) |
64.7 (95.2) |
0.012 |
| Current smoker (%) |
14 (10.8) |
17 (14.5) |
25 (20.8) |
0.274 |
Significant differences are highlighted in bold (p < 0.05).
MET, metabolic equivalent of task.
Data are presented as mean (±standard deviation) and n (%). Physical activity (metabolic equivalent of task [MET] hour/week) as median (interquartile range).
a Na/K ratio was calculated as sodium intake (mg/day) divided by potassium intake (mg/day).
Supplementary Table 1 presents the binary logistic regression analysis of associations between quartiles of rice intake and obesity for all participants, including age- and sex-stratified results. To mitigate the possibility that the statistically significant results from the tertile analysis are due to chance, we included a quartile analysis. We found a statistically significant positive association between the highest quartile of rice intake and obesity in older women (95% CI: 1.183–6.680, p trend = 0.003) in the model adjusted for age, energy intake, lifestyle factors, diabetes duration, and diabetes treatment status (Model 3). In Model 4, which additionally adjusted for fiber density, this association became non-significant, consistent with the tertile analysis findings. Conversely, in younger women, an inverse association emerged for the highest quartile after fiber density adjustment (95% CI: 0.182–0.918, p trend = 0.055).
Binary logistic regression of associations between rice intake and obesity (stratified by physical activity)
Supplementary Table 2 presents the binary logistic regression analysis of associations between tertiles of rice intake and obesity for all participants, stratified by median split of physical activity level as measured by MET. This subsequent analysis was conducted to test the hypothesis that physical activity could influence the association between rice intake and obesity. Compared to the lowest tertile of rice intake (T1), the highest tertile (T3) showed no significant association with obesity in both low (95% CI: 0.853–1.800, p trend = 0.532) and high physical activity groups (95% CI: 0.607–1.296, p trend = 0.762).
Discussion
In our analysis of 1,565 outpatients with type 2 diabetes, older women (>65 years) in the highest tertile of rice intake (mean 372.1 g/day) showed a higher prevalence of obesity. This association remained significant after adjusting for lifestyle factors but became non-significant after additional adjustment for fiber density. In younger women (≤65 years), an inverse association between the highest quartile of rice intake (mean 386.4 g/day) and obesity became evident after fiber density adjustment in the supplementary analysis. No associations were found in men.
Previous studies in Japan have indicated that excessive rice consumption can contribute to the onset of obesity [18], type 2 diabetes [19], and metabolic syndrome incidence [20]. However, while higher rice consumption was associated with metabolic syndrome among individuals aged 40 to 59 years, no such association was observed among those aged 60 to 74 years [20]. This was attributed to the higher physical activity levels and increased opportunities to undergo medical examinations and receive dietary advice in the older age group. In our study, the older women subgroup demonstrated the lowest levels of physical activity, which may increase their susceptibility to obesity. However, our stratified analysis of participants with low and high physical activity in the entire cohort showed no significant association between rice intake and obesity, suggesting a limited effect modification.
Rice intake was negatively associated with energy from protein, fiber density, and dairy products, suggesting poor overall diet quality and limited dietary diversity. In older women, the initially observed positive association between rice intake and obesity became non-significant after adjusting for fiber density. The dietary pattern associated with rice intake, marked by poor overall diet quality, rather than rice itself, may contribute to the increased obesity observed in older women. Conversely, in younger women, despite lower fiber with higher rice intake, increased fish and soy intake and reduced sweets were observed; rice was inversely associated with obesity after fiber density adjustment in the quartile analysis (Supplementary Table 1). The reversed association by age suggests that in younger women, rice intake aligns with a healthier traditional Japanese dietary pattern, while in older women, rice consumption may lack complementary nutritious foods, such as soybeans, fish, and vegetables.
Consistent with our findings, adopting a rice dietary pattern with limited consumption of other food groups was associated with obesity prevalence in Korean adults [21] and elderly Hispanic individuals in the US [10]. On the other hand, a 4-week randomized controlled trial showed that individuals adhering to a diet reflective of the 1975 Japanese dietary pattern, centered on rice and soup meals with greater food diversity, experienced significant reductions in body weight, body fat, and favorable alterations in lipid metabolic parameters compared to those following the contemporary Japanese diet (based on the 2015 Japan National Health and Nutrition Survey). Higher rice intake in a Chinese cohort was associated with mitigated weight gain and reduced hypertension incidence, although it was also positively associated with elevated fasting blood glucose levels [11]. Rice consumption was positively associated with the intake of zinc, sodium, animal foods, and vegetables in this cohort. Consuming carbohydrates at more than 70% of total caloric intake may be linked to a higher risk of type 2 diabetes, with Asian populations being particularly susceptible [22]. The increased risk in Asian populations is due to the relatively higher intake of carbohydrates from refined sources, including white rice and noodles [23], coupled with a lower capacity for insulin secretion compared to Western populations [24].
Recently, there has been growing interest in dietary energy density (DED), which measures the amount of energy contained in a given weight of food, as an indicator of overall diet quality. Studies have suggested a positive association between DED, weight gain, and metabolic syndrome [25-28]. In this group of outpatients with type 2 diabetes, we have previously shown significant inverse associations between DED and various food groups: vegetables not classified as green or yellow, fruits, green and yellow vegetables, steamed white rice, fish and seafood, potatoes, dairy products, beans, noodles, and eggs [29]. Lower energy dense foods are typically rich in essential nutrients, dietary fiber and/or water content, which can promote satiety, support weight management, and potentially improve metabolic outcomes. This suggests that these food groups could potentially be incorporated into dietary recommendations for managing obesity.
Sex disparities have emerged in chronic disease patterns, requiring their consideration in dietary research. A study in China demonstrated a continuous rise in the prevalence of metabolic syndrome with age among women, but not men [30]. This sex-specific pattern was attributed to the improved triglyceride profile observed in older male individuals, contributing to the stability of metabolic syndrome prevalence in that group. A meta-analysis on weight-loss outcomes reported there were no significant differences in weight loss between premenopausal and postmenopausal women; however, premenopausal women tended to lose more fat mass in diet-only interventions [31]. One plausible avenue for further exploration is to investigate the hormonal and metabolic changes that occur during the aging process in both men and women. Hormonal shifts, particularly during menopause in women, have been associated with alterations in lipid metabolism, insulin sensitivity, and body composition [32]. Unfortunately, our study did not collect data on menopausal status or hormonal therapy use, which may have influenced the associations observed in older women. Examining how hormonal changes interact with dietary and lifestyle factors could shed light on the divergent patterns observed in the prevalence of metabolic syndrome between sexes.
This study has several strengths, such as the inclusion of both men and women, a wide age range, and adjustments for lifestyle and dietary factors. The inclusion of participants from clinics located throughout Japan contributes to a more representative sample. However, certain limitations should be acknowledged: the sensitivity of the BMI for detecting obesity decreases with age, which may limit its accuracy in older adults. However, the ≥25 kg/m2 cutoff performed better than the ≥30 kg/m2 cutoff, with improved sensitivity and relatively high specificity [33]. Therefore, the use of the ≥25 kg/m2 cutoff in our study may have reduced misclassifications in older adults. The cross-sectional design of the study limits its capacity to determine causation between variables and outcomes. Reverse causality is an inherent issue in this design, and obesity may lead to changes in dietary behaviors, including rice intake, potentially confounding the observed associations. However, the loss of significance after fiber density adjustment may imply that dietary habits, rather than obesity-related changes in rice consumption, explain the association. The accuracy of nutrient intake calculations faces constraints due to the inherent limitations of the food frequency questionnaire, which do not provide absolute nutrient intake data. Additionally, sociodemographic variables such as education level and socioeconomic status were not available and the study did not account for the prevalence of chronic diseases or health-related behaviors, such as sleep duration and eating patterns (including late dinners and eating speed), known to be linked with obesity. Furthermore, although our data does not differentiate between white and brown rice, it is worth noting that within the Japanese population, white rice is the most commonly consumed variety, minimizing the effect of this limitation.
Our findings suggest that higher rice intake in Japanese outpatients with type 2 diabetes is associated with lower dietary quality, and that its relationship with obesity differs by age and sex (Graphical Abstract). Specifically, in older women, higher rice intake was initially associated with greater obesity risk, but the association disappeared after adjusting for fiber density. In younger women, an inverse association emerged after fiber density adjustment in a supplementary analysis. No significant associations were found in men. These results suggest that the association between rice intake and obesity is influenced by overall dietary quality rather than rice consumption alone. This highlights the importance of considering overall dietary patterns rather than single food groups when addressing obesity risk. Promoting greater dietary diversity while maintaining traditional staples like rice may be a strategy to improve diet quality in Japan. Prospective studies in Japanese and other populations are needed to confirm these associations.
Graphical Abstract
Acknowledgments
We thank the physicians, staff, and participants of JDDM for their generous contributions to this study.
Disclosure
The authors declare no potential conflicts of interest related to this research. Hirohito Sone discloses membership on the Editorial Board of Endocrine Journal.
Funding
This study was supported by the Japan Society for the Promotion of Science (22H03529).
Authors’ Contributions
Conceptualization: Efrem d’Avila Ferreira, Mariko Hatta and Hirohito Sone; Methodology: Efrem d’Avila Ferreira, Chika Horikawa, Mariko Hatta, Yasunaga Takeda, Sakiko Yoshizawa Morikawa and Hirohito Sone; Validation: Mariko Hatta and Hirohito Sone; Formal analysis: Efrem d’Avila Ferreira, Mariko Hatta, Laymon Khin, Izumi Ikeda and Hirohito Sone; Investigation: Efrem d’Avila Ferreira, Mariko Hatta, Yasunaga Takeda, Sakiko Yoshizawa Morikawa, Mizuki Takeuchi, Noriko Kato, Hiroshi Maegawa, Kazuya Fujihara and Hirohito Sone; Resources: Sakiko Yoshizawa Morikawa, Noriko Kato, Hiroshi Maegawa and Hirohito Sone; Data curation: Efrem d’Avila Ferreira, Mariko Hatta, Yasunaga Takeda, Sakiko Yoshizawa Morikawa, Mizuki Takeuchi, Noriko Kato, Hiroshi Maegawa, Kazuya Fujihara and Hirohito Sone; Writing—original draft: Efrem d’Avila Ferreira, Mariko Hatta, Laymon Khin, Izumi Ikeda, Kazuya Fujihara and Hirohito Sone; Writing—review and editing: Chika Horikawa, Yasunaga Takeda, Mizuki Takeuchi, Kazuya Fujihara and Hirohito Sone; Visualization: Chika Horikawa and Hirohito Sone; Supervision, Project administration and Funding acquisition: Hirohito Sone.
Ethical Standards Disclosure
This study was conducted according to the guidelines laid down in the Declaration of Helsinki and all procedures involving research study participants were approved by the Niigata University and Health Research Involving Human Subjects in Japan (approval number: 1927). Informed consent was obtained from all study participants.
References
- 1 Mohidem NA, Hashim N, Shamsudin R, Che Man H (2022) Rice for food security: revisiting its production, diversity, rice milling process and nutrient content. Agriculture (Switzerland) 12: 741.
- 2 Saleh ASM, Wang P, Wang N, Yang L, Xiao Z (2019) Brown rice versus white rice: nutritional quality, potential health benefits, development of food products, and preservation technologies. Compr Rev Food Sci Food Saf 18: 1070–1096.
- 3 Rathna Priya TS, Eliazer Nelson ARL, Ravichandran K, Antony U (2019) Nutritional and functional properties of coloured rice varieties of South India: A review. J Ethn Foods 6: 11.
- 4 Hurrell RF (2021) The potential of iodine and iron double-fortified salt compared with iron-fortified staple foods to increase population iron status. J Nutr 151: 47s–63s.
- 5 Saneei P, Larijani B, Esmaillzadeh A (2017) Rice consumption, incidence of chronic diseases and risk of mortality: meta-analysis of cohort studies. Public Health Nutr 20: 233–244.
- 6 Kim HJ, Park HA, Cho YG, Kang JH, Kim KW, et al. (2011) Gender difference in the level of HDL cholesterol in Korean adults. Korean J Fam Med 32: 173–181.
- 7 Ahn Y, Park SJ, Kwack HK, Kim MK, Ko KP, et al. (2013) Rice-eating pattern and the risk of metabolic syndrome especially waist circumference in Korean Genome and Epidemiology Study (KoGES). BMC Public Health 13: 61.
- 8 Ren G, Qi J, Zou Y (2021) Association between intake of white rice and incident type 2 diabetes—an updated meta-analysis. Diabetes Res Clin Pract 172: 108651.
- 9 Kolahdouzan M, Khosravi-Boroujeni H, Nikkar B, Zakizadeh E, Abedi B, et al. (2013) The association between dietary intake of white rice and central obesity in obese adults. ARYA Atheroscler 9: 140–144.
- 10 Lin H, Bermudez OI, Tucker KL (2003) Dietary patterns of Hispanic elders are associated with acculturation and obesity. J Nutr 133: 3651–3657.
- 11 Shi Z, Taylor AW, Hu G, Gill T, Wittert GA (2012) Rice intake, weight change and risk of the metabolic syndrome development among Chinese adults: the Jiangsu Nutrition Study (JIN). Asia Pac J Clin Nutr 21: 35–43.
- 12 Kobayashi M, Yamazaki K, Hirao K, Oishi M, Kanatsuka A, et al. (2006) The status of diabetes control and antidiabetic drug therapy in Japan—a cross-sectional survey of 17,000 patients with diabetes mellitus (JDDM 1). Diabetes Res Clin Pract 73: 198–204.
- 13 Takahashi K, Yoshimura Y, Kaimoto T, Kunii D, Komatsu T, et al. (2001) Validation of a food frequency questionnaire based on food groups for estimating individual nutrient intake. Jpn J Nutr Diet 59: 221–232.
- 14 Horikawa C, Yoshimura Y, Kamada C, Tanaka S, Tanaka S, et al. (2014) Dietary sodium intake and incidence of diabetes complications in Japanese patients with type 2 diabetes: analysis of the Japan Diabetes Complications Study (JDCS). J Clin Endocrinol Metab 99: 3635–3643.
- 15 Craig CL, Marshall AL, Sjöström M, Bauman AE, Booth ML, et al. (2003) International physical activity questionnaire: 12-country reliability and validity. Med Sci Sports Exerc 35: 1381–1395.
- 16 Murase N (2002) Validity and reliability of Japanese version of international physical activity questionnaire. J Health Welfare Stat 49: 1–9.
- 17 Examination Committee of Criteria for ‘Obesity Disease’ in Japan, Japan Society for the Study of Obesity (2002) New criteria for ‘obesity disease’ in Japan. Circ J 66: 987–992.
- 18 Sawada K, Takemi Y, Murayama N, Ishida H (2019) Relationship between rice consumption and body weight gain in Japanese workers: white versus brown rice/multigrain rice. Appl Physiol Nutr Metab 44: 528–532.
- 19 Nanri A, Mizoue T, Noda M, Takahashi Y, Kato M, et al. (2010) Rice intake and type 2 diabetes in Japanese men and women: the Japan Public Health Center-based Prospective Study. Am J Clin Nutr 92: 1468–1477.
- 20 Watanabe Y, Saito I, Asada Y, Kishida T, Matsuo T, et al. (2013) Daily rice intake strongly influences the incidence of metabolic syndrome in Japanese men aged 40–59 years. J Rural Med 8: 161–170.
- 21 Kim J, Jo I, Joung H (2012) A rice-based traditional dietary pattern is associated with obesity in Korean adults. J Acad Nutr Diet 112: 246–253.
- 22 Hosseini F, Jayedi A, Khan TA, Shab-Bidar S (2022) Dietary carbohydrate and the risk of type 2 diabetes: an updated systematic review and dose-response meta-analysis of prospective cohort studies. Sci Rep 12: 2491.
- 23 Lim CGY, Tai ES, van Dam RM (2022) Replacing dietary carbohydrates and refined grains with different alternatives and risk of cardiovascular diseases in a multi-ethnic Asian population. Am J Clin Nutr 115: 854–863.
- 24 Li L, Zou X, Huang Q, Han X, Zhou X, et al. (2021) Do east asians with normal glucose tolerance have worse β-cell function? A meta-analysis of epidemiological studies. Front Endocrinol (Lausanne) 12: 780557.
- 25 Stelmach-Mardas M, Rodacki T, Dobrowolska-Iwanek J, Brzozowska A, Walkowiak J, et al. (2016) Link between food energy density and body weight changes in obese adults. Nutrients 8: 229.
- 26 Mendoza JA, Drewnowski A, Christakis DA (2007) Dietary energy density is associated with obesity and the metabolic syndrome in U.S. adults. Diabetes Care 30: 974–979.
- 27 Du H, van der AD, Ginder V, Jebb SA, Forouhi NG, et al. (2009) Dietary energy density in relation to subsequent changes of weight and waist circumference in European men and women. PLoS One 4: e5339.
- 28 Pour-Abbasi MS, Nikrad N, Farhangi MA, Vahdat S, Jafarzadeh F (2023) Dietary energy density, metabolic parameters, and blood pressure in a sample of adults with obesity. BMC Endocr Disord 23: 3.
- 29 Takeda Y, Fujihara K, Nedachi R, Ikeda I, Morikawa SY, et al. (2021) Comparing associations of dietary energy density and energy intake, macronutrients with obesity in patients with type 2 diabetes (JDDM 63). Nutrients 13: 3167.
- 30 Jiang B, Zheng Y, Chen Y, Chen Y, Li Q, et al. (2018) Age and gender-specific distribution of metabolic syndrome components in East China: role of hypertriglyceridemia in the SPECT-China study. Lipids Health Dis 17: 92.
- 31 Thomson ZO, Kelly JT, Sainsbury A, Reeves MM (2020) Weight loss outcomes in premenopausal versus postmenopausal women during behavioral weight loss interventions: a systematic review and meta-analysis. Menopause 28: 337–346.
- 32 Kodoth V, Scaccia S, Aggarwal B (2022) Adverse changes in body composition during the menopausal transition and relation to cardiovascular risk: a contemporary review. Womens Health Rep (New Rochelle) 3: 573–581.
- 33 Batsis JA, Mackenzie TA, Bartels SJ, Sahakyan KR, Somers VK, et al. (2016) Diagnostic accuracy of body mass index to identify obesity in older adults: NHANES 1999–2004. Int J Obes (Lond) 40: 761–767.
