Hostname: page-component-788cddb947-m6qld Total loading time: 0 Render date: 2024-10-15T09:23:11.959Z Has data issue: false hasContentIssue false

Rice and noodle consumption is associated with insulin resistance and hyperglycaemia in an Asian population

Published online by Cambridge University Press:  14 November 2013

Yasmin L. M. Zuñiga
Affiliation:
Division of Endocrinology, Department of Medicine, National University Hospital, 1E Kent Ridge Road, NUHS Tower Block Level 10, Singapore119228, Singapore
Salome A. Rebello*
Affiliation:
Saw Swee Hock School of Public Health, National University of Singapore, National University Health System, Block MD3, #03-17, 16 Medical Drive, Singapore117597, Singapore
Puay Leng Oi
Affiliation:
Saw Swee Hock School of Public Health, National University of Singapore, National University Health System, Block MD3, #03-17, 16 Medical Drive, Singapore117597, Singapore
Huili Zheng
Affiliation:
Saw Swee Hock School of Public Health, National University of Singapore, National University Health System, Block MD3, #03-17, 16 Medical Drive, Singapore117597, Singapore
Jeannette Lee
Affiliation:
Saw Swee Hock School of Public Health, National University of Singapore, National University Health System, Block MD3, #03-17, 16 Medical Drive, Singapore117597, Singapore
E. Shyong Tai
Affiliation:
Department of Medicine, Yong Loo Lin School of Medicine, National University of Singapore, National University Health System, 1E Kent Ridge Road, NUHS Tower Block Level 10, Singapore119228, Singapore
Rob M. Van Dam
Affiliation:
Saw Swee Hock School of Public Health, National University of Singapore, National University Health System, Block MD3, #03-17, 16 Medical Drive, Singapore117597, Singapore Department of Medicine, Yong Loo Lin School of Medicine, National University of Singapore, National University Health System, 1E Kent Ridge Road, NUHS Tower Block Level 10, Singapore119228, Singapore Department of Nutrition, Harvard School of Public Health, Boston, MA02115, USA
*
*Corresponding author: S. A. Rebello, fax +65 67791489, email ephsar@nus.edu.sg
Rights & Permissions [Opens in a new window]

Abstract

High consumption of refined grains, particularly white rice, has been reported to be associated with a higher risk of type 2 diabetes. Therefore, in the present study, we evaluated the association between rice and noodle consumption and markers of glucose homeostasis, inflammation and dyslipidaemia in an Asian population. We carried out a population-based cross-sectional study in 2728 Singaporean Chinese men and women aged between 24 and 92 years. Rice and noodle intake was assessed using a validated FFQ and studied in relation to glycaemic (fasting glucose, glycated Hb, homeostasis model assessment (HOMA) index for insulin resistance (HOMA-IR) and HOMA index for β-cell function (HOMA-β)), inflammatory (plasma adiponectin and C-reactive protein (CRP)) and lipid (fasting TAG and HDL-cholesterol (HDL-C)) markers. We used multiple linear regression analyses with adjustment for total energy intake and sociodemographic, anthropometric (BMI and waist:hip ratio) and lifestyle factors. Higher rice consumption was found to be associated with higher fasting glucose concentrations (0·81 % higher values per portion increment; 95 % CI 0·09, 1·54) and HOMA-IR (4·62 %; 95 % CI 1·29, 8·07). Higher noodle consumption was also found to be significantly associated with higher fasting glucose concentrations (1·67 %; 95 % CI 0·44, 2·92), HOMA-IR (6·17 %; 95 % CI 0·49, 12·16) and fasting TAG concentrations (9·17 %; 95 % CI 3·44, 15·22). No significant association was observed between rice and noodle consumption and adiponectin, CRP and HDL-C concentrations or HOMA-β in the fully adjusted model. These results suggest that high consumption of rice and noodles may contribute to hyperglycaemia through greater insulin resistance and that this relationship is independent of adiposity and systemic inflammation.

Type
Full Papers
Copyright
Copyright © The Authors 2013 

More than half of all individuals with diabetes mellitus in the world reside in Asia, and a large increase in the number of diabetes cases has been predicted for this continent( Reference Wild, Roglic and Green 1 ). Grains in the form of rice and noodles are the primary carbohydrate source for most Asian populations, with rice providing as much as 60 % of total energy in Southeast Asia( 2 , Reference Ahn, Park and Kwack 3 ). Rice and other grains are predominantly consumed as refined grains by most Asian populations( Reference Spiller 4 , Reference Misra, Rastogi and Joshi 5 ). Accounts as to when polished rice started being consumed in Asia are limited; in Japan, consumption of polished white rice began at the end of the seventeenth century and was initially limited to the urban upper class due to its high price( Reference Ohnuki-Tierney 6 ). Currently, mechanised steel roller mills and automated sifting devices are being used to efficiently refine grains, resulting in a huge loss of vitamins, minerals, essential fatty acids and phytochemicals( Reference Spiller 4 , Reference Slavin, Martini and Jacobs 7 , Reference Slavin, Jacobs and Marquart 8 ).

Recently published data suggest that there is an association between high white rice consumption and a higher risk of type 2 diabetes in both Western( Reference Sun, Spiegelman and van Dam 9 ) and Asian( Reference Villegas, Liu and Gao 10 Reference Hu, Pan and Malik 12 ) populations. However, the biological mechanisms that may underlie this association are unclear. In addition, data on the association between consumption of noodles, another major source of refined grains in Asia, and risk of type 2 diabetes are sparse( Reference Nanri, Mizoue and Noda 11 ). High consumption of white rice and noodles may lead to high dietary glycaemic index( Reference Jenkins, Wolever and Taylor 13 ) and glycaemic load, which have been suggested to increase the risk of type 2 diabetes through excess postprandial variations in blood glucose and insulin concentrations( Reference Salmeron, Ascherio and Rimm 14 Reference Sluijs, van der Schouw and van der 19 ). Such variations in insulin concentrations may increase circulating NEFA concentrations( Reference Ludwig 20 ) and reduce the number of insulin receptors, both of which can contribute to insulin resistance( Reference Esfahani, Wong and Mirrahimi 21 ). Furthermore, in contrast to whole grains, refined grains contain less amounts of fibre and phytochemicals that may lower the risk of type 2 diabetes( Reference Nanri, Mizoue and Noda 11 , Reference Radhika, Van Dam and Sudha 18 , Reference Sluijs, van der Schouw and van der 19 , Reference Liu, Manson and Stampfer 22 Reference Fung, Hu and Pereira 24 ). Studies on the association between refined grain consumption and metabolic risk factors can provide insights into the potential mechanisms that may contribute to a higher risk of type 2 diabetes. However, few of these studies have been carried out in Asian populations in which rice and noodles are the main carbohydrate source( Reference Woo, Ho and Sham 25 Reference Zhang, Pan and Zong 28 ). Insulin resistance and impaired β-cell function are key contributors to the development of type 2 diabetes, and chronic inflammation and dyslipidaemia can contribute to these conditions( Reference Li, Shin and Ding 29 , Reference Mineo and Shaul 30 ). To better understand the mechanisms by which rice or noodle intake may be involved in the pathogenesis of type 2 diabetes and its impact on the metabolic traits associated with type 2 diabetes, we evaluated the relationship between rice and noodle consumption and markers of insulin resistance, insulin secretion, inflammation and dyslipidaemia in a cross-sectional study in Singaporean Chinese.

Methods

Study population

The present study used cross-sectional data of the Singapore Prospective Study Program-2 (SP2) cohort. These data were collected from 2004 to 2007. SP2 was a follow-up study of the participants of four population-based studies carried out in Singapore during 1982–8: the Thyroid and Heart Study; the National Health Survey (1992); the National University of Singapore Heart Study; the National Health Survey (1998). These studies selected participants by stratified random sampling of individuals aged 18–69 years with oversampling of ethnic minority groups (i.e. Malays and Asian-Indians)( Reference Yim-Lui Cheung, Wong and Lamoureux 31 ). The present study was conducted according to the guidelines laid down in the Declaration of Helsinki, and all procedures involving human subjects were approved by the National University of Singapore and the Singapore General Hospital Institutional Review Boards. Written informed consent was obtained from all the participants.

For SP2, 10 747 study subjects qualified for participation, and 7744 completed interviewer-administered questionnaires on demographics, lifestyle and medical history. Of these participants, 5163 made a clinic visit during which blood was drawn and anthropometric parameters and blood pressure were measured. Details on participant recruitment and study methodology have been reported elsewhere( Reference Nang, Khoo and Tai 32 ). As the type of refined grain foods consumed by Indians and Malays can be substantially different and may be less comprehensively represented in our questionnaire, we focused only on the Chinese subpopulation in the present study. Of the 5163 participants who visited the clinic, 3439 were Chinese. From this subpopulation, we excluded subjects with pre-existing CVD (n 109), diabetes diagnosis or known diabetes (n 232), current cancer (n 47) or a current pregnancy (n 2), as these conditions may affect diet or reporting of dietary intake potentially leading to reverse causation. From the remaining 3088 participants, we also excluded individuals who changed their diet in the month preceding the interview (n 203) and those with extremes of energy intake (>29 288 or < 2092 kJ (>7000 or < 500 kcal)) and extreme energy intake based on the ratio of energy intake:energy expenditure (lowest or highest 2·5 percentage of the ratio, n 138)( Reference Nang, Gitau Ngunjiri and Wu 33 ). From the remaining cohort comprising 2758 participants, we further excluded individuals with missing covariates (n 30). As a result, 2728 persons remained for the analysis.

Assessment of dietary intakes

A semi-quantitative 169-item FFQ developed by the Institute of Health, Singapore, was used to assess the dietary intake of participants during the month preceding the interview. Food items were selected if they were significant contributors to the intake of energy and selected nutrients in adult Singaporeans based on the National Nutrition Survey. The questionnaire has been validated previously( Reference Deurenberg-Yap, Li and Tan 34 ). The participants were asked to estimate the frequency of consumption of each food group based on a reference portion size and to indicate consumption on a per-d, per-week or per-month basis or as never/rarely. Daily energy and nutrient intakes were subsequently calculated by the Health Promotion Board based on an in-house database of energy and nutrient values of local foods.

Data on rice intake were obtained by asking the participants about the frequency of consumption of a standard bowl or portion of rice. Household measures such as bowls of different sizes or visual aids for various food groups and individual food items were used to help the participants estimate amounts consumed as fractions or multiples of the illustrated reference portions. In total, ten rice dishes and thirteen noodle dishes were included in the questionnaire. Options for the type of rice consumed included plain rice or plain rice porridge, flavoured rice (e.g. fried rice and chicken rice) and flavoured rice porridge. For noodle intake (mainly rice noodles, wheat noodles, bean noodles or pasta), the questionnaire assessed the consumption of different types of noodles in soup, dry noodles, fried noodles, noodles in lemak gravy (with coconut milk) and other noodles (including instant noodles). One portion of cooked white rice weighs approximately 200 g and one portion of cooked noodles weighs approximately 275 g.

Assessment of outcome variables

Fasting venous blood samples were sent to the National University Hospital Reference Laboratory for analysis on the day the blood was drawn for biochemical testing of fasting plasma glucose (FPG; ADVIA 2400, Siemens), glycated Hb (Biorad Variant II analyser, Bio-Rad Laboratories), HDL-cholesterol (HDL-C) and TAG. Fasting serum samples were analysed for total cholesterol, TAG and HDL-C (Siemens Medical Solutions Diagnostics), high-sensitivity C-reactive protein (CRP, Roche Diagnostics), total adiponectin (Sekisui Medical Company Limited) and insulin (microparticle enzyme immunoassay, Abbott AXSYM, Abbott Laboratories). LDL-cholesterol concentrations were calculated using the Friedewald equation. The respective intra-batch and inter-batch CV for the variables were as follows: 0·93–1·15 and 0·56–0·65 % for TAG; 0–3·85 and 1·18–2·00 % for HDL-C; 1·27–3·40 and 2·50–6·60 % for glucose; 4·00–4·50 % for insulin; 0–2·00 and 0·85–1·54 % for glycated Hb, 0·60–1·30 and 2·30–3·10 % for CRP; 18·10 and 15·90 % for adiponectin. Homeostasis model assessment (HOMA) indices were used as measures of insulin resistance and β-cell function. The HOMA insulin resistance (HOMA-IR) index was computed as FPG (mmol/l) × fasting serum insulin (mU/l)/22·5, where for insulin 1 mU = 6·00 pmol. The HOMA-β-cell (HOMA-β) index was computed as 20 × fasting serum insulin (mU/l)/FPG (mmol/l) − 3·5( Reference Matthews, Hosker and Rudenski 35 ).

Assessment of covariates

Height was measured using a wall-mounted measuring tape and weight using a digital scale. BMI was computed as weight (kg) divided by height (m2). Data on alcohol and coffee intake were obtained using the FFQ and those on cigarette smoking status, education level and physical activity the main questionnaire. Total physical activity, expressed as metabolic equivalents of task-h/week, was assessed using a validated questionnaire based on activities in four domains (household, occupational, leisure-time and transport)( Reference Nang, Gitau Ngunjiri and Wu 33 ). Data on history of hypertension or dyslipidaemia were obtained from self-reports of physician diagnosis or use of medications to treat these conditions.

Statistical analyses

The participants were classified into quintiles of rice and noodle consumption, and we compared the characteristics of the participants across the quintiles using ANOVA (for continuous variables with a normal distribution), Kruskal–Wallis tests (for continuous variables with a non-normal distribution) or χ2 tests (for categorical variables). Rice and noodle consumption is expressed as portions per 8368 kJ (2000 kcal). All the response variables were transformed using natural logarithms to achieve normality. Geometric means and 95 % CI were obtained by exponentiation of means and 95 % CI of values on the logarithmic scale. Sensitivity analyses were carried out, in which we truncated response variables that were more than 4 sd from the mean to reduce the potential impact of outliers. As this did not materially affect the results, we report results for data without truncation.

We used multiple linear regression analyses to study rice and noodle consumption in relation to biological risk factors. Sociodemographic factors and type 2 diabetes risk factors that were identified from the literature were adjusted for in the analyses as potential confounders. Variables were incorporated into four multivariable models as follows: (1) adjusted for age (years), sex and total energy intake (kJ/d); (2) further adjusted for physical activity (metabolic equivalents of task-h/week), alcohol intake (non-drinker, < 1 serving/d, ≥ 1 serving/d), cigarette smoking status (non-smoker, ex-smoker, current smoker < 10 cigarettes/d, current smoker ≥ 10 cigarettes/d), education level (primary/below, secondary, polytechnic/diploma, university), dyslipidaemia diagnosis or known dyslipidaemia (yes/no) and hypertension diagnosis or known hypertension (yes/no); (3) further adjusted for BMI (kg/m2) and waist:hip ratio; (4) further adjusted for dietary confounders, specifically consumption of coffee (never/rarely, < 1 cup daily, 1–2 cups daily, ≥ 3 cups daily), protein (% energy), wholemeal bread (0, 0·1 to < 1, ≥ 1 serving daily), white bread (0, 0·1 to < 1, ≥ 1 serving daily), fruits (servings per 8368 kJ (2000 kcal)), vegetables (servings per 8368 kJ (2000 kcal)), cholesterol (mg/8368 kJ (2000 kcal)), ratio of PUFA:SFA and rice (portions per 8368 kJ (2000 kcal)) or noodles (portions per 8368 kJ (2000 kcal)) when appropriate. Rice and noodle intakes were fitted as quintiles; bread and coffee intakes were fitted in the categorical scale, while the other dietary variables were modelled in the continuous scale. Analyses with HOMA-β as the outcome variable were further adjusted for HOMA-IR. Participants with diagnosed dyslipidaemia were excluded from the analyses with HDL-C or TAG as the outcome variable (remaining sample size n 1978). We also carried out analyses modelling rice and noodle intakes as continuous variables. The regression coefficients are expressed as percentage changes in the outcome variables for each portion increment in rice or noodle intake( Reference Lopez-Garcia, van Dam and Qi 36 ).

Secondary analyses were carried out to evaluate whether associations were similar for men and women and for overweight (BMI ≥ 23 kg/m2) and non-overweight (BMI < 23 kg/m2) individuals using a commonly accepted cut-off value for overweight in Asians( Reference Expert WHO Consultation 37 ). A multiplicative interaction term with sex or overweight as a dichotomous variable and rice or noodle intake as a continuous variable was included in the multivariable models. We observed no significant interaction by sex and therefore present the results for men and women together. All the data were analysed using the Statistical Analysis System version 9.2 software (SAS Institute, Inc.). Two-sided P values < 0·05 were considered statistically significant.

Results

The median intake of rice was 1·57 portions/d (1·75 portions/8368 kJ (2000 kcal)) and the median intake of noodles was 0·57 portions/d (0·63 daily portions/8368 kJ (2000 kcal)). Rice and noodle consumption explained 64·6 % of starch intake in the study population independent of total energy intake. The characteristics of the study population according to rice and noodle consumption are given in Table 1. The study participants were aged between 24·6 and 91·8 years (mean 48·7 (sd 11·5) years). Participants with higher rice intake tended to be older and male, have a low education level, smoke, be less physically active and have hypertension diagnosis or known hypertension. Higher rice consumption was associated with higher intakes of carbohydrate and starch, but with lower intakes of protein, fat, mono- and disaccharides and noodles. Participants with higher noodle intake tended to be younger and be less physically active. Higher noodle consumption was associated with higher intakes of protein and starch, but with lower intakes of carbohydrates, mono- and disaccharides, and fibre and a lower ratio of PUFA:SFA.

Table 1 Characteristics of the study participants according to quintiles of rice and noodle consumption (Mean values and standard deviations; median values, interquartile ranges and percentages)

WHR, waist:hip ratio; MET-h/week, metabolic equivalents of task-h/week.

* Expressed in portions per 8368 kJ (2000 kcal).

Table 2 summarises the concentrations of metabolic markers according to quintiles of rice consumption. Higher rice consumption was not significantly associated with fasting glucose concentrations and HOMA-IR in the basic model adjusted for age, sex and energy intake. However, after further adjustment for potential confounders, particularly BMI and protein and noodle intake, higher rice consumption was strongly associated with higher fasting glucose concentrations and HOMA-IR. Rice consumption was not significantly associated with glycated Hb, CRP, adiponectin, HDL-C, and fasting TAG concentrations and HOMA-β.

Table 2 Metabolic risk factors according to quintiles of rice consumption (Geometric means and 95 % confidence intervals; median values and interquartile ranges)

FPG, fasting plasma glucose; HbA1c, glycated Hb; HOMA-IR, homeostasis model assessment of insulin resistance; HOMA-β, homeostasis model assessment of β-cell function; CRP, C-reactive protein.

* Percentage changes (and 95 % CI) in plasma concentrations of metabolic risk factors for each portion increment of rice, approximately 200 g of cooked rice.

From multiple linear regression models for the relationship between rice consumption and log-transformed markers.

Adjusted for age, sex and energy intake. HOMA-β models were further adjusted for HOMA-IR.

§ Further adjusted for physical activity, alcohol intake, smoking status, education level, diagnosed dyslipidaemia and diagnosed hypertension.

Further adjusted for BMI and waist:hip ratio.

Further adjusted for coffee consumption and intake of protein, noodles, wholemeal bread, white bread, fruits, vegetables, cholesterol and ratio of PUFA:SFA.

** Participants with diagnosed dyslipidaemia were excluded from HDL-cholesterol and TAG analyses (remaining sample size n 1978, whereby n 1 411, n 2 395, n 3 390, n 4 397 and n 5 385 for quintiles 1–5, respectively).

Higher noodle consumption was associated with higher fasting glucose, glycated Hb, CRP, and fasting TAG concentrations and HOMA-IR, but with lower adiponectin and HDL-C concentrations in the basic model (Table 3). After further adjustment for additional potential confounders, noodle consumption remained significantly directly associated with fasting glucose and TAG concentrations and HOMA-IR. Associations between noodle consumption and other biomarkers were not statistically significant in the fully adjusted model.

Table 3 Metabolic risk factors according to quintiles of noodle consumption (Geometric means and 95 % confidence intervals; median values and interquartile ranges)

FPG, fasting plasma glucose; HbA1c, glycated Hb; HOMA-IR, homeostasis model assessment of insulin resistance; HOMA-β, homeostasis model assessment of β-cell function; CRP, C-reactive protein.

* Percentage changes (and 95 % CI) in plasma concentrations of metabolic risk factors for each portion increment of noodles, approximately 275 g of cooked noodles.

From multiple linear regression models for the relationship between noodle consumption and log-transformed markers.

Adjusted for age, sex and energy intake. HOMA-β models were further adjusted for HOMA-IR.

§ Further adjusted for physical activity, alcohol intake, smoking status, education level, diagnosed dyslipidaemia and diagnosed hypertension.

Further adjusted for BMI and waist:hip ratio.

Further adjusted for coffee consumption and intake of protein, rice, wholemeal bread, white bread, fruits, vegetables, cholesterol and ratio of PUFA:SFA.

** Participants with diagnosed dyslipidaemia were excluded from HDL-cholesterol and TAG analyses (remaining sample size n 1978, whereby n 1 373, n 2 407, n 3 405, n 4 405 and n 5 388 for quintiles 1–5, respectively).

In secondary analyses, we examined the odds of having hyperglycaemia using the American Diabetes Association FPG cut-off ≥ 5·6 mmol/l( 38 ) across the tertiles of rice and noodle consumption (Supplementary Table 1, available online). In the fully adjusted model, we observed higher odds of having high FPG concentrations with higher noodle consumption (OR per portion increment: 1·55; 95 % CI 1·09, 2·20; P trend = 0·01), but not with higher rice consumption (0·99; 95 % CI 0·79, 1·22; P trend = 0·89).

We also examined the association between total refined grain consumption (the sum of rice and noodle intakes) and metabolic risk factors that were significantly associated with either rice or noodle consumption. Each portion increment in total refined grain consumption was associated with higher FPG concentrations (percentage change 0·92; 95 % CI 0·22, 1·62; P trend = 0·01), higher fasting TAG concentrations (3·42; 95 % CI 0·33, 6·60; P trend = 0·03) and a higher insulin resistance index (4·96; 95 % CI 1·75, 8·27; P trend = 0·002).

It has been suggested that the association between refined grain consumption and markers of hyperglycaemia may be more pronounced among overweight individuals( Reference Jeppesen, Schaaf and Jones 39 , Reference Willett, Manson and Liu 40 ). Therefore, we evaluated the possible interaction of rice and noodle consumption with overweight status (BMI ≥ 23 kg/m2) in relation to metabolic risk factors for which we observed significant associations in the main analysis (Supplementary Tables 2 and 3, available online). The direct association between noodle consumption and HOMA-IR was stronger in the overweight group (change per portion 13·15 %, 95 % CI 3·02, 24·29) than in the leaner group (1·49 %, 95 % CI − 5·02, 8·45; P interaction = 0·001). However, the direct association between rice consumption and HOMA-IR tended to be weaker in the overweight participants (1·88 %, 95 % CI − 3·69, 7·78) than in the leaner participants (6·87 %, 95 % CI 2·80, 11·10, P interaction = 0·33).

Discussion

In the present population-based study of 2728 Singaporean Chinese adult men and women, high consumption of rice and noodles was found to be associated with higher fasting glucose concentrations and a higher HOMA index for insulin resistance. In contrast, rice and noodle consumption was found to be not associated with adiponectin or CRP concentrations after adjustment for potential confounders. The present results thus suggest that high consumption of refined grains may contribute to hyperglycaemia through greater insulin resistance, rather than through increased systemic inflammation.

The results obtained for rice consumption in the present study are consistent with the findings of several other studies carried out in Asian populations in which white rice intake is high. Higher consumption of rice was found to be significantly associated with a higher risk of type 2 diabetes in a cohort of older Shanghai women (relative risk 1·78; 95 % CI 1·48, 2·15 for ≥ 300 v. < 200 g/d)( Reference Villegas, Liu and Gao 10 ). In a Japanese cohort, higher rice consumption was also reported to be associated with a higher risk of type 2 diabetes in women (relative risk 1·65; 95 % CI 1·06, 2·57 for ≥ 420 v. < 200 g/d), but not in men( Reference Nanri, Mizoue and Noda 11 ). In a cross-sectional study carried out in India, high refined grain consumption (of which white rice comprised 75·8 %) was found to be associated with a higher prevalence of newly diagnosed type 2 diabetes( Reference Mohan, Radhika and Sathya 27 ). In the same study, higher refined grain consumption was found to be associated with significantly 7·9 % higher fasting glucose concentrations, 13·6 % higher HOMA-IR values, 36·5 % higher serum TAG concentrations and 10·1 % lower HDL-C concentrations for the highest (median 449 g/d) v. the lowest (median 218 g/d) quartile( Reference Radhika, Van Dam and Sudha 18 ). In a cross-sectional study in Japanese female farmers, white rice was found to be the major contributor to the dietary glycaemic load, which is associated with higher fasting TAG and glucose concentrations and lower HDL-C concentrations( Reference Murakami, Sasaki and Takahashi 41 ).

Few studies on noodle consumption in relation to metabolic risk factors or risk of type 2 diabetes have been carried out. In a study in Hong Kong Chinese, higher consumption of rice, noodles and pasta was found to be associated with a significantly higher prevalence of newly diagnosed type 2 diabetes. However, specific results for noodles were not reported( Reference Woo, Ho and Sham 25 ). In a Japanese cohort, consumption of noodles was found to be not associated with the risk of type 2 diabetes( Reference Nanri, Mizoue and Noda 11 ). It is possible that differences in the type of noodles consumed (e.g. whole-grain buckwheat noodles are commonly consumed in Japan but not in Singapore) are responsible for these differences in association. In the present study, we observed stronger associations between noodle consumption and metabolic risk markers than between rice consumption and the markers. In Singapore, both rice and wheat-based noodles typically prepared from refined grains are commonly consumed. Also, noodle-based dishes are often prepared with lard and typically contain high amounts of Na and cholesterol. It is possible that addition of these ingredients to noodles during their preparation exacerbates the adverse effects of refined grain on metabolic outcomes. In a cohort of older Singaporean-Chinese, noodle consumption was found to be part of a dietary pattern comprising other less healthy food choices including red meats and deep-fried foods, which is associated with an increased risk of type 2 diabetes( Reference Odegaard, Koh and Butler 42 ). In the present study, noodle consumption was found to be correlated with a higher intake of cholesterol, a lower intake of total fibre and vegetables and a lower PUFA:SFA ratio, suggesting that noodle consumption may serve as a marker of unhealthy dietary choices in this population. In the present study, rice consumption was found to be associated with a lower consumption of vegetables, but not with the consumption of other unfavourable food items. However, associations of both rice and noodle consumption with metabolic risk factors remained after adjustment for dietary risk factors.

It has been postulated that the adverse metabolic effects of high carbohydrate intakes on glucose and lipid metabolism are stronger in insulin-resistant individuals than in insulin-sensitive individuals( Reference Murakami, Sasaki and Takahashi 41 ) and may thus be stronger in overweight individuals than in lean individuals( Reference Jeppesen, Schaaf and Jones 39 ). In Asian populations, however, both stronger( Reference Villegas, Liu and Gao 10 ) and weaker( Reference Nanri, Mizoue and Noda 11 , Reference Sakurai, Nakamura and Miura 43 ) associations for glycaemic index, glycaemic load and high rice consumption have been observed for overweight individuals than for leaner individuals. In the present study, the association between rice consumption and HOMA-IR tended to be more pronounced among non-overweight participants, whereas the association between noodle consumption and HOMA-IR was more pronounced among overweight participants. Taken together, results regarding possible effect modification of the association between refined grain consumption and metabolic health outcomes by overweight status have not been consistent.

The specific biological mechanisms by which increased consumption of refined grains such as white rice and refined grain noodles may contribute to insulin resistance remain to be elucidated. Metabolic studies have shown that both the glycaemic index of foods and the amount of carbohydrates consumed contribute to the glycaemic response to a meal( Reference Willett, Manson and Liu 40 ). The glycaemic load reflects both the quality and quantity of carbohydrate intake. Commonly consumed rice varieties in this population such as Jasmine rice and glutinous rice typically have high glycaemic index values( Reference Atkinson, Foster-Powell and Brand-Miller 44 ). Data on the glycaemic index values of Asian types of noodles are limited. Published values are usually in the intermediate range (40–66 units, using glucose as a standard), although some types of noodles such as fresh wheat noodles have a high glycaemic index of 82( Reference Atkinson, Foster-Powell and Brand-Miller 44 , Reference Lok, Chan and Chan 45 ). Given the high carbohydrate content of both rice and noodles, their consumption can contribute to a higher dietary glycaemic load, which has been reported to be associated with a higher risk of type 2 diabetes in several( Reference Villegas, Liu and Gao 10 , Reference Salmeron, Ascherio and Rimm 14 Reference Liu 16 ) but not in all( Reference Meyer, Kushi and Jacobs 23 ) studies. Consumption of foods with a high glycaemic load leads to high postprandial glucose and insulin concentrations. The resulting peak in postprandial glucose concentrations is typically followed by a rapid decline in glucose concentrations, which may trigger the secretion of counter-regulatory hormones and reduce the suppression of circulating NEFA concentrations. Chronic elevations in postprandial concentrations of insulin, counter-regulatory hormones and NEFA may contribute to insulin resistance( Reference Ludwig 20 ).

Another possible explanation for the association between rice and noodle consumption and hyperglycaemia lies in the processing of the refined grains. This process removes most of the bran and some of the germ, resulting in the loss of various components that may be beneficial for glucose metabolism such as minerals, lignans and phenolic compounds( Reference Slavin, Martini and Jacobs 7 ). Although we controlled for other dietary risk factors in the multivariable models in the present study, the possibility remains that high rice and noodle consumption contributes to metabolic disturbances by replacing foods that have a beneficial effect on glucose metabolism.

The strengths of the present study include the population-based selection of participants, the reasonably large sample size, and the detailed information on dietary intakes and potential confounders based on validated questionnaires. However, the study also has several limitations. First, the cross-sectional nature of the study did not allow us to determine the direction of effects. However, we excluded individuals with known diabetes or CVD. It seems unlikely that the remaining participants were aware of their glycaemic blood markers and that the concentrations of these biomarkers affected food choices. Second, measurement error in the assessment of dietary intakes was inevitable. However, this is unlikely to explain the observed associations in the present study, as this would have weakened rather than strengthened the associations. We did not distinguish between whole-grain noodles and brown rice in our questionnaire. However, in the Singaporean population, as in most Asian populations( Reference Spiller 4 , Reference Misra, Rastogi and Joshi 5 ), consumption of brown rice and whole-grain noodles was rare at the time of the study. Our findings for rice consumption and metabolic risk factors can, therefore, be assumed to reflect associations for high white rice and refined grain noodle consumption. Finally, although we considered a wide range of potential confounders, residual confounding by unmeasured or imperfectly measured confounders may still have affected the results of the study.

In several cohort studies, high white rice consumption has been reported to be associated with a higher risk of type 2 diabetes. The results of the present study suggest that this association may be mediated by detrimental effects of high rice consumption on insulin sensitivity, rather than by the effects on systemic inflammation. We observed that high consumption of noodles may be at least as detrimental for metabolic risk factors as high consumption of white rice. Previous studies in Asian population have mainly focused on white rice, but attention on noodles, which are frequently consumed by many East Asian and South East Asian populations, is also warranted. If these detrimental effects of high white rice and noodle consumption on glucose metabolism are confirmed in further epidemiological studies and randomised trials, it could have major implications for Asian populations that have very high refined grain intakes and a rapidly increasing burden of type 2 diabetes( Reference Wild, Roglic and Green 1 ).

Supplementary material

To view supplementary material for this article, please visit http://dx.doi.org/10.1017/S0007114513003486

Acknowledgements

The authors gratefully acknowledge the Health Promotion Board, Singapore, for use of their FFQ and database for dietary intake analyses.

The present study was supported by grants from the Biomedical Research Council (grant 03/1/27/18/216) and National Medical Research Council (grants 0838/2004, 1111/2007 and CSI/0002/2005), Singapore. The funders had no role in the design and analysis of the study or in the writing of this article.

The authors' contributions are as follows: Y. L. M. Z. had the primary responsibility for writing the manuscript; S. A. R. contributed to the analytical design and revised and edited the manuscript; P. L. O. and H. Z. analysed the data and edited the manuscript; J. L. and E. S. T. collected the data; R. M. V. D. developed the analytical plan, interpreted the results, contributed to manuscript writing, reviewed the manuscript, and directed the study. All the authors read and approved the final version of the manuscript.

Y. L. M. Z., S. A. R., P. L. O., H. Z., J. L. and R. M. V. D. do not have any conflicts of interest. E. S. T. has served as a member of advisory panels for Astra Zeneca (S) Private Limited, Merck Sharp and Dohme Asia Pacific Services Private Limited, Novo Nordisk Pharma (S) Private Limited, Novartis (S) Private Limited, Unilever and Bristol-Myers Squibb (S). He has lectured at Bristol-Myers Squibb Singapore, Residual Risk Reduction Initiative, Astra Zeneca (S) Private Limited, Abbott Manufacturing, Singapore, MSD Technology Singapore and Glaxo Smith-Kline, Singapore.

References

1 Wild, S, Roglic, G, Green, A, et al. (2004) Global prevalence of diabetes: estimates for the year 2000 and projections for 2030. Diabetes Care 27, 10471053.CrossRefGoogle ScholarPubMed
2 Chang T-T (2000) Rice, Staple Foods: Domesticated Plants and Animals from The Cambridge World History of Food. Cambridge: Cambridge University Press. http://www.cambridge.org/us/books/kiple/rice.htm (accessed September 2011). Google Scholar
3 Ahn, Y, Park, SJ, Kwack, HK, 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.CrossRefGoogle Scholar
4 Spiller, GA (2002) Whole Grain, Whole Wheat and White Flours in History. St Paul, MN: Eagan Press.Google Scholar
5 Misra, A, Rastogi, K & Joshi, SR (2009) Whole grains and health: perspective for Asian Indians. J Assoc Physicians India 57, 155162.Google Scholar
6 Ohnuki-Tierney, E (1993) Rice as Self: Japanese Identities Through Time. Princeton, NJ: Princeton University Press.Google Scholar
7 Slavin, JL, Martini, MC, Jacobs, DR, et al. (1999) Plausible mechanisms for the protectiveness of whole grains. Am J Clin Nutr 70, 459S463S.CrossRefGoogle ScholarPubMed
8 Slavin, JL, Jacobs, D, Marquart, L, et al. (2001) The role of whole grains in disease prevention. J Am Diet Assoc 101, 780785.CrossRefGoogle ScholarPubMed
9 Sun, Q, Spiegelman, D, van Dam, RM, et al. (2010) White rice, brown rice, and risk of type 2 diabetes in US men and women. Arch Intern Med 170, 961969.CrossRefGoogle Scholar
10 Villegas, R, Liu, S, Gao, YT, et al. (2007) Prospective study of dietary carbohydrates, glycemic index, glycemic load, and incidence of type 2 diabetes mellitus in middle-aged Chinese women. Arch Intern Med 167, 23102316.Google Scholar
11 Nanri, A, Mizoue, T, Noda, 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, 14681477.CrossRefGoogle ScholarPubMed
12 Hu, EA, Pan, A, Malik, V, et al. (2012) White rice consumption and risk of type 2 diabetes: meta-analysis and systematic review. BMJ 344, e1454.CrossRefGoogle ScholarPubMed
13 Jenkins, DJ, Wolever, TM, Taylor, RH, et al. (1981) Glycemic index of foods: a physiological basis for carbohydrate exchange. Am J Clin Nutr 34, 362366.CrossRefGoogle Scholar
14 Salmeron, J, Ascherio, A, Rimm, EB, et al. (1997) Dietary fiber, glycemic load, and risk of NIDDM in men. Diabetes Care 20, 545550.CrossRefGoogle ScholarPubMed
15 Salmeron, J, Manson, JE, Stampfer, MJ, et al. (1997) Dietary fiber, glycemic load, and risk of non-insulin-dependent diabetes mellitus in women. JAMA 277, 472477.CrossRefGoogle ScholarPubMed
16 Liu, S (2002) Intake of refined carbohydrates and whole grain foods in relation to risk of type 2 diabetes mellitus and coronary heart disease. J Am Coll Nutr 21, 298306.Google Scholar
17 Halton, TL, Liu, S, Manson, JE, et al. (2008) Low-carbohydrate-diet score and risk of type 2 diabetes in women. Am J Clin Nutr 87, 339346.CrossRefGoogle ScholarPubMed
18 Radhika, G, Van Dam, RM, Sudha, V, et al. (2009) Refined grain consumption and the metabolic syndrome in urban Asian Indians (Chennai Urban Rural Epidemiology Study 57). Metabolism 58, 675681.Google Scholar
19 Sluijs, I, van der Schouw, YT, van der, AD, et al. (2010) Carbohydrate quantity and quality and risk of type 2 diabetes in the European Prospective Investigation into Cancer and Nutrition-Netherlands (EPIC-NL) study. Am J Clin Nutr 92, 905911.CrossRefGoogle ScholarPubMed
20 Ludwig, DS (2002) The glycemic index: physiological mechanisms relating to obesity, diabetes, and cardiovascular disease. JAMA 287, 24142423.CrossRefGoogle ScholarPubMed
21 Esfahani, A, Wong, JM, Mirrahimi, A, et al. (2009) The glycemic index: physiological significance. J Am Coll Nutr 28, Suppl., 439S445S.Google Scholar
22 Liu, S, Manson, JE, Stampfer, MJ, et al. (2000) A prospective study of whole-grain intake and risk of type 2 diabetes mellitus in US women. Am J Public Health 90, 14091415.Google Scholar
23 Meyer, KA, Kushi, LH, Jacobs, DR, et al. (2000) Carbohydrates, dietary fiber, and incident type 2 diabetes in older women. Am J Clin Nutr 71, 921930.CrossRefGoogle ScholarPubMed
24 Fung, TT, Hu, FB, Pereira, MA, et al. (2002) Whole-grain intake and the risk of type 2 diabetes: a prospective study in men. Am J Clin Nutr 76, 535540.CrossRefGoogle ScholarPubMed
25 Woo, J, Ho, SC, Sham, A, et al. (2003) Diet and glucose tolerance in a Chinese population. Eur J Clin Nutr 57, 523530.Google Scholar
26 Kim, K, Yun, SH, Choi, BY, et al. (2008) Cross-sectional relationship between dietary carbohydrate, glycaemic index, glycaemic load and risk of the metabolic syndrome in a Korean population. Br J Nutr 100, 576584.CrossRefGoogle Scholar
27 Mohan, V, Radhika, G, Sathya, RM, et al. (2009) Dietary carbohydrates, glycaemic load, food groups and newly detected type 2 diabetes among urban Asian Indian population in Chennai, India (Chennai Urban Rural Epidemiology Study 59). Br J Nutr 102, 14981506.Google Scholar
28 Zhang, G, Pan, A, Zong, G, et al. (2011) Substituting white rice with brown rice for 16 weeks does not substantially affect metabolic risk factors in middle-aged Chinese men and women with diabetes or a high risk for diabetes. J Nutr 141, 16851690.Google Scholar
29 Li, S, Shin, HJ, Ding, EL, et al. (2009) Adiponectin levels and risk of type 2 diabetes: a systematic review and meta-analysis. JAMA 302, 179188.CrossRefGoogle ScholarPubMed
30 Mineo, C & Shaul, PW (2012) Novel biological functions of high-density lipoprotein cholesterol. Circ Res 111, 10791090.Google Scholar
31 Yim-Lui Cheung, C, Wong, TY, Lamoureux, EL, et al. (2010) C-reactive protein and retinal microvascular caliber in a multiethnic Asian population. Am J Epidemiol 171, 206213.CrossRefGoogle Scholar
32 Nang, EE, Khoo, CM, Tai, ES, et al. (2009) Is there a clear threshold for fasting plasma glucose that differentiates between those with and without neuropathy and chronic kidney disease?: the Singapore Prospective Study Program. Am J Epidemiol 169, 14541462.CrossRefGoogle Scholar
33 Nang, EE, Gitau Ngunjiri, SA, Wu, Y, et al. (2011) Validity of the International Physical Activity Questionnaire and the Singapore Prospective Study Program physical activity questionnaire in a multiethnic urban Asian population. BMC Med Res Methodol 11, 141.Google Scholar
34 Deurenberg-Yap, M, Li, T, Tan, WL, et al. (2001) Validation of a semiquantitative food frequency questionnaire for estimation of intakes of energy, fats and cholesterol among Singaporeans. Asia Pac J Clin Nutr 9, 282288.Google Scholar
35 Matthews, DR, Hosker, JP, Rudenski, AS, et al. (1985) Homeostasis model assessment: insulin resistance and beta-cell function from fasting plasma glucose and insulin concentrations in man. Diabetologia 28, 412419.Google Scholar
36 Lopez-Garcia, E, van Dam, RM, Qi, L, et al. (2006) Coffee consumption and markers of inflammation and endothelial dysfunction in healthy and diabetic women. Am J Clin Nutr 84, 888893.Google Scholar
37 Expert WHO Consultation, (2004) Appropriate body-mass index for Asian populations and its implications for policy and intervention strategies. Lancet 363, 157163.Google Scholar
38 American Diabetes Association (2011) Diagnosis and classification of diabetes mellitus. Diabetes Care 34, Suppl. 1, S62S69.Google Scholar
39 Jeppesen, J, Schaaf, P, Jones, C, et al. (1997) Effects of low-fat, high-carbohydrate diets on risk factors for ischemic heart disease in postmenopausal women. Am J Clin Nutr 65, 10271033.CrossRefGoogle ScholarPubMed
40 Willett, W, Manson, J & Liu, S (2002) Glycemic index, glycemic load, and risk of type 2 diabetes. Am J Clin Nutr 76, 274S280S.Google Scholar
41 Murakami, K, Sasaki, S, Takahashi, Y, et al. (2006) Dietary glycemic index and load in relation to metabolic risk factors in Japanese female farmers with traditional dietary habits. Am J Clin Nutr 83, 11611169.Google Scholar
42 Odegaard, AO, Koh, WP, Butler, LM, et al. (2011) Dietary patterns and incident type 2 diabetes in chinese men and women: the Singapore Chinese Health Study. Diabetes Care 34, 880885.Google Scholar
43 Sakurai, M, Nakamura, K, Miura, K, et al. (2012) Dietary glycemic index and risk of type 2 diabetes mellitus in middle-aged Japanese men. Metabolism 61, 4755.Google Scholar
44 Atkinson, FS, Foster-Powell, K & Brand-Miller, JC (2008) International tables of glycemic index and glycemic load values: 2008. Diabetes Care 31, 22812283.Google Scholar
45 Lok, KY, Chan, R, Chan, D, et al. (2010) Glycaemic index and glycaemic load values of a selection of popular foods consumed in Hong Kong. Br J Nutr 103, 556560.CrossRefGoogle ScholarPubMed
Figure 0

Table 1 Characteristics of the study participants according to quintiles of rice and noodle consumption (Mean values and standard deviations; median values, interquartile ranges and percentages)

Figure 1

Table 2 Metabolic risk factors according to quintiles of rice consumption (Geometric means and 95 % confidence intervals; median values and interquartile ranges)

Figure 2

Table 3 Metabolic risk factors according to quintiles of noodle consumption (Geometric means and 95 % confidence intervals; median values and interquartile ranges)

Supplementary material: File

Zuñiga et al. Supplementary Material

Tables

Download Zuñiga et al. Supplementary Material(File)
File 90.6 KB