Abstract

Introduction

Asians, especially those from East Asia, are known to have a genetic predisposition to poor insulin secretion by pancreatic β-cells and their insulin secretion is prone to decline with β-cell exhaustion.^{1, 2} Asians have traditionally consumed white rice (WR) as a staple food that provides more than 30% of daily energy intake.³ High intake of WR is reported to be associated with an increased risk of type 2 diabetes and metabolic syndrome,⁴ and it has also been reported that this relation is stronger for Asians than Westerners.⁵

In subjects with metabolic syndrome, eating brown rice (BR) led to weight loss, improved endothelial dysfunction and improved glucose and lipid metabolism,⁶ while intake of BR also decreased hemoglobin A1c (HbA1c) in subjects with diabetes mellitus.⁷ When 50 g of WR, corresponding to one-third of the daily intake, is replaced by BR, the risk of developing type 2 diabetes is reduced by 16%.⁸ However, it may be difficult for people to continue eating BR on a daily basis due to its taste and texture, even though BR could prevent or improve diabetes compared with WR.

Rice can be classified as glutinous or non-glutinous. Glutinous rice is stickier and it is mainly used to make rice cakes in Japan, while non-glutinous rice is generally boiled and eaten with the daily meals. The starch in non-glutinous rice (including WR) consists of 20% amylose and 80% amylopectin, whereas the starch in glutinous rice is 100% amylopectin. Since glutinous rice is widely accepted by Japanese people as having a good taste and texture, we expected that patients might prefer glutinous brown rice (GBR) over BR and continue to eat it in daily life. We previously demonstrated that eating GBR for just 1 day improved the whole-day glucose profile compared with eating WR or BR, mainly by reducing postprandial glucose excursion, and GBR also overcame the problem of poor palatability of BR.⁹

Accordingly, the present study was performed to evaluate whether eating GBR twice a day for 8 weeks could improve glycemic control in Japanese patients with type 2 diabetes.

Materials and methods

Results

Among the 18 subjects registered in this study, 2 were excluded from analysis because of personal problems (n=1) and emergency hospitalization for colorectal cancer (n=1). The other 16 subjects (12 men and 4 women) completed the study and formed the per protocol set for analyses (Figure 1). None of the subjects dropped out of the study because of disliking the taste of GBR.

Open in a separate window

Figure 1

Flow diagram of subject disposition.

Baseline characteristics of the 16 subjects are shown in Table 1. Their average age was 64.0±8.8 (45–74) years and average body mass index was 25.7±5.6 (14.9–36.2) kg m⁻². The mean duration of diabetes was 14.7±10.3 (2–33) years and the mean dose of insulin was 34.4±16.0 (12–60) units per day. The average energy intake per meal from GBR and WR was 277±75 kcal (1157±312 kJ).

In the group eating GBR first, the baseline HbA1c was 7.5±0.5%. After 8 weeks of eating GBR twice daily, HbA1c showed a decrease to 7.1±0.5%. After switching to WR for 8 weeks, the HbA1c of this group remained at 7.1±0.5%. In the group that ate WR first, HbA1c was 7.4±0.7% at baseline, 7.5±0.9% after 8 weeks of eating WR twice daily and 7.3±0.8% after switching to GBR for the subsequent 8 weeks. A crossover order effect was not detected by one-way analysis of variance (P=0.409). When pooled data for GBR and WR were compared, there was a significant decrease of HbA1c from 7.5±0.2 to 7.2±0.2% (P=0.014) when the subjects were eating GBR (Figure 2a), while there was no significant change while they were eating WR (7.2±0.2–7.3±0.2%, P=0.848; Figure 2b). Dietary intake of GBR also resulted in a significant decrease of glycoalbumin from 20.4±0.8 to 19.4±0.9% (P=0.029), while there was no change with WR (from 20.0±0.7 to 20.1±0.9%, P=0.875). The profiles of plasma glucose and serum C-peptide during the mixed meal tolerance test are displayed in Figure 3. In comparison with baseline, plasma glucose levels were lower when the subjects were eating GBR, and there was a significant reduction from 193.8±8.1 mg dl⁻¹ (10.7±0.4 mmol l⁻¹) to 172.3±8.2 mg dl⁻¹ (9.5±0.4 mmol l⁻¹; P=0.031) at 30 min. Although the serum C-peptide level did not differ significantly from baseline at each time point of assessment, the incremental AUC_C-peptide was significantly smaller when the subjects were eating GBR from 31.3±5.6 to 22.1±4.5 ng min ml⁻¹ (P=0.023). On the other hand, there were no significant changes of plasma glucose and serum C-peptide (including the total and incremental AUC) while the subjects were eating WR. Furthermore, there was no significant change of body mass index throughout the entire 16-week study period. Other parameters also showed no significant changes during the study period, including plasma levels of T-C, high-density lipoprotein, triglycerides and BUN, and the estimated glomerular filtration rate. No adverse events occurred in any of the 16 subjects who completed the study.

Post hoc analysis showed a significant decrease of HbA1c (from 7.9±0.8 to 7.4±0.8%, P=0.013) and glycoalbumin (from 20.2±4.3 to 17.9±3.8%, P=0.002) when eating GBR in patients using DPP-4 inhibitors (n=7; Figures 4a and b), while there were no significant decreases when eating GBR without DPP-4 inhibitor therapy (n=9; Figures 4c and d). In contrast, there was no difference in the effect of WR on glycemic control between the subgroups with or without DPP-4 inhibitors (data not shown). Plasma levels of active GLP-1 and PYY also increased in patients using DPP-4 inhibitors when eating GBR (Supplementary Figure). With respect to insulin sensitizers, there were no significant differences in the effects of GBR or WR between patients with (n=8) and without (n=8) metformin (data not shown).

Discussion

We previously reported that eating GBR for even 1 day reduced postprandial glucose excursion compared with intake of WR. In the present study, postprandial plasma glucose, glycoalbumin and HbA1c were significantly decreased (by 21.5 mg dl⁻¹, 0.3% and 1.1%, respectively) after 8-week ingestion of GBR twice a day, demonstrating that the single-day improvement of the plasma glucose profile could be maintained for at least 8 weeks. The incremental AUC_C-peptide was also significantly smaller while the subjects were eating GBR, despite no change of weight. These results suggested that improvement of glycemic control by intake of GBR may not have been due to increased insulin secretion. However, the mechanisms involved are presently unknown.

It was interesting that this study demonstrated a significant difference in the effect of GBR between patients with and without concomitant use of DPP-4 inhibitors, which inhibit the breakdown of GLP-1. Why was there greater improvement of glycemic control by GBR in patients using DPP-4 inhibitors? A possible explanation is that GBR might stimulate GLP-1 secretion secondary to an increase of short-chain fatty acids produced from dietary fiber by the gut microbial flora. A previous study showed that eating whole grains for 6 weeks improved gut flora and increased short-chain fatty acid levels.¹⁰ It was also reported that the short-chain fatty acid propionate stimulates secretion of GLP-1 and PYY via free fatty acid receptor 2 in rodents.¹¹ GBR has a higher content of dietary fiber compared with WR (Supplementary Table). When we compared plasma levels of active GLP-1 and PYY before and after each period, both were increased by intake of GBR (Supplementary Figure), although the changes were not significant due to the small sample size. Therefore, DPP-4 inhibitor therapy could have an additive effect on the benefits of GBR.

GBR consists of bran, embryo Buda germ and endosperm. Bran also contains some substances, such as magnesium and γ-oryzanol, that may improve glycemic control.^{12, 13} Insufficient dietary intake of magnesium may contribute to exacerbation of insulin resistance due to inhibition of the autophosphorylation of insulin receptors.^{14, 15} One pack of GBR (113 g) contains 72.5 mg of magnesium, while the same amount of WR only contains 2.3 mg. Bran is also rich in γ-oryzanol.¹² It has been reported that a single oral dose of γ-oryzanol improves glycometabolic parameters and insulin resistance in non-insulin-dependent obese patients with type 2 diabetes.^{12, 16} We confirmed that one pack of GBR contains 31 mg of γ-oryzanol, although endosperm does not contain γ-oryzanol. These features of GBR may partly explain the improvement of glycemic control by intake of this type of rice in the present study.

Eating GBR overcomes the problem of poor palatability associated with BR. We previously reported that the palatability of GBR was similar to that of WR, based on the results of a questionnaire survey after ingestion of the different types of rice for 1 day each.⁹ In that study, the subjects reported striking differences of taste, texture and consumability between BR and GBR in spite of the similar appearance of these two types of rice. In the present study, there were no dropouts because of unacceptability of the taste of GBR. This is the first study to demonstrate that intake of GBR is tolerable and maintains efficacy for at least 8 weeks.

The present study had several limitations including a small sample size. Although a double-blind, randomized study would have been desirable, we used an open-label, crossover design because of the difficulty of blinding the different types of rice. Accordingly, we cannot exclude the possible influence of bias on our results. In addition, given the limitations of post hoc analysis, a randomized prospective study is needed to confirm the better response to GBR of patients treated with DPP-4 inhibitors. Furthermore, we did not assess insulin sensitivity, although a change of insulin sensitivity might have been associated with the improvement of glycemic control by GBR. Since we detected a decrease of the incremental AUC_C-peptide during the GBR period of the present study that suggested improvement of insulin sensitivity, this warrants further evaluation. Despite these limitations, the present findings could be useful for assisting patients with diabetes to make decisions about their diet.

In conclusion, eating GBR twice a day for 8 weeks was well tolerated with respect to palatability and led to sustained improvement of glycemic control in patients with type 2 diabetes.

Footnotes

References