OBJECTIVE

To consider the effects of whole-grain processing, specifically milling, on glycemic control in free-living adults with type 2 diabetes.

RESEARCH DESIGN AND METHODS

Participants of this crossover trial were randomized to two interventions of 2 weeks, separated by washout. They were advised to replace the grain foods they normally consumed with intervention foods. Intervention foods were nutrient-matched whole-grain products of wheat, oats, and brown rice that differed in their degree of processing. No other lifestyle advice was given. Continuous glucose monitoring systems were worn. Other cardiometabolic risk factors and alkylresorcinols (a biomarker of whole-grain intake) were measured pre- and postintervention.

RESULTS

Thirty-one adults with type 2 diabetes (63 ± 13 years old, BMI 32.4 ± 7 kg/m2, HbA1c 7.5 ± 3.4% [59 ± 14 mmol/mol]) commenced the trial; 28 (90%) completed both interventions. The increase in alkylresorcinols did not differ between interventions, and there was no difference in reported energy intake. Postprandial responses were 9% (95% CI 3–15) lower following breakfast and 6% (1–10) lower following all meals of less-processed whole grains when compared with finely milled grains. Day-long glycemic variability also was reduced when measured by 24-h SD (−0.16 mmol/L [95% CI −0.25 to −0.06]) and mean amplitude of glycemic excursion (−0.36 [95% CI −0.65 to −0.08]). Mean change in body weight differed by 0.81 kg (95% CI 0.62–1.05) between interventions, increasing during the finely milled intervention and decreasing during the less-processed whole-grain intervention. This was not a mediating factor for the glycemic variables considered.

CONCLUSIONS

Consuming less-processed whole-grain foods over 2 weeks improved measures of glycemia in free-living adults with type 2 diabetes compared with an equivalent amount of whole-grain foods that were finely milled. Dietary advice should promote the consumption of minimally processed whole grains.

Diets rich in whole grains are associated with reduced incidence of type 2 diabetes, coronary heart disease, and colorectal cancer (1). Randomized controlled trials of increasing whole-grain intakes have demonstrated their potential to improve glycemic control, body weight, the lipid profile, and other cardiometabolic risk factors in adults with diabetes (2). Furthermore, their production has a lesser adverse effect on planetary health than is the case for many other nutrient-rich foods (3). Unsurprisingly, many national dietary guidelines (4) and recommendations for the management of diabetes (5,6) encourage whole-grain consumption and suggest replacing refined grains with whole grains.

Current definitions of the term whole grain permit the inclusion of highly processed and reconstituted fractions of whole grains (7,8). Furthermore, such finely ground products may be added into ultraprocessed foods, the consumption of which can lead to increased calorie intake and weight gain (9). There is evidence that the acute glycemic response to whole-grain foods is greater when grains have been finely milled (10). More recently, we have shown that the glycemic response to four different whole-grain wheat breads is related to the degree of grain processing in adults with type 2 diabetes (11).

To date, all relevant studies have examined the acute effects of whole-grain processing on postprandial glycemia in controlled settings. We have undertaken a 2-week randomized crossover trial in free-living adults with type 2 diabetes to determine whether the structural integrity of whole grains as a result of differences in milling is a determinant of overall glycemic control. We provided participants with their whole-grain foods and used continuous glucose monitoring (CGM) systems to assess glycemic control.

This randomized crossover trial compared two dietary interventions, each of 2-week duration, separated by a washout period of at least 2 weeks. The trial was conducted between October 2018 and April 2019 in Dunedin, New Zealand. The trial protocol was reviewed and approved by the Health and Disability Ethics Committee of the Ministry of Health, New Zealand (reference 18/STH/172). All participants provided written informed consent. The trial protocol was prospectively registered with the Australia New Zealand Clinical Trials Registry.

Participant Eligibility Criteria

Adults diagnosed with type 2 diabetes and aged between 18 and 80 years were eligible. Presence of comorbidities did not exclude participation; however, pregnancy, lactation, or a change in medication potentially influencing blood glucose control in the past 3 months did. Participants were recruited locally through general practices, fliers in supermarkets, online advertisements, and participation in earlier studies.

Randomization and Masking

The intervention order was determined by a computer-generated 1:1 block randomization protocol. Each intervention order was stored in a separate opaque envelope and accessed sequentially once each participant provided written consent. Participants were blinded to CGM data during both interventions and did not receive study feedback until both interventions were complete. The author who ran the analysis was blinded to the interventions for the initial comparisons. Given the nature of the interventions, participants could not be masked. They were aware of the research question this trial sought to address, and there were visible differences in the foods provided.

Interventions

Given that the aim of the trial was to examine the effect of food processing of whole grains on glycemic control and other cardiometabolic risk factors, we provided participants with commonly consumed whole-grain foods that differed in the extent of their milling. In one intervention, participants were provided with intact oats, brown rice, and whole-grain bread made with coarsely ground flour and kibbled wheat kernels. They were instructed to replace their current grain intake, determined by a preintervention 4-day diet record, with the foods provided. In the other intervention, participants were provided with instant oats, brown rice pasta, and whole-grain bread made with finely milled flour, and the same instructions to replace their current grain intake with the foods provided. No advice was given to change the amount of grains consumed. No further dietary or other lifestyle advice was given. Whole-grain foods were commercially available, were 100% whole grain (7,8), were matched for macronutrients and fiber, and met the American Association of Cereal Chemists International characterization of a whole-grain product (12).

Before the first intervention, participants completed a 4-day semiquantitative food diary to estimate their usual intakes. Metric cups were provided to each participant for use when weighing scales were not available. Participants attended the clinic on days 1, 7, and 14 during both interventions. On day 1, a baseline fasting blood sample and anthropometric measurements were taken, the intervention was explained, and participants received their whole-grain foods. CGM systems were fitted to the upper arm and activated. Participants were provided with daily checklists to record when and what intervention foods were consumed. On day 7, participants were provided with more intervention foods if required and given a 4-day semiquantitative food diary with instructions to capture 3 weekdays and 1 weekend day in the coming week. On day 14, the intervention ended, a fasting blood sample and anthropometric measurements were taken, and the CGM system removed.

Measurements

Anthropometric measurements (height, weight, and body composition) were recorded in duplicate, and resting blood pressure was measured three times. Participants wore a CGM system (FreeStyle Libre Pro; Abbot Laboratories) to measure interstitial glucose every 15 min for the duration of the interventions. This model was blinded so that data could not be viewed during the intervention. This model was also factory calibrated and did not need additional calibration against capillary blood measures. Fasting plasma samples were collected and stored at −80°C until analysis. Glycated hemoglobin A1c (HbA1c), cholesterol (total, LDL, and HDL), triglycerides, C-reactive protein, and α-1-acid glycoprotein were measured on an automated analyzer (Roche Diagnostics) after calibration with the appropriate standards. Fasting insulin was measured with a Bio-Plex magnetic bead array (Bio-Rad Laboratories), with concentrations calculated on standard curve data using the manufacturer’s software. Alkylresorcinols were measured as an objective marker of whole-grain intake (13).

Alkylresorcinol homologs C17:0, C19:0, and C21:0 were measured with liquid chromatography-high-resolution mass spectrometry (14) and then summed. Grain particle size for each of the foods provided in the interventions was characterized by sieve analysis (15), using 12 sieves from 63 to 5,600 μm.

The primary outcome investigated was a change in blood glucose control following meals and over the day. Postprandial glycemia was measured by the blood glucose incremental area under the curve (iAUC) with CGM data in the 3 h after breakfast, lunch, and dinner. Participants’ food diary and habitual mealtimes provided the initial basis for identifying breakfast, lunch, and dinner time frames through a hierarchical decision-making process. Where the time of meal commencement could not be established, the data were excluded from the analysis. iAUC calculations for the 2-week interventions were based on 730 breakfasts, 734 lunches, and 764 dinners. Day-long variables were time spent in range (3.9–10.0 mmol/L), time spent below range (<3.9 mmol/L), time spent above range (>10.0 mmol/L) (16), and measures of glycemic variability. The daily measures of glycemic variability were the SD of the mean (16), the continuous overall net glycemic action (17), and the mean amplitude of glycemic excursion (MAGE) (18). Values of glycemic variability were calculated (19). Nutrition information was obtained from manufacturers. Dietary data were analyzed with FoodWorks 9 (Xyris Software) using the New Zealand FOODfiles 2016, supplemented by AusFoods 2017.

Statistical Analysis

The sample size estimate was based on a power calculation with an α of 0.05 and power of 0.80 to detect within-group differences in primary outcome variables, a 20% difference in mean postprandial glycemia as measured by iAUC. Twenty-eight participants were required to complete both interventions. Data were analyzed according to intention to treat. Analyses of CGM data were performed with a mixed model accounting for intervention order. An interaction between each glycemic variable and weight change during the intervention period was considered. For pre- and postintervention measures, we compared the difference in one intervention with the difference in the other intervention. The data for CGM variables, triglycerides, HDL cholesterol, alkylresorcinols, and markers of inflammation were log-transformed to address skew. Analyses were performed using Stata 15 (StataCorp, College Station, TX). Results are mean ± SD unless otherwise stated.

The flow of participants through the trial is shown in Fig. 1. Of the 31 participants with type 2 diabetes randomized, 28 (90%) completed both interventions. Fourteen (45%) of the 31 participants were female. Mean age was 63 ± 13 years. Mean HbA1c was 7.5 ± 3.4% (59 ± 14 mmol/mol), and mean diabetes duration was 11.4 ± 9 years. Participants self-identified as being of New Zealand European (n = 25, 81%), Māori (n = 3, 10%), European (n = 2, 6%), or Indian (n = 1, 3%) ethnicity. The majority of participants were on oral hypoglycemic agents metformin and/or gliclazide (n = 19, 61%), with nine (29%) on both oral hypoglycemic agents and insulin and three (10%) able to control their blood glucose with diet alone. Three (10%) participants were current smokers.

Figure 1

Flowchart of participants through the study.

Figure 1

Flowchart of participants through the study.

Close modal

Characteristics of the intervention foods are shown in Table 1, with full details of the sieve analysis available in the Supplementary Material. Dietary intakes before and during the interventions are shown in Table 2. During both intervention periods, total carbohydrate, starch, and dietary fiber increased at the expense of fat intake compared with baseline levels. Intakes of all macronutrients and dietary fiber were comparable between the two interventions. The increase in alkylresorcinols during the less-processed whole-grain intervention (from 55.8 ± 63 to 141.2 ± 295 nmol/L) was comparable (P = 0.403) with the increase during the intervention of finely milled whole grains (from 49.5 ± 41 to 171.9 ± 336 nmol/L). During the interventions, participants consumed a mean intake of 5.5 ± 1.3 servings of less-processed whole grains or 5.5 ± 1.4 servings of finely milled whole grains each day. The most servings of whole grains were consumed at breakfast (2.3 ± 0.9 and 2.2 ± 0.9) and then at lunch (1.8 ± 0.8 and 1.8 ± 0.7) and dinner (1.4 ± 0.8 and 1.4 ± 0.8) for less-processed whole grains and finely milled whole grains, respectively.

Table 1

Nutrient information of intervention foods

Less-processed whole-grain interventionFinely milled whole-grain intervention
Traditional oats (cooked)Brown rice (cooked)Coarsely milled breadInstant oats (cooked)Brown rice pasta (cooked)Finely milled bread
Nutrients per 100 g       
 Energy (kJ) 512 850 862 512 822 852 
 Carbohydrates (g) 18.4 39.5 35.77 18.4 40.5 36.49 
 Protein (g) 4.8 4.9 6.99 4.8 4.3 7.03 
 Fat (g) 2.3 2.2 2.45 2.3 1.6 1.79 
 Fiber (g) 4.0 1.6 9.2 4.0 1.4 9.1 
 Sodium (mg) <5 <5 290 <5 6.5 290 
Retention of whole grains on particle-size sieves (μm), %       
 >2,800 93 23 40 0* 
 1,000–2,799 100 39 52 0* 
 180–999 16 6* 59 
 <180 22 94* 41 
Less-processed whole-grain interventionFinely milled whole-grain intervention
Traditional oats (cooked)Brown rice (cooked)Coarsely milled breadInstant oats (cooked)Brown rice pasta (cooked)Finely milled bread
Nutrients per 100 g       
 Energy (kJ) 512 850 862 512 822 852 
 Carbohydrates (g) 18.4 39.5 35.77 18.4 40.5 36.49 
 Protein (g) 4.8 4.9 6.99 4.8 4.3 7.03 
 Fat (g) 2.3 2.2 2.45 2.3 1.6 1.79 
 Fiber (g) 4.0 1.6 9.2 4.0 1.4 9.1 
 Sodium (mg) <5 <5 290 <5 6.5 290 
Retention of whole grains on particle-size sieves (μm), %       
 >2,800 93 23 40 0* 
 1,000–2,799 100 39 52 0* 
 180–999 16 6* 59 
 <180 22 94* 41 
*

These measurements were made on brown rice flour as the only listed ingredient in brown rice pasta.

Table 2

Nutrient intakes pretrial and during both interventions

NutrientPrestudyLess-processed whole grainsFinely milled whole grainsP value difference between interventions
Energy (kJ) 8,820 ± 1,936 9,027 ± 2,669 9,141 ± 2,171 0.858 
Fat (%TE) 35 ± 6 30 ± 5 31 ± 6 0.455 
Carbohydrates (%TE) 46 ± 7 52 ± 6 51 ± 7 0.641 
Protein (%TE) 19 ± 4 17 ± 4 17 ± 3 0.914 
Saturated fat (%TE) 14.0 ± 3.5 12.2 ± 3 12.4 ± 3 0.790 
Carbohydrates (g/day) 241.7 ± 55 285.6 ± 84 287.4 ± 65 0.921 
Sugar (g/day) 83.1 ± 30 78.0 ± 34 80.6 ± 29 0.603 
Starch (g/day) 133.9 ± 38 166.0 ± 54 167.3 ± 46 0.937 
Fiber (g/day) 25.8 ± 9 40.0 ± 11 38.3 ± 11 0.300 
Carbohydrates at breakfast (g/day) 55.8 ± 36.6 61.6 ± 31.9 60.9 ± 31.0 0.616 
Carbohydrates at lunch (g/day) 49.9 ± 30.3 55.0 ± 32.1 55.8 ± 33.1 0.841 
Carbohydrates at dinner (g/day) 67.4 ± 35.5 75.3 ± 45.4 77.9 ± 42.7 0.951 
NutrientPrestudyLess-processed whole grainsFinely milled whole grainsP value difference between interventions
Energy (kJ) 8,820 ± 1,936 9,027 ± 2,669 9,141 ± 2,171 0.858 
Fat (%TE) 35 ± 6 30 ± 5 31 ± 6 0.455 
Carbohydrates (%TE) 46 ± 7 52 ± 6 51 ± 7 0.641 
Protein (%TE) 19 ± 4 17 ± 4 17 ± 3 0.914 
Saturated fat (%TE) 14.0 ± 3.5 12.2 ± 3 12.4 ± 3 0.790 
Carbohydrates (g/day) 241.7 ± 55 285.6 ± 84 287.4 ± 65 0.921 
Sugar (g/day) 83.1 ± 30 78.0 ± 34 80.6 ± 29 0.603 
Starch (g/day) 133.9 ± 38 166.0 ± 54 167.3 ± 46 0.937 
Fiber (g/day) 25.8 ± 9 40.0 ± 11 38.3 ± 11 0.300 
Carbohydrates at breakfast (g/day) 55.8 ± 36.6 61.6 ± 31.9 60.9 ± 31.0 0.616 
Carbohydrates at lunch (g/day) 49.9 ± 30.3 55.0 ± 32.1 55.8 ± 33.1 0.841 
Carbohydrates at dinner (g/day) 67.4 ± 35.5 75.3 ± 45.4 77.9 ± 42.7 0.951 

Data are mean ± SD. Variables that differed in both interventions from prestudy intakes were fat and carbohydrate (%TE) as well as carbohydrates, starch, and fiber (g/day). %TE, percentage of total energy provided to the diet by that macronutrient.

Measures of glycemia are shown in Table 3. The mean iAUC of the 3 h following all the breakfast meals was 9% (95% CI 3–15) lower during the 2 weeks of consuming less-processed whole grains than during the weeks when finely milled whole grains were consumed. The mean postprandial response to the average of all meals showed a similar trend, being 6% (1–10) lower. Glycemic variability reduced when participants consumed the less-processed whole grains compared with the finely milled whole grains as measured by MAGE (−0.36 [95% CI −0.65 to −0.08]) and the SD of the mean glucose value (−0.16 mmol/L [95% CI −0.25 to −0.06]). The study was not powered to consider differences in response between participants on the basis of sex or ethnicity.

Table 3

Measures of glycemia calculated from CGM

MeasureLess-processed whole grainsFinely milled whole grainsP value difference between interventionsP value interaction with weight change
Meal responses (mmol/L/min)     
 All-meal iAUC 423 ± 210 466 ± 192 0.022 0.555 
 Breakfast iAUC 449 ± 256 525 ± 248 0.007 0.984 
 Lunch iAUC 412 ± 287 440 ± 304 0.614 0.321 
 Dinner iAUC 391 ± 293 415 ± 277 0.117 0.118 
24-h measures     
 Hours spent in range (3.9–10 mmol/L) 15.49 ± 6.77 15.18 ± 6.74 0.466 0.383 
 Hours spent above range (>10 mmol/L) 7.90 ± 7.02 8.22 ± 7.12 0.736 0.305 
 Hours spent below range (<3.9 mmol/L) 0.61 ± 1.65 0.82 ± 2.30 0.602 0.812 
Measures of glycemic variability     
 MAGE 5.61 ± 2.75 5.94 ± 2.60 0.014 0.193 
 CONGA 8.07 ± 2.49 8.20 ± 2.85 0.496 0.699 
 SD of daily mean (mmol/L) 2.33 ± 1.07 2.51 ± 1.10 0.002 0.803 
MeasureLess-processed whole grainsFinely milled whole grainsP value difference between interventionsP value interaction with weight change
Meal responses (mmol/L/min)     
 All-meal iAUC 423 ± 210 466 ± 192 0.022 0.555 
 Breakfast iAUC 449 ± 256 525 ± 248 0.007 0.984 
 Lunch iAUC 412 ± 287 440 ± 304 0.614 0.321 
 Dinner iAUC 391 ± 293 415 ± 277 0.117 0.118 
24-h measures     
 Hours spent in range (3.9–10 mmol/L) 15.49 ± 6.77 15.18 ± 6.74 0.466 0.383 
 Hours spent above range (>10 mmol/L) 7.90 ± 7.02 8.22 ± 7.12 0.736 0.305 
 Hours spent below range (<3.9 mmol/L) 0.61 ± 1.65 0.82 ± 2.30 0.602 0.812 
Measures of glycemic variability     
 MAGE 5.61 ± 2.75 5.94 ± 2.60 0.014 0.193 
 CONGA 8.07 ± 2.49 8.20 ± 2.85 0.496 0.699 
 SD of daily mean (mmol/L) 2.33 ± 1.07 2.51 ± 1.10 0.002 0.803 

Data are mean ± SD. All values have been log-transformed to address skew. CONGA, continuous overall improvement in net glycemic action.

Anthropometric measures and cardiometabolic risk factors measured preintervention and after each intervention are shown in Table 4. Body weight increased during the finely milled whole-grain intervention and decreased when less-processed whole grains were consumed, resulting in a statistically significant mean difference in weight change (0.81 kg [95% CI 0.62–1.05]). This change in body weight was not a mediating factor for the glycemic variables considered (Table 3 ). There were no differences observed between interventions for blood lipids, insulin, inflammatory markers, or blood pressure.

Table 4

Anthropometric, blood pressure, blood lipid, and inflammation measures

Intervention
Less-processed whole grainsFinely milled whole grains
MeasurePrePostPrePostP difference between interventions
Weight (kg) 92.9 ± 21.1 92.4 ± 20.8 93.6 ± 21.0 94.0 ± 21.3 0.002 
BMI (kg/m232.5 ± 7.0 32.2 ± 6.8 32.8 ± 6.9 32.9 ± 7.0 0.001 
Fat mass (%) 36.1 ± 10.9 35.6 ± 11.2 37.2 ± 10.7 37.1 ± 11.0 0.312 
Systolic BP (mmHg) 130 ± 15 134 ± 17 134 ± 17 134 ± 15 0.161 
Insulin* (pg/mL) 497 ± 553 425 ± 373 414 ± 324 452 ± 513 0.234 
Total cholesterol (mmol/L) 4.17 ± 1.11 3.94 ± 1.16 4.38 ± 1.35 4.10 ± 1.22 0.346 
LDL cholesterol (mmol/L) 2.11 ± 0.90 1.96 ± 0.89 2.22 ± 0.84 2.06 ± 0.86 0.803 
HDL cholesterol* (mmol/L) 1.32 ± 0.42 1.26 ± 0.40 1.30 ± 0.38 1.27 ± 0.38 0.312 
Triglycerides* (mmol/L) 1.64 ± 0.85 1.61 ± 0.82 1.58 ± 0.87 1.55 ± 0.99 0.954 
CRP* (mg/L) 3.73 ± 3.95 3.66 ± 4.31 3.75 ± 3.53 3.27 ± 3.11 0.052 
AGP* (g/L) 0.74 ± 0.24 0.82 ± 0.28 0.66 ± 0.26 0.64 ± 0.25 0.131 
Intervention
Less-processed whole grainsFinely milled whole grains
MeasurePrePostPrePostP difference between interventions
Weight (kg) 92.9 ± 21.1 92.4 ± 20.8 93.6 ± 21.0 94.0 ± 21.3 0.002 
BMI (kg/m232.5 ± 7.0 32.2 ± 6.8 32.8 ± 6.9 32.9 ± 7.0 0.001 
Fat mass (%) 36.1 ± 10.9 35.6 ± 11.2 37.2 ± 10.7 37.1 ± 11.0 0.312 
Systolic BP (mmHg) 130 ± 15 134 ± 17 134 ± 17 134 ± 15 0.161 
Insulin* (pg/mL) 497 ± 553 425 ± 373 414 ± 324 452 ± 513 0.234 
Total cholesterol (mmol/L) 4.17 ± 1.11 3.94 ± 1.16 4.38 ± 1.35 4.10 ± 1.22 0.346 
LDL cholesterol (mmol/L) 2.11 ± 0.90 1.96 ± 0.89 2.22 ± 0.84 2.06 ± 0.86 0.803 
HDL cholesterol* (mmol/L) 1.32 ± 0.42 1.26 ± 0.40 1.30 ± 0.38 1.27 ± 0.38 0.312 
Triglycerides* (mmol/L) 1.64 ± 0.85 1.61 ± 0.82 1.58 ± 0.87 1.55 ± 0.99 0.954 
CRP* (mg/L) 3.73 ± 3.95 3.66 ± 4.31 3.75 ± 3.53 3.27 ± 3.11 0.052 
AGP* (g/L) 0.74 ± 0.24 0.82 ± 0.28 0.66 ± 0.26 0.64 ± 0.25 0.131 

Data are mean ± SD. AGP, α-1-acid glycoprotein; BP, blood pressure; CRP, C-reactive protein.

*

Log-transformed to address skew.

Consuming less-processed whole-grain foods over 2 weeks improved measures of glycemia in free-living adults with type 2 diabetes compared with an equivalent amount of whole-grain foods that were finely milled to reduce their particle size. All whole-grain foods met the definition of being 100% whole grain and had a similar composition in terms of dietary fiber and macronutrients. Whole-grain foods derived from wheat, oats, and brown rice were provided ad libitum to participants, with the amount consumed balanced between interventions. The reduction in glycemia was particularly striking after breakfast, when more servings of whole-grain foods were eaten than at other times during the day. Measures of day-long glycemic variability also improved, suggesting potential benefits in terms of reduced protein glycation and cardiovascular risk (2022). Although the only difference between the two intervention periods related to whole-grain structure as determined by milling and the reported energy intake did not differ, a different pattern of weight change was observed. Consumption of the less-processed whole-grain foods was associated with a mean reduction of body weight, whereas there appeared to be a modest increase during the 2-week period where finely milled whole-grain foods were provided. There was no indication that the weight difference explained the improvements in glycemia associated with the less-processed whole-grain foods characterized by larger particle size.

Several plausible mechanisms could explain the overall observation that the degree of whole-grain processing, as characterized by whole-grain particle size as a result of milling, alters starch digestibility. First, less-processed whole grains may pass through the small intestine to be digested in the colon by the microbiome into short-chain fatty acids (23), which are absorbed without altering circulating blood glucose levels. Second, smaller whole-grain particle size associated with milling allows water and enzymes easier access to the starch and enables catalytic activity, which alters the rate of glucose absorption (24). Third, the milling of whole grains to flour removes the compartmentalization of starch, affecting starch and protein interactions, which may further affect digestion rate (25). The differing pattern in body weight change between the interventions was unexpected. However, there are several possible explanations. A higher proportion of the intact whole grain may have escaped digestion in the small intestine to be either metabolized in the large bowel or have avoided anaerobic metabolism entirely. It is also conceivable that digestion of less-processed whole grains may be associated with a greater thermic effect than the digestion of whole grains that are more finely ground (26).

A body of work supports benefits in postprandial glycemia when comparing whole-grain consumption with refined grains (27). Fewer studies have considered glycemia following the consumption of intact or processed whole grains (11,2830). To our knowledge, the longest intervention relevant to this issue by Järvi et al. (28) reported on two different diets followed for 24 days by adults with type 2 diabetes. Diets were nutrient matched and provided the same foods; however, in one diet, the carbohydrate-dense foods, including whole grains, were milled, reducing particle size to produce a difference in glycemic index. Following the interventions, glucose was assessed over 1 day in a controlled setting. As with the current study, Järvi et al. reported improved glycemia following the intervention with less-processed foods of larger particle size. Older acute studies considering the structural integrity of whole grains on postprandial glycemia also confirm our findings (29,31,32). More recently, we have shown an inverse trend between postprandial glycemia and whole-grain particle size in breads in a controlled setting (11).

Current definitions of whole grains permit a wide range of processing methods and require that reconstitution of grain components is to the same ratio as they exist in an intact grain (7,8). The findings of this trial have immediate relevance to clinical guidelines for diabetes management, which currently recommend diets high in whole-grain foods without reference to grain structure or particle size (5,6,33). Given that other health benefits of whole-grain foods (1,9) may also be influenced by their processing, our data provide evidence for a review of the definition of the term whole grain by relevant authorities. A revised definition may reduce the number of health or content claims made for a range of ultraprocessed foods by virtue of their whole-grain content. Such products can also have high contents of sodium, free sugars, or saturated fats as well as total calories.

This trial has a number of strengths. Participants were free living and consuming the provided whole-grain foods within the context of their usual diet, increasing the generalizability of these results. As far as we are aware, this is the first study of this topic to use CGM systems. This has enabled more intensive blood glucose monitoring than has previously been possible and consideration of day-long glycemic variability and daily time spent in range and while hyperglycemic and hypoglycemic. Furthermore, this study used alkylresorcinols in plasma, providing an objective measure of adherence to the whole-grain interventions. The absence of objective measures of intervention adherence is a recognized limitation of previous studies. The trial does have some limitations. It was of insufficient duration to observe potential changes in HbA1c or the lipid profile. The trial was insufficiently powered to consider subgroup analyses according to participant glycemic control, medication type, sex, or ethnicity. The trial was also insufficiently powered to consider specific whole-grain foods or other variables that may influence the outcomes measured. Finally, the participants of this trial were volunteers who are not necessarily representative of the population of people with type 2 diabetes and, therefore, may be more likely to adhere to the trial interventions (34). Future larger and longer studies are planned to resolve these questions.

Consuming less-processed whole-grain foods achieved an improvement in postprandial glycemia and other indices of glycemic control in adults with type 2 diabetes compared with consuming whole-grain foods where the grain particle size was further reduced through milling. These findings suggest that maintaining the structural integrity of whole grains in foods available for consumption will likely have long-term health benefits. Our findings are highly relevant to nutritional guidelines for diabetes management and suggest practical means for those wishing to improve their blood glucose control. They may also contribute to discussion regarding a revision of the definition of the term whole grains and, consequently, use of the term in front-of-pack labeling claims.

Clinical trial reg. no. ACTRN12618001285246, www.anzctr.org.au

This article contains supplementary material online at https://doi.org/10.2337/figshare.12114120.

This article is featured in a podcast available at https://www.diabetesjournals.org/content/diabetes-core-update-podcasts.

Acknowledgments. The authors sincerely thank the participants of this trial.

Funding. This trial was funded by a Laurenson Award of the Otago Medical Research Foundation (LA384) and The Riddet Centre of Research Excellence. J.M. was supported by the Healthier Lives National Science Challenge. Harraway & Sons (Dunedin, New Zealand) provided the oats used in this trial.

Funders did not play a role in the design of the study or in conducting the study, the data analyses, or the interpretation of these results.

Duality of Interest. No potential conflicts of interest relevant to this article were reported.

Author Contributions. S.Å. conducted the study and collected the data. J.M. and A.N.R. designed the study. S.N. and A.B.R. measured cardiometabolic risk factors and biomarkers in plasma samples. All authors contributed to interpretation of the data, revision of drafts, and approval of the final manuscript. A.N.R. is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

1.
Reynolds
A
,
Mann
J
,
Cummings
J
,
Winter
N
,
Mete
E
,
Te Morenga
L
.
Carbohydrate quality and human health: a series of systematic reviews and meta-analyses
.
Lancet
2019
;
393
:
434
445
2.
Reynolds
AN
,
Akerman
AP
,
Mann
J
.
Dietary fibre and whole grains in diabetes management: systematic review and meta-analyses
.
PLoS Med
2020
;
17
:
e1003053
3.
Clark
MA
,
Springmann
M
,
Hill
J
,
Tilman
D
.
Multiple health and environmental impacts of foods
.
Proc Natl Acad Sci U S A
2019
;
116
:
23357
23362
4.
U.S. Department of Health and Human Services
.
Dietary Guidelines for Americans 2015-2020
.
New York
,
Skyhorse Publishing
,
2017
5.
American Diabetes Association
.
Lifestyle management: standards of medical care in diabetes
.
Diabetes Care
2019
;
42
:
S46
S60
6.
Dyson
PA
,
Twenefour
D
,
Breen
C
, et al.
Diabetes UK evidence-based nutrition guidelines for the prevention and management of diabetes
.
Diabet Med
2018
;
35
:
541
547
7.
Food Standards Australia New Zealand. Food Standards Code: Standard 2.1.1. Cereal and cereal products. Commonwealth of Australia Gazette FSC 96, 2015
8.
American Association of Cereal Chemists International
.
Whole grain definition
.
Cereal Foods World
1999
;
45
:
79
9.
Hall
KD
,
Ayuketah
A
,
Brychta
R
, et al.
Ultra-processed diets cause excess calorie intake and weight gain: an inpatient randomized controlled trial of ad libitum food intake [published correction appears in Cell Metab 2019;30:226]
.
Cell Metab
2019
;
30
:
67
77
.
e3
10.
Tosh
SM
,
Chu
Y
.
Systematic review of the effect of processing of whole-grain oat cereals on glycaemic response
.
Br J Nutr
2015
;
114
:
1256
1262
11.
Reynolds
AN
,
Mann
J
,
Elbalshy
M
, et al.
Wholegrain particle size influences postprandial glycemia in type 2 diabetes: a randomized crossover study comparing four wholegrain breads
.
Diabetes Care
2020
;
43
:
476
479
12.
AACCI's Whole Grains Working Group unveils new whole grain products characterization [21 May 2013]. Available from https://www.cerealsgrains.org/about/newsreleases/Pages/WholeGrainProductCharacterization.aspx. Accessed 4 February 2020
13.
Ross
AB
.
Present status and perspectives on the use of alkylresorcinols as biomarkers of wholegrain wheat and rye intake
.
J Nutr Metab
2012
;
2012
:
462967
14.
Rodríguez-Morató
J
,
Jayawardene
S
,
Dolnikowski
G
,
Galluccio
J
,
Lichtenstein
AH
,
Matthan
NR
.
Development of a simplified method for the measurement of plasma alkylresorcinols as a biomarker of whole grain intake and Application to a human clinical trial evaluating the effect of carbohydrate quality on cardiometabolic risk factors (Abstract)
.
Circulation
2019
;
139
:
AP285
15.
Folk
RL
.
Petrology of Sedimentary Rocks
.
Austin, TX
,
Hemphill Publishing Company
,
1980
16.
Battelino
T
,
Danne
T
,
Bergenstal
RM
, et al.
Clinical targets for continuous glucose monitoring data interpretation: recommendations from the international consensus on time in range
.
Diabetes Care
2019
;
42
:
1593
1603
17.
McDonnell
CM
,
Donath
SM
,
Vidmar
SI
,
Werther
GA
,
Cameron
FJ
.
A novel approach to continuous glucose analysis utilizing glycemic variation
.
Diabetes Technol Ther
2005
;
7
:
253
263
18.
Service
FJ
,
Molnar
GD
,
Rosevear
JW
,
Ackerman
E
,
Gatewood
LC
,
Taylor
WF
.
Mean amplitude of glycemic excursions, a measure of diabetic instability
.
Diabetes
1970
;
19
:
644
655
19.
Hill
NR
,
Oliver
NS
,
Choudhary
P
,
Levy
JC
,
Hindmarsh
P
,
Matthews
DR
.
Normal reference range for mean tissue glucose and glycemic variability derived from continuous glucose monitoring for subjects without diabetes in different ethnic groups
.
Diabetes Technol Ther
2011
;
13
:
921
928
20.
Ceriello
A
,
Hanefeld
M
,
Leiter
L
, et al.
Postprandial glucose regulation and diabetic complications
.
Arch Intern Med
2004
;
164
:
2090
2095
21.
Monnier
L
,
Lapinski
H
,
Colette
C
.
Contributions of fasting and postprandial plasma glucose increments to the overall diurnal hyperglycemia of type 2 diabetic patients: variations with increasing levels of HbA(1c)
.
Diabetes Care
2003
;
26
:
881
885
22.
Bergenstal
RM
,
Beck
RW
,
Close
KL
, et al.
Glucose management indicator (GMI): a New term for estimating A1C from continuous glucose monitoring
.
Diabetes Care
2018
;
41
:
2275
2280
23.
Stephen
AM
,
Cummings
JH
.
Mechanism of action of dietary fibre in the human colon
.
Nature
1980
;
284
:
283
284
24.
Singh
J
,
Dartois
A
,
Kaur
L
.
Starch digestibility in food matrix: a review
.
Trends Food Sci Technol
2010
;
21
:
168
180
25.
Anderson
IH
,
Levine
AS
,
Levitt
MD
.
Incomplete absorption of the carbohydrate in all-purpose wheat flour
.
N Engl J Med
1981
;
304
:
891
892
26.
Calcagno
M
,
Kahleova
H
,
Alwarith
J
, et al.
The thermic effect of food: a review
.
J Am Coll Nutr
2019
;
38
:
547
551
27.
Slavin
J
.
Why whole grains are protective: biological mechanisms
.
Proc Nutr Soc
2003
;
62
:
129
134
28.
Järvi
AE
,
Karlström
BE
,
Granfeldt
YE
,
Björck
IE
,
Asp
NG
,
Vessby
BO
.
Improved glycemic control and lipid profile and normalized fibrinolytic activity on a low-glycemic index diet in type 2 diabetic patients
.
Diabetes Care
1999
;
22
:
10
18
29.
Jenkins
DJ
,
Wesson
V
,
Wolever
TM
, et al.
Wholemeal versus wholegrain breads: proportion of whole or cracked grain and the glycaemic response
.
BMJ
1988
;
297
:
958
960
30.
Haber
GB
,
Heaton
KW
,
Murphy
D
,
Burroughs
LF
.
Depletion and disruption of dietary fibre. Effects on satiety, plasma-glucose, and serum-insulin
.
Lancet
1977
;
2
:
679
682
31.
Jenkins
DJ
,
Wolever
TM
,
Jenkins
AL
, et al.
Low glycemic response to traditionally processed wheat and rye products: bulgur and pumpernickel bread
.
Am J Clin Nutr
1986
;
43
:
516
520
32.
Heaton
KW
,
Marcus
SN
,
Emmett
PM
,
Bolton
CH
.
Particle size of wheat, maize, and oat test meals: effects on plasma glucose and insulin responses and on the rate of starch digestion in vitro
.
Am J Clin Nutr
1988
;
47
:
675
682
33.
Mann
JI
,
De Leeuw
I
,
Hermansen
K
, et al
.
Diabetes and Nutrition Study Group (DNSG) of the European Association. Evidence-based nutritional approaches to the treatment and prevention of diabetes mellitus
.
Nutr Metab Cardiovasc Dis
2004
;
14
:
373
394
34.
Kennedy-Martin
T
,
Curtis
S
,
Faries
D
,
Robinson
S
,
Johnston
J
.
A literature review on the representativeness of randomized controlled trial samples and implications for the external validity of trial results
.
Trials
2015
;
16
:
495
Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More information is available at https://www.diabetesjournals.org/content/license.