Impact of Dietary Fiber Fermentation from Cereal Grains on Metabolite Production by the Fecal Microbiota from Normal Weight and Obese Individuals

Abstract

Gut bacteria may influence obesity through the metabolites produced by dietary fiber fermentation (mainly, short-chain fatty acids [SCFA]). Five cereal grain samples (wheat, rye, maize [corn], rice, and oats) were subjected to in vitro digestion and fermentation using fecal samples from 10 obese and nine normal weight people. No significant differences in total SCFA production between the normal weight and obese groups were observed [279 (12) vs. 280 (12), mean (standard error), respectively; P=.935]. However, the obese microbiota resulted in elevated propionate production compared with that of normal weight [24.8(2.2) vs. 17.8(1.9), respectively; P=.008]. Rye appeared to be particularly beneficial among grain samples due to the lowest propionate production and highest butyrate production during fermentation. These data suggest that the dietary fibers from cereal grains affect bacterial metabolism differently in obese and normal weight classes and that certain grains may be particularly beneficial for promoting gut health in obese states.

Introduction

T he role that human colonic bacteria, or the gut microbiota, plays in health and disease is an area of research with intense current interest. This is fueled by provocative reports suggesting that the gut microbiota contribute to the cause, prevention, or exacerbation of many diseases that plague modern society, including obesity, diabetes, colon cancer, and other diseases.¹ Thus, maintaining a healthy gut microbiota composition is desirable.

Considerable efforts are currently devoted to identifying what constitutes a healthy gut microbiota; however, reports are controversial. For instance, with regard to obesity, some have shown that the ratio of the two major phyla of bacteria in the gut, Firmicutes and Bacteroidetes, shifts in favor of Firmicutes in obese individuals compared with normal weight,^2,3 whereas others have shown increased proportions of Bacteroidetes in obese individuals,^4,5 and still others have shown no significant differences.^6,7 Thus, defining the most desirable human gut microbiota composition is difficult. An important factor to consider may be the influence that metabolites produced by the microbiota have on hormones and cell signaling.⁸

The major microbial metabolites produced by gut bacteria upon dietary fiber fermentation are short-chain fatty acids (SCFA), acetate, propionate, and butyrate, which contribute to various human metabolic processes, including the regulation of hormones involved in energy metabolism, such as leptin,⁹ peptide YY,¹⁰ and glucagon-like peptide-1.¹¹ Branched-chain fatty acids (BCFA) are also metabolic end products of bacterial fermentation but arise from protein fermentation. They are generally regarded as detrimental to health, since they are often accompanied by ammonia and phenol production.¹² Thus, one way the gut microbiota may influence host energy metabolism and health is through their metabolites.

While a wealth of research has focused on isolated dietary fibers and how they contribute to gut health, little research has been conducted on how dietary fibers from whole foods affect the gut microbiota. Cereal grains, for instance, can be rich sources of dietary fibers and other bioactive compounds that may affect the gut microbiota in a positive manner.^13
–15 Thus, the purpose of this study was to compare SCFA production from the fecal microbiota obtained from normal weight and obese individuals upon fermentation of the dietary fibers from several grain samples in an effort to identify which cereals may induce particularly beneficial fermentation profiles relative to reducing obesity.

Materials and Methods

Cereal grain samples

Whole grain brown rice (Oryza sativa L.) was obtained from Riceland Foods, Inc. (Stuttgart, AR, USA). Maize (Zea mays L., referred to herein as corn) and hard red winter wheat (Triticum aestivum L.) were gifts from Stephen Mason (University of Nebraska-Lincoln) and P. Stephen Baenziger (University of Nebraska-Lincoln), respectively. Dark rye flour (Secale cereale L.) containing about 10% added bran and oat groats (Aveba sativa L.) were provided by ConAgra Mills (Omaha, NE, USA) and Grain Millers, Inc. (Minneapolis, MN, USA), respectively. Oats, wheat, and brown rice were milled with a cyclone sample mill (UDY, Boulder, CA, USA) equipped with a 1-mm screen. Corn was milled on a microhammer mill (Glen Mills, Clifton, NJ, USA) equipped with a 2-mm screen. Samples were analyzed for moisture (according to Approved Method 44-15A),¹⁶ starch (K-TSTA, Megazyme, Bray, Ireland), and total dietary fiber (see Total dietary fiber and total carbohydrates).

Fecal samples

Using the variation among fecal fermentations among individuals reported in Kim and White,¹⁷ eight fecal fermentations per group were estimated to detect a 10% difference in SCFA production between fermentations using fecal samples from normal weight individuals (body mass index, BMI, ≤25 kg/m²) and fermentations using fecal samples from obese individuals (BMI ≥30 kg/m²) with 82% power and α=0.05. With an anticipated 20% fecal donor dropout rate, 20 fecal donors were therefore recruited from the Rush University Medical Center's patient database as having no known gastrointestinal disorders and having avoided antibiotics for 3 months before the study (Table 1).

Table 1.

Fecal Sample Donor Information

Patient	BMI	Assigned group	Age	Gender	Race
19	18.4	Normal weight	46	Male	Caucasian
20	20.7	Normal weight	24	Female	Caucasian
3	21.8	Normal weight	32	Female	Caucasian
8	22.1	Normal weight	31	Male	African American
14	22.1	Normal weight	51	Female	African American
2	24.1	Normal weight	44	Female	African American
29	24.1	Normal weight	31	Female	African American
18	25.1	Normal weight	56	Male	African American
7	25.3	Normal weight	56	Female	African American
10	26.1	Excluded	23	Male	African American
9	31.4	Obese	54	Male	African American
15	31.7	Obese	23	Female	African American
21	33.2	Obese	51	Female	Hispanic
5	36.7	Obese	50	Female	African American
4	36.9	Obese	44	Male	African American
6	40.1	Obese	52	Male	African American
16	40.6	Obese	31	Female	Hispanic
17	40.9	Obese	46	Female	Caucasian
11	42.8	Obese	29	Female	African American
23	45.4	Obese	33	Female	African American

BMI: body mass index.

Fecal samples were collected through a specimen hat inserted under the toilet seat, and then immediately transferred to gas-tight bags containing an Anaerocult C strip (BD GasPak, NJ, USA), which created an anaerobic environment. Stools were stored at –80°C until collection was complete and then shipped overnight to the University of Nebraska-Lincoln on dry ice, where they were further stored (–80°C) until fermentation experiments were performed. Using frozen fecal samples collected in this manner produces a similar fermentation profile compared with using fresh fecal samples.¹⁸ Collection of these samples was approved by the Rush University's Institutional Review Board (#10062307-IRB02).

After fecal sample collection, it was found that three donors fell within the overweight category (BMI >25 and <30; subjects 7, 10, and 18; Table 1). Based on our prepower calculation, an n of 3 was not sufficient to merit a separate overweight category. Therefore, subjects 7 and 18 were assigned to the normal weight group because their BMIs rounded to the cutoff value of 25 kg/m², and subject 10 was excluded from the study. Thus, fermentations were performed with nine fecal samples from the normal weight group (BMI 18.4–25.3) and 10 fecal samples from the obese group (BMI 31.4–45.4).

In vitro digestion

The digestion process consisted of a simulated gastric digestion followed by a small intestinal phase following the method of Mishra and Monro¹⁹ with some modifications. Samples (25 g) were boiled for 20 min with 300 mL distilled water in a 500-mL beaker inside another larger beaker that was filled with boiling water. This double boiler set up was necessary to prevent scorching of the flour–water porridge. After cooling to room temperature, ∼8 mL of 1 M HCl was added to the sample to reduce the pH to 2.5. Ten milliliters of 100 mg/mL pepsin (P-700; Sigma, St. Louis, MO, USA) dissolved in 50 mM HCl was added and the mixture was placed on an orbital shaker (MaxQ 7000, 150 rpm; Barnstead, Dubuque, IA, USA) at 37°C for 30 min to achieve the gastric phase. The small intestinal phase was initiated with the addition of 50 mL of 0.1 M sodium maleate buffer (pH=6, containing 1 mM CaCl₂) and ∼20 mL of 1 M NaHCO₃ to bring the pH to 6.9. Fifty milliliters of 125 mg/mL pancreatin (P-7545, Sigma) dissolved in the sodium maleate buffer and 2 mL of amyloglucosidase (3260 U/mL; Magazyme) were then added and samples were incubated in a shaking water bath at 37°C for 6 h. Digested contents were then poured into dialysis tubing (molecular weight cutoff 12,000–14,000) and dialyzed for 3 days against distilled water with changing of the water every 12 h. The retentate was then frozen (–20°C) overnight and then freeze dried. The freeze dried material was analyzed for total starch (K-TSTA, Megazyme) and total carbohydrates (see Total dietary fiber and total carbohydrates).

In vitro fermentation

Samples after in vitro digestion containing 40 mg of total carbohydrates (sum of neutral sugar and uronic acid residues) were suspended in a sterile nutrient basal medium²⁰ to a final concentration of 10 mg/mL and then hydrated overnight at 4°C. Fecal samples were then taken from the freezer and defrosted in an anaerobic hood (Bactron IV; Sheldon Manufacturing, Cornelius, OR, USA) containing 5% H₂, 5% CO₂, and 90% N₂. Defrosted samples were weighed and mixed with sterile phosphate-buffered saline at 100 mg/mL (w/v),²⁰ blended for 30 sec using a kitchen blender (2774 heritage series; Sunbeam Company, Boca Raton, FL, USA), and then filtered through four layers of cheesecloth. Tubes were then inoculated with 0.4 mL of fecal slurry, capped, placed at a 45° angle, and incubated at 37°C with shaking (MaxQ 7000, 140 rpm). A sample containing only the basal medium and fecal suspension was included as a control and a portion of fecal slurry was retained for a zero time reading. Samples were taken at 12 h by plunging in an ice bath and then freezing (–80°C) until analysis.

Due to the large number of samples, only one time point during fermentation was analyzed. Twelve hours of fermentation was selected based on previous recommendations as a reasonable time to evaluate the effects of different carbohydrates on bacterial metabolism and composition.²⁰ Before taking samples, the volume of gas produced by the bacteria was measured by inserting a lubricated glass syringe with a needle through a septum in the cap of the tube.

Total dietary fiber and total carbohydrates

Total dietary fiber and total carbohydrates were analyzed before and after in vitro digestion, respectively. The dietary fiber content was determined as the sum of neutral sugar residues, uronic acid residues, and Klason lignin; total carbohydrates were the sum of neutral sugar and uronic acid residues. Neutral sugars and uronic acids were quantified as described.¹⁶ Klason lignin was determined as described by Theander et al. ²¹

Short-chain fatty acids

SCFA were quantified based on Campbell et al.²² Briefly, 1 mL of sample was removed from storage at –80°C and thawed at room temperature. The samples were then acidified with 0.25 mL of 50 mg/mL metaphosphoric acid containing 5–10 mM 4-methylvaleric acid (Alfa Aesar) as an internal standard. The acidified tubes were vortex mixed and centrifuged for 10 min at 16,100 g. The supernatant was stored overnight at –20°C, and then thawed and centrifuged in the same condition as before. Four microliters of supernatant was then injected onto a gas chromatograph (Clarus 580; PerkinElmer, Waltham, MA, USA), and SCFA were separated on a capillary column (Elite-FFAP, 15 m×0.25 mm inner diameter×0.25 μm film thickness, PerkinElmer) and detected with a flame ionization detector. For quantification, response factors for acetate, propionate, butyrate, iso-butyrate, and iso-valerate relative to 4-methyl valeric acid were determined by injecting pure standard mixtures (Sigma).

Statistical analysis

Data were analyzed using SAS software (version 9.2; SAS Institute, Cary, NC, USA). Sample size calculations were performed using proc POWER with means and standard deviations from in vitro fermentations of whole grain oats¹⁷ and assuming a 10% difference between obese and normal weight. Each of the grain samples and a blank (5+1) were fermented with each of the nine (normal weight) or ten (obese) fecal samples. Analyses were performed in duplicate. A general linear analysis of variance (PROC GLM) was used with grain type, weight group, and their interaction as factors with correction for gender, race, and age. The Fisher's least significant difference test was performed to determine significant differences among factors with significant F-tests. Where significant differences between the normal weight and obese groups were detected, the obese group was further divided into class I (BMI: 30–34.9), class II (BMI: 35–39.9), and class III (BMI: ≥40), and the analysis was rerun to determine the differences among classes of obesity. Partial correlation of SCFA and BCFA with BMI was performed using the “proc CORR” command in SAS, with corrections for age, gender, and race. Significance was defined as P<.05.

Results and Discussion

Sample composition and in vitro digestion

The dietary fiber contents of the five cereal grains are shown in (Table 2). These values were similar to those reported in the National Nutrient Database for Standard Reference provided by the US Department of Agriculture.²³ The starch contents were typical of those reported previously for whole grains,^24
–26 except for the rye that was lower due to the added bran (∼10%). Other components in the starting materials were assumed to be protein (9–18%), lipid (2–8%), lignin (2–10%), ash (2–4%), and other compounds typically associated with cereal grains.²⁶

Table 2.

Polysaccharide Composition of Cereal Grain Samples

		Dietary fiber
Sample	Starch	Neutral sugars	Uronic acids	Lignin	Total
Corn	63.3±0.1	6.60±0.82	0.43±0.04	1.49±0.23	8.51±0.92
Oats	59.5±0.8	7.24±0.10	1.14±0.08	2.38±0.00	9.90±0.16
Rice	75.4±0.3	2.35±0.21	0.58±0.04	1.51±0.20	4.01±0.37
Rye	33.6±0.7	23.7±1.6	0.51±0.08	4.74±0.20	28.9±1.56
Wheat	59.5±0.1	9.22±0.72	1.29±0.10	2.04±0.03	11.6±0.71

Values are reported as % dry basis. Mean±standard deviation of two replicates. All samples were received as whole grain (oats and rice were dehulled) except rye; rye was obtained as a flour that contained about 10% added bran.

After in vitro digestion, the starch content was <4.5% on all five substrates (data not shown). Considering that the resistant starch content of whole grain is about 2–10%,^24,25 it was concluded that the in vitro digestion procedure was successful at removing the digestible starch that would have otherwise interfered with the in vitro fermentation results.²⁷

The nonstarch carbohydrate content after in vitro fermentation in each of the five cereal grain samples was (% dry basis) corn, 33.1±0.7; oats, 28.5±0.1; rice, 15.1±1.3; rye, 47.5±1.6; and wheat, 39.7±0.6. These values were similar to Karppinen et al.,²⁸ who reported 33.3–51.0% nonstarch carbohydrate in digestion residues before in vitro fermentation. The low nonstarch carbohydrate content in the digested rice sample may have been due to the low starting dietary fiber content (Table 2). The remainder of the digestion residues was likely concentrated amounts of noncarbohydrate grain components (undigested proteins, lipids, lignins, and ash) as well as residual proteins from the in vitro digestion procedure.²⁹

In vitro fermentation

Total SCFA production varied substantially among individuals and grain types (Fig. 1). A slight increase in SCFA was observed in the obese group as a function of BMI (P=.0298); however, after correcting for age, gender, and race, SCFA were no longer correlated with BMI (P=.0972). When categorized into weight groups (normal weight and obese), the fecal bacteria collected from normal weight and obese individuals produced similar amounts of total SCFA (Table 3). Others have shown enhanced SCFA concentrations in the ceca of obese mice³⁰ and in the feces of humans⁴ compared to normal weight. These reports suggest that SCFA are undesirable because they are more prevalent as BMI increases. This, however, is unlikely since many studies have shown the benefits of SCFA^10,11 and dietary fibers^31,32 on human health and reduction of obesity.

FIG. 1.

Short-chain fatty acid produced during in vitro fermentation of selected cereal grain samples as a function of the body mass index (BMI) of the fecal donor.

Table 3.

Short- and Branched-Chain Fatty Acid Production During 12 h of In Vitro Fecal Fermentation

		SCFA (μmol/100 mg CHO)				BCFA (μmol/100 mg CHO)
Main effect	Gas (mL/tube)	Acetate	Propionate	Butyrate	Total	Iso-butyrate	Iso-valerate	Total
Weight group
Normal	10.8 (0.5)	253 (11)	17.8 (1.9)^b	8.57 (0.98)	279 (12)	3.34 (0.37)	11.0 (0.87)	14.3 (1.2)
Obese	10.8 (0.5)	249 (11)	24.8 (2.2)^a	6.68 (0.90)	280 (12)	2.79 (0.27)	10.6 (0.88)	13.4 (1.1)
Grain
Blank	3.77 (0.16)^d	115 (5)^d	8.80 (1.61)^d	6.01 (0.88)^c	130 (6)^d	3.19 (0.53)^ab	7.13 (0.95)^d	10.3 (1.4)^c
Corn	10.4 (0.6)^c	207 (10)^c	23.3 (3.9)^bc	2.88 (0.20)^c	234 (11)^c	2.39 (0.46)^b	8.61 (1.19)^cd	11.0 (1.6)^bc
Oats	11.7 (0.7)^b	267 (11)^b	25.9 (4.0)^b	6.37 (1.63)^bc	299 (11)^b	2.72 (0.49)^b	12.7 (1.4)^ab	15.5 (1.7)^ab
Rice	15.2 (0.9)^a	414 (20)^a	40.8 (4.5)^a	6.11 (1.08)^bc	461 (21)^a	4.19 (0.78)^a	15.4 (2.4)^a	19.6 (3.1)^a
Rye	12.0 (0.6)^b	250 (10)^b	11.9 (1.6)^d	16.4 (2.6)^a	278 (12)^b	3.35 (0.55)^ab	11.2 (1.3)^bc	14.5 (1.7)^a–c
Wheat	11.9 (0.7)^b	251 (9)^b	17.0 (2.3)^cd	8.04 (1.40)^b	276 (11)^b	2.54 (0.41)^b	9.73 (1.11)^b–d	12.3 (1.4)^bc

Reported as mean (standard error); values carrying different superscript letters within column and effect are significantly different (P<.05); weight group×grain interaction was not significant (P>.05) for any of the response variables; data were corrected for differences in age, gender, and race.

SCFA, short-chain fatty acid; BCFA, branched-chain fatty acid; CHO, carbohydrate.

Rather, the types of SCFA produced during fermentation may be important to consider. The obese group showed enhanced propionate production compared with the normal weight group (Table 3). No significant differences in acetate, butyrate, or BCFA production were evident. The increase in propionate production in the obese group was reflected in a significant increase in the molar propionate: total SCFA ratio in the obese group compared with normal weight [0.09 (0.01) vs. 0.06 (0.01), mean (standard error), respectively, P=.0003], accompanied by a proportional decrease in the acetate: SCFA ratio [0.88 (0.01) vs. 0.90 (0.01), respectively, P=.016]. Propionate production was also correlated with BMI, which remained significant even after correcting for age, gender, and race (P=.0069). When the obese group was further divided into class I, class II, and class III, propionate production in the type III obese group was 28.1 μmol/100 mg carbohydrates versus 18.1 μmol/100 mg carbohydrates in the normal weight group (P=.0015).

In agreement with our results, others have also shown that propionate is enhanced in the feces of obese individuals compared to normal weight.⁴ This was surprising, as propionate is generally considered to be a desirable metabolite.³³ The changes in microbiota upon fermentation were not quantified; however, Schwiertz et al. ⁴ attributed the enhanced propionate concentrations in the feces of obese individuals to an increase in Bacteroides. Interestingly, Bacteroides possesses far more genes that encode for carbohydrate-degrading enzymes than other gut bacteria and can metabolize host-derived mucins when host dietary polysaccharides (i.e., dietary fibers) are lacking.³⁴ Considering that epidemiological reports consistently show that obese adults consume less dietary fiber-rich foods than their normal-weight counterparts,³⁵ it may be that Bacteroides thrive in an obese situation because they are best equipped to adapt to inadequate food supplies. Importantly, fermentative processes that take place under such circumstances (low dietary fiber intake) result in undesirable fermentation products and may have a negative impact on health.¹² Thus, a microbial composition that results in high propionate production may be undesirable not because propionate itself is detrimental, but because the bacteria that induce this environment or thrive therein result in undesirable effects on the host.

Given these observations, to help reduce obesity, our goal may be to reduce propionate production or to reduce the number of propiogenic bacteria in the gut. Additionally, although our results did not show impaired butyrate production by the fecal microbiota from obese individuals (Table 3), it would likely be of interest to increase the production of this metabolite due to its established trophic effects,³⁶ particularly against inflammation in the case of obesity, which is now considered an inflammatory disease as a consequence of metabolic syndrome.⁸

Accordingly, rye appears to be among the more beneficial grains examined in this study, resulting in the highest butyrate production accompanied by the lowest propionate production (Table 3). The observation that rye may be particularly beneficial against obesity is in support of a mouse study showing that rye induces greater reduction in body weight and adiposity compared with wheat.³⁷ One unique aspect of rye is its high water-extractable arabinoxylan content relative to other grains.³⁸ Isolated water-extractable arabinoxylan has been shown to reduce body weight gain, fat storage, and cholesterol in high-fat-fed mice.³²

It is worth mentioning that the rye used in this study contained ∼10% added bran, whereas the other samples were used as whole grains. The added bran resulted in the highest total carbohydrates (i.e., dietary fiber) in the sample going into the in vitro fermentation (Table 2). However, our data were corrected for total initial carbohydrates (Fig. 1; Table 3); therefore, the effects of this were likely minimal. Moreover, even if analyzed on a total weight basis, rather than a carbohydrate basis, these conclusions are still valid (rye resulting in the highest butyrate production and the lowest propionate production; data not shown).

These data provide useful information for designing studies on the effects of cereal grains on obesity. More research is needed on how cereal grains, rye in particular, influence gut microbiota populations and how this impacts physiology of the host.

Footnotes

Acknowledgment

This work was supported by a grant from the Agricultural Research Division of the University of Nebraska-Lincoln.

Author Disclosure Statement

The authors declare no conflict of interests.

References

Nicholson

, Holmes

, Kinross

et al. Host-gut microbiota metabolic interactions. Science, 2012; 336:1262–1267.

Ley

, Turnbaugh

, Klein

, Gordon

. Microbial ecology: human gut microbes associated with obesity. Nature, 2006; 444:1022.

Turnbaugh

, Hamady

, Yatsunenko

et al. A core gut microbiome in obese and lean twins. Nature, 2008; 457:480–484.

Schwiertz

, Taras

, Schäfer

et al. Microbiota and SCFA in lean and overweight healthy subjects. Obesity, 2009; 18:190–195.

Zhang

, DiBaise

, Zuccolo

et al. Human gut microbiota in obesity and after gastric bypass. Proc Natl Acad Sci USA, 2009; 106:2365–2370.

Duncan

, Lobley

, Holtrop

et al. Human colonic microbiota associated with diet, obesity and weight loss. Int J Obesity, 2008; 32:1720–1724.

Arumugam

, Raes

, Pelletier

et al. Enterotypes of the human gut microbiome. Nature, 2011; 473:174–180.

Delzenne

, Cani

. Interaction between obesity and the gut microbiota: relevance in nutrition. Annu Rev Nutr, 2011; 31:15–31.

Soliman

, Ahmed

, Salah-eldin

et al. Butyrate regulates leptin expression through different signaling pathways in adipocytes. J Vet Sci, 2011; 12:319–323.

10.

Zhou

, Hegsted

, McCutcheon

et al. Peptide YY and proglucagon mRNA expression patterns and regulation in the gut. Obesity, 2006; 14:683–689.

11.

Tolhurst

, Heffron

, Lam

et al. Short-chain fatty acids stimulate glucagon-like peptide-1 secretion via the G-protein–coupled receptor FFAR2. Diabetes, 2012; 61:364–371.

12.

Russell

, Gratz

, Duncan

et al. High-protein, reduced-carbohydrate weight-loss diets promote metabolite profiles likely to be detrimental to colonic health. Am J Clin Nutr, 2011; 93:1062–1072.

13.

Costabile

, Klinder

, Fava

et al. Whole-grain wheat breakfast cereal has a prebiotic effect on the human gut microbiota: a double-blind, placebo-controlled, crossover study. Brit J Nutr, 2008; 99:110.

14.

Carvalho-Wells

, Helmolz

, Nodet

et al. Determination of the in vivo prebiotic potential of a maize-based whole grain breakfast cereal: a human feeding study. Brit J Nutr, 2010; 104:1353–1356.

15.

Connolly

, Lovegrove

, Tuohy

. In vitro fermentation characteristics of whole grain wheat flakes and the effect of toasting on prebiotic potential. J Med Food, 2012; 15:33–43.

16.

AACC International. Approved Methods of Analysis, 11th. AACC International: St. Louis, MO, 2012methods 32–25.01 and 44–15.02.

17.

Kim

, White

. In vitro fermentation of oat flours from typical and high β-glucan oat lines. J Agric Food Chem, 2009; 57:7529–7536.

18.

Rose

, Venema

, Keshavarzian

, Hamaker

. Starch-entrapped microspheres show a beneficial fermentation profile and decrease in potentially harmful bacteria during in vitro fermentation in faecal microbiota obtained from patients with inflammatory bowel disease. Brit J Nutr, 2010; 103:1514.

19.

Mishra

, Monro

. Digestibility of starch fractions in wholegrain rolled oats. J Cereal Sci, 2009; 50:61–66.

20.

Sanz

, Gibson

, Rastall

. Influence of disaccharide structure on prebiotic selectivity in vitro. J Agric Food Chem, 2005; 53:5192–5199.

21.

Theander

, Aman

, Westerlund

et al. Total dietary fiber determined as neutral sugar residues, uronic acid residues, and Klason lignin (the Uppsala method): collaborative study. J AOAC Int, 1995; 78:1030.

22.

Campbell

, Fahey

, Wolf

. Selected indigestible oligosaccharides affect large bowel mass, cecal and fecal short-chain fatty acids, pH and microflora in rats. J Nutr, 1997; 127:130–136.

23.

USDA: National Nutrient Database for Standard Reference. http://ndb.nal.usda.gov/. 2012 November 30.

24.

Englyst

, Liu

, Englyst

. Nutritional characterization and measurement of dietary carbohydrates. Eur J Clin Nutr, 2007; 61:S19–S39.

25.

Englyst

, Englyst

. Carbohydrate bioavailability. Brit J Nutr, 2005; 94:1–11.

26.

Welch

. Nutrient composition and nutritional quality of oats and comparisons with other cereals. Oats: Chemistry and Technology, 2nd Webster

, Wood

. AACC International: St. Louis, MO, 2011; 95–107.

27.

Lebet

, Arrigoni

, Amado

. Digestion procedure using mammalian enzymes to obtain substrates for in vitro fermentation studies. LWT-Food Sci Tech, 1998; 31:509–515.

28.

Karppinen

, Liukkonen

, Aura

et al. In vitro fermentation of polysaccharides of rye, wheat and oat brans and inulin by human faecal bacteria. J Sci Food Agric, 2000; 80:1469–1476.

29.

Zhang

, Ao

, Hamaker

. Nutritional property of endosperm starches from maize mutants: A parabolic relationship between slowly digestible starch and amylopectin fine structure. J Agric Food Chem, 2008; 56:4686–4694.

30.

Turnbaugh

, Ley

, Mahowald

et al. An obesity-associated gut microbiome with increased capacity for energy harvest. Nature, 2006; 444:1027–1131.

31.

Parnell

, Reimer

. Weight loss during oligofructose supplementation is associated with decreased ghrelin and increased peptide YY in overweight and obese adults. Am J Clin Nutr, 2009; 89:1751–1759.

32.

Neyrinck

, Possemiers

, Druart

et al. Prebiotic effects of wheat arabinoxylan related to the increase in Bifidobacteria, Roseburia and Bacteroides/Prevotella in diet-induced obese mice. PLoS One, 2011; 6:e20944.

33.

Hosseini

, Grootaert

, Verstraete

et al. Propionate as a health-promoting microbial metabolite in the human gut. Nutr Rev, 2011; 69:245–258.

34.

Sonnenburg

, Xu

, Leip

et al. Glycan foraging in vivo by an intestine-adapted bacterial symbiont. Science, 2005; 307:1955–1959.

35.

Mckeown

, Hruby

, Saltzman

et al. Weighing in on whole grains: a review of evidence linking whole grains to body weight. Cereal Food World, 2012; 57:20–27.

36.

Hamer

, Jonkers

, Venema

et al. Review article: the role of butyrate on colonic function. Aliment Pharm Therap, 2008; 27:104–119.

37.

Andersson

, Rosén

, Östman

et al. Metabolic effects of whole grain wheat and whole grain rye in the C57BL/6J mouse. Nutrition, 2010; 26:230–239.

38.

Rose

. Autohydrolytic production of feruloylated arabinoxylan hydrolysates from cereal processing coproducts for food applications. Advances in Cereal Science: Implications to Food Processing and Health Promotion. Awika

, Piironen

, Bean

. American Chemical Society: Washington, DC, 2011; 111–130.