Molecular Weight of β‐Glucan Affects Physical Characteristics,In Vitro Bile Acid Binding,and Fermentation of Muffins

Muffins containing different amounts and molecular weights (MW) of β‐glucan were evaluated for the effect of β‐glucan on the physical characteristics of the muffins and on in vitro bile acid binding and fermentation with human fecal flora. Wheat flour muffins were prepared with the addition of β‐glucan extracts with high‐, medium‐, or low‐MW. For oat flour muffins, the native oat flour contained high‐MW β‐glucan; the oat flours were treated to create medium‐ and low‐MW β‐glucan within the prepared muffin treatments. For each 60‐g muffin, the amounts of β‐glucan were 0.52, 0.57, and 0.59 g for high‐, medium‐, and low‐MW β‐glucan wheat flour muffins, and 2.38, 2.18, and 2.23 g for high‐, medium‐, and low‐MW β‐glucan oat flour muffins, respectively. The lower the MW of the β‐glucan in muffins, the lower the height and volume of the muffins. The oat flour muffins were less firm and springy than the wheat flour muffins as measured on a texture analyzer; however, MW had no effect on muffin texture. The oat flour muffins bound more bile acid than did the wheat flour muffins. The muffins with high‐MW β‐glucan bound more bile acid than did those with low‐ and medium‐MW β‐glucan. Muffin treatment affected the formation of gas and total short‐chain fatty acids (SCFA) compared with the blank without substrate during in vitro fermentation. There were no differences in pH changes and total gas production among muffin treatments. The high‐MW β‐glucan wheat flour muffins produced greater amounts of SCFA than did the wheat flour muffin without β‐glucan and the oat flour muffins; however, there were no differences in SCFA production among muffins with different MW. In general, the β‐glucan MW affected the physical qualities of muffins and some potential biological functions in humans.     

In Vitro Binding of Bile Acids by Rice Bran,Oat Bran,Barley and β‐Glucan Enriched Barley

The in vitro bile acid binding by rice bran, oat bran, dehulled barley, and β‐glucan enriched barley was determined using a mixture of bile acids at a duodenal physiological pH of 6.3. Six treatments and two blank incubations were conducted testing substrates on an equal protein basis. The relative in vitro bile acid binding of the cereal brans on an equal total dietary fiber (TDF) and insoluble dietary fiber (IDF) basis considering cholestyramine as 100% bound was rice bran 45 and 49%; oat bran 23 and 30%; dehulled barley 33 and 57%; and β‐glucan enriched barley 20 and 40%, respectively. Bile acid bindings on equal protein basis for the respective cereals were 68, 26, 41, and 49%. Bile acid binding by rice bran may account to a great extent for its cholesterol‐lowering properties, while bile acid binding by oat bran suggests that the primary mechanism of cholesterol lowering by oat bran is not due to the bile acid binding by its soluble fiber. Bile acid binding was not proportional to the soluble fiber content of the cereal brans tested. Except for dehulled barley, bile acid binding for rice bran, oat bran, and β‐glucan enriched barley appear to be related to their IDF content. Highest relative bile acid binding values for rice bran and β‐glucan enriched barley were observed on an equal protein basis, whereas highest values for dehulled barley were based on IDF. Data suggest that of all four cereals tested, bile acid binding may be related to IDF or protein anionic, cationic, physical and chemical structure, composition, metabolites, or their interaction with active binding sites.     

Gastrointestinal Release of β‐Glucan and Pectin Using an In Vitro Method

The release of soluble dietary fiber is a prerequisite for viscous effects and hence beneficial health properties. A simple in vitro method was adapted to follow the release during gastrointestinal digestion, and the percentage of solubilized fiber was measured over time. β‐Glucan from oat bran was mainly released during gastric digestion while the release of pectin from sugar beet fiber continued in the small intestine. Unmilled fractions of sugar beet fiber released more soluble fiber than oat bran flakes, probably due to the porous structure of sugar beet fiber as a result of manufacturing processes, but also due to differences in source. Milling to smaller fiber particles significantly improved releasability (from 20 to 55% released β‐glucan and from 50 to 70% released pectin, respectively, after digestion). When milled fibers were included in individual food matrices, the release was reduced by protein and starch matrices (5% β‐glucan and 35% pectin released, respectively) and slowed by fat (45% β‐glucan and 60% pectin released). This may result in a too low or too late release in the upper small intestine to be able to interfere with macronutrient uptake. The method may be suitable for predicting the gastrointestinal release of soluble dietary fibers from food matrices in the development of healthy food products.     

Whole Grains and Digestive Health

Whole grains contain all parts of the grain: the endosperm, germ, and bran. Whole grains are rich in fermentable carbohydrates that reach the gut: dietary fiber, resistant starch, and oligosaccharides. Most research that supports the importance of grains to gut health was conducted with isolated fiber fractions, rather than whole grains. Whole grains are an important source of dietary fiber and grain fibers such as wheat, oats, barley, and rye increase stool weight, speed intestinal transit, get fermented to short chain fatty acids, and modify the gut microflora. Wheat bran is particularly effective in increasing stool weight; wheat bran increases stool weight by a ratio of 5:1. In contrast, many novel fibers that are easily incorporated into beverages and foods increase stool weight only on a ratio of 1:1. In vitro fermentation studies with whole grains have been published. Carbohydrates of oat bran (rich in β‐glucan) were consumed by bacteria faster than those of rye and wheat brans (rich in arabinoxylan). Grain fibers were fermented more slowly than inulin, causing less gas production. Wheat is particularly high in fructo‐oligosaccharides, while wheat germ is high in raffinose oligosaccharides. Some in vivo studies show the prebiotic potential of whole grains. Whole grain breakfast cereal was more effective than wheat bran breakfast cereal as a prebiotic, increasing fecal bifidobacteria and lactobacilli in human subjects. Wheat bran consumption increased stool frequency. Thus, the gut enhancing effects of cereal fibers are well known. Limited data exist that whole grains alter gut health.     

Molecular Weight Distribution of β‐Glucan in Oat‐Based Foods

Oats, different oat fractions as well as experimental and commercial oat‐based foods, were extracted with hot water containing thermostable α‐amylase. Average molecular weight and molecular weight distributions of β‐glucan in extracts were analyzed with a calibrated high‐performance size‐exclusion chromatography system with Calcofluor detection, specific for the β‐glucan. Oats, rolled oats, oat bran, and oat bran concentrates all had high Calcofluor average molecular weights (206 × 10⁴ to 230 × 10⁴ g/mol) and essentially monomodal distributions. Of the oat‐containing experimental foods, extruded flakes, macaroni, and muffins all had high average molecular weights. Pasteurized apple juice, fresh pasta, and teacake, on the other hand, contained degraded β‐glucan. Calcofluor average molecular weights varied from 24 × 10⁴ to 167 × 10⁴ g/mol in different types of oat bran‐based breads baked with almost the same ingredients. Large particle size of the bran and short fermentation time limited the β‐glucan degradation during baking. The polymodal distributions of β‐glucan in these breads indicated that this degradation was enzymatic in nature. Commercial oat foods also showed large variation in Calcofluor average molecular weight (from 19 × 10⁴ g/mol for pancake batter to 201 × 10⁴ g/mol for porridge). Boiling porridge or frying pancakes did not result in any β‐glucan degradation. These large differences in molecular weight distribution for β‐glucan in different oat products are very likely to be of nutritional importance.     

Composition and Utilization of Barley Pearling By‐Products for Making Functional Pastas Rich in Dietary Fiber and β‐Glucans

Pearling by‐products and the pearled products of two commercial stocks of hulled barley, pearled according to an industrial process consisting of five consecutive pearling steps, were analyzed for β‐glucans, dietary fiber (total, soluble, and insoluble), protein, lipid, ash, and digestible carbohydrate. The data showed that the pearling flour fractions, abraded in the fourth and fifth hullers, contained interesting amounts of β‐glucans (3.9–5.1% db) from a nutritional point of view. These fractions were subsequently enriched in β‐glucans using a milling‐sieving process to double β‐glucan content (9.1–10.5% db). Functional pastas, enriched with β‐glucans and dietary fiber, were produced by substituting 50% of standard durum wheat semolina with β‐glucan‐enriched barley flour fractions. Although darker than durum wheat pasta, these pastas had good cooking qualities with regard to stickiness, bulkiness, firmness, and total organic matter released in rinsing water. The dietary fiber (13.1–16.1% wb) and β‐glucan (4.3–5.0% wb) contents in the barley pastas were much higher than in the control (4.0 and 0.3% wb, respectively). These values amply meet the FDA requirements of 5 g of dietary fiber and 0.75 g of β‐glucans per serving (56 g in the United States and 80 g in Italy). At present, the FDA has authorized the health claim “may reduce the risk of heart disease” for food containing β‐glucans from oat and psyllium only.     

Oat Bran Fermentation by Rye Sourdough

Hydration of oat bran including fermentation by rye sourdough was studied. Three types of oat bran suspensions were prepared (a control, one with whole meal rye flour added, and one with rye starter added). The suspensions were incubated for 1, 2, 3 and 4 hr. β‐Glucan content and solubilities of protein and β‐glucan were analyzed. Viscosity of the supernatants of oat bran suspensions was determined. Neither the rye sourdough nor the rye flour alone had a significant effect on the total β‐glucan content of oat bran suspensions. However, the addition of rye, either as whole meal rye flour or as sourdough starter, markedly increased the solubility of β‐glucan and proteins and simultaneously decreased the viscosity of the water‐soluble fraction of oat bran suspension. This suggests that a hydrolysis of β‐glucan had occurred that could change the rheological properties of oat bran in baking and the physiological potential of oat bran in nutrition.     

Prediction of β‐Glucan Concentration Based on Viscosity Evaluations of Raw Oat Flours from High β‐Glucan and Traditional Oat Lines

Rheological properties of raw oat flour slurries were determined in experimental high β‐glucan (≤7.8%) and traditional oat lines (4–5% β‐glucan) grown in two consecutive years. Three different media were used to disperse oat flours: deionized water, silver nitrate solution (to inactivate endogenous enzymes), and alkali solution (to solubilize both water‐soluble and water‐insoluble β‐glucans). Significant correlations (P < 0.05) between viscosity of slurries and β‐glucan concentration obtained in either deionized water (r = 0.833), silver nitrate (r = 0.940), or alkali (r = 0.896) solutions showed that β‐glucans were the main contributor to oat extract viscosity. The highest correlation was obtained in silver nitrate solution, suggesting that inactivating endogenous enzymes is important to obtain high correlations. Predictive models of oat β‐glucan concentration based on the viscosity profile were developed using partial least squares (PLS) regression. Prediction of β‐glucan concentration based on viscosity was most effective in the silver nitrate solution (r = 0.949, correlation coefficient of predicted vs. analyzed β‐glucans) and least effective in the alkali solution (r = 0.870). These findings demonstrate that the β‐glucan in oat could be predicted by measuring the viscosity of raw flours in silver nitrate solution, and this method could be used as a screening tool for selective breeding.     

Extraction and Physicochemical Characterization of Rye β‐Glucan and Effects of Barium on Polysaccharide Molecular Weight

Use of saturated Ba(OH)₂ to extract rye β‐glucan led to a depolymerized product. Similar depolymerization of β‐glucan was observed when oat bran was extracted with this reagent. Isolated oat β‐glucan, detarium xyloglucan, guar galactomannan, and wheat and rye arabinoxylan were also depolymerized by treatment with the barium reagent. The degree of depolymerization was related to time of contact with, and concentration of, the barium. Rye β‐glucan of two different molecular weights (MW) were isolated and characterized. The structure of rye β‐glucan, as evaluated from the ratio of (1→3)‐linked cellotriosyl to (1→3)‐linked cellotetraosyl primary structural units, most closely resembles barley β‐glucan. Analytical variability of this ratio is discussed. A freshly prepared solution (2%) of the higher MW sample showed shear thinning behavior typical of cereal β‐glucans. The lower MW sample at 2% was not shear thinning, but on further purification, after storage for seven days, a 6% solution had gelled as shown by the mechanical spectrum.     

Content and Molecular Weight Distribution of Oat β‐Glucan in Oatrim,Nutrim, and C‐Trim Products

The soluble fiber, β‐glucan, in oat products is an active hypolipidemic component that is responsible for lowering plasma lipids. Quantitative analysis of β‐glucan in oat hydrocolloids such as Oatrim, Nutrim, and C‐Trim was performed to measure the total β‐glucan content and molecular weight distribution. For the measurement of total β‐glucan content, both modified flow‐injection analysis (FIA) method and the standard AACC enzymatic method were employed. FIA method uses the enhanced fluorescence produced when β‐glucan forms complexes with Calcofluor. Experimental results of both the modified FIA method and the standard AACC enzymatic method revealed very good coincidence with each other. This result confirms the applicability of either technique for the quantitative evaluation of β‐glucan in hydrocolloids. Molecular weight (MW) distribution of β‐glucan was determined by size‐exclusion chromatography with postcolumn detection. Experimental results revealed that the molecular weight of β‐glucan in the Trim products was decreased during the manufacturing process. This result was ascribed to the rigorous processing condition of jet‐cooking.     

REVIEW: Oat and Rye β‐Glucan: Properties and Function

Over the years, the β‐glucan of oats and barley has been the subject of study either because of the importance of the cholesterol‐lowering potential to health claims (FDA 1997, 2005) or, in the case of barley, because of the role of β‐glucan and β‐glucan‐rich endosperm cell walls in malting and brewing. β‐Glucan is also present in rye and in much lesser amounts in wheat. The most striking difference in these latter two sources is the difficulty in extractability; alkali rather than water is required for significant release from the cell walls. This review will discuss physicochemical properties of oat and rye β‐glucan and, where information allows, relate these to physiological effects. Viscosity, or more generally rheology, plays a central role in discussions of cereal β‐glucan functionality and physiological effects and will be the focus of this review.     

Pasting and Thermal Properties of Flours from Oat Lines with High and Typical Amounts of β‐Glucan

Experimental oat lines high in β‐glucan (6–7.8%) and traditional lines (3.9–5.7% β‐glucan) were used to evaluate the effect of β‐glucan on pasting (by rapid viscoanalysis) and thermal properties (by differential scanning calorimetry) of oat flours. Significant correlations established between β‐glucan concentration and the pasting parameters after amylolysis demonstrated the role of β‐glucans in oat pasting. The relative decrease of peak viscosity (PV) observed after enzymatic removal of β‐glucans was correlated with β‐glucan concentration (r = 0.880, P < 0.010) and reconfirmed their contribution to pasting. A significant increase of PV with β‐glucan concentration obtained under conditions of either autolysis (deionized water used for dispersion) (r = 0.89, P < 0.010) or inhibition (silver nitrate solution used for dispersion) (r = 0.91, P < 0.001) might be explained by an increase in water retention capacity caused by the β‐glucans. Predictive models of β‐glucan concentration based on the whole pasting profile or selected profile regions were developed using partial least squares (PLS) regression.Prediction of β‐glucan based on the whole profile obtained in the silver nitrate solution was the most effective (r = 0.93, correlation coefficient of predicted vs. analyzed β‐glucans, P < 0.050). No correlations were observed between the thermal properties of oat flours and the β‐glucan concentration.     

Effect of Processing on Viscosity and Molecular Weight of (1,3)(1,4)‐β‐Glucan in Western Australian Oat Cultivars

Six Australian milling oat cultivars grown over two growing seasons were characterized for differences in (1,3)(1,4)‐β‐glucan (β‐glucan) viscosity, solubility, molecular weight (M_w), and the effect of processing. Oat cultivars grown in 2012 had significantly higher extracted β‐glucan viscosity from oat flour than the same oat cultivar grown in 2011 (P < 0.05, mean 137 and 165 cP, respectively). Noodle β‐glucan mean viscosity for 2012 (147 cP) was significantly higher than for 2011 (128 cP). β‐Glucan from ‘Williams’ and ‘Mitika’ oats had the highest viscosity (P < 0.05) in flour (5.92 and 5.25%, respectively) and noodles (1.64 and 1.47%, respectively) for both years, compared with the other oat cultivars. β‐Glucan (M_w) of Williams for 2012 and ‘Kojonup’ for both years were the least affected by processing, with an average drop of 33% compared with a maximum of 63% for other cultivars. Therefore, Williams showed superior β‐glucan properties to other oat cultivars studied, and can potentially provide improved health benefits. High and low β‐glucan M_w populations were found in the same elution peak after processing. Oat cultivars chosen for processing should be those with β‐glucans that are more resistant to processing, and that maintain their physiochemical properties and, therefore, bioactivity.     

Distribution of β‐Glucan in the Grain of Hull‐less Barley

Nine hull‐less barley (HB) containing waxy (0–7% amylose), normal (≈25% amylose), or high amylose (≈42% amylose) starch with normal or fractured granule make‐up and 4–9% (1→3)(1→4)‐β‐d ‐glucans (β‐glucan) were pearled to remove 70% of the original grain weight in 10% intervals. The pearled fractions were analyzed for β‐glucan distribution within HB grain. Protein content of the pearled fractions indicated that the three outermost fractions contained pericarp and testa, aleurone, and subaleurone tissues, respectively. For all HB, β‐glucan and acid‐extract viscosity were very low in the outermost 20% of the kernel. For low β‐glucan HB, β‐glucan content was the greatest in the subaleurone region and declined slightly toward inner layers. For high β‐glucan HB, however, more than 80% of grain β‐glucan was distributed more evenly throughout the endosperm. Acid extract viscosity was significantly (P < 0.01) correlated with total (r = 0.75) and soluble (r = 0.87) β‐glucan content throughout the kernel of all HB. Growing conditions, location and year, had significant effects on the concentration of protein, starch and β‐glucan. However, protein, starch, and β‐glucan distribution patterns were not affected by growing conditions. The difference in β‐glucan distribution between low and high β‐glucan HB may explain the difference in milling performance of HB with low or high β‐glucan.     

Effect of Health Information on Consumer Acceptability of Bread Fortified with β‐Glucan and Effect of Fortification on Bread Quality

Fortifying bread with β‐glucan reduces bread sensorial properties, though fortification using β‐glucan concentrates of low solubility under the conditions of dough preparation has not been investigated. This study investigated the consumer acceptability and purchase intent of bread fortified with a less soluble β‐glucan concentrate at levels corresponding to 0, 0.75, and 1.5 g β‐glucan/serving bread in relation to the provision of health information, gender, and whole wheat bread consumption. The effect of β‐glucan concentration on the physical properties of the bread produced under pilot plant settings was also investigated. β‐Glucan addition decreased (P < 0.05) loaf volume, increased firmness, and produced a darker, redder bread (P < 0.05), though fortification at 1.5 g β‐glucan/serving bread decreased height as well (P < 0.05). Consumer evaluation (n = 122) revealed that health information increased liking of appearance, flavor, and overall acceptability of the 1.5 g/serving bread to levels similar to or exceeding that of the control. Liking of the 1.5 g β‐glucan/serving bread appearance increased more in women than in men and for consumers who regularly consumed whole wheat bread for perceived health benefits when β‐glucan health information was provided. The provision of β‐glucan health information may be the key to increasing consumer acceptability of bread fortified with β‐glucan.     

Glycemic Response to Oat Bran Muffins Treated to Vary Molecular Weight of β‐Glucan

Oat bran muffins, containing 4 or 8 g of β‐glucan per two‐muffin serving, were prepared with or without β‐glucanase treatment to produce a range of β‐glucan molecular weights from 130,000 to just over 2 million. Following an overnight fast, the glycemic responses elicited by the untreated and treated muffins was measured in 10 healthy subjects and compared with a control whole wheat muffin. Taken all together, the 4‐g β‐glucan/serving muffins reduced blood glucose peak rise (PBGR) by 15 ± 6% compared with the control. The 8‐g β‐glucan/serving muffins had a significantly greater effect (44 ± 5% reduction compared with the control, P < 0.05). The efficacy of the muffins decreased as the molecular weight was reduced from a 45 ± 6% reduction in PBGR (P < 0.05) for the untreated muffins (averaged of both serving sizes) to 15 ± 6% (P < 0.05) for muffins with the lowest molecular weight. As the molecular weight was reduced from 2,200,000 to 400,000, the solubility of the β‐glucan increased from a mean of 44 to 57%, but as the molecular weight was further decreased to 120,000, solubility fell to 26%. There was a significant correlation (r² = 0.729, P < 0.001) between the peak blood glucose and the product of the extractable β‐glucan content and the molecular weight of the β‐glucan extracted.     

Isolation,Fractionation, and Structural Characteristics of Alkali‐Extractable β‐Glucan from Rye Whole Meal

The main nonstarch polysaccharide of rye is arabinoxylan (AX), but rye contains significant levels of (1→3)(1→4)‐β‐d ‐glucan, which unlike oat and barley β‐glucan, is not readily extracted by water, possibly because of entrapment within a matrix of AX cross‐linked by phenolics. This study continues objectives to improve understanding of factors controlling the physicochemical behavior of the cereal β‐glucans. Rye β‐glucan was extracted by 1.0N NaOH and increasing concentrations of ammonium sulfate were used to separate the β‐glucan from AX and prepare a series of eight narrow molecular weight (MW) distribution fractions. Composition and structural characteristics of the isolated β‐glucan and the eight fractions were determined. High‐performance size‐exclusion chromatography (HPSEC) with both specific calcofluor binding and a triple detection (light scattering, viscometry, and refractive index) system was used for MW determination. Lichenase digestion followed by high‐performance anion exchange chromatography of released oligosaccharides, was used for structural evaluation. The overall structure of all fractions was similar to that of barley β‐glucan.     

Effect of Formulation and Processing Treatments on Viscosity and Solubility of Extractable Barley β‐Glucan in Bread Dough Evaluated Under In Vitro Conditions

β‐Glucan shows great potential for incorporation into bread due to its cholesterol lowering and blood glucose regulating effects, which are related to its viscosity. The effects of β‐glucan concentration, gluten addition, premixing, yeast addition, fermentation time, and inactivation of the flour enzymes on the viscosity of extractable β‐glucan following incorporation into a white bread dough were studied under physiological conditions, as well as, β‐glucan solubility in fermented and unfermented dough. β‐Glucan was extracted using an in vitro protocol designed to approximate human digestion and hot water extraction. The viscosity of extractable β‐glucan was not affected by gluten addition, the presence of yeast, or premixing. Fermentation produced lower (P ≤ 0.05) extract viscosity for the doughs with added β‐glucan, while inactivating the flour enzymes and increasing β‐glucan concentration in the absence of fermentation increased (P ≤ 0.05) viscosity. The physiological solubility of the β‐glucan concentrate (18.1%) and the β‐glucan in the unfermented dough (20.5%) were similar (P > 0.05), while fermentation substantially decreased (P ≤ 0.05) solubility to 8.7%, indicating that the reduction in viscosity due to fermentation may be highly dependent on solubility in addition to β‐glucan degradation. The results emphasize the importance of analyzing β‐glucan fortified foods under physiological conditions to identify the conditions in the dough system that decrease β‐glucan viscosity so that products with maximum functionality can be developed.     

Effects of Different Cereal Fibers on Cholesterol and Bile Acid Metabolism in the Syrian Golden Hamster

This study examined the effects of various cereal fibers and various amounts of β-glucan on cholesterol and bile acid metabolism. Hamsters were fed semisynthetic diets containing 0.12% cholesterol, 20% fat, and either 16% total dietary fiber (TDF) from wheat bran (control) or 10% TDF from oat bran, 13% TDF from oat bran concentrate or barley grains, 16% TDF from oat fiber concentrate, barley flakes, or rye bran. After five weeks, plasma total cholesterol and liver cholesterol concentrations were significantly lower (20 and 50%, respectively) only in hamsters fed rye bran. Diets containing any of the oat ingredients or barley had no effect on total cholesterol. Changes in the pattern of biliary bile acids occurred in hamsters fed 16% TDF from barley flakes or 10% TDF from oat bran. Hamsters fed rye bran had a significantly higher fecal bile acid excretion when compared with controls fed wheat bran. Because rye bran caused the most pronounced lowering effect of total cholesterol despite the lowest content of β-glucan and soluble fibers, components other than β-glucan and soluble fibers seem to be involved in its hypocholesterolemic action. Since the effects of the oat and barley ingredients were not solely correlated to the β-glucan content, structural changes occurring during processing and concentrating of the products may have impaired the hypocholesterolemic potential of the β-glucans, and other factors such as solubility and viscosity of the fiber components seem to be involved.     

In vitro fermentation of bacterial cellulose composites as model dietary fibers

Plant cell walls within the human diet are compositionally heterogeneous, so defining the basis of nutritive properties is difficult. Using a pig fecal inoculum, in vitro fermentations of soluble forms of arabinoxylan, mixed-linkage glucan, and xyloglucan were compared with the same polymers incorporated into bacterial cellulose composites. Fermentation rates were highest and similar for the soluble polysaccharides. Cellulose composites incorporating those polysaccharides fermented more slowly and at similar rates to wheat bran. Bacterial cellulose and cotton fermented most slowly. Cellulose composite fermentation resulted in a different short-chain fatty acid profile, compared with soluble polysaccharides, with more butyrate and less propionate. The results suggest that physical form is more relevant than the chemistry of plant cell wall polysaccharides in determining both rate and end-products of fermentation using fecal bacteria. This work also establishes bacterial cellulose composites as a useful model system for the fermentation of complex cell wall dietary fiber.