J Sci Food Agric 1998, 77, 179È183 Barley b-Glucan is Eþective as a Hypocholesterolaemic Ingredient in Foodss Kari D Hecker,” Mary L Meier, Rosemary K Newman* and C Walter Newman Montana State University, Department of Plant, Soil and Environmental Sciences, Bozeman, MT 59717, USA (Received 21 August 1997 ; accepted 6 October 1997) Abstract : Barley contains high levels of soluble dietary Ðbre, including mixed linked 1 ] 3, 1 ] 4b-D-glucans (b-glucan). An extract of b-glucan from waxy, hulless barley containing 56% total dietary Ðbre (TDF) was incorporated into Ñour tortillas, cornstarch pudding and apple granola bars to provide 2 g soluble Ðbre as b-glucan per serving. The foods were tested for objective functional properties. Flour tortillas with b-glucan were incorporated into rat diets and compared to diets containing an equivalent amount of cellulose, to test the Ðbre e†ect on growth and lipid metabolism parameters. Rats fed b-glucan tortillas had lower feed consumption and body weight (P \ 0É05) compared to those fed the cellulose tortillas, although feed/gain ratios were not di†erent (P [ 0É05). Plasma LDL-cholesterol of rats fed b-glucan was lower (P \ 0É05) than cellulose-fed controls, although total cholesterol and triglycerides did not di†er (P [ 0É05). Rats fed b-glucan tortillas had higher (P \ 0É05) faecal fat excretion, suggesting impairment of intestinal fat absorption. Liver composition data showed lower (P \ 0É05) levels of total lipid and cholesterol in b-glucan-fed rats. The results suggest that the barley b-glucan concentrate has potential as a food ingredient to provide supplemental soluble Ðbre which may be beneÐcial in reducing plasma LDL-cholesterol in humans. ( 1998 SCI. J Sci Food Agric 77, 179È183 (1998) Key words : barley ; b-glucans ; cholesterol ; soluble Ðbre INTRODUCTION cosity in the intestine, and e†ect on the bile acidÈ cholesterol cycle (Story 1986 ; Wang et al 1992). Barley contains high levels of soluble dietary Ðbre, particularly mixed linked 1 ] 3, 1 ] 4 b-D-glucans, and has been demonstrated to have hypocholesterolaemic e†ects in experimental animals and humans (Fadel et al 1987 ; Newman et al 1989 ; McIntosh et al 1991). Utilisation of barley as a food grain has been reviewed and there is a potential for many forms of barley as ingredients in a wide variety of foods (Newman and Newman 1991). Incorporation of large percentages of barley grain to produce high Ðbre foods, however, may not always be practical or desirable. Concentrated b-glucan extracts from barley and oats have been demonstrated to have hypocholesterolaemic properties (Newman et al 1993). The present study was conducted to evaluate the use of a barley b-glucan concentrate as a Ðbre-enriching food ingredient. The speciÐc objectives were to increase Interest in dietary Ðbre has increased recently due to its beneÐcial e†ects on blood cholesterol levels, glycaemic responses, and associated risks for chronic diseases (Pilch 1987). Total dietary Ðbre (TDF) consists of insoluble fractions, both of which are resistant to digestion by the alimentary enzymes of humans (Slavin 1987). The insoluble fraction in cereal grain contains a large proportion of cellulose and is noted for its beneÐcial e†ects in the gastrointestinal tract (Jenkins et al 1986). The soluble fraction, which contains mostly pectin, arabinoxylan, and b-glucan, has the ability to lower blood serum cholesterol, through its tendency to increase vis* To whom correspondence should be addressed. ¤ Contribution No J-5057, Montana Agricultural Experiment Station, Bozeman 59717. ” Present address : University of Missouri, Department of Human Environmental Sciences, Columbia, MO 65211, USA. 179 ( 1998 SCI. J Sci Food Agric 0022È5142/98/$17.50. Printed in Great Britain K D Hecker et al 180 soluble Ðbre by 2 g per serving in three di†erent food products. Serving sizes were one Ñour tortilla, one-half cup of cornstarch pudding and one 8 cm apple granola bar. These products, previously tested for acceptability, were selected as representative foods that might be consumed either as snacks or, in the case of tortillas, breadstu† for a variety of menu items. The Ñour tortillas were tested for cholesterol-lowering properties in a rat experimental model and all three foods were tested for functional quality. MATERIALS AND METHODS Fibre concentrate b-Glucan concentrate, a Ñavourless white powder, was obtained from Centennial Foods Inc (Dillon, MT, USA). The composition was : 7É11% moisture ; 0É95% ash ; 0É10% fat ; 3É59% protein ; and 55É60% b-glucan, the remainder consisting of oligosaccharides. Product preparation and objective evaluation Flour tortillas were prepared using a standard formula containing 446 g Ñour (Gold Medal, General Mills Inc, Minneapolis, MN, USA), 5 g salt (Albertsons Inc, Boise, ID, USA), and 28 g baking powder (Calumet, White Plains, NY, USA). For the experimental tortillas, 50 g of b-glucan concentrate was substituted for an equivalent amount of Ñour and added with the dry ingredients. Thirty grams of lard (Rex Manteca, Bar-S Foods Co, Phoenix, AZ, USA) was then added to the dry ingredients and the mixture was stirred to the consistency of peas. Warm (40¡C) tap water (336 g) was gradually added while stirring ingredients to form a ball of dough. After the ball was formed, it was kneaded for 5 min and placed in a bowl, covered with a towel and let stand for 20 min at room temperature. Eighteen equal portions were then weighed out and rolled on a pasta press (Berarducci Bros Mfg Inc, level 4). The tortillas were placed on a hot griddle (West Bend Sensa Temp, West Bend, WI, USA), at 121¡C and cooked for 45 s on each side. Tenderness of the tortillas was tested using the Warner-Bratzler Shear Apparatus (G-R Elec Mfg Co, Manhattan, KS, USA). Three samples of each product, both control and experimental, were tested once and the average of the three results used to determine tenderness. Storage stability, determined by rollability of stored tortillas, was evaluated. Three samples of control and experimental tortillas were tested every other day for 7 days, and scored according to Friend et al (1992). The rollability scale ranged from 1È5, as follows : (1) no cracking (best) ; (2) signs of cracking but no breaking ; (3) cracking and breaking beginning on one surface ; (4) cracking and breaking imminent on both sides ; and (5) unrollable, breaks easily (poorest). Since sample tortillas were stored in a refrigerator, they were heated in a microwave on full power for 6 s before each test, then rolled around a 0É64 cm dowel and judged according to the above scale. Cornstarch pudding was prepared by a standard method (Morr and Irmiter 1990), the formula containing 18 g cornstarch (KingsfordÏs, Best Foods CPC International Inc, Englewood Cli†s, NJ, USA), 48 g granulated sugar, 1 g salt (AlbertsonÏs Inc, Boise, ID, USA), 15 g margarine (AlbertsonÏs Inc, Boise, ID, USA), 8 g vanilla (Schilling, McCormick & Co Inc, Hunt Valley, MD, USA), and 53 g beaten egg yolk. For the experimental pudding, 5 g b-glucan concentrate was added with the dry ingredients. For the control pudding, 239 g of skim-milk (Darigold Inc, Seattle, WA, USA) was added, and the amount for the experimental pudding was increased to 270 g to compensate for the water-holding capacity of the b-glucan. Pudding viscosity was determined using a line-spread apparatus (McWilliams 1987), measured in consistometers. Duplicate samples of 100 g each of control and experimental puddings were measured at 55 and 64¡C. The measured samples were placed in a bottomless styrofoam cup with a diameter of 4É6 cm, on measured concentric rings and allowed to Ñow for 2 min. At the end of each 2 min trial, the spread at each 90¡ increment of the circle was read. Line-spread values are the means of the four values obtained. Control and experimental apple granola bars were prepared as follows : 223 g brown sugar (California and Hawaiian Sugar Co, Concord, CA, USA), 57 g creamy peanut butter (Simply Jif, Procter and Gamble, Cincinnati, OH, USA), 36 g light corn syrup (Karo, Best Foods, CPC International Inc, Englewood Cli†s, NJ, USA), 53 g whole egg, 117 g margarine (AlbertsonÏs, Boise, ID, USA), and 260 g apple Ðlling (Solo, Sokol & Company, Countryside, IL, USA) were combined in a large bowl and blended. The remaining ingredients, 7 g vanilla (Schilling, McCormick & Co Inc, Hunt Valley, MD, USA), 1 g salt (AlbertsonÏs Inc, Boise, ID, USA), 0É35 g ground cinnamon (Schilling, McCormick & Co Inc, Hunt Valley, MD, USA), 269 g quick-cooking oats (Quaker Oats Company, Chicago, IL, USA), 107 g whole wheat Ñour (Gold Medal, General Mills Inc, White Plains, NY, USA), 57 g Ñaked coconut (Kraft General Foods Inc, White Plains, NY, USA), 25 g sunÑower nuts (Gold Shield, PaciÐc Foods Inc, Kent, WA, USA), 28 g wheat germ, and 22 g sesame seeds were added in sequence, blending well between each addition. The mixture was then pressed evenly into a well greased 33 ] 23 ] 5 cm pan and baked at 177¡C for 35 min until golden brown. The experimental bars were prepared using the same ingredients with the exception of using 88 g corn syrup, 378 g apple Ðlling, 87 g whole wheat Ñour and 59 g b-glucan concentrate. Because the b-glucan concentrate has high liquid-holding capacity, Hypocholesterolaemic e†ects of barley b-glucan 181 the increased amounts of corn syrup and apple Ðlling were added and the Ñour quantity was adjusted to achieve a comparable consistency to the control product. The apple granola bars were tested for tenderness using the same methods as for the Ñour tortillas. Physiological and cholesterol-lowering properties The cholesterol-lowering property of the b-glucan enriched tortillas was tested in a rat feeding study. The control tortilla used for the rat experiment was made using the same formula as used in objective evaluation, except for the substitution of 27 g of cellulose (to equalise TDF) for this amount of wheat Ñour. Rat diets (Table 3) met all nutrient requirements of laboratory rats (NRC 1978). Twenty Sprague-Dawley rats, 22È26 days old, were acclimated in an approved animal facility and fed a standard pelletised diet with a 12 : 12 light : dark schedule for 5 days. All conditions conformed to the NRC standards (NRC 1985). Rats were weighed and assigned to two groups with equalised weight ranges between 133 and 147 g. Animals were placed in individual cages and fed diets containing tortillas with either cellulose (control) or b-glucan for 25 days. Rats were weighed weekly, diet consumption reported every four days, and feed waste estimates made twice weekly. Total faecal collections were taken on days 17 to 19, and samples were freeze-dried and stored at [20¡C prior to analysis. On day 25, rats were fasted for 12 h before blood samples were taken by heart puncture. Rats were killed with carbon dioxide and livers excised, weighed and frozen at [20¡C for later analysis. Chemical analysis The three food products and the test diets were analysed for content of soluble and insoluble dietary Ðbre (Prosky et al 1988), protein (N ] 6.25), ether extract and ash (AOAC 1984). Plasma was analysed for total cholesterol (Allain et al 1974), triglycerides (Tietz 1987), and HDL-cholesterol (Finley et al 1978) on a Kodak DT 60 analyser (Eastman Kodak, Rochester, NY, USA). Plasma LDLcholesterol was calculated as described by Friedwald et al (1972) and faecal total lipid content was analysed (Anon 1971). Livers were analysed for lipids (Folch et al 1957) and total cholesterol (Warnick et al 1982). Statistical analysis Data were analysed using SAS General Linear Models Procedure and di†erences between means were analysed by Least Square Means (SAS 1988). RESULTS AND DISCUSSION Chemical analysis Table 1 shows the analysed composition of food products. Control and experimental versions of each product had comparable fat, ash and protein content, except for the apple granola bars, which had considerably lower fat in the b-glucan version due to added carbohydrate ingredients. Fibre content varied between control and experimental products relative to the added b-glucan. Objective evaluations Results of objective tests on food products are shown in Table 2. In tenderness tests, the control tortillas used an average of 0É15 kg of pressure compared to 0É127 kg for those containing b-glucan, indicating slightly more tenderness for the latter. In rollability, control tortillas scored “1Ï on the initial trial (day 0), while b-glucan tortillas scored “2Ï. Thereafter, there were no di†erences in rollability scores between control and b-glucan tortillas. Decreasing scores (higher numbers) over time indicated gradual deterioration in consistency. Storage stability of TABLE 1 Composition of food products (DM%) Item Tortillas Control b-Glucan Pudding Control b-Glucan Apple granola bars Control b-Glucan Fat Ash Protein Fibre Insoluble Soluble T otal 5É3 5É7 4É1 3É6 9É5 9É9 2É0 2É3 1É7 10É2 3É7 12É5 12É0 10É5 2É8 2É9 8É1 8É6 0É5 0É4 0É1 2É0 0É6 2É4 18É1 12É6 1É7 1É5 9É9 8É2 3É6 3É1 1É4 3É3 5É0 6É4 K D Hecker et al 182 TABLE 2 Objective tests of food productsa Product T ype test Control TABLE 3 Composition of rat diets (% of diet) b-Glucan Tortillas Tenderness (kg pressure) Rollability scores Day 0 Day 2 Day 4 Day 6 0É15 0É127 1 3 4 5 2 3 4 5 Pudding Viscosity (linespread cm)b 0 minc 5 min 7É1 6É5 10É9a 8É3a Tenderness (kg pressure) 0É69 1É12 Apple granola bars a Values followed by di†erent letters are statistically di†erent than control products at a signiÐcance level of P \ 0É05. b Viscosity : cm per 2 min spread. c Cool time. the tortillas could be improved by ways such as those shown in Friend et al (1992) and Suhendro et al (1993) who reported a reversal of poor storage stability in high Ðbre and protein-enriched tortillas, respectively, by addition of wheat gluten. Line spread tests showed the control pudding to be more viscous (smaller area spread) than the pudding with b-glucan at 0 and 5 min cooling times. This is an interesting phenomenon, in that uncooked b-glucan is known to possess high viscosity in grain extract analysis and in the intestine of experimental animals (Wang et al 1992). In tenderness testing of the apple granola bars, the control bars were more tender with an average of 0É69 kg of pressure compared to 1É12 kg for bars with b-glucan. Physiological properties Feed consumption of rats, body weight gain and feed/ gain ratios are shown in Table 4. Rats fed b-glucan consumed less feed and gained signiÐcantly less body weight (P \ 0É05) than those fed the control tortilla, although there was not a signiÐcant di†erence in feed/ gain ratio. The LDL-cholesterol was signiÐcantly lower (P \ 0É05) in rats fed b-glucan tortillas, but there were no di†erences in total cholesterol, HDL-cholesterol or triglycerides between the two groups. Rats fed b-glucan tortillas had higher (P \ 0É05) faecal excretion of fat, suggesting that b-glucan interfered with fat absorption. These results agree with earlier results reported for barley (Fadel et al 1987), Ingredient Control Experimental b-Glucan tortilla (13% TDF) Cellulose tortilla (13% TDF) Casein Peanut oil AIN mineral mix AIN vitamin mix Cholesterol Sodium cholate Antioxidant DL-Methionine Lysine Cornstarch Dye Analysed composition (DM%) Total dietary Ðbre Soluble Ðbre Insoluble Ðbre Total protein Total fat È 80É00 80É00 È 7É7 4É4 3É5 1É0 1É0 0É20 0É01 0É31 0É04 1É8 0É01 7É3 4É7 3É5 1É0 1É0 0É20 0É01 0É33 0É01 1É9 0É01 10É4 1É3 9É1 14É8 10É1 9É5 7É8 1É7 13É9 10É5 barley b-glucan concentrate and oat concentrate (Newman et al 1993). Livers of rats fed b-glucan tortillas had lower (P \ 0É05) levels of total lipid and cholesterol than the controls, suggesting a metabolic shift toward reduced total blood cholesterol. These results are supportive of earlier studies reporting the cholesterolTABLE 4 Feed consumption, weight gain and metabolic parameters of rats tortilla dietsa T ortillas Diet group with tortillas with cellulose b-glucan Totals for 25 days Food consumption (g) Body weight gain (g) Feed/gain Plasma lipids (mmol litre~1) Total cholesterol HDL-cholesterol LDL-cholesterol Triglycerides Faecal fat (DM%) Livers Total lipids (g g~1 tissue) Cholesterol (mg g~1 tissue) 430É0 162É0 2É66 2É82 1É45 1É13 0É38 7É8 15É0 34É7 392É5a 146É2a 2É69 SEb 6É8 2É8 0É04 2É38 1É63 0É68a 0.33 13É3a 0É22 0É12 0É12 0É03 0É40 11.4a 25É1a 0É80 1É60 a Values followed by di†erent letters are statistically di†erent than control group at a signiÐcance level of (P \ 0É05). b Standard error of the least square means. Hypocholesterolaemic e†ects of barley b-glucan lowering properties of barley grain (Newman et al 1989), and are particularly important as a demonstration of the use of extracted b-glucan concentrate as a soluble Ðbre supplement in palatable food products. CONCLUSIONS Current dietary recommendations for Americans specify 25È35 g of total dietary Ðbre per day, of which onequarter should be soluble (Pilch 1987). The best common sources of soluble Ðbre are beans, other vegetables, fruits, oats and barley. Barley b-glucan extracted from the grain is a tasteless, odourless white powder which can be incorporated into snack foods and breadstu†. The inclusion of 2 g soluble Ðbre per serving in pudding, granola bars and tortillas is an example of Ðbre enrichment that can contribute to consumer consumption approaching the recommended levels of Ðbre. In this study, one of these foods was shown to have hypocholesterolaemic properties. ACKNOWLEDGEMENTS This research was partially supported by the Montana Wheat and Barley Committee, Box 3024, Great Falls, MT 59403. Petrea Hofer assisted with chemical analysis and care of animals. REFERENCES Allain C C, Poon L S, Chan C S G, Richmond W, Fu P C 1974 Enzymatic determination of total cholesterol in serum. Clin Chem 20 470È476. Anon 1971 Dosage des matières grasses brut [Determination of crude fat]. Journal Officiel des Communautès Europèennes, L egislation [Offic J Eur Comm] L279 17È18. AOAC 1984 Official Methods of Analysis (14th edn). Association of Official Analytical Chemists, Washington, DC, USA. Fadel J G, Newman R K, Newman C W, Barnes A E 1987 Hypocholesterolemic e†ects of b-glucans in di†erent barley diets fed to broiler chicks. Nutr Rep Int 35 1049È1058. Finley P R, Shifman R B, Williams R J, Lichti D A 1978 Cholesterol in high density lipoprotein : Use of Mg2` dextran sulfate in its enzymic measurement. Clin Chem 24 931È933. Folch J, Lees M, Sloane Stanley G H 1957 A sample method for the isolation and puriÐcation of total lipids from animal tissues. J Biol Chem 226 497È509. 183 Friedwald W T, Levy R I, Fredrickson D S 1972 Estimation of the concentration of low-density lipoprotein cholesterol in plasma without the use of preparative ultracentrifuge. Clin Chem 18 499È502. Friend C P, Serna-Saldivan S O, Waniska R D, Rooney L W 1992 Increasing the Ðber content of wheat tortillas. Cereal Foods W orld 37 325È328. Jenkins D J, Jenkins A L, Wolever T M S, Rao A V, Thompson L V 1985 Fiber and starchy foods : Gut function and implications in disease. Am J Gastroenterol 81 920È930. McIntosh G H, Whyte J, McArthur T, Nestel P J 1991 Barley and wheat foods inÑuence on plasma cholesterol concentrations in hypercholesterolemic men. Am J Clin Nutr 53 1205È1209. McWilliams M 1987 Foods : Experimental Perspectives. Macmillan, New York, USA. Moor M, Irmiter T F 1990 Introductory Foods. A L aboratory Manual of Food Preparation and Evaluation. Macmillan, New York, USA. Newman R K, Newman C W 1991 Barley as a food grain. Cereal Foods W orld 36 801È805. Newman R K, Newman C W, Graham H 1989 The hypocholesterolemic function of barley b-glucans. Cereal Foods W orld 34 883È886. Newman R K, Inglett G E, Newman C W, Hofer P J, Wang D L 1993 b-Glucan concentrates from oats and barley lower cholesterol in chicks. FASEB J 38 601. NRC (National Research Council) 1978 Nutrient Requirements of L aboratory Animals (3rd edn). National Academy of Sciences, Washington, DC, USA. NRC (National Research Council) 1985 Guide for the Care and Use of L aboratory Animals (Publ No 85-23). NIH, Bethesda, MD, USA. Pilch S M 1987 Ad hoc expert panel in dietary Ðber. In : Physiologic and Health Consequences of Dietary Fiber. FASEB, Rockville, MD, USA. Prosky L H, Asp G, Schweizer T F, DeVries J W, Furda I 1988 Determination of insoluble, soluble and total dietary Ðber in foods and food products : Interlaboratory Study J Assoc Anal Chem 71 1017È1023. SAS 1988 Statistical Analysis Systems UserÏs Guide (Basic Version 6.03 ed). SAS Institute Inc, Cary, NC, USA. Slavin J L 1987 Dietary Ðber : ClassiÐcation, chemical analysis and food sources. J Am Diet Assoc 87 1164È1171. Story J A 1986 ModiÐcation of steroid excretion in response to dietary Ðber. In : Dietary Fiber Basic and Clinical Aspects, eds Vahouny G V & Kritchevsky D. Plenum Press, New York, USA. Suhendo E L, Waniska R D, Rooney L W 1993 E†ects of added proteins in wheat tortillas. Cereal Chem 70 412È416. Tietz N W 1987 Fundamentals of Clinical Chemistry. WB Saunders Co, Philadelphia, USA. Wang L, Newman R K, Newman C W, Hofer P J 1992 Barley b-glucans alter intestinal viscosity and reduce plasma cholesterol in chicks. J Nutr 122 2292È2297. Warnick G R, Benderson J, Albers J J 1982 Dextran sulfateMg2` precipitation for quantitation of high density lipoprotein cholesterol. Clin Chem 28 1379È1388.