Effect of the Source of Fiber in Bread on Intestinal Responses and Nutrient Digestibilities,

Cereal Chem. 65(l):9-12 Young rats were fed one fiber-free and 12 isofibrous diets for three weeks. to colonic degradation than naturally occurring insoluble fiber; e) fecal Fiber originated from 12 different types of breads. Five (white, oatmeal, density was lower on cellulose-containing, bran-containing, and corn, and two multigrain) of these contained one-fifth or more of the total wholewheat breads than on other breads; (f) fiber degradation averaged the fiber as soluble fiber. The following results were obtained: a) fiber caused a highest (82. 1%) on white bread and the least (15.9%) on cellulosesignificant increase in stomach and colon weights; b) fiber increased the dry containing bread; g) apparent digestibility of nitrogen and fat was adversely fecal weight threeto sixfold; c) insoluble fiber was positively correlated affected by fiber; and h) fiber caused a significant reduction in fecal pH. with fecal output (r= 0.56); d) cellulose added to bread was more resistant Human epidemiological studies show a lower incidence of colorectal cancer in population groups consuming diets high in fiber (Trowell et al 1985). Fiber is postulated to protect against colorectal cancer by decreasing the transit time of contents, altering the bile acid metabolism, affecting the consequences of fermentation, providing a surface for adsorption and by physical dilution (bulking effect) of lumen contents (Burkitt et al 1972, Cummings 1985). The fecal bulking capacity of fiber sources varies. Sources that are high in water-insoluble fiber fractions tend to provide more bulk. Whole wheat and other wheat-based variety breads are an important source of insoluble fiber in our diet. Several of these and related bread products, however, also contain flours and fractions from nonwheat cereals, noncereal grains, and vegetables. Some products also contain highly purified and/ or modified fiber fractions. These various ingredients, usually carried through the use of added gluten, may contain a sizable portion of fiber as water-soluble fiber. Soluble fiber is reported to be highly fermentable in the large intestine (Nyman and Asp 1982, Van Dokkum et al 1983, Nyman 1985, Cheng et al 1987) and thus may provide little bulk. This, and the understanding that food processing (baking, for example) may modify the physiological properties generally attributed to plant fibers (Nyman 1985, Eastwood et al 1986, Schneeman 1987), prompted these studies. Fiber in 12 types of widely consumed variety breads was examined for its behavior along the intestinal tract. Rats, which show a good correlation with man in fiber degradation and in bulking capacity (Nyman 1985), were used as the test model. MATERIALS AND METHODS Test Breads and Diets Breads tested (Table I) were purchased locally but represent national brands. They were air-dried, finely ground, and then used to formulate diets (Table I). Diets were formulated to be equal in nitrogen, fat, and moisture (Table I). With the exception of the control (fiber-free) diet, all diets also contained the same, 4.1%, level of total dietary fiber (TDF) and about the same level of available carbohydrates. Diets were complete in all micronutrients required by the rat (NAS/NRC 1978). 'This paper was presented at the AACC 72nd Annual Meeting, Nashville, TN, November 1987. 2Nutrition Research, American Institute of Baking, 1213 Bakers Way, Manhattan, KS 66502. Animals and Feeding Male weanling rats (six rats per diet) of the Sprague-Dawley strain (Harlan Sprague-Dawley, Indianapolis, IN) were housed individually in mesh bottom stainless steel cages in a controlled environment. Sliding fecal collection trays were provided under each cage. Distilled water was offered to the animals ad libitum, but the amount of diet fed was restricted; each rat was fed the same amount, however, which was adequate and was gradually increased during the three-week test period. Rats were weighed at weekly intervals. Intestinal Measurements At the end of the test period, all animals were sacrificed and their intestinal tract was removed. Stomach, small intestine, and colon (cecum included) were separated, thoroughly cleaned of the lumen contents, blotted dry, and immediately weighed. Fecal Measurements Feces were collected quantitatively twice daily for the entire three-week period, pooled, air-dried, weighed, and stored under refrigeration. Density was calculated by dividing the fecal weight by volume. Fecal volume was determined in a long-stem graduated cylinder using fine sand as the embedding medium. Feces, recovered from the sand, were finely ground and analyzed for TDF, nitrogen, fat, ash, and pH. Fiber Fermentation and Nutrient Digestibilities Fiber fermentation and the apparent digestibility of nitrogen, fat, and ash were estimated by the difference between amounts ingested and excreted over the three-week test period. Analytical Finely ground breads and casein were analyzed for moisture, protein (Kjeldahl), fat (acid hydrolysis), and ash using standard AACC methods (1983). TDF in breads was determined by the recently approved method of Prosky et al ( 1985); the incorporation of additional steps (filtration and precipitation) in the method allowed the determination of insoluble and soluble fiber components. The same methods (AACC 1983, Prosky et al 1985) were used to determine TDF, nitrogen, fat, and ash in feces. Fecal pH was also determined by the standard AACC method (1983). Statistical Averages, as are listed in Tables II and III, were compared by Duncan's (1955) multiple range test. RESULTS AND DISCUSSION Twelve different types of breads were tested (Table 1). Whole wheat bread, corn tortillas, and rye bread represent three different grains used almost in totality. Two of the four mixed grain/ multiVol. 65, No. 1, 1988 9 This article is in the public domain and not copyrightable. It may be freely reprinted with customary crediting of the source. American Association of Cereal Chemists, Inc., 1988. grain breads (breads K and L) also contained powdered vegetables and a high level of calcium. Four other bread types contained a purified (cellulose) or a natural (wheat bran, cracked wheat, oats) fiber source. Only white bread was a refined product (made entirely with low-extraction wheat flour). Fiber in Breads and Diets On an as-purchased basis, TDF in test breads ranged between 3.0 and 9.9%. In five test breads, over 20% of the TDF was soluble fiber (Table I). For white bread, this may reflect the compositional makeup of the fiber fractions or the contribution of resistant starch formed during baking (Berry 1986). White bread was lowest (4.4%, dry basis) in TDF content and permitted only 4.1% TDF in the diet (diet A) with a nitrogen content of 1.88%. Other bread-based diets (diets B-L) were formulated to also contain only 4.1% TDF. Casein, corn oil, and sucrose were used to equalize the content of nitrogen, fat, available carbohydrates, and energy among diets. Growth Response of Animals Rats were fed the same amount of the test diets (159 g) during the three-week period. The growth response of animals (Table II), however, differed significantly (P< 0.01) among groups, primarily because the amount of casein (Table I) and, thus, the quality of protein in diets varied. The effect of fiber on energy utilization (discussed later) was likely not a significant factor affecting growth. Intestinal Measurements The weight of the gastrointestinal tract (GIT) of rats fed the test diets averaged between 1.65 and 3.07 g, being significantly higher in animals fed the fiber-free diet than the fiber-containing diets except for diet B (Table II). GIT weights were only weakly correlated (r = 0.53) with body weight gains. As a percentage of the GIT weight, colon weight was significantly higher in rats fed the fiber-containing diets than the fiber-free diet. This agrees with results reported by Jacobs and Schneeman (1981), who attribute this effect to proliferation of colonic mucosa and possible increase in colonic muscle mass. The health consequences of this observation remain uncertain, however. Stomach weights also tended to be significantly higher on several, but not all, of the fiber-containing diets. In contrast, weight of the small intestine (as percent of GIT weight) was consistently and significantly higher in rats fed the fiber-free diet than diets A-L. Thus two anatomical segments of the GIT, separated by the small intestine, responded to fiber in a morphologically similar manner. Fecal Measurements Unless feces are collected immediately as expelled (this was not done), fecal moisture loss may vary between groups. For this reason, fecal measurements (based on three-week collections) are expressed on dry basis only (Table II). This, no doubt, distracts from a more realistic assessment of physiological effects where measurements are based on wet feces. Although a low fecal weight is not necessarily cancer promoting, a high fecal weight may protect against colon cancer (Cummings 1985). Compared to the fiber-free diet, fecal weights increased threeto sixfold as breads were included in the diet. This increase represents primarily the bacterial mass, the undegraded fiber, and, in some cases, the excreted mineral matter. The increase in bacterial mass due to colonic degradation of fiber may be sufficient to increase the fecal weight. This is evident when fiber-free diet M is compared with diet A, which contained a highly degradable fiber source (Table III). Where fiber is resistant to bacterial degradation (diet B, for example), the increase in fecal weight is primarily due to undegraded fiber. Exceptions to this are noted for diets D, K, and L. Rats fed these diets also excreted a substantial amount of mineral matter that originated from limetreated tortillas (diet D) or super-fortified (with calcium) breads K and L. Soluble fiber components are readily degraded by the colonic bacteria (Nyman 1985, McLean Ross et al 1983). For white bread, oatmeal bread, and corn tortillas, this is apparent when relevant data in Tables I and III are examined. Breads K and L, probably because of the high ash content (Table I), defy this trend, however. Disregarding the error introduced by the high ash content on diets K and L, fecal output was the highest on cellulose-containing bread (bread B). Breads F-J, which like bread B were also high in insoluble fiber (Table I), yielded significantly lower fecal outputs. Thus, although dietary insoluble fiber (diets D, K, and L not considered) appears to be correlated with fecal weight (r = 0.56), the source of insoluble fiber may be another determining factor of bacterial degradation and, therefore, fecal output. Eastwood et al (1986), however, suggest there is no correlation between chemical composition and structure of the fibers and their physiological effects. Fecal volumes, which highly correlated with fecal weights, well exceeded the fecal weights in some cases. This is particularly true in animals fed diets made with breads containing cellulose (bread B), TABLE I Composition of Test Breads and Diets Bread Compositiona (%) Diet Compositionb (g/100 g) Protein Total Corn Diet Bread Moisture (N X 5.7) Fat Ash Fiberc Bread Casein Oil Sucrose A White 36.7 7.8 2.8 1.8 3.0 (34.1) 93.8 ... 1.84 0.1 B Cellulose-containing 41.5 8.2 3.4 2.1 9.9 (11.4) 26.3 9.45 4.22 51.0 C Oatmeal 36.3 9.1 5.1 2.3 3.7 (32.6) 75.0 1.02 -17.8 D Tortillas, corn 43.8 5.0 3.7 1.4 4.3 (21.7) 58.1 7.64 2.13 25.9 E Rye, German 36.8 10.2 3.8 2.2 8.3 (17.9) 33.6 7.27 3.76 46.5 F Cracked wheat 35.2 10.0 3.6 2.1 6.7 (14.1) 42.3 5.96 3.43 40.0 G Bran-containing 39.1 9.2 3.6 2.1 5.4 (15.1) 49.7 4.87 2.88 34.8 H Whole wheat, 100% 39.7 10.5 5.3 2.1 8.1 (13.6) 32.7 6.98 2.97 48.4 I Mixed grain 40.3 9.9 3.3 1.9 5.6 (10.7) 46.6 4.58 3.20 37.6 J Multigrain 39.5 10.6 3.5 2.4 9.6 (13.1) 27.8 8.00 4.12 50.9 K Multigrain + DPV-ld 38.3 7.4 3.1 4.6 3.4 (40.6) 81.4 2.25 1.87 9.7 L Multigrain + DPV-II d 38.8 9.0 3.4 4.5 4.9 (24.4) 55.2 4.22 2.81 31.0 M None (control) ... ... ... ... ... ... 13.75 5.43 70.5 'Available carbohydrate values (not listed) equal the remainder of the sum of components listed subtracted from 100. 'All diets contained 1.888% nitrogen, 5.6% fat, 4.1 % total dietary fiber (diet M was a fiber-free control), and 7.0% moisture. Water added to equalize the moisture content among diets varied from 0 to 6.11%. Diet components not listed included: vitamin mix (AIN mix 76 from U.S. Biochemicals, Cleveland, OH), I g; and mineral mix (in sucrose base), 3.2g. Mineral mix contained (mg): Ca, 500; P, 400; Fe, 3.5; Cu, 0.5; 1, 0.015; Mg, 40; Mn, 5; Se, 0.01; Zn, 1.2; and K, 360. 'Total dietary fiber. Values within parentheses are soluble fiber contents expressed as percentage of total fiber. dDPV = Dried powdered vegetable. DPV-1 is a light and DPV-I1 is a dark product.