![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
The Journal of Nutrition Vol. 128 No. 3 March 1998,
pp. 651-660
Department of Applied Nutrition and Food Chemistry, Center for Chemistry and Chemical Engineering, Lund University, S-221 00 Lund, Sweden
 |
ABSTRACT |
|---|
Abstract
Introduction
Methods
Results & Discussion
References
|
|---|
The purpose of this work was to develop a method for measurement of the major forms of resistant starch (RS) in foods. The analytical procedure was chosen to mimic physiologic conditions, and included chewing as a prestep before incubation with pepsin, pancreatin and amyloglucosidase. The undigestible polysaccharides, including RS, were recovered by ethanol precipitation and subsequent filtration. RS was analyzed as total starch in the filter residue. The residues were also used for gravimetric determination of dietary fiber after correcting for remaining protein, ash and RS. The potentially available starch fraction was determined from analysis of glucose in the filtrate. The foods included were prepared to resemble products for which RS figures were available from in vivo measurements, and/or from analysis with other current in vitro methods. For six of these foods, and for three additional starchy materials, RS figures were compared with in vivo and/or in vitro data for identical products. The pooled standard deviation for the suggested RS method was 2.9%. A high correlation was obtained with in vivo figures from the literature for 19 realistic foods (r = 0.97; y = 0.77x + 0.45). After correction for RS, dietary fiber figures corresponded well with conventional gravimetric dietary fiber analysis for 14 starchy foods (r = 0.97). It is concluded that the procedure described here provides a convenient way to estimate RS content of realistic foods, allowing parallel determination of the potentially available starch fraction and dietary fiber.
KEY WORDS: resistant starch · potentially available starch · dietary fiber · humans · chewing
At the beginning of the 1980s, it was discovered that a fraction of dietary starch escaped digestion and absorption in the human small intestine (Anderson et al. 1981 As yet, very few products have been characterized as to their RS content, which makes reliable estimations of the dietary RS intake difficult. Consequently, such estimations fall into a wide range, from ~4 g/d (Dysseler and Hoffem 1994a Like dietary fiber, RS will reach the colon where it will be fermented by the microflora (Björck et al. 1987 Several direct and indirect methods have been proposed for evaluation of the amount of RS delivered to the large bowel. Direct methods include analysis of starch in ileal effluents from ileostomy patients (Englyst and Cummings 1985 The in vivo methods mentioned above are laborious and thus not suitable for product evaluation. Several methods have therefore been developed for in vitro prediction of RS. The original procedures included milling and boiling (Englyst et al. 1982 In all of the procedures referred to here, the samples are milled before analysis. Consequently, with products containing RS 1, these methods are likely to underestimate the true RS content. To address this problem, Englyst et al. (1992) Yet another, more physiologic approach for in vitro determination of RS was suggested by Muir and O'Dea (1992) There is a lack of data concerning the RS content of realistic food items. The aim of this study was to develop a convenient and reliable in vitro method for the determination of RS. The method should, as far as possible, mimic physiologic conditions, including chewing before incubation with proteolytic and amylolytic enzymes. It should also be capable of including the major forms of RS likely to be present in foods, including physically encapsulated starch (RS 1). An additional objective was to design the method in a way that would allow parallel determination of dietary fiber. The test products were selected to include not only the commonly used raw B-type starches, but also to focus on realistic intact foods.
Test products
INTRODUCTION
Abstract
Introduction
Methods
Results & Discussion
References
, Stephen et al. 1983). Such starch, or starch degradation products delivered to the large bowel of healthy subjects, has been defined as resistant starch (RS) by EURESTA (1994). The first RS fraction to be identified in vitro was retrograded amylose, which appeared as dietary fiber during gravimetric (Asp et al. 1983) or gas-chromatographic determination of dietary fiber (Englyst et al. 1982). According to the terminology introduced later by Englyst et al. (1992), this type of RS is referred to as RS 3 and is found in corn flakes and white wheat bread, for example (Englyst and Cummings 1985). Other RS-fractions include starch that is resistant because of physical encapsulation such as that found in whole or partly milled grains or seeds (RS 1) and enzyme resistant B-type starches found in foods such as unripe bananas or native potato starch (RS 2) (Englyst and Cummings 1986, Langkilde and Andersson 1994). Although other starch fractions, such as thermally and chemically modified starch (Björck et al. 1989, Siljeström and Björck 1989), also may escape digestion and absorption in the small intestine, RS 1-RS 3 are likely to form the major sources in a mixed diet.
) to as much as 20-30 g/d, assuming a daily intake of 200-300 g starch (Cummings and Englyst 1989).
). During fermentation, short-chain fatty acids such as acetic, propionic and butyric acid are formed. It has been suggested that, in comparison with dietary fiber, RS yields a larger proportion of butyric acid (Scheppach et al. 1988, Silvester et al. 1995, Weaver et al. 1992). Because butyric acid is the main energy source for the colonocytes (Roediger et al. 1980), this may indicate a preventive role of certain forms of RS in relation to the development of colonic disease. As suggested by Annison and Topping (1994), however, RS from different sources might vary in the proportions of the major acids formed.
, 1986 and 1987, Jenkins et al. 1987a, Langkilde and Andersson 1994, Muir and O'Dea 1993, Steinhart et al. 1992), ileal intubation experiments (Noah et al. 1995, Stephen et al. 1983), balance experiments in rats with suppressed hind-gut microflora (Granfeldt et al. 1993) and analysis of the ileal excreta in colectomized rats (Marlett and Longacre 1996). Indirect estimations include analysis of different markers of fermentative activity in the colon such as breath H2 (Cummings and Englyst 1989, Muir et al. 1995b) and blood acetate (Cummings and Englyst 1989, Muir et al. 1995b).
, Johansson et al. 1984) and were thus not capable of including all forms of RS. In 1986, Berry introduced a modified procedure of the Englyst method from 1982. The most important alteration was that the boiling step was omitted, hence making it possible also to include RS 2 in the analysis. The original Berry method was further modified by Champ (1992) and later by Faisant et al. (1995).
proposed a method based on a more physiologic approach. Instead of milling, the sample is minced before enzymatic incubation in the presence of glass balls and guar gum, to resemble the disintegration occurring during chewing and passage through the gastrointestinal tract. The incubation is performed in a shaking water-bath with an enzyme mixture of amyloglucosidase, pancreatin and invertase. Aliquots are withdrawn after 20 min and 120 min for analysis of rapidly and slowly digested starch, respectively. The quantity of RS is then calculated as the amount of total starch from which the rapidly and slowly digested starch are subtracted. The method can also be used to analyze the three main forms of RS separately. In comparison with data obtained in ileostomy models, the method has shown good agreement for the limited number of products tested to date, mainly containing RS 3 (Englyst and Cummings 1985, 1986 and 1987, Englyst and Kingman 1990, Englyst et al. 1996). A good correlation was also shown when comparing the RS content in a mixed diet as measured in ileostomists with the RS content calculated from analysis of the individual foods according to the Englyst method (Silvester et al. 1995). The disadvantages of the method, however, are that it is complicated and time consuming. Moreover, collaborative studies have indicated that it appears difficult to obtain reproducible results in different laboratories (Dysseler and Hoffem 1994b).
. This method includes chewing as well as incubation with pepsin followed by further incubation with a mixture of porcine pancreatic enzyme and amyloglucosidase. The method has been validated by comparisons with ileostomy studies for a few realistic food products (Muir and O'Dea 1993, Muir et al. 1995a). Although this method was among the first to include chewing instead of mechanical disintegration of test products before in vitro testing, a problem with this method is that an amount of a sample, corresponding to only 0.1 g chewed material, is withdrawn for further enzyme incubations. In this way it is difficult to obtain a representative and homogenous sample, especially when the sample is coarse in texture and contains large amounts of RS 1, for which the method is intended.
MATERIALS AND METHODS
Abstract
Introduction
Methods
Results & Discussion
References
Bread products.
Six bread products were analyzed. Two of the breads were Arepas, a Latin American flat bread, based on flour from either ordinary (25% amylose), or high amylose corn (70-75% amylose). Recipes and baking conditions are available elsewhere (Granfeldt et al. 1993). Breads were also made from 80% barley, added either as scalded intact kernels or as a whole-meal flour, and 20% white wheat flour. The recipes and baking conditions have been described previously (Liljeberg et al. 1992). A commercial German rye pumpernickel bread (Gro
bäckerei Wendel, D-4594 Garrel, Germany) was obtained at the local market. Finally, a white wheat bread was included and baked according to Liljeberg and Björck (1994). All of the bread products, except for the Arepas, were sliced and the crust removed. The slices were wrapped in aluminum foil, put into plastic bags and stored in a freezer until used. The frozen bread products were thawed overnight at ambient temperature before analysis of RS.
Biscuits.
Three different biscuits were analyzed as follows: 1) 100 g cornstarch (Amioca, containing mainly amylopectin) was mixed with 0.25 L water. The starch was pregelatinized at 90-95°C for 5 min before being mixed with 50 g white wheat flour, 55 g egg, 30 g butter and 15 g sucrose. The biscuits were baked at 150°C for 1 h; 2) 100 g Amioca cornstarch was mixed with 0.2 L water (the starch was not submitted to pregelatinization). The starch-water mixture was then mixed with ingredients as described above. The biscuits were baked at 300°C for 5 min; 3) 100 g Hylon VII cornstarch (70% amylose) was mixed with 0.2 L water and prepared as described for biscuit no. 2. The biscuits were kindly provided by Dr. Furio Brighenti, University of Milan, Italy, and RS data for these products, obtained from balance experiments in ileostomy patients, have already been reported (Brighenti et al. 1996).
Breakfast cereals. Corn flakes were purchased locally (Kellogg Company, Svendborg, Denmark). Oat grains (Vårgårda, Sweden) were roasted in a continuous double-shell cylindrical drum, heated with steam (0.11 MPa, 120°C). The processing time was 20-25 min and the final temperature of the grains was ~95°C. The grains were steamed (0.11 MPa, 120°C) for 17-18 min to a final temperature of ~104 °C. Finally, the grains were flaked to a thickness of 0.5 mm.
Barley and rice products.
The barley used was from a commercial mixture provided by Kungsörnen (Järna, Sweden). Pearled kernels were boiled either for 85 min with a water/kernel ratio of 4:1 (wt/wt), or for 20 min with a water/kernel ratio of 2:1 (wt/wt) (Liljeberg et al. 1992). The added water was removed during boiling. Regular long-grain white rice (Österberg & Löfquist, Stockholm, Sweden) was purchased at the local market. The rice was boiled as is or as a flour (milled with a laboratory mill Cyclotec, Tecator, Sweden to a particle size 
0.5 mm). The intact rice was boiled with a water/rice ratio of 2:1 (wt/wt) for 15 min during which essentially all of the free water disappeared, and the rice flour was cooked for 2 min 30 s as a porridge with a water/rice flour ratio of 3:1 (wt/wt).
Spaghetti. Dried spaghetti made of 100% durum wheat (Turelli, Kungsörnen, Järna, Sweden) was added to boiling water (water/pasta ratio 9:1, wt/wt) and was boiled for 9 min according to instructions from the manufacturer.
Potato products. Potatoes (Bintje) of ~100 g size were peeled, put into boiling water in an amount to cover the potatoes and boiled for 30 min. Some potatoes were analyzed directly after boiling, whereas the remainder were further stored in a refrigerator at 5°C for 24 h before analysis.
Legume products.
Boiled red lentils (Friggs, Bromma, Sweden), autoclaved white beans (Generale Conserve, Theix, France), boiled white beans (S. C. Galec, Issy-les-Moulineaux, France) and a bean flake porridge (Nestlé, Vevey, Switzerland) were included. The red lentils were boiled for 10 min with a water/lentil ratio of 2:1 (wt/wt). All of the added water was removed during boiling. The autoclaved white beans, purchased at the local market, were heated for 5 min in the canning water. After the beans were heated, the remaining water was discarded. The boiled white beans, kindly provided by Dr. Martine Champ (INRA, Nantes, France), were prepared according to Noah et al. (1995). Dry beans were soaked for 4 h with a water/bean ratio of 3:1(wt/wt) and then boiled for 1 h 45 min with a soaked bean/water ratio of 1:6 (wt/wt). Finally, they were frozen at °C overnight before analysis. The bean flakes, prepared from beans according to Tappy et al. (1986), were boiled as a porridge. The porridge, containing a water/flake ratio of 3:1 (wt/wt), was boiled for 2 min 30 s.
Green banana. A few totally unripe green bananas were obtained from the local market and were analyzed as is.
Native and processed starches.
Although the intention was to analyze the RS-content in realistic foods, some starches were included as well. Two cornstarches were analyzed, native and extruded retrograded high amylose cornstarch (Hylon VII, 70% amylose, Cerestar, Vilvoorde, Belgium). The extruded, retrograded cornstarch was identical to the one included in EURESTA (1994) and was prepared as follows: Hylon VII starch was extruded, milled, stored at 4°C for 48 h and then dried and milled (Würsch and Delcour 1994). Two potato starches were also analyzed, native potato starch (Lyckeby Stärkelsen, Nöbbelöv, Sweden), purchased at the local market, and pregelatinized potato starch (Roquette, Lestrem, France), obtained from the EURESTA project (EURESTA 1994).
Chemical analysis
Total starch.
Total starch was analyzed in milled test products by using sequential incubations with Termamyl and amyloglucosidase after solubilization in 2 mol/L KOH (Siljeström et al. 1988).
Separation of total starch into potentially available starch and RS. In Figure 1, the schedule for the analysis of RS is shown. Six healthy subjects participated, five females and one male, mean age 33 ± 6 y. The subjects were told to brush their teeth, without using any toothpaste, before chewing the test products under standardized conditions. An amount of product, corresponding to 1 g total starch, was chewed 15 times for ~15 s. The chewing procedure was adopted from Granfeldt et al. (1992), who originally introduced this approach for substrate disintegration when attempting to predict glycemic responses from the rate of in vitro starch hydrolysis. Products that had been cooked were allowed to reach room temperature before chewing to obtain a stable weight. After chewing, the product was expectorated into a beaker containing 5 mL of distilled H2O and 1 mL pepsin solution [(pepsin) = 50 g/L, 2000 FIB-U/g, Merck, Darmstadt, Germany]. The subjects rinsed their mouths with another 5 mL of distilled H2O for 60 s. The rinsing solution was also expectorated into the beaker. The pH was adjusted to 1.5 with 2 mol/L HCl and the samples were incubated at 37°C for 30 min. During the incubation, the samples were stirred every 10 min. With powdery materials, glass beads were "chewed" instead to obtain saliva to be used in the analysis. This was performed to prevent powdery materials such as native potato starch from remaining between the teeth. The glass beads and the rinsing solution were expectorated into the same beaker; then the water-saliva solution was transferred into a beaker containing the sample, pepsin and distilled water.
|
Analysis of potentially available starch.
The filtrates were diluted to a concentration suitable for analysis. Because they contained a high concentration of ethanol, they were diluted on a weight basis instead of on a volume basis. After centrifugation at 1500 × g, the glucose concentration was analyzed spectrophotometrically at 450 nm with the use of a a glucose oxidase-peroxidase reagent. The possible effect of the ethanol on the oxidase-peroxidase reagent was examined and was found to be negligible. The filtrates were checked for the possible presence of starch fragments by comparing the glucose concentration in the filtrate with and without acid hydrolysis by using 0.75 mol/L HCl, a concentration previously used for hydrolysis of neosugar (Berggren 1996). The incubation was performed for 10 min at 60°C and then for 30 min at 20°C before the sample was neutralized with 2 mol/L NaOH. No differences were found, suggesting that the potentially available starch fraction had been degraded to glucose and thus did not consist of soluble low molecular weight dextrins, i.e., soluble RS.
Analysis of RS.
The filter residues were ground in a mill for 1 min (IKA A10, IKA Labor Technik, Germany). This was done to release physically and botanically encapsulated starch to the analytical amylases. An amount corresponding to 2 × 0.1 g of each replicate was pooled for analysis of protein and ash, respectively. The rest of the filter residue, but not more than 0.5 g, was used for determination of total starch after a 30-min solubilization in 2 mol/L KOH as described above (Siljeström et al. 1988).
Dietary fiber.
The total fiber content was determined gravimetrically as the total weight of the filter residue, obtained with the chewing procedure, from which the amount of celite, protein, ash and RS was subtracted. Nitrogen content was determined by the Kjeldahl method, and a factor of 6.25 was used to calculate the protein content. The ash content of the filter residues was measured after an overnight (16 h) incineration in a muffle furnace at 500°C. For a comparison, the total fiber content was also analyzed gravimetrically according to Asp et al. (1983). To allow comparison with the dietary fiber values obtained with the "chewing" procedure, these dietary fiber values were also corrected for total starch remaining in the residues.
Statistical analysis
Linear regression was done using the calculation program Microsoft Excel (Microsoft Corporation, Redmond, CA). The relationships between this method and in vivo data, or the in vitro methods according to Muir and O'Dea (1993) and Englyst et al. (1992), respectively, were calculated for different food products. In addition, the correlation between in vitro data obtained by the Englyst method (1992) and in vivo data was calculated for the native and processed starches. Finally, the correlation between the amount of dietary fiber, as analyzed with the proposed method and according to Asp et al. (1983), was calculated. The P-value for the correlations was estimate using the SPSS/PC+ program (SPSS, Chicago, IL).
The standard deviation and standard error of the mean was derived using Microsoft Excel (Microsoft Corporation).
| |
RESULTS AND DISCUSSION |
|---|
|
Abstract
Introduction
Methods
Results & Discussion
References
|
|---|
RS and "potentially available" starch
In Table 1, the RS and potentially available starch contents are listed for the different food products. For comparison, RS values of similar or identical products, as determined in ileostomy patients with intubation experiments or in antibiotic-treated rats, are included. The range in RS concentration was considerable and varied from <1 g/100 g in white wheat bread to 72 g/100 g in raw green banana (starch basis).
|
1.4 g/100 g. A slight overestimation was seen with boiled white beans, rice flour porridge and barley kernels boiled for 20 min. In contrast, an underestimation of ~8 g/100 g was seen with the high amylose Arepas. A possible explanation for this might be that the filtrate contained some of the starch as dextrins, which are not determined with the glucose oxidase-peroxidase reagent. However, no such soluble dextrins could be detected after acid hydrolysis with 0.75 mol/L HCl and analysis of the filtrate obtained from this product.
RS correlation with in vivo data
In Figure 2, RS figures for 19 composite and/or intact foods analyzed with this method have been plotted vs. in vivo data for similar or identical products obtained from the literature. Most of the in vivo data came from ileostomy studies. In the case of six products, biscuits made of high amylose cornstarch, waxy cornstarch and pregelatinized waxy cornstarch, Arepas made of high amylose and normal corn, and boiled white beans, in vivo and in vitro figures were obtained from analysis of identical products. The other foods included were prepared to resemble as much as possible the conditions reported in the literature during in vivo evaluation of RS content. The suggested in vitro method showed a very good agreement with corresponding in vivo data, resulting in a correlation coefficient of r = 0.97, P < 0.001, (y = 0.77x + 0.45) (Fig. 2).
|
RS content in different realistic foods and pure starchy materials
Bread products.
The flour-based bread products, i.e., the bread made from 80% whole-meal barley flour and 20% white wheat flour or 100% white wheat flour, contained low levels, i.e., 1.5 g/100 g RS (starch basis), possibly in the form of RS 3, that is, starch resistant due to retrogradation of amylose. RS values for white wheat bread and whole-meal bread obtained by the Englyst method (Englyst et al. 1992) were similar to those obtained by this method, i.e., 1.3 and 1.7 g/100 g RS (starch basis), respectively. In contrast, if the integrity of the botanical structure was maintained, as in the kernel-based bread, the amount of RS increased significantly. Consequently, in the bread with 80% barley kernels, as much as 26 g/100 g RS (starch basis) was found. This is a considerable amount compared with the range reported for common flour-based bread in ileostomy studies, about 2-5 g/100 g (Englyst and Cummings 1985, Jenkins et al. 1987a, Steinhart et al. 1992). The commercial pumpernickel bread, based on whole-meal flour and a low concentration of cracked rye kernels, also contained significant amounts of RS, 11 g/100 g (starch basis). This figure corresponds well with ileostomy data from similar products (Table 1). A major cause for the high RS content in pumpernickel bread is probably that the particular baking conditions used (20 h, 120°C) promote amylose retrogradation. Consequently, the RS content of this product was high, ~7 g/100 g, when analyzed according to Johansson et al. (1984) and Liljeberg and Björck (1994), that is, after milling and boiling of the sample.
Biscuits.
The RS concentration varied from 1 g/100 g for the biscuits made of waxy cornstarch to 33 g/100 g for the biscuits made from high amylose cornstarch. The in vitro RS figures for the three biscuits were within the range of RS recovered for identical products in ileostomy patients as recently reported by Brighenti et al. (1996).
Breakfast cereals. The steamed oat flakes contained low amounts of RS, <1 g/100 g (starch basis). This is a level similar to that in white wheat bread. The result from this study agrees well with ileostomy data with oat flakes (Table 1). Consequently, despite comparatively mild conditions for heat treatment, steaming and rolling obviously render the starch more or less completely digestible.
and Muir and O'Dea (1993), RS content of corn flakes as predicted in vitro was 4 or 3 g/100 g, respectively, which is similar to the yield with this method.
Barley and rice products.
The boiled barley kernels contained important amounts of RS, or 28 g/100 g (starch basis) in the case of kernels boiled for 20 min. This is similar to the results with bread with 80% scalded barley kernels, which contained 26 g/100 g RS (starch basis). A prolongation of the boiling time to 85 min significantly reduced RS content to 16 g/100 g. This was probably related to a more extensive disintegration of the botanical tissue, hence increasing the availability of the starch for enzymatic attack. Consequently, the kernels boiled for 85 min were more disrupted in appearance. According to Muir and O'Dea (1993), the RS yield in barley kernels boiled for 85 min, as estimated in ileostomists, was 6 g/100 g (starch basis), with an even lower in vitro RS content, 3 g/100 g. The discrepancy between the RS values obtained in vivo by Muir and O'Dea (1993) and that obtained by this in vitro method might be explained by the fact that the barley kernels had been subjected to different techniques for pearling. Hence, differences in the extent of removal of the botanical structure can be expected to have an effect on the RS 1 content of the finished product.
and were found to be lower (3 and 1 g/100 g, respectively) than those obtained by this method. These differences might have emanated not only from differences in methodology, but also from differences in the amylose content of rice. Consequently, the concentration of RS (mainly in the form of RS 3) varied from <1 g/100 g in a waxy rice genotype to 5 g/100 g in a high amylose genotype with 36% amylose (Björck et al. 1995).
Spaghetti.
The RS concentration in spaghetti, 5 g/100 g (starch basis), is in accordance with data based on studies in ileostomy patients (Jenkins et al. 1987a) (Table 1). Englyst et al. (1992) found a similar RS content in vitro (6 g/100 g). Thus, in comparison with many other flour-based cereal products, the RS concentration of spaghetti is high, which is probably related to the physical structure of the product.
Potato products.
The two potato products analyzed were either boiled, or boiled and stored at 5°C for 24 h. The stored potatoes had a significantly higher concentration of RS (7 g/100 g, starch basis) than the potatoes tested immediately after boiling (4 g/100 g), showing that the RS yield can be increased by the choice of processing conditions, in this case storage at low temperature. The increase of the RS yield during storage is probably due to the fact that retrogradation of the amylopectin fraction was favored at these conditions. Previously, Englyst and Cummings (1987) reported RS values of 3 and 12 g/100 g (starch basis) for boiled and cool-stored potatoes, respectively, as determined in ileostomists. Hence, the RS concentration in the boiled potatoes was similar to that found in this study, whereas the amount of RS in the stored product was higher. Possibly, the temperature profile during storage may affect the yield of retrograded amylopectin and hence the enzyme resistance in the stored product. In vitro data by Englyst et al. (1992), however, have indicated high levels also in boiled potatoes, 7 g/100 g, suggesting a variability due to variety and/or sample handling.
|
Legume products.
Among the legume products, the highest RS value was seen with the boiled red lentils (23 g/100 g, starch basis). In comparison with in vivo and in vitro data for lentils (14-17 g/100 g), this is a higher value, which might be due to that the fact that different kinds of lentils were analyzed (red lentils in the ileostomy studies by Steinhart et al. 1992 and Jenkins et al. 1987a, and unspecified types in the in vitro study by Englyst et al. 1992). The RS concentrations of boiled and autoclaved white beans were 17 and 14 g/100 g (starch basis), respectively. In the case of the boiled white beans, there are RS data from intubation studies available for an identical product. In addition, RS values with the Englyst method (Englyst et al. 1992) are also available for this particular product (Noah et al. 1995). The results showed good agreement with the RS concentration obtained with this method, 17 g/100 g, based on the Englyst method, vs. 16 g/100 g in vivo (Noah et al. 1995). The RS concentration of the autoclaved beans, however, was higher (14 g/100 g) than that reported previously in vivo and in vitro for a similar product (6 and 7 g/100 g, respectively, Muir and O'Dea 1993). The bean flake porridge contained 12 g/100 g RS. This value is within the same range reported previously for bean flakes tested without boiling according to the Englyst procedure (11 g/100 g, Englyst et al. 1992), or in balance experiments in antibiotic-treated rats (8 g/100 g, Ekwall et al., unpublished data). The explanation for the high RS levels in legume products is that the cotyledon cell walls in legumes retain their integrity during food processing, thus reducing the enzyme availability through physical encapsulation.
Green banana.
The intact totally green bananas contained a substantial amount of RS, 72 g/100 g (starch basis), in this case RS 2. Ileostomy data for green bananas have indicated a lower level, or 57 g/100 g (starch basis) (Englyst and Cummings 1986). The reason for this discrepancy might be related to different degrees of maturity of the bananas tested. Hence, during ripening, the starch in bananas is converted into sucrose, a process that may also affect the enzyme availability of the remaining starch.
Native and processed starches. RS values of native and processed corn and potato starches are listed in Table 2, together with ileostomy and in vitro data from the literature. The cornstarch was prepared from high amylose genotypes analyzed either in a native form, yielding 73 g/100 g RS (starch basis), or after extrusion and retrogradation to alter the amount of RS, giving 26 g/100 g RS (starch basis). Native potato starch is highly enzyme resistant because it is of B-type. However, when the starch is gelatinized, it becomes readily available to amylolytic enzymes and the amount of RS decreases. This is also evident from Table 2, which shows RS values of 66 and 1 g/100 g (starch basis) for native and pregelatinized potato starch, respectively.
showed a high correlation (r = 0.997). The reason for the underestimation of RS in the pure starches with this method might be due to an incomplete hydrolysis of the starch in the filter residue. Modifications similar to those suggested by Faisant et al. (1995) for improved determination of RS in high RS products analyzed according to Champ (1992) might solve the problem.
Dietary fiber
The dietary fiber contents obtained gravimetrically with the developed method and those obtained with the original gravimetric procedure introduced by Asp et al. (1983) are shown in Figure 3. When correcting for RS in the residues, good agreement was obtained (r = 0.97). However, apparent dietary fiber values as estimated with the chewing procedure, increased importantly when including RS as fiber. Consequently, for green banana, the concentration of undigestible carbohydrates, i.e., "dietary fiber," rose from 11 to 66 g/100 g (dry matter basis). In pumpernickel bread and barley kernels boiled for 85 min, the contribution of undigestible carbohydrates was prominent both from fiber (16 g/100 g for both) and RS (8 and 10 g/100 g, respectively).
|
|
Reproducibility
The pooled standard deviation for the developed method was 2.9%. This shows that the reproducibility of the method was high and that the method provides a reliable tool for in vitro estimation of RS contents in foods.General discussion
As judged from Figure 2, this method, which is based on chewing, appears to estimation RS content in realistic food products. Few in vitro data are available, particularly for food products with a more or less intact botanical integrity. An evaluation of data for five food items (barley kernels boiled for 85 min, white autoclaved beans, corn flakes and rice boiled either intact or as a flour) vs. in vitro RS figures obtained according to Muir and O'Dea (1993), resulted in a significant correlation (y = 0.34x + 1.09, r = 0.92, P = 0.025). However, a higher correlation (r = 0.99, P < 0.001) was obtained with the Englyst procedure (Englyst et al. 1992) (Fig. 4). In this case, data from the literature for eight food products and four pure starchy materials were included. For three of the pure starches, this method was applied on identical materials obtained from the EURESTA program (EURESTA 1994).
). Apart from chewing, in the gastrointestinal tract, the food is also subjected to the amalgamation of the stomach. This form of disintegration is of course difficult to simulate by artificial means. Although it is frequently suggested that the particle size of food components leaving the stomach is ~2-4 mm in diameter (Holt et al. 1982), this has been demonstrated for only a few realistic test meals, and to our knowledge not with foods such as starchy foods with botanical integrity, e.g., products with intact cereal kernels for which the chewing procedure was actually introduced. It should also be acknowledged that because food particles are irregular in shape, the actual size is probably underestimated with procedures using glass beads as a tool for measurement of particle size distribution. We therefore believe that the chewing procedure introduced here more closely mimics in vivo conditions than does the method of Faisant et al. (1996), which suggests milling the sample to a particle size <3 mm, or the method of Englyst et al. (1992), which uses mincing and further mechanical disintegration with glass balls.
). In addition, the RS fraction will provide substrate to the colonic microflora, thus promoting short-chain fatty acid production in the colon with potential health benefits (Roediger et al. 1980, Wolever et al. 1991).
| |
FOOTNOTES |
|---|
Manuscript received 11 April 1997. Initial reviews completed 16 May 1997. Revision accepted 14 October 1997.
| |
LITERATURE CITED |
|---|
|
Abstract
Introduction
Methods
Results & Discussion
References
|
|---|
in vitro and in vivo studies.
Cereal Chem.
1988;
65:1-8 This article has been cited by other articles:
|
|
 |
|
  A. C. Nilsson, E. M. Ostman, J. J. Holst, and I. M. E. Bjorck Including Indigestible Carbohydrates in the Evening Meal of Healthy Subjects Improves Glucose Tolerance, Lowers Inflammatory Markers, and Increases Satiety after a Subsequent Standardized Breakfast J. Nutr., April 1, 2008; 138(4): 732 - 739. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. C Nilsson, E. M Ostman, Y. Granfeldt, and I. M. Bjorck Effect of cereal test breakfasts differing in glycemic index and content of indigestible carbohydrates on daylong glucose tolerance in healthy subjects Am. J. Clinical Nutrition, March 1, 2008; 87(3): 645 - 654. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K.-H. Han, M. Sekikawa, K.-I. Shimada, K. Sasaki, K. Ohba, and M. Fukushima Resistant Starch Fraction Prepared from Kintoki Bean Affects Gene Expression of Genes Associated with Cholesterol Metabolism in Rats Experimental Biology and Medicine, September 1, 2004; 229(8): 787 - 792. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Henningsson, I. M. E. Bjorck, and E. M. G. L. Nyman Combinations of Indigestible Carbohydrates Affect Short-Chain Fatty Acid Formation in the Hindgut of Rats J. Nutr., October 1, 2002; 132(10): 3098 - 3104. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. L. Topping and P. M. Clifton Short-Chain Fatty Acids and Human Colonic Function: Roles of Resistant Starch and Nonstarch Polysaccharides Physiol Rev, July 1, 2001; 81(3): 1031 - 1064. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. G. Liljeberg, A. K. Akerberg, and I. M. Bjorck Effect of the glycemic index and content of indigestible carbohydrates of cereal-based breakfast meals on glucose tolerance at lunch in healthy subjects Am. J. Clinical Nutrition, April 1, 1999; 69(4): 647 - 655. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||
