The glycaemic index (GI) is the measure of the power of food to raise the concentration of B-glucose in the blood after a meal. GI is defined as the ratio between the incremental area under the B-glucose response curve of the tested meal containing 50 g digestible carbohydrates and the incremental area under the B-glucose response curve of the standard food, i.e. 50 g pure glucose. Overall, carbohydrates that break down quickly during digestion have a high GI since their B-glucose response is fast and high. Carbohydrates that break down slowly have a low GIReference Jenkins, Kendall, Augustin and Vuksan1.
Several large-scale, observational studies from Harvard University (Cambridge, MA, USA) indicate that long-term consumption of a diet with high glycaemic load (GI × dietary carbohydrate content) is a significant independent predictor of the risk of developing type 2 diabetesReference Salmeron, Ascherio, Rimm, Colditz, Spiegelman, Jenkins, Stampfer, Wing and Willett2, Reference Salmeron, Manson, Stampfer, Colditz, Wing and Willett3 and CVDReference Liu, Willett and Stampfer4. More recently, evidence has been accumulating that a low-GI diet might protect against development of obesityReference Ludwig5, colon cancer and breast cancerReference Austin, Gallus and Bosetti6. The mechanisms of protection of the low-GI diet may derive from: (1) the slow release of carbohydrates in the upper gastrointestinal tract; (2) the lowering of the insulin demandReference Jenkins, Wolever and Ocana7; and (3) the generally high concentration of indigestible carbohydrates such as dietary fibre and resistance starch (RS), which, in turn, increase the fermentative activity at the colon level. This latter fermentative activity increases the synthesis of propionic acid which is considered the moderator of hepatic glucoseReference Venter, Vorster and Cummings8 and lipid metabolismReference Wolever, Brighenti and Royall9. Therefore, the low-GI diet has emerged as an interesting tool in decreasing diseases related to the metabolic syndrome and cereal fibres have been shown to possess a preventive effect (for reviews see Jenkins et al. Reference Jenkins, Kendall, Augustin, Francesci, Hamidi, Marchie, Jekins and Axelsen10 and Björck & LiljebergReference Björck and Liljeberg11).
Based on experimental and epidemiological evidence, the development of low-GI cereal products seems to be particularly relevant from a metabolic perspectiveReference Björck and Liljeberg11. Bread biotechnology as well some bread components may have an influence on glycaemia. Bread leavening by sourdough or the direct addition of acetic, propionic and lactic acids during bread making decreased the postprandial blood glucose and insulin responses (for review see Björck & LiljebergReference Björck and Liljeberg11). Cereal grains are mainly composed of starch and NSP such as β-glucan, polyfructan and arabinoxylan. Some of these polysaccharides are partially digestible, e.g. starch, while others are believed to serve as dietary fibres, e.g. arabinoxylanReference Campbell, Fahey and Bryan12. These dietary fibres exert beneficial effects on the reduction of the level of cholesterol and on decreasing the risk of colon cancer (for review see Jenkins et al. Reference Jenkins, Axelsen, Kendall, Augustin, Vuksan and Smith13). Nevertheless, white bread, the most commonly consumed, is manufactured from wheat flours which have a very low concentration of these dietary fibres. Currently, the biotechnology of white bread is considering the manufacture of baked goods enriched with dietary fibres. When fibres or wholemeal are included in the formulas of baked goods, the protocol of manufacture needs the set up of several technological parameters in order to get high-quality and consumer-acceptable breadsReference Björck and Liljeberg11, Reference Wang, Rosell and Benedito de Barber14.
The present work was aimed at decreasing the GI of white wheat flour (WF) bread by using selected sourdough lactobacilli and dietary fibres. The GI of selected bread was determined in vivo for fifteen healthy volunteers, and the structural and sensory properties of the resulting bread were characterized.
Materials and methods
Strains and culture conditions
Lactobacillus sanfranciscensis DA110, L. plantarum DC400 and P1, and L. brevis P2, belonging to the Culture Collection of the Dipartimento di Protezione delle Piante e Microbiologia Applicata, Università degli Studi di Bari, Italy, were used in the present study. Strains were selected previously based on the acidification capacity. Strains were routinely propagated for 24 h at 30°C in modified MRS broth (Oxoid, Basingstoke, UK) with the addition of fresh yeast extract (5 % v/v) and 28 mm-maltose at final pH 5·6.
Wheat flour breads fermented with sourdough lactobacilli: preliminary screening
The white WF (Triticum aestivum, var. S. Pastore) used had the characteristics reported in Table 1. L. sanfranciscensis DA110, L. plantarum DC400, L. plantarum P1 and L. brevis P2 were used for dough fermentation to select the most appropriate starter based on the formation of resistant starch (RS). Cells of each strain (24 h old) cultivated in modified MRS were harvested by centrifugation (9000 g, 10 min, 4°C), washed twice with 50 mm-sterile potassium phosphate buffer (pH 7·0) and resuspended in sterile distilled water to a final optical density at 620 nm of 2·5 (109 colony-forming units/ml). WF (312 g), 133 g sterilized distilled water and 50 g of the cellular suspension, containing 109 colony-forming units/ml of DA110, DC400, P1 or P2 strains (final cell number of approximately 5 × 108 colony-forming units/g dough), and 5 g baker's yeast were used to produce 500 g dough (dough yield, dough weight × 100/flour weight, 160) with a continuous high-speed mixer (60 g; dough mixing time, 5 min; Chopin & Co., Boulogne, France). A control dough without bacterial inoculum was prepared also. All the doughs were fermented at 30°C for 3 h and baked at 220°C for 50 min. All experiments were carried out using wholemeal flour (WMF) also (Table 1).
Table 1 Chemical composition of wheat flours and dietary fibres used in the present study
Bread making with dietary fibres, chemical acidification or sourdough
Three different dietary fibres (dextran, sugarbeet and oat fibres) were kindly supplied by Puratos n.v. (Groot-Bijgaarden, Belgium). The chemical composition of the dietary fibres is shown in Table 1. The following flours and fibres were used: (1) WF; (2) WMF; (3) WF and WMF in the ratio of 1:1 (WF/WMF); (4) WF/WMF and dextran; (5) WF/WMF and sugarbeet fibre; and (6) WF/WMF and oat fibre (Table 2).
Table 2 Ingredients used for bread making*
WF, white wheat flour bread; WF/WMF, WF and WMF in the ratio of 1:1; WMF, wholemeal flour bread; ΔpH, decrease of the initial pH of the dough.
* All dough contained 1·0 % of baker's yeast. For details about bread making, see Materials and Methods.
To study the influence of chemical acidification, the following ingredients were used: (1) WF/WMF; (2) WF/WMF acidified with acetic acid to decrease the initial pH of the dough by 0·5 unit (ΔpH 0·5); (3) WF/WMF acidified with a mixture of lactic and acetic acids (ratio of 1:4) to ΔpH 0·5; and (4) WF/WMF acidified with lactic acid to ΔpH 0·5, 1·0 or 1·5 (Table 2).
L. plantarum P1 and L. brevis P2 were selected and used as sourdough to ferment dietary fibre-enriched breads. The initial cell density in the dough varied according to the ΔpH to be reached after 3 h of fermentation at 30°C (1 × 107, 5 × 107 or 5 × 108 colony-forming units/g for ΔpH 0·5, 1·0 or 1·5, respectively). All WF, WMF and WF/WMF were fermented by sourdough to ΔpH 0·5, 1·0 or 1·5. Sourdough WF/WMF, containing dextran, sugarbeet fibre or oat fibre, was also fermented to ΔpH 0·5–1·5 (Table 2).
All the doughs were fermented at 30°C for 3 h and baked at 220°C for 50 min. For each condition, three independent experiments were carried out.
Determination of pH and physico-chemical analyses
The values of pH were determined by a Foodtrode electrode (Hamilton, Bonaduz, Switzerland). Moisture was determined according to the standard American Association of Cereal Chemists method15. Nitrogen content was measured by the semimicro-Kjeldahl method. Protein was calculated by multiplying the nitrogen concentration × 5·7. Carbohydrates were determined by HPLCReference Wang, Rosell and Benedito de Barber14. Soluble, insoluble and total dietary fibres were determined according to the American Association of Cereal Chemists method16.
Total starch was analysed as described by Holm et al. Reference Holm, Björck, Drews and Asp17. Factor conversion from glucose to starch was 0·9. The RS and potentially available starch was determined using the in vitro method described by Åkerberg et al. Reference Åkerberg, Liljeberg and Björck18. Factor conversion from glucose to starch was 0·9.
In vitro starch hydrolysis
The analysis of starch hydrolysis simulates the in vivo digestion of starchReference Liljeberg, Akerberg and Björck19. All bread portions contained 1 g starch and were given in randomized order to ten volunteers. Subjects rinsed their mouths with tap water and subsequently chewed the bread products for 15 s (approximately fifteen times). They were told not to eat within 1 h prior to the experiment. The products were then expectorated into a beaker containing 50 mg pepsin in 6 ml of 0·05 m-Na,K-phosphate buffer (containing 0·4 g/l NaCl) adjusted to pH 1·5 with 2 m-HCl. Finally, the subjects rinsed their mouths with 5 ml 0·05 m-Na,K-phosphate buffer (pH 6·9) for 60 s and expectorated the rinsing solution into the beaker. The contents were stirred and pH adjusted to 1·5. Each sample was incubated at 37°C for 30 min with gentle mixing three times during incubation. The pH was re-adjusted to 6·9 with NaOH before incubation with porcine pancreatin α-amylase (110 U; Sigma, St Louis, MO, USA). The enzyme was dissolved in 10 ml buffer, and 1 ml of this solution was added to the beaker. The sample was brought to volume (30 ml) with 0·05 m-Na,K-phosphate buffer, and transferred to the dialysis tubing (cut-off 12 400). Each bag was incubated at 37°C for 3 h in a beaker with 0·05 m-Na,K-phosphate buffer (800 ml). The beaker was placed in a stirred water-bath. Every 30 min, aliquots (2 ml) from the dialysate were removed and refrigerated. Then 3 ml 0·4 m-sodium acetate buffer (pH 4·75) were added to each aliquot, and 80 μl amyloglucosidase (140 U/ml) were used to hydrolyse the digested starch into glucose after 45 min at 60°C in a shaking water-bath. The glucose content was measured with Enzy Plus D-Glucose (Diffchamb, Italia srl, RM, Italy). The factor conversion from glucose to starch was 0·9. The rate of starch digestion was expressed as the percentage of potentially available starch hydrolysed at different times (30, 60, 90, 120 and 180 min). The hydrolysis curves were obtained with the following equation, using the program Statistica 6·0 (Statsoft Inc., Tulsa, OK, USA). Hydrolysis curves follow a first-order equation: C = C ∞(1 − e− kt), where C is the concentration at t time, C ∞ is the equilibrium concentration, k is the kinetic constant and t is the chosen time.
Glycaemic index in vivo test
Fifteen healthy non-smoking volunteers, eight women and seven men, aged 23–55 years, with normal BMI (21·5 kg/m2) and without drug therapy participated in the present study. Anhydrous glucose and two breads were used for the in vivo test: WF (control) and WF/WMF enriched with oat fibre (Table 2). Available carbohydrates of each bread and anhydrous glucose (50 g) were consumed in a random order on separate mornings after 10–12 h overnight fast. The analysis was carried out in triplicate: each volunteer consumed the same sample during a period of 3 d. Samples were consumed over 6 min with tap water. After anhydrous glucose or bread ingestion, capillary finger-prick blood samples (three or four drops) were taken from volunteers fasting (0 min) and after 15, 30, 45, 60, 75, 90, 105 and 120 min. Blood samples were collected into tubes and frozen at − 20°C before analysis by the glucose oxidase (EC 1.1.3.4) methodReference El20.
The areas under the glucose-response curves for each sample were calculated geometrically, excluding beneath the fasting level. GI was calculated by expressing the glycaemic response area for the WF bread and sourdough WF/WMF bread enriched with oat fibre as the percentage of the mean response area of the 50 g anhydrous glucose ingested by the same volunteers. The resulting individual GI values were averaged to obtain the GI value for WF bread and sourdough WF/WMF bread.
Bread characteristics
The determinations of bread mass and specific volume were carried out according to the American Association of Cereal Chemists 10–10 official method (AACC, 2000, St Paul, MN, USA).
The bread crumb grain was evaluated by using image analysis technology. Images of the sliced breads were captured using an Image Scanner (Amersham Pharmacia Biotech, Uppsala, Sweden). Images were scanned full scale at 300 dots per inch and analysed in grey scale (0–255). Image analysis was performed using the UTHSCSA ImageTool program (version 2.0; University of Texas Health Science Centre, San Antonio, Texas, available by anonymous FTP from maxrad6.uthscsa.edu). A threshold method was used for differentiating gas cells and non-cells. The threshold was determined according to the method of Crowley et al. Reference Crowley, Grau and Arendt21. Analysis was carried out on two sub-images of 500 × 500 pixels selected from within the bread slice. Two slices were analysed per treatment. Gas cell to total area ratio was recovered.
The laboratory acceptance panel was used to give an indication of consumer acceptance of the products under studyReference Knuckles, Hudson, Chiu and Sayre22. Breads, baked the day before sensory testing, were served at room temperature and under normal (daylight) illumination. A serving of each bread, identified by code numbers, on a single tray was served to each panelist. Ten volunteers from laboratory staff evaluated each product for quality attributes: colour, flavour/aroma, elasticity, sweetness and dryness. Acceptability of each quality attribute was rated with a score from 0 (lowest) to 100 (highest)Reference Haglund, Johansson and Dahlstedt23.
Statistical analysis
Experimental data were subjected to ANOVA and pair-comparison of treatment means was achieved using Tukey's procedure at P < 0·05 using the statistical software Statistica Windows version 6.0). For the GI in vivo test, the analyses were carried out in triplicate and the capillary finger-prick blood samples (three or four drops) were taken from volunteers fasting (0 min) and after 15, 30, 45, 60, 75, 90, 105 and 120 min after ingestion of the sample.
Results
Wheat flour breads fermented with sourdough lactobacilli: preliminary screening
Breads were preliminary fermented with baker's yeast alone or with sourdough lactobacilli and baker's yeast at 30°C for 3 h. The concentrations of RS and potentially available starch are shown in Table 3. Values of RS for breads made with white WF or WMF and fermented with baker's yeast were similar; they ranged from approximately 1·4 to 1·7 % (starch basis). The concentration of RS in breads made with WF and WMF and fermented by sourdough lactobacilli and baker's yeast was approximately double (3·36–5·04 %). The highest formation of RS was found (P < 0·05 %) in breads fermented with strains of L. plantarum and L. brevis. As a consequence, L. plantarum P1 and L. brevis P2 were selected for further studies. No significant differences (P < 0·05 %) for RS were found when WF or WMF were fermented with sourdough lactobacilli. The same was found when the mixture (1:1) of WF and WMF (WF/WMF) was used (data not shown). Further experiments were carried out by using WF/WMF flour.
Table 3 Potentially available starch and resistant starch in white wheat flour (WF) and wholemeal flour (WMF) breads fermented with baker's yeast alone or sourdough lactobacilli and baker's yeast*
(Mean values of three independent fermentations)
a,b,c Mean values within a column with unlike superscript letters were significantly different (P < 0·05).
* For details of ingredients and bread making, see Table 2 and Materials and Methods.
† Fermented with baker's yeast alone.
‡ Fermented with sourdough lactobacilli and baker's yeast.
Rate of in vitro starch hydrolysis of breads made with dietary fibres, chemical acidification and sourdough
The rate of in vitro starch hydrolysis is considered to be a presumptive measure of the GI in healthy subjectsReference Åkerberg, Liljeberg and Björck18. Starch hydrolysis at 30–180 min and values of the hydrolysis index (HI) are reported in Table 4 and Fig. 1, respectively. In agreement with the concentration of RS, the rate of starch hydrolysis did not vary between breads made with WF or WMF and fermented by baker's yeasts alone. Except for oat fibre, no decreases of the starch hydrolysis were found when the other fibres were used. The HI of the bread made with WF/WMF and enriched with oat fibre was 90·84 %.
Table 4 Degree of starch hydrolysis at various times*
(Mean values of three independent fermentations)
WF, white wheat flour bread; WF/WMF, WF and WMF in the ratio of 1:1; WMF, wholemeal flour bread; ΔpH, decrease of the initial pH of the dough.
a–e Mean values within a column with unlike superscript letters were significantly different (P < 0·05).
* For details of ingredients and bread making, see Table 2 and Materials and Methods.
† Fermented with baker's yeast alone.
‡ Fermented with sourdough lactobacilli and baker's yeast.
Fig. 1 Percentage of hydrolysis index of different breads calculated from 180 min areas. For details of bread making and ingredients, see Table 2 and Materials and Methods. Data are the mean of three independent fermentations. Sourdough WF/WMF, bread fermented with selected sourdough lactobacilli; WF, white wheat flour bread; WF/WMF, WF and WMF in the ratio of 1:1; WMF, wholemeal flour bread; ΔpH, decrease of the initial pH of the dough.
The effect of organic acids and related decreases of pH (ΔpH) were determined (Table 4; Fig. 1). Lactic and acetic acids were used since they represent the major fermentation end-products during sourdough fermentationReference Gobbetti, De Angelis, Corsetti and Di Cagno24. Overall, the rate of starch hydrolysis of breads containing acetic and lactic acids was significantly (P < 0·05) lower than non-acidified breads. Differences were found depending on the type and concentration of organic acids. In particular, the bread acidified with acetic acid to ΔpH 0·5 showed values of starch hydrolysis always higher than the corresponding bread acidified with lactic acid (Table 4). As expected, the values of the starch hydrolysis found with the mixture of lactic and acetic acids were intermediate. Besides, the rate of starch hydrolysis and HI decreased proportionally with the increase of ΔpH as set by lactic acid. After 180 min, the percentage of hydrolysed starch was 41·16, 35·72 and 29·27 for WF/WMF breads with ΔpH 0·5, 1·0 and 1·5, respectively. In agreement, the HI ranged between 84·07 and 66·05 % (Fig. 1).
Chemical acidification was compared to the microbial acidification by selected sourdough lactobacilli. The cell density of lactobacilli after 3 h of fermentation was approximately 5 × 108 (ΔpH 0·5), 2 × 109 (ΔpH 1·0) and 5 × 109 (ΔpH 1·5) colony-forming units/g dough. For the same values of ΔpH, all breads fermented with sourdough lactobacilli showed a rate of starch hydrolysis markedly lower than the corresponding chemically acidified breads (Table 4). After 180 min, the percentage of hydrolysed starch in WF/WMF was 38·10 (ΔpH 0·5), 34·0 (ΔpH 1·0) and 26·86 (ΔpH 1·5). In agreement, the HI ranged between 74·54 and 59·62 %.
Selected sourdough lactobacilli were further used for bread making with enriched dietary fibres. The growth of sourdough lactobacilli was not affected by the addition of fibres (data not shown). Compared to sourdough fermented breads, the addition of dextran or sugarbeet fibre did not modify the degree of starch hydrolysis (Table 4). For instance, the bread made with sourdough WF/WMF and enriched with dextran had values of HI of 70·58 and 57·38 % when ΔpH 0·5 and 1·5 were reached, respectively (Fig. 1). On the contrary, the addition of oat fibre to WF/WMF fermented with sourdough lactobacilli to ΔpH 1·5 gave the lowest value (23·47 %) of starch hydrolysis (Table 4). In agreement, the HI was 47·65 % (Fig. 1).
Glycaemic index in vivo test
Bread made with sourdough WF/WMF enriched with oat fibre with ΔpH 1·5 was used for the in vivo determination of the blood glucose level and GI since it was characterized by the lowest value of HI. The bread made with WF and fermented with baker's yeast alone was used as the control since it was made under the same conditions, except for the use of WMF, sourdough lactic acid bacteria and oat fibre. Anhydrous glucose was used as reference. All volunteers consumed anhydrous glucose and breads according to the experimental protocol. The mean incremental blood glucose response curves during 120 min after sample ingestion are shown in Fig. 2. The GI of WF was approximately 72 % of the reference (anhydrous glucose, GI = 100). During 120 min after bread ingestion, the blood glucose increment of sourdough WF/WMF bread was significantly (P < 0·05) lower than that of WF bread. In agreement, the GI of sourdough WF/WMF bread enriched with oat fibre was approximately 53·7 %.
Fig. 2 Blood glucose increment (mmol/l) in healthy volunteers following ingestion of anhydrous glucose (reference; ●), white wheat flour bread (WF; ■) and WF/WMF (1:1) bread enriched with oat fibre and fermented with sourdough lactobacilli to ΔpH 1·5 (sourdough WF/WMF; 
). Values are means with their standard errors depicted by vertical bars (n 15). Mean sourdough WF/WMF values were significantly different from those of the WF group: *P < 0·05. ΔpH 1·5, decrease of the initial pH of the dough by 1·5 units.
Bread characteristics
Breads made with WF or WMF and fermented with baker's yeast alone, and sourdough WF/WMF bread were analysed for volume, bread mass, specific volume, crumb grain and sensory characteristics. WF bread had a higher volume (approximately 1340 ml) than WMF bread (670 ml; Table 5). The volume of sourdough WF/WMF bread enriched with oat fibre was intermediate (997 ml). On the contrary, no significant (P < 0·05 %) differences were found for bread masses. Consequently, sourdough WF/WMF showed an increased specific volume (73·91 %) compared to WMF bread (49·16 %).
Table 5 Volume, mass and sensory characteristics of white wheat flour (WF), wholemeal flour (WMF) and sourdough WF/WMF breads*
(Mean values of three independent fermentations)
WF/WMF, WF and WMF in the ratio of 1:1.
a,b,c Mean values within a row with unlike superscript letters were significantly different (P < 0·05).
* Details about ingredients and bread making are described in Table 2 and Materials and Methods.
† Fermented with baker's yeast alone.
‡ Enriched with oat fibre (5 %) and fermented with sourdough lactobacilli and baker's yeast.
The bread crumb grain was determined by image analysis technology. Digital images were pre-processed to detect crumb cell-total area by a binary conversion (black/white pixels; Fig. 3). The cell-total area of WF bread (corresponding to the black pixel total area) was 46 %. Crumb cell detection for WMF bread slices showed smaller and less numerous gas cells, as shown by the visual inspection of the binarized images of the breads. Indeed, a considerable variation of the cell-total area with respect to WF bread (17 % v. 46 %) was found. Substantial differences were found in the crumb features of sourdough WF/WMF bread enriched with oat fibre. The addition of WF and sourdough fermentation resulted in a larger gas cell-total area (38 %) with respect to WMF bread (17 %).
Fig. 3 Representative images of breads made with white wheat flour (WF) (A, B) or wholemeal flour (WMF) (panel C, D) and fermented with baker's yeast alone, and 1:1 WF/WMF enriched with oat fibre and fermented with sourdough lactobacilli to ΔpH 1·5 (E, F). (A, C, E), Digital images of bread showing the original grey level images; (B, D, F), computed binary results from grey level thresholding at the two-cluster. ΔpH 1·5, decrease of the initial pH of the dough by 1·5 units.
Overall, sourdough WF/WMF bread enriched with oat fibre showed a higher acceptability score than WMF bread (Table 5). As expected, the presence of WMF in sourdough WF/WMF bread influenced the colour. It was darker than WF bread. Due to the lactic acid fermentation, the sourdough WF/WMF bread enriched with oat fibre showed a more pronounced acidulous smell and taste, and a more intense aroma with respect to WF or WMF breads.
Discussion
Sourdough fermentation with selected lactobacilli improved the nutritional features of starch in white WF breads. All breads fermented with lactobacilli contained higher concentrations of RS than breads started with baker's yeasts alone. Based on its definition, RS includes starches and their partial hydrolysis products that are not digested or absorbed in the small intestinal lumen of healthy personsReference Asp, Tovar and Bairoliya25. The main physiological influence of RS is concerning the: (1) inhibition of the postprandial increase of blood glucose and insulinReference Yamada, Hosoya, Nishimura, Tanaka, Kajimoto, Nishimura and Kajimoto26, Reference Morita, Hayashi, Motoi, Yagishita, Takeya, Sugiyama and Kiriyama27; (2) decrease of the blood lipid concentration; and (3) decrease of the concentrations of ammonia, phenol and secondary bile acid in the colonReference Muir, Yeow, Keogh, Pizzey, Bird, Sharpe, O'Dea and Macrae28. In agreement with previous studiesReference Holm and Björck29–Reference Granfeldt, Björck and Björck31, the concentration of RS ( < 2·0 % starch) in white WF and WMF breads started with baker's yeast alone was too low to significantly interfere with the glycaemic response. An increase of the RS was found in organic acid added or sourdough fermented pumpernickel breadsReference Liljeberg, Akerberg and Björck19. Indeed, selected sourdough L. plantarum and L. brevis strains caused the highest increase of RS in breads made with WF and WMF. Overall, the use of WMF did not influence the concentration of RS.
Consumption of foods naturally rich and/or enriched in cereal fibres is recommended to prevent several diseasesReference Murtaugh, Jacobs, Steffen, Jacob and Marquart32–Reference Larsson, Giovannucci, Bergkvist and Wolk34. Nevertheless, a large gap between dietary recommendations and daily intake of fibres is currently present in the diet of a large part of the populationReference Lang and Jebb35. To enrich cereal baked goods with fibres by keeping optimal sensory properties is probably the most desirable goalReference Björck and Liljeberg11, Reference Wang, Rosell and Benedito de Barber14. The present study considered the use of WF in a mixture with WMF (1:1) and the addition (5 %) of different fibres (dextran, sugarbeet fibre or oat fibre) as ingredients. The mixture of WF/WMF was chosen to increase the cereal fibre content in breads. Overall, the rate of in vitro starch hydrolysis was similar in all the breads fermented with baker's yeast. As shown by the HI, only the addition of oat fibre decreased the starch availability in the bread made with WF/WMF. A large and steadily expanding body of clinical evidence attests to the extraordinary therapeutic significance of oat soluble fibre. The primary health benefits include the reduction of cholesterolReference Behal, Scholfield and Hallfrisch36 and other risk factors for heart disease, also characterized as the metabolic syndrome (Syndrome X)Reference Ginsberg and Stalenhoef37. In 1997, the Food and Drug Administration enacted a health claim addressing the cardiovascular benefits of oat β-glucan38. A 100 g portion of bread made with WF/WMF and enriched with oat fibre contained approximately 750 mg β-glucan. This amount still represents the minimum effective daily sub-dose to obtain positive effects38. As also discussed later, the results of the present study suggest an effect of oat fibre also on the decrease of the postprandial glycaemic response.
Chemical acidification decreased the rate of starch hydrolysis and HI in breads. Acetic, propionic and lactic acids have the capacity to lower the postprandial blood glucose and insulin responses in man when included in the bread formulaReference Liljerberg and Björck39, Reference Östman, Nilsson, Liljeberg Elmstahl, Molin and Björck40. Breads made with the mixture of WF/WMF showed the highest decrease of HI when lactic acid was used. The decrease of HI seemed to be concentration-dependent and the lowest values of HI were found for ΔpH 1·5. Overall, the lowering of glycemia and insulinaemia by lactic acid are not attributed to a decreased gastric emptying rateReference Liljerberg and Björck39 but to interactions between starch and gluten that limited the starch bioavailabilityReference Östman, Nilsson, Liljeberg Elmstahl, Molin and Björck40. For the same values of ΔpH, the biological acidification of WF/WMF by sourdough L. plantarum P1 and L. brevis P2 showed a higher decrease of the rate of starch hydrolysis and HI with respect to chemical acidification. This agreed with a previous study on wholemeal barley flour breads produced by sourdough, lactic acid or sodium propionateReference Mendosa41 (www.gilisting.com/2004/05/glycemic-index-of-breads.html).
When WF/WMF was enriched with dietary fibres and fermented by sourdough lactobacilli only oat fibre caused a marked decrease in the rate of starch hydrolysis and showed the lowest HI. The in vivo challenge on fifteen healthy volunteers showed that this bread caused a lower blood glucose increment compared to bread made with WF and fermented with baker's yeast alone. This bread was included in the analysis since it is consumed in large amounts in southern Europe. The value of GI calculated from glycaemic responses for sourdough WF/WMF bread enriched with oat fibre was significantly (P < 0·05) lower than that of WF bread (approximately 53·7 and 72 %, respectively).
The sourdough WF/WMF bread enriched with oat fibre was also characterized for volume, mass, specific volume, crumb grain and sensory properties. Almost all these characteristics indicated a superior quality of sourdough WF/WMF bread enriched with oat fibre compared to bread made with WMF and fermented with baker's yeast alone. In particular, sourdough WM/WMF showed an increased specific volume and a better cell crumb structure. In agreement with previous resultsReference Sapirstein, Roller and Bushuk42, Reference Gallagher, Gormley and Arendt43, the cell-total area of the bread made with WF was 46 %. Total cell-area of sourdough WF/WMF bread enriched with oat fibre bread was only 8 % smaller than that of WF bread. Besides, sourdough WF/WMF bread enriched with oat fibre showed a more pronounced acidulous smell and taste, and an intense and more appreciated aroma with respect to WF or WMF breads. These sensory attributes are probably related to the complementary fermentative activities of L. plantarum P1 (facultative heterofermentative lactic acid bacterium) and L. brevis P2 (obligate heterofermentative lactic acid bacterium)Reference Gobbetti, De Angelis, Corsetti and Di Cagno24.
In most European countries, bread constitutes the major source of dietary carbohydrates. Bread with decreased GI could be included in a low-GI diet. The higher dietary fibre content frequently associated with low-GI foods may also add to the metabolic merits of a low-GI diet. Consequently, the exchange of common high-GI bread for sourdough WF/WMF bread enriched with oat fibre, as the only dietary modification, may improve the insulin economyReference Björck and Liljeberg11. The present study proposed a new biotechnology for the manufacture of a functional white WF bread enriched with oat fibre and fermented with selected sourdough lactobacilli which combined the: (1) physiologically significant supply of β-glucans; (2) decrease of the GI to 53·7 % as estimated by the in vivo challenge; and (3) high standard structure and sensory characteristics.
Acknowledgement
This work was supported by the Italian Ministry of University and Research, project no. 12819, D.D. 1801 (31 December 2004).
