Development of a baking procedure for the production of oat-supplemented wheat bread

Introduction

Recently, considerable interest has been generated in the consumption of nutritionally rich functional foods such as composite bread. It is well known that the unique bread-making properties of wheat flour can be attributed to the ability of its gluten proteins to form a viscoelastic network when kneaded with water. The supplementation of wheat flour with other flours containing no-gluten-forming proteins reduces the bread-making potential of the mixtures as a consequence of gluten dilution. This results in difficulties in dough handling, lower loaf volume and worsening of crumb grain and softness (Wang et al., 2002). On the other hand, the nutritional value of bread can be enhanced through the addition of a large number of flours of different origin, as reviewed by Dendy (2001) who also indicated that acceptable bread can be produced with a wheat flour substitution of up to 25–30%. Oat, for instance, was used to improve the protein content of bread (Oomah, 1983) or to increase the soluble fibre level (Krishnan et al., 1987). Integration of wheat flour with barley or oat enhances the β-glucan content of bread and may have a significant effect on human health. Potential benefits of soluble dietary fibre include reduction of bowel transit time, reduction in the risk of colorectal cancer, lowering of serum blood cholesterol, regulation of glucose metabolism and promotion of the growth of beneficial gut microflora (Welch & McConnell, 2001; Brennan & Cleary, 2005).

Various bread-making protocols have been used to produce bread from composite flours. In general, a straight dough baking method was adopted, but the length of the bulk fermentation and the number and the length of the proofing periods were extremely variable. Literature reports that composite breads have been baked after fermentation/proofing periods shorter than 90 min (Dervas et al., 1999; Hallen et al., 2004) or with longer fermentations times up to 3 h (D'Appolonia & Youngs, 1978; Oomah, 1983; Krishnan et al., 1987; Zhang et al., 1998). Other differences in process were related to dough production and in particular to the use of low- or high-speed mixers as in the Chorleywood Process (Dobraszczyk, 2001).

The purpose of this investigation was to develop a baking process suitable for the production of wheat-supplemented bread, by comparing the performances of different baking conditions on the characteristics of bread prepared with wheat–oat composite flours at different substitution levels.

Materials and methods

Raw materials

Dehulled oat seeds of commercial origin (Molino Lameri S.p.A., Cremona, Italy) were grounded with a hammer mill in order to reach a flour particle size lower than 500 μm. A commercial wheat flour (W = 344 × 10⁻⁴J; P/L = 0.75) supplied by Molino Pagani (Borghetto Lodigiano, Italy) was used for bread-making. Three different wheat–oat blends were prepared with a substitution level of 20%, 30% and 40%.

Chemical analyses

The moisture content of flours was determined according to the official method (44–15A; AACC, 2000). The total nitrogen content was determined according to the official method (920.87; AOAC, 1995) and the protein content was calculated adopting 6.25 as a conversion factor for oat and 5.70 for wheat. Lipid content was analysed according to the official method (n.136; ICC, 1992) and total fibre according to the official method (991.43; AOAC, 1995). All these determinations were made at least in duplicate and the average result is presented. Total starch and β-glucan content and the amount of damaged starch were determined using the ‘Total Starch Assay Kit’, the ‘Mixed-linkage β-glucan Assay Procedure’ and the ‘Starch Damage Assay Kit’, respectively, by Megazyme International Ireland Ltd (Bray Business Park, Bray, Co., Wicklow, Ireland). The results given for the enzymatic tests are the mean of four replicates.

Rheological measurements

The dough mixing properties of the different wheat–oat blends were examined with the Brabender farinograph (Brabender, Duisburg, Germany) according to the constant flour weight procedure (n. 115/1; ICC (1992).

Dough development and the gas volume from the yeast activity were measured with a rheofermentometer (Chopin, Tripette & Renaud, Villeneuve La Garenne Cedex, France). The dough was prepared in the farinograph bowl by mixing, for 6 min at 30 °C, the same ingredients used for bread production. Wheat flour or wheat–oat blends (300 g) were mixed with 10.5 g compressed yeast (previously dissolved in water), 9 g shortening, 6 g salt, 18 g sucrose, 2.4 mg ascorbic acid and water according to the farinographic absorption value. The rheofermentographic test was performed for 3 h at 30 °C on 315 g portion of the dough, by placing the weight support (254 g) of the instrument on the sample without adding any extra weight.

Bread-making techniques

The same formula reported for the rheofermentographic test was used. Flour (or wheat–oat blends) was stirred for 3 min in the mixer. After this period, yeast, sucrose, salt and ascorbic acid previously dissolved in some water, melted shortening and the remaining water were added. Water was added according to the farinographic absorption value. The kneading time varied according to the mixer used: 3–4.5 min for Swanson-Working pin-type mixer (National Manufacturing Co., Lincoln, NE, USA) and 10 min for spiral mixer (model N-50G; Hobart Foster International, Norfolk, UK).

Three different bread-making procedures were adopted. In the first one, the Baking Test (10–10B; AACC, 2000) (STD), dough prepared with pin-type mixer was scaled to a mass of 168 g, placed in baking pans and proofed at 30 °C and 80% RH. After 105 min, the dough was rounded and proofed again for 50 min. Before baking at 220 °C for 25 min, dough was rounded and proofed other two times, for 25 and 50 min respectively. During baking, some water was vaporised in the oven to avoid excess drying of the bread crust.

The second bread-making method (medium length procedure, ML) consisted in two proofed periods, the first of 40 min at 30 °C at 80% RH carried out on bulk dough and the second performed on rounded pieces for 60 min at the same conditions. Baking was performed as reported previously. For this procedure, besides using the pin-type mixer, dough was also prepared with a spiral mixer. In the third method (fast length procedure, FL), dough prepared with the pin-type mixer was directly moulded, fermented for 60 min at 30 °C at 80% RH, and baked at conditions similar to that of the other protocols.

After baking, the loaves were removed from the pan and cooled for 60 min at room temperature before being characterised for weight, height and volume (determined by rapeseed displacement on three separate loaves). The specific volume (mL g⁻¹) was calculated. Baking trials were repeated twice and the coefficient of variation for the specific loaf volume was less than 3%.

Image analysis of crumb grain

Images of sliced bread were captured using a scanner (Scanjet 6300c; Hewlett Packard, Palo Alto, CA, USA). Images were scanned full scale in 256 grey level at 300 dots per inch, as reported by Crowley et al. (2000). The number of cells, cell area, number of cells cm⁻² and cell to total area ratio were considered as crumb grain features. The size distribution of cells was performed by counting the percentage of cells that fell into the following categories: X < 0.8 mm²; 0.8 < X < 4.0 mm²; X > 4.0 mm². The contribution of the different size categories to the total cell area was also evaluated and presented as area of each class size to total cell area (%).

Statistical analysis

Analysis of variance (Anova) was performed using Duncan's multiple-range test to compare treatment means; differences were considered significant at P < 0.05.

Results and discussion

Flour characterisation

The chemical composition of the two flours is reported in Table 1. Wheat flour showed a higher amount of starch and proteins and a lower fibre and fat content. Oat protein does not complement wheat gluten and interfere with gluten formation. The wheat milling process, that includes many breaking and reduction steps, induced a more severe mechanical damage to starch.

Dough rheological properties

The replacement of the wheat flour with increasing amounts of oat was associated with the rise of farinograph water absorption, mainly due to the high fibre (rich in β-glucans) content of the oat flour (Fig. 1). Lee et al. (1995) reported similar effects for the addition of 1% barley β-glucan into a wheat flour dough system. Comparable results have also been obtained with formulations containing 25% oat flour (Oomah, 1983) or 10–15% oat bran (Krishnan et al., 1987). Dough stability sharply decreased from 15.3 min in wheat flour to 5.9 min for the 20%, 6.1 min for the 30% and 4.8 min for the 40% substitution levels. A similar effect was noticed when using 20% roasted oat flour while steaming, or a combination of roasting and steaming of oat grain, significantly improving the bread-making potential of oat flour (Zhang et al., 1998). The loss of gluten elasticity is pointed out by the decrease in the width of the farinograph curve.

The fermentation behaviour of dough containing different amounts of oat flour was continuously monitored by the rheofermentometer (Table 2). Dough height (H_m) was negatively affected by oat flour, with a 40% decrease at the higher supplementation level. The time of maximum dough development (T₁) also decreased, increasing the level of oat flour. These results were in agreement with the loss of stability and elasticity measured by the farinograph. The time at which gas starts to escape from the dough (T_x) was decreased by wheat substitution, revealing an increase of dough permeability to CO₂ which was produced to a greater extent, but more easily released. These data indicate that short-term fermentation helps avoid dough structure breakdown and excessive CO₂ release.

Bread-making properties

The possibility of using a bread-making protocol mainly depends on dough characteristics. A 20% wheat replacement level induced such strong alteration on the gluten structure that made bread production adopting the AACC Baking Test unfeasible (Table 3). In fact, using this procedure, the dough was prepared with a pin-type mixer working at high speed (100–125 rpm) that subjected gluten to high stress and allowed the incorporation of a large amount of air. The presence of oat flour promoted a physical interference to the gluten matrix that did not permit the production of bread even at a wheat substitution level of only 20%. Indeed the four leavening periods spaced by three rounded phases made the dough very sticky and difficult to handle and form.

The use of shorter baking methods allowed the production of bread with 20% and 30%integration levels (Fig. 2). For both baking methods, dough containing 20% oat flour appeared properly developed and easy to handle; on the contrary, dough pieces at 30% integration level slightly collapsed during baking when the ML protocol was used. No bread was obtained with the higher substitution level as, after the first mixing, the dough appeared loose because of the excessive mechanical stress applied.

The effect of the different protocols on the loaf characteristics is evident from the data regarding 100% wheat bread (Table 3, Fig. 2). Longer mixing and proofing time resulted in an increase in bread specific volume from 3.4 to 5.6 mL g⁻¹. The same trend was observed for the 20% integration level, while no difference was noticed for the 30% integration level. The high-speed mixer, subjecting dough to great shear stress, is not suitable for making dough from composite flours if an integration level over 30% is pursued. On the other hand, a single leavening phase did not give a proper increase of dough pieces volume when low integration was used. The use of a spiral mixer, adopting a medium length bread-making method, also allowed the production of bread from 40% integrated flour. In this case longer kneading times (10 min) were required to obtain a coherent dough.

Data reported in Table 3 indicate that, while wheat and 20% integrated bread present higher specific volumes when the Swanson mixer is used, for the other mixtures better results are obtained with the Hobart mixer. Moreover, under the same bread-making protocol (ML), 40% supplemented bread obtained with the spiral mixer showed higher specific volume than 30% supplemented bread produced with the pin-type mixer. It is also interesting to note that for each loaf prepared with the high-speed mixer the number of cells was nearly uniform, while, under the same bread-making conditions, cell number decreased as the supplementation level increased. By preparing the dough with a spiral mixer, wheat and 20% supplemented bread showed lower cell numbers whereas the opposite was found for bread containing 30% oat flour. Both the mixing process and the dough composition are therefore critical factors for bread structure. A different situation emerged from the results of the cell total area that appeared related to the length of proving for wheat and 20% supplemented bread when the pin-mixer was used. Bread produced with 30% oat flour, when adopting the ML procedure, presented a loaf volume and cell total area lower than bread produced with the FL procedure, revealing that air bubbles collapsed and coalesced during the longer leavening phase. The air content and the bubble size distribution depends on the balance between the rates of entrainment and escape of air during mixing and on bubble break-up (Campbell, 2003). These phenomena are affected by the rheological properties of dough, which are strongly influenced by the presence of interfering materials such as non-gluten-forming proteins and fibre. All loaves prepared with the spiral mixer presented higher values of mean cell area, whose size sharply increased from 0.65 mm² for wheat bread to 0.97 mm² for 40% supplementation. These data demonstrate that during dough formation with a low-speed mixer, a small number of nuclei are generated for developing gas cells that make up the air phase in the bread crumb (Cauvain et al., 1999; Campbell, 2003). The resulting crumb texture of the bread appeared coarser due to the higher surface belonging to cell having an area larger than 4 mm² that accounts for 30% in wheat bread and about 40% in the other breads (Fig. 3).

In conclusion, the results show that a medium length bread-making method appears to be a good tool for comparing the baking performance of composite flours under a wide range of integration levels. Using this method and mixing the dough in a low-speed spiral mixer, bread from 40% integrated flour was also obtained.

Acknowledgment

The authors wish to thank Mr Lorenzo Fongaro for his skilful technical assistance.

References