Final Report Summary - GLUTENFREE (Tasty and healthy gluten-free bakery products and pasta - improved products for wide consumer acceptance)
Executive Summary:
Common problems of currently available gluten-free products include high starch and low fibre contents, reduced shelf life, strong off-flavours as well as dense, crumbly and dry texture (bread) and untypical colouring, low protein and fibre contents, texture issues such as reduced elasticity and increased cooking losses (pasta). The scientific and technological development activities carried out in the GLUTENFREE project tackled these issues from different angles and delivered a number of valuable results for the involved SMEs.
A baseline sensory study on gluten-free bread and pasta with consumers in Germany, Italy and Ireland at project start revealed information on consumer preferences, acceptance and needs, which provided the basis for target-oriented product development and optimization in the project. An in depth characterisation of gluten-free flours showed that some of these flours represent an interesting alternative to the commonly used wheat flour with regard to their nutritional properties. However, their potential for the production of bread is certainly limited due to the lack of the structure-forming proteins. Therefore, the use of complex formulations combining different gluten-free flours and ingredients such as hydrocolloids and proteins are necessary to get products of similar quality to wheat containing counterparts (i.e. bread or pasta).
With regard to structure forming proteins, processes for the production of new lupin and broad bean protein isolates were optimized regarding the product yield and viscoelastic properties as well as subsequent scale-up and testing on a technical scale. Fundamental research then has generated new knowledge and insights on the effects of a range of proteins, hydrocolloids and fibres on the technological properties of breads and pasta.
Based on these results a wide range of recipes for new gluten-free prototypes were developed in collaboration between the RTD and the industry partners. These included oat bread, toasting bread, ciabatta, French petit pain, German style sourdough bread and pizza base. In addition promising recipes for gluten-free spaghetti enriched with protein and fibre were generated. Finally, the outcomes of the above mentioned research were used for the implementation of new gluten-free food prototypes into the production lines of the involved SMEs and the production of high-quality gluten-free bakery products and pasta for a final consumer acceptance study in the end of the project. The results of this study showed that some breads (oat bread, linseed bread) achieved high consumer acceptance, whereas some other breads and especially the pizza bases need some final improvements prior to their market launch. With regard to pasta, most of the new spaghetti prototypes were well accepted by consumers and one product was comparable to the gluten-free market leader. This prototype was considered as being ready for market launch and was commercialized by the respective pasta producer by the end of the project.
As a conclusion, these results allow the participating food producers to place new or improved products on the market and thus have the potential to generate additional turnover and employment after the end of the project. Celiac as well as healthy consumers will also benefit by having access to a wider range of tasty gluten-free foods.
Project Context and Objectives:
At present up to two per cent of European consumers are suffering from celiac disease and therefore have to adhere to a strict gluten-free diet. Gluten is part of the common cereals wheat, rye and barley and is therefore contained in basic foods like bakery products and pasta. The market for gluten-free products, which is predominately served by small and medium sized enterprises (SMEs), showed considerable growth in the last years. However, the consumer acceptance of respective products is still limited by their insufficient taste, texture and mouth-feel.
Therefore, the overall objective of this research project was to enable SMEs to produce gluten-free bakery products and pasta of high consumer acceptance in order to participate in this profitable market sector and ultimately to significantly improve their competitiveness. This aim was achieved by utilising carefully selected raw materials and ingredients derived from these, namely plant proteins and hydrocolloids, which are able to effectively replace gluten and gluten-containing ingredients in bakery products and pasta. The interaction of different components in the recipes during baking and pasta processing and their influence on the texture and flavour formation was analysed in order to allow specific improvements. Sensory improvement should be ensured by using selected raw materials, aromatic malts or sourdoughs. Evaluation of consumer preferences, acceptance and needs was an integral part of the research and allowed target-oriented food developments and promised well accepted products.The corresponding research, technological development and innovation related activities focused on the following work packages and specific objectives:
WP 1: Baseline study on gluten-free foodstuff
- To evaluate existing gluten-free bakery products and pasta and to identify consumer preferences and needs.
- To derive targets for research and development with respect to product improvement.
WP 2: Screening of raw materials and selection of ingredients
- To select promising ingredients for the development of gluten-free baked goods and pasta.
- To identify promising alternative raw materials from the range of grain legumes and pseudo-cereals suited for the production of protein ingredients with viscoelastic and film-forming properties for gluten-free bakery products and pasta.
WP 3: Process development for the production of protein ingredients with viscoelastic properties
- To identify key parameters influencing the yield and properties of protein ingredients and derive optimised processing conditions.
- To develop economic feasible processes for the production of plant based protein ingredients considering the best combination of techno-functionality, sensory properties and price.
WP 4: Bio-engineering structural and nutritional value of gluten-free foods
- To identify starter cultures suitable for the fermentation of gluten-free sourdough.
- To isolate and identify exopolysaccharides (EPS) from specific lactic acid bacteria (LAB) strains.
- To isolate lactic acid bacteria (LAB) producing antifungal compounds.
- To apply these LAB in sourdough and identify the positive effects on final product quality.
- To identify the causes of these positive effects on product quality.
WP 5: Interactions of ingredients in baking and pasta processing
- To use advanced rheological tools to rapidly evaluate the strength and functionality of the functional ingredients selected in WP 2.
- To evaluate the effect of the functional ingredients on pasta and bread quality.
- To develop predictive models for the effect of ingredient functionality on product quality for further product development.
- To evaluate the importance of protein, starch and other functional ingredients interactions in pasta and bread.
WP 6: Development of prototype bakery products
- To develop recipes and process parameters for the production of high quality gluten-free bread products typical of western, central and southern Europe.
- To evaluate the effect of functional and nutritive ingredients on bread quality.
- To implement the recipes and production procedures for gluten-free breads on industrial scale.
WP 7: Development of prototype pasta
- To develop gluten-free pasta having firm texture, good taste and high nutritional value.
- Scale-up: Transfer the recipes and production procedures for different model foods to production scale.
- To implement process conditions and recipes, identified and characterised in the development stage, to the production lines of the SME.
WP 8: Aroma analysis and improvement
- Sensory and analytical characterisation of prototype aroma and aroma development.
WP 9: Scale-up and production of food prototypes for the consumer acceptance study
- To implement processes for the production of new gluten-free food prototypes into the production lines of SMEs.
- To evaluate the process conditions and recipes identified on a technical scale.
- To produce sufficient amounts of high-quality gluten-free bakery products and pasta for the final consumer acceptance study (WP 10).
WP 10: Evaluation of consumer acceptance of prototype foods
- To evaluate the developed gluten-free bakery and pasta products in several hedonic tests with consumers with and without celiac disease.
- Sensory Profiling of developed prototypes.
- Preference mapping: combining data with PCA to find preference driving factors.
WP 11: Dissemination and Training
- To disseminate the outcomes of the RTD activities to SMEs, consumers and research institutes.
- To enable the SMEs to apply the technological solutions realised in the project by means of specific training in the application of new ingredients and sourdough technique for the production of improved bakery and pasta products.
Project Results:
Baseline study on gluten-free foodstuff
As a first step of the research a baseline sensory study was carried out in work package 1 (WP 1) aiming to obtain a market overview of gluten-free white bread, dark bread and pasta in four European countries (Germany, Ireland, Italy and Sweden). From the breads and pasta available within these markets, samples were selected for sensory evaluation, with the intention to cover a wide product range and thus a wide range of sensory characteristics.
Sensory product properties and differences were assessed by a trained panel. This data was combined with preference data of consumers from three countries (celiac patients, family members, and non-affected/ healthy consumers in Germany, Ireland and Italy) to reveal factors, which drive, but also adversely affect consumer liking via external preference mapping and penalty analysis. This information provided the basis of the improvement and development of new gluten-free breads and pasta, aiming at achieving high consumers' acceptance.
Characterization of the breads:
Selected white bread, dark bread and pasta varieties, available on the German, Italian, Irish or Swedish markets, were sensorily profiled by a trained panel, and hedonically evaluated by consumers in Germany, Italy and Ireland. For each country the consumers were segmented into three groups: celiac patients, family members, and healthy / non-affected persons. The profiling and hedonic datasets were correlated via multivariate statistical methods to extract preference driving and decreasing factors, which provided directions for product development and optimization.
The white bread samples selected for this research differed significantly in nearly all evaluated product characteristics. This indicated that the selected breads represented a broad range of the products available within the markets. This in turn allowed for identifying a broad range of preference driving and decreasing factors. The German breads showed most of these product characteristics which were recommended to be focused on in product optimization.
The dark bread samples also covered a large variety of breads, as indicated by significant product differences in all assessed attributes, particularly in the mouthfeel. Despite of being different types of bread as such, white and dark bread shared numerous preference driving and decreasing factors, indicating that for the consumer bread needs to have certain characteristics, irrespective of bread type or composition.
Characterization of the pasta:
In contrast to the breads, the pasta samples did not show big sensory differences. In accordance with observations reported in literature (Moskowitz, 2003), there were no relevant overall liking differences between markets and between consumer segments. For future consumer tests this opened the opportunity of conducting the tests with easier-to-reach consumer samples and focusing only on one market (Germany).
Screening of raw materials and selection of ingredients
As part of WP 2, an extensive raw material characterisation was carried out on seven commercial gluten free flours (teff, rice, sorghum, maize, quinoa, buckwheat, oat) and results were compared to wheat and wholemeal wheat flour. Furthermore WP 2 focused on the screening and evaluation of hydrocolloids and protein isolates to identify the most suitable ingredients for gluten-free bread and pasta development.
The results of the analytical characterisation of the gluten-free flours showed that, although most commercial bread and pasta products are based on rice and maize, more nutritious gluten-free raw materials would be available. While endosperm-derived white wheat flour, maize and rice flour contain only 3.4%, 2.6% and 0.4% fibre, quinoa flour contains a much higher amount (7.1%). Also regarding protein content, maize and rice flour are poor when compared to teff or pseudocereal flours such as buckwheat and quinoa.
The seven flours were used to produce gluten-free breads using basic recipes simply consisting of flour, water, sugar, salt and yeast. While the majority of existing publications give information on complex gluten-free formulations, this more fundamental approach was taken to understand the baking properties of the individual flour raw materials. Quality evaluation of resulting loaves showed big differences not only in visual appearance but also in final loaf volume, crumb firmness, crumb structure, shelf life and taste attributes.
When comparing white wheat bread to all other breads it became apparent that crumb structure and cell characteristics were very distinct. Quinoa, buckwheat, maize, teff and wholemeal wheat breads showed very dense crumbs while oat and rice breads had a more open texture compared to wheat bread. As the same mixing regime was followed for the production of all gluten-free breads, it can be assumed that this is due to the differences in dough consistency.
Obtaining an acceptable shelf life that would allow selling fresh gluten-free bread represents one of the biggest challenges. Due to high aw values and the high moisture contents resulting from the increased water addition levels, mould growth commences earlier. The results of shelf life trials showed that on average, mould growth commences on day 4 for most gluten-free breads and for rice bread already on day 3. In order to place a bread product in the supermarket, 8 to 10 days of shelf life are required. This issue was addressed in work package 4, which aimed at prolonging microbial shelf life by using sourdough fermented with lactic acid bacteria producing antifungal compounds.
Apart from microbial deterioration, shelf life of bread is also determined by the staling behaviour. Crumb firming, attributed mainly to re-crystallization of amylopectin and water redistribution between crumb and crust, was especially pronounced in oat, white wheat and maize breads. The staling rate of teff bread was significantly lower than for the other gluten-free flours. This was expected since teff starch has a lower tendency to retrograde as shown in the Rapid Viscoanalyser tests.
The flours also showed significant differences regarding their sensory properties. Each of them exhibited a very specific aroma profile due to characteristic odour attributes and different odour intensities which can be influenced by fermentation and baking (details see WP 8, 'Aroma analysis and improvement').
Overall, the gluten-free flours screened represent an interesting alternative to the commonly used wheat flour but their potential for the production of bread is certainly limited due to the lack of the structure-forming proteins. In order to get a product of similar quality to wheat bread, it is probably most practicable to use a complex formulation combining different gluten-free flours and ingredients such as hydrocolloids and proteins. Their different properties with regard to bread making, allow choosing suitable flours according to the desired bread characteristics.
With regard to proteins a series of experiments was carried out to obtain protein isolates with viscoelastic properties from seeds of different grain legumes (lupin, soy, pea, broad bean), pseudocereals (buckwheat, quinoa, amaranth) and cereals (oat). The process used was established for lupin and consists of a salt induced protein extraction and a dilution precipitation as the main steps. The feasibility of this process for the other raw materials was evaluated on basis of the overall and protein yields, the protein content and the rheological characteristics of respective protein isolates as well as their performance in baking tests.
With regard to the overall as well as the protein yield the extraction of grain legumes revealed the best results. The highest overall yield was achieved for soy protein isolate followed by broad bean, lupin and pea protein isolates.
Oscillation tests performed with a rheometer revealed significant differences in the viscoelastic properties of the respective protein isolates. While lupin protein isolate showed fluid-like characteristics and a smooth, honey-like texture, all other protein isolates showed gel-like characteristics and more stiff texture. The structure strength of the protein isolates was highest for quinoa and soy and lowest for broad bean, lupin, pea and oat.
The application of the protein isolates in a complex formulation generally increased the stiffness of the dough compared to the reference. But the rheological characteristics of the different types of dough did not generally correlate with the viscoelastic properties of the protein isolates. In baking tests only lupin protein isolate improved the specific loaf volume, which was similar to wheat flour bread. In addition, the lupin breads had the best crumb texture. The breads with the other legume protein isolates achieved loaf volumes comparable to the control gluten-free bread.
Process development for the production of protein ingredients with viscoleastic properties
In WP 3 a systematic variation of the key processing parameters using response surface methodology (RSM) was carried out in order to identify the best conditions for the production of viscoelastic protein isolates from lupin and broad bean regarding yield and technological properties. Both raw materials were selected by the consortium on basis of the results achieved in WP 2. The resulting isolates were analytically characterized for their rheological, techno-functional (water binding, emulsifying) and baking properties.
For lupin seeds, the standard process consisting of a salt-induced protein extraction and dilution precipitation was optimised regarding the product yield. A face centred central composite design was applied to determine the operating conditions due to the product yield. The parameters pH value, NaCl-concentration, solid to salt solution ratio, extract to water ratio during precipitation, pH value of the extract before precipitation, centrifugation and sedimentation time were chosen as factors varying in a specific range and were identified to be significant. Finally, an experimental plan of 50 combinations was obtained and evaluated, while the order of experiments was fully randomized to avoid any systematic bias. As a result the optimum conditions to maximize the product yield could be established. In a further step, the protein isolates achieved at different parameter settings were analysed with regard to their techno-functional and their rheological properties as well as their performance in gluten-free baking.
Applied in a complex gluten-free formulation the different lupin protein isolates led to a fluid dough structure with low complex modulus but the addition of LPI generally increased the stiffness of the dough compared to the control dough without protein. The viscoelastic properties of the dough however did not generally correlate to that of the protein isolates. Regarding the specific loaf volume the isolate samples produced at low salt level and neutral pH achieved the lowest volume. The highest loaf volume was observed with the isolate recovered at standard conditions. This isolate also achieved the softest bread crumb. Breads produced with the other protein isolates did not improve the specific volume compared to the control bread without protein addition. Taking into account the product yield of LPI, it can be concluded that the product yield is contrary to product properties: isolates achieved at highest yield (18-22 %) have lowest protein content, highest fat content and worst baking properties, but highest emulsifying and water hydration capacities and highest resistance to deformation.
As a further raw material, broad bean was chosen to produce an alternative protein isolate for gluten replacement. Similar to the procedure used for lupin seeds, a face centred central composite design was applied to determine the optimised parameter settings by only varying the factors pH of protein extraction, salt concentration in the extract and pH of protein precipitation. As for lupin, the statistical variation of process conditions allowed to identify optimised process conditions to gain a satisfactory amount of protein isolate. Thus, a maximum product yield of ~20 % was achieved, which is equal to the yield obtained when using lupin seeds. The investigated broad bean protein isolates (BPI) on average showed a higher water hydration capacity but lower emulsifying capacity than the lupin protein isolates.
In the rheological analysis all broad bean isolates showed the characteristics of viscoelastic gels (G'>G'') with large differences in their stiffness and resistance to deformation (G*). The protein isolate showing the highest loaf specific volume within the different isolates was achieved with the highest protein yield (56.1%) and at a low salt level, which makes it the most interesting product with regard to the technical implementation. However, further conclusions regarding the correlation of parameter settings in the production of the different BPI with their product properties cannot be derived from the results.
Finally a scale-up of the processes for lupin and broad bean protein processing from the laboratory scale to the facilities of Alberts, the involved SME, was realised. However, the implementation of the process required a few adjustments, e.g. the seed pre-treatment was carried out at the Fraunhofer institute. The protein isolates was successfully prepared at Alberts starting from lupin flakes or broad bean flour respectively. Since Alberts did not have the facilities for drying the protein isolate, the protein isolates were pasteurized and spray-dried at the Fraunhofer institute.
An economic feasibility study showed that with further optimization of the process (higher protein yield, energy saving, water purification, etc.) the protein isolates may be commercialized successfully after the end of the project.
Bio engineering structural and nutritional value of gluten-free foods
As part of WP 4 ('Bio-engineering structural and nutritional value of gluten-free foods') strains of LAB haven been:
a) isolated from different gluten-free flours,
b) evaluated for their ability to successfully ferment and survive in a gluten-free system and
c) screened for their ability to produce EPS and antimicrobial compounds.
Finally, the bread quality parameters have been evaluated on bread samples containing 20% of sourdough fermented with the selected LAB strains. Suitable protocols for the applications of selected starter cultures in gluten-free sourdough were developed. The positive effects of sourdough are mainly due to the metabolic activities of the lactic acid bacteria performing the fermentation. A strain with antifungal properties (Lactobacillus plantarum ZA 7) and an exopolysaccharide producing strain were chosen (Weisella cibaria ZA 13). Their effects on aroma, structure and shelf life of baked goods was studied. Quality of breads produced from different gluten-free flours was influenced to different extent by sourdough fermentation. Apart from oat and wheat sourdough bread the application of both sourdough types did not affect the specific volume in gluten-free breads compared to control breads. In terms of crumb hardness the trend that breads baked with strain ZA 13 give lower values for crumb hardness is apparent in most breads, but not in oat and sorghum sourdough bread.
Hardness values for ZA 7 breads were not significantly different from those of control breads. ZA 13 reduced crumb hardness after 5 days of storage to a higher extent than ZA 7. This shows that the effect seems to be independent from the sourdough acidity, as sourdoughs from ZA 7 show higher acid values (TTA) but not higher crumb hardness than the non-acidified control breads. Even though oat and wheat sourdough bread had a soft crumb on day 0, the rate of staling was high. In general, the rate of staling was slightly delayed for some gluten-free breads but compared to control only significantly in oat and wheat bread. The tendency of reducing crumb chewiness compared to gluten-free control breads could be found in both sourdough trials but was only significant for breads fermented with ZA 13.
Both strains are capable to reduce chewiness in buckwheat, quinoa and teff breads compared to control breads with the trend of larger influence by ZA 13 than ZA 7. Differing from that, either no significant (with ZA 13 and oat or wheat bread) or the adverse effect (for ZA 7 and oat bread, and for ZA 13 and ZA 7 in sorghum sourdough bread) was found in certain breads. Furthermore, the only influence of sourdough on springiness of bread crumb was found for wheat bread with a significantly increased springiness for both sourdough breads.
Overall, based on standard bread-quality parameters, oat flour resulted in good quality breads, indicated by high loaf specific volume and a soft crumb. Though the other gluten-free flours had poorer bread making properties, quinoa and buckwheat resulted in agreeable breads with good crumb structure. The ratio of pore area to slice area and the cell volume of the bread crumb were either higher or equal compared to gluten-free control breads. This characteristic of crumb structure was improved by sourdough application in general and especially when ZA 13 was used as a starter culture. Concerning shelf life, it was possible to observe an extension of shelf life only for wheat sourdough bread when ZA 7 was used. The application of ZA 7 sourdough in gluten-free formulation showed no difference to the fungal spoilage of the control breads.
The application of sourdough ZA 13 on hardness of buckwheat bread was higher than for the other gluten-free breads. Quinoa and teff breads had also an acceptable crumb structure, but the loaf volume was low. Sorghum flour on its own was not suitable for bread making, due to poor structure, high crumb hardness and rate of staling. Regarding the influence of the chosen sourdough strains, it can be said, that what applies to the wheat system, does not necessarily apply to the gluten-free system. The improvement of gluten-free bread by lactic acid bacteria was much less pronounced than in wheat bread.
Interactions of ingredients in baking and pasta processing
WP 5 aimed at the evaluation of the effect of functional ingredients (starch, proteins, hydrocolloids) on bread and pasta quality in more detail. Starch plays an important role in pasta as well as bread, especially in gluten-free systems. The Rapid Viscoanalyser and differential scanning calorimetry represent two important tools for the study of starch behaviour. They were applied in this work package to get information on pasting properties and gelatinisation temperatures of the nine flours described in WP 1. Viscosity behaviour during heating from 50°C to 95°C, which gives an indication of the capacity of different starches to retain water and swell, was shown to be different for all flours. It was concluded that pseudocereal starches are more resistant to heating and shear stress. Results showed that, except for oat and quinoa, gelatinisation temperatures were significantly higher for the gluten-free flours compared to wheat flour.
In addition dough characteristics of the different flours were investigated in more detail by means of fermentability trials (Rheofermentometer) and rheological tests as well as confocal laser scanning microscopy and scanning electron microscopy. White wheat and oat samples reached a maximum dough development height unmatched by the gluten-free batters or the wholemeal wheat flour dough, indicating superior viscoelastic properties of oat. Gas production was highest in teff, buckwheat and quinoa batters, indicating that these flours have a more favourable sugar composition for yeast fermentation. Scanning electron micrographs of bread crumbs revealed that distinct starch granules are still visible in wheat and oat breads, while starch is gelatinized to a much wider extent in case of all other gluten-free flours. The observation of the ultrastructure of bread showed the difference between wheat samples and samples where the structure forming gluten network is absent. These observed pictures visualized on a microscopic level the reduced structural quality, associated with gluten-free products. Thus microscopy is an excellent tool to visualise processes occurring during baking.
Another part of WP 5 was the evaluation of cereals other than wheat and further ingredients on pasta quality. Due to outstanding nutritional characteristics, teff and oat flour were chosen. As mentioned above, commercial gluten-free pasta is generally low in protein. Therefore different protein ingredients were evaluated (egg white powder, whole egg and egg yolk powder and prototype lupin protein isolate). Results showed that egg white powder not only increased the nutritional value of gluten-free pasta, but also improved textural properties such as elasticity, firmness and stickiness.
Many commercial products contain emulsifiers and several publications report on the reduction of cooking loss upon incorporation of emulsifiers. Looking at the response surface graphs of the current study, this effect was not observed. Looking at the results of texture profile analysis as well as sensory evaluation by a trained panel, it could be seen that when using oat flour, spaghetti comparable to wheat based products can be obtained. Teff pasta was characterised by high cooking loss and dryness. Differences in pasting properties and gelatinisation temperatures of oat, teff and wheat were measured. The developed oat and teff pasta are nutritionally superior to many gluten-free pasta products currently on the market.
Scanning electron micrographs of pasta produced in this study showed in case of wheat samples a strong outer layer of coagulated protein that entraps the ungelatinized starch granules. This layer was less pronounced in case of gluten-free samples, which explains the higher cooking loss and stickiness observed in these products. Confocal Laser Scanning Microscopy is an interesting tool for ultra-structure evaluation of foods as components can be stained selectively. Changes in the starch structure occurring during cooking of pasta were observed using a Periodic Schiff's acid kit as staining technique. Significant changes were observed upon cooking. Pasta represents a limited-water system and hence, even after cooking for 12 min, a great proportion of starch is still present in its granular form. By using different hydrocolloids and protein ingredients (i.e. guar gum, soy and whey protein) also a formulation for dough suitable for the production of e.g. lasagna sheets could be developed with a visual appearance and consistency similar to that of the wheat containing counterpart.
In WPs 6 and 7 the results achieved in the previous research work were used to develop different new recipes for gluten-free breads and pasta in cooperation with the involved industry partners.
Development of prototype bakery products
WP 6 dealt with the development of different gluten-free bread formulations suitable for Western, Central and Southern Europe. As first step commercial wheat containing soda bread, toasting bread and wholemeal oat bread were purchased, to get an idea on the characteristics and special features of these bread types. In a next step gluten-free counterparts were developed. A wide product range including soda bread, oat bread, toasting bread, ciabatta, German style sourdough bread, pizza base and petit pain were developed. Soda bread is a traditional bread variety of Ireland, prepared using sodium bicarbonate as raising agent instead of yeast. The soda bread dough developed in this study was formable and hence the characteristic soda bread shape was achieved without the need of tins. Crumb structure of the resulting bread was similar to the gluten containing counterpart.
The toasting bread developed under this task was described as soft coarse textured product with a slightly yellow crumb and a relatively low staling rate. Two commercial flours (i.e. wholegrain oat flour and endosperm oat flour) as well as flour milled from gluten-free oats supplied by Semper (Sweden) were used for the development of oat bread. All three breads showed lower values for crumb hardness than the wheat containing wholemeal bread, with the lowest being bread produced from wholegrain oat flour. When comparing the crumb grain to a wheat wholegrain product, the developed breads under this task were more favourable as the structure was more open and aerated.
Using a dried sourdough supplied by Bocker (Sarrasin, buckwheat-based) gluten-free bread suitable for the central European market was successfully developed. Based on the flour mixture used for gluten-free breads a sheetable pizza base recipe has been developed. Fat was added in form of olive oil. Compared to the commercial wheat-based recipe, the gluten-free pizza base developed in WP 6 showed a higher resistance to bending, resulting in a higher value for firmness. Gluten-free French Petit Pain was developed using the same recipe as for gluten-free toasting bread. The structure of the developed gluten-free Ciabatta Style bread was denser than the wheat containing ciabatta.
For all developed gluten-free bread formulations, shelf life was evaluated and enhanced with calcium propionate as well as sorbate. In addition to the development of gluten-free breads, the procedure for the addition of nutritive and technologically favourable ingredients to these formulations was optimised. The darkening effect is desirable since gluten-free breads are usually considerably lighter than their wheat containing counterparts. Incorporation of oat b beta-glucan had a remarkable effect of softening the crumb, as indicated by lower hardness values. The addition of inulin resulted in higher crumb hardness.
Hence it was concluded, that the two grades of citrus fibre can be interchanged without effect on product quality. Malted gluten-free grains may be another interesting potential ingredient for the gluten-free baking industry, as malt enzymes have the potential to increase the rate of fermentation, enhance crust colour formation, volume and soften the crumb. As the effect of enzymes on gluten-free bread quality depends strongly on the formulation and ingredients, four complex and one simple oat recipe were used for these trials. The use of oat malt at concentrations higher than 1% led to a significant increase in the specific volume in three of the four recipes. However, a 2.5% malt addition deteriorated the crumb structure of two of the formulations, leading to the formation of large holes and a hollow loaf.
Quinoa malt had no effect on the specific volume in any of the recipes. In an attempt to further improve the German style sourdough bread, the originally used sourdough was replaced by two different, dried, gluten-free sourdoughs, based on maize and buckwheat. The use of different sourdoughs in the formulation of German style bread had no influence on the specific volume. Crumb hardness on day 0 was highest for German style bread baked with buckwheat, while the original- and corn-sourdough breads showed lower crumb hardness. Gluten-free toasting bread was baked containing xanthan, dextran and a combination thereof as well as without hydrocolloid addition.
There was a significant increase of loaf volume upon hydrocolloid addition, especially if xanthan gum was added. Crumb hardness was reduced and crumb springiness increased by xanthan, dextran and a combination of both. Therefore it was concluded, that xanthan and dextran are able to improve the quality of gluten-free toasting bread. Therefore, the effect of xanthan and/or dextran on all other developed formulations was evaluated as well. The effect of these hydrocolloids on the properties of different gluten-free bread recipes was very versatile.
Summarising these results it can be stated that fibre ingredients are not only beneficial from a nutritional point of view, but can also have a positive influence on bread quality as shown for oat beta-glucan, inulin or dextran. However, it has to be kept in mind that an adjustment of water level can be necessary. Malt and sourdough addition can also have a positive influence on gluten-free breads, especially in terms of loaf volume and flavour improvement. The effect of these ingredients depends strongly on the bread matrix and ingredients have to be chosen carefully in order to achieve the desired effects.
In a collaboration between UCC, Fraunhofer and the industry partners, these bread prototypes were then up-scaled as part of WP 9 ('Scale-up and production of food prototypes for the consumer acceptance study').
Development of prototype pasta
WP 7 aimed to develop high quality gluten-free pasta with firm texture, good taste and nutritional value on a lab scale and the transfer of the new recipes and production conditions to the production lines of the SMEs. The focus thereby was on the improvement of protein and fibre content of a basic pasta mixture of the involved pasta producer consisting of corn and rice flour without impairing the sensory properties and the texture.
In the first step several gluten-free commercial pasta samples were investigated with regard to their technological and sensory properties. The technological parameters analysed by using standard methods were the cooking loss, water absorption, firmness, stickiness, tensile strength and elasticity The analytical values achieved were used as references for the development of new gluten-free pasta samples and their evaluation in the second phase of the study. In addition, the technological and sensory properties of the gluten-free samples were compared with the values of two commercial gluten containing spaghetti, both representing best quality in Italy.
Thereafter proteins and fibres were added in varying amounts to the pasta matrix and the effects on the main technological quality parameters like cooking time, water absorption, cooking loss, firmness, stickiness, tensile strength, elasticity and sensory quality (see WP 8) were evaluated and compared to commercial references. The ingredients used were lupin protein isolate developed in WP 3, commercial pea and potato protein isolates and gelling and non-gelling citrus fibres. The selection of ingredients and the concentrations used were adapted to the specific objectives of the pasta producer involved in this WP (i.e. plant based proteins, no allergens, natural ingredients, no E-numbers, increase of the protein and/or the fibre content, respectively).
A general observation was that protein and/or fibre enrichment mainly had positive effects on the technological properties of gluten-free spaghetti but only a few ingredients or mixtures were able to improve all of the technological properties. Sensory properties were not or only slightly influenced by the ingredients, and there were no negative effects on aroma and taste. With regard to the objective of the study it was therefore possible to prepare a number of recipes with increased protein and fibre content and sufficient texture and sensory properties that could be used for the scale-up to technical scale.
Regarding the cooking loss most of the ingredients had a positive effect reducing this parameter compared to the basic mixture. Thus, some of them even reached the values of the gluten containing references. The biggest effect was obtained by combinations of lupin protein, citrus fibres and emulsifier. A number of other samples also reached remarkable reductions especially the samples containing lupin or potato protein isolates or these proteins in combination with fibres.
By applying standard processing conditions for pasta production, the utmost increase of the firmness was reached by adding HPMC or potato protein. The firmness of these samples was even higher as the durum wheat reference sample. Other ingredients which also significantly increased the firmness where pea protein isolate, potato protein isolate, citrus fibre and combinations thereof. The addition of lupin protein generally deteriorated the firmness.
The pasta stickiness was negatively influenced (increase) by the majority of mixtures, especially by the samples consisting of pea or lupin protein isolates and their combinations with citrus fibres and HPMC. The tensile strength and the elongation were only improved by a few samples. Most effective were pasta samples containing pea protein isolate, citrus fibre and emulsifier. The elongation was improved by the addition of lupin protein isolate and by the addition of gelling fibres respectively. The technological variations carried out with two types of citrus gelling fibres revealed some interesting perspectives for industrial use. Despite reductions of the cooking loss some combinations were able to increase the firmness and decrease the stickiness very effectively.
Based on the results, several recipes could be derived for the further development of protein and fibre enriched pasta on an industrial scale using standard production conditions. The most promising recipes were based on combinations of potato protein, pea protein, citrus fibres and emulsifier. Moreover, a combination of lupin protein isolate with citrus fibre and emulsifier showed very good results, but is not in accordance with the company's philosophy to use no potential allergens. However, if the company wants to produce pasta either enriched with protein or enriched with fibre the best ingredients appeared to be potato protein isolate or non-gelling citrus fibre. On basis of these results Bioalimenta was able to produce a number of new gluten-free spaghetti using a new production line which was established in June 2012 (see below, WP 9). Samples of these recipes were sent to research partners for sensory (HAW) and technological (Fraunhofer) evaluation.
Aroma analysis and improvement
Investigations in WP 8 were run in parallel to the basic investigations of the effects of functional ingredients on product properties (WPs 2-5) and prototype bread and pasta development (WP 6 and 7) and focused on the sensory and analytical characterisation of prototype aroma and aroma development. As mentioned previously, sensory investigations of amaranth, buckwheat, oat, quinoa, rice, sorghum, teff and wheat flours demonstrated that each of gluten-free flour exhibits a very specific aroma due to characteristic odour attributes.
Moistening intensified in many flours the intensities of several attributes, which may be caused by a high degree of macromolecular constituents to adsorb odour-active compounds. The flours consist in part of high odorant loads, which are the basis for the final bread aroma. Hedonic test revealed oat, rice and teff as flours with the highest sensory liking. The preference of e.g. oat was correlated with the oat flakes-like attribute, which was very intense in oat flour. In contrast, pea and quinoa showed a low acceptance in hedonic tests, which was caused by intense pea-like odour notes.
The aroma of yeast breads, which were prepared from the flours, showed distinct changes in the aroma profiles and demonstrated the high influence of fermentation and baking on the odorant composition. Interestingly, some characteristic flour attributes were not perceivable or their intensities were decreased in the crumbs of several breads. For instance, the oat flakes-like attribute was reduced in the crumb of oat bread, but a positive crumb aroma was evaluated by the hedonic test. In consequence, degradation of the key odorant to a certain extent and generation of further odorants with characteristic attributes, which were detected in the aroma profile of oat bread, can be concluded for this observation. The undesirable pea-like aroma of quinoa flour could not be removed after fermentation and baking.
The developed pasta basic mixture and all improved recipes for pasta showed no influence on taste in comparison to commercial wheat based reference pasta. The aroma of the gluten-free pasta samples differed from the reference pasta in particular in a reduced intensity of the typical pasta-like and buttery attribute and a dominant cooked rice-like note. The texture properties juiciness, firmness and stickiness were influenced by the tested protein ingredients, emulsifier and fibres. In particular, stickiness was in part much higher than the references. Regarding juiciness and firmness, a combination of lupin protein or gel fibre with the basic mixture showed comparable properties. The investigated pasta samples showed that lupin, gel fibres and emulsifier in combination with the developed basic mixture lead to pasta with comparable textural properties than a commercial reference pasta. However, the aroma should be optimised.
Scale-up and production of food prototypes for the consumer acceptance study
In WP 9 'Scale-up and production of food prototypes for the consumer acceptance study' the outcomes of the previous research was used for the implementation of processes for the production of new gluten-free food prototypes into the production lines of the involved SMEs, the evaluation of the process conditions and recipes on a technical scale and finally the production of sufficient amounts of high-quality gluten-free bakery products and pasta for the consumer study.
The recipes for gluten-free bread products which have been developed in the course of WP 6 could be used by the industrial partners as a basis to create a range of new breads which are adaptable to consumer requirements. A part of these recipes, namely toasting bread, wholemeal oat bread and pizza were used by the involved bakery companies, adapted and implemented on a technical scale and evaluated regarding texture and sensory properties. Two of the SMEs used the project knowledge to develop new or modify existing breads such as a potato and a linseed bread, white bread and country bread. After product optimisation the SMEs provided sufficient amounts of the new bread prototypes for the final consumer study which was conducted at Hamburg University of Applied Sciences.
In parallel the most promising recipes of pasta which have been developed in WP 7 were implemented by the involved pasta producer on an industrial scale. Various trials were carried out using the new spaghetti line of the company and by applying pea protein isolate, citrus fibre and emulsifier in different combinations. The tests revealed that the addition of emulsifier to the basic recipe gave the final pasta product a better colour, better taste, sufficient weight gain, good texture and the least cooking loss among the different recipes. Weight gain and texture of samples without emulsifier were only poor whereas their grittiness was lower than that of the samples with emulsifier. Finally, six spaghetti recipes, all of them showing good potential for commercialization were developed and produced for the final consumer study in WP 10. Two of the recipes performed very well and the company decided to commercialize one of them on the Italian market.
Evaluation of consumer acceptance of prototype foods
The sensory research conducted in the final research work package (WP 10) at Hamburg University of Applied Sciences aimed at examining the sensory properties of the products from the up-scaling process, and their consumer acceptance. These results may be used to conduct some final improvements of these products beyond this research project, so that they can be successfully launched to the market.
Bread:
All participating consumers highly liked bread, as the mean overall liking score was 7.9 on a nine-point scale. For the celiac patients and the family members, one type of oat bread almost met their expectations towards bread. This was also reflected in their comments on the sample, with the moistness, the crust, and the soft / fluffy texture being most often positively commented on. Some consumers even described it as 'very much like gluten-containing bread', which clearly indicated that in case of this sample, the objectives of this research project have been met. This bread only needed optimization regarding taste (too little intensive), which can be easily achieved by adding more salt, and maybe by increasing roasty aromas through, for example, slightly modifying the baking process.
With the neutral taste being the sample's only main point of consumers' criticism, and the only area of relevant penalty points, this bread should catch up to the consumers' category liking of bread, once the formula is optimized. Linseed bread also achieved comparably high liking ratings. Similar to the above mentioned oat bread, the consumers positively commented on the bread's moistness, its crust, and its consistency, i.e. that it was not crumbly, like many gluten-free breads. They found that the taste was like a typical German Schwarzbrot (dark bread) and some consumers even called it a 'new gluten-free bread experience'.
While the bread did not receive any relevant penalty points, it should be optimized particularly regarding taste, and colour (in these areas, it obtained the highest penalty points). This was also reflected by the consumers criticism, which mainly focused on the bread being too bitter, the aroma being too 'malty', 'like chocolate', 'too intensive compared to taste', and the colour being too dark. The bitterness may be diminished by reducing the linseed content, whereas the malty chocolate-like aroma might be traced back to the use of malt in the formula. However, as the industry partner did not provide any insight into the formula, this may be pure speculation.
With the toasting bread sample, there was still considerable room for optimization, as being reflected by the acceptance score of 5.4 compared to the category liking of 7.9 by the penalty points, and by the consumers' comments. Many consumers expressed that this bread would be good for toasting. They complimented its fluffy, soft texture, the crust, and the fact that it did not crumble. However, the taste was perceived as 'dough-like', 'like flour / starch', and as being too sweet. Furthermore, the consumers disliked the colour being 'yellow', and the bread being too dry. The aroma reminded many consumers of 'typical gluten-free bread'.
The second type of oat bread was clearly disliked. Although healthy consumers (who were the only consumer group to rate this bread, as it could not be produced according to gluten-free standards) generally have lower liking scores for gluten-free samples compared to celiacs and family members, this sample would also have obtained distinctively low liking scores from the latter two consumer groups. With such a low overall liking, penalty analysis only partly indicates where urgent areas for optimization are.
Pizza
The pizza bases did not receive any relevant penalty points. However, this did not indicate that there was no requirement for optimization as these samples urgently needed more development work. With one pizza sample, there was a severe variation in sensory properties between the batches delivered in October 2012 and November 2012. While the October batch was dry and foam-like in texture, the November batch had a rubbery interior, presumably due to an insufficient baking process during manufacturing. The other pizza sample generally was too flat and too hard due to a lack of fine to medium-sized pores. Pasta:
The new spaghetti showed a continuous improvement during the product development phases, in consumer acceptance and sensory perception. In the end, the final prototype did not obtain any penalty points and was well accepted by consumers and comparable to the market leader. The acceptance could be further increased by further reducing the grittiness of the pasta. However, taking into account that pasta usually will be consumed with a sauce, this rest of grittiness is not really significant. This prototype was considered as being ready for market launch.
For all products, sensory profiling was conducted to obtain information about sensory characteristics of different breads, pizza bases and pasta. Significant differences could be seen between the products in several sensory attributes. The products also differed significantly from the gluten-free bench mark products obtained from the market, in several attributes. With the pasta, the cooking time had a significant influence on the gritty mouthfeel, with the grittiness decreasing with increasing cooking time.
Potential Impact:
GLUTENFREE will enable SMEs to produce premium gluten-free bakery products and pasta well accepted by the consumer. This will allow the SMEs to participate within this profitable market sector, and thus significantly improve their competitiveness. The new products have the potential to generate extra annual turnover and additional employment at the SMEs involved. Society will also benefit because celiac patients as well as healthy consumers will have access to a wider variety of tasty gluten-free foods, which will make the diet and life of celiac disease patients and their family members easier.
Main dissemination activities
The dissemination activities in the project were part of WP 11 and comprised four main tasks.
As a first step a project website (Task 11.1) was set-up at the beginning of the project. The website was established by the project coordinator and was updated regularly during the project term. It consists of a public part informing on the project background, objectives and all partners and also links to different celiac societies. Dates of project meetings, workshops and relevant conferences were announced in the News-section. An internal section was also established and used as a communication platform between the beneficiaries. All relevant project documents like deliverables, reports, presentations, meeting minutes and publications were filed on the internal platform and available for project partners and the project officers.
During the project period the RTD partners actively published the obtained knowledge (as far as compatible with the IP policy) in scientific, peer reviewed papers as part of Task 11.2 in order to make these results available to a broad public. In total eight articles were published in prestigious international periodicals. An article in Journal of Cereal Science presents results of the investigation of nutrient composition of seven commonly used commercial gluten-free flours (oat, rice, sorghum, maize, teff, buckwheat and quinoa) compared to wheat and wholemeal wheat flour (Hager et al., 2012a). In a publication in the journal entitled 'European Food Research and Technology', the bread making potential of seven gluten-free flours, wheat and wholemeal wheat flour was compared (Hager et al., 2012b). The study using systematic baking trials based on a central composite design to investigate the influence of HPMC and xanthan as well as their combination on gluten-free model systems was published in Food Hydrocolloids (Hager and Arendt, 2012c).
Results of the development of gluten-free pasta based on oat and teff flour are presented in a paper published in 'European Food Research and Technology' as well as in a paper submitted to Journal of Cereal Science (Hager et al. 2012d; Hager et al. 2012e). Two reviews on the applications of microbial fermentations for production of gluten-free products were published by Arendt et al. (2011) and Zannini et al (2012). Wolter et al. (2012) submitted a paper on the evaluation of the EPS producing strain Weisella cibaria MG1 as starter culture for the production of sourdough from buckwheat, oat, quinoa, teff and wheat flour to the Journal of Food Microbiology.
Other dissemination activities included posters (5), presentations at international scientific conferences (5) and several press releases in national and international media or marketing activities by the SMEs. A detailed summary of all dissemination activities is outlined in the Final plan for the use and dissemination of the knowledge and in section 4.2 in this final report. In addition, the website was used as dissemination tool and a dissemination workshop was organized at the end of the project.
Within Task 11.3 training courses for the SMEs were planned and carried out in order to apply the technological solutions realised in the project by means of specific training in the application of new ingredients and sourdough technique for the production of improved bakery and pasta products. Initially four training events were planned during the project term.
Three of the trainings were carried out as part of internal workshops as follows:
1) 'Sensory evaluation of existing market products', Thursday, 14th April 2010, Freising, Germany;
2) Gluten-free sourdough technology' Wednesday 12th October 2011, Minden, Germany; and
3) 'Technologies for improved gluten-free pasta' 8th May 2012, Fara San Martino, Italy.
The last training 'Technology for improved gluten-free breads' planned for December 2012 was cancelled due to scheduling difficulties by the bakery SMEs.
Initially two workshops were planned, one at mid-term and one at the end of the project. Due to organisational reasons the mid-term workshop was cancelled. Instead, an internal workshop on sourdough technology was organised at project partner Bocker along with the third project meeting in October 2011 in Minden, Germany. In the end of the project an industry workshop entitled 'Gluten-free Foods' was held on 14th of December 2012 at the University College Cork as main dissemination activity. Around 60 participants from industry, universities and other stakeholders from different European countries attended the workshop. Twelve speakers from academia and industry gave presentations and the programme covered all areas of gluten-free foods such as ingredients, texture, sensory aspects, and marketing as well as consumer research.
Exploitation results
The GLUTENFREE project generated a number of valuable results which can be exploited by the involved industry partners. The main results are described below. Regarding IP exploitation the consortium decided in the final project meeting that insufficient new IP was produced from the project to enable the development of a license package or to form the basis of a patent. All new knowledge will therefore be protected via 'confidentiality'. To date the involved companies intend to use these results individually as outlined in the Final Plan of the Use and Dissemination of the Knowledge. They did not agree on a common strategy yet but discussions are ongoing beyond the project.
List of exploitation results:
- Results on the consumer acceptability and consumer preferences of breads;
- Results on functional properties of proteins, fibres, hydrocolloids, their interactions and models to describe their application potential in gluten-free bread and pasta making;
- Protocols for the processing of new plant based protein ingredients with viscoelastic properties to be used as gluten replacers;
- Report on gluten-free sourdough technology: strains, optimal processing conditions;
- Knowledge on aroma optimisation by technological processes, fermentation and ingredient selection;
- Recipes for the production of new gluten-free breads;
- Protocols and recipes for the production of new gluten-free spaghetti;
- Application of gelling-fibre in gluten-free pasta;
- Prototypes of gluten-free oat bread;
- Toasting bread;
- Pizza base;
- Gluten-free spaghetti; and
- Pre-existing sourdough breads (two recipes of gluten-free sourdough breads, potato bread and linseed bread) improved in the project.
References
- Arendt, E.K. Moroni, A., and Zannini, E. 2011. Medical nutrition therapy: use of sourdough lactic acid bacteria as a cell factory for delivering functional biomolecules and food ingredients in gluten free bread. Microbial cell factories 10 Suppl 1:S15
- Hager, A-S., Wolter, A., Jacob, F., Zannini, E., Arendt, E.K. 2012a. Nutritional properties and ultra-structure of commercial gluten free flours from different botanical sources compared to wheat flours, Journal of Cereal Science 56:239-247
- Hager, A-S., Wolter, A., Czerny, M., Bez, J., Zannini, E., Arendt, E.K. Czerny, M. 2012b. Investigation of product quality, sensory profile and ultrastructure of breads made from a range of commercial gluten-free flours compared to their wheat counterparts, European Food Research and Technology, 235(2):333-344
- Hager, A.-S. Arendt, E.K. 2012c. Influence of hydroxypropylmethylcellulose (HPMC), xanthan gum and their combination on loaf specific volume, crumb hardness and crumb grain characteristics of gluten-free bread. Food Hydrocolloids, In Press.
- Hager, A.-S. Lauck, F., Zannini, E., Arendt, E.K. 2012d. Development of gluten-free fresh egg pasta based on oat and teff flour. European Food Research and Technology, 235(5):861-871
- Hager, A.-S. Zannini, E., Bez, J., Czerny, M., Arendt, E.K. 2012e. Sensory properties, microstructure and in vitro digestibility of egg pasta produced from oat, teff and wheat flour. Journal of Cereal Science, Submitted.
- Wolter, A., Hager, A.-S. Zannini, E., Galle, S., Ganzle, M., Arendt, E.K. 2012 Evaluation of the EPS producing strain Weisella cibaria MG1 as starter culture for the production of sourdough from buckwheat, oat, quinoa, teff and wheat flour a comparative study. Journal of Food Microbiology, Submitted.
- Zannini, E., Pontonio, E., Waters, D.M. and Arendt, E.K. 2012. Applications of microbial fermentations for production of gluten-free products and perspectives. Applied Microbiology and Biotechnology 93:473-48
List of Websites:
Adress of the project website:
http://www.glutenfree-project.eu/
Project Coordinator:
Jurgen Bez
Fraunhofer-Institut fur Verfahrenstechnik und Verpackung,
Giggenhauser Str. 35
85354 Freising
Germany
Phone: +49 8161 491-430
E-Mail: juergen.bez@ivv.fraunhofer.de
Project officer:
Raluca Iagher
REA - Research Executive Agency
16 Place Rogier
1210 Brussels
Belgium
Phone: +32 229 99146
E-Mail: Raluca.IAGHER@ec.europa.eu
Common problems of currently available gluten-free products include high starch and low fibre contents, reduced shelf life, strong off-flavours as well as dense, crumbly and dry texture (bread) and untypical colouring, low protein and fibre contents, texture issues such as reduced elasticity and increased cooking losses (pasta). The scientific and technological development activities carried out in the GLUTENFREE project tackled these issues from different angles and delivered a number of valuable results for the involved SMEs.
A baseline sensory study on gluten-free bread and pasta with consumers in Germany, Italy and Ireland at project start revealed information on consumer preferences, acceptance and needs, which provided the basis for target-oriented product development and optimization in the project. An in depth characterisation of gluten-free flours showed that some of these flours represent an interesting alternative to the commonly used wheat flour with regard to their nutritional properties. However, their potential for the production of bread is certainly limited due to the lack of the structure-forming proteins. Therefore, the use of complex formulations combining different gluten-free flours and ingredients such as hydrocolloids and proteins are necessary to get products of similar quality to wheat containing counterparts (i.e. bread or pasta).
With regard to structure forming proteins, processes for the production of new lupin and broad bean protein isolates were optimized regarding the product yield and viscoelastic properties as well as subsequent scale-up and testing on a technical scale. Fundamental research then has generated new knowledge and insights on the effects of a range of proteins, hydrocolloids and fibres on the technological properties of breads and pasta.
Based on these results a wide range of recipes for new gluten-free prototypes were developed in collaboration between the RTD and the industry partners. These included oat bread, toasting bread, ciabatta, French petit pain, German style sourdough bread and pizza base. In addition promising recipes for gluten-free spaghetti enriched with protein and fibre were generated. Finally, the outcomes of the above mentioned research were used for the implementation of new gluten-free food prototypes into the production lines of the involved SMEs and the production of high-quality gluten-free bakery products and pasta for a final consumer acceptance study in the end of the project. The results of this study showed that some breads (oat bread, linseed bread) achieved high consumer acceptance, whereas some other breads and especially the pizza bases need some final improvements prior to their market launch. With regard to pasta, most of the new spaghetti prototypes were well accepted by consumers and one product was comparable to the gluten-free market leader. This prototype was considered as being ready for market launch and was commercialized by the respective pasta producer by the end of the project.
As a conclusion, these results allow the participating food producers to place new or improved products on the market and thus have the potential to generate additional turnover and employment after the end of the project. Celiac as well as healthy consumers will also benefit by having access to a wider range of tasty gluten-free foods.
Project Context and Objectives:
At present up to two per cent of European consumers are suffering from celiac disease and therefore have to adhere to a strict gluten-free diet. Gluten is part of the common cereals wheat, rye and barley and is therefore contained in basic foods like bakery products and pasta. The market for gluten-free products, which is predominately served by small and medium sized enterprises (SMEs), showed considerable growth in the last years. However, the consumer acceptance of respective products is still limited by their insufficient taste, texture and mouth-feel.
Therefore, the overall objective of this research project was to enable SMEs to produce gluten-free bakery products and pasta of high consumer acceptance in order to participate in this profitable market sector and ultimately to significantly improve their competitiveness. This aim was achieved by utilising carefully selected raw materials and ingredients derived from these, namely plant proteins and hydrocolloids, which are able to effectively replace gluten and gluten-containing ingredients in bakery products and pasta. The interaction of different components in the recipes during baking and pasta processing and their influence on the texture and flavour formation was analysed in order to allow specific improvements. Sensory improvement should be ensured by using selected raw materials, aromatic malts or sourdoughs. Evaluation of consumer preferences, acceptance and needs was an integral part of the research and allowed target-oriented food developments and promised well accepted products.The corresponding research, technological development and innovation related activities focused on the following work packages and specific objectives:
WP 1: Baseline study on gluten-free foodstuff
- To evaluate existing gluten-free bakery products and pasta and to identify consumer preferences and needs.
- To derive targets for research and development with respect to product improvement.
WP 2: Screening of raw materials and selection of ingredients
- To select promising ingredients for the development of gluten-free baked goods and pasta.
- To identify promising alternative raw materials from the range of grain legumes and pseudo-cereals suited for the production of protein ingredients with viscoelastic and film-forming properties for gluten-free bakery products and pasta.
WP 3: Process development for the production of protein ingredients with viscoelastic properties
- To identify key parameters influencing the yield and properties of protein ingredients and derive optimised processing conditions.
- To develop economic feasible processes for the production of plant based protein ingredients considering the best combination of techno-functionality, sensory properties and price.
WP 4: Bio-engineering structural and nutritional value of gluten-free foods
- To identify starter cultures suitable for the fermentation of gluten-free sourdough.
- To isolate and identify exopolysaccharides (EPS) from specific lactic acid bacteria (LAB) strains.
- To isolate lactic acid bacteria (LAB) producing antifungal compounds.
- To apply these LAB in sourdough and identify the positive effects on final product quality.
- To identify the causes of these positive effects on product quality.
WP 5: Interactions of ingredients in baking and pasta processing
- To use advanced rheological tools to rapidly evaluate the strength and functionality of the functional ingredients selected in WP 2.
- To evaluate the effect of the functional ingredients on pasta and bread quality.
- To develop predictive models for the effect of ingredient functionality on product quality for further product development.
- To evaluate the importance of protein, starch and other functional ingredients interactions in pasta and bread.
WP 6: Development of prototype bakery products
- To develop recipes and process parameters for the production of high quality gluten-free bread products typical of western, central and southern Europe.
- To evaluate the effect of functional and nutritive ingredients on bread quality.
- To implement the recipes and production procedures for gluten-free breads on industrial scale.
WP 7: Development of prototype pasta
- To develop gluten-free pasta having firm texture, good taste and high nutritional value.
- Scale-up: Transfer the recipes and production procedures for different model foods to production scale.
- To implement process conditions and recipes, identified and characterised in the development stage, to the production lines of the SME.
WP 8: Aroma analysis and improvement
- Sensory and analytical characterisation of prototype aroma and aroma development.
WP 9: Scale-up and production of food prototypes for the consumer acceptance study
- To implement processes for the production of new gluten-free food prototypes into the production lines of SMEs.
- To evaluate the process conditions and recipes identified on a technical scale.
- To produce sufficient amounts of high-quality gluten-free bakery products and pasta for the final consumer acceptance study (WP 10).
WP 10: Evaluation of consumer acceptance of prototype foods
- To evaluate the developed gluten-free bakery and pasta products in several hedonic tests with consumers with and without celiac disease.
- Sensory Profiling of developed prototypes.
- Preference mapping: combining data with PCA to find preference driving factors.
WP 11: Dissemination and Training
- To disseminate the outcomes of the RTD activities to SMEs, consumers and research institutes.
- To enable the SMEs to apply the technological solutions realised in the project by means of specific training in the application of new ingredients and sourdough technique for the production of improved bakery and pasta products.
Project Results:
Baseline study on gluten-free foodstuff
As a first step of the research a baseline sensory study was carried out in work package 1 (WP 1) aiming to obtain a market overview of gluten-free white bread, dark bread and pasta in four European countries (Germany, Ireland, Italy and Sweden). From the breads and pasta available within these markets, samples were selected for sensory evaluation, with the intention to cover a wide product range and thus a wide range of sensory characteristics.
Sensory product properties and differences were assessed by a trained panel. This data was combined with preference data of consumers from three countries (celiac patients, family members, and non-affected/ healthy consumers in Germany, Ireland and Italy) to reveal factors, which drive, but also adversely affect consumer liking via external preference mapping and penalty analysis. This information provided the basis of the improvement and development of new gluten-free breads and pasta, aiming at achieving high consumers' acceptance.
Characterization of the breads:
Selected white bread, dark bread and pasta varieties, available on the German, Italian, Irish or Swedish markets, were sensorily profiled by a trained panel, and hedonically evaluated by consumers in Germany, Italy and Ireland. For each country the consumers were segmented into three groups: celiac patients, family members, and healthy / non-affected persons. The profiling and hedonic datasets were correlated via multivariate statistical methods to extract preference driving and decreasing factors, which provided directions for product development and optimization.
The white bread samples selected for this research differed significantly in nearly all evaluated product characteristics. This indicated that the selected breads represented a broad range of the products available within the markets. This in turn allowed for identifying a broad range of preference driving and decreasing factors. The German breads showed most of these product characteristics which were recommended to be focused on in product optimization.
The dark bread samples also covered a large variety of breads, as indicated by significant product differences in all assessed attributes, particularly in the mouthfeel. Despite of being different types of bread as such, white and dark bread shared numerous preference driving and decreasing factors, indicating that for the consumer bread needs to have certain characteristics, irrespective of bread type or composition.
Characterization of the pasta:
In contrast to the breads, the pasta samples did not show big sensory differences. In accordance with observations reported in literature (Moskowitz, 2003), there were no relevant overall liking differences between markets and between consumer segments. For future consumer tests this opened the opportunity of conducting the tests with easier-to-reach consumer samples and focusing only on one market (Germany).
Screening of raw materials and selection of ingredients
As part of WP 2, an extensive raw material characterisation was carried out on seven commercial gluten free flours (teff, rice, sorghum, maize, quinoa, buckwheat, oat) and results were compared to wheat and wholemeal wheat flour. Furthermore WP 2 focused on the screening and evaluation of hydrocolloids and protein isolates to identify the most suitable ingredients for gluten-free bread and pasta development.
The results of the analytical characterisation of the gluten-free flours showed that, although most commercial bread and pasta products are based on rice and maize, more nutritious gluten-free raw materials would be available. While endosperm-derived white wheat flour, maize and rice flour contain only 3.4%, 2.6% and 0.4% fibre, quinoa flour contains a much higher amount (7.1%). Also regarding protein content, maize and rice flour are poor when compared to teff or pseudocereal flours such as buckwheat and quinoa.
The seven flours were used to produce gluten-free breads using basic recipes simply consisting of flour, water, sugar, salt and yeast. While the majority of existing publications give information on complex gluten-free formulations, this more fundamental approach was taken to understand the baking properties of the individual flour raw materials. Quality evaluation of resulting loaves showed big differences not only in visual appearance but also in final loaf volume, crumb firmness, crumb structure, shelf life and taste attributes.
When comparing white wheat bread to all other breads it became apparent that crumb structure and cell characteristics were very distinct. Quinoa, buckwheat, maize, teff and wholemeal wheat breads showed very dense crumbs while oat and rice breads had a more open texture compared to wheat bread. As the same mixing regime was followed for the production of all gluten-free breads, it can be assumed that this is due to the differences in dough consistency.
Obtaining an acceptable shelf life that would allow selling fresh gluten-free bread represents one of the biggest challenges. Due to high aw values and the high moisture contents resulting from the increased water addition levels, mould growth commences earlier. The results of shelf life trials showed that on average, mould growth commences on day 4 for most gluten-free breads and for rice bread already on day 3. In order to place a bread product in the supermarket, 8 to 10 days of shelf life are required. This issue was addressed in work package 4, which aimed at prolonging microbial shelf life by using sourdough fermented with lactic acid bacteria producing antifungal compounds.
Apart from microbial deterioration, shelf life of bread is also determined by the staling behaviour. Crumb firming, attributed mainly to re-crystallization of amylopectin and water redistribution between crumb and crust, was especially pronounced in oat, white wheat and maize breads. The staling rate of teff bread was significantly lower than for the other gluten-free flours. This was expected since teff starch has a lower tendency to retrograde as shown in the Rapid Viscoanalyser tests.
The flours also showed significant differences regarding their sensory properties. Each of them exhibited a very specific aroma profile due to characteristic odour attributes and different odour intensities which can be influenced by fermentation and baking (details see WP 8, 'Aroma analysis and improvement').
Overall, the gluten-free flours screened represent an interesting alternative to the commonly used wheat flour but their potential for the production of bread is certainly limited due to the lack of the structure-forming proteins. In order to get a product of similar quality to wheat bread, it is probably most practicable to use a complex formulation combining different gluten-free flours and ingredients such as hydrocolloids and proteins. Their different properties with regard to bread making, allow choosing suitable flours according to the desired bread characteristics.
With regard to proteins a series of experiments was carried out to obtain protein isolates with viscoelastic properties from seeds of different grain legumes (lupin, soy, pea, broad bean), pseudocereals (buckwheat, quinoa, amaranth) and cereals (oat). The process used was established for lupin and consists of a salt induced protein extraction and a dilution precipitation as the main steps. The feasibility of this process for the other raw materials was evaluated on basis of the overall and protein yields, the protein content and the rheological characteristics of respective protein isolates as well as their performance in baking tests.
With regard to the overall as well as the protein yield the extraction of grain legumes revealed the best results. The highest overall yield was achieved for soy protein isolate followed by broad bean, lupin and pea protein isolates.
Oscillation tests performed with a rheometer revealed significant differences in the viscoelastic properties of the respective protein isolates. While lupin protein isolate showed fluid-like characteristics and a smooth, honey-like texture, all other protein isolates showed gel-like characteristics and more stiff texture. The structure strength of the protein isolates was highest for quinoa and soy and lowest for broad bean, lupin, pea and oat.
The application of the protein isolates in a complex formulation generally increased the stiffness of the dough compared to the reference. But the rheological characteristics of the different types of dough did not generally correlate with the viscoelastic properties of the protein isolates. In baking tests only lupin protein isolate improved the specific loaf volume, which was similar to wheat flour bread. In addition, the lupin breads had the best crumb texture. The breads with the other legume protein isolates achieved loaf volumes comparable to the control gluten-free bread.
Process development for the production of protein ingredients with viscoleastic properties
In WP 3 a systematic variation of the key processing parameters using response surface methodology (RSM) was carried out in order to identify the best conditions for the production of viscoelastic protein isolates from lupin and broad bean regarding yield and technological properties. Both raw materials were selected by the consortium on basis of the results achieved in WP 2. The resulting isolates were analytically characterized for their rheological, techno-functional (water binding, emulsifying) and baking properties.
For lupin seeds, the standard process consisting of a salt-induced protein extraction and dilution precipitation was optimised regarding the product yield. A face centred central composite design was applied to determine the operating conditions due to the product yield. The parameters pH value, NaCl-concentration, solid to salt solution ratio, extract to water ratio during precipitation, pH value of the extract before precipitation, centrifugation and sedimentation time were chosen as factors varying in a specific range and were identified to be significant. Finally, an experimental plan of 50 combinations was obtained and evaluated, while the order of experiments was fully randomized to avoid any systematic bias. As a result the optimum conditions to maximize the product yield could be established. In a further step, the protein isolates achieved at different parameter settings were analysed with regard to their techno-functional and their rheological properties as well as their performance in gluten-free baking.
Applied in a complex gluten-free formulation the different lupin protein isolates led to a fluid dough structure with low complex modulus but the addition of LPI generally increased the stiffness of the dough compared to the control dough without protein. The viscoelastic properties of the dough however did not generally correlate to that of the protein isolates. Regarding the specific loaf volume the isolate samples produced at low salt level and neutral pH achieved the lowest volume. The highest loaf volume was observed with the isolate recovered at standard conditions. This isolate also achieved the softest bread crumb. Breads produced with the other protein isolates did not improve the specific volume compared to the control bread without protein addition. Taking into account the product yield of LPI, it can be concluded that the product yield is contrary to product properties: isolates achieved at highest yield (18-22 %) have lowest protein content, highest fat content and worst baking properties, but highest emulsifying and water hydration capacities and highest resistance to deformation.
As a further raw material, broad bean was chosen to produce an alternative protein isolate for gluten replacement. Similar to the procedure used for lupin seeds, a face centred central composite design was applied to determine the optimised parameter settings by only varying the factors pH of protein extraction, salt concentration in the extract and pH of protein precipitation. As for lupin, the statistical variation of process conditions allowed to identify optimised process conditions to gain a satisfactory amount of protein isolate. Thus, a maximum product yield of ~20 % was achieved, which is equal to the yield obtained when using lupin seeds. The investigated broad bean protein isolates (BPI) on average showed a higher water hydration capacity but lower emulsifying capacity than the lupin protein isolates.
In the rheological analysis all broad bean isolates showed the characteristics of viscoelastic gels (G'>G'') with large differences in their stiffness and resistance to deformation (G*). The protein isolate showing the highest loaf specific volume within the different isolates was achieved with the highest protein yield (56.1%) and at a low salt level, which makes it the most interesting product with regard to the technical implementation. However, further conclusions regarding the correlation of parameter settings in the production of the different BPI with their product properties cannot be derived from the results.
Finally a scale-up of the processes for lupin and broad bean protein processing from the laboratory scale to the facilities of Alberts, the involved SME, was realised. However, the implementation of the process required a few adjustments, e.g. the seed pre-treatment was carried out at the Fraunhofer institute. The protein isolates was successfully prepared at Alberts starting from lupin flakes or broad bean flour respectively. Since Alberts did not have the facilities for drying the protein isolate, the protein isolates were pasteurized and spray-dried at the Fraunhofer institute.
An economic feasibility study showed that with further optimization of the process (higher protein yield, energy saving, water purification, etc.) the protein isolates may be commercialized successfully after the end of the project.
Bio engineering structural and nutritional value of gluten-free foods
As part of WP 4 ('Bio-engineering structural and nutritional value of gluten-free foods') strains of LAB haven been:
a) isolated from different gluten-free flours,
b) evaluated for their ability to successfully ferment and survive in a gluten-free system and
c) screened for their ability to produce EPS and antimicrobial compounds.
Finally, the bread quality parameters have been evaluated on bread samples containing 20% of sourdough fermented with the selected LAB strains. Suitable protocols for the applications of selected starter cultures in gluten-free sourdough were developed. The positive effects of sourdough are mainly due to the metabolic activities of the lactic acid bacteria performing the fermentation. A strain with antifungal properties (Lactobacillus plantarum ZA 7) and an exopolysaccharide producing strain were chosen (Weisella cibaria ZA 13). Their effects on aroma, structure and shelf life of baked goods was studied. Quality of breads produced from different gluten-free flours was influenced to different extent by sourdough fermentation. Apart from oat and wheat sourdough bread the application of both sourdough types did not affect the specific volume in gluten-free breads compared to control breads. In terms of crumb hardness the trend that breads baked with strain ZA 13 give lower values for crumb hardness is apparent in most breads, but not in oat and sorghum sourdough bread.
Hardness values for ZA 7 breads were not significantly different from those of control breads. ZA 13 reduced crumb hardness after 5 days of storage to a higher extent than ZA 7. This shows that the effect seems to be independent from the sourdough acidity, as sourdoughs from ZA 7 show higher acid values (TTA) but not higher crumb hardness than the non-acidified control breads. Even though oat and wheat sourdough bread had a soft crumb on day 0, the rate of staling was high. In general, the rate of staling was slightly delayed for some gluten-free breads but compared to control only significantly in oat and wheat bread. The tendency of reducing crumb chewiness compared to gluten-free control breads could be found in both sourdough trials but was only significant for breads fermented with ZA 13.
Both strains are capable to reduce chewiness in buckwheat, quinoa and teff breads compared to control breads with the trend of larger influence by ZA 13 than ZA 7. Differing from that, either no significant (with ZA 13 and oat or wheat bread) or the adverse effect (for ZA 7 and oat bread, and for ZA 13 and ZA 7 in sorghum sourdough bread) was found in certain breads. Furthermore, the only influence of sourdough on springiness of bread crumb was found for wheat bread with a significantly increased springiness for both sourdough breads.
Overall, based on standard bread-quality parameters, oat flour resulted in good quality breads, indicated by high loaf specific volume and a soft crumb. Though the other gluten-free flours had poorer bread making properties, quinoa and buckwheat resulted in agreeable breads with good crumb structure. The ratio of pore area to slice area and the cell volume of the bread crumb were either higher or equal compared to gluten-free control breads. This characteristic of crumb structure was improved by sourdough application in general and especially when ZA 13 was used as a starter culture. Concerning shelf life, it was possible to observe an extension of shelf life only for wheat sourdough bread when ZA 7 was used. The application of ZA 7 sourdough in gluten-free formulation showed no difference to the fungal spoilage of the control breads.
The application of sourdough ZA 13 on hardness of buckwheat bread was higher than for the other gluten-free breads. Quinoa and teff breads had also an acceptable crumb structure, but the loaf volume was low. Sorghum flour on its own was not suitable for bread making, due to poor structure, high crumb hardness and rate of staling. Regarding the influence of the chosen sourdough strains, it can be said, that what applies to the wheat system, does not necessarily apply to the gluten-free system. The improvement of gluten-free bread by lactic acid bacteria was much less pronounced than in wheat bread.
Interactions of ingredients in baking and pasta processing
WP 5 aimed at the evaluation of the effect of functional ingredients (starch, proteins, hydrocolloids) on bread and pasta quality in more detail. Starch plays an important role in pasta as well as bread, especially in gluten-free systems. The Rapid Viscoanalyser and differential scanning calorimetry represent two important tools for the study of starch behaviour. They were applied in this work package to get information on pasting properties and gelatinisation temperatures of the nine flours described in WP 1. Viscosity behaviour during heating from 50°C to 95°C, which gives an indication of the capacity of different starches to retain water and swell, was shown to be different for all flours. It was concluded that pseudocereal starches are more resistant to heating and shear stress. Results showed that, except for oat and quinoa, gelatinisation temperatures were significantly higher for the gluten-free flours compared to wheat flour.
In addition dough characteristics of the different flours were investigated in more detail by means of fermentability trials (Rheofermentometer) and rheological tests as well as confocal laser scanning microscopy and scanning electron microscopy. White wheat and oat samples reached a maximum dough development height unmatched by the gluten-free batters or the wholemeal wheat flour dough, indicating superior viscoelastic properties of oat. Gas production was highest in teff, buckwheat and quinoa batters, indicating that these flours have a more favourable sugar composition for yeast fermentation. Scanning electron micrographs of bread crumbs revealed that distinct starch granules are still visible in wheat and oat breads, while starch is gelatinized to a much wider extent in case of all other gluten-free flours. The observation of the ultrastructure of bread showed the difference between wheat samples and samples where the structure forming gluten network is absent. These observed pictures visualized on a microscopic level the reduced structural quality, associated with gluten-free products. Thus microscopy is an excellent tool to visualise processes occurring during baking.
Another part of WP 5 was the evaluation of cereals other than wheat and further ingredients on pasta quality. Due to outstanding nutritional characteristics, teff and oat flour were chosen. As mentioned above, commercial gluten-free pasta is generally low in protein. Therefore different protein ingredients were evaluated (egg white powder, whole egg and egg yolk powder and prototype lupin protein isolate). Results showed that egg white powder not only increased the nutritional value of gluten-free pasta, but also improved textural properties such as elasticity, firmness and stickiness.
Many commercial products contain emulsifiers and several publications report on the reduction of cooking loss upon incorporation of emulsifiers. Looking at the response surface graphs of the current study, this effect was not observed. Looking at the results of texture profile analysis as well as sensory evaluation by a trained panel, it could be seen that when using oat flour, spaghetti comparable to wheat based products can be obtained. Teff pasta was characterised by high cooking loss and dryness. Differences in pasting properties and gelatinisation temperatures of oat, teff and wheat were measured. The developed oat and teff pasta are nutritionally superior to many gluten-free pasta products currently on the market.
Scanning electron micrographs of pasta produced in this study showed in case of wheat samples a strong outer layer of coagulated protein that entraps the ungelatinized starch granules. This layer was less pronounced in case of gluten-free samples, which explains the higher cooking loss and stickiness observed in these products. Confocal Laser Scanning Microscopy is an interesting tool for ultra-structure evaluation of foods as components can be stained selectively. Changes in the starch structure occurring during cooking of pasta were observed using a Periodic Schiff's acid kit as staining technique. Significant changes were observed upon cooking. Pasta represents a limited-water system and hence, even after cooking for 12 min, a great proportion of starch is still present in its granular form. By using different hydrocolloids and protein ingredients (i.e. guar gum, soy and whey protein) also a formulation for dough suitable for the production of e.g. lasagna sheets could be developed with a visual appearance and consistency similar to that of the wheat containing counterpart.
In WPs 6 and 7 the results achieved in the previous research work were used to develop different new recipes for gluten-free breads and pasta in cooperation with the involved industry partners.
Development of prototype bakery products
WP 6 dealt with the development of different gluten-free bread formulations suitable for Western, Central and Southern Europe. As first step commercial wheat containing soda bread, toasting bread and wholemeal oat bread were purchased, to get an idea on the characteristics and special features of these bread types. In a next step gluten-free counterparts were developed. A wide product range including soda bread, oat bread, toasting bread, ciabatta, German style sourdough bread, pizza base and petit pain were developed. Soda bread is a traditional bread variety of Ireland, prepared using sodium bicarbonate as raising agent instead of yeast. The soda bread dough developed in this study was formable and hence the characteristic soda bread shape was achieved without the need of tins. Crumb structure of the resulting bread was similar to the gluten containing counterpart.
The toasting bread developed under this task was described as soft coarse textured product with a slightly yellow crumb and a relatively low staling rate. Two commercial flours (i.e. wholegrain oat flour and endosperm oat flour) as well as flour milled from gluten-free oats supplied by Semper (Sweden) were used for the development of oat bread. All three breads showed lower values for crumb hardness than the wheat containing wholemeal bread, with the lowest being bread produced from wholegrain oat flour. When comparing the crumb grain to a wheat wholegrain product, the developed breads under this task were more favourable as the structure was more open and aerated.
Using a dried sourdough supplied by Bocker (Sarrasin, buckwheat-based) gluten-free bread suitable for the central European market was successfully developed. Based on the flour mixture used for gluten-free breads a sheetable pizza base recipe has been developed. Fat was added in form of olive oil. Compared to the commercial wheat-based recipe, the gluten-free pizza base developed in WP 6 showed a higher resistance to bending, resulting in a higher value for firmness. Gluten-free French Petit Pain was developed using the same recipe as for gluten-free toasting bread. The structure of the developed gluten-free Ciabatta Style bread was denser than the wheat containing ciabatta.
For all developed gluten-free bread formulations, shelf life was evaluated and enhanced with calcium propionate as well as sorbate. In addition to the development of gluten-free breads, the procedure for the addition of nutritive and technologically favourable ingredients to these formulations was optimised. The darkening effect is desirable since gluten-free breads are usually considerably lighter than their wheat containing counterparts. Incorporation of oat b beta-glucan had a remarkable effect of softening the crumb, as indicated by lower hardness values. The addition of inulin resulted in higher crumb hardness.
Hence it was concluded, that the two grades of citrus fibre can be interchanged without effect on product quality. Malted gluten-free grains may be another interesting potential ingredient for the gluten-free baking industry, as malt enzymes have the potential to increase the rate of fermentation, enhance crust colour formation, volume and soften the crumb. As the effect of enzymes on gluten-free bread quality depends strongly on the formulation and ingredients, four complex and one simple oat recipe were used for these trials. The use of oat malt at concentrations higher than 1% led to a significant increase in the specific volume in three of the four recipes. However, a 2.5% malt addition deteriorated the crumb structure of two of the formulations, leading to the formation of large holes and a hollow loaf.
Quinoa malt had no effect on the specific volume in any of the recipes. In an attempt to further improve the German style sourdough bread, the originally used sourdough was replaced by two different, dried, gluten-free sourdoughs, based on maize and buckwheat. The use of different sourdoughs in the formulation of German style bread had no influence on the specific volume. Crumb hardness on day 0 was highest for German style bread baked with buckwheat, while the original- and corn-sourdough breads showed lower crumb hardness. Gluten-free toasting bread was baked containing xanthan, dextran and a combination thereof as well as without hydrocolloid addition.
There was a significant increase of loaf volume upon hydrocolloid addition, especially if xanthan gum was added. Crumb hardness was reduced and crumb springiness increased by xanthan, dextran and a combination of both. Therefore it was concluded, that xanthan and dextran are able to improve the quality of gluten-free toasting bread. Therefore, the effect of xanthan and/or dextran on all other developed formulations was evaluated as well. The effect of these hydrocolloids on the properties of different gluten-free bread recipes was very versatile.
Summarising these results it can be stated that fibre ingredients are not only beneficial from a nutritional point of view, but can also have a positive influence on bread quality as shown for oat beta-glucan, inulin or dextran. However, it has to be kept in mind that an adjustment of water level can be necessary. Malt and sourdough addition can also have a positive influence on gluten-free breads, especially in terms of loaf volume and flavour improvement. The effect of these ingredients depends strongly on the bread matrix and ingredients have to be chosen carefully in order to achieve the desired effects.
In a collaboration between UCC, Fraunhofer and the industry partners, these bread prototypes were then up-scaled as part of WP 9 ('Scale-up and production of food prototypes for the consumer acceptance study').
Development of prototype pasta
WP 7 aimed to develop high quality gluten-free pasta with firm texture, good taste and nutritional value on a lab scale and the transfer of the new recipes and production conditions to the production lines of the SMEs. The focus thereby was on the improvement of protein and fibre content of a basic pasta mixture of the involved pasta producer consisting of corn and rice flour without impairing the sensory properties and the texture.
In the first step several gluten-free commercial pasta samples were investigated with regard to their technological and sensory properties. The technological parameters analysed by using standard methods were the cooking loss, water absorption, firmness, stickiness, tensile strength and elasticity The analytical values achieved were used as references for the development of new gluten-free pasta samples and their evaluation in the second phase of the study. In addition, the technological and sensory properties of the gluten-free samples were compared with the values of two commercial gluten containing spaghetti, both representing best quality in Italy.
Thereafter proteins and fibres were added in varying amounts to the pasta matrix and the effects on the main technological quality parameters like cooking time, water absorption, cooking loss, firmness, stickiness, tensile strength, elasticity and sensory quality (see WP 8) were evaluated and compared to commercial references. The ingredients used were lupin protein isolate developed in WP 3, commercial pea and potato protein isolates and gelling and non-gelling citrus fibres. The selection of ingredients and the concentrations used were adapted to the specific objectives of the pasta producer involved in this WP (i.e. plant based proteins, no allergens, natural ingredients, no E-numbers, increase of the protein and/or the fibre content, respectively).
A general observation was that protein and/or fibre enrichment mainly had positive effects on the technological properties of gluten-free spaghetti but only a few ingredients or mixtures were able to improve all of the technological properties. Sensory properties were not or only slightly influenced by the ingredients, and there were no negative effects on aroma and taste. With regard to the objective of the study it was therefore possible to prepare a number of recipes with increased protein and fibre content and sufficient texture and sensory properties that could be used for the scale-up to technical scale.
Regarding the cooking loss most of the ingredients had a positive effect reducing this parameter compared to the basic mixture. Thus, some of them even reached the values of the gluten containing references. The biggest effect was obtained by combinations of lupin protein, citrus fibres and emulsifier. A number of other samples also reached remarkable reductions especially the samples containing lupin or potato protein isolates or these proteins in combination with fibres.
By applying standard processing conditions for pasta production, the utmost increase of the firmness was reached by adding HPMC or potato protein. The firmness of these samples was even higher as the durum wheat reference sample. Other ingredients which also significantly increased the firmness where pea protein isolate, potato protein isolate, citrus fibre and combinations thereof. The addition of lupin protein generally deteriorated the firmness.
The pasta stickiness was negatively influenced (increase) by the majority of mixtures, especially by the samples consisting of pea or lupin protein isolates and their combinations with citrus fibres and HPMC. The tensile strength and the elongation were only improved by a few samples. Most effective were pasta samples containing pea protein isolate, citrus fibre and emulsifier. The elongation was improved by the addition of lupin protein isolate and by the addition of gelling fibres respectively. The technological variations carried out with two types of citrus gelling fibres revealed some interesting perspectives for industrial use. Despite reductions of the cooking loss some combinations were able to increase the firmness and decrease the stickiness very effectively.
Based on the results, several recipes could be derived for the further development of protein and fibre enriched pasta on an industrial scale using standard production conditions. The most promising recipes were based on combinations of potato protein, pea protein, citrus fibres and emulsifier. Moreover, a combination of lupin protein isolate with citrus fibre and emulsifier showed very good results, but is not in accordance with the company's philosophy to use no potential allergens. However, if the company wants to produce pasta either enriched with protein or enriched with fibre the best ingredients appeared to be potato protein isolate or non-gelling citrus fibre. On basis of these results Bioalimenta was able to produce a number of new gluten-free spaghetti using a new production line which was established in June 2012 (see below, WP 9). Samples of these recipes were sent to research partners for sensory (HAW) and technological (Fraunhofer) evaluation.
Aroma analysis and improvement
Investigations in WP 8 were run in parallel to the basic investigations of the effects of functional ingredients on product properties (WPs 2-5) and prototype bread and pasta development (WP 6 and 7) and focused on the sensory and analytical characterisation of prototype aroma and aroma development. As mentioned previously, sensory investigations of amaranth, buckwheat, oat, quinoa, rice, sorghum, teff and wheat flours demonstrated that each of gluten-free flour exhibits a very specific aroma due to characteristic odour attributes.
Moistening intensified in many flours the intensities of several attributes, which may be caused by a high degree of macromolecular constituents to adsorb odour-active compounds. The flours consist in part of high odorant loads, which are the basis for the final bread aroma. Hedonic test revealed oat, rice and teff as flours with the highest sensory liking. The preference of e.g. oat was correlated with the oat flakes-like attribute, which was very intense in oat flour. In contrast, pea and quinoa showed a low acceptance in hedonic tests, which was caused by intense pea-like odour notes.
The aroma of yeast breads, which were prepared from the flours, showed distinct changes in the aroma profiles and demonstrated the high influence of fermentation and baking on the odorant composition. Interestingly, some characteristic flour attributes were not perceivable or their intensities were decreased in the crumbs of several breads. For instance, the oat flakes-like attribute was reduced in the crumb of oat bread, but a positive crumb aroma was evaluated by the hedonic test. In consequence, degradation of the key odorant to a certain extent and generation of further odorants with characteristic attributes, which were detected in the aroma profile of oat bread, can be concluded for this observation. The undesirable pea-like aroma of quinoa flour could not be removed after fermentation and baking.
The developed pasta basic mixture and all improved recipes for pasta showed no influence on taste in comparison to commercial wheat based reference pasta. The aroma of the gluten-free pasta samples differed from the reference pasta in particular in a reduced intensity of the typical pasta-like and buttery attribute and a dominant cooked rice-like note. The texture properties juiciness, firmness and stickiness were influenced by the tested protein ingredients, emulsifier and fibres. In particular, stickiness was in part much higher than the references. Regarding juiciness and firmness, a combination of lupin protein or gel fibre with the basic mixture showed comparable properties. The investigated pasta samples showed that lupin, gel fibres and emulsifier in combination with the developed basic mixture lead to pasta with comparable textural properties than a commercial reference pasta. However, the aroma should be optimised.
Scale-up and production of food prototypes for the consumer acceptance study
In WP 9 'Scale-up and production of food prototypes for the consumer acceptance study' the outcomes of the previous research was used for the implementation of processes for the production of new gluten-free food prototypes into the production lines of the involved SMEs, the evaluation of the process conditions and recipes on a technical scale and finally the production of sufficient amounts of high-quality gluten-free bakery products and pasta for the consumer study.
The recipes for gluten-free bread products which have been developed in the course of WP 6 could be used by the industrial partners as a basis to create a range of new breads which are adaptable to consumer requirements. A part of these recipes, namely toasting bread, wholemeal oat bread and pizza were used by the involved bakery companies, adapted and implemented on a technical scale and evaluated regarding texture and sensory properties. Two of the SMEs used the project knowledge to develop new or modify existing breads such as a potato and a linseed bread, white bread and country bread. After product optimisation the SMEs provided sufficient amounts of the new bread prototypes for the final consumer study which was conducted at Hamburg University of Applied Sciences.
In parallel the most promising recipes of pasta which have been developed in WP 7 were implemented by the involved pasta producer on an industrial scale. Various trials were carried out using the new spaghetti line of the company and by applying pea protein isolate, citrus fibre and emulsifier in different combinations. The tests revealed that the addition of emulsifier to the basic recipe gave the final pasta product a better colour, better taste, sufficient weight gain, good texture and the least cooking loss among the different recipes. Weight gain and texture of samples without emulsifier were only poor whereas their grittiness was lower than that of the samples with emulsifier. Finally, six spaghetti recipes, all of them showing good potential for commercialization were developed and produced for the final consumer study in WP 10. Two of the recipes performed very well and the company decided to commercialize one of them on the Italian market.
Evaluation of consumer acceptance of prototype foods
The sensory research conducted in the final research work package (WP 10) at Hamburg University of Applied Sciences aimed at examining the sensory properties of the products from the up-scaling process, and their consumer acceptance. These results may be used to conduct some final improvements of these products beyond this research project, so that they can be successfully launched to the market.
Bread:
All participating consumers highly liked bread, as the mean overall liking score was 7.9 on a nine-point scale. For the celiac patients and the family members, one type of oat bread almost met their expectations towards bread. This was also reflected in their comments on the sample, with the moistness, the crust, and the soft / fluffy texture being most often positively commented on. Some consumers even described it as 'very much like gluten-containing bread', which clearly indicated that in case of this sample, the objectives of this research project have been met. This bread only needed optimization regarding taste (too little intensive), which can be easily achieved by adding more salt, and maybe by increasing roasty aromas through, for example, slightly modifying the baking process.
With the neutral taste being the sample's only main point of consumers' criticism, and the only area of relevant penalty points, this bread should catch up to the consumers' category liking of bread, once the formula is optimized. Linseed bread also achieved comparably high liking ratings. Similar to the above mentioned oat bread, the consumers positively commented on the bread's moistness, its crust, and its consistency, i.e. that it was not crumbly, like many gluten-free breads. They found that the taste was like a typical German Schwarzbrot (dark bread) and some consumers even called it a 'new gluten-free bread experience'.
While the bread did not receive any relevant penalty points, it should be optimized particularly regarding taste, and colour (in these areas, it obtained the highest penalty points). This was also reflected by the consumers criticism, which mainly focused on the bread being too bitter, the aroma being too 'malty', 'like chocolate', 'too intensive compared to taste', and the colour being too dark. The bitterness may be diminished by reducing the linseed content, whereas the malty chocolate-like aroma might be traced back to the use of malt in the formula. However, as the industry partner did not provide any insight into the formula, this may be pure speculation.
With the toasting bread sample, there was still considerable room for optimization, as being reflected by the acceptance score of 5.4 compared to the category liking of 7.9 by the penalty points, and by the consumers' comments. Many consumers expressed that this bread would be good for toasting. They complimented its fluffy, soft texture, the crust, and the fact that it did not crumble. However, the taste was perceived as 'dough-like', 'like flour / starch', and as being too sweet. Furthermore, the consumers disliked the colour being 'yellow', and the bread being too dry. The aroma reminded many consumers of 'typical gluten-free bread'.
The second type of oat bread was clearly disliked. Although healthy consumers (who were the only consumer group to rate this bread, as it could not be produced according to gluten-free standards) generally have lower liking scores for gluten-free samples compared to celiacs and family members, this sample would also have obtained distinctively low liking scores from the latter two consumer groups. With such a low overall liking, penalty analysis only partly indicates where urgent areas for optimization are.
Pizza
The pizza bases did not receive any relevant penalty points. However, this did not indicate that there was no requirement for optimization as these samples urgently needed more development work. With one pizza sample, there was a severe variation in sensory properties between the batches delivered in October 2012 and November 2012. While the October batch was dry and foam-like in texture, the November batch had a rubbery interior, presumably due to an insufficient baking process during manufacturing. The other pizza sample generally was too flat and too hard due to a lack of fine to medium-sized pores. Pasta:
The new spaghetti showed a continuous improvement during the product development phases, in consumer acceptance and sensory perception. In the end, the final prototype did not obtain any penalty points and was well accepted by consumers and comparable to the market leader. The acceptance could be further increased by further reducing the grittiness of the pasta. However, taking into account that pasta usually will be consumed with a sauce, this rest of grittiness is not really significant. This prototype was considered as being ready for market launch.
For all products, sensory profiling was conducted to obtain information about sensory characteristics of different breads, pizza bases and pasta. Significant differences could be seen between the products in several sensory attributes. The products also differed significantly from the gluten-free bench mark products obtained from the market, in several attributes. With the pasta, the cooking time had a significant influence on the gritty mouthfeel, with the grittiness decreasing with increasing cooking time.
Potential Impact:
GLUTENFREE will enable SMEs to produce premium gluten-free bakery products and pasta well accepted by the consumer. This will allow the SMEs to participate within this profitable market sector, and thus significantly improve their competitiveness. The new products have the potential to generate extra annual turnover and additional employment at the SMEs involved. Society will also benefit because celiac patients as well as healthy consumers will have access to a wider variety of tasty gluten-free foods, which will make the diet and life of celiac disease patients and their family members easier.
Main dissemination activities
The dissemination activities in the project were part of WP 11 and comprised four main tasks.
As a first step a project website (Task 11.1) was set-up at the beginning of the project. The website was established by the project coordinator and was updated regularly during the project term. It consists of a public part informing on the project background, objectives and all partners and also links to different celiac societies. Dates of project meetings, workshops and relevant conferences were announced in the News-section. An internal section was also established and used as a communication platform between the beneficiaries. All relevant project documents like deliverables, reports, presentations, meeting minutes and publications were filed on the internal platform and available for project partners and the project officers.
During the project period the RTD partners actively published the obtained knowledge (as far as compatible with the IP policy) in scientific, peer reviewed papers as part of Task 11.2 in order to make these results available to a broad public. In total eight articles were published in prestigious international periodicals. An article in Journal of Cereal Science presents results of the investigation of nutrient composition of seven commonly used commercial gluten-free flours (oat, rice, sorghum, maize, teff, buckwheat and quinoa) compared to wheat and wholemeal wheat flour (Hager et al., 2012a). In a publication in the journal entitled 'European Food Research and Technology', the bread making potential of seven gluten-free flours, wheat and wholemeal wheat flour was compared (Hager et al., 2012b). The study using systematic baking trials based on a central composite design to investigate the influence of HPMC and xanthan as well as their combination on gluten-free model systems was published in Food Hydrocolloids (Hager and Arendt, 2012c).
Results of the development of gluten-free pasta based on oat and teff flour are presented in a paper published in 'European Food Research and Technology' as well as in a paper submitted to Journal of Cereal Science (Hager et al. 2012d; Hager et al. 2012e). Two reviews on the applications of microbial fermentations for production of gluten-free products were published by Arendt et al. (2011) and Zannini et al (2012). Wolter et al. (2012) submitted a paper on the evaluation of the EPS producing strain Weisella cibaria MG1 as starter culture for the production of sourdough from buckwheat, oat, quinoa, teff and wheat flour to the Journal of Food Microbiology.
Other dissemination activities included posters (5), presentations at international scientific conferences (5) and several press releases in national and international media or marketing activities by the SMEs. A detailed summary of all dissemination activities is outlined in the Final plan for the use and dissemination of the knowledge and in section 4.2 in this final report. In addition, the website was used as dissemination tool and a dissemination workshop was organized at the end of the project.
Within Task 11.3 training courses for the SMEs were planned and carried out in order to apply the technological solutions realised in the project by means of specific training in the application of new ingredients and sourdough technique for the production of improved bakery and pasta products. Initially four training events were planned during the project term.
Three of the trainings were carried out as part of internal workshops as follows:
1) 'Sensory evaluation of existing market products', Thursday, 14th April 2010, Freising, Germany;
2) Gluten-free sourdough technology' Wednesday 12th October 2011, Minden, Germany; and
3) 'Technologies for improved gluten-free pasta' 8th May 2012, Fara San Martino, Italy.
The last training 'Technology for improved gluten-free breads' planned for December 2012 was cancelled due to scheduling difficulties by the bakery SMEs.
Initially two workshops were planned, one at mid-term and one at the end of the project. Due to organisational reasons the mid-term workshop was cancelled. Instead, an internal workshop on sourdough technology was organised at project partner Bocker along with the third project meeting in October 2011 in Minden, Germany. In the end of the project an industry workshop entitled 'Gluten-free Foods' was held on 14th of December 2012 at the University College Cork as main dissemination activity. Around 60 participants from industry, universities and other stakeholders from different European countries attended the workshop. Twelve speakers from academia and industry gave presentations and the programme covered all areas of gluten-free foods such as ingredients, texture, sensory aspects, and marketing as well as consumer research.
Exploitation results
The GLUTENFREE project generated a number of valuable results which can be exploited by the involved industry partners. The main results are described below. Regarding IP exploitation the consortium decided in the final project meeting that insufficient new IP was produced from the project to enable the development of a license package or to form the basis of a patent. All new knowledge will therefore be protected via 'confidentiality'. To date the involved companies intend to use these results individually as outlined in the Final Plan of the Use and Dissemination of the Knowledge. They did not agree on a common strategy yet but discussions are ongoing beyond the project.
List of exploitation results:
- Results on the consumer acceptability and consumer preferences of breads;
- Results on functional properties of proteins, fibres, hydrocolloids, their interactions and models to describe their application potential in gluten-free bread and pasta making;
- Protocols for the processing of new plant based protein ingredients with viscoelastic properties to be used as gluten replacers;
- Report on gluten-free sourdough technology: strains, optimal processing conditions;
- Knowledge on aroma optimisation by technological processes, fermentation and ingredient selection;
- Recipes for the production of new gluten-free breads;
- Protocols and recipes for the production of new gluten-free spaghetti;
- Application of gelling-fibre in gluten-free pasta;
- Prototypes of gluten-free oat bread;
- Toasting bread;
- Pizza base;
- Gluten-free spaghetti; and
- Pre-existing sourdough breads (two recipes of gluten-free sourdough breads, potato bread and linseed bread) improved in the project.
References
- Arendt, E.K. Moroni, A., and Zannini, E. 2011. Medical nutrition therapy: use of sourdough lactic acid bacteria as a cell factory for delivering functional biomolecules and food ingredients in gluten free bread. Microbial cell factories 10 Suppl 1:S15
- Hager, A-S., Wolter, A., Jacob, F., Zannini, E., Arendt, E.K. 2012a. Nutritional properties and ultra-structure of commercial gluten free flours from different botanical sources compared to wheat flours, Journal of Cereal Science 56:239-247
- Hager, A-S., Wolter, A., Czerny, M., Bez, J., Zannini, E., Arendt, E.K. Czerny, M. 2012b. Investigation of product quality, sensory profile and ultrastructure of breads made from a range of commercial gluten-free flours compared to their wheat counterparts, European Food Research and Technology, 235(2):333-344
- Hager, A.-S. Arendt, E.K. 2012c. Influence of hydroxypropylmethylcellulose (HPMC), xanthan gum and their combination on loaf specific volume, crumb hardness and crumb grain characteristics of gluten-free bread. Food Hydrocolloids, In Press.
- Hager, A.-S. Lauck, F., Zannini, E., Arendt, E.K. 2012d. Development of gluten-free fresh egg pasta based on oat and teff flour. European Food Research and Technology, 235(5):861-871
- Hager, A.-S. Zannini, E., Bez, J., Czerny, M., Arendt, E.K. 2012e. Sensory properties, microstructure and in vitro digestibility of egg pasta produced from oat, teff and wheat flour. Journal of Cereal Science, Submitted.
- Wolter, A., Hager, A.-S. Zannini, E., Galle, S., Ganzle, M., Arendt, E.K. 2012 Evaluation of the EPS producing strain Weisella cibaria MG1 as starter culture for the production of sourdough from buckwheat, oat, quinoa, teff and wheat flour a comparative study. Journal of Food Microbiology, Submitted.
- Zannini, E., Pontonio, E., Waters, D.M. and Arendt, E.K. 2012. Applications of microbial fermentations for production of gluten-free products and perspectives. Applied Microbiology and Biotechnology 93:473-48
List of Websites:
Adress of the project website:
http://www.glutenfree-project.eu/
Project Coordinator:
Jurgen Bez
Fraunhofer-Institut fur Verfahrenstechnik und Verpackung,
Giggenhauser Str. 35
85354 Freising
Germany
Phone: +49 8161 491-430
E-Mail: juergen.bez@ivv.fraunhofer.de
Project officer:
Raluca Iagher
REA - Research Executive Agency
16 Place Rogier
1210 Brussels
Belgium
Phone: +32 229 99146
E-Mail: Raluca.IAGHER@ec.europa.eu
