- Email: [email protected]
Antioxidants and shelf life of whole wheat bread
Journal of Cereal Science 53 (2011) 291e297
Contents lists available at ScienceDirect
Journal of Cereal Science journal homepage: www.elsevier.com/locate/jcs
Antioxidants and shelf life of whole wheat bread Sidsel Jensen a, c, Henrik Ostdal b, Leif H. Skibsted c, Anette K. Thybo a, * a
Department of Food Science, Faculty of Agricultural Sciences, Aarhus University, Kirstinebjergvej 10, DK-5792 Aarslev, Denmark Novozymes A/S, Krogshoejvej 36, DK-2880 Bagsvaerd, Denmark c Department of Food Science, Faculty of Life Sciences, University of Copenhagen, Rolighedsvej 30, DK-1958 Frederiksberg C, Denmark b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 May 2010 Received in revised form 21 January 2011 Accepted 24 January 2011
Industrially produced bread normally operates with a shelf life of several weeks at room temperature and indications of storage-related off-flavour development as a consequence of lipid oxidation have been suggested. The present study has tested enrichment of whole wheat bread with a-tocopherol or commercially used rosemary extracts for production of bread loaves with higher oxidative stability and hence better overall sensory quality. Bread quality was evaluated by sensory profiling, determination of antioxidative capacity, determination of lipid hydroperoxides as primary oxidation products, and analysis of volatile compounds including secondary lipid oxidation products. Enrichment of bread with atocopherol resulted in higher degrees of rancid aroma and flavour in fresh samples, which was explained by higher levels of secondary oxidation products, whereas enrichment of bread with rosemary extracts did not have any effect. Accordingly, application of antioxidants cannot, based on the current data, be recommended for achievement of bread with improved sensory properties during storage. Ó 2011 Elsevier Ltd. All rights reserved.
Keywords: Whole wheat bread Storage Antioxidants
1. Introduction Modern food production necessitates a certain shelf life and longer shelf life is often considered as a competitive advantage due to more flexible and centralised production, longer display time during retail, and increased convenience for the consumer. Extended shelf life is achieved by control of key reactions decisive for shelf life and therefore other reactions normally considered to be of less importance may become important in determining shelf life. Shelf life of bread products are normally limited by bread firming and microbial growth (Fernandez et al., 2006). Currently, bread producers can, by controlling these parameters, achieve products with a shelf life of up to 4 weeks (Sargent, 2008). Management of firming rate and microbial activity means that other parameters like flavour and aroma may become the limiting factor for shelf life of baked products (Holtekjoelen et al., 2008; Jensen et al., 2010, in press). The aroma and flavour of bread has long been acknowledged to influence the desirability of baked products and a vast number of studies employing sensory analysis and chemical characterisation
Abbreviations: ANOVA, Analysis of variance; DM, Dry matter; GCeMS, Gas chromatography mass spectroscopy; ORAC, Oxygen radical absorbance capacity; PC, Principal component; PLS, Partial least squares; POV, Peroxide value; TEAC, Trolox equivalent antioxidant capacity. * Corresponding author. Tel.: þ45 8999 3405; fax: þ45 8999 3495. E-mail address: [email protected] (A.K. Thybo). 0733-5210/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.jcs.2011.01.010
of volatiles have been performed (Chiavaro et al., 2008; Schieberle and Grosch, 1992; Seitz et al., 1998). Most of these studies concern flavour and aroma of fresh bread, and only very few studies touch upon changes of these parameters in bread products as a consequence of storage. Chiavaro et al. (2008) found a decrease in the total content of volatiles after 4 days of storage and Schieberle and Grosch (1992) detected a decrease in the concentration of selected volatiles after 8 days of storage, and concluded this decrease to be of utmost importance for the acceptability of the bread product in question. Significant changes in aroma, flavour, and taste of wheat bread and whole wheat bread during 3 weeks of storage were previously reported together with a build up of hexanal and heptanal (Jensen et al., accepted for publication). Formation of primary lipid oxidation products as precursors for these volatiles indicates that oxidative reactions probably have a high impact on the flavour and aroma profile of bread with extended shelf life (Jensen et al., 2011b). Effective control of lipid oxidation in food includes reduction in loss of antioxidants naturally present in the raw materials by modification in processing steps, elimination of metal contamination, addition of antioxidants, and implementation of appropriate packaging technologies (Fernandez et al., 2006; Frankel, 1996; Paradiso et al., 2008). Use of synthetic antioxidants in food products has for many years been an effective and relatively cheap way to achieve more oxidative stable products (Pokorny, 1991). However, synthetic antioxidants such as butylated hydroxyanisole and butylated hydroxytoluene may present a risk of carcinogenesis
292
S. Jensen et al. / Journal of Cereal Science 53 (2011) 291e297
which is why the interest in natural antioxidants has been intensified (Yanishlieva and Marinova, 1992; IARC, 1986a, 1986b). Apart from ascorbic acid normally considered being an antioxidant, but which in bread is used to achieve increased dough stability and bread volume (Grosch and Wieser, 1999), addition of antioxidants is not common for bread products. Enrichment of bread dough with antioxidants represents a strategy for production of bread loaves with a higher degree of oxidative stability and hence better overall sensory quality during storage, but only few studies concerning addition of antioxidants to bread products are available. Peng et al. (2010) studied the effect of grape seed extracts on the antioxidative capacity in bread made from refined wheat flour and found the oxidative status of the enriched bread to be greatly enhanced. These bread loaves were further evaluated to have a more desirable crust colour when compared with control bread prepared without the addition of grape seed extract. Other than the composition of the bread matrix, the effect of an antioxidative extract in food products depends on the type of antioxidants present in the extract and their solubility in the specific food matrix. Furthermore, the concentration of the individual antioxidants and their combined effect are important as well as the compatibility of the antioxidants with the processing and packaging methods employed. Benefits in bread production with one type of antioxidative extract is, due to the above mentioned factors, not necessarily valid for other antioxidative extracts or for other types of bread. The present study investigates the effect of adding naturally occurring antioxidants, a-tocopherol and two different rosemary extracts, to bread and monitoring the effect of these on bread quality during extended storage. a-Tocopherol is a lipophilic antioxidant naturally present in wheat, particularly in wheat bran (Zhou et al., 2004), while the antioxidative effect of rosemary extracts have been assigned to the high content of carnosol and carnosic acid (Aruoma et al., 1992). Bread quality was evaluated by sensory profiling, changes in antioxidative capacity, formation of primary oxidation products, and analysis of volatiles. 2. Materials and methods 2.1. Types of bread and preparation Whole wheat bread was made from stone milled fine whole wheat flour (min. 14% protein) and the ingredients listed in Table 1. All ingredients were obtained from commercial suppliers in EU except from stone milled fine whole wheat flour and azodicarbonamide, which was obtained from commercial suppliers in the US.
The bread was prepared according to a sponge and dough procedure and the four different bread varieties were prepared differing with respect to antioxidative additives: ‘a-tocopherol’, ‘fat soluble rosemary extract’, ‘water dispersible rosemary extracts’ and ‘no additive’. The fourth bread prepared without any antioxidative additive was included as a control sample. a-Tocopherol (>97) was purchased from SigmaeAldrich (SigmaeAldrich Inc., Steinheim, Germany) and fat soluble and water dispersible rosemary extracts (GUARDIANTM Rosemary Extract 201 and GUARDIANTM Rosemary Extract 202) were kindly donated by Danisco (Danisco A/S, Brabrand, Denmark). The rosemary extracts were practically flavourless and contained 96% natural rosemary extract, 4% phenolic diterpenes (compounds present in rosemary extract), together with propylene glycol, E1520. In addition, GUARDIANTM Rosemary Extract 202 also contained polyoxyethylene (20) sorbitan monooleate, E433. The list of ingredients for each of the bread varieties are given in Table 1. Novamyl 10.000 BG and Fungamyl Super MA were both from Novozymes (Novozymes A/S, Bagsvaerd, Denmark). Novamyl is a maltogenic amylase and Fungamyl Super is a blend of fungal a-amylase and xylanase. Microbial growth was prevented by addition of propanoic acid and by surface treatment with a 6% calcium sorbate solution of the baked bread. After surface treatment, the bread was further baked for 1 min to evaporate excess water from the sorbate treatment. Moisture loss during baking was not determined in the present baking experiment, but for this type of bread a moisture loss of 11e13% is normal. After cooling, the bread was packed in commercially used plastic bags (polyethylene, 25 mm, Multiline A/S, Soroe, Denmark) with high oxygen-permeability. Bread was prepared over a five week period providing bread stored for 0, 1, 2, 3, 4, and 5 weeks. All samples were analysed by sensory analysis and GC measurements at the same dates. Bread loaves were stored at room temperature (approximately 22 C) and samples stored for 4 and 5 weeks were only subjected to chemical analysis due to safety reasons (microbial status was not evaluated, but no visible mould growth was observed). For analysis of the overall antioxidative capacity and peroxide value, approximately 40 g of bread crumb was frozen rapidly by immersing samples directly into liquid nitrogen. The frozen bread pieces were ground and vacuum packed in non-transparent oxygen-impermeable aluminium foil bags (PETP12/ALU9/LLDPE75, Dansico Flexible, Horsens, Denmark) in sample portions of approximately 3 g. The foil bags were stored at 80 C until the dates of analyses. Samples were frozen on the same dates as the sensory analyses and the GCeMS measurements were carried out. All bread loaves were produced by the same baker to ensure uniformity in the baking procedure.
Table 1 Ingredients per kilo whole wheat flour [g]. Ingredients
[g/kg whole wheat flour]
Control
a-Tocopherol
Rosemary (lipophilic)
Rosemary (hydrophilic)
Water Salt Glucose syrup Cane syrup Soy oil Yeast sodium stearoyl-2-lactylate Ascorbic acid Calcium propionate Mono and di-glycerides Azodicarbonamide Novamyl 10.000 BG Fungamyl Super MA a-Tocopherol Rosemary extract (fat soluble) Rosemary extract (water dispersible)
629.5 20.0 80.0 40.0 20.0 50.0 5.0 0.06 2.5 10.0 0.04 0.25 0.04 1.0 1.0 1.0
x x x x x x x x x x x x x
x x x x x x x x x x x x x x
x x x x x x x x x x x x x
x x x x x x x x x x x x x
x x
S. Jensen et al. / Journal of Cereal Science 53 (2011) 291e297
293
2.2. Sensory analysis
2.4. Evaluation of the oxidative status
The samples of bread crumb used for sensory analysis consisted of bread sliced to a thickness of approximately 1.2 cm using a slicing machine (Gourmet Food Slicer, MAS 6108, Bosch, Stuttgart, Germany) and divided into quarters. The end pieces of each bread loaf, approximately 5 cm, were discarded. One serving consisted of two quarters of a bread slice, free of crust, put in a transparent plastic container and labelled with a three-digit number. The sensory evaluation was conducted by a sensory panel consisting of 8 experienced assessors (7 females/1 male, aged from 42 to 55). The sensory profiling was performed in a sensory evaluation laboratory, which fulfil requirements according to international standards (ASTM, 1986; ISO, 1988). The sensory evaluation started with a panel discussion, where the assessors agreed on a consensus list of attributes and on a definition of each attribute. Fifteen attributes were included in the list: 6 aroma related attributes (nasal perception), 6 flavour related attributes (oral and retro nasal perception), and 3 taste and mouth-feel related attributes (oral perception and physical sensation in the oral cavity). The attributes and matching descriptions are listed in Table 2. All assessors were experienced in evaluating plant based food products on a weekly basis, and most of the assessors had experience in evaluating whole wheat bread and therefore the training sessions were limited to four times 2 h sessions. After each training session, assessors received feedback on their performance with the aim of improving and standardising the panel’s discriminative power. For each of the four bread varieties (control, addition of atocopherol, addition of fat soluble rosemary extract or addition of water dispersible rosemary extract) the assessors evaluated bread stored for: ‘0 weeks’, ‘1 week’, ‘2 weeks’, and ‘3 weeks’ in triplicates over a period of three days giving a total of 16 samples per test day. The assessors evaluated the samples at individual speed on an unstructured 15.0 cm line scale anchored on the left by ‘low’ and on the right with the term ‘high’. Ratings were registered directly on a computerised registration system (Fizz software, 2.30C, Biosystemes, Couternon, France).
The overall antioxidative capacity analysed by the TEAC and ORAC assay together with analysis of lipid hydroperoxides (POV) were performed as described by Jensen et al. (2011b).
2.3. Analysis of volatile compounds Head space collection of volatile compounds from whole wheat bread crumb was conducted and analysed by GCeMS as described by Jensen et al. in press. Verification of compounds suggested by the MS database (NIST, 1998) was made by comparison of the relative retention indices and MS of authentic reference compounds. Authentic compounds were supplied by SigmaeAldrich (Sigmae Aldrich Inc., Steinheim, Germany). Table 2 Sensory attributes and descriptions. Category
Attribute
Description
Aroma
Dough Bran Rancid Acid Spice Dust Dough Sweet aromatic Rancid Dust Bran Spice Bitter Prickly Astringency
Freshly made wheat dough Bran, cereal-like Rancid linseed Propanoic acid Mild spicy, roasted Dry Freshly made wheat dough Toffee, malt-like Rancid linseed Dry Bran, cereal-like Mild spicy, roasted, cocoa Bitter Prickly sensation on the tip of the tongue Pungent feeling on the side of the tongue and pain-like feeling in the jaw
Flavour
Taste and Mouth-Feel
2.5. Statistics Data were analysed for significant effects of type of antioxidant added and storage time using univariate analysis of variance (ANOVA) (SAS Institute, version 9.1, Cary, NC). For the sensory data analysis, the interaction between sample and assessor was used as random effect, and thereby, level and range differences between individual assessors were modelled (Brockhoff and Skovgaard, 1994). Statistical significance was defined at P < 0.05. Principal component analysis (PCA) was used to interpret the structure in the large amount of volatile compounds while partial least squares (PLS) regression was used to study the prediction of sensory data (Y-data) from chemical measurements (X-data). Uncertainty testing by a jack-knife procedure was applied to identify variables of highest importance in the predictive model (Martens and Martens, 2000). Multivariate data analyses were performed using Unscrambler (version 9.6, CAMO Software A/S, Oslo, Norway) and LatentiX (version 2.00, Latent5 Aps, Copenhagen, Denmark). Full cross validation was used as validation criterion and all data were auto-scaled. 3. Results and discussion 3.1. Sensory analysis Results from the univariate ANOVA showed a significant effect of assessors for all tested attributes, which is a common observation in sensory evaluation and can be ascribed to level differences in the scaling (Brockhoff and Skovgaard, 1994; Naes and Langsrud, 1998) (data not shown). The panel was able to distinguish between samples according to storage time and less pronounced between types of antioxidant added. Main effect of storage time was obtained for 80% of the attributes, emphasising that the sensory quality of bread is highly dependent on storage time and that aroma, flavour, and taste are in fact important characteristics when shelf life of bread is to be determined. Fig. 1 illustrates sensory mean data across replicates and assessors for each bread variety. It is obvious that storage affects the sensory profiles and that especially the sensory profile of non-stored (0 weeks) bread displays a huge difference compared to bread stored for 1, 2, and 3 weeks. A significant effect of antioxidant addition was displayed for the aroma attributes: ‘dough’, ‘bran’, and ‘rancid’, and for rancid flavour. Especially bread enriched with a-tocopherol differed from the remaining samples (Fig. 1) as these samples were generally evaluated to have a lower intensity of dough, bran, and dust aroma and a higher intensity of rancid aroma and flavour. a-Tocopherol has been found to enhance the sensory and chemical stability in various food products (Paradiso et al., 2008; Warner and Moser, 2009). In a study concerning the stability of vitamin E in wheat flour, wheat flour with high vitamin E content was found to delay the formation of hexanal compared to flour with a lower content of vitamin E (Nielsen and Hansen, 2008). Hexanal is commonly used as an indicator of lipid oxidation as it contributes to the rancid flavour. The antioxidant/pro-oxidant nature of a-tocopherol depends, however, on the concentration, and a pro-oxidative effect has been observed especially when a-tocopherol has been added in relatively high concentrations (Huang et al., 1994). The rancid flavour of the present bread enriched with a-tocopherol suggests a prooxidative effect of a-tocopherol and might be a result of the
294
S. Jensen et al. / Journal of Cereal Science 53 (2011) 291e297
Fig. 1. Spider plot of sensory scores (0 (low intensity)-15 (high intensity)) of control bread and bread enriched with a-tocopherol, and with a fat soluble (L) and water dispersible (H) rosemary extract, respectively. Means are calculated across replicates and assessors. A_: Aroma attributes; F_: Flavour attributes.
relatively high concentration of a-tocopherol (1 g kg1 flour). This conclusion is supported by analyses performed by Park et al. (2002) and Ochi et al. (1993), who observed a pro-oxidative effect of tocopherols when added in concentrations of 100e200 mg kg1 to heat treated cereal products such as cookies and rice snacks. 3.2. Analysis of volatiles Thirty five different volatiles were identified and verified in the present study. The concentrations of all verified volatile compounds for the four bread varieties are given in Table 3. The volatile compounds can be divided into seven classes: ‘acids’, ‘alkanes’, alcohols’, ‘aldehydes’, ‘esters’, ‘furans’, ‘ketones’, and ‘terpenes’. Statistical analysis showed significant effect of the interaction between storage and bread variety for the total content of acids (P ¼ 0.048). Furthermore, a significant main effect of storage was found for the decreasing content of alcohols, aldehydes, alkanes, furans, and terpenes (P < 0.0001). In addition, aldehydes were found to be significantly affected by bread variety (P ¼ 0.041). Higher concentrations of aldehydes were found in fresh samples of bread enriched with a-tocopherol compared to the other bread varieties. Especially, the secondary oxidation products hexanal (V12) and heptanal (V15) were higher in fresh samples of these bread loaves indicating a higher degree of lipid oxidation already during baking of bread enriched with a-tocopherol.
A multivariate interpretation of the volatile components (Fig. 2) confirmed the results from the ANOVA as it is evident that bread samples could be separated according to effect of storage time along PC1 (PC1 explains 51% of variance) and less pronounced according to effect of antioxidative additives along PC2 (PC2 explains 10% of variance). This differentiation of samples in accordance with storage was in good agreement with results from the sensory analysis. Fresh samples of bread are positioned in the opposite direction of PC1 compared to the more stored samples, with a few exceptions (a-tocopherol stored for 1 and 2 weeks). A differentiation of bread enriched with a-tocopherol and bread enriched with rosemary extracts are seen along PC2. No obvious differences in volatile profiles between fat soluble and water dispersible rosemary extracts were found. The efficiency of fat soluble and water dispersible antioxidants is known to differ significantly when tested in similar food systems (Frankel, 1996) and a differentiation of samples according to type of rosemary extract could therefore be expected. However, the present data showed no difference between the two tested rosemary extracts which could be explained by the active ingredients in the rosemary extracts being the same, and that these ingredients generally did not have an effect on the chemical stability of bread. The difference in solubility of the two extracts is exclusively obtained by addition of the emulsifier polyoxyethylene (20) sorbitan monooleate, E433, to the water dispersible rosemary extract.
S. Jensen et al. / Journal of Cereal Science 53 (2011) 291e297
295
Table 3 Maximum and minimum values for the concentration [ng g1 of bread] of verified volatile compounds in control bread and bread enriched with a-tocopherol and a fat soluble (L) and water dispersible (H) rosemary extracts. No Acid
Alkane and alcohol
Aldehyde
Ester
Furan
Ketone
Terpene
28 31 32 33 1 8 10 13 17 24 25 35 2 6 7 12 15 26 30 4 19 23 27 3 18 20 29 34 5 9 11 14 21 22 16
Volatiles Acetic acid Propanoic acid 2-Methylpropanoic acid Butanoic acid Octane Ethanol 1-Propanol 2-Methyl-1-propanol 3-Methyl-1-butanol Hexanol 3-Ethoxy-1-propanol Phenylethanol 2-Methylpropanal 2-Methylbutanal 3-Methylbutanal Hexanal Heptanal Nonanal Benzaldehyde Ethylacetate Ethylhexanoate Ethyllactate Ethyloctanoate 2-Methylfuran 2-Pentylfuran 2-Methyltetrahydro3furanone Furfural Furfuryl alcohol 2-Butanone 2,3-Butandione 2,3-Pentandione 2-Heptanone 3-Hydroxy-2-butanone 1-Hydroxy-2-propanone Limonene
Control
a-Tocopherol
Rosemary (L)
Rosemary (H)
0.129e0.336 25.618e43.362 2.039e3.087 0.145e0.211 0.059e0.571 731.713e1352.534 4.011e6.748 93.071e180.345 28.235e77.777 1.006e2.114 0.212e0.349 1.904e2.192 0.196e2.100 0.156e3.617 0.698e11.597 1.845e3.867 0.271e1.050 0.337e0.518 0.746e1.013 1.274e3.809 0.210e0.502 0.166e0.608 0.418e0.896 0.107e0.204 0.158e0.853 0.090e0.188 0.330e1.255 0.187e0.613 0.414e0.778 4.95e6.291 0.137e0.698 0.177e0.235 16.596e27.173 0.763e1.823 0.048e0.145
Min-Max 0.150e0.263 25.366e39.995 2.127e2.997 0.118e0.196 0.078e0.691 618.571e1406.702 3.995e6.777 95.596e180.886 27.892e77.834 0.856e2.141 0.218e0.296 1.627e2.089 0.237e2.582 0.187e5.954 1.055e14.375 2.690e4.758 0.318e1.157 0.138e0.505 0.601e1.099 1.126e2.141 0.375e0.656 0.312e0.662 0.376e0.708 0.105e0.281 0.175e0.955 0.099e0.247 0.267e1.593 0.240e0.471 0.343e0.760 3.637e7.481 0.067e0.898 0.242e0.390 23.925e28.379 0.577e1.494 0.050e0.173
0.118e0.225 28.737e37.492 2.134e2.762 0.176e0.292 0.060e0.541 654.456e1508.114 4.055e6.930 93.648e179.107 28.529e73.020 0.917e1.974 0.213e0.284 1.704e2.260 0.184e1.657 0.155e2.920 0.781e8.213 1.616e3.747 0.229e0.499 0.277e0.680 0.678e0.956 1.250e2.512 0.253e0.468 0.151e0.667 0.440e0.777 0.073e0.261 0.146e0.726 0.082e0.156 0.287e0.925 0.217e0.470 0.366e0.727 4.470e5.706 0.123e0.659 0.096e0.559 21.732e25.685 2.395e3.576 0.044e0.174
0.130e0.234 27.112e40.623 1.995e2.851 0.164e0.216 0.044e0.426 592.502e1491.608 3.726e7.331 70.941e204.239 22.172e84.544 0.669e2.163 0.206e0.283 1.654e2.152 0.188e1.478 0.164e3.071 0.900e8.729 1.555e4.006 0.192e1.062 0.310e0.523 0.558e0.963 1.228e2.863 0.248e0.548 0.296e0.610 0.340e0.722 0.050e0.129 0.141e0.757 0.074e0.144 0.210e0.857 0.182e0.431 0.282e0.646 4.357e7.244 0.108e0.706 0.136e0.264 21.569e25.945 1.155e1.947 0.070e0.191
Sign. variety
**
*** * *
*
*
***
Fig. 2. Principal component analysis bi-plot on volatile compounds in whole wheat bread (control) and whole wheat bread enriched with a-tocopherol and fat soluble (L) and water dispersible (H) rosemary extracts. Volatiles (V) are designated by the number of the individual compound given in Table 3. The digits at the end of the sample name states the time of storage [0e5 weeks].
296
S. Jensen et al. / Journal of Cereal Science 53 (2011) 291e297
A substantial part of the volatiles were associated with the fresh samples of bread, independent of antioxidant added. Furfuryl alcohol (V34) and butanoic acid (V33) increased in concentration during storage while all other volatiles decreased as a consequence of storage. Furfuryl alcohol was found to increase in whole wheat bread crumb and decrease in whole wheat bread crust during a 3 week period of storage in a previous study conducted by the present authors (Jensen et al., accepted for publication), suggesting that furfuryl alcohol migrates from crust to crumb during storage. 3.3. Oxidative status The measurement of lipid hydroperoxides (POV) was performed to follow the progress in lipid oxidation. No changes in POV was observed during storage, but a significant difference (P < 0.0001) was observed for the type of antioxidant added. A substantially higher level of hydroperoxides was present in bread enriched with a-tocopherol compared to the three other bread varieties (0.7e0.9 mE H2O2 g1 of bread enriched with a-tocopherol, based on dry matter content (DM) compared to 0.0e0.2 mE H2O2 g1 of bread for control bread and bread enriched with rosemary extracts, DM). Similar results were obtained in a study performed by Terao and Matsushita (1986), who found an increase in the concentration of hydroperoxides for edible oil enriched with a-tocopherol. The elevated level of hydroperoxides can be explained by breakdown of lipid hydroperoxides induced by reactive species of a-tocopherol (a-tocopherol radical). The accumulation of hydroperoxides observed in the present study right after baking can thus be explained by an increase in the concentration of a-tocopheroxyl as a result of the elevated temperatures during the baking process. Comparison of Fig. 1 and Table 3 shows that the increased level of lipid hydroperoxides in bread with added a-tocopherol was also accompanied by increased rancidity score in the sensory evaluation. A study by Haakansson and Jägerstad (1990) showed that vitamin E loss in whole wheat flours started immediately after flour and water were mixed due to lipoxygenase activity and the loss increased as the temperature of the flour-water slurry increased. Another study determined the content of tocopherols to decrease by one third during baking (Ranhotra et al., 2000). Vitamin E loss is equal to formation of a-tocopheroxyl radicals and a-tocopheroxyl
radicals have been shown to act as pro-oxidants as mentioned above. The high concentration of lipid hydroperoxides in bread enriched with a-tocopherol is accordingly suggested to be a result of a pro-oxidative activity of oxidized tocopherols during bread making. The overall antioxidative capacity was evaluated by the TEAC assay and the ORAC assay. The antioxidative capacity in bread samples when measured by ORAC assay was between 17.4 and 19.8 mg trolox equivalents g1 crumb, DM. This was significantly higher than the level measured by TEAC, which showed the antioxidative capacity to be between 4.8 and 6.8 mg trolox equivalents g1 crumb, DM. Both assays exhibited significance for the interaction between storage time and bread variety, emphasising that the development in overall antioxidative capacity during storage depended on bread variety. 3.4. Correlation of sensory and chemical results To study the correlation between sensory and chemical (GCeMS, POV, ABTS, and ORAC) data a multivariate PLS regression, based on prediction of the sensory data (Y-variables) from the chemical data (X-variables) was performed (Fig. 3). The regression plot shows that 1) most of the aroma attributes are spanned by PC1 together with a large amount of volatile compounds, 2) a few flavour attributes are spanned by PC2 together with fewer volatile compounds, and 3) another large group of attributes related to flavour, taste, and mouth-feel are located near origio and are thus not explained by any of the volatile compounds. For PC1, the dough aroma, mainly presented in the fresh bread samples, was positively correlated with a large number of volatiles like ethanol (V8), 1-propanol (V10), 2,3-pentadione (V11), 2-methyl-1-propanol (V13), limonene (V16), 3-methyl-1-butanol (V17), hexanol (V24), and ethyloctanoate (V27) and inversely correlated to furfuryl alcohol (V34) which, on the contrary, correlated with the aroma attributes: ‘rancid’, ‘spicy’, and ‘aromatic sweet’. This means that most of the sensory aroma attributes are correlated to volatile compounds, and that volatility has a great impact on the perceived changes of bread aroma. In PC2, a positive correlation was found between two flavour attributes: ‘bran’ and ‘dough’, mainly found in samples added rosemary extract, and
Fig. 3. Partial least squares (PLS) regression with only significant important variables of 4 different bread varieties (Control bread, bread enriched with a-tocopherol, fat soluble (L) and water dispersible (H) rosemary extracts) and with 4 different storage times. Score plot (a) and loading plot (b) with chemical data as X-variables and sensory data as Y-variables. Sensory aroma attributes (normal), flavour attributes (italic), and taste and mouth-feel (underlined). The digit in sample name indicates storage time [0e3 weeks].
S. Jensen et al. / Journal of Cereal Science 53 (2011) 291e297
ethyloctanoate (V27), which was inversely correlated to ‘rancid’ flavour. The remaining flavour, taste, and mouth-feel attributes being: ‘prickly’, ‘bitter’, ‘spicy’, ‘astringent’, ‘acid’, and ‘dust’ are less correlated to any of the aroma compounds and the TEAC. This indicates that the volatile compounds are not the main contributors to the perceived flavour, taste, and mouth-feel of bread and probably non-volatile compounds are responsible for changes of these attributes. 4. Conclusion The present study showed that the tested antioxidative additives did not improve the sensory quality and stability of bread. Changes in the sensory profile and in the content of volatile compounds in bread enriched with a-tocopherol or rosemary extracts were, like for the control bread, mainly ascribed to the effect of storage time whereas the effect of bread enrichment was less pronounced. Most volatile compounds were present in higher concentrations in the fresh samples of bread, emphasising that volatility is an important contributor to changes, especially changes related to aroma, during storage. Fresh samples of bread enriched with a-tocopherol had higher concentrations of hydroperoxides and secondary lipid oxidation products. These samples shared some of the same sensory characteristics as stored control samples indicting that lipid oxidation is responsible for these less favourable sensory notes like rancid aroma and flavour, bitter taste, and astringency. The analysis of the antioxidative capacity showed that the relatively small changes in antioxidative capacity during storage depended on the type of additive included in the bread recipe. Acknowledgements The authors gratefully acknowledge the excellent technical assistance from Nina Eggers and Birgitte Foged, the great help from Lone R. Borum in the sensory lab as well as the work performed by Bo B. Madsen, who baked the large number of breads forming the basis of the present work. The investigation is a part of a Ph.D. study supported financially by Novozymes, University of Copenhagen, and the Research School FOOD Denmark. References Aruoma, O.I., Halliwell, B., Aeschbach, R., Löligers, J., 1992. Antioxidant and prooxidant properties of active rosemary constituents: carnosol and carnosic acid. Xenobiotica 22, 257e268. ASTM, 1986. Physical Requirements, Guidelines for Sensory Evaluation Laboratories, STP 913. American Society for Testing and Materials, PA. Brockhoff, P.M., Skovgaard, I.M., 1994. Modelling individual differences between assessors in sensory evaluations. Food Quality and Preferences 5, 215e224. Chiavaro, E., Vittadini, E., Musci, M., Bianchi, F., Curti, E., 2008. Shelf-life stability of artisanally and industrially produced durum wheat sourdough bread (‘‘Altamura bread’’). LWT 41, 58e70. Fernandez, U., Vodovotz, Y., Courtney, P., Pascall, M.A., 2006. Extended shelf life of soy bread using modified atmosphere packaging. Journal of Food Protection 69, 693e698. Frankel, E.N., 1996. Antioxidants in lipid foods and their impact on food quality. Food Chemistry 57, 51e55.
297
Grosch, W., Wieser, H., 1999. Redox reactions in wheat dough as affected by ascorbic acid. Journal of Cereal Science 29, 1e16. Haakansson, B., Jägerstad, M., 1990. The effect of thermal inactivation of lipoxygenase on the stability of vitamin E in wheat. Journal of Cereal Science 12, 177e185. Holtekjoelen, A.K., Baevre, A.B., Roedbotten, M., Berg, H., Knutsen, S.H., 2008. Antioxidant properties and sensory profiles of breads containing barley flour. Food Chemistry 110, 414e421. Huang, S.-W., Frankel, E.N., German, J.B., 1994. Antioxidant activity of a- and g-tocopherols in bulk oils and in oil-in-water emulsions. Journal of Agricultural and Food Chemistry 42, 2108e2114. IARC, 1986a. Butylated hydroxyanisole (BHA). IARC Monographs on the evaluation of the Carcinogenic risks of chemicals to Humans. Some naturally occurring and synthetic food components. Furocoumarins and Ultraviolet Radiation 40, 123e159. IARC, 1986b. Butylated hydroxytoluene (BHT). IARC Monographs on the evaluation of the Carcinogenic risk of chemicals to Humans., some naturally occurring and synthetic food components. Furocoumarins and Ultraviolet Radiation 40, 161e206. ISO, 1988. International Standard 8589, Sensory Analysis - General Guidance for the Design of Test Rooms. Ref. no. ISO 8589:1988 (E). International Organization for Standardization, Genéve, Switzerland. Jensen, S., Oestdal, H., Thybo, A.K., 2010. Sensory profiling of changes in wheat and whole wheat bread during a prolonged period of storage. Journal of Sensory Studies 25, 231e245. Jensen, S., Oestdal, H., Skibsted, L. H., Larsen, E., Thybo, A. K., Chemical changes in wheat pan bread during storage and how it affects the sensory perception of aroma, flavor, and taste. Journal of Cereal Science, in press. Jensen, S., Oestdal, H., Clausen, M.R., Andersen, M., Skibsted, L.H., 2011b. Oxidative stability in whole wheat bread during storage. LTW e Food Science and Technology 44, 637e642. Martens, H., Martens, M., 2000. Modified Jack-knife estimation of parameter uncertainty in bilinear modelling by partial least squares regression (PLSR). Food Quality and Preference 11, 5e16. Naes, T., Langsrud, O., 1998. Fixed or random assessors in sensory profiling? Food Quality and Preferences 9, 145e152. Nielsen, M.M., Hansen, Aa, 2008. Stability of vitamin e in wheat flour and whole wheat flour during storage. Cereal Chemistry 85, 716e720. NIST, 1998. Mass Spectral Database, Standard Reference Database 1. National Institute of Standards and Technology, Gaithersburg, Maryland, US. Ochi, T., Tsuchiya, K., Ohtsuka, Y., Aoyama, M., Maruyama, T., Niiya, I., 1993. Synergistic antioxidant effects of organic acids and their derivatives with tocopherols on cookies. Nippon Shokuhin Kogyo Gakkaishi 40, 393e399. Paradiso, V.M., Summo, C., Trani, A., Caponio, F., 2008. An effort to improve the shelf life of breakfast cereals using natural mixed tocopherols. Journal of Cereal Science 47, 322e330. Park, Y.S., Kim, Y.S., Shin, D.H., 2002. Antioxidative effects of ethanol extracts from Rhus verniciflua Stoke on Yukwa (Oil popped rice snack) base during storage. Journal of Food Science 67, 2474e2479. Peng, X., Ma, J., Cheng, K.-W., Jiang, Y., Chen, F., Wang, M., 2010. The effects of grape seed extract fortification on the antioxidant activity and quality attributes of bread. Food Chemistry 119, 49e53. Pokorny, J., 1991. Natural antioxidants for food use. Trends in Food Science & Technology 2, 223e227. Ranhotra, G.S., Gelroth, J.A., Okot-Kotber, B.M., 2000. Stability and dietary contribution of vitamin E added to bread. Cereal Chemistry 77, 159e162. Sargent, K., 2008. A softer“ approach to improving the quality of refrigerated bakery products. Cereal Foods World 53, 301e305. Schieberle, P., Grosch, W., 1992. Changes in the concentration of potent odorants during storage of white bread. Flavour and Fragrance Journal 7, 213e218. Seitz, L.M., Chung, O.K., Rengarajan, R., 1998. Volatiles in selected commercial breads. Cereal Chemistry 75, 847e853. Terao, J., Matsushita, S., 1986. The peroxidizing effect of a-tocopherol on autoxidation of methyl linoleate in bulk phase. Lipids 21, 255e260. Warner, K., Moser, J., 2009. Frying stability of purified mid-oleic sunflower oil triacylglycerols with added pure tocopherols and tocopherol mixtures. Journal of the American Oil Chemists’ Society 86, 1199e1207. Yanishlieva, N.V., Marinova, E.M., 1992. Inhibited oxidation of lipids. 1. Complex estimation and comparison of the antioxidative properties of some natural and synthetic antioxidants. Fett Wissenschaft Technologie 94, 374e379. Zhou, K., Su, L., Yu, L.L., 2004. Phytochemicals and antioxidant properties in wheat bran. Journal of Agri-cultural and Food Chemistry 52, 6108e6114.
Contents lists available at ScienceDirect
Journal of Cereal Science journal homepage: www.elsevier.com/locate/jcs
Antioxidants and shelf life of whole wheat bread Sidsel Jensen a, c, Henrik Ostdal b, Leif H. Skibsted c, Anette K. Thybo a, * a
Department of Food Science, Faculty of Agricultural Sciences, Aarhus University, Kirstinebjergvej 10, DK-5792 Aarslev, Denmark Novozymes A/S, Krogshoejvej 36, DK-2880 Bagsvaerd, Denmark c Department of Food Science, Faculty of Life Sciences, University of Copenhagen, Rolighedsvej 30, DK-1958 Frederiksberg C, Denmark b
a r t i c l e i n f o
a b s t r a c t
Article history: Received 25 May 2010 Received in revised form 21 January 2011 Accepted 24 January 2011
Industrially produced bread normally operates with a shelf life of several weeks at room temperature and indications of storage-related off-flavour development as a consequence of lipid oxidation have been suggested. The present study has tested enrichment of whole wheat bread with a-tocopherol or commercially used rosemary extracts for production of bread loaves with higher oxidative stability and hence better overall sensory quality. Bread quality was evaluated by sensory profiling, determination of antioxidative capacity, determination of lipid hydroperoxides as primary oxidation products, and analysis of volatile compounds including secondary lipid oxidation products. Enrichment of bread with atocopherol resulted in higher degrees of rancid aroma and flavour in fresh samples, which was explained by higher levels of secondary oxidation products, whereas enrichment of bread with rosemary extracts did not have any effect. Accordingly, application of antioxidants cannot, based on the current data, be recommended for achievement of bread with improved sensory properties during storage. Ó 2011 Elsevier Ltd. All rights reserved.
Keywords: Whole wheat bread Storage Antioxidants
1. Introduction Modern food production necessitates a certain shelf life and longer shelf life is often considered as a competitive advantage due to more flexible and centralised production, longer display time during retail, and increased convenience for the consumer. Extended shelf life is achieved by control of key reactions decisive for shelf life and therefore other reactions normally considered to be of less importance may become important in determining shelf life. Shelf life of bread products are normally limited by bread firming and microbial growth (Fernandez et al., 2006). Currently, bread producers can, by controlling these parameters, achieve products with a shelf life of up to 4 weeks (Sargent, 2008). Management of firming rate and microbial activity means that other parameters like flavour and aroma may become the limiting factor for shelf life of baked products (Holtekjoelen et al., 2008; Jensen et al., 2010, in press). The aroma and flavour of bread has long been acknowledged to influence the desirability of baked products and a vast number of studies employing sensory analysis and chemical characterisation
Abbreviations: ANOVA, Analysis of variance; DM, Dry matter; GCeMS, Gas chromatography mass spectroscopy; ORAC, Oxygen radical absorbance capacity; PC, Principal component; PLS, Partial least squares; POV, Peroxide value; TEAC, Trolox equivalent antioxidant capacity. * Corresponding author. Tel.: þ45 8999 3405; fax: þ45 8999 3495. E-mail address: [email protected] (A.K. Thybo). 0733-5210/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.jcs.2011.01.010
of volatiles have been performed (Chiavaro et al., 2008; Schieberle and Grosch, 1992; Seitz et al., 1998). Most of these studies concern flavour and aroma of fresh bread, and only very few studies touch upon changes of these parameters in bread products as a consequence of storage. Chiavaro et al. (2008) found a decrease in the total content of volatiles after 4 days of storage and Schieberle and Grosch (1992) detected a decrease in the concentration of selected volatiles after 8 days of storage, and concluded this decrease to be of utmost importance for the acceptability of the bread product in question. Significant changes in aroma, flavour, and taste of wheat bread and whole wheat bread during 3 weeks of storage were previously reported together with a build up of hexanal and heptanal (Jensen et al., accepted for publication). Formation of primary lipid oxidation products as precursors for these volatiles indicates that oxidative reactions probably have a high impact on the flavour and aroma profile of bread with extended shelf life (Jensen et al., 2011b). Effective control of lipid oxidation in food includes reduction in loss of antioxidants naturally present in the raw materials by modification in processing steps, elimination of metal contamination, addition of antioxidants, and implementation of appropriate packaging technologies (Fernandez et al., 2006; Frankel, 1996; Paradiso et al., 2008). Use of synthetic antioxidants in food products has for many years been an effective and relatively cheap way to achieve more oxidative stable products (Pokorny, 1991). However, synthetic antioxidants such as butylated hydroxyanisole and butylated hydroxytoluene may present a risk of carcinogenesis
292
S. Jensen et al. / Journal of Cereal Science 53 (2011) 291e297
which is why the interest in natural antioxidants has been intensified (Yanishlieva and Marinova, 1992; IARC, 1986a, 1986b). Apart from ascorbic acid normally considered being an antioxidant, but which in bread is used to achieve increased dough stability and bread volume (Grosch and Wieser, 1999), addition of antioxidants is not common for bread products. Enrichment of bread dough with antioxidants represents a strategy for production of bread loaves with a higher degree of oxidative stability and hence better overall sensory quality during storage, but only few studies concerning addition of antioxidants to bread products are available. Peng et al. (2010) studied the effect of grape seed extracts on the antioxidative capacity in bread made from refined wheat flour and found the oxidative status of the enriched bread to be greatly enhanced. These bread loaves were further evaluated to have a more desirable crust colour when compared with control bread prepared without the addition of grape seed extract. Other than the composition of the bread matrix, the effect of an antioxidative extract in food products depends on the type of antioxidants present in the extract and their solubility in the specific food matrix. Furthermore, the concentration of the individual antioxidants and their combined effect are important as well as the compatibility of the antioxidants with the processing and packaging methods employed. Benefits in bread production with one type of antioxidative extract is, due to the above mentioned factors, not necessarily valid for other antioxidative extracts or for other types of bread. The present study investigates the effect of adding naturally occurring antioxidants, a-tocopherol and two different rosemary extracts, to bread and monitoring the effect of these on bread quality during extended storage. a-Tocopherol is a lipophilic antioxidant naturally present in wheat, particularly in wheat bran (Zhou et al., 2004), while the antioxidative effect of rosemary extracts have been assigned to the high content of carnosol and carnosic acid (Aruoma et al., 1992). Bread quality was evaluated by sensory profiling, changes in antioxidative capacity, formation of primary oxidation products, and analysis of volatiles. 2. Materials and methods 2.1. Types of bread and preparation Whole wheat bread was made from stone milled fine whole wheat flour (min. 14% protein) and the ingredients listed in Table 1. All ingredients were obtained from commercial suppliers in EU except from stone milled fine whole wheat flour and azodicarbonamide, which was obtained from commercial suppliers in the US.
The bread was prepared according to a sponge and dough procedure and the four different bread varieties were prepared differing with respect to antioxidative additives: ‘a-tocopherol’, ‘fat soluble rosemary extract’, ‘water dispersible rosemary extracts’ and ‘no additive’. The fourth bread prepared without any antioxidative additive was included as a control sample. a-Tocopherol (>97) was purchased from SigmaeAldrich (SigmaeAldrich Inc., Steinheim, Germany) and fat soluble and water dispersible rosemary extracts (GUARDIANTM Rosemary Extract 201 and GUARDIANTM Rosemary Extract 202) were kindly donated by Danisco (Danisco A/S, Brabrand, Denmark). The rosemary extracts were practically flavourless and contained 96% natural rosemary extract, 4% phenolic diterpenes (compounds present in rosemary extract), together with propylene glycol, E1520. In addition, GUARDIANTM Rosemary Extract 202 also contained polyoxyethylene (20) sorbitan monooleate, E433. The list of ingredients for each of the bread varieties are given in Table 1. Novamyl 10.000 BG and Fungamyl Super MA were both from Novozymes (Novozymes A/S, Bagsvaerd, Denmark). Novamyl is a maltogenic amylase and Fungamyl Super is a blend of fungal a-amylase and xylanase. Microbial growth was prevented by addition of propanoic acid and by surface treatment with a 6% calcium sorbate solution of the baked bread. After surface treatment, the bread was further baked for 1 min to evaporate excess water from the sorbate treatment. Moisture loss during baking was not determined in the present baking experiment, but for this type of bread a moisture loss of 11e13% is normal. After cooling, the bread was packed in commercially used plastic bags (polyethylene, 25 mm, Multiline A/S, Soroe, Denmark) with high oxygen-permeability. Bread was prepared over a five week period providing bread stored for 0, 1, 2, 3, 4, and 5 weeks. All samples were analysed by sensory analysis and GC measurements at the same dates. Bread loaves were stored at room temperature (approximately 22 C) and samples stored for 4 and 5 weeks were only subjected to chemical analysis due to safety reasons (microbial status was not evaluated, but no visible mould growth was observed). For analysis of the overall antioxidative capacity and peroxide value, approximately 40 g of bread crumb was frozen rapidly by immersing samples directly into liquid nitrogen. The frozen bread pieces were ground and vacuum packed in non-transparent oxygen-impermeable aluminium foil bags (PETP12/ALU9/LLDPE75, Dansico Flexible, Horsens, Denmark) in sample portions of approximately 3 g. The foil bags were stored at 80 C until the dates of analyses. Samples were frozen on the same dates as the sensory analyses and the GCeMS measurements were carried out. All bread loaves were produced by the same baker to ensure uniformity in the baking procedure.
Table 1 Ingredients per kilo whole wheat flour [g]. Ingredients
[g/kg whole wheat flour]
Control
a-Tocopherol
Rosemary (lipophilic)
Rosemary (hydrophilic)
Water Salt Glucose syrup Cane syrup Soy oil Yeast sodium stearoyl-2-lactylate Ascorbic acid Calcium propionate Mono and di-glycerides Azodicarbonamide Novamyl 10.000 BG Fungamyl Super MA a-Tocopherol Rosemary extract (fat soluble) Rosemary extract (water dispersible)
629.5 20.0 80.0 40.0 20.0 50.0 5.0 0.06 2.5 10.0 0.04 0.25 0.04 1.0 1.0 1.0
x x x x x x x x x x x x x
x x x x x x x x x x x x x x
x x x x x x x x x x x x x
x x x x x x x x x x x x x
x x
S. Jensen et al. / Journal of Cereal Science 53 (2011) 291e297
293
2.2. Sensory analysis
2.4. Evaluation of the oxidative status
The samples of bread crumb used for sensory analysis consisted of bread sliced to a thickness of approximately 1.2 cm using a slicing machine (Gourmet Food Slicer, MAS 6108, Bosch, Stuttgart, Germany) and divided into quarters. The end pieces of each bread loaf, approximately 5 cm, were discarded. One serving consisted of two quarters of a bread slice, free of crust, put in a transparent plastic container and labelled with a three-digit number. The sensory evaluation was conducted by a sensory panel consisting of 8 experienced assessors (7 females/1 male, aged from 42 to 55). The sensory profiling was performed in a sensory evaluation laboratory, which fulfil requirements according to international standards (ASTM, 1986; ISO, 1988). The sensory evaluation started with a panel discussion, where the assessors agreed on a consensus list of attributes and on a definition of each attribute. Fifteen attributes were included in the list: 6 aroma related attributes (nasal perception), 6 flavour related attributes (oral and retro nasal perception), and 3 taste and mouth-feel related attributes (oral perception and physical sensation in the oral cavity). The attributes and matching descriptions are listed in Table 2. All assessors were experienced in evaluating plant based food products on a weekly basis, and most of the assessors had experience in evaluating whole wheat bread and therefore the training sessions were limited to four times 2 h sessions. After each training session, assessors received feedback on their performance with the aim of improving and standardising the panel’s discriminative power. For each of the four bread varieties (control, addition of atocopherol, addition of fat soluble rosemary extract or addition of water dispersible rosemary extract) the assessors evaluated bread stored for: ‘0 weeks’, ‘1 week’, ‘2 weeks’, and ‘3 weeks’ in triplicates over a period of three days giving a total of 16 samples per test day. The assessors evaluated the samples at individual speed on an unstructured 15.0 cm line scale anchored on the left by ‘low’ and on the right with the term ‘high’. Ratings were registered directly on a computerised registration system (Fizz software, 2.30C, Biosystemes, Couternon, France).
The overall antioxidative capacity analysed by the TEAC and ORAC assay together with analysis of lipid hydroperoxides (POV) were performed as described by Jensen et al. (2011b).
2.3. Analysis of volatile compounds Head space collection of volatile compounds from whole wheat bread crumb was conducted and analysed by GCeMS as described by Jensen et al. in press. Verification of compounds suggested by the MS database (NIST, 1998) was made by comparison of the relative retention indices and MS of authentic reference compounds. Authentic compounds were supplied by SigmaeAldrich (Sigmae Aldrich Inc., Steinheim, Germany). Table 2 Sensory attributes and descriptions. Category
Attribute
Description
Aroma
Dough Bran Rancid Acid Spice Dust Dough Sweet aromatic Rancid Dust Bran Spice Bitter Prickly Astringency
Freshly made wheat dough Bran, cereal-like Rancid linseed Propanoic acid Mild spicy, roasted Dry Freshly made wheat dough Toffee, malt-like Rancid linseed Dry Bran, cereal-like Mild spicy, roasted, cocoa Bitter Prickly sensation on the tip of the tongue Pungent feeling on the side of the tongue and pain-like feeling in the jaw
Flavour
Taste and Mouth-Feel
2.5. Statistics Data were analysed for significant effects of type of antioxidant added and storage time using univariate analysis of variance (ANOVA) (SAS Institute, version 9.1, Cary, NC). For the sensory data analysis, the interaction between sample and assessor was used as random effect, and thereby, level and range differences between individual assessors were modelled (Brockhoff and Skovgaard, 1994). Statistical significance was defined at P < 0.05. Principal component analysis (PCA) was used to interpret the structure in the large amount of volatile compounds while partial least squares (PLS) regression was used to study the prediction of sensory data (Y-data) from chemical measurements (X-data). Uncertainty testing by a jack-knife procedure was applied to identify variables of highest importance in the predictive model (Martens and Martens, 2000). Multivariate data analyses were performed using Unscrambler (version 9.6, CAMO Software A/S, Oslo, Norway) and LatentiX (version 2.00, Latent5 Aps, Copenhagen, Denmark). Full cross validation was used as validation criterion and all data were auto-scaled. 3. Results and discussion 3.1. Sensory analysis Results from the univariate ANOVA showed a significant effect of assessors for all tested attributes, which is a common observation in sensory evaluation and can be ascribed to level differences in the scaling (Brockhoff and Skovgaard, 1994; Naes and Langsrud, 1998) (data not shown). The panel was able to distinguish between samples according to storage time and less pronounced between types of antioxidant added. Main effect of storage time was obtained for 80% of the attributes, emphasising that the sensory quality of bread is highly dependent on storage time and that aroma, flavour, and taste are in fact important characteristics when shelf life of bread is to be determined. Fig. 1 illustrates sensory mean data across replicates and assessors for each bread variety. It is obvious that storage affects the sensory profiles and that especially the sensory profile of non-stored (0 weeks) bread displays a huge difference compared to bread stored for 1, 2, and 3 weeks. A significant effect of antioxidant addition was displayed for the aroma attributes: ‘dough’, ‘bran’, and ‘rancid’, and for rancid flavour. Especially bread enriched with a-tocopherol differed from the remaining samples (Fig. 1) as these samples were generally evaluated to have a lower intensity of dough, bran, and dust aroma and a higher intensity of rancid aroma and flavour. a-Tocopherol has been found to enhance the sensory and chemical stability in various food products (Paradiso et al., 2008; Warner and Moser, 2009). In a study concerning the stability of vitamin E in wheat flour, wheat flour with high vitamin E content was found to delay the formation of hexanal compared to flour with a lower content of vitamin E (Nielsen and Hansen, 2008). Hexanal is commonly used as an indicator of lipid oxidation as it contributes to the rancid flavour. The antioxidant/pro-oxidant nature of a-tocopherol depends, however, on the concentration, and a pro-oxidative effect has been observed especially when a-tocopherol has been added in relatively high concentrations (Huang et al., 1994). The rancid flavour of the present bread enriched with a-tocopherol suggests a prooxidative effect of a-tocopherol and might be a result of the
294
S. Jensen et al. / Journal of Cereal Science 53 (2011) 291e297
Fig. 1. Spider plot of sensory scores (0 (low intensity)-15 (high intensity)) of control bread and bread enriched with a-tocopherol, and with a fat soluble (L) and water dispersible (H) rosemary extract, respectively. Means are calculated across replicates and assessors. A_: Aroma attributes; F_: Flavour attributes.
relatively high concentration of a-tocopherol (1 g kg1 flour). This conclusion is supported by analyses performed by Park et al. (2002) and Ochi et al. (1993), who observed a pro-oxidative effect of tocopherols when added in concentrations of 100e200 mg kg1 to heat treated cereal products such as cookies and rice snacks. 3.2. Analysis of volatiles Thirty five different volatiles were identified and verified in the present study. The concentrations of all verified volatile compounds for the four bread varieties are given in Table 3. The volatile compounds can be divided into seven classes: ‘acids’, ‘alkanes’, alcohols’, ‘aldehydes’, ‘esters’, ‘furans’, ‘ketones’, and ‘terpenes’. Statistical analysis showed significant effect of the interaction between storage and bread variety for the total content of acids (P ¼ 0.048). Furthermore, a significant main effect of storage was found for the decreasing content of alcohols, aldehydes, alkanes, furans, and terpenes (P < 0.0001). In addition, aldehydes were found to be significantly affected by bread variety (P ¼ 0.041). Higher concentrations of aldehydes were found in fresh samples of bread enriched with a-tocopherol compared to the other bread varieties. Especially, the secondary oxidation products hexanal (V12) and heptanal (V15) were higher in fresh samples of these bread loaves indicating a higher degree of lipid oxidation already during baking of bread enriched with a-tocopherol.
A multivariate interpretation of the volatile components (Fig. 2) confirmed the results from the ANOVA as it is evident that bread samples could be separated according to effect of storage time along PC1 (PC1 explains 51% of variance) and less pronounced according to effect of antioxidative additives along PC2 (PC2 explains 10% of variance). This differentiation of samples in accordance with storage was in good agreement with results from the sensory analysis. Fresh samples of bread are positioned in the opposite direction of PC1 compared to the more stored samples, with a few exceptions (a-tocopherol stored for 1 and 2 weeks). A differentiation of bread enriched with a-tocopherol and bread enriched with rosemary extracts are seen along PC2. No obvious differences in volatile profiles between fat soluble and water dispersible rosemary extracts were found. The efficiency of fat soluble and water dispersible antioxidants is known to differ significantly when tested in similar food systems (Frankel, 1996) and a differentiation of samples according to type of rosemary extract could therefore be expected. However, the present data showed no difference between the two tested rosemary extracts which could be explained by the active ingredients in the rosemary extracts being the same, and that these ingredients generally did not have an effect on the chemical stability of bread. The difference in solubility of the two extracts is exclusively obtained by addition of the emulsifier polyoxyethylene (20) sorbitan monooleate, E433, to the water dispersible rosemary extract.
S. Jensen et al. / Journal of Cereal Science 53 (2011) 291e297
295
Table 3 Maximum and minimum values for the concentration [ng g1 of bread] of verified volatile compounds in control bread and bread enriched with a-tocopherol and a fat soluble (L) and water dispersible (H) rosemary extracts. No Acid
Alkane and alcohol
Aldehyde
Ester
Furan
Ketone
Terpene
28 31 32 33 1 8 10 13 17 24 25 35 2 6 7 12 15 26 30 4 19 23 27 3 18 20 29 34 5 9 11 14 21 22 16
Volatiles Acetic acid Propanoic acid 2-Methylpropanoic acid Butanoic acid Octane Ethanol 1-Propanol 2-Methyl-1-propanol 3-Methyl-1-butanol Hexanol 3-Ethoxy-1-propanol Phenylethanol 2-Methylpropanal 2-Methylbutanal 3-Methylbutanal Hexanal Heptanal Nonanal Benzaldehyde Ethylacetate Ethylhexanoate Ethyllactate Ethyloctanoate 2-Methylfuran 2-Pentylfuran 2-Methyltetrahydro3furanone Furfural Furfuryl alcohol 2-Butanone 2,3-Butandione 2,3-Pentandione 2-Heptanone 3-Hydroxy-2-butanone 1-Hydroxy-2-propanone Limonene
Control
a-Tocopherol
Rosemary (L)
Rosemary (H)
0.129e0.336 25.618e43.362 2.039e3.087 0.145e0.211 0.059e0.571 731.713e1352.534 4.011e6.748 93.071e180.345 28.235e77.777 1.006e2.114 0.212e0.349 1.904e2.192 0.196e2.100 0.156e3.617 0.698e11.597 1.845e3.867 0.271e1.050 0.337e0.518 0.746e1.013 1.274e3.809 0.210e0.502 0.166e0.608 0.418e0.896 0.107e0.204 0.158e0.853 0.090e0.188 0.330e1.255 0.187e0.613 0.414e0.778 4.95e6.291 0.137e0.698 0.177e0.235 16.596e27.173 0.763e1.823 0.048e0.145
Min-Max 0.150e0.263 25.366e39.995 2.127e2.997 0.118e0.196 0.078e0.691 618.571e1406.702 3.995e6.777 95.596e180.886 27.892e77.834 0.856e2.141 0.218e0.296 1.627e2.089 0.237e2.582 0.187e5.954 1.055e14.375 2.690e4.758 0.318e1.157 0.138e0.505 0.601e1.099 1.126e2.141 0.375e0.656 0.312e0.662 0.376e0.708 0.105e0.281 0.175e0.955 0.099e0.247 0.267e1.593 0.240e0.471 0.343e0.760 3.637e7.481 0.067e0.898 0.242e0.390 23.925e28.379 0.577e1.494 0.050e0.173
0.118e0.225 28.737e37.492 2.134e2.762 0.176e0.292 0.060e0.541 654.456e1508.114 4.055e6.930 93.648e179.107 28.529e73.020 0.917e1.974 0.213e0.284 1.704e2.260 0.184e1.657 0.155e2.920 0.781e8.213 1.616e3.747 0.229e0.499 0.277e0.680 0.678e0.956 1.250e2.512 0.253e0.468 0.151e0.667 0.440e0.777 0.073e0.261 0.146e0.726 0.082e0.156 0.287e0.925 0.217e0.470 0.366e0.727 4.470e5.706 0.123e0.659 0.096e0.559 21.732e25.685 2.395e3.576 0.044e0.174
0.130e0.234 27.112e40.623 1.995e2.851 0.164e0.216 0.044e0.426 592.502e1491.608 3.726e7.331 70.941e204.239 22.172e84.544 0.669e2.163 0.206e0.283 1.654e2.152 0.188e1.478 0.164e3.071 0.900e8.729 1.555e4.006 0.192e1.062 0.310e0.523 0.558e0.963 1.228e2.863 0.248e0.548 0.296e0.610 0.340e0.722 0.050e0.129 0.141e0.757 0.074e0.144 0.210e0.857 0.182e0.431 0.282e0.646 4.357e7.244 0.108e0.706 0.136e0.264 21.569e25.945 1.155e1.947 0.070e0.191
Sign. variety
**
*** * *
*
*
***
Fig. 2. Principal component analysis bi-plot on volatile compounds in whole wheat bread (control) and whole wheat bread enriched with a-tocopherol and fat soluble (L) and water dispersible (H) rosemary extracts. Volatiles (V) are designated by the number of the individual compound given in Table 3. The digits at the end of the sample name states the time of storage [0e5 weeks].
296
S. Jensen et al. / Journal of Cereal Science 53 (2011) 291e297
A substantial part of the volatiles were associated with the fresh samples of bread, independent of antioxidant added. Furfuryl alcohol (V34) and butanoic acid (V33) increased in concentration during storage while all other volatiles decreased as a consequence of storage. Furfuryl alcohol was found to increase in whole wheat bread crumb and decrease in whole wheat bread crust during a 3 week period of storage in a previous study conducted by the present authors (Jensen et al., accepted for publication), suggesting that furfuryl alcohol migrates from crust to crumb during storage. 3.3. Oxidative status The measurement of lipid hydroperoxides (POV) was performed to follow the progress in lipid oxidation. No changes in POV was observed during storage, but a significant difference (P < 0.0001) was observed for the type of antioxidant added. A substantially higher level of hydroperoxides was present in bread enriched with a-tocopherol compared to the three other bread varieties (0.7e0.9 mE H2O2 g1 of bread enriched with a-tocopherol, based on dry matter content (DM) compared to 0.0e0.2 mE H2O2 g1 of bread for control bread and bread enriched with rosemary extracts, DM). Similar results were obtained in a study performed by Terao and Matsushita (1986), who found an increase in the concentration of hydroperoxides for edible oil enriched with a-tocopherol. The elevated level of hydroperoxides can be explained by breakdown of lipid hydroperoxides induced by reactive species of a-tocopherol (a-tocopherol radical). The accumulation of hydroperoxides observed in the present study right after baking can thus be explained by an increase in the concentration of a-tocopheroxyl as a result of the elevated temperatures during the baking process. Comparison of Fig. 1 and Table 3 shows that the increased level of lipid hydroperoxides in bread with added a-tocopherol was also accompanied by increased rancidity score in the sensory evaluation. A study by Haakansson and Jägerstad (1990) showed that vitamin E loss in whole wheat flours started immediately after flour and water were mixed due to lipoxygenase activity and the loss increased as the temperature of the flour-water slurry increased. Another study determined the content of tocopherols to decrease by one third during baking (Ranhotra et al., 2000). Vitamin E loss is equal to formation of a-tocopheroxyl radicals and a-tocopheroxyl
radicals have been shown to act as pro-oxidants as mentioned above. The high concentration of lipid hydroperoxides in bread enriched with a-tocopherol is accordingly suggested to be a result of a pro-oxidative activity of oxidized tocopherols during bread making. The overall antioxidative capacity was evaluated by the TEAC assay and the ORAC assay. The antioxidative capacity in bread samples when measured by ORAC assay was between 17.4 and 19.8 mg trolox equivalents g1 crumb, DM. This was significantly higher than the level measured by TEAC, which showed the antioxidative capacity to be between 4.8 and 6.8 mg trolox equivalents g1 crumb, DM. Both assays exhibited significance for the interaction between storage time and bread variety, emphasising that the development in overall antioxidative capacity during storage depended on bread variety. 3.4. Correlation of sensory and chemical results To study the correlation between sensory and chemical (GCeMS, POV, ABTS, and ORAC) data a multivariate PLS regression, based on prediction of the sensory data (Y-variables) from the chemical data (X-variables) was performed (Fig. 3). The regression plot shows that 1) most of the aroma attributes are spanned by PC1 together with a large amount of volatile compounds, 2) a few flavour attributes are spanned by PC2 together with fewer volatile compounds, and 3) another large group of attributes related to flavour, taste, and mouth-feel are located near origio and are thus not explained by any of the volatile compounds. For PC1, the dough aroma, mainly presented in the fresh bread samples, was positively correlated with a large number of volatiles like ethanol (V8), 1-propanol (V10), 2,3-pentadione (V11), 2-methyl-1-propanol (V13), limonene (V16), 3-methyl-1-butanol (V17), hexanol (V24), and ethyloctanoate (V27) and inversely correlated to furfuryl alcohol (V34) which, on the contrary, correlated with the aroma attributes: ‘rancid’, ‘spicy’, and ‘aromatic sweet’. This means that most of the sensory aroma attributes are correlated to volatile compounds, and that volatility has a great impact on the perceived changes of bread aroma. In PC2, a positive correlation was found between two flavour attributes: ‘bran’ and ‘dough’, mainly found in samples added rosemary extract, and
Fig. 3. Partial least squares (PLS) regression with only significant important variables of 4 different bread varieties (Control bread, bread enriched with a-tocopherol, fat soluble (L) and water dispersible (H) rosemary extracts) and with 4 different storage times. Score plot (a) and loading plot (b) with chemical data as X-variables and sensory data as Y-variables. Sensory aroma attributes (normal), flavour attributes (italic), and taste and mouth-feel (underlined). The digit in sample name indicates storage time [0e3 weeks].
S. Jensen et al. / Journal of Cereal Science 53 (2011) 291e297
ethyloctanoate (V27), which was inversely correlated to ‘rancid’ flavour. The remaining flavour, taste, and mouth-feel attributes being: ‘prickly’, ‘bitter’, ‘spicy’, ‘astringent’, ‘acid’, and ‘dust’ are less correlated to any of the aroma compounds and the TEAC. This indicates that the volatile compounds are not the main contributors to the perceived flavour, taste, and mouth-feel of bread and probably non-volatile compounds are responsible for changes of these attributes. 4. Conclusion The present study showed that the tested antioxidative additives did not improve the sensory quality and stability of bread. Changes in the sensory profile and in the content of volatile compounds in bread enriched with a-tocopherol or rosemary extracts were, like for the control bread, mainly ascribed to the effect of storage time whereas the effect of bread enrichment was less pronounced. Most volatile compounds were present in higher concentrations in the fresh samples of bread, emphasising that volatility is an important contributor to changes, especially changes related to aroma, during storage. Fresh samples of bread enriched with a-tocopherol had higher concentrations of hydroperoxides and secondary lipid oxidation products. These samples shared some of the same sensory characteristics as stored control samples indicting that lipid oxidation is responsible for these less favourable sensory notes like rancid aroma and flavour, bitter taste, and astringency. The analysis of the antioxidative capacity showed that the relatively small changes in antioxidative capacity during storage depended on the type of additive included in the bread recipe. Acknowledgements The authors gratefully acknowledge the excellent technical assistance from Nina Eggers and Birgitte Foged, the great help from Lone R. Borum in the sensory lab as well as the work performed by Bo B. Madsen, who baked the large number of breads forming the basis of the present work. The investigation is a part of a Ph.D. study supported financially by Novozymes, University of Copenhagen, and the Research School FOOD Denmark. References Aruoma, O.I., Halliwell, B., Aeschbach, R., Löligers, J., 1992. Antioxidant and prooxidant properties of active rosemary constituents: carnosol and carnosic acid. Xenobiotica 22, 257e268. ASTM, 1986. Physical Requirements, Guidelines for Sensory Evaluation Laboratories, STP 913. American Society for Testing and Materials, PA. Brockhoff, P.M., Skovgaard, I.M., 1994. Modelling individual differences between assessors in sensory evaluations. Food Quality and Preferences 5, 215e224. Chiavaro, E., Vittadini, E., Musci, M., Bianchi, F., Curti, E., 2008. Shelf-life stability of artisanally and industrially produced durum wheat sourdough bread (‘‘Altamura bread’’). LWT 41, 58e70. Fernandez, U., Vodovotz, Y., Courtney, P., Pascall, M.A., 2006. Extended shelf life of soy bread using modified atmosphere packaging. Journal of Food Protection 69, 693e698. Frankel, E.N., 1996. Antioxidants in lipid foods and their impact on food quality. Food Chemistry 57, 51e55.
297
Grosch, W., Wieser, H., 1999. Redox reactions in wheat dough as affected by ascorbic acid. Journal of Cereal Science 29, 1e16. Haakansson, B., Jägerstad, M., 1990. The effect of thermal inactivation of lipoxygenase on the stability of vitamin E in wheat. Journal of Cereal Science 12, 177e185. Holtekjoelen, A.K., Baevre, A.B., Roedbotten, M., Berg, H., Knutsen, S.H., 2008. Antioxidant properties and sensory profiles of breads containing barley flour. Food Chemistry 110, 414e421. Huang, S.-W., Frankel, E.N., German, J.B., 1994. Antioxidant activity of a- and g-tocopherols in bulk oils and in oil-in-water emulsions. Journal of Agricultural and Food Chemistry 42, 2108e2114. IARC, 1986a. Butylated hydroxyanisole (BHA). IARC Monographs on the evaluation of the Carcinogenic risks of chemicals to Humans. Some naturally occurring and synthetic food components. Furocoumarins and Ultraviolet Radiation 40, 123e159. IARC, 1986b. Butylated hydroxytoluene (BHT). IARC Monographs on the evaluation of the Carcinogenic risk of chemicals to Humans., some naturally occurring and synthetic food components. Furocoumarins and Ultraviolet Radiation 40, 161e206. ISO, 1988. International Standard 8589, Sensory Analysis - General Guidance for the Design of Test Rooms. Ref. no. ISO 8589:1988 (E). International Organization for Standardization, Genéve, Switzerland. Jensen, S., Oestdal, H., Thybo, A.K., 2010. Sensory profiling of changes in wheat and whole wheat bread during a prolonged period of storage. Journal of Sensory Studies 25, 231e245. Jensen, S., Oestdal, H., Skibsted, L. H., Larsen, E., Thybo, A. K., Chemical changes in wheat pan bread during storage and how it affects the sensory perception of aroma, flavor, and taste. Journal of Cereal Science, in press. Jensen, S., Oestdal, H., Clausen, M.R., Andersen, M., Skibsted, L.H., 2011b. Oxidative stability in whole wheat bread during storage. LTW e Food Science and Technology 44, 637e642. Martens, H., Martens, M., 2000. Modified Jack-knife estimation of parameter uncertainty in bilinear modelling by partial least squares regression (PLSR). Food Quality and Preference 11, 5e16. Naes, T., Langsrud, O., 1998. Fixed or random assessors in sensory profiling? Food Quality and Preferences 9, 145e152. Nielsen, M.M., Hansen, Aa, 2008. Stability of vitamin e in wheat flour and whole wheat flour during storage. Cereal Chemistry 85, 716e720. NIST, 1998. Mass Spectral Database, Standard Reference Database 1. National Institute of Standards and Technology, Gaithersburg, Maryland, US. Ochi, T., Tsuchiya, K., Ohtsuka, Y., Aoyama, M., Maruyama, T., Niiya, I., 1993. Synergistic antioxidant effects of organic acids and their derivatives with tocopherols on cookies. Nippon Shokuhin Kogyo Gakkaishi 40, 393e399. Paradiso, V.M., Summo, C., Trani, A., Caponio, F., 2008. An effort to improve the shelf life of breakfast cereals using natural mixed tocopherols. Journal of Cereal Science 47, 322e330. Park, Y.S., Kim, Y.S., Shin, D.H., 2002. Antioxidative effects of ethanol extracts from Rhus verniciflua Stoke on Yukwa (Oil popped rice snack) base during storage. Journal of Food Science 67, 2474e2479. Peng, X., Ma, J., Cheng, K.-W., Jiang, Y., Chen, F., Wang, M., 2010. The effects of grape seed extract fortification on the antioxidant activity and quality attributes of bread. Food Chemistry 119, 49e53. Pokorny, J., 1991. Natural antioxidants for food use. Trends in Food Science & Technology 2, 223e227. Ranhotra, G.S., Gelroth, J.A., Okot-Kotber, B.M., 2000. Stability and dietary contribution of vitamin E added to bread. Cereal Chemistry 77, 159e162. Sargent, K., 2008. A softer“ approach to improving the quality of refrigerated bakery products. Cereal Foods World 53, 301e305. Schieberle, P., Grosch, W., 1992. Changes in the concentration of potent odorants during storage of white bread. Flavour and Fragrance Journal 7, 213e218. Seitz, L.M., Chung, O.K., Rengarajan, R., 1998. Volatiles in selected commercial breads. Cereal Chemistry 75, 847e853. Terao, J., Matsushita, S., 1986. The peroxidizing effect of a-tocopherol on autoxidation of methyl linoleate in bulk phase. Lipids 21, 255e260. Warner, K., Moser, J., 2009. Frying stability of purified mid-oleic sunflower oil triacylglycerols with added pure tocopherols and tocopherol mixtures. Journal of the American Oil Chemists’ Society 86, 1199e1207. Yanishlieva, N.V., Marinova, E.M., 1992. Inhibited oxidation of lipids. 1. Complex estimation and comparison of the antioxidative properties of some natural and synthetic antioxidants. Fett Wissenschaft Technologie 94, 374e379. Zhou, K., Su, L., Yu, L.L., 2004. Phytochemicals and antioxidant properties in wheat bran. Journal of Agri-cultural and Food Chemistry 52, 6108e6114.