Fiber enriched reduced sugar muffins made from iso-viscous batters

LWT - Food Science and Technology 65 (2016) 32e38

Contents lists available at ScienceDirect

LWT - Food Science and Technology journal homepage: www.elsevier.com/locate/lwt

Fiber enriched reduced sugar muffins made from iso-viscous batters Susanne Struck*, Linda Gundel, Susann Zahn, Harald Rohm €t Dresden, 01062 Dresden, Germany Chair of Food Engineering, Technische Universita

a r t i c l e i n f o

a b s t r a c t

Article history: Received 12 May 2015 Received in revised form 2 July 2015 Accepted 20 July 2015 Available online 22 July 2015

The application of dietary fiber produced from fruit or vegetable by-products in baked foods is of growing interest for the food industry, providing the possibility for delivering reduced energy/sugar products. The aim of this study was to analyze the potential of sucrose replacement by a combination of rebaudioside A and wheat, apple or pea fiber, and the respective influence on batter and product characteristics. The main focus was on the production of muffins from batters with similar viscosity, the formulation of which was realized by the adaption of water levels in the recipes. Small strain oscillatory measurements of muffin batter with temperature sweeps from 25 to 100
C revealed that thermally induced structure modifications caused by protein denaturation and starch gelatinization were delayed with increasing replacement of sucrose by fiber. Volume, water activity and crumb firmness increased and in vitro starch digestion decreased with increasing level of fiber incorporation and sugar replacement. Sucrose replacement of 30% by wheat fiber and rebaudioside A resulted in products close to the reference muffin. © 2015 Elsevier Ltd. All rights reserved.

Keywords: Baking Fiber Muffins Rebaudioside A Rheology Starch digestion

1. Introduction Overweight and obesity have been gaining relevance in industrial countries, especially in the part of the society which is susceptible to malnutrition. The increase in diet-associated chronic diseases has been related to factors such as inappropriate dietary patterns, decreased physical activity and increased tobacco use (Nishida, Uauy, Kumanyika, & Shetty, 2004). To promote a healthier diet, the food industry frequently focuses on the production of reduced fat/sugar/energy foods with a sensory quality that is comparable to that of conventional products. Quick bread muffins for example are popular breakfast or snack products with high consumer acceptance. Because muffins are high in sugar and fat, several attempts were made to increase their nutritional value by, e.g., replacing sucrose with high intensive sweeteners, or by incorporating dietary fiber. Glycosides from Stevia rebaudiana Bertoni are heat stable natural sweeteners with a relative sweetness of approx. 300 and were recently applied as sugar replacers in sweet baked foods (Manisha, Soumya, & Indrani, 2012; Zahn, Forker, Krügel, & Rohm, 2013). In these products sucrose is one of the main ingredients; it is responsible for sweetness, flavor, and texture formation, and it

* Corresponding author. E-mail address: [email protected] (S. Struck). http://dx.doi.org/10.1016/j.lwt.2015.07.053 0023-6438/© 2015 Elsevier Ltd. All rights reserved.

contributes to volume increase, crust color, shelf life and moisture retention (Nip, 2006). Sucrose reduction causes reduced batter viscosity which results in low product volume and poor cell structure (Manisha et al., 2012). Because of its technofunctional properties and the high proportion in most recipes replacement of sucrose is challenging when product characteristics should be similar to those of full sugar bakery products (Struck, Jaros, Brennan, & Rohm, 2014). As sugar substitutes need to replace all major functions of sucrose, the most promising approach is to use sweeteners with different sweetening intensity, and to combine them with bulking agents that act as structure building substances (e.g., inulin, polydextrose, dietary fiber) (Esteller, Amaral, & Lannes, 2004; Martínez-Cervera, Sanz, Salvador, & Fiszman, 2012; Riedel, €hme, & Rohm, 2015; Zahn et al., 2013; Zoulias, Piknis, & Bo Oreopoulou, 2000). To increase dietary fiber in baked goods, products from different sources can be incorporated into the recipe for partially replacing flour, sugar or fat. Fibers that were used in recent studies were from mango or potato peels (Ajila, Leelavathi, & Prasada Rao, 2008; Arora & Camire, 1994), apple pomace (Masoodi, Sharma, & Chauhan, 2002; Sudha, Baskaran, & Leelavathi, 2007), and orange or grape pomace (Mildner-Szkudlarz, Bajerska, Zawirska-Wojtasiak, &  recka, 2013; O'Shea, Doran, Auty, Arendt, & Gallagher, 2013). Go The incorporation of dietary fiber in sweet baked goods is associated with an increase in batter viscosity, which has a great influence on product volume and texture (Lebesi & Tzia, 2011;

S. Struck et al. / LWT - Food Science and Technology 65 (2016) 32e38

Martínez-Cervera, Salvador, Muguerza, Moulay, & Fiszman, 2011). The increase in batter viscosity is related to physicochemical properties of the fiber such as its high water binding capacity  mez, & Rosell, 2012), which reduces the (Gularte, de la Hera, Go water available for other ingredients and, consequently, affects product characteristics such as crumb hardness and chewiness, or volume (Grigelmo-Miguel, Carreras-Boladeras, & Martín-Belloso, 1999; Sudha et al., 2007). It was our hypothesis that the impact of sucrose replacement and fiber addition on batter viscosity can be balanced by adjusting the water level in the recipes. The aim of this study therefore was to evaluate the integration of apple, pea and wheat fiber in muffin batters of similar viscosity (further denoted as iso-viscous batters), and to examine the effects of sugar replacement by rebaudioside A (RebA) and these fibers on batter characteristics and product properties. 2. Materials and methods 2.1. Analysis of fiber composition Moisture content of wheat (WF 600-30), apple (AF 401-30) and pea (EF 100) fiber (JRS Rettenmaier, Rosenberg, Germany) was determined by drying in an oven at 103
C to constant mass. Fat content was analyzed by Soxhlet extraction, protein by the Kjeldahl procedure (conversion factor 6.25). Ash was determined gravimetrically (5 h, 550
C) in a muffle furnace. Carbohydrates were calculated by difference. 2.2. Water binding capacity (WBC) The determination of fiber WBC was based on centrifugation (Chen, Piva, & Labuza, 1984). 4 g fiber was mixed with 40 g water and kept at room temperature for 30 min. Samples were centrifuged at 2000 g for 10 min, and the supernatant was decanted and weighed. Supernatant dry matter, giving the amount of fiber that remains in the aqueous phase, was determined using an MA 30 € ttingen, Germany) at 95
C. WBC moisture analyzer (Sartorius, Go refers to the amount of water bound per gram dry fiber that remained in the centrifugation sediment.

33

2.4. Rheological analysis Rotational and oscillatory measurements of muffin batters, separately produced without baking powder, were performed with a Physica MCR 300 rheometer (Anton Paar, Germany) equipped with a 50 mm plate-plate-geometry. After preparation, each batter was kept at 25
C for 1 h and loaded to the geometry. The gap was adjusted to 1.025 mm as trimming position to carefully remove excess batter and, after batter relaxation for 10 min, finally adjusted to 1 mm. Experiments were performed from two batches in duplicate.

2.4.1. Flow properties Apparent viscosity ha was measured at 25
C as a function of shear rate (0.1  g_  100 s1). Recipe development to create isoviscous batters was carried out by adapting the water level. ha of the reference (22.7 Pa s at g_ ¼ 10 s1) was the basis for an iterative approach to adjust viscosity of fiber-containing batter. For each fiber, 30% or 60% (100% for WF only) sucrose was replaced with a mixture of water and fiber. The fiber/water ratio was adjusted until ha at 10 s1 did not differ significantly from reference ha. The batter flow curves were power-law fitted

ha ¼ k$g_ n1

(1)

where k (Pa sn) refers to the viscosity factor, and n is the flow behavior index.

2.4.2. Viscoelastic properties All oscillatory measurements were performed in the linear viscoelastic region at a strain of g ¼ 0.0005. The gap was covered with vaseline oil to prevent dehydration. A frequency sweep was performed at 25
C were f decreased logarithmically from 10 to 0.1 Hz. A subsequent temperature sweep from 25 to 100
C was conducted at f ¼ 1 Hz and a heating rate of 5 K min1, previously determined as approximate heating rate in the center of a muffin during baking. After holding at 100
C for 10 min, a final frequency sweep was performed (10  f  0.1 Hz). 2.5. Product characteristics

2.3. Muffin formulation and preparation The reference recipe comprised 200.0 g wheat flour, 120.0 g sucrose, 10.0 g baking powder, 15.0 g skim milk powder, 15.0 g whole egg powder mixed with 45.0 g water, 2.4 g salt, 80.0 g canola oil and 120.0 g water. Reduced sugar muffins were prepared by replacing 30, 60 or 100% sucrose with a combination of fiber, water and steviol glycosides. Apple fiber (AF), wheat fiber (WF) or pea fiber (PF) and water were incorporated to achieve batters isoviscous to the reference. RebA was added to ensure iso-sweetness that has been determined previously (Zahn et al., 2013). To replace 36.0 g (¼30%) sucrose, 0.14 g RebA and 35.86 g waterefibermixture were added; to replace 60% it were 0.28 g and 71.72 g, respectively. For full replacement, 0.48 g RebA and 119.52 g waterefiber-mixture were used and analyzed solely for WF since it gave the best product characteristics and batter processibility among the fibers. Batter preparation was performed as described previously (Zahn, Pepke, & Rohm, 2010), and the fiber was added with the dry ingredients. Finally, 43 ± 0.1 g batter were filled in paper cups, placed in a muffin baking tray and baked at 200
C top heat and 180
C bottom heat for 24 min (MIWE condo, Arnstein, Germany). Products from each recipe were produced, baked and analyzed in two independent batches.

Maximum height and average diameter of ten muffins were measured using a digital caliper. After weighing, relative baking loss was calculated. Volume was determined in triplicate using the canola seed displacement method. Water activity of milled muffins was measured in triplicate with a thermoconstanter (Novasina, Lachen, Switzerland, Germany) at 25
C. Crumb texture was analyzed using a XT2i Texture Analyzer (Stable Micro Systems, Surrey, UK). After storing the muffins in polythene bags at room temperature for 1 d, cylinders (d ¼ 22.5 mm) were cut from the crumb of five muffins from each batch (28 mm length). Using an acrylic plate (d ¼ 40 mm), crumb cylinders were compressed to 50% height at 1 mm/s, decompressed and compressed again; off-time between compressions was 5 s. Crumb firmness was the peak force during first compression, and cohesiveness refers to the ratio of areas under the 2nd to the 1st compression cycle in forceetime diagrams (Zahn et al., 2013). A Luci 100 spectral colorimeter (D65 Xenon lamp, 10
standard observer; Hach Lange, Düsseldorf, Germany) was used to measure crumb and crust color from three muffins in triplicate. The measurement was based on the CIE-Lab color space, and lightness L*, chroma C* and hue angle hab were further considered as color descriptors (Rohm & Jaros, 1996).

34

S. Struck et al. / LWT - Food Science and Technology 65 (2016) 32e38

2.6. Sensory analysis

3.2. Rheological properties of muffin batter

The standard muffins, and the three samples with 30% sucrose replacement were subjected to flash profiling as described recently (Zahn et al., 2013). The raw data collected from 12 panelists (five female, all with experience in sensory analysis) in two individual sessions were analyzed using General Procrustes Analysis (GPA) in the Senstools.Net software (OP & P Product Research BV, Utrecht, Netherlands).

Table 1 summarizes the ratios of water to fiber for iso-viscous batters (i.e., ha ¼ 22.7 Pa s at g_ ¼ 10 s1). These ratios were 5.0e6.2 at 30% sucrose replacement, 4.6e5.0 at 60%, and 4.0 when all sucrose was replaced by a mixture of WF and water. These values were lower than WBC for WF, AF and PF, indicating that an experimentally determined water binding capacity does not necessarily reflect real conditions in a batter. Power-law fitting of the flow curves (R2;  0.99) showed pronounced shear thinning. The power-law parameter k, which refers to ha at 1 s1, ranged from 66.0 ± 2.4e83.5 ± 2.1 Pa s (P > 0.05), and the flow exponent n from 0.273 to 0.397. In small strain oscillatory measurements, the complex modulus G* of batters at 25
C increased with increasing frequency, and a loss tangent (tan d) < 1 in the entire frequency domain indicates a soft gel behavior (Baixauli, Sanz, Salvador, & Fiszman, 2008; Sanz, Salvador, & Fiszman, 2008). For the WF muffins, G* increased with increasing sucrose replacement (Fig. 1); tan d at f ¼ 1 Hz was 0.61 (reference), 0.52 (30% less sucrose), 0.48 (60% less sucrose), and 0.39 (full sucrose replacement). Despite all muffin batters were adjusted to similar shear viscosity, similar trends concerning the relation between G* or tan d and sucrose replacement were observed for AF and EF batters (data not shown). This is in line with stiffening effects of different fiber types reported for wheat dough and gluten-free bread dough (Ktenioudaki, O'Shea, & Gallagher, 2013; O'Shea et al., 2013). Temperature-induced reactions observed at constant g and f are helpful to understand structural changes in batter during heating. The impact of thermal energy leads to (1) a decrease to a stiffness minimum from 25 to 70
C, (2) a subsequent increase in stiffness to a global maximum, and (3) a final stiffness decrease until a constant value is reached, regardless of whether standard batter or batter fortified with functional ingredients is considered (Chen, Kang, & Chen, 2008; Peressini, Foschia, Tubaro, & Sensidoni, 2015). Stiffness reduction in phase (1) is because of the thermal activation of molecules and reduced viscosity, and temperature at which G* minimum occurs depends on batter sucrose content (Table 2), being 66.2
C (reference) and 49.9
C (batter with full sucrose replacement). Sucrose and fiber lower water activity so that less water remains available for the starch granules, causing a delay in swelling, an increased swelling rate, and a larger diameter of swollen starch granules (Bean & Yamazaki, 1978). In the subsequent phase (2) G* increases as a result of structure formation because of protein denaturation and starch swelling. Water enters the starch granule's amorphous regions and causes increasing solubilization of amylose and amylopectin. In phase (3), G* reaches a maximum at a specific temperature, and decreases thereafter because of the completion of structural modifications. The temperature that is associated with the maximum modulus decreases with increasing sucrose replacement, because sucrose raises starch gelatinization temperature and protein denaturation temperature (Manisha et al., 2012; Mizukoshi, 1997); temperature at G*max was 101.0
C for the reference and 78.8
C for WF-100% (Table 2). Donovan (1977) reported that sucrose raises protein denaturation temperature in cake batter by 13
C, and increases starch gelatinization temperature by 30
C. The sucrose-induced delay in starch gelatinization and protein denaturation can be explained by the competition for water (the high hygroscopicity of sucrose reduces the availability of water for starch and proteins), and by intermolecular interactions of sucrose with starch chains in the amorphous regions of the starch granule which lead to the stabilization of that regions. Consequently, more energy is required to break the starch granules, and starch swelling is restrained (Spies & Hoseney, 1982). This effect plays an important

2.7. In vitro starch digestibility In vitro starch digestibility of muffins was analyzed according to Brennan, Merts, Monro, Woolnough, and Brennan (2008) in triplicate. Grinded muffin was mixed with distilled water, HCl and pepsin, and the system was pre-digested at 37
C. After buffering to pH 6, the time 0 aliquot was taken, and amyloglucosidase and pancreatin was added. Samples were incubated under agitation (120 min, 37
C), aliquots were removed after 20, 60 and 120 min digestion. For the amount of reducing sugars, samples were digested with invertase (Sigma, grade V, 30 U/L) and amyloglucosidase (Megazyme, 3260 U/mL) in acetate buffer at 55
C. Measurement of reducing sugars was conducted with 3,5dinitrosalicylic acid (Miller, 1959). The amount of digestible starch expressed as mg reducing sugars per g muffin was calculated. A stoichiometric factor of 0.9 was used to calculate starch from glucose contents. The hydrolysis curves were fitted with a nonlinear-model following Gularte et al. (2012) with modification. The model applied was

  c ¼ c0 þ c∞ $ 1  ekt

(2)

where c0 is reducing sugar concentration at time t ¼ 0 min, c∞ is equilibrium concentration, and k is a kinetic constant.

2.8. Statistical analysis Analysis of variance including StudenteNewmaneKeuls Post hoc-tests at P  0.05 was conducted using Systat 12 (Systat, Erkrath, Germany). Fitting of flow curves and starch hydrolysis curves was performed with Systat Table Curve 2D v4.

3. Results and discussion 3.1. Dietary fiber characteristics The fibers differed significantly in composition and water binding capacity (Table S1). Wheat fiber dry matter was composed of 97 g 100 g1 insoluble dietary fiber, comprising 70% cellulose, 22% hemicellulose and 5% lignin (data provided by supplier). Apple fiber dry matter consisted of 55 g 100 g1 dietary fiber, including 10 g 100 g1 soluble components, mainly pectin. The content of carbohydrates other than dietary fiber was highest in AF, pointing on the high amount of mono- and disaccharides present in the fruit. A protein content at least as twice as high as in other fibers (6.73 g 100 g1) was found in pea fiber. PF further contained 2 g 100 g1 soluble and 68 g 100 g1 insoluble dietary fiber (mainly cellulose, hemicellulose and lignin) and approx. 91 g 100 g1 total carbohydrates. WBC was 3.82 g H2O per g fiber for WF, and approx. 6 g H2O per g fiber in case of AF and PF (P < 0.05). During baking, water absorption of fibers might increase because of temperatureinduced swelling (Kaack & Pedersen, 2005).

S. Struck et al. / LWT - Food Science and Technology 65 (2016) 32e38

35

Table 1 Replacement of sucrose by fiber/water combinations resulting in iso-viscous batters. k and n are power-law fitting parameters at 25
C. Codea

Fiber (g)

Water (g)

Reference WF-30 AF-30 PF-30 WF-60 AF-60 PF-60 WF-100

e 4.96 5.97 5.80 12.78 12.14 11.93 23.86

e 30.90 29.89 30.05 58.94 59.58 59.79 95.66

a b

Water/fiber ratio ()

Consistency index k (Pa. sn)b

Flow behavior index n ()

6.2 5.0 5.2 4.6 4.9 5.0 4.0

74.8a 66.0a 75.2a 70.8a 68.8a 83.5a 78.2a 79.3a

0.397a 0.368ab 0.355ab 0.367ab 0.342ab 0.273b 0.304ab 0.353ab

± ± ± ± ± ± ± ±

8.2 2.4 1.8 3.7 5.9 2.1 5.8 6.1

± ± ± ± ± ± ± ±

0.022 0.038 0.016 0.042 0.053 0.038 0.009 0.077

WF, wheat fiber; AF, apple fiber; PF, pea fiber. Numbers refer to sucrose reduction in %. Mean values (±standard deviation, n ¼ 4) in a column with different superscripts differ significantly (P < 0.05).

Fig. 1. Frequency dependency of batter stiffness at 25
C (left), stiffness development during simulated baking (middle), and frequency dependency of the stiffness of thermally treated batter at 100
C (right). Data are mean values (n ¼ 4), standard deviations are given in the middle plot. Symbol color: black, reference; dark gray, wheat fiber (WF) in water replaces 30% sucrose; light gray, WF in water replaces 60% sucrose; white, WF in water replaces 100% sucrose.

Table 2 Minimum and maximum values of the complex modulus G* and the respective temperatures T during batter heating in oscillatory experiments (f ¼ 1 Hz, heating rate ¼ 5 K min1). Codea Reference WF-30 AF-30 PF-30 WF-60 AF-60 PF-60 WF-100

G*min (Pa)b c

186.0 206.5c 252.5b 257.5b 234.5b 326.0a 330.0a 330.0a

± ± ± ± ± ± ± ±

11.3 3.5 3.5 29.0 9.2 14.1 4.2 11.3

T at G*min (
C) a

66.2 61.6b 59.1c 58.0c 58.6c 55.0d 54.5d 49.9e

± ± ± ± ± ± ± ±

0.0 0.7 0.0 0.0 0.8 0.0 0.7 1.4

G*max (kPa) c

17.3 18.3bc 18.9bc 19.5ab 19.8ab 19.9ab 18.4bc 21.2a

± ± ± ± ± ± ± ±

0.0 0.4 0.4 1.1 0.2 0.5 0.3 0.1

T at G*max (
C) 101.0a 93.4b 93.7ab 92.9b 86.5c 86.6c 86.7c 78.8d

± ± ± ± ± ± ± ±

0.0 0.4 0.0 0.4 0.1 0.1 0.7 0.9

a WF, wheat fiber; AF, apple fiber; PF, pea fiber. Numbers refer to sucrose reduction in %. b Mean values (±standard deviation, n ¼ 4) in a column with different superscripts differ significantly (P < 0.05).

role for sweet bakery products because a delay in starch gelatinization and protein denaturation supports air bubble expansion in the batter due to evaporation and carbon dioxide formation from baking powder before the structure sets and the air cell walls harden. The volume of the batter is at maximum when starch and protein denature almost simultaneously during baking (Donovan, 1977). Therefore, a right amount of sucrose ensures a sufficient product rise, a factor which has to be compensated when sucrose is replaced by non-sugar ingredients. The fiber specific effects on G*min, G*max and the respective temperatures are summarized in Table 2. There is no significant difference in G*min for AF-30% and PF-30% whereas G*min for WF30% is significantly lower. Since the stiffness minimum in the

temperature sweep is related to starch swelling it can be concluded that WF restrains water absorption of starch granules more than AF and PF, presumably because of a more intense competition for water between WF and starch granules. This observation is in accordance with the highest water/fiber ratio of WF for an isoviscous batter to the reference (Table 1), and might result from fiber composition (WF almost exclusively contained insoluble fiber). The same behavior was observed for 60% sugar reduction. As regards G*max, there is no significant effect in both sucrose replacement levels that may be attributed to fiber type. The delay in starch gelatinization temperature in our study is therefore mainly generated by the sucrose replacement rather than by fiber integration into the batter. Finally, G* measured at 100
C showed a significantly reduced dependency on frequency (Fig. 1). G* also decreased with increasing sucrose replacement, indicating that fiber addition causes a weaker gel structure after heating. tan d varied from 0.099 to 0.121 and was higher when more sucrose was replaced, indicating that reduced sugar batter exhibits a lower ability to retain gas due to rupture of cell membrane before setting of muffin structure. 3.3. Product characteristics It is evident from muffin geometrical properties (Table 3) that mainly product height is responsible for the significantly higher density of reduced sugar products. At both 30% and 60% sugar replacement WF performed better than AF and PF, presumably because of the high amount of insoluble fiber which provides more structure. It is also evident that mainly differences in muffin rise during baking contribute to variations in volume and density,

36

S. Struck et al. / LWT - Food Science and Technology 65 (2016) 32e38

Table 3 Characteristics of fiber enriched reduced sugar muffins. Codea

Height (mm)b

Diameter (mm)

Volume (mL)

Density (g cm3)

Water activity ()

Baking loss (%)

Crumb firmness (N)

Cohesiveness (N)

Reference WF-30 AF-30 PF-30 WF-60 AF-60 PF-60 WF-100

52.1a 45.7c 46.0b 45.1d 43.5e 43.6e 43.3f 36.1g

63.0a 62.4b 62.0c 62.5b 61.3d 61.2d 61.2d 59.3e

98.2a 82.6b 81.1c 81.2c 78.5d 77.5de 76.9e 65.5f

0.38f 0.45e 0.46d 0.46d 0.47c 0.48b 0.49b 0.57a

0.85g 0.91f 0.92d 0.91e 0.95b 0.95b 0.95c 0.98a

12.4d 13.9a 13.4b 13.5b 13.4b 13.0c 13.1c 13.0c

1.6e 2.4d 2.7cd 2.6cd 3.0bc 3.0bc 3.1b 5.4a

0.58a 0.51b 0.48b 0.51b 0.51b 0.47b 0.46b 0.50b

a b

± ± ± ± ± ± ± ±

0.58 0.39 0.26 0.25 0.19 0.18 0.25 0.25

± ± ± ± ± ± ± ±

0.34 0.18 0.31 0.31 0.11 0.15 0.08 0.13

± ± ± ± ± ± ± ±

1.31 0.36 1.15 0.89 0.63 0.87 0.60 1.50

± ± ± ± ± ± ± ±

0.01 0.00 0.01 0.01 0.00 0.01 0.00 0.01

± ± ± ± ± ± ± ±

0 0 0 0 0 0 0 0

± ± ± ± ± ± ± ±

0.38 0.21 0.36 0.35 0.36 0.38 0.48 0.53

± ± ± ± ± ± ± ±

0.29 0.29 0.43 0.33 0.22 0.25 0.41 0.48

± ± ± ± ± ± ± ±

0.01 0.08 0.06 0.05 0.05 0.04 0.05 0.08

WF, wheat fiber; AF, apple fiber; PF, pea fiber. Numbers refer to sucrose reduction in %. Mean values (±standard deviation, n ¼ 4) in a column with different superscripts differ significantly (P < 0.05).

whereas muffin diameter and baking loss showed reduced variation. Expressed as highest to lowest value, respective ratios are 1.44 (muffin height), 1.06 (diameter), and 1.12 (baking loss). Regardless of fiber type, average product height for 30% and 60% sucrose reduction was 45.6 ± 0.51 mm and 43.5 ± 0.39 mm, respectively. Sucrose free WF muffins had the lowest height and highest density because of earlier structure setting during baking, presumably caused by a lower starch gelatinization temperature. Generally, the volume of the final product depends on the amount of air incorporated in the batter and then on the ability to retain that air during baking (Frye & Setser, 1991), which in turn is influenced by destabilization mechanisms such as bubble rise, coalescence and disproportionation (Wilderjans, Luyts, Brijs, & Delcour, 2013). Water activity ranged between 0.85 and 0.98 and was highest for WF-100, resulting from high water level in the recipe and baking loss of 13%. Fig. 2A shows the interrelation (R2; ¼ 0.92, P < 0.05) between muffin height and stiffness of the heated batter (G* at 1 Hz and 100
C), and a clear trend towards lower values at higher sucrose replacement. A volume reduction in reduced sugar baked goods with high intensity sweeteners is already reported in literature (e.g., Manisha et al., 2012; Martínez-Cervera et al., 2012; Zahn et al., 2013). Lower stiffness of reduced sugar batters explains this effect, as the replacement of sucrose weakens air bubble stability and leads to an earlier rupture of their cell walls so that the gas holding ability during baking is reduced. It can be concluded that the temperature sweeps, and the subsequent frequency sweeps at 100
C reflect realistically batter behavior during baking so that they are helpful for simulating baking processes in small scale. Further advantages of rheological analyzes is small sample size and a reproducibility much better than that of cake baking experiments (Chesterton, Wilson, Sadd, & Moggridge, 2015). Product texture is also significantly (P < 0.05) affected by sucrose replacement: compared to the reference, crumb hardness increased, and cohesiveness decreased (Table 3). Crumb firmness is significantly (P < 0.05) interrelated with muffin density (Fig. 2B), indicating that the fiber affects air cell incorporation and contributes to mechanical resistance during compression. Arora and Camire (1994) showed comparable results for muffins with potato peels, and Sudha et al. (2007) observed a harder cake crumb with increasing amounts of incorporated apple pomace. In our case, the consequences on muffin texture were independent of fiber type. The effects of sucrose replacement on crust and crumb color are depicted in Fig. 3. Additionally, photographic images of muffin crust and crumb are shown in Fig. S1. For the reference, crust L* was 53.3. C* ¼ (a*2 þ b*2)0.5 was 39.7, and hab ¼ arctan (b* a*1) was 66.5
. Except for AF, L* of the crust increased with increasing replacement up to L* ¼ 71.3 for WF-100. Crust C* was significantly lower (P < 0.05) for muffins containing apple fiber, and WF-100 muffins showed the lowest saturation. With increasing fiber incorporation,

Fig. 2. Interrelation between selected properties of muffins with varying amount of sucrose replacement by fiber/water mixtures. AF, apple fiber; PF, pea fiber; WF, wheat fiber. Numbers refer to the amount of sucrose replacement (%).

hab shifted towards higher values and was approx. 85
for the WF100 muffin. The differences are caused by caramelization and Maillard reactions of reducing sugars (Nip, 2006), that result from thermal hydrolysis of sucrose (Ameur, Mathieu, Lalanne, Trystram, & Birlouez-Aragon, 2007). Because of much lower temperature in product center during baking, sucrose induced browning does not play a major role for crumb color formation. The majority of muffins including the reference was characterized by 66.2  L*  69.3, 21.3  C*  22.5 and 88.0
 hab  90.3
(Fig. 3). Crumb color coordinates and characteristics of AF muffins were however significantly different because of the beige-brown color of the fiber itself (WF and PF were

S. Struck et al. / LWT - Food Science and Technology 65 (2016) 32e38

37

measurements. 3.4. In vitro starch digestibility

Fig. 3. Color coordinates of the crust (gray circles) and crumb (white circles) of muffins with varying amount of sucrose replacement by fiber/water mixtures. AF, apple fiber; PF, pea fiber; WF, wheat fiber. Numbers refer to the amount of sucrose replacement (%). Lightness L* is proportional to circle area as indicated. Two values each for chroma C* and hue hab are also indicated.

whitish). Crumb L* of the muffins was 55.6 (AF-30) and 52.3 (AF60), crumb C* was 22.5 (AF-30) and 25.6 (AF-60), and crumb hue was 74.9
(AF-30) and 71.9
(AF-60). This is a difference that can clearly be noticed by the human eye (Sudha et al., 2007). Principal component analysis on the consensus matrix revealed that dimension 1 and dimension 2 account for 41.6% and 31.8% of total variance, respectively. In the GPA group average plot (Fig. 4), the reduced sugar muffins are described as being harder in crumb and crust. An accumulation of sweetness expressions was found for the reference, whereas expressions for pungent that were attributed to the fiber containing muffins may be related to the RebA. Finally, expressions referring to crumb and, partially, crust color are obviously in context with fiber color (dark for apple, almost white for pea and wheat), and reflect the results of analytical color

The reducing sugars released over time in in vitro starch digestion assays fitted to Eq. (2) resulted in coefficients summarized in Table 4. In vitro digestion was analyzed in all WF recipes to demonstrate the sugar reduction effect, and in all 30% reduced sucrose recipes to demonstrate the influence of the fiber. Coefficient c0 refers to reducing sugar concentration before digestion (264.9 mg g1 reference) decreased with increasing sucrose replacement to 28.9 mg g1 (WF-100%). At 30% reduction, AF muffins showed a significantly higher c0 than WF and PF products. This results from fiber composition (see Table 1): whereas AF contains approx. 38.44 g 100 g1 non-dietary fiber carbohydrates (mainly mono- and disaccharides), WF and PF has 1.1 and 20.7 g 100 g1, respectively. Consequently, WF-30 exhibited the lowest amount of reducing sugars before digestion. Coefficient c∞ represents the increase of reducing sugars after 120 min in vitro digestion. The final concentration is therefore c0 þ c∞, and the increase is independent of c0. In the WF muffins, c∞ decreased with increasing sucrose replacement, indicating that the amount of starch resistant to digestion is higher when more fiber is present. Upon incorporating guar galactomannan in bread, Brennan, Blake, Ellis, and Schofield (1996) observed a decreased starch hydrolysis and concluded that the fiber acts as a physical barrier for the digestion enzymes. At 30% sucrose reduction, the effect of fiber type was insignificant. The observation that a higher water holding capacity of a fiber reduces starch gelatinization (Symons & Brennan, 2004) was not verified in this study. The kinetic constant k was neither influenced by sucrose replacement level nor by fiber type. This is consistent with Gularte et al. (2012) who added inulin and oat fiber to gluten-free layer cakes. 4. Conclusions Dietary fiber application in sweet bakery products is associated with an increase in batter viscosity and reduced product volume caused by high water binding capacity of fibers. Iso-viscous batters in this study allowed a more detailed analysis of the temperatureinduced effects during baking of reduced sucrose batters with apple, wheat or pea fiber, and their effects on product characteristics. Sucrose reduction of 30% with application of fiber and RebA produced muffins with satisfactory product characteristics; here, the use of wheat fiber resulted in a volume and density closest to the reference muffin. Crumb firmness increased with the sucrose reduction level, as did crust color lightness. Crumb color was mainly influenced by fiber color. In vitro starch digestibility Table 4 Reducing sugars released during in vitro starch digestion over time fitted to a nonlinear model. Model parameters: c0, initial concentration of sugars; c∞, equilibrium concentration; k, kinetic constant. Codea Reference WF-30 AF-30 PF-30 WF-60 WF-100

Fig. 4. GPA group average plots for muffin descriptors. Muffin samples in the consensus space are reference R, and made by replacing 30% sucrose by apple fiber (AF), wheat fiber (WF), or pea fiber (PF).

c0 (mg g1)b a

264.9 193.6c 219.1b 201.6c 93.4d 28.9e

± ± ± ± ± ±

17.8 12.0 6.3 7.4 11.3 8.6

c∞ (mg g1) a

332.5 312.0ab 289.0b 295.3b 289.7b 236.6c

± ± ± ± ± ±

27.0 16.2 15.2 14.3 10.4 15.2

k (min1) 0.13a 0.10ab 0.09ab 0.10ab 0.08b 0.09ab

± ± ± ± ± ±

0.05 0.01 0.02 0.01 0.02 0.02

a WF, wheat fiber; AF, apple fiber; PF, pea fiber. Numbers refer to sucrose reduction in %. b Mean values (±standard deviation, n ¼ 4) in a column with different superscripts differ significantly (P < 0.05).

38

S. Struck et al. / LWT - Food Science and Technology 65 (2016) 32e38

decreased with increasing level of sucrose replacement. Energy density calculated from the ingredients diminished with an increasing amount of sugar reduction and fiber addition, and ranged between 379.4 kcal 100 g1 (reference) and 291.9 kcal 100 g1 (WF-100). The highest reduction of energy density of 23.1% was determined for WF-100%. Since the replacement of 30% sucrose with RebA and fiber gave satisfactory product characteristics, a reduction in energy density by 5.5% would be possible. Acknowledgment The fibers used in this study were kindly provided by J. €hne GmbH þ Co KG, Rosenberg, Germany. Rettenmaier & So Rebaudioside A was provided by Rudolf Wild GmbH & Co KG, Berlin, Germany, and egg powder by EISA-Sachsen GmbH, BrandErbisdorf, Germany. Appendix A. Supplementary data Supplementary data related to this article can be found at http:// dx.doi.org/10.1016/j.lwt.2015.07.053. References Ajila, C. M., Leelavathi, K., & Prasada Rao, U. J. S. (2008). Improvement of dietary fiber content and antioxidant properties in soft dough biscuits with the incorporation of mango peel powder. Journal of Cereal Science, 48, 319e326. Ameur, L. A., Mathieu, O., Lalanne, V., Trystram, G., & Birlouez-Aragon, I. (2007). Comparison of the effects of sucrose and hexose on furfural formation and browning in cookies baked at different temperatures. Food Chemistry, 101, 1407e1416. Arora, A., & Camire, M. E. (1994). Performance of potato peels in muffins and cookies. Food Research International, 27, 15e22. Baixauli, R., Sanz, T., Salvador, A., & Fiszman, S. M. (2008). Muffins with resistant starch: baking performance in relation to the rheological properties of the batter. Journal of Cereal Science, 47, 502e509. Bean, M. M., & Yamazaki, W. T. (1978). Wheat starch gelatinization in sugar solutions. I. Sucrose: microscopy and viscosity effects. Cereal Chemistry, 55, 936e944. Brennan, C. S., Blake, D. E., Ellis, P. R., & Schofield, J. D. (1996). Effects of guar galactomannan on wheat bread microstructure and on the in vitro and in vivo digestibility of starch in bread. Journal of Cereal Science, 24, 151e160. Brennan, M. A., Merts, I., Monro, J., Woolnough, J., & Brennan, C. S. (2008). Impact of guar and wheat bran on the physical and nutritional quality of extruded €rke, 60, 248e256. breakfast cereals. Starch e Sta Chen, H.-H., Kang, H.-Y., & Chen, S.-D. (2008). The effects of ingredients and water content on the rheological properties of batters and physical properties of crusts in fried foods. Journal of Food Engineering, 88, 45e54. Chen, J. Y., Piva, M., & Labuza, T. P. (1984). Evaluation of water binding capacity (WBC) of food fiber sources. Journal of Food Science, 49, 59e63. Chesterton, A. K. S., Wilson, D. I., Sadd, P. A., & Moggridge, G. D. (2015). A novel laboratory scale method for studying heat treatment of cake flour. Journal of Food Engineering, 144, 36e44. Donovan, J. W. (1977). A study of the baking process by differential scanning calorimetry. Journal of the Science of Food and Agriculture, 28, 571e578. Esteller, M. S., Amaral, R. L., & Lannes, S. C. S. (2004). Effect of sugar and fat replacers on the texture of baked goods. Journal of Texture Studies, 35, 383e393. Frye, A. M., & Setser, C. S. (1991). Optimizing texture of reduced-calorie yellow layer cakes. Cereal Chemistry, 69, 338e343. Grigelmo-Miguel, N., Carreras-Boladeras, E., & Martín-Belloso, O. (1999). Development of high-fruit-dietary-fibre muffins. European Food Research and Technology, 210, 123e128. mez, M., & Rosell, C. M. (2012). Effect of different Gularte, M. A., de la Hera, E., Go

fibers on batter and gluten-free layer cake properties. LWT e Food Science and Technology, 48, 209e214. Kaack, K., & Pedersen, L. (2005). Low energy chocolate cake with potato pulp and yellow pea hulls. European Food Research and Technology, 221, 367e375. Ktenioudaki, A., O'Shea, N., & Gallagher, E. (2013). Rheological properties of wheat dough supplemented with functional by-products of food processing: Brewer's spent grain and apple pomace. Journal of Food Engineering, 116, 362e368. Lebesi, D. M., & Tzia, C. (2011). Effect of the addition of different dietary fiber and edible cereal bran sources on the baking and sensory characteristics of cupcakes. Food and Bioprocess Technology, 4, 710e722. Manisha, G., Soumya, C., & Indrani, D. (2012). Studies on interaction between stevioside, liquid sorbitol, hydrocolloids and emulsifiers for replacement of sugar in cakes. Food Hydrocolloids, 29, 363e373. Martínez-Cervera, S., Salvador, A., Muguerza, B., Moulay, L., & Fiszman, S. M. (2011). Cocoa fibre and its application as a fat replacer in chocolate muffins. LWT e Food Science and Technology, 44, 729e736. Martínez-Cervera, S., Sanz, T., Salvador, A., & Fiszman, S. M. (2012). Rheological, textural and sensorial properties of low-sucrose muffins reformulated with sucralose/polydextrose. LWT e Food Science and Technology, 45, 213e220. Masoodi, F. A., Sharma, B., & Chauhan, G. S. (2002). Use of apple pomace as a source of dietary fiber in cakes. Plant Foods for Human Nutrition, 57, 121e128. recka, D. (2013). Mildner-Szkudlarz, S., Bajerska, J., Zawirska-Wojtasiak, R., & Go White grape pomace as a source of dietary fibre and polyphenols and its effect on physical and nutraceutical characteristics of wheat biscuits. Journal of the Science of Food and Agriculture, 93, 389e395. Miller, G. L. (1959). Use of dinitrosalicylic acid reagent for determination of reducing sugar. Analytical Chemistry, 31, 426e428. Mizukoshi, M. (1997). Baking mechanism in cake production. In S. E. Friberg, & K. Larsson (Eds.), Food emulsions (pp. 549e575). New York: Marcel Dekker. Nip, W.-K. (2006). Sweeteners. In Y. H. Hui (Ed.), Bakery products: Science and technology (pp. 137e159). Ames, Iowa: Blackwell. Nishida, C., Uauy, R., Kumanyika, S., & Shetty, P. (2004). The joint WHO/FAO expert consultation on diet, nutrition and the prevention of chronic diseases: process, product and policy implications. Public Health Nutrition, 7, 245e250. O'Shea, N., Doran, L., Auty, M., Arendt, E., & Gallagher, E. (2013). The rheology, microstructure and sensory characteristics of a gluten-free bread formulation enhanced with orange pomace. Food & Function, 4, 1856e1863. Peressini, D., Foschia, M., Tubaro, F., & Sensidoni, A. (2015). Impact of soluble dietary fibre on the characteristics of extruded snacks. Food Hydrocolloids, 43, 73e81. €hme, B., & Rohm, H. (2015). Development of formulations for reducedRiedel, R., Bo sugar and sugar-free agar-based fruit jellies. International Journal of Food Science and Technology. http://dx.doi.org/10.1111/ijfs.12787. Rohm, H., & Jaros, D. (1996). Colour of hard cheese. 1. Description of colour properties and effects of maturation. Zeitschrift für Lebensmittel-Untersuchung und -Forschung, 203, 241e244. Sanz, T., Salvador, A., & Fiszman, S. M. (2008). Evaluation of four types of resistant starch in muffin baking performance and relationship with batter rheology. European Food Research and Technology, 227, 813e819. Spies, R. D., & Hoseney, R. C. (1982). Effect of sugars on starch gelatinization. Cereal Chemistry, 59, 128e131. Struck, S., Jaros, D., Brennan, C. S., & Rohm, H. (2014). Sugar replacement in sweetened bakery goods. International Journal of Food Science & Technology, 49, 1963e1976. Sudha, M. L., Baskaran, V., & Leelavathi, K. (2007). Apple pomace as a source of dietary fiber and polyphenols and its effect on the rheological characteristics and cake making. Food Chemistry, 104, 686e692. Symons, L. J., & Brennan, C. S. (2004). The effect of barley b-glucan fiber fractions on starch gelatinization and pasting characteristics. Journal of Food Science, 69, FCT257eFCT261. Wilderjans, E., Luyts, A., Brijs, K., & Delcour, J. A. (2013). Ingredient functionality in batter type cake making. Trends in Food Science & Technology, 30, 6e15. Zahn, S., Forker, A., Krügel, L., & Rohm, H. (2013). Combined use of rebaudioside A and fibres for partial sucrose replacement in muffins. LWT e Food Science and Technology, 50, 695e701. Zahn, S., Pepke, F., & Rohm, H. (2010). Effect of inulin as a fat replacer on texture and sensory properties of muffins. International Journal of Food Science & Technology, 45, 2531e2537. Zoulias, E. I., Piknis, S., & Oreopoulou, V. (2000). Effect of sugar replacement by polyols and acesulfame-K on properties of low-fat cookies. Journal of the Science of Food and Agriculture, 80, 2049e2056.