Sweetness potency and sweetness synergism of sweeteners in milk and coffee systems

Food Research International 74 (2015) 168–176

Contents lists available at ScienceDirect

Food Research International journal homepage: www.elsevier.com/locate/foodres

Sweetness potency and sweetness synergism of sweeteners in milk and coffee systems Ji-hye Choi, Seo-jin Chung ⁎ Department of Nutritional Science and Food Management, Ewha Womans University, 52 Ewhayeodae-gil, Seodaemun-Gu, Seoul 120-750, South Korea

a r t i c l e

i n f o

Article history: Received 23 February 2015 Received in revised form 20 April 2015 Accepted 22 April 2015 Available online 28 April 2015 Keywords: Sweetener Concentration–response curve Sweetness potency Sweetener synergism Sensory analysis

a b s t r a c t This study investigated the presence of sweetness synergism in milk and instant coffee systems. It consists of three parts: 1) modeling concentration–sweetness intensity curves of sweeteners (stevia, sucralose, xylose, tagatose and erythritol); 2) measuring the sweetness potencies of sweeteners compared to sucrose at wide concentration range; and 3) investigating the presence of sweetness synergisms in binary sweetener mixtures. The panelists evaluated sweetness and other sensory characteristics of sweeteners using descriptive analysis. Based on the modeled curve derived from step 1, the concentration of each sweetener with sweetness intensity equal to 2.5% or 2.8% sucrose was calculated for milk and coffee systems, respectively. For the sweetness synergism study, one type of intense sweetener was mixed with one type of bulk sweetener, each eliciting 2.5% or 2.8% equi-sweetness to sucrose, and compared with 5% sucrose added to a milk system or 5.6% sucrose added to a coffee system. The sweetness potencies of bulk sweeteners generally increased whereas the sweetness potencies of intense sweeteners decreased as the concentration increased. The binary sweetener mixtures mostly showed additivity in milk and suppression in coffee system rather than synergism when the concentration dependent nature of sweetness potency for each sweetener was taken into account. © 2015 Elsevier Ltd. All rights reserved.

1. Introduction Taste synergism refers to eliciting higher taste intensity when two or more taste substances are mixed compared to the taste intensity of each individual substance summed together (Hutteau, Mathlouthi, Portmann, & Kilcast, 1998). The synergism present when MSG and IMP/GMP are mixed is one well-known example. Synergism between taste substances has commercial significance in food production, as reducing the amount of ingredients added to target food systems can reduce production costs (Lawless, 1998). Low-calorie sweeteners have been widely used for decades by consumers concerned about health and calories in their daily diet. Food product developers frequently apply multiple low-calorie sweeteners rather than use a single low-calorie sweetener in formulating low-calorie products. Synergism in sweetness and improvement of sensory quality are expected advantages of mixing sweeteners (Portmann & Kilcast, 1998). Studies have reported that using multiple sweeteners in a food system can suppress off-flavors (e.g., bitterness, astringencies) elicited by sweeteners such as stevia or erythritol (Belščak-Cvitanović et al.,

⁎ Corresponding author at: Department of Nutritional Science and Food Management, College of Health Science, Ewha Womans University, 52 Ewhayeodae-gil, SeodaemunGu, Seoul 120-750, South Korea. Tel.: +82 10 9108 7213; fax: +82 2 3277 2862. E-mail address: [email protected] (S. Chung).

http://dx.doi.org/10.1016/j.foodres.2015.04.044 0963-9969/© 2015 Elsevier Ltd. All rights reserved.

2015; Beyts & Latymer, 1985; Heikel, Krebs, Köhn, & Busch‐Stockfisch, 2012; Mona & Wafaa, 2005; Portmann & Kilcast, 1998; Wolf, Bridges, & Wicklund, 2010). Schiffman, Sattely-Miller, and Bishay (2007) showed that the time to reach maximum sweetness is reduced with multiple sweeteners compared to a single sweetener. Finding the optimal combination of low-calorie sweeteners can be an effective strategy to substitute sugar in foods. Stevia is a popular, natural low-calorie sweetener extracted from Stevia rebaudiana Bertoni leaves. Using stevia as a single sweetener in food is difficult due to sesquiterpene lactones responsible for the distinctive bitterness (Soejarto, Compadre, Medon, Kamath, & Kinghorn, 1983). Sucralose is an intense sweetener produced from chlorination of parts of sucrose with a relatively similar sweetness and timeintensity profile to that of sucrose (De, Medeiros, Bolini, André, & Efraim, 2007). Compared to intense sweeteners, carbohydrate-based low-calorie bulk sweeteners have better sensory quality at a higher sweetness level (Gwak, Chung, Kim, & Lim, 2012). Tagatose and xylose were recently introduced to Korean consumers as low-calorie bulk sweeteners. Tagatose is an isomer of galactose eliciting relatively high sweetness intensity (0.85–0.92 time sweetness to sucrose) compared to the sweetness potency of other low-calorie bulk sweeteners. It is relatively stable in heat and exhibits similar taste and texture characteristics to that of sucrose (Roh et al., 1999). Xylose, a material known to produce xylitol, has become an attractive low-calorie bulk sweetener due to its potential health functionality of lowering blood glucose

J. Choi, S. Chung / Food Research International 74 (2015) 168–176

level (Bae et al., 2011; Moon, Lee, Jung, Park, & Yang, 2012). Another low-calorie bulk sweetener erythritol has wide application in food systems with its stability in heat and acid and a relatively high osmotic pressure, making it an appealing material for pickling. Determining the sweetness potency of a sweetener is a critical first step to investigate the presence of sweetness synergisms. One of the 3 approaches is generally taken to determine the sweetness potency in synergism studies: 1) use the “general” sweetness potency reported in previous literature or the sweetness potency value provided by the sweetener producer (George, Arora, Wadhwa, Singh, & Sharma, 2010; Wolf et al., 2010); 2) measure the sweetness potency of a sweetener anew at a specific concentration level and food system using magnitude estimation, rating or two-alternative forced-choice method (Alcaire et al., 2014; Batterman, Beyts, & Lillard, 1995; Cardello, Da Silva, & Damasio, 1999; Heikel et al., 2012); or 3) utilize a concentration–response (C–R) curve of sweeteners to calculate the sweetness potency of a sweetener at a specific concentration level (Hutteau et al., 1998; Parke, Birch, Portmann, & Kilcast, 1999; Portmann & Kilcast, 1998; Schiffman, Sattely-Miller, Graham, Booth, & Gibes, 2000; Schiffman, Booth, Carr, Losee, Sattely-Miller, et al., 1995, 2007). Studies have shown that the sweetness potency of a sweetener depends on the concentration level of the sweetener, applied food matrix, and tasting condition such as temperature (Esmerino et al., 2013; Fry, Yurttas, & Biermann, 2011; Fujimaru, Park, & Lim, 2012; Moraes & Bolini, 2010; Paixão, Rodrigues, Esmerino, Cruz, & Bolini, 2014). Hence, modeling the C–R curve of a sweetener in a target food matrix would be an effective procedure to accurately determine the sweetness potency to be applied in a synergism study. Choi and Chung (2014) proposed a sucrose–sweetener combined (SSC) method, which can accurately model the C–R curve of sweeteners by minimizing contextual bias that may occur during the measurement. The SSC method essentially measures the sweetness intensities of target sweetener and sucrose at various concentration levels together in one test. Then the method models each of the C–R curves of the target sweetener and sucrose based on the sweetness intensity measured in the same test set. Finally, the sweetness potency at a specific sweetener level is calculated based on the two generated C–R curves. Thus, the sweetness potency of a sweetener in this respect is the relative sweetness to sucrose at various concentrations. Many sweetness synergism studies were carried out in an aqueous system (Cardello et al., 1999; Heikel et al., 2012; Hutteau et al., 1998; Parke et al., 1999; Portmann & Kilcast, 1998; Schiffman, Booth, Carret, al., 1995, 2000). More recently, efforts are being made to investigate the sweetness synergisms in a real food matrix, such as fruit drinks (Mona & Wafaa, 2005) and lassi (George et al., 2010). In the present study, skimmed milk and vegetable creamer added to instant coffee were used to investigate synergisms of different sweeteners since reducing sucrose levels in these systems are currently in demand among Korean consumers. One type of highly intense sweetener was mixed with one type of bulk sweetener (each eliciting 50% equi-sweetness to sucrose) and compared with sucrose added to skim milk or instant coffee. The overall experimental design is: 1. model concentration–sweetness intensity curves of sweeteners using the SSC method; 2. measure sweetness potency of sweeteners compared to sucrose at wide concentrations (milk system 1%, 2%, 3.5%, 5% and 7%; coffee system 0.9%, 2.3%, 3.7%, 5.6%, and 7.9% sucrose equivalent range); and 3. investigate the presence of sweetness synergism between two types of sweeteners based on the sweetness potency values calculated in step 2. The C–R curve of sweetener in a milk system using the SSC method was published by Choi and Chung (2014), whose results will be adapted to investigate the presence of synergism between sweeteners. Procedure and results on the C–R curve of sweeteners in milk were previously described in detail (Choi & Chung, 2014). Therefore, only a brief description of the method and results on sweeteners in milk will be provided in the present study.

169

2. Materials and methods 2.1. Stimuli Three types of bulk sweeteners (xylose, tagatose, erythritol) and two types of highly intense sweeteners (sucralose and enzyme-treated stevia) were sweeteners of interest. Sucrose was used as a control sample. All bulk sweeteners and stevia were purchased from CJ CheilJedang (Seoul, South Korea), and sucralose from ESFood (Anyang, Gyeonggi, South Korea). The samples were prepared by adding an adequate amount of sweetener to skimmed milk (Seoul Milk, Seoul, South Korea) or to vegetable cream (Dongsuh Food, Seoul, South Korea) added instant coffee (Dongsuh Food, Seoul, South Korea). 2.2. Panel Eight to ten panelists participated as descriptive panelists. All panelists were females with ages ranging from 20 to 25. They had previously participated and been trained to evaluate sweeteners in milk and coffee systems (Choi, Kim, & Chung, 2013). 2.3. Modeling concentration–sweetness intensity curves of sweeteners 2.3.1. Sample preparation In a previous study (Choi et al., 2013), the sweetness potency values of xylose, tagatose, erythritol, sucralose and stevia were 0.63, 0.85, 0.60, 556, and 25 times that of sucrose, respectively, when compared to 5% sucrose added to skim milk and 5.6% sucrose added to a coffee system. For the milk system, five concentration levels of each sweetener corresponding to the sweetness equivalent value (SEV) of 1%, 2%, 3.5%, 5%, and 7% sucrose were calculated based on the sweetness potency values mentioned above. Samples were prepared 4 h prior to the sensory evaluation session and served at room temperature (18 ± 5 °C) to the panelists. A 40 ml sample was served in a solo cup labeled with a random 3 digit code. For the coffee system, five concentration levels of each sweetener corresponding to the SEV of 0.9%, 2.3%, 3.7%, 5.6%, and 7.9% sucrose were calculated. Coffee was formulated by mixing 1.9% instant coffee, 5.6% vegetable cream, an adequate amount of sweetener, and boiling water. Samples were prepared 2 h prior to the evaluation and served at 80 ± 5 °C to the panelists. A 150 ml sample was served in a thermos (Zhejiang Wuyi Hongyun Cups Co., Zhejiang, China) labeled with a random 3 digit code. The panelists were instructed to pour the sample into a 50 ml ceramic cup, taste and evaluate. The concentration of sweeteners used to model the C–R curve in milk and coffee systems is shown in Table 1. 2.3.2. Sensory evaluation procedure The experiments on milk system and coffee system were conducted separately and independently. The sucrose–sweetener combined (SSC) method developed in a previous study (Choi & Chung, 2014) was used to model the C–R curve, as it was shown to minimize the context effect that may be present during the evaluation of the samples for C–R curve modeling. The SSC method measures the sweetness of 5 levels of a specific sweetener and 5 levels of sucrose in one test set (e.g., xylose 1.6%, 3.2%, 5.6%, 7.9%, 11.1% and sucrose 1%, 2%, 3.5%, 5% and 7% in the milk system). The regression models are produced based on the C–R curve measured for both sucrose and target sweetener. Then, sweetness potencies of a sweetener at various concentrations were calculated from the regression models of sweetener and sucrose. Reference standards for sweetness as well as other sensory attributes were used to train the panelists and evaluate the samples to accurately measure the sweetness intensity of the samples and avoid the dumping effect (Tables 2 and 3). The reference standards for sweetness intensity were provided at different levels (3%, 5%, 7% and 10% sucrose solution anchored as 4, 7, 10 and 14 points). The sensory intensity of

170

J. Choi, S. Chung / Food Research International 74 (2015) 168–176

Table 1 Concentration range1 of 5 sweeteners applied in the C–R modeling experiment. System

Sweetener concentration level

Sucrose

Xylose

Tagatose

Erythritol

Sucralose

Stevia

Milk

Level 1 Level 2 Level 3 Level 4 Level 5 Level 1 Level 2 Level 3 Level 4 Level 5

1% 2% 3.5% 5% 7% 0.9% 2.3% 3.7% 5.6% 7.9%

1.6% 3.2% 5.6% 7.9% 11.1% 1.4% 3.7% 5.9% 8.9% 12.5%

1.2% 2.4% 4.1% 5.9% 8.2% 1.1% 2.7% 4.4% 6.6% 9.3%

1.7% 3.3% 5.8% 8.3% 11.7% 1.5% 3.8% 6.2% 9.3% 13.1%

0.002% 0.004% 0.006% 0.009% 0.013% 0.002% 0.004% 0.007% 0.01% 0.014%

0.04% 0.08% 0.14% 0.2% 0.28% 0.04% 0.09% 0.15% 0.22% 0.32%

Coffee

1

Concentration of each sweetener in each sweetness level was calculated based on the sweetness potency reported in Choi et al. (2013).

the sample was rated on a 15-point scale numbered 0 to 14. All evaluations were repeated 3 times. Warm water at 55 ± 5 °C and unsalted crackers were provided between samples to cleanse the palate. The panelists were asked to rest for 6 min between samples. The serving order of the samples was determined by Williams Latin Square design (Williams, 1949). Two sessions were required to complete one test set, and a single session required approximately 1 to 1.5 h to complete. 2.3.3. Statistical analysis The sweetness potencies of sweeteners at various concentration levels were calculated using the method described by Choi and Chung (2014). The C–R curve of each sweetener was modeled separately for milk and coffee systems using regression analysis. Regression analysis was applied to the mean data of the samples varying in sweetener concentration. The C–R curve of sucrose was newly generated for different sweeteners, as the sweetness intensities of sucrose and sweeteners were measured concurrently. The sweetness potency of a sweetener calculated using two C–R curves generated from the same test set will have a value that accounts and calibrates for the potential context effect. The sweetener and sucrose concentrations that elicit equal sweetness intensity were interpolated, and that specific concentration ratio was calculated as the sweetness potency of the sweetener. Statistical analyses were performed using SPSS 21 (SPSS Inc., Chicago, IL, USA) and Microsoft Excel 2010 (Microsoft, Redmond, WA, USA). 2.4. Investigating the presence of sweetness synergisms between 2 types of sweetener 2.4.1. Operational definition of taste synergism Although the definition of taste synergism was mentioned in the first part of the Introduction, demonstrating the presence of synergism among sweeteners is a challenge since comparing the mixture intensity and the sum of individual intensities is experimentally not simple (Lawless, 1998). In the present study, sweetness intensities of various sweeteners were converted to sucrose equivalent values based on the relative sweetness of sweeteners to sucrose at various concentration levels. We operationally defined sweetness synergism as the intensity

of two sweetener mixtures, each sweetener at 50% level of sucrose, being higher than the intensity of 100% level of sucrose. The 50% level of sucrose is not concentration but intensity based. For example, synergism exists if the intensity of tagatose at 3% sucrose equivalent value (SEV) and sucralose at 3% SEV mixture shows higher sweetness intensity rating than 6% sucrose solution. Additivity is described as showing no significant difference between the sweetness intensities of 50% + 50% mixture and 100% sucrose. 2.4.2. Sample preparation For the sweetness synergism study, one type of intense sweetener was mixed with one type of bulk sweetener. Based on the modeled curves, the concentration level of each sweetener eliciting a sweetness intensity equal to 2.5% and 2.8% sucrose was calculated for milk and coffee systems, respectively. Two sweeteners, each eliciting 2.5% or 2.8% equi-sweetness to that of sucrose in milk or coffee, were mixed and compared with 5% or 5.6% sucrose added to milk or coffee samples. Sweetener mixture concentrations are shown in Table 4. 2.4.3. Sensory evaluation procedure Separate descriptive analyses were conducted on the milk and coffee samples to analyze the sensory characteristics and the presence of sweetness synergism in samples. The same attributes from previous experiments were used again (Tables 2 and 3). Samples were evaluated on a 0–14-point numerical scale and compared with the sweetness intensity of 5% sucrose in milk or 5.6% sucrose in coffee. The sample serving and evaluation methods were identical to the method described in Section 2.3.2. All evaluations were repeated 3 times. 2.4.4. Statistical analysis Data from the milk and coffee systems were analyzed separately. General linear model (GLM) analysis was conducted to evaluate the presence of significant differences in sensory attribute intensities of milk or coffee samples containing different sweetener combinations. The model applied was [intensity of sensory attribute = sweetener mixture + panel + sweetener mixture × panel]. Significance level was set at α = 0.05, and Duncan's multiple range test was applied in

Table 2 Sensory descriptors, definitions, reference materials, and reference scores of the descriptive attributes developed for milk system. Categories

Descriptors

Definitions

Reference

Reference score

Flavor/taste attributes

Overall intensity Sweetness

Intensity of overall flavor Fundamental taste sensation of which sucrose is typical

Bitterness Milk flavor Goso flavor

Fundamental taste sensation of which caffeine is typical Flavor associated with typical milk Complex flavor associated with mixture of dairy, fatty, nutty, and roasted carbohydrate flavor Flavor associated with low calorie beverages Flavor associated with whipped cream Aftertaste associated with sucrose Aftertaste associated with astringent

– Sugar solution 5% Sugar solution 7% Sugar solution 10% Caffeine solution 0.05% Milk (Seoul Milk Co., Korea) –

– 7 10 14 7 9 –

Coca-Cola zero (Coca-Cola Co., USA) Whipped cream (Mail Dairy Co., Korea) Sugar solution –

12 10

Residual flavor

Artificial sweetness Fresh cream flavor Residual sweetness Residual astringent

–

J. Choi, S. Chung / Food Research International 74 (2015) 168–176

171

Table 3 Sensory descriptors, definitions, reference materials, and reference scores of the descriptive attributes developed for coffee system. Categories

Descriptors

Definitions

Reference

Reference score

Flavor/taste attributes

Overall intensity Sweetness

Intensity of overall flavor Fundamental taste sensation of which sucrose is typical

Bitterness Sourness Coffee flavor Vegetable cream flavor Caramel flavor Astringent/tuptup

Fundamental taste sensation of which caffeine is typical Fundamental taste sensation of which citric acid is typical Flavor associated with coffee Flavor associated with vegetable cream Flavor associated with milk caramel Complex mouth feel associated with dry, roughness, and lingering residuals after swallowing in the mouth Degree of thickness of the fluid Aftertaste associated with sucrose Aftertaste associated with caffeine

– Sugar solution 3% Sugar solution 5% Sugar solution 7% Sugar solution 10% Caffeine solution 0.05% Citric acid solution 0.07% Coffee (Dongsuh Food Co., Korea) Vegetable cream (Dongsuh Food Co., Korea) Milk caramel (Orion Co., Korea) –

4 7 10 14 7 7 10 9 8 –

Texture/mouth feel attributes Residual flavor

Viscosity Residual sweetness Residual bitterness

post-hoc analysis when the sweetener mixture effect was significant. SPSS 21 software was used to analyze all data. 3. Results and discussion 3.1. Modeling concentration–sweetness intensity curves of sweeteners When the C–R curve was modeled for sweeteners in milk and coffee systems using regression analysis, most models showed a regression coefficient higher than 0.95, indicating a good fit (Figs. 1 and 2). In both milk and coffee systems, all bulk sweeteners showed a linear relationship between concentration and sweetness intensity, whereas intense sweeteners showed a logarithmic relationship. Using the regression model, the sweetness potencies of each sweetener at various concentration levels (1–7.5% SEV for milk, 0.9–7.9% SEV for coffee) were calculated and graphed in Fig. 3. The sweetness potency of bulk sweeteners (xylose, tagatose, erythritol) presented a tendency to increase as the concentration level increased. Erythritol showed the sharpest increase in sweetness potency among the bulk sweeteners. In the coffee system, sweetness potency of erythritol increased from 0.38 at 0.9% SEV to 0.59 at 7.9% SEV. The changes in sweetness potency of intense sweeteners exhibited different patterns from bulk sweeteners. In most cases, sucralose and stevia showed an increase in sweetness potency and reached a peak at a certain concentration level, then decreased as concentration increased. This result was similar to previous studies (Cardello et al., 1999; Schiffman, Booth, Losee, Pecore, & Warwick, 1995) that reported a decrease in sweetness potency at high sweetener concentration levels due to the significant increase of bitterness. For example, sucralose elicited a sweetness potency of 864 at 3.7% SEV but dropped to 491 at 7.9% SEV in the coffee system (Fig. 3).

– Sugar solution Caffeine solution 0.05%

– 7

The sweetness potency of the same types of sweetener was relatively similar between milk and coffee systems for bulk sweeteners but not for intense sweeteners. At a lower concentration range (1–2% SEV), the sweetness potency of stevia in the milk system was almost two times higher than that of the coffee system. In the case of sucralose at lower range, the sweetness potency was somewhat higher in coffee than milk system. Food system-dependent sweetness potency of sweetener has been observed in other studies (Kim et al., 2005; Moraes & Bolini, 2010). Protein and lactose in skim milk and polyphenols, caffeine, and fat in vegetable creamer added coffee are some possible components that can interact with sweeteners chemically or at perceptual level. Chemical interactions between stevia and other substances such as water soluble vitamins, organic acids or sweeteners have not been observed (Kroyer, 1999, 2010). The intrinsic bitterness of coffee or fat in added creamer may have suppressed the sweetness of stevia more than other sweeteners, resulting in a marked drop in sweetness potency in the coffee system. It has been reported that the sweetness potency of stevia dropped when creamer was added to instant coffee (Choi et al., 2013). Sweetness potency of sucralose was shown to be affected by the applied food matrix and temperature. Sweetness potency of sucralose decreased by the presence of fat (Paixão et al., 2014; Wiet, Ketelsen, Davis, & Beyts, 1993) and increased temperature (Paixão et al., 2014). Although care should be taken for direct comparison, the reported sweetener potency of sucralose in chocolate milk beverage (509) was lower than that of coffee (635.87) system (Moraes & Bolini, 2010; Paixão et al., 2014). The concentration and food systemdependent nature of sweetness potency shown in the present study imply that the sweetness potency of a sweetener should be measured anew for the appropriate concentration and food in which it is intended, rather than using a generic iso-sweetness potency.

Table 4 Concentration of each sweetener mixture used in synergy study. System

Sample

Set 1

Milk

Coffee

Sucrose Xylose Tagatose Erythritol Sucralose Stevia Sucrose Xylose Tagatose Erythritol Sucralose Stevia

2

3

4

5

6

7

5% 4.5%

4.5% 3.3%

3.3% 5% 0.003%

0.03%

0.03%

0.003%

5% 0.003%

0.03%

5.6% 5.2%

5.2% 4.0%

4.0% 5.4% 0.003%

0.057%

0.057%

0.057%

0.003%

5.4% 0.003%

172

J. Choi, S. Chung / Food Research International 74 (2015) 168–176

a

Sucrose-Tagatose (M)

14

Sweetness intensity

Sweetness intensity

b

Sucrose-Xylose (M)

14 12 10 8 6 4 2

12 10 8 6 4 2 0

0

0

1

2

3

4

5

6

7

0

8

1

2

Sucrose concentration(%)2 1y

sucrose sucrose

c

= 1.451x + 3.742 R² = 0.957

d

Sucrose-Erythritol (M)

Sweetness intensity

Sweetness intensity

5

6

7

8

12 10 8 6 4 2

tagatose tagatose

y = 1.178x + 3.623 R² = 0.973

Sucrose-Sucralose (M)

14

14

12 10 8 6 4 2 0

0 0

1

2

3

4

5

6

7

0

8

y = 1.439x + 3.752 R² = 0.973

sucrose sucrose

e

1

2

3

4

5

6

7

8

Sucrose concentration(%)2

Sucrose concentration(%)2 y = 0.895x + 2.951 R² = 0.975

erythritol erythritol

sucrose sucrose

y = 1.425x + 3.761 R² = 0.949

sucralose sucralose

y = 4.159ln(x) + 31.01 R² = 0.991

Sucrose-Stevia (M)

14

Sweetness intensity

4

y = 1.419x + 3.912 R² = 0.933

sucrose sucrose

y = 0.965x + 3.086 R² = 0.954

xylose xylose

3

Sucrose concentration(%)2

12 10 8 6 4 2 0

0

1

2

3

4

5

6

7

8

Sucrose concentration(%)2 sucrose sucrose

y = 1.364x + 3.798 R² = 0.979

stevia stevia

y = 2.213ln(x) + 14.87 R² = 0.881

Fig. 1. Concentration–response curve of sucrose–xylose (a), sucrose–tagatose (b), sucrose–erythritol (c), sucrose–sucralose (d), and sucrose–stevia (e) in milk system. 1x denotes the actual percent concentration of the target sweetener; y denotes the perceived sweetness intensity. 2 The actual sweetener concentration levels corresponding to the sucrose concentration levels are indicated in Table 1.

3.2. Investigating the presence of sweetness synergisms between 2 types of sweetener 3.2.1. Skim milk system Descriptive analysis was performed on skim milk samples with different sweetener combinations to investigate the presence of sweetness synergism between sweeteners and to characterize their sensory properties. The mean attribute intensities of each sample are shown in Table 5. The sweetness intensity of most sweetener combinations did not show a significant difference to sucrose control sample, except stevia–erythritol. Thus, additivity rather than synergism was observed for most sweetener interactions. Significant suppression interaction was observed in the stevia–erythritol combination, as its sweetness intensity was significantly lower than milk with sucrose. Samples showed

significant differences in the intensities of bitterness, goso flavor, artificial sweetness and residual astringency. In general, the types of intense sweetener affected the sensory characteristics of milk more than the types of bulk sweetener. That is, the sensory characteristics of the samples with sucralose were more similar to sucrose control samples than the samples with stevia. The samples with stevia elicited significantly higher bitterness, artificial sweetness and residual astringency but lower goso flavor than the samples with sucrose or sucralose. Stevia is known to carry a distinct bitterness (Prakash, DuBois, Clos, Wilkens, & Fosdick, 2008), which was not completely suppressed in the milk system and affected other sensory attributes. Overall, the sensory characteristics of milk samples with sucralose–bulk sweetener mixtures were more similar to the sucrose control sample than the milk samples with stevia–bulk sweetener mixtures. Studies have shown that

J. Choi, S. Chung / Food Research International 74 (2015) 168–176

a

Sucrose-Tagatose (C) 14

Sweetness intensity

Sweetness intensity

b

Sucrose-Xylose (C)

14 12 10 8

6 4 2

0

12 10 8 6 4 2 0

0

2

4

6

8

0

10

2

sucrose sucrose

1y

= 1.320x - 0.215 R² = 0.991

c

sucrose sucrose

y = 0.841x - 0.894 R² = 0.978

d

Sucrose-Erythritol (C)

6

8

10

12 10 8

6 4 2

y = 1.268x + 0.004 R² = 0.993

tagatose tagatose

y = 0.966x - 0.292 R² = 0.990

Sucrose-Sucralose (C)

14

Sweetness intensity

14

Sweetness intensity

xylose xylose

4

Sucrose concentration(%)2

Sucrose concentration(%)2

12 10 8

6 4 2

0

0 0

2

4

6

8

10

0

Sucrose concentration(%)2 sucrose sucrose

y = 1.335x - 0.471 R² = 0.980

e

4

6

8

10

Sucrose concentration(%)2 y = 0.855x - 1.308 R² = 0.981

erythritol erythritol

2

sucrose sucrose

y = 1.307x - 0.214 R² = 0.996

sucralose sucralose

y = 4.148ln(x) + 27.24 R² = 0.980

Sucrose-Stevia (C)

14

Sweetness intensity

173

12 10 8 6 4 2 0 0

2

4

6

8

10

Sucrose concentration(%)2 sucrose sucrose

y = 1.273x - 0.601 R² = 0.991

stevia stevia

y = 3.396ln(x) + 12.68 R² = 0.997

Fig. 2. Concentration–response curve of sucrose–xylose (a), sucrose–tagatose (b), sucrose–erythritol (c), sucrose–sucralose (d), and sucrose–stevia (e) in coffee system. 1x denotes the actual percent concentration of the target sweetener; y denotes the perceived sweetness intensity. 2 The actual sweetener concentration levels corresponding to the sucrose concentration levels are indicated in Table 1.

sucralose elicits sensory properties (i.e., similar sensory quality and time-intensity profile) similar to sucrose (Dutra & Bolini, 2013; Palazzo & Bolini, 2014; Palazzo, Carvalho, Efraim, & Bolini, 2011). This study confirmed that the close resemblance of sucralose and sucrose on sweetness quality is maintained when sucralose is mixed with other bulk sweeteners in a milk system. 3.2.2. Coffee system When GLM was applied to the descriptive analysis data of coffee system, the type of sweetener mixture significantly affected sweetness, sourness, vegetable cream flavor and residual sweetness. The mean intensity values of coffee samples with different sweetener combinations are presented in Table 6. Results showed that all sweetener combinations except sucralose–xylose exhibited significantly lower sweetness intensity than the sucrose control sample, implying that the suppression

effect was observed in most sweetener combinations. Sweetness intensity of the sucralose–xylose mixture did not significantly differ from the sucrose control sample, suggesting the presence of additivity. Synergisms between sweeteners were not observed in the milk or coffee systems when the concentration level of the sweetener applied to the sample was adjusted based on the changes of sweetness potency at different concentrations using C–R curves. Unlike the distinctive bitterness elicited by stevia–bulk sweetener mixture in the milk system, bitterness intensity did not significantly differ between samples in the coffee system. The intrinsic bitterness of stevia may have congruently blended with the bitterness present in coffee. Another observation different from the milk system was sweetener mixtures with sucralose tended to elicit significantly higher levels of sourness compared to other coffee samples. Stevia–bulk sweetener mixture samples were similar in sensory characteristics to coffee with sucrose, except sweetness and residual

174

J. Choi, S. Chung / Food Research International 74 (2015) 168–176

Relative sweetness

b

Xylose

0.9

Tagatose

0.9

0.8 0.7 0.6

milk

0.5

coffee

0.4

0.8

Relative sweetness

a

0.7 0.6

milk

0.5

coffee

0.4 0.3

0.3 0

2

4

6

8

0

10

Equi sweetness corresponding to sucrose concentration(%)

c

d

Erythritol

4

6

8

10

Sucralose 900

0.8 0.7 0.6

milk

0.5

cofffe

0.4

Relative sweetness

Relative sweetness

0.9

600

milk 300

0.3

coffee

0

0

2

4

6

8

0

10

Equi sweetness corresponding to sucrose concentration(%)

e

2

4

6

8

10

Equi sweetness corresponding to sucrose concentration(%)

Stevia

100

Relative sweetness

2

Equi sweetness corresponding to sucrose concentration(%)

75 50

milk

coffee

25 0

0

2

4

6

8

10

Equi sweetness corresponding to sucrose concentration(%) Fig. 3. Comparing the sweetness potency values of xylose (a), tagatose (b), erythritol (c), sucralose (d), and stevia (e) in milk and coffee systems.

sweetness. Sucralose–bulk sweetener mixture samples were similar in attributes to sucrose coffee sample, except sourness. As the results shown above, synergisms between binary sweetener combinations were not observed in either milk or coffee systems. This finding contradicts previous studies on synergism of sweetener mixtures. Batterman et al. (1995) observed synergisms of sweetness between sucralose and other sweeteners in tap water and soft drink systems. Synergism was reported when stevia was mixed with four

other sweeteners in fruit drinks (Mona & Wafaa, 2005). Mixtures of aspartame, acesulfame-K and sucralose in lassi and mixtures of aspartame and acesulfame-K in peach nectar showed synergism effects in sweetness and increased acceptance (George et al., 2010; Melo, Cardoso, Battochio, & Bolini, 2013). One of the main reasons of the contrasting results on the sweetness synergism between the current and previous studies is the difference in the sweetness potency value adapted for investigation. As mentioned

Table 5 The mean intensity values of attributes that showed significant sample effect for sweetener combinations in milk system.

1 2

Sample

Overall I

Sweet

Bitter

Milk F

Goso F

Artificial sweet

Fresh cream F

Res_sweet

Res_astringent

Sucrose Sucralose–xylose Sucralose–tagatose Sucralose–erythritol Stevia–xylose Stevia–tagatose Stevia–erythritol p-Value (sign.)

10.5 ± 2 10.7 ± 2 10.6 ± 1.6 10.8 ± 1.5 10.9 ± 2.1 11 ± 1.8 10.4 ± 1.7 0.802 ns2

9.8 ± 2.2b1 9.9 ± 1.9b 10 ± 1.5b 9.9 ± 1.8b 9.3 ± 2.8ab 9.6 ± 2.1b 8.8 ± 2.4a 0.034*

0.7 ± 0.8a 1.2 ± 1.5b 0.9 ± 1.2ab 1.1 ± 1.4ab 1.9 ± 1.5c 1.8 ± 1.6c 1.9 ± 1.4c 0.000***

4.5 ± 2 5±2 4.8 ± 1.9 4.5 ± 1.8 4.1 ± 2 4.2 ± 2 4.1 ± 1.9 0.181 ns

3.9 ± 2.3abc 4.5 ± 2.2c 4.2 ± 2.6bc 4.4 ± 2.4c 3.7 ± 2.3ab 3.6 ± 2.5ab 3.4 ± 2.7a 0.002**

4.8 ± 3.1ab 4.6 ± 2.7a 4.9 ± 2.7ab 5 ± 2.9ab 6.1 ± 2.8c 6 ± 2.8c 5.6 ± 2.3bc 0.011*

3.3 ± 2.4 3.4 ± 2.3 3.4 ± 2.3 2.9 ± 2.4 2.9 ± 2.4 3.1 ± 2.5 2.9 ± 2.5 0.156 ns

6.8 ± 3.3 6.8 ± 3.2 7.1 ± 3.1 6.9 ± 3.1 6.7 ± 2.8 7±3 6.3 ± 2.8 0.595 ns

1.7 ± 1.6a 1.7 ± 1.6a 2 ± 1.8ab 1.6 ± 1.7a 2.5 ± 2.1b 2 ± 1.5ab 2 ± 1.7ab 0.006**

Duncan's test: different letters in superscripts indicate significant differences within a column. * indicates significance at p-value b 0.05, ** indicates significance at p-value b 0.01, *** indicates significance at p-value b 0.001, ns indicates no significant difference.

J. Choi, S. Chung / Food Research International 74 (2015) 168–176

175

Table 6 The mean intensity values of attributes that showed significant sample effects for sweetener combinations in coffee system.

1 2

Sample

Overall I

Sweet

Bitter

Sour

Vegetable cream F

Caramel F

Coffee F

Astringent

Viscosity

Re_sweet

Re_bitter

Sucrose Sucralose–xylose Sucralose–tagatose Sucralose–erythritol Stevia–xylose Stevia–tagatose Stevia–erythritol p-Value (sign.)

9 ± 1.6 8.8 ± 1.8 8.7 ± 1.8 8.5 ± 1.9 8.3 ± 1.6 8.5 ± 1.9 8.4 ± 1.8 0.364 ns

7.1 ± 1.9c1 6.7 ± 1.7bc 6 ± 1.8ab 6 ± 2.3ab 5.8 ± 1.8ab 6 ± 2ab 5.7 ± 2a 0.02*2

5.7 ± 2.5 5.8 ± 2.2 5.9 ± 2.2 5.7 ± 2.2 6 ± 1.8 5.6 ± 2.3 5.9 ± 2 0.834 ns

2.1 ± 1.4a 2.9 ± 1.4b 2.9 ± 1.9b 2.3 ± 1.8a 2.6 ± 1.6ab 2.5 ± 1.7ab 2.5 ± 1.5ab 0.039*

6 ± 2.1ab 6.4 ± 1.8b 5.9 ± 2.2ab 6.1 ± 2.1ab 5.8 ± 2.1a 5.8 ± 2.1a 5.6 ± 2.3a 0.046*

4.3 ± 2.1 3.9 ± 2.3 3.7 ± 2.2 3.8 ± 2.3 3.6 ± 2.3 3.6 ± 2 3.4 ± 2.2 0.053 ns

5.9 ± 1.9 6.1 ± 1.7 6.1 ± 1.6 6.3 ± 1.6 5.9 ± 1.7 6 ± 1.8 5.7 ± 1.9 0.445 ns

6.2 ± 2.1 6.1 ± 2.1 6 ± 2.2 5.7 ± 2.1 5.5 ± 2 5.6 ± 2.2 5.8 ± 2.2 0.301 ns

5.5 ± 1.8 5.3 ± 2.1 5.1 ± 2.1 4.9 ± 1.9 5.2 ± 1.8 4.8 ± 1.9 4.7 ± 2 0.062 ns

5.1 ± 2.3bc 5.2 ± 2.3bc 4.4 ± 2.6ab 4.2 ± 2.4a 4 ± 1.9a 4.3 ± 2.3a 4.1 ± 2.1a 0.002**

5.2 ± 2.6 5.2 ± 2.2 5.5 ± 2.1 5.4 ± 2.1 5.2 ± 2 5.1 ± 2.3 5.3 ± 2.2 0.858 ns

Duncan's test: different letters in superscripts indicate significant differences within a column. * indicates significance at p-value b 0.05, ** indicates significance at p-value b 0.01, *** indicates significance at p-value b 0.001, ns indicates no significant difference.

previously, the determination method of sweetness potency is a critical first step in a sweetness synergism experiment. Different interpretations of synergism can result if one assumes that the sweetness potency of a sweetener is identical throughout a wide range of concentrations, which was not the case in the present study, and uses this “universal” sweetness potency value for the sweetness synergism experiment. For example, a researcher may randomly pick and measure the sweetness potencies of sucralose and stevia at 6% SEV and assumes that these potency values are universal for the two sweeteners. These values will then be applied to investigate sweetness synergism by mixing one half concentration of both sucralose 6% SEV + stevia 6% SEV and compared with a 6% sucrose sample. It is highly likely that the sweetness intensity of the mixture will be significantly higher than the sucrose sample since the sweetness potency of the two sweeteners at 3% SEV level is in fact higher than the 6% SEV level. As shown in Fig. 3, the sweetness potency of intense sweeteners changes drastically with concentration. The hypothetical sweetness synergism observed in the above example can be explained by an increase in sweetness due to an actual increase in sweetness potency of sweeteners at lower concentration levels. In the present study, additivity or suppression instead of synergism between sweetener mixtures was observed in both milk and coffee systems when the concentration dependent nature of sweetness potency was considered. An interesting hypothesis from the results of the present study can be suggested. If the observed synergisms in the previous studies are mainly due to the increase in individual sweetener's sweetness potency, plotting the changes of sweetness potency in broad concentration level as such in Fig. 3 can be useful to identify the most cost effective sweetener combination. Simply choosing the SEV level of sweeteners that exhibit the highest sweetness potency will result in the most effective sweetness eliciting combination. So for tagatose–sucralose pair in milk system, 7% SEV tagatose (sweetness potency of 0.81) and 3.5% SEV sucralose (sweetness potency of 739) may show maximal efficiency in eliciting sweetness. However, further validation is required for this hypothesis. 4. Conclusion This study explored the presence of synergism between low-calorie sweeteners by applying sweetness potency values obtained from a C–R curve measured by the SSC method. A C–R curve measured by the SSC method showed that the sweetness potency of sweeteners is concentration-dependent. For each sweetener, the concentration level that shows high sweetness potency has been identified. Thus, these concentration levels are suggested for the sweeteners to bring out their maximal efficiencies when applying them to skim milk or cream added instant coffee system. Synergisms between sweeteners were not observed in coffee or milk systems when the sweetness potency value of sweeteners used in the synergism study was adjusted for applied concentration levels and context effect. There is also a possibility that the previously reported sweetener synergisms are due to a possible increase in sweetness potency of individual sweeteners at lower

concentration levels. Based on the present study, sucralose with any of the low calorie bulk sweetener (xylose, tagatose, and erythritol) showed relatively similar sensory characteristics to that of sucrose added milk sample. In coffee system, sensory characteristics of sucralose–xylose combination were compatible to that of sucrose. Our investigation was limited to binary combinations of intense and bulk sweeteners. The presence of synergism in tertiary or quaternary combinations of various low-calorie sweeteners should be looked into in the future to find an ideal sweetener combination to substitute sucrose in food system. Acknowledgment This work was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT, and Future Planning (NFR 2011-0014820). References Alcaire, F., Zorn, S., Cadena, R.S., AntÚnez, L., Vidal, L., Giménez, A., et al. (2014). Application of survival analysis to estimate equivalent sweet concentration of low-calorie sweeteners in orange juice. Journal of Sensory Studies, 29(6), 474–479. Bae, Y.J., Bak, Y. -K., Kim, B., Kim, M. -S., Lee, J. -H., & Sung, M. -K. (2011). Coconut-derived D-xylose affects postprandial glucose and insulin responses in healthy individuals. Nutrition Research and Practice, 5(6), 533–539. Batterman, C. K., Beyts, P. K., & Lillard, D. W. (1995). Sucralose compositions: Google patents. Belščak-Cvitanović, A., Komes, D., Dujmović, M., Karlović, S., Biškića, M., Brnčić, M., et al. (2015). Physical, bioactive and sensory quality parameters of reduced sugar chocolates formulated with natural sweeteners as sucrose alternatives. Food Chemistry, 167(15), 61–70. Beyts, P. K., & Latymer, Z. (1985). Sweetening agents containing chlorodeoxysugar: Google patents. Cardello, H., Da Silva, M., & Damasio, M. (1999). Measurement of the relative sweetness of stevia extract, aspartame and cyclamate/saccharin blend as compared to sucrose at different concentrations. Plant Foods for Human Nutrition, 54(2), 119–129. Choi, J. h., & Chung, S.J. (2014). Optimal sensory evaluation protocol to model concentration–response curve of sweeteners. Food Research International, 62, 886–893. Choi, J.H., Kim, K.P., & Chung, S.J. (2013). Relative sweetness and sweetness quality of low calorie sweeteners in milk and coffee model system. Korean Journal of Food Science and Technology, 45(6), 754–762. De, M., Medeiros, L.L.M., Bolini, H., André, M., & Efraim, P. (2007). Equisweet milk chocolates with intense sweeteners using time-intensity method. Journal of Food Quality, 30(6), 1056–1067. Dutra, M.B.D.L., & Bolini, H.M.A. (2013). Sensory and physicochemical evaluation of acerola nectar sweetened with sucrose and different sweeteners. Food Science and Technology (Campinas), 33(4), 612–618. Esmerino, E.A., Cruz, A.G., Pereira, E.P.R., Rodrigues, J.B., Faria, J.A.F., & Bolini, H.M.A. (2013). The influence of sweeteners in probiotic Petit Suisse cheese in concentrations equivalent to that of sucrose. Journal of Dairy Science, 96, 5512–5521. Fry, J.C., Yurttas, N., & Biermann, K.L. (2011). Sweetness concentration–response behavior of Rebiana at room and refrigerator temperatures. Journal of Food Science, 76(9), S545–S548. Fujimaru, T., Park, J.H., & Lim, J. (2012). Sensory characteristics and relative sweetness of tagatose and other sweeteners. Journal of Food Science, 77(9), S323–S328. George, V., Arora, S., Wadhwa, B.K., Singh, A.K., & Sharma, G.S. (2010). Optimisation of sweetener blends for the preparation of lassi. International Journal of Dairy Technology, 63(2), 256–261. Gwak, M.J., Chung, S.J., Kim, Y.J., & Lim, C.S. (2012). Relative sweetness and sensory characteristics of bulk and intense sweeteners. Food Science and Biotechnology, 21(3), 889–894.

176

J. Choi, S. Chung / Food Research International 74 (2015) 168–176

Heikel, B., Krebs, E., Köhn, E., & Busch‐Stockfisch, M. (2012). Optimizing synergism of binary mixtures of selected alternative sweeteners. Journal of Sensory Studies, 27(5), 295–303. Hutteau, F., Mathlouthi, M., Portmann, M., & Kilcast, D. (1998). Physicochemical and psychophysical characteristics of binary mixtures of bulk and intense sweeteners. Food Chemistry, 63(1), 9–16. Kim, M., Cho, H., Park, J., Lee, S., Suh, D., Chung, S., et al. (2005). Relative sweetness of sucralose in beverage systems and sensory properties of low calorie beverages containing sucralose. Korean Journal of Food Science and Technology, 37(3), 425–430. Kroyer, G.T. (1999). The low calorie sweetener stevioside: Stability and interaction with food ingredients. LWT — Food Science and Technology, 32(8), 509–512. Kroyer, G. (2010). Stevioside and stevia–sweetener in food: Application, stability and interaction with food ingredients. Journal für Verbraucherschutz und Lebensmittelsicherheit, 5(2), 225–229. Lawless, H.T. (1998). Theoretical note: Tests of synergy in sweetener mixtures. Chemical Senses, 23(4), 447–451. Melo, L., Cardoso, J.M.P., Battochio, J.R., & Bolini, H.M.A. (2013). Using response surface methodology and high-intensity sweeteners' positive synergy to optimize peach nectar acceptability. Food and Nutrition Sciences, 4, 503–509. Mona, I.M., & Wafaa, A.A. (2005). Synergistic effects of some alternative sweeteners on the unpleasant attributes of stevia sweetener and its application in some fruit drinks. Food Science & Technology, 2(2), 1–10. Moon, S.H., Lee, K.S., Jung, S.W., Park, Y.J., & Yang, C.K. (2012). Study on the proper dxylose concentration in sugar mixture to reduce glycemic index (GI) value in the human clinical model. Korean Journal of Food Science and Nutrition, 25(4), 787–792. Moraes, P.C.B.T., & Bolini, H.M.A. (2010). Different sweeteners in beverages prepared with instant and roasted ground coffee: Ideal and equivalent sweetness. Journal of Sensory Studies, 25(s1), 215–225. Paixão, J.A., Rodrigues, J.B., Esmerino, E.A., Cruz, A.G., & Bolini, H.M.A. (2014). Influence of temperature and fat content on ideal sucrose concentration, sweetening power, and sweetness equivalence of different sweeteners in chocolate milk beverage. Journal of Dairy Science, 97(12), 7344–7353. Palazzo, A., & Bolini, H. (2014). Multiple time-intensity analysis: Sweetness, bitterness, chocolate flavor and melting rate of chocolate with sucralose, rebaudioside and neotame. Journal of Sensory Studies, 29(1), 21–32.

Palazzo, A., Carvalho, M., Efraim, P., & Bolini, H. (2011). The determination of isosweetness concentrations of sucralose, rebaudioside and neotame as sucrose substitutes in new diet chocolate formulations using the time‐intensity analysis. Journal of Sensory Studies, 26(4), 291–297. Parke, S.A., Birch, G.G., Portmann, M.O., & Kilcast, D. (1999). A study of the solution properties of selected binary mixtures of bulk and intense sweeteners in relation to their psychophysical characteristics. Food Chemistry, 67(3), 247–259. Portmann, M.O., & Kilcast, D. (1998). Descriptive profiles of synergistic mixtures of bulk and intense sweeteners. Food Quality and Preference, 9(4), 221–229. Prakash, I., DuBois, G., Clos, J., Wilkens, K., & Fosdick, L. (2008). Development of rebiana, a natural, non-caloric sweetener. Food and Chemical Toxicology, 46(7), S75–S82. Roh, H.J., Kim, S.Y., Kim, S.S., Oh, D.K., Han, K.Y., & Noh, B.S. (1999). Physicochemical properties of a low calorie sweetener, tagatose. Korean Journal of Food Science and Technology, 31(1), 24–29. Schiffman, S.S., Booth, B., Carr, B., Losee, M., Sattely-Miller, E., & Graham, B. (1995a). Investigation of synergism in binary mixtures of sweeteners. Brain Research Bulletin, 38(2), 105–120. Schiffman, S.S., Booth, B.J., Losee, M.L., Pecore, S.D., & Warwick, Z.S. (1995b). Bitterness of sweeteners as a function of concentration. Brain Research Bulletin, 36(5), 505–513. Schiffman, S.S., Sattely-Miller, E.A., & Bishay, I.E. (2007). Time to maximum sweetness intensity of binary and ternary blends of sweeteners. Food Quality and Preference, 18(2), 405–415. Schiffman, S.S., Sattely-Miller, E.A., Graham, B.G., Booth, B.J., & Gibes, K.M. (2000). Synergism among ternary mixtures of fourteen sweeteners. Chemical Senses, 25(2), 131–140. Soejarto, D., Compadre, C., Medon, P., Kamath, S., & Kinghorn, A. (1983). Potential sweetening agents of plant origin. II. Field search for sweet-tasting Stevia species. Economic Botany, 37(1), 71–79. Wiet, S.G., Ketelsen, S.M., Davis, T.R., & Beyts, P.K. (1993). Fat concentration affects sweetness and sensory profiles of sucrose, sucralose, and aspartame. Journal of Food Science, 58(3), 599–602. Williams, E. (1949). Experimental designs balanced for the estimation of residual effects of treatments. Australian Journal of Chemistry, 2(2), 149–168. Wolf, P., Bridges, J., & Wicklund, R. (2010). Application of agonist-receptor modeling to the sweetness synergy between high fructose corn syrup and sucralose, and between high-potency sweeteners. Journal of Food Science, 75(2), S95–S102.