DISCRIMINATION AND SENSORY CHARACTERIZATION OF STEVIOL GLYCOSIDES 

(REBAUDIOSIDE A, D, AND M) BY CONSUMERS AND ELECTRONIC TONGUE 

 
By 
 

Ran Tao 

 
 
 
 
 

 
 
 
 
 
 
 
 
 
 
 
 
 

 

 

 

A THESIS 

Submitted to 

Michigan State University 

in partial fulfillment of the requirements 

for the degree of 

Food Science – Master of Science 

2020 

 

 

ABSTRACT 

DISCRIMINATION AND SENSORY CHARACTERIZATION OF STEVIOL GLYCOSIDES 

(REBAUDIOSIDE A, D, AND M) BY CONSUMERS AND ELECTRONIC TONGUE 

 
By 

Ran Tao 

Rebaudioside (Reb) D and M are the recent focus of the food industry to address the 

undesirable aftertaste of Reb A, which is the most commonly used steviol glycoside in natural 
sweetener stevia. The first study evaluated the sensory characteristics of Reb A, D, and M, 
compared to 14% (w/v) sucrose, using a consumer panel and explored the relationship between 
6-n-Propylthiouracil (PROP) taster status (i.e., non-tasters, medium tasters, supertasters) and the 
perceived intensity of sweet and bitter tastes of the three steviol glycosides. The results showed 
that Reb D and M had sensory profiles that were closer to sucrose, compared to Reb A, but were 
associated with negative sensation, such as artificial, and Reb M was higher in lingering 
sweetness than sucrose (P < 0.001), which may cause negative perception toward Reb D and M. 
No significant differences were found among the PROP taster groups on the perceived sweetness 
and bitterness of Reb A, D, and M, suggesting that supertasters may not report aversive 
sensations from stevia. The second study was aimed to develop a new protocol for the electronic 
tongue (E-tongue), which is an analytical instrument for the sensory evaluation of taste, to 
discriminate stevia leaves for its potential use in stevia breeding programs as a method of testing 
flavor quality. With the new protocol, the E-tongue successfully separated Reb A, D, and M and 
discriminated among stevia leaf samples. This suggests that E-tongue has the potential to be used 
in stevia breeding programs for flavor selection. 

 
 

 

 

 

 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

This thesis is dedicated to my parents, Xiaodong Tao and Jie Gu.  

Thank you for the love and support.

iii 

 

 

ACKNOWLEDGEMENTS 

 

I am extremely grateful for having Dr. Sungeun Cho as my advisor. Thank you for your 

guidance and support in these two years, which allowed me to explore different research topics 

and gain invaluable experiences through research and extracurricular activities. I would also like 

to thank Dr. Zeynep Ustunol for being my second advisor and Dr. Bridget Behe for serving as 

my guidance committee. Your professionalism and expertise in the field have helped me improve 

my research a lot and also thank you for the personal support when Dr. Cho is away.  

 

Additionally, I would like to thank Dr. Alisa Doan, who provided me insights from the 

sensory industry and was very supportive of my research. I would also like to thank all of those 

who helped along with the development of TryDough, which has been the highlight of my 2 

years at the MSU. I am especially grateful to Edward Szczygiel, who guided me through the IFT 

product development competition; Dr. Cho, Dr. Janice Harte, and Dr. Deirdre Ortiz for their 

expertise in the field; Dr. Jeffery Swada, who helped me with the patent filing; Jason Hofman 

and Tina Conklin, who made the scale-up TryDough kits possible during this difficult pandemic 

time. To Karl and Kaylan, thank you for being my best teammates ever. Our product could not 

make this far without your creativity, passion, and commitment.  

 

Finally, I would like to thank my parents for supporting me to study abroad and pursue 

my passion. Thank you for the unconditional love and encouragement. To my wonderful lab 

mates, Ed and Shelby, thank you for the accompany during my first year of graduate school. I am 

also very grateful to Srishti, Maisie, Ellie, and Hampton for assisting with my research. 

 

 

 

iv 

 

TABLE OF CONTENTS 

 
LIST OF TABLES ...................................................................................................................... vii 
 
LIST OF FIGURES ................................................................................................................... viii 
 
INTRODUCTION ......................................................................................................................... 1 
1. Background ............................................................................................................................ 1 
2. Hypotheses and Objectives .................................................................................................... 3 
2.1. Hypotheses ................................................................................................................... 3 
2.2. Objectives .................................................................................................................... 3 
REFERENCES .............................................................................................................................. 5 
 
CHAPTER 1: LITERATURE REVIEW .................................................................................. 10 
1. Stevia .................................................................................................................................... 10 
1.1. History and Regulation .............................................................................................. 10 
1.2. Metabolism, Chemical Structure and Sweet Potency ................................................ 10 
1.3. Taste Profile of Steviol Glycosides ............................................................................ 13 
1.4. Next Generation Steviol Glycosides (Reb D & M) ................................................... 13 
1.5. Food Applications ...................................................................................................... 15 
2. Sensory Evaluation ............................................................................................................... 16 
2.1. Consumer Tests .......................................................................................................... 16 
2.2. Descriptive Analysis .................................................................................................. 16 
3. Electronic Tongue ................................................................................................................ 17 
3.1. Electronic Tongue Components ................................................................................. 17 
3.2. Benefits and Limitations ............................................................................................ 18 
3.3. Food Applications ...................................................................................................... 19 
4. PROP (6-n-Propylthiouracil) Taster Status .......................................................................... 19 
REFERENCES ............................................................................................................................ 21 
 
CHAPTER 2: CONSUMER-BASED SENSORY CHARACTERIZATION OF STEVIOL 
GLYCOSIDES (REBAUDIOSIDE A, D, AND M) .................................................................. 30 
Abstract .................................................................................................................................... 30 
1. Introduction .......................................................................................................................... 31 
2. Materials and Methods ......................................................................................................... 33 
2.1. Materials .................................................................................................................... 33 
2.2. PROP Status Determination ....................................................................................... 34 
2.3. Subjects Demographics .............................................................................................. 35 
2.4. Consumer Testing ...................................................................................................... 36 
2.5. Statistical Analyses .................................................................................................... 37 
3. Results .................................................................................................................................. 38 
3.1. Participant Characteristics ......................................................................................... 38 
3.2. Sensory Characteristics .............................................................................................. 40 
3.3. PROP Bitterness ......................................................................................................... 43 
4. Discussion ............................................................................................................................ 45 
5. Conclusions .......................................................................................................................... 50 

 

v 

Author Contributions ............................................................................................................... 51 
Funding .................................................................................................................................... 51 
Acknowledgments .................................................................................................................... 52 
Conflicts of Interest .................................................................................................................. 52 
REFERENCES ............................................................................................................................ 53 
 
CHAPTER 3: NEW PROTOCOL FOR THE #6 SENSOR ARRAY OF ASTREE 
ELECTRONIC TONGUE TO DISCRIMINATE STEVIA REBAUDIANA LEAVES ........ 60 
Abstract .................................................................................................................................... 60 
1. Introduction .......................................................................................................................... 61 
2. Samples and Sample Preparation ......................................................................................... 64 
2.1. Part 1: Protocol Validation ......................................................................................... 64 
2.2. Part 2: Application for Stevia Leaves in Breeding Programs .................................... 65 
3. E-Tongue Analysis ............................................................................................................... 66 
3.1. Analysis Sequence ..................................................................................................... 66 
3.2. E-Tongue Data Analysis ............................................................................................ 67 
3.3. Statistical Analysis ..................................................................................................... 68 
4. Validation of Analytical Protocol ........................................................................................ 68 
4.1. Sensor Stability .......................................................................................................... 68 
4.2. Discrimination of Reb A, D, and M ........................................................................... 71 
5. Discrimination of Stevia Leaf Samples ................................................................................ 73 
5.1. Fort Valley State University Samples ........................................................................ 73 
5.2. Michigan State University Samples ........................................................................... 73 
5.3. Future Potential of E-tongue for Flavor Selection in Stevia Breeding Program ....... 75 
6. Conclusions .......................................................................................................................... 76 
REFERENCES ............................................................................................................................ 77 
 
CONCLUSIONS ......................................................................................................................... 81 

 

 

 

 

 

vi 

 

LIST OF TABLES 

Table 1. R-groups, molecular formulas, molecular weights and potencies of the Stevia 
sweeteners. (Prakash et al., 2014) ................................................................................................. 12 
 
Table 2. Current Reb D and M products. ..................................................................................... 14 
 
Table 3. Reported maximum use-level of steviol glycosides (FAO, 2016) ................................. 15 
 
Table 4. Socio-demographic characteristics of participants (n = 126). ........................................ 39 
 
Table 5. Participants’ sweetener consumption behavior (n = 126). ............................................. 40 
 
Table 6. Mean intensity scores (± SEM) of sweetener solutions by participants (n=126). ......... 41 
 
Table 7. Total counts of check-all-that-apply attributes for sweetener solutions. ....................... 42 
 
Table 8. Relative standard deviation of electronic tongue sensors in the first and second run on 
Day 1 (using 3rd, 4th, and 5th data points), with Reb A, D, and M samples. ................................. 70 
 
Table 9. Relative standard deviation of electronic tongue sensors in the first and second run on 
Day 2 (using 3rd, 4th, and 5th data points), with Reb A, D, and M samples. ................................. 70 
 
Table 10. Relative standard deviation of electronic tongue sensors in the first and second run on 
Day 3 (using 3rd, 4th, and 5th data points), with Reb A, D, and M samples. ................................. 70 
 
Table 11. Discrimination power of the 7 electronic tongue sensors on Day 1, 2, and 3. ............. 71 
 
Table 12. Steviol glycosides profile1 of stevia leaf samples from Michigan State University. ... 75 
 
Table 13. Classification of stevia lines based on E-tongue and steviol glycoside profiles. ......... 75 

 

 
 
 

 

vii 

 

LIST OF FIGURES 

 
Figure 1. Steviol backbone of all glycosides (Geuns, 2003) ....................................................... 12 
 
Figure 2. The components of E-tongue from AlphaMOS (Toulouse, France) ............................ 18 
 
Figure 3. Correspondence analysis (CA) of sweeteners. Gray color indicates non-significant 
attributes; Red color indicates significant attributes; Samples are in blue. .................................. 43 
 
Figure 4. The influence of PROP taster groups on the perceived intensities (± SEM), with 
sweeteners combined (Reb A, D, and M). .................................................................................... 44 
 
Figure 5. E-tongue analysis sequence for protocol validation using Reb A, D, and M. Yellow 
color indicates cleaning steps; Purple color indicates samples. .................................................... 67 
 
Figure 6. Sensor values of electronic tongue in the second run on Day 1. Blue indicates sample 
Reb A. Red indicates sample Reb D. Green indicates sample Reb M. ......................................... 69 
 
Figure 7. (a) Principal component analysis of steviol glycoside samples in the second run on 
Day 1 as evaluated by electronic tongue and accounting for 96.31% of the variability. The 
discrimination index is 99. (b) Principal component analysis of steviol glycoside samples in the 
second run on Day 2 as evaluated by electronic tongue and accounting for 98.88% of the 
variability. The discrimination index is 94. (c) Principal component analysis of steviol glycoside 
samples in the second run on Day 3 as evaluated by electronic tongue and accounting for 96.10% 
of the variability. The discrimination index is 94. ........................................................................ 72 
 
Figure 8. Principal component analysis of stevia leaf samples from Fort Valley State University 
as evaluated by electronic tongue and accounting for 93.10% of the variability. The 
discrimination index is 95. ............................................................................................................ 73 
 
Figure 9. Principal component analysis of stevia leaf samples from Michigan State University as 
evaluated by electronic tongue and accounting for 80.86% of the variability. The discrimination 
index is 90. .................................................................................................................................... 74 

 

 

viii 

 

 

1. Background 

INTRODUCTION 

Stevia, which is a natural high-intensity non-nutritive sweetener derived from stevia plant 

(Stevia Rebaudiana Bertoni), has been used by the food industry to respond to the consumers’ 

demand for natural sugar substitutes with low/zero calories. Stevia is the source of many 

different types of steviol glycosides, which are the sweetening compounds in stevia leaves 

(Kinghorn, 2002). Stevioside and rebaudioside (Reb) A are the two major steviol glycosides 

(Kinghorn, 2002) and are the most widely used steviol glycosides on the market (Mintel Global 

New Products Database (GNPD), 2020). However, many studies reported the bitter and licorice 

off-taste of stevioside and Reb A (Gwak et al., 2012; Jenner, 1989; Kim et al., 2015; Li et al., 

2015; Medeiros et al., 2019; I. Prakash et al., 2008), which pose challenges to consumer 

acceptance. To address the taste challenges of stevioside and Reb A, the researchers and food 

industry have investigated Reb D and M, which are two minor steviol glycosides, as they have 

been shown to elicit significantly less bitterness with better sweetness than Reb A (Hellfritsch et 

al., 2012; Prakash et al., 2014). Most of the studies investigating sensory characteristics of 

steviol glycosides were conducted at a relatively low sweetness equivalency related to sucrose 

(SE) (e.g. at 5-10% SE) (Gwak et al., 2012; Kim et al., 2015; Li et al., 2015; Prakash et al., 2008; 

Prakash et al., 2014). The sweetness potency of stevia heavily depends on the SE (Prakash et al., 

2008), however, little research was done at high concentrations for high-sugar applications such 

as frozen desserts, which generally contain 13-22% sucrose w/v (Goff, 2015).  

Supertasters are a group of people who perceive the intense bitterness from 

Phenylthiocarbamide (PTC) and 6-n-Propylthiouracil (PROP) bitter-tasting compounds, while 

 

1 

 

those who barely detect the bitterness of them are classified as non-tasters (Bartoshuk et al., 

1994). It has been reported that individuals have different sensitivity to the aftertaste of high-

intensity sweeteners (Simons et al., 2008), and thus researchers have long been interested in 

understanding the relationship between PROP status (e.g. non-tasters vs supertasters) and 

perceived taste intensities of high-intensity sweeteners. However, the influence of PROP status 

on the perceived intensity has been controversial. Some researcher found the difference between 

non-tasters and supertasters in perceiving the bitterness of artificial sweeteners (Bartoshuk et al., 

1994; Drewnowski et al., 1997; Zhao & Tepper, 2007), while some did not (Horne et al., 2002; 

Rankin et al., 2003). Risso et al. (2014) found that the bitter taste receptor for PROP did not 

predict the bitterness perception of stevioside. However, little research was done to investigate 

the influence of PROP status on the perceived sweet and bitter taste intensities of the three 

popular steviol glycosides Reb A, D, and M. 

Electronic tongue (E-tongue) is an analytical instrument developed for the sensory 

analysis of taste, which mimics human sensations and evaluates the taste by analyzing dissolved 

compounds in a liquid matrix. It has the potential to be used for monitoring quality (Hruškar et 

al., 2009; Winquist et al., 2005), detecting adulteration (Dias et al., 2009), classification and even 

predicting attributes of unknown samples (Bleibaum et al., 2002; Dong et al., 2017; Kirsanov et 

al., 2012; Waldrop & Ross, 2014). Current papers using the E-tongue from AlphaMOS all used 

old sensor arrays, such as sensor array #1 (Bleibaum et al., 2002; Dong et al., 2017) and #5 

(Barnett et al., 2019; Jung et al., 2017; Lee et al., 2019; Lipkowitz et al., 2018; Schlossareck & 

Ross, 2019; Waldrop & Ross, 2014), but they were discontinued in late 2018. No paper has 

published using the most updated #6 array to date. Besides, the protocols for E-tongue analysis 

sequence using old arrays were vague and were different by papers (Bleibaum et al., 2002; Dong 

 

2 

 

et al., 2017; Lee et al., 2019; Lipkowitz et al., 2018; Schlossareck & Ross, 2019; Waldrop & 

Ross, 2014).  

Stevia plant breeders have worked on developing new varieties of stevia plants with 

better tasting profile to increase the amount of Reb D and M in the stevia leaves (Watson, 2015). 

Few breeding programs evaluate the flavor profile of stevia varieties due to the time and cost 
constraint to train a human panel, however, it is important to know the sensory profile of 
different stevia varieties to select the desirable ones. High-performance liquid chromatography 
(HPLC) is one way to determine the steviol glycoside content of stevia leaves (Ahmed & 
Dobberstein, 1982; Bondarev et al., 2003; Gardana et al., 2010; Hashimoto et al., 1978; Kolb et 
al., 2001; Makapugay et al., 1984), but the sample preparation and analysis procedure are rather 
complex (Dong et al., 2017; Kirsanov et al., 2012). If E-tongue could quickly differentiate 
glycoside profiles of different stevia varieties, it could save a lot of time for breeding programs. 

2. Hypotheses and Objectives 

2.1. Hypotheses 

 

Researchers hypothesized that there would be significant differences in the sensory 

profiles of Reb A, D, and M at 14% (w/v) sucrose equivalency. Furthermore, there would not be 

significant differences between non-tasters, medium tasters, and supertasters on the perceived 

sweetness and bitterness of the three steviol glycosides. Also, researchers hypothesized the 

electronic tongue would be able to discriminate steviol glycosides and stevia leaf samples and 

increase the potential use in stevia breeding programs. 

2.2. Objectives 

1) Determine sensory characteristics of Reb A, D, and M, compared to 14% (w/v) 

sucrose, using a consumer panel (Chapter 2).  

 

3 

 

2) Determine if there is a relationship between PROP taster status and the perceived 

intensities of three steviol glycosides (Chapter 2).  

3) Develop a protocol for E-tongue to discriminate stevia leaf samples for its potential 

use in stevia breeding programs as a method of testing flavor quality (Chapter 3).   

 

 

 

4 

 

 

 

 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

REFERENCES 

 

5 

 

REFERENCES 

 

Ahmed, M. S., & Dobberstein, R. H. (1982). Stevia rebaudiana: III. High-performance liquid 

chromatographic separation and quantitation of rebaudiosides B, D, and E, dulcoside A, and 
steviolbioside. Journal of Chromatography A, 245(3), 373–376. 
https://doi.org/10.1016/S0021-9673(00)88027-9 

 
Barnett, S. M., Diako, C., & Ross, C. F. (2019). Identification of a Salt Blend: Application of the 
Electronic Tongue, Consumer Evaluation, and Mixture Design Methodology: Identification 
of a salt blend…. Journal of Food Science, 84(2), 327–338. https://doi.org/10.1111/1750-
3841.14440 

 
Bartoshuk, L. M., Duffy, V. B., & Miller, I. J. (1994). PTC/PROP tasting: Anatomy, 

psychophysics, and sex effects. Physiology & Behavior, 56(6), 1165–1171. 
https://doi.org/10.1016/0031-9384(94)90361-1 

 
Bleibaum, R. N., Stone, H., Tan, T., Labreche, S., Saint-Martin, E., & Isz, S. (2002). Comparison 

of sensory and consumer results with electronic nose and tongue sensors for apple juices. 
Food Quality and Preference, 13(6), 409–422. https://doi.org/10.1016/S0950-
3293(02)00017-4 

 
Bondarev, N. I., Sukhanova, M. A., Reshetnyak, O. V., & Nosov, A. M. (2003). Steviol 

Glycoside Content in Different Organs of Stevia rebaudiana and Its Dynamics during 
Ontogeny. Biologia Plantarum, 47(2), 261–264. 
https://doi.org/10.1023/B:BIOP.0000022261.35259.4f 

 
Dias, L. A., Peres, A. M., Veloso, A. C. A., Reis, F. S., Vilas-Boas, M., & Machado, A. A. S. C. 
(2009). An electronic tongue taste evaluation: Identification of goat milk adulteration with 
bovine milk. Sensors and Actuators B: Chemical, 136(1), 209–217. 
https://doi.org/10.1016/j.snb.2008.09.025 

 
Dong, W., Zhao, J., Hu, R., Dong, Y., & Tan, L. (2017). Differentiation of Chinese robusta 

coffees according to species, using a combined electronic nose and tongue, with the aid of 
chemometrics. Food Chemistry, 229, 743–751. 
https://doi.org/10.1016/j.foodchem.2017.02.149 

 
Drewnowski, A., Henderson, S. A., & Shore, A. B. (1997). Genetic sensitivity to 6-n-

propylthiouracil (PROP) and hedonic responses to bitter and sweet tastes. Chemical Senses, 
22(1), 27–37. https://doi.org/10.1093/chemse/22.1.27 

 
Gardana, C., Scaglianti, M., & Simonetti, P. (2010). Evaluation of steviol and its glycosides in 
Stevia rebaudiana leaves and commercial sweetener by ultra-high-performance liquid 
chromatography-mass spectrometry. Journal of Chromatography A, 1217(9), 1463–1470. 
https://doi.org/10.1016/j.chroma.2009.12.036 

 

6 

 

Goff, H. D. (2015). Ice Cream and Frozen Desserts. In Ullmann’s Encyclopedia of Industrial 

Chemistry (pp. 1–15). American Cancer Society. 
https://doi.org/10.1002/14356007.a13_563.pub2 

 
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. https://doi.org/10.1007/s10068-012-0115-0 

 
Hashimoto, Y., Moriyasu, M., Nakamura, S., Ishiguro, S., & Komuro, M. (1978). High-

performance liquid chromatographic determination Stevia components on a hydrophilic 
packed column. Journal of Chromatography A, 161, 403–405. 
https://doi.org/10.1016/S0021-9673(01)85262-6 

 
Hellfritsch, C., Brockhoff, A., Stähler, F., Meyerhof, W., & Hofmann, T. (2012). Human 

electronic tongue. 8. 

Psychometric and Taste Receptor Responses to Steviol Glycosides. Journal of Agricultural 
and Food Chemistry, 60(27), 6782–6793. https://doi.org/10.1021/jf301297n 

Acesulfame-K. Chemical Senses, 27(1), 31–38. https://doi.org/10.1093/chemse/27.1.31 

 
Horne, J., Lawless, H. T., Speirs, W., & Sposato, D. (2002). Bitter Taste of Saccharin and 
 
Hruškar, M., Major, N., & Krpan, M. (2009). Evaluation of milk and dairy products by 
 
Jenner, M. R. (1989). Progress in sweeteners (T. H. Grenby, Ed.). Elsevier. 
https://agris.fao.org/agris-search/search.do?recordID=US201300648441 
 
Jung, H. Y., Kwak, H. S., Kim, M. J., Kim, Y., Kim, K.-O., & Kim, S. S. (2017). Comparison of 
a descriptive analysis and instrumental measurements (electronic nose and electronic 
tongue) for the sensory profiling of Korean fermented soybean paste (doenjang). Journal of 
Sensory Studies, 32(5), e12282. https://doi.org/10.1111/joss.12282 

 
Kim, M.-J., Yoo, S.-H., Jung, S., Park, M.-K., & Hong, J.-H. (2015). Relative sweetness, 

sweetness quality, and temporal profile of xylooligosaccharides and luo han guo (Siraitia 
grosvenorii) extract. Food Science and Biotechnology, 24(3), 965–973. 
https://doi.org/10.1007/s10068-015-0124-x 

http://www.crcnetbase.com/isbn/9780415268301 

 
Kinghorn, A. D. (2002). Stevia: The genus Stevia. Taylor & Francis. 
 
Kirsanov, D., Mednova, O., Vietoris, V., Kilmartin, P. A., & Legin, A. (2012). Towards reliable 
estimation of an “electronic tongue” predictive ability from PLS regression models in wine 
analysis. Talanta, 90, 109–116. https://doi.org/10.1016/j.talanta.2012.01.010 

 
Kolb, N., Herrera, J. L., Ferreyra, D. J., & Uliana, R. F. (2001). Analysis of Sweet Diterpene 

Glycosides from Stevia rebaudiana: Improved HPLC Method. Journal of Agricultural and 
Food Chemistry, 49(10), 4538–4541. https://doi.org/10.1021/jf010475p 

 

7 

 

Lee, H. J., Yoon, J. W., Kim, M., Oh, H., Yoon, Y., & Jo, C. (2019). Changes in microbial 

composition on the crust by different air flow velocities and their effect on sensory 
properties of dry-aged beef. Meat Science, 153, 152–158. 
https://doi.org/10.1016/j.meatsci.2019.03.019 

 
Li, X. E., Lopetcharat, K., & Drake, M. A. (2015). Parents’ and Children’s Acceptance of Skim 

Chocolate Milks Sweetened by Monk Fruit and Stevia Leaf Extracts. Journal of Food 
Science, 80(5), S1083–S1092. https://doi.org/10.1111/1750-3841.12835 

 
Lipkowitz, J. B., Ross, C. F., Diako, C., & Smith, D. M. (2018). Discriminating aging and 

protein-to-fat ratio in Cheddar cheese using sensory analysis and a potentiometric electronic 
tongue. Journal of Dairy Science, 101(3), 1990–2004. https://doi.org/10.3168/jds.2017-
13820 

 
Makapugay, H. C., Nanayakkara, N. P. D., & Kinghorn, A. D. (1984). Improved high-

performance liquid chromatographic separation of the Stevia rebaudiana sweet diterpene 
glycosides using linear gradient elution. Journal of Chromatography A, 283, 390–395. 
https://doi.org/10.1016/S0021-9673(00)96278-2 

 
Medeiros, A. C., Filho, E. R. T., & Bolini, H. M. A. (2019). Impact of Natural and Artificial 

Traditional, Lactose-Free, and Vegan Frozen Desserts of Chocolate Flavor. Journal of Food 

Sweeteners Compounds in the Sensory Profile and Preference Drivers Applied to 
Science, 84(10), 2973–2982. https://doi.org/10.1111/1750-3841.14806 

 
Mintel Global New Products Database (GNPD). (2020). https://www.mintel.com/global-new-
 
Prakash, I., DuBois, G. E., Clos, J. F., Wilkens, K. L., & Fosdick, L. E. (2008). Development of 

products-database 

rebiana, a natural, non-caloric sweetener. Food and Chemical Toxicology, 46(7, 
Supplement), S75–S82. https://doi.org/10.1016/j.fct.2008.05.004 

 
Prakash, Indra, Markosyan, A., & Bunders, C. (2014). Development of Next Generation Stevia 
Sweetener: Rebaudioside M. Foods, 3(1), 162–175. https://doi.org/10.3390/foods3010162 
 
Rankin, K. M., Godinot, N., Tepper, B. J., Kirkmeyer, S. V., & Christensen, C. M. (2003). 
Assessment of different methods for PROP status classification. In J. Prescott & B. J. 
Tepper (Eds.), Genetic Variation in Taste Sensitivity (pp. 63–88). CRC Press. 

 
Risso, D., Morini, G., Pagani, L., Quagliariello, A., Giuliani, C., De Fanti, S., Sazzini, M., 
Luiselli, D., & Tofanelli, S. (2014). Genetic signature of differential sensitivity to stevioside 
in the Italian population. Genes & Nutrition, 9(3). https://doi.org/10.1007/s12263-014-
0401-y 

 
Schlossareck, C., & Ross, C. F. (2019). Electronic Tongue and Consumer Sensory Evaluation of 

Spicy Paneer Cheese. Journal of Food Science, 84(6), 1563–1569. 
https://doi.org/10.1111/1750-3841.14604 

 

8 

Simons, C. T., Adam, C., LeCourt, G., Crawford, C., Ward, C., Meyerhof, W., & Slack, J. P. 
(2008). The “Bitter-Sweet” Truth of Artificial Sweeteners. In D. K. Weerasinghe & G. E. 
DuBois (Eds.), Sweetness and Sweeteners (Vol. 979, pp. 335–354). American Chemical 
Society. https://doi.org/10.1021/bk-2008-0979.ch021 

 
Waldrop, M. E., & Ross, C. F. (2014). Sweetener Blend Optimization by Using Mixture Design 
Methodology and the Electronic Tongue: Sweetener blend optimization…. Journal of Food 
Science, 79(9), S1782–S1794. https://doi.org/10.1111/1750-3841.12575 

 
Watson, E. (2015). GLG: We’ll be able to breed stevia plants with more Reb D, Reb M. 

FoodNavigator. https://www.foodnavigator.com/Article/2015/09/16/GLG-We-ll-be-able-to-
breed-stevia-plants-with-more-Reb-D-Reb-M 

 
Winquist, F., Bjorklund, R., Krantz-Rülcker, C., Lundström, I., Östergren, K., & Skoglund, T. 
(2005). An electronic tongue in the dairy industry. Sensors and Actuators B: Chemical, 
111–112, 299–304. https://doi.org/10.1016/j.snb.2005.05.003 

 
Zhao, L., & Tepper, B. J. (2007). Perception and acceptance of selected high-intensity 

sweeteners and blends in model soft drinks by propylthiouracil (PROP) non-tasters and 
super-tasters. Food Quality and Preference, 18(3), 531–540. 
https://doi.org/10.1016/j.foodqual.2006.07.004 

 

 
 

 

 

 

9 

 

CHAPTER 1: LITERATURE REVIEW 

 

1. Stevia 

1.1. History and Regulation 

 

Stevia Rebaudiana Bertoni is a herbaceous shrub that is native to Paraguay (Brandle et 

al., 1998). The leaves of stevia, which contain sweetening compound steviol glycoside, has been 

used to sweeten teas for hundreds of years in Paraguay and Brazil (Brandle et al., 1998; Geuns, 

2003). Stevia extract was first commercialized in Japan in the 1970s and was widely used to 

sweeten food and beverages (Abe & Sonobe, 1977; Akashi, 1977). The use of stevia comes later 

to western countries. In the United States, FDA approved certain high purity steviol glycosides 

(>95% pure glycosides) for generally recognized as safe (GRAS) status since 2008, but crude 

extracts or raw stevia leaf are excluded (FDA, 2018a). In the European Union, the use of steviol 

glycoside as a sweetener was approved in 2011 (SGS, 2011). Joint FAO/WHO Expert 

Committee on Food Additives (JECFA) established an acceptable daily intake (ADI) of steviol 

glycoside up to 4mg/ kg bw/day. 

1.2. Metabolism, Chemical Structure and Sweet Potency 

Stevia contains zero-calorie because the body does not absorb steviol glycosides. In vitro 

studies showed that steviol glycosides were hydrolyzed to steviol by human microflora and 

steviol is not degraded anymore (Gardana et al., 2003; Koyama et al., 2003). Steviol glucuronide 

(SVG), the major metabolite of steviol, is excreted through urine (Wang et al., 2015). 

Stevia leaves contain multiple steviol glycosides. All glycosides share the same steviol 

backbone (ent-13-hydroxykaur-16-en-19-oic acid) (Figure 1). R1 and R2 groups, which contain 

different numbers of glycosidic molecules, differentiate the glycosides identified in the stevia 

(Prakash et al., 2008, 2014). The R groups of each steviol glycosides are shown in Table 1. 

 

10 

 

Generally, the highly branched sugar chains at the R1 site make the glycoside sweeter than those 

with unbranched chains, the number of glucose molecules attached to the backbone positively 

correlates to the sweetness of the glycoside, and the rhamnose reduces the sweetness (Kinghorn, 

Fullas, & Hussain, 1995). For example, Reb D and Reb M have highly branched R1 groups and 

have five and six glucose molecules attached to the steviol backbone, respectively. Based on 

Table 1, they are sweeter than most of the glycosides. The relative sweetness of Reb D was 

higher than any other steviol glycosides, except Reb M (not included in the study) (Hellfritsch et 

al., 2012). Prakash et al. (2014) listed Reb M as the sweetest steviol glycosides.  

Steviol glycosides consist of 4 - 20% of the dry leaves by weight (Geuns, 2003). 

Stevioside, the most abundant steviol glycosides (4-13% w/w) (Makapugay, Nanayakkara, & 

Kinghorn, 1984; Momtazi-Borojeni et al., 2017), has been reported to be 150-250 times sweeter 

than sucrose (Carakostas et al., 2012). Rebaudioside A, the most widely used glycoside (2-4% 

w/w) (Makapugay et al., 1984; Momtazi-Borojeni et al., 2017), has a sweet potency about 200-

300 (Carakostas et al., 2012). The sweet potency has a wide range because it heavily depends on 

the sweetness equivalency (SE). The sweet potency is higher at low SE levels than at high SE 

levels. DuBois et al. (1991) reported that the sweet potency of Reb A is about 200 at 6% SE. It is 

also important to indicate the medium, temperature, and pH because the sweet potency changes 

when one of the factors changes. The 68% Reb A had a sweet potency of 348 at 5% SE and a 

181 at 15% SE in water solution at refrigerated temperature (Wee, Tan, & Forde, 2018). A stevia 

leaf extract (81% stevioside, 7.7% Reb A, and 0.6% Reb C) had a sweet potency of 97 when pH 

= 7 and a 109 when pH = 3 in water solutions at room temperature (Cardello, Da Silva, & 

Damasio, 1999).  

 
 

 

11 

 

Figure 1. Steviol backbone of all glycosides (Geuns, 2003) 

 

Table 1. R-groups, molecular formulas, molecular weights and potencies of the Stevia 
sweeteners. (Prakash et al., 2014) 

Sweetener 

Rebaudioside 

A 

Rebaudioside 

B 

Rebaudioside 

C 

Rebaudioside 

D 

Rebaudioside 

E 

Rebaudioside 

F 

 Rebaudioside 

M 

Stevioside 

Steviolbioside 
Rubusoside 
Dulcoside A 

R-Groups in Backbone Figure 

Above 

R1 

R2 

Formula 

Molecular 

Weight (g/mol) 

Potency 

* 

β-glc- 

(β-glc)2-β-glc- 

C44H70O23 

H 

(β-glc)2-β-glc- 

C38H60O18 

β-glc- 

(β-glc, α-rha-)-β-

glc- 

C44H70O22 

967.01 

804.88 

951.01 

β-glc-β-glc- 

(β-glc)2-β-glc- 

C50H80O28 

1129.15 

β-glc-β-glc- 

β-glc-β-glc- 

C44H70O23 

β-glc- 

(β-glc)2-β-

glc- 
β-glc- 

H 

β-glc- 
β-glc- 

(β-glc, β-xyl)-β-

glc- 

C43H68O22 

(β-glc)2-β-glc- 

C56H90O33 

β-glc-β-glc- 
β-glc-β-glc- 

β-glc- 

α-rha-β-glc- 

C38H60O18 
C32H50O13 
C32H50O13 
C38H60O17 

967.01 

936.99 

1291.3 

804.88 
642.73 
642.73 
788.87 

200 

150 

30 

221 

174 

200 

250 

210 
90 
114 
30 

glc = glucose; rha = rhamnose; xyl = xylose;  
* Potency from Kinghorn, 1999; Prakash, 2008; JECFA, 2008. 
 

 

12 

 

1.3. Taste Profile of Steviol Glycosides 

Stevioside and Reb A have a clean sweet taste at low SE levels (i.e., SE≤6), but have an 

undesirable bitterness and licorice aftertaste at high SE levels (Young & Wilkens, 2007). 

Stevioside has even stronger bitterness than Reb A, so stevioside is rarely used in consumer 

products.  

Recent focuses on Reb D and Reb M, the two minor glycosides in the stevia plant, show 

that Reb D and Reb M have a better flavor profiles compared to Reb A. Reb D exhibits a very 

low level of bitterness, compared to other glycosides (Hellfritsch et al., 2012). Reb D was 

discovered in the 1970s (Kohda et al., 1976), but Reb M was a more recent discovery (Prakash, 

Chaturvedula, & Markosyan, 2013). A trained descriptive panel compared the sensory attributes 

of Reb M and Reb A at 8% SE and reported Reb M had less bitterness and astringency than Reb 

A (Prakash et al., 2014). Prakash et al. (2014) also reported that Reb M had a similar sweetness 

level in water but higher sweetness perception in acidified solution (pH 3.2) than Reb A at 8% 

SE.  

1.4. Next Generation Steviol Glycosides (Reb D & M) 

 

Due to the limitation of the bitter aftertaste from Reb A, the food ingredient industry is 

moving to develop Reb D and M products in recent years to provide better sugar reduction 

alternatives. Table 2 shows the Reb D and M products currently available. Splenda® stevia 

sweetener is the only product that is accessible by consumers.  

 
 
 
 
 
 
 

 

13 

 

Table 2. Current Reb D and M products. 
Product 
EverSweet™ (Reb D & M) 
BESTEVIA® Reb M 
BESTEVIA® Reb D 
TASTEVA® M 
SPLENDA® (Reb D & Erythritol) 
AVANSYA™ Reb M 

 

Company 
Cargill 
Ingredion 
Ingredion 
Tate & Lyle 
Tate & Lyle 
Royal DSM 

It is not commercially possible to produce Reb D and M by extracting from the plant 

currently, because they only constitute approximately 0.4-0.5% by dry leaf weight (Jackson et 

al., 2009; Prakash et al., 2014). However, the ingredient companies use yeast to convert sugars 

into Reb D and M via a microbial fermentation process, which makes the production of these 

desirable rebaudioside more cost-effective and enables the companies to bring them to the 

market. For example, Avansya™ Reb M sweetener is produced via a genetically modified yeast 

Yarrowia lipolytica, which expresses the same metabolic pathway of steviol glycosides (FDA, 

2018b).  

Even though the genetically modified yeast is filtered out and the final product is GMO-

free, it cannot be labeled as “non-GMO”. Besides that, fermentative rebaudioside cannot be 

labeled as “stevia leaf extract” because it is not plant-derived and can only be labeled as Reb 

D/M, Rebaudioside D/M, or steviol glycosides (Watson, 2018). From a marketing standpoint, the 

inability to label as a leaf extract could be a downside because consumers may not be able to 

associate these terms with stevia. Furthermore, stevia sweetener produced by fermentation may 

be unattractive to consumers who only want naturally sourced sweetener, said PureCircle 

(Watson, 2017). Thus companies such as Cargill and PureCircle, are working on breeding stevia 

with higher levels of Reb D and M (Watson, 2018). 

 

14 

 

1.5. Food Applications 

The leading sources of added sugar in the United States are beverages, desserts, candy, 

baked goods, and dairy products (HHS & USDA, 2015) and stevia has been used as a sweetener 

in all of these food categories. Table 3 shows the reported maximum steviol glycosides use-

levels in foods (FAO, 2016).  

PureCircle (2019) stated the new products sweetened with stevia launched in 2017 

increased 11% and then additionally increased by 31% in 2018, based on Mintel Global New 

Products Database. There was a 36% increase for beverages and 27% for foods containing stevia 

in 2018. The reason stevia is used more in beverages could be that beverages do not have much 

texture and mouthfeel. However, in baked goods, the formulators need to consider browning, 

softness, and all other sensory attributes of the product as sugar not only provides sweetness, but 

also contributes to texture, mouthfeel, and viscosity (Samuel et al., 2018). In addition, beverages 

are the top source for added sugar (HHS & USDA, 2015) and also the top concern of added 

sugar source by consumers (Mintel, 2018). 

Table 3. Reported maximum use-level of steviol glycosides (FAO, 2016) 
Food type  
Beverages (soft drinks, fruit drinks)  
Desserts  
Yogurt  
Cold confectionery  
Sauces  
Pickles  
Delicacies  
Sweet corn  
Bread  
Biscuits  

Reported maximum use-level (mg/kg)  
600  
500  
500  
500  
1000  
1000  
1000  
200  
160  
300  

 

 

15 

 

2. Sensory Evaluation 

2.1. Consumer Tests 

Consumer tests are primarily used to measure consumer acceptance or preference toward 

products. Representative consumers, generally more than 100, are selected to represent the 

targeted larger population. Some selection criteria include user group, age, gender, and income 

level. A 9-point hedonic scale [extremely dislike (1) to extremely like (9)] is a commonly used 

scale to determine the acceptable level of a sensory attribute of a product or hedonic impression 

on a product (Peryam & Pilgrim, 1957). A check-all-that-apply (CATA) question, which 

consumers select terms from a given list, is a simple way to understand consumer perception on 

certain attributes of the product (Adams et al., 2007; Ares et al., 2010). 

Consumers are also capable of evaluating intensities (Moskowitz, 1996; Husson, Le 

Dien, & Pagès, 2001; Worch, Lê, Punter, 2010; Ares, Bruzzone, & Giménez, 2011). Two 

consumer panels (n1 = 218, n2 = 124) evaluated 10 attributes of 28 grape/raspberry beverages 

and the results showed that consumers could discriminate the products and the data was 

reproducible (Husson et al., 2001). Perfumes were evaluated by naïve consumers (n=103) and a 

trained panel and both groups gave similar results in terms of discrimination and reproducibility 

(Worch et al., 2010). Ares et al. (2011) reported that the intensity scores of consumers (n=86) 

and trained panel are similar when evaluating the five texture attributes of milk desserts.  

2.2. Descriptive Analysis 

 

Quantitative descriptive analysis (QDA) method (Stone et al., 1974; Stone & Sidel, 1992) 

usually consists of a small panel of 8-10 subjects for typical products and the panel is trained to 

objectively measure sensory attributes of a product. After insensitive training, the trained panel 

should be able to detect and describe the perceived sensory attributes. A 15-cm line scale is used 

 

16 

 

to evaluate intensities. Trained panel data can be used to aid in the interpretation of consumer 

data. Researchers can discover the relationship between product attributes and consumer 

acceptance by relating the two data and explain why consumers like or dislike a certain product. 

Even though the panel is trained to respond like an instrument, they are still humans. 

Human errors, such as health conditions, emotion, and stress, can affect their ability to smell and 

taste. Fatigue limits the number of samples that members of a trained panel can taste in once. It is 

also time-consuming to screen and train panelists.  

3. Electronic Tongue 

3.1. Electronic Tongue Components   

Electronic tongue (E-tongue) is an analytical instrument developed for the sensory 

analysis of taste. It mimics human sensations and evaluates the taste by analyzing dissolved 

compounds in a liquid matrix. AlphaMOS (Toulouse, France) is one of the leading E-tongue 

companies in the world. Its E-tongue includes an autosampler system, seven sensors, a reference 

electrode, a stirring rod, an electronic unit, and computer software for statistical analysis. The 

components are shown in Figure 2.  

The sensors are potentiometric sensors, which measure the voltage difference between 

the sensor membrane and the reference electrode and send electric signals to the computer for 

analysis. The sensors have different coated membranes that give each sensor different selectivity 

and sensitivity. The most updated sensor array #6 consists of AHS, PKS, CTS, NMS, CPS, ANS, 

and SCS sensors that each sensor measures multiple tastes. The old sensor array #5 had each 

sensor measures a single sensation: sweetness (SWS), sourness (SRS), saltiness (STS), bitterness 

(BRS), umami (UMS), metallic (GPS), and spiciness (SPS), but the production of these sensors 

was discontinued in late 2018. The old sensory array #1 for soft drinks, beers, and flavors 

 

17 

 

application and #2 for pharmaceutical formulation were discontinued and replaced by #6 as well. 

Current papers that used AlphaMOS’s E-tongue all used old sensor arrays (Bleibaum et al., 

2002; Hruškar, Major, & Krpan, 2009; Waldrop & Ross, 2014; Jung et al., 2017; Barnett, Diako, 

& Ross, 2019; Lee et al., 2019; Schlossareck & Ross, 2019) and no paper published using #6 

array to date.  

Figure 2. The components of E-tongue from AlphaMOS (Toulouse, France) 

 

3.2. Benefits and Limitations  

One of the benefits of using E-tongue is its rapidity and high throughput of samples. It 

can evaluate a large number of samples in a relatively short amount of time without fatigue. 

Moreover, it assures assessors’ safety by evaluating potentially dangerous substances or 

substances that are not allowed for human consumption yet, which strongly benefits the 

pharmaceutical industry. Besides that, it can be a more reliable sensory evaluation as it is 

automated and objective.  

However, E-tongue also has some limitations. E-tongue can only analyze liquid without 

particles, so solid samples must be homogenized and extracted in order to analyze. For example, 

the supernatants of homogenized and centrifuged ground meat samples were used for analysis 

(Lee et al., 2019). In addition, samples with high fat content could damage the sensors. 

 

18 

 

Therefore, lipid-free cheese extracts were made to analyze cheese samples (Lipkowitz et al., 

2018).  

3.3. Food Applications 

E-tongue has been used to evaluate many foods in many ways, such as to differentiate 

coffees (Dong et al., 2017) and milk and yogurts (Hruškar et al., 2009), to detect bitterness and 

astringency of green tea (Zou et al., 2018), to evaluate apple juices (Bleibaum et al., 2002), 

wines (Kirsanov et al., 2012), Korean fermented soybean paste (doenjang) (Jung et al., 2017), 

dry-aged beef (Lee et al., 2019), cheeses (Lipkowitz et al., 2018; Schlossareck & Ross, 2019), 

and sweet and salt solutions (Waldrop & Ross, 2014; Barnett et al., 2019).   

3.4. Statistical Analysis for Electronic Tongue Data 

 

Principal component analysis (PCA) is a visualization plot of the data that presents data 

in clusters for easier interpretation. It gives information on the similarity and differences among 

samples. Discrimination index (DI) is determined by how well groups are separated and the size 

of each group. DI is close to 100 when the groups are clearly separated and the size of each 

group is small. However, when groups overlap, the DI is negative.  

DI is strongly depended on the discrimination power of the sensors, which ranges from 0 

to 1. A number closer to 1 means that the sensor has a good discrimination ability on samples 

and a number lower than 0.5 means that the sensor can hardly differentiate the samples.  

4. PROP (6-n-Propylthiouracil) Taster Status 

 

Phenylthiocarbamide (PTC) and 6-n-Propylthiouracil (PROP) are bitter-tasting 

compounds that have been used to test people’s sensitivity to bitter taste. Supertasters are a group 

of people that are very sensitive to PTC and PROP, while non-tasters can barely detect the 

bitterness of them. Fox (1932) estimated that about 30% of the Caucasian population is non-

 

19 

 

tasters, as the proportion varies among races and ethnicity (Guo & Reed, 2001). Within the 70% 

tasters, about 45% are medium tasters and 25% are supertasters (Zhao, Kirkmeyer, & Tepper, 

2003). 

 

Some studies showed that supertasters perceive not only stronger bitterness, but also 

other sensory attributes, such as sweetness (Drewnowski, Henderson, & Shore, 1997). Sucrose 

and saccharin solutions were perceived sweeter by medium and supertasters than non-tasters, 

(Drewnowski et al., 1997). Zhao and Tepper (2007) found that in addition to bitterness and 

persistence of bitterness, supertasters perceived more sweetness, aftertaste, thickness, and overall 

flavor than non-tasters in carbonated soft drinks sweetened with high-intensity sweeteners. A 

study on coffee showed that supertasters rated the sourness, bitterness, and astringency higher 

than non-tasters (Masi et al., 2015). However, there were other studies that did not find a 

relationship between PROP taster status and bitterness and/or sweetness (Horne et al., 2002; 

Rankin et al., 2003). Risso et al. (2014) investigated the effect of genetic variations on stevioside 

and found that the bitter taste receptor for PROP did not predict the bitterness perception of 

stevioside. Little research was done to investigate the effect of PROP taster status on the popular 

steviol glycosides, such as Reb A, D, and M. 

 

 

 

20 

 

 

 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

REFERENCES 

 

21 

 

 

REFERENCES 

Abe, K., & Sonobe, M. (1977). Use of stevioside in the food industry. New Food Industry 19,  

Adams, J., Williams, S., Lancaster, B., & Foley, M. (2007). Advantages and uses of check-all- 

that-apply response compared to traditional scaling of attributes for salty snacks. 7th 
Pangborn sensory science symposium, Hyatt Regency, Minneapolis, MN, USA  

 
Akashi, H. (1977). Present status and prospect for stevioside for utilization. Shokuhin Kogyo  

67–72. 
 

20(24), 20–26. 
 

Alizadeh, M., Azizi-Lalabadi, M., & Kheirouri, S. (2014). Impact of Using Stevia on  

Physicochemical, Sensory, Rheology and Glycemic Index of Soft Ice Cream. Food and 
Nutrition Sciences, 05(04), 390–396. https://doi.org/10.4236/fns.2014.54047 

 
Ares, G., Deliza, R., Barreiro, C., Giménez, A., & Gámbaro, A. (2010). Comparison of two  

sensory profiling techniques based on consumer perception. Food Quality and 
Preference, 21, 417–426. 

 
Ares, G., Bruzzone, F., & Giménez, A. (2011). Is a Consumer Panel Able to Reliably Evaluate 
the Texture of Dairy Desserts Using Unstructured Intensity Scales? Evaluation of Global 
and Individual Performance. Journal of Sensory Studies, 26, 363–370. 

 
Arnold, D.L., Moodie, C.A., Stavric, B., Stoltz, D.R., Grice, H.C., & Munro, I.C. (1977). 

Canadian Saccharin Study. Science, 197, 320–320. 

 
Barnett, S.M., Diako, C., & Ross, C.F. (2019). Identification of a Salt Blend: Application of the 
Electronic Tongue, Consumer Evaluation, and Mixture Design Methodology: Identification 
of a salt blend…. Journal of Food Science, 84, 327–338. 

 
Bleibaum, R.N., Stone, H., Tan, T., Labreche, S., Saint-Martin, E., & Isz, S. (2002). Comparison 

of sensory and consumer results with electronic nose and tongue sensors for apple juices. 
Food Quality and Preference, 13, 409–422. 

 
Brandle, J.E., Starratt, A.N., & Gijzen, M. (1998). Stevia rebaudiana : Its agricultural, biological, 

and chemical properties. Canadian Journal of Plant Science, 78, 527–536. 

 
Carakostas, M., Prakash, A., Kinghorn, D., Wu, C.D., & Soerjato, D.D. (2012). Steviol  

glycosides. In Alternative Sweeteners, 4th Ed. ( L.O.B. Nabors, ed.) pp. 159– 180, CRC 
Press, London, UK 

 

 

22 

Cardello, H.M.A.B., Da Silva, M.A.P.A., & Damasio, M.H. (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, 119–129. 

 
Chen, T.-H., Chen, S.-C., Chan, P., Chu, Y.-L., Yang, H.-Y., & Cheng, J.-T. (2005). Mechanism 

of the Hypoglycemic Effect of Stevioside, a Glycoside of Stevia rebaudiana. Planta 
Medica, 71, 108–113. 

 
DeNoon, D.J. (2019). Drink More Diet Soda, Gain More Weight? [Internet document] . 

WebMDURL https://www.webmd.com/diet/news/20050613/drink-more-diet-soda-gain-
more-weight. Accessed 26/11/2019. 

 
Deshmukhan, YRK., Sirsat, A., Pritamk, H., Zele, SS., & More, KD. (2014). Preparation of ice- 

cream using natural sweetener stevia. Food Sci Res J 5(1):30–33 

 
Dong, W., Zhao, J., Hu, R., Dong, Y., & Tan, L. (2017). Differentiation of Chinese robusta 

coffees according to species, using a combined electronic nose and tongue, with the aid of 
chemometrics. Food Chemistry, 229, 743–751. 

 
Drewnowski, A., Henderson, S.A., & Shore, A.B. (1997). Genetic sensitivity to 6-n-

propylthiouracil (PROP) and hedonic responses to bitter and sweet tastes. Chemical Senses, 
22, 27–37. 

 
DuBois et al., (1991). Sweeteners: Discovery, Molecular Design, and Chemoreception. 
Symposium Sponsored by the Division of Agricultural and Food Chemistry at 
the 199th National Meeting of the American Chemical Society, April 22–27, 
1990, Boston. ACS Symposium Series, vol. 450. American Chemical Society 
(ACS), Washington, DC. pp. 261–276. 
 

FAO. (2016). 82nd JECFA - Chemical and Technical Assessment (CTA) Steviol Glycosides.  

Retrieved from FAO website: http://www.fao.org/3/a-br566e.pdf 
 

 

 

 

FDA. (2018a). Additional Information about High-Intensity Sweeteners Permitted for Use in  

Food in the United States [WebContent]. From U.S. Food & Drug Administration  
website: https://www.fda.gov/food/food-additives-petitions/additional-information-about-
high-intensity-sweeteners-permitted-use-food-united-states 

 
FDA. (2018b). GRAS Notice No. GRN 000759. Retrieved from FDA website:  

https://www.fda.gov/media/113875/download 

 
Fox, A. L. (1932). The Relationship between Chemical Constitution and Taste. Proceedings of  

the National Academy of Sciences of the United States of America, 18(1), 115–120.  
doi:10.1073/pnas.18.1.115 

23 

 

Gardana, C., Simonetti, P., Canzi, E., Zanchi, R., & Pietta, P. (2003). Metabolism of Stevioside 

and Rebaudioside A from Stevia rebaudiana Extracts by Human Microflora. Journal of 
Agricultural and Food Chemistry, 51, 6618–6622. 

 
Geuns, J.M.C. (2003). Stevioside. Phytochemistry, 64, 913–921. 
 
Goff, H.D. (2015). Ice Cream and Frozen Desserts. In: Ullmann’s Encyclopedia of Industrial 

Chemistry. Pp. 1–15. American Cancer Society. 

 
Green, B. G., Shaffer, G. S., & Gilmore, M. M. (1993). Derivation and evaluation of a semantic  

scale of oral sensation magnitude with apparent ratio properties. Chemical Senses, 18,  
683–702. 

 
Green, B.G., Dalton, P., Cowart, B., Shaffer, G., Rankin, K., & Higgins, J. (1996). Evaluating 
the ‘Labeled Magnitude Scale’ for Measuring Sensations of Taste and Smell. Chemical 
Senses, 21, 323–334. 

 
Guo, S. W., & Reed, D. R. (2001). The genetics of phenylthiocarbamide perception. Annals of  

Human Biology, 28, 111–142. 

 
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, 889–
894. 

 
Hampton, T. (2008). Sugar Substitutes Linked to Weight Gain. JAMA, 299, 2137–2138. 
 
Hellfritsch, C., Brockhoff, A., Stähler, F., Meyerhof, W., & Hofmann, T. (2012). Human 

Psychometric and Taste Receptor Responses to Steviol Glycosides. Journal of Agricultural 
and Food Chemistry, 60, 6782–6793. 

 
Horne, J., Lawless, H. T., Speirs, W., & Sposato, D. (2002). Bitter Taste of Saccharin and 

Acesulfame-K. Chemical Senses, 27(1), 31–38. https://doi.org/10.1093/chemse/27.1.31 

 
Hruškar, M., Major, N., & Krpan, M. (2009). Evaluation of milk and dairy products by electronic 

tongue, 8. 

 
Husson, F., Le Dien, S., & Pagès, J. (2001). Which value can be granted to sensory profiles 

given by consumers? Methodology and results. Food Quality and Preference, 12, 291–296. 

 
Jung, H.Y., Kwak, H.S., Kim, M.J., Kim, Y., Kim, K.-O., & Kim, S.S. (2017). Comparison of a 
descriptive analysis and instrumental measurements (electronic nose and electronic tongue) 
for the sensory profiling of Korean fermented soybean paste (doenjang). Journal of Sensory 
Studies, 32, e12282. 

 
Kennedy, S. (2018). Ice cream’s healthy future. Retrieved from Dairy Foods website:  

https://www.dairyfoods.com/articles/92778-ice-creams-healthy-future 

 

24 

 

Kim, M.-J., Yoo, S.-H., Jung, S., Park, M.-K., & Hong, J.-H. (2015). Relative sweetness,  

sweetness quality, and temporal profile of xylooligosaccharides and luo han guo (Siraitia 
grosvenorii) extract. Food Science and Biotechnology, 24, 965–973. 

 
Kinghorn, A.D., Fullas, F., & Hussain, R.A. (1995). Structure-activity relationship of highly 

sweet natural products. In: Studies in Natural Products Chemistry, Structure and chemistry 
(Part C) (edited by Atta-ur-Rahman). Pp. 3–41. Elsevier. 

 
Kirsanov, D., Mednova, O., Vietoris, V., Kilmartin, P.A., & Legin, A. (2012). Towards reliable 
estimation of an “electronic tongue” predictive ability from PLS regression models in wine 
analysis. Talanta, 90, 109–116. 

 
Kohda, H., Kasai, R., Yamasaki, K., Murakami, K., & Tanaka, O. (1976). New sweet diterpene 

glucosides from Stevia rebaudiana. Phytochemistry, 15, 981–983. 

 
Koyama, E., Kitazawa, K., Ohori, Y., Izawa, O., Kakegawa, K., Fujino, A., & Ui, M. (2003). In 
vitro metabolism of the glycosidic sweeteners, stevia mixture and enzymatically modified 
stevia in human intestinal microflora. Food and Chemical Toxicology, 41, 359–374. 

 
Lee, H.J., Yoon, J.W., Kim, M., Oh, H., Yoon, Y., & Jo, C. (2019). Changes in microbial 
composition on the crust by different air flow velocities and their effect on sensory 
properties of dry-aged beef. Meat Science, 153, 152–158. 

 
Li, X.E., Lopetcharat, K., & Drake, M.A. (2015). Parents’ and Children’s Acceptance of Skim 

Chocolate Milks Sweetened by Monk Fruit and Stevia Leaf Extracts. Journal of Food 
Science, 80, S1083–S1092. 

 
Lipkowitz, J.B., Ross, C.F., Diako, C., & Smith, D.M. (2018). Discriminating aging and protein-

to-fat ratio in Cheddar cheese using sensory analysis and a potentiometric electronic 
tongue. Journal of Dairy Science, 101, 1990–2004. 

 
Makapugay, H., Nanayakkara, N., & Kinghorn, A. (1984). Improved high performance liquid  
chromatographic separation of the Stevia rebaudiana sweet diterpene glycosides using  
linear gradient elution. J Chromatogr A; 283:390–5. 

 
Masi, C., Dinnella, C., Monteleone, E., & Prescott, J. (2015). The impact of individual variations 

in taste sensitivity on coffee perceptions and preferences. Physiology & Behavior, 138, 
219–226. 

 
Medeiros, A.C., Filho, E.R.T., & Bolini, H.M.A. (2019). Impact of Natural and Artificial 

Sweeteners Compounds in the Sensory Profile and Preference Drivers Applied to 
Traditional, Lactose‐Free, and Vegan Frozen Desserts of Chocolate Flavor. Journal of Food 
Science, 84, 2973–2982. 

 
Mintel. (2018). Sugar and Alternative Sweeteners – US – December 2018 [Market Research  

Report]. Mintel Academic. 

 

25 

Mintel. (2019). Ice Cream and Frozen Novelties – US – May 2019 [Market Research Report].  

Mintel Academic. 

 
Momtazi-Borojeni, AA., Esmaeili, S-A., Abdollahi, E., & Sahebkar, A. (2017). A review on the  

pharmacology and toxicology of steviol glycosides extracted from Stevia rebaudiana.  
Curr Pharm Des; 23:1616–22 
 

Moskowitz, H.R. (1996). Experts Versus Consumers: A Comparison. Journal of Sensory Studies, 

11, 19–37. 

 
Ortenberg, C. (2019). Halo Top Sold to Blue Bunny Producer. Retrieved from nosh website:  

https://www.nosh.com/news/2019/halo-top-sold-to-blue-bunny-bunny-producer 

 
Ozdemir, C., Arslaner, A., Ozdemir, S., & Allahyari, M. (2015). The production of ice cream 

using stevia as a sweetener. Journal of Food Science and Technology, 52, 7545–7548. 

 
Pandiyan, R., Velu, R.K., Chandrasekaran, S.K., & Bashiyam, P. (2009). Effect on extracts of 

Stevia rebaudiana bertoni. in ethanol induced gastric ulcer by using wister rats. Recent 
Research in Science and Technology. 

 
Peryam, D. R., & Pilgrim, F. J. (1957). Hedonic scale method of measuring food   
        preferences. Food Technology, 11, Suppl., 9–14. 
 
Pon, S.Y., Lee, W.J., & Chong, G.H. (2015). Textural and rheological properties of stevia ice  

cream. International Food Research Journal, 22 (4), pp. 1544-1549 

 
Prakash, I., Chaturvedula, V.S.P., & Markosyan, A. (2013). Isolation, Characterization and 
Sensory Evaluation of a Hexa β-D-Glucopyranosyl Diterpene from Stevia rebaudiana. 
Natural Product Communications, 8, 1934578X1300801106. 

 
Prakash, I., DuBois, G.E., Clos, J.F., Wilkens, K.L., & Fosdick, L.E. (2008). Development of 

rebiana, a natural, non-caloric sweetener. Food and Chemical Toxicology, Rebaudioside A: 
An Assessment of Safety, 46, S75–S82. 

 
Prakash, I., Markosyan, A., & Bunders, C. (2014). Development of Next Generation Stevia 

Sweetener: Rebaudioside M. Foods, 3, 162–175. 

 
PureCircle. (2019). Stevia Use In Food and Beverages Accelerated Significantly in 2018.  

Retrieved from PureCircle website: https://purecircle.com/news/stevia-use-in-food-and-
beverages-accelerated-significantly-in-2018/ 

 
Rankin, K. M., Godinot, N., Tepper, B. J., Kirkmeyer, S. V., & Christensen, C. M. (2003). 
Assessment of different methods for PROP status classification. In J. Prescott & B. J. 
Tepper (Eds.), Genetic variation in taste sensitivity (pp. 63–88). New York: Marcel Dekker.  

 

 

 

26 

 

 
 
 

 

 
Stone, H., Sidel, J., Oliver, S., Woolsey, A., & Singleton, R. C. (1974). Sensory evaluation by  

quantitative descriptive analysis. Food Technol 28(11): 24-34 

 
Swithers, S.E., & Davidson, T.L. (2008). A role for sweet taste: calorie predictive relations in 

energy regulation by rats. Behavioral Neuroscience, 122, 161–173. 

 
Tandel, K.R. (2011). Sugar substitutes: Health controversy over perceived benefits. Journal of 

Pharmacology & Pharmacotherapeutics, 2, 236–243. 

 
U. Jackson, A., Tata, A., Wu, C., H. Perry, R., Haas, G., West, L., & Graham Cooks, R. (2009). 

Direct analysis of Stevia leaves for diterpene glycosides by desorption electrospray 
ionization mass spectrometry. Analyst, 134, 867–874. 

Risso, D., Morini, G., Pagani, L., Quagliariello, A., Giuliani, C., De Fanti, S., Sazzini, M., 

Luiselli, D., & Tofanelli, S. (2014). Genetic signature of differential sensitivity to stevioside 
in the Italian population. Genes & Nutrition, 9(3). https://doi.org/10.1007/s12263-014-
0401-y 

 
Samuel, P., Ayoob, K.T., Magnuson, B.A., Wölwer-Rieck, U., Jeppesen, P.B., Rogers, P.J., 

Rowland, I., & Mathews, R. (2018). Stevia Leaf to Stevia Sweetener: Exploring Its Science, 
Benefits, and Future Potential. The Journal of Nutrition, 148, 1186S-1205S. 

 
Saravanan, R., Vengatash babu, K., & Ramachandran, V. (2012). Effect of Rebaudioside A, a 
diterpenoid on glucose homeostasis in STZ-induced diabetic rats. Journal of Physiology 
and Biochemistry, 68, 421–431. 

 
Savita, S.M., Sheela, K., Sunanda, S., Shankar, A.G., Ramakrishna, P., & Sakey, S. (2004). 

Health Implications of Stevia rebaudiana. Journal of Human Ecology, 15, 191–194. 

 
Schlossareck, C., & Ross, C.F. (2019). Electronic Tongue and Consumer Sensory Evaluation of 

Spicy Paneer Cheese. Journal of Food Science, 84, 1563–1569. 

 
SGS. (2011). European Commission Approves Stevia as a Sweetener and Improves  

Transparency for Additives. Retrieved from SGS website: 
https://www.sgs.com/en/news/2011/11/european-commission-approves-stevia-as-a-
sweetener-and-improves-transparency-for-additives 

 
Shukla, S., Mehta, A., Bajpai, V.K., & Shukla, S. (2009). In vitro antioxidant activity and total 
phenolic content of ethanolic leaf extract of Stevia rebaudiana Bert. Food and Chemical 
Toxicology, 47, 2338–2343. 

 
Stone, H., & Sidel, J. L. (1992). Sensory evaluation practices, 2nd Ed. Orlando, FL: Academic  

Press. 

27 

 

U.S. Department of Health and Human Services and U.S. Department of Agriculture. 2015 –  

2020 Dietary Guidelines for Americans. 8th Edition. December 2015. Available  
at https://health.gov/dietaryguidelines/2015/guidelines/. 

 
Waldrop, M.E., & Ross, C.F. (2014). Sweetener Blend Optimization by Using Mixture Design 

Methodology and the Electronic Tongue. Journal of Food Science, 79, S1782–S1794. 

 
Wang, M., Qi, H., Li, J., Xu, Y., & Zhang, H. (2015). Transmembrane transport of steviol 

glucuronide and its potential interaction with selected drugs and natural compounds. Food 
and Chemical Toxicology, 86, 217–224. 

 
Watson, E. (2018). A match made in heaven: Cargill and DSM team up to bring fermentation- 

based sweeteners to market. Retrieved from Foodnavigator website: 
https://www.foodnavigator-usa.com/Article/2018/11/08/Cargill-and-DSM-to-establish-
JV-to-bring-fermentation-based-high-potency-sweeteners-to-market 

 
Watson, E. (2017). Consumers want stevia from the leaf, says PureCircle: “They want naturally- 

sourced plant-based ingredients”. Retrieved from Foodnavigator website: 
https://www.foodnavigator-usa.com/Article/2017/03/20/Consumers-want-stevia-from-
the-leaf-says-PureCircle 

 
Wee, M., Tan, V., & Forde, C. (2018). A Comparison of Psychophysical Dose-Response 

Behaviour across 16 Sweeteners. Nutrients, 10, 1632. 

 
Weihrauch, M.R., & Diehl, V. (2004). Artificial sweeteners—do they bear a carcinogenic risk? 

Annals of Oncology, 15, 1460–1465. 

 
Worch, T., Lê, S., & Punter, P. (2010). How reliable are the consumers? Comparison of sensory 

profiles from consumers and experts. Food Quality and Preference, Sensometrics 2008 
(TBC), 21, 309–318. 

 
Young, N.D., & Wilkens, K. (2007). Study of descriptive analysis of rebaudioside A, 
aspartame and sucrose in water at room temperature, Unpublished results. The 
Coca-Cola Company, Atlanta, GA, USA. 

 
Zhang, H., Yang, H., Zhang, M., Wang, Y., Wang, J., Yau, L., Jiang, Z., & Hu, P. (2012). 

Identification of flavonol and triterpene glycosides in Luo-Han-Guo extract using ultra-high 
performance liquid chromatography/quadrupole time-of-flight mass spectrometry. Journal 
of Food Composition and Analysis, 25, 142–148. 

 
Zhao, L., Kirkmeyer, S.V., & Tepper, B.J. (2003). A paper screening test to assess genetic taste 

sensitivity to 6-n-propylthiouracil. Physiology & Behavior, 78, 625–633. 

 
Zhao, L., & Tepper, B.J. (2007). Perception and acceptance of selected high-intensity sweeteners 

and blends in model soft drinks by propylthiouracil (PROP) non-tasters and super-tasters. 
Food Quality and Preference, 18, 531–540. 

 

28 

 

	

 

Zou, G., Xiao, Y., Wang, M., & Zhang, H. (2018). Detection of bitterness and astringency of 

green tea with different taste by electronic nose and tongue. PLoS ONE, 13. 

 

 

29 

 

CHAPTER 2: CONSUMER-BASED SENSORY CHARACTERIZATION OF STEVIOL 

GLYCOSIDES (REBAUDIOSIDE A, D, AND M) 

Published in Foods Journal on July 31st. doi:10.3390/foods9081026 

 

Abstract   

Rebaudioside (Reb) D and M are the recent focus of the food industry to address the 
bitter taste challenge of Reb A, which is the most commonly used steviol glycoside in natural 
sweetener stevia. This study evaluated the sensory characteristics of Reb A, D, and M, compared 

to 14% (w/v) sucrose, using a consumer panel and explored the relationship between 6-n-

Propylthiouracil (PROP) taster status (i.e., non-tasters, medium tasters, supertasters) and the 

perceived intensity of sweet and bitter tastes of the three steviol glycosides. A total of 126 

participants evaluated the intensities of in-mouth, immediate (5 seconds after expectorating), and 

lingering (1 minute after expectorating) sweetness and bitterness of 0.1% Reb A, D, M, and 14% 

sucrose and described the aftertaste of the sweeteners by using a check-all-that-apply (CATA) 

question. The results showed that in-mouth sweetness and bitterness of Reb D and M were not 

significantly different from sucrose, unlike Reb A which showed significant bitterness. However, 

Reb D and M showed more intense lingering sweetness than sucrose. The CATA analysis 

resulted that Reb D and M were closer to positive attribute terms and also to sucrose than Reb A, 

but Reb D and M were still considered artificial, which may cause negative perception. When 

comparing among PROP taster groups, no significant differences in the perceived sweetness and 

bitterness of the three steviol glycosides were found. This study generates important information 

about Reb A, D, and M for the food industry, especially working with products formulated to 
deliver reductions in sugar using a natural high-intensity sweetener, stevia.  

 

 

30 

 

1. Introduction 

Artificial sweeteners are widely used in a variety of foods and beverages as a sugar 

substitute that mimics the effect of sugar on taste without adding calories. However, consumers 

have a negative perception of artificial sweeteners not only due to aversive sensations such as 

bitter off-taste [1,2] but also due to potential health risks and demand more natural options [3]. 

To respond to the consumers’ demand for natural sugar substitutes with low/zero calories, the 

food industry has focused on stevia, which is a natural high-intensity non-nutritive sweetener. 

Stevia (Stevia Rebaudiana Bertoni) is a shrub native to Paraguay, and the leaves of stevia have 

been used to sweeten teas for hundreds of years in Paraguay and Brazil [4,5]. Stevia is the source 

of many different types of steviol glycosides, which are the sweetening compounds in stevia 

leaves [6]. Stevioside and rebaudioside (Reb) A are the major sweet compounds among the 

steviol glycosides [6] and are the most widely used steviol glycosides on the market according to 

a Mintel Global New Products Database (GNPD) product search [7]. However, stevioside and 

Reb A exhibit bitter and licorice off-taste [8–13], which pose challenges to product formulation.  

To overcome the taste challenges of stevioside and Reb A, the researchers and food 

industry have looked into other minor steviol glycosides in the stevia leaves to provide better 

sugar reduction solutions. Several studies have reported that the two minor steviol glycosides, 

Reb D and M elicit significantly less bitterness with better sweetness than Reb A and also work 

well in products without sacrificing the taste [14–19]. Prakash et al. [16] reported that Reb M had 

less bitterness and astringency than Reb A. Most of the studies investigating sensory 

characteristics of steviol glycosides were conducted within a specific range of 5-10% sweetness 

equivalency related to sucrose (SE) [8,10,11,13,16]. Little research was done at high 

 

31 

 

concentrations for high-sugar applications such as frozen desserts, which generally contain 13-

22% sucrose w/v [20], although sweetness potency of stevia heavily depends on the SE [13].  

For sensory characterization of food products, sensory descriptive analysis using trained 

assessors is the most widely used method, but it is time-consuming to train panel. Less time 

consuming and more flexible methodologies such as check-all-that-apply (CATA) or intensity 

scales using consumers have been discussed in the last two decades [21]. It has been reported 

that consumers were capable of evaluating sensory attributes of various products, showing good 

agreement between consumers and trained assessors in terms of discrimination, reproducibility, 

and consensus [22–25]. Although Worch et al. [24] found that the trained panel showed greater 

consensus among each other, the larger sample size of consumers compensated for the higher 

variability. Moskowitz [26] suggested that a minimum of 40-50 people was needed to get stable 

averages, and the averages would not be affected by the base size much once the participant 

number exceeded 80. Ares et al. [27] also indicated that 80 consumers would be sufficient to get 

stable results when samples had large differences, but caution would be needed if samples had 

smaller differences or more complex attributes.  

CATA is also often used to determine the characteristics of a product from a consumer 

perspective, which allows the consumers to describe a product by selecting terms from a given 

list that would match the product [28]. CATA questions have been used for a variety of foods 

and beverages [28–32], and these studies showed that CATA was a simple way to understand 

consumer perception on the sensory profile of a product.  

Phenylthiocarbamide (PTC) and 6-n-Propylthiouracil (PROP) are bitter-tasting 

compounds that have been used to test people’s sensitivity to bitter taste. Supertasters are a group 

of people who perceived intense bitter taste from PTC and PROP, while non-tasters barely detect 

 

32 

 

the bitterness of them [33]. It has been reported that individuals have different sensitivity to the 

aftertaste of high-intensity sweeteners [34], and thus researchers have long been interested in 

understanding the relationship between PROP status (e.g. non-tasters vs supertasters) and 

perceived taste intensities of high-intensity sweeteners. Bartoshuk [35] and Drewnowski et al. 

[36] found a significant difference between non-tasters and supertasters in the bitterness of 

saccharin at low concentrations. Zhao and Tepper [37] also suggested that supertasters perceived 

more bitterness and sweetness than non-tasters in carbonated soft drinks with artificial 

sweeteners, including sucralose, aspartame, acesulfame-K. However, Horne et al. [38] did not 

find a relationship between PROP taster status and the sweetness and bitterness of saccharin and 

acesulfame-K. Rankin et al. [39] failed to find any significant difference in bitterness between 

supertasters and non-tasters in cola drink sweetened with artificial sweeteners either. Risso et al. 

[40] looked into the effect of genetic variations on stevioside and found that the bitter taste 

receptor for PROP did not predict the bitterness perception of stevioside. However, little research 

was done to investigate the influence of PROP status on the perceived sweet and bitter taste 

intensities of novel steviol glycosides such as Reb D and M. 

 

The primary objective of this study was to determine sensory characteristics of Reb A, D, 

and M, compared to 14% (w/v) sucrose, using a consumer panel. A secondary objective was to 

determine if there is a relationship between PROP taster status and the perceived intensities of 

the three steviol glycosides.  

2. Materials and Methods 

2.1. Materials 

Sweeteners used in the study were 95% Reb A (ENLITEN® 30000015 High Intensity 

Sweetener, Ingredion, Westchester, IL), 95% Reb D (BESTEVIA® Reb D stevia leaf sweetener, 

 

33 

 

Ingredion, Westchester, IL), 95% Reb M (BESTEVIA® Reb M stevia leaf sweetener, Ingredion, 

Westchester, IL), and sucrose (Smidge & SpoonTM, Kroger, Cincinnati, OH). PROP (6-n-pro- 

pyl-2-thiouracil, #P3755, Sigma-Aldrich, St. Louis, MO), NaCl (Sigma-Aldrich, St. Louis, MO), 

and filter papers (1.5 dia. cm, VWR Scientific Products, West Chester, PA) were used to make 

paper disks for supertaster screening. 

2.2. PROP Status Determination 

The paper disks for PROP status determination were prepared following the method 

described by Zhao et al. [41]. Blank, NaCl, and PROP disks were prepared. Blank disks were 

used as the control. NaCl disks were made by placing filter papers in 1.0 mol/l NaCl solution for 

30 seconds at room temperature and oven-dried for 1 hour at 121 °C (250 °F). 50-mmol/l PROP 

solution at boiling temperature was used for PROP disks.  

PROP testing and classification were based on Zhao et al. [41] and Zhao and Tepper  

[37]. Michigan State University SONA Paid Research Pool (https://msucas-paid.sona-

systems.com) was used to recruit participants with age between 18 and 55. Participants were 

instructed to rinse their mouth with water, taste the paper disk for 15 seconds or until the disk is 

wet, discard the paper disk, and then rate the perceived intensity of the taste on the labeled 

magnitude scale (LMS). The participants would taste a blank, a NaCl, and a PROP disk in order 

with a 30-second break in between samples to minimize fatigue and carryover. The set was 

repeated after a 5-minute break. 

The LMS is a 100 mm quasi-logarithmic spacing vertical scale with verbal labels from 

“barely detectable” to “strongest imaginable” [42]. The scale set up was “no sensation” = 0, 

“barely detectable” = 1.5, “weak” = 6, “moderate” = 17, “strong” = 35, “very strong” = 52, and 

“strongest imaginable” = 100 [43,44]. The PROP score of participants was calculated based on 

 

34 

 

the mean of the two replicates. Because the LMS is not equal in spacing, the difference between 

two scores when both ratings are at the higher end is less than when ratings are at the lower end. 

If the difference between two ratings was bigger than 30 mm, or bigger than 40 mm when both 

ratings were higher than “very strong”, the participant would be considered having bad 

reproducibility and would not be invited to the following water solution testing. Out of 224 

participants, 27 were excluded.  

Initially, “moderate” or below (≤ 17 mm on the LMS) and “very strong” or above (≥ 52 

mm on the LMS) of PROP score were used to group participants into non-tasters and 

supertasters. The group means and 95% confidence interval were then calculated to set new cut-

off scores. The new cut-off score for non-tasters was 10.3 and for supertasters was 70.7. 

Participants with scores in between were classified as medium tasters. When the PROP score of a 

participant was at a borderline, the NaCl score was used to help classify the person [41]. A 

participant would be classified as a non-taster if the person gave a non-taster borderline score 

and rated NaCl much higher than PROP (~ 30 mm difference on the LMS). When a participant 

was at the supertaster borderline and gave a much lower NaCl score than PROP, the person 

would be classified as a supertaster. Out of 197 remaining participants, 25 were identified as 

non-tasters and 55 were supertasters.  

2.3. Subjects Demographics  

Following the PROP test, participants were asked to provide some basic demographic 

information, including age, gender, ethnicity, educational level, weight, height, health condition, 

consumption frequency of low/zero sugar added products, consumption of sweeteners on a 

regular basis (at least once a month), and familiarity with stevia.  

 

35 

 

2.4. Consumer Testing 

2.4.1. Samples and Sample Preparation 

All solutions were prepared using deionized water and the concentration of the sample is 

expressed in g/L (w/v). Sucrose at 14% was chosen as the control. Reb A, D, and M at 0.09% 
were used in a preliminary test (n = 31) to determine the relative sweetness to 14% sucrose. The 
result showed that 0.09% Reb M were not statistically different from 14% sucrose in sweetness 
intensity (P = 0.16), but there was still a 1.1 difference on a marked 15-cm line scale with 
descriptors of “not at all” and “extremely” as endpoint anchors. Another preliminary test (n = 65) 
was then conducted to prove the sweetness equivalency of Reb M to sucrose, using 0.09% and 
0.12% Reb M and 10% and 14% sucrose. The result indicated that both 0.09% and 0.12% Reb M 
were not significantly different from 14% sucrose (P = 0.34 and P = 0.11, respectively), with 
0.09% Reb M closer to 14% sucrose at 0.5 difference on a 15-cm line scale, comparing to a 1.0 
difference between 0.12% Reb M and 14% sucrose. Since 0.09% Reb M was again lower in 
intensity on the scale, Reb M at 0.10% was chosen for the consumer testing. Reb A and D at 
0.10% were used to compare the sensory characteristics of the three steviol glycosides at the 
same concentration. Thus, samples used for the testing were Reb A, D, and M at 0.10%, and 
14% sucrose. The consumer test lasted four days and fresh samples were made 1 day before 
testing each day. 10 ml of each solution was measured into a 1 oz soufflé cup and stored in the 
refrigerator (4 °C) prior to serving.  

2.4.2. Testing Procedure 

This study was approved by the University Institutional Review Board of the Michigan 

State University (East Lansing, MI) [Study ID: STUDY00004019]. SIMS 2000 software (SIMS 

Sensory Software, Morristown, NJ, USA) was used to create and administer the questionnaire.  

 

36 

 

Consumers were instructed to rate the sweetness and bitterness intensities of the solutions 

on a 15-cm line scale three times, which were while the solution was in the mouth, 5 seconds 

after expectorating it, and 1 minute after expectorating it. Consumers were asked to pinch their 

nose while holding the solution in the mouth to focus on the taste. The sweet and bitter tastes 

perceived at this time would be called in-mouth sweetness and bitterness throughout this paper. 

The perceived intensities of sweet and bitter tastes 5 seconds after expectorating would be 

referred to as immediate sweetness and bitterness. A check-all-that-apply (CATA) question on 

the aftertaste was followed after evaluating the immediate tastes, including terms collected from 

an open-ended question in the two preliminary tests (n1 = 31 and n2 = 65), asking if the 

consumers noticed any aftertaste. The term pleasant was added to the list as a positive word, and 

spicy was added as an attention check to identify careless respondents and would be removed 

from the correspondence analysis. The final list of CATA consisted of 15 terms, which were 

artificial, bitter, chemical, honey, licorice, metallic, minty, pleasant, pungent, spicy, sweet, 

tangy, tart, tingling, and vanilla, and the terms were listed in alphabetical order. A 45-second 

break was enforced after the CATA question, which was before evaluating the sweet and bitter 

tastes 1 minute after expectorating. The perceived intensities would be considered as lingering 

sweetness and bitterness. Water and crackers were provided as palate cleansers in between 

samples.  

2.5. Statistical Analyses 

Data analysis was performed using XLSTAT (AddinSoft, New York, NY). Intensity data 
were analyzed using a one-factor ANOVA model. For CATA analysis, the frequencies of each 
attribute were counted. Cochran’s Q test was performed for each attribute to compare the 
difference among samples. Multiple pairwise comparisons using critical difference (Sheskin) 

 

37 

 

were performed when the attribute was significant (P < 0.05). Correspondence analysis (CA) 
was generated to visually show the relationship between sensory attributes and samples. A two-
way ANOVA model was used to determine the effect of PROP taster status, sweetener, and their 
interaction. Fisher’s least significant difference (LSD) post hoc test was performed when P < 
0.05. Agglomerative hierarchical clustering (AHC) was used as a second way to classify PROP 
groups. Pearson correlation test was performed, and correlation coefficients were calculated 
between PROP bitterness and sweet and bitter tastes of Reb A, D, and M combined over time 
(in-mouth, immediate, lingering sweetness and bitterness).  

3. Results 

3.1. Participant Characteristics 

A total of 126 naïve consumers completed the study, with an average age of 23 ± 1.7 

years and an average BMI of 24.7 ± 4.6 kg/m2 based on self-reported height and weight. None of 

the participants had heart disease, cancer, or diabetes. The socio-demographics of participants 

are shown in Table 4. The majority were female (72.2%) and 60.3% of the participants identified 

themselves as white. Table 5 listed out the responses of sweetener consumption behavior 

questions. Sucrose (81.0%) and honey (69.8%) were the most commonly consumed sweetener 

on a regular basis (at least once a month), followed by stevia (19.8%), sucralose (19.0%), and 

aspartame (19.0%), which were high-intensity sweeteners. Other sweeteners consumed (8.7%) 

included maple syrup, brown sugar, xylitol, high fructose corn syrup, and acesulfame K. Sixty-

seven percent of participants consumed low or zero sugar added products at least once a month. 

More than half of the participants (54.8%) said they were somewhat or very familiar with stevia. 

 
 
 

 

38 

Table 4. Socio-demographic characteristics of participants (n = 126). 
Variable  Definition 
Gender 
 

 
Male 

Frequency  % 
 
35 
91 
 
76 
5 
33 
7 
0 
3 
2 
 
0 
29 
4 
50 

 
27.8% 
72.2% 
 
60.3% 
4.0% 
26.2% 
5.6% 
0.0% 
2.4% 
1.6% 
 
0.0% 
23.0% 
3.2% 
39.7% 

43 

34.1% 

  Hispanic or Latino 
  Asian or Pacific Islander 
  Black or African American 
  Native American or American Indian 
  Other 
  Prefer not to respond 

Less than high school 
High school diploma or GED 

  2-year college degree 
  4-year college degree 

Graduate degree (Master`s, Doctorate, 
etc.) 

  Female 

Ethnicity 
 

 
White 

Education 
 
 

 

  

 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

 

39 

 

Table 5. Participants’ sweetener consumption behavior (n = 126). 
Characteristic 
Sweetener consumption1 
 
 
 
 
 
 
 
 
 
 
Low/zero sugar added product 
consumption frequency 
 

Definition 
 
Agave nectar 
Aspartame 
Erythritol 
Honey 
Monk fruit extract 
Saccharin 
Stevia 
Sucralose 
Sucrose 
Others 

 
More than 3 times a week 
  1-2 times a week 
  2-3 times a month 
  Once a month 
  Every other month 
  1-2 times per 6 months 
  Less than once a year 
  Almost never 
 
Very unfamiliar 
  Somewhat unfamiliar 
  Neutral 
  Somewhat familiar 
Very familiar 

Familiarity with stevia 
 

  
1 This is a check-all-that-apply question.  

Frequency  % 
 
16 
24 
10 
88 
9 
12 
25 
24 
102 
11 
 

 
12.7% 
19.0% 
7.9% 
69.8% 
7.1% 
9.5% 
19.8% 
19.0% 
81.0% 
8.7% 

16 
29 
27 
12 
6 
15 
7 
14 
 
25 
21 
11 
55 
14 

 
12.7% 
23.0% 
21.4% 
9.5% 
4.8% 
11.9% 
5.6% 
11.1% 
 
19.8% 
16.7% 
8.7% 
43.7% 
11.1% 

3.2. Sensory Characteristics 

3.2.1 Intensities of Sweet and Bitter Tastes 

Table 6 summarizes the mean intensity ratings (± SEM) for four sweetener solutions 

evaluated by all participants. The decreasing trend in sweetness and bitterness intensities from 

in-mouth to immediate (5 seconds after expectorating the samples) to lingering (1 minute after 

 

40 

 

expectorating the samples) indicated that consumers followed the directions and evaluated the 

samples correctly, since a fading in intensity over time was expected.  

 

 

Table 6. Mean intensity scores (± SEM) of sweetener solutions by participants (n=126). 
Sweetener 
Sucrose 
Reb A 
Reb D 
Reb M 

Bitterness2 
Sweetness1 
Immediate  Lingering 
In-mouth 
Immediate 
In-mouth 
0.8 ± 0.1 b  0.6 ± 0.1c  0.4 ± 0.1b 
8.3 ± 0.33 a4  7.1 ± 0.3 b 
7.2 ± 0.3 b  6.5 ± 0.3 b 
3.5 ± 0.3 a  3.5 ± 0.3 a  1.6 ± 0.2 a 
7.8 ± 0.3 ab  7.2 ± 0.3 b  4.5 ± 0.3 ab  1.1 ± 0.2 b  1.3 ± 0.2 b  0.6 ± 0.1 b 
0.6 ± 0.1 b 
8.6 ± 0.3 a 

Lingering 
3.6 ± 0.3 b 
4.3 ± 0.3 b 

0.9 ± 0.1 

8.2 ± 0.3 a 

5.3 ± 0.3 a 

1.0 ± 0.2 b 

bc 

1,2In-mouth tastes (sweetness and bitterness) were evaluated when the solution was in the mouth; Immediate tastes 
were evaluated 5 seconds after expectorating the sample; Lingering tastes were evaluated 1 minute after expectorating 
the sample.  3Intensities were evaluated on a marked 15-cm line scale anchored with “not at all” to “extremely”. 
4Different letters in the same column show the significant differences between sample means at P < 0.05 by Fisher’s 
LSD. 

The in-mouth sweetness of 14% sucrose and 0.1% Reb M were not significantly different 
(P  =  0.55).  The  in-mouth  sweetness  of  Reb  D  was  slightly  lower  than  sucrose  but  was  still 
considered to be not different from sucrose (P = 0.19). Reb A showed significantly less in-mouth 
sweetness than Reb M and sucrose (P < 0.01 and P < 0.05, respectively). Reb M had the highest 
immediate  sweetness  among  the  samples  and  was  significantly  different  from  others.  The 
sweetness of Reb M remained the highest after one minute. The lingering sweetness of Reb M 
(intensity = 5.3) was higher than Reb D (intensity = 4.5), but the difference was not significant (P 
= 0.05). Reb D was higher in lingering sweetness than sucrose (intensity = 3.6), but it was not 
significantly different (P = 0.05). The participants rated the in-mouth bitterness of sucrose, Reb D, 
and Reb M around 1, while the rating of Reb A was at 3.5 on a 15-cm line scale. The bitterness of 
Reb A persisted after 5 seconds (intensity = 3.5). Reb D was perceived to have more immediate 
bitterness  than  sucrose  (P  <  0.05),  and  there  was  no  significant  difference  in  the  immediate 
bitterness between Reb M and sucrose (P = 0.27). While the lingering bitterness of sucrose, Reb 
D, and Reb M was at a minimum, Reb A still had detectable bitterness remaining (intensity = 1.6).  

 

41 

 

3.2.2. CATA 

Table 7 summarizes the total counts of CATA attributes selected by the consumer panel 

(n=126) to describe the aftertaste of each sweetener solution. The term sweet was the most 

frequently used term, and spicy was the least, which were as expected. Significant differences 

among samples were found in 10 out of 15 attributes (P < 0.05). Reb A, D, and M were 

described as artificial more frequent than sucrose. The bitter and chemical tastes of Reb A were 

significantly higher than other sweeteners, and fewer participants considered Reb A as sweet and 

pleasant. Honey and vanilla were checked the most for sucrose, followed by Reb D and M, while 

Reb A was rarely associated with these two terms. Licorice, metallic, minty, pungent, spicy, 

tangy, tart, and tingling were rarely selected by participants, with no more than 15 counts for 

each sample. Among those 8 less-checked terms, licorice, pungent, spicy, tangy, and tingling 

were not significantly different among samples.  

Sucrose  Reb A 

Table 7. Total counts of check-all-that-apply attributes for sweetener solutions. 
Attribute 
Artificial***  38 b 
Bitter*** 
8 b 
Chemical***  9 b 
Honey*** 
41 a 
5 
Licoricens 
6 a 
Metallic* 
Minty** 
7 ab 
65 a 
Pleasant*** 
2 
Pungentns 
Spicyns 
0 
110 a 
Sweet*** 
1 
Tangyns 
Tart* 
2 ab 
4 
Tinglingns 
Vanilla*** 
27 a 
* indicates P < 0.05, ** indicates P < 0.01, *** indicates P < 0.001, and ns indicates no significant differences 
among samples. Different letters in the same row indicate the significant differences between sample means at P < 
0.05 by Critical Difference (Sheskin). 
 

Reb M 
69 a 
12 b 
18 b 
24 b 
6 
6 a 
3 ab 
57 a 
6 
1 
114 a 
8 
0 b 
7 
15 ab 

83 a 
66 a 
42 a 
8 c 
8 
15 a 
0 b 
25 b 
7 
0 
83 b 
5 
6 a 
7 
7 b 

Reb D 
64 a 
17 b 
17 b 
25 b 
4 
6 a 
9 a 
49 a 
2 
2 
110 a 
4 
1 ab 
6 
15 ab 

 

42 

 

The sensory attributes of sweeteners were summarized visually in Figure 3. The first two 

dimensions explained 96% of the variation. Honey and vanilla were associated with sucrose. Reb 

A was close to metallic, bitter, and chemical. Reb D and M were similar to each other and were 

closer to sucrose as compared to Reb A. Reb D and M were mostly associated with the positive 

words, but artificial was between Reb A and Reb D and M.  

Figure 3. Correspondence analysis (CA) of sweeteners. Gray color indicates non-significant 
attributes; Red color indicates significant attributes; Samples are in blue.  
 

)

%
 
2
7
.
9
(
 
2
F

0.6

0.4

0.2

0

-0.2

-0.4

-0.6

 

Tart

Bitter

Licorice

Metallic

Reb A

Chemical

Vanilla

Honey

Sucrose

Minty

Pleasant

Sweet
Reb D
Reb M

Artificial

Tingling

Pungent

Tangy

-1

-0.8

-0.6

-0.4

-0.2

0

0.2

0.4

0.6

0.8

1

1.2

F1 (87.01 %)

 

3.3. PROP Bitterness 

3.3.1. PROP Taster Groups 

Out of 126 participants who completed the consumer test, there were 15 non-tasters, 81 

medium tasters, and 30 supertasters. The interaction between taster groups and samples was not 

significant for all taste evaluations of each Reb A, D, and M solutions. There was also no 

significant difference when examining the main effect of taster groups on perceived intensity 

 

43 

 

scores of sweet and bitter tastes of the three sweeteners combined over time (in-mouth, 

immediate, and lingering) (Figure 4).  

Figure 4. The influence of PROP taster groups on the perceived intensities (± SEM), with 
sweeteners combined (Reb A, D, and M).  

9
8
7
6
5
4
3
2
1
0

)
e
l
a
c
s
 

m
c
-
5
1
(
 
y
t
i
s
n
e
t
n
I

In-mouth
sweetness

Immediate
sweetness

Lingering
sweetness

In-mouth
bitterness

Immediate
bitterness

Lingering
bitterness

Attributes
Medium tasters

Non-tasters

Supertasters

 

Due to the disproportional ratio of people in each taster groups, agglomerative 

hierarchical clustering (AHC) was used to group people based on their dissimilarity on the PROP 

rating (data not shown). Three groups were generated with 55, 44, and 30 people, corresponding 

to low, medium, and high-sensitive clusters, respectively. However, no significant difference in 

the main effect of clusters was found.  

3.3.2. Relationships with Perceived Intensities of Reb A, D, and M 

Pearson  correlation  tests  were  conducted  to  determine  the  association  between  PROP 
bitterness and perceived intensities of sweet and bitter tastes of Reb A, D, and M over time (in-
mouth, immediate, lingering). No significant relationships existed between PROP bitterness and 
the rated intensities of the three steviol glycosides (P > 0.05).  

 

 

 

44 

 

4. Discussion 

The present study investigated the sweetness and bitterness of Reb A, D, and M compared 
to sucrose at a high sucrose equivalent level (14% w/v) using consumers. To compare with 14% 
(w/v) sucrose solution, the solution concentration of the three steviol glycosides was determined 
by two small scale consumer tests as a preliminary test (see 2.4.1 for details). Briefly, the sweetness 
of 0.1% (w/v) Reb M was proved to be not significantly different from a 14% sucrose solution, 
and the same concentration was used for Reb A and D to compare the sensory characteristics of 
the three steviol glycosides at the same concentration. Prakash et al. [16] estimated that Reb M 
was about 200-350 times sweeter than sucrose, and the sweet potency at 10% SE was calculated 
to be 159. In the present study, the sweet potency of Reb M at 14% SE was calculated as 140. This 
is in line with the model from Prakash et al. [16], sweet potency of high-intensity sweeteners 
tended to decrease as the sucrose sweetness equivalent level increased [45].  

The three steviol glycosides showed significant differences in sweetness and bitterness at 
the same concentration (0.1% w/v). The in-mouth sweetness of Reb D and M at 0.1% were not 
statistically different from sucrose at 14%, while 0.1% Reb A was less sweet than sucrose. Reb A 
was significantly less sweet than Reb M as well (P < 0.01) but was not significantly different from 
the in-mouth sweetness of Reb D (P = 0.24) with a tendency of being less sweet. Reb D and M 
were not significantly different in in-mouth sweetness (P = 0.06), but there was also a clear 
tendency of Reb D to be less sweet than Reb M. These results were consistent with the previous 
studies investigating the sweetness of Reb A, D, and/or M at different concentrations showing that 
Reb M was the sweetest sweetener and Reb A was the least sweet sweetener among Reb A, D, and 
M at the same concentration [15,16,46]. 

 

45 

 

The sweetness temporal profile of Reb M was studied by Prakash et al. [16], who compared 
the sweetness appearance time and extinction time to examine the change in perception over 3 
minutes. The sweetness of Reb M elicited later and persisted longer than sucrose at 10% SE in 
water. The descriptive panel rated the lingering sweetness of Reb M higher than that of sucrose as 
well [16]. In the present study, Reb M had a similar in-mouth sweetness to sucrose, but the 
lingering  sweetness  was  significantly  higher  than  that  of  sucrose,  which  corresponded  with 
Prakash’s finding. Reb A was also found to have a longer extinction time than sucrose [13] and 
exhibited persistent flavor duration in the mouth [8]. When at a similar sweetness level (i.e., at 8% 
SE), the lingering sweetness of Reb A and Reb M were not different [16]. Even though, in this 
study, there was no significant difference in lingering sweetness between sucrose and Reb A (P = 
0.12), the lingering sweetness of Reb A became higher than sucrose after being rated less sweet 
in-mouth, which suggested that if the sweetness of Reb A was at the same level as sucrose, the 
lingering sweetness might be significantly higher than sucrose. Reb D, like Reb M, also had a 
similar in-mouth sweetness to sucrose, but had marginally higher lingering sweetness than sucrose 
(P = 0.05). When comparing Reb D to Reb M, the lingering sweetness of Reb D was marginally 
less than that of Reb M (P = 0.05) similar to the in-mouth sweetness of Reb D that was almost 
significantly less than Reb M (P = 0.06). Although the lingering sweetness of Reb M seemed to 
be stronger than Reb D, it may due to its higher initial sweetness than Reb D. 

The bitterness of Reb A stood out among the samples when consumers first tasted the 
sample and the bitterness continued to be significantly different from others even after one minute. 
Many other researchers have reported the bitterness of Reb A [13,15]. On the other hand, Reb D 
and M did not show much in-mouth bitterness and had a similar intensity to sucrose. Even though 
Reb D exhibited a significantly higher immediate bitterness than sucrose, it was still considered 

 

46 

 

low. A trained panel did not detect any significant bitter taste of Reb M when comparing to sucrose 
at 10% SE [16]. Hellfritsch et al. [15] and Ko et al. [47] indicated that Reb D elicited a lot less 
bitterness than Reb A. Our results confirmed that naïve consumers like trained assessors did not 
detect much bitterness from Reb D and M. 

Based on the total counts of CATA and the CA, Reb A was associated with some negative 
perception terms, such as bitter, chemical, and artificial. The bitter and chemical attributes were 
significantly more selected for Reb A than Reb D, Reb M, and sucrose. The bitter attribute was in 
agreement with the bitterness intensity rating. The bitterness and chemical sensation of Reb A was 
reported by Fujimaru et al. [48] as well. Significantly less sweet was checked for Reb A than the 
other sweeteners, suggesting that the bitterness and chemical sensation might overshadow the 
sweetness of it. Reb D and M appeared to have good taste profiles because they were close to 
positive terms and sucrose. However, even though Reb D and M had low citations for bitter and 
higher citations for pleasant than Reb A, many participants still checked artificial significantly 
more frequent than sucrose. Waldrop and Ross [49] reported that consumers did not like stevia 
because of its association with artificial flavor. Thus, the artificial attribute may cause negative 
consumer perception of Reb D and M even though they are natural sweeteners without the aversive 
bitter  aftertaste.  Interestingly,  the  artificial  attribute  was  also  selected  for  sucrose  by  38 
participants. It is not common to drink pure sugar water in daily life, so the participants may not 
be familiar with the taste of sucrose solutions, and thus might select artificial for sucrose solution.  
Licorice, pungent, spicy, tangy, and tingling were rarely selected by the participants and 
were not significant to discriminate the samples. Thus, these five terms may not be appropriate 
terms  for  consumers  to  describe  the  three  steviol  glycosides,  even  though  licorice  has  been 
commonly used to describe the aftertaste of Reb A by researchers [10,13,50] and media [18]. The 

 

47 

 

licorice taste of Reb A did not exhibit at low SE levels, but was elicited at higher SE levels [13], 
and this was further proved by Reyes et al. [50] that Reb A at 0.1% had more notable licorice taste 
than at 0.012%. In this study, we did not find the correlation between licorice and Reb A at 0.1%, 
since only 8 people out of 126 selected it, which suggested that licorice may not be an appropriate 
term for consumers to describe the aftertaste of Reb A or to discriminate Reb A, D, and M.  

The CATA analysis also found that vanilla and honey were associated with sucrose. A 
consumer survey showed that honey was the most popular sugar alternative, which was natural, 
and natural sweeteners were perceived better than artificial sweeteners in general [3]. In this study, 
those who checked honey for steviol glycosides might imply that the sample gave them a sense of 
natural. As for vanilla, Lavin and Lawless [51] showed that an added vanilla flavor enhanced the 
perception of sweetness in milk, and Wang et al. [52] also indicated a taste-aroma interaction 
between  perceived  sweetness  and  vanilla  flavor  in  skim  milk.  Vanilla  was  congruent  with 
sweetness, so participants might choose the term even though the attribute was not presented in 
the solution.  

A secondary objective of this study was to investigate the influence of consumers’ PROP 
taster status on the sweetness and bitterness of Reb A, D, and M. We found that there were no 
significant differences in the perceived sweetness and bitterness of Reb A, D, and M (in-mouth, 
immediate, and lingering) among PROP taster groups. Risso et al. [40] reported that there was no 
correlation between PROP bitterness and stevioside bitterness. Humans have about 25 bitter taste 
receptors  from  the  taste  2  receptors  (hTAS2Rs)  gene  family  [53].  Each  receptor  responds  to 
different  compounds  but  may  have  overlapped  molecular  range  [54].  The  sensitivity  to  the 
bitterness of PROP/PTC is mainly associated with TAS2R38 bitter taste receptor [55,56]. TAS2R4 
and TAS2R14 responded to the bitterness of stevioside and Reb A in vitro, while TAS2R38 did 

 

48 

 

not react [15]. Meyerhof et al. [54] sorted receptors into 4 groups, and both TAR2S4 and TAS2R14 
were not in the same group as TAS2R38. The different responses in bitter taste receptors might 
explain  why  no  relationship  was  found  between  PROP  taster  status  and  perceived  bitterness 
intensity of Reb A, D, and M. 

Some studies suggested that PROP bitterness sensitivity influenced other oral sensations, 
such as sweetness [36,37,57-61]. Drewnowski et al. [36] found a weak and marginal significant 
difference in sweetness perception of sucrose and saccharin between PROP tasters and non-tasters, 
and the difference was more significant at lower concentrations. Allen et al. [59] reported that the 
sweetness  of  acesulfame  potassium  was  positively  associated  with  PROP  bitterness.  A  large 
sample size study (n > 1,500) found a weak association between sweetness and PROP bitterness, 
suggesting that a bigger size sample is required to detect weak association with PROP [60]. A 
recent study confirmed that PROP bitterness was positively correlated with sweetness of sucrose 
[61]. However, some of the previous studies also indicated that there was no relationship between 
PROP  sensitivity  and  sweet  taste  responsiveness  [62-64].  Here,  we  found  no  significant 
differences  in  perceived  sweetness  intensity  among  PROP  groups  and  further,  no  correlation 
between PROP bitterness and sweetness of the steviol glycosides. In this study, the test stimuli 
were  singles  (i.e.,  each  sweetener  solution),  but  the  aftertaste  of  the  three  steviol  glycosides, 
especially the sweet-bitter Reb A at a high concentration, might cause difficulties for participants 
to evaluate intensities of sweet and bitter tastes. Horne et al. [38] reported that sweet-bitter stimuli 
might  be  more  difficult  to  evaluate  than  single  taste  stimuli  due  to  taste-taste  interactions. 
Expansive, linear, and compressive phases of psychophysical functions could be used to predict 
how taste stimuli would behave when mixed at low, medium, and high intensity/concentration 
[65]. For example, perceptual enhancement and suppression has been extensively reported at low 

 

49 

 

and high intensity/concentration mixtures, corresponding to the expansive phase and compressive 
phase  of  the  psychophysical  function,  respectively  [65].  Ly  and  Drewnowski  [63]  showed  a 
reduced difference in bitterness between PROP taster groups was found when the caffeine solution 
was sweetened, even though PROP tasters rated caffeine solution without sweetener as more bitter 
than non-tasters [63]. The perceptual suppression as a result of sweet-bitter interaction at a high 
intensity/concentration may explain no differences in perceived sweetness intensity among PROP 
groups and no correlations between PROP bitterness and perceived sweetness of Reb A, D, and 
M.  

One possible limitation of the study was that consumers did not swallow the solution, 
which limited the number of taste buds utilized for the evaluation. Taste buds are distributed not 
only in the mouth but also in the throat [66]. Consumers were asked to expectorate the sample to 
reduce  fatigue,  however,  the  swallowing  sensation  could  be  different  and  might  impact  the 
perceived intensities. No hedonic question was asked in this study because it might be difficult for 
naïve  consumers  to  rate  the  likings  of  pure  sweetener  solutions  when  the  solutions  were  not 
regularly consumed in daily life. However, no association could be drawn between the negative 
CATA attributes and the likings of the sweeteners. Another limitation was the disproportional size 
of PROP groups, which only had 15 non-tasters. The data from non-tasters might be less variable 
if more non-tasters were recruited. 

5. Conclusions 

The present study investigated the sensory profile of Reb A, D, and M at 14% SE using a 
consumer panel, and the influence of PROP taster groups on the perceived sweet and bitter tastes 
of the three glycosides. Reb D and M had sensory profiles that were closer to sucrose, compared 
to Reb A, but were still associated with negative sensation, such as artificial, which may cause 

 

50 

 

negative perception toward Reb D and M. The lingering sweetness of Reb D and M was also a 
concern. The sensory characteristics of Reb A, D, and M in this study can be used as a reference 
for the food industries working with steviol glycosides in high-sugar applications, such as frozen 
desserts. Furthermore, there were no significant differences among non-tasters, medium tasters, 
and supertasters on the perceived sweetness and bitterness of Reb A, D, and M as well as no 
significant correlations between PROP bitterness and perceived sweet and bitter tastes, 
suggesting that supertasters who experience more intense taste sensations may not report 
aversive sensations from stevia. Further studies on the consumer acceptance of Reb A, D, and M 
are needed to determine if these characteristics would affect the likings of these sweeteners. 
Since the sweeteners may perform differently in a food matrix than in aqueous solution, more 
research using steviol glycosides in final food products are needed to determine the sensory 
profile and acceptance of them.  

 

Author Contributions 

R.T. contributed to study design, data collection, data analysis, and manuscript writing. 

S.C. contributed to study design, data collection, data analysis, and critical revision of the 

manuscript. S.C. supervised the project.  

 

Funding 

This work was supported by the United States Department of Agriculture Specialty Crop 

Research Initiative [grant number 2017-5181-26828]. 

 

 
 

 

51 

 

Acknowledgments 

The authors thank Ingredion Incorporated for providing Reb A, D, and M. 

Conflicts of Interest 

The authors declare no conflict of interest. 

 

 

52 

 

 
 

 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

REFERENCES 

 

53 

 

REFERENCES 

 

1.  Chattopadhyay, S.; Raychaudhuri, U.; Chakraborty, R. Artificial sweeteners – a review. J. 

Food Sci Technol 2014, 51, 611–621, doi:10.1007/s13197-011-0571-1. 

 

203-16594-2. 

2.  Ott, D.B.; Edwards, C.L.; Palmer, S.J. Perceived taste intensity and duration of nutritive and 

non-nutritive sweeteners in water using time-intensity (T-I) evaluations. Journal of Food 
Science 1991, 56, 535–542, doi:10.1111/j.1365-2621.1991.tb05319.x. 

 
3.  Mintel. Sugar and alternative sweeteners - US - December 2018 - Market research report. 
Available online: https://reports-mintel-com.proxy2.cl.msu.edu/display/860879/ (accessed 
on Jun 25, 2020). 

 
4.  Brandle, J.E.; Starratt, A.N.; Gijzen, M. Stevia rebaudiana : Its agricultural, biological, and 
chemical properties. Can. J. Plant. Sci. 1998, 78, 527–536, doi:10.4141/P97-114. 
 
5.  Geuns, J.M.C. Stevioside. Phytochemistry 2003, 64, 913–921, doi:10.1016/S0031-
 
6.  Kinghorn, A.D. Stevia: The genus Stevia; Taylor & Francis: London, 2002; ISBN 978-0-
 
7.  Mintel Global New Products Database (GNPD). Available online: 
 
8.  Gwak, M.-J.; Chung, S.-J.; Kim, Y.J.; Lim, C.S. Relative sweetness and sensory 

https://www.mintel.com/global-new-products-database (accessed on Jun 25, 2020). 

9422(03)00426-6. 

characteristics of bulk and intense sweeteners. Food Sci Biotechnol 2012, 21, 889–894, 
doi:10.1007/s10068-012-0115-0. 

978-1-85166-364-4. 

 
9.  Jenner, M.R. Progress in sweeteners; Grenby, T.H., Ed.; Elsevier: London, 1989; ISBN 
 
10.  Kim, M.-J.; Yoo, S.-H.; Jung, S.; Park, M.-K.; Hong, J.-H. Relative sweetness, sweetness 
quality, and temporal profile of xylooligosaccharides and luo han guo (Siraitia grosvenorii) 
extract. Food Sci Biotechnol 2015, 24, 965–973, doi:10.1007/s10068-015-0124-x. 

 
11.  Li, X.E.; Lopetcharat, K.; Drake, M.A. Parents’ and children’s acceptance of skim chocolate 

milks sweetened by monk fruit and stevia leaf extracts. Journal of Food Science 2015, 80, 
S1083–S1092, doi:10.1111/1750-3841.12835. 

 
12.  Medeiros, A.C.; Filho, E.R.T.; Bolini, H.M.A. Impact of natural and artificial sweeteners 
compounds in the sensory profile and preference drivers applied to traditional, lactose-free, 
and vegan frozen desserts of chocolate flavor. Journal of Food Science 2019, 84, 2973–
2982, doi:10.1111/1750-3841.14806. 

 

54 

 

 

13.  Prakash, I.; DuBois, G.E.; Clos, J.F.; Wilkens, K.L.; Fosdick, L.E. Development of rebiana, 

a natural, non-caloric sweetener. Food and Chemical Toxicology 2008, 46, S75–S82, 
doi:10.1016/j.fct.2008.05.004. 

 
14.  Gray, N. Evolva and Cargill publish patent application for ‘next generation’ fermented 

sweeteners. Available online: https://www.foodnavigator.com/Article/2014/08/18/Evolva-
and-Cargill-publish-patent-application-for-next-generation-fermented-sweeteners (accessed 
on Jun 25, 2020). 

 
15.  Hellfritsch, C.; Brockhoff, A.; Stähler, F.; Meyerhof, W.; Hofmann, T. Human psychometric 

and taste receptor responses to steviol glycosides. J. Agric. Food Chem. 2012, 60, 6782–
6793, doi:10.1021/jf301297n. 

 
16.  Prakash, I.; Markosyan, A.; Bunders, C. Development of next generation stevia sweetener: 
 
17.  PureCircle. Stevia use in food and beverages accelerated significantly in 2018. Available 

Rebaudioside M. Foods 2014, 3, 162–175, doi:10.3390/foods3010162. 

online: https://purecircle.com/news/stevia-use-in-food-and-beverages-accelerated-
significantly-in-2018/ (accessed on Jun 25, 2020). 

 
18.  Watson, E. GLG: We’ll be able to breed stevia plants with more Reb, D., Reb, M. Available 

online: https://www.foodnavigator.com/Article/2015/09/16/GLG-We-ll-be-able-to-breed-
stevia-plants-with-more-Reb-D-Reb-M (accessed on Jun 25, 2020). 

 
19.  Watson, E. Amyris bids for 30% slice of stevia sweetener market by 2022 with Reb, M. 

Available online: https://www.foodnavigator.com/Article/2018/12/03/Amyris-bids-for-30-
slice-of-stevia-sweetener-market-by-2022-with-Reb-M (accessed on Jun 25, 2020). 
 
20.  Goff, H.D. Ice cream and frozen desserts. In Ullmann’s Encyclopedia of Industrial 
Chemistry; American Cancer Society, 2015; pp. 1–15 ISBN 978-3-527-30673-2. 
 
21.  Varela, P.; Ares, G. Sensory profiling, the blurred line between sensory and consumer 

science. A review of novel methods for product characterization. Food Research 
International 2012, 48, 893–908, doi:10.1016/j.foodres.2012.06.037. 

 
22.  Moskowitz, H.R. Experts versus consumers: A comparison. Journal of Sensory Studies 
 
23.  Husson, F.; Le Dien, S.; Pagès, J. Which value can be granted to sensory profiles given by 

1996, 11, 19–37, doi:10.1111/j.1745-459X.1996.tb00030.x. 

consumers? Methodology and results. Food Quality and Preference 2001, 12, 291–296, 
doi:10.1016/S0950-3293(01)00014-3. 

 
24.  Worch, T.; Lê, S.; Punter, P. How reliable are the consumers? Comparison of sensory 
profiles from consumers and experts. Food Quality and Preference 2010, 21, 309–318, 
doi:10.1016/j.foodqual.2009.06.001. 

 

55 

 

25.  Ares, G.; Bruzzone, F.; Giménez, A. Is a consumer panel able to reliably evaluate the texture 

of dairy desserts using unstructured intensity scales? Evaluation of global and individual 
performance. Journal of Sensory Studies 2011, 26, 363–370, doi:10.1111/j.1745-
459X.2011.00352.x. 

 
26.  Moskowitz, H.R. Base size in product testing: A psychophysical viewpoint and analysis. 
Food Quality and Preference 1997, 8, 247–255, doi:10.1016/S0950-3293(97)00003-7. 
 
27.  Ares, G.; Tárrega, A.; Izquierdo, L.; Jaeger, S.R. Investigation of the number of consumers 
necessary to obtain stable sample and descriptor configurations from check-all-that-apply 
(CATA) questions. Food Quality and Preference 2014, 31, 135–141, 
doi:10.1016/j.foodqual.2013.08.012. 

 
28.  Dooley, L.; Lee, Y.; Meullenet, J.-F. The application of check-all-that-apply (CATA) 
consumer profiling to preference mapping of vanilla ice cream and its comparison to 
classical external preference mapping. Food Quality and Preference 2010, 21, 394–401, 
doi:10.1016/j.foodqual.2009.10.002. 

 
29.  Adams, J.; Williams, A.; Lancaster, B.; Foley, M. Advantages and uses of check-all-that-
apply response compared to traditional scaling of attributes for salty snacks. 7th Pangborn 
sensory science symposium 2007, 16. 

 
30.  Ares, G.; Barreiro, C.; Deliza, R.; Giménez, A.; Gámbaro, A. Application of a check-all-

that-apply question to the development of chocolate milk desserts. Journal of Sensory 
Studies 2010, 25, 67–86, doi:10.1111/j.1745-459X.2010.00290.x. 

 
31.  Ares, G.; Varela, P.; Rado, G.; Giménez, A. Are consumer profiling techniques equivalent 
for some product categories? The case of orange-flavoured powdered drinks. International 
Journal of Food Science & Technology 2011, 46, 1600–1608, doi:10.1111/j.1365-
2621.2011.02657.x. 

 
32.  Plaehn, D. CATA penalty/reward. Food Quality and Preference 2012, 24, 141–152, 
 
33.  Bartoshuk, L.M.; Duffy, V.B.; Miller, I.J. PTC/PROP tasting: Anatomy, psychophysics, and 

doi:10.1016/j.foodqual.2011.10.008. 

sex effects. Physiology & Behavior 1994, 56, 1165–1171, doi:10.1016/0031-
9384(94)90361-1. 

 
34.  Simons, C.T.; Adam, C.; LeCourt, G.; Crawford, C.; Ward, C.; Meyerhof, W.; Slack, J.P. 

The “bitter-sweet” truth of artificial sweeteners. In Sweetness and Sweeteners; Weerasinghe, 
D.K., DuBois, G.E., Eds.; ACS Symposium Series; American Chemical Society: 
Washington, DC, 2008; Vol. 979, pp. 335–354 ISBN 978-0-8412-7432-7. 

 
35.  Bartoshuk, L.M. Bitter taste of saccharin related to the genetic ability to taste the bitter 
 

substance 6-n-Propylthiouracil. Science 1979, 205, 934–935. 

 

56 

36.  Drewnowski, A.; Henderson, S.A.; Shore, A.B. Genetic sensitivity to 6-n-propylthiouracil 
(PROP) and hedonic responses to bitter and sweet tastes. Chem. Senses 1997, 22, 27–37, 
doi:10.1093/chemse/22.1.27. 

 
37.  Zhao, L.; Tepper, B.J. Perception and acceptance of selected high-intensity sweeteners and 
blends in model soft drinks by propylthiouracil (PROP) non-tasters and super-tasters. Food 
Quality and Preference 2007, 18, 531–540, doi:10.1016/j.foodqual.2006.07.004. 

 
38.  Horne, J.; Lawless, H.T.; Speirs, W.; Sposato, D. Bitter taste of saccharin and acesulfame-K. 
 
39.  Rankin, K.M.; Godinot, N.; Tepper, B.J.; Kirkmeyer, S.V.; Christensen, C.M. Assessment of 

Chem Senses 2002, 27, 31–38, doi:10.1093/chemse/27.1.31. 

different methods for PROP status classification. In Genetic Variation in Taste Sensitivity; 
Prescott, J., Tepper, B.J., Eds.; CRC Press, 2003; pp. 63–88 ISBN 978-0-203-02343-3. 

 
40.  Risso, D.; Morini, G.; Pagani, L.; Quagliariello, A.; Giuliani, C.; De Fanti, S.; Sazzini, M.; 

Luiselli, D.; Tofanelli, S. Genetic signature of differential sensitivity to stevioside in the 
Italian population. Genes Nutr 2014, 9, s12263–s014. 

 
41.  Zhao, L.; Kirkmeyer, S.V.; Tepper, B.J. A paper screening test to assess genetic taste 

sensitivity to 6-n-propylthiouracil. Physiology & Behavior 2003, 78, 625–633, 
doi:10.1016/S0031-9384(03)00057-X. 

 
42.  Green, B.G.; Dalton, P.; Cowart, B.; Shaffer, G.; Rankin, K.; Higgins, J. Evaluating the 

‘labeled magnitude scale’ for measuring sensations of taste and smell. Chem Senses 1996, 
21, 323–334, doi:10.1093/chemse/21.3.323. 

 
43.  Bartoshuk, L.M. Comparing sensory experiences across individuals: Recent psychophysical 

advances illuminate genetic variation in taste perception. Chemical Senses 2000, 25, 447–
460, doi:10.1093/chemse/25.4.447. 

 

 

 
44.  Low, J.Y.Q.; McBride, R.L.; Lacy, K.E.; Keast, R.S.J. Psychophysical evaluation of 

sweetness functions across multiple sweeteners. Chem Senses 2017, 42, 111–120, 
doi:10.1093/chemse/bjw109. 

 
45.  Wee, M.; Tan, V.; Forde, C. A comparison of psychophysical dose-response behaviour 
 
46.  Kinghorn, A.D.; Kim, N.C.; Kim, D.H.L. Terpenoid glycoside sweeteners. In Naturally 

across 16 sweeteners. Nutrients 2018, 10, 1632, doi:10.3390/nu10111632. 

Occurring Glycosides; Ikan, R., Ed.; John Wiley & Sons, 1999; pp. 399–429 ISBN 978-0-
471-98602-7. 

 
47.  Ko, W.-W.; Kim, S.-B.; Chung, S.-J. Effect of concentration range on the accuracy of 
measuring sweetness potencies of sweeteners. Food Quality and Preference 2020, 79, 
103753, doi:10.1016/j.foodqual.2019.103753. 

 

57 

48.  Fujimaru, T.; Park, J.-H.; Lim, J. Sensory characteristics and relative sweetness of tagatose 
and other sweeteners. Journal of Food Science 2012, 77, S323–S328, doi:10.1111/j.1750-
3841.2012.02844.x. 

 
49.  Waldrop, M.E.; Ross, C.F. Sweetener blend optimization by using mixture design 

methodology and the electronic tongue. Journal of Food Science 2014, 79, S1782–S1794, 
doi:10.1111/1750-3841.12575. 

 
50.  Reyes, M.M.; Castura, J.C.; Hayes, J.E. Characterizing dynamic sensory properties of 

nutritive and nonnutritive sweeteners with temporal check-all-that-apply. Journal of Sensory 
Studies 2017, 32, e12270, doi:10.1111/joss.12270. 

 
51.  Lavin, J.G.; Lawless, H.T. Effects of color and odor on judgments of sweetness among 
children and adults. Food Quality and Preference 1998, 9, 283–289, doi:10.1016/S0950-
3293(98)00009-3. 

 
52.  Wang, G.; Hayes, J.E.; Ziegler, G.R.; Roberts, R.F.; Hopfer, H. Dose-response relationships 

for vanilla flavor and sucrose in skim milk: evidence of synergy. Beverages 2018, 4, 73, 
doi:10.3390/beverages4040073. 

 
53.  Chandrashekar, J.; Mueller, K.L.; Hoon, M.A.; Adler, E.; Feng, L.; Guo, W.; Zuker, C.S.; 

Ryba, N.J.P. T2Rs function as bitter taste receptors. Cell 2000, 100, 703–711, 
doi:10.1016/S0092-8674(00)80706-0. 

 
54.  Meyerhof, W.; Batram, C.; Kuhn, C.; Brockhoff, A.; Chudoba, E.; Bufe, B.; Appendino, G.; 
Behrens, M. The molecular receptive ranges of human TAS2R bitter taste receptors. Chem 
Senses 2010, 35, 157–170, doi:10.1093/chemse/bjp092. 

 
55.  Duffy, V.B.; Davidson, A.C.; Kidd, J.R.; Kidd, K.K.; Speed, W.C.; Pakstis, A.J.; Reed, 
D.R.; Snyder, D.J.; Bartoshuk, L.M. Bitter receptor gene (TAS2R38), 6-n-Propylthiouracil 
(PROP) bitterness and alcohol intake. Alcoholism: Clinical and Experimental Research 
2004, 28, 1629–1637, doi:10.1097/01.ALC.0000145789.55183.D4. 

 
56.  Kim, U.; Jorgenson, E.; Coon, H.; Leppert, M.; Risch, N.; Drayna, D. Positional cloning of 

the human quantitative trait locus underlying taste sensitivity to phenylthiocarbamide. 
Science 2003, 299, 1221–1225, doi:10.1126/science.1080190. 

 
57.  Hayes, J.E.; Bartoshuk, L.M.; Kidd, J.R.; Duffy, V.B. Supertasting and PROP bitterness 

depends on more than the TAS2R38 gene. Chem Senses 2008, 33, 255–265, 
doi:10.1093/chemse/bjm084. 

 
58.  Tepper, B.J. Nutritional implications of genetic taste variation: The role of PROP sensitivity 

 

 

 

and other taste phenotypes. Annu. Rev. Nutr. 2008, 28, 367–388, 
doi:10.1146/annurev.nutr.28.061807.155458. 

58 

59.  Allen, A.L.; McGeary, J.E.; Knopik, V.S.; Hayes, J.E. Bitterness of the non-nutritive 

sweetener acesulfame potassium varies with polymorphisms in TAS2R9 and TAS2R31. 
Chem Senses 2013, 38, 379–389, doi:10.1093/chemse/bjt017. 

 
60.  Hwang, L.-D.; Breslin, P.A.S.; Reed, D.R.; Zhu, G.; Martin, N.G.; Wright, M.J. Is the 

Association Between Sweet and Bitter Perception due to Genetic Variation? Chem Senses 
2016, 41, 737–744, doi:10.1093/chemse/bjw083. 

 
61.  Nolden, A.A.; McGeary, J.E.; Hayes, J.E. Predominant Qualities Evoked by Quinine, 
Sucrose, and Capsaicin Associate With PROP Bitterness, but not TAS2R38 Genotype. 
Chem Senses 2020, 45, 383–390, doi:10.1093/chemse/bjaa028. 

 
62.  Drewnowski, A.; Henderson, S.A.; Shore, A.B.; Barratt-Fornell, A. Nontasters, tasters, and 
supertasters of 6-n-Propylthiouracil (PROP) and hedonic response to sweet. Physiology & 
Behavior 1997, 62, 649–655, doi:10.1016/S0031-9384(97)00193-5. 

 
63.  Ly, A.; Drewnowski, A. PROP (6-n-Propylthiouracil) Tasting and sensory responses to 

caffeine, sucrose, neohesperidin dihydrochalcone and chocolate. Chem Senses 2001, 26, 41–
47, doi:10.1093/chemse/26.1.41. 

perception. Chem Senses 2008, 33, 493–501, doi:10.1093/chemse/bjn016. 

 
64.  Lim, J.; Urban, L.; Green, B.G. Measures of individual differences in taste and creaminess 
 
65.  Keast, R.S.J.; Breslin, P.A.S. An overview of binary taste–taste interactions. Food Quality 
 
66.  Kikut-Ligaj, D.; Trzcielińska-Lorych, J. How taste works: Cells, receptors and gustatory 

and Preference 2003, 14, 111–124, doi:10.1016/S0950-3293(02)00110-6. 

perception. Cell. Mol. Biol. Lett. 2015, 20, 699–716, doi:10.1515/cmble-2015-0042. 

 

 

 

 

 

59 

 

CHAPTER 3: NEW PROTOCOL FOR THE #6 SENSOR ARRAY OF ASTREE 
ELECTRONIC TONGUE TO DISCRIMINATE STEVIA REBAUDIANA LEAVES  

 

Abstract 

 

Electronic tongue (E-tongue), which is an analytical instrument for the sensory evaluation 

of taste, has been used to analyze a variety of foods and beverages. However, a new protocol is 

needed for the E-tongue analysis sequence and data selection with the most updated #6 sensor 
array. The objective of this study was to develop a protocol for E-tongue to discriminate Stevia 
rebaudiana leaves for its potential use in stevia breeding programs as a method of testing flavor 
quality. The E-tongue analysis sequence was set to repeat 6 times to obtain 6 data points and 
remove the first two and last data points because of sensor error. The run would be repeated, and 
the second run would be used for analysis. The relative standard deviation (%RSD) needed to be 
lower than 5% to be considered good for further analysis and the sensor selection was depended 
on the discrimination power. Rebaudioside (Reb) A, D, and M, which are three steviol 
glycosides in the stevia leaves, were used for the protocol validation. Seven stevia leaf samples 
from each of the two stevia breeding programs were analyzed by E-tongue to test its ability to 
discriminate stevia leaf samples from different lines. The results showed that E-tongue 
successfully separated Reb A, D, and M and distinguished among stevia leaf samples by using 
the new protocol, which meant that E-tongue has the potential to be used in stevia breeding 
programs for flavor selection. Further analyses, such as human panel, are needed to correlate to 
E-tongue data and determine if E-tongue can be used as a fast and accurate way to determine the 
flavor profile of stevia leaf samples. 

 
 

 

 

60 

 

1. Introduction 

Electronic tongue (E-tongue) is an analytical instrument developed for the sensory 

analysis of taste. It mimics human sensations and evaluates the taste by analyzing dissolved 

compounds in a liquid matrix. AlphaMOS (Toulouse, France) is one of the leading E-tongue 

companies in the world. Its E-tongue includes an autosampler system, seven sensors, a reference 

electrode, a stirring rod, an electronic unit, and computer software for statistical analysis. The 

sensors are potentiometric sensors, which measure the voltage difference between the sensor 

membrane and the reference electrode and send electric signals to the computer for analysis. The 

sensors have different coated membranes that give each sensor different selectivity and 

sensitivity (AlphaMOS, 2020).  

E-tongue has been used in foods and beverages in many different ways, such as to 

differentiate coffees (Dong et al., 2017) and milk and yogurts (Hruškar et al., 2009), to detect 

bitterness and astringency of green tea (Zou et al., 2018), to evaluate apple juices (Bleibaum et 

al., 2002), wines (Kirsanov et al., 2012), Korean fermented soybean paste (doenjang) (Jung et al., 

2017), dry-aged beef (Lee et al., 2019), cheeses (Lipkowitz et al., 2018; Schlossareck & Ross, 

2019), and sweet and salt solutions (Waldrop & Ross, 2014; Barnett et al., 2019). One of the 

limitations of the E-tongue is that it can only analyze liquid without particles. However, solid 

samples can still be analyzed by homogenization and extraction. For example, ground meat 

samples were homogenized and centrifuged to get the supernatants for analysis (Lee et al., 

2019). The E-tongue sensors could also be damaged if the sample is high in fat. Therefore, lipid-

free cheese extracts were made to analyze cheese samples (Lipkowitz et al., 2018; Schlossareck 

& Ross, 2019). When E-tongue is used solely, it does not tell much other than differentiating the 

samples. For example, Hruškar et al. (2009) differentiated 5 different brands of milk and yogurt 

 

61 

 

samples using E-tongue only and stated that E-tongue could serve as a fast and accurate analysis 

instrument for evaluation of samples in the dairy industry. On the other hand, when the E-tongue 

is used with other analytical methods, researchers can correlate quality parameters or sensory 

data to E-tongue data and even predict attributes (Bleibaum et al., 2002; Dong et al., 2017; 

Kirsanov et al., 2012; Waldrop & Ross, 2014). For example, Dong et al. (2017) used E-tongue to 

classify the varieties of 126 roasted coffee beans successfully and the R2 of the predictive model 

based on the E-tongue data ranged from 0.879 to 0.933 for calibration set and from 0.855 to 

0.892 for prediction set when correlating to flavor quality parameters (pH, total solids (TS), total 

titratable acidity (TA), total soluble solids (TSS), and TSS/TA ratio). Waldrop and Ross (2014) 

evaluated sweetener solutions using E-tongue, a trained panel, and consumers and showed that 

the E-tongue data was strongly correlated with 10 of the 16 intensity scores rated by consumers 

and trained panel, with R2 ranging from 0.79 to 0.95, suggesting that E-tongue has a potential to 

be used to predict sensory characteristics of unknown samples.  

Previous studies using AlphaMOS’s E-tongue used earlier generation sensor arrays, such 

as sensor array #5 (Barnett et al., 2019; Jung et al., 2017; Lee et al., 2019; Lipkowitz et al., 2018; 

Schlossareck & Ross, 2019; Waldrop & Ross, 2014). The sensor array #5 has each sensor 

measured a single sensation: sweetness (SWS), sourness (SRS), saltiness (STS), bitterness 

(BRS), umami (UMS), metallic (GPS), and spiciness (SPS), but it was discontinued by the 

manufacturer in late 2018. The sensory array #1 for soft drinks, beers, and flavors application, 

which was used by Bleibaum et al. (2002) and Dong et al. (2017), and #2 for pharmaceutical 

formulation were discontinued and replaced by the most updated sensor array #6 as well. The #6 

consists of AHS, PKS, CTS, NMS, CPS, ANS, and SCS sensors that each sensor measures 

 

62 

 

multiple tastes at the same time to increase measurement accuracy. To our knowledge, no 

research study to date has used the sensor array #6 yet. 

The protocols using earlier generation arrays (sensory arrays #1, 2, 5) were vague in most 

of the previous studies. Some researchers evaluated samples in triplicate and took the average 

(Bleibaum et al., 2002; Dong et al., 2017). Waldrop and Ross (2014) ran the samples in triplicate 

and replicated to get 6 data points, and then removed 3 outliers, which were the first and/or 

second samplings due to sensor error. Lipkowitz et al. (2018) and Schlossareck and Ross (2019) 

evaluated the samples in 6 loops but used only 3 data points in the result, assuming the other 3 

data points were removed. These studies indicate that the sensors may not be stable all the time. 

For the new sensory array #6, a new evaluation protocol is needed not only for the E-tongue 

analysis sequence but also for a data selection basis to improve repeatability and reproducibility.    

 

Stevia, which is a natural high-intensity non-nutritive sweetener, has been the focus of 

the food industry to respond to the consumers’ demand for natural sugar substitutes with 

low/zero calories. Steviol glycosides are the sweetening compounds in stevia leaves and 

stevioside and rebaudioside (Reb) A are the major steviol glycosides (Kinghorn, 2002). Plant 

breeders have worked on developing new varieties of stevia plants with higher Reb A and less 

stevioside content because stevioside has bitter aftertaste whereas Reb A has better flavor profile 

(Yadav et al., 2011). However, studies showed that Reb A exhibited bitter off-taste as well 

(Gwak et al., 2012; Kim et al., 2015; Li et al., 2015; Medeiros et al., 2019). In recent years, Reb 

D and M, which are two minor steviol glycosides, have been reported to elicit significantly less 

bitterness with better sweetness than Reb A (Hellfritsch et al., 2012; Prakash et al., 2014). GLG, 

a company based in Canada, has worked on breeding stevia plants with more Reb D and M 

contents (Watson, 2015). Few breeding programs evaluate the flavor profile of stevia varieties. It 

 

63 

 

could be time-consuming to develop a panel as a descriptive panel usually requires 40 to 120 
hours of training (Meilgaard et al., 2015). A large number of stevia samples will need to be 
evaluated. Besides, human panels get sensory fatigue very easily and need to take breaks after 
evaluating 4 or 5 samples (Gwak et al., 2012; Kim et al., 2015; Waldrop & Ross, 2014). 
However, it is important to know the sensory profile of different stevia varieties to select the 
desirable ones. Many studies determined the steviol glycosides content of stevia leaves by high-
performance liquid chromatography (HPLC) (Ahmed & Dobberstein, 1982; Bondarev et al., 
2003; Gardana et al., 2010; Hashimoto et al., 1978; Kolb et al., 2001; Makapugay et al., 1984), 
but the sample preparation and analysis were rather complex (Dong et al., 2017; Kirsanov et al., 
2012). If E-tongue could quickly differentiate glycoside profiles of different stevia varieties, it 
could save a lot of time for breeding programs. 

Thus, the objective of this study was to develop a protocol for E-tongue to discriminate 
Stevia rebaudiana leaves for the potential of using the E-tongue in stevia breeding programs as a 
method of testing flavor quality.   

2. Samples and Sample Preparation 

All sample solutions were made on the day of analysis and kept at room temperature. 

About 25 ml of each sample was used to fill up the sample beaker to the marked line. 

2.1. Part 1: Protocol Validation  

Three steviol glycosides were used to validate our new protocol for the #6 sensory array, 

which were 95% Reb A (ENLITEN® 30000015 High Intensity Sweetener, Ingredion, 

Westchester, IL), 95% Reb D (BESTEVIA® Reb D stevia leaf sweetener, Ingredion, 

 

64 

 

Westchester, IL), and 95% Reb M (BESTEVIA® Reb M stevia leaf sweetener, Ingredion, 
Westchester, IL) at 0.1% (w/v).  

2.2. Part 2: Application for Stevia Leaves in Breeding Programs 

Dried and ground stevia leaf samples were obtained from the Dr. Bipul Biswas’ 

laboratory in the Department of Agricultural Sciences Academic Department at Fort Valley State 
University (FVSU) in Georgia and the Dr. Ryan Warner’s laboratory in the Department of 
Horticulture at Michigan State University (MSU). The seven FVSU stevia line samples were 
selected based on ‘cold tolerance’, which showed different levels of tolerance to cold in the 
production field. The MSU stevia line samples were selected from an open-pollinated 
population. To confirm the E-tongue’s ability to discriminate the stevia leaf samples, seven MSU 
stevia lines (11-223, 11-547, 12-05-140, 12-05-005, 12-11-041, Kenya, Paraguay) were used for 
this study, which showed the difference in steviol glycoside profiles analyzed following the 
method described by Shafii et al. (2012). 

The stevia leaf samples were extracted by the microwave extraction method described by 
Teo et al. (2009); the microwave power was set at 700 W and the temperature was kept at 100°C 
for 20 min. The FVSU samples were prepared by placing 0.078 g of stevia leaves in 20 mL of 
deionized water for extraction and then diluted to 50 ml for analysis. For the MSU samples, 0.2 g 
of each sample in 20 mL deionized water was used for extraction and then diluted to 50 mL for 
analysis. 

 

 

 

 

65 

 

3. E-Tongue Analysis  

3.1. Analysis Sequence 
 
The sensors consisting of 7 sensors (AHS, PKS, CTS, NMS, CPS, ANS, and SCS) were 
immersed in water to hydrate for at least 30 minutes before running the E-tongue. Conditioning, 

calibration, and diagnostic, using 0.01M hydrochloric acid, sodium chloride, and sodium 

glutamate standard solutions, were conducted prior to analysis. The analysis sequence was 

repeated six times to get six data points. The number of beakers with deionized water in between 

samples for cleaning purposes was depended on the type of samples. One cleaning was sufficient 

for aqueous solutions, and 3 cleanings would be needed for thicker samples, such as melted ice 

cream. If too many cleanings were placed in between samples (i.e., 3 cleanings for aqueous 

solutions), some drift could happen, which would influence the accuracy of the sensors, since the 

sensors were constantly changing. In this study, one cleaning was used in between steviol 

glycoside solutions or stevia leaf extract samples. The acquisition time was 120 seconds and the 

cleaning step was 30 seconds. Figure 5 showed the analysis sequence of protocol validation 

using Reb A, D, and M on the E-tongue software. The run was replicated, and the second run 

was used for data analysis since the sensors did not have good stability in the first run.  

 
 
 
 
 
 
 
 
 
 
 
 
 
 

 

66 

 

Figure 5. E-tongue analysis sequence for protocol validation using Reb A, D, and M. Yellow 
color indicates cleaning steps; Purple color indicates samples. 

 

It is important to note that the sample must be at room temperature for E-tongue to 

analyze accurately. When possible, it is better to run all the samples in the same sequence 

because the sensors will not be in the same state. The stirrer should be periodically checked if it 

stirs properly. If the stirrer rotates slower than normal, adjusting the cable connected to the stirrer 

should fix the problem. 

3.2. E-Tongue Data Analysis 

For general data analysis, the first two and last data points were removed due to sensor 
error, and the 3rd, 4th, and 5th data points were used. Relative standard deviation (%RSD) and 
discrimination power were used to determine the quality of data. The preciseness of the data was 
determined by %RSD, which equaled to standard deviation divided by mean and multiply by 
100. If the %RSD was less than 5%, the data were considered good. If it was greater than 5%, 
the samples would either need to be rerun or the sensor with greater than 5% RSD would be 
deleted, depending on the discrimination power. The discrimination power, which ranged from 0 

to 1, indicated the discrimination ability of the sensors on samples. A number closer to 1 meant 

 

67 

 

that the sensor could separate the samples, while a number lower than 0.5 meant that the sensor 

could hardly differentiate the samples. If one of the sensors had poor discrimination power when 
analyzing the samples, the sensor would be removed from the analysis.  

3.3. Statistical Analysis  

 

The data was analyzed using Astree AlphaSoft software (Version 14, AlphaMOS). 

Principal component analysis (PCA) was generated to visually show the data and discriminate 

the samples. The discrimination was assessed by the discrimination index and the spread of the 

samples on the graph (AlphaMOS, 2020). The discrimination index was calculated based on the 

surface area of groups. When the groupings of the samples were distinct, the discrimination 

index was calculated using the first equation. When the samples were overlapped, the 

discrimination index was calculated using the second equation. 

1.  Di = 100 * [1-[(Surface(A)+Surface(B)+Surface(C)) / (Total Surface)]] 

2.  Di = - (∑ Intersection Surface / Total surface) * 100  

4. Validation of Analytical Protocol 

 

Reb A, D, and M were analyzed 7 times (2 runs on Day 1, 3 runs on Day 2, and 2 runs on 

Day 3) by using the analysis sequence mentioned in Section 3.1. To compare the 1st and 2nd run 

each day, the 3rd run on Day 2 will not be discussed further. 

4.1. Sensor Stability 

The sensors values of the second run on Day 1 were shown in Figure 6. There were 

differences in sensor values among the 6 data points, especially sensor SCS. Previous studies 

using sensory array #5 suggested that the first and/or second data points should be removed for 

data analysis due to sensor instability (Waldrop & Ross, 2014). We also verified that the first two 

 

68 

 

and the last data points were not stable with sensory array #6, so they were removed and the rest 

of the three were used for further analysis.  

Figure 6. Sensor values of electronic tongue in the second run on Day 1. Blue indicates sample 
Reb A. Red indicates sample Reb D. Green indicates sample Reb M. 

 

Table 8 shows the %RSD values of the first and second run with selected data points on 

Day 1. The %RSD of sensor PKS was higher than 5% in the first run, which meant that the first 

run could not be used for further analysis. However, the second run was improved, in which 6 

out of the 7 sensors had %RSD lower than 1%. A huge improvement in sensor PKS and CPS 

was observed. The comparison of the first and second run in Day 2 and 3 are shown in Table 9 

and 10, respectively. Both days showed the same trend that the second run had lower %RSD than 

the first run. Even though %RSD below 5% is acceptable, our recommendation is to have %RSD 

less than 1% to get better repeatable data. The discrimination power of the sensors is shown in 

Table 11. Comparing the second run to the first run of each day, the discrimination power was 

highly improved in the second run. 

 
 
 
 
 
 

 

69 

 

Sensors 
AHS 
PKS 
CTS 
NMS 
CPS 
ANS 
SCS 

Sensors 
AHS 
PKS 
CTS 
NMS 
CPS 
ANS 
SCS 

Sensors 
AHS 
PKS 
CTS 
NMS 
CPS 
ANS 
SCS 

Reb A 
1.28 
5.10 
0.55 
0.07 
4.15 
0.42 
3.33 

Reb A 
0.54 
1.40 
1.23 
0.96 
1.01 
1.64 
0.55 

Reb A 
0.88 
0.55 
1.44 
0.69 
0.60 
0.05 
4.00 

First run 
Reb D 
1.13 
5.18 
0.27 
0.02 
3.95 
0.35 
3.06 

First run 
Reb D 
0.57 
1.57 
1.46 
1.27 
1.14 
1.62 
0.67 

First run 
Reb D 
1.03 
0.53 
1.14 
0.68 
0.58 
0.13 
3.66 

Reb M 
0.94 
5.78 
0.20 
0.07 
3.68 
0.23 
2.94 

Reb M 
0.70 
1.52 
1.45 
1.35 
1.28 
1.56 
0.80 

Reb M 
0.65 
0.41 
0.86 
0.40 
0.47 
0.01 
3.09 

Reb A 
0.65 
0.12 
0.64 
0.20 
0.39 
0.10 
2.11 

Reb A 
0.04 
0.18 
0.10 
0.24 
0.15 
0.06 
0.09 

Reb A 
0.44 
0.40 
0.53 
0.98 
0.37 
0.04 
0.05 

Second run 
Reb D 
0.87 
0.22 
0.78 
0.29 
0.58 
0.23 
1.10 

Second run 
Reb D 
0.11 
0.25 
0.16 
0.20 
0.24 
0.09 
0.03 

Second run 
Reb D 
0.36 
0.25 
0.40 
0.67 
0.37 
0.07 
0.04 

Reb M 
0.64 
0.04 
0.51 
0.03 
0.42 
0.03 
1.64 

Reb M 
0.11 
0.20 
0.08 
0.11 
0.19 
0.09 
0.08 

Reb M 
0.35 
0.26 
0.43 
0.40 
0.39 
0.07 
0.06 

Table 8. Relative standard deviation of electronic tongue sensors in the first and second run on 
Day 1 (using 3rd, 4th, and 5th data points), with Reb A, D, and M samples.  

Bold indicates sensors that had big difference (>3.00 % RSD) in %RSD between the two runs.  

Table 9. Relative standard deviation of electronic tongue sensors in the first and second run on 
Day 2 (using 3rd, 4th, and 5th data points), with Reb A, D, and M samples.  

 
Table 10. Relative standard deviation of electronic tongue sensors in the first and second run on 
Day 3 (using 3rd, 4th, and 5th data points), with Reb A, D, and M samples.  

Bold indicates sensors that had big difference (>3.00 % RSD) in %RSD between the two runs.  

 
 
 
 

 

70 

 

Table 11. Discrimination power of the 7 electronic tongue sensors on Day 1, 2, and 3. 

Sensors 
CPS 
SCS 
PKS 
AHS 
CTS 
ANS 
NMS 

First run 
0.057 
0.064 
0.150 
0.200 
0.786 
0.864 
0.996 
*Compare within the day. 

Day 1* 

Second run 

0.824 
0.064 
0.949 
0.525 
0.752 
0.982 
0.984 

Day 2 

First run  Second run  First run 
0.541 
0.394 
0.036 
0.471 
0.231 
0.272 
0.758 
0.128 
0.060 
0.405 
0.989 
0.540 
0.157 
0.669 

0.851 
0.903 
0.649 
0.978 
0.899 
0.977 
0.983 

Day 3 

Second run 

0.353 
0.984 
0.288 
0.498 
0.373 
0.977 
0.397 

4.2. Discrimination of Reb A, D, and M 

Figure 7 (a), (b), and (c) show the PCA of Reb A, D, and M from the second run on Day 

1, 2, and 3, respectively. For Figure 7 (a), SCS with a low discrimination power of 0.064 was 

removed. The first two dimensions explained 96.31% of the variation. The discrimination index 

was 99, which indicated a good separation among the three steviol glycosides. On Day 2 and 3, 

98.88% and 96.10% of the variation were explained by the first two dimensions, respectively. 

The discrimination index was 94 for both days. The second run on all 3 days showed similar 

positions of Reb A, D, and M in relation to PC1 and PC2, and the clear separation among Reb A, 

D, and M was found, which suggests that E-tongue is capable of discriminating Reb A, D, and 

M.  

 
 
 
 
 
 
 
 
 
 
 

 

71 

 

Figure 7. (a) Principal component analysis of steviol glycoside samples in the second run on 
Day 1 as evaluated by electronic tongue and accounting for 96.31% of the variability. The 
discrimination index is 99. (b) Principal component analysis of steviol glycoside samples in the 
second run on Day 2 as evaluated by electronic tongue and accounting for 98.88% of the 
variability. The discrimination index is 94. (c) Principal component analysis of steviol glycoside 
samples in the second run on Day 3 as evaluated by electronic tongue and accounting for 96.10% 
of the variability. The discrimination index is 94. 
 

(b)  

(c)  

 

 

 

 

 

72 

 

5. Discrimination of Stevia Leaf Samples 

Among the seven sensors (AHS, PKS, CTS, NMS, CPS, ANS, and SCS), the CTS sensor 
had a poor discrimination power when evaluating stevia leaf samples, which suggested that CTS 
was not capable of distinguishing the stevia leaf samples. Thus, CTS was removed from the 
analysis.  

5.1. Fort Valley State University Samples 

Figure 8 shows the PCA of stevia samples from the FVSU. The first two dimensions 

explained 93.1% of the variation. The discrimination power of the six sensors (AHS, PKS, NMS, 
CPS, ANS, SCS) ranged from 0.946 to 0.998 and RSD were all below 0.5%; CTS with a low 
discrimination power of 0.129 was removed. The discrimination index was 95, which indicated 
that a clear separation among the samples was observed. This application shows the E-tongue’s 
ability to discriminate stevia leaf samples. 

Figure 8. Principal component analysis of stevia leaf samples from Fort Valley State University 
as evaluated by electronic tongue and accounting for 93.10% of the variability. The 
discrimination index is 95. 

 

5.2. Michigan State University Samples 

Figure 9 showed the PCA of stevia line samples from the MSU. The first two dimensions 
explained 80.9% of the variation. The discrimination power of the six sensors (AHS, PKS, NMS, 

 

73 

 

CPS, ANS, SCS) ranged from 0.672 to 0.948 and RSD were all below 1%; CTS was removed 
due to a low discrimination power (0.212). The discrimination index was 90, meaning that the 
samples were well-separated from each other. Sample Paraguay and 11-223 were grouped 
together. Kenya and 12-11-041 were together. 12-05-140 and 11-547 were similar and 12-05-005 
was closest to them, comparing to other groups.  

Figure 9. Principal component analysis of stevia leaf samples from Michigan State University as 
evaluated by electronic tongue and accounting for 80.86% of the variability. The discrimination 
index is 90. 
 

 

 
Table 12 shows the steviol glycoside profile of the stevia leaf samples. Agglomerative 

hierarchical clustering (AHC) was conducted using XLSTAT (AddinSoft, New York, NY) and 

was used to group stevia lines into 4 groups based on their similarities on the steviol glycoside 

profiles (Table 13). Both similarities and discrepancies between the clustering methods based on 
E-tongue and steviol glycosides were observed (Table 13). For example, Paraguay and 11-223 
were similar based on PCA acquired from E-tongue but were in different classes according to 
AHC. This might be due to the different levels of both total glycosides and individual glycoside. 
12-05-140 and 11-547 were not in the same class according to the steviol glycoside profile 

 

74 

 

clustering but were close to each other on PCA. 12-05-005 was relatively close to 11-547 on the 
PCA and was also grouped with it in AHC based on the steviol glycoside profile.  

0.00 
0.00 
4.25 
0.00 
3.85 
4.44 
0.00 

Reb C  Reb D  Reb M  Total  
1.16 
29.12 
1.42 
22.78 
85.34 
5.42 
12.30 
167.26 
4.03 
56.11 
6.47 
81.56 
84.46 
4.45 

Table 12. Steviol glycosides profile1 of stevia leaf samples from Michigan State University. 
Stevioside  Reb A  Reb B 
Samples 
0.37 
17.542 
11-223 
0.28 
2.99 
11-547 
5.42 
12-05-005  7.26 
5.30 
Kenya 
20.92 
12.39 
12-05-140  10.26 
14.96 
12-11-041  15.95 
Paraguay 
0.69 
29.76 
1 Analyzed by QTRAP 3200 mass spectrometer (Shafii et al., 2012)  
2 Amounts are in mg/g. 
 
Table 13. Classification of stevia lines based on E-tongue and steviol glycoside profiles. 
Samples 
11-223 
Paraguay 
Kenya 
12-11-041 
11-547 
12-05-140 
12-05-005 
1 Classification on the PCA map (Figure 9). 
2 Classification from agglomerative hierarchical clustering 

Steviol glycoside profile2 
Class 1 
Class 4 
Class 2 
Class 3 
Class 2 
Class 3 
Class 2 

9.72 
17.24 
62.99 
109.50 
25.57 
40.93 
37.16 

0.33 
0.84 
0.00 
19.24 
0.00 
3.23 
12.40 

E-Tongue1 
Class 1 
Class 1 
Class 2 
Class 2 
Class 3 
Class 3 
Class 4 

5.3. Future Potential of E-tongue for Flavor Selection in Stevia Breeding Program  

It is important to understand not only chemical analysis of steviol glycosides but also 

sensory data to provide a complete range of flavor information. The chemical analysis by mass 
spectrometer provides quantitative data of each steviol glycoside that is somewhat predictable for 
the flavor of stevia. However, because mass spectrometer detects each taste substance separately, 
it cannot reveal taste-substance interaction (i.e., synergistic and suppression effects), and thus, 
the overall taste of stevia leaves cannot be fully explained (Kobayashi et al., 2010). Sensory 
evaluation can provide integrated, direct measurements of flavor, but the sensory evaluation is 
time-consuming and expensive. Therefore, E-tongue as a quick and simple analytical tool has a 

 

75 

 

great potential for quick flavor discrimination of stevia lines in the breeding programs since it 
mimics human taste perception without sensory fatigue. In order to develop a predictive model 
of stevia flavors in breeding programs using E-tongue, the relationships between chemical 
analysis, sensory evaluation, and E-tongue should be well understood to correlate the E-tongue 
data with chemical analysis and human sensory perception. 

6. Conclusions 

The protocol developed for #6 sensor array worked well to produce more repeatable data 

for stevia samples. E-tongue successfully discriminated the stevia leaf samples, which was the 
first step to determine if E-tongue could be used for flavor selection in stevia breeding programs. 
Further analyses, such as human panels, are needed to evaluate the sensory characteristics of the 
stevia leaf samples, in order to understand which sample has a better flavor profile such as clean 
sweetness or less bitterness. E-tongue data could then be correlated to human panel data and 
determine if E-tongue can be used as a fast and accurate way to determine the flavor profile of 
stevia leaf samples for the stevia breeding programs. 

 

 

 

76 

 

 
 
 
 
 
 
 
 
 
 

 
 

 

 

REFERENCES 

 

77 

 

REFERENCES 

 
Ahmed, M. S., & Dobberstein, R. H. (1982). Stevia rebaudiana: III. High-performance liquid 

chromatographic separation and quantitation of rebaudiosides B, D, and E, dulcoside A, and 
steviolbioside. Journal of Chromatography A, 245(3), 373–376. 
https://doi.org/10.1016/S0021-9673(00)88027-9 

 
Alpha MOS. (2020). Astree electronic tongue hardware user’s manual. 
 
Barnett, S. M., Diako, C., & Ross, C. F. (2019). Identification of a Salt Blend: Application of the 
Electronic Tongue, Consumer Evaluation, and Mixture Design Methodology: Identification 
of a salt blend…. Journal of Food Science, 84(2), 327–338. https://doi.org/10.1111/1750-
3841.14440 

 
Bleibaum, R. N., Stone, H., Tan, T., Labreche, S., Saint-Martin, E., & Isz, S. (2002). Comparison 

of sensory and consumer results with electronic nose and tongue sensors for apple juices. 
Food Quality and Preference, 13(6), 409–422. https://doi.org/10.1016/S0950-
3293(02)00017-4 

 
Bondarev, N. I., Sukhanova, M. A., Reshetnyak, O. V., & Nosov, A. M. (2003). Steviol 

Glycoside Content in Different Organs of Stevia rebaudiana and Its Dynamics during 
Ontogeny. Biologia Plantarum, 47(2), 261–264. 
https://doi.org/10.1023/B:BIOP.0000022261.35259.4f 

 
Dong, W., Zhao, J., Hu, R., Dong, Y., & Tan, L. (2017). Differentiation of Chinese robusta 

coffees according to species, using a combined electronic nose and tongue, with the aid of 
chemometrics. Food Chemistry, 229, 743–751. 
https://doi.org/10.1016/j.foodchem.2017.02.149 

 
Gardana, C., Scaglianti, M., & Simonetti, P. (2010). Evaluation of steviol and its glycosides in 
Stevia rebaudiana leaves and commercial sweetener by ultra-high-performance liquid 
chromatography-mass spectrometry. Journal of Chromatography A, 1217(9), 1463–1470. 
https://doi.org/10.1016/j.chroma.2009.12.036 

 
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. https://doi.org/10.1007/s10068-012-0115-0 

 
Hashimoto, Y., Moriyasu, M., Nakamura, S., Ishiguro, S., & Komuro, M. (1978). High-

performance liquid chromatographic determination Stevia components on a hydrophilic 
packed column. Journal of Chromatography A, 161, 403–405. 
https://doi.org/10.1016/S0021-9673(01)85262-6 

 

 

78 

 

Hellfritsch, C., Brockhoff, A., Stähler, F., Meyerhof, W., & Hofmann, T. (2012). Human 

psychometric and taste receptor responses to steviol glycosides. Journal of Agricultural and 
Food Chemistry, 60(27), 6782–6793. https://doi.org/10.1021/jf301297n 

electronic tongue. 8. 

 
Hruškar, M., Major, N., & Krpan, M. (2009). Evaluation of milk and dairy products by 
 
Jung, H. Y., Kwak, H. S., Kim, M. J., Kim, Y., Kim, K.-O., & Kim, S. S. (2017). Comparison of 
a descriptive analysis and instrumental measurements (electronic nose and electronic 
tongue) for the sensory profiling of Korean fermented soybean paste (doenjang). Journal of 
Sensory Studies, 32(5), e12282. https://doi.org/10.1111/joss.12282 

 
Kim, M.-J., Yoo, S.-H., Jung, S., Park, M.-K., & Hong, J.-H. (2015). Relative sweetness, 

sweetness quality, and temporal profile of xylooligosaccharides and luo han guo (Siraitia 
grosvenorii) extract. Food Science and Biotechnology, 24(3), 965–973. 
https://doi.org/10.1007/s10068-015-0124-x 

http://www.crcnetbase.com/isbn/9780415268301 

 
Kinghorn, A. D. (2002). Stevia: The genus Stevia. Taylor & Francis. 
 
Kirsanov, D., Mednova, O., Vietoris, V., Kilmartin, P. A., & Legin, A. (2012). Towards reliable 
estimation of an “electronic tongue” predictive ability from PLS regression models in wine 
analysis. Talanta, 90, 109–116. https://doi.org/10.1016/j.talanta.2012.01.010 

 
Kobayashi, Y., Habara, M., Ikezazki, H., Chen, R., Naito, Y., & Toko, K. (2010). Advanced 

taste sensors based on artificial lipids with global selectivity to basic taste qualities and high 
correlation to sensory scores. Sensors, 10(4), 3411–3443. 
https://doi.org/10.3390/s100403411 

 
Kolb, N., Herrera, J. L., Ferreyra, D. J., & Uliana, R. F. (2001). Analysis of Sweet Diterpene 

Glycosides from Stevia rebaudiana: Improved HPLC Method. Journal of Agricultural and 
Food Chemistry, 49(10), 4538–4541. https://doi.org/10.1021/jf010475p 

 
Lee, H. J., Yoon, J. W., Kim, M., Oh, H., Yoon, Y., & Jo, C. (2019). Changes in microbial 

composition on the crust by different air flow velocities and their effect on sensory 
properties of dry-aged beef. Meat Science, 153, 152–158. 
https://doi.org/10.1016/j.meatsci.2019.03.019 

 
Li, X. E., Lopetcharat, K., & Drake, M. A. (2015). Parents’ and children’s acceptance of skim 

chocolate milks sweetened by monk fruit and stevia leaf extracts. Journal of Food Science, 
80(5), S1083–S1092. https://doi.org/10.1111/1750-3841.12835 

 
Lipkowitz, J. B., Ross, C. F., Diako, C., & Smith, D. M. (2018). Discriminating aging and 

protein-to-fat ratio in Cheddar cheese using sensory analysis and a potentiometric electronic 
tongue. Journal of Dairy Science, 101(3), 1990–2004. https://doi.org/10.3168/jds.2017-
13820 

 

79 

 

Makapugay, H. C., Nanayakkara, N. P. D., & Kinghorn, A. D. (1984). Improved high-

performance liquid chromatographic separation of the Stevia rebaudiana sweet diterpene 
glycosides using linear gradient elution. Journal of Chromatography A, 283, 390–395. 
https://doi.org/10.1016/S0021-9673(00)96278-2 

 
Medeiros, A. C., Filho, E. R. T., & Bolini, H. M. A. (2019). Impact of natural and artificial 
sweeteners compounds in the sensory profile and preference drivers applied to traditional, 
lactose-free, and vegan frozen desserts of chocolate flavor. Journal of Food Science, 
84(10), 2973–2982. https://doi.org/10.1111/1750-3841.14806 

 
Meilgaard, M. C., Civille, G. V., & Carr, B. T. (2015). Sensory Evaluation Techniques. CRC 
 
Prakash, I., Markosyan, A., & Bunders, C. (2014). Development of next generation stevia 
sweetener: Rebaudioside M. Foods, 3(1), 162–175. https://doi.org/10.3390/foods3010162 
 
Schlossareck, C., & Ross, C. F. (2019). Electronic Tongue and Consumer Sensory Evaluation of 

Press. 

Spicy Paneer Cheese. Journal of Food Science, 84(6), 1563–1569. 
https://doi.org/10.1111/1750-3841.14604 

 
Shafii, B., Vismeh, R., Beaudry, R., Warner, R., & Jones, A. D. (2012). Large-scale profiling of 

diterpenoid glycosides from Stevia rebaudiana using ultrahigh performance liquid 
chromatography/tandem mass spectrometry. Analytical and Bioanalytical Chemistry, 
403(9), 2683–2690. https://doi.org/10.1007/s00216-012-6071-y 

 
Teo, C. C., Tan, S. N., Yong, J. W. H., Hew, C. S., & Ong, E. S. (2009). Validation of green-

solvent extraction combined with chromatographic chemical fingerprint to evaluate quality 
of Stevia rebaudiana Bertoni. Journal of Separation Science, 32(4), 613–622. 
https://doi.org/10.1002/jssc.200800552 

 
Waldrop, M. E., & Ross, C. F. (2014). Sweetener blend optimization by using mixture design 
methodology and the electronic tongue. Journal of Food Science, 79(9), S1782–S1794. 
https://doi.org/10.1111/1750-3841.12575 

 
Watson, E. (2015). GLG: We’ll be able to breed stevia plants with more Reb D, Reb M. 

FoodNavigator. https://www.foodnavigator.com/Article/2015/09/16/GLG-We-ll-be-able-to-
breed-stevia-plants-with-more-Reb-D-Reb-M 

 
Yadav, A. K., Singh, S., Dhyani, D., & Ahuja, P. S. (2011). A review on the improvement of 
stevia [ Stevia rebaudiana (Bertoni)]. Canadian Journal of Plant Science, 91(1), 1–27. 
https://doi.org/10.4141/cjps10086 

 
Zou, G., Xiao, Y., Wang, M., & Zhang, H. (2018). Detection of bitterness and astringency of 

green tea with different taste by electronic nose and tongue. PLoS ONE, 13(12). 
https://doi.org/10.1371/journal.pone.0206517 

 

80 

 

 

CONCLUSIONS 

The findings of the first study generate important information about Reb A, D, and M for 

the food industry. Reb D and M can be considered as good sucrose substitutes due to their 
similar sensory profiles to the sucrose, comparing to Reb A. However, the use of Reb D and M 
for products formulated to deliver a reduction in sugar should be careful with the artificial 
attribute and the lingering sweetness of Reb D and M. The sensory characteristics of Reb A, D, 
and M identified in this study can be used as a reference for the food industries, especially 
working in high-sugar applications such as frozen desserts. The individual’s sensitivity to the 
perceived intensity of sweet and bitter tastes of the three steviol glycosides was not predicted by 
the PROP taster groups. This suggests that supertasters who experience more intense taste 
sensations may not report aversive sensations from stevia. Future studies using steviol glycosides 
in food matrix are needed to determine the sensory profile and acceptance of them.  

The second study provides a new protocol for E-tongue analysis using the most updated 

sensor array #6 and shows that E-tongue has the potential to be used as an analytical tool for 
flavor discrimination of stevia lines in the breeding programs. Understanding both the chemical 
components and sensory data are important to build a complete range of flavor information. 
Since E-tongue is relatively quick and simple comparing to a human panel, it is important to 
understand the relationship between E-tongue and human sensory perception in order to develop 
a predictive model of stevia flavors in breeding programs using E-tongue. Further analyses, such 
as human panels, are needed to evaluate the sensory characteristics of the stevia leaf samples and 
correlate to E-tongue data. 

 

 

81