INFORMATION TO USERS This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer. The quality of this reproduction is dependent upon the quality of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction. In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion. Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand corner and continuing from left to right in equal sections with small overlaps. Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6" x 9" black and white photographic prints are available for any photographs or illustrations appearing in this copy for an additional charge. Contact UMI directly to order. ProQuest Information and Learning 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 800-521-0600 ® UMI
197
Embed
Categorization and Sensory Profiling of Functional Beverages
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
INFORMATION TO USERS
This manuscript has been reproduced from the microfilm master. UMI films
the text directly from the original or copy submitted. Thus, some thesis and
dissertation copies are in typewriter face, while others may be from any type of
computer printer.
The quality of this reproduction is dependent upon the quality of the
copy submitted. Broken or indistinct print, colored or poor quality illustrations
and photographs, print bleedthrough, substandard margins, and improper
alignment can adversely affect reproduction.
In the unlikely event that the author did not send UMI a complete manuscript
and there are missing pages, these will be noted. Also, if unauthorized
copyright material had to be removed, a note will indicate the deletion.
Oversize materials (e.g., maps, drawings, charts) are reproduced by
sectioning the original, beginning at the upper left-hand corner and continuing
from left to right in equal sections with small overlaps.
Photographs included in the original manuscript have been reproduced
xerographically in this copy. Higher quality 6" x 9" black and white
photographic prints are available for any photographs or illustrations appearing
in this copy for an additional charge. Contact UMI directly to order.
ProQuest Information and Learning 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA
800-521-0600
®
UMI
CATEGORIZATION AND SENSORY PROFILING OF FUNCTIONAL BEVERAGES
BY
LAUREN CHIEMI TAMAMOTO
B.S., University of Hawaii at Manoa, 2003 M.S., University of Queensland, 2004
DISSERTATION
Submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Food Science and Human Nutrition
in the Graduate College University of Illinois at Urbana-Champaign, 2009
Urbana, Illinois
Doctoral Committee:
Professor Keith R, Cadwallader, Chair Associate Professor Soo-Yeun Lee, Co-Director of Research Professor Shelly J. Schmidt, Co-Director of Research Associate Professor Elvira de Mejia
UMI Number: 3395512
All rights reserved
INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
JLIMT^ Dissertation Publishing
UMI 3395512 Copyright 2010 by ProQuest LLC.
All rights reserved. This edition of the work is protected against unauthorized copying under Title 17, United States Code.
Chapter 3 - Categorization of Commercially-Available Functional Beverages by Chemical, Physical, and Sensory Commonalities 33 3.1 Abstract .' ,...-,- 33 3.2 Introduction 34 3.3 Materials and Methods 36 3.4 Results and Discussion 42 3.5 Conclusions 48 3.6 References 49 3.7 Tables and Figures 52
v
Chapter 4 - Validation and Reproducibility Study of a Two-Step Sensory Sorting Method to Categorize Functional Beverages 64 4.1 Abstract 64 4.2 Introduction 65 4.3 Materials and Methods : 67 4.4 Results and Discussion 75 4.5 Conclusions 81 4.6 References 82 4.7 Tables and Figures 85
Chapter 5 - Sensory Profile of a Model Energy Drink with Varying Levels of Functional Ingredients-Caffeine, Ginseng, and Taurine , 104 5.1 Abstract 104 5.2 Introduction 105 5.3 Materials and Methods 108 5.4 Results and Discussion 113 5.5 Conclusions 118 5.6 References 118 5.7 Tables and Figures 121
Chapter 6 - Sensory Properties of Ginseng Solutions Modified by Masking Agents ....129 6.1 Abstract 129 6.2 Introduction 130 6.3 Materials and Methods 132 6.4 Results and Discussion 139 6.5 Conclusions 145 6.6 References 146 6.7 Tables and Figures 149
Table 3.2: Functional beverage category names generated through visual observation of the beverage ingredient commonalities in the ingredient inventory spreadsheet 53
Table 3.3: Viscosities of Newtonian Functional Beverages measured at 20°C 54
Table 3.4: Viscosities of non-Newtonian Functional Beverages at a shear rate of 50 s"1 55
Table 4.2: Adjusted Rand Index values of the comparison of clusters generated through free and fixed sorting tasks by Panels 1 to 4 86
Table 4.3: Compilation of Panel 2 and 3's validation study results of commercially available functional beverages sorted into categories by visual and
visual-oral fixed sorts compared to Panel l 's results 87
Table 4.4: Compilation of Panel 1 and 4's reproducibility study results comparing commercially-available functional beverages sorted into categories by visual and visual-oral free and fixed sorting task results 89
Table 5.1: Amount of functional ingredients listed on Nutritional Facts labels of a sampling of popular commercially-available energy drinks 121
Table 5.2: Amount of functional ingredients (caffeine, ginseng, and taurine) in 100 mL model energy drink solutions 122
Table 5.3: Terms, definitions, references, and ratings for scale anchors of the descriptive attributes for the model energy drink solutions 123
Table 5.4: Analysis of Variance on 13 descriptive attributes rated for model energy drink solutions 124
Table 5.5: Mean intensity scores of sensory attributes of varying levels of functional ingredients 125
viii
Table 5.6: Correlation analysis on significant sensory attributes for 27 combinations of functional ingredients in model energy drink solutions 126
Table 6.1: Solution treatment codes and corresponding levels of y-, p-CDs, and their combinations in both 100 mL water base and 100 mL model energy drink base solutions 149
Table 6.2: Bitterness intensity rankings (l=least bitter to 6=most bitter) of the bitterness minimizing treatments incorporated in a 0.0529 g ginseng/100 mL water solution 150
Table 6.3: Mean bitterness intensity rating scores (0 to 9) of bitterness minimizing treatment levels incorporated in a 0.0529 g ginseng/100 mL water solution 151
Table 6.4: Analysis of Variance on descriptive attributes rated for ginseng solutions containing varying levels of y- and P-CDs 152
Table 6.5: Mean quinine bitter and caffeine bitter aftertaste attribute intensity scores (0 to 15) across all 21 solution treatments combining water base and model energy drink base solutions and with and without nose clips usage data 152
Table 6.6: Mean" quinine bitter and caffeine bitter aftertaste attribute intensity scores (0 to 15) across all 7 y-CD solution treatments in water base or model energy drink base, without nose clips and with nose clips usage data 153
Table 6.7: Mean" quinine bitter and caffeine bitter aftertaste attribute intensity scores (0 to 15) across all 7 P-CD solution treatments in water base or model energy drink base, without nose clips and with nose clips usage data 154
ix
LIST OF FIGURES
Figure 2.1: Chemical Structure of Caffeine 10
Figure 2.2: Chemical Structure of Ginsenoside 11
Figure 2.3: Chemical Structure of Taurine 12
Figure 3.1: Multidimensional Scaling of the results of the ingredient inventory categorization with stress = 0.227 56
Figure 3.2: Agglomerative hierarchical clustering of 50 functional beverages (including decarbonated beverages) by viscosity measurement using .the ARES RFS III 57
Figure 3.3: Agglomerative hierarchical clustering of 33 Newtonian functional beverages by viscosity measurement using the ARES RFS III 58
Figure 3.4: Agglomerative hierarchical clustering of 17 non-Newtonian functional beverages by viscosity measurement at 50 sec"1 shear rate using the ARES RFS III 59
Figure 3.5: Multidimensional Scaling of a visual free sort (Part 1) of 50 functional beverages plotted in two dimensions with stress = 0.265 and functional beverage categories generated through the free visual sorting method 60
Figure 3.6: Multidimensional Scaling of a visual-oral free sort (Part 1) of 50 functional beverages plotted in two dimensions with stress = 0.290 and functional beverage categories generated through the free visual-oral sorting method 61
Figure 3.7: Multidimensional Scaling of a visual fixed sort (Part 2) of 50 functional beverages plotted in two dimensions with stress = 0.289 and corresponding beverage categories 62
Figure 3.8: Multidimensional Scaling of a visual-oral fixed sort (Part 2) of 45 functional beverages plotted in two dimensions with stress = 0.283 and corresponding beverage categories 63
Figure 4.1: Flow chart of studies and panels, categories, types of sorting task, and results of the conducted two-step sensory sorting method 91
Figure 4.2: Multidimensional Scaling Panel l's visual free sort (Part 1) of 50 functional beverages plotted in two dimensions with stress = 0.265 and functional beverage categories generated through the free visual sorting method
Figure 4.3: Multidimensional Scaling of Panel l 's visual-oral free sort (Part 1) of 50 functional beverages plotted in two dimensions with stress = 0.290 and functional beverage categories generated through the free visual-oral sorting method 93
Figure 4.4: Multidimensional Scaling of Panel l 's visual fixed sort (Part 2) of 50 functional beverages plotted in two dimensions with stress = 0.289 and corresponding functional beverage categories 94
Figure 4.5: Multidimensional Scaling of Panel l 's visual-oral fixed sort (Part 2) of 45 functional beverages plotted in two dimensions with stress = 0.283 And corresponding functional beverage categories 95
Figure 4.6: Multidimensional Scaling of Panel 2's visual fixed sort of 50 functional beverages plotted in two dimensions with stress = 0.301 and corresponding fixed functional beverage categories a) Multidimensional Scaling of Panel 4's visual free sort data of 50 96
Figure 4.7: Multidimensional Scaling of Panel 2's visual-oral fixed sort of 46 functional beverages plotted in two dimensions with stress = 0.271 and corresponding fixed functional beverage categories 97
Figure 4.8: Multidimensional Scaling of Panel 3's visual fixed sort of 50 functional beverages plotted in two dimensions with stress = 0.241 and corresponding fixed functional beverage categories 98
Figure 4.9: Multidimensional Scaling of Panel 3's visual-oral fixed sort of 46 functional beverages plotted in two dimensions with stress = 0.284 and corresponding fixed functional beverage categories 99
Figure 4.10: Multidimensional Scaling of Panel 4's visual free sort (Part 1) of 46 functional beverages plotted in two dimensions with stress = 0.260 and functional beverage categories generated through the free visual sorting method 100
Figure 4.11: Multidimensional Scaling of Panel 4's visual-oral free sort (Part 1) of 46 functional beverages plotted in two dimensions with stress = 0.232 and functional beverage categories generated through the free visual-oral sorting method 101
Figure 4.12: Multidimensional Scaling of Panel 4's visual fixed sort (Part 2) of 46 functional beverages plotted in two dimensions with stress = 0.261 and corresponding fixed functional beverage categories 102
xi
Figure 4.13: Multidimensional Scaling of Panel 4's visual-oral fixed sort (Part 2) of 46 functional beverages plotted in two dimensions with stress = 0.262 and corresponding fixed functional beverage categories 103
Figure 5.1: Principal component analysis biplot of covariance matrix of mean sensory attributes of 27 combinations of functional ingredients in model energy drink solutions with varimax rotation 127
Figure 5.2: Agglomerative hierarchical clustering (AHC) of attribute ratings for 27 combinations of functional ingredients in model energy drink solutions on the dissimilarity scale by Euclidean distance and agglomeration by Ward's method , 128
Figure 6.1: Effect of y-CD levels on (a) quinine bitter and (b) caffeine bitter aftertaste intensity ratings of ginseng solution treatments with and without nose clips and in water base or model energy drink base solutions 155
Figure 6.2: Effect of p-CD levels on (a) quinine bitter and (b) caffeine bitter aftertaste intensity ratings of ginseng solution treatments with and without nose clips and in water base or model energy drink base solutions 156
Figure 6.3: Agglomerative hierarchical clustering (AHC) of quinine bitter and caffeine bitter aftertaste attribute mean intensity ratings for 21 ginseng solution treatments containing varying levels of y-CD and p-CD on the dissimilarity scale by Euclidean distance and agglomeration 157
xii
CHAPTER 1- INTRODUCTION
1.1 Motivation
Functional beverages are a booming market, with hundreds of new beverages
introduced each year (Packaged Facts 2009). The popularity of functional beverages
stems from the fact that they are a healthy alternative to soft drinks because they provide
additional health benefits. The origins of functional beverages began with the
introduction of beverages geared to replace fluids and electrolytes after exercise (Leveen
2007). Energy drinks were then developed for consumers who desired an extra boost of
energy in a beverage. The functional beverage market is rapidly expanding, and it is
necessary to understand the characteristics associated with specific types of functional
beverages, to meet consumers' expectations. Therefore, the final central hypothesis of
this thesis is that there are distinct functional beverage categories, and that the functional
ingredients incorporated into beverage formulations affect the sensory properties of
beverages.
One of the initial working hypotheses of this research was that there are particular
mouthfeel and sensory expectations that correspond to different types of functional
beverages. The questions we wanted to answer included: is there a particular mouthfeel
associated with an isotonic drink that is unique in comparison to an energy drink? Or
could a highly viscous solution be considered a tea? To answer these questions, a
literature review was conducted, and it was found that there was no set of standard
definitions of functional beverage categories. The rapid introduction of new functional
beverages into the market raised the need to define the specific categories of these
beverages. The expansion of the functional beverage market resulted in the creation of
1
numerous hybrid-type beverages, which combine multiple concepts. Therefore, the
initial research focus was then expanded to answer the question, "What are the different
categories of functional beverages and what makes each beverage category distinct?"
The first hypothesis of this research was that functional beverage can be classified into
well-defined categories through categorization methods.
In the categorization research, energy drinks were found to be one of the most
distinctive and popular categories of the beverage market. The ingredients of thirteen
commercial carbonated energy drink products were inspected, and it was found that
caffeine, ginseng, and taurine were some of the most common functional ingredients
contained in energy drinks. In 2007, ginseng, caffeine, and taurine were all considered
one of the top 15 functional ingredients consumers seek in functional beverages (Lai
2007). Yet, limited research had been conducted to determine the sensory effects
associated with the addition of these functional ingredients in food products. Consumers
want products that provide health benefits and have a pleasant taste (Drewnowski and
Gomez-Carneros 2000), and the inclusion of ingredients to a beverage solely based on
functional properties may result in a product rich in bioactives, but with an unacceptable
taste.
Therefore, it is necessary to investigate the resulting effects from the
incorporation of functional ingredients into a beverage matrix. This is important because
understanding the interaction between the base solution and ingredients will help in
creating pleasant tasting products (Granato 2002). The next hypothesis in this study was
that the functional ingredients in a model energy drink solution will have synergistic
effects on the sensory properties of the drink. Understanding the effects of the addition
2
of specific functional ingredients into a food matrix will aid in the development of new
food products by providing potential solutions to flavor issues such as reducing negative
flavor attributes with the addition of a masking agent. It could also help in selecting
concentrations of functional ingredients that are acceptable to consumers.
The results from the sensory study on the synergistic effects of caffeine, ginseng,
and taurine into a model energy drink solution suggested that ginseng was the most
dominant functional ingredient in the solution and that it imparts a bitter taste. The •
bitterness in functional foods often reduces the liking of a product (Tuorila and Cardello
2002), and identifying effective methods of minimizing the bitterness in functional
beverages will allow formulators to produce products that have health benefits as well as
acceptable sensory qualities. Thus, the last part of this research focuses on minimizing
bitterness of ginseng in a model energy drink solution. The hypothesis was that the use
of cyclodextrins will aid in minimizing bitterness of ginseng in model energy drink
solutions while still maintaining ginseng's health benefits. This research can be utilized
in the development of acceptable energy drinks, and also to predict changes in sensory
characteristics when reformulating functional ingredients in energy drinks.
1.2 Objectives
The two main hypotheses of this research were that 1) functional beverages can
be classified into defined categories through relatively quick and easy methods, and 2)
the inclusion of functional ingredients has an effect on the sensory properties of energy
drinks. The first objective of this research was to categorize commercially-available
functional beverages using three different methods: 1) ingredient inventory, 2) flow
3
behavior comparison, and 3) two-step sensory sorting. The second objective was to
develop and validate the two-step sensory sorting method to categorize a large number of
samples. The third objective was to determine and describe the effects of functional
ingredients (caffeine, taurine, and ginseng) on the sensory characteristics of model energy
drink solutions. The fourth and final objective was to identify effective treatments and
levels of bitterness minimizers to reduce the bitterness of ginseng in water base and
model energy drink base solutions. The findings from this research will be beneficial to
the development of acceptable functional beverage formulations.
1.3 References
Drewnowski A, Gomez-Carneros C. 2000. Bitter taste, phytonutrients, and the consumer: a review. Am J Clin Nutr 72(6): 1424-35.
Granato H. 2002. Manipulating Flavor Perception in Functional Products. Natural Products Insider [serial online]. Available from Posted 8 April 2002 2002.
Lai GG. 2007. Getting Specific with Functional Beverages. Food Technology [serial online]. 61 (12):Available from Posted 2007.
Leveen T. 2007. Functional Beverages: Market Evolution. Natural Products Marketplace [serial online]. Available from Posted 2007.
Packaged Facts. 2009. Functional Foods and Beverages in the U.S. 1-210.
Tuorila I-I, Cardello AV. 2002. Consumer responses to an off-flavor in juice in the presence of specific health claims. Food Qual Pref 13(7-8):561-9.
4
CHAPTER 2 - LITERATURE REVIEW
2.1 Functional Beverages
In the US, concerns about disease and health have increased the popularity of
functional food products (Schmidl and Labuza 2000, Urala and LMhteenmaki 2003).
Functional foods are consumed because they are thought to provide more benefits than
ordinary foods. It is also more convenient to consume a beverage providing health
benefits rather than swallow vitamins or pills for the same health benefits (Leveen 2007).
These "total health" and weight management concerns have prompted growth in the
number of functional food and beverages available on the market (Lai 2007) and
consumers are now seeking products which provide an added health benefit to ordinary
food products. In 2007, there were over $10.1 billion in functional beverage sales in the
US, and by 2010 functional beverage sales are projected to increase to over $12 billion
(Mintel 2008).
There is currently no legal definition of a "functional beverage" or universally
accepted categories of functional beverages in the United States. The Institute of Food
Technologists (2005), however, has defined functional beverages as beverages that
provide health benefits beyond basic nutrition. Therefore, functional beverage categories
may encompass beverages containing probiotics, stimulants, or additional vitamins and
minerals. The inclusion of ingredients based on functional needs is a greater driving
force in product development than category boundaries (Humphries 2007), and has
resulted in numerous new products introduced each year. The significant increase in
functional beverage popularity has led to an influx of new products and the introduction
of numerous "hybrid-type" beverages to the market. These hybrid-type beverages fall
5
under the "functional beverage" definition because they provide unique health benefits,
and at the same time incorporate multiple beverages concepts. For example, there are
now diet energy drink teas and dairy-based diet beverages on the market.
Some benefits of functional ingredients include: replenishing electrolytes in the
body, providing an extra boost of energy, or aiding with digestion. A few popular
functional ingredients included in beverage formulations are: antioxidants, stimulants,
botanicals, vitamins, and minerals (Lai 2007). Some beverages on the market incorporate
ingredients naturally containing functional benefits, while other beverages include
synthetic ingredients which provide the same benefits. These functional ingredients
affect the sensory experience, and may sometimes result in unpleasant sensory .
characteristics, such as bitterness or chalkiness. Studies, however, have shown that
consumers are more tolerant of unpleasant flavors if the beverages deliver additional
health benefits (LeClair 2000).
The functional beverage boom has led to an increase in new products as
companies are attempting to capitalize on this market (Wright 2008). Consumers desire
functional beverages that encompass multiple concepts and provide specific health
benefits. The functional beverage market is driven by consumers, thus there has been a
shift to create more hybrid-type products. Therefore, there is a need to understand the
influence of functional ingredients on the sensory properties to create belter tasting
functional beverages.
2.1.1 Functional Beverage Categories
The constant introduction of functional beverages to the market makes it difficult
to pinpoint the major functional beverage categories and corresponding definitions.
6
Initially, energy drinks and sports drinks were considered functional beverages, but now
the inclusion of botanicals and probiotics are considered part of the functional beverage
market (Alldrick 2006). Functional beverages with claims of managing appetite or aiding
in younger-looking skin are currently part of the influx of new functional beverages
introduced to the market (Leveen 2007). Publications often refer to loosely defined
categories, but the categories that are considered part of the functional beverages segment
are never consistent. The lack of functional beverage categories leads to an absence .of
requirements and standards of identity. The general beverages market, which has been
established for many years, has guidelines and regulations concerning ingredient amounts
and package labeling. The absence of functional beverage categories also may hinder the
purchase intent of consumers and could lead to improper positioning of a product (Lord
2000).
Mintel Reports (2008) segmented the functional beverage market into six
categories: juices and juice drinks; smoothies and yogurt drinks; teas; soy-based drinks;
energy drinks and enhanced water; and sports drinks. Rehydrating sports drinks and meal
or diet drinks were not considered functional beverage categories in the Mintel Reports
(2008). Mintel reports, however, are not available to the general public, therefore, these
functional beverage categories are not common knowledge. In the Packaged Facts -
Functional Foods and Beverages in the US report (2009), the eight functional beverage .
categories included in the report were: shelf-stable bottled juice drinks, blends and
Cyclodextrins (CDs) are large ring-shaped molecules created through the
enzymatic conversion of starch. They are an odorless, white powder that are soluble in
water. Common types of CDs are a, [3-, and y-CDs, which consist of 6, 7, and 8
glucosidic units, respectively. This ring-shaped structure allows CDs to trap smaller
molecules, thus forming inclusion compounds. This unique structure gives CDs many
useful food-related applications.
Cyclodextrins can be used as an emulsifier, a stabilizer for fragile compounds
such as flavors, colors, amino acids, and vitamins (Dodziuk 2006), a solubilizer for
pharmaceuticals (Loftsson and Brewster 1996), and a stabilizer in foods such as cookies
and chewing gums (Dodziuk 2006). CDs have also been used to create lower cholesterol
products by forming inclusions of cholesterol in CDs and then taking the CD
complexalions out of the product; this method has been used to lower cholesterol egg
products (Smith and others 1995). According to the FDA, all a, P-, and y-cyclodextrins
(CDs) are generally recognized as safe (GRAS) for use as a stabilizer, emulsifier, carrier,
and formulation in foods at levels between 1 to 5%.
Cyclodextrins have been used to minimize the bitterness of compounds. P-CDs at
a 0.4% concentration was able to reduce the bitterness of a 0.05% caffeine solution by
90% (Binello and others 2004). p-CDs have been shown to reduce the bitterness of
ginseng tea (Takeuchi and Naae 1992). The success of P-CDs as a bitterness minimize!-
was more than the use of a- and y-CDs, possibly because high amounts of p-CDs impart a
sweet taste (Binello and others 2004). It was also found that CDs remove the bitter taste
of sweeteners such as stevioside (Astray and others 2009). A study on y-CDs found that
18
there was no difference between consuming yogurt containing 8 g y-CDs/100 g versus no
cyclodextrins present in yogurt (Koutsou and others 1999).
2.4 Sorting and Categorization Mclhods
Sorting methods have been used in many fields to understand the relationship
between products or concepts (Rugg and McGeorge 1997, Viswanathan and Childers
1999). These methods allow researchers to gather information about panelists'
perceptions of a large group of products (Viswanathan and Childers 1999) and provides
structured information about the relationship among products (Mervis and Rosch 1981).
Sorting tasks have also been commonly used as marketing tools as a way to obtain
consumer insight on product perception and for comparison to competitor's products.
Sorting tasks can range from having panelists sort products into predefined categories, to
allowing panelists to freely sort the products. Different types of sorting procedures range
from sorting cards containing descriptor words to having panelists evaluate and sort food
products.
Sorting methods can be used as a quick and simple method to gather descriptors
and relationship information about products and are much less time-consuming compared
to other methods of sensory evaluation. These methods require minimal panelist training
and can often be conducted in one session. Sorting methods are useful when gathering
information on large groups of products (Cartier and others 2006), while other methods
such as quantitative descriptive analysis (QDA) produce more detailed information on a
smaller group of products.
19
Prior to their applications to food, these sorting methods have been applied to
nonfood materials such as car fabrics (Giboreau and others 2007), colored plastic chips
(Faye and others 2004), and oral health care products (Bertino and Lawless 1993).
Sorting methods were first introduced to sensory work in 1995 when Lawless and others
conducted work on different cheeses (Lawless and others 1995). This method lias been
successful in sorting other food items such as types of water (Falahee and MacRae 1997),
snack bars (King and others 1998), red wine (Gawel and others 2000), novel food
products (Woolf and others 2002), and yogurts (Saint-Eve and others 2004).
Free sorting is a sorting method which is simple and only requires objects to be
sorted, criteria for objects to be sorted, a record sheet, and instructions (Coxon 1999).
This method has few restrictions which include that more than one category must be
created and that all objects must be sorted into mutually-exclusive categories (Lim and
Lawless 2005). Free sorting incorporates panelists' normal thought processes into the
categorization of products and results in candid opinions. Another type of sorting method
is fixed sorting, which involves a sorting products based on predefined criteria (Coxon
1999). Panelists are confined in their sorting by either set categories or definitions.
Another sorting method is the projective mapping, which involves placing
products on a blank sheet of paper by similarities and differences in attributes in two
dimensions (Perrin and others 2008). The more similar the products are, the closer they
are placed on the sheet of paper. The paper is later divided and marked into uniform
squares to determine the distance of the products from one another. This method has
been used to sort ewe's milk cheeses (Barcenas and others 2004), wines (Pages 2005) and
orange juice (Nestrud and Lawless 2008). Flash profiling is another sensory method in
20
which panelists evaluate the products and come up with their own descriptive terms to
describe the products and cluster similar products together (Delarue and Sieffermann
2004, Tarea and others 2007).
When analyzing sorting data, results are based on consensus of the panel and do
not account for individual differences. Sorting data is analyzed through patterns seen in
the data and the relationships between the products sorted. A multidimensional scaling
(MDS) plot is used to understand the relationship between objects when underlying
dimensions are unknown (Schiffman and others 1981). MDS plots are used as an
analysis tool, as well as to visually understand the relationships among products and are
based on comparing data in a similarity or dissimilarity matrix. This similarity matrix is
based on the frequency in which products are sorted and placed together. Stress values
and correlations are used to determine the reliability of a sorting method. In a MDS plot,
the stress value indicates the goodness of the fit of all the data into the two-dimensional
plot. The more objects to be compared and plotted will result in larger stress values
(Borg and Groenen 2003).
There is currently no universal standard method to compare the generated clusters
from MDS plots or sorting methods. The RV coefficient (Risvik and others 1994, Szejtli
and Szente 2005, Abdi 2007) has been used to calculate the agreement in categories
generated by two different methods. The RV coefficient is calculated by comparing
configurations of the categories on a plot; therefore, the data must be transformed prior to
analysis such that each object has an X and Y coordinate point. The RV coefficient
ranges from 0 to 1, and a value close to 1 means that the two configurations have
excellent agreement. This method has been used as a means to validate categories
21
generated in various sorting research (Faye and others 2004, Carder and others 2006,
Lelicvre and others 2008). Tang and Heymann (2002) also calculated an RV coefficient
to compare the similarity between multidimensional sorting, similarity scaling and a free
choice profiling of grape jellies.
Another method to assess the agreement of the clusters generated between
methods is the Rand Index (RI) or Rand'Measure (Rand 1971) which has been used to
compare clustered data to determine the level of agreement in categories. The Rl (Rand
1971) examined the similarities between the agreements and disagreements through the
comparison of the results from the two sorts. The RI compares the generated categories
from two different sorting methods or generated category sorting results to established
categories. Unlike the RV coefficient, the data being compared does not have to be
transformed into a plot. A RI value ranging from 0 to 1 was calculated based on the
number of agreements and disagreements were seen through the comparison of the results
from the two sorts. An ARI value ranges from 0 to 1 and explains the correspondence
between the categories of the two compared sorts, with a value of 1 signifying that the
two sorts were exactly the same. An Adjusted Rand Index (ARI) (Hubert and Arabie
1985) was then developed to adjust for chance agreement between the two sorts. In
Steinly's research (2004), the same clustering method was compared and the RI was a
higher value than the ARI.
Therefore, the ARI is a more sensitive scale than RI for determining the degree of
correspondence between two sorts. The validity of the quality of the cluster recovery of
the ARI value is: >0.86 is in the 95th percentile, 0.77 is in the 90lh percentile, and 0.67 is
in the 85th percentile (Steinley 2004). An ARI less than 0.65 is considered poor recovery
22
of data, a value greater than 0.80 is considered good recovery, while a value greater than
0.90 is considered excellent recovery (Steinley 2004). The ARI is more commonly used
as an analysis tool in other fields of study than Food Science and Sensory Science
(Soufflet and others 2004), although there are a few food-related studies that used the RI
to compare groupings of products (Cartier and others 2006).
Categories provide insight on the attributes most representative of a group of
similar products (Mervis and Rosch 1981). If there are no specific criteria, panelists
generate their own context to compare the similarities of products to identify underlying
similarities to tie the products together. Research conducted on sorting and
categorization methods have been studied to validate Icnown categories (Viswanathan and
Childers 1999).
2.5 Concluding Remarks
Based on the works mentioned, it appears that there are many unanswered
questions regarding the ever-changing functional beverage market. There is no apparent
set of standard functional beverage categories in the US that is easily accessible to the
general public. The development of functional beverage categories is vital to
understanding and regulating the functional beverage market to lessen the confusion in
product marketing, consumer expectations, and legal guidelines. Defined functional
beverage categories would be useful during the development of products because it will
aid in targeting specific characteristics that match the category.
Few studies have been conducted on the individual effects of the incorporation of
functional ingredients into food or beverage matrices. It is important to investigate the
23
sensory effects of the inclusion of functional ingredients to create successful products.
This research seeks to investigate the creation of functional beverage categories. It also
aims to help create better tasting functional beverages through the research of the effects
of incorporating functional ingredients into the matrix and investigating a solution to
minimize the bitter taste of ginseng.
2.6 References
Anonymous 2008. U.S. Functional Beverages Market: A Young Market with Growing Popularity. Food Industry News [serial online]. Available from Posted 28 May 2008 2008.
Anonymous 1975. Minutes of Division Business Meeting. Institute of Food Technologists-Sensory Evaluation Division, IFT.
Abdi H. 2007. The RV coefficient and the congruence coefficient. Encyclopedia of measurement and statistics.Thousand Oaks (CA): Sage
Alldrick AJ. 2006. Functional beverages: satisfying consumers' needs. Agro Food Industry Hi Tech 17(4):41.
Allison ANNMA, Chambers E. 2000. The Importance Of Caffeine As A Flavor Component in Beverages. J Sens Stud 15(4):449-57.
Asao M, Iwamura I-I, Akamatsu M, Fujita T. 1987. Quantitative structure-activity relationships of the bitter thresholds of amino acids, peptides, and their derivatives. J Med Chem 30(10): 1873-9.
Astray G, Gonzalez-Barreiro C, Mejuto JC, Rial-Otero R, Simal-Gandara J. 2009. A review on the use of cyclodextrins in foods. Food Hydrocoll 23(7): 1631-40.
Barcenas P, Elortondo FJP, Albisu M. 2004. Projective mapping in sensory analysis of ewes milk cheeses: A study on consumers and trained panel performance. Food Res Inl 37(7):723-9.
Berry D. 2009. Powerful Beverage Performers. Food Product Design [serial online]. 32-50. Available from Posted 2009.
Bertino M, Lawless HT. 1993. Understanding mouthfeel attributes: a multidimensional scaling approach. J Sens Stud 8(2): 101-14.
24
Binello A, Cravotto G, Nano GM, Spagliardi P. 2004. Synthesis of chitosan-cyclodextrin adducts and evaluation of their bitter-masking properties. Flavour Fragrance J 19(5):394-400.
Borg I, Groenen P. 2003. Modern Multidimensional Scaling: Theory and Applications. Journal of Educational Measurement 40(3):277-80.
Brandt LA. 2001. Flavor Masking: Strategies for Success. Prepared Foods [serial online]. (July 2001):Available from Posted 2001.
Burg JC. 1998. High-Energy Beverages. Food Product Design
Camire ME. 2000. Dietary Supplements. In: M. K. Schmidl, T. P. Labuza, editors. Essentials of Functional Foods. Gaithersburg: Aspen Publishers, Inc. pi 65-180.
Cartier R, Rytz A, Lecomte A, Poblete F, Krystlik J, Belin E, Martin N. 2006. Sorting procedure as an alternative to quantitative descriptive analysis to obtain a product sensory map. Food Qual Pref 17(7-8):562-71/
Coon .IT, Ernst E. 2002. Panax ginseng: a systematic review of adverse effects and drug interactions. Drug Saf 25(5):323-44.
Cosgrove J. 2008. Caffeinated Snacks. Nutraceuticals World [serial online], July/August 2008Available from Posted July 2008.
Court WE. 2000a. Ginseng: The Genus Panax. Singapore: CRC Press. 266 p.
Court WE. 2000b. The Pharmacology and Therapeutics of Ginseng. In: Anonymous Ginseng: The Genus Panax. Singapore: Harwood Academic Publishers, pi 17-197.
Coxon APM. 1999. Sorting Data: Collection and Analysis. Sage Publications Inc.
Cravotto G, Binello A, Baranelli E, Carraro P, Trotta F. 2006. Cyclodextrins as Food Additives and in Food Processing. Current Nutrition & Food Science 2(4):343-50.
de Sousa PHM, Maia GA, de Azeredo HMC, de Souza Filho MSM, Garruti DS, de Freilas CAS. 2007. Mixed tropical fruit nectars with added energy components, hit J Food Sci Tech 42(11): 1290-1296.
Delarue J, Sieffermann JM. 2004. Sensory mapping using Flash profile. Comparison with a conventional descriptive method for the evaluation of the flavour of fruit dairy products. Food Qual Pref 15(4):383-92.
Delwiche J. 2004. The impact of perceptual interactions on perceived flavor. Food Qual Pref 15(2): 137-46.
25
Dodziuk I-I. 2006. Cyclodextrins and their complexes: chemistry, analytical methods, applications. Weinheim: Wiley-VCH. 489 p.
Drewnowski A, Gomez-Carneros C. 2000. Bitter taste, phylonutrienls, and the consumer:" a review. Am J Clin Nutr 72(6): 1424-35.
Falahee M, MacRae AW. 1997. Perceptual variation among drinking waters: the reliability of sorting and ranking data for multidimensional scaling. Food Qual Pref 8(5):389-94.
Farrell J, Ries NM, Kachan N, Boon H. 2009. Foods and natural health products: Gaps and ambiguities in the Canadian regulatory regime. Food Policy
Faye P, Bremaud D, Durand Daubin M, Courcoux P, Giboreau A, Nicod H. 2004, Perceptive free sorting and verbalization tasks with naive subjects: an alternative to descriptive mappings. Food Qual Pref 15(7-8):781-91.
Food and Drug Administration. 2003. Code of Federal Regulations Title 2 Sec. 182.1180.
Frank RA. 2003. Response context affects judgments of flavor components in foods and beverages. Food Quality and Preference 14(2): 139-45.
Gawel R, Oberholster A, Francis IL. 2000. A 'mouth-feel wheel': terminology for communicating the mouth-feel characteristics of red wine. Australian journal of grape and wine research.6(3):203-07.
Giboreau A, Dacremont C, Egoroff C, Guerrand S, Urdapilleta I, Candel D, Dubois D. 2007. Defining sensory descriptors: Towards writing guidelines based on terminology. Food Qual Pref 18(2):265-74.
Gnadt T. 2007. Effect of Masking Agents and Lactones on the Aroma and Flavor of Soymilk. [dissertation]. University of Illinois at Urbana-Champaign.
Granato H. 2002, Manipulating Flavor Perception in Functional Products. Natural Products Insider [serial online]. Available from Posted 8 April 2002 2002.
Hubert L, Arabie P. 1985. Comparing partitions. Journal of classification 2(1):193-218.
Humphries G. 2007. Nutraceutical Soft Drinks: Innovation in sports, energy, dairy, and functional beverages.
Institute of Food Technologists. 2005. Functional Foods: Opportunities and Challenges.
Jung MY, Jeon BS, Bock JY. 2002. Free, esterified, and insoluble-bound phenolic acids in while and red Korean ginsengs (Panax ginseng CA Meyer). Food Chem 79(1): 105-11.
26
Katan MB, Roos NM. 2004. Promises and Problems of Functional Foods. Crit Rev Food SciNutr44(5):369-77.
Keast RSJ. 2008. Modification of the bitterness of caffeine. Food Qual Pref 19(5):465-72.
Keast RSJ, Breslin PAS. 2003. An overview of binary taste-taste interactions. Food Qual Pref 14(2):111-24.
Kim . 1985. Effects of temperature and sugar addition on the flavor of ginseng lea. Korean Journal of Food Science and Technology. 17(4):304.
King MC, Cliff MA, Hall JW. 1998. Comparison of Projective Mapping and Sorting • Data Collection and Multivariate Methodologies for Identification of Similarity-Of-Use Snack Bars. J Sens Stud 13(3):347-58.
Kitts D, Hu C. 2000. Efficacy and safety of ginseng. Public Health Nutr 3(4A):473-85.
Koutsou GA, Storey DM, Bar A. 1999. Gastrointestinal tolerance of y-cyclodextrin in humans. Food Additives & Contaminants 16(7):313-7.
Lai GG. 2007. Getting Specific with Functional Beverages. Food Technology [serial online], 61(12):Available from Posted 2007.
Lawless HT, ShengN, Knoops SSCP. 1995. Multidimensional scaling of sorting data applied to cheese perception. Food Qual Pref 6(2):91-8.
LeClair K. 2000. Breaking the Sensory Barrier for Functional Foods. Food Product Design [serial online], 6 November 2006. Available from http://www.foodproductdesign.com/articles/462/462 0297DE.html. Posted 1 September 2000.
Lelicvre M, Chollet S, Abdi H, Valentin D. 2008. What is the validity of the sorting task for describing beers? A study using trained and untrained assessors. Food Qual Pref
Lesschaeve I, Noble AC. 2005. Polyphenols: factors influencing their sensory properties and their effects on food and beverage preferences. Am J Clin Nutr 81(1 Suppl):330S-5S.
Leveen T. 2007. Functional Beverages: Market Evolution. Natural Products Marketplace [serial online]. Available from Posted 2007.
Ley JP, Blings M, Paetz S, Krammer GE, Bertram HJ. 2006. New bitter-masking compounds: hydroxylated benzoic acid amides of aromatic amines as structural analogues of homoeriodictyol. J Agric Food Chem 54(22):8574-9.
Ley JP, Krammer G, Reinders G, Gatfield IL, Bertram HJ. 2005. Evaluation of bitter masking flavanones from Herba Santa (Eriodictyon californicum (H. & A.) Ton-., Hydrophyllaceae). J Agric Food Chem 53(15):6061-6.
Lim J, Lawless HT. 2005. Qualitative differences of divalent salts: multidimensional scaling and cluster analysis. Chem Senses 30(9):719-26.
Loftsson T, Brewster ME. 1996. Pharmaceutical applications of cyclodextrins. 1. Drug solubilization and stabilization. J Pharm Sci 85(10):
Lord JB. 2000. New Product Failure and Success. In: A. L. Brody, J. B. Lord, editors. Developing New Food Products for a Changing Marketplace. Lancaster: Technomic • Publishing Co, Inc. p55-86.
Lourenco R, Camilo ME. 2002. Taurine: a conditionally essential amino acid in humans? An overview in health and disease. Nutr Hosp 17(6):262-70.
Mattes RD. 2007. Effects of linoleic acid on sweet, sour, salty, and bitter taste thresholds and intensity ratings of adults. AJP-Gastrointestinal and Liver Physiology 292(5):G1243.
Mervis CB, Rosch E. 1981. Categorization of Natural Objects. Annual Reviews in Psychology 32(1):89-115.
Minor LJ, Pearson AM, Stine CM. 1966. Chicken Flavor Studies. J Agric Food Chem 14(4):416-9.
Mintel. 2009. Energy Drinks and Energy Shots-US.
Mintel. 2008. Functional Beverages-US August 2008. Mintel Reports.
Mobini S, Elliman TD, Yeomans MR. 2005. Changes in the pleasantness of caffeine-associated flavours consumed at home. Food Qual Pref 16(8):659-66.
Nestrud MA, Lawless HT. 2008. Perceptual mapping of citrus juices using projective mapping and profiling data from culinary professionals and consumers. Food Qual Pref 19(4):431-38.
Ohki K, Nakamura Y, Takano T. 2004. The Japanese Nutritional Health Beverage Market. In: Anonymous Beverages in Nutrition and Health. Totowa: Humana Press. p377-387.
Packaged Facts. 2009. Functional Foods and Beverages in the U.S. 1-210.
• 28
Pages J. 2005. Collection and analysis of perceived product inter-distances using multiple factor analysis: application to the study of 10 white wines from the Loire Valley. Food Qual Pref 16(7):642-9.
Perrin L, Symoneaux R, Maitre I, Asselin C, Jourjon F, Pages J. 2008. Comparison of three sensory methods for use with the Napping procedure: Case often wines from Loire valley. Food Qual Pref 19(1): 1-11.
Ramos de Armas R, Gonzalez Diaz H, Molina R, Perez Gonzalez M, Uriarle E. 2004. Stochastic-based descriptors studying peptides biological properties: modeling the bitter lasting threshold of dipeptides. Bioorg Med Chem 12(18):4815-22.
Rand WM. 1971. Objective criteria for the evaluation of clustering methods. Journal of the American Statistical Association 846-50.
Reid AW, Becker CH, REID AW. 1956. The use of cocoa syrups for masking the taste of quinine hydrochloride. J Am Pharm Assoc Am P.harm Assoc 45(3): 151-2.
Reineccius GA. 2004. Flavoring Systems for Functional Foods, In: T. Wilson, N. J. Temple, editors. Beverages in Nutrition and Health. Totowa: Humana Press. p89-97.
Risvik E, Mcewan JA, Colwill JS, Rogers R, Lyon DM. 1994. Projective mapping: a tool for sensory analysis and consumer research. Food Qual Pref 5(4):263-9.
Rouseff RL. 1990. Bitterness in Food Products and Beverages. Amsterdam: Elsevier. 356 P-
Roy G, Roy GM. 1997. Modifying Bitterness: Mechanism, Ingredients, and Applications. CRC Press.
Rugg G, McGeorge P. 1997. The sorting techniques: a tutorial paper on card sorts, picture sorts and item sorts. Expert Syst 14(2):80-93.
Saint-Eve A, Paci Kora E, Martin N. 2004. Impact of the olfactory quality and chemical complexity of the flavouring agent on the texture of low fat stirred yogurts assessed by three different sensory methodologies. Food Qual Pref 15(7-8):655-68.
Schiffman SS, Reynolds ML, Young FW. 1981. Introduction to Multidimensional Scaling: Theory, Methods, and Applications. New York: Academic Press. 413 p.
Schiffman SS, Sattely-Miller EA, Graham BG, Zervakis J, Butchko HH, Stargel WW. 2003. Effect of repeated presentation on sweetness intensity of binary and ternary mixtures of sweeteners. Chem Senses 28(3):219-29.
Schmidl MK, LabuzaTP. 2000. Essentials of Functional Foods. Gaithersburg: Aspen Pub. 395 p.
Smit HJ, Rogers PJ. 2002, Effects of'energy' drinks on mood and mental performance: critical methodology. Food Qual Pref 13(5):317-26.
Smith DM, Awad AC, Bennink MR, Gill JL. 1995. Cholesterol reduction in liquid egg yolk using p-cyclodextrin. J Food Sci 60(4):691-4.
Soldo T, Hofmann T. 2005. Application of Hydrophilic Interaction Liquid Chromatography/Comparative Taste Dilution Analysis for Identification of a Bitter Inhibitor by a Combinatorial Approach Based on Maillard Reaction Chemistry, J Agric Food Chem 53(23):9165-71.
Soufflet I, Calonnier M, Dacremont C. 2004. A comparison between industrial experts' and novices' haptic perceptual organization: a tool to identify descriptors of the handle of fabrics. Food Quality and Preference 15(7-8):689-99.
Spiller MA. 1998. Caffeine. Boca Raton: CRC Press. 363 p.
Steinley D. 2004. Properties of the Hubert-Arabie adjusted Rand index. Psychol Methods 9(3):386-96.
Stevens JC. 1997. Detection of very complex taste mixtures: generous integration across constituent compounds. Physiol Behav 62(5): 1137-43.
Stone H, Sidel JL. 2004. Sensoiy Evaluation Practices. Academic Press.
Szejtli J, Szente L. 2005. Elimination of bitter, disgusting tastes of drugs and foods by cyclodextrins. Eur J Pharm Biopharm 61 (3): 115-25.
Takeuchi T, Naae N. 1992. Stationary phase effect on enantioselectivity of dansyl phenylalanine in microcolumn liquid chromatography with y-cyclodextrin as mobile phase additive. J Chromatogr 595(1-2): 121-4.
Tamura M, Mori N, Miyoshi T, Koyana S, Kohri H, Okai H. 1990. Practical Debittering Using Model Peptides and Related Compounds. Agric Biol Chem 54(1):41-51.
Tang C, Heymann I-I. 2002. Multidimensional Sorting, Similarity Scaling and Free-Choice Profiling of Grape Jellies. J Sens Stud 17(6):493-509.
Tarea S, Cuvelier G, Sieffermann JM. 2007. Sensory Evaluation of the Texture of 49 Commercial Apple and Pear Purees. J Food Qual 30(6): 1121-31.
Tuorila I-I, Cardello AV. 2002. Consumer responses to an off-flavor in juice in the presence of specific health claims. Food Qual Pref 13(7-8):561-9.
Urala N, Lahteenmaki L. 2003. Reasons behind consumers functional food choices. Nutr Food Sci 33(4):148-58.
30
Viswanathan M, Childers TL. 1999. Understanding How Product Attributes Influence Product Categorization: Development and Validation of Fuzzy Set-Based Measures of Gradedness in Product Categories. J Market Res 36(l):75-94.
V.ogler BK, Pittler MH, Ernst E. 1999. The efficacy of ginseng. A systematic review of randomised clinical trials. Eur J Clin Pharmacol 55(8):567-75.
Vuksan V, Sievenpiper JL. 2005. Herbal remedies in the management of diabetes: Lessons learned from the study of ginseng. Nutrition, Metabolism and Cardiovascular Diseases 15(3): 149-60.
Warskulat U, Flogel U, Jacoby C, Hartwig HG, Thewissen M, Merx MW, Molojavyi A, Heller Stilb B, Schrader J, Haussinger D. 2004. Taurine transporter knockout depletes muscle taurine levels and results in severe skeletal muscle impairment but leaves cardiac function uncompromised. FASEB Journal 18(3):577.
Woolf A, Wright R, Amarasiriwardena C, Bellinger D. 2002. A child with chronic manganese exposure from drinking water. Environ Health Perspect 110(6):613-6.
Wright R. 2008. Nutraceuticals Coast in the Beverage Market. Nutraceuticals World [serial online]. Available from Posted July 2008.
Yeomans MR, Mobini S, Chambers L. 2007. Additive effects of flavour-caffeine and flavour-flavour pairings on liking for the smell and flavour of a novel drink. Physiol Behav92(5):831-9.
Yu J, Kim AK. 2009. Effect of Taurine on Antioxidant Enzyme System in B16F10 Melanoma Cells. In: Anonymous Taurine 7. New York: Springer. p491.
31
2.7 Tables
Table 2.1: Amount of Functional Ingredients in Commercially-available Energy Drinks purchased in Sept
Product Name Cocaine Full Throttle MDX No Fear NOS Red Bull Red Jak Rockstnr Sobe Adrenaline Rush Tab Eiicray
and Figure 3.3) and 2) Non-Newtonian products (Table 3.4 and Figure 3.4). Eight
categories were generated through the combination of both AHC analyses, although
there were no obvious similarities that could be used to characterize the generated
beverage clusters. The AHC of the Newtonian beverages show that most of the
decarbonated energy drinks were clustered together along with a few other beverages,
while diet and low calorie beverages were clustered together.
It is possible that the apparent viscosity at a shear rate of 50 s"1 was not the best
representative shear rate to categorize the non-Newtonian pseudoplastic functional
beverages. Pseudoplastic fluids have varying viscosities at different shear rates;
therefore, calculating the apparent viscosity at a lower set shear rate may result in a more
accurate representation of the apparent viscosity of the non-Newtonian functional
beverages than at a shear rate of 50 s'1.
The combination of other physical property measurements, such as pH level, Brix,
and titratable acidity, may provide additional criteria for the beverages to be categorized
in a more meaningful way. Diet and full calorie cola and lemon-lime flavored
commercial carbonated beverages were separately clustered through an AHC based on
44
these physical properties (Kappes 2007), which provided a more meaningful
interpretation of the clusters.
Two-Step Sensory Sorting Categorization
The visual free sort generated six beverage categories (Figure 3.5), while the
visual-oral free sorting method generated seven beverage categories (Figure 3.6). Six of
the beverage groups were the same and included: Enhanced Waters, Energy Drinks, Fruit
Smoothies, Nutritional Drinks, Sports Drinks, and Teas. The visual-oral sort included an
additional category termed, "Yogurt Smoothies". The resulting categories and definitions
remained consistent between the two sorting procedures, which suggest that visual and
visual-oral sorts could be combined into one task to reduce sorting time.
Panelists commented that it was difficult to place beverages encompassing
multiple concepts into a single category. Beverages such as Fuze® Slenderize (14),
Sobe® Tsunami (40), and Sobe® Lean Energy Diet Citrus (41) were a few of the hybrid-
type beverages that panelists had difficulty sorting. The results from the MDS plot
mirror the panelists' uncertainty because these "hybrid-type" beverages were not
consistently placed in the same category by the panelists, thus these beverages were not
grouped on the MDS plot. From this data, it can be concluded that there are functional
beverages on the market that lack a clear concept and are difficult for consumers to
define and characterize.
The fixed sort MDS plots (Part 2), resulted in fewer ungrouped beverages
(Figures 3.7 and 3.8) than the free sort MDS plots (Part 1). These results were expected
since the categories were fixed and panelists were forced to sort each beverage into a pre
defined category despite, perhaps, a non-ideal fit. The fixed sort, however, helped to
45
categorize some beverages such as Capri Sun Sport (4), Sobe Power (39), Sobe
Tsunami (40), and Sobe® Lean Energy Diet Citrus (41) (Figure 3.3) possibly because the
characteristics of these beverages matched the definitions of the fixed functional
beverage categories (Figures 3.7 and 3.8). Capri Sun® Sport IM (4) was placed in the
Sports Drinks category, while Sobe® Power (39), Sobe® Tsunami (40), and Sobe® Lean
Energy Diet Citrus (41) were placed in the Energy Drinks category. The three previously
mentioned beverages did not completely match the characteristics defining the categories,
but the results suggest that the panelists thought that these functional beverages best
belonged in those categories. The results from the fixed sort also suggest that panelists
were still uncertain about which categories to sort the hybrid-type beverages. Panelists
also commented that they had difficulty placing beverages incorporating multiple
concepts into a single category. Beverages such as Fuze® Refresh (13), Fuze® Slenderize
(14), Propel®'Propel® Fitness Water (33),and Fitness Water (34), were a few of the
hybrid-type beverages that panelists had difficulty sorting. Some of these beverages may
not have been categorized because the key characteristics of fixed functional beverage
categories may have limited inclusion into a categoiy.
The two-step sensory sorting method aids in developing categories based on
similarities among products. The fixed sort confirms the initial beverage categories
created during the free sort, through the general comprehension of the categories by the
panelists. Overall, the two-step sensory sorting method provides insight on the similarity
and dissimilarity relationships between products and aids in category development based
on the grouping of beverages with similar characteristics.
Comparison of Categorization Methods
Generated beverage categories were not consistent across the three methods. The
methods each resulted in the clustering of beverages; however the two-step sensory
sorting method generated distinct categories with distinguishing characteristics that are
meaningful compared to the ingredient inventory and flow behavior comparison
categorization methods. Categorization using the ingredient inventory and flow behavior
comparison methods was based on only one aspect of the beverages, while the two-step
sensory sorting method allowed panelists to categorize beverages using multiple senses
and a variety of information. Therefore, the two-step sensory sorting method was the
only method which included consumer insight in the categorization and definitions of
beverage categories.
Although the ingredient inventory and flow behavior comparison categorizations
were not as successful in categorizing functional beverages as the two-step sensory
sorting method, the information obtained about each beverage could possibly be used to
aid in describing the categories generated by the two-step sensory sorting categorization.
Comparing the ingredient and flow behavior data of individual beverages grouped
together into a category through the two-step sensory sort may result in additional
common characteristics distinctive of a category. For example, the beverages categorized
in the Sports Drinks category all contained multiple sugars and had Newtonian
viscosities; which could be the major chemical and physical characteristics of the Sports
Drinks category. This could possibly result in further defined functional beverage
categories.
47
One limitation of this study's design, regarding all three categorization methods,
is that too few beverages were selected from each of the potential categories. For
example, there was only one low-calorie carbonated energy drink and only two fruit-
based smoothies. If there had been more of those beverage types included in this study, it
is possible that other categories would have been generated. The beverages selected for
this categorization method, however, represented the assortment and quantities of
functional beverages available in the market reflecting the current market trend.
After comparing the results of all three methods, in general there are some basic
major categories that were apparent regardless of the categorization method applied:
Energy Drinks,, Nutritional Drinks,, Sports Drinks, Teas, and Yogurt Smoothies. There
are still some beverages which were difficult to categorize such as Fuze® "Refresh" and
Fuze® "Slenderize". Reasons for the difficulty could be attributed to consumers'
uncertainty or unfamiliarity of the key characteristics of these beverages. In addition, as
more hybrid-type products are introduced into the functional beverage market, a new
beverage category may be generated.
3.5 Conclusions
Strategically marketing a product to meet consumer expectations requires
knowledge of a product's categorical membership. The inclusion of a beverage into a
category provides a quick snapshot of the expected attributes of a particular beverage. It
is also important to determine and define functional beverage categories to aid in the
regulation of the beverages. Regulations would protect and inform consumers about
ingredient compositions specific to functional beverage categories. Therefore, it is
necessary to determine and define categories for the expanding functional beverage
market, which currently lacks distinct categories.
Categorizing functional beverages by ingredient inventory did not result in
categories, possibly due to the expansive list of ingredients compared. The flow behavior
comparison also did not result in distinct functional beverage categories with
distinguishing characteristics. This suggests that flow behavior and measured viscosity at
a shear rate of 50 s"1 may not be the best means of categorizing functional beverages.
The two-step sensory sorting method has potential as a method to categorize functional
beverages because the incorporation of sensory evaluation aids in generating distinct
functional beverage categories. Future research includes investigating the validity and
reproducibility of the two-step sorting method as a rapid method to categorize large
groups of products. Additional research may include an investigation of a free sort
focusing only on an oral sensory evaluation sort of the blinded products, to determine if
the functional beverages could be categorized by only oral sensations and tastes without
the influence of packaging or product identification. The level of sweetness, mouthfeel,
or other oral sensations may play a significant role in generating different functional
beverage categories.
3.6 References
Amendola C, lannilli I, Restuccia D, Santini I, Vinci G. 2004. Multivariate statistical analysis comparing sport and energy drinks. Innovative Food Science & Emerging Technologies 5(2):263-7.
Cook DJ, Hollowood TA, Linforth RST, Taylor A J. 2003. Oral shear stress predicts flavour perception in viscous solutions. Chem Senses 28(1):11-23.
Fraley C, Raftery AE. 1998. How Many Clusters? Which Clustering Method? Answers Via Model-Based Cluster Analysis. The Computer Journal 41(8):578-88.
Hollowood TA, Linforth RST, Taylor AJ. 2002. The Effect of Viscosity on the Perception of Flavour. Chem Senses 27583-91.
Humphries G. 2007. Nutraceutical Soft Drinks: Innovation in sports, energy, dairy, and functional beverages.
Kappes SK, Schmidt SJ, Lee SY. 2006. Mouthfeel Detection Threshold and Instrumental Viscosity of Sucrose and High Fructose Corn Syrup Solutions. J Food Sci 71(9):S597-602.
Kappes SM, Schmidt SJ, Lee SY. 2007. Relationship between Physical Properties and Sensory Attributes of Carbonated Beverages. J Food Sci 72(1):S001-11.
Krzanowski W.T, Marriott FHC. 1995. Multivariate analysis. New York: Halsted Press. 280 p.
Lai GG. 2007. Getting Specific with Functional Beverages. Food Technology [serial online]. 61(12):Available from Posted 2007.
Lawless HT, Glatter S. 1990. Consistency of multidimensional scaling models derived from odor sorting. J.Sensory Studies 5(2):217-30.
Lawless HT, Sheng N, Knoops SSCP. 1995. Multidimensional scaling of sorting data applied to cheese perception. Food Qual Pref 6(2):91-8.
Lim J, Lawless HT. 2005. Qualitative differences of divalent salts: multidimensional scaling and cluster analysis. Chem Senses 30(9):719-26.
Lord JB. 2000. New Product Failure and Success. In: A. L. Brody, J. B. Lord, editors. Developing New Food Products for a Changing Marketplace. Lancaster: Technomic Publishing Co, Inc. p55-86.
MacQueen JB. 1966. Some Methods for Classification and Analysis of Multivariate Observations. Proc. Fifth Berkeley Symp. on Math. Statist, and Prob.281-297.
Mintel. 2008. Functional Beverages-US August 2008. Mintel Reports.
Orth UR, de Marchi R. 2005. Advertising's influence on product experience and purchase intention. Fruit Processing 6:372-76.
Packaged Facts. 2009. Functional Foods and Beverages in the U.S. 1-210.
Rugg G, McGeorge P. 1997. The sorting techniques: a tutorial paper on card sorts, picture sorts and item sorts. Expert Syst 14(2):80-93.
50
Schmidl MK, Labuza TP. 2000. Essentials of Functional Foods. Gaithersburg: Aspen Pub. 395 p.
Tang C, Heymann H. 2002. Multidimensional Sorting, Similarity Scaling and Free-Choice Profiling of Grape Jellies. J Sens Stud 17(6):493-509.
Urala N, Lahteenmaki L. 2003. Reasons behind consumers functional food choices. Nutr Food Sci 33(4): 148-58.
Viswanathan M, Childers TL. 1999. Understanding How Product Attributes Influence Product Categorization: Development and Validation of Fuzzy Set-Based Measures of Gradedness in Product Categories. J Market Res 36(l):75-94.
Ward JH. 1963. Hierarchical grouping to optimize a quantitative function. J Am Stat Assoc 58(301):236-44.
Wright R. 2008. Nutraceuticals Coast in the Beverage Market. Nutraceuticals World [serial online]. Available from Posted July 2008.
Yamauchi T, Markman AB. 2000. Inference using categories. Journal of Experimental Psychology Learning Memory and Cognition 26(3):776-95.
51
3.7 Tables and Figures
Tabic 3.1: Fifty commercially-available functional beverages and corresponding numerical codes.
No. Code
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Commercially-Available Functional Beverages
Arizona® Pomegranate Green Tea
Bolthouse® Farms Fruit Smoothie
Boost®
Capri Sun® Sport ™
Dannon™ - Danimals®
Dannon™ - Light 'n Fit® Smoothie
Dannon™ - Frusion®
Dasani® Flavored Water
Elements Energy®
Ensure® Shake
Fruit20®
Full Throttle®
Fuze® "Refresh"
Fuze® "Slenderize"
Fuze® Green Tea
Gatorade® Endurance
Gatorade® Lemonade
Gatorade® Original
Gatorade® Rain
Glucerna®
Gold Peak™ Iced Tea
Honest Tea®
Lifeway® Lowfat Kefir
Liplon® Original White Tea
MDX
No. Code
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Commercially-Available Functional Beverages
Metromint®
Minute Maid® Fruit Falls™
Naked® Fruit Smoothie
NOS®
Pediasure®
Powerade™
PoweraderM - Advance
Powerade™ Option
Propel® Fitness Water
Rockstar®
Slimfast Optima®
Snapple® White Tea
Sobe® - NoFear
Sobe® - Power
Sobe® - Tsunami
Sobe® Lean Energy Diet Citrus
Sobe Life Water®
Stonyfield Farm® Organic Smoothie
Sweet Leaf™ Tea
TAB® Energy
Tazo® Iced Tea
Trinity® Water
Whitney's® Yo on the Go™
Yoplait® Go-GURT® Smoothie
Yoplait® Nouriche® SiiperSmoothie
52
Tabic 3.2: Functional beverage category names generated through visual observation of the beverage ingredient commonalities in the ingredient inventory spreadsheet.
Category Names Carbonated Energy Drinks
Energy Drinks
Nutritional Drinks
Sports Drinks
Teas
Waters
Yogurt Smoothies
Miscellaneous Drinks '
Beverages per
Category 6
4
5
8
8
7
8
4
Ingredient Commonalities B vitamins Carbonated Water Natural Extracts (Ginseng) Stimulants (Caffeine, Taurine, or D-ribose) MFCS Natural Extracts Stimulants (Caffeine, Taurine, or D-ribosc) Gums Protein Vitamins and Minerals Electrolytes (Na, K) Natural Sugars Natural Extracts Tea or water infused with tea
B Vitamins Natural Flavors Modified Starches Yogurt None
53
Tabic 3.3: Viscosities of Newtonian Functional Beverages measured at 20 C.
Code
11
27
8
45
26
33
34
42
41
16
47
17
14
18
19
32
37
15
4
44
21
31
24
9 46
35 25
1 12 29 38
39 13
Commercially-Available Beverages
Fruit20®
Minute Maid® Fruit FallsTM
Dasani® Flavored Water
TAB® Energy
Metromint®
PoweradeTM Option
Propel® Fitness Water
Sobe Life Water®
Sobe® Lean Energy Diet Citrus
Gatorade® Endurance
Trinity® Water
Gatorade® Lemonade
Fuze® "Slenderize"
Gatorade® Original
Gatorade® Rain
PoweradeTM - Advance
Snapple® White Tea
Fuze® Green Tea
Capri Sun® Sport TM
Sweet LeafTM Tea
Gold PeakTM Iced Tea
PoweradeTM
Lipton® Original White Tea
Elements Energy® Tazo® Iced Tea Rockstar® MDX
Arizona® Pomegranate Green Tea Full Throttle® NOS® Sobe® - NoFear Sobe® - Power
2.430 ±0.079 R2 values ranged from 0.989 to 1.000 witli an average and standard deviation of 0.999±0,002
Table 3.4: Viscosities of non-Newtonian Functional Beverages at a shear rate of 50 s .
Code
22
40
30
3
20
10
36
2
43
5
7
28
6
48
49
23
50
Commercially-Available Functional Beverages
Honest Tea®
Sobe® - Tsunami
Pediasure®
Boost®
Glucerna®
Ensure® Shake
Slimfast Optima®
Bolthouse® Farms Fruit Smoothie
Stonyfield Farm® Organic Smoothie
DannonTM - Danimals®
DannonTM - Frusion®
Naked® Fruit Smoothie
DannonTM - Light 'n Fit® Smoothie
Whitney's® Yo on the GoTM
Yoplait® Go-GURT® Smoothie
Lifeway® Lovvfat Kefir
Yoplait® Nouriche® SiiperSmoothie
Viscosity (mPa*s)
1.287 ±0.01
3.340 + 0.06
4.757 + 0.24
9.810±0.12
11.70± 0.77
15.75 ±0.15
32.81 ±2,60
42.03 ± 1.62
47.13 ±0.81
49.12 ±5.61
50.77 ± 1.01
52.45 ± 1.50
54.40 ±7.10
57.93 ±3.78
77.84 ± 2.27
100.6 ±5.0
121.4 ±28.3
55
50
40
30
20 '
•10 •
0 L-
-10
-20
-30
•40
-50
• so • 20
t 42
• 35 • 33 0
30 • 13 • 48
• 15 „ u • • 47
• 12 • ' 3 1 • ,34
• 40 * 20
•Ji 48 44'
• 41 monf^mm # • 25 « « 7 1<T # " g 5 3 10'
• 38 • • 2 7 1 ^ . 5
• 7 • 8
• 3
-50 -40 •30 -20 •10 • 0
Dim1
10 20 '30
• 30
30
40 50
Figure 3.1: Multidimensional Scaling of the results of the ingredient inventory categorization with stress = 0.227. • Refer to Table 3.1 for product names and corresponding number codes.
Figure 3.2: Agglomerative hierarchical clustering of 50 functional beverages by viscosity measurement using the ARES RFS III on the dissimilarity scale by Euclidean distance and agglomeration by Ward's Method. Viscosities of non-Newtonian beverages were calculated at a 50 sec"1 shear rate. The dotted line was computed using the software and truncates groups
• based on the largest relative increase in dissimilarity.
- 4
Figure 3.3: Agglomerative hierarchical clustering of 33 Newtonian functional beverages by viscosity measurement using the ARES RFS III on the dissimilarity scale by Euclidean distance and agglomeration by Ward's Method. The dotted line was computed using the software and truncates groups based on'the largest relative increase in dissimilarity.
Figure 3.4: Agglomerative hierarchical clustering of 17 non-Newtonian functional beverages by. viscosity measurement at 50 sec'1 shear rate using the ARES RFS III on the dissimilarity scale by Euclidean distance and agglomeration by Ward's Method. The dotted line was computed using the software and truncates groups based on the largest relative increase in dissimilarity.
-12 0 .
Dim1
12
Group
1 .
2
•• 3
. 4
5
6
Category
Energy Drinks
Enhanced Waters
Nutritional Drinks
Smoothies
Sports Drinks
Teas
Characteristics
Contains stimulants such as caffeine arid taurine; provides extra energy
Contains minimal Calories, lightly flavored, clear liquid Contains many nutrients could serve' as a meal-replacement beverage
Contains fermented dairy products; opaque ' Mineral or vitamin enhanced; does not contain caffeine. Labeling targets athletes Contains Tea or Tea Extracts
Figure 3.5: Multidimensional Scaling of a visual free sort (Part 1) of 50 functional beverages plotted in two dimensions with stress = 0.265 and functional beverage categories generated through the free visual sorting method. Functional beverages .. were grouped based on K-means clustering of beverage coordinate points on the MDS plot. Refer to Table 3.1 for product names and corresponding number codes.. -
12
CM
= 0 a
-4
-12
Group 6 ^ "
Group 1
Group 5
Group 4
Group £
/43 Group 3 ^ s 2 3
QD
-12 0
Dim1
12
Group
1
2
3
4
5
6
7 •
Category
Energy Drinks
Enhanced Waters
Fruit Smoothies
Nutritional Drinks
Sports Drinks
Teas
Yogurt Smoothies,
Characteristics
Contains stimulants such as caffeine and taurine; provides extra energy
Contains 100% Fruit,; non-clear liquid; all natural
Contains many nutrients could serve • as a meal-replacement beverage
Mineral or vitamin enhanced; does not contain caffeine. Labeling targets
Contains Tea or Tea Extracts
Contains fermented dairy products;
Figure 3.6:.Multidimensional Scaling of a visual-oral free sort (Part 1) of 50 functional beverages plotted in two dimensions with stress = 0.290 and functional beverage categories generated through the free visual-oral sorting method. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot. Refer to Table 3.1 for. product names and corresponding number codes.
12
£ 0
-12
Group 6
Group 2
Group 1
Group 4
o
Dim1
Group
1
2
3
4
5
6
7
Category -
Energy Drinks
Enhanced Waters
Fruit Smoothies
Nutritional Drinks
Sports Drinks
Teas
Yogurt Smoothies
Characteristics
Contains stimulants such as caffeine and taurine;provides extra energy '
Contains minimal Calories, lightly
flavored, clear liquid '
Contains 100% Fruit,; non-clear liquid;
all natural"
Contains many nutrients could serve as a meal-replacement beverage
Mineral or vitamin enhanced; does not contain caffeine. Labeling targets
Contains Tea or Tea Extracts
Contains fermented dairy products;
Figure 3.7: Multidimensional Scaling of a visual fixed sort (Part 2) of 50 functional beverages plotted in two dimensions with stress = 0.289 and corresponding functional beverage categories. .Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot. Refer to Table 3.1 for product names and corresponding number codes.
to
-12
Group 6 -j.
Group 2
Group 1
Group 4
Group 3
-8 0
Dim1
12
Group
1
2
3
4
5 •
6.
7
Category
Energy Drinks
Enhanced Waters
Fruit Smoothies-
Nutritional Drinks
Sports Drinks
Teas
Yogurt Smoothies
Characteristics
Contains'stimulants such as caffeine and taurine; provides extra energy
Contains minimal Calories, lightly flavored, clear liquid Contains 100% Fruit,; non-clear liquid; all natural Contains many nutrients could serve as a meal-replacement beverage
Mineral or vitamin enhanced; does not contain caffeine. Labeling targets
Contains Tea or Tea Extracts
Contains fermented dairy products;
Figure 3.8: Multidimensional Scaling of a visual-oral fixed sort (Part 2) of 45 functional beverages plotted in two dimensions with stress = 0.283 and corresponding functional beverage categories. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot. Refer to Table 3.1 for product names and corresponding number ;
codes.
CHAPTER 4 - VALIDATION AND REPRODUCIBILITY STUDY OF A TWO-STEP SENSORY SORTING METHOD TO CATEGORIZE FUNCTIONAL BEVERAGES
4.1 Abstract
Sorting and categorizing are quick and useful methods, which aid in identifying
product traits and highlighting important attributes. The aim of this study was to
determine the validity and reproducibility of a two-step sensory sorting method used to
categorize functional beverages based on qualitative judgments.
A validation study involved two groups of naive panelists sorting forty-six
functional beverages into the fixed categories pre-generated from the initial two-step
sensory sorting method. This was done to confirm that the fixed categories were
understandable and that the beverages could be similarly sorted. To determine if the two-
step sorting task was reproducible, the method was replicated with another group of nai've
panelists.
Adjusted Rand Index (ARI) values greater than 0,90 showed that there was
excellent correspondence between the fixed sorts conducted in the validation study.
Beverages that were difficult to sort in the initial two-step sensory sorting task, however,
were still not consistently categorized by the other two panels. Six functional beverage
categories were generated in the reproducibility study, with the major difference between
the initial sort and the replicated sort being that the Yogurt Smoothie and Fruit Smoothie
categories were combined into one category encompassing both types of beverages in the
replicate. The high ARI values from the validation study (ARI >0.88) and the similar
functional beverage categories generated through the reproducibility study (ARI >0.77)
64
suggest that the two-step sensory sorting method can be used to consistently create
and Trinity® Water (47)) had become unavailable on the market between conducting the
free and fixed sorting tasks and thus were not included in the visual-oral fixed sort.
Panelists were forced to sort the forty-six available beverages into the seven defined
categories generated from the combined results of the visual and visual-oral free sorting
tasks. Panelists were given a set of worksheets containing the list of the fixed functional
beverage categories and definitions, and sticker labels with the functional beverage
product names, Data from the fixed sorting tasks were analyzed using the same methods
as previously described in the free sort. Eight classes (seven categories plus an extra
class) were selected as the K-means value. The extra class was added to account for the
beverages that were not observed to be in a cluster on the MDS plot. The fixed sorting
task data were compared to the functional beverage categories generated through the free
sorting task, by calculating the Adjusted Rand Index (ARI).
Validation Study
A validation study was conducted on the categories generated through the two-
step sensory sorting method. The forty-six functional beverages were sorted into seven
previously generated functional beverage categories by two panels of naive panelists
(Panel 2 and Panel 3). The seven previously-generated functional beverage categories
were based on the combined results of both the visual and visual-oral free sorting tasks.
The panelists were not exposed to the free sorting task of the initial panel (Panel 1) and
were forced to sort beverages into the fixed categories generated by Panel 1. Panel 2
73
consisted often untrained panelists (5 males, 5 females, 18 to 50 years old) and Panel 3
consisted of thirteen untrained panelists (4 males, 9 females, 18 to 50 years old). The
panels followed the same instructions and guidelines as the fixed sort from the two-step
sensory sorting method of Panel 1. The same four functional beverages were not
(available for evaluation by Panel 2, while Sobe® Lean Energy Diet Citrus (41) became
available again, but Yoplait® Nouriche® Smoothie (50) became unavailable in the market
for evaluation by Panel 3.
The fixed sort data from the validation study (Panels 2 and 3) were analyzed using
the same methods previously described in the free sort. An ARI was conducted to
compare Panel 2 and Panel 3's sorting data to the functional beverage categories
generated through the initial two-step sensory sorting method. Since four of the
beverages were not available during the validation fixed sorts, only functional beverages
present in both sorts were used in the analyses. Therefore, forty-six beverages were
compared in this validation study.
Reproducibility Study
Panel 4 consisted of thirteen untrained panelists (4 males, 9 females, 18 to 50
years old), who replicated the two-step sensory sort (free sorting task followed by fixed
sorting task) on forty-six functional beverages. Four beverages (Elements Energy® (9),
Powerade™ Advance (32), Sobe® Lean Energy Diet Citrus (41), and Trinity® Water
(47)) had become unavailable on the market between conducting the Panel 1 's free and
fixed sorts. Sobe® Lean Energy Diet Citrus (41) had become available while Yoplait®
Nouriche®SuperSmoothie (50) had become unavailable between conducting the initial
two-step sensory sort and the reproducibility study. In total, there were five beverages
that were not sorted during compared fixed sorting tasks. Therefore, forty-five functional
74
beverages were compared in the reproducibility study. The data were analyzed using the
same statistical analyses as the initial two-step sensory sorting method (Panel 1). The
reproducibility of the method was analyzed using the ARI by comparing the categories
generated from Panel 1 and Panel 4's two-step sensory sorts.
4.4 Results and Discussion
Initial Two-Step Sensory Sorting Method (Panel 1)
Panel l 's visual free sort resulted in six functional beverage categories which
included: Energy Drinks, Enhanced Waters, Fruit Smoothies, Nutritional Drinks, Sports
Drinks, and Teas (Figure 4.6). The visual-oral free sort resulted in seven categories, six
that were the same categories and an additional category named, "Yogurt Smoothies"
(Figure 4.7). The ARI between both free sorts (Table 4.2) was 0.80, which is good
recovery between the sorts. The categories and definitions remained consistent between
the two free sorts, which suggest that visual and visual-oral sorts could be combined into
one task to reduce sorting time. The visual-oral sort incorporates both visual and oral
evaluations of the products which provides a complete sensory experience, thus is the
recommended task to choose to reduce the sorting time of the two-step sensory sorting
method.
The stress value of the MDS plot for visual free sort was 0.265 and 0.290 for the
visual-oral free sort, which are above the acceptable stress level of less than 0.10
(Krzanowski and Marriott 1995). The stress values were high and suggest that the
configuration on the dimensions may not be acceptable. However, the larger the number
of objects compared, the greater the stress value (Kruskal and Wish 1978). In previous
research, stress values have been lower because fewer objects were being compared
75
(Lawless and others 1995, Faye and others 2006). Plotting a large number of products
increases the number of comparisons and makes it difficult to display all the product
relationships in only two dimensions. When analyzing data using an MDS plot, it is
expected that all the relationships between objects will appear on one plot. The
relationships between products on a plot that are further apart are much more accurate
than the relationships between products that are plotted closer together on a MDS plot.
To reduce the high stress value, three dimensions could be plotted on a three-dimensional
MDS plot.
Panelists commented that it was difficult to place beverages incorporating
multiple concepts into a single category. Beverages such as Fuze® Slenderize (14),
Sobe® Tsunami (40), and Sobe® Lean Energy Diet Citrus (41) were a few of the hybrid-
type beverages that panelists had difficulty sorting. The results from the MDS plot
mirror the panelists' uncertainty because these hybrid-type beverages were not
consistently placed in the same category, thus they were not clustered into a particular
group of beverages (Figures 4.2 and 4,3). It can be concluded that there are functional
beverages on the market that lack clear concepts and are difficult for consumers to
categorize.
Both the visual and visual-oral fixed sort MDS plots, resulted in fewer ungrouped
beverages (Figures 4.4 and 4.5) compared to the free sorts. These results were expected
since the categories were fixed and panelists were forced to sort each beverage into pre
defined categories. The results from the fixed sort also suggest that panelists still had
difficulty sorting hybrid-type beverages such as products Fuze® "Refresh" (13), Fuze®
"Slenderize"(14), Powerade™ Option (33), and Propel® Fitness Water (34) into the fixed
categories because of the disagreement in the placement of beverages. Some beverages
may not have been categorized because the defined fixed functional beverage categories
may have been too restrictive, which excluded beverages from categories. The fixed sort,
however, helped to categorize some beverages such as Capri Sun® Sport1 M (4), Sobe®
Power (39), Sobe® Tsunami (40), and Sobe® Lean Energy Diet Citrus (41) (Figure 4.4
and 4.5), possibly because the characteristics of these beverages matched the definitions
of the fixed functional beverage categories.
The two-step sensory sorting method aids in developing categories based on
panelists' judgments of the similarities among products. Categories were generated
through the free sort, while the fixed sort was instrumental in verifying the generated
categories through the similar placement of beverages into pre-generated categories.
Descriptive analysis studies and free sorts often result in similar product descriptor results
(Faye and others 2004), which suggests that the two-step sensory sorting method may be
a faster method to obtain product descriptors.
The two-step sensory sorting method has the potential to be used as a rapid
sorting method to categorize a large quantity of products. The two-step sensory sorting
method consisted of two approximately one-hour sessions, which makes it a relatively
quick method to gather information about product relationships. There was no group
consensus in the generation of categories; therefore, the development of categories relied
heavily on a series of statistical analyses. Overall, the two-step sensory sorting method
provides insight on the similarity and dissimilarity among products and aids in category
development based on the grouping of beverages with similar characteristics.
77
Validation Study
In general, Panels 2 and 3 sorted the functional beverages into the same categories
as Panel 1 (Table 4.5). Panel 1 and 2 had an ARI of 0.91 for the comparison of the fixed
visual sort and an ARI of 0.93 comparing the fixed visual-oral sorts (Table 4.4). Panels 1
and 3 had an ARI of 0.88 for the fixed visual sort and 0.93 for the fixed visual-oral sort
(Table 4.2). Across all panels, fixed visual and visual-oral sorts, Fuze® "Refresh"(13)
and Fuze® "Slenderize"(14) were not sorted into the same categories by the panelists;
therefore, they did not fall into a specific functional beverage category. In the visual sort,
however, these two functional beverages were placed in the "Enhanced Water" category.
A reason for the placement of beverages into different categories could be due to the
varying judgments of the panelists.
Propel® Fitness Water (34) and Powerade™ - Option (33) were not categorized
by Panel 1; however, Propel® Fitness Water (34) was placed in the "Enhanced Water"
category in Panel 2's visual-oral fixed sort and in Panel 3's visual and visual-oral fixed
sort. A reason the beverages may not have been sorted into a category is that they were
clear in color, unlike other sports drink types, which made panelists uncertain about
which category these beverages belonged. Powerade™ - Option (33) was placed in the
"Sports Drinks" category in both Panel 2 and 3's visual-oral fixed sort, but was not sorted
into a specific category by visual sorting tasks. A possible reason for these results may
be attributed to panelists evaluating that Powerade™ - Option (33) was similar in taste
and mouthfeel but dissimilar in visual cues to other sports drink-type beverages.
Overall, the functional beverage categories generated from the two-step sensory
sorting method were understandable to naive panelists who had not participated in
78
creating the categories. In Tang and Heymann's (2002) research, naive subjects have
comparable product positioning as an expert panel. Therefore, untrained panelists have
similar ability as trained panelists in describing and categorizing functional beverages.,
The results suggest that the Teas, Yogurt Smoothies, Fruit Smoothies, and Nutritional
Drinks were understandable, well-defined categories. The functional beverages chosen in
this study were placed in consistent categories for these four categories by all three panels
(Table 4.3).
Categories in which there were a few discrepancies in beverage placement
included Energy Drinks and Enhanced Waters. The validation study results suggest that
the functional beverage categories generated by panelists in the two-step sensory sorting
method can be understood and used by a naive group of panelists who did not participate
in the free sort.
Reproducibility Study
The purpose of the reproducibility study was to determine if the same functional
beverage categories could be reproduced by a different panel. The categories generated
and replicated may aid in the creation of functional beverage categories that are
universally-understood by the general population. Panel 4's two-step sensory sorting
method resulted in six functional beverage categories (Figure 4.10). The categories
generated by Panel 4 were consistent with Panel l 's categories except that a "Smoothies"
group was created and defined to encompass both dairy and fruit-based smoothies and a
"Carbonated Energy Drink" category replaced Panel l's "Energy Drink" category.
Similar functional beverage categories were expected and observed through the
reproducibility study. The ARI values were all above 0.80 (Table 4.2) which means that
79
there was a good recovery of the compared data. The ARI comparing the visual-oral
fixed sorts between Panel 1 and 4 was 0.77. Possible reasons for the lower value could
be primarily attributed to the different number of fixed functional beverage categories,
and that functional beverages (Powerade™ Option (33), Propel® Fitness Water (34),
Sobe® No Fear (38), Sobe® Power (39), and Sobe® Tsunami (40)) were not sorted into a
category by either panel (Table 4.4). The similar functional beverage categories and high
ARI values suggest that the two-step sensory sorting method is reproducible when
attempting to categorize a set of 45 to 50 functional beverages.
In both the visual and visual-oral free sorts, Panel 4 had difficulty categorizing
Fuze® MRefrcsh"(13), Fuze® MSlenderize"(14), Sobe® Power (39), Sobe® Tsunami (40),
and Sobe® Lean Energy Diet Citrus (41), which were the same functional beverages not
sorted into categories by Panel 1 (Figures 4.10 and 4.11). The fixed sort resulted in
Sobe® Power (39) and Sobe® Tsunami (40) not being placed into a category because the
definition of the newly created "Carbonated Energy Drinks" category excluded the two
beverages. Compared to Panel 1 's data, Panel 4 could not sort Powerade™ Option (33),
and Propel® Fitness Water (34) into a category, but these beverages were placed in the
"Fitness and Sports Drinks" and "Flavored Waters", respectively (Table 4.4, Figures 4.12
and 4.13).
The results from the reproducibility study suggest that there may be functional
beverage categories commonly used and understood. Two sets of panelists were able to
generate a set of similar functional beverage categories with good correspondence. Since
the results from the reproducibility study showed similar generated categories and
80
placement of beverages into categories, a minimum of only thirteen panelists may be
necessary to complete the two-step sensory sorting method,
4.5 Conclusions
We have described a two-step sensory sorting method based on panelists' sorting
of 45 to 50 commercially-available functional beverages. This approach combines
sensory observation of products and the use of multivariate statistics in the development
of functional beverage categories. The two-step sensory sorting method is still in its
initial stages and needs more research to be considered a valid categorization method.
Future studies include combining the results of the sorting research with a
consumer test on the same functional beverages. An external preference map could-be
generated based on the functional beverage categories and the beverages preferred by
consumers. This could aid in the prediction of potential opportunities in the market for
new product development based on the relationship between functional beverage
categories and consumers' preferred beverage characteristics.
Other possible studies include testing the two-step sensory sorting method on
other products such as cereals, candies, and food bars. One study would be to have a
panel sort a set of products that fall into well-known, defined categories to determine if
the method accurately categorizes the products into predetermined well-defined
categories. Another study could be conducted by sorting based on only oral perceptions
using coded samples, to determine if the functional beverages could be categorized by
only oral sensations and tastes without the influence of packaging or product
identification. The level of sweetness, mouthfeel, and other oral sensations may play a
role in different functional beverage categories. Lastly, it would be interesting to conduct
the two-step sensory sorting method on a group of functional beverages including newly
81
introduced products to determine if new functional beverage categories would be created
or if the categories generated in this study would suffice.
4.6 References
Bertino M, Lawless HT. 1993. Understanding mouthfeel attributes: a multidimensional scaling approach. J Sens Stud 8(2): 101-14.
Delarue J, Sieffermann JM. 2004. Sensory mapping using Flash profile. Comparison with a conventional descriptive method for the evaluation of the flavour of fruit dairy products. Food Qual Pref 15(4):383-92.
Falahee M, MacRae AW. 1997. Perceptual variation among drinking waters: the reliability of sorting and ranking data for multidimensional scaling. Food Qual Pref 8(5):389-94.
Faye P, Bremaud D, Durand Daubin M, Courcoux P, Giboreau A, Nicod H. 2004. Perceptive free sorting and verbalization tasks with naive subjects: an alternative to descriptive mappings. Food Qual Pref 15(7-8):781-91.
Faye P, Bremaud D, Teillet E, Courcoux P, Giboreau A, Nicod I-I. 2006. An alternative to external preference mapping based on consumer perceptive mapping. Food Qual Pref 17(7-8):604-14.
Fraley C, Raftery AE. 1998. How Many Clusters? Which Clustering Method? Answers Via Model-Based Cluster Analysis. The Computer Journal 41(8):578-88.
Gawel R, Oberholster A, Francis IL. 2000. A 'mouth-feel wheel1: terminology for communicating the mouth-feel characteristics of red wine. Australian journal of grape and wine research,6(3):203-07.
Giboreau A, Dacremont C, Egoroff C, Guerrand S, Urdapilleta I, Candel D, Dubois D. 2007. Defining sensory descriptors: Towards writing guidelines based on terminology. Food Qual Pref 18(2):265-74.
Hubert L, Arabic P. 1985. Comparing partitions. Journal of classification 2(1):193-218.
King MC, Cliff MA, Hall JW. 1998. Comparison of Projective Mapping and Sorting Data Collection and Multivariate Methodologies for Identification of Similarity-Of-Use Snack Bars. J Sens Stud 13(3):347-58.
Kruskal JB, Wish M. 1978. Multidimensional scaling Quantitative applications in the Social Sciences. Sage Publications Newbury Park. 93p.
Krzanowski WJ, Marriott FHC. 1995. Multivariate analysis. New York: Halsted Press. 280 p.
Lawless HT, Sheng N, Knoops SSCP. 1995. Multidimensional scaling of sorting data applied to cheese perception, Food Qual Pref 6(2):91-8.
Lim J, Lawless HT. 2005. Qualitative differences of divalent salts: multidimensional scaling and cluster analysis. Chem Senses 30(9):719-26.
MacQueen JB. 1966. Some Methods for Classification and Analysis of Multivariate Observations. Proc. Fifth Berkeley Symp. on Math. Statist, and Prob.281-297,
Moreau CP, Markman AB, Lehmann DR. 2001. "What is it?" Categorization flexibility and consumers' responses to really new products.-Journal of Consumer Research 27(4):489-98.
Murphy GL, Ross BH. 1994. Predictions from uncertain categorizations. Cogn Psychol 27(2): 148-93.
Nestrud MA, Lawless HT. 2008. Perceptual mapping of citrus juices using projective mapping and profiling data from culinary professionals and consumers. Food Qual Pref 19(4):431-38.
Pages J. 2005. Collection and analysis of perceived product inter-distances using multiple factor analysis: application to the study of 10 white wines from the Loire Valley. Food Qual Pref 16(7):642-9,
Perrin L, Symoneaux R, Maitre I, Asselin C, Jourjon'F, Pagfis J. 2008. Comparison of three sensory methods for use with the Napping® procedure: Case often wines from Loire valley. Food Qual Pref 19(1): 1-11.
Rand WM. 1971. Objective criteria for the evaluation of clustering methods. Journal of the American Statistical Association 846-50.
Rugg G, McGeorge P. 1997. The sorting techniques: a tutorial paper on card sorts, picture sorts and item sorts. Expert Syst 14(2):80-93.
Saint-Eve A, Paci Kora E, Martin N. 2004. Impact of the olfactory quality and chemical complexity of the flavouring agent on the texture of low fat stirred yogurts assessed by three different sensory methodologies. Food Qual Pref 15(7-8):655-68.
Steinley D. 2004. Properties of the Hubert-Arable adjusted Rand index. Psychol Methods 9(3):386-96.
83
Tang C, Heymann H. 2002. Multidimensional Sorting, Similarity Scaling and Free-Choice Profiling of Grape Jellies. J Sens Stud 17(6):493-509.
Tarea S, Cuvelier G, Sieffermann JM. 2007. Sensory Evaluation of the Texture of 49 Commercial Apple and Pear Purees. J Food Qual 30(6): 1121-31.
Viswanathan M, Childers TL. 1999. Understanding How Product Attributes Influence Product Categorization: Development and Validation of Fuzzy Set-Based Measures of Gradedness in Product Categories. J Market Res 36(l):75-94.
Woolf A, Wright R, Amarasiriwardena C, Bellinger D. 2002. A child with chronic manganese exposure from drinking water. Environ Health Perspect 110(6):613-6.
Wright R. 2008. Nutraceuticals Coast in the Beverage Market. Nutraceuticals World [serial online]. Available from Posted July 2008.
4.7 Tables and Figures
Table 4.1: Fifty commercially-available functional beverages and corresponding numerical codes.
No. Code
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
Commercially-Available Functional Beverages
Arizona® Pomegranate Green Tea
Bolthouse® Farms Fruit Smoothie
Boost®
Capri Sun® Sport ™
Dannon™ - Danimals®
Dannon™ - Light 'n Fit® Smoothie
Dannon™ - Frusion®
Dasani® Flavored Water
Elements Enerpy®
Ensure® Shake
Fruit20®
Full Throttle®
Fuze® "Refresh"
Fuze® "Slenderize"
Fuze® Green Tea
Gatorade® Endurance
Gatorade® Lemonade
Gatorade® Original
Gatorade® Rain
Glucerna®
Gold Peak™ Iced Tea
Honest Tea®
Lifeway® Lowfat Kefir
Lipton® Original White Tea
MDX
No. Code
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
Commercially-Available Functional Beverages
Metromint®
Minute Maid® Fruit Falls™
Naked® Fruit Smoothie
NOS®
Pcdiasure®
Powerade™
Powerade™ - Advance
Powerade™ Option
Propel® Fitness Water
Rockstar®
Slimfast Optima®
Snapple® White Tea
Sobe® - NoFcnr
Sobe® - Power
Sobe® - Tsunami
Sobe® Lean Energy Diet Citrus
Sobe Life Water® Stonyfield Farm® Organic Smoothie
Sweet Leaf™ Tea
TAB® Energy
Tazo® Iced Tea
Trinity® Water
Whitney's® Yo on the Go™
Yoplait® Go-GURT® Smoothie
Yoplait® Nouriche® SiiperSmoothie
85
Table 4.2: Adjusted Rand Index values of the comparison of clusters generated through free and fixed sorting tasks by Panels 1 to 4.
Table 4.3: Compilation of Panel 2 and 3's validation study results of commercially-available functional beverages sorted into categories by visual and visual-oral fixed sorts compared to Panel l 's results. The categories were determined based on K-means clustering. ED=Energy Drinks, EW=Enhanced Waters, FRU=Fruit Smoothies, M=MisceIIancous, NUT=Nutritional Drinks, SP=Sports Drinks, T=Tcas, and YOG=Yogurt Smoothies. # denotes beverages that were not available for purchase during the time of the sorting tasks.
87
Table 4.3 (cont.)
No, Code
12 25
29 35
38 45 39
40
41 9
32
8
11 26
27 42 47
13 14 2
28 34 33
3 10
20 30 36
4 16
17 18
19 31
1 15
21 22 24
37 44 46
5 6
7 23
43 48
49 50
Commorclally-Avallablo Beverages
Full Throttle'1'
MDX NOS®
Rockstar® Sobe®-NoFear TAB® Energy
Sobe® - Power
Sobe®-Tsunami Sobe® Lean Energy Diet Citrus
Elements Energy® Powerade™ - Advance
Dasani® Flavored Water Fruit20® Metromint®
Minute Maid® Fruit Falls™ Sobe Life Water® Trinity® Water
Fuze® "Refresh" Fuze® "Slenderize" Bolthouse® Farms Fruit Smoothie
Naked® Fruit Smoothie Propel® Fitness Water Powerade™ Option Boost® Ensure® Shake
Tabic 4.4: Compilation of Panel 1 and 4's reproducibility study results comparing commercially-available functional beverages sorted into categories by visual and visual-oral free and fixed sorting task results. The categories were determined based on K-mcans clustering. ED=Energy Drinks, EW=Enhanccd Waters, FRU=Fruit Smoothies, M=Misccllaneous, NUT=Nutritional Drinks, SP=Sports Drinks, T=Tcas, and YOG=Yogurt Smoothies. # denotes beverages that were not available for purchase during the time of the sorting tasks.
Tab
No. Code
12
25
29
35
38
45
9
32
39
40
41
8
11
26
27
42
47
13
14
2
28
34
33
3
10
20
30
36
4
16
17
18
19
31
1
15
21
22
24
37
44
46
5
6
7
23
43
48
49
50
c 4.4 (cont.)
Commerc ia l l y -Ava i l ab le Boveraqes
Full Throttle®
MDX
NOS®
Rockstar®
Sobe* - NoFear
T A B " Energy
Elements Energy*
Powerade™ - Advance
Sobe® - Power
Sobe®-Tsunami
Sobe® Lean Energy Diet Citrus
Dasani® Flavored Waler
Frult20®
Metromint®
Minute Maid® Fruit Falls™
Sobe Lite Water®
Trinity® Waler
Fuze® "Refresh"
Fuze® "Slenderize"
Bolthouse® Farms Fruit Smoothie
Naked® Fruit Smoothie
Propel® Fitness Water
Powerade™ Option
Boost®
Ensure® Shake
Glucerna®
Pediasuro®
Slimfast Optima®
Capri Sun® Sport ™
Gatorade® Endurance
Gatorade® Lemonade
Gatorade® Original
Gatorade® Rain
Powerade™
Arizona® Pomegranate Green Tea
Fuze® Green Tea
Gold Peak™ Iced Tea
Honest Tea®
Lipton® Original Whi te Tea
Snapple® White Tea
Sweet Leaf™ Tea
Tazo® Iced Tea
Dannon™ - Danlmals®
Dannon™ - Light 'n Fit® Smoothie
Dannon™ - Frusion®
Lifeway® Lowfat Kefir
Stonyfield Farm® Organic Smoothie
Whitney's® Yo on the Go™
Yoplai t* Go-GURT® Smoothie
Yoplait® Nouriche® SuperSmoolhle
Froo V isua l
Panot 1
ED
ED
ED
ED
ED
ED
ED
ED
''";f M" •
"•.•.•> M ' •
• '••'"M .
EW
EW
EW
EW
EW
EW
;•• M
''• M
YOG
. Y O G •
EW
SP
WWr!® &RUT& •v?'Ninv*s ^'NUfRl $f t iUT$
••"M '• SP
SP
SP
SP
SP
T
T
T
T
T
T
T
T
•."•YOG.
i YOG
' . .YOG
' Y O G . ' '
V . Y O G :
• ; Y O G
;'.'YOG
YOG
Freo Visual-Oral Panel 1
ED
ED
ED
ED
ED
ED
ED
" . M
M
' .M
M
EW
EW
EW
M
EW
EW
M
• M ••
FRU
FRU
SP
SP
i.J&iNMTi'-ti-i;
; ! ! #NUT&$
^ f i N U T ^ j -
:WNLIT)#; :
foiSNlntfvSv
M
SP
SP
SP
SP
SP
T
T
. T
T
T
T
T
T
• ' " . ' lYOG' '
. YOG •
. YOG •
' . . ' .YOG •'
.• -.YOG •
- . . Y O G •
"YOG.
YOG
FIxod V isua l Panol 1
ED
ED
ED
ED
ED
ED
ED
ED
ED
ED
ED
EW
EW
EW
EW
EW
EW
EW
EW
FRU
FRU
^ M : . <
'$>.'M';3:.'
&NUTV?
ittNUTft SflltJTvff
'MU& #N.Ut&
SP
SP
SP
SP
SP
SP
T
'•t'T
:';' T ' ' X
X
.-. X T
T
i.-YOG.
• ' . 'YOG.
' ;YOG •
l l Y O G
. i 'YOG •,
l uYOG. ;
kTtae „• •:;.y.o'G
FIxod V isua l -Ora l
Panol 1
ED
ED
ED
ED
ED
ED
it
# ED
ED
it
EW
EW
EW
EW
EW
it
'.• M >
' ' M '
FRU
FRU
.::MM*&'
?^MiiS::s ISiiNUT®
>'JSM(JT;®i •S'f.ROt;;̂ ',' v.&iftU.T-Htt
^ N U t l ^ ! SP
SP
SP
SP
SP
SP
• T
X
X
• T
X •
• T
• X
T
;'' : r.,YOG/.
. ' . Y O G . . .
:.. YOG
• ..YOG
•••'.-YOG--...-
. ' • ;YOG. ' ',
, ••:•..'YOG v".
.'•AYOG'. '
Froo V isua l Panol 4
ED
ED
ED
ED
ED
ED
ED
..VM . •
V'M ' ' . J M . .
"••.' M .
EW
EW
EW
EW
EW
EW
• ' M
V'"M . ' Y O G
.; YOG •
r.^M-'-:. SP
• » * ^'NUjrii':'
mm H.WT&: piNUT,N
SP
SP
SP
SP
SP
SP
T
T .
' T
• -T
'• T
T
T
T
"YOG
''.'.YOG
•.•yoG.
'::'.!YOG
'5YOG
J'YO'G'':
;;;,YOG' •
" Y O G '
Free V isua l -
Ora l Panol 4
ED
ED
ED
ED
ED
ED
it
it
' ." ' M '
' M '
M
EW
EW
EW
EW
EW
if
M
' M '
YOG
" •YOG; . ;
EW
SP
SkNfthJK ?#NUTv&
ft'N'UM'f
^ N U T * ! : v
.."NUTi?1'
• M
SP
SP
SP
SP
SP
T
T
X
• T
T
T
T
r Y O G .
YOG
.YOG
. YOG
. .YOG •
• vYOG •
t O G
it
FIxod V i sua l Panel 4
ED
ED
ED
ED
ED
ED
ED
SP
ED
ED
ED
EW
EW
EW
EW
EW
EW
. M
• M-- '
• Y O G '
. Y O G .
'y>M-
SP
^IKUTA!
'>'!Nmxi< y^ura
tf.NU-tf;
SNUT*: SP
SP
SP
SP
SP
SP
T
X
T
T
T
T
T
T
YOG
YOG
Y O G .
'•' iYOG
.YOG, . .
YOG' . -
.-•YOG .
• Y O G
F ixed V i sua l -
Oral Panol 4
ED
ED
ED
ED
M
ED
it
it
•'•' M
'. M •
. 'M
EW
EW
• EW
EW
EW
it
•' M .
'• " M
YOG
-YOG
EW
SP
SNUtiS .^tipf/Vi
v?,N.UT./.:;
• ^ N U J ; * ?
•*$Wff-\
SP
SP
SP
SP
SP
SP
T
T
T
T
T
T
X
T
YOG
:• YOG
' Y O G • '
YOG
YOG.
/•YOG
• YOG
it
Two-step Sensory Sorting Method
Panel 1 (PI) 13 panelists (3 MT OF)
Validation Study Panel 2 (P2)
10 panelists (5 M6 F)
Validation Studv . Panel3 (P3)
13 panelists (3 MTOF)
Reproducibilrtv Studv Panel (P4)
13 panelists (3 MT OF)
Figure 4.1: Flow chart of studies and panels, categ sorting method. M=MaIes and F=FemaIes
PI Fixed Categories
PI Fixed
Categories
PI Filed Categories
2*0^^
P4 Fixed
V
Y
V
V
V
V
Msual
Msual-Oral
Msual
Msual-Oral
Msual
Msual-Oral
Visual
Mstnl-Oral
Visual
Msual-Oral
Msual
Msual-Oral
^ PI Free Sort
Generated Fixed Categories
^
P4 Free Sort Generated Fixed
Categories
ies, types of sorting task, and results of the conducted two-step sensor}'
12
Q 0
-4
-12
Group6
Group 1,
Group 5,
-12 -8 0
Dim1
12
Group 1
2
3
4
•'5
6
C^egory
Energy Drinks
Enhanced Waters
Nutritional Drinks
Smoothies
Sports Drinks
Teas
Characteristics
Contains stimulants such as caffeine and taurine; provides extra energy
Contains minimal Calories, lightly flavored, clear liquid Contains many nutrients could serve as a meal-replacement beverage
Contains fermented dairy products; opaque Mineral or vitamin enhanced; does not contain caffeine. Labeling targets athletes Contains Tea or Tea Extracts
Figure 4.2: Multidimensional Scaling Panel l 's visual free sort (Part 1) of 50 functional beverages plotted in two dimensions with stress = 0.265 and functional beverage categories generated through the free visual sorting method. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot Refer to Table 4.1 for product names and corresponding number codes.
12
-4
-8
-12
Group 5
-12 0
Dim1
12
Group
1
2
3
4
5
6
7
Category
Energy Drinks
Enhanced Waters
Fruit Smoothies
Nutritional Drinks
Sports Drinks
Teas
Yogurt Smoothies
Characteristics
Contains stimulants such as caffeine and taurine; provides extra energy
Contains minimal Calories, lightly
flavored, clear liquid
Contains 100% Fruit,; non-clear liquid; all natural
Contains many nutrients could serve as a meal-replacement beverage
Mineral or vitamin enhanced; does not contain caffeine. Labeling targets
Contains Tea or Tea Extracts
Contains fermented dairy products;
Figure 4.3: Multidimensional Scaling of Panel l 's visual-oral free sort (Part 1) of 50 functional beverages plotted in two dimensions with stress = 0.290 and functional beverage categories generated through the free visual-oral sorting method. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot. . Refer to Table 4.1 for product names and corresponding number codes.
Group 2
Group 1
Group4
o
Dim1
12
Group
1
2 .
3
4
5
6
7
Category
Energy Drinks
Enhanced Waters
Fruit Smoothies
Nutritional Drinks
Sports Drinks
Teas
Yogurt Smoothies
Characteristics
Contains stimulants such as caffeine and taurine; provides extra energy
Contains 100% Fruit,; non-clear liquid; all natural
Contains many nutrients could serve as a meal-replacement beverage
Mineral or vitamin enhanced; does not
contain caffeine. Labeling targets
Contains Tea or Tea Extracts
Contains fermented dairy products;
Figure 4.4: Multidimensional Scaling of Panel l 's visual fixed sort (Part 2) of 50 functional beverages plotted in two dimensions with stress = 0.289 and corresponding functional beverage categories. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot Refer to Table 4.1 for product names and corresponding number codes.
12
-4
-12
-12 0
Dim1
12
Group 1
2
3
4
5
6
7
Category
Energy Drinks
Enhanced Waters
Fruit Smoothies
Nutritional Drinks
Sports Drinks
Teas
Yogurt Smoothies
Characteristics
Contains stimulants such as caffeine and taurine; provides extra energy
Contains many nutrients could serve as a meal-replacement beverage
Mineral or vitamin enhanced; does not contain caffeine. Labeling targets
Contains Tea or Tea Extracts
Contains fermented dairy products;
Figure 4.5: Multidimensional Scaling of Panel l 's visual-oral fixed sort (Part 2) of 45 functional beverages plotted in two dimensions with stress = 0.283 and corresponding functional beverage categories. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot. Refer to Table 4.1 for product names and corresponding number codes.
12
-12
Group 5 Group 4
-12
j-JSroup 7
Group 3
o
Diml
Group
1
2
3
4
5
6
7
Category
Energy Drinks
Enhanced Waters
Fruit Smoothies
Nutritional Drinks
Sports Drinks
Teas
Yogurt Smoothies
Characteristics
Contains stimulants such as caffeine and taurine; provides extra energy
Contains many nutrients could serve as a meal-replacement beverage
Mineral or vitamin enhanced; does not contain caffeine. Labeling targets
Contains Tea or Tea Extracts
Contains fermented dairy products;
Figure 4.6: Multidimensional Scaling of Panel 2's visual fixed sort of 50 functional beverages plotted in two dimensions with stress = 0.301 and corresponding fixed functional beverage categories. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot. Refer to Table 4.1 for product names and corresponding number codes.
-12
-12 0
Diml
12
Group
1
2 .
3
4
5
6
7
Category
Energy Drinks
Enhanced Waters
Fruit Smoothies
Nutritional Drinks
Sports Drinks
Teas
Yogurt Smoothies
Characteristics
Contains stimulants such as caffeine and taurine; provides extra energy
Contains 100% Fruit,; non-clear liquid; all natural
Contains many nutrients could serve as a meal-replacement beverage
Mineral or vitamin enhanced; does not
contain caffeine. Labeling targets
Contains Tea or. Tea Extracts
Contains fermented dairy products;
Figure 4.7: Multidimensional Scaling of Panel 2's visual-oral fixed sort of 46 functional beverages plotted in two dimensions with stress = 0.271 and corresponding fixed functional beverage categories. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot Refer to Table 4.1 for product names and corresponding number codes.
a
-12
Group
1
2
3
4
5
6
7
Category
Energy Drinks
Enhanced Waters
Fruit Smoothies
Nutritional Drinks
Sports Drinks
Teas
Yogurt Smoothies
Characteristics
Contains stimulants such as caffeine and taurine; provides extra energy
Contains 100% Fruit,; non-clear liquid; all natural
Contains many nutrients could serve as a meal-replacement beverage
Mineral or vitamin enhanced; does not contain caffeine. Labeling targets
Contains Tea or Tea Extracts
Contains fermented dairy products;
-12 12
Diml
Figure 4.8: Multidimensional Scaling of Panel 3's visual fixed sort of 50 functional beverages plotted in two dimensions with stress = 0.241 and corresponding fixed functional beverage categories. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot. Refer to Table 4.1 for product names and corresponding number codes.
o 00
Group
1
2
3
4
5
6
7
Category
Energy Drinks
Enhanced Waters
Fruit Smoothies
Nutritional Drinks
Sports Drinks
Teas
Yogurt Smoothies
Characteristics
Contains stimulants such as caffeine and taurine; provides extra energy
Contains many nutrients could serve as a meal-replacement beverage
Mineral or vitamin enhanced; does not contain caffeine. Labeling targets
Contains Tea or Tea Extracts
Contains fermented dairy products;
Figure 4.9: Multidimensional Scaling of Panel 3's visual-oral fixed sort of 46 functional beverages plotted in two dimensions with stress = 0.284 and corresponding fixed functional beverage categories. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot Refer to Table 4.1 for product names and corresponding number codes.
12
-12
-12
Diml
12
Group
1
2
3
4
5
6
Category
Carbonated Energy Drinks
Flavored Waters
Nutritional Healthy Drink
Smoothies
Sports and Fitness Drinks
Teas
Characteristics
Drinks containing caffeine or other
stimulants and is bubbly; provides a sense of
restoring energy
Water drinks with added flavors or vitamins; tastes like sweetened flavored water; no stimulants added
Meal replacements which are related to health; contains a lot of added nutrients; has a milk-like texture and is thick Beverages made with yogurt and/or fruit; thick and creamy and has fruit flavors Marketed to refuel body, contains electrolytes; not too sweet; thirst-quenching, and very light
Tea-based drinks, and is labeled with "tea" or "iced tea"; the main flavor is tea
Figure 4.10: Multidimensional Scaling of Panel 4's visual free sort (Part 1) of 46 functional beverages plotted in two dimensions with stress = 0.260 and functional beverage categories generated through the free visual sorting method. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot. Refer to Table 4.1 for product names and corresponding number codes.
12
-12
-12 0
Diml
Group
1
2
3
4
5
6
Category
Carbonated Energy Drinks
Flavored Waters
Nutritional Healthy Drink
Smoothies
Sports and Fitness Drinks
Teas
Characteristics
Drinks containing caffeine or other stimulants and is bubbly; provides a sense of restoring energy
Water drinks with added flavors or vitamins;
tastes like sweetened flavored water; no
stimulants added
Meal replacements which are related to
health; contains a lot of added nutrients; has
a milk-like texture and is thick
Beverages made with yogurt and/or fruit;
thick and creamy and has fruit flavors
Marketed to refuel body, contains electrolytes; not too sweet; thirst-quenching, and very light
Tea-based drinks, and is labeled with "tea" or "iced tea"; the main flavor is tea
Figure 4.11: Multidimensional Scaling of Panel 4's visual-oral free sort (Part 1) of 46 functional beverages plotted in two dimensions with stress = 0.232 and functional beverage categories generated through the free visual-oral sorting method. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot Refer to Table 4.1 for product names and corresponding number codes.
Group
1
2
4
5
6
Category
Carbonated Energy Drinks
Flavored Waters
Nutritional Healthy Drink
Smoothies
Sports and Fitness Drinks
Teas
Characteristics
Drinks containing caffeine or other
stimulants and is bubbly; provides a sense of
restoring energy Water drinks with added flavors or vitamins; tastes like sweetened flavored water; no stimulants added
Meal replacements which are related to
health; contains a lot of added nutrients; has
a milk-like texture and is thick
Beverages made with yogurt and/or fruit;
thick and creamy and has fruit flavors
Marketed to refuel body, contains electrolytes; not too sweet; thirst-quenching, and very light
Tea-based drinks, and is labeled with "tea" or "iced tea"; the main flavor is tea
Figure 4.12: Multidimensional Scaling of Panel 4's visual fixed sort (Part 2) of 46 functional beverages plotted in tvvo dimensions with stress = 0.261 and corresponding fixed functional beverage categories. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot Refer to Table 4.1 for product names and corresponding number codes.
Group 1
2
3
4
5
6
Category
Carbonated Energy Drinks
Flavored Waters
Nutritional Healthy Drink
Smoothies
Sports and Fitness Drinks
Teas
Characteristics
Drinks containing caffeine or other stimulants and is bubbly; provides a sense of restoring energy Water drinks with added flavors or vitamins; tastes like sweetened flavored water; no stimulants added Meal replacements which are related to health; contains a lot of added nutrients; has a milk-like texture and is thick Beverages made with yogurt and/or fruit; thick and creamy and has fruit flavors Marketed to refuel body, contains electrolytes; not too sweet; thirst-quenching, and very light Tea-based drinks, and is labeled with "tea" or "iced tea"; the main flavor is tea
Figure 4.13: Multidimensional Scaling of Panel 4's visual-oral fixed sort (Part 2) of 46 functional beverages plotted in two dimensions with stress = 0.262 and corresponding fixed functional beverage categories. Functional beverages were grouped based on K-means clustering of beverage coordinate points on the MDS plot. Refer to Table 4.1 for product names and corresponding number codes.
CHAPTER 5 - SENSORY PROFILE OF A MODEL ENERGY DRINK WITH VARYING LEVELS OF FUNCTIONAL INGREDIENTS-CAFFEINE, GINSENG, AND TAURINE
5.1 Abstract
Energy drinks have increased in popularity in recent years due to the claimed
energy boost provided by functional ingredients. A multitude of functional ingredients
have been utilized; however, there is limited research on their sensory effects in energy
drink formulations. Descriptive analysis was conducted to investigate the effects on the
sensory properties of three common functional ingredients - caffeine, ginseng, and
taurine - in a non-carbonated model energy drink solution. Combinations of these
functional ingredients at three levels (low, medium, high) were added to create a total of
27 different solutions (3x3x3 factorial design). Analysis of variance was performed to
evaluate the sensory effects of the varying concentrations of functional ingredients in
solution. Principal component analysis (PCA) was performed to summarize the
relationship among the attributes and solutions. In general, high levels of caffeine in
solution resulted in low ratings of fruity attributes and high ratings of bitter attributes.
The high level of ginseng in solution was characterized by high ratings of bitter
attributes. A horns effect was observed as the sweet, artificial lemon-lime, pear, mango,
and pineapple attributes were rated lower in intensity with increased ginseng levels.
Taurine levels of up to 416 mg/100 mL had no significant effect on the sensory attribute
ratings of the model energy drink solutions. These findings can be utilized to predict the
changes in sensory characteristics when formulating energy drinks containing these
popular functional ingredients.
104
Key Words: Descriptive Analysis, Energy Drinks, Functional Ingredients, Caffeine,
Ginseng
5.2 Introduction
Energy Drinks
Energy drinks are one of the fastest growing segments of the functional beverage
market, with over 200 drinks introduced into the market between 2006 and 2007 (Reissig
and others 2009). They have gained popularity for the extra energy they provide via a
large concentration of stimulants. Energy drinks are an alternative to coffee as a source
of caffeine, and also contain other functional ingredients such as antioxidants, ginseng,
taurine, and B vitamins. In 2007, there were over $10.1 billion in functional beverage
sales in the US, and by 2010 functional beverage sales are projected to increase to over
$12 billion (Mintel 2008),
Functional Ingredients
Three of the most common functional ingredients in energy drinks are caffeine,
ginseng, and taurine. Caffeine is a methylxanthine with the chemical formula
C8H10N4O2. It is a white odorless powder with low solubility and is usually combined
with other chemicals, such as purines and pyrimidines, to increase its solubility (Spiller
1998). Caffeine is commonly found in cola products and has been incorporated into
snack foods, such as cereal bars and sunflower seeds (Cosgrove 2008). Caffeine is on the
US Food and Drug Administration (FDA)'s Generally Recognized as Safe (GRAS) list
and is limited to no more than 0.02% by volume in cola-type products (Food and Drug
Administration 2003). Currently, there are no regulations regarding the maximum
105
amount of caffeine allowed in energy drinks. The typical amount of caffeine in energy
drinks ranges from 21 to 112 mg/100 mL (wt/vol) (Table 5.1).
Ginseng is from the Araliaceae family and contains ginsenosides, which are active
steroid-like compounds (Spiller 1998). These active compounds of ginseng are
triterpenoid saponin glycosides, which also are responsible for the bitter taste of ginseng
(Court 2000a). Ginseng is known to have antioxidant properties (Jung and others 2002)
and may aid in alleviating some health conditions, such as diabetes and cognitive
function (Coon and Ernst 2002). Some research has been conducted on the efficacy of
consuming ginseng for increased energy, help with indigestion, and overall improvement
of health (Court 2000b). However, there has been limited research validating these
medicinal benefits attributed to ginseng (Kitts and Hu 2000, Vogler and others 1999).
Taurine is a derivative of the amino acid cysteine with the chemical formula of
C2M7NO3S, and is present in the tissues of humans and animals. It aids with bile acid
conjugation, retinal development, and central nervous system function (Lourenco and
Camilo 2002). Taurine has also been found to aid in immunity and may have antioxidant
properties (Yu and Kim 2009). Taurine is commonly incorporated in energy drinks and
in muscle-building supplements because of its suggested benefits such as improved
athletic performance and increased energy. Research suggests that higher levels of
taurine in muscle tissues may improve optimal exercise performance in rats (Yatabe and
others 2009). Taurine supplementation in humans, however, has not been shown to
improve exercise performance (Galloway and others 2008).
106
Sensory Analysis of Energy Drinks
Studies have been conducted on the acceptability of new functional beverages, but
limited research has been done on the effects of functional ingredients on the sensory
properties of model functional beverage solutions. Luckow and Delahunty (2004)
conducted research on the addition of probiotics and prebiotics in orange juice, while
Smit and Rogers (2002) added different levels of caffeine and vitamins to an energy drink
to determine the difference in preference with the addition of stimulants. They found that
the energy drinks containing higher concentrations of caffeine, vitamins, and stimulants
were not liked as much as the energy drinks containing lower concentrations of
functional ingredients. Panelists also preferred pure water over both energy drinks.
Another sensory test found that the concentration of 100 mg/L caffeine in a mixed
tropical fruit juice nectar had acceptable ratings (de Sousa and others 2007). While
Qimire (2000) determined that there was no significant difference between orange juice
with and without 600 mg of ginseng (20% ginsenosides) per liter, and ginseng
concentrations of 1000 mg/L of orange juice resulted in a medicinal taste.
Research related to the sensory effects of the addition of ingredients has suggested
that there is a synergistic effect of including multiple ingredients into a formulation.
Previous studies suggest that mixtures of tastants result in an increase in overall intensity
ratings of the compound mixture (Delwiche 2004). Therefore, the more functional
ingredients added to a beverage formulation, the more likely the tastes will be noticed.
The objective of this study was to investigate the effects on the sensory properties of
three common functional ingredients - caffeine, ginseng, and taurine - in a non-
carbonated model energy drink solution.
107
5.3 Materials and Methods
Model Energy Drink Formulation
The base model energy drink solution was composed of 1106.45 g spring water
(Absopure, Plymouth, MI), 285.00 g high fructose corn syrup (Isosweel 5500, Tate &
Lyle, Decatur, IL), 4.04 g sodium citrate (Tate & Lyle, Decatur, IL), 3.80 g citric acid
(Tate & Lyle, Decatur, IL), and 0.70 g potassium citrate (Tale & Lyle, Decatur, IL). The
model energy drink base solution was developed to have a Brix of 12.20°B and pH of
3.0, which fall in the range of commercially-available energy drink values. Non-
carbonated "still" solutions were used in this study to eliminate interference between the
carbonation and the actual changes due to the different levels of functional ingredients
tested in the model energy drink solutions.
Sample Preparation
Functional ingredients tested included caffeine (Fisher Scientific, Fair Lawn, NJ,
energy drink solution and mixed for another five minutes with a magnetic stir bar on a
stir plate.
The solutions were stored overnight in sealed wide-mouth glass mason jars .
(14400-67000 Ball®, Alltrista, Munice, IN) at ~5°C in a commercial grade refrigerator.
On the same day as evaluation, approximately 35 mL samples were poured into 73.9 mL
plastic souffle cups (Solo Cup Company, Urbana, IL) labeled with random 3-digit codes.
The samples were stored in the refrigerator until 10 minutes prior to evaluation.
Panelists Selection and Screening
The panelist recruitment and selection process included a questionnaire, a test for
6-n-propyl-2-thiouracil (PROP) status, and a basic tastes test (sour, sweet, bitter, salty).
The questionnaire asked volunteers about basic demographic information, allergies,
smoker status, frequency of functional beverage consumption, and schedule availability.
PROP taster status was determined by presenting volunteers pieces of filter paper
impregnated with PROP following Zhao and others (2003) paper disc method. If the
volunteers could not taste anything on the paper they were considered a non-taster. If the
volunteers could taste a bitter taste, they were labeled a taster.
The basic taste test consisted of presenting volunteers with 20 mL of basic taste
solutions in 59.2 mL plastic souffle" cups (Solo Cup Company, Urbana, IL), Basic taste
solutions labeled A through F (sweet, sour, bitter, waler, salty, and sour, respectively)
were presented to volunteers. The basic taste solutions tested included: 0.70% sucrose
(C&H Sugar Company, Inc. Crockett, CA) solution for the sweet solution, a 0.05% citric
acid (Tate & Lyle, Decatur, IL) solution for the sour solution, a 0.02% caffeine (Fisher
Scientific, Fair Lawn, NJ) solution for the bitter solution, and a 0.10%o sodium chloride
(Morton , Chicago, IL) solution for the salty solution. All solutions were prepared with
spring water (Absopure, Plymouth, MI). Two sour solutions were presented to minimize
the chance of blind guessing by the volunteers.
Thirteen panelists (4 males, 9 females, 18 to 50 years old) were selected based on
non-smoker and positive PROP taster status, and coirectly identifying two or more of the
basic taste solutions. Four panelists correctly identified all the solutions, three panelists
identified four of the six solutions, five panelists identified three of the six solutions, and
one panelist identified two of the six solutions. Panelists' frequency of energy drink
usage ranged from rarely to daily consumption.
Panelist Training
Panel training consisted of sixteen 1-hour sessions, which included evaluating
three complete replications of the 27 solution set (Table 5.2), Initial training sessions
were conducted at a round table setting under incandescent lighting. The first two days.
included an introduction to the descriptive analysis method to be used in the study and
familiarization with sample solutions. During the next two days, panelists generated
descriptor terms and term definitions for nine samples. For each term and definition
generated, panelists selected and refined a reference. References were chosen to reflect
the sensory attributes of the solutions.
Once the panel generated the terms, developed definitions, and selected
references, the terms were narrowed to the thirteen terms that best represented the
sensory attributes of the solutions. The thirteen terms included: artificial lemon-lime
flavor, citrus, mango, pineapple, pear, sweet, tart, bitter tea, fruit bitter, astringent, bitter
tea afterfeel, fruit bitter afterfeel, and moutheoating (Tabic 5.3). Panelists had a difficult
110
time pinpointing the bitter attributes perceived in the samples. To determine the specific
bitter taste perceived in the samples, bitter references presented to the panelists included a
0.25% caffeine solution, lemon seeds, brewed black tea solution (100 mL brewed tea and
200 mL water), a 0,16% PROP solution, a 0.44% naringin solution, and quinine solutions
(0.013%, 0.067%, 0.05%, 0.097%). Panelists agreed upon the brewed black lea solution
and the 0.44% naringin solution as the references that best matched the bitter tastes
detected in the solutions.
The references for each term were then rated on a 16-point categorical scale (0 to
15) to generate anchors for each attribute. Panelists rated the solutions for each attribute
against group-determined reference anchors. The rinse protocol determined by the
panelists was a warm water (~40°C) rinse followed by a cold water (~20°C) rinse.
Panelists were instructed to follow the rinsing protocol prior to evaluating the first sample
and between samples.
The sampling protocol consisted of sipping one-third of the sample (~12 mL) and
moving it to contact all sides of the tongue and mouth for about 5 seconds before rating
the attributes. The first third of the sample was used to evaluate aroma-by-mouth
attributes, the second third to evaluate taste, and the last third to evaluate mouthfeel and
afterfeel.
Three practice sessions and six data collection sessions were conducted in
individual sensory booths under red light to mask the slight color difference among
solutions. The color difference was due to the different levels of ginseng in each
solution. Each session consisted of monadically presenting nine samples; five samples
then a two-minute break, followed by the four remaining samples.
I l l
References were made 20 to 24 hours in advance and stored in lidded 73.9 mL
plastic souffle cups (Solo Cup Company, Urbana, IL, 61802) at ~5°C in a commercial
grade refrigerator. Panelists familiarized themselves with the references and reference
intensity scores prior to each session. Panelists then entered the individualized booths to
evaluate the samples. The data were collected by the Compusense//ve version 4.2
(Compusense, Inc. Guelph, Ontario, Canada) program. A modified Williams design
(1950) was used to randomize samples among the panelists to balance out first order
carry over effects (Macfie and others 2007).
Flow Behavior and Viscosity Measurements
The flow behavior and viscosities of the solutions were measured using the ARES
RFS III Rheometer (TA Instruments, New Castle, DE) following the method outlined by
Kappes and others (2006), 1,11 mL of solution was placed between two parallel plates
and a shear rate sweep program was run to measure the flow behavior of the solutions.
The shear rate sweep program was run in log mode in a clockwise direction from a rate of
0.6 to 200 s'1. Solutions were measured at 5°C in triplicate and measured on the same
day panelists evaluated them. All solutions exhibited Newtonian behavior, and viscosity
was calculated by the slope of the linear plot of shear rate (s"1) by shear stress (dyn/cm"2).
Data Analyses
Descriptive analysis and viscosity measurement data were analyzed using the
Statistical Analysis Software (SAS) version 9.1 (SAS Institute, Inc., Cary, NC). Analysis
of Variance (ANOVA) tested for significant difference of mean scores of the solutions,
panelists, and interactions for each attribute. ANOVA was also clone to determine the
effects of the three levels of the different functional ingredients. Fisher's Least
112
Significant Difference (LSD) was conducted to determine the difference among sample
means and by levels of functional ingredients. Principal component analysis (PCA) on
the covariance data matrix with varimax rotation was done using XLStat 2008,4,2
(Addinsoft, New York, NY). Agglomerative hierarchical clustering (AHC) by Ward's
method (1963) was done on the sensory data to observe groupings among the solutions
using XLStat 7.5.3 (Addinsoft, New York, NY).
5.4 Results and Discussion
The ANOVA results of the model energy drink solutions (Table 5.4) show that for
all attributes, except citrus and moutheoating, the panelist and sample factors were highly
significant (pO.001). There was a significant difference in panelist ratings, which
indicate that panelists used different parts of the scale when rating samples, which is
typically found in descriptive analysis panels. For all attributes, except moutheoating, the
interaction of panelists by sample was also highly significant. Therefore, an adjusted F-
value was calculated to account for the variability of the interaction between panelists
and samples as a source of error. Eleven of the thirteen attributes were still significantly
different across samples after the adjusted F-tesl.
An explanation for the lack of difference in moutheoating rating among the
samples is that each solution contained the same amount of high fructose corn syrup
(MFCS) with a concentration of 20% (wt/vol), which was the dominant ingredient
contributing to the moutheoating of the solutions. Results from Kappes and others (2006)
suggested that the possible mouthfeel detection threshold of MFCS in water falls in the
range of 2.81% to 20.15% (wt/vol). The consistent moutheoating ratings of all the model
113
energy drink solutions suggest that panelists could not detect a moutheoating difference
by the varying levels of functional ingredients.
The mean measured viscosity of the solutions was 2.35±0.06 mPa*s with a range
of 2.24 to 2.51 mPa*s, which was not significantly correlated to any attributes. This
suggests that viscosity measurements were not affected by the levels of functional
ingredients added to the model energy drink, although significant sensory attribute
differences were perceived.
The comparison of mean intensity ratings of the three taurine levels used in our
study did not exhibit a significant difference in ratings of any of the attributes (Table 5.5),
with levels of up to 416 mg/100 mL taurine in solution having no significant effects on
sensory properties. The comparison of mean intensity ratings showed that solutions
containing high levels of caffeine had low ratings for fruity attributes and high ratings for
bitter attributes (Table 5.5). High ginseng and high caffeine levels both increased the
intensity ratings of bitter attributes. It is known that caffeine and ginseng arc two
functional ingredients which have negative sensory characteristics that are difficult to
mask (Backas 2009), which is supported by our findings.
An interesting observation was the horns effect that ginseng and caffeine levels
had on the sweet and fruity attributes (pear, artificial lemon-lime, mango, and pineapple).
The greater the level of ginseng and caffeine in solution, the lower the ratings for the fruit
attributes. Caffeine and ginseng levels may have had a horns effect on the sweet and
fruity attributes due to the bitterness it imparted on the solutions. Results from other
studies have shown that increasing amounts of caffeine reduced the sweetness intensity of
solutions (Pangborn 1960, Calvino and others 1990). Ginseng levels were seen to have a
114
more prominent effect than caffeine levels on the intensity ratings of bitter attributes
(Figures 5.1 and 5.2). A possible explanation for the horns effect is that ginseng is not a
familiar taste for the western consumer, therefore the unfamiliar taste of ginseng may
have played a role in decreasing the perception of the fruit flavors. When unfamiliar
lastes are added to a product, it increases the intensity rating of the taste (Kang and others
2007, Labbe and others 2006). The inclusion of ginseng into the formulation was not
expected with the fruity flavors and may have contributed to the decrease in fruit and
sweet attribute intensity ratings.
The bitter attributes were highly correlated (Table 6.1). The highest correlation
was seen between the fruit bitter and fruit bitter afterfeel attributes (r=0.97, p<0.05) and
the tea bitter and tea bitter afterfeel attributes (r=0.98, p<0.05). The highest negative
correlation was between the sweet and tea bitter and fruit bitter (r=-0.91, p<0.05). The
bitter taste and afterfeel attributes were all highly positively correlated (r=0.95, p<0.05).
This suggests that the bitter attributes might have been measuring the same bitter
perception in the solutions, although in panelist training and term generation, the
panelists determined that there were four distinct bitter attributes identified in the model
energy drink solutions.
Bitter tea has a complex flavor containing both astringent and bitter sensations. In
Drobna's (2004) research, the descriptors for bitter tea included alum for bitterness and
tannic acid for astringency. The main compounds which contribute to the taste and
bitterness in tea are catechins, caffeine, and saponins (Rouseff 1990). Caffeine and
ginseng saponins were present in the energy drink solution and could be the reason
panelists selected the tea reference. During the term generation and reference refinement
115
sessions, a 0.25% caffeine solution was tested as a possible bitter reference. The panel as
a whole did not feel that the bitterness of the caffeine solution could be detected, but
instead the bitterness of brewed black tea was detected in the solutions.
Although there were no fruit flavorings added to the model energy drink
solutions, panelists detected fruity notes in the model energy drink solutions. Fruits are
generally sweet and associated with sweet flavors, which could explain why the pear,
pineapple, and mango attributes were selected. King and others (2007) found that pear
attribute was perceived to be higher in an apple-flavored beverage with a Brix level of
12°B versus an apple-flavored beverage with a Brix level 8°B. The Brix level in the base
model energy drink solution was 12.20°B, which could have contributed to the panelists'
perception of the presence of fruity attributes in the energy drink solutions. Lemon flavor
was found to increase in a beverage solution when the acidity of the solution was
increased (King and others 2007). The citric acid in the base model energy drink
formulation could explain the artificial lemon-lime attributes identified in the solutions.
The low, medium, and high levels of caffeine, ginseng, and taurine provide a
representation of the range of these ingredients in commercially-available energy drinks.
To the researchers' knowledge, there are currently no products on the market which
contain the lowest levels of all of the ingredients or the highest levels of all of the
ingredients. If a product contained high amounts of ginseng (i.e. Red Jak), there was a
moderate level of caffeine in the beverage. Therefore, current commercially-available
beverages may not be as bitter as the model solutions that were studied in this
experiment.
116
Principal component analysis (PCA) on the covariance matrix with varimax
rotation described 65.6% of the variance on Factor 1 and 12.4% on Factor 2 (Figure 5.1).
Factor 1 was defined by the astringent, bitter, and fruity attributes. The astringent, bitter,
and bitter afterfeel attributes were all highly positively correlated, while all negatively
correlated to the artificial lemon-lime, pear, pineapple, and sweet attributes (Table 5.6
and Figure 5.1). Sweet and bitter attributes have been found to be negatively correlated
in Calvino's (1990) research on solutions. Also in Keast's (2008) work, an increased
amount of caffeine in solution decreased the sweetness ratings. The mango and tart
attributes defined Factor 2, which accounted for 12.4% of the variation of data. Factor 2,
however, does show that solutions with lower levels of caffeine and ginseng are located
in the opposite area of the tart attribute, suggesting that there was a higher perception of
tart at higher caffeine and ginseng levels.
Cluster analysis was conducted based on the significant attributes of the 27
solutions, which were clustered into four groups, The clusters were generally
characterized by the different levels of ginseng in solution (Figure 5.2). This suggests
that the high bitter attribute ratings and the low fruit attribute ratings had the most
prominent effect in clustering of solutions, which was mainly caused by the ginseng
level. These findings can be utilized to predict the changes in sensory characteristics
when formulating energy drinks containing these specific functional ingredients.
117
5.5 Conclusions
The more caffeine and ginseng were added to solution, the higher the bitter
attribute ratings of the model energy drink solutions. Determining ways of minimizing
the bitterness in functional beverages will allow manufacturers to produce products that
have more health benefits as well as lower level of objectionable sensory properties.
There was no significant difference in sensory attribute ratings across the taurine levels
(208 to 416 mg/100 mL) added to the model energy drink solutions. The findings from
this study can be used when selecting palatable amounts of caffeine, ginseng, and taurine
to incorporate in an energy drink formulation. The findings can also be utilized to predict
the changes in sensory characteristics when reformulating functional ingredients in
energy drinks.
Future studies may include: 1) determining methods to effectively minimize
ginseng bitterness in energy drinks and 2) identifying acceptable bitterness levels of
energy drinks via a consumer test, which will aid in determining optimal functional
ingredient levels and bitterness masking agents to incorporate into energy drink
formulations.
5.6 References
Backas N. 2009. Brimming Opportunities for Nutraceutical Beverages, Food Product Design [serial online], 19(2):50-68. Available from Posted January 21,2009.
Calvino AM, Garcia-Medina MR, Comelto-Muniz JE. 1990. Interactions in caffeine-sucrose and coffee-sucrose mixtures: evidence of taste and flavor suppression. Chem Senses 15(5):505-19.
Camire ME. 2000. Dietary Supplements. In: M. K. Schmidl, T. P. Labuza, editors. Essentials of Functional Foods. Gaithersburg: Aspen Publishers, Inc. pi 65-180.
Coon JT, Ernst E. 2002. Panax ginseng: a systematic review of adverse effects and drug interactions. Drug Saf 25(5):323-44.
118
Cosgrove J. 2008. Caffeinated Snacks. Nutraceuticals World [serial online], July/August 2008Available from Posted July 2008.
Court WE. 2000a. Ginseng: The Genus Panax. Singapore: CRC Press. 266 p.
Court WE. 2000b. The Pharmacology and Therapeutics of Ginseng. In: Anonymous Ginseng: The Genus Panax. Singapore: Harwood Academic Publishers, pi 17-197.
de Sousa PI-IM, Maia GA, de Azeredo HMC, de Souza Filho MSM, Gamiti DS, de Freitas CAS. 2007. Mixed tropical fruit nectars with added energy components. Int J Food Sci Tech 42(11): 1290-1296.
Delwiche J. 2004. The impact of perceptual interactions on perceived flavor. Food Qual Pref 15(2): 137-46.
Drobna Z, Wismer WV, Goonewardene LA. 2004. Selection of an Astringcncy Reference Standard for the Sensory Evaluation of Black Tea. J Sens Stud 19(2): 119-32.
Food and Drug Administration. 2003. Code of Federal Regulations Title 2 Sec. 182.1180.
Galloway DR, Talanian JL, Shoveller AK, Heigenhauser GJF, Sprict LL, 2008. Seven days of oral taurine supplementation does not increase muscle taurine or alter substrate metabolism during prolonged exercise in humans. J Appl Physiol 105(2):643-51.
Jung MY, Jeon BS, Bock JY, 2002. Free, esterified, and insoluble-bound phenolic acids in white and red Korean ginsengs (Panax ginseng CA Meyer). Food Chem 79(1): 105-11.
Kang MW, Chung SJ, Lee I-IS, Kim Y, Kim KO. 2007. The sensory interactions of organic acids and various flavors in ramen soup systems. J Food Sci 72(9):S639-47.
Kappes SM, Schmidt SJ, Lee SY. 2006, Color halo/horns and halo-attribute clumping effects within descriptive analysis of carbonated beverages. J Food Sci 71(8):S590-5.
Keast RSJ. 2008. Modification of the bitterness of caffeine. Food Qual Pref 19(5):465-72.
King BM, Duineveld CAA, Arents P, Meyners M, Schroff SI, Soekhai ST. 2007. Retronasal odor dependence on tastants in profiling studies of beverages, Food Qual Pref 18(2):286-95.
Kitts D, I-Iu C. 2000. Efficacy and safety of ginseng. Public Health Nutr 3(4A):473-85.
Labbe D, Damevin L, Vaccher C, Morgenegg C, Martin N. 2006. Modulation of perceived taste by olfaction in familiar and unfamiliar beverages. Food Qual Pref 17(7-8):582-9.
119
Lourenco R, Camilo ME. 2002. Taurine: a conditionally essential amino acid in humans? An overview in health and disease. Nutr Hosp 17(6):262-70.
Luckow T, Delahunty C. 2004. Consumer acceptance of orange juice containing functional ingredients. Food Res Int 37(8):805-14.
Macfie HJ, Bratchell N, Greenhoff K, Vallis LV. 2007. Designs to Balance the Effect of Order of Presentation and First-Order Carryover Effects in Hall Tests. J Sens Stud 4(2):129-148.
Mintel. 2008. Functional Beverages-US August 2008. Mintel Reports.
Reissig CJ, Strain EC, Griffiths RR. 2009. Caffeinated energy drinks—A growing problem. Drug Alcohol Depend 9(1): 1-10.
Rouseff RL. 1990. Bitterness in Food Products and Beverages. Amsterdam: Elsevier. 356 P-
Smit HJ, Rogers PJ. 2002, Effects of'energy' drinks on mood and mental performance: critical methodology. Food Qual Pref 13(5);317-26.
Spiller MA. 1998. Caffeine. Boca Raton: CRC Press. 363 p.
Vogler BK, Pittler MH, Ernst E. 1999. The efficacy of ginseng. A systematic review of randomized clinical trials. Eur J Clin Pharmacol 55(8):567-75,
Ward JH. 1963. Hierarchical grouping to optimize a quantitative function. J Am Stat Assoc 58(301):236-44.
Williams EJ. 1950, Experimental designs balanced for pairs of residual effects. Australian Journal of Scientific Research 3(3):351-363.
Yatabe Y, Miyakawa S, Ohmori H, Mishima H, Adachi T. 2009. Effects of Taurine Administration on Exercise. In: Anonymous Advances in Experimental Medicine and Biology: Taurine 7. New York: Springer New York. p245-252,
Yu J, Kim AK. 2009. Effect of Taurine on Antioxidant Enzyme System in B16F10 Melanoma Cells. In: Anonymous Taurine 7. New York: Springer. p491.
Zhao L, Kirkmeyer SV, Tepper BJ. 2003. A paper screening test to assess genetic taste sensitivity to 6-n-propylthiouracil. Physiol Behav 78(4-5):625-33.
120
5.7 Tables and Figures
Tabic 5.1: Amount of functional ingredients listed on Nutritional Facts labels of a sampling of popular commereia
Product Name Cocaine Full Throttle IMDX No Fear NOS Red Bull Red Jak Rockstnr Sobe Adrcnnliiic Rush Tab Energy // denotes that ingredient
The aroma of artificial lemon-lime soda while in the mouth The aroma of diluted Gatorade lemon-lime drink while in the mouth The aroma of diluted mango juice while in the mouth
The aroma of diluted pineapple juice while in the mouth. The aroma of diluted 100% canned pear juice.
Taste of 7% Fructose Solution
The tart and sourness of passion fruit juice.
Bitter Tea is the taste of unsweetened Black Tea while in the mouth
Fruit Bitter is the taste of Naringin Solution while in the mouth
Bitter Tea is the taste of unsweetened Black Tea after it is expectorated.
Fruit Bitter Afterfeel is the taste of Naringin Solution after it is expectorated. The moutheoating of sensation perceived on the teeth, tongue, and sides of the mouth after expectorating the samples
Astringent is a mouthdrying sensation felt on the tongue and sides of the mouth
Decarbonated Sierra Mist lemon-lime soda Diluted Lemon-lime Gatorade
Canned Mango Nectar
Diluted Dole Canned 100% Pineapple Juice Diluted 100% Pear Juice from Canned Pear
Fructose Solution
Passion Fruit Juice
Lipton Black Tea
Naringin Solution
Black Tea
Naringin Solution
HFCS and water solution
Diluted Pomegranate Juice
Decarbonated Sierra Mist
125 mL lemon-lime Gatorade + 125 mL water 200 mL juice+50 mL water
125 mL juice+125 mL water
150 mLjuice+100 mL water
21 g fructose + 300 mL water
125 mL juice+125 mL water
1 Lipton tea bag+100 mL hot water; steeped 5 min. then add 200 mL cold water 0.11 g Naringin + 250 mL water
1 Lipton tea bag+100 mL hot water; steeped 5 min. then add 200mL cold water 0.1 lg Naringin + 250 mL water
30 g Isosweet 55 +300 mL water
8
10
8
10
10
9
10
9
10
9
10
100 mL juice +150 mL water
Table 5.4: Analysis of Variance on 13 descriptive attributes rated for model energy drink solutions. Adjusted F-values with judge x sample interaction as the error term.
Attribute Aroma-by-Mouth Artificial Lemon-Lime Citrus Mango Pineapple Pear Taste Sweet Tart Bitter Tea Fruit Bitter Mouthfeel / Afterfeel Bitter Tea Afterfeel Fruit Bitter Afterfeel Moutheoating Astringent
Replication
4.27 2.26 17.56"' 17.70*" 35.35'"
1.12 37.02'" 1.35 1.31
1.73 1.86 1.08 25.58"*
Panelist
33.87'" 43.83'" 77.56"' 57.48*" 100.49*"
67.60"' 75.87"* 51.90"* 40.31"*
71.16"* 63.83"' 91.97"* 87.13"*
Samples (Solutions)
6.92 1.42 5.04*" 3.45"' 5.21"*
11.99"* 3.42"* 16.46*** 21.68*"
22.83"* 28.73"* 1.08 4.60*"
RxP
0.65 1.76 6.8*" 5.28"' 5.40*"
2.2 5.57*" 5.48"* 2.87*"
2.84*" • 3.33*" 2.83"* 6.43"*
RxS
1.36 0.91 1.53* 1.37 1.66'
1.89 1.33 2.04 2.47'"
2.10" 3.01*" 1.3 1.79*
PxS
1.38" 1.40" 1.47"* 1.57"* 1.28*
1.56*" 1.41" 2.49*" 1.95"*
2.16"* 1.80*" 0.94 1.25*
Adjusted F
5.01*** 1.01 3.43*" 2.2*** 4.07***
7.69*** 2.43"* 6.61"* 11.12*"
10.57"* 15.96*"
3.68*" Statistical significance at p<0.05, p<0.01 and RxS=Replication by Sample Interaction; P*S
p<0.001 are denoted by , , and =Panelist by Sample Interaction.
, respectively.RxP= Replication by panelists interaction;
Table 5.5: Mean intensity scores of sensory attributes of varying levels of functional ingredients.
Attribute Sweet
Artificial lemon-
lime Citrus Mango Pineapple Pear Tart
Bitter Bitter Tea Tea Afterfeel
Fruit Fruit Bitter Mouth-Bitter Afterfeel coating Astringent
CAFFEINE LEVEL
High
Medium Low
GINSENG LEVEL High
Medium Low
TAURINE LEVEL High
Medium Low
6.2a
7.0b
7.6C
6.3a
6.9b
l.T
7.0a
6.9a
7.0*
5.3a
5.8b
6.4C
5.6*
5.8a
6.1"
5.9a
5.9a
5.8a
5.1a
5.1a
5.3b
5.r 5.2a
5.3a
5.0a
5.2a
5.3a
3.0a
3.4b
3.7C
3.1a
3.3b
3.7C
3.4a
3.4s
3.3a
3.6a
3.8b
4.1c
3.6a
3.8b
4.1e
3.8a
3.9a
3.9a
3.8a
4.0b
4.3C
3.7a
4.1b
4.4C
5.2a
5.1ab
4.9b
5.4a
5.1b
4.7C
5.0a
4.3b
3.9C
5.3a
4.5b
3.4C
5.0a
4.2b
3.9b.
5.6a
4.5b
3.0C
4.4a
3.5b
3.0C
4.6a
3.6b
2.6C
4.6a
3.6b
3.3C
5.3a
3.8b
2.4°
6.9a
7.0a
7.0a
6.9a
6.9ab
7.1b
4.4a
4.2b
4.2b
4.7a
4.3b
3.7C
4.2b
4.1 ab
5.0a
5.0a
5.2b
4.4a
4.3a
4.5a
4.3a
4.4a
4.4a
3.5a
3.7a
3.7a
3.7a
3.9a
3.9a
6.9a
6.9a
7.0a
4.2' 4.2!
4.4=
Means within a column per treatment (caffeine, ginseng, and taurine levels) with the same superscript are not significantly different (p<0.05, Fisher's Least Significant Difference Test).
Table 5.6: Correlation analysis on significant sensory attributes for 27 combinations of functional ingredients in model energy drink solutions. Bolded values are significant at p<0.05.
<D _ 3 ,— U
. 3 O rO — ,o V-
o ° a
~ - o •- « o u n •£ .
Sweet Artificial Lemon-Lime
Citrus
Mango
Pineapple
Pear
Tart
Tea Bitter
Bitter Tea Afterfeel
Fruit Bitter
Fruit Bitter Afterfeel
Moutheoating
Astringent Viscosity
1.00
0.91
0.47
0.10
0.86
0.83
-0.28
-0.91
-0.90
-0.91
-0.88
0.39
-0.77 -0.11
.5 "o
<
1.00
0.50
0.30
0.81
0.79
-0.23
-0.77
-0.74
-0.79
-0.74
0.44
-0.60 -0.05
—
u
1.00
0.13
0.56
0.50
-0.02
-0.40
-0.40
-0.40
-0.33
0.27
-0.19 -0.08
o
a
1.00
0.30
0.18
0.22
0.04
0.09
-0.03
0.05
0.17
0.17 -0.20
a. a o
i£
1.00
0.90
-0.07
-0.81
-0.81
-0.84
-0.82
0.35
-0.69 0.01
5 cu
1.00
-0.28
-0.80
-0.82
-0.83
-0.81
0.22
-0.68 0.04
E-
1.00
0.28
0.30
0.25
0.24
-0.20
0.31 0.15
5 E-
•
1.00
0.98
0.96
0.96
-0.42
0.90 0.00
o -4~"
3
1.00
0.97
0.97
-0.38
0.93 -0.01
5
1.00
0.98
-0.36
0.88 0.09
5
1.00
-0.35
0.92 0.02
o o 3 O
2
1.00
-0.28 -0.06
o CO
*n: <
1.00 -0.08
"co O O to
>
1.00
--Factor 1 (65.6%)-->
Figure 5.1: Principal component analysis biplot of covariancc matrix of mean scnsorj' attributes of 27 combinations of functional ingredients in model cnergj' drink solntions with varimax rotation. Factor 1 represents 65.6% of the variation and Factor 2 represents 12.4% of the variation. L=Low, M=Medium, II=High, C=Caffcinc, G=Ginscng, and T=Taurinc.
127
1/4
145-
116"
87"
58
29"
n J
1 2
t— i 1
3
•
i — ? = - , ~ J U I - J - I I - J - I
4
i | , ' , •
- i 1 , - 1 . 1 , r-J—i .
5 5 5 CD CD CD
I O
CD
"̂ _J — I T T T —1 ^ —I -^ —J "̂ ^ T T t= — I T
g CD C5 § $ $ CD _i - 1 ^ I X 2 ^
O I CD
O _ i CD I
2? _ 2 X X o o o o I o °, O O
§° t> i §o cj %i i I- \_J CD
2 ; 2 X
r- H E- b
Q o g g I '«, CD CD
-J 2
to co
Figure 5.2: Agglomerative hierarchical clustering (AHC) of attribute ratings for 27 combinations of functional ingredients in model energj' drink solutions on the dissimilarity scale by Euclidean distance and agglomeration by Ward's method. The dotted line was computed using the software and truncates the groups based on the largest relative increase in dissimilarity. L=Low, M=Medium, H=High, C=Caffeine, G=Ginseng, and T=Taurine
CHAPTER 6: SENSORY PROPERTIES OF GINSENG SOLUTIONS MODIFIED BY MASKING AGENTS
6.1 Abstract
Ginseng is one of the most popular functional ingredients found in energy drink
formulations. Although ginseng is Icnown for its health benefits, ginseng is also
notorious for imparting a bitter taste. Incorporating ginseng into beverages without the
bitterness, while still maintaining its health benefits, is necessary for developing an
acceptable product. Thus, the objectives of this study were to: 1) identify effective
treatments for minimizing the bitterness of ginseng in water base and model energy drink
base solutions and 2) determine the sensory effects of incorporating different treatment
levels to minimize the bitterness of ginseng.
Based on the results of a series of pilot studies, which investigated bitterness
reducing treatments including congruent flavor addition, bitterness blocking agent
incorporation, enzymatic modification, ingredient interaction, and complexation, y-
cyclodextrins (y-CDs) and |3-cyclodexlrins (P-CDs) complexing agents were identified as
having the most potential. Descriptive analysis was conducted on the effects of the
inclusion of y-CDs, P-CDs and combinations of y-CDs and P-CDs in solutions containing
0,052 g 80% ginsenosides panax ginseng in 100 mL water and in 100 mL model energy
Ustundag and Mazza 2007). Although the ginsenosides-have smaller molecular weights
than both y- and P-CDs, the different configuration of the ginsenosides could be the
reason why some may have been less effective in forming a complex with the P-CDs, yet
could complex and fit in the cavity of the larger y-CD molecule.
The chemical interactions between CDs and the compounds in solution, and the
sensory effects of these interactions must also be considered when formulating beverages.
For example, research should be conducted investigating whether the added CDs
complex with only the bitter compounds and not the flavor compounds. Flavorings are
143
often the most expensive ingredient included in a functional beverage formulation.
Understanding the chemistry of the interactions between compounds and CDs will aid in
the selection of the most effective type and amounts of CDs to incorporate into
formulations.
Cluster A nalysis
Cluster analysis was conducted based on the two rated attributes, quinine bitter
and caffeine bitter aftertaste, of the 21 solutions, which were truncated into three groups.
Clusters were generally characterized by levels of both y- and P-CDs (Figure 6.3).
Solutions containing none or the lowest tested level of y-CDs (0.030 g/100 mL) and P~
CDs (0.250 g/100 mL) were clustered together and were rated high in quinine bitter
intensity. A second cluster grouped ginseng solutions incorporating y-CD levels ranging
from 0.030 to 0.060 g and p-CD levels ranging from 0.250 to 0.500 g. All other solution
treatments were clustered together; possibly due to the higher levels of P-CD and y-CD
present in these treatments. This suggests that ginseng solutions containing more than
0.750 g p-CDs/100 mL had similar quinine bitter and caffeine bitter aftertaste intensity
ratings as treatments containing more than 0.090 g y-CDs/100 mL of solution. The
clustering results also suggest that the Combo 7 solution treatment (0.030 y-CDs g/100
mL and 0.875 P-CDs g/100 mL) had similar quinine bitter and caffeine bitter aftertaste
intensity ratings as both the 0.75 g p-CDs treatment and the 0.090 g y-CDs treatment.
These findings may be utilized to formulate ginseng-containing energy drinks with
minimal bitterness.
144
Cost Analysis
Although p-CDs cost five-fold less than y-CDs by weight, less y-CDs are required
to comparably reduce the bitterness of ginseng in solution. y-CDs in the amount of 0.090
g in 100 mL of solution is the most cost effective level of y-CDs treatment at $0.08/L of
solution. The most cost effective level of p-CDs was 1.00 g, which would cost $0.17/L
of solution. This is the level needed to minimize the quinine bitterness to the same
intensity as using 0.090 g y-CDs. It also costs $0.17/L of solution to utilize Combo 7
solution treatment (0.030 y-CDs g/100 mL and 0.875 P-CDs g/100 mL) to reduce the
quinine bitter and caffeine bitter aftertaste attribute intensity levels comparable to the
intensity as the 0.090 g y-CDs treatment. Knowing the acceptable level of bitterness in a
functional beverage would aid in optimizing the amount of cyclodextrins necessary to
produce a less bitter tasting and more appealing product. This can be done by conducting
a consumer acceptability test on a set of model energy drink solutions containing varying
levels of CDs.
6.5 Conclusions
Energy drinks have many functional ingredients incorporated in the formulation
to provide energy and other benefits. These drinks often have medicinal tastes, due to
functional ingredients, such as the bitter perceptions of ginseng, Treatments to minimize
the bitterness of ginseng in solution are important in the creation of more palatable and
acceptable products. The challenge of incorporating ginseng into food products is
retaining the healthful properties of ginseng, while minimizing the bitter tastes.
145
The results of this study suggest that 0.090 g y-CDs significantly reduces the
bitter taste and aftertaste of ginseng in 0.052 g ginseng/100 mL solution. Consumers,
however, may accept higher levels of bitterness intensity in energy drinks through
familiarity of the taste associated with these drinks, and therefore, less bitterness
minimizers may be necessary in formulations. Future research should include conducting
a consumer acceptance test on a ginseng model energy drink base solution incorporating
the different types and levels of CDs to determine the bitterness level that is acceptable to
consumers. This research will be useful in selecting the minimum amount of CDs
necessary to produce an acceptable energy drink.-
Other future studies include investigating the chemistry behind the complexations
occurring between the bitter compounds in ginseng and CDs. The study could be
directed at studying the size and other physical properties of the bitter compounds.
Research focusing on the chemical interactions between CDs and bitter compounds can
then be compared and correlated to sensory studies on bitterness perception.
6.6 References
Binello A, Cravotto G, Nano GM, Spagliardi P. 2004. Synthesis of chitosan-cyclodextrin adducls and evaluation of their bitter-masking properties. Flavour Fragrance J 19(5):394-400.
Breslin PAS. 1996, Interactions among salty, sour and bitter compounds. Trends Food SciTechnol7(12):390-9.
Coon JT, Ernst E. 2002. Panax ginseng: a systematic review of adverse effects and drug interactions. Drug Saf 25(5);323-44.
Court WE. 2000a. Ginseng: The Genus Panax. Singapore: CRC Press. 266 p.
Court WE. 2000b. The Pharmacology and Therapeutics of Ginseng. In: Anonymous Ginseng: The Genus Panax. Singapore: Harwood Academic Publishers, pi 17-197.
146
Cravotto G, Binello A, Baranelli E, Carraro P, Trotta F. 2006. Cyclodextrins as Food • Additives and in Food Processing. Current Nutrition & Food Science 2(4):343-50.
Dodziuk I-I. 2006. Cyclodextrins and their complexes: chemistry, analytical methods, applications. Wcinheim: Wiley-VCH. 489 p.
Granato I-I. 2002. Masking Agents Maximize Functional Foods' Potential. Natural Products Insider [serial online]. Available from Posted 14 January 2002 2002.
Giiclu-Usliindag 6, Mazza G. 2007. Saponins: Properties, Applications and Processing. Crit Rev Food Sci Nutr 47(3):231-58.
Hamilton RM, Heady RE, inventors; 1970. Eliminating Undesirable Taste From Coffee And Tea Extracts And Products. U.S. patent United States Patent 3528819.
Huang KC. 1999. The Pharmacology of Chinese Herbs. Boca Raton: CRC Press. 512 p.
Kappes SM, Schmidt SJ, Lee SY. 2006. Color halo/horns and halo-attribute dumping effects within descriptive analysis of carbonated beverages. J Food Sci 71(8):S590-5.
Katan MB, Roos NM. 2004. Promises and Problems of Functional Foods. Crit Rev Food Sci Nutr 44(5):369-77.
Konno A, Misaki M, Toda J, Wada T, Yasumatsu K. 1982. Bitterness Reduction of Naringin and Limonin by P-Cyclodextrin. Agric Biol Chem 46(9):2203-8.
LeClair K. 2000. Breaking the Sensory Barrier for Functional Foods. Food Product Design [serial online]. 6 November 2006. Available from http://www.foodproductdesign.com/articles/462/462 0297DE.html. Posted 1 September 2000.
Lee SK, Yu HJ, Cho NS, Park JI-I, Kim TH, Abdi H, Kim KM, Lee SK, inventors; October 2008. A Method for Preparing the Inclusion Complex of Ginseng Extract with Gamma-Cyclodextrin, and the Composition Comprising the Same. U.S. patent WO/2008/127063.
Lesschaeve I, Noble AC. 2005. Polyphenols: factors influencing their sensory properties and their effects on food and beverage preferences. Am J Clin Nutr 81(1 Suppl):330S-5S.
Lelhuaut L, Brossard C, Meynier A, Rousseau F, Llamas G, Bousseau B, Genot C. 2005. Sweetness and aroma perceptions in dairy desserts varying in sucrose and aroma levels and in lextural agent. Int Dairy J 15(5):485-93.
Mojet J, Koster EP, Prinz JF. 2005. Do tastants have a smell? Chem Senses 30(1):9-21.
Muether AT, Lee SY. 2005. Halo effect on bitterness and astringency by flavor attributes in soy protein isolate (SPI) model solutions [dissertation]. University of Illinois at Urbana-Champaign.
Reineccius GA. 2004. Flavoring Systems for Functional Foods, In: T. Wilson, N. J, Temple, editors. Beverages in Nutrition and Health. Totowa: Humana Press. p89-97.
Roy G, Roy GM. 1997. Modifying Bitterness: Mechanism, Ingredients, and Applications. CRC Press.
Szejtli J. 1988. Cyclodextrin Technology. Boston: Kluwer Academic Publishers. 450 p.
Szejtli J, Szente L. 2005. Elimination of bitter, disgusting tastes of drugs and foods by cyclodextrins. Eur J Pharm Biopharm 61(3): 115-25.
Szente L, Szejtli J. 2004. Cyclodextrins as food ingredients. Trends Food Sci Technol 15(3-4): 137-42.
Tamamoto,L.C, Schmidt.S.J., Lee S. 2009. Sensory Profile of a Model Energy Drink with Varying Levels of Functional Ingredients-Caffeine, Ginseng, and Taurine. J Food Sci In Submission.
Tamura M, Mori N, Miyoshi T, Koyana S, Kohri I-I, Okai I-I. 1990. Practical Debittering Using Model Peptides and Related Compounds. Agric Biol Chem 54(1):41-51.
Vuksan V, Sievenpiper JL. 2005. Herbal remedies in the management of diabetes: Lessons learned from the study of ginseng. Nutrition, Metabolism and Cardiovascular Diseases 15(3): 149-60.
Wansink B, Payne CR, North J. 2007. Fine as North Dakota wine: Sensory expectations and the intake of companion foods. Physiol Behav 90(5):712-716.
Ward JH. 1963. Hierarchical grouping to optimize a quantitative function. J Am Stat Assoc 58(301):236-44.
Yu KK, inventor; 1993. Method for removing bitter taste of Ginseng. Korea Patent 930,005,196 B.
Zhao L, Kirkmeyer SV, Tepper BJ. 2003. A paper screening test to assess genetic taste sensitivity to 6-n-propyllhiouracil. Physiol Behav 78(4-5):625-33.
148
6.7 Tables and Figures
Table 6.1: Solution treatment codes and corresponding levels of y-, [1-CDs, and their combinations in both 100 mL water base and 100 mL model cnergj' drink base solutions
Sample Code yCDl
yCD2
yCD3
yCD4
yCD5
yCD6
yCD7
Cone y-CDs
(g/100 mL) 0
0.030
0.060
0.090
0.120
0.150
0.180
y-CDs Molarity (mol/L)
0
2.31xl0-5
4.63 xlO"5
6.94x10-5
9.25xl0-5
1.16x10-"
1.39X10"1
Cone p-CDs
(g/100 mL) ~
~
~
~
~
~
~
|l-CDs Molarity )(moI/L)
~
~
~
~
~
~
~
PCD1 PCD2
PCD3
PCD4
PCD5
PCD6
PCD7
~ ~
~
~
~
~
~
~ ~
~
~
~
~
~
0 0.250
0.500
0.750
1.000
1.250
1.500
0 2.20x10-"
4,41x10-"
6.61x10-"
8.81x10-"
LlOxlO'3
1.32X10'3
Combo* 1
Combo 2
Combo 3 Combo 4
Combo 5
Combo 6
Combo 7
0.030
0.105
0.180 0.030
0.180
0.105
0.030
2.31 xlO-5
S.lOxlO'5
1.39xl0-5
2.31x10°
1.39xl0-5
S.lOxlO-5
2.31xl0"5
0.250
0.875
1.500 1.500
0.250
0.250
0.875
2.20x10-"
7.71 xlO-4
1.32xl0'3
1.32xl0-J
2.20x10-"
2.20x10-"
7.71X10'4
Combo indicates combination of y- and P-CDs.
149
Tabic 6.2: Bitterness intensity rankings (l=least bitter to 6=most bitter) of the bitterness minimizing treatments incorporated in a 0.0529 g ginscng/100 mL water solution.
Treatment
y-Cyclodextrins
Resolver®
Rapidase
Taurine
Control
Citrus Flavoring
Concentration (g/100 mL)
0.090
0.800
0.200
15.160
0.000
0.001
Mean Ranking"
1.00a
3.09b
3.45bc
4.09 bc
4.64 bc
4.73° Mean rating scores with a superscript letter are not significantly different (p<0,05, Least Significant
Ranked Difference Test).
150
Table 6.3: Mean bitterness intensity rating scores (0 to 9) of bitterness minimizing treatment levels incorporate!:
Treatment (g/lOOmL)
in a 0 Mean Intensity
Rating"
y-Cyclodextrins
0
0.03
0.06
0.09
0.12
0.15
0.18
7.2 r ± 5.79b ±
3.04c ±
1.59c ±
1.64c ±
1.59c ±
2.68'1 ±
1.72
1.76
1.31
0.93
1.08
0.93
1.41
Resolver®
0
0.2
0.4
0.6
0.8
1
1.2
' 6.18a ±
5.46a ±
5.79a ±
4.61a ±
5.79a ±
5.68a ±
6.21° ±
2.11
2.37
2.36
1.94
1.81
2.28
1.72 P-Cyclodextins
0
0.03
0.06
0.09
0.12
0.15
0.18
6.90a ±
6.20ab ±
5.00nbc ±
4.60bc ±
5.50b0 ±
4.50bc ±
5.20° ±
1.60
1.40
2.16
1.58
1.78
2.17
2.10 Mean rating scores within each treatment with a superscript letter are not significantly
different (p<0.05, Least Significant Difference Test).
151
Table 6.4: Analysis of Variance on descriptive attributes rated for ginseng solutions containing varying levels of y- and P-CDs. Adjus ted F-values a wi th j u d g e x sample interact ion as t he e r r o r
Attribute Replication Panelist
Quinine Bitter 0.10 46.44***
Caffeine Bitter Aftertaste 0.14 59.76***
Samples
(Solutions)
106.63***
111.15***
t e rm.
RxP
0.96
0.87
RxS
1.14
1.46*
PxS I 44***
1.61***
Adjusted
F
74.05***
69.04*** Statistical significance at p<0.05, p<0.01 and p<0.001 are denoted by *, " , and "*, respectively. Rxp= Replication by panelists interaction; R*S=Replication by Sample Interaction; PxS=Panelist by Sample Interaction.
Table 6.5: Mean" quinine bitter and caffeine bitter aftertaste attribute intensity scores (0 to 15) across all 21 solution treatments combining water base and model energy drink base solutions and with and without nose clips usage data. CDs=CycIodextrins.
Treatment (y-CDs
g/100 mL)
y- l (0)
y-2 (0.03)
7-3 (0.06)
y-4 (0.09)
y-5 (0.12)
y-6 (0.15)
7-7(0.18)
Quinine Bitter
12.91a
11.18b
6.63c
3.99d
3.35e
3.55de
2.79f
Caffeine Bitter
Aftertaste
12.80a
10.53b
6.13c
3.91d
3.44d
3.48d
2.72e
Treatment (p-CDs
g/100 mL)
P-1(0)
P-2 (0.25)
p-3 (0.50)
0-4 (0.75)
P-5 (1.00)
p-6 (1.25)
P-7(1.5)
Quinine Bitter
12.59a
9.58b
7.14c
4.95d
3.83e
3.55e
3.55e
Caffeine Bitter
Aftertaste
12.35a
9.13b
6.89c
4.73d
4.09e
3.59c
3.61"
Treatment (7- and P-CDs
g/100 mL)
Combo 1
Combo 2
Combo 3
Combo 4
Combo 5
Combo 6
Combo 7
Quinine • Bitter
8.54a
3.40bc
3.18c
3.69b
2.97c
3.68b
3.81b
Caffeine Bitter
Aftertaste
7.65a
3.28bcd
3.24^
3.69 te
2.84d
3.27bcd
3.71b
"Means within a column per treatment with the same superscript letter are not significantly different (p<0.05, Fisher's Least Significant Difference Test). * Combo indicates combination of 7- and P-CDs. For 7- and p-CD amounts in the Combination treatments, refer to Table 1.
Table 6.6: Mean quinine bitter and caffeine bitter aftertaste attribute intensity scores (0 to 15) across all 7 y-CD solution treatments in water base or model energy drink base, without nose clips and with nose clips usage data. The data are plotted in Figure 6.1. CDs=CycIodextrins.
Treatment
(y-CDs g/100 mL)
Water Base, without nose
clips Water Base,
with nose clips
Model Energy Drink Base,
without nose clips
Model Energy Drink Base,
with nose clips
Quinine Bitterness
7-1(0)
7-2 (0.03)
7-3 (0.06)
7-4 (0.09)
y-5 (0.12)
7-6 (0.15)
7-7 (0.18)
13.54a
11.38b
7.42°
4.33d
3.42d
3.29d
3.08d
± 1.64
±2.30
±3.09
±2.39
±1.91
± 1.78
±2.32
12.54a
1121"
6.92b
3.75c
2.54cd
2.50^
2.00d
± 2.65
± 3.06
± 4.05
± 2.59
± 1.79
± 1.93
± 1.53
12.50a
11.00b
6.46c
4.00d
3.96d
4.58d
3.46d
± 2.64
± 2.75
± 3.46
± 3.31
± 3.01
± 3.09
± 3.09
13.04a
11.13b
5.71c
3.88d
3.50de
3.83e
2.63e
± 1.52
± 2.03
± 2.63
± 2.38
± 2.15
± 2.46
± 2.08
Caffeine Bitter Aftertaste
7-1(0)
y-2 (0.03)
7-3 (0.06)
y-4 (0.09)
7-5 (0.12)
7-6 (0.15)
7-7 (0.18)
13.67a
11.25b
7.08c
4.79d
3.83dc
3.33e
3.08e
±1.31
±1.92
±2.57
±2.28
± 1.93
±1.88
±1.25
13.29a
10.71b
6.67c
3.25d
2.42de
2.17e
1.88e
± 2.01
± 2.68
± 3.62
± 2.19
± 1.61
± 1.66
± 1.92
11.83a
9.88b
6.17c
4.17d
4.50d
4.46d
3.42d
± 2.93
± 3.04
± 3.64
± 3.13
± 3.08
± 3.08
± 2.72
12.42*
10.29b
4.58c
3.42dc
3.00de
3.96cd
2.50ef
± 1.67
± 1.76
± 2.48
± 2.43
± 2.13
± 2.77
± 2.02 Means within a column per treatment with the same superscript letter are not significantly different (p<0.05, Fisher's Least Significant Difference Test).
Table 6.7: Mean" quinine bitter and caffeine bitter aftertaste attribute intensity scores (0 to 15) across all 7 P-CD solution treatments in -water base or model energy drink base, without nose clips and with nose clips usage data. The data are plotted in Figure 6.2. CDs=Cyclodextrins.
Treatment
(p-CDs g/100 mL)
Water Base, without nose clips
Water Base, with nose clips
Model Energy Drink Base,
without nose clips
Model Energy Drink Base,
with nose clips
Quinine Bitterness
P-1 (0)
P-2 (0.25)
p-3 (0.50)
P-4 (0.75)
p-5 (1.00)
P-6 (1.25)
p-7 (1.50)
13.17a
10.58b
8.00°
6.21d
4.58E
3.79e
3.96e
± 1.81
± 2.65
± 3.39
± 2.73 .
± 2.43
± 2.11
± 2.20
12.38*
10.54b
7.3 8C
5.00d
3.42e
3.13e
3.17e
± 2.43
± 3.28
± 3.23
± 2.98
± 2.81
± 2.40
± 2.06
12.63a
8.29b
6.29c
4.00d
3.75d
3.54d
3.96d
^
±
1.84
2.85
± 3.53
±
±
2.78
2.69
± 2.89
± 2.84 •
12.21a
8.92b
6.88c
4.58d
3.58de
3.42d=
3.13e
± 1.96
± 2.32
± 2.46
± 2.38
± 2.26
± 2.89
± 2.54
Caffeine Bitter Aftertaste
P-1 (0)
p-2 (0.25)
p-3 (0.50)
p-4 (0.75)
P-5 (1.00)
p-6 (1.25)
P-7 (1.50)
13.38a
10.25b
8.3 8C
6.29d
5.04de
4.63e
4.58e
± 1.64
± 2.31
± 2.62
± 2.49
± 2.22
± 2.14
± 2.06
12.46a
10.7 lb
7.33c
4.75d
3.33e
3.00e
2.75e
± 2.23
± 2.77
± 2.78
± 2.54
± 2.75
± 2.09
± 2.17
12.25a
8.00b
6.2 l c
3.88d
4.25d
3.71d
4.13d
J.
±
j .
-L
±
±
±
1.80
2.98
3.60
2.98
3.22
2.74
3.00
11.33a
7.54b
5.63c
4.00d
3.75d
3.04d
3.00d
± 2.18
± 2.84
± 2.37
± 2.62
± 2.54
± 2.68
± 2.27
"Means within a column per treatment with the same superscript letter are not significantly different (p<0.05, Fisher's Least Significant Difference Test).
(a)
en c
• * - » CO tr.
w c a> *̂ r.
14 •
12 '
10 i i
8 i
6 •: ; ! 4 .. i
2 -<
ot
(b) r
14 •!
12 H
en ! .S 10 " •*->
CO
£ = 6 -0)
0.00 0.03 0.06 0.09 0.12 0.15 0.18
4
2i 0 -I r
0.00 0.03 0.06 0.09 0.12 0.15
grams y-CDs in 100 ml of so lu t ion 0,18
Figure 6.1: Effect of y-CD levels on (a) quinine bitter and (b) caffeine bitter aftertaste intensity ratings of ginseng solution treatments with and without nose clips and in water base or model energy drink base solutions. (•) without nose clips in water base, (•) with nose clips in water base, (A) without nose clips in model energy drink base, and (x) with nose clips in model cnergj' drink base. The data arc presented in Table 6.6.
155
(a)
en c *-• co (£ & in c n> •»-»
c
14 -
12 i ^
10 < •
. 8 1
i 6 •
4 •
2
0 -
0.00 0.25 0.50 0.75 1.00 1.25 1.50
(b)
D) C
* J
CO cc & w c 0)
• J
c
10
8
6
4
14 -:
12
0.00 0.25 0.50 0.75 1.00 1.25 grams p-CDs in 100 ml of solution
1.50
Figure 6.2: Effect of P-CD levels on (a) quinine bitter and (b) caffeine bitter aftertaste intensity ratings of ginseng solution treatments with and without nose clips and in water base or model cnergj' drink base solutions. (•) without nose clips in water base, (•) with nose clips in water base, (A) without nose clips in model cnergj' drink base, and (x) with nose clips in model energy drink base. The data arc presented in Table 6.7.
156
350
100
en to
b 50
Q o
•LT ,
CD
7 om
bo
> - <->
*=r Q O st
un o o 1
ex
I') u s o o
Q O >-
M o
£ n o
•*t o
E o
E 1 o ° o <->
u j t D N c N r - ' - m m r N 2 Q Q Q Q Q Q Q Q C J - 2
o o o o o o o o o E i i i i i i i i • o
> - 0 2 . C O . > - > - CO. D J - Q 3 . C O . o
Figure 6.3: Agglomerative hierarchical clustering (AHC) of quinine bitter and caffeine bitter aftertaste attribute mean intensity ratings for 21 ginseng solution treatments containing varying levels of y-CD and p-CD on the dissimilarity scale by Euclidean distance and agglomeration by Ward's method. The dotted line was computed using the XLStat 7.5.3 software and truncates the clusters based on the largest relative increase in dissimilarity. Refer to Tabic 6.1 for the corresponding amounts of y- and p-CDs in each solution treatment.
157
CHAPTER 7 - SUMMARY
The significant influx of a wide variety of commercially-available functional
beverages into the market has resulted in a beverage category that is not clearly defined
or understood. The rapid increase in functional beverages has also resulted in the lack of
understanding of the sensory, chemical, and physical effects of functional ingredients in
these products.
Three categorization methods: 1) ingredient inventory, 2) flow behavior
comparison, and 3) two-step sensory sorting were used to categorize fifty commercially-
available functional beverages. The categorization is important in formulating a
successful product based on consumer perception and expectation. Of the three methods,
the two-step sorting method produced the most well-defined categories.
Validation and reproducibility studies were then conducted on the two-step
sensory sorting method. Adjusted Rand Index (ARI) values greater than 0.90 showed
that there was excellent correspondence between the fixed sorts conducted in the
validation study. Six functional beverage categories were generated in the reproducibility
study, with the major difference between the initial sort and the replicated sort being that
the Yogurt Smoothies and Fruit Smoothies categories were combined into one category
encompassing both types of beverages in the replicate. The high ARI values from the
validation study (ARI >0.88) and the similar functional beverage categories generated
through the reproducibility study (ARI >0.77) suggest that the two-step sensory sorting
method can be used to consistently create similar functional beverage categories. This
research suggests that the two-step sorting method is a valid and reproducible method to
relatively quickly categorize a large number (~50) of functional beverages and is useful
158
to the beverage industry for the formulation of acceptable functional beverages. Taste is
a key component in the acceptability of food products and the more information known
about the effects of the inclusion of ingredients into aTood matrix, the belter we can
develop successful products.
The two-step sensory sorting method has the potential as a quick categorization
method that could be useful in product development and marketing. Future research
includes investigating the validity and reproducibility of the two-step sensory sorting
method as a rapid process to categorize large groups of products. One such study would
involve sorting a set of products that fall into well-known, defined categories to
determine if the method accurately categorizes products, It would be interesting to
investigate the effectiveness of the two-step sensory sorting method by using it to
categorize other products such as cereals, candies, and food bars. Another study should
focus only on an oral sensory evaluation sort to determine if the functional beverages
could be categorized by only oral sensations and tastes without the influence of
packaging and prior advertising. The level of sweetness, mouthfeel, or other oral-
sensations may play a significant role in defining and generating functional beverage
categories. Lastly, it would be interesting to conduct the two-step sensory sorting method
on a group of functional beverages including newly introduced products to determine if
new functional beverage categories would be generated. If the two-step sensory sorting
method has the ability to effectively categorize other products, it may be a valuable and
inexpensive tool that aids in the development of product descriptors and understanding
product relationships.
159
To determine the effects of functional ingredients on the sensory properties, we
focused on the "energy drink" category, because it is one of the largest sectors in the
functional beverages market and it was one of the most clearly defined categories
generated from the categorization study. The most common functional ingredients found
in energy drinks are caffeine, ginseng, and taurine. The combinations of the three
functional ingredients were added to a model energy drink solution, and descriptive
analysis (DA) was conducted on 27 combinations (3x3x3 factorial design) of the three
functional ingredients at three concentrations (low, medium, high) to determine the
synergistic effects on the sensory properties of the solutions.
A horns effect was observed as the sweet, artificial lemon-lime, pear, mango, and
pineapple attributes were rated lower in intensity with increased ginseng levels. Taurine
levels of up to 416 mg/100 mL had no significant effect on the sensory attribute ratings of
the model energy drink solutions. The results from the DA research on the effects of
functional ingredients suggested that ginseng contributes predominantly to the bitter
attributes. Therefore, a DA was conducted on 21 ginseng solutions with the use of
masking agents (3 masking agent treatments x 7 levels x 2 bases) to reduce the bitter
attributes contributed by 0.052 g ginseng in 100 mL water base and model energy drink
base solutions. It was found that y- and P-cyclodextrins (CDs) showed the most promise
in reducing the bitterness of ginseng in these solutions. Results showed that 0.09 g y-CD
in 100 mL solution and 1 g P-CD in 100 mL solution both reduced the bitterness intensity
of the solutions by half.
In regards to developing energy drinks containing functional ingredients,
specifically ginseng, future research should include running a consumer acceptance test
160
on a ginseng model energy drink base solution incorporating the different types and
levels of CDs to determine the bitterness level that is acceptable to consumers. Also,
conducting a study on the effect of a specific amount of CDs in solutions containing
varying concentrations of ginseng could be done to determine the degree of effectiveness
CDs have on reducing bitterness in regards to ginseng levels. This research will be useful
in selecting the minimum amount of CDs necessary to produce an acceptable energy
drink.
161
CHAPTER 8 - REFERENCES
Anonymous 2008. U.S. Functional Beverages Market: A Young Market with Growing Popularity. Food Industry News [serial online]. Available from Posted 28 May 2008 2008.
Amendola C, Iannilli I, Restuccia D, Santini I, Vinci G. 2004. Multivariate statistical analysis comparing sport and energy drinks. Innovative Food Science & Emerging Technologies 5(2):263-7.
Anonymous 1975. Minutes of Division Business Meeting. Institute of Food Technologists-Sensory Evaluation Division, IFT.
Abdi I-I, 2007. The RV coefficient and the congruence coefficient. Encyclopedia of measurement and statistics.Thousand Oaks (CA): Sage
Alldrick AJ. 2006. Functional beverages: satisfying consumers' needs. Agro Food Industry Hi Tech 17(4):41.
Allison ANNMA, Chambers E. 2000. The Importance Of Caffeine As A Flavor Component in Beverages. J Sens Stud 15(4):449-57.
Asao M, Iwamura I-I, Akamalsu M, Fujita T. 1987. Quantitative structure-activity relationships of the bitter thresholds of amino acids, peptides, and their derivatives. J Med Chem 30(10): 1873-9.
Astray G, Gonzalez-Barreiro C, Mejuto JC, Rial-Otero R, Simal-Gandara J. 2009. A review on the use of cyclodextrins in foods. Food Hydrocoll 23(7): 1631-40.
Backas N, 2009. Brimming Opportunities for Nutraceutical Beverages. Food Product Design [serial online], 19(2):50-68. Available from Posted January 21, 2009.
Barcenas P, Elortondo FJP, Albisu M. 2004. Projective mapping in sensory analysis of ewes milk cheeses: A study on consumers and trained panel performance. Food Res Int 37(7):723-9.
Berry D. 2009. Powerful Beverage Performers, Food Product Design [serial online], 32-50. Available from Posted 2009.
Berlino M, Lawless HT. 1993. Understanding mouthfeel attributes: a multidimensional scaling approach. J Sens Stud 8(2): 101-14.
Binello A, Cravotto G, Nano GM, Spagliardi P. 2004. Synthesis of chitosan-cyclodextrin adducts and evaluation of their bitter-masking properties, Flavour Fragrance J 19(5):394-400.
162
Borg I, Groenen P. 2003. Modern Multidimensional Scaling: Theory and Applications. Journal of Educational Measurement 40(3):277-80.
Brandt LA. 2001. Flavor Masking: Strategies for Success. Prepared Foods [serial online], (July 2001):Available from Posted 2001. -
Breslin PAS. 1996. Interactions among sally, sour and bitter compounds. Trends Food SciTechnol7(12):390-9.
Burg JC. 1998. High-Energy Beverages. Food Product Design
Calvino AM, Garcia-Medina MR, Cometlo-Muniz JE. 1990. Interactions in caffeine-sucrose and coffee-sucrose mixtures: evidence of taste and flavor suppression. Chem Senses 15(5):505-19.
Camire ME. 2000. Dietary Supplements. In: M. K. Schmidl, T. P. Labuza, editors. Essentials of Functional Foods. Gaithersburg: Aspen Publishers, Inc. pi 65-180.
Cartier R, Rytz A, Lecomte A, Poblete F, Krystlik J, Belin E, Martin N. 2006. Sorting procedure as an alternative to quantitative descriptive analysis to obtain a product sensory map. Food Qual Pref 17(7-8):562-71.
Cook DJ, Hollowood TA, Linforth RST, Taylor AJ. 2003. Oral shear stress predicts flavour perception in viscous solutions. Chem Senses 28(1):11-23.
Coon JT, Ernst E. 2002. Panax ginseng: a systematic review of adverse effects and drug interactions. Drug Saf 25(5):323-44.
Cosgrove J. 2008. Caffeinated Snacks. Nutraceuticals World [serial online]. July/August 2008Available from Posted July 2008.
Court WE. 2000a. Ginseng: The Genus Panax. Singapore: CRC Press. 266 p.
Court WE. 2000b. The Pharmacology and Therapeutics of Ginseng. In: Anonymous Ginseng: The Genus Panax. Singapore: Harwood Academic Publishers, pi 17-197,
Coxon APM. 1999. Sorting Data: Collection and Analysis. Sage Publications Inc.
Cravotto G, Binello A, Baranelli E, Carraro P, Trotta F. 2006. Cyclodextrins as Food Additives and in Food Processing. Current Nutrition & Food Science 2(4):343-50,
de Sousa PHM, Maia GA, de Azeredo HMC, de Souza Filho MSM, Garruti DS, de Freitas CAS. 2007. Mixed tropical fruit nectars with added energy components. Int J Food Sci Tech 42(11):1290-1296.
163
Delarue J, Sieffermann JM, 2004. Sensory mapping using Flash profile. Comparison with a conventional descriptive method for the evaluation of the flavour of fruit dairy products. Food Qual Pref 15(4):383-92.
Delwiche J. 2004. The impact of perceptual interactions on perceived flavor. Food Qual Pref 15(2): 137-46.
Dodziuk I-I. 2006. Cyclodextrins and their complexes: chemistry, analytical methods, applications. Weinheim: Wiley-VCH. 489 p.
Drobna Z, Wismer WV, Goonewardene LA. 2004. Selection of an Aslringency Reference Standard for the Sensory Evaluation of Black Tea, J Sens Stud 19(2): 119-32,
Falahee M, MacRae AW. 1997. Perceptual variation among drinking waters: the reliability of sorting and ranking data for multidimensional scaling. Food Qual Pref 8(5):389-94.
Farrell J, RiesNM, Kachan N, Boon H. 2009. Foods and natural health products: Gaps and ambiguities in the Canadian regulatory regime. Food Policy
Faye P, Bremaud D, Durand Daubin M, Courcoux P, Giboreau A, Nicod H. 2004. Perceptive free sorting and verbalization tasks with naive subjects: an alternative to descriptive mappings. Food Qual Pref 15(7-8):781-91.
Faye P, Bremaud D, Teillet E, Courcoux P, Giboreau A, Nicod H. 2006. An alternative to external preference mapping based on consumer perceptive mapping. Food Qual Pref 17(7-8):604-14.
Food and Drug Administration. 2003. Code of Federal Regulations Title 2 Sec. 182.1180.
Fraley C, Raftery AE. 1998. How Many Clusters? Which Clustering Method? Answers Via Model-Based Cluster Analysis. The Computer Journal 41(8):578-88.
Frank RA. 2003. Response context affects judgments of flavor components in foods and beverages. Food Quality and Preference 14(2): 139-45.
Galloway DR, Talanian JL, Shoveller AK, Heigenhauser GJF, Spriet LL. 2008. Seven days of oral taurine supplementation does not increase muscle taurine or alter substrate metabolism during prolonged exercise in humans. J Appl Physiol 105(2):643-51.
Gawel R, Oberholster A, Francis IL. 2000. A 'mouth-feel wheel': terminology for communicating the mouth-feel characteristics of red wine. Australian journal of grape and wine research. 6(3):203-07.
164
Giboreau A, Dacremont C, Egoroff C, Guerrand S, Urdapillela I, Candel D, Dubois D. 2007. Defining sensory descriptors: Towards writing guidelines based on terminology. Food Qual Pref 18(2):265-74.
Gnadt T. 2007. Effect of Masking Agents and Lactones on the Aroma and Flavor of Soymilk. [dissertation]. University of Illinois at Urbana-Champaign.
Granato H. 2002. Manipulating Flavor Perception in Functional Products. Natural Products Insider [serial online]. Available from Posted 8 April 2002 2002.
Giiclii-Ustundag O, Mazza G, 2007. Saponins: Properties, Applications and Processing. Crit Rev Food Sci Nutr 47(3):231-58.
Hamilton RM, Heady RE, inventors; 1970. Eliminating Undesirable Taste From Coffee And Tea Extracts And Products. U.S. patent United States Patent 3528819.
Hollowood TA, Linforth RST, Taylor AJ. 2002. The Effect of Viscosity on the Perception of Flavour. Chem Senses 27583-91.
Huang KC. 1999. The Pharmacology of Chinese Herbs. Boca Raton: CRC Press. 512 p.
Hubert L, Arabie P. 1985. Comparing partitions. Journal of classification -2(1): 193-218.
Humphries G. 2007. Nutraceutical Soft Drinks: Innovation in sports, energy, dairy, and functional beverages.
Institute of Food Technologists. 2005. Functional Foods: Opportunities and Challenges.
Jung MY, Jeon BS, Bock JY. 2002. Free, esterified, and insoluble-bound phenolic acids in white and red Korean ginsengs (Panax ginseng CA Meyer). Food Chem 79(1): 105-11,
Kang MW, Chung SJ, Lee MS, Kim Y, Kim KO. 2007. The sensory interactions of organic acids and various flavors in ramen soup systems. J Food Sci 72(9):S639-47.
Kappes SM, Schmidt SJ, Lee SY. 2006. Color halo/horns and halo-attribute dumping effects within descriptive analysis of carbonated beverages. J Food Sci 71(8):S590-5.
Kappes SK, Schmidt SJ, Lee SY, 2006, Mouthfeel Detection Threshold and Instrumental Viscosity of Sucrose and High Fructose Corn Syrup Solutions. J Food Sci 71(9):S597-602.
Kappes SM, Schmidt SJ, Lee SY. 2007. Relationship between Physical Properties and Sensory Attributes of Carbonated Beverages. J Food Sci 72(1):S001-11.
Katan MB, Roos NM. 2004. Promises and Problems of Functional Foods. Crit Rev Food Sci Nutr 44(5):369-77.
165
Keast RSJ. 2008. Modification of the bitterness of caffeine. Food Qual Pref 19(5);465-72.
Keast RSJ, Breslin PAS. 2003. An overview of binary taste-taste interactions. Food Qual Pref 14(2): II1-24.
Kim . 1985. Effects of temperature and sugar addition on the flavor of ginseng tea. Korean Journal of Food Science and Technology. 17(4):304.
King MC, Cliff MA, Hall JW. 1998. Comparison of Projective Mapping and Sorting Data Collection and Multivariate Methodologies for Identification of Similarily-Of-Use Snack Bars. J Sens Stud 13(3):347-58.
King BM, Duineveld CAA, Arents P, Meyners M, Schroff SI, Soekhai ST. 2007. Retronasal odor dependence on tastants in profiling studies of beverages. Food Qual Pref 18(2):286-95.
Kitts D, Hu C. 2000. Efficacy and safety of ginseng, Public Health Nutr 3(4A):473-85.
Konno A, Misaki M, Toda J, Wada T, Yasumatsu K. 1982. Bitterness Reduction of Naringin and Limonin by fi-Cyclodextrin. Agric Biol Chem 46(9):2203-8.
Koutsou GA, Storey DM, Bar A. 1999. Gastrointestinal tolerance of y-cyclodextrin in humans. Food Additives & Contaminants 16(7):313-7.
Kruskal JB, Wish M. 1978. Multidimensional scaling Quantitative applications in the Social Sciences. Sage Publications Newbury Park. 93p.
Krzanowski WJ, Marriott FHC, 1995. Multivariate analysis. New York: Halsted Press. 280 p.
Labbe D, Damevin L, Vaccher C, Morgenegg C, Martin N. 2006. Modulation of perceived taste by olfaction in familiar and unfamiliar beverages. Food Qual Pref 17(7-8):582-9.
Lai GG. 2007. Getting Specific with Functional Beverages. Food Technology [serial online], 61(12):Available from Posted 2007.
Lawless HT, Glatter S. 1990, Consistency of multidimensional scaling models derived from odor sorting. J.Sensory Studies 5(2):217-30.
Lawless HT, Sheng N, Knoops SSCP, 1995. Multidimensional scaling of sorting data applied to cheese perception. Food Qual Pref 6(2):91-8.
166
LeClair K. 2000, Breaking the Sensory Barrier for Functional Foods. Food Product Design [serial online], 6 November 2006. Available from http://www.foodproductdesign.com/articles/462/462 0297DE.html, Posted 1 September 2000.
Lee SK, Yu HJ, Clio NS, Park JH, Kim TH, Abdi H, Kim KH, Lee SK, inventors; October 2008. A Method for Preparing the Inclusion Complex of Ginseng Extract with Gamma-Cyclodextrin, and the Composition Comprising the Same. U.S. patent WO/2008/127063.
Lelicvre M, Chollet S, Abdi I-I, Valentin D. 2008. What is the validity of the sorting task for describing beers? A study using trained and untrained assessors. Food Qual Pref
Lesschaeve I, Noble AC. 2005. Polyphenols: factors influencing their sensory properties and their effects on food and beverage preferences. Am J Clin Nutr 81(1 Suppl):330S-5S.
Lelhuaut L, Brossard C, Meynier A, Rousseau F, Llamas G, Bousseau B, Genot C. 2005. Sweetness and aroma perceptions in dairy desserts varying in sucrose and aroma levels and in textural agent. Int Dairy J 15(5)^485-93.
Leveen T. 2007. Functional Beverages: Market Evolution. Natural Products Marketplace [serial online]. Available from Posted 2007.
Ley JP, Blings M, Paetz S, Krammer GE, Bertram HJ, 2006. New bitter-masking compounds: hydroxylated benzoic acid amides of aromatic amines as structural analogues of homoeriodictyol. J Agric Food Chem 54(22):8574-9. •
Ley JP, Krammer G, Reinders G, Gatfield IL, Bertram HJ. 2005. Evaluation of bitter masking flavanones from Herba Santa (Eriodictyon californicum (H. & A.) Torr., Hydrophyllaceae). J Agric Food Chem 53(15):6061-6.
Lim J, Lawless HT. 2005. Qualitative differences of divalent salts: multidimensional scaling and cluster analysis. Chem Senses 30(9):719-26.
Loftsson T, Brewster ME. 1996. Pharmaceutical applications of cyclodextrins. 1. Drug solubilization and stabilization. J Pharm Sci 85(10):
Lord JB, 2000, New Product Failure and Success. In: A. L. Brody, J. B. Lord, editors. Developing New Food Products for a Changing Marketplace. Lancaster: Technomic Publishing Co, Inc. p55-86.
Lourenco R, Camilo ME. 2002. Taurine: a conditionally essential amino acid in humans? An overview in health and disease, Nutr Hosp 17(6):262-70.
Luckow T, Delahunty C. 2004. Consumer acceptance of orange juice containing functional ingredients. Food Res Int 37(8):805-14,
Macfie HJ, Bratchell N, Greenhoff K, Vallis LV. 2007. Designs to Balance the Effect of Order of Presentation and First-Order Carryover Effects in Hall Tests. J Sens Stud 4(2): 129-148.
MacQueen JB. 1966. Some Methods for Classification and Analysis of Multivariate Observations. Proc. Fifth Berkeley Symp. on Math. Statist, and Prob.281-297.
Mattes RD. 2007. Effects of linoleic acid on sweet, sour, salty, and bitter taste thresholds and intensity ratings of adults. AJP-Gastrointestinal and Liver Physiology 292(5):G1243.
Mervis CB, Rosch E. 1981. Categorization of Natural Objects. Annual Reviews in Psychology 32(1 ):89-l 15.
Minor LJ, Pearson AM, Stine CM. 1966. Chicken Flavor Studies. J Agric Food Chem 14(4):416-9.
Mintel. 2009. Energy Drinks and Energy Shots-US.
Mintel. 2008. Functional Beverages-US August 2008. Mintel Reports.
Mobini S, Elliman TD, Yeomans MR. 2005, Changes in the pleasantness of caffeine-associated flavours consumed at home. Food Qual Pref 16(8):659-66.
Mojet J, Koster EP, Prinz JF. 2005. Do tastants have a smell? Chem Senses 30(1):9-21.
Moreau CP, Markman AB, Lehmann DR. 2001. "What is it?" Categorization flexibility and consumers' responses to really new products. Journal of Consumer Research 27(4):489-98.
Murphy GL, Ross BH. 1994. Predictions from uncertain categorizations. Cogn Psychol 27(2): 148-93.
Muether AT, Lee SY. 2005. Halo effect on bitterness and astringency by flavor attributes in soy protein isolate (SPI) model solutions [dissertation]. University of Illinois at Urbana-Champaign.
Nestrud MA, Lawless HT. 2008. Perceptual mapping of citrus juices using projective mapping and profiling data from culinary professionals and consumers. Food Qual Pref 19(4):431-38.
168
Orth UR, de Marchi R. 2005. Advertising's influence on product experience and purchase intention. Fruit Processing 6:372-76.
Ohki K, Nakamura Y, Takano T. 2004. The Japanese Nutritional Health Beverage Market. In: Anonymous Beverages in Nutrition and Health. Totowa: Humana Press, p377-387.
Packaged Facts, 2009. Functional Foods and Beverages in the U.S. 1-210.
Pages J. 2005. Collection and analysis of perceived product inter-distances using multiple factor analysis: application to the study of 10 white wines from the Loire Valley. Food Qual Pref 16(7):642-9.
Perrin L, Symoneaux R, Maitre I, Asselin C, Jourjon F, Pages J. 2008. Comparison of three sensory methods for use with the Napping® procedure: Case often wines from Loire valley. Food Qual Pref 19(1):1-11.
Ramos de Armas R, Gonzalez Diaz I-I, Molina R, Perez Gonzalez M, Uriarte E. 2004. Stochastic-based descriptors studying peptides biological properties: modeling the bitter tasting threshold of dipeptides. Bioorg Med Chem 12(18):4815-22,
Rand WM. 1971. Objective criteria for the evaluation1 of clustering methods. Journal of the American Statistical Association 846-50.
Reid AW, Becker CH, REID AW. 1956. The use of cocoa syrups for masking the taste of quinine hydrochloride. J Am Pharm Assoc Am Pharm Assoc 45(3):151-2.
Reineccius GA. 2004. Flavoring Systems for Functional Foods. In: T. Wilson, N. J. Temple, editors, Beverages in Nutrition and Health. Totowa: Humana Press. p89-97.
Reissig CJ, Strain EC, Griffiths RR. 2009. Caffeinated energy drinks—A growing problem. Drug Alcohol Depend 9(1):1-10.
Risvik E, Mcewan JA, Colwill .IS, Rogers R, Lyon DH. 1994. Projective mapping: a tool for sensory analysis and consumer research. Food Qual Pref 5(4):263-9.
Rouseff RL. 1990. Bitterness in Food Products and Beverages. Amsterdam: Elsevier. 356 P-
Roy G, Roy GM. 1997. Modifying Bitterness: Mechanism, Ingredients, and Applications. CRC Press.
Rugg G, McGeorge P. 1997. The sorting techniques: a tutorial paper on card sorts, picture sorts and item sorts. Expert Syst 14(2):80-93.
169
Saint-Eve A, Paci Kora E, Martin N. 2004. Impact of the olfactory quality and chemical complexity of the flavouring agent on the texture of low fat stirred yogurts assessed by three different sensory methodologies. Food Qual Pref 15(7-8):655-68.
Schiffman SS, Reynolds ML, Young FW. 1981. Introduction to Multidimensional Scaling: Theory, Methods, and Applications. New York: Academic Press. 413 p.
Schiffman SS, Sattely-Miller EA, Graham BG, Zervakis J, Butchko HH, Stargel WW. 2003. Effect of repeated presentation on sweetness intensity of binary and ternary mixtures of sweeteners. Chem Senses 28(3):219-29.
Smit HJ, Rogers PJ. 2002. Effects of'energy' drinks on mood and mental performance: critical methodology. Food Qual Pref 13(5):317-26.
Smith DM, Awad AC, Bennink MR, Gill JL. 1995. Cholesterol reduction in liquid egg yolk using P-cyqlodextrin. J Food Sci 60(4):691-4.
Soldo T, Hofmann T. 2005. Application of Hydrophilic Interaction Liquid Chromatography/Comparative Taste Dilution Analysis for Identification of a Bitter Inhibitor by a Combinatorial Approach Based on Maillard Reaction Chemistry. J Agric Food Chem 53(23):9165-71.
Soufflet I, Calonnier M, Dacremont C. 2004. A comparison between industrial experts' and novices' haptic perceptual organization: a tool to identify descriptors of the handle of fabrics. Food Quality and Preference 15(7-8):689-99.
Spiller MA. 1998. Caffeine. Boca Raton: CRC Press. 363 p.
Steinley D. 2004. Properties of the Hubert-Arabie adjusted Rand index. Psychol Methods 9(3):386-96.
Stevens JC. 1997. Detection of very complex taste mixtures: generous integration across constituent compounds. Physiol Behav 62(5): 1137-43.
Stone I-I, Sidel JL, 2004. Sensory Evaluation Practices. Academic Press,
Szejtli J. 1988. Cyclodextrin Technology. Boston: Kluwer Academic Publishers. 450 p.
Szejtli J, Szente L. 2005. Elimination of bitter, disgusting tastes of drugs and foods by cyclodextrins. Eur J Pharm Biopharm 61 (3): 115-25.
Szente L, Szejtli J. 2004. Cyclodextrins as food ingredients. Trends Food Sci Technol 15(3-4): 137-42.
170
Takeuchi T, Naae N, 1992. Stationary phase effect on enantioselectivity of dansyl phenylalanine in microcolumn liquid chromatography with y-cyclodextrin as mobile phase additive. J Chromatogr 595(l-2):121-4.
Tamamoto,L.C, Schmidt,S.J., Lee S. 2009. Sensory Profile of a Model Energy Drink with Varying Levels of Functional Ingredients-Caffeine, Ginseng, and Taurine. J Food Sci In Submission
Tamura M, Mori N, Miyoshi T, Koyana S, Kohri I-I, Okai H. 1990. Practical Debittering Using Model Peptides and Related Compounds. Agric Biol Chem 54(1):41-51.
Tang C, Heymann H. 2002. Multidimensional Sorting, Similarity Scaling and Free-Choice Profiling of Grape Jellies. J Sens Stud 17(6):493-509.
Tarea S, Cuvelier G, Sieffermann JM. 2007. Sensory Evaluation of the Texture of 49 Commercial Apple and Pear Purees. J Food Qual 30(6): 1121-31.
Tuorila H, Cardello AV. 2002. Consumer responses to an off-flavor in juice in the presence of specific health claims. Food Qual Pref 13(7-8):561-9.
Urala N, Lahteenmaki L. 2003, Reasons behind consumers functional food choices. Nutr Food Sci 33(4):148-58.
Viswanathan M, Childers TL. 1999. Understanding How Product Attributes Influence Product Categorization: Development and Validation of Fuzzy Set-Based Measures of Gradedness in Product Categories. J Market Res 36(l):75-94.
Vogler BK, Pittler MH, Ernst E. 1999. The efficacy of ginseng. A systematic review of randomised clinical trials. Eur J Clin Pharmacol 55(8):567-75.
Vuksan V, Sievenpiper JL. 2005. Herbal remedies in the management of diabetes: Lessons learned from the study of ginseng. Nutrition, Metabolism and Cardiovascular Diseases 15(3): 149-60.
Wansink B, Payne CR, North J. 2007. Fine as North Dakota wine: Sensory expectations and the intake of companion foods. Physiol Behav 90(5):712-716.
Ward JH. 1963. Hierarchical grouping to optimize a quantitative function. J Am Stat Assoc 58(301):236-44.
Warskulat U, Flogel U, Jacoby C, Hartwig MG, Thewissen M, Merx MW, Molojavyi A, Heller Stilb B, Schrader J, Haussinger D. 2004. Taurine transporter knockout depletes muscle taurine levels and results in severe skeletal muscle impairment but leaves cardiac function uncompromised. FASEB Journal 18(3):577.
Williams EJ. 1950, Experimental designs balanced for pairs of residual effects. Australian Journal of Scientific Research 3(3):351-363.
Woolf A, Wright R, Amarasiriwardena C, Bellinger D. 2002. A child with chronic manganese exposure from drinking water. Environ Health Perspecl 110(6):613-6.
Wright R. 2008. Nutraceuticals Coast in the Beverage Market. Nutraceuticals World [serial online]. Available from Posted July 2008.
Yamauchi T, Markman AB. 2000. Inference using categories. Journal of Experimental Psychology Learning Memory and Cognition 26(3):776-95,
Yatabe Y, Miyakawa S, Ohmori H, Mishima H, Adachi T. 2009, Effects of Taurine Administration on Exercise. In: Anonymous Advances in Experimental Medicine and Biology: Taurine 7, New York: Springer New York. p245-252.
Yeomans MR, Mobini S, Chambers L. 2007, Additive effects of flavour-caffeine and flavour-flavour pairings on liking for the smell and flavour of a novel drink. Physiol Behav 92(5):831-9.
Yu J, Kim AK. 2009. Effect of Taurine on Antioxidant Enzyme System in B16F10 Melanoma Cells. In: Anonymous Taurine 7. New York: Springer. p491.
Yu KK, inventor; 1993 Method for removing bitter taste of Ginseng. Korea Patent 930,005,196 B. •
Zhao L, Kirkmeyer SV, Tepper BJ. 2003. A paper screening test to assess genetic taste sensitivity to 6-n-propylthiouracil. Physiol Behav 78(4-5):625-33.
MD ma MM
JW:
ml mm M ui. um M MDI UM l l l l l l l l i M nui 11 «wM<jjU|iiim!
nn l l l : , | l l l l ! i l l l l l l l ! i n l i n l : f l l l l l l l l l l l : ' i 1 1 ' ^ TOT irnm M II I II l l l l l l l i l l l tesmnnmnw
l l l i i l II m um MfMrnTTiii i iinffrnirr i I I 1
IBM HI I Ml iiiiiiiii WMHWt iptwr-iK w lllllllllllllllllHIl n ITTTTITI rrnTmr lili VW1WI)
I I > H > « mm in i l l l i lllllllilll
l i lHIMl l l l i l i l i i l i i SHI Ml \m i l l l i l l l l l l i l i l l i l i l l l lwl l l l l l iHl l i l l i lQi i
i nil |iii|i||i|||iii|iii i, WWP M i i i i i i i wmw \m
I ' l l l l l l lWll l l l l l l l l l l l l l lDI i i Nlll | i l | i i | | l | |h WflW
ill ilmilm
OUT I h i iiiHiiiiiiiiiiiiiiniw I HI I
TIM mm I I I m ram mm HUH m \m lllllllllllllllll m l i l l l l l l l l l l l l l l l l i l M l
I IJPj j l •ii ' • ' i ; - , i - i | :• i HI M
III|I'1IIII||I1IIIIHI!II lllllillllllil | l | | H | | | | l l l l l l | | | | | MM. "IPS l l l l l l l
arawmms MM l l l l l l i l l l l l l i l TgwPw l l l l l i l l l i l i Mm n i l fflT
ni i i i i
MP" MDI mm M JM II IB I B WWHK
M m m iniiiiiiiini .MM MDI
HE ml iiiiiiiiiiiiiiiiiryiiiiiiiiiiiiiiiiiiiiiiii
I I I MR WW&tfBi M M H I W IIIIIMHIIII iiiiiiiiiiiiiiii
SB
lff||W|||!ll|l|'lll|l||QM iiTntnt fflffl
IIMIT WW
iMM Willi II
wrrrrrmt i mnv*-.
mb Bi l l
I l | IB M i l I Hi i
l l l l l l l Mil OTI «ai«JHtk
l l l i I m i;-g|tti;<ig IM I llli j liljl llli II I lllllll MM 1 1 1 l l j l l l ! ffifc
HIM lllllll
TrnjitrnTrrrT M K
.wimmiMUft M liHIH iWilM
ri!lllllii|l |l,ill|IIIBllll|ll |lil |i |ii | |ll |l | ,,,np, M W i l l HI I
|IIHI| l i l i DP W B iiinTHiHinn i i i i i I i j i l i
JTfllllii l|iini||HliMTI'liTIHli|illHli|||ill HIHIIIIIIIIB jgisi; iiimiiriiiiMn i l l hi Mil HIIIM i lit IIII i i l l IMll mm II limn in WWW!
UJ fflmrnTi in iniiiiiiiiiii
HIr KSW§
WLffl I III :.lll|i|i| I lljlll I'M Will lllllllilll IUIIIIB SMWIB
ni | iH'i i i i i i i inii i i i i i i i i i i i i i i i i i i i | i i i i i i i i in | i i i i i i i i i i i i i : i i i | i i | i i i i | iw UK II
lllll;»lll|l|l|'ai MM WW\ M DM DTI lEQliOWTOB
Ifflifflfl JIM UL raTiliiIII11ii ii I l l l l l l l l l i l l l i l l l l l l l l l b lWI I l l
HIIIIHT
m 1 m i n
in mum
mum M i l rffui m IfflMi msm ~Bm iiniiiiH MMTl mm JUffi
nwsn lllllllllllllllll|llllll|fl| 1IHH
Mimr llillllllFillltll'llll'lll^MF LhH 111 ixLLi j Hi j i flflTP
im in 11 iiiln iiiiii|i ui 11 MI IHTHiJI I IIIlM lllllll 11
H m
MLUlll H T M I
IE um iJlJIHTT ma i i ln i i i in i i l i i i i i i i i i i i i i i i i i i i i i i i l i in i i i i i i i i i i i i i i i i l i i i i i i l i i i i i i i i i i l i i i i i hllHIIllllllillllllillllll 1 H HHI1 I I H I' rTTTTTTBIIHTHil
l l l | i | i lHlll l l i lHlll i l i i l l l l l lHlii[ | | | l i l i | | | | | l l i l l i | lirnTnTT I lllllllli anaaMaawa
Directions: Please read the categories and the characteristics of the categories. Then VISUALLY evaluate the beverages and place each beverage into a category (stickers are provided for this task). You may sample the beverages in any order and may go back and sample the beverages as many times as you would like to,
PLEASE MAKE SURE TO PLACE EACH BEVERAGE INTO A CATEGORY!!!!
Energj' Drinks
Characteristics: Contains stimulants such as caffeine and taurine; provides extra energy
Are able to participate on all the scheduled dates?
Questionnaire:
1. Gender (check one):
a Male " p Female
2. Age Group (check one): a Under 21 a 21-30 a 31-40 a 41-50
3. Ethnicity (check one): a White a Black a Hispanic
Yes/No
Yes/No
a 51-60 a 61-70 a 71-80 a over 80
a Asian/Pacific Islander a American Indian/Alaska Native a Other (please describe)
4. Are you currently on a restricted diet? If yes, please explain. Yes/No
5. Do you have any of the following? • Dentures a Diabetes a Food Allergies
a Oral or Gum Disease a Hypoglycemia • Hypertension
6. If you have allergies please list:
177
APPENDIX E (cont.)
7. Are there any foods or beverages that you hate?
8. Do you consume products containing ginseng? Yes / No
If yes, what products do you consume?
9. Do you smoke? Yes / No
178
APPENDIX F: DESCRIPTIVE ANALYSIS RECRUITMENT PRESCREENING TASTE IDENTIFICATION TEST
Name: Date:
SOLUTION TESTS Your task is to recognize the basic taste of each sample solution (sweet, salty, sour or bitter). Write in the blank which taste you perceive. When the sample tastes like water mark with a "0". If your recognition is questionable, write a question mark "?". Re-tasting is allowed.
For each sample, take the sample into the mouth in sips and move it around in such a way that it touches all parts of the tongue, Do not swallow the sample; use spit cups. Rinse between samples with spring water.
PAPER TEST Place the piece of filter paper on your tongue, close your mouth, and wet the paper with saliva for 10 seconds. Do you perceive a taste? If so, what do you taste?
On a scale of 1-10 (ten being the strongest possible taste), circle the number that represents how strong the taste you perceive is (if you didn't perceive anything, leave this blank).
1 Very weak 2 3 4 5 6 7 8 9
10 Very Strong
179
APPENDIX G: INFORMED CONSENT FORM FOR SENSORY EVALUATION STUDIES
INFORMED CONSENT FORM FOR SENSORY EVALUATION PANELISTS
"EFFECTIVE MASKING AGENT TASTE TESTING"
You are invited to participate in a study involving sensory evaluation of ginseng solutions. The goal of this research is to establish the perceived bitter levels of ginseng solutions containing bitterness modifiers. These solutions will be evaluated using a hybrid descriptive analysis method. You will be asked to taste each sample and rank the samples by bitterness intensity. You will also be asked to taste each sample and rate the bitterness intensity of each sample. You are free to withdraw from the study at any time for any reason.
The study will be conducted at Bevier Hall Room # 376 (Sensory lab). We anticipate that there will be two evaluations over a span of two weeks. Each evaluation session will last about 10 minutes. Participation in the study will be voluntary.
Your performance in this study is confidential. Responses are coded to be anonymous and any publications or presentations of the results of the research will only include information about group performance.
You will be able to withdraw at any time during the course of the study. The experimenter(s) also reserve the rights to terminate the study of an individual subject at any time during the course of the whole study.
You are encouraged to ask any questions that you might have about this study whether before, during, or after your participation. However, specific questions about the samples that could influence the outcome of the study will be deferred to the end of the experiment. Questions can be addressed to Dr. Soo-Yeun Lee (217-244-9435, [email protected]) or Lauren C. Tamamoto (217-333-9795, [email protected]). You may also contact the IRB Office (217-333-2670, [email protected]) for any question about the rights of research subjects. If you live outside the local calling area, you may also call collect.
I understand the above information and voluntarily consent to participate in the study described above. I have been offered a copy of this consent form.