Plant Foods for Human Nutrition 54: 119–130, 1999.
© 1999 Kluwer Academic Publishers. Printed in the Netherlands.
Measurement of the relative sweetness of stevia
extract, aspartame and cyclamate/saccharin blend as
compared to sucrose at different concentrations
H.M.A.B. CARDELLO1, M.A.P.A. DA SILVA2 and M.H. DAMASIO3
1 Department of Food and Nutrition, FCF-UNESP, CP 502, CEP 14801-902 Araraquara, SP,
Brazil (e-mail: [email protected]); 2 Department of Food Technology, FEA-UNICAMP,
Campinas, SP, Brazil; 3 Instituto de Agroquímica y Tecnología de Alimentos – C.S.I.C.,
Burjassot, Valencia, Spain
Received 28 April 1998; accepted in revised form 4 May 1999
Abstract. Special diets are used to mitigate many human diseases. When these diets require
changes in carbohydrate content, then sweetness becomes an important characteristic. The
range of low-calorie sweeteners available to the food industry is expanding. It is essential
to have an exact knowledge of the relative sweetness of various sweeteners in relation to
different sucrose concentrations. The objective of this study was to determine the variation
on the relative sweetness of aspartame (APM), stevia [Stevia rebaudiana (Bert.) Bertoni] leaf
extract (SrB) and the mixture cyclamate/saccharin – two parts of cyclamate and one part of
saccharin – (C/S) with the increase in their concentrations, and in neutral and acid pH in equisweet concentration to 10% sucrose, using magnitude estimation. Sweetness equivalence of
SrB in relation to sucrose concentrations of 20% or higher and of APM and C/S to sucrose
concentrations of 40% or higher could not be determined, because a bitter taste predominated.
The potency of all sweeteners decreased as the level of sweetner increased. In equi-sweet
concentration of sucrose at 10%, with pH 7.0 and pH 3.0, the potency was practically the
same for all sweeteners evaluated.
Key words: Aspartame, Cyclamate, Saccharin, Sensory analysis, Stevia, Sweetness
Introduction
Alternatives to sucrose serve a number of purposes. They are used to expand
food and beverage choices for those who must or want to control caloric,
carbohydrate, or sugar intake; assist weight control or reduction; aid in the
management of diabetes; assist the control of dental caries; enhance the usability of pharmaceuticals and cosmetics; provide sweetness when sugar is
not available; and assist the cost-effective use of limited resources. There
are a number of sweetening agents permitted in various countries. Because
of factors such as cost, availability, legislation, exports or reduced energy
intake, sometimes food companies think that is necessary or appropriate to
120
substitute one sweetener for another. It is important that such alterations not
cause significant changes in the sensory characteristics of the product. Thus, it
is essential to have a clear understanding of the various sweeteners which are
perceived as equi-sweet to sucrose at acid and neutral pH value. Many foods
in which it is desired to replace the sucrose have an acid nature. Therefore,
it is necessary to know the equivalence of sweetness in acid pH in order to
maintain the sensory quality of foods and beverages. These data are essential
for qualitative and quantitative comparisons of the sensory characteristics of
sweeteners.
Various sweeteners have been used in foods with success [1], for example,
the extract of stevia leaves [2] and a mixture of cyclamate/saccharin [3]. A
large number of studies looking at the sweetness equivalent of aspartame
(APM) have been done; however, there is disagreement in the studies about
the sweetness equivalents of APM. Homler [4] found that APM is 400 times
sweeter than sucrose at 3.4% and 215 times at 4.3%. Cloninger & Baldwin [5]
reported APM was considered 182 times sweeter than sucrose at 2% and 43
times sweeter at 30%. Tunaley et al. [6] considered APM 128 times sweeter
than a 5% sucrose solution. In all these studies, the sweetness potential decreased with increase in concentration. Some authors have reported a residual
bitter taste with increased APM concentration [7, 8], while others did not find
this [4, 9].
Stevia is composed of several natural, heat-stable entkaurene glycosides
(steviol glycosides) whose intensities of sweetness and flavor profiles differ
from each other and vary according to concentration and environment. Collectively, they give stevia 100 to 300 times the sweetness of sucrose. Extracted
from a cultivar of chrysanthemum, Stevia rebaudiana, the sweet principle, is
believed to have been used by the Paraguayan Indians for centuries. Since
the 1950s, the Japanese have overcome problems of refining to eliminate
undesirable flavors. Following a slow start in the 1970s, stevia is now used in
a wide range of food applications in Japan. Currently it is being evaluated for
approval in the West. Use in man and data from animal feeding trials indicate
that it is safe for human consumption [11]. Extracts of stevia, leaves [Stevia
rebaudiana (Bert.) Bertoni] (SrB) have been used in products in Japan for
10 years. This sweetener is considered about 300 times sweeter than sucrose
in a 0.4% solution, decreasing to 150 times as sweet in a 10% solution. The
potency can vary depending on the purity and the proportion of stevioside
and rebaudioside found in the extract [10].
Although human consumption of stevia began before the Spanish settlement of the country we now know as Paraguay, improved versions have been
developed only recently. Variation in the profile and total concentration of
the diterpene glycosides is well illustrated by Kinghorn & Soejarto [12] who
121
summarized the results of analyses on a total of 67 samples of leaves of Stevia
rebaudiana (Bert.) Bertoni from Japan, Korea, Paraguay and Brazil which
appeared in 17 separate reports in Japan between 1975 and 1982. Of the
other leaf components, Kinghom & Soejarto [12] listed triterpene, labdane
diterpene, sterol, flavonoid glycoside, tannins and 31 volatile oils. Soejarto
et al. [13] suggested that the bitter taste, common to many Stevia species, is
probably due to sesquiterpene lactones.
A mixture of cyclamate/saccharin is widely used in different proportions
with the potency varying as a function of the ratio. For the ratio of 2:1 of
cyclamate/saccharin (C/S), that is commercially used in Brazil, no study on
the sweetness equivalent to sucrose has been found.
The potency of the various non-nutritive and nutritive sweetners can vary
with the methodology used to determine sweetness and methods used to obtain this information vary in the literature. For example, paired comparisons
[14], variations of the constant stimulus method [15] and the much applied
method of magnitude estimation and graphic representation of the normalized
results using Stevens Law, or Power Function [16–18] have been employed.
The psychophysical relationship between the increase in perceived sweetness intensity as a function of concentration provides practical formulation
information. Magnitude estimation, a ratio scaling technique, has been frequently employed to generate dose-response functions of sweeteners [7, 16,
19, 20]. This method has several advantages over other scaling procedures
when studying psychophysical relationships. It yields a scale that has true
ratio properties, has no arbitrarily limited endpoints and provides data, which
can be converted to percentages and compared between studies. It may also
provide a psychophysical curve frequently described by a power function.
The exponent or slope of the power function curve denotes the rate of change
in perceived sweetness as a function of concentration.
The objectives of this study were to evaluate the sweetness of the aspartame (APM), stevia leaf extract (SrB) and a mixture of cyclamate/saccharin
– two parts of cyclamate and one part of saccharin – (C/S) against sucrose at
concentrations of 3, 10, 20, 30, 40 and 50% (w/v) and to evaluate the effect
of pH 3.0 and pH 7.0 on the sweetness of the alternative sweeteners against a
10% sucrose solution.
Materials and methods
Materials. The sweeteners studied were: stevia [Stévia rebaudiana (Bert.)
Bertoni] leaf extract supplied by Stéviafarma Industrial S/A, Brazil; aspartame
supplied by NutrasweetTM Co.; sodium cyclamate (Brasfanta Indústria e Comércio Ltda. Brazil); sodium saccharin (Choheung Chemical Industrial Co.,
122
Ltd. supplied by Vepê Indústria Alimentícia Ltda.; and sucrose (Sigma Chemical Co.). The sodium cyclamate and sodium saccharin were mixed in a 2:1
ratio which is the most common combination in the commercial market in
Brazil. The composition of the stevia leaf extract was: 81.0% stevioside,
0.6% rebaudioside C and 7.7% rebaudioside A. These components were determined by high-perfornance liquid chromatography and ultraviolet detection [21].
Samples and procedure. Deionized water was used to prepare the solutions
and was available to the 10 trained subjects for rinsing purposes. All the
solutions were prepared 24 hours before testing and held in the dark at room
temperature (25 ± 2 ◦ C) in volumetrics flasks. No solution was retained
for more than 48 hours. For each series of experiments, the panelists were
provided with a solution of sucrose (the reference sweetener) at 3, 10, 20, 30,
40 or 50% (marked with ‘R’), and samples presented in 50 ml beakers coded
with three digit random numbers. The experiment on equivalence of sweetness in sucrose at 10%, was determined also at pH 3.0. The acid solutions
of sweeteners were prepared using citrate-phosphate buffer, pH 3.0. Sensory
tests were carried out in a standardized room. The study was performed by a
panel of 10 trained judges (5 women and 5 men, staff members from the Faculty of Pharmaceuticals Ciences of Araraquara, SP, Brazil) who were between
35 and 45 years of age, selected for their ability to discriminate sweet taste in
sequential analyses with triangular tests.
Relative sweetness measurement. Measurements of the relative sweetness
of the high intensity sweeteners and sweetener mixtures were accomplished
using the method of magnitude estimation [22], which yields a direct quantitative measure of the subjective intensity of sweetness. The subjects were
informed that they would be presented a reference sample with a arbitrary
sweetness value of 100, followed by a random series of solutions with intensities both less than and greater than the reference intensitity. Their task was to
estimate the sweetness intensity of the unknowns relative to the reference. For
example, a value of 200 should indicate a sample two fold sweeter than the
reference, whereas a value of 50 should be half as sweet as the reference. They
were instructed to use integers that seemed appropriate (but not zero) and to
judge each solution separately. Judges also were informed that the reference
would be presented periodically. In this experiment, the reference was placed
at the geometric mean of the series. The test concentrations utilized are listed
in Table 1. The serial dilution differed from one another by a factor of 1.6 for
determining sweetness equivalent for sucrose at 3, 10 (at both pH values), 20
and 30% and a factor of 2 at 40 and 50%.
123
Table 1. Concentrations of sucrose, aspartame (APM), stevia leaf
extract (SrB) and cyclamate/saccharin (C/S), tested to determine
sucrose equivalent sweetness
Stimuli
Sucrose
APM
SrB
C/S
Sucrose
APM
SrB
C/S
Sucrose
APM
SrB
C/S
Sucrose
APM
SrB
C/S
Sucrose
APM
SrB
C/S
Sucrose
APM
Srb
C/S
Concentrations for equivalency at 3%1
1.1700 1.8700 3.0000
4.8000
0.0073 0.0120 0.0190
0.0300
0.0780 0.0125 0.0200
0.0320
0.0040 0.0650 0.01034 0.01654
Concentrations for equivalency at 10%2
3.9100 6.2500 10.00
16.00
0.0200 0.0340 0,0550
0.0880
0.0391 0.0625 0.1000
0.1600
0.0141 0.0225 0.0360
0,0576
Concentrations for equivalency at 20%
7.8125 12.50
20.00
32.00
0.0781 0.1250 0.2000
0.3200
0.0390 0.0625 0.1000
0.1600
0.0434 0.0694 0.1111
0.1778
Concentrations for equivalency at 30%
11.72
18.75
30.00
48.00
0.3910 0.6250 1.000
1.6000
0.2300 0.3750 0.600
0.9600
0.1172 0.1875 0.300
0.4800
Concentrations for equivalency at 40%
27.77
33.33
40.00
48.00
0.9200 1.1100 1.3300
1.6000
0.9200 1.1100 1.3300
1.6000
0.6944 0.8333 1.0000
1.2000
Concentrations for equivalency at 50%
26.40
31.25
50.00
60.00
1.7400 2.0800 2.500
3.0000
1.7400 2.0800 2.500
3.0000
1.7361 2.0833 2.500
3.0000
1 Concentrations on percentage (w/v%).
2 Concentrations utilized at both pH 3.0 and pH 7.0.
7.6800
0.0480
0.0500
0.0265
25.60
0.1408
0.2560
0.0923
51.20
0.5120
0.2560
0.2844
76.48
2.5600
1.5300
0.7680
57.60
1.9200
1.9200
1.4400
72.00
3.6000
3.6000
3.6000
124
Table 2. Slope values, variances, standard deviations, y-intercepts, correlations coeficients (r) and power function for the test stimuli reported in Table 1 for sucrose (SUC),
aspartame (APM), cyclamate/saccharin (C/S) and stevia leaf extract (SrB)
Stimuli
Slope Variance Standard YR1
deviation intercept
Power function
SUC 3%
APM SES 3%
SrB SES 3%
C/S SES 3%
1.5572
0.7983
0.8481
0.5528
0.2574
0.1382
0.0735
0.0322
0.4538
0.3325
0.2426
0.1604
–0.7452
1.4330
1.4409
1.0914
0.9850
0.9908
0.9998
0.9982
P=0.1798, S1.5572
P=27.11, S0.7983
P=27.5907, S0.8481
P=12.34, S0.5528
SUC 10% pH 7.0
APM SES 10%
SrB SES 10%
C/S SES 10%
1.3364
0.9481
0.7770
0.6724
0.1930
0.1031
0.0688
0.0476
0.3929
0.2872
0.2346
0.1951
–1.3358
1.2009
0.7664
0.9660
0.9860
0.9843
0.9554
0.9935
P=0.046, S1.3364
P=15.8830, S0.94112
P=5.84, S0.7770
P=9.2472, S0.6724
SUC 10% pH 3.0
APM SES 10%
SrB SES 10%
C/S SES 10%
1.2976
1.2048
0.6146
0.7864
0.1045
0.1676
0.0416
0.0659
0.4205
0.4094
0.2040
0.2567
–1.2966
1.5262
0.6376
1.1405
0.9975
0.9812
0.9719
0.9876
P=0.0505, S1.2976
P=33.5892, S1.2048
P=4.3411, S0.6146
P=13.8197, S0.7864
SUC 20%
APM SES 20%
SrB SES 20%2
C/S SES 20%
1.7336
1.1426
–
1.1422
0.3159
0.1301
–
0.1378
0.5027
0.3323
–
0.3320
–2.2415
0.7849
–
1.0916
0.9961
0.9930
–
0.9980
P=0.0057, S1.7300
P=6.093, S1.1426
–
P=12.3482, S1.1412
SUC 30%
APM SES 30%
SrB SES 30%2
C/S SES 30%
2.0262
1.0242
–
1.3105
0.4371
0.1310
–
0.2258
0.5913
0.3237
–
0.4250
–3.0475
–8.1×10
–
0.5869
0.9900 P=0.0009, S2.0262
0.9128 P=0.99981, S1.0242
–
0.8902 P=3.8621, S1.3105
1 R refers to Pearson, the correlation coefficient.
2 Results are not practically valid for the calculation of the sweetness equivalence.
SES = Sweetness Equivalent of Sucrose.
Data analysis. Data were normalized and magnitude estimates were converted to logarithmic values and expressed using the geometric mean. Dose
response curves for each sweetener were fitted to the power function of S=aCn ,
where S was the stimuli perceived, C was the concentration of the stimuli, a
was the antilog of the value of the y-intercept and n was the slope.
125
Figure 1. Sweetness power functions for aspartame (APM) (), stevia leaf extract (SrB) (#),
cyclamate/saccharin (C/S) () and sucrose (SUC) ( ) in aqueous solutions. A = sucrose
equivalence at 3%; B = sucrose equivalence at 10% at pH 7.0; C = sucrose equivalence at
10% at pH 3.0; D = sucrose equivalence at 20%; E = sucrose equivalence at 30%.
Results and discussion
Psychophysical function. When the logs of the concentrations (C) of each
sweetener were plotted against the logs of the magnitude estimates of the
sensations (S), a regression line could be fitted to the points, indicating that a
simple power function S=aCn (or log S= log a+n log C) could describe the
data, which is presented in Table 2 for each sweetener and each concentra-
126
tion and illustrated in Figure 1 (A to E). The relationship between sweetness
intensity and concentration for each sweetener was represented graphically
on log-log coordinates, for each sweetness concentration equivalence. Log
concentrations for the high potency sweeteners are in % w/v. APM, SrB, C/S
and sucrose shared different exponent values (slope) for each concentration
and reference. Although exponent values can vary depending on how samples
were evaluated, the values for sucrose and APM were generally consistent
with previously published data [16, 23].
Potency. Sweetness potency curves, derived from the power curves, were
developed for the three high potency sweeteners relative to sucrose (Figure
1 A to E). Potency was defined as the number of times sweeter a compound,
on a weight basis, was than an isosweet concentration of sucrose. All curves
exhibited a decrease in sweetness potency relative to sucrose as sweetner level
increased. The decreased potency for APM and SrB have been previously
published [4, 24]. C/S showed the same property as APM and SrB. C/S displayed the greatest potency among the three high intensity sweeteners. C/S
was about 288 times more potent than sucrose at a concentration equi-sweet
to 3% sucrose. At a 10% sucrose sweetness equivalency, C/S was about 275
times more potent than sucrose, where at pH 3.0 it was 281 times, and at
20% and 30%, 178 and 93 times, respectively. APM was about 190 times
more potent than sucrose at a concentration equi-sweet to 3% sucrose. At
10% (at pH 3.0 and pH 7.0), 20%, and 30% sucrose sweetness equivalencies,
APM was about 185, 96 and 34 times more potent than sucrose, respectively.
These values were consistent with previously published potency data [4, 8,
22]. SrB was about 152 times more potent than sucrose at a concentration
equi-sweet to 3% sucrose. At a 10% sucrose sweetness equivalency, SrB was
about 97 times more potent than sucrose at pH 7.0, and 109 times at pH
3.0, therefore, the acid pH favored the increase of the potency of SrB in that
sweetness equivalence.
Sweetness equivalence. Based on the potency curves for APM, SrB and
C/S, a graph was developed to show the sweetness equivalence values on an
equi-sweet level of sucrose. Figure 3 data illustrate the relationship between
sweetness of APK SrB, C/S and sucrose when plotted on a log-log scale.
The same Figure shows that the magnitude of the sucrose equivalent of
the three sweeteners decreased as level of sweetness increased.
APM at a 0.016 g level promoted the same level of sweetness as 3.0 g
of sucrose in aqueous solution, and 0.054, 0.208 and 0.88 g were needed for
10.0, 20.0 and 30.0 g of sucrose, respectively. Higher (0.133%) [25], lower
(0.050%) [26] and the same (0.053%) [8] aspartame values have been cited
at sweetness equivalence for 10% sucrose. This discrepancy could be due to
127
Figure 2. Potencies of aspartame (APM), stevia leaf extract (SrB) and cyclamate/saccharin
(C/S) over a wide range of sucrose sweetness equivalencies (pH 7.0).
Figure 3. Relationship of sweetness of aspartame (APM), stevia leaf extract (SrB), cyclamate/saccharin (C/S) and sucrose (pH 7.0).
methodological differences. Off-tastes have been detected at relatively high
saccharin and aspartame sweetener concentrations [27]. C/S was equi-sweet
at 0.01, 0.0365, 0.1122 and 0.324 g as sucrose at 3.0, 10.0, 20.0 and 30 g in
aqueous solution. SrB was equi-sweet at 0.0197 and 0.103 g as sucrose at 3.0
and 10.0 g, respectively.
The panelists observed that SrB had an increased bitter residual taste with
increased concentration, nearly covering the sweet taste starting at concentra-
128
Figure 4. Comparison between the potency sweetness of the sweeteners in equi-sweet
concentration as sucrose in aqueous solution 10% at pH 3.0 and 7.0.
tion equi-sweet to 20% sucrose. For APM and C/S the appearance of other
tastes impeded the determination of the sweetness equivalence starting at 40%
sucrose.
Table 2 data include the values utilized for the calculations of the sweetener
power functions for the equivalent sweetness determination. Based on the potency for APM, SrB and C/S, a graph was developed showing the sweetness
equivalence values on an equi-sweet level of sucrose (Figure 2). This graph
illustrates the relationship between sweetness of APM, SrB, C/S and sucrose.
Figure 4 information shows that the magnitude of the sucrose equivalent of
the three sweeteners was the same in pH 3.0 and 7.0, except for SrB, which
had a slight increase with the decrease in pH value. A concentration of 0.054
g of APM was needed to induce equivalent sweetness to 10.0 g of sucrose in
aqueous solutio at both pH 3.0 and 7.0. C/S was equi-sweet at 0.0365 g as
sucrose at 10.0 g in aqueous solution at pH 7.0, and 0.0356 g at pH 3.0 at the
same concentration. SrB was equi-sweet at 0.103 g and 0.092 g as sucrose at
10.0 g in pH 7.0 and 3.0, respectively. The sensory panelists observed that SrB
had decreased residual taste with decreased pH value. It is very important to
remember that these levels should be regarded as estimates. Sweetness equivalence values for the high potency sweeteners are highly system-dependent
and may vary in different food products [6].
129
Acknowledgments
We gratefully acknowledge the financial support provided by FAPESP – Fundação de Amparo à Pesquisa do Estado de São Paulo – Brazil (Process No.
945859-7).
References
1.
Hanger LY, Lotz A, Lepeniotis S (1996) Descriptive profiles of selected high intensity
sweeteners (IRS) HIS blends and sucrose. J Food Sci 61(2): 456–864.
2. Higginbotham JD (1982) L’état présent des aspartame. stévioside et autres édulcorants.
La Sucrerie Belge 101: 235.
3. Crosby GA (1976) New sweeteners. CRC in Food Sci Nutr Jun: 297.
4. Homler BE (1988) In Birch CG, Lindle MG (eds), Low calorie products. London:
Elsevier Applied Science.
5. Cloningert MR, Baldwin RE (1974) L-aspartyl-L-phenylalanine methyl ester (aspartame) as a sweetener. J Food Sci 39: 347.
6. Tunaley A, Thomson DMH, McEwan JA (1987) Determination of equi-sweet concentrations of nine sweeteners using a relative rating technique. Int J Food Sci Tech 22(6):
627.
7. Redlinger PA, Setser CS (1987) Sensory quality of selected sweeteners: aqueous and
lipid model systems. J Food Sci 52: 45.
8. Wie SG, Beyts PK (1992) Sensory characteristics of sucralose and other high intensity
sweeteners. J Food Sci 57(4): 1014.
9. Baldwin RE, Korschgen BM (1979) Intensification of fruit flavors by aspartame. J Food
Sci 44: 938.
10. Bakal AI, O’Brien Nabors L (1986) In O’Brien Nabors L, Gelardi RC (eds), Alternative
sweeteners. New York: Marcel Dekker.
11. DuBois E (1981) Diterpenoid sweeteners: Syntesis and sensory evaluation of stevioside
analogues nondegradable to steviol. J Med Chem 24: 1269.
12. Kinghorn AD, Soejarto DD (1985) Economic and medicinal plant research. London:
Academic Press.
13. Soejarto DD, Kinghorn AD, Farnsworth NR (1982) Potential sweetening agents of plant
origin. III. Organoleptic evaluation of stevia leaf herbarium samples for sweetness. J Nat
Prod 45: 590.
14. Yamaguchi S, Yoshikawa T, Ikeda S, Ninomyia T (1970) Studies on the taste of some
sweet substances part I. Inter-relationships among them. Agric Biol Chem 34: 181.
15. Ennis DM (1990) Chemical Senses 14(4): 597.
16. Stone H, Oliver SM (1969) Measurement of the relative sweetness of selected sweeteners
and sweetener mixtures. J Food Sci 34: 215.
17. Giovanni ME, Pangborn RM (1983) Measurement of taste intensity and degree of linking
of beverages by graphic scales and magnitude estimation. J Food Sci 48: 1175.
18. Wiseman JJ, McDaniel MR (1991) Modification of fruit flavors by aspartame and
sucrose. J Food Sci 56: 1668.
19. Moskowitz HR (1970) Ratio scales of sugar sweetness. Percep Psychoph 7(5): 315.
20. Bartoshuk LM, Renert K, Rodin J, Stevens C (1982) Effects of temperature on the
perceived sweetness of sucrose. Physiol Behav 28: 905.
130
21.
22.
23.
24.
25.
26.
27.
Hashimoto Y, Moriyasu M (1978) Determination of sweet components in Stevia rebaudiana by high-perfomance liquid chromatograph. Ultraviolet detection. Shoyakugaku
Zasshi 32 (2): 209.
Stevens SS (1957) On the psychophysical law. Psychol Rev 64: 153.
Gremby TH (1991) Intense sweeteneres for the food industry: an overview. Trends Food
Sci Tech jan 1991: 2.
Isima N, Kakayama 0 (1976) Sensory evaluation of stevioside as a sweetener. Rep Nat
Food Res Inst 31: 80.
Inglett GE (ed) (1974) Symposium: Sweeteners. Westport, CT: AVI Publishing Co .
Beck CI (1974) Sweetness, character, and applications of aspartic acid-based sweeteners.
In Inglett GE (ed), Symposium: Sweeteners. Westport, CT: AVI Publishing Co.
Harrison SK, Bernhardt RA (1984) Time-intensity sensory characteristics of sacharin,
xylitol and galactose and their effect on the sweetness of lactose. J Food Sci 49: 780.