Chem. Senses 27: 583-591,
2002
© Oxford University Press 2002
The Effect of Viscosity on the Perception of Flavour
Samworth Flavour Laboratory, Division of Food Sciences, University of Nottingham, Sutton Bonington Campus, Loughborough LE12 5RD, UK
Correspondence to be sent to: Tracey Hollowood, Samworth Flavour Laboratory, Division of Food Sciences, University of Nottingham, Sutton Bonington Campus, Loughborough LE12 5RD, UK. e-mail: tracey.haddock{at}nottingham.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
A trained sensory panel assessed flavour and sweetness intensity in solutions containing varying concentrations of hydroxy propyl methylcellulose (HPMC), sugar and flavour volatile. The flavour and sweetness of the viscous solutions were rated using magnitude estimation with a controlled modulus. In addition, the concentration of volatile released on the breath was measured using MS NoseTM. For low concentrations of HPMC (<0.5 g/100 g), perceived flavour intensity remained the same; however, a steady decrease was noted at higher concentrations (>0.6 g/100 g). The change in perceived intensity occurred at the point of random coil overlap (c*) for this hydrocolloid. The perceived sweetness of the solution showed a similar pattern with increasing HPMC concentration, although the inflection at c* was not so obvious. Despite the change in perceived flavour intensity, the actual concentration of volatile measured on the breath was not affected by the change in HPMC concentration. Low-order polynomial models were produced to describe perceived flavour intensity and sweetness in viscous solutions containing HPMC and potential explanations for the changes in perception are discussed.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
Flavour is defined as the combined perception of mouth-feel, texture, taste and aroma (British Standards Institute, 1975
). In application, it is important to understand how varying
these flavour components affects flavour quality so that foods can be
formulated to maximize consumer acceptability. Hydrocolloid thickeners are common ingredients in many food products. Utilized for their thickening properties at low concentration, they have a profound effect on both food texture and flavour. Reformulation of food flavour using empirical, trial-and-error methodology can be commercially inefficient. A fundamental understanding of how changes in matrix influence flavour release would be of great benefit to the food industry. Furthermore, understanding the relative contribution of hydrocolloid, non-volatile and volatile components to flavour perception could allow changes in perception to be predicted for a modified recipe.
It is generally understood that increasing viscosity through the addition
of thickeners results in a decrease in perceived intensity of volatile and
non-volatile components (Vaisey et
al., 1969; Moskowitz and
Arabie, 1970; Pangborn et
al., 1973; Christensen,
1980; Baines and Morris,
1987; Malkki et al.,
1993). Furthermore, the decrease can be dependent on thickener
type (Pangborn et al.,
1973, Paulus and Haas,
1980)
Previous studies showed that the perception of sweetness and strawberry
flavour was greatly affected by the addition of guar gum at concentrations
above the point of random coil overlap (c*) (Baines and
Morris, 1987,
1988). For any given
hydrocolloid, c* is the concentration at which individual
polymer chains interpenetrate and start to form an entangled network
(Morris et al.,
1981). It is dependent on the number and space occupancy of the
polymer molecules and is associated with a sharp increase in viscosity. Below
this concentration, the individual polymer chains are free to move
independently. Baines and Morris discovered that guar gum had no significant
effect on perception of sweetness or flavour below c*, but
above this concentration the perceived intensity of both attributes decreased
steadily with increasing polymer concentration. They concluded that the
decrease in flavour perception was due to inefficient mixing, as the polymer
chains became obstacles to diffusion, rather than direct binding of flavour
molecules to the polymer (Morris,
1987).
Contrary to the view that no binding occurs, an investigation into the
effect of polymer composition [oat gum, carboxy-methyl cellulose (CMC), guar
gum] on sensory perception revealed that the nature of the hydrocolloid had
more effect on perceived sweetness than viscosity
(Malkki et al., 1993).
Guar had the greatest effect on sweetness and oat gum the least. The reduction
in sweetness due to the addition of thickeners was dependent on the sweetener
used (aspartame, fructose, sucrose). The same study looked at the effect of
thickener type on the perception of flavour intensity. The physicochemical
properties of the compounds used were more important than the type of polymer.
Any differences in perception from equiviscous solutions of oat, guar and CMC
were determined to be evidence of binding or interaction between the polymer
and the flavour compounds.
To study potential binding effects, static equilibrium headspace was used
to study the behaviour of seven volatile compounds in water and in 1% CMC
solution (De Roos, 1997). At
equilibrium, viscosity effects are nullified and, therefore, any differences
in static equilibrium headspace between water and 1% CMC would be due to
binding. No differences were found, indicating that no binding occurred with
the biopolymer. However, CMC concentration did affect the release rate of the
volatile compounds during dynamic headspace studies. It was concluded that,
although flavour molecules do not bind to the polymer, the increase in
viscosity has a physical effect on the movement of flavour molecules.
In a comprehensive study investigating the effect of thickener composition
and viscosity on dynamic flavour release
(Roberts et al.,
1996), a decrease in the release of highly volatile compounds was
reported as viscosity increased. Less volatile compounds showed little or no
effect with increasing viscosity. The extent of the decrease was dependent on
both thickener type and viscosity, which the author suggested was because of
some sort of binding mechanism and the physical inhibition of volatile
mobility.
Much of the previous work investigating the effect of viscosity on flavour release and perception focuses on either the dynamics of the release mechanism or, alternatively, the sensory properties of the viscous solutions. Rarely have the two effects been studied together. Furthermore, the flavour release studies tend to simulate dynamic in-mouth conditions with the use of heated vessels, stirrers and a gas flow to represent breathing. They do not always recreate the dilution with saliva, swallowing, continuously changing volume and surface area or different chewing patterns typical of real eating or drinking. In the following study, the volatile release was measured using the MS NoseTM (Micromass, Manchester, UK). This non-invasive method allows the in-nose volatile signal to be measured, close to the nasal receptors, in human subjects rather than in model systems.
This paper investigates the effect of hydroxy propyl methylcellulose (HPMC) concentration on volatile release from viscous solutions and the perceived intensity of flavour and taste. In addition, it attempts to use low-order polynomial models to explain the perceptual responses in terms of HPMC, flavour and sugar composition of the samples.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
Experiment 1—effect of viscosity on release and perception of strawberry flavour
Sample preparation
Liquid samples were prepared containing HPMC (Methocel; DOW Germany) at
concentrations of 0.0625, 0.125, 0.25, 0.5, 0.75, 1.0, 1.5 and 2.0 g/100 g.
Each sample contained 2 g/100 g sugar (Tate & Lyle) and 200 p.p.m. of a
strawberry flavour (Firmenich SA, Geneva, Switzerland).
Samples were prepared by weighing appropriate quantities of distilled water and sucrose into a beaker and heating to 55-60°C. The hot sugar solution was stirred, without turbulence, using a motorized paddle and the HPMC powder carefully added to the side of the vortex. The solution was then cooled, with continual stirring, to 4°C. A flavour concentrate was prepared by mixing 800 µl strawberry flavour with 200 µl of carmoisine food colour in a 10 ml volumetric flask and making up to volume with 100% absolute ethanol. The flavour concentrate was added to a pre-weighed quantity of the cooled viscous solution such that the final concentration was 200 p.p.m. This was mixed using a roller bed (SRT2; Stuart Scientific, Redhill, UK) for 6-10 h prior to ingestion by the panel. The carmoisine acted as a marker for complete mixing.
Experimental design
Samples were presented in a randomized complete block design. Each assessor
consumed all eight samples in duplicate. The presentation order was randomized
using simple random number generation in order to reduce sample order effects.
Samples were presented as groups of three to minimize sensory fatigue.
Sensory panel training
A group of 13 trained assessors was selected on the basis of their sensory
acuity, in particular their ability to distinguish between concentrations of
the same stimulus and their ability to perform magnitude estimation
(Stevens, 1957;
Moskowitz, 1977).
Sensory evaluation
A trained sensory panel used magnitude estimation with a controlled modulus
to rate the intensity of sweetness and strawberry flavour for each of the
prepared samples. The modulus, or reference, which contained 0.25 g/100 g
HPMC, 2 g/100 g sugar and 200 p.p.m. strawberry flavour, was assigned an
arbitrary score of 100. The sweetness and strawberry flavour intensities of
each sample were rated relative to the perceived intensity of the modulus.
Assessments were carried out in individual booths designed to international
standards (ISO 8589—Design of Sensory Test Facilities) with northern
hemisphere daylight lighting at 750-1070 lux.
Samples were presented at room temperature (18-23°C) in sealed containers. Assessors were instructed to place a level dessert spoonful (10 ml) into the mouth, to allow the liquid to pass over the tongue and to swallow. They were advised not to hold the sample in the mouth for longer than a few seconds, as it would become diluted with saliva and make rating difficult. A break of 15 min was given between each set of three samples to prevent fatigue. Plain crackers and still mineral water were used as palate cleansers between each sample.
Instrumental analysis—volatile release during consumption.
The release of ethyl butyrate onto the breath was measured using the MS
NoseTM interface fitted to a platform LCZ mass spectrometer (Micromass,
Manchester, UK). Ethyl butyrate was selected as a marker for the strawberry
flavour, which contained several fruit esters with similar release profiles.
Each assessor consumed all eight samples in a single session, with a break of
at least 15 min between each sample. Plain crackers and water were used as
palate cleansers. The method of consumption was standardized; assessors were
asked to take a normal breath in, place 10 ml of sample in their mouth and
close, place their nose over the sampling tube, swallow the liquid and exhale
normally, thereafter continuing to breath regularly into the tube. The
sampling tube, which was attached to the MS Nose transfer line, allowed
exhaled air to be sampled in real time at a rate of 30 ml/min. Volatile
molecules were ionized (4 kV corona discharge, sample cone voltage 18 V) and
the volatile release followed by monitoring the appropriate MH+ ion
(ethyl butyrate: m/z 117, dwell time 0.05 s). The concentration of
ethyl butyrate on the first and second breaths was determined against the
signal from an ethyl butyrate standard in hexane
(Taylor et al.,
2000).
Rheological studies
Seventeen samples were prepared at HPMC concentrations of 0.025, 0.05,
0.075, 0.1, 1.15, 1.2, 0.25, 0.3, 0.35, 0.4, 0.45, 0.5, 0.6, 0.8, 1.0, 1.5 and
2.0 g/100 g. They were prepared using the method detailed for experiment 1,
without addition of sugar and flavour. The flow characteristics of each
solution were determined usign a CS10 Controlled Stress Rheometer (Bohlin
Instruments, Lund, Sweden) at 25°C, for a range of shear rates (5-100
s-1). Double gap geometry was used for low concentrations of HPMC,
whereas cone and plate geometry was used for the higher concentrations. For
each sample, the viscosity at zero shear [
0], was extrapolated from the
data and used to produce a Huggins-Kraemer plot from which the intrinsic
viscosity [] could be calculated. The value of c* was
then estimated from a plot of log(specific viscosity) versus
log(c*[]) (Figure
2).
|
Experiment 2—effect of viscosity on release and perception of almond flavour
Sample preparation
In a further experiment, samples were prepared containing HPMC at
concentrations of 0, 0.3, 0.6, 0.9 and 1.2 g/100 g. At each concentration of
HPMC, samples were prepared containing 2, 5 and 8 g/100 g sucrose. For each
combination of HPMC and sucrose, samples were prepared containing 10, 55 and
100 p.p.m. benzaldehyde (Firmenich SA, Geneva, Switzerland). This produced a
total of 45 samples (Figure
1).
|
Low, medium and high intensity flavour concentrates were prepared by mixing 40, 220 and 400 µl of benzaldehyde with 200 µl of carmoisine in a 10 ml volumetric flask and making up to volume with 100% absolute ethanol. The appropriate flavour concentrate was added to a pre-weighed quantity of the cooled viscous solution such that the final concentration was 10, 55 or 100 p.p.m. This was mixed for 6-10 h prior to ingestion by the panel.
Experimental design
A three factorial response surface design was used to investigate the
effect of HPMC, sugar and volatile concentration on the perception of
sweetness and almond flavour, and the release of benzaldehyde on the breath.
The experiment was designed with the aid of Design Expert 5.0 (Statease,
Minneapolis, MN; Figure 1).
Within the experimental design, samples containing 0, 0.3, 0.9 and 1.2 g/100 g HPMC were duplicated, samples containing 0.6 g/100 g were replicated four times and the centre point (0.6 g/100 g HPMC, 5 g/100 g sugar, 55 p.p.m. benzaldehyde) was replicated an additional 24 times. This resulted in a grand total of 132 samples presented overall. The design was split into 12 blocks, each containing 11 samples. Of these 11 samples, nine were of different composition and two were replicate samples of the centre point. The samples selected for any single block were orthogonal, thus creating a design in which the variables were not correlated with each other or with the blocks. This is important as it allows the results to be modelled using independently assessed design variables. Each block represented the set of samples presented to any one assessor. The orthogonality and blocking structure allowed any variation in results due to assessors to be separated from the main effects and residual error when analysing the data. All 11 samples in a block were prepared separately, including the two centre point replicates. Any variation in these results provided a measure of pure error for the experiment.
Sensory panel training
Due to the complexity of this experiment, the panel was given additional
training in magnitude estimation of sweet and almond flavour solutions. This
involved familiarizing the panel with sugar solutions of differing
concentrations (1, 2, 3, 4.5, 5, 6.5 and 8 g/100 ml) and then, in a further
exercise, asking individuals to score their perceived intensity of sweetness
against a modulus, given an arbitrary score of 100. The samples were presented
randomly, in triplicate and included internal references. This exercise was
repeated using solutions containing a fixed concentration of sugar (2 g/100
ml) but differing concentrations of benzaldehyde (10, 55, 75, 100 and 200
p.p.m.), with assessors asked to score sweetness and almond flavour (results
not shown).
Sensory evaluation
The panel used magnitude estimation with a controlled modulus to rate the
intensity of sweetness and almond flavour for each sample within their block.
The modulus, or reference, which contained 0.6 g/100 g HPMC, 5 g/100 g sugar
and 55 p.p.m. benzaldehyde, was assigned an arbitrary score of 100. The
sweetness and almond flavour intensities of each sample were rated relative to
the perceived intensity of the modulus. The tasting protocol was as described
for experiment 1.
Static equilibrium headspace
The concentration of benzaldehyde in the headspace at static equilibrium
was determined for the 45 almond flavour samples. Approximately 100 ml of each
sample were placed in a 250 ml bottle (Schott bottle; Fisher Scientific,
Loughborough, UK). Samples were allowed to equilibrate for 60 min at room
temperature (22°C), after which the headspace was sampled using the MS
NoseTM fitted to a platform LCZ mass spectrometer (Micromass, Manchester,
UK). The headspace was sampled at a rate of 10 ml/min. Compounds present in
the gas phase were ionized (4 kV corona discharge, sample cone voltage 18 V)
and the resulting MH+ ion was monitored (benzaldehyde: m/z
107, dwell time 0.05 s). Headspace concentrations were calibrated against a
signal from a benzaldehyde standard in hexane at 100 p.p.b.v.
(Taylor et al.,
2000).
Instrumental analysis—volatile release during consumption
The release of benzaldehyde onto the breath was measured using the MS
NoseTM interface fitted to a platform LCZ mass spectrometer (Micromass,
Manchester, UK), as detailed in experiment 1. Each assessor consumed all 11
samples in a single session with a break of at least 15 min between each
sample. Volatile molecules released on the breath were ionized (4 kV corona
discharge, sample cone voltage 18 V) and the volatile release followed by
monitoring the appropriate MH+ ion (benzaldehyde: m/z 107,
dwell time 0.05 s). The concentration of benzaldehyde on the first breath was
determined against the signal from a benzaldehyde standard in hexane
(Taylor et al.,
2000).
|
Results and discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
Determination of c* for HPMC
Rheological studies of the HPMC solution confirmed that c* (the point of random coil overlap), occurred at a concentration 0.57 g/100 g (Figure 2).
The effect of viscosity on volatile release and perception of strawberry flavour and sweetness intensity
Analysis of variance (two factor, repeated measures, with interaction) showed a significant difference in perceived strawberry flavour intensity and perceived sweetness intensity between samples containing increasing concentrations of HPMC (P < 0.001). Fisher's LSD (P = 0.05) showed that, for strawberry intensity, samples containing >0.5 g/100 g HPMC were significantly different to all others, whereas lower concentrations were not significantly different (Table 1). Similarly, for sweetness intensity, many significant differences were evident between samples containing increasing concentrations of HPMC. Generally, the higher the thickener concentration, the more differences were observed (Table 2).
|
|
Results also showed a significant difference between assessors (P < 0.001) and a significant interaction between samples and assessors (P < 0.001) for flavour and sweetness. Despite the use of a controlled modulus, individuals used a varying range of scale values to score the flavour properties. These differences show a lack of consistency across the panel and may be due to a poor understanding of `strawberry flavour' and `sweetness', or confusion associated with experiencing different viscosities in mouth.
The results for strawberry flavour and sweetness intensity (arbitrary
units) were averaged for the panel and plotted against HPMC concentration
(g/100 g). Whilst averaging sensory data is never recommended, in this context
it was used to illustrate the general trend in the data. Initially, perception
of strawberry flavour is constant below a HPMC concentration of 
0.5 g/100
g, after which point the perception of flavour intensity decreases steadily
with increasing HPMC concentration (Figure
3). The minimum concentration of HPMC at which flavour perception
is reduced is consistent with the value of c*, determined
to be 0.57 g/100 g.
|
The results for sweetness intensity (Figure 4) showed a similar reduction with increased HPMC concentration; however, the intensity tended to decrease steadily rather than show a sharp decline at the concentration corresponding to c*.
|
This pattern of results is similar to those obtained in previous studies
(Morris et al.,
1981). To explain the decrease in perception, Morris hypothesized
that, at c*, the hydrocolloid molecules begin to overlap
and, as a result, there is a decrease in volatile mobility through the matrix.
As discussed previously, the decreased mobility would affect the dynamics of
volatile flavour release and, since eating and swallowing are dynamic
processes, we would expect to see a reduction in concentration of volatile on
the breath.
Analysis of variance (two factor, repeated measures, without interaction) showed no significant effect of HPMC concentration on the release of ethyl butyrate onto the breath. Large differences were seen between assessors, reflected in a significant difference in their results (P < 0.001; Figure 5). These are a consequence of differing physiology (e.g. size and shape of buccal cavity, size and movement of tongue, diameter of airway, size of nasal cavity) and are a common feature of flavour release studies involving human subjects.
|
Furthermore, neither the concentration of ethyl butyrate released on the
second breath after swallowing nor the persistence of ethyl butyrate (ratio of
first to second breaths) were affected by the concentration of HPMC. This is
consistent with previous studies, which have shown no effect of HPMC on the
persistence of several volatile compounds, regardless of physicochemical
properties (Linforth and Taylor,
2000).
The effect of viscosity on release and perception of almond flavour
Static equilibrium headspace
Static equilibrium headspace concentrations of benzaldehyde were calculated
for each sample. There was no significant effect of HPMC or sugar
concentration on the headspace concentration of benzaldehyde. As expected,
there was a significant effect of volatile concentration on the headspace
values (P < 0.001). For illustration
(Figure 6), headspace
concentrations (mg/m3) were averaged across the different sugar
concentrations to give a mean result for each volatile level in 0, 0.6 and 1.2
g/100 g HPMC. The lack of an effect due to HPMC concentration suggested that
no binding or chemical interaction occur between the hydrocolloid and volatile
molecule.
|
Sensory Perception and In-nose Volatile Release
The data for perceived almond intensity, sweetness intensity and
benzaldehyde release were analysed using multiple linear regression (Design
Expert 6.0). Low-order polynomial models were derived to explain the variation
in the data and to predict volatile release (equation 1), sweetness intensity
(equation 2) and almond flavour intensity (equation 3) in terms of sample
composition.
 | (1) |
 | (2) |
0.0001), with adjusted
R2 and predicted R2 values of 0.97 and
an `adequate precision' of 57.42. The interaction term indicates that the
relationship between sweetness and sugar concentration is dependent on HPMC
and, conversely, that the relationship between sweetness and HPMC
concentration is also dependent on sugar level.
|
A further illustration of the model for perceived sweetness intensity is
shown in Figure 8. This graph
represents a slice through a three-dimensional model at benzaldehyde = 55
p.p.m. Each contour represents a sweetness value (60, 80, 100, 120, etc.). As
would be expected, the contour `sweetness = 100' passes through the point 0.6%
HPMC, 5% sugar and 55 p.p.m. benzaldehyde. The ability of the assessors to
rate a blind coded sample (identical to the modulus) as `100', gives a good
indication of their consistency. The shapes of the contours indicate that, for
any given sweetness intensity, the concentration of sugar must be increased to
compensate for an increase in thickener. The model for predicting sweetness
was robust and well described the variation in the results. Predicted values
from the model plotted against the experimental values gave an
R2 = 0.97 (Figure
9)
 | (3) |
0.0001). It included terms for HPMC, sugar and
benzaldehyde concentration with quadratic terms for each variable and an
interaction between sugar and benzaldehyde. All terms included had a
significant effect on the model. The adjusted and predicted
R2 values for the model were 0.89 and 0.85, respectively
and the `adequate precision' was 27.85. The model for predicting almond
flavour was robust and well described the variation in the results. Predicted
values from the model plotted against the experimental values gave an
R2 = 0.97 (Figure
10).
|
|
|
The inclusion of the interaction term suggests that, for any given level of HPMC, the relationship between perceived almond intensity and volatile concentration is dependent on sugar level. A further illustration of the model is shown in Figure 11. Each contour represents almond flavour intensity; the contour shape illustrates the effect of HPMC and sugar concentration at 55 p.p.m. benzaldehyde. For HPMC values >0.5% and for any given almond flavour intensity, the sugar level can be increased to maintain perceived flavour. This holds true until a level of 6-6.5% sugar, after which point an increase in sugar results in a decrease in flavour perception. This effect is most dramatic at low levels of HPMC and may, in part, be due to the intense sweetness masking the almond flavour.
|
Interactions between volatile and non-volatile stimuli are well documented
(Noble, 1996). Davidson et
al. (Davidson et al.,
1999) reported that a decrease in perception of mint flavour
correlated closely with decrease in sugar release from chewing gum, despite
constant delivery of mint volatiles to the nasal receptors. It follows,
therefore, that the decrease in flavour perception observed from these results
may be due to the effect of HPMC on stimulation of taste receptors by sugar
molecules, rather than volatile stimulation of nasal receptors.
One possible explanation may be the effect of HPMC on the mobility of free
water in solution (particularly at concentrations >c*).
Studies (Mathlouthi, 1984;
Mathlouthi et al.,
1986; Mathlouthi and Seuvre,
1988) have shown that sweetness increases as water mobility
increases. Conformational changes in sucrose molecules in solution enhance
sweetness intensity. Furthermore, dissociation of free water molecules
arranged around the periphery of the sugar molecule produces a high membrane
potential across the taste cell, thereby enhancing sweetness perception.
A possible alternative hypothesis is that the perception of viscosity
itself affects overall flavour perception. Interactions may occur at a
neurological level where gustatory and trigeminal inputs converge, or even at
a perceptual level where previous dietary experiences could influence taste
judgements in thick and thin solutions
(Christensen, 1980).
|
Conclusion |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
The perception of flavour and sweetness is greatly reduced when HPMC is added to sugar/flavour solutions at concentrations >c*. However, the concentration of volatile released onto the breath is not affected by the increase in viscosity. Significant, robust statistical models were derived to describe the results and to predict the intensity of perceived flavour, sweetness and the release of volatile from the thickened liquids.
A possible explanation for the decrease in perception may be the effect of increasing HPMC on the free water available in solution, resulting in a decrease in sweetness intensity and, therefore, a decrease in flavour intensity. Investigation of this would require nuclear magnetic resonance studies of the mobility of water and conformation of sweeteners in thickened solutions.
Alternatively, the perception of a thickened solution in the mouth may have an impact on the perception of tastants and, consequently, overall flavour.
|
Acknowledgments |
|---|
The authors wish to thank Firmenich SA Geneva, Nestlé Lausanne, Mars UK and BBSRC for supporting this work.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResults and discussionConclusionReferences |
|---|
Baines, Z.V. and Morris, E.R. (1987) Flavour/taste perception in thickened systems: the effect of guar gum above and below c*. Food Hydrocoll.,1 , 197-205.
Baines, Z.V. and Morris, E.R. (1988) Effects of polysaccharide thickeners on organoleptic attributes. In Phillips, G.O., Williams, P.A. and Wedlock, D.J. (eds), Gums and Stabilisers for the Food Industry 4. IRL Press, Oxford, pp.192 -201.
British Standards Institute (1975)British Standards Glossary of Terms Relating to the Sensory Analysis of Food . British Standards Institute, London.
Christensen, C.M. (1980) Effects of solution viscosity on perceived saltiness and sweetness. Percept. Psychophys., 28,347 -353.[Web of Science][Medline]
Davidson, J.M., Hollowood, T.A., Linforth, R.S.T. and Taylor, A.J. (1999) The effect of sucrose on the perceived flavour intensity of chewing gum. J. Agric. Food Chem., 47,4336 -4340.[Web of Science][Medline]
De Roos, K.B. (1997) How lipids influence food flavour. Food Technol., 51,60 -62.
Linforth, R.S.T. and Taylor, A.J. (2000) Persistence of volatile compounds in the breath after their consumption in aqueous solutions. J. Agric. Food Chem.,48 , 5419-5423.[Web of Science][Medline]
Malkki, Y., Heinio, R.L. and Autio, K. (1993) Influence of oat gum, guar gum and carboxymethyl cellulose on the perception of sweetness and flavour. Food Hydrocolloids,6 , 525-532.
Mathlouthi, M. (1984) Relationship between the structure and the properties of carbohydrates in aqueous-solutions. 2. Solute solvent interactions and the sweetness of D-fructose, D-glucose and sucrose in solution. Food Chem.,13 , 1-16.
Mathlouthi, M. and Seuvre, A.M. (1988) Solution properties and the sweet taste of small carbohydrates.J. Chem. Soc. Faraday Trans. I , 84,2641 -2650.
Mathlouthi, M., Cholli, A.L. and Koenig, J.L. (1986) Spectroscopic study of the structure of sucrose in the amorphous state and in aqueous solution. Carbohydr. Res.,147 , 1-9.
Morris, E.R. (1987) Organoleptic properties of food polysaccharides in thickened systems. In Yalpani, M. (ed.),Industrial Polysaccharides: Genetic Engineering, Structure/Property Relations and Applications . Elsevier Science, Amsterdam, pp225 -238.
Morris, E.R., Cutler, A.N., Ross-Murphy, S.B., Rees, D.A. and Price, J. (1981) Concentration and shear rate dependence of viscosity in random coil polysaccharide solutions.Carbohydr. Polymers , 1,5 -11.
Moskowitz, H.R. (1977) Magnitude estimation: notes on what, how and why to use it. J. Food Qual., 1,195 -228.
Moskowitz, H.R. and Arabie, P. (1970) Taste intensity as a function of stimulus concentration and solvent viscosity. J. Texture Stud., 1,502 -510.
Noble, A.C. (1996) Taste—aroma interactions. Trends Food Sci. Technol.,7 , 439-444.
Pangborn, R.M., Trabue, I.M. and Szczesniak, A.S. (1973) Effect of hydrocolloids on oral viscosity and basic taste intensities. J. Texture Stud.,4 , 224-241.
Paulus, K. and Haas, E.M. (1980) The
influence of solvent viscosity on the threshold values of primary tastes.Chem. Senses
, 5,23
-32.
Roberts, D.D., Elmore, J.S., Langley, K.R. and Bakker, J. (1996) Effects of sucrose, guar gum, and carboxymethylcellulose on the release of volatile flavour compounds under dynamic conditions. J. Agric. Food Chem.,44 , 1321-1326.
Stevens, S.S. (1957) On the psychophysical law. Psychol. Rev., 64,153 -181.[Web of Science][Medline]
Taylor, A.J., Linforth, R.S.T., Harvey, B.A. and Blake, A. (2000) Atmospheric pressure chemical ionisation for monitoring of volatile flavour release in vivo. Food Chem., 71,327 -338.
Vaisey, M., Brunon, R. and Cooper, J. (1969) Some sensory effects of hydrocolloid sols on sweetness. J. Food Sci., 34,397 -400.
Accepted April 5, 2002
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D Labbe, A Rytz, C Morgenegg, S Ali, and N Martin Subthreshold Olfactory Stimulation Can Enhance Sweetness Chem Senses, April 2, 2007; (2007) bjl040v2. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. E. Hayes and V. B. Duffy Revisiting Sugar-Fat Mixtures: Sweetness and Creaminess Vary with Phenotypic Markers of Oral Sensation Chem Senses, March 1, 2007; 32(3): 225 - 236. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Frasnelli, S. van Ruth, I. Kriukova, and T. Hummel Intranasal Concentrations of Orally Administered Flavors Chem Senses, September 1, 2005; 30(7): 575 - 582. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. J. Cook, T. A. Hollowood, R. S.T. Linforth, and A. J. Taylor Oral Shear Stress Predicts Flavour Perception in Viscous Solutions Chem Senses, January 1, 2003; 28(1): 11 - 23. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||