Chemical Senses Vol. 29 No. 7 © Oxford University Press
2004; all rights reserved
‘Thermal Taste’ Predicts Higher Responsiveness to Chemical Taste and Flavor
1 The John B. Pierce Laboratory, New Haven, CT 06519, USA and 2 Department of Surgery (Otolaryngology), Yale School of Medicine, New Haven, CT 06510, USA
Correspondence to be sent to: Barry G. Green, The John B. Pierce Laboratory, 290 Congress Avenue, New Haven, CT 06519, USA. e-mail: green{at}jbpierce.org
 |
Abstract |
|---|
 Top Abstract IntroductionGeneral methodsExperiment 1: perception of...Experiment 2: thermal perception...Experiment 3: taste...Experiment 4: taste...Experiment 5: whole mouth...DiscussionAcknowledgementReferences |
|---|
Individual differences in taste perception have been explained in part by variations in peripheral innervation associated with the genetic ability to taste the bitter substances PTC and PROP. In the present study we report evidence of another source of individual differences that is independent of taste stimulus, taste quality, or gustatory nerve. Individuals who perceived taste from thermal stimulation alone (thermal taste) gave significantly higher taste ratings to chemical stimuli—often by a factor of >2:1—than did individuals who perceived no taste from thermal stimulation. This was true for all taste stimuli tested (sucrose, saccharin, sodium chloride, citric acid, quinine sulfate, MSG and PROP), for all three gustatory areas of the mouth (anterior tongue, posterior tongue and soft palate) and for whole-mouth stimulation. Moreover, the same individuals reported stronger sensations from the olfactory stimulus vanillin, particularly when it was sensed retronasally. The generality of the thermal-taster advantage and its extension to an olfactory stimulus suggests that it arises from individual differences in CNS processes that are involved in perception of both taste and flavor.
Key words: human, individual differences, psychophysics, retronasal olfaction, taste
|
Introduction |
|---|
|
TopAbstractIntroductionGeneral methodsExperiment 1: perception of...Experiment 2: thermal perception...Experiment 3: taste...Experiment 4: taste...Experiment 5: whole mouth...DiscussionAcknowledgementReferences |
|---|
Striking individual differences in taste perception were first demonstrated in experiments with the bitter-tasting chemicals phenylthiocarbamide (PTC) and 6-n-propylthiouracil (PROP) (Fox, 1931
;
Harris and Kalmus, 1949b;
Fischer and Griffin, 1963;
Hall et al., 1975) and
subsequent studies uncovered associated differences in perception of other taste stimuli
as well (Gent and Bartoshuk, 1983;
Bartoshuk et al., 1988, 1994;
Bartoshuk, 1993;
Lucchina et al., 1998;
Prescott et al., 2001). PROP
sensitivity has been shown to be correlated with the number and density of taste buds and
taste pores on the anterior tongue (Miller and
Reedy, 1990) and sensitivity to PTC was recently linked to a gene that
controls expression of a specific taste receptor protein (Kim et al., 2003). However, the failure in some
studies to find a strong association between sensitivity to PROP/PTC and other tastes
(e.g.
Harris and Kalmus, 1949a;
Schifferstein and Frijters, 1991;
Yokomukai et al., 1993;
Delwiche et al., 2001;
Horne et al., 2002) has
implied that additional, unknown factors contribute to individual differences in taste
perception.
We recently uncovered evidence of a source of individual differences that appears
unrelated to genetic expression of a specific taste receptor or to differences in
innervation density of the chorda tympani nerve. The evidence came unexpectedly from a
test of the hypothesis that the ability to perceive sweetness from thermal stimulation
alone (thermal taste;
Cruz and Green, 2000) might be
associated with a high responsiveness to sweet-tasting carbohydrates, such as sucrose.
Thermal sweetness is the most common of the thermally induced tastes experienced on the
tongue tip and is perceived by 
50% of individuals when the tongue is rapidly
re-warmed after being briefly cooled to 15–20°C. The mechanism of thermal taste
is unknown, though it has been hypothesized to result from a temperature-sensitive
process related to chemical taste transduction (Cruz and Green, 2000). Prior to discovery of thermal taste,
the principal effect of temperature on taste that had been observed was a modulation of
perceived sweetness and bitterness: both tastes were reduced by cooling and enhanced by
warming (Stone et al., 1969;
Paulus and Reisch, 1980;
Bartoshuk et al., 1982;
Calvino, 1986;
Green and Frankmann, 1987). However,
a subsequent study showed that the enhancement occurred for the sweetness of sucrose,
glucose or fructose, but not for the sweetness of saccharin (Green and Frankmann, 1988). The greater temperature
dependence of carbohydrate sweetness raised the possibility that thermal sweetness is a
byproduct of a heat-induced increase in the excitability of taste receptors or cells that
are more sensitive to carbohydrates than to saccharin. Arguing against this possibility
is the recent evidence that the sensitivity to carbohydrate and noncarbohydrate
sweeteners depends upon the same two taste receptor proteins (Adler et al., 2000;
Max et al., 2001;
Zhao et al., 2003;
Inoue et al., 2004).
However, the greater thermal lability of sucrose sweetness, together with earlier
psychophysical evidence of asymmetric cross-adaptation of sweetness between saccharin and
sucrose (McBurney, 1972;
Lawless and Stevens, 1983), suggested
that differences may nevertheless exist in the way the two classes of sweeteners are
transduced. We therefore hypothesized that perception of thermal sweetness caused by
warming should be correlated with a higher responsiveness to sucrose compared to
saccharin. The results of the first experiment, which indicated that individuals who
perceived thermal sweetness were more responsive to both sweeteners as well as to a third
taste stimulus, led to four additional experiments designed to discover more about the
nature and extent of the association between thermal taste and perception of chemical
taste and flavor. The results suggest that in addition to previously identified
peripheral factors, variation in central neural processes may also contribute to
individual differences in perception of both taste and flavor.
|
General methods |
|---|
|
TopAbstractIntroductionGeneral methodsExperiment 1: perception of...Experiment 2: thermal perception...Experiment 3: taste...Experiment 4: taste...Experiment 5: whole mouth...DiscussionAcknowledgementReferences |
|---|
Subjects
A total of 95 subjects (63 females and 32 males) between the ages of 18 and 45 (average age = 22.8 years) participated in the study. Of this total, 51.8% of females and 34.6% of males were categorized as ‘thermal tasters’ (TTs). The subjects were recruited from public postings on the Yale University Medical School and Yale College campuses. All were self-reported healthy nonsmokers who had no known taste or smell disorders or deficiencies and who were not taking prescription pain or allergy medication. Each person gave informed consent and was paid for their participation. Approximately one-third of the subjects served in two or more of the experiments, though none were aware of the hypotheses under test.
Thermal stimulation
Temperature stimuli were delivered to the oral test sites using a 0.64 cm2
Peltier thermoelectric module (referred to hereafter as the thermode), designed and built
in the Pierce Laboratory Machine and Electronics shop. The thermode was epoxied to the
top of a closed-loop, water-circulated heat sink that also served as a handle for
positioning and holding the thermode against the tongue. Temperature was computer
controlled and monitored via a 40 ga copper constantan thermocouple recessed in the
stimulating face of the Peltier module. The base and target temperatures were set by the
experimenter on each trial and the rate of temperature change was constant at
±1.5°C/s. The thermode was tightly wrapped in clean cellophane before
each experimental session to provide hygienic protection without significantly altering
thermal conduction. A second, larger Peltier thermode (4.84 cm2) powered from
the same device and temperature-controlled in the same manner as the oral thermode, was
used to measure thermal responsiveness of the hand in experiment 2.
Chemical stimulation
The chemical stimuli were prepared as aqueous solutions using deionized water. In
experiment 1, 3 and 4 the stimuli were delivered using sanitary 6 in. cotton tipped swabs
(McKesson) saturated with the test solution just prior to application. The swabs were
rubbed gently onto the test site for 5 s. In experiment 5, which used a sip-and-spit
procedure, the stimuli were delivered in 10 ml samples that were sipped from medicine
cups, gently swished in the mouth for 5 s, then expectorated. In all cases subjects
rinsed at least twice between trials with 37°C deionized water after each taste
stimulus.
Practice procedure and thermal taste screening
Every experiment began with a practice session that acquainted subjects with the
general Labeled Magnitude Scale (gLMS;
Green et al., 1993, 1996;
Bartoshuk et al., 2003) and
gave them experience using it to rate the strength of taste sensations prior to actual
data collection. Subjects were read instructions about how to use the computerized gLMS
which emphasized that intensity ratings were to be made relative to the strongest
imaginable stimulus of any kind. Ratings were made using a mouse to move a cursor along
the scale to the desired location before clicking the mouse to register the response.
After these initial instructions, subjects were asked to use the gLMS to rate a wide
range of imagined oral sensations that are commonly experienced in daily life (e.g. the
sweetness of cotton candy; the burn of cinnamon gum). This part of the practice session
emphasized that intensity ratings should be made in the context of normal oral sensations
and relative to the strongest imaginable sensation of any kind. Subjects were then given
a practice series of actual taste stimuli (1.0 mM QSO4, 0.056 M citric acid,
1.0 M sucrose, 0.56 M sodium chloride, 5.6 mM saccharin and dH2O) which they
rated on four separate gLMS scales (for sweetness, sourness, saltiness and bitterness)
presented successively on a computer monitor. These warm-up stimuli provided practice
with actual tastes before subjects were tested for possible perception of thermal
taste.
Following the practice stimuli, subjects were given a 5 min break during which they
rinsed the mouth repeatedly to eliminate residual tastes. We then tested the subjects for
perception of thermal taste on the tongue tip. Under the guidance of the experimenter,
the subject used a mirror to place the stimulating surface of the thermode (set to
35°C) against the tongue tip. Once in place, the temperature of the thermode was
decreased to 15°C and then immediately re-warmed to 35°C (warming trial).
Subjects were instructed to ‘attend now’ as soon as warming began and then to
‘rate’ the taste sensation (if one had been perceived) when the thermode
reached 35°C. Ratings were made on the gLMS exactly as they had been with the
chemical stimuli. If no taste sensations were perceived, subjects were instructed to rate
‘no sensation’ for all four taste qualities. After completion of the warming
trial the thermode was reapplied to the tongue tip and cooled to 15°C, where it was
held constant for 10 s and the taste ratings repeated. Finally, the thermode was applied
a third time at 9°C and a final set of taste ratings were made. Because thermal taste
is not always perceived to be strongest at the tongue tip, the stimuli were presented
sequentially to three sites on the anterior edge of the tongue: the tongue tip followed
by two sites 1 cm to the left and right of the midline. Warming trials always
preceded cooling trials to avoid possible contamination from adaptation due to the
intense, sustained cold stimulation.
To avoid inducing a potential bias toward reporting thermal taste, subjects were told that not everyone perceives taste during thermal stimulation of the tongue and that we were interested in testing people who do not perceive tastes as well as those who do. Subjects were classified as TTs if they reported a taste that was at least ‘weak’ in intensity on the gLMS during either warming or cooling. When a taste was reported, the site was retested to confirm the reliability of the sensation. In experiments with multiple sessions, subjects were reclassified as ‘thermal non-tasters’ (TnTs) and their data analyzed accordingly if they failed to report thermal taste during testing sessions. This occurred for <10% of the subjects who were initially identified as TTs; no subjects who reported no thermal taste at the time of screening reported thermal taste in later sessions.
Statistical analyses
Perceived intensity ratings were averaged arithmetically across replicates (when they occurred). Because data from the gLMS is typically log-normally distributed across subjects, the within-subject means were converted to log10 before calculating across-subjects means and conducting parametric statistics. Prior to converting to logs, all zeros (ratings of ‘no sensation’) were replaced with 0.24, the lowest rating possible (one pixel) on the computerized gLMS. All data analyses were carried out using StatisticaTM 6.1. Differences between groups and across stimuli and concentrations were assessed using repeated-measures MANOVA’s with multiple factors. The default significance level was P < 0.05. All statistical correlations were calculated using Pearson rs.
|
Experiment 1: perception of sucrose and saccharin |
|---|
|
TopAbstractIntroductionGeneral methodsExperiment 1: perception of...Experiment 2: thermal perception...Experiment 3: taste...Experiment 4: taste...Experiment 5: whole mouth...DiscussionAcknowledgementReferences |
|---|
This experiment was designed to test the hypothesis that perception of thermal sweetness was associated with a higher responsiveness to sucrose compared to saccharin. Measurements were therefore made of the perceived intensity of taste for sucrose and saccharin, as well as for a third, non-sweet-tasting stimulus, NaCl, which was included as a control stimulus.
Procedure
Following the screening and practice session, 28 subjects (14 TTs and 14 TnTs) returned for two separate experimental sessions. At the beginning of each session subjects received three concentrations of three chemical taste stimuli: 1.0, 0.32 and 0.1 M sucrose (J.T. Baker); 0.56, 0.18 and 0.056 M NaCl (Sigma); and 5.6, 1.8 and 0.56 mM sodium saccharin (Janssen). The stimuli were applied in one of six pseudorandom sequences to the area of the tongue previously found to be most sensitive to thermal taste (TnTs were always tested on the tongue tip). Immediately following each stimulus application subjects rated the peak intensity of sweet, salty, sour and bitter sensations. Between trials subjects rinsed at least twice with de-ionized water to remove residual tastes. After rating the taste stimuli, subjects rinsed extensively to remove any residual stimulus and were given a 10 min break before being tested again in the thermal taste paradigm. After completing the thermal taste series, a 5 min break was given before a final thermal warming stimulus (15–35°C) was given to obtain an intensity rating for perceived warmth. This rating was used as an additional control stimulus to see if TTs rated warmth intensity higher than did TnTs. Warmth intensity was rated on a separate trial because we sought to make the temperature rating as independent as possible of thermal taste.
Results
Contrary to the hypothesis that TTs would perceive more intense sweetness compared to TnTs for sucrose but not saccharin, TTs rated the tastes of all three gustatory stimuli significantly higher than did TnTs [main effect of group, F(1,26) = 21.0, P < 0.001)]. Figure 1 shows that TTs rated the sweetness of sucrose and the saltiness of NaCl an average of 2.1 times stronger (an average difference in log-means of +0.32). The difference in ratings was even greater for saccharin [group x stimulus interaction, F(2,52) = 8.6, P < 0.001], which TTs rated an average of 3.7 times sweeter (+0.57 log units) than did TnTs (Figure 1b). TTs also rated warmth sensations significantly higher [t-test for independent groups, t(26) = 3.7, P < 0.005] than did TnTs (Figure 1d), but the size of the difference between groups was much smaller than the differences for the three taste stimuli.
|
|
Experiment 2: thermal perception at nongustatory sites |
|---|
|
TopAbstractIntroductionGeneral methodsExperiment 1: perception of...Experiment 2: thermal perception...Experiment 3: taste...Experiment 4: taste...Experiment 5: whole mouth...DiscussionAcknowledgementReferences |
|---|
The purpose of this experiment was to test whether TTs also rated nongustatory sensations (i.e. thermal) higher than TnTs. The across the board higher ratings of taste and temperature by TTs led us to consider the possibility that the difference between groups resulted from a response bias rather than from a true difference in sensory perception (Ekman et al., 1968
). In particular, it was possible that TTs rated all sensations higher on
the response scale (the gLMS;
Green et al., 1993, 1996;
Bartoshuk et al., 2003),
regardless of sensory modality. Procedure
To test this hypothesis we invited back 22 of the 28 subjects of experiment 1 (11 TTs
and 11 TnTs) to rate the perceived intensity of five heating (36–44°C) and five
cooling (25–5°C) stimuli on two non-lingual sites: the vermilion border of the
lip and the palm of the hand. The lip was chosen because it is a nongustatory site which,
like the tongue, is innervated by the trigeminal nerve; the hand was chosen because it is
remote from the oral cavity and served by different cutaneous nerves. The subjects’
task was to rate the intensity of thermal sensations using the gLMS with the same
instructions as experiment 1, except that no taste ratings were made. To standardize the
initial hand temperature before testing began, subjects wore a waterproof glove
(Flexigloves; Recombinant Technologies LLC) on the left hand and submerged the hand up to
the wrist for 5 min in a 30°C water bath. Subjects then removed the hand from the
bath, took off the glove and placed the hand in a cotton-lined (oven) mitten to help keep
hand temperature stable throughout testing. Because warmth sensitivity is spatially
heterogeneous (Green and Cruz, 1998),
it was important to reduce spatial differences in sensitivity as a source of
inter-individual variability by locating and testing only the most sensitive sites on the
hand. Sensitivity was surveyed by obtaining heat intensity ratings for the four quadrants
of the palm (the thenar eminence, hypothenar eminence and the pads at the base of the
first and fourth digits) in response to 5 s applications of the 4.84 cm2
thermode set to 40°C. Stimuli were separated by a 1 min inter-stimulus interval (ISI)
during which the hand was placed back inside the mitten. The two sites that yielded the
highest average intensity ratings were used as test sites in the experiment proper. The
40°C stimulus was then delivered to two sites on the lip (the left and right sides
adjacent to the midline) in successive trials using the 0.64 cm2 oral
thermode, followed by 15°C delivered to the same sites in the same manner. Finally,
the cold stimulus was applied to the four quadrants of the hand using the 4.84
cm2 thermode. In addition to identifying the most sensitive palmar sites, this
testing provided practice in rating cutaneous thermal stimuli on both the hand and
lip.
In the main part of the experiment five warm temperatures (36, 38, 40, 42 and 44°C) and five cold temperatures (25, 20, 15, 10 and 5°C) were applied to the palm of the hand and the lip. Stimuli were delivered in blocks by temperature, with the warm stimuli presented first. Within each block the temperatures were presented in one of three pseudorandom orders with a 1 min ISI. Testing was alternated between the palm and lip across trials such that stimulation occurred on the same site no more often than once every 2 min. The procedure for adjusting and stabilizing hand skin temperature was the same as in the practice session. All temperature ratings were made on the gLMS using the same instructions as for the taste measurements.
Results
There was no significant difference in thermal responsiveness between groups (Figure 2). The perceived intensity of warmth and cold did not differ between TTs and TnTs [F(1,20) = 0.07, P = 0.80] at either anatomical site (Figure 2a,b). Log-mean ratings for TnTs were actually slightly higher for both warmth and cold on the palm and for warmth on the lip, though not significantly so [group x temperature x site, F(1,180) = 1.61, P = 0.11]. Accordingly, the results rule out the possibility that the group differences in taste perception of experiment 1 were the result of a difference in scale use between the two groups.
|
|
Experiment 3: taste responsiveness of the back of the tongue |
|---|
|
TopAbstractIntroductionGeneral methodsExperiment 1: perception of...Experiment 2: thermal perception...Experiment 3: taste...Experiment 4: taste...Experiment 5: whole mouth...DiscussionAcknowledgementReferences |
|---|
After response bias was eliminated as a significant factor, we investigated whether differences in taste perception between TTs and TnTs might be limited to the tongue tip, where the criterion for thermal tasting was determined. It was possible that the heightened ability to taste both thermal and chemical taste at that site resulted primarily from more abundant taste innervation by the chorda tympani nerve, which has been implicated before as a source of individual differences in taste perception (Miller and Reedy, 1990
;
Bartoshuk et al., 1994;
Doty et al., 2001). If so,
no difference should be found between TTs and TnTs on the back of the tongue, where taste
is mediated by a different cranial nerve (glossopharyngeal;
Zotterman, 1935;
Bradley et al., 1986). Procedure
A total of 19 TTs and 27 TnTs served in the experiment. Each experimental session
began with a brief practice sequence in which subjects rated the taste intensity of five
of the six stimuli to be used in the experiment [0.18 M sucrose, 0.10 M NaCl, 0.18
mM QSO4 (J.T. Baker), 0.018 M citric acid, 0.056 mM PROP (Aldrich) and 29 mM
DL-monosodium glutamate (MSG; AccentTM)]. PROP was not given as
practice because of concerns about inducing context effects in subjects who might be
exceptionally sensitive to it (Bartoshuk et
al., 1998). MSG was added in this experiment because it is known to be
an effective taste stimulus in the back of the mouth. To accommodate the unusual taste
quality (for non-Asian subjects) of MSG, we included the response category
‘other’ in addition to the taste qualities of sweet, sour, salty and bitter.
Instructions directed subjects to use the ‘other’ category to rate any
sensations that could not be described in terms of the four qualities. The five stimuli
were applied once each via cotton-tipped swabs to the tongue-tip and to the
posterior-lateral region of the tongue (on separate trials) to stimulate the
circumvallate papillae. Stimulation of some foliate papilla on the rear edge of the
tongue may also have occurred. Following the practice block and extensive rinsing,
subjects were tested for thermal taste in the manner of preceding experiments.
Categorization as TTs and TnTs were based on the results of this screening test.
In the experimental session, subjects again received the chemical taste stimuli on both the front and back of the tongue via cotton swabs. To minimize taste adaptation, stimuli were applied to the front and back of the tongue and to the left and right sides, in alternating fashion. Thus each lingual site was stimulated on every fourth trial. This procedure was repeated until all of the chemicals were applied once to the front of the tongue and once to the back of the tongue. The stimuli were themselves presented in a pseudorandom sequence, with PROP again given last to avoid a potential context effect and complications from lingering bitterness in highly sensitive individuals.
Results
Taste responsiveness to sucrose, NaCl, citric acid, QSO4, PROP and MSG was again significantly higher for TTs [effect of group, F(1,44) = 12.4, P < 0.001], with the magnitude of the difference between groups at least as large on the back of the tongue as on the front (Figure 3). A trend toward an even larger difference on the back of the tongue fell just short of statistical significance [group x site interaction, F(1,44) = 3.7, P = 0.061]. This trend was strongest for PROP, where TTs rated its bitterness on the back of the tongue to be an average of 5.1 times stronger than did TnTs, compared to two times stronger on the tongue tip. MSG was not reliably perceived on the tongue tip in the concentration we tested (29 mM) and was only marginally perceived on the back of the tongue. Nevertheless, TTs also showed a slight tendency to rate the very weak taste of MSG higher on the back of the tongue than did TnTs (Figure 3f).
|
|
Experiment 4: taste responsiveness of the palate |
|---|
|
TopAbstractIntroductionGeneral methodsExperiment 1: perception of...Experiment 2: thermal perception...Experiment 3: taste...Experiment 4: taste...Experiment 5: whole mouth...DiscussionAcknowledgementReferences |
|---|
After replicating the group difference on the back of the tongue we decided to compare taste perception of TTs and TnTs on the soft palate, where a third taste nerve, the superficial petrosal nerve, mediates taste perception (Rollin, 1977
;
Harada et al., 1997). Procedure
A total of 12 TTs and 12 TnTs were screened and tested in the main part of the experiment. As in experiment 3, subjects were again familiarized with five of the six taste stimuli (all but PROP) during the practice session. Following practice, thermal taste was measured on the tongue tip as before to enable subjects to be grouped according to thermal tasting. The chemical stimuli (0.18 M sucrose, 0.18 M NaCl, 0.18 mM citric acid, 0.18 mM QSO4, 180 mM MSG and 0.1 mM PROP) were swabbed onto alternate sides of the soft palate across trials in pseudorandom sequence with the exception of PROP, which was always delivered last. Subjects again rated the intensity of sweetness, sourness, saltiness, bitterness and ‘other’ on the gLMS, then rinsed at least twice before the next trial. The site of stimulation was just anterior to the uvula on the roof of the mouth; exact localization of the taste-sensitive region of the palate is not otherwise possible. However, swabbing an area a few square centimeters in area proved adequate to produce reliable taste sensations in all subjects. To avoid spreading the stimulus to the tongue, subjects were asked to keep the tongue resting on the bottom of the mouth until they had completed their intensity ratings. Stimuli were presented twice each.
Results
The results showed that average log-mean ratings of taste intensity were once again higher for TTs [F(1,22) = 7.6, P < 0.05] for all six stimuli tested (group x stimulus interaction [F(1,22) = 0.66, P > 0.05, n.s.]. However, the magnitude of the TT advantage varied considerably for the different stimuli (Figure 4), ranging from a ratio of 1.8:1 for the (again) weakly perceived MSG, to 4.2:1 for NaCl. Perception of bitterness from PROP differed between groups by a factor of 2.3 to 1, which was similar to the difference found previously on the tongue tip.
|
|
Experiment 5: whole mouth taste and flavor perception |
|---|
|
TopAbstractIntroductionGeneral methodsExperiment 1: perception of...Experiment 2: thermal perception...Experiment 3: taste...Experiment 4: taste...Experiment 5: whole mouth...DiscussionAcknowledgementReferences |
|---|
The large and reliable group differences across all three gustatory regions indicated that TTs were more responsive to tastes throughout the mouth and thus should show a similar advantage in perception of whole-mouth taste stimulation. The gustatory ubiquity of the individual differences also raised the possibility that TTs might be more responsive to odors perceived retronasally. Retronasal odors are routinely misperceived as tastes and can become perceptually associated with tastes (Murphy et al., 1977
;
Frank and Byram, 1988;
Frank et al., 1993;
Stevenson et al., 1999).
This close perceptual association with taste implies that retronasal olfactory and taste
stimulation are integrated centrally. An obvious question is whether this integration
occurs at a level in the CNS that is prior to the source of the TT advantage. If so,
retronasal odors and odor-taste mixtures (i.e. flavors) should also be perceived more
strongly by TTs. Procedure
The perception of whole-mouth taste stimulation was compared for the two groups using a sip-and-spit-paradigm. The taste stimuli were 0.56, 0.32, 0.18, 0.10 and 0.056 M sucrose and 18, 10, 5.6, 3.2 and 1.8 mM citric acid. A single concentration of PROP (0.056 mM) was also included. The odorant vanillin (1.8 mM) was presented retronasally in aqueous solution by itself and with two concentrations of sucrose (0.056 and 0.56 M). A brief practice block of trials was given in which subjects were familiarized with tasting and rating the sucrose and citric acid stimuli in the form of 10 ml aqueous samples. Subjects practiced rating the sweetness, saltiness, sourness, bitterness and ‘other’ sensations on the gLMS. Just as the category ‘other’ was added in experiment 3 to accommodate the taste of MSG, it was used in this experiment to accommodate the odor of vanillin (vanilla), which some subjects might not be able to identify. Following the practice trials and water rinses, thermal taste was assessed as in previous experiments.
The procedure in the main part of the experiment was as follows: subjects tasted five
concentrations of sucrose and five concentrations of citric acid presented as 10 ml
samples in one of six pseudo-randomized orders. The order tested in a given session was
randomly selected at the beginning of each session. Each taste sample was sipped and
gently swished for 3 s, after which it was expectorated. Intensity ratings of
sweetness, saltiness, sourness, bitterness and ‘other’ were made immediately
thereafter, with instructions to base ratings on the peak intensity experienced while the
solution was in the mouth. The subjects were also instructed that while tasting the
samples it was important to inhale and exhale normally through the nose at least once.
This practice ensured retronasal delivery of vanillin on trials when it was
presented.
After all of the taste samples except PROP had been presented, vanillin was presented
as an orthonasal stimulus by holding a 5 ml sample of 1.8 mM vanillin directly under the
subjects nose for 3 s, during which time they inhaled once through the nose.
Subjects then rated the total perceived odor intensity on the gLMS. This procedure was
repeated with a fresh sample after a 1 min ISI. Subjects were unaware they were receiving
the same solution twice.
Following the orthonasal odor rating, subjects sipped the PROP solution, expectorated it and rated its taste intensity in the same manner as the other chemical tastes.
Results
Consistent with the data from local taste stimulation, TTs rated whole-mouth stimulation from sucrose and citric acid significantly stronger than did TnTs [group effect, F(1,40) = 8.0, P < 0.01]. Figure 5a,b shows that the group effect was consistent throughout the concentration range for both stimuli, although the mean difference was much larger for sucrose than for citric acid. Whole-mouth responsiveness to PROP (Figure 6) was also markedly higher for TTs [t(40) = 2.54, P > 0.05]. Perceived bitterness for the single concentration we tested was rated 3.9 times stronger by TTs than by TnTs (a log-mean difference of +0.59).
|
|
More surprising was the group difference in perception of vanillin sensed retronasally. Figure 7a shows that when vanillin was sipped and held briefly in the mouth, either by itself or in mixture with sucrose, ratings of ‘other’ (which were assumed to include the vanilla odor or flavor) were much higher for TTs than for nontasters [F(1,40) = 7.7, P < 0.01]. Sweetness ratings (Figure 7b) were also significantly higher for TTs in all three conditions [F(1,40) = 21.5, P < 0.001]. The latter outcome indicates that TTs perceived more ‘sweetness’ in the vanillin odor, which translated into higher ratings of sweetness for the mixtures as well as for vanillin itself (i.e. odor-taste enhancement; Frank et al., 1993
). TTs
also rated the odor of vanillin higher when it was sniffed orthonasally
[t(40) = 2.4, P < 0.05], although the proportional
difference between groups (1.8:1) was less than half what was found for retronasal
delivery (3.8:1).
|
|
Discussion |
|---|
|
TopAbstractIntroductionGeneral methodsExperiment 1: perception of...Experiment 2: thermal perception...Experiment 3: taste...Experiment 4: taste...Experiment 5: whole mouth...DiscussionAcknowledgementReferences |
|---|
The occurrence of the TT advantage for all taste qualities, for all three gustatory sites and for olfactory stimulation, strongly suggests that individual differences in thermal taste perception are associated with a generally higher responsiveness to both gustatory and olfactory stimulation. Assuming that thermal taste results from stimulation of a subset of temperature-sensitive gustatory fibers, individuals who are more responsive to gustatory stimulation would be more likely to perceive taste from this relatively weak sensory signal. A peripheral factor, such as higher innervation density, is unlikely to be the primary cause of the TT advantage, since the advantage holds for stimulation of three different cranial nerves: the facial nerve (VII) (of which the chorda tympani and greater superficial petrosal nerves are separate branches), the glossopharyngeal nerve (IX), and the olfactory nerve (I). Moreover, if the TT advantage in perception of vanillin had a peripheral origin, the difference between groups would have been independent of the mode of stimulus delivery. Instead, retronasal stimulation yielded a two-fold larger group difference than did orthonasal stimulation.
A more parsimonious hypothesis is that the TT advantage arises from differences in
the sensitivity or excitability of brain regions where olfactory and gustatory
stimulation converge. Greater excitability in such regions would effectively produce a
higher ‘gain’ within the afferent system, leading to stronger perceptual
responses for a given level of gustatory or olfactory stimulation. CNS factors have been
discussed as possible contributors to individual differences in normal taste perception
only with respect to response bias (Ekman and
Akesson, 1965;
Ekman et al., 1968).
However, there is no a priori reason to assume that CNS sensory processes vary less
across individuals than do PNS sensory processes, and studies have begun to identify
brain regions in humans where differences in excitability could potentially affect
perception of both taste and flavor. A recent fMRI study in humans found that
taste-olfactory mixtures activated the caudal orbitofrontal cortex, insular and anterior
cingulate cortex, and the amygdala (de Araujo
et al., 2003). Although it has yet to be determined with certainty
which brain areas are most important for intensity perception (Small et al., 2003a), together these regions
appear to form a functional system for flavor perception. The idea that the sensitivity
of all or part of this ‘flavor system’ could vary across individuals receives
indirect support from studies in other sensory systems which have shown that changes in
the availability of and sensitivity to neuromodulators can affect cortical activation
levels. Brain serotonin levels have been hypothesized to affect the intensity-dependence
of auditory event-related potentials (Hegerl and
Juckel, 1993;
Hegerl et al., 1995) and in
visual attention tasks the level of activation in the anterior cingulate cortex, a region
also involved in taste and flavor perception (de
Araujo et al., 2003), has been linked to genetic differences in
expression of dopamine receptors (Fan et
al., 2003). Interestingly, recent evidence has implicated dopamine in
modifying either the motivation to eat palatable foods (Yamamoto et al., 1998) or their taste pleasantness
(Small et al., 2003b).
However, the significant group difference in perception of vanillin sensed
orthonasally could be taken as evidence against a flavor-system explanation of the TT
advantage. Orthonasal odor sensations arise from stimuli outside the mouth, and thus are
not part of the flavor signal per se. This seeming contradiction may be
explained both by the ability of olfactory stimuli to become perceptually associated with
gustatory stimuli (Murphy et al.,
1977;
Frank and Byram, 1988;
Frank et al., 1993;
Stevenson et al., 1999) and
by evidence that orthonasal stimuli can be integrated with taste stimulation (Rozin, 1982). Vanillin is normally experienced
in connection with foods and has a perceived sweetness (Figure
7) that gives its odor a complex,
flavor-like quality. This combination of factors makes it less surprising that TTs
perceived the vanillin odor to be stronger (though less so) when sensed orthonasally.
Similarly, the small but significant difference between groups in perception of warmth on
the tongue tip (Figure
1d) may reflect the ability of warming
to evoke sweetness in TTs. Even though subjects were instructed to attend only to
temperature on trials in which they rated warmth, the spatially and temporally correlated
gustatory quality may have influenced the way the thermal stimulus was perceived. The
absence of a difference in warmth (or cold) perception between groups when the stimuli
were presented to nongustatory surfaces (Figure
2a,b) supports this
interpretation.
Whatever its specific cause, centrally-mediated differences in taste and flavor
responsiveness could help to explain inconsistencies in the published data on the
relationship of PROP/PTC sensitivity to perception of other tastes. Classification of
individuals as PROP/PTC ‘tasters’ and ‘nontasters’ has not been
disputed, but the idea that tasters can be further subdivided into
‘supertasters’ and ‘medium-tasters’ and that responsiveness to
all tastes can be predicted from responsiveness to PROP/PTC (Bartoshuk, 1993;
Bartoshuk et al., 1994, 1996,
1998;
Tepper and Nurse, 1998;
Prescott et al., 2001), has
not always been supported (Schifferstein and
Frijters, 1991;
Yokomukai et al., 1993;
Delwiche et al., 2001;
Horne et al., 2002). Some of
the negative findings are undoubtedly attributable to use of psychophysical methods not
well-suited for measuring individual differences (Lucchina et al., 1998;
Bartoshuk, 2000;
Bartoshuk et al., 2003).
However, variation in a CNS process that is independent of PROP/PTC receptor expression
would tend to lower the correlation between perception of PROP/PTC and other tastes.
Although individuals with an abundance of PROP receptors and a high central
responsiveness should be highly responsive to all tastes (‘supertasters’),
individuals with fewer PROP receptors but a high central sensitivity should be also be
highly responsive to other tastes. By this reasoning, perceived intensity should be more
highly correlated among taste stimuli for which individual variation in receptor
expression is not as extreme as it is for PROP/PTC. The present data are consistent with
this expectation: sucrose sweetness on the front of the tongue in experiment 3 was
significantly correlated (Pearson’s r, P < 0.05) with ratings
for the primary taste qualities of all other taste stimuli, whether they were delivered
to the front or back of the tongue (average r = 0.44), except for PROP
(r = 0.26, n.s.). In contrast, PROP bitterness on the front of the tongue
was significantly correlated only with citric acid sourness and quinine bitterness on the
front of the tongue, and PROP bitterness on the back of the tongue. The same general
pattern held for the soft palate (experiment 4) and for whole mouth stimulation
(experiment 5). On the soft palate, correlations between ratings for sucrose and NaCl,
citric acid, and quinine were 0.72, 0.85, and 0.68, respectively (Pearson’s
r, all Ps < 0.05), compared to 0.03, 0.20, 0.13 (all Ps
> 0.05) between PROP and the same three stimuli. In the whole mouth, the average
correlation between intensity ratings for the five concentrations of sucrose and citric
acid was 0.44, compared to only 0.20 between the same stimuli and the bitterness of
PROP.
These relatively high correlations in intensity perception across taste stimuli raise
an obvious question: why have they not been reported before? As was mentioned in the
context of individual differences in PROP/PTC perception, the primary reason may be the
prevalence in the past of psychophysical methods that are not well suited for identifying
and quantifying individual differences. The two most widely used methods, magnitude
estimation and category scales, both tend to obscure individual differences, but in
different ways: magnitude estimation by confounding differences in perceived intensity
with differences in number usage, and category scales by introducing ceiling effects that
limit the ability to differentiate between moderately and highly responsive individuals.
The gLMS is a category ratio scale (Borg,
1982) that was developed specifically for measuring individual differences in
taste and somatosensation (Green et al.,
1993, 1996). Subsequent studies have demonstrated the scale’s
effectiveness for measuring individual and group differences within certain constraints
(Lucchina et al., 1998;
Bartoshuk, 2000;
Bartoshuk et al., 2003). One
of the chief constraints is the need to rule out differences in scale use that could be
misinterpreted as evidence of sensory differences. This was accomplished in the present
study by employing the same scale and instructions to measure temperature perception on
the lip and hand, which revealed no differences in ratings between groups.
It is reasonable to ask whether the differences we have found here for perceived
intensity are reflected in a greater sensitivity to threshold-level tastes and odors.
While finding lower thresholds in TTs would buttress the suprathreshold findings and
indicate that the TT advantage affects perception throughout the full perceptual range,
not finding a difference would not contradict the suprathreshold results. This is because
differences in suprathreshold responsiveness are not always associated with differences
in threshold sensitivity (Bartoshuk et
al., 1994;
Bartoshuk, 2000). Indeed, a difference
in central gain might be expected to have a lesser effect on the threshold for detection
than on perceived intensity. A higher central gain should increase the background noise
in the flavor system to about the same extent that it boosts a weak chemosensory signal.
Without a significant change in the signal-to-noise ratio, little change in the detection
threshold would be expected. This hypothesis is currently being tested.
In summary, the present results show that the ability to perceive thermal taste on the tongue tip is positively correlated with the responsiveness to chemical taste stimuli of all kinds throughout the mouth. The data also provide the first evidence of a significant association between the ability to perceive taste and smell at suprathreshold levels, particularly when olfactory stimulation is sensed retronasally. This implies that individuals differ in the ability to perceive the flavor of food as well as its taste and that these differences may arise in part from variation in the sensitivity or ‘gain’ of CNS processes that are involved in perception of the chemosensory attributes of foods.
|
Acknowledgement |
|---|
|
TopAbstractIntroductionGeneral methodsExperiment 1: perception of...Experiment 2: thermal perception...Experiment 3: taste...Experiment 4: taste...Experiment 5: whole mouth...DiscussionAcknowledgementReferences |
|---|
This research was supported in part by a grant from the National Institutes of Health (RO1 DC05002).
|
References |
|---|
|
TopAbstractIntroductionGeneral methodsExperiment 1: perception of...Experiment 2: thermal perception...Experiment 3: taste...Experiment 4: taste...Experiment 5: whole mouth...DiscussionAcknowledgementReferences |
|---|
Adler, E., Hoon, M.A., Mueller, K.L., Chandrashekar, J., Ryba, N.J. and Zuker, C. (2000) A novel family of mammalian taste receptors. Cell, 100, 693–702.[CrossRef][ISI][Medline]
Bartoshuk, L.M. (1993) The biological basis of food perception and acceptance. Food Qual. Pref., 4, 21–32.
Bartoshuk, L.M. (2000) Comparing sensory experiences across individuals: recent psychophysical advances illuminate genetic variation in taste perception. Chem. Senses, 25, 447–460.
Bartoshuk, L.M., Rennert, K., Rodin, H. and Stevens, J.C. (1982) Effects of temperature on the perceived sweetness of sucrose. Physiol. Behav., 28, 905–910.[CrossRef][Medline]
Bartoshuk, L.M., Rifkin, B., Marks, L.E. and Hooper, J.E. (1988) Bitterness of KCl and benzoate: related to genetic status for sensitivity to PTC. Chem. Senses, 13, 528.
Bartoshuk, L.M., Duffy, V.B. and Miller, I.J. (1994) PTC/PROP tasting: anatomy, psychophysics, and sex effects. Physiol. Behav., 56, 1165–1171.[CrossRef][Medline]
Bartoshuk, L.M., Duffy, V.B., Reed, D. and Williams, A. (1996) Supertasting, earaches and head injury: genetics and pathology alter our taste worlds. Neurosci. Biobehav. Rev., 20, 79–87.[CrossRef][ISI][Medline]
Bartoshuk, L.M., Duffy, V.B., Lucchina, L.A., Prutkin, J. and Fast, K. (1998) PROP (6-n-propylthiouracil) supertasters and the saltiness of NaCl. Ann. N. Y. Acad. Sci., 855, 793–796.
Bartoshuk, L.M., Duffy, V.B., Fast, K., Green, B.G., Prutkin, J. and Snyder, D.J. (2003) Labeled scales (e.g., category, Likert, VAS) and invalid across-group comparisons: what we have learned from genetic variation in taste. Food Qual. Pref., 14, 125–138.[CrossRef]
Borg, G. (1982) A category scale with ratio properties for intermodal and interindividual comparisons. In Geissler, H.-G. and Petxold, P. (eds), Psychophysical Judgment and the Process of Perception. VEB Deutscher Verlag der Wissenschaften, Berlin, pp. 25–34.
Bradley, R.M., Mistretta, C.M. and Nagai, T. (1986) Demonstration of sensory innervation of rat tongue with anterogradely transported horseradish peroxidase. Brain Res., 367, 364–367.[CrossRef][ISI][Medline]
Calvino, A.M. (1986) Perception of sweetness: the effects of concentration and temperature. Physiol. Behav., 36, 1021–1028.[CrossRef][Medline]
Cruz, A. and Green, B.G. (2000) Thermal stimulation of taste. Nature, 403, 889–892.[CrossRef][Medline]
de Araujo, I.E., Rolls, E.T., Kringelbach, M.L., McGlone, F. and Phillips, N. (2003) Taste–olfactory convergence, and the representation of the pleasantness of flavour, in the human brain. Eur. J. Neurosci., 18, 2059–2068.[CrossRef][ISI][Medline]
Delwiche, J.F., Buletic, Z. and Breslin, P.A. (2001) Relationship of papillae number to bitter intensity of quinine and PROP within and between individuals. Physiol. Behav., 74, 329–337.[CrossRef][Medline]
Doty, R.L., Bagla, R., Morgenson, M. and Mirza, N. (2001) NaCl thresholds: relationship to anterior tongue locus, area of stimulation, and number of fungiform papillae. Physiol. Behav., 72, 373–378.[CrossRef][Medline]
Ekman, G. and Akesson, C. (1965) Saltness, sweetness, and preference. A study of quantitative relations in individual subjects. Scand. J. Psychol., 6, 241–253.[ISI][Medline]
Ekman, G., Hosman, B., Lindman, R., Ljungberg, L. and Akesson, C.A. (1968) Interindividual differences in scaling performance. Percept. Mot. Skills, 26, 815–823.[ISI][Medline]
Fan, J., Fossella, J., Sommer, T., Wu, Y. and Posner, M.I. (2003) Mapping the genetic variation of executive attention onto brain activity. Proc. Natl Acad. Sci. USA, 100, 7406–7411.
Fischer, R. and Griffin, F. (1963) Quinine dimorphism: a cardinal determinant of taste sensitivity. Nature, 200, 343–347.[CrossRef][Medline]
Fox, A.L. (1931) Six in ten ‘tasteblind’ to bitter chemical. Sci. News Lett., 92, 115–120.
Frank, R.A. and Byram, J. (1988) Taste–smell interactions are tastant and odorant dependent. Chem. Senses, 13, 445–455.
Frank, R.A., Van der Klaauw, N.J. and Schifferstein, H.N. (1993) Both perceptual and conceptual factors influence taste–odor and taste–taste interactions. Percept. Psychophys., 54, 343–354.[ISI][Medline]
Gent, J.F. and Bartoshuk, L.M. (1983) Sweetness of sucrose, neohesperidin dihydrochalcone, and saccharin is related to genetic ability to taste the bitter substance 6-n-propylthioracil. Chem. Senses, 7, 272.
Green, B.G. and Cruz, A. (1998) ‘Warmth-insensitive fields’: evidence of sparse and irregular innervation of human skin by the warmth sense. Somatosens. Mot. Res., 15, 269–275.[CrossRef][ISI][Medline]
Green, B.G. and Frankmann, S.P. (1987) The effect of cooling the tongue on the perceived intensity of taste. Chem. Senses, 12, 609–619.
Green, B.G. and Frankmann, S.P. (1988) The effect of cooling on the perception of carbohydrate and intensive sweeteners. Physiol. Behav., 43, 515–519.[CrossRef][Medline]
Green, B.G., Shaffer, G.S. and Gilmore, M.M. (1993) Derivation and evaluation of a semantic scale of oral sensation magnitude with apparent ratio properties. Chem. Senses, 18, 683–702.
Green, B.G., Dalton, P., Cowart, B.J., Shaffer, G.S., Rankin, K.M. and Higgins, J. (1996) Evaluating the ‘Labeled Magnitude Scale’ for measuring sensations of taste and smell. Chem. Senses, 21, 323–334.
Hall, M.J., Bartoshuk, L.M., Cain, W.S. and Stevens, J.C. (1975) PTC taste blindness and the taste of caffeine. Nature, 253, 442–443.[CrossRef][Medline]
Harada, S., Yamamoto, T., Yamaguchi, K. and Kasahara, Y. (1997) Different characteristics of gustatory responses between the greater superficial petrosal and chorda tympani nerves in the rat. Chem. Senses, 22, 133–140.
Harris, H. and Kalmus, H. (1949a) Chemical specificity in genetical differences of taste sensitivity. Ann. Eugen., 15, 32–45.
Harris, H. and Kalmus, H. (1949b) The measurement of taste sensitivity to phenylthiourea (P.T.C.) . Ann. Eugen., 15, 24–31.
Hegerl, U. and Juckel, G. (1993) Intensity dependence of auditory evoked potentials as an indicator of central serotonergic neurotransmission: a new hypothesis. Biol. Psychiatry, 33, 173–187.[CrossRef][ISI]