Sensitization and desensitization to capsaicin and menthol in the oral cavity: Interactions and individual differences

Physiology & Behavior, Vol. 59, No. 3, 487-494, 1996 Copyright © 1996 Elsevier Science Inc. Printed in the USA. All rights reserved C031-9384/96 $15.00 + .00

ELSEVIER

0031-9384(95)02089-E

Sensitization and Desensitization to Capsaicin and Menthol in the Oral Cavity: Interactions and Individual Differences M A R G A R E T A. CLIFF* AND BARRY G. G R E E N t I

*Agriculture & Agri-Food Canada, Research Centre, Summerland, BC VOH1ZO Canada and ~Monell Chemical Senses Center, 3500 Market Street, Philadelphia, PA 19104-3308 USA. Received 14 November 1994 CLIFF, M. A. AND B. G. GREEN. Sensitization and desensitization to capsaicin and menthol in the oral cavity: Interactions and individual differences. PHYSIOL BEHAV 59(3) 487-494, 1996.--It was reported in a recent study that, like capsaicin, menthol is capable of producing a desensitization to sensory irritation in the oral cavity. Whereas capsaicin is known to be able to cross-desensitize with other chemical irritants, no such information exists for menthol. To address this question, the first experiment was designed to reveal whether cross-desensitization would occur between menthol and capsaicin. After a pretest on the tongue tip in which subjects rated the intensity of irritation and cold produced by 3.5 ppm capsaicin or 0.3% /-menthol, five samples of the same stimuli were sipped and swished at 1-min intervals for 5 min. Fifteen minutes later subjects were tested on the tongue tip with either capsaicin or menthol. The results l) confirmed self-desensitization for both chemicals, 2) demonstrated cross-desensitization of menthol by capsaicin, and 3) revealed cross-sensitization of capsaicin by menthol. This series of outcomes suggests that menthol produces much of its sensory irritation via capsaicin-sensitive pathways, but that it excites and/or desensitizes those pathways via different mechanisms than does capsaicin. Analysis of the individual data revealed large differences in sensitization and desensitization that were significantly correlated across chemicals, which suggests the possibility that the perceptual response to repeated exposures to irritants may be idiosyncratic. Contrary to earlier findings, the first experiment also revealed apparent self- and cross-desensitization of the menthol sensation of coolness. The latter outcome was investigated in a second experiment in which the effect of capsaicin desensitization an the perception of physical as well as chemical (menthol) cooling was measured when the stimuli were presented as oral rinses. No desensitization was found for either form of stimulation, which implied the apparent desensitization of coolness in Experiment 1 may have been due to the difficulty of discriminating sensations of cold from sensation,,; of chemical irritation. The overall findings are discussed in terms of the complex sensory and perceptual interactions that take place within the chemesthetic modality. Capsaicin Human

Menthol

Desensitization

Sensitization

MENTHOL and capsaiciin are two of the most commonly experienced oral irritants. Menthol is known best for its ability to produce distinct sensations of coolness, and capsaicin is chiefly responsible for the sensations of burning and "heat" caused by chili peppers. In additioa to these distinctive characteristics, the two chemicals have other important sensory effects. Most notable is capsaicin's ability to produce chemical desensitization, which is evidenced by a reduced sensory response to subsequent exposures to itself (6,17,19,28,30) and to other chemical irritants

Irritation

Cold

Oral

(14,18,27,29,34). However, when the time between successive exposures is sufficiently brief, capsaicin can also produce the opposite effect of sensitization (i.e., it can enhance the response to subsequent stimulation) (21,22,31,33,37). Recently, menthol was discovered to have similar properties when presented to the mucous membranes of the mouth (8). Mean ratings of sensations of irritation produced by a high concentration of racemic menthol (0.3% w / v ) decreased significantly over repeated exposures, even when the time between stimuli was as long as 5 rain. As

I To w h o m requests for reprints should be addressed.

487

488

CLIFF AND GREEN

with capsaicin, a decline in perceived irritation after an interstimulus interval (ISI) of several minutes suggests the occurrence of a neural desensitization process rather than an adaptation or fatigue-like process (17,19,21). Perusal of the data from individual subjects showed, however, that when stimulation occurred as frequently as once per minute, a few subjects reported that irritation increased whereas most others reported that it decreased (8). The tendency for some individuals to experience sensitization to menthol irritation was similar to what had been reported for a majority of subjects when capsaicin was delivered at the same and shorter ISis (19). Menthol's coolness, which is mediated by a different population of nerve fibers (cold fibers) than is its irritation (nociceptors), remained remarkably constant under the same experimental conditions. Thus, the neurochemical processes that underlie menthol's sensory irritation appear to share important temporal characteristics with the processes that underlie capsaicin's sensory irritation. The main goal of the study reported here was to compare the temporal properties of menthol and capsaicin directly by testing both chemicals using the same experimental paradigm and the same subjects. Central to the comparison was the extent to which cross-desensitization would occur. Symmetrical cross-desensitization would indicate that the two chemicals stimulate and desensitize many of the same trigeminal nerve fibers; little or no cross-desensitization would imply capsaicin and menthol stimulate largely different populations of nerve fibers; and asymmetrical desensitization would imply the two chemicals stimulate many of the same nerve fibers, but in different ways. EXPERIMENT 1 This experiment was designed to measure the relative amounts of self-desensitization and cross-desensitization produced by capsaicin and menthol in the oral cavity. Because we had previously observed marked individual differences in the response to repeated menthol exposures, another objective of the experiment was to assess and compare individual differences for both chemicals. METHOD

Subjects Fifteen adults (nine females, six males) between 24 and 34 years of age were paid to participate in the experiment. All had served in other psychophysical experiments at the Monell Center.

Stimuli Solutions of 11.6 /zM (3.5 ppm) natural capsaicin (8-methyln-vanillyl-6-nonenamide, 98%) were prepared by diluting a 1% ( w / v ) stock solution of capsaicin in 100% ethanol with an appropriate amount of deionized water (dH20). Solutions of 0.30% ( w / v ) /-menthol were prepared by dissolving menthol crystals in 4% ( w / v ) ethanol, mixed with 1% ( w / v ) polysorbate 80 (Tween 80), and brought to volume with dH20. To avoid recrystallization of menthol at the highest concentrations, solutions were prepared on a daily basis. The stimuli were delivered to subjects in two ways: as 10-ml, sip-and-spit samples poured into 1-oz plastic cups just prior to each trial, and as focal, tongue tip stimuli via 1.27 cm 2 filter paper disks that were wetted with 20 /zl of one of the two solutions just prior to application.

Procedure The basic design of the experiment was to apply a filter paper pretest to the tongue tip, follow the pretest with five whole-mouth treatment stimuli, and conclude with a filter paper posttest. Of primary interest was the difference in perceived sensory irritancy or coolness between the pretest and posttest stimuli; of secondary interest was whether irritancy and coolness would change during the treatment phase. The specific procedure was as follows. A session began with the subject rinsing three times with dH20. A filter paper stimulus was then placed on the dorsal tip of his or her tongue, using forceps. The subject sat with the tongue retracted and immobile within the closed mouth both to reduce evaporative cooling (15,36) and to avoid touching the filter paper to other parts of the oral mucosa. After 15 s the subject was asked to rate the intensity of irritation and cooling (if any), and to indicate which of the following sensation qualities was present: burning ["a sensation produced by extreme temperatures or by chemical irritants (e.g., alcohol) that may or may not be accompanied by a thermal sensation"], stinging/pricking ["a sharp sensation similar to that produced by a pinprick or an insect bite (other than itch) which may be constant or intermittent"], itching ["a sensation that induces the desire to scratch"], tingling ["a lively pins-and-needles sensation"], numbness [not the absence of sensation but rather the feeling commonly experienced during the onset and offset of local anesthesia (e.g., as with novocaine)], and ache ["a dull, uncomfortable sensation that fluctuates in strength and is often difficult to localize"]. Pain ["any sensation that hurts"] was also listed to determine whether irritation reached noxious levels. The quality descriptors and their definitions were visible to the subject throughout testing. Ten seconds after instructing the subject to rate the sensation, the experimenter removed the filter paper stimulus from the tongue. No water rinse was allowed. Thirty-five seconds later (1 min after application of the pretest) the subject sipped the first of five whole-mouth stimuli, which they were told to hold toward the front of the mouth to ensure the tip of the tongue would be fully immersed in the stimulus. The sensations experienced throughout the mouth were rated after 15 s, and the solution was expectorated after 25 s. This procedure was repeated four more times, and following expectoration of the fifth treatment stimulus the subject rested for 15 min with the mouth closed. At the end of this hiatus, the filter paper posttest was applied to the tip of the tongue and ratings of sensation intensity and quality were made as before. Ratings of sensation intensity were made using the labeled magnitude scale (LMS) devised previously in this laboratory (23). Briefly, the LMS is a form of category-ratio scale (5) labeled with six verbal descriptors--no sensation, barely detectable, weak, moderate, strong, very strong, and strongest imagi n a b l e - a t numerical locations empirically determined via a semantic scaling task. Subjects that instructed them to rate the perceived intensity of both irritation and cooling within the same perceptual context of "all possible oral sensations." Responses were made by writing the numerical value corresponding to the location on the scale that best described the intensity of the sensation on a response sheet. Subjects also checked as few or as many descriptors as were necessary to describe the sensation fully. The subject was given a separate response sheet on each trial. There were four different versions of the response sheet, each with a different spatial ordering of the descriptors. The experiment had four conditions: 1. menthol self-desensitization: a menthol pretest followed by menthol treatment stimuli and a menthol posttest; 2. menthol cross-desensitization: a menthol pretest followed by menthol treatment stimuli and a capsaicin posttest;

CROSS-DESENSITIZATION TO CAPSAICIN AND MENTHOL

3. capsaicin self-desensitization: a capsaicin pretest followed by capsaicin treatment stimuli and a capsaicin posttest; and 4. capsaicin cross-desensitization: a capsaicin pretest followed by capsaicin treatment stimuli and a menthol posttest. Note that in the cross-desensitization conditions the pretest and posttest stimuli were not the same, as would be desirable based solely on design considerations. However, because we discovered in pilot testing that asymmetrical interactions occurred between capsaicin and menthol that could affect the intensity and, possibly, the effectiveness of the treatment stimuli, it was decided to precede each treatment series with a pretest of the same chemical. Assessment of cross-desensitization was therefore made between rather than within experimental sessions. Subjects received treamaents in duplicate, with a minimum of 24 h between sessions. The order of testing was randomized within subject.

Data Analysis Preliminary evaluation of the data indicated the intensity ratings were distributed log-normally. Consequently, the raw scores were transformed to logarithms before statistical analysis. Differences between pre- and posttest means were evaluated using t-tests with Bonfen'oni correction for multiple comparisons based on a p-value of 0.05, and changes in irritation and coolness during the treatment phase were assessed with separate repeated-measures ANOVAs using the Greenhouse-Geisser E to correct for lack of sphericity (STATISTICA®). Reports of sensation quality were evaluat~,=d by calculating the percentages of the total possible responses (:30; 15 subjects × 2 repetitions), and the Cochran Q-test for categorical data was used to assess statistical significance. RESULTS

Irritation Self-Desensitization. "]~e results from the two conditions (1 and 3) designed to produce self-desensitization of menthol and capsaicin are shown in Fig. 1. Consistent with previous findings, both chemicals produced significant desensitization after exposure to five treatment stimuli [t(14)= 5.3 for menthol, 8.5 for capsaicin; p < 0.008]. The amount of desensitization was approx-

Serf-Desensitization 25

.

.

.

.

!

.

.

.

.

!

.

.

.

|

• --0-• --~--

¢-

O~ r~ ~

'_E

.

.

.

.

.

I

Cap~icin: Filter Paper Capsaicin: Rinse Menthol: Filter Paper Menthol: Rinse

"o 15 r-

IO

. . . . . . . . . . .":_._,~ D. i

,

i

i

I

5

,

-

•

i

10

.

.

.

.

i

15

.

.

.

.

i

20

Time of Report (min) FIG. 1. The perceived intens~ityof irritation is shown as a function of time for the self-desensitization conditions of Experiment 1. The dotted lines connect data points obtained sequentially in time, but using different modes of stimulation (filter paper vs. anterior rinses).

489

Cmu-Desensitization/Sensitiz~lon 25

.

.

.

.

i

.

.

.

.

i

C:

o 2o

.

.

.

.

i

.

.

.

•

Menthol:Filter Paper Menthol: Rinse

•

Capsaidn: Filter Paper

.

--0-- Capsaldn: Rinse

".

i

•

!

.

..0

.............>""-........

c-

a.

0

i

0

=

I

,

I

5

=

=

=

I

I

i

i

i

•

10

I

15

.

.

.

.

20

Time of Report (min)

FIG. 2. Same as Fig. 1, but under conditions capable of producing cross-desensitization (i.e., the menthol rinses were followed by a capsaicin test stimulus, and vice versa). Note the cross-sensitization of capsaicin irritation by menthol.

imately the same for menthol and capsaicin even though the intensity of irritation produced by the menthol treatment stimuli was somewhat less than that produced by the capsaicin treatment stimuli. Compared to pretreatment levels, menthol irritation was reduced by 70.6% and capsaicin irritation by 74.8%. In addition to desensitization following treatment, there was also a tendency for irritation ratings to decline during the course of treatment. However, this trend was significant only for menthol, F(I, 16) = 7.56, p < 0.02. It is notable that the difference in irritation for both chemicals between the pretest and the first rinse trial was likely the result of spatial summation (1,13,16): the rinse stimuli contacted a much larger region of the oral cavity than did the filter paper disks. Cross-Desensitization. Figure 2 contains the results from the cross-desensitization condition. Five exposures to capsaicin significantly reduced the perceived irritation produced by the menthol test stimulus, t(14) = 8.3, p < 0.008 (i.e., cross-desensitization occurred). In contrast, five exposures to menthol significantly increased the perceived irritation produced by the capsaicin stimulus, t(l 4) = 5.1, p < 0.008 (i.e., cross-sensitization occurred). The magnitude of this asymmetrical interaction can be appreciated by comparing the perceived intensities of the menthol and capsaicin posttreatment stimuli in Figs. 1 and 2: whereas menthol irritation was at nearly the same low level after treatment with either menthol or capsaicin, the irritation from capsaicin was more than five times stronger after treatment with menthol than after treatment with capsaicin. Individual Data. Analyses of the irritation ratings for the test stimuli revealed large individual differences in the response to repeated exposures to both chemicals. Whereas some individuals experienced nearly complete self-desensitization, others reported reductions in irritation of less than 50%. Figure 3 illustrates this variation, and shows that it tended to be consistent between chemicals. That is, subjects who experienced pronounced self-desensitization to menthol were likely to experience pronounced self-desensitization to capsaicin (Pearson r = 0.78, p < 0.0001). The single conspicuous exception to this trend was an individual, indicated by the arrow in Fig. 3, who reported nearly 50% desensitization to capsaicin but almost 30% sensitization to menthol. Omitting this subject from the regression analysis raises the correlation to 0.87.

490

CLIFF AND GREEN

too

,

i

•

I

•

I

'

i

,

I

•

I

•

I

'

;

' ~/~

/

Individual Data 40

v /

20

"5 o

|

•

i

,

!

-

i

'iI

,

;

,

~

,

~

:

;

40

,

|

,

•

"

m N U) '- 40 ¢n

9

•

Regltpssi0¢l Y=-16.0 + 1.15X r= 0.87

/ ............

.............................

~0~!

............

C

-20 , S

"400/// 10

. . . . . . 20

30

411

ot" °,~'. 50

"_E '5

. . . . . . . 60

70

80

90

1~

CapsaicinSelf-Desensitization[%1

¢,n (,-

FIG. 3. The relationship between capsaicin and menthol self-desensitization is shown with the best-fit linear regression line. The data are plotted in terms of percent desensitization, where + 100% would equal complete desensitization. The single data point (indicated by the arrow) lying below the dashed line is from a single subject who showed self-sensitization. This datum was not included in the linear regression shown in the figure.

o;

-~

~ f CapsaicinRinse

15

10

Figure 4 shows that large individual differences were also present during the treatment phase of the experiment. The curves shown are for individual subjects and are the mean of four sessions, two replicates each in the two menthol and two capsaicin treatment conditions. The proportional changes in irritation that took place over trials have been illustrated by standardizing the data to an arbitrary irritation rating of 20 for the first exposure. The top graph shows that during exposure to the five menthol treatment stimuli, 5 of the 15 subjects exhibited an immediate and continuous decline in irritation, three others reported a net sensitization, and the remaining seven showed slight-to-moderate desensitization. Overall, the responses during the treatment phase ranged from 50% sensitization to 95% desensitization. The bottom graph shows a similar range of responses to capsaicin. However, fewer subjects experienced pronounced desensitization to the rinses, which is consistent with the absence of a significant effect of trial for capsaicin. These striking individual differences demonstrate that, in the short term, repeated exposure to menthol or capsaicin can have contrasting effects on sensory irritation in different people. It is important to emphasize, however, that 15 min after the last rinse stimulus, all subjects exhibited desensitization to capsaicin, and 14 of 15 subjects exhibited desensitization to menthol (Fig. 2). As was true of the individual differences in posttreatment effects, changes in irritation during the treatment phase were significantly correlated between chemicals (r = 0.85, p < 0.0001). This relationship can be seen in Fig. 5, which plots the standardized irritation reported on the last treatment trial for menthol vs. the standardized irritation reported on the last treatment trial for capsaicin for each of the 15 subjects. The dotted lines drawn through 20 on the x- and y-axes indicate the boundaries between sensitization and desensitization for the two chemicals: data points exceeding 20 indicate sensitization, data points below 20 indicate desensitization. Thus, by the fifth exposure to capsaicin (y-axis), seven subjects were to some degree sensitized and eight subjects were to some degree desensitized. Menthol (x-axis) showed more uniform desensitization, with 12 of 15 subjects

5

0

2'

3'

,

s

o

Time of Report (min) FIG. 4. Individual functions for perceived intensity over time are shown for the treatment phase of Experiment 1. The data, which are the means of four replicates, have been arbitrarily standardized to an intensity of 20 on the first trial to facilitate comparisons among subjects.

40 m -t-

I-

35

to t--

30

=.

25

. . . .

6

i . . . .

i . . . .

i . . . .

! . . . .

i . . . .

S

e

i . . . .

i . . . .

~

i

-J 20

C .....................................

~:'""D

....................................

,

C

F

L)

i

5 t= o[

................

0

5

y:8.04÷o.~x :

i

r--0.~

i ........

10

15

20

~o.~

25

30

....

35

40

MentholIrritationon LastRinse Trial [Standardized] FIG. 5. Plotted is the standardized irritation on the last menthol rinse trial vs. the standardized irritation on the last capsaicin rinse trim for the 15 subjects of Experiment 1. Ratings > 20 represent sensitization and ratings < 20 represent desensitization. The quadrants indicate the four possible outcomes of (A) sensitization to capsaicin and desensitization to menthol, (B) sensitization to both chemicals, (C) desensitization to both chemicals, and (D) sensitization to menthol and desensitization to capsaicin.

CROSS-DESENSITIZATION TO CAPSAICIN AND MENTHOL

Menthol Coolness

25

0 0

•

•

!

•

•

•

•

i

•

•

•

i

i

•

--~--Rinse Stimuli: Whole-Mouth • Probe Stimuli: Tongue-tip

20

ProbeStimulus:AfterCapsatcin

•

o

C

-'-----.~.,

i~

"-"'"----t

D. ,

,

,

,

I

=

.

i

i

5

I

i

•

•

•

10

i

.

.

.

15

.

i

,

20

Time of Report (rain) FIG. 6. Mean ratings of perceived coolness produced by menthol in the self- and cross-desensitization conditions (cf. Figs. 1 and 2).

491

Individual Data. Although the difference in coolness ratings between the pretest and posttest stimuli was significant, not all of the subjects reported this trend. Thirteen of 15 subjects gave lower coolness ratings after menthol and only 10 of 15 gave lower coolness ratings after capsaicin. This result was consistent with the smaller mean reduction in coolness than in irritation. Coolness ratings were reduced 53.5% following menthol and 41.8% following capsaicin; in the same conditions, irritation was reduced by 70.6% and 73.6%. A correlation analysis showed a positive but nonsignificant relationship between the reductions in coolness and irritation for the menthol test stimuli ( r = 0.47, p = 0.077). Figure 7 shows that large individual differences were also present during the treatment phase; subjects exhibited increases and decreases in intensity similar to those seen for capsaicin irritation (Fig. 4). For cooling the extremes ranged from apparent sensitization of 82.5% to apparent desensitization of 70%. These tendencies were significantly correlated (r = 0.59, p < 0.05) with the percentage changes in menthol irritation that were reported during the same rinse trials. Sensation Quality

reporting less irritation by the fifth exposure. Additional information about the effects of the treatment trials on individuals comes from the distribution of responses within the quadrants delineated by the dotted lines: the raost common outcome was desensitization to both chemicals (eight subjects, C), followed by sensitization to capsaicin and de,,;ensitization to menthol (fours subjects, A), and sensitization to I:~th chemicals (three subjects, B). None of the subjects sensitized to menthol and desensitized to capsaicin (D).

The data of Fig. 8 indicate that the quality of irritation varied between chemicals and across treatments. The percentages shown are based upon the total number of trials on which a sensation of irritation of any kind was reported. Itching, aching, and pain were not included in the figure because they were each reported less than 10% of the time. First, it can be seen that when presented as a whole-mouth rinse, menthol and capsaicin tended to evoke somewhat different patterns of sensation quality. The diamonds

Menthol Coolness Group Data. The av,eraged data (geometric means) for the perceived coolness of menthol are shown in Fig. 6. Unlike menthol irritation, average coolness ratings did not decline significantly during the treatment phase. However, ratings of the coolness produced by the filter paper test stimuli were significantly lower following Ixeatment with either menthol, t(14)= 3.57, p < 0.008, or capsaicin, t(14) = 5.33, p < 0.008. This result was surprising both because of the absence of a significant decline during the treatment phase and because in a previous study (6) coolness had not been observed to decrease after repeated exposures to menthol.

Menthol Treatment

Rinse m

Tongue-tip: P r e - T H t

m

Tongue-Ut~ Poet-Mer
e

0 C1. 4)

n," _=

f:

=

.r"

I4)

Menthol Coolness C

Individual Data 4)

n

8-6g

'il .........

Burning

Stinging

Tingling

Numbness

Sensation Quality "~

0

'

2

3

4

5

6

Time of Report (min) FIG. 7. Individual functions for the perceived coolness of menthol over time are shown for the treatment phase of Experiment 1.

FIG. 8. The perceived qualities of irritation produced by menthol and capsaicin for the different conditions of Experiment 1. OSignificant differences ( p < 0.003) in the frequency of reports of qualities between chemicals; * significant differences (p < 0.003) in the frequency of reports of qualities between the pretests (light gray bars) and posttests (dark gray and black bars) in particular conditions.

492

CLIFF AND GREEN

above the rinse trial data (open bars) indicate the presence of significant differences between chemicals in the number of times a particular quality was reported (Cochran Q-test, corrected p < 0.003). Capsaicin was more often reported to produce burning, and menthol was more often reported to produce tingling and numbness; the chemicals produced similar numbers of stinging responses. Second, tests between the pretreatment data (light gray bars) and the two sets of posttreatment data (dark gray and black bars) indicated that exposure to capsaicin rinses resulted in significantly ( p < 0.003) fewer reports of buming and stinging for both capsaicin and menthol, whereas exposure to menthol rinses reduced reports of the same two qualities only when menthol was the test stimulus. These results are consistent with the symmetrical desensitization produced by capsaicin, and the self-desensitization and cross-sensitization produced by menthol.

1.4

1.4 Menthol

1.2

-6 0

1.2

~.0 Vehicle

"o •> o.8 ~ 0.6 I1) Q.

1.0

~iz !i! i~i!

..I 0.2

o.o Pre

Post

Pre

Post

Measurement Re: Capsaicin Treatment

EXPERIMENT 2 As noted above, the significant decrease in the perceived coolness of menthol after exposure to rinses of either menthol or capsaicin (Fig. 6) conflicts with the absence in the group data of a change in coolness ratings during the rinse phase, and with data from our previous study of menthol desensitization (8). Moreover, because menthol coolness apparently results from activation of low-threshold cold fibers rather than nociceptors (11,26), and because capsaicin has not previously been found to affect the perception of innocuous cold (12,40), there was no reason to expect the coolness of menthol to change after exposure to capsaicin. Consequently, we considered the possibility that the apparent desensitization of coolness following either menthol or capsaicin was attributable to a difficulty in differentiating between sensations of coolness and irritation. The significant positive correlation between the percent change in irritation and coolness during the rinse trials, and a similar (nonsignificant) trend for the posttest stimuli, were consistent with this hypothesis. A second experiment was therefore run in which subjects were instructed to ignore irritation and to rate only coolness. Capsaicin was used as the treatment stimulus, and its effects on menthol coolness were compared to its effects on the perception of actual cooling produced by a room-temperature liquid.

FIG. 9. Mean (log) perceived coolness is shown for menthol and the vehicle before and after exposure to five capsaicin treatment stimuli in Experiment 2. Vertical bars represent SEs. Pretreatment with capsaicin had no effect on the perceived coolness, whether produced by thermal cooling alone (vehicle) or thermal plus chemical cooling (menthol stimulus).

five capsaicin treatment stimuli, vehicle (cool) posttest, menthol posttest. The two pretest stimuli and the treatment stimuli were all presented at 1-min interstimulus intervals. However, as in Experiment 1, a 15-rain hiatus was inserted between the last treatment stimulus and the first posttest stimulus. The vehicle and capsaicin stimuli were sipped and held in the anterior part of the mouth for 15 s; the menthol stimuli were sipped and held for 25 s. The subject's task was to rate the intensity of sensations of coolness produced while the pretest and posttest stimuli were in the mouth and to ignore as best they could any accompanying sensations of irritation. Coolness ratings were requested 2 s after sipping the vehicle and 15 s after sipping the menthol. The two different rating times were dictated by the more rapid time course of thermal cooling compared to the development of chemical coolness. To maintain the subject's focus on sensations of coolness, the irritation produced by the capsaicin treatment stimuli was not rated. Coolness ratings were made on the LMS as before.

METHOD RESULTS

Subjects Fifteen individuals (eight females and seven males) between the ages of 20 and 34 were paid to participate. Eight of the subjects had participated in Experiment 1, but did not know its outcome.

Stimuli The test stimuli were 10-ml samples of 0.3% ( w / v ) l-menthol dissolved in the same vehicle of 4% ethanol and 1% polysorbate 80 used in Experiment 1. The treatment stimuli were 10-ml solutions of 5 ppm (16.5 /zM) capsaicin prepared in the same vehicle. For both stimuli, the chemicals were first dissolved in 100% ethanol then brought to volume with appropriate amounts of d H 2 0 and polysorbate 80. Samples (10 ml) of the room-temperature (approximately 23°C) vehicle were also used as control stimuli.

Procedure After rinsing with dH20, subjects received nine solutions in the following sequence: vehicle (cool) pretest, menthol pretest,

Figure 9, which contains the mean (log) ratings of coolness during the pre- and posttests, clearly shows that repeated exposure to capsaicin under conditions that normally produce desensitization of sensory irritation did not affect the perceived coolness of either the vehicle or the menthol test stimulus. A repeatedmeasures ANOVA revealed a main effect of stimulus, F(1, 14) = 65.37, p < 0.0001, but found no effect of time of testing (pre vs. post) and no interaction between stimulus and time of testing (both p > 0.05). The main effect of stimulus reflects the enhancement of coolness for which menthol is well known. DISCUSSION Experiment 1 confirmed the earlier finding that the sensory irritation produced by menthol desensitizes over time, and that the amount and rate of desensitization varies markedly across individuals (8). For some people, irritation actually rises over the first few exposures, falling below initial levels only after a hiatus in stimulus application. The experiment also yielded two new, contrasting findings: that desensitizing the tongue tip to capsaicin

CROSS-DESENSITIZATION TO CAPSAICIN AND MENTHOL

causes a comparable desensitization to menthol irritation, but that desensitizing the tongue tip to menthol causes sensitization to capsaicin irritation. The former result demonstrates that menthol's irritancy is mediated by capsaicin-sensitive fibers; the latter result implies the two irritants stimulate a common sensory pathway, but that they do so via different excitatory mechanisms. Although it is unclear from the present data why desensitization between menthol and capsaicin should be asymmetrical, this result is consistent with capsaicin's nonspecific blocking effect (6,27,34), and implies that menthol's desensitization effect is more specific. Current knowledge of the cellular effects of the two chemicals also suggests a plausible hypothesis as to why pretreatment with menthol heightens the response to capsaicin. Whereas capsaicin is known to cause an influx of Ca and Na across neural membranes (27,34), menthol is known to block Ca channels in some cells (38,39). It is therefore conceivable that menthol causes a build-up of extracellular Ca, which then amplifies excitatory effects of capsaicin during posttreatment. The large individual differences observed for both capsaicin and menthol are simila:r in magnitude to those obtained in a previous study with menthol (8), and intersubject variability in the response to repeated exposures to capsaicin has been reported previously (9,35). In the: present study, the correlation between irritants of the tendencie.,; toward sensitization and desensitization suggests that individual differences arise in part, at least, from factors that are not stimulus specific. Whether the differences arise solely from properties of the sensory system, or whether they can be attributed in some measure to higher level cognitive or emotional factors, cannot be determined from the present data. Further experimentation with a wider variety of irritants and stimulus concentrations would help to clarify the characteristics of these differences and might lead to testable hypotheses about their origins. The data on sensation quality are consistent with the data on sensation intensity in that frequencies of reports of the two clearly nociceptive qualities of burning and stinging paralleled changes in intensity ratings observed between the pretreatment and posttreatment stimuli. This outcome has been seen before for capsaicin desensitization (18), and provides strong indirect evidence of the activation and subsequent desensitization of chemically sensitive nociceptors by menthol. However, the relatively high frequency of tingling and numbness evoked by menthol during the rinse trials suggests menthol affects other somatosensory fibers, and in a more pronounced way than does capsaicin. Tingling and numbness are enigmatic, however, insofar as they tend to resist desensitization (18). In addition, the frequency of reports of tingling and numbness produced by both ethanol (unpublished data) and CO 2 (20) has been observed to remain largely independent of concentration and temperature. These traits, plus the associatiion of tingling and numbness with the diminished sensation during local anesthesia, have prompted the

493

hypothesis (18,20) that these sensations result from peripheral blocking effects (24,25) that disrupt the normal pattern of somatosensory input. That is, tingling and numbness may result from inhibition rather than from stimulation. The lack of effect of capsaicin exposure on perceptions of coolness in Experiment 2 is consistent with previous data showing capsaicin does not affect the sensitivity to innocuous cold. Changing the task to eliminate simultaneous judgments of irritation appears to have reduced confusion by enabling subjects to attend fully to coolness. In addition, assessing the intensity of sensations of coolness is made more difficult when the sensations arise from a small, homogeneous area like the tip of the tongue. Casual observation demonstrates that when a large portion of the oral cavity is stimulated, as during an oral rinse, coolness and irritation appear to be separable in part because they have somewhat different spatial distributions throughout the mouth. However, the correlation between ratings of irritation and coolness for the menthol rinses in Experiment 1 suggests that spatial factors alone cannot account for the difficulty that subjects have discriminating between these two qualities. Indeed, there is evidence that the sensation of cold has an intrinsic nociceptive quality (e.g., "stinging cold") that is offset at moderate temperatures by inhibition from low-threshold cold fibers (4,7,10), but which intensifies at lower temperatures as activity in cold-sensitive nociceptors increases and activity in low-threshold cold fibers decreases. In this view the strength of menthol's illusory cooling depends in part on the level of irritation it produces, hence desensitization should influence ratings of menthol coolness. The design of Experiment 2 may have reduced this influence by giving subjects a prototypical cool sensation (room-temperature water) to which they could compare the more complex sensations produced by menthol. In addition to questions of quality coding and discrimination, the present results pose several questions for future research. First, the occurrence of cross-sensitization to capsaicin raises questions about the generalizability of menthol s hypoalgesic effect. Additional irritants need to be tested. Second, how long does menthol desensitization last? Informal tests in our laboratory indicate that it lasts for at least an hour, but more careful experimentation is needed to map its time course. Third, as noted above, individual differences in the response to both menthol and capsaicin deserve further study. Given that chemical sensitivity plays a crucial role in inflammatory pain and hyperalgesia (2,3,28,32), the existence of differences in sensory responsiveness to exogenous chemicals like menthol and capsaicin raises the possibility that similar differences exist in the responsiveness of the nociceptive system to endogenous algogenic chemicals. ACKNOWLEDGEMENT This research was funded in part by a grant from the National Institutes of Health, DC00249.

REFERENCES 1. Armstrong, D.; Dry, R. M. L.; Keele, C. A.; Markham, J. W. Observations on chemical excitants of cutaneous pain in man. J. Physiol. 120:326-351; 1953. 2. Barth6, L.; Stein, C.; Herz, A. Involvement of capsaicin-sensitive neurones in hyperalgesia and enhanced opioid antinociception in inflammation. Naunyn Schmiedeberg's Arch. Pharmacol. 342:666670; 1990. 3. Baumann, T. K.; Simone, D. A.; Shain, C. N.; LaMone, R. H. Neurogenic hyperalgesia: The search for the primary cutaneous afferent fibers that contribulLeto capsaicin-induced pain and hyperalgesia. J. Neurophysiol. 66:212-227; 1991. 4. Bini, G.; Cruccu, G.; Hagbarth, K.-E.; Shady, W.; Torebjork, E.

5. 6. 7. 8.

Analgesic effect of vibration and cooling on pain induced by intraneural electrical stimulation. Pain 18:239-248; 1984. Borg, G. Psychophysical scaling with applications in physical work and the perception of exertion. Scand. J. Work Environ. Health 16:55-58; 1990. Buck, S. H.; Burks, T. F. The neuropharmacology of capsaicin: Review of some recent observations. Pharmacol. Rev. 38:179-226; 1986. Chapman, C. E. Can the use of physical modalities for pain control be rationalized by the research evidence?. Can. J. Physiol. Pharmacol. 69:704-712; 1991. Cliff, M. A.; Green, B. G. Sensory irritation and coolness produced

494

9. 10. 11.

12.

13. 14.

15. 16. 17.

18.

19. 20.

21. 22.

23.

24.

CLIFF AND GREEN

by menthol: Evidence for selective desensitization of irritation. Physiol. Behav. 56:1021-1029; 1994. Craft, R. M.; Porreca, F. Treatment parameters of desensitization to capsaicin. Life Sci. 51:1767-1775; 1992. Craig, A. D.; Bushnell, M. C. The thermal grill illusion: Unmasking the burn of cold pain. Science 265:252-255; 1994. Dodt, E.; Skouby, A. P.; Zotterman, Y. The effect of cholinergic substances on the discharge from thermal receptors. Acta Physiol. Scand. 28:101-114; 1953. Foster, R. W.; Ramage, A. G. The action of some chemical irritants on somatosensory receptors of the cat. Neurophannacology 20:191198; 1981. Garcia Medina, M. R.; Cain, W. S. Bilateral integration in the common chemical sense. Physiol. Behav. 29:349-353; 1982. Gilmore, M. M.; Green, B. G. Sensory irritation and taste produced by NaCI and citric acid: Effects of capsaicin desensitization. Chem. Senses 18:257-272; 1993. Green, B. G. Sensory interactions between capsaicin and temperature. Chem. Senses 11:371-382; 1986. Green, B. G. Spatial and temporal factors in the perception of ethanol irritation on the tongue. Percept. Psychophys. 44:108-116; 1988. Green, B. G. Capsaicin sensitization and desensitization on the tongue produced by brief exposures to a low concentration. Neurosci. Lett. 107:173-178; 1989. Green, B. G. Capsaicin cross-desensitization on the tongue: Psychophysical evidence that oral chemical irritation is mediated by more than one sensory pathway. Chem. Senses 16:675-689; 1991. Green, B. G. Temporal characteristics of capsaicin sensitization and desensitization on the tongue. Physiol. Behav. 49:501-505; 1991. Green, B. G. The effects of temperature and concentration on the perceived intensity and quality of carbonation. Chem. Senses 17:435-450; 1992. Green, B. G. Evidence that removal of capsaicin accelerates desensitization on the tongue. Neurosci. Lett. 150:44-48; 1993. Green, B. G.; Shaffer, G. S. The sensory response to capsaicin during repeated topical exposures: Differential effects on sensations of itching and pungency. Pain 53:323-334; 1993. Green, B. G.; Shaffer, G. S.; Gilmore, M. M. Derivation and evaluation of a semantic scale of oral sensation magnitude with apparent ratio properties. Chem. Senses 18:683-702; 1993. Hellekant, G. Action and interaction of ethyl alcohol and some other substances on the receptors of the tongue. In: Hayashi, T., ed. Olfaction and Taste II. New York: Pergamon Press; 1967:465-479.

25. Hellekant, G. The effect of menthol on taste receptors. Acta Physiol. Scand. 76:361-368; 1969. 26. Hensel, H.; Zotterman, Y. The effect of menthol on thermoreceptors. Acta Physiol. Scand. 24:27-34; 1951. 27. Holzer, P. Capsaicin: Cellular targets, mechanisms of action, and selectivity for thin sensory neurons. Pharmacol. Rev. 43:143-201; 1991. 28. Jancso, N. Role of the nerve terminals in the mechanism of inflammatory reactions.. Bull. Millard Fillmore Hosp. 7:53-77; 1960. 29. Jancso, N.; Jancso-Gabor, A.; Takats, I. Pain and inflammation induced by nicotine, acetylcholine and structurally related compounds and their prevention by desensitizing agents.. Acta Physiol. Acad. Sci. Hung. 19:113-132; 1961. 30. Karrer, T.; Bartoshuk, L. Capsaiein desensitization and recovery on the human tongue. Physiol. Behav. 49:757-764; 1991. 31. Kenins, P. Responses of single nerve fibres to capsaicin applied to the skin. Neurosci. Lett. 29:83-88; 1983. 32. Kilo, S.; Schmelz, M.; Koltzenburg, M.; Handwerker, H. O. Different patterns of hyperalgesia induced by experimental inflammation in human skin. Brain 117:385-396; 1994. 33. LaMotte, R. H.; Shain, C. N.; Simone, D. A.; Tsai, E.-F. P. Neurogenic hyperalgesia: Psychophysical studies of underlying mechanisms. J. Neurophysiol. 66:190-211; 1991. 34. Maggi, C. A. Capsaicin and primary afferent neurons: From basic science to human therapy. J. Auton. Nerv. Syst. 33:1-14; 1991. 35. Nasrawi, C. W.; Pangborn, R. M. Temporal effectiveness of mouthrinsing on capsaicin mouth-bum. Physiol. Behav. 47:617-623; 1990. 36. Sizer, F.; Harris, N. The influence of common food additives and temperature on threshold perception of capsaicin. Chem. Senses 10:279-286; 1985. 37. Stevens, D. A.; Lawless, H. T. Enhancement of responses to sequential presentation of oral chemical irritants. Physiol. Behav. 39:63-65; 1987. 38. Swandulla, D.; Carbone, E.; Shafer, K.; Ix, H. D. Effect of menthol on two types of calcium currents in cultured sensory neurons of vertebrates. Pflugers Arch. 409:52-59; 1987. 39. Swandulla, D.; Schafer, K.; Ix, H. D. Calcium channel current inactivation is selectively modulated by menthol. Neurosci. Lett. 68:23-28; 1986. 40. Szolcsanyi, J. A pharmacological approach to elucidation of the role of different nerve fibres and receptor endings in mediation of pain. J. Physiol. (Paris) 73:251-259; 1977.