Archs oral Bid. Vol. 29, ho. 1, pp. 17-23, 1984 0003-9969/84 $3.00 + 0.

00
Printed in Great Britain. All rights reserved Copyright 0 1984 Pergamon Press Ltd

EFFECTS OF PHARMACOLOGIC REDUCTIONS IN

SALIVARY FLOW ON TASTE THRESHOLDS
IN MAN
CAROL M. CHRISTENSEN*?, M. NAVA’ZESH*t and V. J. BRIGHTMAN?
*Monell Chemical Senses Center, 3500 Market Street, Philadelphia, PA 19104, and tSchool of Dental
Medicine, University of Pennsylvania, Philadelphia, Pennsylvania, U.S.A.

Summary-The effects of short-term salivary flow reductions on human taste thresholds were measured.
Recognition and detection thresholds were obtained from 65 subjects during periods of both normal and
reduced salivary flow. Decreased salivary flow was achieved by oral administration of either Elavil”,
Benadryl” or atropine. Thresholds were measured for NaCl, citric acid, quinine sulphate and sucrose with
a traditional series of aqueous solutions as well as with a series of dry taste stimuli using a filter-paper
base. Whole mouth resting flow and stimulated salivary flow were measured before and after taste testing.
The pharmacologic agents produced depressions in salivary flow ranging between 30 and 75 per cent of
normal levels. The large decreases in flow produced no measurable changes in taste thresholds with the
exception that an increased sensitivity to aqueous and dry citric acid stimuli consistently was observed
following atropine administration. Changes in salivary bicarbonate levels, produced by atropine, may have
mediated the c’bserved shifts in oral sensitivity to citric acid.

INTRODUCTION Our purpose was to measure whether salivary flow

per se affected taste perception by investigating
It is reasonable to expect that saliva affects the sense whether short-term reductions in salivary flow pro-
of taste. The relationship of taste perception to either duced changes in taste thresholds. Both an aqueous
salivary composition or salivary volume (flow rate) and dry series of taste stimuli were used. It was
has been examined in but few studies. NaCl taste hypothesized that individuals might become more
sensitivity in man has been linked to salivary-sodium sensitive to taste stimuli in an aqueous base following
levels (McBurney and Pfaffmann, 1963; Bartoshuk, salivary-flow reductions because less saliva would be
1978; Morino and Langford, 1978); the effects on available to dilute the tastants. The opposite predic-
taste of other salivary constituents have not been tion was made for the dry series; decreased salivary
examined. flow may possibly reduce the rate of release as well
Animal models used to examine the effects of as the total quantity of tastant leached from the dry
salivary gland excision on taste (e.g. Catalanotto and matrix by saliva and thereby reduce taste sensitivity.
Sweeney, 1972; Cat.slanotto and Leffingwell, 1979) Decreases in flow were produced by separately
have shown shifts in taste preferences which suggest administering several different pharmacologic agents
that animals lose taste acuity. Interpretation of these known to induce xerostomia. The use of several
findings is complicated by probable damage to the medications permits some identification of taste
oral mucosa and possible changes in taste receptor changes attributable to unique, drug-specific effects
cells as a result of desalivation (Cano and Rodriguez- rather than to salivary flow reductions. Pharma-
Enchandia, 1980; Nanda and Catalanotto, 1981). cological manipulation of salivary flow rate offers the
However, preferences for certain taste substances most useful and direct approach to examining the
were altered in rats following short-term reductions effects of salivary flow on taste perception because the
in salivary flow produced by pharmacologic means effects are reversible and short-lived. Further, there
(Galilli, Maller and Brightman, 1978) which suggests are a large number of prescribed and over-the-
that salivary flow per se may influence taste percep- counter medications reported to produce xerostomia
tion. and so there is clinical relevance in ascertaining
Research on the role of salivary flow in human whether pharmacologically-induced xerostomia pro-
taste perception has also been complicated by several duces significant shifts in taste perception. A
possible factors. Taste thresholds have been obtained repeated-measures design, whereby subjects were
from patients irradia.ted in the head and neck regions tested during periods of both normal and reduced
for treatment of ca.ncer (Conger, 1973) and from salivary flow, was used because large individual
patients with Sjdgren’s syndrome. Irradiation caused differences in taste thresholds and salivary flow rate
pronounced xerostomia but also apparently damaged make comparisons of taste and salivary measures
innervation to taste buds; all taste acuity was rapidly between different groups of subjects difficult to make.
lost; a IO-fold recovery was evident 2660 days
following the end of treatment. Sjiigren’s syndrome is MATERIALS AND METHODS
identified by a triad of symptoms: xerostomia, kera-
toconjunctivis sicca and a connective tissue disorder; Experiment 1. Detection thresholdri/Elavil and Ben-
there is elevated recognition- and detection-taste adryl
thresholds but also evidence of abnormal and smaller Subjects. Twenty subjects were recruited and 18
numbers of taste cells (Henkin et al., 1972). completed the experiment. Subjects were University
17
18 CAROL M. CHRISTENSEN et 01

students and staff (approximately equal numbers of Pharmacologic agents. Subjects were given a single
males and females). The criteria for participation oral dose of either a placebo (lactose capsule), Ben-
were that subjects were non-smokers between the adryl(50 mg) or Elavil(25 mg). The medications were
ages and I8 and 35 years, weighed between I05 and chosen because they are reported to produce xero-
180 lb, had no medical problems and no dietary stomia and because other physiological effects of
restrictions, took no medications on a regular basis, these drugs are relatively mild at the levels used.
had no reported history of drug abuse and, if female, Diphenhydramine hydrochloride (Benadryl ‘I,
were neither pregnant nor lactating. Subject screening Parke-Davis, Inc.) is an antihistamine. Most subjects
included administration of the Cornell Medical In- reported feeling a dry mouth and some subjects also
dex. reported that the medication caused mild drowsiness.
Salivary$ow measurements. Whole-mouth saliva Amitriptyline HCI (Elvail’, Merck, Sharp & Dohme,
was collected in pre-weighed test tubes fitted with a Inc.) is one of the tricyclics commonly used as an
plastic funnel. First, subjects rinsed several times with antidepressant. Subjects did not report drowsiness
deionized water and rested quietly for 5 min. The with this medication but did report dry mouth.
collection period began by instructing subjects to Testkg regimen. There were 4 morning testing
swallow in order to empty the mouth of saliva. The sessions held at weekly intervals. The first was a
subject then tilted the head forward, held the mouth practice session designed to acquaint subjects with
over the funnel and allowed saliva to drain into the the staff and with the salivary and taste procedures.
test tube. Subjects were cautioned to remain station- It was hoped that the practice session would prevent
ary during the collection period and not to swallow. the occurrence of anxiety-provoked xerostomia and
At the end of the collection period, subjects ex- reduce the variability in taste-threshold measures.
pectorated any remaining saliva. This collection tech- For the subsequent 3 sessions, subjects fasted over-
nique was both simple and reliable (Navazesh and night. When they arrived at a session, they were given
Christensen, 1982). Weight rather than volume deter- one of 3 medications and then given a standard
minations of flow rate were calculated because, al- breakfast. Every subject received Benadryl, Elavil
though weight and volume measurements are highly and the placebo and the session order in which
correlated (+0.99), weight measures are more re- subjects received the medications was randomized.
liable. Subjects were not told what medications they were
Salivary flow was assessed during resting (absence receiving at a session.
of any obvious salivary stimulation) and stimulated Taste testing began immediately after the initial
states immediately before taste-threshold testing be- salivary collections. Thresholds for aqueous and dry
gan. Resting flow was measured first. To stimulate taste stimuli were determined in separate but se-
salivary flow, a I .2 cm diameter filter paper saturated quential tests. For aqueous tests, subjects were given
with a 0.15 M citric acid solution and shaken to 10 ml quantities of taste soiutions which were ex-
remove excess liquid, was placed on the tongue for pectorated. For dry tests, subjects were instructed to
the first 30 s of a collection period. For half the manipulate each paper with the tongue to express the
subjects, restmg flow was measured in 4 sequential taste substance before removing it. Subjects were
3-min collection periods punctuated by I-min rest tested with only 2 taste qualities and the same
periods and stimulated flow was measured in 4 qualities were used for both the dry and wet tests and
sequential 2-min collection periods punctuated by for all 4 testing sessions.
2-min rest periods. Salivary collections began 45 min Detection thresholds were measured using a forced
after this group of subjects received medication. For choice procedure. Subjects chose the taste stimulus
the other subjects, resting flow was measured in 3 from a pair of stimuli: a blank (deionized water) and
sequential 4-min collection periods and stimulated a taste stimulus. Thresholds for 2 taste qualities were
flow was measured in 3 sequential 2-min collection simultaneously determined using a modified staircase
periods. Salivary collections began 30 min after sub- procedure (Cornsweet, 1962). An ascending series
jects received the placebo or Benadryl and 90 min continued until 2 correct responses were given and a
after subjects received Elavil. All flow rates (g/min) descending series continued until I incorrect response
were averaged from each collection series to obtain a was given. Five reversals were obtained and a thresh-
single representative measure for each subject. old was computed from the geometric mean of the
Taste stimuli. A series of aqueous and dry taste last 4 reversals. There was no rinse or time delay
stimuli were prepared. Taste substances representing between each stimulus of a pair. Subjects rinsed
the four basic tastes were used: NaCl (salt), citric acid several times after each test pair and then a 30-set
(sour), quinine sulphate (bitter) and sucrose (sweet). period was timed before the presentation of the next
For the aqueous series, taste stimuli were dissolved in pair of stimuli. A compromise was necessary between
deionized water using 2-fold serial dilutions. The an interval long enough to restore some resting
highest concentrations were: 1.6 x 10m2M NaCl, quantity of saliva in the mouth and an interval short
2.5 x 10m4M citric acid, 1.0 x 10d5 M quinine and enough to permit testing of sufficient numbers of taste
4.0 x 10m2M sucrose. Dry taste stimuli were pre- stimuli before a drug’s xerostomic effects diminished.
pared by squirting 50~1 quantities of taste solution
on the surface of filter papers (2.1 cm diameter Experiment 2. Recognition thresholds/Elavil and Ben-
hardened, ashless grade) and allowing them to air- adryi
dry. The dry series was also prepared using 2-fold The experimental protocol was similar to that
dilution steps; the highest concentrations were: described above, apart from differences that are
0.1 M NaCl, 1.2 x 10m2M citric acid, 1.0 x 10-j M mentioned. Ten subjects were recruited on the same
quinine and 7.5 x 10m2M sucrose. criteria; 9 completed the experiment. Resting and
 Salivary flow reductions and taste 19

stimulated salivary flow was assessed both before and who completed this experiment received 0.8 mg of
after taste threshold testing. The salivary collections atropine. The dosage was increased to 1.O mg for the
began 45 min after subjects received the medications. other subjects. Resting saliva was measured in 2
Measurement of flow rate following testing occurred sequential 3-min collection periods before and after
between 2 to 3 h after administration of the medica- taste testing. Two stimulated measures were obtained
tions. Resting salivary flow was measured first in 3 after taste tests were completed and only for the
sequential 3-min collection periods punctuated by group receiving 1.0 mg of atropine. Salivary col-
l-min rest periods. Stimulated flow was measured in lections and taste tests began 75 min after the drug
3 sequential 2-min collection periods punctuated by was administered.
2-min rest periods.
Recognition thresholds were determined simulta-
neously for 2 taste qualities using a modified double- Experiment 4. Recognition threshob (5 ml)/atropine
staircase procedure. A blank stimulus (deionized To test the salivary dilution hypothesis, 5 ml
water) and 2 stimuli of each taste quality were quantities of aqueous taste solutions were substituted
randomly presented in each block of 5 trials. After for the 10 ml quantities used in all the previously
each taste stimulus, subjects rinsed with deionized described threshold testing procedures. The rationale
water and waited 30 s before being presented with the was that salivary dilution would have a greater
next stimulus. Thresholds were obtained for NaCl impact with smaller quantities of taste solution and,
(n = 9) and either citric acid (n = 5) or quinine consequently, the effect of atropine in reducing sali-
sulphate (n = 4). !jubjects were given a supra- vary dilution would be more apparent.
threshold solution of each taste substance before a Twenty subjects selected by the criteria referred to
test session to familiarize them with the different taste in Experiment 1 were tested. The experimental
qualities. Subjects responded on each trial that the protocol was similar to the previous experiment
test stimulus was Ieither salty, sour or bitter, or except that salivary collections and taste testing be-
neither. An ascending concentration series was re- gan 60 min after the drug was administered.
versed when subjects correctly identified the taste
quality and a descending concentration series was
reversed when subjects either incorrectly identified or RESULTS
could not taste the test stimulus. Testing was termi-
nated when 5 reversals were obtained for each taste Experiment 1
quality. The geometric mean of the last 4 reversals Pharmacologic reductions in salivary flow failed to
was considered to ble the threshold concentration for affect taste detection thresholds. Using log threshold
a subject. Any taste quality responses to blanks were data, separate repeated measures analyses of variance
recorded as measures of response bias. Aqueous and for each taste quality failed to yield significant
dry recognition thresholds were determined in sepa- differences between the 3 drug treatments. The F
rate but sequential tests with the order of the 2 tests values (2 and 14-18 degrees of freedom) ranged from
randomized. 0.15 to 1.99 for the main effect of drug treatment and
The aqueous taste stimuli were a series of solutions from 0.03 to 1.02 for the interaction term
prepared by 2-fold (NaCl) or 4-fold (citric acid and (treatment/wet-dry tastant). Table 1 lists the mean
quinine) serial dilutions. The highest concentrations threshold values. Thresholds for dry taste stimuli
in the series 1.3 x IO-’ M NaCI, 4.1 x lo-’ M citric were consistently and, with the exception of NaCl,
acid and 1.0 x 10e3 M quinine. The dry taste stimuli significantly higher (F values were greater than 12.7
were prepared from 2-fold serial dilutions. The high- for 1 and 8-9 degrees of freedom) than for the
est concentrations aere 0.8 M NaCl, 0.2 M citric acid aqueous stimuli, perhaps because the entire quantity
and 4.0 x lO-3 M quinine. of taste solution was not dissolved from the paper
matrix.
Single oral doses of Benadryl and Elavil produced
Experiment 3. Recognition thresholdsjatropine significant (p < 0.001) decreases in salivary flow
It is possible that salivary-flow reductions greater (Table 2): F(2,22) = 7.93 (resting) and 12.50 (stimu-
than those obtained with Benadryl and Elavil would lated). Prior to tast testing, resting salivary flow
affect taste thresholds. Rather than increasing the reductions were measured to be 45 per cent for Elavil
dosage levels of the 2 medications, atropine sulphate and 40 per cent for Benadryl. Stimulated flow was
(Merck, Sharp & Dohme, Inc.), a powerful anti- reduced 45 per cent with Elavil and 32 per cent with
cholinergic, was chosen as the pharmacologic agent. Benadryl. Five subjects did not show salivary-flow
In this experiment and the next, subjects received the reductions following Benadryl; only 1 subject failed
placebo and atropine separately in weekly test ses- to respond to Elavil. Their salivary data are not
sions. Altogether subjects participated in three ses- included in the calculation of salivary-flow reduc-
sions: a practice and 2 test sessions. tions. It is likely that more subjects responded to
Recognition thresholds were obtained for all 4 Benadryl than was recorded because 4 of the 5
taste qualities. For a single subject, thresholds were subjects came from the group where salivary flow was
determined simultaneously for only 2 taste qualities. measured only 30 min after drug administration.
The procedures were similar to those described above Based on subsequent observations with this parad-
except that 2 correct responses were required before igm, salivary-flow reductions could be recorded 45 min
a reversal and a I-min interval was imposed between after administration of this dosage of Benadryl. It is
intertrial rinsing and the presentation of the next also possible that the stimulatory effects of the recent
taste stimulus. Approximately half of the 18 subjects meal were still present (Dawes and Chebib, 1972).
 Table I. Taste detection thresholds (mean + SEM) for wet and dry stimuli during periods of normal and pharmacologically-reduced salivary flow*
NaCl Sucrose Citric acid Quinine

Placebo Dry 2.62 x 1O-3 f 1.22 x 1O-3 2.07 x lo-‘If: 1.10 x 10-j 3.41 x 10-4* 1.88 x 10-d 2.92 x IO-’ k 1.46 x 1O-5
Wet 9.66 x 10m4 f 3.85 x 10m4 4.49 x 10m4 + 2.69 x 1O-4 1.13 x 10~5f0.49 X 10-s 6.87 x IO-’ + 2.60 x lo-’

Elavil Dry 2.95 x IO-‘* 1.52 x lo--’ 1.84 x IO-’ k 0.88 x tom3 7.66 x 10m4 k 2.88 x lO-4 3.91 x 10-5If: 1.50 X 10-S
Wet 8.86 x 10m4 + 2.46 x 10m4 4.82 x 1O-4 k 2.72 x 1O-4 9.46 x 1O-6 k 3.62 x 10m6 9.27 x lo-’ + 3.89 x lo-’

Dry 6.22 x IO-’ k 4.29 x 10m3 9.51 x lo-4k6.16 X IO-4 7.44 x 10-d f 4.44 x IO-4 2.18 x 1O-4 f 1.52 x lo-“
Benadryl
Wet 2.67 x IO-‘+ 1.66 x 10-l 3.43 x lo-“* 1.89 x 10m4 7.44 x 10m6 k 3.84 x 10m6 6.23 x 10m6 f 5.04 x 10mh
(n = 8) (n = 10) (n = 9) (n = 9)

*Thresholds are expressed in molar concentration and represent the geometric mean.

Table 2. Resting and stimulated salivary-flow rates (g/min) during different test and pharmacologic conditionst
Resting flow Stimulated flow
(X & SE) (X k SE)
k!
Test/Drug Condition Before After Before After
4
Experiment 1: Placebo 0.38 f 0.11 0.98 kO.16 9
(n = 12) Benadryl 0.24+0.11** 0.66 * 0.15**
%
Elavil 0.21 + 0.08** 0.50 * 0.10**

Experiment 2: Placebo 0.46 k 0.10 0.54 * 0.10 1.11 k0.18 1.28+0.17

(n = 9) Benadryl 0.28 k 0.08** 0.34 * 0.07** 0.87 k 0.16* 0.91 + 0.17**
Elavil 0.30 + 0.09** 0.19f0.07** 0.84 + 0.15* 0.62 f 0.13**

Experiment 3: Placebo 0.27 k 0.05 0.40~0.10 0.95 + 0.22

(n = 9) Atropine 0.09 * 0.04** 0.11 & 0.05** 0.63 f 0.18*
(1 .O mg dosage only)

Experiment 4: Placebo 0.36 k 0.06 0.39 k 0.06 0.83 If: 0.08

(n = 19) Atropine 0.12 * 0.03** 0.09 * 0.02** 0.44 * 0.07**
*p < 0.05 in post-hoc tests (Dunnett test or f-tests for difference scores). **p -C 0.01 in post-hoc tests (Dunnett test or t-tests
for difference scores). tMeasures of salivary flow before and after taste tests occurred at different times in each
experiment.
 Salivary flow reductions and taste 21

Experiment 2 The 0.8 mg dosage of atropine reduced resting

Taste recognition thresholds for NaCl were not salivary flow rates by an average of 56 per cent before
affected by pharmacologic reductions in salivary and after taste testing. The higher dosage of atropine
flow. The threshold values (mM mean + SEM) were: (1 .O mg) reduced resting salivary flow 67 per cent (75
placebo 17.8 f 4.0 (dry), 5.49 f 1.82 (wet); Elavil min) and 72 per cent (- 3 h) before and after taste
14.3 f 2.5 (dry), 4.23 + 1.79 (wet); and Benadryl testing and reduced stimulated flow 34 per cent (* 3
19.4 f 3.7 (dry), 7.76 f 1.97 (wet). A repeated mea- h). As indicated in Table 2, salivary-flow reductions
sures analysis of variance revealed no significant main with atropine were significant compared to placebo
effect of pharmacologic treatment, F(2,16) = 1.69, or measures (t(8) = 3.2w.50). Though atropine was
treatment/dry-wet tastant interaction, F(2,16) = more effective than Elavil and Benadryl in reducing
0.27. The sample size was too small for meaningful resting flow rates, it was not more effective in reduc-
analyses of threshold data with citric acid and qui- ing acid-stimulated saliva (Table 2). Whereas during
nine. As with detection thresholds, the recognition control conditions gustatory stimulation produced
thresholds were higher when the stimuli were approx. a 2-fold increase in salivary flow compared
presented in a dry form. to resting levels, more than a 5-fold increase was
Both medications significantly reduced resting and measured in individuals receiving atropine. This pat-
stimulated salivary flow rates whether measured be- tern indicates that acid stimulation can overcome a
fore or after taste testing (Table 2). The F(2,16) portion of the depressive effects of atropine on sali-
values ranged from 4.52 to 20.67. Compared to vary flow. Gustatory stimulation also overcame some
placebo levels, resting flow rates were reduced 35 per of the xerostomic effects of Benadryl and Elavil but
cent with Elavil and 47 per cent with Benadryl before the effects were much smaller (Table 2). The asym-
taste testing began; the reductions after taste testing metry in blocking resting and stimulated salivary flow
were 65 and 43 per cent respectively for Elavil and may be characteristic of atropine or may be a func-
Benadryl. Stimulated flow rates were not reduced as tion of the time after administration of the drug.
much by the medications. With Elavil, salivary-flow Stimulated flow was measured approx. 3 h after oral
reductions were 24 -per cent before and 52 per cent administration of atropine.
after taste testing and with Benadryl were 25 per cent Experiment 4
before and 35 per cent after taste testing. It is evident
from these data that the onset and duration of Once again, individuals receiving atropine demon-
salivary reductions are different for the two medica- strated a significant increase in sensitivity to wet and
tions. The maximum xerostomic effect was achieved dry sour taste stimuli compared to placebo levels;
sooner with Benadryl but, at the dosage used, a F(1,8) = 72.26, p 0.001 (see Fig. I). Aqueous sour
greater degree of xerostomia was achieved with thresholds shifted more in this experiment than the
Elavil. last where larger volumes of solution were used. This
finding would appear to support a dilution hypothe-
sis except that no other significant shifts occurred for
Experiment 3 the other aqueous taste thresholds (see Table 3). F
A significant increase was observed in sensitivity to values for the main effect and interaction term in-
both wet and dry citric acid taste stimuli following volving pharmacologic treatment ranged from 0.00 to
atropine administration (Fig. 1). This was reflected in 2.59 (1 and 8-9 degrees of freedom).
a significant main effect for the drug condition;
F(l,8) = 5.34, p < 0.05. A trend was evident towards
1 Oml 5ml
increased sensitivity to aqueous sucrose solutions
following administration of atropine (Table 3) but it
was not statistically significant; drug/wet-dry tastant
interaction, F(1,8) = 4.44, p = 0.07. No other
significant changes in threshold were found (Table 3).
The F values for the main effect and interaction term
involving pharmacologic treatment for bitter and
salty taste stimuli ranged from 0.00 to 0.05 (1 and 8
degrees of freedom).
The direction of the threshold shifts for the 2
aqueous taste qualities suggest that salivary dilution
may underlie the diflerences between the atropine and
placebo conditions. Following atropine adminis-
tration, less saliva may be present to dilute the
aqueous test solutions resulting in increased taste
sensitivity. In support of this hypothesis, all subjects
PA PA PA PA
receiving 1.0 mg of atropine showed a large shift in
aqueous sour thresholds whereas, on average, sub- TESTING CONDITION

jects receiving 0.8 mg demonstrated no shift. As

Fig. 1. Recognition thresholds (mean + standard error) for
described below, the 1.0 mg dosage reduced salivary aqueous and dry citric acid stimuli during placebo (P) and
flow more than either the lower dosage of atropine or atropine (A) conditions. The left portion of the figure
the other tested medications and perhaps, a critical depicts the results when 10 ml quantities of taste solution
level of salivary-flow reduction was obtained for the were used for taste testing; and the right portion, when 5 ml
manifestation of taste changes. quantities were used.
22 CAROL M. CHRISTENSEN et al.

The consistent effect of atropine on only wet and

dry sour thresholds pointed to a possible drug-related
change in salivary composition. The effects of atro-
pine on salivary composition have been examined
(Shannon, Suddick and Chauncey, 1969). Of the
salivary constituents measured, bicarbonate levels
were most affected by atropine. In that report, the
concentration of salivary bicarbonate dropped from
0.77 to 0.24 m-equiv./l following an oral dose of l/50
grain (1.3 mg) of atropine sulphate. Bicarbonate is
the principal buffer in saliva and changes in its
concentration would be expected to most affect taste
sensitivity to acids. Bicarbonate levels rise when
salivary flow is stimulated (Dawes, 1974). This effect
may be responsible for the greater changes in the
aqueous compared to dry sour thresholds because the
oral manipulation required to express the taste sub-
stance from the dry filter papers would stimulate
salivary flow.
The pattern of salivary flow reductions following
atropine administration was similar to the previous
study (Table 2). Resting salivary flow was reduced 67
per cent (60 min) and 77 per cent (_ 2f h) by atropine
and stimulated flow (measured in only 10 subjects)
was reduced 47 per cent (~2; h). As indicated in
Table 2, salivary-flow reductions with atropine were
significant (t(9) = 4.78; t(18) = 5.266.00). Only one
subject showed no salivary flow reduction with atro-
pine and that observation was excluded from all
analyses.

DISCUSSION

These experiments demonstrate that large, short-

term decreases in salivary flow have little impact on
taste thresholds. A similar conclusion was drawn
from studies with supra-threshold taste stimuli
(Christensen, Navazesh and Brightman, 1984) where
pharmacologic reductions in salivary flow did not
produce shifts in the perceived intensity of a range of
aqueous- and dry-taste stimuli. We predicted that
reductions in salivary flow would affect the percep-
tion of dry-taste stimuli by reducing the rate of
release as well as the total quantity of tastant leached
from the dry matrix resulting in a decrease in taste
sensitivity. The opposite result was predicted for
aqueous solutions. Increased sensitivity to dissolved
tastants could follow from reductions in flow rate
because less saliva would be available to dilute taste
solutions in the mouth. Neither prediction was
confirmed. The selective shift in sour taste sensitivity
that was observed is more likely due to alterations in
salivary composition than to flow rate because the
predicted effects of flow rate on taste should occur
with all tastants.
The results cast doubt on the interpretation that
taste changes following surgical desalivation are di-
rectly mediated by large salivary-flow reductions. Of
course, the two methodologies for reducing salivary
flow are not completely comparable. Surgical and
pharmacologic desalivation would be expected to
produce different magnitudes and patterns of
salivary-flow changes. Neither method produces
complete xerostomia and residual salivary func-
tioning in the rested and stimulated states may be
quite different. In this study, pharmacologic reduc-
 Salivary flow reductions and taste 23

tion of flow rate did not eliminate the salivary Catalanotto F. A. and Sweeney E. A. (1972) The effects of
response to stimulation (see Table 2). In fact, when surgical desalivation of the rat upon taste acuity. Archs
subjects received the test drugs, particularly atropine, oral Biol. II, 1455-1465.
they showed an exaggerated salivary increase with Christensen C. M., Navazesh M. and Brightman V. J. (1984)
Effects of pharmacologic reductions in salivary flow on
stimulation. Thus, gustatory stimulation could over-
suprathreshold judgments of taste stimuli. Chem. Senses
come some of the depressive effects of the pharma- (in press).
cologic agents though flow rates still remained below Conger A. D. (1973) Loss and recovery of taste acuity in
normal. The capacity to respond to stimulation dur- patients irradiated to the oral cavity. Radiut. Res. 53,
ing pharmacologic depression of flow rate may have 338-347.
prevented a taste decrement from being observed Cornsweet T. N. (1962) The staircase method in psycho-
with the dry taste stimuli because they would be physics. Am. J. Psychol. IS, 485491.
expected to stimulate salivary flow. Dawes C. (1974) The effects of flow rate and duration of
The role of salivary bicarbonate in oral perception stimulation on the concentrations of protein and the main
electrolytes in human submandibular saliva. Arch oral
of acids has never been explored. The selective
Biol. 19, 887-895.
changes in sour sensitivity observed following admin- Dawes C. and Chebib F. S. (1972) The influence of previous
istration of atropine, which is known to reduce stimulation and the day of the week on the concentrations
salivary bicarbonate levels, suggest that salivary bi- of protein and the main electrolytes in human parotid
carbonate may affect oral sensitivity to acids. Al- saliva. Archs oral Biol. 17, 1289-1301.
though this set of ex.periments was designed to mea- Galili D., Maller 0. and Brightman V. J. (1978) Effects of
sure the effect of salivary flow on taste thresholds, the drug-desalivation on feeding and taste preferences in the
pattern of selective taste changes suggest that possible rat.-Archs oral Biol. 23, 459-464.
drug-induced alterations in salivary composition Henkin R. I.. Talal N.. Larson A. L. and Mattern C. F. T.
(1972) Abnormalities of taste and smell in Siogren’s
rather than alterations in flow mediated the observed
syndrome. Am. int. Med. 76, 375-383. - -
changes in taste thresholds. Mavhall C. W. (1975) The ohvsioloaical roles of saliva.
klnbama J. me>. SC>. 12, 45-63. -
Acknowledgemenfs-This research was supported in part by McBurney D. H. and Pfaffmann C. (1963) Gustatory adap-
N.I.H. grant DE-05545 to the first author. Subjects were tation to saliva and sodium chloride. J. exp. Psychol. 65,
tested at the W. D. Miller Clinical Research Center, Univer- 523-529.
sity of Pennsylvania. Appreciation is expressed to Mary Morino T. and Langford H. G. (1978) Salivary sodium
Mahoney and Lisa C.asper for their technical assistance. correlates with salt recognition threshold. Percept. Psy-
chophys. 21, 45-48.
REFERENCES Nanda R. and Catalanotto F. A. (1981) Long-term effects
of surgical desalivation upon taste acuity, fluid intake,
Bartoshuk L. M. (I 978) The psychophysics of taste. Am. J. and taste buds in the rat. J. dent. Res. 60, 69-76.
clin. Nutr. 31, 1068-1077. Navazesh M. and Christensen C. M. (1982) A comparison
Cano J. and Rodriguez-Enchandia E. L. (1980) Degener- of whole mouth resting and stimulated salivary mea-
ating taste buds in sialectomized rats. Acfa unar. 106, surement procedures. J. dent. Res. 61, 1158-l 162.
487492. Shannon I. L., Suddick R. P. and Chauncey H. H. (1969)
Catalanotto F. A. and Leffingwell C. (1979) Early effects of Effect of atropine-induced flow rate depression on the
desalivation upon fluid consumption and taste acuity in composition of unstimulated human parotid fluid. Archs
the rat. Behau. neurtrl Biol. 25, 19&205. oral Biol. 14, 761-770.