Sie sind auf Seite 1von 8

Physiology & Behavior 75 (2002) 435 442

Paradoxical effects of hedonic disparities in negative


anticipatory contrast
Nathan D. Mossa,1, J. Christopher Clarkeb, E. James Kehoeb,*
a

School of Applied Vision Sciences, University of Sydney, Sydney, NSW, Australia


School of Psychology, University of New South Wales, Sydney, NSW 2052, Australia

Received 12 March 2001; received in revised form 14 August 2001; accepted 12 September 2001

Abstract
Negative anticipatory contrast (NAC) occurs when the presentation of two solutions in sequence over a number of trials leads to the
suppression in consumption of the first solution. A hedonic disparity between the two solutions may lead to this suppression. However,
what the relative role of the gustatory properties versus the nutricaloric loads of the two solutions is in determining the acquisition of NAC is
uncertain. Previous experiments have typically used saccharine and sucrose solutions, which resemble each other in gustatory properties yet
differ in their nutricaloric loads. In contrast, the present experiments used soy milk and sucrose solutions, which are both highly nutritive but
differ in their gustatory properties. Soy milk was found to have a higher hedonic value than a 16% sucrose solution as measured by both
choice and absolute consumption. According to a hedonic disparity hypothesis, NAC should have occurred using a sucrose soy sequence
but not a soy sucrose sequence. Paradoxically, the sucrose soy sequence failed to yield NAC, but the soy sucrose sequence did yield a
repeatable, significant NAC. A consideration of the available theory and research indicated that NAC may be explained by a strong
conditioning of a satiety response produced by sucrose that opposes the conditioning of sucroses positive hedonic response. D 2002 Elsevier
Science Inc. All rights reserved.
Keywords: Negative anticipatory contrast; Hedonic disparity

1. Introduction
The consumption of food by rats is affected by foods that
they anticipate to come. For example, consumption of a
saccharine solution becomes suppressed when it is repeatedly followed by a sucrose solution [4 6,8,9]. A similar
suppression of a saccharine solution has also been acquired
when it has been followed by a range of other foods (e.g. lab
chow, chocolate milk, skim milk) [14]. This finding, known
as negative anticipatory contrast (NAC), relies on the
mechanisms of Pavlovian conditioning. For example, the
magnitude of the suppression depends on the degree of
contiguity between the two solutions [4,14].
NAC, however, presents a paradox for Pavlovian conditioning theories, because the pairings of the saccharine
solution with a second solution should cause an increase in
the consumption of the first solution, not a decrease
* Corresponding author. Tel./fax: +61-2-9385-3826.
E-mail address: j.kehoe@unsw.edu.au (E.J. Kehoe).
1
Present address: School of Psychology and Counselling, Queensland
University of Technology, Beams Road, Carseldine, Qld 4034, Australia.

[2,12,13]. To add to the paradox, pairings of plain water


with a sucrose solution do lead to an increase in the
consumption of water, a positive anticipatory contrast, which
is more in keeping with the conventional outcome of
Pavlovian conditioning [10].
The key determinants of NAC remain unclear. A disparity in the hedonic qualities of the first and second solutions
may be the key factor [3,16]. That is, after several pairings,
the animals compare the hedonic value of the saccharine
solution as they consume it with the expected value of the
more preferred, sucrose solution. This unfavorable comparison leads to a reduction in the consumption of the saccharine solution. However, this theoretical approach runs into
immediate difficulty in explaining the positive anticipatory
contrast seen when water is paired with a sucrose solution.
In the absence of a well-operationalized definition of hedonic value, water would seem to have at least as a large
hedonic disparity relative to sucrose as saccharine does;
hence, NAC should have occurred.
Even if a hedonic disparity hypothesis is restricted to
real and artificial sugar solutions, a hedonic disparity

0031-9384/02/$ see front matter D 2002 Elsevier Science Inc. All rights reserved.
PII: S 0 0 3 1 - 9 3 8 4 ( 0 1 ) 0 0 6 5 2 - 7

436

N.D. Moss et al. / Physiology & Behavior 75 (2002) 435442

hypothesis cannot explain the available results. Recently,


solutions of two sugarsfructose and glucosehave been
paired in a factorial fashion [7]. Given a choice between
solutions of equal concentrations, rats do not show a
consistent preference for one or the other, suggesting they
have equal hedonic value. Yet, contrast effects did appear
when the solutions were repeatedly presented in succession, but the results did not fit any simple pattern. NAC
appeared only when a low concentration of a sugar was
paired with a greater concentration of the same sugar.
When the two different sugars were paired, glucose
enhanced the intake of fructose, whether it was presented
before or after the fructose. To add to the confusion,
Gomez and Grigson [11] failed to find NAC when a
low-concentration sucrose solution was paired with a
higher-concentration sucrose solution.
A more analytic approach to anticipatory contrast effects
has been proposed, in which the relative hedonic value of
two foods can be viewed as a global psychophysical judgement measured behaviorally in a preference test [16]. However, a disparity in this global judgement could arise from
any or all of three sources. First, two foods may differ in their
nutricaloric load, that is, their nutrient and caloric potentials
that are realized during digestion and absorption. Second,
two foods may differ in their gustatory properties, for
example, taste, smell, and texture, which are detected rapidly
as they reach the appropriate sensory receptors. Third, as
with all psychophysical judgements, the hedonic value of
two foods may also depend on the context in which they are
presented, including the animals previous history with those
foods and other foods, in the immediate and/or remote past.
The present experiments attempted to delineate the
relative contributions of hedonic and nutricaloric differences to NAC. In particular, a saccharine solution
(Experiment 1A) was replaced in subsequent experiments
by soy milk. As will be shown, soy milk has a measurably
greater hedonic value than the sucrose solutions as used
here. Furthermore, soy milk has a similar nutricaloric load
to the sucrose solutions. Specifically, the soy milk used in
the present experiments was fairly balanced in macronutrients, with each 10 ml containing 0.47 g carbohydrates
(0.16 g as sugars), 0.34 g protein, 0.34 g fat, and yielding
6.2 cal. Four solutions of sucrose were used, which had
nutricaloric loads that bracketed that of the soy milk.
Specifically, the solutions contained sucrose concentrations
of 4%, 8%, 16%, and 32%. Each 10 ml of the 4%
solution contained 0.42 g of sucrose and 1.72 cal; the
8% solution contained 0.87 g of sucrose and 3.57 cal; the
16% solution contained 1.9 g of sucrose and 7.79 cal; and
the 32% solution contained 4.7 g of sucrose and 19.27 cal.

2. Experiment 1A
Experiment 1A was aimed at delineating the effects of
manipulating the concentration of the sucrose solution in a

conventional anticipatory contrast paradigm, in which the


first solution (S1) was saccharine and the second solution
(S2) was sucrose. This experiment was run to ascertain the
reliability of NAC prior to using soy milk as the first
solution in Experiment 1B. All four sucrose concentrations
were used (4%, 8%, 16%, and 32%) and compared to a
control group that received only the saccharine solution.
According to both hedonic and nutricaloric disparity hypotheses, the amount of saccharine consumed should
decrease as the sucrose concentration is increased.
2.1. Method
2.1.1. Subjects
The subjects were 40 male Wistar rats aged 160 days at
the start of the experiment. All the rats had been used in
previous experiments unrelated to eating behavior. The rats
had 18-h access to water and were fed for a 1-h period each
day. Subjects were maintained on a 12-h light 12-h dark
cycle, in which the lights were on from 6 a.m. to 6 p.m.
2.1.2. Apparatus
The rats were housed in groups of eight. Training trials
occurred in individual cages, which were 30-cm cubes. With
the exceptions of the faceplate and the floor, the test cages
were of all metal construction, with a 10-cm diameter fan in
the back wall to provide ventilation and background noise.
This fan was protected by a metal grill. The base plate of
each test cage was a metal grill. There was a tray of wood
shavings placed below the cage. The faceplate consisted of a
transparent Perspex sheet. The upper left-hand corner of the
faceplate formed a 20  20-cm door. At 3 cm left and 4 cm
up from the bottom right-hand corner of the faceplate was a
4-cm diameter plug in which a small hole had been drilled
for the insertion of a sipper tube. Sipper tubes were attached
to 16 feeding tubes for the presentation of the saccharine
and the sucrose solutions. A fresh 0.15% saccharine solution
(by weight) and 4%, 8%, 16%, and 32% sucrose solutions
(by weight) were prepared on the day prior to each experimental day. The 32% sucrose solution required mild heating
to achieve complete mixing. All solutions were maintained
at room temperature (approx. 20 C).
2.1.3. Procedure
Subjects were randomly placed into one of five experimental groups (n = 8) labeled as: Sacch, 4%, 8%, 16%, and
32%. The labels refer to the concentration of sucrose in the
S2 solution. Group Sacch received no solution as S2, which
is the conventional control condition in demonstration of
NAC. Subjects were restricted to a 1-h/day feeding schedule
from 4 to 5 p.m. throughout the experiment and for 4 days
prior to its start. Ambient training took place on the last 2
days before experimentation, in which subjects were placed
in the test cages without solutions being present. For the
following 13 consecutive days, one experimental trial was
given each day.

N.D. Moss et al. / Physiology & Behavior 75 (2002) 435442

437

Fig. 1. Consumption of a saccharine solution as S1 (left panel) and a sucrose solution as S2 (right panel) as a function of Days in Experiment 1.

Water was removed at 10 a.m., 1 h prior to testing and


replaced at the feeding time at 4 p.m. For each group, a
0.15% saccharine solution was presented for 5 min and then
withdrawn for 5 min. After this delay, either a 4%, 8%,
16%, or 32% sucrose solution was presented for 5 min.
Group Sacch remained in the cages for a further 5 min, but
with no solution or sipper tube present.
2.1.4. Measurement and statistical analysis
The amount of saccharine solution and the amount of
sucrose solution consumed was recorded to the nearest
0.5 ml. Inspection of the data revealed that all rats
reliably consumed both the saccharine and sucrose solutions in measurable amounts. Statistical analyses were
conducted using planned contrasts for repeated measure
designs as recommended by OBrien and Kaiser [17].
This type of analysis allowed the concentration of statistical power on testing relevant hypotheses concerning
specific differences among groups.
2.2. Results
Fig. 1 (left panel) shows the mean consumption of the
saccharine solution (S1) across days in Experiment 1A.
Consumption after Day 1 has been averaged across blocks
of three trials for all groups, for graphical purposes only.
To decrease the variance introduced by neophobia, the
results of Day 1 were excluded from all analyses.
The consumption of saccharine (S1) on Day 1 was
uniformly low for all groups. Thereafter, consumption of
saccharine rose in all groups but with distinct differences. For
Days 2 4, consumption of saccharine was greater in Group
Sacch than in Groups 4%, 8%, and 16%, but was lower than
Group 32%. From Days 5 to 7 onward, the consumption of

saccharine in Group Sacch increased while the consumption


of saccharine in the other groups decreased in absolute
measure to levels below that of Group Sacch, for all but
Group 4%, which remained at a level equal to that of Group
Sacch on Days 11 13. The overall mean consumption of
saccharine for Groups Sacch, 4%, 8%, 16%, and 32% across
Days 2 13 were 5.3, 5.8, 5.1, 4.3, and 4.0 ml, respectively.
These means differed significantly in a linear fashion, with
Group Sacch serving as a 0 point, F(1,35) = 13.78, P < .01.
Consumption increased for all groups in a linear fashion
across Days 2 13, F(1,35) = 11.91, P < .01.
Fig. 1 (right panel) shows the mean consumption of the
sucrose solution (S2) across days. Overall differences in
sucrose consumption occurred among the groups in an
inverted U-shaped pattern. That is, consumption of 4%
and 32% sucrose was lower than consumption of 8% and
16% sucrose. The mean consumption across Days 2 13
for Groups 4%, 8%, 16%, and 32% were 8.3, 9.5, 9.7, and
7.9 ml, respectively. The inverted U (quadratic) pattern of
differences across groups was confirmed by statistical
analysis, F(1,28) = 15.92, P < .01. Consumption across days
increased in a curvilinear fashion, which was confirmed by
both significant linear F(1,28) = 48.39, P < .01, and quadratic trends, F(1,28) = 41.08, P < .01.
2.3. Discussion
The results of Experiment 1A confirmed the findings in
other laboratories that NAC could be obtained reliably with
a saccharine sucrose sequence. As predicted by both the
hedonic and the nutricaloric disparity hypotheses, the
suppression of consumption of the saccharine solution
varied as a direct function of the concentration of the
sucrose solution. Moreover, the absolute consumption of

438

N.D. Moss et al. / Physiology & Behavior 75 (2002) 435442

the sucrose solution generally rose with its concentration. In


an exception to this finding, the 32% sucrose solution was
consumed less than the 8% and 16% sucrose solutions. At a
32% concentration, the sucrose solution became viscous.
Although the solution flowed freely through the sipper
tubes, it was very sticky to the touch. Observation of the
rats indicated that they consumed the 32% solution in
shorter bouts, and spent more time preening their whiskers
(cf. Ref. [15]).

3. Experiment 1B
While NAC emerged reliably using saccharine as S1 and
sucrose as S2, the two solutions differed in their nutricaloric
load, their known hedonic value, and to some extent, their
gustatory properties. It has yet to be shown whether anticipatory contrast occurs between two solutions that are both
high in nutricaloric load but are distinguishable in other
respects. Accordingly, Experiment 1B investigated the
results of pairing two solutions, both high in nutricaloric
load, namely, soy milk (S1) and sucrose (S2). As in Experiment 1A, this study parametrically varied the concentration of the sucrose solution.
3.1. Method
3.1.1. Subjects
The subjects were 40 male Wistar rats, aged 160 days.
All subjects were obtained and maintained as in Experiment 1A.
3.1.2. Apparatus and procedure
The apparatus and procedure were identical to those of
Experiment 1A, except soy milk replaced the saccharine

solution as S1 for all groups. Thus, there were five groups


(n = 8) labeled as: Soy, 4%, 8%, 16%, and 32%. The labels
refer to the concentration of sucrose in the S2 solution.
Group Soy received only presentations of the S1 soy milk
without the S2 sucrose solution, similar to Group Sacch in
the previous experiment.
3.2. Results
Fig. 2 (left panel) shows the mean consumption of the
S1 soy milk. All groups increased their consumption of
soy milk from Days 2 4 to 5 7. From that point,
consumption appeared to plateau for all groups except
Group Soy. By Days 11 13, all the groups that received
pairings of soy milk with sucrose consumed less soy milk
than Group Soy. The overall group means for Days 2 13
were 12.3, 12.0, 11.9, 11.3, and 11.2 ml for Groups Soy,
4%, 8%, 16%, and 32%, respectively. Statistical comparisons confirmed there was a significant downward linear trend in soy milk consumption across groups,
F(1,35) = 4.58, P < .05. In addition, soy milk consumption
increased across experimental days in a curvilinear fashion,
linear F(1,35) = 172.84, P < .01, quadratic F(1,35) = 36.06,
P < .01.
Fig. 2 (right panel) shows the mean consumption of the
S2 sucrose solution. As in Experiment 1A, overall differences in sucrose consumption occurred between the groups
in an inverted U-shaped pattern. That is, consumption of
the 4% and 32% sucrose solutions was again lower than
consumption of the 8% and 16% sucrose solutions. The
mean consumption of sucrose across Days 2 13 were 6.2,
7.5, 7.7, and 5.6 ml for Groups 4%, 8%, 16%, and 32%,
respectively. This quadratic pattern was statistically significant, F(1,28) = 13.21, P < .01. Consumption also
increased significantly for all groups across days in a

Fig. 2. Consumption of soy milk as S1 (left panel) and a sucrose solution as S2 (right panel) as a function of Days in Experiment 1.

N.D. Moss et al. / Physiology & Behavior 75 (2002) 435442

curvilinear fashion, linear F(1,28) = 64.46, P < .01, quadratic F(1,28) = 4.82, P < .05.
3.3. Discussion
Experiment 1B revealed that NAC appeared for two
solutions that were both high in nutricaloric load. The
suppression of soy milk consumption was a direct function
of the sucrose concentration. Moreover, the size of the
effect was about the same as in Experiment 1A. For
example, in Experiment 1A, Group 32% drank 1.3 ml less
of the saccharine solution than Group Sacch, that is, 4.0
versus 5.3 ml. Similarly, in Experiment 1B, Group 32%
drank 1.1 ml less soy milk than Group Soy, that is, 11.2
versus 12.3 ml. However, the gross consumption of soy
milk was greater than that of the sucrose solutions. All the
groups that received soy sucrose pairings drank more than
11 ml of soy milk averaged across Days 2 13, but no
group drank more than 8 ml of their sucrose solution.
The animals consumption of the S1 soy milk was
sensitive to the concentration of the S2 sucrose solution,
but not in the way anticipated by the caloric disparity
hypothesis. According to the caloric disparity hypothesis,
NAC should only have occurred when the S2 sucrose
solution had a greater caloric load than the S1 soy milk. In
the present experiment, only the 16% and 32% solutions
had a greater caloric load than the soy milk, but NAC was
also obtained with the 4% and 8% sucrose solutions,
which had a lower caloric load than the soy milk.
There are two alternative hypotheses that might
explain the greater consumption of S1 soy milk versus
the S2 sucrose solution. First, the soy milk may have a
lower hedonic value than the sucrose solutions. However,
the absolute consumption of soy milk was greater than
that of the sucrose solution, which suggests that soy
milk had a greater hedonic value. If so, the suppression
in soy milk consumption as a result of pairing with the
sucrose solution would present a paradox for a hedonic
disparity hypothesis. Second, since the soy milk was
always presented before the sucrose solution, there may
have been suppression of sucrose consumption due to
satiation arising from the considerable consumption of
the soy milk.

439

4.1. Method
4.1.1. Subjects
The subjects were eight rats of the same type used in
Experiments 1A and 1B.
4.1.2. Apparatus and procedure
The rats were maintained as in previous experiments, with
1 h feeding at 4 p.m. each day, and water being removed at
10 a.m. The rats were placed into single cages for testing at
11 a.m. on three successive days. For 15 min, each rat was
given access to two tubes, one containing soy milk, the other
containing the 16% sucrose solution. The left/right positions
of the solutions were counterbalanced across rats and alternated across days for each rat in an ABA design.
4.2. Results and discussion
The rats consumed significantly more soy milk
(mean = 9.4 ml) than the 16% sucrose solution (mean = 2.6
ml), F(1,7) = 16.56, P < .01. Therefore, on this basis, the
subjects preferred soy milk to the 16% sucrose solution. In
conjunction with the greater consumption of soy milk
relative to sucrose in Experiment 1B, soy milk appears to
have a substantially greater hedonic value than the sucrose
solution. However, as noted previously, soy milk always
occurred before the sucrose solution in Experiment 1B,
which may have lowered the consumption of sucrose
through satiation.

5. Experiment 3
Experiment 3 was conducted with five aims: (1) to
replicate the NAC seen in Experiment 1B when soy milk
preceded sucrose solutions, (2) to determine whether a
contrast effect would appear when sucrose preceded soy
milk, (3) to determine whether a contrast effect would
appear when the same two solutions are presented in
sequence, (4) to determine the effects of consumption of
the first solution on the second solution, and (5) to determine the absolute consumption of the sucrose solution in
isolation from pairings with soy milk.
5.1. Method

4. Experiment 2
In Experiment 1B, the absolute consumption of soy
milk exceeded that of sucrose, suggesting that soy milk
may have the greater hedonic value. Experiment 2 was
undertaken to provide a converging measure of the relative
hedonic value of soy milk versus the sucrose solutions by
using a preference test. Specifically, rats were given a
choice between soy milk and the 16% sucrose solution.
The 16% solution was chosen for comparison, because its
caloric load was closest to that of the soy milk.

5.1.1. Subjects
The subjects were 56 rats of the same type used in the
previous experiments.
5.1.2. Apparatus and procedure
Unless otherwise stated, the apparatus and the procedure
were the same as in Experiments 1A and 1B. Subjects were
randomly placed into one of six experimental groups (n = 8),
which received either soy milk sucrose pairings (Soy
Suc), sucrose soy milk pairings (Suc Soy), soy milk soy

440

N.D. Moss et al. / Physiology & Behavior 75 (2002) 435442

milk pairings (Soy Soy), sucrose sucrose pairings (Suc


Suc), soy milk alone (Soy), or sucrose alone (Suc). The 16%
sucrose solution was used in all the relevant groups.
5.2. Results
Fig. 3 (left panels) shows consumption of S1 across days.
From a low level of consumption on Day 1, consumption of
soy as S1 increased across Days 2 13 for all groups, with
consumption leveling off from approximately Days 5 to 7
(upper left panel). From Days 2 to 4 onwards, NAC
appeared in Group Soy Suc. That is, it consumed significantly less soy milk than either Group Soy or Soy Soy,
F(1,42) = 4.23, P < .05. Hence, the NAC seen in Experiment
1B was replicable. In contrast, the increase in consumption
of sucrose as the S1 was relatively uniform across groups
(lower left panel). In particular, Group Suc Soy showed no
evidence of a contrast effect in either direction relative to
either Group Suc or Group Suc Suc.
Statistical analyses confirmed that consumption of S1
increased in a curvilinear fashion across days, linear
F(1,42) = 155.48, P < .01, quadratic F(1,42) = 58.12,
P < .01. More soy milk as S1 was consumed than the 16%
sucrose solution as S1, F(1,42) = 4.63, P < .05. Mean consumption of S1 soy milk in Groups Soy Suc, Soy Soy,
and Soy were 10.4, 11.9, and 12.6 ml, respectively, while
mean consumption of the sucrose solution as S1 in Groups
Suc Soy, Suc Suc, and Suc were 10.7, 9.8, and 10.4 ml,
respectively. A comparison of Groups Soy and Suc, for
which there was no possible contrast effect, confirmed that

soy milk was consumed more than the sucrose solution


F(1,42) = 4.25, P < .05.
Fig. 3 (right panels) shows the mean consumption for
both soy milk and sucrose when serving as S2. Consumption
of S2 increased for all groups until Days 5 7, after which
consumption remained level, linear F(1,28) = 37.47, P < .01,
quadratic F(1,28) = 45.81, P < .01. Analysis of the overall
means revealed that soy milk as S1 and/or S2 had additive
effects on the consumption of S2. Group Soy Soy showed
the greatest consumption of S2 (13.2 ml). Groups Soy Suc
(10.0 ml) and Suc Soy (10.8 ml) showed intermediate levels
of S2 consumption, and finally, Group Suc Suc (8.6 ml)
showed the lowest level of consumption. Statistical comparisons confirmed that there were main effects on the consumption of S2 by both the S1 solution, F(1,28) = 8.91,
P < .01, and the S2 solution, F(1,28) = 18.29, P < .01.
5.3. Discussion
As was seen in Experiment 1B, suppression in the
consumption of the S1 soy milk appeared when a soy
sucrose sequence was used. No positive or negative contrast, however, appeared when any other sequence was used.
The failure to see a contrast effect when using the sucrose
sucrose sequence is consistent with previous findings [6,11].
More importantly, the sucrose soy sequence failed to
produce any discernible suppression in the consumption of
the S1 sucrose solution. This pattern of results adds to the
dilemma for a hedonic disparity hypothesis. As soy milk is
both preferred to sucrose and consumed in greater quantity

Fig. 3. Consumption of soy milk and a 16% sucrose solution as S1 (left panels) and/or S2 (right panels) as a function of Days in Experiment 3.

N.D. Moss et al. / Physiology & Behavior 75 (2002) 435442

than sucrose, then a hedonic disparity hypothesis would


predict results exactly the reverse to those currently found.
That is, NAC should have appeared for the sucrose soy
sequence, and perhaps a positive anticipatory contrast
should have appeared for the soy sucrose sequence.
The pattern of results suggests that prior consumption of
sucrose had a negative effect on the consumption of S2. For
example, Group Suc Suc showed the lowest level of S2
consumption among the four groups that received two
solutions. In particular, Group Suc Suc (8.6 ml) showed a
lower level of S2 consumption in comparison to Group
Soy Suc (10.0 ml) and also in comparison to Group Suc
(10.4 ml), which received only one solution per day. A
similar impact of S1 sucrose appeared in the consumption of
S2 soy milk. Group Suc Soy (10.8 ml) showed less
consumption of soy milk than in comparison to either S2
consumption in Group Soy Soy (13.2 ml) or in comparison
to Group Soy (12.6 ml).
Although sucrose consumption appeared to have a satiating effect, there was no support for the hypothesis proposed in the discussion of Experiment 1B that prior
consumption of soy milk satiates the animal. Group Soy
Soy showed the greatest consumption of S2 (13.2 ml)
compared to the other two-solution groups. Moreover,
Group Soy Soys consumption was greater than that of
Group Soy (12.6 ml). Similarly, prior consumption of soy
milk had little if any impact on the sucrose solution as S2.
Group Soy Sucs consumption of S2 sucrose (10.0 ml) was
only slightly less than that of Group Suc (10.4 ml).
The present results provide converging evidence with the
results of Experiments 1B and 2 that soy milk had greater
hedonic value. Specifically, the comparison of Groups Soy
and Suc, for which there was no possible contrast effect,
revealed that soy milk was consumed in greater quantity
than the sucrose solution. Hence, it would not be possible to
explain the NAC seen when soy milk precedes sucrose by
asserting that soy milk has the lower hedonic value.

6. General discussion
The present experiments examined the contribution of
the caloric load and hedonic value of successive solutions to
the acquisition of NAC. First, a disparity in caloric load did
not appear to be a necessary or sufficient condition for
NAC. That is, NAC appeared whether the second solution
had either a lower or higher caloric load than the first
solution. Second, the present experiments revealed that
soy milk had a greater hedonic value as measured both by
a preference test (Experiment 2) and absolute consumption
(Experiments 1B and 3). Despite this greater hedonic value,
repeated presentations of a soy sucrose sequence depressed
consumption of the soy milk. This NAC paralleled that seen
using a saccharine sucrose sequence in Experiment 1A.
Conversely, a sucrose soy sequence failed to produce either
positive or negative contrast.

441

The pattern of findings presents a paradox for the caloric


disparity hypothesis. Across Experiments 1B and 3, NAC
was obtained whether the caloric load of the second solution
was less than or greater than the caloric load of the first
solution. Although NAC did not require a caloric disparity,
the animals nevertheless were sensitive to the caloric load of
the second solution. Specifically, in Experiment 1B, the
absolute consumption of the second solution as well as
the size of the NAC depended on the concentration of the
sucrose solution.
The present findings also present a paradox for the
hedonic disparity hypothesis. A solution with less hedonic
value produced NAC when preceded by a more highly
valued solution. A hedonic disparity hypothesis may be
defended but only by discounting the conventional methods
for measuring hedonic value. In fact, an inspection of the
effects of the sucrose solution does give some reason for
questioning whether the conventional methods pertain to
sucrose consumption. The results of Experiment 3 revealed
that consumption of the 16% sucrose solution depressed
consumption of the next solution, whether it was soy milk or
sucrose. In contrast, soy milk, when it was the first solution,
did not suppress consumption of the second solution.
Such a rapid satiating effect of sucrose may have reduced
its absolute consumption in a way that could have masked
its apparent hedonic value in both one- and two-bottle tests.
In the one-bottle test, rapid satiation may have curtailed
consumption within the 5-min used for presentation of the
sucrose solution as S1 or S2. In the two-bottle preference
test used in Experiment 2, the same satiating effect may
have also reduced overall consumption of either solution as
well as reducing specific consumption of sucrose. In the
two-bottle test, the absolute consumption for both sucrose
(2.6 ml) and soy milk (9.4 ml) was comparatively small
across the 15-min test period relative to the 5-min one-bottle
test periods. In the one-bottle tests in Experiment 3, for
example, Group Suc consumed 10.4 ml of sucrose, and
Group Soy consumed 12.6 ml of soy milk.
Apart from the conundrum for a hedonic disparity
hypothesis and measurement of hedonic value, the present
results do suggest that NAC may depend crucially on the
rapid satiety produced by sucrose consumption. A survey
of the literature reveals that this satiety depends on the
postingestive nutricaloric consequences of sugars rather
than their gustatory factors [1,18,19]. Conversely, the
positive hedonic value of sugars may depend largely on
their gustatory properties [21]. If sucrose has a positive
gustatory effect and a negative satiety effect, then NAC
may depend on the interplay between conditioning of both
these effects [21]. If so, then, in future research, the
opponent types of conditioning can be independently
manipulated. For example, the positive gustatory effect of
a sucrose solution can be reduced by the addition of
quinine. Combined with stomach loading of a sucrose
solution, it may be possible to enhance the acquisition of
NAC. Conversely, NAC might be reduced and positive

442

N.D. Moss et al. / Physiology & Behavior 75 (2002) 435442

anticipation exposed if the level of satiation is reduced by


the manipulation of blood sugar levels [18,20] or by sham
feeding [21].

[10]
[11]

Acknowledgments
Preparation of this manuscript was supported in part by
Australian Research Council Grant No. A79800067 to
E.J. Kehoe.

[12]
[13]

[14]

References
[1] Berthoud HR, Baettig K. Effects of nutritive and nonnutritive stomach
loads on plasma glucose level and lateral hypothalamic eating threshold in the rat. Physiol Behav 1974;12:1015 9.
[2] Capaldi ED, Sheffer JD. Contrast and reinforcement in consumption.
Learn Motiv 1992;23:63 79.
[3] Flaherty CF. Incentive contrast: a review of behavioral changes
following shifts in reward. Anim Learn Behav 1982;10:409 40.
[4] Flaherty CF, Checke S. Anticipation of incentive gain. Anim Learn
Behav 1982;10:177 82.
[5] Flaherty CF, Grigson PS. From contrast to reinforcement: role of
response contingency in anticipatory contrast. J Exp Psychol, Anim
Behav Processes 1988;14:165 76.
[6] Flaherty CF, Grigson PS, Checke S, Hnat KC. Deprivation state and
temporal horizons in anticipatory contrast. J Exp Psychol, Anim
Behav Processes 1991;17:503 18.
[7] Flaherty CF, Mitchell C. Absolute and relative rewarding properties of
fructose, glucose, and saccharin mixtures as reflected in anticipatory
contrast. Physiol Behav 1999;66:841 53.
[8] Flaherty CF, Rowan GA. Anticipatory contrast: within-subjects
analysis. Anim Learn Behav 1985;13:2 5.
[9] Flaherty CF, Rowan GA. Successive, simultaneous, and anticipatory

[15]

[16]

[17]

[18]
[19]

[20]

[21]

contrast in the consumption of saccharin solutions. J Exp Psychol,


Anim Behav Processes 1986;12:381 93.
Flaherty CF, Turovsky J, Krauss KL. Relative hedonic value modulates anticipatory contrast. Physiol Behav 1994;55:1047 54.
Gomez F, Grigson PS. The suppressive effects of LiCl, sucrose, and
drugs of abuse are modulated by sucrose concentration in fooddeprived rats. Physiol Behav 1999;67:351 7.
Holman EW. Immediate and delayed reinforcers for flavor preferences
in rats. Learn Motiv 1975;6:91 100.
Lucas GA, Timberlake W. Negative anticipatory contrast and preference conditioning: flavor cues support preference conditioning, and
environmental cues support contrast. J Exp Psychol, Anim Behav
Processes 1992;18:34 40.
Lucas GA, Timberlake W, Gawley DJ, Drew J. Anticipation of future
food: suppression and facilitation of saccharin intake depending on
delay and type of future food. J Exp Psychol, Anim Behav Processes
1990;16:169 77.
Mook DG, Culberson R, Gelbart RJ, McDonald K. Oropharyngeal
control of ingestion in rats: acquisition of sham-drinking patterns.
Behav Neurosci 1983;97:574 84.
Moss ND. The effects of stimulus disparities on negative anticipatory
contrast in food consumption. Sydney (Australia): School of Psychology, University of New South Wales, 1994.
OBrien RG, Kaiser MK. MANOVA method for analyzing repeated
measures designs: an extensive primer. Psychol Bull 1985;97:
316 33.
Oomura Y. Chemical and neuronal control of feeding motivation.
Physiol Behav 1988;44:555 60.
Sclafani A, Fanizza LJ, Azzara AV. Conditioned flavor avoidance,
preference, and indifference produced by intragastric infusions of galactose, glucose, and fructose in rats. Physiol Behav 1999;67:227 34.
VanderWeele DA, Haraczkiewicz E, Vasselli JR. Tolbutamide affects
food ingestion in a manner consistent with its glycemic effects in the
rat. Physiol Behav 1988;44:679 83.
Warwick ZS, Weingarten HP. Flavor postingestive consequence associations incorporate the behaviorally opposing effects of positive
reinforcement and anticipated satiety: implications for interpreting
two-bottle tests. Physiol Behav 1996;60:711 5.

Das könnte Ihnen auch gefallen