University of Southern California Dissertations and Theses

(USC Thesis Other)

Age-related changes in cognitive functioning and their implications for the reacquisition of conditioned taste avoidances

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
AGE-RELATED CHANGES IN COGNITIVE FUNCTIONING AND THEIR
IMPLICATIONS FOR THE REACQUISITION OF CONDITIONED TASTE
AVOIDANCES

by

Alexandra E. Ycaza

A Thesis Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF ARTS
(PSYCHOLOGY)

December 2009

Copyright 2009 Alexandra E. Ycaza

ii
Table of Contents

List of Tables iv

List of Figures v

Abstract vi

Chapter 1: Age-Related Changes in Cognitive Functioning 1
1.1 Introduction 1
1.2 Differences in the Acquisition of a New Behavior 1
1.3 Differences in the Retention of a New Behavior 9
1.4 Glucocorticoids: The Possible Impact on Learning in Aged Animals 14
1.5 Concluding Remarks 23

Chapter 2: Age-Related Changes in Conditioned Taste Avoidances 24
2.1 Introduction 24
2.2 Acquisition of CTAs in Aged Animals 26
2.2.1 Effects of US Preexposure 26
2.2.2 Effects of Increased CS-US Interval 31
2.2.3 Effects of Increased Interval between Acquisition and the First
Extinction Test 34
2.3 Extinction of CTAs in Aged Animals 38
2.4 A Summary of the Results 45
2.5 Possible Explanations for the Differences in CTA Learning Observed
in Aged Animals 48
2.5.1 Physiological Changes 48
2.5.2 A Glucocorticoid Hypothesis 50
2.5.3 An Estrogen Hypothesis 55
2.6 Concluding Remarks 60

Chapter 3: Reacquisition of Conditioned Taste Avoidances 62
3.1 Aging and Reacquisition of a CTA 62
3.2 Experiments Investigating Reacquisition of a CTA in Young
Animals 64
3.2.1 Introduction 64
3.2.2 General Methods 66
3.2.2.1 Subjects 66
3.2.2.2 Surgical Procedures 67
3.2.2.3.Conditioning Procedures 67
3.2.2.4 Statistical Methods 68
3.3 Experiment 1 71
3.3.1 Introduction 71
3.3.2 Methods 71
3.3.3 Results 72
iii
3.3.3.1 Development of First CTA and Second CTA 72
3.3.3.2 Extinction of First CTA and Second CTA 78
3.3.3.3 Comparisons between Acquisition and Last Extinction
Tests of First CTA and Second CTA 81
3.3.4 Discussion 84
3.4 Experiment 2 93
3.4.1 Introduction 93
3.4.2 Methods 95
3.4.3 Results 97
3.4.3.1 Development of First CTA and Second CTA 97
3.4.3.2 Extinction of First CTA and Second CTA 103
3.4.3.3 Comparisons between Acquisition and Last Extinction
Tests of First CTA and Second CTA 105
3.4.4 Discussion 109
3.5 General Discussion 112
3.5.1 Comparisons between Experiment 1 and Experiment 2:
CTA1 versus CTA2 112
3.5.2 Comparisons between Experiment 1 and Experiment 2:
Retention of Avoidance 119
3.5.3 The Reacquisition of CTAs in Aged Animals 121

References 124

iv
List of Tables

TABLE 1. Results for Experiments 1 and 2 of Misanin, Hoefel,
Riedy, and Hinderliter (1997) 29

TABLE 2. CTA Formation Over Different CS-US Intervals and Doses
(Misanin, Goodhart, Anderson, and Hinderliter, 2002) 33

TABLE 3. Results of Guanowsky, Misanin, and Riccio (1983)
and Martinez and Rigter (1983) 36

TABLE 4. Results for Ingram and Peacock (1980) 40

TABLE 5. Results for Springer and Fraley (1981) 43

TABLE 6. Results of Misanin, Goodhart, Anderson, and Hinderliter
(2002); Guanowsky, Misanin, and Riccio (1983); Ingram and
Peacock (1980); Springer and Fraley (1981) 46

TABLE 7. Hormone Levels and Expressions of IEGs in Senba
and Ueyama (1997) 52

TABLE 8. Hypothesized Hormone and IEG Patterns in Young and Aged
Animals During Remote US Preexposure 54

TABLE 9. Experimental Design for Experiment 1 72

TABLE 10. Experimental Design for Experiment 2 96
v
List of Figures

FIGURE 1. Experiment 1: Development of First CTA and Second CTA 74

FIGURE 2. Experiment 1: Comparisons between the First CTA
and Second CTA 75

FIGURE 3. Experiment 1: Comparisons between the First CTA
and Second CTA Based on ACQ1 Consumption 78

FIGURE 4. Experiment 1: Consumption During Extinction of the
First CTA and Second CTA 79

FIGURE 5. Experiment 1: Comparisons between Extinction of the
First CTA and Second CTA 81

FIGURE 6. Experiment 1: Consumption During ACQ and LET of the
First CTA and Second CTA 83

FIGURE 7. Experiment 2: Development of First CTA and Second CTA 99

FIGURE 8. Experiment 2: Comparisons between the First CTA
and Second CTA 101

FIGURE 9. Experiment 2: Comparisons between the First CTA
and Second CTA Based on ACQ1 Consumption 103

FIGURE 10. Experiment 2: Consumption During Extinction of the
First CTA and Second CTA 105

FIGURE 11. Experiment 2: Consumption During ACQ and LET of the
First CTA and Second CTA 107
vi
Abstract
Both the human and animal literature report observations of marked age-related
declines in cognitive functioning. The existing evidence indicates that the larger deficit
involves the retention of newly learned behaviors, rather than the acquisition of those
behaviors. However, one learning and memory paradigm, conditioned taste avoidances,
suggests that aged animals show improvement in both learning and memory. Rather than
displaying the usual decline in acquiring and maintaining a new behavior, aged animals
show more robust learning and retention of these food-illness associations than younger
cohorts. Following is a review of the age-related changes observed in the more
traditional learning and memory paradigms, the age-related changes observed in the
conditioned taste avoidance paradigm, hypotheses to explain the observed differences in
both types of paradigms, and two preliminary experiments examining the impact of
extinction on the reacquisition of a conditioned taste avoidance.

1

Chapter 1
Age-Related Changes in Cognitive Functioning
1.1 Introduction
Although age-related declines in cognitive functioning have been well documented in
humans and animals, the specific tasks and functions that are most affected by the
process of aging have only relatively recently been methodically examined. Through the
use of the many different learning paradigms it has been possible to identify some of the
more specific features of learning and memory that are affected by this natural process.
In this paper, some of the age differences that have been observed are discussed and a
possible explanation is proposed. Interestingly, despite the popular notion that aging is
accompanied by general declines in all areas of learning and memory, there seem to be
some learning paradigms that are more affected than others and even one where learning
seems to take place more readily and retention appears to be improved.
One of the problems in comparing the results of the different studies is that
investigators define the age constituting young, middle-aged, and old differently. Thus,
for the purposes of this paper, I will use the following definitions: 2 to 8 months are
young, greater than 8 months to 19 months are middle-aged, greater than 19 months to 30
months are old.
1.2 Differences in the Acquisition of a New Behavior
In general, it has been observed that aged animals require more training before
meeting the set criteria said to indicate that learning has taken place. Acquisition deficits
have been noted in a number of different paradigms, including classically conditioned
2

tasks such as eyeblink conditioning and instrumental tasks such as acquired immobility,
shock-avoidance, and spatial learning (for review see Kubanis and Zornetzer, 1981) and
have been observed in a number of different strains of the rat. Differences in rates of
acquisition between aged and young animals have been observed in the classical
conditioning paradigm, eyeblink conditioning. In one study (Weiss and Thompson,
1991), non-food-or-water-deprived young (3 months), young-middle-aged (12 months),
old-middle-aged (18 months), and old (30 months) male Fischer 344 rats were trained to
associate a tone with a shock to the eye. The tone in this study was an 85db, 350ms burst
of white noise and it was accompanied by a 100ms, coterminating, periorbital shock.
Compared to the young and young-middle-aged groups, the old-middle-aged and old
groups required significantly more trials to reach the criteria for learning, which was
defined as displaying 8 consecutive eyeblinks in response to the tone when given 9 trials.
Age-related deficits also were found in the acquisition of an immobility task (Rigter,
Martinez, and Crabbe 1980). In this study, the authors indicated that the animals were
placed in a water bath from which they could not escape. Although not clearly specified,
the use of the same apparatus used by Porsolt (1977), who examined an animal model of
depression, was implied. In the Porsolt study, animals were placed in a 40cm tall
cylinder filled with 15cm of water for 15 minutes. Rigter and colleagues found that after
some time of fruitless struggling, both young (4 months) and aged (25 months) non-
deprived male Wistar rats learn to become and stay immobile, however, young animals
learned to become immobile more quickly than the old animals.
3

Three different active shock avoidance tasks have revealed age-related deficits in
acquisition. In one study (McNamara, Benignus, Benignus, and Miller Jr., 1977), non-
deprived old-middle-aged male and female Sprague-Dawley rats (18 months) were
compared to non-food-or-water-deprived post-weanlings (1 month), young (6 months),
and young-middle-aged (12 months) male and female Sprague-Dawley rats in a one-way
step-through active avoidance task. The animals were placed into one chamber of a
shuttle-box, with a closed door preventing the animal from entering the second chamber.
Once the separating door was opened, the animals had 5-seconds to escape to the newly
accessible chamber before shock was initiated. The observed deficit in acquisition was
that old-middle-aged rats required more trials than the post-weanlings, young, and young-
middle-aged adult rats before learning to escape to the safe chamber in order to avoid
shock. Rigter, Martinez, and Crabbe (1980) also reported deficits in another strain of rat
that was trained and tested in the one-way step-up shock-avoidance task. In this
experiment, young (4 months) and aged (25 months) non-deprived male Wistar rats were
placed in an apparatus in which the floor was set to deliver a shock. In order to avoid or
escape the shock, the animals had to climb atop a small platform in the center of the
apparatus. After 10 acquisition trials the young animals made significantly more escapes
onto the platform than the aged animals, indicating that the younger age group had
achieved a greater amount of learning in less time than was required by the aged animals.
The observation that more trials are required for aged animals to learn the same task
also has been reported in two-choice aversive maze conditioning. In a study by Vasquez,
Martinez, Jensen, Messing, Rigter, and McGaugh (1983), non-deprived young (2 months)
4

male and old (24 months) male and female Fischer 344 rats were trained in a shock
avoidance procedure set in a Y-maze. Animals were placed in the start alley and 10-
seconds later an 800µA shock was delivered to their paws. On the first trial, animals
were able to escape shock by running into either arm of the Y-maze. After this initial
trial, they were trained to avoid shock by running into the arm opposite of their initial
choice on trial 1. The aged animals required more trials to reach the criterion level than
the young animals. This difference has been reported across species as well. Dean,
Scozzafava, Goas, Regan, Beer, and Bartus (1981) made the same observations in non-
deprived male and female C57BL/6J mice who were trained in a similar procedure with
the exception that the apparatus was a T-maze and the arms of the T-maze were
differentiated by one arm being dark and the other being lit (a 60w bulb was placed
15.2cm above the arm). The results of the Dean et al. study (1981) are interesting in that
one would expect learning to be facilitated through the use of the additional cue of
brightness. The observation that aged animals still display difficulty in learning despite
the additional cue may speak to additional deficits in the ability to recognize spatial cues
in their environment. In fact, although some differences have been found in the above-
mentioned studies, aged animals do seem to have exceptional trouble with tasks that
require them to use spatial mapping strategies, with respect to both learning and selecting
the most effective strategy.
Spatial learning also is effected by age; studies using the Morris water maze task to
examine spatial learning capabilities of aged animals have found a minor deficit in the
initial leaning of the task. It appears that aged animals typically reach levels of learning
5

that are comparable to their young counterparts, however, they seem to require a larger
number of trials. In general, animals navigate and remember a location’s place-in-space
by using local cues that mark the location of the goal and the movements they made in
order to arrive at the goal (i.e., which turns were made and from where in the
environment they started). As animals age there is a decline in the use of these place-
learning strategies, further compounding the slower acquisition rate by making learning
even more difficult when the environment is altered in order to make the animals change
their search strategy (for review see Gallagher and Pelleymounter, 1988). When this
methodology is employed aged animals tend to show even greater deficits in learning.
For instance, in a series of experiments, Rapp, Rosenberg, and Gallagher (1987) found
that aged animals not only display deficits in acquisition of spatial learning tasks but they
also show a compromised ability to change strategies in a spatial learning situation
(animals were non-deprived during all experiments). In their first experiment, young (6
months), middle-aged (12 months), and aged (23 to 28 months) male Long-Evans rats
were trained in a water maze task. In one condition the tub had a visible black platform
protruding 1-cm above the water, while the other condition used the traditional
submerged platform; both conditions made use of many extra-maze cues consisting of
normal laboratory items, such as cabinets, sink, shelves, and the experimenter. Animals
were initially trained and tested using the submerged platform, after 31 training trials the
platform was removed and the animals were placed in the tub and allowed to swim for
90-seconds. For the first 16 training trials, the aged animals took longer to arrive at the
platform than the young and middle-aged animals, however, by the second half of the
6

trials (17-31) all animals performed equally; middle-aged animals, on the other hand,
performed equal to young animals on all trials. During the 90-second free swim test, all
animals showed preferences for the quadrant where the platform had been located,
indicating that all animals learned the platform’s location in space. The animals were
then trained using the visible black platform, and again no age differences were found. In
the second experiment, the visible platform was used for both phases. Young (4 months)
and aged (23 to 25 months) male Long-Evans rats were trained over 13 trials using the
visible platform, which again revealed no difference in the latencies to arrive at the
platform by the end of the trials. On trial 14, the platform was removed and animals were
given another free swim trial for 90-seconds. It was observed that the aged animals
showed no preference for the quadrant that had contained the platform, as they spent an
equal amount of time swimming in all quadrants, whereas the young animals spent most
of the time swimming in the former location of the platform. The animals then were
given a second set of learning trials in which the visible platform was moved to the
quadrant opposite to the one where the platform had been for the first 13 trials (e.g. from
the NE quadrant to the SW quadrant). The change in location led to longer arrival
latencies for aged animals on all training trials, when compared to the young animals. On
the 90-second free swim trial, the aged group again showed no preference for any
quadrant, whereas the young animals showed a preference for the quadrant housing the
new location of the platform. The results of their study indicate that aged animals learn
the Morris water maze task more slowly than their young counterparts, but depending on
the training procedure can eventually reach levels comparable to those shown by younger
7

groups. However, larger deficits become apparent when the aged animals are required to
change the strategy they use in order to locate the platform.
Aged animals also show deficits in another spatial task, the circular holeboard task
(Barnes, 1979; Barnes, Nadel, and Honig, 1980; Barnes and McNaughton, 1985). In this
task animals are trained to escape from an illuminated circular platform to a tunnel that is
connected to one of 18 holes distributed around the perimeter of the platform. There are
no cues as to which is the correct tunnel to enter within the maze; the animals must rely
on the extra-maze cues (if provided) and remembering which tunnels have been explored.
Young and aged male Long-Evans rats (Barnes, 1979; Barnes, Nadel, and Honig, 1980)
and young and aged male Fischer 344 rats (Barnes and McNaughton, 1985) perform
similarly on the first few trials, however, as trials progress, young animals reach criteria
significantly sooner and perform better on retention tests than aged animals. A potential
error in the interpretation that this is an acquisition deficit is that this procedure may be
measuring retention, and thus may be revealing a retention deficit. It is possible that the
aged animals may have trouble remembering which holes they have already explored,
causing them to require more time to learn which hole is the correct escape tunnel, since
they repeatedly enter already explored tunnels.
There also are several caveats that must be considered when making conclusions
based on the data presented as acquisition deficits in aged animals. First, some age-
related deficits in acquisition of a new behavior could be due to an increased incidence of
perseverative behaviors displayed by aged animals. Perseveration is commonly thought
of as the exhibition of behavioral rigidity, such that an animal insistently follows the
8

same course of behavior even in situations where a change in behavior would prove
beneficial. For instance, in the above-mentioned Vasquez et al. (1983) and Dean et al.
(1981) studies animals had trouble learning to escape shock by going into the arm
opposite of their initial choice on the first trial. However, when reversal training ensued,
and animals were trained to once again enter their initial arm of choice, aged rats and
mice learned as quickly as the young animals, indicating an inclination toward the initial
choice that was much stronger in the aged animals than in the younger animals. A
second issue that must be taken into account is the observation that rats, in general, are
agoraphobic, and this characteristic may become more prominent in aged animals. For
instance, in the Rigter, Martinez, and Crabbe (1980) study, where animals were trained to
escape a shock by climbing onto a platform in the center of the apparatus, it was noticed
that the aged animals rarely ventured out to the middle of the floor, rather they remained
up against the wall of the apparatus. Upon making this observation the researchers
moved the platform to a corner of the apparatus and found that aged animals learned just
as quickly as the young animals. Although some may interpret this as evidence of
perseveration in that the aged animals do not alter their behavior of remaining near the
wall, it also may be interpreted as evidence of increased severity of agoraphobia in aged
rodents. A third caveat to the interpretation of acquisition deficits is the possible
involvement of retention deficits. As was indicated with the circular holeboard task, it is
important to take care when making conclusions that deficits in acquisition are actually
deficits in the learning of the behavior rather than deficits in the retention of the learned
behavior or retention of events that would affect the learning of the behavior.
9

1.3 Differences in the Retention of a New Behavior
Age deficits in the retention of learned behaviors have also been observed in a
number of paradigms. In the above acquired immobility study performed by Rigter,
Martinez, and Crabbe (1980), where animals were placed in the water filled cylinders, it
was noted that the young Wistar rats learned to become immobile more quickly than the
aged animals. Two and six days after acquisition training, the animals were tested on
how soon they became immobile once placed in the cylinder and how long they remained
immobile. Both the aged and young animals improved on both measures on both of the
test days, meaning that they became immobile sooner and remained immobile for a
longer period of time. However, the young animals still became immobile more quickly,
upon placement within the apparatus, and remained immobile for a longer period of time
compared to the aged animals on both test days. This suggests that despite the fact that
aged animals showed some retention of the experience as they became immobile sooner
with each test, the aged animals did not remember the training upon reexposure to the
apparatus as quickly as their younger counterparts did. In another study, Martinez and
Rigter (1983) expanded on the earlier Rigter, Martinez, and Crabbe (1980) immobility
study by following the same methodology but adding longer retention intervals. In this
study, the non-deprived young (3 to 6 months) Wistar males again learned more quickly
on the acquisition trial than the non-deprived aged (24 to 27 months) Wistar males and
both aged and young animals improved on the measures of latency and duration 1, 3, and
6 weeks after acquisition. However, when tested 13 weeks after acquisition, the aged
animals showed a marked decrease in retention as indicated by increased latency to
10

immobility and shorter duration of immobility; the young animals did not show any
decrease in retention (Martinez and Rigter, 1983). Taken together, these studies give the
impression that the aged animals not only require more time to learn such a task, but also,
the memory trace is subject to degradation much sooner than the memory trace of
younger animals.
In the same paper by Martinez and Rigter (1983), another study trained non-deprived
rats of the same ages, sex, and strain in a one-trial passive-avoidance task. In this task
animals were placed on a pedestal facing away from a dark shock chamber. Once the
chamber door was opened animals were allowed to enter. Upon complete entry, the door
was closed and animals were trapped for 3-seconds during which time a 600µA shock
was delivered to the paws. Testing was done 1, 3, and 22 days after acquisition. During
testing animals were placed on the pedestal and latency to enter the shock chamber was
measured. Young and aged animals were similar in the amount of time taken to enter the
chamber on the acquisition trial, and their latencies were similar on test days falling 1 and
3 days after acquisition. However, on the test day falling 22 days after acquisition the
aged animals had a significantly shorter latency to enter the chamber than the young
animals (Martinez and Rigter, 1983). As with the acquired immobility task, we are
seeing that the memory of the shock experience, as it exists in aged animals, is subject to
degradation sooner than a young animal’s memory of the experience, leading to a shorter
retention span for such events.
In a series of three experiments by Zornetzer, Thompson, and Rogers (1982) non-
deprived aged animals displayed retention deficits in three other paradigms. In their first
11

experiment, animals were tested in a spontaneous alternation task set in a T-maze. It is
believed that animals alternate between which arm is entered on every trial because the
animal remembers exploring the first arm and opts to explore the novel environment.
Since the motivator in this instance is thought to be the novel environment, no food
reward is required. The researchers used young (3 to 6 months) and aged (24 to 26
months) male Sprague-Dawley rats as subjects. At the beginning of each trial the animal
was placed in the start alley of the maze and was allowed to enter either arm of the maze.
Once inside an arm, the animal was trapped for 5-seconds, and was then placed in a
holding cage for one of three predetermined intertrial intervals (ITI). At the end of the
ITI the animal was placed back in the start alley and was classified as alternating if he
entered the unexplored arm, and as not alternating if he explored the same arm as the
previous trial. Testing was done for 3 consecutive days, with one set of trials per day;
each set consisted of 2 trials. On the first day the ITI was set at 30-seconds, on the
second day the ITI was set at 4-minutes, and on the third day the ITI was set at 10-
minutes. Ninety-three percent of the young animals displayed alternation when the ITI
was 30-seconds, while 100% of the aged animals showed alternation, making the two
groups the same when the ITI was only 30-seconds. However, fewer of the aged animals
showed alternation when the ITI was 4 or 10 minutes. With the 4-minute ITI, 100% of
the young animals displayed alternation, while only 73% of the aged animals alternated
on trials. Although more of the young animals showed no retention of which arm had
already been explored at the 10-minute interval, as only 70% alternated, the aged were
12

still significantly lower, as only 45% showed alternation. Thus, fewer and fewer aged
animals alternated as the intertrial interval increased.
In the second experiment, non-deprived young and old animals were trained on a
step-down passive-avoidance task. These animals were the same subjects that had been
used in the first experiment, and were separated into 4 groups; 1 group of young and 1
group aged animals were tested 2 hours after training, and 1 group of young and 1 group
of aged animals were tested 24 hours after training. The apparatus was a chamber that
had a grid floor to allow for shock delivery, with a shelf along the wall, where the animal
could escape/avoid the shock. At the beginning of the trial, the animals were placed on
the shelf and the latency to step down from the shelf on the first exposure was recorded.
Once the animal reached the floor they received a 1.0mA shock, which was not
terminated until the animal climbed back on the safe shelf. After climbing back onto the
shelf, the animal was returned to his home cage. The animals were then tested either 2-
or 24-hours after acquisition, depending on their group designation. The initial learning
latencies were the same for the young and aged groups. The latencies to step down off
the shelf for young and old animals tested 2 hours after acquisition also were statistically
equivalent (600 and 581 seconds, respectively). However, the aged animals tested 24
hours after the training trial showed significantly shorter latencies (158 seconds) than the
aged animals tested 2 hours after acquisition (581 seconds) and the 24-hour young
animals (600 seconds). It is important to point out that a ceiling of 600-seconds was set
in place, such that if an animal did not step down by the time 600 seconds had passed
they were returned to their home cage. Therefore, it is possible that a significant
13

difference could have been detected at the 2-hour test if animals had been given more
time to step down.
In the third experiment, a new set of non-deprived young and aged male animals, of
the same age and strain as the previous 2 experiments, were trained in a spatial reversal
task. At the start of each trial, the animals were placed in the start box of a T-maze,
which had a grid floor wired for shock. Their task was to run into one of the arms to
escape shock. During training, animals received a 1.0-to-1.4mA shock delivered to the
paws, which began once they were placed in the start box. On the first trial, animals were
allowed to run into either arm on the T-maze to escape the shock. Then the animals were
trained to run into the opposite arm when placed in the start box, e.g., if they selected the
right arm on the first trial, they were required to select the left arm on the subsequent
trial, each time alternating arms until they reached criteria, which was defined as the first
appearance of a “perfect” or “near-perfect” series of correct responses having a
probability of less than .05” (Zornetzer, Thompson, and Rogers, 1982). They continued
to be shocked if they entered the incorrect arm and escaped shock upon entry into the
correct arm. Forty-eight hours later, they were tested for retention of this reversal, e.g.,
for entry into the left arm. After that test, the animals again were trained to run into the
opposite arm, e.g., the right arm. This procedure was done for 8 days, with 48 hours
between each training session, totaling 8 reversals. There were two age-related deficits.
First, the aged animals required more trials to acquire the reversal during the first two
days of reversal training. For the last six reversals, however, young and aged animals
learned the reversal in a similar number of trials. Second, aged animals showed a deficit
14

in retaining the reversal that was learned 48-hours prior to that day’s test, meaning that
for the test trial of any session, the aged animals selected the correct arm significantly
less often than the young animals, who consistently selected the correct arm.
Thus far we have seen that aged animals show deficits in the acquisition and retention
of a number of different behaviors. Additionally, it has been noted that deficits reported
in the acquisition of a behavior must be carefully examined for other explanations, such
as perseveration, before any meaningful conclusions may be made. Thus, although
differences in learning a new behavior or strategy have been observed, it seems that the
larger deficit lies in the retention of the memory, or memories of events essential for
learning, regardless of whether the information must be pulled from short- or long-term
stores, more so than with the acquisition of a behavior. Understandably, there is much
interest in the formation of the deficits and a great deal of research has been done in an
attempt to understand why and how such changes occur. One of the major theories that
has come about from this research is centered on changes in the stress-response system of
aged animals.
1.4 Glucocorticoids: The Possible Impact on Learning in Aged Animals
Several studies have shown that with aging comes an increase in the activity of the
hypothalamic-pituitary-adrenal (HPA) axis, and thus a hypersecretion of glucocorticoids
(GC). One model, aiming to account for both HPA hyperactivity and the witnessed
decline in cognitive function, suggests that hypersecretion of GC leads to neuronal
atrophy in the hippocampus. The resulting damage then impairs the negative feedback
cycle, which would normally shut down the HPA axis and stop GC release. Instead, an
15

even greater release of GC occurs, which in turn, causes more neuronal damage in the
hippocampus (for review see Pedersen, Wan, and Mattson 2001).
In the early studies of age-related differences in the stress response, the application of
a stressor typically induced a lesser release of stress hormones by aged animals compared
to young animals (Britton, Rotenberg, and Adelman, 1975). Such results led to
hypotheses that aged animals were less responsive to significant events encountered in
the environment (for review see Kubanis and Zornetzer, 1981). It was only when
researchers began investigating the response of the HPA axis to chronic stressors that
some of the currently accepted major age differences began to be uncovered. In one
study in which animals were subjected to chronic restraint stress, the decrease in
corticosterone release usually observed over time in chronic stress situations was only
observed in the young animals (Riegle, 1973). This observed inability for aged animals
to shutdown HPA activity under conditions of chronic stress is consistent with the
hypothesis that aging is accompanied by defects in the negative feedback system of the
stress response. In another study, Riegle and Hess (1972) administered dexamethasone, a
synthetic glucocorticoid, or placebo to young (4-6 months) and aged (22-32 months) rats
for 28 days and then exposed them to the acute stressor ether. Although this acute
stressor elicited an increase in corticosterone levels in all animals, both the young and
aged animals that had been receiving the dexamethasone treatment showed a smaller
release of corticosterone than control animals. This suggests that their HPA system was
compensating for already increased levels of stress hormone due to the dexamethasone
treatment. However, the young animals appeared better able to compensate for their
16

already elevated levels, as their corticosterone release was much less dramatic than the
release measured in the aged animals. These data also are consistent with the hypothesis
that chronic exposure to stress hormones causes a malfunction of the negative feedback
system and consequently renders aged animals hyperresponsive to stress.
That the stress response of aged animals is defective was substantiated by studies
examining recovery from stress. These studies showed that aged animals require more
time to recover from stress after the removal of the stimulus. In their 1983 study,
Sapolsky, Krey, and McEwen, found that aged animals (24 months) exposed to one of a
number of acute stressors or 3 weeks of chronic stress experienced a delayed return to
basal corticosterone levels when compared to a group of young animals (3 months).
Specifically, blood collected 1 hour after termination of the stressor showed that the
young animals’ corticosterone levels had returned to their basal values, while blood
collected at 90 minutes showed that the aged animals’ hormone levels were still well
above their basal levels, and remained high for up to 4 hours after the removal of the
stressor(s). The researchers then examined the clearance rate of corticosterone in both
the young and aged animals; they found that the clearance rate was similar for the two
age cohorts. The authors interpreted the similarity in clearance rates as indicating that
aged animals were unable to terminate GC secretion after the stressor is removed.
Interestingly, although aged animals show a prolonged response to stressors
(Sapolsky, Krey, and McEwen, 1983), the peak corticosterone levels seen during the
application of the stressor does not appear to differ from that seen in middle-aged or
young animals. However, this apparent similarity in corticosterone response is deceptive.
17

During the release of stress-related hormones that continues after the cessation of the
stressor in aged animals, the levels of these hormones keep rising such that they
eventually reach higher levels than the peak levels seen in the young animals during the
application of the stressor (Issa, Rowe, Gauthier, and Meaney, 1990). In the experiment
demonstrating this effect, aged male Long-Evans rats (23 to 27 months) were first tested
for their ability to learn the Morris water maze task, and then were separated into groups
designated as either cognitively-impaired or cognitively-unimpaired rats (Issa, Rowe,
Gauthier, and Meaney, 1990). Once the groups were formed, these animals and a group
of young control rats (6-7 months) were subjected to a stressor while measurements to
assess hormonal responses to stress were taken. At the end of the study, possible
differences in hippocampal morphology and functioning for all groups were determined.
In order to assess the circadian rhythm and basal levels of stress hormones, the
researchers measured corticosterone in 6 blood samples that were taken over a 24-hour
period; the time that the first sample was taken was not specified by the authors. They
also measured adrenocorticotropin hormone (ACTH) in two blood samples that were
taken 2 hours after light and dark onset. After these measurements were collected, all of
the animals were subjected to restraint stress by placing them in a restrainer for 20-
minutes; all animals experienced restraint stress somewhere between 3 and 5 hours after
lights were turned on. In addition to the 6 blood samples mentioned above, pre-stress
basal plasma levels were determined from a blood sample taken immediately before
placement in the restrainer, peak plasma levels were assessed from a blood sample taken
just before the animals were released, and post-stress levels were assessed from blood
18

samples taken every 30 minutes for the first 180 minutes after the stressor was removed.
Basal measurements of the hormones revealed: (1) there was a circadian rhythm of
corticosterone, with the peak onset in the dark phase of the diurnal cycle; (2) the
cognitively-unimpaired aged group and the young control group had similar levels of
corticosterone at every point in the 24-hour cycle; (3) the cognitively-impaired aged
group exhibited higher levels of corticosterone 4 hours before dark onset, 4 hours after
dark onset, and 4 hours before light onset; (4) all groups exhibited similar levels of
ACTH during the light phase, however, the cognitively-impaired aged group had higher
plasma levels of ACTH during the dark phase of the light/dark cycle. Pre-stress basal
plasma levels of corticosterone were found to be higher in the cognitively-impaired aged
animals compared to the cognitively-unimpaired aged and the young control animals.
Peak levels of corticosterone, however, were similar for all three groups. At every point
after the termination of stress application, however, the cognitively-impaired aged group
exhibited higher plasma levels of corticosterone compared to the cognitively-unimpaired
aged and young control groups. With respect to hippocampal neuron density, both of the
aged groups were found to have a lower cell density than the young group, however, the
cognitively-impaired aged group had an even lower neuron density than the unimpaired
group, with the most affected regions being CA1 and CA3. The results of this study lend
support to the above-presented hypothesis in that the aged animals exhibiting cognitive
impairments had a more prolonged response to stress and that both severe cognitive
deficits and a prolonged response to the stressor were correlated with greater neuronal
atrophy in the hippocampal formation.
19

Other studies have reported corroborating results. One such study was able to
establish a more substantial causal relationship between the increased corticosterone
levels to the cognitive deficits and hippocampal defects (Arbel, Kadar, Silbermann, and
Levy, 1994). In this experiment, middle-aged (12 months) male Fischer 344 rats were
divided into cognitively-impaired and cognitively-unimpaired groups on the basis of their
ability to learn the Morris water maze task. In addition to these groups, there also was a
young (4 months old) control group, that also had been tested in the Morris water maze
and were cognitively unimpaired; the young animals did not undergo any other
behavioral or drug treatments. Half of each of the middle aged groups then was
implanted with either a corticosterone pellet or a placebo pellet for 9 weeks; a total of 3
pellets were implanted, as each pellet only lasted for 3 weeks. One week following the
termination of the last pellet (i.e., 4 weeks after implantation of the last pellet), changes in
their cognitive ability were assessed through performance on the radial-arm maze. The
other half of each of the middle aged groups was not subjected to hormone or placebo
treatment or to radial-arm testing. The hippocampus of all animals was examined in
order to assess what effects corticosterone treatment had on this structure.
In the radial arm task, animals are usually given the same number of runs as there
are arms; correct responses on each run are scored as entering a previously unexplored
arm at each run. Thus, performance on the radial arm maze was scored by the number of
correct entries and the number of errors (i.e. entering the same arm more than once
during the testing session) out of eight. Comparisons of the middle-aged animals
revealed that the unimpaired-placebo group had more correct entries on the first 8 runs
20

than the impaired-corticosterone, unimpaired-corticosterone, and impaired-placebo
groups, whereas the impaired-placebo group performed similarly to the impaired-
corticosterone and unimpaired-corticosterone groups.
Examination of the hippocampus of middle-aged impaired and unimpaired animals
who had not undergone corticosterone treatment or radial-arm testing revealed that
impaired animals showed a higher density of damaged cells in the CA1 and CA3 regions
compared to unimpaired group and the young controls. Additional comparisons of the
tissue from middle-aged animals implanted with placebo pellets revealed that the
unimpaired-placebo group showed significantly less damage to the CA1 and CA3 regions
of the hippocampus than the impaired-placebo group. The unimpaired-corticosterone
group showed significantly more damage to CA3 and the dentate gyrus regions compared
to the unimpaired-placebo group, whereas the impaired-corticosterone group showed
significantly more damage to both CA1 and CA3 compared to the impaired-placebo
group. Furthermore, when the subareas of CA1 and CA3 were examined, it was found
that the only group to show no abnormal cell damage was the unimpaired-placebo group.
Unlike the first study, which could only speak to a correlation between corticosterone
levels and apparent hippocampal damage, this study provides a more causal relationship
between the two events. Although the unimpaired group already had damage to the
structure, their corticosterone cohorts experienced further damage after long-term
application of the hormone.
The next conundrum to figure out was what the formed defect was that was causing
the inhibition of the negative feedback. One avenue of research that was followed was
21

the examination of possible alterations in the number of functioning receptors. A 1974
study by Roth, did indeed find that the concentration of corticosterone receptors in the rat
decreases between the ages of 3 and 25 months. A slightly more recent study was able to
uncover some more specific changes (Eldridge, Brodish, Kute, and Landfield, 1989).
Generally, the number of hippocampal glucocorticoid receptors is up-regulated in
animals who have been adrenalectomized and down-regulated in intact animals
administered glucocorticoids or a chronic stressor, suggesting an adjustment to the
extremes in glucocorticoid concentration that persist for extended periods of time. It is
also the case that with age there is a decrease in the number of receptors and an inability
to up-regulate in response to adrenalectomy. Such facts make it quite odd that the
increased response to stress observed in aged animals would have such detrimental
effects on the hippocampus, as such factors would make it seem as though less
corticosterone is able to enter a neuron to cause damage to it. One possible explanation is
that there are differences in the longevity of the different types of glucocorticoid
receptors. Currently, we know of 2 types of receptors that will take up the stress
hormone, type 1 (T1) and type 2 (T2). T1 receptors are usually fully occupied at basal
corticosterone levels, whereas the T2 receptors only become occupied at more elevated
corticosterone levels. In experiment 1 of a series of studies (Eldridge, Brodish, Kute, and
Landfield, 1989), male Fischer 344 rats, ages 4, 12, and 18 months at commencement,
received chronic stress training and then were examined for age differences in
hippocampal glucocorticoid receptor density. Chronic stress training consisted of a 2-
way shuttle-escape task for 4-hours a day/5 days a week for 6 months; a 2-second
22

auditory cue was given before the onset of the 0.75mA shock, such that they could learn
to avoid the shock; the animals were terminated 1-day or 3-weeks following the end of
the 6-month shock treatment. The young animals, 10-months at time of death, showed
the expected down-regulation of the T2 receptors 1 day after the conclusion of the shock
treatment, but 3 weeks after treatment, the T2 receptor count had returned to normal
levels. The middle aged (18-months at time of death) and aged (24-months at time of
death) animals had no significant decrease in T2 receptors 1 day or 3 weeks following
termination of the chronic stress, although the middle aged animals had a nonsignificant
trend to down-regulate when tissue was collected 1 day after termination of treatment. In
experiment 2, animals underwent the same chronic stress treatment as the animals in
experiment 1; however, shock treatment was given for 2, 4, or 6 months and they were
terminated 1 day following the termination of the stressor. An additional difference
between this experiment and experiment 1 was that only middle-aged animals were used
as subjects. The number of T2 receptors in the hippocampus of the 2-month and 6-month
stressed animals was similar to that of control animals but for the 4-month stressed males,
the number was higher than the controls. The results of the two experiments suggests
that unlike young animals, older animals are unable to decrease the number of T2
receptors and in some cases older animals may actually up-regulate the number of T2
receptors during the presence of a chronic stressor, since the 4-month group of
experiment 2 had a greater number of receptors than either the 2 or 6 month groups. This
may make the older animals more susceptible to the toxic effects of the increased levels
of glucocorticoids that occur during chronic stress. Young animals are able to avoid
23

some of the damage inflicted by elevated levels of these hormones because they are able
to decrease the number of entry points. On the other hand, the aged animals are not only
unable to decrease the points of entry, but may actually increase the number of entry
points, making them more prone to damage. Taken together, these studies show that
aged animals develop an increased difficultly in dealing with stressors due to biological
changes that seem to occur with aging, and these changes have been linked to the
cognitive impairments that are often seen in aged animals.
1.5 Concluding Remarks
Thus far we have discussed some of the cognitive deficits observed in aging animals
through the study of traditional learning paradigms as well as one possibility for such
declines. Evidence has been provided lending support to the proposition that deficits exist
in the acquisition of new behaviors, although caution must be made in the interpretation
of such observations. We also have reviewed some of the evidence supporting the notion
of a dramatic decrease in the short- and long-term memory stores of aged animals, as
demonstrated through studies on the retention abilities of those animals. Finally, we have
reviewed a major hypothesis, focused on malfunctions in the HPA axis, aiming to
uncover why organisms share this seemingly universal decline in such abilities.
However, there seems to be at least one paradigm in which aged animals show an
increased ability to make and maintain an association, the conditioned taste aversion
paradigm.
24

Chapter 2
Age-Related Changes in Conditioned Taste Avoidances
2.1 Introduction
The general tendency to develop cognitive deficits with age has prompted an
examination into possible age differences of a rather different type of learning: food-
illness associations or conditioned taste aversions. Conditioned taste aversion refers to
the phenomenon whereby an animal consumes a novel flavor, experiences feelings of
malaise, and subsequently avoids further ingestion of that flavor. John Garcia discovered
conditioned taste aversions, somewhat fortuitously, while his lab was investigating the
behavioral effects of ionizing radiation. Since that time, it has been determined that taste
aversions can be acquired (1) through a single pairing of the conditioned stimulus (CS,
flavor) and unconditioned stimulus (US, illness-inducing agent), (2) with the presentation
of the CS and the US separated by a relatively long interval, and (3) by using drug doses
that would, upon first thought, seem too weak to induce an aversion to a paired substance
(for review, see Freeman and Riley, 2006). The implications of these features for
learning are profound. Learning in most other conditioning paradigms cannot occur
within such confines; temporal contiguity and repeated pairings are usually required.
Yet, we do see learning within the conditioned taste aversion paradigm. We observe the
behavioral expression of the animal’s association that the recently ingested substance
made it ill, through the animal’s initial investigation of the substance, upon the first re-
exposure, and then ardent avoidance of it. We also observe the retention of that memory
through the animal’s, at times, everlasting avoidance of the substance. However, most
25

animals will begin to drink the once avoided substance upon repeated exposure to the
substance without becoming ill, displaying another phase where learning occurs. These
behavioral events are referred to as acquisition and extinction, respectively, and are the
means by which the extent of learning and memory are evaluated. Both measures can be
manipulated in order to better understand each process.
There is an important distinction that must be made before further discussion of
conditioned taste aversions can take place. The concept of a conditioned taste aversion
implies that the animal has exhibited negative reactions that are indicative of an aversion
to a taste, such as gaping mouth (a gag-like response) and passive drip, when the taste is
introduced into the mouth. Studies examining such behaviors as a measure of aversion
typically use apparatuses that allow for forced entry of the substance into the mouth, such
as a cheek fistula. All of the studies discussed here, however, have not assessed such
behaviors; rather, they have measured the amount consumed by the animal as a result of
the animal approaching the substance and subsequently ingesting the CS. Since this
method of assessing conditioning is based on the animal’s avoidance of the substance,
rather than negative reactions to the substance, the term CTA will stand for conditioned
taste avoidance throughout the rest of the discussion. This distinction is important
because there are some USs that induce conditioned taste avoidance but not aversion
(Parker, 1995).
A number of studies looking into CTA formation and memory have found that aged
animals more readily develop, and less quickly extinguish, CTAs. First, some studies
have found that when the traditional CTA procedure is used (novel CS and novel US
26

presented with a 0 minute interval in between presentations), aged animals tend to
develop strong avoidances regardless of dose, unlike young animals who will develop
avoidances in a dose dependent manner (Misanin, Goodhart, Anderson, and Hinderliter
2002). Aged animals also appear to retain the memory of avoidance formation longer
than younger animals, in that they drink significantly less of the CS than younger animals
on the first extinction test, whether that test is 1 week or 12 weeks after acquisition day
(Guanowsky, Misanin, and Riccio 1983). Second, when certain manipulations of
acquisition are carried out, even more differences between age cohorts appear. For
instance, it has been shown that aged animals can make food-illness associations over
longer CS-US intervals than younger animals (Misanin, Greider, and Hinderliter 1988;
Misanin, Goodhart, Anderson, and Hinderliter 2002; Misanin, Collins, Rushanan,
Anderson, Goodhart, Hinderliter 2002). In addition, prior exposure to the conditioning
agent (or other agents that produce similar feelings of malaise) fails to attenuate the
strength of the developed avoidance in aged animals, whereas such exposure significantly
attenuates the strength of the developed avoidance in younger age groups (Valliere,
Misanin, and Hinderliter 1988; Misanin, Hoefel, Riedy, and Hinderliter 1997).
2.2 Acquisition of CTAs in Aged Animals
2.2.1 Effects of US Preexposure
Acquisition represents an initial learning period, where the animal first makes the
association that the illness (US) was caused by the consumed substance (CS). Within
acquisition there are of course multiple processes that must work in accordance with each
other. First, the animal must process the many qualities of the CS. Second, she must
27

remember the CS until she becomes symptomatic from the US. Third, the information
from the CS and US must intersect in order for the association to be made. These
qualities make the acquisition of CTAs an ideal candidate for studying the mechanisms of
learning, as well as a means to examine learning deficits that are commonly associated
with aging.
As we have seen above, aged animals show deficits in both acquisition and retention,
and one of the manipulations of CTA acquisition that can be used to examine retention is
the US preexposure effect. The US preexposure effect is where prior experience with a
US attenuates the strength of the formed avoidance, and it can be utilized in two different
ways, distally or proximally. When preexposure is distal, administration of the US is
given 24-hours or more before conditioning. In the case of proximal preexposure, the US
is given less than 24-hours prior to conditioning. Traditionally, the same agent, of equal
or higher dose, is used as both the preexposure US and the CTA US; however, it has been
shown that two different agents also may be used (Rabin, Hunt, and Lee 1989).
Regardless, in a preexposure paradigm, US exposure precedes CS exposure, which
precedes exposure once again to a US.
When young animals are preexposed to the US agent, they develop a CTA that is of a
much weaker strength than the CTA displayed by those animals who have had no prior
experience with the US at the time of conditioning. Employing preexposure paradigms
allows one to examine retention because in order for preexposure to be successful, the
animal must remember experiencing the feelings of malaise prior to the ingestion of the
novel substance. Assuming that the animal does remember feeling sick prior to
28

experiencing the new substance they will be less likely to associate those same (or
similar) feelings of malaise to the new flavor. Recognizing that preexposure is a memory
test of sorts, it seems reasonable to infer that preexposure would be less effective in aged
animals, as age is commonly accompanied by retention deficits. Inasmuch, memory
deficits in the aged animals would make it less likely that they remember their prior
experiences with the US; thus, preexposure would be ineffective and rather than
eliminating or attenuating the CTA, the aged animals would develop a CTA as potent as
one developed without having any prior experiences with the US agent.
As it turns out, a study performed by Valliere, Misanin, and Hinderliter (1988)
showed that aged female Wistar rats who received both remote and proximal preexposure
did not show the same abolition of CTA formation as was displayed by two younger age
groups. This development of CTAs, despite preexposure to the US, is what one would
expect in animals with a memory deficit. These results prompted Misanin, Hoefel,
Riedy, and Hinderliter (1997) to determine the independent effects of the two types of
preexposure in young adults (3 to 3.5 months old) and aged (21 to 24 months old) female
Wistar rats. In order to separate the two forms of preexposure, the researchers completed
two identical experiments, with the exception of type of preexposure. Both experiments
consisted of the same age groups, and each age group was further divided into four
treatment groups: given LiCl during both preexposure and conditioning (LL), given LiCl
during preexposure and NaCl during conditioning (LN), given NaCl during preexposure
and LiCl during conditioning (NL), and given NaCl during both preexposure and
29

conditioning (NN). During the course of both experiments, female Wistar rats were
maintained on 23-hour a day water deprivation.
In Experiment 1, which examined remote preexposure, LiCl (1% body weight of
0.15M) or NaCl was injected each day for 6 days prior to the acquisition test, which was
given the following day. In Experiment 2, which investigated proximal preexposure,
animals received an injection of the preexposure agent 60 minutes before the acquisition
test. In both experiments, acquisition testing involved injecting animals with the
conditioning agent immediately after a 20 minute exposure to a saccharin solution and
post-acquisition testing entailed giving animals a six hour two-bottle test (water and
saccharin). The post-acquisition test was administered the day following the acquisition
test. For both experiments, preference scores were calculated at the end of the six-hour
test (calculated by [saccharin/water + saccharin] x 100). The results for both studies are
presented in table 1.
TABLE 1. Results for Experiments 1 and 2 of Misanin, Hoefel, Riedy, and Hinderliter (1997)
Effect of remote preexposure to LiCl or NaCl
when LiCl or NaCl were used to induce CTA

Effect of proximal preexposure to LiCl or
NaCl when LiCl or NaCl were used to induce
CTA

CS-US
Interval
ACQ-EXT
Interval
Young
Adults
Aged

CS-US
Interval
ACQ-EXT
Interval
Young
Adults
Aged
LL 0-min 1 day No CTA CTA

LL 0-min 1 day No CTA No CTA
LN 0-min 1 day No CTA No CTA

LN 0-min 1 day No CTA No CTA
NL 0-min 1 day CTA CTA

NL 0-min 1 day CTA CTA
NN 0-min 1 day No CTA No CTA

NN 0-min 1 day No CTA No CTA
Abbreviations: ACQ-EXT Interval = Acquisition-Extinction interval CTA= Acquired a CTA; L= LiCl;
N=NaCl; No CTA= Did not acquire a CTA

As is shown in the table, aged animals still formed a CTA in the remote preexposure
treatment (LL), whereas the younger groups did not. However, unlike the remote
condition, proximal preexposure succeeded in blocking the formation of a CTA in all
30

three age groups. The authors suggested that the reason the aged animals were able to
learn a CTA in the remote preexposure treatment, is that they were more familiar with
their home environment than the young animals. Therefore, the aged animals were less
likely to associate the illness with some stimulus that was present within their home
environment, rendering them more likely to associate the illness with the novel saccharin
solution. On the other hand, because the younger age groups did not have as much
experience with their home environment, they were more likely to associate the illness
with some stimulus that was present within this environment, making them less likely to
associate the illness with the saccharin. However, it seems that the difference aged
animals showed in the effectiveness of remote and proximal preexposure may lie in the
deterioration of retention abilities that occurs during the aging process. Perhaps, a 24-
hour interval was simply too long for the aged animals to remember the feelings of
malaise, and when the interval was diminished to only 60-minutes, they could remember
the feelings. Another possibility stems from the previously described chronic stress
experiments in Chapter 1 (Glucocorticoids: The Possible Impact on Learning in Aged
Animals). It is possible that the increased stress response exhibited by aged animals may
have led to the aged animals experiencing a more severe reaction to the repeated
administration of the LiCl, making it more difficult for them to habituate to the repeated
stress encountered in the remote preexposure condition. This possibility will be
discussed in more detail in the section below, Possible Explanations for the Differences
in CTA Learning Observed in Aged Animals: A Glucocorticoid Hypothesis.

31

2.2.2 Effects of Increased CS-US Interval
Another method in which to examine the components of acquisition is to manipulate
the interval between CS exposure and US administration. This method again requires
that the animal remember one of the stimuli before the application of the other. In this
case, the stimulus that must be remembered is the CS and the CS trace must last until the
symptoms of the US manifest. Again, because of the well-documented troubles with
retention experienced by older animals, it is easy to assume that young animals would be
able to form a CTA over a longer CS-US interval than would aged animals. Surprisingly
enough, a 1988 study performed by Misanin, Greider, and Hinderliter showed that among
two age groups (young adults and aged), the aged animals were able to form a CTA over
a longer interval compared to the younger age group. Specifically, the young adults were
able to form a CTA when the US was administered either immediately, 45 minutes, or 90
minutes following CS exposure, while the aged animals were able to form a CTA when
the US was administered immediately, 45, 90, and 180 minutes following CS exposure.
These unexpected results prompted Misanin, Goodhart, Anderson, and Hinderliter
(2002) to explore whether a more intense (i.e. stronger) US would support the CS trace
for an even longer time interval, based on his hypothesis that as a CS trace becomes
weaker a stronger US dose would be required to support CTA formation. In order to test
this, the researchers used two age groups (2 to 3.5 months old young adult and 24 to 26
months old aged adult) of female Wistar rats, 4 time intervals (0 minutes, 90 minutes, 180
minutes, and 270 minutes), and 3 LiCl doses (1% percent body weight of either 0.075M,
0.15M, and 0.30M). For both ages, there were a total of 13 groups; one of these groups
32

included a control group that received NaCl at 0 minutes following 10-minutes of
saccharin exposure. For the three days leading up to acquisition and the day of
acquisition, all females were maintained on a 23 hour water deprivation schedule. The
day following acquisition, the animals underwent a post-acquisition test in which they
had continuous access to saccharin and water for 24 hours. Preference scores were
computed at the end of the 24-hour post-acquisition test (calculated by [saccharin/water +
saccharin] x 100).
The results are illustrated in table 2. Briefly, the authors found that aged animals
formed more potent avoidances to saccharin with a traditional 0-minute CS-US interval
and they were able to learn an avoidance over longer CS-US intervals than young
animals. These results were explained by the authors as a deficit in an internal pacemaker
of aged animals. Particularly, they proposed that there is a pacemaker whose function is
to dictate how long a taste memory trace should survive and that in aged animals, this
pacemaker has formed some dysfunction that slows it down, allowing for the longer
survival of a taste memory. For example, suppose that the pacemaker requires the
passing of 400 ticks before the memory decays. Then, in young animals, 400 ticks might
pass in 60 minutes, but in aged animals, 400 ticks would pass in over 300 minutes.
33

TABLE 2. CTA Formation Over Different CS-US Intervals and Doses (Misanin, Goodhart, Anderson, and
Hinderliter, 2002)
Interval ACQ-EXT Interval Dose Young Adult Aged
0 min 1-day NaCl No CTA No CTA
0 min 1 day 0.075 CTA** CTA
0 min 1 day 0.15 CTA* CTA
0 min 1 day 0.30 CTA CTA
90 min 1day 0.075 No CTA CTA
90 min 1day 0.15 No CTA CTA
90 min 1 day 0.30 No CTA CTA
180 min 1 day 0.075 No CTA CTA**
180 min 1 day 0.15 No CTA CTA**
180 min 1 day 0.30 No CTA CTA
270 min 1 day 0.075 No CTA No CTA
270 min 1 day 0.15 No CTA No CTA
270 min 1 day 0.30 No CTA CTA*
Abbreviations: ACQ-EXT interval = Acquisition-Extinction interval; CTA = Acquired a CTA; No CTA =
Did not acquire a CTA. Statistical Comparisons: To assess if experimental animals acquired a CTA,
comparisons were made between the experimental groups and control group within a given age group. For
each age group, comparisons among the experimental groups within a given CS-US interval were made.
No statistical comparisons between young adult and aged animals or across the different CS-US intervals
were reported. Therefore, in order to better compare the age groups and CS-US intervals, the following
designations were made for the strongest to weakest avoidances based on the figures provided in Misanin,
Goodhart, Anderson, and Hinderliter (2002): CTA = percent preference for saccharin fell below 25%,
CTA* = percent preference fell between 26% and 50%, and CTA** = percent preference fell between 51%
and 70%. According to statistical analyses, animals with a percent preference for saccharin above 75%
were found to not have developed an avoidance.

In explaining the results in the above fashion, Misanin, Goodhart, Anderson, and
Hinderliter (2002) discredited another possibility, that aged animals perceive the US as
more intense (as though it were a higher dose) than younger animals. He based this on
the results of a previous study in which the same amount of LiCl (2.3 ml of a 0.15M
solution) was administered to each age group, rather than the normal 1% body weight
amount. In that study, aged animals still formed a stronger CTA. However, the cited
study does not seem to disprove the possibility that aged animals are more sensitive to the
effects of the US. Instead, it seems to support this possibility. A dose of 2.3 ml of LiCl is
34

equivalent to a 1% body weight dose for an animal that weighs 230 g, a weight that is on
average surpassed by the time Wistar females reach 65 days (see the Harlan Sprague-
Dawley, Inc., website). Aged rats can weigh at least 100g more than that, which means
that the 2.3 ml dose of LiCl was in fact a smaller relative dose in the aged animals
compared to the young animals, and yet, the aged still formed a stronger CTA. This issue
is discussed in more detail in the below section Possible Explanations for the Differences
in CTA Learning Observed in Aged Animals: Physiological Changes.
As a final comment, it seems unclear as to whether the 24-hour test really examined
CTA formation or extinction. Twenty-four hours of access to the CS is enough time for
an animal to discover that the substance no longer makes her ill and thus to extinguish her
CTA. Since the consumption amount was only recorded at the end of the 24-hour period,
it is possible that some of the groups who were said not to form a CTA simply
extinguished. It also leaves the door open to the possibility that aged animals are more
resistant to the extinction of formed CTAs, rather than better equipped to form the CTA
in the first place. If this is the case, it could be that aged animals develop stronger
avoidances, which may lead to lengthened extinction, or it is possible that the aged are
showing perseveration, in that they cannot change their course of behavior from not
drinking the substance to drinking the substance.
2.2.3 Effects of Increased Interval between Acquisition and the First Extinction Test
There also is evidence that aged animals can retain the memory of avoidance
formation for a much longer time than younger animals. The first study to be presented
was completed by Guanowsky, Misanin, and Riccio in 1983. This experiment used
35

female Wistar rodents that were young (approximately 4.5 months) and the old (19 to 21
months). On the day of acquisition, animals were given 10-minutes of saccharin
exposure and then were immediately injected with saline or 1% body weight of 0.15M
LiCl. Either 1 day or 28 days following conditioning, animals were given a 24-hour two-
bottle test (saccharin and water). On the test day, consumption was recorded and
preference scores were calculated (calculated by [saccharin/water + saccharin] x 100) at 5
different intervals; 0.5, 1, 6, 12, and 24 hours after the test had begun. The results are
summarized in table 3, below. It was found that all experimental groups differed from
their respective control groups at all intervals of the 24-hour test. Whether tested 1 day
or 28 days after acquisition it was found that both young and old female Wistar rats
retained the memory of avoidance formation. Furthermore, no difference in strength
between the two age groups was found at either testing interval.
In the same year, an experiment by Martinez and Rigter (1983) found that aged
animals developed more potent avoidances at all, but one, retention intervals. In their
study, young (3 to 6 months) and old (24 to 27 months) male Wistar rats were trained in a
CTA paradigm. Animals were maintained on a deprivation schedule that allowed only
45-minutes per day of water access. On conditioning day, all animals were given a 5%
glucose solution for 15 minutes and 30 minutes later they were given an injection of
0.15M LiCl, 10 ml/kg. They were then tested for retention 1, 3, 6, and 12 weeks
following conditioning by giving them access to a glucose solution and water (a 2-bottle
test) for 15 minutes. At the end of the 15-minute retention test, the amount of glucose
consumed was recorded. It was found that the old age animals exhibited more potent
36

avoidances than the young animals when tested 1, 6, and 12 weeks after conditioning.
These results also are depicted in table 3.
TABLE 3. Results of Guanowsky, Misanin, and Riccio (1983) and Martinez and Rigter (1983)
Sex/Strain LiCl Dose CS-US Interval ACQ-EXT Interval Acquisition Strength
Guanowsky, Misanin, and Riccio (1983)
Female/Wistar 0.15M, 1% BW 0-min 1 day A=Y
Female/Wistar 0.15M, 1% BW 0-min 28 days (4 weeks) A=Y
Martinez and Rigter (1983)
Male/Wistar 0.15M, 10/ml/kg 30-min 1 week A > Y
Male/Wistar 0.15M, 10/ml/kg 30-min 3 weeks A = Y
Male/Wistar 0.15M, 10/ml/kg 30-min 6 weeks A > Y
Male/Wistar 0.15M, 10/ml/kg 30-min 12 weeks A > Y
Abbreviations: ACQ-EXT interval = Acquisition-Extinction interval; A = Aged; BW = Body weight; Y =
Young

The finding of Martinez and Rigter (1983) that aged animals show evidence of
prolonged memory retention for avoidance formation compared to younger groups is
interesting since for most other learning paradigms aged animals tend to show declines in
retention capabilities. However, Guanowsky, Misanin, and Riccio (1983) failed to find
age-related differences in retention and this inconsistency needs to be discussed. These
investigators used female Wistar rats in their study, whereas Martinez and Rigter (1983)
used male Wistar rats. Taking into account the fact that both studies used the same dose
of LiCl, we may be seeing a sex difference. Some studies have shown that male animals
develop stronger conditioned taste avoidances than females (Chambers, Sengstake,
Yoder, Thornton, 1981; De Beun, Jansen, Smeets, Niesing, Slangen, and Van de Poll,
1991; Gustavson, Gustavson, Young, Pumariega, and Nicolaus, 1989). As such, it may
be the case that differences between young and aged animals may be more dramatic when
the subjects are males versus females. That is, it may be that aged animals are generally
37

more likely to make food-illness associations. However, because aged males are already
more prone to makes these associations, the changes that occur in all aging animals,
coupled with their existing increased sensitivity, may lead to greater differential effects in
aged male animals.
It is also perplexing that young and aged males in the Martinez and Rigter (1983)
study show avoidances of equal strength at 3 weeks after the acquisition test, yet aged
males exhibited stronger avoidances when tested 1, 6, and 12 weeks after acquisition.
With respect to the 3-week interval, it may be that some unrelated event led the young
males to consume less of the glucose solution on the test day, thus making the lack of
difference an anomaly. When looking at the Guanowsky, Misanin, and Riccio (1983)
study, however, we see that young and aged females consumed similar amounts at 4
weeks, leaving the possibility open that something else is occurring. It may be that for
both males and females, the time course for the strengthening and then weakening of the
memory trace changes during the aging process, one day is not long enough for retention
differences to appear, by one week the memory trace is stronger in aged animals than
young, by 3-4 weeks the strengthening of the memory trace in young animals catches up
to the strength found in old animals, and then the weakening of the memory trace in
young animals precedes that of old animals whose memory trace remains strong for at
least 12 weeks. The fact that the (Martinez and Rigter, 1983) study did not include 1 day
intervals and the Guanowsky, Misanin, and Riccio (1983) study did not extend past the 4-
week mark makes it difficult to make any conclusions.
38

However, it is important to point out that the lack of age differences in the
Guanowsky, Misanin, and Riccio (1983) study does not discount the possibility aged
animals show better retention of CTAs compared to other types of learning. It seems that
comparable retention in young and aged animals is still an improvement over other
learning paradigms. Recall that Rigter, Martinez, and Crabbe (1980) found that aged
animals showed decrements in retention of acquired immobility as soon as 2 days
following acquisition. In addition, Martinez and Rigter (1983) found that aged animals
showed decrements in retention of a passive avoidance task when tested 22 days
following acquisition. Thus, it seems that even with a lack of difference between age
groups the Guanowsky, Misanin, and Riccio (1983) study may still speak to a different
pattern of cognitive decline in aged animals with respect to food-illness associations
versus other types of learning.
2.3 Extinction of CTAs in Aged Animals
The length of time it takes an animal to extinguish from a CTA is thought to speak to
the strength of the formed avoidance; the weaker the avoidance, the sooner the animal
should try the substance again and subsequently learn that the substance no longer makes
them ill. Extinction is more complicated because it not only contains components of the
strength of the avoidance learning and its retention but of another kind of learning as
well. The avoidance of the substance upon initial reexposure is the behavioral
manifestation that avoidance learning has occurred and that the memory for the food-
illness association has been retained, and the eventual normal consumption of the
substance is the manifestation of the next phase of learning where the animal discovers
39

that the substance is safe to consume. This poses an interesting duality in terms of what
to expect in aging animals. Because of the documented trouble with retention, it might
be assumed that aged animals would extinguish much sooner than young animals, on the
other hand, because of the trouble with acquisition, one might expect extinction to take
longer since the animals would have trouble learning that the substance is in fact safe.
However, when considering the above evidence showing that aged animals can retain the
memory of an avoidance for longer periods of time, it seems reasonable to assume that
aged animals might show an increased resistance to extinction. As such, there is
evidence that older animals do show greater resistance to extinction.
A study conducted in 1980 by Ingram and Peacock looked at extinction in young (3
to 4 months), middle aged (12 to 13 months), and aged (24 to 25 months) male Fischer
344 rats. On conditioning day, all groups were given access to saccharin for 15 minutes
and then were given an injection LiCl (0.4M, 7.5 ml/kg) or saline with delays of 15, 60,
or 240 minutes following saccharin exposure. The initial strengths of avoidance were
based on comparisons of consumption on acquisition day and the first post-acquisition
test. Extinction testing began 2 days after the acquisition trial. The first extinction test
consisted of a 15-minute window where animals were given water and saccharin. On the
following day, animals began an extinction procedure in which all animals had constant
access to both water and saccharin for 32 days. On all days prior to the 32-day extinction
procedure, animals were on a 23¾-hour water deprivation schedule. Preference scores
during extinction were calculated (calculated by [saccharin/water + saccharin] x 100)
every 96 hours, resulting in eight blocks of 4-day measurements.
40

The initial strengths of avoidance were found to be equal for all age groups within a
given CS-US interval. With respect to extinction, the authors found there was an increase
in resistance to extinction with an increase in age, however, the resistance to extinction
decreased with the increase in CS-US interval. Specifically, for the 15-minute interval,
the aged males showed greater resistance compared to the mature and young animals, for
the 60-minute interval the aged and middle aged groups showed greater resistance
compared to the young, and for the 240-minute interval no age difference was found.
TABLE 4. Results of Ingram and Peacock (1980)
CS-US
Interval
Acquisition-Extinction
Interval
Results

Acquisition Strength Extinction Resistance
15 minutes 2 days No age difference A > (MA = Y)
60 minutes 2 days No age difference (A = MA) > Y
240 minutes 2 days No age difference No age difference
Abbreviations: A = Aged, MA = Middle-aged, and Y = Young
The authors offered three possible explanations for their results. First it was
proposed, as speculated earlier, that the aged rats may be showing perseveration. A
second possibility, also proposed earlier, was that there are US sensitivity differences
between young and aged animals for the same dose of any given US, that is, aged are
more sensitive to the US than young. Finally, they proposed that although the aged
animals may show normal learning in the acquisition of the CTA, they might still have
learning deficits. The learning deficit would present itself through extinction as the
inability to learn something new (i.e. the safety of the substance), since extinction
requires the animal to learn a “new incompatible response” (Ingram and Peacock, 1980).
An analysis of the differences among the three CS-US intervals in the ability to
acquire a CTA suggests that there is an inherent problem with the perseveration
41

hypothesis. Although all age groups developed stronger CTAs with the 15- and 60-
minute intervals, they still developed a CTA with the 240-minute interval, which would
mean that the older rats with this interval should still show perseveration and thus slower
extinction rates. The same is true for the learning deficit. The animals with the 240-
minute interval still form an avoidance, why then would they show no trouble performing
and learning the “new incompatible response”? Both of these proposals would require
the existence of a mechanism that codes for avoidance strength, as well as affecting
subsequent behavior based on the determined strength, in all age groups. For example,
such a mechanism may function by determining that the avoidances formed at the 15- and
60-minute intervals were strong enough to cause true impact on the animal, thus
prolonging extinction. Whereas the avoidance formed with the 240-minute interval
would be coded as less salient, and the connections that inhibit consumption of the
conditioned substance may be weaker, thus allowing for quicker extinction rates. Both
the perseveration and learning proposals, then, would require some malfunction in the
translation of avoidance strength to subsequent behavior in aged animals, such that
although the expression of the initial avoidance is the same for all age groups within each
CS-US interval, the aged mechanism translates the avoidance into a much stronger one
thus prolonging extinction. As stated above in response to the perseveration and learning
proposals, this seems unlikely as such a mechanism should still result in the aged animals
exhibiting prolonged extinction at the 240-minute interval.
Three other experiments have examined extinction of CTA in young (2 months),
middle aged (11 months), and aged (20 months) male C57/BL6 mice (Springer and
42

Fraley, 1981). In all of these experiments, animals experienced one preexposure to a
milk solution (CS). The following day they underwent conditioning which consisted of a
15-minute access period to the CS and an injection of 0.4M LiCl, 15 mg/kg either 0
(immediately), 1, or 3 hours following CS exposure, depending on the experiment. All
animals experienced one 24-hour period of water deprivation on the day before
preconditioning. After this 24-hour period, water was available ad libitum except for the
15-minute periods where they had access to the CS during acquisition and the post-
acquisition test. For experiment 1, the post-acquisition test was given 2 days after
acquisition, while for experiments 2 and 3 a four-day interval was used. Each post-
acquisition test lasted for 15-minutes, at the end of which percent change scores were
calculated (calculated by [extinction test consumption – acquisition
consumption/acquisition consumption] x 100); initial avoidance strength was determined
by comparing consumption on acquisition day to consumption on the first post-
acquisition test.
With respect to the 0-minute delay, it was found that aged and middle-aged mice
developed stronger avoidances and exhibited greater resistance to extinction than the
young when their post-acquisition test was 2 days after acquisition; whereas, the middle-
aged animals displayed the strongest avoidances and exhibited the greatest resistance to
extinction when the post-acquisition test was 4 days after acquisition. When looking at
the pattern of results within the 4-day acquisition-extinction interval, it was found (1)
middle-aged animals developed the strongest avoidances and displayed the greatest
resistance to extinction with the 0-minute CS-US interval, (2) middle-aged mice again
43

displayed the strongest avoidances and greatest resistance to extinction in the 1-hour CS-
US interval, and (3) no age differences were found within the 3-hour CS-US interval. In
addition, for all age groups with a 4-day acquisition-extinction interval, animals with a 0-
minute CS-US interval developed stronger avoidances than those with a 3-hour interval.
The results are depicted in table 5.
TABLE 5. Results of Springer and Fraley (1981)
Experiment

CS-US
Interval
Acquisition-Extinction
Interval
Results

Acquisition Strength Extinction Resistance
Experiment 1 0hr delay 2 days (A = MA) > Y (A = MA) > Y
0hr delay MA > A > Y MA > A > Y Experiment 2
3hr delay
4 days
No age differences No age differences
0hr delay MA > (Y = A) MA > (Y = A) Experiment 3
1hr delay
4 days
MA > (Y = A) MA > (Y = A)
Abbreviations: A = Aged, MA = Middle-aged, and Y = Young
The experiments from the above study show that when extinction begins 2 days after
conditioning, aged and middle-aged animals display slower extinction rates. When
extinction tests begin 4-days after conditioning, middle-animals show the greatest
resistance to extinction. Since the differences in extinction rates for middle-aged and
aged mice appears with prolonged acquisition-extinction intervals, the authors propose
age related differences in retention capacities rather than in ability to learn. Specifically,
it may be as an animal ages they become more resistant to food avoidance extinction,
however, as the animal becomes increasingly old their ability to remember the formation
of an avoidance decreases. This assertion seems to imply that with a longer acquisition-
extinction interval (i.e. four-day interval) the strength of an avoidance decreases, leading
to faster extinction rates. Whereas, with the more commonly used two-day interval, not
enough time has passed to significantly decrease avoidance strength, leading to the
44

slower extinction rates. Such a supposition does fall into line with other evidence
showing that aged animals have retention deficits. However, the authors’ discussion of
the extinction rate results in relation to retention fails to take into account a more obvious
reason for the difference, initial avoidance strength. Notice, from table 5, that the
extinction rates mimic the strengths of acquisition, making it reasonable to assume that
the rate of extinction is reflective of the initial strength of avoidance, not differences in
retention.
Comparison of Ingram and Peacock (1980) and Springer and Fraley (1981) reveal
both similarities and differences. First, for both studies, no age-related differences
occurred with CS-US intervals less than 3-4 hours. For the shorter CS-US intervals, aged
rats consistently showed greater resistance to extinction than young rats, while
differences between middle-aged and young rats was less predictable. On the other hand,
the opposite was true of mice; middle-aged mice consistently showed greater resistance
to extinction than young mice, while sometimes differences between aged and young
mice were found and sometimes they were not. These data suggest that there are species
differences in the pattern of changes that occur during aging. If this is the case, there are
two possibilities that may explain these species differences: (1) It might be that in rats an
increased proclivity to develop and remember CTAs begins as they enter middle-age and
continues into old age, whereas in mice this increased proclivity begins prior to middle-
age and dissipates with old age as other malfunctions develop throughout the aging
process. (2) It also might be the case that the ages used for the three age groups were not
comparable for across-species comparisons. That is, perhaps the age of the aged Fischer
45

344 rats was younger than the age of the aged C57/BL6 mice, and the same for the other
two age groups. If this is the case then we might find the same pattern of results across
the two species if comparable ages had been used.
2.4 A Summary of the Results
Both of the extinction studies described above also looked at strength of acquisition.
Thus it is instructive to compare the results of these studies with those studies described
in the other sections (Misanin, Goodhart, Anderson, and Hinderliter, 2002; Guanowsky,
Misanin, and Riccio, 1983). Because the other studies did not test middle-aged animals,
only differences between aged and young animals will be discussed. With a 0 to 15
minute CS-US interval and acquisition-extinction intervals of 1 to 4 days, the four studies
were inconsistent in finding age-related differences (see Table 6 for all comparisons).
However, when differences were found, they supported the notion that for both females
and males, aged animals acquire stronger avoidances than younger animals. When
looking at CS-US intervals of 60-90 minutes and acquisition-extinction intervals of 1 to 2
days, aged female rats were able to develop avoidances over longer CS-US intervals
when three different US intensities were used but younger female rats were not. On the
other hand, neither aged male rats and mice nor their young counterparts were able to
acquire CTAs. Lastly, when the CS-US interval was extended to 180 to 270 minutes and
acquisition-extinction intervals of 1 to 2 days, it was again the case that young and aged
male mice and rats failed to differ. Aged female rats were able to acquire CTAs, but
unlike the 60-90 minute intervals, only with the highest dose of LiCl.
46

TABLE 6. Results of Misanin, Goodhart, Anderson, and Hinderliter (2002); Guanowsky, Misanin, and
Riccio (1983); Ingram and Peacock (1980); Springer and Fraley (1981)
CS-US INTERVAL: 0 to 15 minutes
Author Species/Strain Sex LiCl Dose ACQ-EXT
Interval
Result
Misanin,
(2002)
Rat/Wistar Female 1% BW of 0.075M,
0.15M, and 0.30M
LiCl
1-day Y: strength of avoidance dose-
dependent
A: strong avoidances for all
doses
Guanowsky,
(1983)
Rat/Wistar Female 1% BW of 0.15M
LiCl
1-day A=Y
Ingram,
(1980)
Rat/Fischer 344 Male 7.5 ml/kg 0.40 LiCl 2-days A=Y
Springer,
(1981)
C57/BL6 Mouse Male 15 ml/kg 0.40 LiCl 2 and 4 days 2-day ACQ-EXT interval:
A>Y
4-day ACQ-EXT interval:
A>Y and A=Y
CS-US INTERVAL: 60 to 90 minutes
Misanin,
(2002)
Rat/Wistar Female 1% BW of 0.075M,
0.15M, and 0.30M
LiCl
1-day Y: No CTA with any dose,
A: CTA with all doses
Ingram,
(1980)
Rat/Fischer 344 Male 7.5 ml/kg 0.40 LiCl 2-days A=Y
Springer,
(1981)
C57/BL6 Mouse Male 15 ml/kg 0.40 LiCl 4-days A=Y
CS-US INTERVAL: 180 to 270 minutes
Misanin,
(2002)
Rat/Wistar Female 1% BW of 0.075M,
0.15M, and 0.30M
LiCl
1-day Y: No CTA with any dose
A: CTA with only 0.30M LiCl
Ingram,
(1980)
Rat/Fischer 344 Male 7.5 ml/kg 0.40 LiCl 2-days A=Y
Springer,
(1981)
C57/BL6 Mouse Male 15 ml/kg 0.40 LiCl 4-days A=Y
Abbreviations: A = aged; ACQ-EXT Interval = acquisition-extinction interval; BW = Body Weight; Y =
young.

Taken together, the data indicate that with short CS-US intervals, aged animals of
both sexes have a tendency to acquire stronger avoidances than young animals while for
the longer CS-US intervals (60 to 270 minutes), similar age-related differences are found
47

in females but not males. The direction of the sex difference with longer CS-US intervals
is surprising given that young males tend to show greater propensity to develop
avoidances than young females (Chambers, Sengstake, Yoder, Thornton, 1981; De Beun,
Jansen, Smeets, Niesing, Slangen, and Van de Poll, 1991; Gustavson, Gustavson, Young,
Pumariega, and Nicolaus, 1989). The tendency for young males to more readily develop
avoidances, however, does not rule out the possibility that the increased propensity
reverses in old age, such that aged females become more likely to develop avoidances
than aged males. Additionally, because the male and female subjects were either
different strains of rat or a different species, it must be noted that the differences between
studies may be a result of an interaction between the strain/species and sex of the
subjects. Yet another possible explanation for the differences in acquisition strength
between males and females at all intervals, is the procedure used in the Misanin,
Goodhart, Anderson, and Hinderliter (2002). Recall that measurements were only taken
at the end of a 24-hour test and thus the results might actually be speaking to extinction
rate rather than acquisition strength. A fourth possible explanation involves the age
ranges selected for each age group. With respect to the 0 to15 minute interval, there were
two studies that used female Wistar rats. For the study that failed to find age differences,
the young subjects were 4.5 months old and the aged subjects were 19 to 21 months old
(Guanowsky, Misanin, and Riccio, 1983), while for the study that found age differences,
the young subjects were 2 to 3.5 months old and the aged subjects were 24 to 26 months
old (Misanin, Goodhart, Anderson, and Hinderliter, 2002). Thus, the failure to find age
differences in the one study, may be a result of a narrower spread in the ages of the young
48

and aged females or the aged animals in that study may not have been old enough to
detect significant differences regardless of the age young group. As discussed earlier,
regardless of the inconsistent findings, the lack of differences between age groups still
may speak to a different pattern of cognitive decline with respect to food-illness
associations as compared to other types of learning and memory.
In conclusion, the preponderance of the data indicate that aged animals form more
potent avoidances with a traditional 0-minute CS-US interval, learn avoidances over
longer CS-US intervals, retain the memory of avoidance formation over longer periods of
time, and exhibit increased resistance to extinction. Furthermore, they are resistant to the
attenuating effects of preexposure to malaise-inducing agents on acquisition of CTAs.
Taken together, these data are consistent in their indication that the aging process
facilitates acquiring and retained CTAs.
2.5 Possible Explanations for the Differences in CTA Learning Observed in Aged
Animals
2.5.1 Physiological Changes
There are a number of physiological changes that occur during aging that could
account for the fact that learning of food-illness associations in aged animals differs from
the same type of learning in young animals. These include decreased liver and/or kidney
function, alteration in adipose tissue composition, and slower metabolism, all of which
likely lead to prolonged residence of the drug in the body and therefore prolonged
sensations of malaise. Although all of these factors may influence a host of
unconditioned stimuli, it is likely that kidney function plays a more important role than
49

any other changes when LiCl is employed as the US. This agent, which was used in all
of the CTA studies described above, is not metabolized and is primarily excreted from
the kidneys (for review see Schou, 1957; Radomski, Fuyat, Nelson, and Smith 1950).
When LiCl is administered through intraperitoneal (ip) injections, which was the case for
all the studies described above, it quickly distributes and is absorbed by, virtually, the
entire body. Widespread distribution of LiCl is observed throughout the total body water,
comparable intracellular and extracellular levels of Li, and is absorbed by multiple
organs, including muscle and bone. When a single dose of LiCl is employed, most of the
Li is excreted by way of the kidneys, through the urine, within the first few days;
however small amounts may still be detected in the urine for 1- to 2-weeks after
administration as the various organs release what was absorbed. As such, it would be no
surprise if age-related changes in kidney function lead to increased and/or prolonged
toxicity from LiCl. Changes in size and adipose tissue composition that typically
accompany aging may also affect the induced toxicity from LiCl. With respect to size
(and presumably fat content of the animals), Levine, Saltsman, Katof, Meister, and
Cooper found that larger animals had higher serum levels of lithium 24-hours after ip or
intravenous (iv) injections. This was found in young male rats of two different ages and
therefore sizes (6- and 9-weeks), as well as across male and female rats of the same age,
but different weights (males being larger than females at the same age). In both
instances, the larger group of rats had higher serum levels. To ensure that the effect was
a result of size and not sex they looked at male and female rats of the same size, but
different ages, and found that the serum levels were comparable, indicating that elevated
50

serum levels were a result of size, and not sex of the animal. The doses and volumes of
LiCl used in this study were selected in order to yield levels comparable to therapeutic
doses in humans, 24 hours after administration; that is, animals received a 0.15M
concentration of LiCl in volumes of 1, 2, or 3 ml/100g of body weight (Levine, Saltsman,
Katof, Meister, and Cooper, 1997). Although the larger animals were injected with a
larger volume of LiCl than the smaller animals, their doses were comparable for their
sizes. Given this, it would be expected that they would be able to excrete LiCl at a
comparable rate, resulting in comparable serum levels 24 hours later. However, the
levels in the larger animals were significantly higher than in the smaller animals after
correction for differences in size and dose. Thus, the higher levels of lithium in larger
rats indicate that they would experience a higher toxicosis from the lithium than would
the smaller rats. Given that as rats age they put on more fatty tissue, it would not be
surprising if they experience a more potent toxicosis after administration of equivalent
doses compared to younger animals. As will be discussed in more detail below, aged
animals develop other pathologies that would further increase the likelihood that they
experience a more potent toxicosis to equivalent, or even the same, doses of LiCl.
2.5.2 A Glucocorticoid Hypothesis
One candidate for the ineffectiveness of remote preexposure, the ability to form
CTAs over longer CS-US intervals, the longer retention of avoidance formation, and the
increased resistance to extinction stems from the glucocorticoid studies discussed above.
It is possible that the aged animals have a stronger response to the stress of being made ill
which then allows longer CS-US intervals, prolonged periods of retention, and increased
51

resistance to extinction. In addition, unlike young animals, aged animals may not
habituate to chronic stressors and this in turn could eliminate the effectiveness of remote
preexposure to LiCl.
As discussed earlier, aged animals experience a prolonged and more severe response
to stressors, but there are a number of other indicators that convey when an animal is
experiencing a more severe stressor than usually encountered. Studies examining the
plasma levels of glucocorticoids (GC), epinephrine (EPI), norepinephrine (NE) and the
expression of different immediate early genes (IEGs), during the application of stressors,
have found that different patterns of plasma levels and expression are seen with acute
stress application as opposed to chronic stress application (see Table 7). IEGs are
proteins that become expressed as down-stream activities are taking place within a cell,
thus the increased expression of these proteins are used as indicators that a particular area
is involved in particular sets of behaviors or events. Although the most commonly
measured IEG is c-fos, other IEGs include fos-b, Jun-b, NGFI-A, and NGFI-B, all of
which indicate that cells are becoming active in response to some event experienced by
the animal. In a particular series of studies that used immobilization (IMO) as the
stressor, it was shown that along with differences in GC levels, there was a difference
between stressed and nonstressed animals in expression of IEGs in the paraventricular
nucleus of the hypothalamus (PVH; Senba and Ueyama 1997). It appeared that as GC
levels increased in the chronic IMO group, there was a suppression of expression of all
but one of the IEGs measured. This finding led to additional studies that applied the
chronic immobilization to animals that had been adrenalectomized and either left
52

untreated or treated with varying doses of corticosterone. It was discovered that chronic
stress, in the absence of corticosterone, led to the expression of all IEGs examined and as
corticosterone levels increased, expression of all but one of the IEGs measured was
suppressed, in a dose dependent manner.
TABLE 7. Hormone Levels and Expressions of IEGs in Senba and Ueyama (1997)
Abbreviations: E = Expressed, S = Suppressed, NE=norepinephrine, EPI=epinephrine,
GC=glucocorticoids, IMO=immobilization, N/A = Since animals were adrenalectomized, they were not
secreting GC. Symbols: ↑=levels elevated above normal, ↑↑=levels even more elevated compared to acute
IMO and remain elevated.

As is displayed above, increased GC release due to chronic stress suppresses the
expression of IEGs, with the exception of NGFI-A. It was concluded, that since the
NGFI-A gene response was similar in both acute and chronically stressed animals, the
animals did not habituate to the stressor. However, it has been shown in other chronic
IMO studies, that prolonged increases in GC levels leads to the suppression of all
examined IEGs, including NGFI-A (Melia, 1994, cited in Senba & Ueyama, 1997). It
Intact Acute
IMO
Intact Chronic
IMO
Adrenalectomized
Chronic IMO
Adrenalectomized
Chronic ↑GC
Hormones
NE ↑ ↑↑ ↑↑ ↑↑
EPI ↑ ↑↑ ↑↑ ↑↑
GC ↑ ↑↑ N/A Varying doses
IEGs
c-fos E S E S
(With higher GC doses)
fos-b E S E S
(With higher GC doses)
Jun-b E S E S
(With higher GC doses)
NGFI-A E E E E
(With all GC doses)
NGFI-B E S E S
(With higher GC doses)
53

was also reported that these animals showed a gradual decrease in GC levels over time,
unlike the group in the above studies, which showed chronically elevated levels, similar
to the witnessed difference between young and aged rats exposed to chronic stress.
Given the difference the difference in the results of the two IMO studies and the proposed
reason for differential NGFI-A expression, it is possible that unlike the animals in the
former IMO study who did not habituate, those in the latter study habituated to the
stressor. So, why did animals habituate in one case but not the other? In fact, the
difference may lie in the method of immobilization used in the two studies. In the study
showing no habituation (NGFI-A expression), immobilization was obtained by taping the
animals, on their backs, to a board, whereas, in the study showing habituation (NGFI-A
suppression), the animals were immobilized in tubes, where they could lie down in a
prone position. Taking the different methods and corresponding findings into account, it
seems as though NGFI-A could be used as a marker for stressor intensity.
Along these lines, remote US preexposure could be viewed as a form of chronic
stress, and the ineffectiveness of the procedure seen in aged animals may be due to an
inability to cope with the chronic illness, and thus habituate. As is discussed above, and
will be further discussed below, there seems to be ample evidence suggesting that
animals may develop an increased sensitivity to LiCl as they age due to various
physiological changes that occur during the aging process. This increased sensitivity to
the drug, may mean that aged animals are experiencing the effects of a higher dose of
LiCl than is experienced by the younger animals. Using the data described above,
perhaps the stress response in the young animals mimics that of the less intense chronic
54

IMO, whereas the stress response of the aged animals mimics that seen in the more
intense IMO.
TABLE 8. Hypothesized Hormone and IEG Patterns in Young and Aged Animals During Remote US
Preexposure
Intact Chronic IMO:
Less Intense
Young Animals:
Effective Remote
Preexposure
Intact Chronic IMO:
More Intense
Aged Animals:
Ineffective Remote
Preexposure
Hormones
NE ↓ Over time? ↓ Over time?
↑↑ ↑↑
EPI ↓ Over time? ↓ Over time?
↑↑ ↑↑
GC ↓ Over time ↓ Over time
↑↑ ↑↑
IEGs
c-fos S S S S
fos-b S S S S
Jun-b S S S S
NGFI-A S S E E
NGFI-B S S S S
Abbreviations: NE=norepinephrine, EPI=epinephrine, GC=glucocorticoids, IMO=immobilization, ↓ Over
time =levels decrease and return to normal over time, ↑↑=levels elevated far above normal and remain
elevated throughout exposure. Symbol: ? for NE and EPI in the “Intact Chronic IMO – Less Intense”
column, indicates that the provided data did not specify the levels of these chemicals over time. However,
since it appears that GC release is dependent upon the levels of these two hormones, I am speculating that
because GC levels are going down over time so are NE and EPI levels.

The increased sensitivity to LiCl would not only be able to account for the observed
ineffectiveness of the preexposure, but it could also account for the ability to aged
animals to form more potent avoidances with a traditional 0-minute interval, to learn
avoidances over longer CS-US intervals, to retain the memory of avoidance formation
over longer periods of time, and to exhibit increased resistance to extinction. With
respect to differences in strength of CTAs, it would be expected that increased sensitivity
to LiCl would lead to a more potent avoidance with the traditional 0-minute interval
versus a lower dose, as is the case for higher doses of LiCl. Such a difference in strength
could be expressed as either a larger drop in consumption from acquisition to the post-
55

acquisition test or the need for more extinction trials before extinction is met, or both. In
relation to the CS-US interval, if the aged animals are more sensitive to the effects of the
drug, then it would be expected that they would be able to form an avoidance over longer
intervals, as is the case for higher doses of LiCl (Misanin, Goodhart, Anderson, and
Hinderliter, 2002), since the US trace would be strong enough to compensate for the
decreased strength of the CS trace at the point of association. In relation to the ability to
retain the memory of avoidance formation over longer periods of time, it would be
expected that greater sensitivity to LiCl (and higher doses of LiCl) would allow better
retention of CTA memory, as measured by the ability to express a CTA after prolonged
intervals between acquisition and post-acquisition testing. It would be in the animal’s
best interest to remember a food-illness experience based on severe malaise for longer
periods than one based on mild malaise because the risk of illness and death would be
greater. And finally, with respect to extinction rate, one would expect that greater
sensitivity to LiCl would lead to a longer amount of time to completely extinguish, as is
the case for higher doses of LiCl, because they have acquired a stronger avoidance.
2.5.3 An Estrogen Hypothesis
Yet another possible explanation for the difference between young and aged female
rats in CTA learning involves the female sex hormone, estrogen. As females age, their
reproductive hormonal cycle becomes erratic and eventually ceases. The average age at
which the estrous cycle becomes irregular ranges from 10 to 12 months, slowly
progressing to a state of constant estrous due to chronic anovulation at approximately 19
months of age (Finch, 1978; Matt, Sarver, and Lu, 1987; Lu, Hopper, Vargo, and Yen,
56

1979). The irregularity and cessation of the cycle is also accompanied (or possibly
partially caused) by deceases in circulating levels of the estrogens (estradiol and estrone).
Specifically, one study found that on diestrous day 2 and the morning of proestrous
young animals had serum levels of 29 pg/ml and 41 pg/ml, respectively, whereas animals
11-13 months of age had serum levels of 17 and 18 pg/ml during the same time of the
cycle. As animals reached 30 months of age, their levels were higher than 17-18 pg/ml,
but were still significantly lower than the 29 or 41 pg/ml found in young regularly
cycling females (Lu, Hopper, Vargo, and Yen, 1979). Two of the consequences of
disruption of this cycle and the subsequent reduction in estrogen levels are an increase in
body weight or adiposity and a decrease in metabolism. As indicated above, it is possible
that the increased adiposity that accompanies decreased estrogen levels may prolong
sensations of malaise. Likewise, deceases in metabolism may indirectly have the same
effect on toxicity, as decreased rates of metabolism usually lead to weight gain,
particularly due to increased fat.
Estradiol is also thought to exert protective factors on the kidney, as declines in renal
function and the occurrence of renal diseases are less frequent in premenopausal females.
This hormone has also been found to upregulate angiotensin receptors, which would
likely allow for better management of water and sodium reabsorption and excretion by
way of the renal tubules (Baiardi, Macova, Armando, Ando, Tyurmin, and Saavedra
2005). There also is evidence that the rate of Li excretion declines with increasingly old
age. One study showed that rate of excretion in the rat progressively increased from an
age of 5 days to an age of 105 days, however, animals that were 240 days old had an
57

excretion rate similar to that of animals between 15 and 33 days old. It also was shown
that, despite the slower rate of Li excretion in younger animals, they survived longer than
the adult animals after the administration of sublethal doses of Li (Kersten, 1981).
Kersten (1979, cited in Kersten 1981) proposed that although Li is absorbed by the CNS
more slowly than other peripheral organs, it seems to be absorbed by the CNS of adult
animals more quickly than in young animals, and therefore death ensues more quickly. It
also has been noted in humans on Li treatments, that elderly individuals are more likely
to develop higher (i.e. toxic) serum levels of the drug than younger patients. The
absorption rate of Li in these patients is normal, however, since the cation is water-
soluble and the amount of water available in the elderly body is decreased, there is less
space for distribution, causing a marked increase in concentration. This is compounded
by the slower excretion rate due to reduced kidney function and a two-fold increase in the
half-life of the drug (Nakra and Grossberg, 1987). It seems reasonable that the discussed
changes would lead to an increased sensitivity to Li in aged animals, and thus a probable
explanation for the observed age differences in CTA learning.
Estradiol also acts on the brain to influence behavior, and reduced levels of estradiol
are known to produce changes in the brain (for review see McEwen and Alves, 1999).
Traditionally, the study of estrogen and its receptors has focused on the hypothalamus
and its role in reproductive behavior, however, the role of estrogen in other brain regions
and behaviors has become of increasing interest. Since the breadth of estrogen research
has widened, the hormone has been implicated in various types of learning, motor
coordination, and affective states (Zurkovsky, Brown, Boyd, Fell, and Korol, 2007;
58

Yune, Kim, Lee, Lee, Oh, Kim, Markelonis, and Oh, 2007; Fink, Sumner, Rosie, Grace,
and Quin, 1996). Estrogen also is known to modulate the activities of other
neurochemicals, such as the neuropeptides vasopressin and oxytocin (De Vries, Wang,
Bullock, Numan, 1994; Coirini, Johnson, McEwen, 1989). As such, it also has been
found that estrogen acts throughout the entire brain, with receptors spanning from the
forebrain to the brainstem (Simerly, Chang, Muramatsu, and Swanson, 1990; Simonian

and Herbison, 1997). Currently, estrogen is thought to have a number of protective
effects throughout the body. With respect to neuroprotective factors, hypothalamic
neurons in cell cultures survived longer in the presence of 17β-estradiol, as did fetal
cortical neurons which not only survived longer but also displayed nuerite outgrowth in
the presence of the hormone (Singer, Rogers, Strickland, and Dorsa, 1996; Diaz-Brinton,
Tran, Proffitt, and Montoya, 1997). The ovarian steroid also has been described as
protecting against oxidative damage (Behl, Widmann, Trapp, and Holsboer 1995). Not
surprisingly then, it appears that the decline in estrogen production that occurs in older
female rats may detrimentally affect various areas of the brain. Specifically, estrogen
loss is thought to be involved in the loss of synaptic connections in the hippocampus, as
well as the decline in cholinergic function in the basal forebrain (Lewis, McEwen, and
Frankfurt, 1995; Luine, 1985; Singh, Meyer, Millard, and Simpkins, 1994). Given the
functions and diseases that are associated with these areas, estrogen is assumed to have
protective effects against declines in learning and memory, Alzheimer’s disease, and
dementia (for review see McEwen and Alves, 1999). It is also likely that the
hypothesized hippocampal damage brought on by the absence of estrogen is further
59

impacted by the documented damage to this structure as a result of the increased GC
levels experienced by older organisms.
Estradiol, as well as LiCl, can serve as a preexposure agent and a US in the CTA
paradigm. It has been shown in the US preexposure design, that estradiol can serve as a
preexposure agent for a LiCl-induced CTA, thus attenuating the formation of the
avoidance (Yuan and Chambers, 1999). Likewise, it has been shown that LiCl can lead
to the attenuation of an estradiol-induced CTA (De Beun, Peeters, and Broekkamp,
1993). The interchangeability of the two substances has led to the supposition that LiCl
and estradiol cause the same pattern of activation in the brain, therefore allowing them to
successfully preexpose for each other (Rabin, Hunt, and Lee, 1989). If LiCl acts
similarly to estradiol in the neural areas mediating the US, then one should expect both
aged males and females, who would both be more sensitive to the effects of estrogen, to
respond differently to the drug than young animals. The reduced levels of estradiol in
these neural components in aged animals may make them more sensitive to LiCl than
younger animals. Support for this hypothesis comes from studies showing that male rats,
with low levels of estradiol, acquire CTAs with lower doses of estradiol and LiCl than
females (Chambers, Sengstake, Yoder, Thornton, 1981; De Beun, Jansen, Smeets,
Niesing, Slangen, and Van de Poll, 1991; Gustavson, Gustavson, Young, Pumariega, and
Nicolaus, 1989). Additionally, animals ovariectomized for 18 weeks prior to
conditioning of an estradiol-induced CTA develop significantly stronger avoidances to a
novel diet than do animals who have only been ovariectomized for 3 weeks at the time of
conditioning (Ganesan and Simpkins, 1991), suggesting that the longer animals are
60

ovariectomized, and thus without estrogen, the more sensitive they are to the effects of
the hormone. If the pattern of activation is the same for the two substances, then we
would also expect that female animals ovariectomized for a longer period of time would
also be more sensitive to the effects of LiCl as well.
2.6 Concluding Remarks
The fact that aged organisms experience defects in the realm of learning and memory
has been known for long time, although the extent and generalization of such shortfalls
may have been exaggerated throughout the years. Studies employing a wide range of
different learning tasks have reported age-related deficits in acquisition and retention.
The evidence suggesting that aged animals show deficits in the acquisition, or learning,
of a new behavior must be considered with caution. Although there have been a number
of studies showing that aged animals do generally require more time to perform at a level
equal to younger age groups, there also is evidence that at least some of the deficits can
attributed to other dysfunctions that develop with age, such as increased tendency to
perseverate and reduced ability to retain information. There are indications that the larger
dysfunction accompanying aging may be related to retention. Studies examining
retention have revealed that aged animals experience a fairly general decline in retention
abilities, as tasks requiring the animal to pull information from both short- and long-term
stores are less often performed successfully when compared to young groups.
These facets of learning also have been examined in the CTA paradigm. Unlike other
learning tasks, however, deficits in acquisition and retention have not been found. On the
contrary, aged animals acquire such avoidances more readily and remembered them for
61

longer periods of time. However, just as with the results from other learning tasks, the
data indicating stronger acquisition and better retention must be carefully interpreted.
Both of these effects could be due to increased toxicosis experienced by older subjects.
62

Chapter 3
Reacquisition of Conditioned Taste Avoidances
3.1 Aging and Reacquisition of a CTA
In many learning tasks, it has been demonstrated that animals are capable of
relearning the task after they have demonstrated what appears to be complete extinction
of the task. Not only can they relearn the task, but they can reacquire the extinguished
learned behavior in a lesser amount of time than the amount of time required to initially
learn that behavior (Napier, Macrae, and Kehoe, 1992; Weidemann and Kehoe, 2003).
This suggests that animals retain memory of the learned behavior despite apparent
extinction. Animals also can reacquire a CTA, using the same CS and US and it has been
suggested that they retain the memory of extinguished CTAs as well (Banerjee and Das,
1980; Hart, Bourne, and Schachtman, 1995; Aguado, de Brugada, and Hall, 2001). As
discussed above, five differences have been found between aged and young animals
when tested in the CTA paradigm: (1) preexposure to the US fails to attenuate the
strength of the developed avoidance in aged animals, whereas such exposure significantly
attenuates the strength of the developed avoidance in younger age groups (Valliere,
Misanin, and Hinderliter 1988; Misanin, Hoefel, Riedy, and Hinderliter 1997); (2) aged
animals tend to develop strong avoidances regardless of dose, unlike young animals who
will develop avoidances in a dose dependent manner (Misanin, Goodhart, Anderson, and
Hinderliter 2002); (3) aged animals can make food-illness associations over longer CS-
US intervals than younger animals (Misanin, Greider, and Hinderliter 1988; Misanin,
Goodhart, Anderson, and Hinderliter 2002; Misanin, Collins, Rushanan, Anderson,
63

Goodhart, Hinderliter 2002); (4) aged animals retain the memory of avoidance formation
longer than younger animals, in that they drink significantly less of the CS than younger
animals on the first extinction test, whether that test is 1 week or 12 weeks after
acquisition day (Guanowsky, Misanin, and Riccio 1983); (5) and aged animals show
increased resistance to extinction of a developed avoidance compared to younger aged
groups (Ingram ad Peacock, 1980; Springer and Fraley, 1981). Given the increased
proclivity towards taste avoidance learning and retention, it seems reasonable to suggest
that aged animals would be more likely to retain the memory of an extinguished CTA.
Although several different hypothesis have been put forward to account for the
various age- related differences in expression of CTAs, there is only one hypothesis that
seems to be able to adequately account for all of these differences, that aged animals
experience increased toxicosis or sensitivity to the US. Support for this hypothesis comes
from studies of young animals, which indicate that the direction of the differences
displayed by aged animals mimics that of young animals given higher doses of the US.
Given that the strength of acquisition and resistance to extinction rises as the intensity of
the US is increased (Nachman and Ashe, 1973; De Buen, Jansen, Smeets, Niesing,
Slangen, and Van de Poll, 1991), one would expect the ability to retain the memory of an
extinguished CTA to improve as the intensity of the US is increased. Given that this is
the case, the increased toxicosis/sensitivity hypothesis would predict that aged animals
would display a greater ability to retain memory of an extinguished CTA. Unfortunately,
unlike other learning tasks, systematic work on the facility with which an animal
reacquires a CTA after she has undergone extinction of that CTA has not been done.
64

Thus, studies were conducted with young rats first in order to determine whether
retention of the memory of an extinguished CTA is improved with higher US intensity.
3.2 Experiments Investigating Reacquisition of a CTA in Young Animals
3.2.1 Introduction
Reacquisition of extinguished behaviors has been observed in several different
classical conditioning paradigms. Several investigators have found that after extinction
trials, rabbits reacquire the nictitating membrane response at a more rapid rate than their
initial learning (Napier, Macrae, and Kehoe, 1992; Weidemann and Kehoe, 2003). In one
study, faster reacquisition occurred even after 1,200 extinction trials were given
(Weidemann and Kehoe, 2003). Salivary conditioning studies completed by Pavlov and
colleagues also have demonstrated that dogs reacquire extinguished conditioned
responses at an accelerated rate as compared to their initial leaning (Pavlov 1927). One
of the hypotheses made concerning these results is that the animal does not forget the
learned behavior, that is, some memory of the behavior is retained making it easier to
relearn it. Thus, the learning of the behavior and the extinction of the behavior are
thought to exist simultaneously, rather than extinction causing forgetting or unlearning of
the original learning experience and consequently replacing that memory. Evidence for
such an assertion comes from observations such as the above discussed reacquisition
studies, spontaneous recovery of the conditioned behavior after extinction training has
been completed, and reinstatement of the extinguished behavior with presentation of the
CS after the US is given alone (for review see Bouton, 1993; see also Delamater, 1996;
Rescorla, 2001). In fact, it has been proposed that extinction, not the original learned
65

behavior, is the memory that is subject to degradation because of events such as
spontaneous recovery and reinstatement (Bouton, 1994).
Although a few studies have shown that animals can reacquire an extinguished CTA,
this problem has not been systematically examined and therefore whether or not
extinction leads to the forgetting of the initial avoidance learning cannot be addressed
(Banerjee and Das, 1980; Hart, Bourne, and Schachtman, 1995; Aguado, de Brugada, and
Hall, 2001). Findings from other learning paradigms suggest that it should be easier to
reacquire a CTA and thus the memory of the extinguished CTA is retained. However,
determining the ability to retain a CTA is important because there are important
differences in the learning mechanism for CTA than other types of classical conditioning;
most notable is the ability of the CS memory trace during acquisition to endure for hours
(Kalat and Rozin, 1971).
In a study that induced a second avoidance, one could examine whether the initial
memory of learning had been destroyed or degraded as a result of extinction, through the
comparison of the strengths of the first and second CTAs and examination of different
patterns of CS consumption during extinction. If the memory of the acquisition of the
first CTA were retained, we would expect to find that animals exhibit a greater decrease
from acquisition to the first extinction test for the second CTA than for the first CTA and
that CS consumption during extinction of the second CTA is less than during the
extinction of the first CTA. In this case, although animals extinguish the first CTA, they
would not have unlearned that the substance makes them ill, thus allowing them to
retrieve their former experience and facilitating their ability to relearn that the substance
66

is not safe. The following experiments were designed to begin to determine whether the
memory of an apparently extinguished CTA is retained by ascertaining whether the
strength of acquisition and resistance to extinction is greater for a reacquired CTA.
3.2.2 General Methods
3.2.2.1 Subjects
Female Sprague-Dawley rats were obtained from Harlan Laboratories (San Diego,
CA). The rats were 60 days old and weighed approximately 190 grams. They were
housed individually in solid bottom cages (58 cm x 38 cm) that had wood chips as
bedding material. The vivarium in which the rats were housed was temperature (21-
22°C), humidity (51%), and light controlled (a 12 hour light and 12 hour dark cycle with
lights off at 0730 and lights on at 1930 h for experiment 1 and lights off at 10:30 and
lights on at 22:30 h for experiment 2). All rats were maintained on ad libitum rat chow
(Rodent Blox; protein 24%, fat 4%, fiber 4.5%; Harlan Teklad 8604) and tap water
throughout the duration of each experiment. They were allowed 1 week to adapt to their
living conditions before ovariectomy and were allowed one week of recovery from
surgery before behavioral procedures were initiated. The experiments were conducted
according to the standards set by the National Institutes of Health Guide for the Care and
Use of Laboratory Animals (DHEW Publication 80-23, Revised 1985, Office of Science
and Health Reports, DRR/NIH, Bethesda, MD) and the institutional guidelines of the
University of Southern California.

67

3.2.2.2 Surgical Procedures
All rats were allowed 1 week to adapt to their living conditions before ovariectomy.
All rats were anesthetized with a ketamine-xylazine solution, which was a mixture of 5
parts of Ketaject (equivalent to 100 mg per ml of ketamine; Phoenix Pharmaceutical, Inc)
to 1 part of Xylazine Sterile Solution (Lloyd Laboratories). This solution was
administered through intraperitoneal (i.p.) injections in doses of 0.12 ml/100 g body
weight. The rats then were ovariectomized, which involved bilateral incisions of the skin
and underlying muscles. Incisions were made 1 to 2 cm from the midline and anterior to
the hip; the uterine horns were ligated with catgut thread, and the ovaries were removed.
Muscle and skin closure were achieved with absorbable monofilament sutures. The ears
of the female rats were punched for individual identification after ovariectomy.
Immediately after surgery, all animals were given s.c. injections of 0.03 ml of the
analgesic buprenex and 0.05 ml of the antibiotic penicillin and were placed individually
in a clean polypropylene shoebox cage that had an insulated heating pad. They were
placed in their home cage after they have begun to show steady locomotion. Buprenex
also was given the day following surgery in order to alleviate any persisting discomfort.
The rats were given a week to recover from surgery before behavioral testing was
initiated.
3.2.2.3 Conditioning Procedures
Preconditioning. A week following ovariectomies, the preconditioning phase of the
experiment began. For 7 days prior to acquisition, animals were preconditioned to the
experimental procedures. Each day, prior to the onset of the dark phase, the animals were
68

weighed and at the start of the dark phase, each animal was given a cylinder of
refrigerated (4˚C) tap water for 1-hour. At the end of the hour the cylinders were
removed.
Acquisition. On the day of acquisition, the sequence of events closely followed the
preconditioning procedures. Once the dark phase began, each rat received a cylinder of
cold (4˚C) aqueous sucrose (10% w/v). After 1-hour, the cylinder was collected and the
remaining volume of sucrose was recorded. Immediately after this, each rat received an
injection of estradiol benzoate (Steraloids, Wilton, NH) or sesame oil vehicle
subcutaneously (s.c.) at the nape of the neck, as determined by its group designation. The
estradiol benzoate was dissolved in sesame oil and was injected in a volume of
0.25ml/10µg per kg or 0.25ml/50µg per kg of body weight. All animals received the
same US for both CTAs.
Extinction. Extinction trials for both the first and second CTAs were initiated 2 days
after acquisition. Similar to the procedure for acquisition, the rats received a cylinder of
the refrigerated sucrose solution with the onset of the dark phase. After an hour had
passed, the cylinders were removed and the remaining volumes were recorded. No,
injections of estradiol benzoate or sesame oil were given.
3.2.2.4 Statistical Methods
Conventional statistical procedures, such as ANOVAs, are laden with problems when
assumptions of normality and equal variances are violated. Violations of these
assumptions lead to poor probability coverage, low power, and inability to control for
Type I errors and bias, which means that the probability of rejecting can drop as the
69

population means become more unequal (Wilcox 2003a; Wilcox 2003b). Traditional
approaches to achieve normality and equality of variances (i.e. transformations) have also
been shown to be inadequate (Wilcox and Keselman 2003). Reliable modern statistical
techniques have been developed to replace such conventional methods of analyses. Due
to the nonnormality and heteroscedasticity of our data, more modern robust statistical
methodologies were implemented to attain reliable results. A percentile bootstrap with
20% trimmed means, which is not predicated on the assumptions of normality or equal
variances, was performed to analyze these data. In general, this method randomly
samples with replacement from the actual data to produce a data set called a bootstrap
sample. Next, the trimmed mean is computed. In the two-sample case, the p-value is the
probability that a bootstrap trimmed mean from the first group is less than the bootstrap
trimmed mean of the second. The value of p is computed by repeatedly generating
bootstrap samples. This procedure can be generalized to a repeated measures design
(Wilcox 2003b). Additionally, this method controls the family-wise error rate using
adjusted p-values. For all analyses computed in the present experiments, the family-wise
error rate was set at 0.05. These statistical methodologies have been found to perform
well among a range of methods when working with real data, including when working
with small sample sizes (Wilcox 2003b; Wu 2002).
Acquisition of a first CTA (CTA1) is typically defined as a significant decrease in
consumption from the acquisition test (ACQ) to the first extinction test (FET). However,
defining how acquisition of a second CTA (CTA2) should be defined is more
complicated because of the neophobic reactions rats express towards novel substances,
70

even when they are inherently palatable. These reactions are manifested as consumption
of small amounts of a novel substance during the first exposures. As the rats are given
more exposures to a palatable substance without experiencing malaise, neophobia is
slowly overcome and their consumption of the substance increases. In the case where
animals suffered malaise after their first exposure to a novel substance, and consequently
acquired a CTA, neophobic levels of consumption also can be surpassed if they are
allowed enough exposure to the substance without the negative consequences. Thus, it is
possible that any subsequent pairing of the substance with malaise could result in a drop
in consumption to a level that is still above the neophobic response. This raises the
question of whether a true CTA is any decrease in consumption or whether the decrease
must fall below neophobic levels. Therefore, in the present experiments, we defined
acquisition of CTA2 in two ways: (1) a significant decrease in consumption from ACQ2
to FET2 and (2) a significant decrease in consumption from ACQ1 to FET2, that is, a fall
in consumption to below the neophobic response.
Extinction of CTA1 was defined as resumption of ACQ1 consumption levels while
extinction of CTA2 was defined in two ways: (1) resumption of ACQ2 consumption
levels and (2) resumption of ACQ1 consumption levels. Analysis of sucrose
consumption across the extinction tests was used to assess differences in extinction. Due
to the nature of the bootstrap analysis, extinction trials were divided into phases.
Experiment 1 consisted of nine phases, while experiment 2 consisted of four phases.
Each of these phases included two extinction days. The value for each phase was
calculated by adding the amount consumed on two tests (i.e., first extinction test (E1) +
71

second extinction test (E2) = Phase 1, E3+E4 = Phase 2). Once the sums were calculated
for each animal the trimmed mean for each phase was computed and group-wise
comparisons were performed across all of the phases. Analyses comparing consumption
during the acquisition test and last extinction test of both CTAs also were completed in
order to assess the level of extinction achieved within each of the CTAs as well as across
the first and second CTA. For each comparison, a critical significance level was reported
along with a significance level. A significant result is obtained when the significance
level is lower than the critical significance value. Trimmed means without standard
errors were presented in all figures (Wilcox 2003b). This type of robust analysis
previously has been used to analyze CTA data (Hintiryan, Hayes & Chambers, 2005,
2006).
3.3 Experiment 1
3.3.1 Introduction
In experiment 1, we investigated the ability of two different doses of estradiol
benzoate to induce two consecutive avoidances to a sucrose solution in ovariectomized
female Sprague-Dawley rats. Because both doses induced a second CTA, we then
determined whether there were dose-related differences in the facility of acquiring and
maintaining the first CTA and the reacquired CTA.
3.3.2 Methods
Thirty-three female Sprague-Dawley rats were randomly divided into three groups
according to the US received (n=11 for each group): (1) sesame oil (Oil), (2) 10 µg/kg
estradiol benzoate (E10), and (3) 50 µg/kg estradiol benzoate (E50; see Table 9). The
72

CTA procedure described in the General Methods was followed. For both CTAs, 19
extinction tests were given. The time interval between the last extinction test of the first
CTA and the acquisition test of the second CTA was 9 weeks. Two animals from group
E10 did not develop a first CTA and one animal from group E50 did not extinguish a first
CTA, therefore they were removed from the statistical analyses.
TABLE 9. Experimental design for Experiment 1
Group CS US Extinction 1 CS US Extinction 2
Oil Sucrose O 19 Days Sucrose O 19 Days
E10 Sucrose E (10µg/kg) 19 Days Sucrose E (10µg/kg) 19 Days
E50 Sucrose E (50µg/kg) 19 Days Sucrose E (50µg/kg) 19 Days
Abbreviations: E = estradiol, O = Oil, Extinction 1 = Extinction of 1
st
CTA, Extinction 2 = Extinction of 2
nd

CTA

3.3.3 Results
3.3.3.1 Development of First CTA and Second CTA
Development of First CTA: As expected, both estradiol groups developed an
avoidance of the sucrose (see Figure 1: CTA1). Both estradiol groups differed from the
Oil group in the change in sucrose consumption across acquisition (ACQ1) and the first
extinction test (FET1) of CTA1 (critical significance = 0.05 and significance < 0.001 for
E10 vs. Oil and E50 vs Oil). Dependent analysis showed that while the Oil group
consumed similar amounts of sucrose during ACQ1 and FET1, groups E10 and E50
displayed less sucrose consumption during FET1 (critical significance = 0.05 and
significance < 0.001 for both cases). Also, independent analysis revealed that although
all groups consumed similar amounts of sucrose during ACQ1, both the E10 and E50
groups consumed significantly less sucrose than the Oil group during FET1 (critical
significance = 0.05 and significance < 0.001 for both cases).
73

The low-dose estradiol group acquired a weaker avoidance than the high-dose group.
The extent of the decrease in sucrose consumption was less for the E10 group than the
E50 group from ACQ1 to FET1 (critical significance = 0.05 and significance = 0.032).
However, these two groups did consume similar amounts of sucrose during both ACQ1
and FET1, suggesting that the difference in the relative strengths of the CTAs expressed
by the two groups was not robust (critical significance = 0.05 and significance = 0.482
for ACQ1; critical significance = 0.05 and significance = 0.055 for FET1).
Development of Second CTA Based on Consumption during the Second Acquisition
Test: When avoidance was defined as a significant decrease in consumption from the
ACQ2 to FET2, both estradiol groups developed a second CTA (see Figure 1). Both
estradiol groups differed from the Oil group in the change in sucrose consumption across
acquisition (ACQ2) and the first extinction test (FET2) of CTA2 (critical significance =
0.05 and significance = 0.008 for E10 vs. Oil and critical significance = 0.05 and
significance < 0.001 for E50 vs. Oil). Dependent analysis showed that while the Oil
group consumed similar amounts of sucrose during ACQ2 and the first extinction test,
groups E10 and E50 displayed less sucrose consumption during FET2 (critical
significance = 0.05 and significance = 0.019 for E10 and critical significance = 0.05 and
significance < 0.001 for E50). Independent analysis once more revealed that although all
groups consumed similar amounts of sucrose during ACQ2, both the E10 and E50 groups
consumed significantly less sucrose than the Oil group during FET2 (critical significance
= 0.05 and significance = 0.002 for E10 vs Oil and critical significance = 0.05 and
significance < 0.001 for E50 vs Oil).
74

The animals given the low dose of estradiol acquired weaker avoidances than the
high-dose animals. The extent of the decrease in sucrose consumption was less for the
E10 group than the E50 group across ACQ2 and FET2 (critical significance = 0.05 and
significance = 0.004). In addition, while the two groups consumed similar amounts of
sucrose during ACQ2, group E10 consumed significantly more sucrose than group E50
during FET2 (critical significance = 0.05 and significance < 0.001).
FIGURE 1. Experiment 1: Development of First CTA and Second CTA

Trimmed mean of sucrose consumption displayed by ovariectomized females during the acquisition test
(ACQ) and the first extinction test (FET) of their first conditioned taste avoidance (CTA1, displayed on the
left) and their second CTA (CTA2, displayed on the right). For both CTAs, acquisition consisted of pairing
a 10% sucrose solution with a subcutaneous injection of either sesame oil vehicle (Oil), 10 µg dose of
estradiol (E10), or a 50 µg dose of estradiol (E50). Those females injected with oil in CTA1 also were
injected with oil in CTA2 and those females injected with E10 or E50 in CTA1 also were injected with the
same dose of estradiol in CTA2. For both CTA 1 and CTA 2, all animals were given 19 daily extinction
tests during which they were given access to the sucrose solution for 1 hour each day beginning 2 days after
the respective ACQ test.
a
Significant change in sucrose consumption between ACQ and FET.
b
Significant
difference between Oil and E10 for the given test.
c
Significant difference between Oil and E50 for the
given test.
d
Significant difference between E10 and E50 for the given test.

Comparisons between First CTA and Second CTA Based on Consumption during the
Second Acquisition Test: For both estradiol groups, the strength of acquisition of CTA1
was similar to the strength of CTA2 (see Figure 2). Dependent analyses revealed that the
extent of the decrease in sucrose consumption across ACQ1 and FET1 was similar to the
decrease across ACQ2 and FET2 for the E10 and E50 groups (critical significance = 0.05
and significance = 0.52 for E10; critical significance = 0.05 and significance = 0.84 for
75

E50). Additional dependent analyses comparing the amount consumed during
acquisition revealed that group E10 consumed less sucrose during ACQ1 than ACQ2
(critical significance = 0.05, significance < 0.001) while group E50 consumed similar
amounts of sucrose (critical significance = 0.05, significance = 0.697). For the first
extinction test, group E10 consumed less sucrose during FET1 than FET2 (critical
significance = 0.05, significance < 0.001) while group E50 consumed more sucrose
during FET1 than FET2 (critical significance = 0.05, significance = 0.001).
The oil group showed the same pattern of change for both CTAs. As indicated above,
the Oil group showed no change in sucrose consumption across ACQ and FET for both
CTA1 and CTA2. However, this group consumed less sucrose during ACQ1 than ACQ2
(critical significance = 0.05, significance = 0.018) and they consumed less sucrose during
FET1 than FET2 (critical significance = 0.05, significance < 0.001).
FIGURE 2. Experiment 1: Comparisons between the First CTA and Second CTA

Trimmed mean of sucrose consumption displayed by ovariectomized females during the acquisition test
(ACQ) and the first extinction test (FET) of their first conditioned taste avoidance (CTA1) and their second
CTA (CTA2). For both CTAs, acquisition consisted of pairing a 10% sucrose solution with a subcutaneous
injection of either sesame oil vehicle (Oil, displayed on the left), 10 µg dose of estradiol (E10, displayed in
the middle), or a 50 µg dose of estradiol (E50, displayed on the right). Those females injected with oil in
CTA1 also were injected with oil in CTA2 and those females injected with E10 or E50 in CTA1 also were
injected with the same dose of estradiol in CTA2. For both CTA 1 and CTA 2, all animals were given 19
daily extinction tests during which they were given access to the sucrose solution for 1 hour each day
beginning 2 days after the respective ACQ test. For each group, the extent of the change between ACQ and
FET was the same for both CTAs.
a
Significant difference between CTA1 and CTA2 for ACQ.
b
Significant difference between CTA1 and CTA2 for FET.

76

Development of a Second CTA Based on Consumption during the First Acquisition
Test: As indicated in the statistical methods, we also compared the amount consumed
during ACQ1 (the first exposure to sucrose) and FET2 in order to determine whether
reductions in consumption after the second pairing of sucrose and estradiol were below
the neophobic response (see Figure 3). Although the high-dose estradiol animals fell
below the neophobic response and thus clearly developed a CTA, the low-dose estradiol
animals did not. Although both estradiol groups differed from the Oil group in the
change in sucrose consumption across ACQ1 and FET2 (critical significance = 0.05 and
significance = 0.008 for E10 vs. Oil and critical significance = 0.05 and significance <
0.001 for E50 vs. Oil), neither the Oil group nor the E10 group showed decreases in
consumption. Dependent analysis showed that the Oil group consumed less sucrose
during ACQ1 than FET2 (critical significance = 0.05, significance < 0.001), the E10
group consumed similar amounts of sucrose during ACQ1 and FET2 (critical
significance = 0.05, significance = 0.463), and the E50 group consumed more sucrose
during ACQ1 than FET2 (critical significance = 0.05, significance < 0.001).
Direct comparisons of the change in sucrose consumption across ACQ1 and FET2 for
the two estradiol groups revealed a significant difference (critical significance = 0.05 and
significance < 0.001). In addition, as stated above, both groups drank similar amounts of
sucrose during ACQ1 but group E10 consumed more sucrose than group E50 during
FET2.
Comparisons Between First CTA and Second CTA Based on Consumption during the
First Acquisition Test: For the estradiol groups, avoidances displayed by the low-dose
77

estradiol animals were stronger for CTA1 than CTA2 while the avoidances of the high-
dose estradiol animals were stronger for CTA2 than CTA1 (see Figure 3). Comparisons
of the extent of change in sucrose consumption from ACQ1 to FET1 with the change
from ACQ1 to FET2 revealed significant differences for both estradiol groups (critical
significance = 0.025, significance < 0.001 for E10; critical significance = 0.05,
significance = 0.001 for E50). The E10 group showed a decrease in sucrose consumption
ACQ1 and FET1 but no change in consumption across ACQ1 and FET2 while the E50
group showed a lesser amount of decrease in sucrose consumption across ACQ1 and
FET1 than across ACQ1 and FET2.
The Oil group also showed different patterns of change for the two CTAs.
Comparisons of the extent of change in sucrose consumption from ACQ1 to FET1 with
the change from ACQ1 to FET2 revealed a significant difference (critical significance =
0.025, significance < 0.001). As indicated above, the Oil group showed no change in
sucrose consumption across ACQ1 and FET1 but an increase in consumption across
ACQ1 and FET2. In addition, this group consumed less sucrose during FET1 than FET2
(critical significance = 0.05, significance < 0.001).
78

FIGURE 3. Experiment 1: Comparisons between the First CTA and Second CTA Based on ACQ1
Consumption

Trimmed mean of sucrose consumption displayed by ovariectomized females during the acquisition test
(ACQ1) of their first conditioned taste avoidance (CTA1) and the first extinction test (FET) of CTA1 and
CTA2. For both CTAs, acquisition consisted of pairing a 10% sucrose solution with a subcutaneous
injection of either sesame oil vehicle (Oil), 10 µg dose of estradiol (E10), or a 50 µg dose of estradiol
(E50). Those females injected with oil in CTA1 also were injected with oil in CTA2 and those females
injected with E10 or E50 in CTA1 also were injected with the same dose of estradiol in CTA2. For both
CTA 1 and CTA 2, all animals were given 19 daily extinction tests during which they were given access to
the sucrose solution for 1 hour each day beginning 2 days after the respective ACQ test.
a
Significant
change in sucrose consumption between ACQ and FET.
b
Significant difference between CTA1 and CTA2
in the extent of the change between ACQ and FET for each group.
c
Significant difference between Oil and
E10 for FET1.
d
Significant difference between Oil and E50 for FET1.
e
Significant difference between Oil
and E10 for FET2.
f
Significant difference between Oil and E50 for FET2.
g
Significant difference between
E10 and E50 for FET2.

3.3.3.2 Extinction of First CTA and Second CTA
Extinction of First CTA: In accordance with the two estradiol groups developing
CTAs, we found that animals in these groups consumed less than the Oil group during
extinction (See Figure 4: CTA1). Group E10 consumed less sucrose during the first and
fourth phase of CTA1 extinction than the Oil group (critical significance = 0.01136 and
significance < 0.001 for phase 1; critical significance= 0.01278 and significance=0.012
for phase 4). Group E50 consumed less sucrose than group Oil throughout each phase of
extinction (critical significances ranged from 0.01136-0.05 and significances ranged from
0-0.02). This group also consumed less sucrose than E10 during the first 8 extinction
79

phases of CTA1 (critical significances ranged from 0.01136-0.05, significances ranged
from 0-0.014).
Extinction of Second CTA: Similar to what was observed in CTA1, the estradiol
animals again consumed less sucrose during extinction (See Figure 4: CTA2). Group
E10 consumed less sucrose than the Oil group during phases one, two, four, and seven
(critical significances ranged from 0.01136-0.01702 and significances ranged from 0-
0.008). Group E50 again differed from the Oil group during all nine extinction phases of
CTA2 (critical significances ranged from 0.01136-0.05 and significances ranged from 0-
0.006). This group also consumed less sucrose than E10 during all phases except the 7
th

phase (critical significances ranged from 0.01136-0.05, significances ranged from 0-
0.012).
FIGURE 4. Experiment 1: Consumption During Extinction of the First CTA and Second CTA

Trimmed mean of sucrose consumption displayed by ovariectomized females during 9 phases of CTA1
(displayed on the left) and CTA2 (displayed on the right). For both CTAs, acquisition consisted of pairing
a 10% sucrose solution with a subcutaneous injection of either sesame oil vehicle (O), 10 µg dose of
estradiol (E10), or a 50 µg dose of estradiol (E50). Those females injected with oil in CTA1 also were
injected with oil in CTA2 and those females injected with E10 and E50 in CTA1 also were injected with
the same dose of estradiol in CTA2. For both CTA 1 and CTA 2, all animals were given 19 daily
extinction tests during which they had access to the sucrose solution for 1 hour each day beginning 2 days
after the respective ACQ test.
a
Significant difference between groups Oil and E10 for the given phase.
b
Significant difference between groups Oil and E50 for the given phase.
c
Significant difference between
groups E10 an E50 for the given phase.

80

Comparisons between First CTA and Second CTA: For the Oil group, consumption
of sucrose during extinction of CTA1 and CTA2 was similar throughout all 9 phases. On
the other hand, the pattern of sucrose consumption for groups E10 and E50 during
extinction of CTA1 differed from that of CTA2 (See figure 5). E10 consumed more
sucrose during phase one and two of CTA2 (critical significance = 0.01702 and
significance = 0 for phase 1; critical significance = 0.02040 and significance = 0.004 for
phase 2). However, from phase three to phase six, this group consumed similar amounts
of sucrose during both CTAs, and during phase seven to nine, they consumed more
sucrose during CTA1 (critical significance = 0.01278 and significance = 0 for phase 8;
critical significance = 0.01136 and significance = 0 for phase 9). Group E50, unlike
group E10, consumed similar amounts of sucrose for the first seven phases of CTA1 and
CTA2 but similar to group E10, the E50 group consumed more sucrose during phases
eight and nine of CTA1 (critical significance = 0.01136 and significance = 0 for phase 8;
critical significance = 0.01278 and significance = 0.002 for phase 9).

81

FIGURE 5. Experiment 1: Comparisons between Extinction of the First CTA and Second CTA

Trimmed mean of sucrose consumption displayed by ovariectomized females during 9 phases of CTA1 and
CTA2. For both CTAs, acquisition consisted of pairing a 10% sucrose solution with a subcutaneous
injection of either sesame oil vehicle (O, (displayed on the left)), 10 µg dose of estradiol (E10, displayed in
the middle), or a 50 µg dose of estradiol (E50, displayed on the right). Those females injected with oil in
CTA1 also were injected with oil in CTA2 and those females injected with E10 and E50 in CTA1 also were
injected with the same dose of estradiol in CTA2. For both CTA 1 and CTA 2, all animals were given 19
daily extinction tests during which they were given access to the sucrose solution for 1 hour each day
beginning 2 days after the respective ACQ test.
a
Significant difference between CTA1 and CTA2 for the
given phase.

3.3.3.3 Comparisons between Acquisition and Last Extinction Tests of First CTA and
Second CTA
The three groups differed in the pattern of sucrose consumption they displayed as
they progressed from ACQ1 to the last extinction test (LET) of CTA2 (see Figure 6).
Pattern of Consumption: The Oil animals, and both estradiol groups (after the initial
drop from ACQ1 to FET1), increased their sucrose consumption across the extinction
tests of CTA1 such that their consumption levels were higher during LET1 than ACQ1
(critical significance = 0.05, significance < 0.001 for all cases).
After a delay of 9 weeks, consumption for all groups dropped during ACQ2 as
compared to LET1 (critical significance = 0.05, significance <0.001 for all cases).
Despite the drop over the 9-week delay, the Oil and E10 animals consumed more during
ACQ2 than ACQ1 (critical significance = 0.05, significance = 0.018 for Oil; critical
significance = 0.05, significance < 0.001 for E10). Unlike the Oil and E10 groups, after
82

the 9-week delay the E50 animals consumed similar amounts of sucrose during ACQ1
and ACQ2 (critical significance = 0.05, significance = 0.697).
Similar to CTA1, the Oil animals, and E10 animals (after the initial drop in
consumption across ACQ2 and FET2), increased consumption across the extinction tests
of CTA1 such that their sucrose consumption during LET2 was greater than during
ACQ2, as well as ACQ1 (critical significance = 0.05, significance < 0.001 for both cases
of group Oil; critical significance = 0.05, significance = 0.011 for E10 LET2 vs ACQ2;
critical significance = 0.05, significance = < 0.001 for E10 LET2 vs ACQ1), but less than
their consumption during LET1 (critical significance = 0.05, significance = 0.007 for Oil;
critical significance = 0.05, significance = < 0.001 for E10). Similar to group E10, after
the drop in consumption across ACQ2 and FET2 the E50 animals increased their
consumption across extinction tests of CTA2. However, unlike the E10 animals, the
consumption levels achieved during LET2 only reached levels similar to those consumed
during ACQ2 and ACQ1 (critical significance = 0.05, significance = 0.561 for LET2 vs
ACQ2; critical significance = 0.05, significance = 0.78 for LET2 vs ACQ1). Similar to
both the Oil and E10 groups, their consumption during LET2 was lower than the level
achieved during LET1 (critical significance = 0.05, significance = 0.004)
Group Comparisons for the Last Extinction Test: The pattern of consumption
between groups during both LET1 and LET2 was similar to the extinction phase results
for CTA1 and CTA2. Independent analysis revealed that during LET1 the Oil animals
consumed similar amounts of sucrose as the E10 animals (critical significance = 0.05,
significance = 0.77), but more than the E50 animals (critical significance = 0.05,
83

significance = 0.006). The same pattern was found between the Oil and estradiol groups
during LET2 (critical significance = 0.05, significance = 0.18 for Oil vs E10; critical
significance = 0.05, significance = 0.005 for Oil vs E50). Comparisons between the E10
and E50 animals revealed that the groups consumed similar amounts of sucrose during
LET1 (critical significance = 0.05, significance = 0.082). Similarly, although the E10
animals consumed more sucrose than the E50 animals during phase 9 of CTA2, they
consumed similar amounts during LET2 (critical significance = 0.05, significance =
0.08).
FIGURE 6. Experiment 1: Consumption During ACQ and LET of the First CTA and Second CTA

Trimmed mean of sucrose consumption displayed by ovariectomized females during acquisition (ACQ)
and the last extinction test (LET) of the first (CTA1) and second (CTA2) conditioned taste avoidance. For
both CTAs, ACQ consisted of pairing a 10% sucrose solution with a subcutaneous injection of either
sesame oil vehicle (Oil), a 10 µg/kg dose of estradiol (E10), or a 50 µg/kg dose of estradiol (E50). Those
females injected with Oil in CTA1 also were injected with oil in CTA2 and those females injected with E10
and E50 in CTA1 also were injected with the same dose of estradiol in CTA2. For both CTA 1 and CTA 2,
FIGURE 6 Caption Con’t: all animals were given 19 daily extinction tests during which they were given
access to the sucrose solution for 1 hour each day beginning 2 days after the respective ACQ test.
a
Significant difference in sucrose consumption between ACQ1 and LET1.
b
Significant difference in
sucrose consumption between LET1 and ACQ2.
c
Significant difference in sucrose consumption between
ACQ2 and LET2.
d
Significant difference in sucrose consumption between ACQ1 and ACQ2.
e
Significant
difference in sucrose consumption between LET1 and LET2.
f
Significant difference in sucrose
consumption between Oil and E50 for given test.
g
Significant difference in sucrose consumption between
E10 and E50 for given test.
84

3.3.4 Discussion
The first aim of this experiment was to determine whether animals could reacquire a
CTA using a sucrose solution as the CS and estradiol as the US in both the first and the
second CTA. Acquisition is typically defined as a significant decrease in consumption
from the acquisition test to the first extinction test. However, when two avoidances are
conditioned, the question arises as to which acquisition test should be used for
comparison to the first extinction test of CTA2 in order to determine whether the CTA
was reacquired. Recall that when acquisition of CTA2 was based on consumption across
ACQ2 and FET2, both estradiol groups reacquired their avoidance to sucrose. However,
when acquisition was based on consumption across the first exposure to sucrose (ACQ1)
and FET2 only the high-dose estradiol group showed a decrease in consumption and thus
developed a second avoidance. As mentioned in the statistical methods, consumption of
a novel palatable substance increases with repeated exposures in rats. This increase is
thought to represent a loss of neophobia and the consumption levels during the first
acquisition test is thought to represent the neophobic response. In the present study, the
reduction in consumption after the second pairing fell below the neophobic response for
group E50 but remained at the neophobic response for group E10. There is no question
that group E50 reacquired the CTA, which is consistent with the results of previous
studies showing that animals can reacquire a CTA when the same CS and US are used in
both CTAs (Banerjee and Das, 1980; Hart, Bourne, and Schachtman, 1995; Aguado, de
Brugada, and Hall, 2001). But the question is raised as to whether the reduction in
consumption to the neophobic response by group E10 represents a true CTA or is simply
85

a resumption of a neophobic (or cautious) type response. This issue is important as
existing evidence suggests that neophobia and CTA are independent mechanisms
(Buresova and Bures, 1980; Reilly and Trifunovic, 2001).
The second aim of this experiment was to determine whether animals retain the first
taste-illness associative experience despite showing behavioral extinction. If as a result
of extinction, the only memory that exists at the time of the second acquisition is that the
sucrose solution is a familiar and safe substance, then one would expect animals not to
acquire a second CTA, or at most, only a weak one. This supposition is based on the
well-established finding that the ability of animals to acquire a first CTA is compromised
after repeated and safe exposures to the CS (Misanin, Guanowsky, and Riccio, 1983; De
La Casa and Lubow, 1995; Rodriguez and Alonso, 2002; Heth and Pierce, 2007). On the
other hand, if retention of the taste-illness associative experience is sufficiently robust,
we would expect a second CTA to be stronger than the first CTA, as the retrieved
memory would facilitate the ongoing associative process. In the Hart, Bourne, and
Schachtman (1995) and Aguado, de Brugada, and Hall (2001) studies mentioned above,
comparisons were made between the experimental and control groups to determine
relative strengths of acquisition of the second CTA. The same CS (saccharin for both
studies) and US (lithium chloride [LiCl] for both studies) were paired for both CTAs in
the experimental group, while in the control group, different CSs (coffee and saccharin
for the Hart study; vinegar and saccharin for the Aguado study) were paired with the
same US (LiCl for both studies) for the two CTAs. The CS (saccharin for both studies)
that was paired with the US in the second CTA was the same for both the experimental
86

and control groups. The results of both studies did not provide evidence for robust
retention of the first CTA memory in the experimental group, as this group developed
weaker avoidances than the control group. However, such a supposition is premature
because it is unclear what kind of effect having had a previous CTA with one CS might
have had on learning a subsequent CTA with a different CS but the same US. Our study
circumvented this problem by pairing the same CS and US for both CTAs and then
comparing the behaviors displayed after each pairing. Unfortunately, the results of our
study do not allow a simple answer as it depends on how one defines acquisition of a
CTA and whether one looks at the initial phases or the final phases of extinction.
Looking first at acquisition, if acquiring a second CTA is defined as a decrease in
sucrose consumption from the amount consumed in the second acquisition trial
(comparing the extent change from ACQ1 to FET1 and ACQ2 to FET2), then for both
estradiol groups, the strengths of acquisition were similar. Such a result suggests that the
taste-illness memory is retained and that it is sufficiently strong to partially obstruct or
erode the extinction memory, thus allowing a taste-illness association to proceed. On the
other hand, if acquiring a second CTA is defined as a decrease from the amount
consumed during the first exposure to the sucrose solution (comparing the extent
decrease from ACQ1 to FET1 and ACQ1 to FET2), then dose effects are found. The
low-dose estradiol group acquired a first CTA but not a second CTA, suggesting that the
taste-illness memory is not retained and thus there is no obstruction or erosion of the
extinction memory. On the other hand, the high-dose estradiol group showed a greater
decrease after the second pairing of the CS and US than after the first, suggesting that the
87

taste-illness memory is retained and that it is sufficiently robust to obstruct or block the
extinction memory and to facilitate the ongoing taste-illness associative process.
With respect to extinction, if the first CTA memory were retained we would expect
animals to initiate extinction at a slower rate and to consume less sucrose during the first
extinction phases of the second avoidance than during the first avoidance. Looking at the
initial phases of extinction for the first and second avoidance, we found that the low-dose
group consumed more sucrose during the first two phases of CTA2 than during the first
two phases of CTA1. This pattern suggests that any retention of the negative memory of
the first CTA was not sufficiently robust enough to impact acquisition of the reacquired
CTA. On the other hand, the high-dose group consumed similar amounts of sucrose
during the first seven phases of CTA1 and CTA2. These results suggest that the first
avoidance memory is retained with sufficient strength to partially obstruct or erode the
extinction memory, thus allowing reacquisition to occur.
Sucrose consumption during the final phases of extinction paints a different picture.
The low-dose and high-dose estradiol groups consumed less sucrose during the last
phases of CTA2 than they did during the last phases of CTA1. Similar results were
found when consumption during the last extinction test for each CTA was analyzed. This
could be evidence that both of the estradiol groups do retain the first negative experience,
and that despite reaching complete extinction, the development of a first CTA exerts
lasting effects on the overall consumption of the target CS after subsequent avoidance
formation and extinction. However, it should be noted that while the Oil group
consumed similar amounts of sucrose during all nine extinction phases for both the first
88

and second avoidance, they did consume less sucrose during the last extinction test of the
CTA2 than CTA1. Taken together, these data raise the question of whether the
suppressed consumption exhibited by the estradiol animals is evidence that they do retain
the negative memory or whether another event resulted in a decrease in consumption by
all groups. Although another event may explain the suppressed consumption displayed
by all groups on the last extinction test of CTA2, such an explanation fails to account for
the consistent suppression in consumption displayed by the two estradiol groups during
four out of the last five extinction tests of the second avoidance (group E10 displayed
suppressed consumption during E16-E19 of CTA2 compared to CTA1 and group E50
showed suppressed consumption during E15, E16, E18 and E19 of CTA2 compared to
CTA1 [critical significances = 0.05 for all cases, range of significances <0.001-0.036]).
In contrast to the estradiol groups, examination of the data revealed that out of the last
five extinction tests the Oil animals only differed on E19 of CTA1 and CTA2
(consumption was the same during E15-E18 of CTA1 and CTA2; critical significance =
0.05 for all cases, significances ranged from 0.154-0.873). Furthermore, while LET2
consumption was similar to E16-E18 of CTA2, consumption during LET1 was aberrantly
high when compared to consumption levels in those extinction tests that immediately
preceded it (E15 and E17; critical significance = 0.05 for both cases, significance <0.001
and 0.024, respectively), suggesting that the difference between LET1 and LET2 for the
oil animals was due simply to random variation. Thus, the possibility remains that the
suppressed consumption displayed by the estradiol groups is indicative of retention of the
first avoidance memory.
89

Putting all of the results together, then, we found no evidence for robust retention of
the first illness experience for the low-dose estradiol animals. There was evidence
suggesting that these animals do retain some memory for the first taste-illness experience.
This was limited to the similar strengths of acquisition displayed during CTA1 and CTA2
(based on ACQ2) and the suppressed consumption during the last extinction phases and
the last extinction test of CTA2 compared to CTA1. The majority of the evidence
suggested that it is the retention of the extinction experience that is robust in the low-dose
estradiol animals. First, the low-dose animals did not develop a second avoidance when
based on ACQ1 consumption and sucrose consumption during the first extinction test of
CTA2 did not fall below the neophobic response. Second, consumption increased across
the acquisition tests of CTA1 and CTA2. Third, initiation of extinction was faster during
CTA2 (based on ACQ2) than CTA1. Finally, consumption was greater during the first
extinction test and the first extinction phase of CTA2 than CTA1. With respect to the
high-dose estradiol group, there was evidence that the avoidance memory is retained and
that it is sufficiently strong to moderately or significantly obstruct or erode the extinction
memory. The majority of evidence suggests a moderate retention of the avoidance
memory. First, the strengths of acquisition of CTA1 and CTA2 (based on ACQ2) were
similar. Second, the levels of consumption during the acquisition tests of CTA1 and
CTA2 were similar. Third, the consumption of sucrose during most of the phases of
extinction of CTA1 and CTA2 was similar. Finally, the suppressed consumption during
the last extinction phases and the last extinction test was similar for CTA1 and CTA2.
There also were two pieces of evidence suggesting that retention of the first conditioning
90

experience was sufficiently robust to significantly impact acquisition and extinction of
the second CTA. When acquisition of CTA2 was based on the first exposure to the CS,
the acquisition of CTA2 was stronger than CTA1 and when extinction criterion for CTA2
was based on ACQ2, animals reached 100% extinction more slowly during CTA2 than
CTA1. Taken together, different conclusions must be made for the two estradiol groups.
The preponderance of data for the low-dose group suggests that the most robust memory
retained was the first extinction experience, while the preponderance of data for the high-
dose group suggests that the first conditioning experience is sufficiently strong to impact
acquisition and extinction of the second conditioning experience.
The results of the present study clearly showed that different doses of estradiol lead to
differences in the strength of acquisition and resistance to extinction of the first CTA.
Recall that group E10 exhibited smaller decreases in consumption after acquisition than
group E50. Furthermore, group E10 initiated extinction sooner than group E50 and
consumed more sucrose during the first 8 extinction phases of CTA1, suggesting that
because the E50 animals developed a stronger avoidance, they required more time to
reach similar levels of consumption. Previous studies also have found that various doses
of estradiol lead to dose dependent avoidances such that the lowest doses lead to the
weakest avoidances and the highest doses lead the strongest avoidances (De Beun,
Jansen, Smeets, Niesing, Slangen, and Van De Poll, 1991). These results from a first
CTA lend credence to the supposition that the differences in acquisition strength and
resistance to extinction of the second CTA also are due to dose dependent differences.
As was true for CTA1, the extent of the decrease in consumption after the second
91

acquisition was greater for group E50 than group E10, regardless of definition, and group
E50 initiated extinction more slowly and consumed less sucrose during extinction than
group E10. In addition, as described above, the two estradiol groups differed in the
patterns of sucrose consumption displayed during CTA1 compared to CTA2. Although
all of these dose differences could simply be due to dose effects, there is a potential
contributing factor that should be considered, especially with respect to the differences
shown when CTA1 was compared to CTA2. It is possible that these differences are tied
to the number of exposures to sucrose that were received after reaching extinction of
CTA1. Because the E10 animals developed weaker avoidances during CTA1 than the
E50 animals, and all animals received 19 extinction tests, the E10 animals received more
exposures to the CS after they had reached their ACQ1 consumption values than the E50
animals. While the number of exposures to sucrose after reaching the criterion for
complete extinction ranged from 1 to 13 for the high-dose group, the values for the low-
dose group ranged from 14 to 18 exposures. A potential consequence of continued
exposure to the CS after extinction, as defined as reaching acquisition levels of
consumption, is a further weakening of the negative memory of a substance that had
triggered malaise and/or strengthening of a trace for a substance that has inherent
palatability and caloric value.
One interesting finding not related to the two aims of this study was that the Oil group
exhibited a decrease in consumption during the 9-week interim between the last
extinction test of CTA1 and the acquisition test of CTA2. One possible reason for this
result is that neophobia was reinstated. The Oil animals had increased their consumption
92

during extinction such that they consumed more sucrose during LET1 than during ACQ1,
indicating an abatement of neophobia. Perhaps with an interval as long as 9 weeks, some
familiarity with the substance is lost, allowing for a reinstatement of neophobia.
However, because sucrose consumption was higher during ACQ2 than ACQ1, these
animals do retain some memory of the substance, otherwise they would have consumed
an equal amount on both acquisition tests. Both of the estradiol groups also showed a
decrease in sucrose consumption during the 9-week interval. For the E10 animals,
sucrose consumption had increased during extinction so that by LET1, they were
consuming the same amounts as the Oil animals. As was true for the Oil animals, the
E10 animals also consumed more sucrose during ACQ2 than ACQ1. These similarities
suggest that the E10 animals experienced a partial reinstatement of neophobia as well.
On the other hand, the decrease from LET1 to ACQ2 is a bit harder to interpret for the
E50 animals. Although sucrose consumption increased during extinction, these animals
consumed less sucrose during LET1 than the Oil animals and they consumed similar
amounts of sucrose during ACQ2 and ACQ1. Two possible explanations for these results
come to mind: (1) they had no retention of their prior experiences with the CS and thus
exhibited a complete reinstatement of neophobia, or (2) the level of extinction achieved
was insufficient to reduce the impact of the negative memory of their prior conditioning
experience with the CS and the retention of this experience was strong enough to prevent
the animals from increasing their consumption from ACQ1 to ACQ2.
In summary, this experiment was designed to determine whether animals had the
ability to reacquire an avoidance with either a low or high dose of estradiol and whether
93

the memory of the prior conditioning experience exerted an influence on the pattern of
consumption during the second CTA. The results of the present study show that animals
conditioned with both low and high doses of estradiol can reacquire an extinguished
avoidance. Furthermore, they suggest that the degree to which the memory of the first
avoidance impacts the second avoidance is dependent on the conditioning dose
employed; the extinction memory is more robust when a low-dose of estradiol is used
while the conditioning memory is of sufficient strength to at least partially obstruct or
erode the extinction memory when a high-dose is used. However, other factors may have
resulted in the suggested dose differences. One of the possible contributors to the
difference is the potential “over-extinction” resulting from the larger number of
exposures E10 received after reaching acquisition levels of CS consumption. With more
safe exposures to the sucrose solution, the animals should have better learned that this CS
no longer makes them ill, is palatable, has a caloric value, and thus is something worth
consuming.
3.4 Experiment 2
3.4.1 Introduction
As indicated in the above discussion, the strength of the first CTA was greater than
the second CTA for the low-dose estradiol animals while there was a tendency for the
strength of the second CTA to be greater than the first CTA for the high-dose estradiol
animals. This difference could simply be a dose effect, as the E10 animals expressed
weaker avoidances then the E50 animals for both CTAs. However, because all animals
were given 19 extinction tests, the E10 animals received more exposure to sucrose after
94

they had extinguished CTA1 than the E50 animals. Thus, at least part of this difference
could have been a consequence of greater exposure to sucrose that was now “safe.” If
this is the case, then it suggests that “over-extinction” could provide some protection
from another negative experience. In order to begin testing the effect of degree of
extinction on subsequent ability to acquire and extinguish a second CTA, the following
experiment was designed to determine the effect of either partial extinction (50% of
acquisition values) or complete extinction (100% of acquisition values) on the
development of a second avoidance.
Another outcome that was noted to have a potential impact on the quality of the
second avoidance was the partial reinstatement of neophobia. Because all groups,
including the Oil group, exhibited decreases in consumption from the last extinction test
of CTA1 to the acquisition test of CTA2, we asserted that the animals had experienced a
partial reinstatement of neophobia (sucrose consumption during acquisition of CTA2 was
higher than during acquisition of CTA1) during the 9-week interim between the first and
second avoidance. However, for the high-dose estradiol group it was difficult to
ascertain whether the decrease across the last extinction test of CTA1 and the acquisition
test of CTA2 was due only to a complete reinstatement of neophobia (sucrose
consumption during acquisition of CTA1 and CTA2 was similar) or whether retention of
the prior negative experience from the first avoidance contributed to this decrease. Thus,
in order to better assess which of these two possibilities is more likely, we shortened the
interim between CTA1 and CTA2 to only 7 days.
95

As such, Experiment 2 aimed at addressing two issues. First, this experiment was
designed to begin to determine the impact of various levels of extinction on the ability of
an animal to reacquire an extinguished avoidance. If the cause of the differential effect
between the two estradiol groups during the second avoidance of Experiment 1 was in
fact due to the extent of extinction reached during CTA1, we would expect that animals
only allowed to reach 50% extinction during CTA1 should develop stronger avoidances
during CTA2 compared to animals allowed to reach 100% extinction during CTA1.
Second, this experiment was designed to determine whether a shorter interim between the
first and second CTA abolishes the observed reinstatement of neophobia. If the high-
dose animals in experiment 1 exhibited the decrease in consumption from the last
extinction test of CTA1 to the acquisition test of CTA2 because they retained the first
negative experience, then we would expect to see a decrease across the same two days
with the shorter interim between the first and second avoidance. However, if it is due to
the reinstatement of neophobia, then we would expect to find that animals consume
similar amounts on the two days.
3.4.2 Methods
Forty female Sprague-Dawley rats were randomly divided into four groups according
to the US and number of extinction tests given (n=10 for each group): (1) 50µg/kg dose
of estradiol benzoate and complete extinction (EC), (2) sesame oil and complete
extinction (OC), (3) 50µg/kg dose of estradiol benzoate and partial extinction (EP), and
(4) sesame oil and partial extinction (OP; see Table 10). The CTA procedure described
in the General Methods was followed. For extinction of the first CTA, animals assigned
96

to the EC group were allowed to completely extinguish, which was defined as 3 days of
sucrose consumption at or above the acquisition trial consumption levels. The animals in
the EP group, however, were only allowed to partially extinguish, which was defined as
reaching 50% of their acquisition day consumption levels (e.g., if an animal drank 12 ml
on acquisition day, she only had to drink 6 ml in order to meet criteria). After each rat
achieved her extinction criterion, she was given chilled tap water for 7 days before
receiving a second acquisition trial. The time interval between the last extinction test of
the first CTA and the acquisition test of the second CTA was 7 days. All animals from
both groups were allowed to completely extinguish from the second CTA. For both
CTAs, animals from the OC group were yoked to animals in group EC, such that the
number of sucrose exposures each animal received depended on the number of extinction
trials its yoked partner from group EC received. In the same manner, animals from the
OP groups were yoked to animals in the EP group. One animal from group EC did not
develop a first CTA so she and her yoked partner from group OC were eliminated from
the statistical analyses.
TABLE 10. Experimental Design for Experiment 2
Group CS US Extinction 1 CS US Extinction 2
EC Sucrose E (50µg/kg) Complete Sucrose E (50µg/kg) Complete
OC Sucrose O Complete Sucrose O Complete
EP Sucrose E (50µg/kg) Partial Sucrose E (50µg/kg) Complete
OP Sucrose O Partial Sucrose O Complete
Abbreviations: E = estradiol, O = Oil, Extinction 1 = Extinction of 1
st
CTA, Extinction 2 = Extinction of 2
nd

CTA, Complete = 100% + 2days of acquisition consumption levels, Partial = 50% of acquisition
consumption levels

97

3.4.3 Results
3.4.3.1 Development of First CTA and Second CTA
Development of First CTA: As expected, both estradiol groups developed an
avoidance of the sucrose (see Figure 7). Each estradiol group differed from its respective
oil group in the change in sucrose consumption across ACQ1 and FET1 (critical
significance = 0.05 and significance < 0.001 for EC vs OC; critical significance = 0.05
and significance < 0.001 for EP vs OP). Dependent analysis showed that group OC
consumed similar amounts during ACQ1 and FET1 while group EC displayed a decrease
in consumption during FET1 compared to ACQ1 (critical significance = 0.05,
significance < 0.001). On the other hand, Group OP exhibited an increase in
consumption from ACQ1 to FET1 (critical significance = 0.05, significance = 0.026),
while group EP exhibited a drop in consumption across the two days (critical significance
= 0.05, significance < 0.001). Independent analysis revealed that while both estradiol
groups consumed similar amounts of sucrose during ACQ1 as their respective controls,
they consumed less sucrose during FET1 than their respective controls (critical
significance = 0.05 and significance < 0.001 for both cases).
Both estradiol groups acquired avoidances that were similar in strength. The two
estradiol groups did not differ in the extent of the decrease from ACQ1 to FET1 (critical
significance = 0.05 and significance = 0.33) or in the amount of sucrose consumed during
ACQ1 and FET1 (critical significance = 0.05 and significance =0.53 for ACQ1; critical
significance = 0.05 and significance =0.419 for FET1). The oil groups also did not differ
in the extent decrease from ACQ1 to FET1 (critical significance = 0.05 and significance
98

= 0.30) or in the amount of sucrose consumed during ACQ1 and FET1 (critical
significance = 0.05 and significance = 0.458 for ACQ1; critical significance = 0.05 and
significance = 0.98 for FET1).
Development of Second CTA Based on Consumption during the Second Acquisition
Test: When avoidance was defined as a significant decrease in consumption from the
ACQ2 to FET2, both estradiol groups developed a second CTA (see Figure 7). Each
estradiol group differed from its respective oil group in the change in sucrose
consumption across ACQ2 and FET2 (critical significance = 0.05 and significance <
0.001 for EC vs OC; critical significance = 0.05 and significance < 0.001 for EP vs OP).
Dependent analysis showed that while the groups OC and OP consumed similar amounts
during ACQ2 as they did during FET2, groups EC and EP displayed a decrease in
consumption on FET2 compared to their consumption during ACQ2 (critical significance
= 0.05 and significance < 0.001 for both cases). Also, independent analysis revealed that
while groups EC and OC consumed similar amounts of sucrose during ACQ2, group EC
consumed less than group OC during FET2 (critical significance = 0.05, significance =
0.52). As expected, group EP consumed less sucrose than group OP during ACQ2
(critical significance = 0.05 and significance = 0.002) and during FET2 (critical
significance = 0.05 and significance < 0.001).
The relative strength of the avoidances of the two estradiol groups is not clearly
discernable. Although, the extent decrease of sucrose consumption from ACQ2 to FET2
was greater for group EC than group EP (critical significance = 0.05 and significance <
0.001), the amount sucrose consumed during both ACQ2 and FET2 was less for group
99

EP than group EC (critical significance = 0.05 and significance < 0.001 for ACQ2;
critical significance = 0.05 and significance = 0.031 for FET2). The oil groups again did
not differ in the extent decrease from ACQ2 to FET2 (critical significance = 0.05 and
significance = 0.351) or in the amount consumed during ACQ2 and FET2 (critical
significance = 0.05 and significance = 0.798 for ACQ2; critical significance = 0.05 and
significance = 0.941 for FET2).
FIGURE 7. Experiment 2: Development of First CTA and Second CTA

Trimmed mean of sucrose consumption displayed by ovariectomized females during the acquisition test
(ACQ) and the first extinction test (FET) of their first conditioned taste avoidance (CTA1, displayed on the
left) and their second CTA (CTA2, displayed on the right). For both CTAs, acquisition consisted of pairing
a 10% sucrose solution with a subcutaneous injection of either sesame oil vehicle (O) or a 50 µg dose of
estradiol (E). Those females injected with oil in CTA1 also were injected with oil in CTA2 and those
females injected with E in CTA1 also were injected with E in CTA2. For CTA1, half of the O and E
females were allowed partial extinction (P) and the other half of the O and E females were allowed
complete extinction (C). In the case of P, extinction testing for each female was terminated when she had
reached at least 50% of her acquisition test consumption and in the case of C, extinction testing was
terminated for each female when she had consumed at least 100% of her acquisition test consumption for 3
days.
a
Significant change in sucrose consumption between ACQ and FET.
b
Significant difference between
OP and EP for the given test.
c
Significant difference between OC and EC for the given test.
d
Significant
difference between EP and EC for the given test.

Comparisons between First CTA and Second CTA Based on Consumption during the
Second Acquisition Test: The two estradiol groups differed in the relative strengths of
the two CTAs (see Figure 8). Dependent analyses revealed that the extent of the decrease
in sucrose consumption across ACQ1 and FET1 was similar to the decrease across ACQ2
and FET2 for group EC (critical significance = 0.05 and significance = 0.20). For group
100

EP, however, the decrease across ACQ1 to FET1 was greater than the decrease across
ACQ2 to FET2 (critical significance = 0.02 and significance < 0.001). Additional
dependent analysis comparing the amounts consumed during acquisition revealed that
group EC consumed more sucrose during ACQ2 than during ACQ1 (critical significance
= 0.05 and significance < 0.001), whereas group EP consumed more sucrose during
ACQ1 than during ACQ2 (critical significance = 0.05, significance = 0.034). For the
first extinction test, group EC consumed more sucrose during FET2 than during FET1
(critical significance = 0.05 and significance = 0.013), whereas group EP consumed
similar amounts of sucrose during FET1 and FET2 (critical significance = 0.05 and
significance = 0.441).
For the oil groups, dependent analyses revealed that the extent change in sucrose
consumption across ACQ1 and FET1 was similar to the change across ACQ2 and FET2
(critical significance = 0.05 and significance = 0.51 for OC; critical significance = 0.05
and significance = 0.33 for OP). Additional dependent analysis comparing the amounts
consumed during acquisition revealed that both oil groups consumed more sucrose during
ACQ2 than during ACQ1 (critical significance = 0.05 and significance < 0.001 for OC;
critical significance = 0.05 and significance = 0.003 for OP). For the first extinction test,
groups OC and OP consumed more sucrose during FET2 than during FET1 (critical
significance = 0.05 and significance < 0.001).
101

FIGURE 8. Experiment 2: Comparisons between the First CTA and Second CTA

Trimmed mean of sucrose consumption displayed by ovariectomized females during the acquisition test
(ACQ) and the first extinction test (FET) of their first conditioned taste avoidance (CTA1) and their second
CTA (CTA2). For both CTAs, acquisition consisted of pairing a 10% sucrose solution with a subcutaneous
injection of either sesame oil vehicle (O) or a 50 µg dose of estradiol (E). Those females injected with O in
CTA1 also were injected with O in CTA2 and those females injected with E in CTA1 also were injected
with E in CTA2. For CTA1, half of the O and E females were allowed partial extinction (P) and the other
half of the O and E females were allowed complete extinction (C). In the case of P, extinction testing for
each female was terminated when she had reached at least 50% of her acquisition test consumption and in
the case of C, extinction testing was terminated for each female when she had consumed at least 100% of
her acquisition test consumption for 3 days.
a
Significant difference between CTA1 and CTA2 for ACQ.
b
Significant difference between CTA1 and CTA2 for FET.

Development of Second CTA Based on Consumption during the First Acquisition
Test: When avoidance was defined as a significant decrease in consumption from ACQ1
to FET2, both estradiol groups acquired a CTA (see Figure 9). They both differed from
their respective oil groups in the change in sucrose consumption across ACQ1 and FET2
(critical significance = 0.05 and significance < 0.001 for EC vs OC; significance = 0.05
and significance < 0.001 for EP vs OC). Dependent analysis showed that both of the oil
groups consumed less sucrose during ACQ1 than FET2 (critical significance = 0.05 and
102

significance < 0.001 for both cases) while both of the estradiol groups consumed more
sucrose during ACQ1 than FET2 (critical significance = 0.05 and significance = 0.048 for
EC; critical significance = 0.05 and significance < 0.001 for EP).
The relative strength of the avoidances of the two estradiol groups were similar.
Independent analysis revealed that the extent of the decrease in sucrose consumption
from ACQ1 to FET2 was similar for the two groups (critical significance = 0.05 and
significance = 0.23). Additional analyses revealed that the amount of sucrose consumed
during ACQ1 was similar for the estradiol groups (critical significance = 0.05 and
significance = 0.53), but group EC consumed mores sucrose than group EP during FET2
(critical significance = 0.05 and significance = 0.0.031). The oil groups also exhibited
similar changes in consumption from ACQ1 to FET2 (critical significance = 0.05 and
significance = 0.73). They also consumed similar amounts of sucrose during ACQ1 and
FET2 (critical significance = 0.05 and significance = 0.80 for ACQ1; critical significance
= 0.05 and significance = 0.94 for FET2).
Comparisons between First CTA and Second CTA Based on Consumption during the
First Acquisition Test: The two estradiol groups differed in the relative strengths of the
two CTAs (see Figure 8). Group EC showed a larger decrease from ACQ1 to FET1 than
the decrease from ACQ1 to FET2 (critical significance = 0.034 and significance = 0.013)
.while group EP, showed a similar amount of decrease from ACQ1 to FET1 and ACQ1 to
FET2 (critical significance = 0.05 and significance = 0.44).
Both of the oil groups expressed a greater increase in sucrose consumption across
ACQ and FET of CTA2 than CTA1. Group OC showed no change across ACQ1 and
103

FET1 but an increase in consumption across ACQ1 and FET2 while group OP showed a
lesser amount of increase in consumption across ACQ1 and FET1 than across ACQ1 and
FET2 (critical significance = 0.025 and significance < 0.001 for both oil groups).
FIGURE 9. Experiment 2: Comparisons between the First CTA and Second CTA Based on ACQ1
Consumption

Trimmed mean of sucrose consumption displayed by ovariectomized females during the acquisition test
(ACQ1) of their first conditioned taste avoidance (CTA1) and the first extinction test (FET) of CTA1 and
CTA2. For both CTAs, acquisition consisted of pairing a 10% sucrose solution with a subcutaneous
injection of either sesame oil vehicle (O) or a 50 µg dose of estradiol (E). Those females injected with O in
CTA1 also were injected with O in CTA2 and those females injected with E in CTA1 also were injected
with E in CTA2. For CTA1, half of the O and E females were allowed partial extinction (P) and the other
half of the O and E females were allowed complete extinction (C). In the case of P, extinction testing for
each female was terminated when she had reached at least 50% of her acquisition test consumption and in
the case of C, extinction testing was terminated for each female when she had consumed at least 100% of
her acquisition test consumption for 3 days.
a
Significant change in sucrose consumption between ACQ and
FET.
b
Significant difference in the extent of the change between ACQ1 and FET1.
c
Significant difference
between OC and EC for FET1.
d
Significant difference between OP and EP for FET1.
e
Significant
difference between OC and EC for FET2.
f
Significant difference between OP and EP for FET2.
g
Significant difference between EC and EP for FET2.

3.4.3.2 Extinction of First CTA and Second CTA2
Because all animals received a variable number of extinction tests, based on their
individual rate of extinction, we only completed analyses on 8 extinction tests, which was
the maximum number of tests during which all animals received sucrose. No extinction
phase analyses were completed for group EP and OP for CTA1 because group EP was
only allowed to reach partial extinction of CTA1 and the majority of animals reached
104

criteria in a less than 8 tests (The number of extinction tests for each group ranged from 1
to 16 exposures for group EC and 1 to 12 exposures for group EP. However, the majority
of animals in group EP reached criteria within 5 extinction tests).
Extinction of First CTA: In accordance with group EC developing a first CTA, we
found that animals in this group consumed less sucrose than group OC during all four of
the extinction phases of CTA1 (critical significances ranged from 0.0254-0.05 and
significances ranged from 0-0.002; see Figure 10).
Extinction of Second CTA: As expected, the estradiol groups consumed less sucrose
than their respective oil groups throughout all four of the extinction phases of CTA2
(critical significances ranged from 0.0254-0.05 and significances ranged from 0-0.05; see
Figure 10).
Groups OC and OP consumed similar amounts of sucrose during each of the four
phases, while group EC consumed more sucrose than group EP during each of the four
phases (critical significances ranged from 0.0254-0.05 and significances ranged from
0.004-0.010; see Figure 10).
Comparisons between CTA1 and CTA2 Extinction: Both the OC and EC groups
showed similar patterns of sucrose consumption during extinction of CTA1 as opposed to
CTA2. These two groups consumed less sucrose during the first two phases of CTA1
than they did during the first two phases of CTA2 (critical significance = 0.0254-0.0338
and significance = 0.0-0.017 for phase 1; critical significance = 0.0254-0.0338 and
significance = 0.007-0.010 for phase 2).
105

FIGURE 10. Experiment 2: Consumption During Extinction of the First CTA and Second CTA

Trimmed mean consumption of sucrose consumption displayed by groups EC and OC during the 4 two-test
phases of CTA1 and CTA2 (panel to the left) and the trimmed mean consumption of sucrose for all groups
during the 4 two-test phases of CTA2 (panel to the right). For both CTAs, acquisition consisted of pairing
a 10% sucrose solution with a subcutaneous injection of either sesame oil vehicle (O) or a 50 µg dose of
estradiol (E). Those females injected with oil in CTA1 also were injected with oil in CTA2 and those
females injected with E in CTA1 also were injected with E in CTA2. For CTA1, half of the O and E
females were allowed partial extinction (P) and the other half of the O and E females were allowed
complete extinction (C). In the case of P, extinction testing for each female was terminated when she had
reached at least 50% of her acquisition test consumption and in the case of C extinction testing was
FIGURE 10 Caption Con’t: terminated for each female when she had consumed at least 100% of her
acquisition test consumption for 3 days.
a
Significant difference between groups OC and EC for the given
phase of CTA1.
b
Significant difference between groups OC and EC for the given phase of CTA2.
c
Significant difference between CTA1 an CTA2 for group OC for the given phase.
d
Significant difference
between CTA1 and CTA2 for group EC for the given phase.
e
Significant difference between groups OP and
EP for the given phase.
f
Significant difference between groups EC and EP for the given phase.

3.4.3.3 Comparisons between the Acquisition Test and the Last Extinction Test of CTA1
and CTA2
Pattern of Consumption for Groups EC and OC during CTA1 and CTA2: The two
groups differed in the pattern sucrose consumption as they progressed from ACQ1 to the
last extinction test of CTA2 (see Figure 11). Group OC steadily increased their sucrose
consumption across the extinction tests of CTA1 such that their consumption levels were
higher during LET1 than ACQ1 (critical significance = 0.05 and significance < 0.0001).
After a delay of 7 days, their consumption dropped during ACQ2 so that it was
significantly lower than LET1 (critical significance = 0.05 and significance = 0.004) but
106

was still higher than ACQ1. Then their sucrose consumption increased again during the
extinction tests of CTA2 such that their sucrose consumption during LET2 was greater
than during ACQ2 (critical significance = 0.05 and significance = 0.011), and was similar
to their consumption during LET1 (critical significance = 0.05 and significance = 0.616).
After the drop in sucrose consumption across ACQ1 and FET1, the EC animals
increased their consumption during the extinction tests of CTA1 so that their
consumption during LET1 was greater than their consumption during ACQ1 (critical
significance = 0.05 and significance < 0.001; see Figure 11). Unlike group OC, the 7-day
delay between CTA1 and CTA2 did not result in a drop in sucrose consumption; rather,
group EC consumed similar amounts of sucrose during ACQ2 and LET1 (critical
significance = 0.05 and significance = 0.571; see Figure 11) and their consumption
during ACQ2 was greater than during ACQ1. Again, their sucrose consumption dropped
across ACQ2 and FET2 and it increased across the extinction testing, such that their
consumption during LET2 was greater than during ACQ2 (critical significance = 0.05
and significance < 0.001). Also unlike group OC, animals in group EC consumed more
sucrose during LET2 than they consumed during LET1 (critical significance = 0.05 and
significance = 0.001).
107

FIGURE 11. Experiment 2: Consumption During ACQ and LET of the First CTA and Second CTA

Trimmed mean consumption of sucrose consumption displayed by ovariectomized females during the
acquisition test of CTA1 and CTA2 (ACQ1 and ACQ2, respectively) and last extinction test of CTA1 and
CTA2 (LET1 and LET2, respectively). For both CTAs, acquisition consisted of pairing a 10% sucrose
solution with a subcutaneous injection of either sesame oil vehicle (O) or a 50 µg dose of estradiol (E).
Those females injected with oil in CTA1 also were injected with oil in CTA2 and those females injected
with E in CTA1 also were injected with E in CTA2. For CTA1, half of the O and E females were allowed
partial extinction (P) and the other half of the O and E females were allowed complete extinction (C). In
the case of P, extinction testing for each female was terminated when she had reached at least 50% of her
acquisition test consumption and in the case of C extinction testing was terminated for each female when
she had consumed at least 100% of her acquisition test consumption for 3 days.
a
Significant difference in
sucrose consumption between ACQ1 and LET1.
b
Significant difference in sucrose consumption between
LET1 and ACQ2.
c
Significant difference in sucrose consumption between ACQ2 and LET2.
d
Significant
difference in sucrose consumption between ACQ1 and LET2.
e
Significant difference in sucrose
consumption between ACQ1 and ACQ2.
f
Significant difference in sucrose consumption between LET1
and LET2.
g
Significant difference in sucrose consumption between OP and EP for given test.
h
Significant
difference in sucrose consumption between EC and EP for given test.

Pattern of Consumption for Groups OP and EP during CTA1 and CTA2: As
expected by virtue of only reaching partial extinction, the pattern of consumption
exhibited by group EP differed from that of group OP during CTA1 and CTA2.
Dependent analysis revealed that group OP steadily increased consumption from ACQ1
to LET1 (critical significance = 0.05 and significance < 0.001), from ACQ2 to LET2
(critical significance = 0.05 and significance = 0.047), and from ACQ1 to LET2 (critical
significance = 0.05 and significance < 0.001). Similar to group EC, group OP consumed
similar amounts during LET1 and ACQ2 (critical significance = 0.05 and significance =
108

0.84), and like group OC, they consumed similar amounts during LET1 and LET2
(critical significance = 0.05 and significance = 0.08; see Figure 11).
As expected by the early termination of extinction tests during CTA1, group EP
consumed less sucrose during LET1 than ACQ1 (critical significance = 0.05 and
significance < 0.001). However, like groups OP, OC, and EC, group EP consumed more
sucrose during LET2 than ACQ1 (critical significance = 0.05 and significance < 0.001)
and ACQ2 (critical significance = 0.05 and significance < 0.001). Also like group EC,
group EP surpassed LET1 consumption during LET2 (critical significance = 0.05 and
significance < 0.001).
Group Comparisons for the Last Extinction Test: Independent analysis revealed that
groups OC and EC consumed similar amounts of sucrose during both LET1 and LET2
(critical significance = 0.05 and significance = 0.06 for LET1; critical significance = 0.05
and significance = 0.576 for LET2). Unlike groups OC and EC, however, group OP
consumed significantly more sucrose during LET1 and LET2 than group EP (critical
significance = 0.05 and significance < 0.001 for LET1; critical significance = 0.05 and
significance < 0.001 for LET2). Likewise, group EC consumed more sucrose during
LET1 and LET2 than group EP (critical significance = 0.05 and significance < 0.001 for
LET1 for both cases). Groups OC and OP consumed similar amounts of sucrose during
LET1 and LET2 (critical significance = 0.05 and significance = 0.14 for LET1; critical
significance = 0.05 and significance = 0.77 for LET2; see Figure 11).

109

3.4.4 Discussion
One of the aims of this study was to determine the impact of complete and partial
extinction on the strength of a reacquired avoidance. While both the complete and partial
extinction groups reacquired the avoidance, such that consumption during FET2 fell
below the neophobic response, the preponderance of the data suggest that a reacquired
CTA is stronger when animals have only been allowed partial extinction of the first CTA.
First, group EP consumed less sucrose than group EC during FET2; second, group EP
consumed less sucrose than group EC during each of the four extinction phases; and
third, group EP consumed less sucrose than group EC during the last extinction test of
CTA2. These data lend support to the suggestion that the weaker reacquired CTA
expressed by the low-dose estradiol group in experiment 1 might have been a result of
this group receiving a larger number of exposures to the CS after reaching acquisition
consumption levels during extinction, and thus they reached a higher level of extinction
than the high-dose group.
In line with the differential effects seen during CTA2 based on the level of CTA1
extinction achieved, we also found that the EC and EP groups showed differential
retention of the first avoidance experience. For group EC, when acquisition of a second
CTA is defined as a decrease in sucrose consumption from the amount consumed during
the second acquisition trial, then the extent of the decrease in consumption was similar
for both CTAs, and thus the strengths of acquisition were similar (comparing the extent
change from ACQ1 to FET1 and ACQ2 to FET2). This finding suggests that retention of
the extinction experience was sufficiently robust enough to impact acquisition of the
110

reacquired CTA. First, if acquisition of a second CTA is defined as a decrease from the
amount consumed during the first exposure to the sucrose solution, then the extent of the
decrease in consumption is greater for CTA1 than CTA2 and thus the animals in this
group developed a weaker second avoidance (comparing the extent decrease from ACQ1
to FET1 and ACQ1 to FET2). Second, consumption was greater during the first two
phases of CTA2 extinction than CTA1 extinction. Third, animals in this group increased
consumption from LET1 to LET2. These results also suggest that retention of the
extinction experience was sufficiently robust enough to impact acquisition of the
reacquired CTA.
For group EP, our results are somewhat misleading. First, when strength of the
second avoidance was based on the extent decrease from ACQ2 to FET2, animals in this
group developed a weaker second avoidance. Taken alone, this evidence suggests that
retention of the first illness experience was not sufficiently robust enough to affect
acquisition of CTA2. However, it is important to remember that these animals were only
allowed to reach 50% extinction during CTA1 and consequently the second acquisition
test could be viewed as the next extinction test for the first CTA. As such, one must take
into consideration that because group EP consumed a lesser amount of sucrose during
ACQ2 than ACQ1, the extent decrease of the second avoidance could not have been as
large as the first avoidance. There is a similar problem in interpretation when
development of a second avoidance was based on the first acquisition test. Although the
strength of CTA1 and CTA2 were similar, floor effects may have obscured any possible
differences. Therefore, any conclusions about retention of conditioning and/or extinction
111

experiences must be made on the basis of consumption levels during extinction of CTA2.
Evidence that disputes the lack of retention of the conditioning experience is the
observation that animals in the EP group consumed less sucrose than group EC during
each phase of CTA2 extinction and during LET2. Because EP animals were allowed to
completely extinguish CTA2, if they had no retention of the first illness experience we
would have expected them to reach consumption levels comparable to the EC group.
This expectation is based on the fact that the only difference between the EP and EC
groups was the level of extinction allowed during CTA1. Based on these different
patterns of consumption during extinction, we must come to different conclusions for
each estradiol group. For group EC, the preponderance of data suggests that the most
robust memory retained was the first extinction experience. With respect to group EP,
the data suggest that the most robust memory retained was the first conditioning
experience.
The second aim of this experiment was to determine whether a shorter interval
between avoidances resulted in an abolition of the observed reinstatement of neophobia
in experiment 1. Although our results were not consistent, the preponderance of our data
leads to one conclusion. First, it is important to note that while group EP consumed
similar amounts during LET1 and ACQ2, this comparison is meaningless as extinction
was terminated early for this group making ACQ2 just another extinction test for CTA1.
Therefore the pattern exhibited by group EP cannot speak to reinstatement of neophobia.
However, recall that while groups EC and OP consumed similar amounts of sucrose
during the last extinction test of CTA1 and the acquisition test of CTA2, group OC
112

decreased consumption across the two tests. This inconsistency might indicate that
shortening the interval had no effect on the restatement of neophobia. However, since
group EC and group OP consumed similar amounts during LET1 and ACQ2, and groups
OC, OP, and EC increased consumption from ACQ1 to ACQ2, it seems more probable
that the decrease in consumption from LET1 to ACQ2 exhibited by the OC group was an
anomaly. Thus, the preponderance of evidence suggests that the shorter interval did
abolish the reinstatement of neophobia.
In summary, this experiment was designed to determine whether the amount of
extinction achieved during the first CTA impacted the strength of the reacquired CTA
and thus influenced the retention of the first avoidance memory. The results of the
present study suggest that animals allowed to reach either partial or complete extinction
of the first CTA are able to reacquire the avoidance, but differences in retention of the
first illness experience do appear to be based on the amount of extinction achieved. The
most robust memory retained for the EC group was the first extinction experience and the
most robust memory retained for the EP group was the first conditioning experience. We
also examined whether a shorter interval between the two avoidances would abolish the
reinstatement of neophobia and found that shortening the interval between the first and
second avoidance did, in fact, abolish the reinstatement of neophobia.
3.5 General Discussion
3.5.1 Comparisons between Experiment 1 and Experiment 2: CTA1 versus CTA2
Although we cannot make direct comparisons between the two experiments we can
indirectly compare how group E50 from Experiment 1 and group EC from Experiment 2
113

acted as a result of the different designs. We would expect the two groups to act
similarly during CTA1 since all aspects leading up to and including acquisition and the
first extinction test of this CTA were the same. However, after this, the two experiments
differed in two ways. First, the procedures used for determining the termination of
extinction testing were different. While extinction testing was terminated after the E50
animals from Experiment 1 had received 19 extinction tests, each of the EC animals were
tested only until they had consumed at least 100% of their acquisition consumption levels
during three tests. Second, the length of the interval imposed between CTA1 and CTA2
was longer for the E50 group than the EC group. This suggests that one of these factors
could account for any difference the two groups displayed during CTA2 (and/or in the
CTA1 versus CTA2 comparisons).
As expected, the two high-dose estradiol groups acted similarly during CTA1. Both
estradiol groups consumed similar amounts of the sucrose as their respective controls
during ACQ1 and consumed less than their controls during FET1, thus both developed
avoidances. With respect to extinction, we could only compare consumption during the
first 4 extinction phases, as analyses for Experiment 2 only could be examined for the
first 8 extinction tests. Due to the development of the first avoidance, both groups
consumed less sucrose than their controls during the first four phases and they both
steadily increased consumption throughout extinction such that their consumption during
LET1 was greater than during ACQ1. However, while group E50 consumed less sucrose
than group Oil during LET1, group EC consumed a similar amount as group OC. Such a
result suggests that although both groups completely extinguished the first avoidance,
114

group EC might have reached a higher level of extinction. This raises the question of
whether this difference is associated with the number of extinction trials allowed once the
animals achieved their acquisition consumption level. For the E50 animals, all of which
were given a total of 19 extinction tests, 2 animals received only 1-2 tests in which they
consumed at least as much as they had consumed during the acquisition trial. However,
it is unlikely that this impacted the statistical comparison with the Oil animals. Only one
of these animals consumed less than the remaining 8 animals. In addition, those 8
animals had received an average of 8 more tests (range of 3 to 13) after achieving
acquisition consumption levels, and consequently, by the end of testing, they were
consuming substantially more than their acquisition consumption levels. Examination of
the change in consumption from ACQ1 to LET1 revealed that both groups increased by
approximately 6 ml. As such, if the amount increase in consumption from the acquisition
test to the end of extinction is indicative of amount of extinction achieved, it appears that
both groups reached comparable amounts of extinction during the CTA1 extinction
period. Thus, rather than asking why the estradiol groups differ, the question becomes
why the difference between the Oil and E50 groups was larger than the difference
between the OC and EC groups and this brings us to an examination of the oil groups.
There is a clear difference in the number of exposures to sucrose the oil groups from the
two studies received. While the OC animals received an average of 9 exposures (range of
3 to 16), the Oil animals received 19 exposures. This allowed the Oil animals to achieve
a higher level of consumption when compared to their first sucrose exposure than the OC
animals achieved. These data suggest that the difference in consumption between the
115

E50 and Oil animals during LET1 was probably a function of a higher level of
consumption in the Oil animals rather than a lower level in the E50 animals. It is worth
noting, however, that on the day preceding LET1 the E50 and Oil groups only differed by
2.9 ml, which was a nonsignificant difference. This raises the question of whether the
difference between groups Oil and E50 was a result of random variability in the
consumption patterns of the two groups and whether the consumption levels of these two
groups would have been similar once again if testing been extended another day.
Regardless of whether the difference during LET1 was due to a higher level of
consumption in the Oil animals because of their extended exposure to the sucrose
solution or simply random fluctuation in consumption, it is highly unlikely that the
difference reflects a lower level of extinction achieved by group E50.
With respect to CTA2, there were both similarities and differences when comparisons
were made between each estradiol group and its oil control. The similarities found
during CTA2 were as follows: both groups (1) consumed comparable amounts as their
controls during ACQ2, (2) consumed less than their controls during FET2, (3) developed
second avoidances whether CTA2 was based on ACQ1 or ACQ2, and (4) consumed less
than their controls during the first four phases of CTA2 extinction. The differences
between the two experiments were restricted to comparisons associated with LET2. First,
during LET2, group E50 consumed less sucrose than group Oil, while group EC
consumed comparable amounts as group OC. Second, whereas group E50 consumed
similar amounts of sucrose during LET2 as they did during ACQ1 and ACQ2, group EC
consumed more sucrose during LET2 than they did during ACQ1 and ACQ2. It is
116

unlikely that these dissimilarities in the patterns of consumption of the E50 and EC
groups is due to differences in the procedures used for determining the termination of
extinction testing during CTA2. It is the case that while all of the EC animals were
allowed to be tested until they had consumed at least 100% of acquisition consumption
levels for three days, only 5 of the 10 animals in group E50 had achieved this level of
consumption at the time extinction testing was terminated. It also is the case that it took
all of the EC animals less than 19 extinction tests to consume at least 100% of acquisition
consumption levels for three days. The mean number of days to reach this criterion was
12.8 days (range of 4-18) for the EC animals and 15.0 days (a truncated value given the
range of 3 to greater than 19) for the E50 animals. This argues against the possibility that
the differences between the E50 and EC groups were an inadvertent result of premature
termination of extinction for the E50 animals. Furthermore, it suggests that the EC
animals exhibited a faster extinction rate than the E50 animals and thus had attained a
higher level of extinction by the end of extinction testing. Two possible reasons for this
difference come to mind. The first possibility is tied to the different lengths of the
intervals imposed between the two CTAs in Experiment 1 and Experiment 2. Perhaps a
longer interval, as was used in Experiment 1, alters the relative strengths of the
conditioning and extinction memories such that conditioning memories become more
robust as the interval increases. The second possibility is that the difference was simply
the result of random variation. It has been the experience of this lab that extinction rate
of CTAs can vary across studies employing the same procedures (see Chambers and
Hayes, 2002).
117

With the exception of one similarity, even more differences between the two
experiments were found when looking at the comparisons between CTA1 and CTA2.
Similar results between the two experiments were limited to the finding that when
strength of CTA2 acquisition was based on ACQ2 consumption, both groups developed
first and second avoidances of the same strength. The first difference was the finding that
group E50 drank less sucrose during FET2 than FET1, whereas group EC consumed
more during FET2 than FET1. Second, when strength of CTA2 acquisition was based on
ACQ1 consumption, group E50 developed a stronger second avoidance, whereas group
EC developed a weaker second avoidance. Third, group E50 consumed similar amounts
of sucrose during all four phases of their first and second CTA, while group EC
consumed more sucrose during the first 2 phases of CTA2, before consuming similar
amounts during the second 2 phases. These dissimilarities consistently indicate that for
the E50 group, the strength of CTA2 was at least as strong, and possibly stronger, than
CTA1 while for the EC group the strength of CTA2 was definitely weaker than CTA1.
Given that the levels of extinction achieved by group E50 and group EC during CTA1
were similar, this difference in acquisition strength is most likely due to the different
CTA1-CTA2 interval lengths employed in each experiment. As discussed in relation to
CTA2, increasing the interval between avoidances may have resulted in greater
degradation of the extinction memory relative to the conditioning memory, leading to a
stronger second CTA. Whereas with the shorter interval, the extinction memory from
CTA1 is still robust enough to overpower the conditioning memory, thus resulting in a
weaker second avoidance.
118

Two additional differences found between the E50 and EC groups were considered in
the discussion of Experiment 2. Group E50 consumed similar amounts on ACQ1 and
ACQ2 and showed a decrease in consumption from LET1 to ACQ2, while group EC
consumed more sucrose during ACQ2 than ACQ1 and showed no change in consumption
across LET1 and ACQ2. With respect to E50, we concluded that the pattern of
consumption was indicative of a reinstatement of neophobia, and was likely a product of
the long interval used in Experiment 1. This conclusion was supported by finding that
shortening the interval between CTAs prevented the observed reinstatement in group EC.
This issue, however, becomes complicated when considering our above contention that
lengthening the interval between avoidances leads to degradation of one memory type
over another. That is, if the differences between the experiments for CTA1 versus CTA2
comparisons are a result of the longer interval experienced by E50 leading to a greater
degradation of the extinction memory than the conditioning memory, then it must be
considered that the pattern of differences in ACQ1, LET1, and ACQ2 consumption
displayed by this group is a result of the retention of the illness memory. Although this
option must be considered, as it seems to have influenced the other aspects of CTA2, it
seems an unlikely explanation for the pattern of consumption across these three tests. If
the retention of the conditioning memory was the reason for the observed decrease across
LET1 and ACQ2, then we should have found that E50 consumed less during ACQ2 than
ACQ1 (i.e. dropped below the neophobic response), not similar amounts. As such, it
might be the case that as the interval between exposures to a tastant increases, the illness
and extinction memories begin to degrade, but the extinction memory degrades more
119

rapidly than the illness memory. As a consequence, animals become more cautious of
ingesting the substance again, which leads to reinstatement of neophobia. In addition, the
second acquisition experience strengthens the remaining illness memory trace from the
first avoidance, potentiating the brain’s representation of the food’s harmful properties.
Whereas, with a short interval the extinction memory is still relatively young and is thus
robust enough to inhibit potential strengthening of such harmful representations in the
brain.
In conclusion, it seems that the different procedures for termination of extinction is an
unlikely explanation for the differences observed between the two experiments for CTA1,
CTA2 and CTA1 versus CTA2 comparisons. In order for this factor to be a plausible
explanation we should have found that group EC reached a higher level of extinction
during CTA1, leading to their more rapid extinction rate during CTA2. The possible
competition of acquisition and extinction memories during the different length intervals,
on the other hand, can explain the differences observed during CTA2 and the CTA1
versus CTA2 comparisons, as well as the reinstatement of neophobia.
3.5.2 Comparisons between Experiment 1 and Experiment 2: Retention of Avoidance
The prevailing view is that after a behavior has been conditioned and then
extinguished, the conditioning memory remains and coexists with the extinction memory
(Bouton, 1994; Delamater, 1996; Rescorla, 2001). In both experiments, it was concluded
that the first illness experience was the strongest memory retained by one group, while
for the other group retention of the extinction experience was stronger. Recall that in
Experiment 1, it was concluded that the most robust memory retained by group E50 was
120

the first illness experience and the same was true for group EP in Experiment 2. On the
other hand, the most robust memory retained by group E10 from Experiment 1 and group
EC from Experiment 2 was the first extinction experience. The results of our two
experiments suggest that the relative strengths of these two memories were altered by the
manipulation of two parameters, the length of the interval between the first and second
CTAs and the degree of extinction of the first CTA that was attained.
Earlier it was suggested that as the length of the interval between the first and second
CTA becomes longer the extinction memory degrades more rapidly than the conditioning
memory. It follows that if the conditioning memory trace is more robust than the
extinction trace at the time of the second acquisition test, CTA2 should be stronger than
CTA1, which was observed for group E50. Conversely, with a short interval, the
extinction memory trace is more robust leading to a weaker second avoidance, which was
observed for group EC.
It seems apparent that as the level of extinction achieved increases, the conditioning
memory should weaken and the extinction memory should strengthen. Consequently one
would expect the facility with which a second CTA is acquired to lessen. The results of
the two experiments are consistent with this expectation. In Experiment 1, the fact that
CTA1 was weaker for group E10 than E50 meant that E10 animals extinguished faster
and thus reached a higher level of extinction by the time extinction testing was
terminated. For E10, the second avoidance was weaker than the first avoidance and
consumption was greater during the first extinction phases of CTA2 than CTA1 while for
E50, the second avoidance was stronger than the first avoidance. In Experiment 2, the
121

group allowed complete extinction (EC) acquired a weaker second avoidance than the
group allowed only partial extinction (EP), as evidenced by higher levels of sucrose
consumption during the first extinction test and the first four extinction phases of CTA2.
Thus, when looking at the within experiment differences, it seems that the extinction
memory is more robust when a high level of extinction is attained, and conversely, the
illness memory is more robust when a lower level of extinction is attained.
It is clear that there is a positive correlation between the strength of the US and the
subsequent strength of the acquired CTA, such that as the dose of the US increases, the
strength of the CTA increases (CTA1 in Experiment 1; Nachman and Ashe, 1973; De
Beun, Jansen, Smeets, Niesing, Slangen, and Van de Poll, 1991). What remains unclear
is whether the dose of estradiol itself altered the likelihood that the retention of one
memory was more robust than the other. More specifically, is a stronger extinction
memory more likely when the dose is low (10 µg/kg) and a stronger conditioning
memory more likely when the US dose is high (50 µg/kg)? If this is the case, then even
when the interval between two CTAs and the level of extinction (as measured by
exhibiting consumption levels equivalent to control animals) are the same for animals
administered low and high doses of estradiol, then the potency of the second CTA should
be weaker than the first for the low dose animals and stronger for the high dose animals.
3.5.3 The Reacquisition of CTAs in Aged Animals
As mentioned in the introduction of this chapter, aged animals develop more potent
avoidances than younger animals to the same dose of the illness agent (Misanin,
Goodhart, Anderson, and Hinderliter, 2002), retain the memory of avoidance formation
122

longer than younger animals, (Guanowsky and Misanin 1983; Martinez and Rigter,
1983), and extinguish from a developed avoidance more slowly than young animals
(Ingram and Peacock 1980; Springer and Fraley 1981). Taking these age differences into
account we would expect to find a different pattern of results than what was observed in
the young animals from the above experiments. First, in young animals, a short interval
resulted in a second avoidance that was weaker than the first avoidance. The suggested
reason for this finding was that with a short interval the extinction memory is still robust
enough to inhibit strengthening of the illness memory upon the second acquisition.
However, since aged animals show a proclivity towards forming and maintaining food-
illness memories, it is questionable whether the extinction memory would be robust
enough to inhibit further strengthening of the conditioning memory even with a short
interval between avoidances. As such, it seems more probable that aged animals would
develop a second avoidance that is as strong or stronger than the first avoidance despite
using a short interval. Second, although young animals allowed to completely extinguish
develop weaker second avoidances, aged animals might require over-extinction in order
to show the same pattern. It seems reasonable to assume, because of their increased
propensity to develop CTAs, that the extinction memory as it exists in aged animals
might be too weak to influence acquisition of a second avoidance, consequently resulting
in a second avoidance that is as strong or stronger than the first avoidance. If this is the
case, aged animals might require prolonged exposure to the CS after reaching complete
extinction in order to make the extinction memory robust enough to inhibit further
strengthening of the conditioning memory. Third, as indicated in the section above, it is
123

unclear whether the dose of estradiol itself alters the relative strengths of CTA1 and
CTA2. First it would have to be determined in young animals whether the potency of the
second CTA is weaker than the first with a low dose (10 µg/kg) of estradiol and stronger
with a high dose (50 µg/kg) when the level of extinction of CTA1 and the interval
between the two CTAs are similar. Assuming that this is the case, we would expect aged
animals to act differently than young animals because of their apparent increased
sensitivity to drugs. Recall in the Misanin, Goodhart, Anderson, and Hinderliter (2002)
study, aged animals given a low dose of LiCl immediately after exposure to the CS
showed a strong avoidance while the avoidance exhibited by the young animals was
weak. This suggests that aged animals would also develop a stronger avoidance with a
10µg dose of estradiol than young animals. If so, we would expect aged animals also to
develop a second avoidance that is stronger than the first avoidance when reconditioned
with this low dose of estradiol while young animals would acquire a second avoidance
that is weaker than the first avoidance.

124

REFERENCES
Aguado, L., De Brugada, I, and Hall, G. (2001). Tests for inhibition after extinction of a
conditioned stimulus in the flavour aversion procedure. Q J Exp Psychol, 54, 201-217.

Arbel, I., Kadar, T., Silberman, M., and Levy, A. (1994). The effects of long-term
corticosterone administration on hippocampal morphology and cognitive performance of
middle-aged rats. Brain Res, 657, 227-235.

Baiardi, G., Macova, M., Armando, I., Ando, H., Tyurmin, D., and Saavedra, J.M.
(2005). Estrogen upregulates renal angiotensin II AT
1
and AT
2
receptors in the rat. Regul
Pept, 124, 7-17.

Banerjee, U. and Das, P. (1980). Conditioned taste aversion by lithium and saline: a
methodological evaluation. Chem Senses, 5, 175-183.

Barnes, C.A. (1979). Memory deficits associated with senescence: A neurophysiological
behavioral study in the rats. J Comp Physiol Psychol, 93, 74-104.

Barnes, C.A., Nadel, L., and Honig, W.K. (1980). Spatial memory deficit in senescent
rats. Can J Psychol, 34, 29-39.

Barnes, C.A., and McNaughton, B.L. (1985). An age comparison of the rates of
acquisition and forgetting of spatial information in relation to long term enhancement of
hippocampal synapses. Behav Neurosci, 99, 1040-1048.

Behl, C., Widmann, M., Trapp, T., and Holsboer, F. (1998). 17ß-Estradiol protects
neurons from oxidative stress-induced cell death in vitro. Biochem Biophys Res Commun,
216, 473–482.

Britton, G.W., Rotenberg, S., Adelman, R.C. (1975). Impaired regulation of
corticosterone levels during fasting in aging rats. Biochem Biophys Res Commun, 64,
184-188.

Bouton, M. (1993). Context, time, and memory retrieval in the interference paradigms of
Pavlovian learning. Psychol Bull, 114, 80-99.

Bouton, M. (1994). Conditioning, remembering, and forgetting. Journal of Experimental
Psychology: Anim Behav Process, 20, 219-231.

Buresova, O. and Bures, J. (1980). Post-ingestion interference with brain function
prevents attenuation of neophobia in rats. Behav Brain Res, 1, 299-312.

125

Chambers, K.C., Sengstake, C.B., Yoder, R.L., and Thornton, J.E. (1981). Sexually
dimorphic acquisition of a conditioned taste aversion in rats: Effects of gonadectomy,
testosterone replacement, and water deprivation. Physiol Behav, 27, 83-88.

Chambers, K.C. and Hayes, U.L. (2002). Exposure to estradiol before but not during
acquisition of LiCl-induced conditioned taste avoidance accelerates extinction. Horm
Behav, 41, 297-305.

Coirini, H., Johnson, A.E., McEwen, B.S. (1989). Estradiol modulation of oxytocin
binding in the ventromedial hypothalamic nucleus of male and female rats.
Neuroendocrinology, 50, 193-198.

Dean, R.L., Scozzafava, J., Goas, J.A., Regan, B., Beer, B, and Bartus, R.T. (1981). Age-
related differences in behavior across the life span of the C57BL/6J mouse. Exp Aging
Res, 7, 427-451.

De Beun, R., Jansen, E., Smeets, M.A.M, Niesing, J., Slangen, J.L., and Van de Poll,
N.E. (1991). Estradiol-induced conditioned taste aversion and place aversion in rats: Sex-
and dose-dependent effects. Physiol Behav, 50, 995-1000.

De Beun, R., Peeters, W.M.M, and Broekkamp, C.L.E. (1993). Stimulus characterization
of estradiol applying a crossfamiliarization taste aversion procedure in female mice.
Physiol Behav, 53, 715-719.

De La Casa, G. and Lubow, R.E. (1995). Latent inhibition in conditioned taste aversion:
the roles of stimulus frequency and duration and the amount of fluid ingested during
preexposure. Neurobiol Learn Mem, 64, 125-132.

Delamater, A.R. (1996). Effects of several extinction treatments upon the integrity of
Pavlovian stimulus-outcome associations. Anim Learn Behav, 24, 437-449.

De Vries, G.J., Wang, Z., Bullock, N.A., and Numan, S. (1994). Sex Differences in the
Effects of Testosterone and Its Metabolites onVasopressin Messenger RNA Levels in the
Bed Nucleus of the Stria Terminalis of Rats. J Neurosci, 14, 1789-1794.

Diaz-Brinton, R., Tran, J., Proffitt, P., and Montoya, M. (1997). 17ß-Estradiol enhances
the outgrowth and survival of neocortical neurons in culture. Neurochem Res, 22, 1339–
1351.

Eldridge, J., Brodish, A., Kute, T., and Landfield, P. (1989). Apparent age-related
resistance to type II hippocampal corticosteroid receptors to down-regulation during
chronic escape training. J Neurosci, 9, 3237-3242.

126

Finch, C.E. (1978). Reproductive senescence in rodents: Factors in the decline of fertility
and loss of regular estrous cycles. In: The aging and reproductive systems (Volume 4).
Schneider, E.L. ed. Raven Press, New York.

Fink, G., Sumner, B.E.H., Rosie, R., Grace, O., and Quin, J.P. (1996). Estrogen control of
central neurotransmission: Effects on mood, mental state, and memory. Cell Mol
Neurobiol, 16, 325‐344.

Freeman, K.B. and Riley, A.L. (2006). The conditioned taste aversion preparation:
practical applications and technical considerations. Tasks and techniques: a sampling of
the methodologies for the investigation of animal learning, behavior and cognition.
Chapter 11.

Gallagher, M. and Pellymounter, M.A. (1988). Spatial learning deficits in old rats: A
model for memory decline in the aged. Neurobiol Aging, 9, 549-556.

Ganesan, R. and Simpkins, J. (1991). Conditioned taste aversion induced by estradiol
pellets. Physiol Behav, 50, 849-852.

Guanowsky, V., Misanin, J. R., and Riccio, D.C. (1983). Retention of conditioned taste
aversion in weanling, adult, and old-age rats. Behav Neural Biol, 37, 173-178.

Gustavson, C.R., Gustavson, J.C., Young, J.K., Pumariega, A.J., and Nicolaus, L.K.
(1989). Estrogen induced malaise; pgs 501-523. In: Neural control of reproductive
function. Alan R. Liss Inc., New York.

Hart, J.A., Bourne, M.J., and Schachtman, T.R. (1995). Slow reacquisition of a
conditioned taste aversion. Anim Learn Behav, 23, 297-303.

Heth, C.D. and Pierce, W.D. (2007). The role of pre-exposure to novel food tastes in
activity-based conditioned taste avoidance. Learn Motiv, 38, 35-43.

Hintiryan, H., Hayes, U.L., and Chambers, K.C. (2005). The role of histamine in
estradiol-induced conditioned consumption reductions. Physiol Behav, 84, 117-128.

Hintiryan, H., Hayes, U.L., and Chambers, K.C. (2006). Intraoral cheek fistulae: a refined
technique. Lab Anim, 40, 456-464.

Hintiryan, H., Foster, N.N., and Chambers, K.C. (In Press) Dissociating the conditioning
and the anorectic effects of estradiol in female rats. Behav Neurosci.

Ingram, D.K. and Peacock, L.J. (1980). Conditioned Taste Aversion as a Function of Age
in Mature Male Rats. Exp Aging Res. 6, 113-123.

127

Issa, A., Rowe, W., Gauthier, S., and Meaney, M. (1990). Hypothalamic-pituitary-adrenal
activity in aged, cognitively impaired and cognitively unimpaired rats. J Neurosci, 10,
3247-3254.

Kalat, J.W. and Rozin, P. (1971). Role of interference in taste-aversion learning. J Comp
Physiol Psychol, 77, 53-58.

Kersten, L. (1981). Toxicity and renal elimination of lithium in rats of different ages.
Arch Toxicol, 47, 135-144.

Kersten, L. (1979). Zur pharmakokinetik von lithium bei ratten verschiedenen alters. Acta
Biol Med Germ, 38, 1331-1345. Cited in: Kersten, L. (1981). Toxicity and renal
elimination of lithium in rats of different ages. Arch Toxicol, 47, 135-144.

Kubanis, P. and Zornetzer, S. (1981). Age-related behavioral and neurobiological
changes: A review with an emphasis on memory. Behav Neural Biol, 31, 115-172.

Levine, S., Saltzman, A., Katof, B., Meister, A., and Cooper, T.B. (1997). Influence of
body size on lithium levels. Psychopharmacology, 129, 23-26.

Lewis, C., McEwen, B.S., and Frankfurt, M. (1995). Estrogen-induction of dendritic
spines in ventromedial hypothalamus and hippocampus: effects of neonatal aromatase
blockade and adult castration. Dev Brain Res, 87, 91–95.

Lu, K.H., Hopper, B.R., Vargo, T.M., and Yen, S.S.C (1979). Chronological changes in
sex steroid, gonadotropin and prolactin secretion in aging female rats displays different
reproductive states. Biology of Reproduction, 21, 193-203.

Luine, V.N. (1985). Estradiol increases choline acetyltransferase activity in specific basal
forebrain nuclei and projection areas of female rats. Exp Neurol, 89, 484–490.

Martinez, J.L. and Rigter, H. (1983). Assessment of retention capabilities in old rats.
Behav Neural Biol, 39, 181-191.

Matt, D.W., Sarver, P.L., and Lu, J.K.H. (1987). Relation of parity and estrous cyclicity
to the biology of pregnancy in aging female rats. Biol Reprod, 37, 421-430.

McEwen, B.S. and Alves, S.E. (1999). Estrogen actions in the central nervous system.
Endocr Rev, 20, 279-307.

McNamara, M.C., Benignus, V.A., Benignus, G., and Miller, A.T. (1977). Active and
passive avoidance in rats as a function of age. Exp Aging Res, 3, 3-16.

128

Melia, K.R., Ryabinin, A.E., Schroeder, R., Bloom, F.E. and Wilson, M.C. (1994).
Induction and habituation of immediate early gene expression in rat brain by acute and
repeated restraint stress. J. Neurosci. 14, 5929–5938. Cited In: Senba, E., and Ueyama,
T. (1997). Stress-induced expression of immediate early genes in the brain and peripheral
organs of the rat. Neurosci Res. 29, 183-207

Misanin, J.R., Guanowsky, V., and Riccio, D.C. (1983). The effect of CS-preexposure on
conditioned taste aversion in young and adult rats. Physiol Behav, 30, 859-862.

Misanin J.R., Greider D.L., and Hinderliter C.F (1988). Age differences in the outcome
of long-delay taste-aversion conditioning in rats. Bull Psychon Soc. 26, 258-260.

Misanin J.R., Hoffel T.D., Reidy C.A., and Hinderliter C.F (1997). Remote and proximal
US preexposure and aging effects in taste aversion learning in rats. Physiol Behav. 61,
221-224.

Misanin J.R., Collins M., Rushanan S., Anderson M.J., Goodhart M., and Hinderliter C.F
(2002). Aging facilitates long-trace taste-aversion conditioning in rats. Physiol Behav.
75, 759-764.

Misanin J.R., Goodhart M.G., Anderson M.J., and Hinderliter C.F. (2002). The
interaction of age and unconditioned stimulus intensity on long-trace conditioned flavor
aversion in rats. Dev Psychobiol. 40, 131-137.

Nachman, M. and Ashe, J.H. (1973). Learned taste aversions in rats as a function of
dosage, concentration, and route of administration of LiCl. Physiol Behav, 10, 73-78.

Nakra, B.R.S. and Grosserg, G.T. (1987). Lithium use in the elderly. J Geriatr Drug
Ther, 2, 47-63.

Napier, R.M., Macrae, M., and Kehoe, E.J. (1992). Rapid reacquisition in conditioning of
the rabbit’s nictitating membrane response. J Exp Psychol: Anim Behav Process, 18, 182-
192.

Parker, L.A. (1995). Rewarding drugs produce taste avoidance, but not taste aversion.
Neurosci Biobehav Rev, 9, 143-151.

Pavlov, I.P. (1927). Conditioned Reflexes. Oxford, England: Oxford University Press.

Pedersen W.A., Wan R., and Mattson M.P. (2001). Impact of aging on stress-responsive
neuroendocrine systems. Mech Ageing Dev. 122, 963-983.

Porsolt, P.E., Barrett, G.V., Sterns, H.L. (1977). Depression: a new animal model
sensitive to antidepressant treatments. Nature, 266, 730-732.
129

Rabin, R.M., Hunt, W.A., and Lee, J. (1989). Attenuation and cross-attentuation in taste
aversion learning in the rat: Studies with ionizing radiation, lithium chloride, and ethanol.
Pharmacol Biochem Behav, 31, 909-918.

Radomski, J.L, Fuyat, H.N., Nelson, A.A., and Smith, P.K. (1950). The toxic effects,
excretion and distribution of lithium chloride. J Pharmacol Exp Ther, 100, 429-444.

Rapp, P., Rosenberg, R., and Gallagher, M. (1987). An evaluation of spatial information
processing in aged rats. Behav Neurosci, 101, 3-12.

Rescorla, R.A. (2001). Retraining of extinguished Pavlovian stimuli. J Exp Psychol:
Anim Behav Process, 27, 115-124.

Reilly, S. and Trifunovic, R. (2001). Lateral parabrachial nucleus lesions in the rat:
neophobia and conditioned taste aversion. Brain Res Bull, 55, 359-366.

Riegle, G.D., and Hess, G.D. (1972). Chronic and acute dexamethasone suppression of
stress activation of the adrenal cortex in young and aged rats. Neuroendocrinology, 9,
175-187.

Riegle, G.D. (1973). Chronic stress effects on adrenocortical responsiveness in young
and aged rats. Neuroendocrinology, 11, 1-10.

Rigter, H., Martinez, J., and Crabbe, J. (1980). Forgetting and other behavioral
manifestations of aging. In: D. Stein, Editor, The Psychobiology of Aging: Problems and
Perspectives, Elsevier/North Holland, New York (1980), pp. 161–175.

Rodriguez, G. and Alonso, G. (2002). Latent inhibition as a function of CS intensity in
taste aversion learning. Behav Processes, 60, 61-67.

Roth, G.S. (1974). Reduced glucocorticoid binding site concentration in cortical neuronal
perikarya from senescent rats. Brain Res, 107, 345-354.

Sapolsy, R.M., Krey, LC., and McEwen, B.S. (1986). The neuroendocrinology of stress
and aging: The glucocorticoid cascade hypothesis. Endocr Rev, 7, 284-301.

Schou, M. (1957). Biology and pharmacology of the lithium ion. Pharmacol Rev, 9, 17-
58.

Senba, E., and Ueyama, T. (1997). Stress-induced expression of immediate early genes
in the brain and peripheral organs of the rat. Neurosci Res. 29, 183-207.

130

Simerly, R.B., Chang, C., Muramatsu, M., and Swanson, L.W. (1990). Distribution of
androgen and estrogen receptor mRNA-containing cells in the rat brain: An in situ
hybridization study. J Comp Neurol, 294, 76-95.

Simonian, S.X. and Herbison, A.E. (1997). Differential expression of estrogen receptor
and neuropeptide Y by brainstem A1 and A2 noradrenaline neurons. Neuroscience, 76,
517-529.

Singer, C.A., Rogers, K.L., Stickland, T.M., and Dorsa, D.M. (1996). Estrogen protects
primary cortical neurons from glutamate toxicity. Neurosci Lett, 212, 13–16.

Singh, M., Meyer, E.M., Millard, W.J., Simpkins, and J.W. (1994). Ovarian steroid
deprivation results in a reversible learning impairment and compromised cholinergic
function in female Sprague-Dawley rats. Brain Res, 644, 305–312.

Springer, A.D. and Fraley, S.M. (1981). Extinction of a Conditioned Taste Aversion in
Young, Mid-Aged, and Aged C57/BL6 Mice. Behav Neural Biol. 32, 282-294.

Yune, T.Y, Kim, S.J., Lee, S.M., Lee, Y.K., Oh, Y.J., Kim, Y.C., Markelonis, G.J., and
Oh, T.H. (2007). Systemic administration of 17β-Estradiol reduces apoptotic cell death
and improves functional recovery following traumatic spinal cord injury in rats. J
Neurotrauma, 21, 293-306.

Valliere W.A., Misanin, J.R., & Hinderliter C.F. (1988). Age-related differences in the
US preexposure effect on conditioned taste aversion in rats. Bull Psychon Soc. 26, 64-66.

Vasquez, B.J., Martinez, J.L, Jensen, R.A., Messing, R.B., Rigter, H., and McGaugh, J.L.
(1983). Learning and memory in young and aged Fischer 344 rats. Arc. Gerontol.
Geriatr., 2, 279-291.

Weidemann, G. and Kehoe, E.J. (2003). Savings in classical conditioning in the rabbit as
a function of extended extinction. Learn and Behav, 31, 49-68.

Weiss, C. and Thompson, R. (1991). The effects of age on eyeblink conditioning in the
freely moving Fischer 344 rat. Neurobiol Aging, 12, 249-254.

Wilcox, R.R. (2003a). Fundamentals of modern statistical methods: substantially
improving power and accuracy. Singer-Verlag, New York.

Wilcox, R.R. (2003b). Applying contemporary statistical techniques. Academic Press,
San Diego.

Wilcox, R.R., and Keselman, H.J. (2003). Modern robust data analysis methods:
measures of central tendency. Psychol Methods 8, 254–274.

131

Wu, PC (2002). Central limit theorem comparing means, trimmed means, one-step M-
estimators and modified one-step M-estimators under non-normality. Unpublished
Dissertation, Los Angeles: University of Southern California.

Yuan, D.L. and Chambers, K.C. (1999). Estradiol accelerates extinction of lithium
chloride-induced conditioned taste aversions through its illness-associated properties.
Horm Behav, 36, 287-298.

Zornetzer, S., Thompson, R., and Rogers, J. (1982). Rapid forgetting aged rats. Behav
Neural Biol, 36, 49-60.

Zurkovsky, L., Brown, S.L., Boyd, S., Fell, J.A., and Korol, D.L. (2007). Estrogen
modulates learning in female rats by acting directly at distinct memory systems.
Neuroscience, 144, 26-37.

Abstract (if available)

Abstract Both the human and animal literature report observations of marked age-related declines in cognitive functioning. The existing evidence indicates that the larger deficit involves the retention of newly learned behaviors, rather than the acquisition of those behaviors. However, one learning and memory paradigm, conditioned taste avoidances, suggests that aged animals show improvement in both learning and memory. Rather than displaying the usual decline in acquiring and maintaining a new behavior, aged animals show more robust learning and retention of these food-illness associations than younger cohorts. Following is a review of the age-related changes observed in the more traditional learning and memory paradigms, the age-related changes observed in the conditioned taste avoidance paradigm, hypotheses to explain the observed differences in both types of paradigms, and two preliminary experiments examining the impact of extinction on the reacquisition of a conditioned taste avoidance.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

The amygdala and conditioned taste aversion

PDF

Role of oxidative stress in age-associated mild cognitive impairment and Alzheimer's disease

PDF

The neural basis of estradiol conditioned taste aversion learning and estradiol anorexia

PDF

The development of electrodermal fear conditioning from ages 3 to 8 years and relationships with antisocial behavior at age 8 years

PDF

Assessing therapeutic change mechanisms in motivational interviewing using the articulated thoughts in simulated situations paradigm

PDF

Cancer of the psyche: antenatal maternal stress accentuates functional brain responses and neuronal endangerment during conditioned fear in adult offspring

PDF

Neuropsychological functioning in older adult long-term cancer survivors

PDF

Aging and emotion regulation in the judgment of facial emotion

PDF

Age-related androgen depletion and the development of Alzheimer's disease

PDF

Biomarkers of age-related health changes: associations with health outcomes and disparities

PDF

Value-based decision-making in complex choice: brain regions involved and implications of age

PDF

Self-perceptions of Aging in the Context of Neighborhood and Their Interplay in Late-life Cognitive Health

PDF

The impact of reduced Src kinase activity on memory-related oscillatory activity in adulthood and memory function in age in a mouse model of schizophrenia

PDF

Experience-dependent neuroplasticity of the dorsal striatum and prefrontal cortex in the MPTP-lesioned mouse model of Parkinson’s disease

Asset Metadata

Creator Ycaza, Alexandra E. (author)

Core Title Age-related changes in cognitive functioning and their implications for the reacquisition of conditioned taste avoidances

School College of Letters, Arts and Sciences

Degree Master of Arts

Degree Program Psychology

Publication Date 10/27/2009

Defense Date 03/23/2009

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag aging,cognitive deficits,conditioned taste avoidances,Learning and Instruction,memory,OAI-PMH Harvest

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Chambers, Kathleen C. (committee chair), Lavond, David G. (committee member), Walsh, John P. (committee member)

Creator Email panda819@me.com,ycaza@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m2695

Unique identifier UC1234366

Identifier etd-Ycaza-3337 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-268249 (legacy record id),usctheses-m2695 (legacy record id)

Legacy Identifier etd-Ycaza-3337.pdf

Dmrecord 268249

Document Type Thesis

Rights Ycaza, Alexandra E.

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu