Arousal and memory

MISSING IMAGE

Material Information

Title:
Arousal and memory effects of aging
Alternate title:
Effects of aging
Physical Description:
vi, 106 leaves : ill. ; 29 cm.
Language:
English
Creator:
Schramke, Carol Joann, 1959-
Publication Date:

Subjects

Subjects / Keywords:
Research   ( mesh )
Mental Recall -- Aged   ( mesh )
Mental Recall -- Adult   ( mesh )
Aging   ( mesh )
Memory, Short-term   ( mesh )
Arousal -- physiology   ( mesh )
Psychophysiology   ( mesh )
Department of Clinical and Health Psychology thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Health Related Professions -- Department of Clinical and Health Psychology -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 1990.
Bibliography:
Bibliography: leaves 97-105.
Statement of Responsibility:
by Carol J. Schramke.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001581152
oclc - 24641677
notis - AHK5059
System ID:
AA00009076:00001


This item is only available as the following downloads:


Full Text














AROUSAL AND MEMORY: EFFECTS OF AGING


By

CAROL J. SCHRAMKE










A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


1990
















ACKNOWLE DGEMENTS


This dissertation is dedicated to my maternal

grandmother, Frances Gendregske Frost, who cannot understand

why my studies are so important to me, but loves me

unconditionally nevertheless. Many thanks must go to my

husband, Carson Lane, who has moved from Michigan to Florida

and then to Oregon with few complaints and who encouraged me

and believed in me whenever I needed his support. I am also

indebted to the many older adults who were enthusiastic

about my project and my interest in aging and who braved the

horrible parking situation at Shands in order to participate

in my research. Finally, I am grateful to Beverly

Funderburk, who was my legs when I was 3,000 miles away,

making the completion of this endeavor possible.

I am also wish to acknowledge my committee members Walt

Cunningham, Jaber Gubruim, Kenneth Heilman, Michael

Robinson, Robin West, and especially Russell Bauer, my

chair, for their patience and their assistance in making

this a better research project. This research was funded,

in part, by NIMH National Research Fellowship Award Number

5F31MH09640-02.
















TABLE OF CONTENTS

Page
ACKNOWLEDGEMENTS........................................ iv

ABSTRACT....................... ...... ......... ....

BACKGROUND. ........................ ....... 1

Introduction........... ... .... .... ...... 1
Aging and Memory................................... 4
Aging and Arousal........... ..... ....... .......... 10
Arousal and Memory: Interactions................... 13
Changes in Brain Physiology with Aging............. 18
Neuropsychological Mechanisms in Aging and Memory.. 20
Aging, Arousal, and Memory ......................... 28

EXPERIMENT 1: METHODS................................... 31

Induced Amnesia for Pictures Before and After a
Critical Event: Overview and Rationale........... 31
Subjects... .... .......... o ............... ... 32
Materials ...... ........ o ................. .... 33
Procedure .... ....... .............. ....... 35

EXPERIMENT 1: RESULTS ................................ 38

Psychometric Test Performance.............. .. 38
List Effects................ .. ........... .... 42
Overall Recall and Recognition Performance........ 43
Effect of Item Position and the
Critical Event on Memory........................ 44
Response Types......... ..... ................. ....... 52
Electrodermal Recognition of Recognized
and Unrecognized Targets............... 55

EXPERIMENT 1: DISCUSSION ............................... 58

EXPERIMENT 2: METHODS ........ ......................... 65

The Interaction of Age and Increased Tonic Arousal
on a Supraspan List Learning Task: Overview and
Rationale ......... ............. ........ 65
Subjects ........ ....................... .. ... 67


iii









Experimental Conditions............................ 67
Materials and Apparatus.................... ...... 67
Procedure .......................................... 69

EXPERIMENT 2: RESULTS................................... 71

Psychometric Test Performance...................... 71
Free Recall, Cued Recall, and Recognition.......... 72
Effects of Exercise, Rest, and Distractor Tasks
on Measures of Arousal .......................... 74
Exercise Condition and Memory ...................... 78

EXPERIMENT 2: DISCUSSION....................................... 81

GENERAL DISCUSSION.................................... 84

APPENDICES
A Screening Questionnaire........................... 93

B Modified California Verbal Learning Test........ 95

REFERENCES..................... ............... ..... 97

BIOGRAPHICAL SKETCH.................................... 106















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy


AROUSAL AND MEMORY: EFFECTS OF AGING

By

Carol J. Schramke

August 1990

Chairman: Russell M. Bauer, Ph.D.
Major Department: Clinical and Health Psychology



These studies examine the influence of phasic and tonic

arousal on memory performance in younger and older adults.

Subjects in Experiment One were asked to learn a list of 25

simple line drawings during Session One and 24 drawings with

a unique item (i.e., the critical event) in the center

position (photograph of either a normal or starving child)

during Session Two. Previous research has demonstrated that

younger adults' memory for a critical event increases, while

memory for items surrounding the critical event is

disrupted, presumably because of phasic arousal elicited by

the distinctive item. This phenomena had not been

replicated with older adults. In this study, while both

younger and older adults demonstrated increased recall of

v









the critical event, only younger adults' recall for

surrounding stimuli was disrupted. Younger adults also

demonstrated greater disruption in memory when the

photograph was subjectively more disturbing (i.e., emaciated

rather than normal child), while older adults showed similar

memory patterns for surrounding items regardless of the

nature of the center item.

In the second study, state dependent learning was

examined by manipulating physiologic arousal. Subjects

either rested or exercised immediately prior to learning a

modified version of the California Verbal Learning Test, and

then engaged in either the same or the alternate activity

immediately prior to delayed recall. Both younger and older

adults showed clear state dependent learning effects, as

well as increased semantic clustering, when in the same

state at both acquisition and recall. Age by type of memory

test interactions were found in both studies, with age

differences on recall but not recognition tasks. No age by

arousal state interactions were suggested; young and old

were equally affected by state dependent learning, and

neither group was affected by overall level of arousal

during the learn or recall phase. Brain areas implicated as

important in explaining the age differences and similarities

found include the amygdala, frontal lobes, and hippocampus.

vi














BACKGROUND

Introduction

Neuropsychology has traditionally studied brain changes

resulting from brain injury or diseases of the brain and

attempted to provide guidelines for separating normal

variation in performance from variations induced by brain

pathology. In studying aging, neuropsychologists can

contribute by clarifying the differences and similarities in

performance between young and old and by speculating as to

what brain changes these behavioral findings suggest.

Normal aging is associated with changes in both the brain

and behavior; exploring these changes offers an opportunity

for understanding how the brain influences behavior.

This paper will focus on aging and how it influences

memory and arousal. This will be accomplished by describing

the behavioral changes that have been associated with aging

in memory and arousal, how memory and arousal are believed

to interact in normal young adults, which brain areas are

believed to be most important for memory and arousal, and

reviewing why the brain changes associated with aging may

alter this interaction. Two studies that examine how memory

performance in young and old is influenced by manipulations

1









in arousal are then described. Finally, the implications of

these studies for understanding the neuropsychology of

memory and aging will be discussed.

Arousal has been conceptualized, and will be discussed

herein, as an index of attention or processing and as an

index of the state of an individual. Arousal, in the

neurological literature, is often associated with cortical

desynchrony that can be measured by electroencephalography.

This activation may be linked with an alteration in

affective state. Although it has been suggested that

physical arousal and an alteration in cognitive state are

necessary for the normal experience of emotion, an

alteration in arousal is not believed to be sufficient to

induce an emotional state (Schacter & Singer, 1962, Heilman,

Watson, & Bowers, 1983).

In the psychological literature arousal has been

conceptualized in a number of different ways. For example,

Eysenck (1976) defines arousal as an elevation in body

function that is both nonspecific and noninformational. He

acknowledges that conceptually arousal can be separated into

electrocortical, autonomic, and behavioral arousal, while

arguing that it can also be treated as a unitary phenomenon

and has been so treated historically. Psychological studies

of arousal have relied on indicators such as changes in

heart rate, skin conductance response, and blood pressure to

verify alterations in activation; no studies to date have

examined what effects this alteration in physical arousal









has on cortical activity, except through task performance.

The physical activation resulting from exercise has been

shown to have effects on performance of many cognitive tasks

(Tomporowski & Ellis, 1986), although there is no

verification in these studies that electrocortical activity

is systematically altered.

Although increasing or decreasing different types of

arousal may produce similar performance effects, there is

sufficient evidence to argue against treating arousal as a

unitary concept. At a minimum arousal can be divided into

tonic versus phasic arousal and electrocortical versus

autonomic arousal. Electrocortical activation may also be

different if it is associated with processing rather than

orienting to a shocking or novel stimulus or if it reflects

an emotional response to a stimulus or situation.

Although these different types of arousal and how they

effect memory have been studied in young adults, patterns of

tonic and phasic arousal may change with age and could

influence memory in older adults differently than they do in

young adults. The age-related neurophysiologic changes

responsible for these alterations in arousal still need to

be determined, but also may be clarified through the study

of the behavioral differences between younger and older

adults.

Memory processes are known to change with age, and

memory studies have come to dominate psychological research

on aging (Poon, 1985). A review of the aging and memory









literature finds many theories about the nature of this

decline. Information-processing models, which examine

specific stages or processes necessary for memory, have been

used to examine the changes in memory of the elderly as well

as brain-injured populations. More general cognitive

functions, such as speed and one's ability to attend, are

also necessary for an individual engaging in the various

stages of information processing, and changes in these

functions may influence memory performance. Both processing

changes and alterations in attention and arousal have been

suggested as possible mechanisms to explain age-related

memory change.

Aging and Memory

Age-related alterations in memory include older adults'

general slowing, poorer performance on supraspan memory

tasks, and much worse performance on tasks requiring recall

rather than recognition when compared to younger adults

(Poon, 1985). Age differences in cross-sectional studies

can be minimized by instructing the elderly on techniques

that can improve memory (e.g., imagery) and by providing

organization of to-be-remembered material (Poon, 1985).

Most models and theories about these changes invoke

psychological or biological factors.

Psychological factors that have been suggested as

possible contributors to this decline in memory performance

include depression, anxiety, task familiarity, and levels of

motivation (Poon, 1985). An increased incidence of









depression among the elderly may interfere with performance

on memory tests or be related to increased memory complaints

(e.g., Niederehe & Camp, 1985; West, Boatwright, & Schleser,

1984). It also has been suggested that the elderly are less

proximally familiar with laboratory procedures or other new

learning tasks. Because of this they may analyze the tasks

differently and use a less efficient approach; or they may

view the tasks as more threatening, which could increase

their anxiety. Either of these factors could negatively

influence performance (Langer, Rodin, Beck, Weinman, &

Spitzer, 1979; Perlmutter, 1978). Although these factors

may contribute to poorer performance of the elderly on

laboratory tests of memory, Burke and Light (1981) argue

that these factors should affect all memory tasks

equivalently and they are not sufficient explanations to

account for the relatively select age differences found with

some memory tasks but not others.

In information-processing models, memory is typically

conceptualized as occurring in three stages: acquisition,

retention, and retrieval of information. Attempts to

isolate the memory difficulties associated with aging have

examined all three levels of this model. Because of the

difficultly in isolating each facet of the model, deficits

can often be explained by invoking more than one stage of

processing.

Evidence for acquisition deficits comes from studies

that show that older subjects do not spontaneously use









mnemonic strategies. These studies have shown that the

performance of older adults can improve, and age differences

can be minimized, when older adults are specifically

instructed to use these strategies (Rankin & Collins, 1985;

Thomas & Ruben, 1973 reported in Poon, 1984). For example,

in the Thomas and Ruben (1973) study, visual imagery

instructions, cartoon mnemonics (in which unusual

relationships between the stimulus and cue were depicted),

and a no-instruction condition were compared in a paired-

associate memory task. They found that the relatively large

differences between young and old seen under the no-

instruction condition were greatly reduced when instructions

were provided. This suggests that older subjects benefit

more than the young when given instructions or organization

to assist in remembering.

Hess (1984) conducted two studies to examine

differences that result from manipulations at the

acquisition phase. In the first study, he presented

semantically related word pairs during the acquisition phase

(e.g., copper-iron) and during the test phase presented

either the same pair, a pair that suggested a new

interpretation of the test word (e.g., clothes-iron), or a

new context word that was similar to the original pair

(e.g., bronze-iron) in a yes/no recognition paradigm. In

the second study, word pairs were also presented but they

were all semantically unrelated. Hess found that older

adults used context information primarily when it activated









old learning (i.e., they did not recognize the target words

when the words were in the new context condition as well as

the younger subjects). Older subjects also had more false

positive errors for cues from the old context, making more

semantic confusion errors. This may reflect more general

and less distinctive encoding.

Results of Rabinowitz, Craik, and Ackerman's (1982)

study are also consistent with the hypothesis of less

distinctive encoding in the elderly. In this study, recall

for words to which subjects had previously generated their

own specific associates were compared to performance when

the associate or a general category cue was given during

retrieval. Older adults did not benefit as much from the

specific retrieval cues when compared with their younger

cohorts, but did equally well when provided with the more

generalized cues.

There is little evidence to support age-related

decreases in retention ability. Generally, researchers

interested in this phenomenon look for increased

interference, either proactive or retroactive, or an

increased rate of forgetting (Poon, 1985). These problems

have not been reliably demonstrated in the elderly (e.g.,

Poon and Fozard, 1980; Craik, 1977).

Proponents of a retrieval deficit theory often cite the

differences between performance on recall and recognition

tasks as supportive evidence. Young subjects do much better

on recall tests, while on recognition tasks older subjects









improve dramatically and the age difference in performance

levels is greatly minimized (Schonfield, 1965, Schonfield

and Robertson, 1966; Erber, 1974). It has been argued that

the older adults' ability to recognize information suggests

that the information is stored, so the problem must

therefore be one of retrieval. Poon (1985) contends it is

nearly impossible to separate retrieval problems from

acquisition difficulties, and differences between free

recall and recognition could also be explained by

inefficient encoding that would subsequently interfere with

retrieval. Poon (1985) notes that age differences can be

minimized if category cues are given at acquisition, but

these cues are not as effective when only given at the

retrieval stage. This finding leads him to argue that,

although retrieval deficits may be present, they are

probably not sufficient to account totally for the age

differences seen.

Other researchers have suggested that an age-related

slowing in performance that accompanies aging can account

for many changes in cognitive functioning, including changes

in performance on many memory tasks (e.g., Birren, 1974;

Salthouse, 1980). For example, it is noted that age

differences typically increase when time limitations are

given and are minimized when unlimited time is allowed

(Poon, 1985). Age-related slowing is noted to affect each

stage of information processing and could be responsible for

the differences seen between young and old at both









acquisition and retrieval (Salthouse and Somberg, 1982).

Supporters of this explanation argue that this provides one

of the most parsimonious explanations for the alterations

seen with aging, is a biologically based explanation that

should not be influenced by cohort effects, and can explain

why more complex cognitive tasks show more pronounced age

differences (Salthouse, 1985). However, although few would

argue that slowing does not influence memory performance, it

probably is not a sufficient explanation for all changes

seen in memory. Although memory differences between young

and old are typically increased when time limitations are

given, differences are maintained in many tasks even with

unlimited time. Speed explanations also have difficulty

accounting for some of the variable differences in different

types of memory tasks (Poon, 1985).

Although current findings are not conclusive, some

research suggests that part of the difference in memory

performance between young and old may be due to a difference

in response bias or response tendencies, with older adults

being more conservative. This response bias has been

demonstrated on both recall and recognition tasks, with

older adults being more likely to make errors of omission on

recall tasks and more likely to have false negative rather

than false positive errors on recognition tasks (Botwinick,

1984). Studies using signal detection analysis for

recognition of sensory stimuli have not consistently

supported the idea that older adults have a different









response bias (Botwinick, 1984). Memory studies, such as

Poon and Fozard's (1980) study on word recognition, have

provided some support for suggestions that older adults are

more cautious when responding in memory tasks, with older

adults in their study exhibiting a lower hit rate and

response sensitivity. The inconsistency of the research has

led some to argue that cautiousness cannot fully explain the

differences in memory performance, but differences in

response tendencies may contribute to differences in some

cases (Burke and Light, 1981).

Information processing and biological explanations for

the differences in memory performance between young and old

should not be viewed as separate from psychological and

cognitive theories. Some researchers have focused on the

patterns of cell loss to explain these deficits. This has

resulted in comparisons between normal aging and dementias,

as well as comparisons between normal aging and more focal

types of brain damage.

Aqing and Arousal

Patterns of physiologic arousal are known to change

with increasing age. Woodruff (1985) and Marsh and Thompson

(1977) provide the most recent reviews of this literature.

A large number of studies have examined EEG changes and

aging. Woodruff (1985) summarizes the major findings from

these studies including changes in frequency and abundance

of alpha rhythm, changes in the incidence of beta activity,










diffuse slowing, focal slowing and abnormal activity in the

temporal lobes.

Autonomic nervous system changes also have been

documented. For example, changes in heart rate and

electrodermal skin conductance response that occur with

aging have frequently been used to argue for an under-

arousal hypothesis of aging (Albert & Kaplan, 1980).

Classical conditioning studies using skin conductance

responses have shown the elderly to condition less readily

and extinguish more quickly (Botwinick & Kornetsky,

1959,1960). Older subjects also showed overall less

responsivity during the habituation period. Heart rate

deceleration in a warned reaction time procedure is believed

to be an index of filtering out irrelevant stimuli when

subjects are alert but relaxed while waiting for the

stimulus (Kahneman, 1973). In older subjects the magnitude

of the heart rate deceleration during the foreperiod is also

reduced, supporting the notion of under-arousal (Thompson &

Nowlin, 1973).

Woodruff (1985) notes, however, that many of these

studies involved passive responses to stimuli, and these

results may be quite different in situations calling for

active intentional processing. Kahneman (1973) points out

that demands on an individual's attention alters arousal,

while the level of arousal also influences how attention is

allocated. He cites the original Yerkes-Dodson studies

completed in 1908 that demonstrated that increased levels of









arousal improve performance on a learning task only up to a

point, and then additional arousal results in declines in

learning performance. Studies with sleep-deprived subjects,

who are considered under-aroused, typically show that

increasing motivation can improve performance to normal

levels (Kahnemen, 1973). This suggests that tonic arousal

and task demands interact to influence performance.

Others have suggested that the elderly are "over-

aroused." Supporters of this hypothesis have relied on

biochemical measures of nervous system functioning such as

free fatty acid (FFA) in blood plasma, which is believed to

be highly correlated with the level of autonomic nervous

system activity (Woodruff, 1985). After stressful tasks the

elderly show greater FFA concentrations in comparison to

baseline measures than do the young, which led Eisdorfer,

Nowlin, and Wilkie (1970) to suggest that over-arousal

interferes with elderly performance. To test this

hypothesis they administered propranolol, an adrenergic

blocking agent, and tested an elderly population's memory

performance. They found better performance of elderly

subjects who received this drug compared with another group

who did not. However, there was no younger group with which

to compare performances, and a subsequent effort to

replicate this study using a within subjects design was

unsuccessful (Froehling, 1974).

Studies attempting to manipulate arousal and to measure

the effects on elderly performance have produced mixed









results. Although not concerned with memory tasks per se,

Falk and Kline (1978) found that white noise selectively

impaired elderly performance on a critical flicker fusion

task. In contrast, Woods' (1980) study, in which body

position was used to increase arousal, found an older

group's performance was selectively enhanced on a reaction

time task. This difference may be due to the difference in

manipulations to increase arousal; arousal induced by

changes in body position may not be equivalent to change

induced by white noise.

Arousal and Memory: Interactions

Research on the interaction between arousal and memory

has a long and sometimes confusing history. Though arousal

has often been treated as a unitary phenomenon in the memory

literature, there are many important distinctions that

should be made. One of the most elementary distinctions in

arousal is the separation of event-related phasicc) arousal

from state (tonic) arousal (i.e. the overall activation

level of an organism). Phasic arousal refers to the

reaction of a organism to particular stimuli, while tonic

arousal refers to the resting (background) state of the

organism. It is also useful, within phasic arousal, to

distinguish between the arousal that indexes cognitive

processing, and the arousal that results from exposure to

highly shocking or emotional stimuli. Within tonic arousal

there may be differences in alterations induced by physical

activation, induction of mood states, and white noise.









Physiological arousal also may be separable from

electrocortical or central arousal.

The influence of phasic arousal on memory functions in

college students has been studied extensively, and fairly

consistent results have been obtained. Eysenck (1976,1977)

and Levonian (1972) provide reviews of this literature. In

the majority of studies the magnitude of phasic arousal at

acquisition has different effects on immediate and long-term

recall. Most studies use electrodermal skin conductance

response as the measure of phasic arousal, and find that low

arousal items are remembered best immediately but are not

recalled as well after a delay; high arousal items are not

recalled as well immediately but are recalled better after a

delay (e.g. Parkin, 1982, McLean, 1969, Levonian, 1967,

Kleinsmith & Kaplan, 1964,1963, Levinger & Clark, 1961). In

the few studies that did not produce this result, no

correction for order of presentation was instituted,

habituation to the testing situation was not achieved prior

to the start of presentation, or individual variation in

response to the stimuli was not taken into account

(Levonian, 1972).

In the earliest studies "arousal" was equated with

"emotion," and this memory pattern was interpreted as

"repression of trauma." In these studies emotional words

were used and suggested to be traumatic, but later studies

demonstrated that arousal from non-emotional stimuli could

also produce this pattern of recall (Eysenck, 1977). Most









recently, arousal during exposure to stimuli has been

suggested to represent the amount or depth of processing

that occurs and increased processing of arousing stimuli is

what causes the increase in delayed recall as well as

increased recognition (Pribram & McGuinness, 1975, Stelmack,

Plouffe, & Winogron, 1983). Information ultimately stored

in memory may not be as accessible immediately if the highly

arousing stimuli requires more time to be consolidated.

Other authors have focused attention on the interfering

effects of arousal on memory performance. The "von Restorff

effect" (Wallace, 1965) refers to the phenomenon in which

one novel stimulus embedded in a group of homogeneous items

is remembered better than other items (e.g. the word

"blood", embedded in a list of furniture items, will be

remembered better than the furniture items). In a variation

of this procedure Tulving (1969) demonstrated that the items

immediately preceding the novel stimulus were not remembered

as well as the other items more distant from the novel

stimulus. Tulving labeled this phenomenon "induced

retrograde amnesia" and likened it to retrograde amnesia

seen after trauma. This effect has been replicated with

both words and pictures, and it has been shown that the

amnesia can be seen for stimuli both immediately preceding

and following the novel stimuli (e.g. Saufley & Winograd,

1970, Schultz, 1971, Erdelyi & Blumenthal, 1973).

The mechanism by which this amnesia occurs continues to

be debated. Many theorists invoke concepts borrowed from









information processing models (see Detterman & Ellis, 1972;

Spear, 1978, Tulving, 1969). Christianson and Nilson (1984)

presented a list of normal faces interrupted by one grossly

deformed face. They measured skin conductance and ratings

of unpleasantness and noted that a large change in arousal

accompanied the critical event. They suggested this phasic

change in arousal interfered with acquisition of surrounding

information (i.e., an "encoding" defect affecting the

acquisition, not the retrieval, stage). Other researchers

argue against this separatist explanation and instead

contend that retrieval and encoding are interdependent

processes (e.g. Crowder, 1982, Cermak, 1982, Schacter &

Tulving, 1982).

The influence of tonic arousal on memory is more

complicated, perhaps in part because it is a more complex

multidimensional phenomenon. Tonic arousal is often

associated with the mood or state of an organism or an

overall orientation of the organism to a specific situation.

The extremes of arousal are seen when the organism is asleep

or awake. Although sometimes viewed as a unitary state,

tonic arousal is influenced by numerous factors and varies

widely across subjects and situations.

Tonic arousal can be manipulated by white noise,

increased physical activity prior to learning, or by mood

induction. It is not clear from the literature whether

these manipulations influence just physiological arousal or

both physiological and psychological aspects of arousal.









Many memory studies report results similar to those seen in

the phasic arousal literature, including the differences in

immediate versus delayed recall of high-arousal and low-

arousal conditions. Eysenck (1976) argues that sampling

only two points (high versus low) of the arousal continuum

may be misleading. He suggests that an inverted U shaped

relationship between immediate memory and arousal may exist,

but this cannot be demonstrated with only two points

sampled. Studies involving recall have at times yielded

conflicting results, and both semantic and phonemic

clustering have also been found to be influenced by tonic

arousal. One study demonstrated a decline in semantic

clustering in a white noise condition (Hormann & Osterkamp,

1966 as quoted in Eysenck, 1976) and Schwartz (1975)

demonstrated that white noise during learning could be used

to decrease semantic clustering in free recall, while

improving performance on phonemically related words.

Being in the same mood or arousal state at retrieval as

in acquisition has also been found to improve memory

performance in college-aged subjects (e.g. Bower, Monteiro,

& Gilligan, 1978, Clark, Milberg, & Ross, 1983). Many

theories continue to be debated, but some authors suggest

physiological arousal may play an important role. Clark,

Milberg, and Ross (1983) manipulated subjects' level of

arousal by having subjects either exercise or relax for

seven minutes, had subjects learn phrases, and then tested

them for recall in either the same or different arousal









state. They found that subjects consistently recalled more

information when they were tested in the same state at test

and acquisition. From this they proposed that one's state

of arousal (i.e., level of tonic arousal) is one of the many

pieces of information retained along with the stimulus

content in any learning situation, and being in the same

state provides additional cues for later accessing that

information.

Changes in Brain Physiology with Aqing

Physiological and biological changes that accompany

senescence are well known and have been documented over the

last fifty years. These include a loss of total brain

weight and reductions in specific populations of neurons, an

increase in senile plaques and neurofibrillary tangles, and

changes in levels of various neurotransmitters (Brizzee,

Ordy, Knox, & Jirge, 1980; Berry, 1975). Although plaques

and tangles are generally present in normal aging, their

significance has been disputed when pathological cases are

excluded (e.g., Tomlinson, 1972). How these changes

influence memory functions remains controversial.

Cotman and Holets (1985) review the human and animal

literature on structural changes in the brain with age.

They delineate some of the problems that have plagued this

area of research and note that variable results are

frequently obtained even when researchers sample the same

brain region. Because of this, links between behavioral

changes and physiological changes must be made with caution.









Some researchers have documented widespread changes such as

overall decreased water content in the brain, decreased

oxygen uptake, decreased glucose utilization, and

circulatory changes, while there is some suggestion that the

cell loss that accompanies aging is probably regionally

specific (Timiras, 1988). Evidence exists for significant

cell loss in the cerebral cortex, the locus coeruleus, the

substantial nigra, the cerebellum, and the hippocampus

(Kubanis & Zornetzer, 1981). Findings of cortical and

hippocampal cell loss may be linked to the changes in both

arousal and memory seen in the elderly.

Neurotransmitters have also been suggested as important

in the study of memory. The information about the role of

acetylcholine in memory performance comes from studies such

as Drachman and Leavitt (1972), who administered

scopolamine, an acetylcholine antagonist, to a group of

healthy young adults. They found scopolamine produced

transient dose-related memory problems similar to those

found in the elderly. These effects in the young could

easily be reversed with the administration of an

acetylcholine agonist. Neurotransmitters changes associated

with aging are difficult to study, but using animal models

and indirect measuring techniques some alterations that

appear to specifically accompany aging have been discovered.

There is little evidence for changes in the overall

level of most neurotransmitters, but there has been more

evidence regarding alterations in both levels of









neurotransmitters and receptors in particular brain regions

(Kubanis & Zornetzer, 1981). The catecholamines have been

widely studied and some age differences have been found.

Choline acetyltransferase (CAT) is the synthesizing enzyme

for acetylcholine and is easier to study because it is much

more stable in postmortem studies than acetylcholine

(Katzman & Terry, 1983). This enzyme decreases in activity

with age, especially in the hippocampus and temporal

neocortex, two brain areas repeatedly suggested as important

in memory (Katzman & Terry, 1983).

The implication of these findings is that acetylcholine

agonists, which increase the amount of acetylcholine

available at the synapse, could be used to improve memory in

the elderly. Experimental administrations of acetylcholine

agonists, however, have failed to produce dramatic changes

in memory functioning in the elderly, suggesting that a

simple lack of acetylcholine is not a sufficient explanation

for the memory problems seen in elderly subjects (Katzman &

Terry, 1983).

Neuropsychological Mechanisms in Arousal and Memory

Brain areas typically associated with the mechanism of

arousal and attention are the reticular activating system,

the diencephalon, and hypothalamus (Pribram & McGuiness,

1975). Research on arousal and the brain stem has a long

history, with earliest studies focusing on animals and sleep

and wake cycles. The brain stem has been of interest since

as early as 1949 when Moruzzi and Magoun noted that









electrical stimulation of the reticular formation would

increase arousal. Additional studies have demonstrated that

animals with lesions of the reticular formation can be

momentarily aroused but are typically unable to maintain

high levels of arousal (Carlson, 1977). Nolte (1988)

describes the reticular formation as being important in many

different psychological functions, diffusely organized, and

characterized by a great deal of convergence and divergence

in patterns of connectivity.

The reticular formation has a lateral portion and a

medial portion (Nolte, 1988). The gigantocellular reticular

nucleus is located in the medial zone in the rostral

medulla. In the pons, the medial zone is divided into the

oral pontine and the caudal pontine reticular nuclei. The

raphe nuclei are in the center of the reticular formation as

it goes through the medulla, pons, and midbrain.

Projections from the reticular formation to the thalamus,

hypothalamus, and basal ganglia are believed to be important

in the regulation of consciousness. These projections

terminate in the intralaminar nuclei, which has connections

throughout the cortex. Bilateral disruption of these fibers

produces prolonged unconsciousness. The pathway to the

thalamus is believed to be especially important in

maintenance of tonic arousal.

Carlson (1977) cites evidence from two animal studies

that suggests that the forebrain, defined as the cortex,

basal ganglia, limbic system, thalamus and hypothalamus, is









capable of producing arousal, independent of brain stem

mechanisms. First, Genovesi, Moruzzi, Palestini, Rossi, and

Zanchetti (1956) demonstrated that when the brain stem was

lesioned gradually over time, the animals were not comatose

but showed periodic signs of arousal. Batsel (1960) is also

cited in which EEG evidence demonstrated a gradual recovery

of desynchronized activity after a single brain stem lesion

in dogs. Hypothalamic lesions have also been shown to

consistently result in lethargy and hypoactivity (Nauta &

Feirtag, 1986).

Pribram and McGuinness (1975) conclude that the

behavioral evidence from both animal and human studies

supports three neurallyy distinct and separate" systems of

attention. They propose three basic attentional control

mechanisms in which arousal after input is mediated

primarily by the amygdala, activation in preparation for

response is mediated by the basal ganglia, and state that

the hippocampal circuit coordinates the two.

Pribram and McGuinness (1975) locate arousal mechanisms

in the brain stem, reticular formation as well as in the

diencephalon and into the hypothalamus, and suggest that the

amygdala and related frontal cortex are important in

attentional control of the structures implicated in basic

arousal. They note that behavioral habituation will not

occur in animals with amygdala or bifrontal damage. Humans

with bilateral amgydala lesions have been noted to exhibit a

decrease in aggressive behavior, and there may be









differences between animals and humans in the effect of

these lesions (Carpenter & Sutin, 1983). Dorsolateral

frontal lesions have been shown to eliminate the

visceroautonomic orienting responses (Carpenter & Sutin,

1983).

Pribram and McGuinness (1975) argue for the existence

of reciprocal systems, one in the dorsolateral frontal

cortex and the other with opposite function in orbitofrontal

region to regulate the orienting response. The existence of

this reciprocal system is used to argue for a locus of

control, which they hypothesize to be located in the

hypothalamic region. This is supported by the results of

electrical stimulation of the hypothalamus resulting in

episodes of fighting or fleeing and lesions which disrupt

the cessation of drinking and eating behavior once

initiated.

Heilman, Watson, and Valenstein (1985) define arousal

as the "physiologic state of preparing to process a

stimuli", which they separate from motor activation which

involves preparing to act or "intention". Heilman, Watson,

and Bowers (1983) summarize evidence that the right

hemisphere may play an important role in the mediation of

activation and arousal, as well as emotional behavior.

Nauta and Feirtag (1986) summarize additional evidence

that suggests that the hypothalamus is also critical in the

mediation of emotional behavior, affect, and motivation.

Their summary includes studies in which electrical









stimulation of the cat's hypothalamus results in increased

agitation and aggressive behavior, while weak stimulation of

the hypothalamus appears to be experienced as extremely

pleasurable or extremely aversive depending on the site

within the hypothalamus. Its connections to the limbic

system, especially the hippocampus, are suggested as

exceedingly important in memory and the determination of

what is memorable.

Hippocampal lesions, which are known to produce

profound amnesia, also influence patterns of arousal

response. The famous H.M. is the most well-studied patient

with severe amnesia that resulted from surgery to relieve

intractable epilepsy. This surgery involved bilateral

destruction of the anterior two-thirds of the hippocampus,

the hippocampal gyrus, and the amygdala (Scoville & Milner,

1957; Penfield & Milner, 1958). Despite relatively intact

old learning and social skills, H.M. is unable to overtly

recall material if there is any significant delay between

learning and recall (Milner, Corkin, & Teuber, 1968). In

addition, H.M. shows no electrodermal response (EDR) to

shock that normal subjects find painful, and he appears

unaware of internal states such as hunger, thirst, pain, and

fatigue (Hebben, Shedlack, Eichenbaum, and Corkin, 1981).

Animal studies suggest that in animals with hippocampal

lesions, phasic EDR returns to normal more rapidly, which

may represent less processing. These animals also seem

abnormally indistractable when absorbed in performing a task









(Pribram & McGuinness, 1975). Stimulation of the

hippocampus in cats has been associated with facial

expression suggestive of increased attention, increased

anxiety, and bewilderment (Brodal, 1981). Just as in

humans, however, overt responses can be differentiated from

covert responses on some learning tasks. Hippocampectomized

monkeys are noted to be perceptually distractible, failing

to habituate to a distractor, while behaviorally able to

cease overt responses to the distractor (Douglas & Pribram,

1969). Pribram and McGuiness (1975) conclude that more

effortful processes are abandoned in favor of more primitive

or simple relationships that do not require central control

operations when the hippocampus has been removed from the

system.

Brain lesions believed to be associated with amnesia or

deficits in the acquisition of new information include the

temporal lobes and the hippocampus, the diencephalon, and

the frontal lobes (Squire, 1982). Deficits in immediate or

short term memory have been associated with lesions in the

parietal lobes (Kolb & Whishaw, 1985). Squire (1982) notes

that investigations of the specific pattern of memory

performance of patients with lesions in different brain

areas reveals that amnesia is not a unitary disorder, that

different brain lesions are associated with different types

of memory deficits, and that some of the behavioral deficits

seen in amnesic subjects may not to be related to the

amnesia at all. Investigations finding different rates of









forgetting among patients with bitemporal versus

diencephalic amnesias have led to suggestions that these

amnesias represent separate types of amnesia (Huppert &

Piercy, 1977); for one group the amnesia may be related to

storage deficits, while the other amnesia may be the result

of a disruption in retrieval abilities. Changes in memory

performance seen after frontal lobe lesions include

increased proactive interference (PI) and a lack of release

from PI, increased errors of intrusions and omissions, and

specific difficulty with memory for temporal order (Kolb &

Whisaw, 1985).

Problems demonstrating the anatomical distinctions

between different types of memory problems are inherent in

the study of brain pathology in humans, since lesions

resulting from diseases or brain insult are rarely limited

to one specific brain area. In addition, limitations in our

ability to evaluate the structural integrity of a living

human brain, despite the advances in brain-imaging

techniques, continue to make specifying an exact extent and

location of areas of pathology or lesions difficult.

Attention and activation are recognized as crucial

components in memory, and both are frequently impaired in

human brain-injured populations. In the elderly, attention

is also thought to be compromised, and Kinsbourne (1982)

suggests this is due both to diffuse neuronal depletion and

to more focalized cell loss. He suggests that neuronal

depletion throughout the cortex may effect one area









important in an opponent process, which will disrupt the

entire system. He contends the diffuse damage should result

in the preservation of specific functions with an overall

slowing of responses, lack of vigilance, and resistance to

changes in mental set. The disruption of opponent

processing could easily interfere with attention in general,

and focalized lesions would determine the nature of an

individual's cognitive deficits. Kinsbourne (1980) suggests

that this model best fits the deficits seen in the elderly,

who show increased variability in performance compared to

younger adults as well as a more generalized decline.

Albert and Kaplan (1980) argue for more focalized

frontal deficits in the elderly, and use comparisons with

brain-damaged patients to support this contention. Two

areas they suggest may illustrate focalized impairment are

the elderly's alteration in attention/arousal and type of

difficulties they demonstrate in visuospatial performance.

They review both the electrophysiological literature and the

cognitive psychology literature for support of the deficits

in attention, and provide a qualitative analysis of the

elderly's performance on visuospatial tasks. They note that

different measures of arousal in the elderly can suggest

either under- or over-arousal and summarize some possible

explanations for these discrepancies. For example, there

may be problems in equating physiologic measures of arousal

in young and old, since it is known that the end organs

change with age. This could result in these measures no









longer accurately reflecting arousal in older adults.

Alternately, they also note that it may not be absolute

arousal, but the degree of congruence of arousal between the

central nervous system and the autonomic nervous system that

may change with age. Behavioral studies cited include

studies of divided attention, studies in which subjects are

required to ignore irrelevant cues, and studies which

measure central nervous functioning while engaging in these

tasks. They argue that all these differences may be related

to changes in frontal lobe functioning.

Aging, Arousal, and Memory

Overall, the arousal, aging, and memory literature

suggests that arousal is altered with increased age and that

these alterations could be important factors in the memory

performance of aging individuals. Comparisons of memory

performance between the young and old subjects suggest that

the elderly may encode information more generally and may

have difficulties in both acquisition and retrieval of new

information (Poon, 1984). Relationships between arousal and

memory have been studied almost exclusively in the young,

and in this population phasic arousal appears to be a useful

index of depth of processing, while tonic arousal has been

suggested as a cue that can be used in retrieving

information. High levels of both item specific-phasic and

tonic arousal have also been shown to interfere with memory.

It has been hypothesized that this occurs by item specific

phasic arousal interfering with acquisition of surrounding









information and by increased tonic arousal decreasing deep

and increasing superficial processing. Brain areas that

have been shown to be important in both arousal and memory

appear to change with increasing age in both animal and

human studies. Both over-arousal and under-arousal theories

have been suggested as important in memory performance of

the elderly.

This relationship between arousal and older adults'

memory is probably not due simply to under- or over-arousal,

but to changes in the overall mechanism of arousal and

arousal regulation. Kubanis and Zornetzer (1981), for

example, suggest that difficulties with organization and

memory strategies may be secondary to over-arousal and

changes in homeostatic control of arousal that accompanies

aging. Research to date suggests that age differences in

phasic and/or tonic autonomic arousal may be related to age

differences in memory ability.

The nature of the memory-arousal relationship is the

focus of this study. The general hypothesis which guides

these studies is that alterations of autonomic and central

arousal may be important mediators of the declining memory

performance that accompanies aging. By manipulating item

specific phasic arousal in Experiment 1 and by manipulating

physiologic arousal in Experiment 2, four questions are

specifically addressed. First, are arousal and memory

performance related the same in the elderly as in the young

with regard to both phasic and tonic arousal? Second, do






30


the effects of arousal on memory performance in elderly

support a particular information processing model of age-

related memory decline? Third, how can this interaction be

used to help understand human memory performance at all

ages? Fourth, how do these performance differences relate

to known alterations in brain pathology that accompanies

aging?














EXPERIMENT ONE: METHODS

Induced Amnesia for Pictures
Before and After a Critical Event:
Overview and Rationale

A distinctive stimulus embedded in a group of

otherwise generic verbal or nonverbal stimuli has been shown

to interfere with retention of the stimuli surrounding the

critical event (Tulving, 1969). This critical event is

known to cause physiological arousal and this arousal, may

play a role in the concomitant interference with memory

(Erdelyi & Blumenthal, 1973). Since it is also known that

older adults tend to have smaller but longer lasting

physiological responses to novelty, this study focuses on

whether this pattern of interference is altered in an older

adult age group. Subjects were asked to remember two sets

of 25 pictures. All items were simple line drawings of

easily identifiable objects, except for the center item or

critical event in the second list, which was a photograph of

either a normal or emaciated child. If the older group's

performance was disrupted significantly more than the

younger group's by the critical event, this would support

the notion that over-arousal is a problem of some cognitive

significance for the elderly as has been suggested by

Eysenck (1977). In contrast, if the older adult's memory

31









performance was less disrupted, this would support the

notion of under-arousal or a more complex model of

regulation of arousal as suggested by Woodruff (1985).

Subjects

30 normal younger adults and 30 normal older adults

were recruited to participate in this study. There were 18

women and 12 men in the younger sample and 21 women and 9

men in the older sample. Older subjects were volunteers

recruited through local volunteer and senior citizen

organizations and were paid $10 for their participation.

Younger subjects either were recruited though the psychology

subject pool and received course credit for their

participation or were recruited through advertisements

asking for volunteers posted at Shands Teaching Hospital and

were paid $10 for their participation. Subjects in the

younger group were between the ages of 18 and 38

(mean=23.56, S.D.=5.78). Subjects in the older group were

between the ages of 60 and 80 (mean=69.56, S.D.=5.82). All

subjects were screened for neurological, psychiatric, and

medical disorders, as well as substance abuse problems or

medications that could alter their performance on memory

tasks via a questionnaire (see Appendix A). Subjects were

excluded if they had a history of stroke, uncontrolled

diabetes, epilepsy, drug abuse, a learning disability, or a

head injury that resulted in loss of consciousness for more

than one hour or hospitalization for concussion.









Younger adults reported having completed between 11 and

19 years of education (Mean=14.63, S.D.=2.22) and older

adults reported having completed between 8 and 20 years of

education (Mean=15.36, S.D.=3.36).

Materials
Materials for this experiment consisted of two lists of

24 line drawings with a "critical" 25th item placed at item

position 13. Both lists contained 24 sketches of easily

identifiable objects selected from a standardized set of

pictures assembled by Snodgrass and Vanderwart (1980) that

have been normed for name agreement, image agreement,

familiarity and visual complexity. The sketches chosen were

of the highest familiarity. Half the subjects saw List A

during Session 1 while the other half saw List B during

Session 1. All subjects saw the alternate list during

Session 2. For all subjects, during Session 1, a line

drawing of a snowman (Control Condition) was shown at

position 13. During Session 2, all subjects saw a

photograph (Critical Event) at item position 13. Half the

subjects were shown a muted color photograph of a normal

child holding a cup (Distinct Critical Event) while the

other half were shown a brightly colored photograph of an

extremely emaciated child (Disturbing Critical Event). The

two lists were presented a minimum of one week but not more

than two weeks apart.

Target and distractor items used for the recognition

test were chosen from the same pool of line drawings and









were balanced for familiarity. Correct recall was based on

acceptance of both dominant and nondominant names so that

subjects were not penalized for naming items differently

(e.g., plane, jet, and airplane were all equally acceptable

for line drawing of an airplane).

Subjects also were administered the Vocabulary Subtest

of the WAIS-R (Wechsler, 1981), the Center for

Epidemiological Studies Depression Scale (CES-D) (Radloff,

1977), and the State Trait Anxiety Inventory (STAI),

(Spielberger, Gorsuch, & Lushene, 1970). The WAIS-R

provided a screening measure which was used to compare the

young and old groups for differences in overall intelligence

as measured by a standardized instrument. The WAIS-R

Vocabulary subtest was chosen because it is relatively

insensitive to age differences, can be administered quickly,

and correlates highly with Full Scale IQ (Matarazzo, 1972).

The CES-D and STAI were administered to rule out increased

anxiety or depression in the older adults as an alternate

explanation for age-related differences in memory

performance. The CES-D was chosen as the measure of

depression because it has been normed on both young and old

adults and does not rely heavily on the somatic symptoms of

depression that may not indicate depression in the elderly

but may be represent normal physiologic changes that

accompany normal aging (Ensel, 1986).










Procedure

After passing initial screening subjects were scheduled

for an appointment. Upon arrival they were seated in a

recliner, given a brief description of what they would be

doing for the study, and then read and signed a consent

form.

Slides were presented by means of a slide projector

with an electronic shutter. Exposure duration was 500

milliseconds with a 500 millisecond inter-trial interval.

Prior to list presentation, subjects were asked to try to

remember what they were about to see, but were not warned

about the distinctive stimulus that was to be presented

during Session 2. Retention was tested with immediate free

recall and, after a thirty minute delay, free recall

followed by a yes/no recognition test. During the 30 minute

delay period the WAIS-R Vocabulary Subtest, the STAI, and

the CES-D were administered, as time allowed. Only two

older adult subjects were unable to finish these tasks

during the delay, and they completed the CES-D at the end of

the testing session. Most subjects completed these measures

and sat and talked with the examiner until a sufficient

delay period had elapsed.

Electrodes and transducers for psychophysiological

recording were also attached during the delay period,

approximately 20 minutes after the target slides were shown.

Electrodermal response (EDR) is frequently used as a measure

of phasic sympathetic arousal (e.g., Kleinsmith & Kaplan,









1963; Corteen, 1969) and has been suggested by Kahneman to

be correlated highly with attention and amount of effortful

processing. In this study EDR was measured to test for

differences in patterns of arousal to targets versus

distractors between the young and old. EDR was monitored by

Beckman instruments standard Ag/AgCl electrodes attached to

thenar and hypothenar eminences of the nondominant palm.

The analog signal was processed by a Coulbourn Model S71-22

Skin Conductance Module. This is a constant voltage system

which passes 0.5 volts across the palm during recording.

The analog signal was sampled and digitized every five

milliseconds by a Data Translation DT-2805 Analog to Digital

Converter (12-bit) housed within an IBM PC/XT microcomputer.

The digital signal was then processed by customized

software, yielding a EDR measure (in micromho units) every

50 milliseconds.

Following a five-minute adaptation period (Meyers &

Craighead, 1978) subjects were told the recognition portion

of the procedure would begin. They were instructed

initially, and reminded just prior to exposure, to avoid

deep breaths or movements that might alter the physiological

measurements. EDR's were averaged every second during the

presentation of each of the lists. The EDR was defined as

the difference in peak conductance from tonic to phasic, and

was calculated by subtracting the tonic mean from the phasic

peak for each picture. Positive scores thus reflect

conductance increases. Each EDR was also corrected for









range of individual responding as recommended by Lykken

(1972). To do this, each EDR was expressed as a proportion

of the magnitude of the subject's largest response during

the session. Using range corrected EDR corrects for

possible group differences in basic differences in phasic

arousal levels and allows a better measure of covert

processing.

Immediately after list presentation and after the

thirty minute delay subjects were asked to report all the

slides that they remembered seeing (i.e., free recall).

During the delayed recognition, 16 of the original target

items (i.e., those shown initially) were randomly

interspersed with 14 distractor items. Subjects were asked

to say "yes" if a slide had been shown before and "no" if it

had not. They were also asked to give a confidence rating,

on a scale from one to five, as to how confident they were

about their answer. A rating of "1" indicated they were not

certain and felt their answer was a "wild guess" while a

rating of "5" indicated that they were totally certain about

their answer.















EXPERIMENT 1: RESULTS

Psychometric Test Performance

On the WAIS-R Vocabulary Subtest younger adults

obtained raw scores between 25 and 66 (Mean=52.9, S.D.=8.3)

with age scaled scores between 8 and 15 (Mean=12.26,

S.D.=2.04). Older adults achieved raw scores between 27 and

67 (Mean=55.63, S.D.=9.95) with age scaled scores between 7

and 17 (Mean=12.8, S.D.=2.64). The two groups were not

significantly different from each other in education

(t(50)=.99, p=.32), Raw Vocabulary Score (t(58)=1.14,

p=.25), or Age Scaled Vocabulary Score (t(58)=87, p=.39).

These data suggest that both the younger and older adults

tended to be of high average intelligence and represent a

highly educated sample.

Results of subjects' responses on the STAI and the CES-

D are presented in Table 1. Some researchers have suggested

that older adults tend to be more anxious than younger

adults and that this anxiety may partially account for the

differences in memory performance (Poon, 1985). The State

Scores from the STAI and CES-D were subjected to two

separate repeated measures ANOVA's. Analysis of the State

scores revealed no Age Group by Session interaction

(F=(1,57)=1.106, p=.30) or main effect for Session

38









(F(1,57)=.136, p=.71), and only a main effect for Age Group

(F(1,57)=7.19, p=.0096, w2=.095). Analysis of the CES-D

showed the same pattern, again with no Group by Session

interaction (F(1,58)=1.6357, p=.2060) or main effect of

Session (F(1,58)=.0906, p=.7645), with only a main effect of

Age Group (F(1,58)=10.16, p=.0023, w2=.132). A t-test on

the Trait portion of the STAI also revealed Age Group

differences on this measure (t(58)=3.0467, p=.0035). The

differences in responses on the STAI and CES-D suggest that

the younger adults consistently rated themselves as more

anxious and more depressed both during the individual

testing sessions (State) and in general (Trait) than did the

older subjects. Neither young nor old changed their ratings

of their level of anxiety or depression appreciably between

Session 1 and Session 2.



Table 1

Anxiety and Depression Scores of
Younger and Older Adults


Session 1 Session 2

State Trait CES-D State CES-D
Young 36 40 13.7 37.1 15.1
(7.7) (9.3) (8.3) (9.3) (8.2)

Old 31.7 32.8 8.8 31 8.0
(10.4) (9.3) (8.6) (8.6) (7.2)



In the literature, memory impairments suggested as

related to increased levels of depression include: decreased









acquisition and recall, increased errors of omission,

transposition errors, miss pairings, and reversals of

stimulus target words, decreased strategy use and

organization of to be remembered material, increased access

to sad memories, altered guessing strategies, and decreased

attention and reaction time (Salzman & Gutfreund, 1986).

Depressed individuals, probably because of their increased

tendency for omission errors, may also perform worse on

recall in comparison to recognition tasks (Niederehe, 1986).

Of these deficits, the impairments that would be most

relevant to these data are the decrease in acquisition and

recall and the larger impairment in recall compared to

recognition memory. The literature also suggests that

increased anxiety is associated with lower performance on

both recall and recognition measures (Seigel & Loftus,

1978).

Since older adults typically perform worse than younger

adults on tests of memory and it has been hypothesized that

a higher level of anxiety and depression may contribute to

these differences, the STAI and CES-D were included to

insure that the older adult sample was not significantly

more anxious or depressed than the younger adults. This was

clearly not the case and the differences seen between young

and old were not large. The CES-D scores were below all

recommended cutoffs scores suggestive of depression (Ensel,

1986) and the mean STAI scores were at approximately the

50th percentile for the younger adults and at the 40th









percentile for the older adults (Speilberger, 1970). This

suggests that although the differences are statistically

significant, both older and younger groups were scoring in

the normal range on both anxiety and depression measures.

The younger adults who were reporting both more

symptoms of anxiety and more symptoms of depression, were

also more likely to perform at a higher level on virtually

all memory measures. Few of the depression studies have

been completed looking at the effect of self report of

depression on memory performance with subjects within the

normal range of depression; the majority of the studies have

compared subjects within the depressed range to subjects

within the normal range on measures of depression (Johnson,

Magraro, 1987; Weingartner, 1986). West, Boatwright, and

Schleser (1984), in examining the correlation between self

report of affective status with memory performance and

ratings of memory performance, found a correlation between

self report of depression and self report of memory

problems, but found no significant relationship between

actual memory performance and self report of depression. A

modest negative correlation was found between self report of

anxiety and memory performance.

It is most probable that these difference in anxiety

and depression are due to a sampling bias secondary to

recruiting a number of younger subjects currently in school

and completing a course requirement, which could be

associated with increased stress, in contrast to recruiting









many older adults who were retired and volunteered to

participate in psychological studies, frequently because

they found such participation interesting. Although the age

difference is statistically significant, it is unlikely to

be practically significant in the context of this study

since these differences are likely to minimize rather than

accentuate any age differences.

List Effects

All subjects were exposed to two lists, half seeing

List A at Session 1 and half seeing List B at Session 1,

with all subjects seeing the alternate list (B or A) at

Session 2. Although these lists were both constructed from

the same pool of items and each item was assigned to one

list or the other randomly, analyses were completed before

combining the data from the two lists to insure that there

were no significant list effects. Three separate Age Group

by Session by Test ANOVA's were performed to test for

effects on overall immediate and delayed recall, overall

recognition, or percent recall of each item position. The

only significant interaction was between List and Session in

the immediate and delayed recall ANOVA (List x Session:

F(1,56)=4.24, p=.04, w2=.052; List: F(1,56)=2.16, p=.1470,

w2=.0180; Session: F(1,56)=1.52, p=.2224, w2=.0083) and

there were no main effects of List in any of the analyses.

The significant interaction between List and Session on

total recall revealed that subjects who were shown List A

first recalled an average of 1.6% more items on immediate









recall and .04% more items on delayed recall from Session 1

to Session 2, while subjects who saw list two first,

recalled an average of 2.3% fewer items on immediate recall

and 4.8% fewer items on delay recall. Although

statistically significant, this difference was very small

and was not practically significant within the context of

this study. Since there were no interactions between the

position effects or age group membership and list, the two

lists were combined for all subsequent analyses.

Overall Recall and Recognition Performance

Performance of Young and Old Adults on tests of recall

and recognition are presented in Table 2. Previous research

with comparisons between younger and older adults on tests

of recall and recognition have found that the older adults

tend to do less well than younger adults on tests of recall,

but tend to do as well as younger adults on tests of

recognition and decay over a delay. Table 2 reveals that

this pattern of performance was also found in this study.

A 2(Age Group) x 2(Session) x 2(Immediate vs. Delayed

Recall) Repeated Measures ANOVA with the Recall Score and

Session as the repeated measures revealed a significant

effect for Immediate versus Delayed Recall (F(1,57)=1.06,

p=.0001, w2=.781) and a significant main effect of Age Group

(F(1,57)=5.24, p=.026, w2=.067). Both young and old

recalled fewer words after a delay and older adults recalled

fewer items than the younger adult groups on both recall









tasks. There were no other significant main effects or

interactions.



Table 2

Performance on Recall and Recognition


Session 1 Session 2


Immediate Delayed Recog- Immediate Delayed Recog-
Recall Recall nition Recall Recall nition

Young 11.0 8.4 24.7 11.0 8.4 24.2
(2.7) (2.9) (2.9) (3.3) (2.9) (2.8)

Old 9.8 7.3 25.5 9.6 6.5 24.5
(2.7) (2.8) (2.5) (2.6) (3.0) (2.4)




In contrast, a 2(Age Group) by 2(Session) Repeated

Measures ANOVA on recognition scores suggested no Age by

Session interaction (F(1,58)=.39, p=.53) and no main effect

of Age Group (F(1,58)=.87, p=.35). The main effect of

Session approached significance (F(1,58)=3.86, p=.054,

w2=.045), with both younger and older adults correctly

recognizing slightly more items during Session 1.


Effect of Item Position and
the Critical Event on Memory

Because of the large number of positions, and because a

priori it was predicted that the distinctive stimulus would

primarily influence recall for the immediately surrounding









groups of items, recall for items was collapsed across item

positions with items 1-3 designated "Primacy" items, items

4-7, 8-11, 15-18, and 19-22 designated "Intermediate" items,

items 12-14 designated "Middle" items, and items 23-25

considered "Recency" items for Session 1. For Session 2,

items 12, 13, and 14 are considered separately in order to

test for the effect of the Critical Event. These points are

plotted for each age group in Figures 1 and 2 A

9(Position) by 2(Age Group) by 2(Session: without vs. with

critical event) repeated measures ANOVA revealed a

significant Position by Age Group by Session interaction

(Position x Age x Session: F(8,51)=2.08, p=.04, w2=.015;

Age: F(1,58)=3.44, p=.0687, w2=.0390; Session: F(1,58)=6.79,

p=.0117, w2=.087) and Session by Position interaction

(Session x Position: F(8,51)=11.94, p=.0001, w2=.152;

Position: F(8,51)=16.11, p=.0001, w2=.197).

Since an a priori prediction was that the Critical

Event would effect items immediately surrounding it, Dunn's

test or a Bonferroni t test was used to determine whether

the Session 2 critical event was significantly better

remembered than Session 1 middle items, and whether Item 12

(immediately preceding the CE), and Item 14 (immediately

following the CE), were remembered less well than Session 1

center items. Since the ANOVA suggested that each age group

was affected differently, each was tested separately and p

values needed for significance were corrected for the number

of comparisons (i.e., 3 comparisons per age group). This

















0'
E
Er-i



0
W 4o
C)







0 U



O 4






U..0
o a








tt
o o co 34 C






















.4 *-4
0 a 0 O
/r**(








/ '-'-4











1 ____-------- 1 --- ---I 0

o O C \






47









LCo
c\2 ,o




S\-1
E
.-4












C4)


oC 0
\\ 44


0 E

0

oCto



14



,-4











I I I I
77-4-


& ow 1.100 Z-) Z 0 mi --









analysis suggested that the CE was remembered significantly

better than Session 1 middle items for both young and old

(Young: t(29)=4.0, p<.01; Old: t(29)=3.61, p< .01), and that

item 12 was remembered significantly less well by the

younger adults (t(29)=4.0, p<.01). Although the decline in

memory for item 14 approached significance for the young

adults (t(29)=1.88, p>.05), older adults' memory for items

12 and 14 was not significantly impaired (Item 12 t(29)=.53,

p>.60, Item 14 t(29)=.12, p>.90).

To avoid a loss of power and to avoid increasing

experimentwise Type I error, only center positions were

tested for age differences. However, it appears that both

younger and older adults showed similar primacy and recency

effects. Both young and old had a significant increase in

their memory for the critical item, but only the younger

adults showed a reduction in memory for items immediately

surrounding the critical item. Only the item immediately

preceding the critical item was significantly effected.

Older adults reported the CE less frequently than did the

younger adults, but this difference was not statistically

significant (t(29)=1.59, p=.ll).

To test for differences between the two different

critical events (i.e., Disturbing: starving child vs.

Distinctive: normal child photograph) an additional ANOVA

was performed. This 2(Age Group) by 2(CE Type: Disturbing

vs. Distinctive) by 3(Item Position: 12, 13, and 14) ANOVA

suggested that there were significant interactions









between Item Position and Age Group (Age Group x Item

Position: F(2,55)=4.48, p=.0135, w2=.033, Age Group:

F(1,55)=.76, p=.3881, w2<.0001; Item Position: F(2,

55)=43.04, p=.0001, w2=.398) as well as between CE Item Type

and Age Group (CE Type x Age Group: F(1,56)=12.11, p=.0010,

w2=.07; CE Type: F(1,55)=.19, p=.9723, w2<.0001). No other

main effects or interactions were significant. The majority

of the variance in this model was accounted for by Item

Position, while Age Group and Position and Age Group and CE

Type interactions account for a statistically significant

but relatively small portion of the variance.

The immediate recall performances of the two age groups

for each of the stimuli are presented in Figures 3 and 4.

Inspection of these figures revealed that although older

adults were less likely than younger adults to remember

either critical event, the younger adults were less likely

than older adults to remember items immediately surrounding

the critical item. In addition, although the starving child

or normal child stimuli were equally likely to be remembered

by the younger subjects, the younger adults were less likely

to remember the items immediately surrounding the picture of

the starving child than the items surrounding the normal

child. Older adults had exactly the opposite pattern, being

more likely to remember the items surrounding the photograph

of the starving child. The pattern seen by the younger

adults is consistent with previous research, in which more

disturbing stimuli resulted in more pronounced amnestic
















4c





4) 0
V'4
E




0




U r

,: 2





0o p





c oo

o--3






0 0
-
0I










I 1 I I I

0 C 0 2


a4 e 0 c 0 0 -) Z5 aQ o o -

























4)
g0



44



Co
4) 4.1


/{I 0-








S 4) 4J









p C
I 0
0 *4

0


U

Cs 0 -14






o \ 0 0 0 3




o C 0 C\









effects for surrounding items (Erdelyi and Blumenthal,

1972). The pattern seen by the older adults suggests that

their memory for surrounding events is not as profoundly

disturbed by distinctive stimuli, and in fact may be less

disrupted by what is more disturbing to younger adults.

Although the sample sizes are small, it is also

interesting to compare the memory performance of older

adults who did recall the critical event with older adults

who did not. Figure 5 compares the memory performance on

items 12, 13, and 14 for the 19 older adult subjects who did

recall the critical item with the 11 subjects who did not.

Although the small numbers preclude any meaningful

statistical analysis, these comparisons do not suggest that

older adults who do remember the critical item are

remarkably different from those who do not on the item

immediately preceding the critical item (t(28)=.18, p=.85).

Those who do not recall the critical event may be more

likely to recall the item immediately following the critical

item (t(28)=1.8, p=.08), although this difference is not

statistically significant.

Response Types
Although most previous studies, and this study as well,

found no difference between young and old on the overall

recognition accuracy, it has been suggested that there may

be differences in response styles and the types of errors

each Age Group is likely to make. It has been suggested






















4 0

0
Aj


0 0










U 0U)

oo
.\ 0 4 4.J






0) 0
SCO o -- .,















)o E-4 0








i I I
0 0 0 0 0 0


12 ou4 4oor--









that older adults are more likely to make false negative

responses (i.e., saying they do not recognize a stimulus

they were exposed to), while younger adults are more likely

to make false positive responses (i.e., saying that they

recognize a stimulus they were not exposed to).

The incidence of True Positives, True Negatives, False

Positives, and False Negatives is presented in Table 3. An



Table 3

Average Number of Each Response Type



True True False False
Positives Negatives Positives Negatives
(32 possible) (28 possible) (32 possible) (28 possible)

Young 27.8 24.4 3.6 4.2
(3.0) (2.7) (2.7) (3.0)

Old 26.3 26.3 1.7 5.7
(3.8) (1.8) (1.8) (3.8)


examination of the response types of young and old suggests

that this may have been the case in this recognition task.

Both d' and Beta were calculated for each subject based

on the 60 responses on the recognition tasks administered at

time one and time two. The measure d' is an index of how

sensitive a subject is to a signal or target, while Beta is

an index of response bias. A higher d' indicates that a

subject needs a stronger signal or perhaps needs to be more

certain that a target has been seen before the subject will









endorse it as a previously exposed item. A Beta higher than

one indicates that a subject is biased towards saying "no",

that they did not see the stimulus previously, while a Beta

less than one suggests that the subject is biased towards

"yes" responses, that they did see the stimulus previously

(McNicol, 1972).

In this study, both young and old averaged Beta's

slightly greater than one, suggesting a small bias towards

caution for both young and old (Young mean Beta =1.21,

S.D.=.37, Old mean Beta=1.56, S.D.=.84). Comparisons of

these means suggested that older adults were significantly

more cautious and likely to say "no" than the younger

subjects (t(58)=4.15, p=.05). The older adults' d' was also

significantly greater than the young adults, suggesting that

older subjects had a higher response criterion and were more

accurate discriminators (Young mean d'=1.75, S.D=.80, Old

mean d'=2.19, S.D.=.84; t(58)=4.18, p=.05).


Electrodermal Response to Recognized
and Unrecognized Targets

During the recognition phase of this study,

electrodermal response (EDR) was measured, as an index of

sympathetic arousal. A 2(Age Group) x 4(Response Type: true

positives, true negatives, false positives, and false

negatives) ANOVA on the range corrected EDR was not

significant for the overall model (F(7)=1.90, p=.066) and

did not suggest an age by response type interaction or an

overall effect for age group. A separate ANOVA examining









the effect of Age Group and Stimulus Type (i.e., target

versus distractor) was significant (F(3,3566)=3.40, p=.017),

suggesting an overall effect of Stimulus Type

(F(1,3566)=9.08, p=.0026, w2=.002), but no effect of Age

Group (F(l,1)=.17, p=.69) or interaction (F(1,l)=.96,

p=.33). These data suggest that regardless of a subject's

overt response or age group membership, they show greater

sympathetic activation to targets than to distractors. This

relationship is shown in Figure 6. However, the omega

squared value suggests that whether the stimulus is a target

or distractor accounts for only a very small proportion of

the variance in EDR.


















00


E4)
0 4
4-)











o o
b40
U4



-44








-4







I I --




o 1 0 a
E--



N CC\2


ft0 C .0 34 O co) ovt C 4 3 G














EXPERIMENT 1: DISCUSSION

This study examines whether exposure to an arousal

inducing Critical Event (CE) has the same effect on young

and old adults' memory for both the CE itself and the

stimuli immediately surrounding the CE.

It was found that these two age groups did not differ

significantly in education or intelligence, as measured by

the Vocabulary subtest of the WAIS-R, so it is unlikely that

differences in memory performance which did emerge could be

attributed to formal schooling or intelligence. Anxiety and

depression measures were included to insure that the older

age group was not significantly more anxious or more

depressed than the younger, since increased anxiety or

depression has been suggested as one possible explanation

for age differences found between younger and older adults.

In this sample the two groups were significantly different

in their responses on the STAI and the CES-D, but it was the

younger age group that was endorsing items suggestive of

higher levels of anxiety and depression. Since virtually

all the literature on depression and anxiety and memory

performance suggests that more anxious and depressed

subjects will do less well on memory measures, and it was

the younger adults who reported more anxiety and depression

58









and who performed better on the memory measures, it is

likely that these differences would minimize rather than

exaggerate age differences.

The pattern of results obtained on recognition and

recall measures, when comparing younger and older adults, is

consistent with previous research in which significant

differences are found on tests of free recall, but small or

insignificant differences are found on tests of recognition.

This supports the notion that the memory patterns seen in

this sample of young and old adults represent reliable,

replicable effects that have been demonstrated across

repeated studies.

Inspection of the effect of item position on recall

shows that, just as in previous research, both younger and

older adults show both primacy and recency effects.

However, the introduction of the CE in the middle of the

list influences the pattern of recall of young and old

differently. While younger and older adults are likely to

remember the critical event, induced retrograde amnesia

appears to be present only in the young adults. One

possible contributor to the lack of amnesia for older adults

was that, although only a trend, fewer older adults recalled

the CE. It was possible that only those older adult

subjects who failed to recall the CE would fail to show

retrograde amnesia, which could have altered the overall

pattern of recall; however, analysis of the recall patterns

of the older adults who did recall the CE compared with









those who did not, failed to find a difference in the effect

of the CE between these two subgroups of older adults. Even

the older adults who, like the younger adults, recalled the

CE did not show evidence of retrograde amnesia. This

suggests that item specific processing sufficient to

increase recall is not necessarily sufficient to induce

amnesia in older adults.

Additionally, young and old are affected differently by

different types of CE's. In young adults both CE's (i.e.,

photograph of a starving child and photograph of a normal

child) resulted in equal recall of the CE, the more

disturbing CE is associated with more pronounced retrograde

and anterograde amnesia. In older adults, even though the

starving child photograph was more likely to be remembered,

there was no evidence that it resulted in a substantial

decline in recall for surrounding items.

This lack of disruption among the elderly is not easy

to explain. These data do not support the notion that older

adults are chronically over-aroused and that their memory

for items are more easily disturbed since they were, in

fact, less likely to exhibit disrupted memory than the

younger adults. It may be that the already lower memory

performance on the center items in the older adults group

may have contributed to a lack of amnesia, since there was

less of a distance for them to decline. But younger adults'

memory for the item immediately preceding the critical event

was even lower than the older adults' memory. Thus, an









explanation based on overall level of recall of middle items

does not seem adequate to account for this data.

How arousing stimuli affects memory performance in

older adults does appear to be clearly different from how it

affects memory performance in the young. The younger adults

seem to overfocus on the distinctive stimulus, to the

detriment of memory for surrounding items. This effect is

even more extreme when stimuli are disturbing. Older

adults, whose memory for distinctive items does not increase

as much as younger adults, do not seem to overfocus and lose

surrounding information. This difference however, does not

appear to influence overall recall, but only which items are

recalled. Since it has been suggested that manipulations in

tonic arousal influences total recall performance, the fact

that only the memory for the middle items was influenced by

the CE supports the notion that this experimental

manipulation primarily influenced phasic and not tonic

arousal.

This study also allowed us to examine response types of

younger and older adults on recognition tasks and to compare

electrodermal recognition of items. The analysis of these

results suggest different response biases and types of

errors likely with young as compared to older adults, with

younger adults more likely to have false positive responses

and older adults more accurate discriminators and being more

cautious than the young in endorsing both targets and

distractors. This response bias does not provide a









sufficient explanation for the item specific memory

differences found between younger and older adults, since it

was specifically the items immediately surrounding the

critical event that showed greatest differences between

young and old adults. It is possible that the older adults

tendency towards caution contributed to their reporting the

CE less frequently (although again, this difference was not

statistically significant) but cannot explain why the older

adults failed to show a decrement in memory for items

surrounding the CE.

Analysis of EDR suggested that both younger and older

adults tended to respond more strongly to targets than to

distractors, regardless of whether the target was explicitly

recognized, although target versus distractor status

accounted for a very small proportion of the variance in

EDR. No age differences or interactions were demonstrated

in this phenomenon. This suggests that in older and younger

adults, relative sympathetic arousal response to previously

exposed stimulus is an unlikely source for memory

differences. However, these measures were taken only during

the recognition portion of this study and it is noteworthy

that the overt, as well as covert, recognition failed to

demonstrate age differences as well. Differences in arousal

during acquisition or recall may still be present and

contribute to age-related memory differences.

The age differences found in reported levels of anxiety

and depression measures may also reflect the lower level of









arousal in older adults and may have contributed to the

different patterns of memory performance found in the two

age groups. The results may be altered with an older adult

group with higher levels of anxiety or depression, or with a

younger adult group with lower levels of anxiety and

depression. Although it is unlikely that the age

differences in anxiety and depression account for the age

differences in memory performance since both groups were

within normal limits on the measures of anxiety and

depression, additional research including samples of

subjects with more variability in tonic arousal would

clarify this issue.

It is also possible for the Yerkes-Dodson curve to

influence age differences in three different ways. First,

older and younger adults, due to cohort differences or

differences in their experience or lifetime experiences or

exposure to disturbing photographs or stimuli may be

responding differently to the photographs presented as the

critical events which could result in arousal levels being

at different points on the curve for the two age groups.

Different photographs or different types of distinctive or

disturbing events may result in arousal levels for young and

old at different places on the Yerkes-Dodson curve that has

been hypothesized to influence performance. Secondly, tonic

arousal differences between young and old may result in the

Yerkes-Dodson curve itself being different for two age

groups. Finally, there could be an interaction between the





64


age-related differences in reaction to the stimuli and the

different curves for the two age groups. It is not possible

to explore this possibility within this study since only two

levels of arousal were sampled, but additional research

sampling more points on the phasic arousal curve could show

a different relationship between arousal and aging.














EXPERIMENT 2: METHODS

The Interaction of Ace and Increased
Tonic Arousal on a Supraspan List Learning Task
Overview and Rationale

Whereas the first experiment manipulated arousal

processes associated with individual items and observed the

resulting effects on item position data, the second study

manipulated overall (tonic) levels of physiologic arousal

and evaluate its effects on general levels of recall and

recognition. Psychophysiological studies of the elderly

have been interpreted to suggest that tonically, the elderly

are both less aroused (e.g., Thompson & Nowlin, 1973) and

more highly aroused than the young (e.g., Woodruff, 1985).

While a low level of white noise and moderate exercise

increase arousal and improved delayed memory performance in

a group of college students (Eysenck, 1977), a white noise

condition has been shown to interfere with the performance

of an elderly group while exercise caused an improvement of

an elderly group's performance on another cognitive task

(Woodruff, 1985). This illustrates that arousal may not be

a single state variable, and that differences may exist

between different kinds of arousal states. Young and old

may be influenced differently by changes in state arousal.

Other studies have suggested that heightened arousal may

65









interfere with semantic encoding of information (Eysenck,

1976; Schwartz, 1975). Since changes in strategy use and

semantic clustering has also been suggested as a possible

source of declining memory in the elderly (Poon, 1985), this

raises the possible role of arousal as a moderating variable

in these effects.

This experiment examines the effect on memory of

activities that have been known to moderately increase or

decrease levels of physiological arousal in both young and

old adults. A modified version of the California Verbal

Learning Test (Delis et al., 1983) was used to measure

immediate and delayed recall, recognition, and semantic

clustering. An exercise and a rest condition were used to

alter physiological arousal.

This experiment was designed to provide information on

how tonic arousal interacts with encoding processes in

memory. If the older group's performance improves more with

moderate exercise than with no exercise, this may be

construed as supportive of the under-arousal hypothesis

while if the opposite pattern results that may be more

supportive of the over-arousal hypothesis. In comparison to

the younger group, if they do not improve as much or at all

under the exercise condition, this would suggest that if

under-arousal is indeed a problem, increasing physiological

arousal is not sufficient to solve it.










Subjects

Subjects in this experiment were recruited and screened

as outlined in Experiment 1. Again there was a young and an

old adult group, each containing 48 subjects. The younger

group in this sample were between 18 and 38 years old

(mean=19.61, S.D.=1.70) and had completed between 13 and 16

years of education (mean=14.2, S.D.=1.0), while the older

group was between 60 and 80 years old (mean=69.35,

S.D.=5.66) and reported completing between 8 and 20 years of

education (mean=15.6, S.D.=3.0).

Experimental Conditions

Immediately prior to exposure to the word list, 24

subjects from each Age Group were asked to exercise for five

to seven minutes and the 24 were asked to rest for five to

seven minutes. 24 subjects in each of the Age Groups were

then asked to engage in the same activity and 24 were asked

to engage in the alternate activity immediately prior to

delayed recall. During the rest condition subjects remained

seated in a recliner. During the exercise condition

subjects walked inside Shands Teaching Hospital.

Materials and Apparatus

As in experiment 1, the STAI, CES-D, and Vocabulary

subtest of the WAIS-R were administered to test for

differences in reported anxiety or depression or general

intelligence. To measure memory performance a modified

version of the California Verbal Learning Test was

constructed. This test consisted of a 20 item word list,









with each item fitting into one of five categories. The

word list was read three times, instead of the standard five

times, and subjects were asked to report all items they

could recall after each presentation. After a thirty minute

delay, free recall, cued recall, and forced choice

recognition were also administered. The forced choice

recognition test consisted or 42 items, the 20 targets and

22 distractors (See Appendix B).

Since it has also been suggested that younger and older

adults may react differently to the testing situation

itself, data was also collected about subject's reactions to

the distractor tasks. In addition to collecting the

physiological measures of arousal before and after engaging

in either rest or exercise, these measures were also taken

at the beginning of the 30 minute delay, and immediately

after learning the word list in order to be compared with

these same measures after engaging in the distractor tasks.

Subjects were also asked to subjectively rate how they

were feeling on the Self Assessment Mannikin or SAM (Hodes,

Cook, & Lang, 1985). This measure asks subjects to place

themselves along a nine point continuum on three different

dimensions: a happy to sad dimension, an excited to calm

dimension, and a feeling in control to feeling out of

control dimension. These measures provide both objective

and subjective information about unintended changes in

arousal produced by participating in this study. Again,

this was used as a manipulation check and to rule out










alternate explanations for the differences seen in memory

performance between young and old.

Both immediately before and immediately after this five

to seven minute period blood pressure and pulse rate were

measured. These measures were taken as a check on the

experimental manipulation because they are very responsive

to increased physical activity (Clark, Milberg, & Ross,

1983). Blood pressure and heart rate were chosen because

they can be measured quickly and easily. These measures

were taken using a Dinamap Adult/Pediatric and Neonatal

Vital Signals Monitor, Model 1846 SX. This monitor

automatically provides a measure of systolic and diastolic

blood pressure and pulse rate.

Procedure

After arriving for their appointment, subjects read and

signed a consent form. Baseline blood pressure was taken

after subjects had read and discussed the consent form with

the examiner. Those assigned to the relaxation condition

sat in a chair and were then asked to relax for five to

seven minutes. Those assigned to the exercise condition

walked around the health center with the examiner for five

to seven minutes.

Immediately after the five to seven minute period blood

pressure was taken and the modified CVLT was administered.

A third blood pressure reading was taken after the modified

CVLT was administered. During the thirty minute delay

subjects were administered the Vocabulary subtest from the









WAIS-R, the CES-D, and the STAI. After a twenty minute

delay, a fourth blood pressure reading was taken and

subjects were again asked to either relax for five to seven

minutes or exercise for five to seven minutes. Half of the

subjects in each group engaged in the same activity as they

did originally and the other half engaged in the alternate

task. After another blood pressure reading, subjects were

asked to recall all items from the original list, first with

no cues (delayed free recall) and then with category cues

(delayed cued recall). Finally, subjects were read a list

of 42 items and asked to say "yes" if an item had been on

the list and "no" if it had not.

The following dependent measures were collected: 1) The

number of CVLT items correctly recalled after the first,

second, and third repetitions, and after a 30 minute delay.

2) The Observed Semantic Clustering Score for each of the

previously described times. Subjects were given one point

each time they recalled one word after another that was from

the same semantic category. Credit was given for a semantic

cluster only if it was the first time each of the words had

been recalled in that trial and if the word was actually a

target. 3) The Cluster Percentage Scores for each of the

above described times. This was calculated by dividing the

Observed Clustering Score by the maximum number of clusters

possible given the number of correctly recalled items on

that trial.















EXPERIMENT 2: RESULTS

Psychometric Test Performance



On the WAIS-R Vocabulary subtest younger adults

achieved an average raw score of 49.0 (S.D.=7.1) and older

adults achieved an average of 57.7 (S.D.=9.0), with mean age

scale corrected scores being in the high average range for

both younger (mean=11.7, S.D.=1.9) and older adults

(mean=13.4, S.D=2.5). On the STAI the younger group had an

average state score of 36.2 (S.D.=8.1) and an average trait

score of 37.5 (S.D.=9.9) compared to the older groups

average state score of 31.1 (S.D.=8.1) and average trait

score of 32.2 (S.D.=7.0). On the CES-D, younger adults had

an average score of 17 (S.D.=10.6), while older adults

averaged 6.7 (S.D.=6.0).

Education, WAIS-R raw and age corrected scaled scores,

State and Trait portions of the STAI, and CES-D scores were

subjected to t-tests to test for differences between age

groups on these measures. Even when reducing significance

to control for the number of comparisons, the two groups

were significantly different from each other on all these

measures. The older group were more highly educated

(t(94)=3.19, p=.0019), achieved a higher score on the WAIS-R

71









vocabulary subtest (raw vocabulary, (t(89)=.15, p=.0001),

and age corrected scaled score, (t(94)=3.66, p=.0004),

reported less anxiety on both state and trait anxiety

measures (state, t(94)=3.06, p=.0029, trait, t(94)=2.97,

p=.0037) and reported fewer symptoms of depression on the

CES-D (t(94)=5.82, p<.0001).

These measures, as in Experiment 1, were for the most

part included to ensure that the older adults were not less

educated, less intelligent, more anxious, or more depressed

since all those factors could contribute to lower

performance on memory measures. As reviewed in the Results

section of Experiment 1, total recall and differences

between recall and recognition have been found to be

influenced in depression, while both total recall and

recognition are thought to be influenced by anxiety. These

age differences however, are likely to have minimized rather

than exaggerated any age differences in memory performance

found in this study.

Free Recall, Cued Recall, and Recognition

The number of items recalled and recognized by the two

age groups are shown in Table 4. A 2(Age Group) by 3(Test)

ANOVA was performed on this data to determine whether the

older and younger adult samples performed similarly to

previously reported studies with largest age differences on

tests of free recall, smaller but still significant

differences on tests of cued recall, and small or no age









differences on tests of recognition. This analysis

suggested that this sample did indeed follow this pattern.


Table 4

Means and Standard Deviations for Recall
Performance on Modified CVLT
(20 minute delay)


Free Recall Cued Recall Recognition

Young 14.43 14.79 19.06
(2.5) (2.5) (0.9)

Old 11.72 12.96 18.45
(3.7) (2.9) (1.5)


Although older adults had lower average scores on all

three memory measures, the ANOVA suggested a significant Age

by Test interaction (F(2,93)=10.09, p=.0003, w2=.02) and

significant main effects of Age group (F(1,94)=15.85,

p=.0001, w2=.134) and Test (F(2,188)=341.32, p=.0001,

w2=.76). Bonferroni t-tests, correcting for the three

comparisons, revealed that younger and older adults are

significantly different on free and cued recall after a 30

minute delay (Free recall: t(83)=4.22, p<.01; Cued Recall:

t(94)=3.32, p<.01). However, the two groups did not perform

differently on the recognition task (t(82)=2.38, p>.05).

Additional analyses, using Bonferroni t-tests, showed that

while both younger and older adults improved significantly

with recognition memory tests in comparison to both free and

cued recall (all p<.01), only the older adults were helped









significantly by cuing (young, t(47)=1.59, p>.ll, old,

t(47)=4.79, p<.01).

Effects of Exercise, Rest, and Distractor
Tasks on Measures of Arousal

The average effects of exercise and rest on pulse rate,

in beats per minute, and diastolic and systolic blood

pressure, in millimeters of Mercury, are presented in Table

5. Both absolute change and percent change from baseline

are presented. These data suggest that the exercise

manipulation had its desired effect both before the learn

trials and before final recall of the list.

All three physiologic measures increased after

exercise, but tended to decrease or remain about the same

after resting for five to seven minutes. Changes for the

young and old adults are also presented separately in Table

5. These data were subjected to two separate 2(Age Group)

by 2(Activity: Exercise or Rest) by 3(Physiologic Measures)

ANOVA's for both acquisition and test conditions, first

using the absolute change scores and then using the percent

change from baseline scores.

For the both acquisition ANOVA's, using the absolute

change scores there was a significant interaction between

Age Group and Acquisition Condition only for Systolic Blood

Pressure (F=(l,95)=6.76, p=.01, w2=.022). Significant main

effects for activity were found on all three physiologic

measures (Pulse Rate: F(1,95)=100.14, p=.0001, w2=.5065,

Systolic Blood Pressure, F(1,95)= 151.98, p=.0001, w2=.5971,












Table 5

Mean Change After Rest and Exercise Conditions


Response Measure


Learn Condition
Exercise


Rest


Test Condition
Exercise


Rest


Learn Condition
Exercise
Young


Old

Rest
Young


Old

Test Condition
Exercise
Young


Old


Rest
Young


Old


Pulse Rate
Chance

+14.9
(+21%)

+.02
(+.2%)


+14.1
(+20%)

-1.3
(-2%)



+16.1
(+23%)

+13.6
(+20%)

+1.1
(+2%)

-1.1
(-1%)


+17.1
(+24%)

+11.2
(+16%)


-0.1
(-.3%)

-2.3
(-3%)


Systolic B.P.
Change

+21.8
(+6%)

-7.3
(-6%)


Diastolic B.P.
Change

+3.7
(+17%)

-4.3
(-6%)


+20.1
(+7%)

-4.9
(-3%)


+4.1
(+17%)


-1.9
(-4%)


+17.5
(+8%)

+26.1
(+5%)

-5.4
(-6%)

-9.1
(-6%)


+20.2
(+9%)

+19.9
(+4%)


-4.0
(-3%)

-5.9
(-3%)


+4.8
(+15%)

+2.5
(+20%)

-4.5
(4%)

-4.0
(-7%)


+5.6
(+18%)

+2.7
(+16%)


-2.0
(-3%)

-1.9
(-4%)









Diastolic Blood Pressure: F(1,95)= 40.65, p=.0001, w2=.2945)

but there was no main effect for Age Group (.11 < p S .58).

The analysis using the percent change scores was essentially

identical with p values of the same magnitude and omega

square values differing by less than three percent.

An identical ANOVA using the change scores from Test

Condition revealed significant main effects of Test

Condition for all three physiologic measures (Pulse Rate:

F(1,95)=157.80, p=.0001, w2=.5946, Systolic Blood Pressure,

F(1,95)= 143.59, p=.0001, w2=.6015, Diastolic Blood

Pressure: F(1,95)= 32.81, p=.0001, w2=.2458) and a main

effect of Age Group for Pulse rate (F(1,95)=10.68, p=.015,

w2=.0367) was also found. There were no significant

interactions between Test Condition and Age Group. Again,

the analysis using the percent change data was essentially

identical, with similar p values and omega squared values

that differed by less than three percent.

These analyses suggest that the exercise manipulation

significantly increased physiologic arousal, while rest did

not. The main effect of Age Group and the Interaction

between Age Group and Arousal Level were not demonstrated

consistently in the physiologic measures and accounted for a

relatively small percentage of the variance compared with

the amount of variance consistently accounted for by

exercise condition. Young and old were both affected in the

predicted directions by the exercise manipulation, but when

there was an Exercise by Age Group or main effect of Age









Group effect, it was the younger adults who tended to show a

greater increase in physiologic arousal.

Because it has also been suggested that older adults

find the act of engaging in psychological tests more

arousing than do younger adults, this was also investigated.

Immediately after learning the word list, the three

physiologic measures were taken and subjects were asked to

complete the Self Assessment Mannikin (SAM) to subjectively

rate their feelings on the valence, arousal, and dominance

dimensions. Change scores were calculated for the three

physiologic measures and three SAM ratings and subjected to

t-tests with Age Group as the between subjects variable.

None of the tests of the physiologic measures suggested

there were significant differences between age groups (.14 <

p : .56), and only on the valence dimension did young and

old differ significantly on SAM (t(75)=2.88, p=.0052), with

younger adults reporting they were happier than older

adults.

From pre- to post- distractor items, the average pulse

rate declined 2.4 beats per minute (bpm) for young and

declined 1.27 bpm for the older adults. Systolic blood

pressure increased an average of 2.8 millimeters of Mercury

(mmHg) for young, and decreased 0.06 (mmHg) for older

adults. Diastolic blood pressure for both young and old

changed less than one mmHg. While younger adults rated

themselves an average of 0.04 points happier, older adults

rated themselves 0.625 points less happy (on a scale of









nine) after engaging in the distractor tasks. On average,

both young and old adults rated themselves as approximately

0.5 points less aroused, and 0.6 (young) and 0.8 (old)

points more dominant after engaging in the distractor task.

Overall, both objectively and subjectively this data

suggests that neither young nor old seemed to be

significantly affected either physiologically or emotionally

by the distractor tasks.

Exercise Condition and Memory

The first effect of interest was whether the exercise

versus no exercise manipulation prior to acquisition

affected recall, observed clustering, or cluster percentage

differently for the young and old. A 2(Acquisition State)

by 2(Age Group) by 3(Repetition) repeated measures ANOVA,

with time as the repeated measure, suggested that there was

no Repetition by Age Group by Acquisition State interaction

(F(2,91)=.078, p=.92) and no Age Group by Acquisition State

interaction (F(1,92)=1.5, p=.22). The Repetition main

effect was significant (F(2,91)=359.96, p=.0001, w2=.79)

with all subjects increasing the number of recalled items

over time. A main effect of Age Group was also found (F(l,

92)=10.84, p=.0014, w2=.0933) with older adults consistently

recalling fewer items than younger adults.

Two additional 2(Age Group) by 2(Acquisition State) by

3(Repetition) Repeated Measures ANOVA's were conducted on

the Observed Cluster scores and Cluster Percentage Scores.

For both these analyses, there were no interactions that









even approached significance. Again, the Repetition main

effect was significant (Cluster Observed, F(2,91)=92.23,

p=.0001, w2=.4892, Cluster Percentage, F(2,91)=26.81,

p=.0001, w2=.2079). Both Cluster Observed and Cluster

Percentage scores increased over time.

The second effect of interest was whether a "state

dependent" learning effect had been demonstrated and whether

this effect differed between age groups. To test this, data

were reorganized and each subject's data were placed in one

of two cells depending on whether their activities before

acquisition and memory testing was the same or different.

"Same" subjects were subjects whose activities prior to list

acquisition and memory testing were either exercise or rest

(i.e., rest-rest or exercise-exercise). "Different"

subjects were those whose activities prior to learning and

memory testing were different (i.e., rest-exercise or

exercise-rest). These data are presented in Table 6.



Table 6

Encoding/Test Correspondence

Same Different

Group Free Rcl Cluster Clust % Free Rcl Cluster Clust %

Younger 14.75 6.95 63.09 1 14.13 5.21 48.79

Older 12.92 5.63 57.70 10.54 3.88 45.21


After being organized in this way, Free Recall,

Observed Clustering, and Cluster Percentage data were each









subjected to separate 2(Age Group) by 2(Learn-Recall

Correspondence) Repeated Measures ANOVA's. All three

ANOVA's had p values less than .02 for the models, but there

were no significant interactions. For both Delayed Recall

and Clustering Observed there were main effects for both Age

Group (Delayed Recall: F(1, 92)=18.99, p=.0001, w2=.1502,

Cluster Observed: F(1,92)=4.05, p=.027, w2=.0379) and Learn-

Recall Correspondence (Delayed Recall: F(1,92)=5.82, p=.018,

w2=.0403, Cluster Observed: F(1,92)=8.7, p=.004, w2=.0721).

However, when Clustering Scores were corrected for the

number of items recalled in the Cluster Percentage analysis,

Age Group as main effect was not significant (F(1,92)=.99,

p=.3218, w2<.0001), but Learn-Recall Correspondence remained

significant (F(l,92)=8.84, p=.0038, w2=.0762). This

analysis suggests that while older adults recalled fewer

items than younger adults after a delay, both younger and

older adults tended to recall more items if they were asked

to recall the items after engaging in the same activity as

during the learn phase. Clustering appeared to be

influenced by age only when not corrected for the total

number of items recalled, but both cluster scores tended to

be higher when subjects had engaged in the same activity

immediately before learning the list and immediately before

recalling the list. This suggests both a state dependent

learning and state dependent clustering effect.















EXPERIMENT 2: DISCUSSION


Just as in Experiment 1, anxiety and depression

measures were included to ensure that older adults were not

more anxious or more depressed than younger adults, since

those conditions have been suggested to decrease memory

performance. In this sample, it was also the younger adults

whose responses suggested higher levels of both anxiety and

depression. In addition, the older adults tended to have

slightly higher levels of education and achieved higher

scores on the WAIS-R Vocabulary subtest. These differences

were not judged to be problematic in this study since these

factors would tend to minimize, rather than maximize, age

differences. Despite this, the previously described

patterns of age differences where still found (i.e., largest

age differences on tests of free recall, older adults

benefitting more from cuing than younger adults, and minimal

or no differences found on recognition tests).

Under-arousal theories of aging and memory would have

predicted that the older adults would have benefitted from

the increased arousal associated with exercising prior to

learning the word list. There also had been some

suggestions that increased arousal is associated with

81









decreased semantic clustering (Woods, 1975). In this study,

despite physiologic evidence of increased arousal from the

exercise, neither young nor old appeared to benefit from

either rest or exercising immediately prior to learning the

word list, but consistently showed age differences in which

older adults recalled fewer list items than did younger

adults. Both younger and older adults increased the number

of items recalled with each successive exposure to the list,

and showed increased semantic clustering with repetition.

The fact the semantic cluster percentage also increased

after each repetition suggests that this increase cannot be

attributed to increased overall recall levels, but suggests

increased list organization with increased exposure. This

increased organization after each repetition is evident with

both younger and older adults.

The discrepancy between the results of this study and

previous research may be due to differences in how arousal

was manipulated and to the degree in which the to-be-

remembered word list could be organized for later recall.

The results may be different if cortical rather than just

physiological arousal is altered since it is likely that

cortical arousal does not influence memory in the identical

way physiological arousal influences memory performance. It

is also possible that the use of semantic organization

strategies influences memory more strongly than does arousal

state and the effect of arousal could be seen if the list

was not organizable into semantic categories.









State dependent learning effects have been demonstrated

repeatedly with younger adults, but have not been

demonstrated previously with older adults. This study

successfully showed that, for both young and old, being in

the same general state of physiological arousal immediately

prior to acquisition and immediately prior to recalling a

list enhances not only learning, but also the amount of

semantic clustering subjects demonstrate. No age

differences or interactions were found, which is consistent

with Hess and Higgin's (1983) study in which older adults

appeared equally competent as younger adults in using

general context to enhance memory performance.

Being in the same state at recall also seems to enhance

semantic clustering for both younger and older adults. If

this was due simply to the increase in number of items

recalled, this increase should disappear in the semantic

cluster percentage (which corrects for the number recalled),

but this was not the case.















GENERAL DISCUSSION


This paper began by describing behavioral changes in

memory and arousal that have been associated with normal

aging, reviewing how memory and arousal are believed to

interact in normal young adults, reviewing the brain areas

believed to be important in memory and arousal, and

suggesting what brain changes associated with aging might

influence this interaction in older adults. Arousal is a

complex phenomenon, comprised of both the background (tonic)

activity and the phasic (stimulus linked) reaction to an

event. The two studies described above examine both phasic

and tonic arousal in an effort to determine if either or

both appear to influence memory differently in older adults.

In these studies, differences between younger and older

adults were found in how phasic (event related) arousal

influenced memory. While a distinctive stimulus was

remembered better by both younger and older adults, memory

for surrounding items was only disrupted in younger adults.

Additionally, a stimulus that was more disturbing to younger

adults and resulted in more profound amnesia (compared to

another distinctive but less disturbing stimulus) did not

affect the memory of older adults at all. Tonic arousal, or

84









the state of an individual, seemed to influence older and

younger adults' memories similarly, since both groups

appeared able to use correspondence between acquisition and

retrieval state as a cue to assist in recall. State

dependent learning was present in younger and older adults,

with being in the same state at learning and recall

enhancing both free recall and semantic clustering in both

age groups.

The results of these two studies do not support a

simple over-arousal theory to explain the differences in

young and old on memory tasks. Increasing arousal did not

interfere with memory performance in old adults in

Experiment 2, as one might predict if the older adults were

already over-aroused, although it is possible that even

higher levels of physiologic arousal or alterations in

cortical arousal may influence the two groups differently.

Findings from Experiment 1 would also argue against an over-

arousal interpretation since fewer older adults remembered

the arousing distinctive stimulus. Secondly, memory for

surrounding items was less, not more, disturbed for older

adults compared to younger adults, despite speculation that

increased arousal plays an important role in induced amnesia

for surrounding items.

Instead, these data could be used to support an under-

arousal explanation for differences between young and old in

memory performance. Although increasing physiologic arousal

also failed to enhance memory in younger adults, it is









possible that increasing cortical rather than physiological

arousal might enhance memory. Since it is impossible to be

certain where on the Yerkes-Dodsen curve the experimental

manipulations sampled when only two points are sampled, it

is also possible that arousal was either not increased

enough or increased too much to facilitate memory. The

lower rate of recall of the distinctive stimulus by the

older adults could be explained by the under-arousal

hypothesis or could be due to a smaller reaction to what is

distinct in the older adults. However, this age difference

was only a trend and not statistically significant. Older

adults were also more likely to recall the more disturbing

or distinctive of the two critical items, perhaps suggesting

that they needed higher arousal, compared to younger adults,

in order to remember items at the same level. The fact that

amnesia for surrounding items was not induced, may also

support the notion of lowered arousal in the older adults.

In this instance, it appears to be advantageous for older

adults to be less highly aroused than the younger adults

since their memory for surrounding items was not disturbed,

while younger adults' memories were disrupted. The older

adults lower reported symptoms of anxiety may also reflect

their lower level of tonic arousal.

There were no age differences in the ability to

autonomically discriminate between targets and distractors

in the context of the recognition task. Both younger and

older reacted more strongly to targets than to distractors









in a recognition task, and for both groups, responses to

targets were larger regardless of whether the subjects

overtly recalled the items. It has been suggested that this

phenomenon may be another measure of implicit memory or

memory without awareness (Bauer, 1984). Previous research

had suggested that some forms of implicit memory in young

and old adults differ (Chiarello & Hoyer, 1988); the results

of this study suggest this is an area in which young and old

do not differ. It appears that young and old are equally

able to use or ignore physiologic response to stimuli as a

cue for recall, since both patterns of response and overt

recognition were not significantly different in these two

populations.

Although it is difficult to be certain how the patterns

of behavior in this study reflect age-related brain changes,

it is possible to speculate based on the literature reviewed

previously. The amygdala was suggested as important in the

regulation of arousal related to the input of information,

while the basal ganglia is associated with differences in

levels of activation. The results of Experiment 1 may

reflect the age-related changes documented in these two

areas since conditions at input appear to influence memory

differently in the younger and older adults, and the most

likely explanation for this is differences in the amount of

activation or processing each group engages in reaction to

the two distinctive stimuli. The hippocampus is believed to

be important in the mediation of arousal and memory and in









determining what is memorable. It is possible that these

studies reflect different hippocampal functions being

variably affected by age. State dependent learning, which

could be considered part of the mediation of memory and

arousal function, is still intact as demonstrated in

experiment 2. Age-related hippocampal differences may be

reflected in Experiment 1 with regard to deciding whether

the CE is memorable relative to surrounding items.

Frontal areas and the orienting response may be changed

with increasing age and this change may be related to the

differences in phasic arousal effects. Changes in the

frontal areas also may be suggested if the failure of older

adults to find the distinctive stimuli as memorable is

attributable to difficulties in releasing from proactive

interference (i.e., older adults are rigidly expecting line

drawings, which would lead to problems in processing the

photograph). One possible explanation for the older adult's

failure to show the same pattern of memory performance as

younger adults in Experiment 1 is that the older subjects

found the Critical Item less distinctive and/or are

distracted from fully processing the Critical Item by the

surrounding items. Either of these explanations could be

related to the frontal lobe changes that Albert and Kaplan

(1980) suggest are at the heart of age-related cognitive

changes.

Age-related changes in the influence of phasic but not

tonic arousal on memory suggest that the previously cited









age-related brain changes in the hippocampus and frontal

cortex may have significant effects on how memory and

arousal interact in older adults. Lesions in the reticular

formation have been most closely associated with changes in

tonic arousal in animals, and this brain area has not been

suggested as particularly affected by aging in human autopsy

studies. The results of this study are consistent with this

since the aging process less strongly affects tonic than

phasic arousal. In these two studies age differences were

found only in how phasic arousal influences memory; this

behavioral evidence is consistent with the physiological

findings of these studies.

It is difficult to be certain which brain areas are

specifically affected by aging and how these changes

influence behavior. However, this study is consistent with

previous psychophysiological, physiological, and

psychological studies which implicate frontal lobe changes

and changes in the hippocampus as important in age-related

memory change. To help determine whether the differences

seen in this study between young and old subjects were due

to diffuse brain changes or more specific focal lesions,

such as frontal lobe lesions as suggested by Albert and

Kaplan (1980), repeating experiment one using frontal lobe

patients compared to patients with more diffuse damage, such

as closed head injury patients might be illuminating.

These results also have implications for psychological

theories that have been invoked to explain behavioral age









differences. Salthouse (1985) has argued that age-related

general slowing is the most parsimonious explanation for

many differences seen between young and old, including

differences in memory. However, if it is merely slowing

that changes memory, in Experiment 1 it would be predicted

that retrograde amnesia would be more profound in older

adults because they should not have completed encoding of

the previously exposed stimulus by the time the distinctive

slide was shown. It would also be predicted that older

adults would take longer to recover, increasing anterograde

amnesia. Neither of these results was seen here, suggesting

that slowing is not a sufficient explanation for these

behavioral differences.

The role of organization and semantic clustering of

information as an explanation of age differences is less

clear in this study. Although a previous study found that

increased arousal with white noise decreased semantic

clustering (Schwartz, 1975), increasing physiologic arousal

through exercise in this study did not influence the

organization of to-be-remembered material. Differences

between young and old were found in the amount of semantic

clustering, but only when the cluster scores were not

adjusted for the total number of items recalled. When

semantic cluster scores did reflect the number of items

recalled, age differences dropped out, which may call into

question whether the differences in clustering cause the

lower memory performance. In addition, although no









instructions to cluster were given during this study, both

young and old increased clustering over time as their recall

increased.

The similarities found in the ability of young and old

to use state as a cue during recall could be seen as

consistent with Hess' (1984) findings that older adults tend

to show more general rather than specific encoding of

information. If the state one is in is viewed as a more

general cue, and this cue was present during encoding, one

would predict no age differences in the ability to use this

information during retrieval. If the phasic changes are

conceptualized as more specific cues, this would also have

predicted greater age differences in the ability of the

older adults to use this as a cue during retrieval.

It is possible that the differences seen between young

and old in their reactions to the Critical Event may be

unique to the Critical Event stimuli chosen or the modality

chosen to test. To examine for these possibilities

additional studies using older adult groups with different

distinctive stimuli would be helpful. Since this pattern of

memory performance is easy to replicate with younger adults

despite variations in stimuli, it is unlikely that a

stimulus change will significantly alter the age differences

and it is most likely that these findings will be

replicated. Replicating this design using verbal rather

than visual stimuli to examine modality specific differences









could be more illuminating since modality has been shown to

be an important variable in influencing age differences.

Together, these two studies illustrate ways in which

memory is both different and stays the same across the

lifespan. Although most studies show how age-related memory

changes are detrimental to overall memory performance,

Experiment 1 provides an example of how the differences in

young and old result in the older adults not losing

information relative to younger adults. Additionally, while

a CE or distinctive stimulus and the surrounding items may

be processed differently by young and old, both young and

old appear to be equally influenced by their reaction to

targets and distractors during recognition and their ability

to use their state of arousal to assist in recall. This

argues against the notion that age-related memory

differences can be attributed solely to differences in

retrieval processes. Overall these studies suggest more

questions about the nature of aging and how it may influence

the interaction between memory and arousal than it answers.

However, it also seems to clearly show that arousal and

memory do not interact in exactly the same way in younger

and older adults and that we may be exploring an important

factor in memory and aging.















APPENDIX A: SCREENING QUESTIONNAIRE


Subject Name:

DOB:


Education Completed:


1) History of head injury?(LOC?)






2) Neurological Problems? (Epilepsy, stroke, periods of

numbness or tingling or visual disturbances suggestive of

TIA)






3) Psychological problems? (Medication or counselling for

depression, psychiatric hospitalization)






4) History of learning disability?









5) History of heart disease?






6) Seeing a physician for any medical problems?






7) Any problem with moderate exercise (e.g. walking) for

five to ten minutes?






8) How much exercise in a typical week?






9) Visual or hearing problems that interfere with reading or

hearing normal conversation?




10) Remind subjects to bring or wear glasses or hearing aids

if they wear them. Provide directions to the lab.




Full Text
xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EM8HLWF86_6R9PBQ INGEST_TIME 2012-02-20T21:26:20Z PACKAGE AA00009076_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES