h

_,
g n;
3'
é

L

«3's 2

. .

.H:

4‘
1
i

I
to"
(1;
3%!

J 3‘

.

it
3E
+
1
‘8‘.
sh.

z
$.dnmwphas
£ﬁ§c .,
2:. 3 ..
54.},
3. i .
. «.5...
not in)...»
. 63%..-.35:
n’ll‘Q-n“=r|3 30“?

.IIQI

JI\'=I¢I.
1:925... :2. a
1 4 .. .
“J1 9:2. to.

, . .

, ..

f...
.ua.-n_..»..x
try-.6.
4.5.... 11.9... .

H . , 3&2
.. mmrgda
Wu“ . ..
’,'I.§¢ I. 7
.ﬁhﬂf gum. .izlmww.
nrlzﬁ. Vang...”
335%. “may,
.1 nu ’ a. "‘P‘Iho‘. .us
. 3.2: .3
:
aiﬁhﬁﬁnafﬂn . kg”...
,

Jun I

. a.» A: I
a .7: H59... .kts:
vitrifii $1. is!!! 3...:
nf,._.....:lr.. .u‘..!rv.idhxun
a (I 13.
that-19H». .{rsxurisa.~
anti! 3L.
.. . .

n 2. ‘

g ...
an 1?.»-
3 2...:

 

Akita-(8.]. 42: .I
)r. A»):

\
sn.ix..<{

1r affil, .
. x...-

 

”52's llllllllllllllllHIUHHNIlllllHNIIHIIIIHIHINllllllll

2000 31293 02106 1852

LIBRARY
Michigan State
University

 

 

 

 

This is to certify that the

dissertation entitled

The Effects of Anxiety on Memory Performance
in the Able Elderly

presented by

Jodi Levy-Cushman

has been accepted towards fulﬁllment
of the requirements for

PhD degree in PSYChOIng

MM

Major professor

 

Date July 10, 2000

 

MS U is an Afﬁrmative Action/Equal Opportunity Institution 0-12771

PLACE IN RETURN BOX to remove this checkout from your record.
To AVOID FINES return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

11/00 chlRCJDateDw.p65-p.14

THE EFFECTS OF ANXIETY ON MEMORY PERFORMANCE IN THE ABLE
ELDERLY
By:

Jodi Suzanne Levy-Cushman

A DISSERTATION

Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Psychology

2000

ABSTRACT
THE EFFECTS OF ANXIETY ON MEMORY PERFORMANCE IN THE ABLE
ELDERLY
By

Jodi Suzanne Levy—Cushman

Anxiety, depression, and memory difﬁculties are probably the most common
mental health concerns facing the elderly today. Differentiating anxiety and depression
has been a long-standing problem in psychopathology research (Nelson & Novy, 1997;
Alexopoulos, 1991). The ﬁrst aim of this study was to differentiate the symptoms of
anxiety and depression in community dwelling older adults (N= 47, Mean age = 71.5).
State anxiety as measured by the STAI-S did not correlate signiﬁcantly with either
depression measure. However, trait anxiety as measure the STAI-T (BDI ;' = .342, p <
.05; GDS r = .387, p < .01) and worry as measured by the Penn State Worry
Questionnaire (BD1 :3 = .411, p < .01; GDS r = .458, p < .01) correlated signiﬁcantly with
depression.

It is a commonly held belief that our affect inﬂuences our cognitive abilities. The
cognitive sequelae of depression in the elderly have attracted great attention (LaRue,
1992); however, the inﬂuence of anxiety on cognitive performance and cognitive
complaints is extremely understudied in, older adults (Salzman & Lebowitz, 1991). This
study examined the inﬂuence of anxiety on memory performance. Before the participants

completed the memory tasks they heard manipulation instructions in an effort to increase

anxiety (anxiety instructions) or decrease anxiety (relaxation instructions). Results
suggest that hearing the anxiety evoking instructions increased anxiety and hindered
memory performance. However, overall state anxiety was not a signiﬁcant predicator of
memory performance. On a more speculative level, although the correlations between
state anxiety and memory performance and the regression models were not statistically
signiﬁcant, there was a consistent negative relationship between state anxiety and
memory performance; does this suggests that state anxiety hinders memory performance?
Notably, this negative relationship was largest on the most complex task (free recall) and
smallest on the least complex task (recognition). It was also predicted that there would be
a differential relationship of strategy use as a function of anxiety with optimal strategy
use occurring at intermediate levels of state anxiety. This hypothesis was not conﬁrmed;
strategy use was not signiﬁcantly correlated with state anxiety levels.

This study also allowed us to investigate the combined effects of state anxiety,
worry and trait anxiety on performance. We expected trait anxiety and state anxiety to
interact to predict memory performance. Contrary to these predictions, individuals who
experience state anxiety at the time of testing and who have high levels of trait anxiety
did not experience exaggerated negative effects of anxiety on memory performance.
Finally, the present study examined the effect of anxiety on memory complaints in a
sample of able elderly. Trait anxiety as measured by the STAI-T correlated signiﬁcantly
with memory complaints (1 = -.411, p.< .01); however, state anxiety and worry (as

measured by the PSWQ) did not correlate signiﬁcantly with memory complaints.

ACKNOWLEDGEMENTS

This work could not have been completed without the support, encouragement,
and guidance of many people. First, I would like to thank my advisor and dissertation
chairperson, Dr. Norman Abeles, whose tireless efforts on behalf of me made this
possible. Without Dr. Abeles’ unﬂagging patience, insightful feedback, constant support
and encouragement I could not have succeeded in graduate school and develop as a
researcher, clinician and a person. Dr. Lester Hyman taught me many lessons on how to
integrate clinical and cognitive theory, and (more importantly) believed in and supported
me throughout my graduate career. Dr. Ray Frankmann generously gave his time day or
night, gently guiding me through the statistics necessary for this study. Many thanks also
go to Dr. Michael Lambert who offered many helpful comments throughout the
completion of this project.

I have also been extraordinarily fortunate to have many friends whose love and
support has help me endure some very challenging times. I would particularly like to
acknowledge Angela McBride for listening, sharing, caring and helping me ﬁnd balance
in my life. Camilla and Carrick Williams for adopting me and making me feel like I had
family just across the ball, when mine was many miles away (thanks also go to Carrick
for helping me understand inhibitory processes). Heather Miles for being the other half of
my brain and for the many hours of statistical consultation throughout the completion of
this project.

I would also like to thank my Mom and Dad who have devoted their life to my

personal growth and happiness. Finally, I’d like to thank my husband, Matt Cushman, for

iv

believing in me when I didn’t believe in myself. Your love is my most treasured gift.

Here’s to ﬁnally beginning our life together!

TABLE OF CONTENTS

LIST OF TABLES ...................................................................................................... vii
LIST OF FIGURES .................................................................................................... viii
INTRODUCTION ...................... q ................................................................................. 1
Anxiety ............................................................................................................. 3
Depression ........................................................................................................ 8
The Relationship Between Anxiety and Depression ........................................ 8
The Effects of Anxiety and Depression on Cognition ................................... 13
Eysenck’s Theoretical Overview of Anxiety ................................................. 14
The Yerkes-Dodson Law ................................................................................ 17
A Review of Studies of Anxiety on Memory Performance in the Elderly ..... l9
Trait Anxiety .................................................................................................. 23
Memory Complaints ....................................................................................... 26
Hypotheses ..................................................................................................... 29
METHOD ................................................................................................................... 32
Participants ..................................................................................................... 32
Pilot Participants ............................................................................................ 33
Procedure ...................................................................................................... 33
Measures ........................................................................................................ 36
RESULTS ................................................................................................................... 41
Pilot Study ...................................................................................................... 41
Hypothesis I .................................................................................................... 43
The effects of the manipulation instructions .................................................. 44
Hypothesis H & HI ......................................................................................... 49
Hypothesis IV ................................................................................................. 54
Hypothesis V .................................................................................................. 55
Hypothesis VI ................................................................................................. 57
DISCUSSION ............................................................................................................ 58
The Relationship Between Anxiety and Depression ...................................... 59
The Manipulation ........................................................................................... 64
The Effects of State Anxiety on Memory Performance ................................. 66
Strategy Use ................................................................................................... 72
Trait Anxiety .................................................................................................. 73
Memory Complaints ....................................................................................... 74
Limitations of Study ....................................................................................... 76
Future Research .............................................................................................. 77
LIST OF REFERENCES ............................................................................................ 91

vi

LIST OF TABLES

Table 1

Coorelations between anxing and depression

Table 2 - pg. 80

Rg_t_ated Component Mgtn'x- F gctor Load_ing§ for Depression and Anxiety
Measures

Table 3 -

CVLT performgnce means by order and condition

Table 4 - pg. 86

Pearson Correlations Between Memog Performance and State Anxiety
Table 5 - pg. 89

Pearson Correlations Between Free Recall Memog Performance and Anxieg

vii

LIST OF FIGURES

Figure 1 - pg. 81

Relationship between state anxieg and free recall performance at the lst manipulation

 

Figure 2 - pg. 82

Relationship between state anxing and free recall performance at the 2nd manipulation

Figure 3 -pg.83

Relationship between state anxiety and cued recall performance at the lst manipulation

Figure 4 - pg.84

Relationship between state anxiety gpd cued recall performance at the 2nd manipulation

 

Figure 5 - pg. 85

Relationship between state anxiety and recogpition performance

 

 

Figure 6 - pg.87

Relationship between state anxiety and semantic cluster strategy use

 

Figure 7 -pg. 89

Relationship between state anxiety 3nd serial cluster strategy use

viii

INTRODUCTION

Every age has a dominant emotion. The ﬁrst half of this century
was the Age of Anxiety, and its emotional tone was captured by Sigmund
Freud. Freud lived through the death throes of the Hapsburg Empire, and
then though the horrors of World War I and its chaotic aftermath. Freud
watched a world order that had stood for hundreds of years dissolve and a
new one struggle to be born. Times when all values crumble and the
ﬁrture is unpredictable are times fraught with anxiety. Anxiety was the
dominant emotion Freud saw in his patients, and it was the dominant
theme of contemporary writing, ﬁlm and painting. Small wonder that
Freud believed that all neuroses and almost all human action stemmed
from anxiety. All the other emotions — depression, awe, anger,

embarrassment, shame and guilt -—- were just footnotes.

— Martin Seligrnan, in What You Can Change and Whpt You Ca_r_n_:t:_

The Complete Guide to Successful Self Improvement

According to recent population estimates, there are over 51 million people in the

United States (21% of the total population) who are at least 55 years of age (US Bureau

of Census, 1987 as cited in LaRue, 1992). By 2030, it is projected that one in every three

Americans will be 55 years old or older. This age group is by far the fastest growing

segment of the population (T ranel, Benton, & Olson, 1997). With the growing number of

older adults, the importance of understanding aging and common disorders of later life is
difﬁcult to exaggerate.

An increasing number of healthy older adults are seeking advice and treatment for
cognitive problems (La Rue, 1992). In particular, memory difﬁculties are often cited
(Williams, Denney & Schadler, 1983). Memory decline with age has become such a
common concern among older adults that researchers and clinicians who work with
geriatric patients have recommended the development of a new diagnostic entity, "Age
Related Cognitive Decline" (ARCD) to be included in the Diagnostic and Statistical
Manual of Mental Disorders (Caine, 1993). The criteria proposed for ARCD includes
both memory dysfunction and subjective complaints (Crook, Bartus, Ferris et al., 1986).

Complaints of anxiety and depression are also encountered more frequently in
this age group (Zarit & Knight, 1996). Symptoms of depression can be found in 15% of
community dwelling older adults (Blazer, 1989). In addition, generalized anxiety disorder
and other anxiety disorders including phobias, obsessions, and compulsions are also
among the most common problems reported in later life (Blazer, George, & Hughes,
1991). However, differentiating anxiety and depression at the mood, symptom, and
syndromal levels has been a long-standing problem in psychopathology research (Nelson
& Novy, 1997; Alexopoulos, 1991). Differentiating anxiety and depression is important
because comorbid conditions like anxiety and depression may or may not be functionally
related, and determining the presence or non-presence of a relationship is a crucial aspect
of treatment planning (Clarkin & Kendall, 1992).

It is also a commonly held belief that our affect inﬂuences our cognitive abilities.

The cognitive sequelae of depression in the elderly have attracted great attention (LaRue,

1992). Eysenck (1979) stated that anxiety also clearly affects the amount of memory
capacity available for task performance. However, the inﬂuence of anxiety on cognitive
performance and cognitive complaints is extremely understudied in older adults (Salzman
& Lebowitz, 1991). Several investigations have been conducted on the relationship
between anxiety and sustained attention and memory in younger adults but there is very
little research of this type with older populations (Rankin, Gilner, Gfeller, Katz, 1994). It
is therefore essential that we investigate the inﬂuence of anxiety on memory performance

and memory complaints in older adults.

Alli—icy

Anxiety can cause signiﬁcant distress. Gurian and Miner (1991) have deﬁned
anxiety as “a subjective state of internal discomfort, dread and foreboding, accompanied
by nervous system arousal. Different from fear, anxiety tends to occur without conscious
or apparent stimulus.” (pp.34) There are eleven DSM-IV categories speciﬁc to anxiety in
adults: panic disorders, agoraphobia, speciﬁc phobias, social phobias, obsessive-
compulsive disorder, post-traumatic stress disorder, acute stress disorder, generalized
anxiety disorder, anxiety disorder due to a general medical condition, substance-induced
anxiety disorder, and anxiety disorder not otherwise speciﬁed.

There is a continuum of anxiety levels. If anxiety is mild, it can be helpful in
planning and coping (Salzman, 1991). On the other hand, in more extreme forms it is
characterized by dread, anticipation of threat, fear, extreme worry, and preoccupation
with the future. When anxiety causes these symptoms it can interfere severely with daily

living (Salzman, 1991). Individuals with moderate to severe levels of anxiety can

experience cognitive, emotional, and physical symptoms. The cognitive symptoms of
anxiety include decreased concentration, memory, and attention. Feelings of dizziness,
impending faintness, heart palpitations, and sensations of “going crazy” are also
associated with anxiety. Difﬁculty falling asleep and changes in appetite are almost
always present. Symptoms of severe anxiety may be similar to those associated with
gastrointestinal, cardiovascular, endocrine, or neurological illness. Symptoms such as
nausea, abdominal burning, belching, constipation, and diarrhea can be signs of anxiety
as well as gastrointestinal disease. Tightness in the chest, breathing difﬁculties and
tachycardia are often misinterpreted as early signs of a heart attack. Poor memory, poor
attention and poor concentration may be feared by the anxious person as symptoms of
Alzheimer’s disease (Salzman, 1991).

The term “anxiety” is currently used to refer to two different but related
constructs. It is often used empirically to describe a distressing emotional state. It is also
used to describe a personality trait (Spielberger, Gorsuch & Lushene, 1970). Cattell and
Scheier ﬁrst introduced the concepts of state and trait anxiety in 1961. State anxiety
refers to transient feelings of tension, apprehension, nervousness, worry, and activation or
arousal of the autonomic nervous system. Trait anxiety refers to more enduring anxiety
levels and anxiety proneness. That is, individual differences in the tendency to perceive
situations as stressful and dangerous. Trait anxiety may reﬂect the probability that an
individual will experience state anxiety in a given situation; for instance individuals who
are high in trait anxiety are more likely to experience elevations in state anxiety when
exposed to threatening situations (Spielberger, Gorsuch & Lushene, 1970).

Do older adults experience anxiety? There is universal agreement among geriatric

psychiatrists that anxiety exists both as a disorder and a symptom in the elderly
(Shamoian, 1991). There have been many studies that have reported high prevalence rates
of anxiety disorders and anxiety symptoms in older adults. In recent epidemiology studies
5% of adults over 65 were determined to have diagnosable anxiety disorders (Flint, 1994;
Regier, Boyd & Burke, 1988). One such study found phobias to be the most commonly
reported disorder in women over sixty-ﬁve, and the second most prevalent disorder for
men over sixty-ﬁve (Myers, Miller, Metzger, & Borkovec, 1984). In another
epidemiological study, Lindesay, Briggs and Murphy (1989) found a 3.7 percent
prevalence rate for generalized anxiety disorder and 10 percent prevalence rate for phobic
disorders in individuals over sixty-ﬁve.

However, anxiety symptoms that do not meet the criteria for speciﬁc syndromes
occur in much larger percentages of the population. Hirnmelfarb and Murell (1984)
reported that 17.1 percent of males and 21.5 percent of females experience symptoms of
anxiety that do not meet clinical levels but are severe enough to warrant some form of
therapeutic intervention. In a nationwide study of 2,460 adults living in private housing
Gurin, Veroff and Feld (1983) found that 10% of elderly adults were anxious; this is
twice the rate in younger adults. Researchers have shown that the rates of anxiety
symptoms increase linearly with age, with rates double that of adolescence observed in
later life (Warheit, Bell, Schwab, & Buhl, 1986).

Studies documenting the distribution of benzodiazepines (minor tranquilizers)
reported rates ranging form 17% to 50% in older adults (Salzman, 1991). In another study
Parry, Baltes, Mellinger, Cisin & Manheirner (1973) found that 19.4% of persons over

the age of 65 had used minor tranquilizers or sedatives in the year preceding the

interview. This makes older individuals the largest consumers of anti-anxiety
medications.

On the other hand, several studies have reported that current and life time
prevalence of anxiety disorders and anxiety symptoms decrease with increasing age.
Blazer, George and Hughes (1991) found the prevalence of anxiety disorders was higher
for the 45 to 64 age group than the 65+ age group. Furthermore, they reported lifetime
prevalence was also lower for the 65+ age group. Other researchers have reported similar
ﬁndings: lower rates of anxiety in later life (Meyer et al., 1984; Robins et al., 1984).

There are many reasons why studies may disagree about the prevalence rates of
anxiety in the elderly. Older adults are thought to under-report anxiety symptoms.
Speciﬁcally, Blazer et a1. (1991), suggested that a higher threshold of discomfort may be
required for reporting of symptoms in later life. Older adults might attribute anxiety
symptoms to medical illness, whereas younger adults might be more prone to report their
symptoms as anxiety. In addition, the somatic manifestation of anxiety makes its
diagnosis difﬁcult in later life. Anxiety is often confused with common geriatric illnesses,
and many anxious adults seek treatment from their internist rather than from a
psychologist or psychiatrist (Katon, 1986). Older adults have different concerns and the
source of the symptoms might therefor be shifted. Older persons also tend to attribute
agitation, fears, and aches and pains to the normal aging process and so deny and under
report anxiety (Gruian and Milner, 1991). Anxiety and depression are often confused in
older adults as the overlap of signs and symptoms makes it difﬁcult to distinguish
between the two syndromes (Alexopoulos, 1991). These factors suggest that the

prevalence rates which have been reported for anxiety disorders, and especially anxiety

symptoms, might be underestimates of the true number of cases.

Elderly women appear to be afﬂicted by late life anxiety more than men. Blazer,
George, and Hughes (1991) reported that Murphy and her colleagues (1986) found the
prevalence rate of generalized anxiety disorder to be 6% in women over forty ﬁve and
only 1% in men over 45 years of age in the Stirling County Study. Furthermore, she
reported that the rates of Generalized Anxiety Disorder increased for women as age
increased but declined in men. Using the Spielberger State-Trait Anxiety Scale,
Himmelfarb and Murrell (1984) also found that anxiety occurs more frequently and
severely in women than in men.

There are many reasons why older adult experience anxiety. For some later life
brings retirement and relaxation. For others it can be a period of extreme loneliness,
worthlessness, and hopelessness (Smith, Sherill, & Colenda, 1995). Elderly adults may
experiences multiple losses that contribute to sadness and anxiety as well as fear of death.
Chronic illness, ﬁnancial limitations, and diminished ﬁmctioning are also likely to take
their toll on older adults. Other common causes of anxiety in later life include late onset
neurotic disorders, life-long recurrent anxiety, late onset acute phobic anxiety,
hypocondraisis, early dementia, paranoid psychotic states, obsessional states, withdrawal
reactions, medicine and caffeine (Shamoian, 1991). Despite the high prevalence rate,
distressing affect and intense worry, anxiety has been relatively neglected and extremely
understudied in older adults (Salzman & Lebowitz, 1991). Depression is also amoung the

most common concerns of older adults group (Zarit & Knight, 1996).

Depression

Depression is generally associated with despondent mood and/or loss of interest
and pleasure in nearly all activities. Unresolved grief, guilt, loneliness, and anger with
symptomatology including insomnia, despair, lethargy, anorexia, loss of interest, and
somatic complaints are often experienced by older adults (Butler & Lewis, 1977).
Depression may refer to a mood, a symptom, or a disorder (Wasylenki, 1987). Depressed
mood refers to feelings of sadness, disappointment, discouragement and related emotions.
The symptom of depression refers to changes in affect that are associated with certain
events, medical illness, drug effects or other psychological disorders. The disorder of
depression can have cognitive, emotional, vegetative, and motor components.

The rate of depression is thought to increase with age (Zarit & Knight, 1996) but
some recent studies have found the opposite (Regier et al., 1988). Older adults may be
more prone to milder forms of depression and less prone to certain severe depressive
syndromes (LaRue, 1992). Prevalence rates of depressive disorders occur in less than 5%
of individuals over sixty-ﬁve (Friedhoff, 1991, Gallo & Lebowitz, submitted). However,
symptoms of depression can be found in 15% of community dwelling older adults
(Blazer, 1994). Rates of depression reach levels of 15% to 25 % among nursing home

residents and those with chronic physical illness (F riedhoff, 1991).

The Relationship Between Anxieg and Depression
Differentiating anxiety and depression at the mood, symptom, and syndromal

levels has been a long-standing problem in psychopathology research (Nelson & Novy,

1997; Alexopoulos, 1991). Three positions have been taken in dealing with the

differentiation of anxiety and depression (Alexopoulos, 1991). Anxiety and depression
have been viewed as part of a continuum with both being symptoms of one disorder: the
unitarian position. The opposite of this position suggests that anxiety disorders and
depressive disorders are independent and can be separated into distinct classes. Finally,
the tripartite position suggests that depression and anxiety have both shared and unique
features (Clark & Watson, 1991).

The unitarian position is supported by studies examining the use of psychotropic
drug distribution. For example, tricyclic antidepressants have been shown to have about
equal effectiveness in treating both anxiety and depressive disorders (Paykel, Rowan,
Roa, & Bhat, 1982; Johnstone et al., 1980). However, it is possible that psychotropic
drugs may not be diagnosis speciﬁc. Alexopoulos (1991) investigated the ﬁ'equency of
anxiety symptoms for in-patient older adults who met DSM-III criteria for depression,
with a signiﬁcant portion of their population meeting criteria ﬁ'om both depression and
dementia (approximately 41%). Sixty-one percent of their sample experienced mild
psychic symptoms of anxiety (e.g., tension, apprehension, fear, and irritability) and ﬁfty-
three percent of the sample population reported mild somatic symptoms of anxiety (e.g.,
dry month, indigestion, heart palpitations). This study suggests that many individuals
with depressive disorders also experience symptoms of anxiety.

The position that suggests that anxiety disorders and depressive disorders are
independent and are two distinct disorders (e. g., cognitive speciﬁcity) has also been
supported. Mash and Barkley (1996) offered the following distinction between anxiety
and depression: “Individuals with a depressive schema for instance, process and interpret

information about themselves, their work, and the future in a negatively biased fashion

(the cognitive triad; Beck, Rush, Shaw & Emery, 1979), whereas persons with an anxiety
schema interpret environmental stimuli with a cognitive focus on future threat” (Kendall,
Howard & Epps, 1988).

Burns and Eidelson (1998) found that the nonspeciﬁc symptoms depression and
anxiety are distinct and cannot be combined into a cluster of negative affect symptoms.
They used structural equation modeling to test the tripartite model of depression and
anxiety (Clark & Watson, 1991) in outpatients seeking treatment for mood disorders,
outpatients seeking treatment for substance abuse, and college students. This tripartite
model proposes that anxiety and depression have both shared and unique characteristics
(the tripartite model is discussed in more detail below). Burn and Eidelson found that the
tripartite model did not ﬁt any of the sample’s data. The nonspeciﬁc symptoms of
depression and anxiety did not load onto a single negative affect factor suggesting that
these two disorders do not have symptoms in common and are best represented as two
separate and distinct syndromes.

Roth, Gurney, Garside, and Kerr (1972) reported that anxiety and depression have
different courses and outcomes. Studies using anxiety and depression scales have also
been successful in separating the two groups (Alexopoulos, 1991). Roth and colleagues
(1972) were able to separate groups of patients with diagnosis of anxiety, depression, or
phobias by using scores on several anxiety and depression measures. They concluded that
anxiety and depression are distinct syndromes. Using the Present State Examination,
Wing, Cooper, and Sartorius (1974) , also were able to distinguish the symptomology of
depression and anxiety and identify two separate groups.

Nelson and Novy (1997) investigated the psychometric distinctiveness of self-

10

reported anxiety and depression in chronic pain suffers. Their ﬁndings demonstrated that
it is possible to distinguish different dimensions of anxiety and depression using the State
Trait Anxiety Inventory -State subscale to measure anxiety (STAI-S) and the Beck
Depression Inventory (BDI) to measure depression. They ﬁrrther suggest that the BDI
and STAI-S scales appear to be well suited for assessing depression and anxiety
respectively in chronic pain sufferers.

Although there is frequent comorbidity between anxiety and depression, a
primary diagnosis of anxiety is more likely to appear alone, than a primary diagnosis of
depression which is usually accompanied by anxiety (Clarkin & Kendell, 1992).
Furthermore, in a review of the literature Stavrakaki and Vargo concluded that anxiety
and depression can be classiﬁed as separate disorders both clinically and statistically
(l 986).

The tripartite model was proposed by Clark and Watson (1991). They suggested
that depression and anxiety have common and unique features. Elevated levels of
negative affect (e. g., distress) characterize both anxiety and depression; however, low
levels of positive affect are unique to depression and psychological hyperarousal is
unique to anxiety. To investigate the common and speciﬁc dimensions of self-reported
anxiety and depression, Clark, Steer, and Beck (1994) administered the Beck Depression
Inventory (BDI) and the Beck Anxiety Inventory (BAI) to 844 outpatients and 420
undergraduates. They found that one second-order general distress or negative affect
factor accounted for more than 40% of the extracted variance in the BDI and BAI. Aﬁer
controlling for the general distress factor, depression and anxiety emerged as ﬁrst order

factors, each accounting for over 20% of the extracted variance. To estimate the

11

generaliziability of these factors, Steer, Beck, Clark, and Ranieri (1995) administered the
Beck Depression Inventory and the Beck Anxiety Inventory to 1,000 outpatients
diagnosed with various types of psychiatric disorders and found similar results.

Comorbid conditions like anxiety and depression may or may not be functionally
related, and determining the presence or non-presence of a relationship is a crucial aspect
of treatment planning (Clarkin & Kendell, 1992). A great majority of research
differentiating anxiety and depression to date has used clinical samples (Paykel, Rowan,
Roa, & Bhat, 1982; Johnstone et a1, 1980; Alexopoulos, 1991; Nelson & Novy, 1997).
These samples are not representative of the population, and so the results might not be
generalizable. Thus, it was important that we examined this relationship in normal
healthy older adults. In addition, most research has tried to distinguish anxiety disorders
from depression disorders. As mentioned previously symptoms of anxiety and depressive
that do not meet the criteria for speciﬁc syndromes occur in much larger percentages of
the population (J este et al.,1999). It is therefore also important to determine if anxiety can
be distinguished from depression at the symptom level.

Two measures of depression, the Beck Depression Inventory (Beck et al., 1961)
and the Geriatric Depression Scale (Y esavage, Brink, Rose et al., 1983), and two
measures of anxiety, the State Trait Anxiety Inventory (Spielberger, Gorsuch & Lushene,
1970) and the Penn State Worry Questionnaire (Meyer, Miller, Metzger, & Borkovec,
1990), were administered to differentiate the symptoms and increase our understanding of
the distinction between depression and anxiety in community dwelling older adults.

Beck, Stanley, and Zebb (1995) examined the psychometric properties of the

Penn State Worry Questionnaire. They reported that the PSWQ signiﬁcantly correlated

12

with the Beck Depression Inventory (r = 0.41 ), the State-Trait Anxiety Inventory- State (r
= 0.40), and the State-Trait Anxiety Inventory-trait (; = 0.58). Meyer, Miller, Metzger,
and Borkovec (1990) reported lower correlations between the PSWQ and the BDI (r_=
.36) in student samples. Molina and Borkovec reported that in clinical samples (e. g.,
Generalized Anxiety Disorder) the PSWQ did not correlate signiﬁcantly with the BDI or
the Hamilton Scale for Depression. Rankin, Gfeller, Gilner (1993) reported that the inter-
correlation between the Geriatric Depression Scale and The State-Trait Anxiety Inventory
for Children was signiﬁcant (; = 0.51). In 1979 the STAI underwent a major revision to
develop a “purer” measure of anxiety that would provide a ﬁrmer basis for discriminating
between anxiety and depresssion, hence several items were eliminated from the STAI-X
when creating the STAI-Y (Spielberger, Gorsuch, Lushene, Vagg, & Jacobs, 1983).
Anxiety and depression clearly affects the amount of memory capacity available
for task performance (LaRue, 1992; Eysenck, 1979). Memory decline with age is one of
the most common concerns among older adults (La Rue, 1992). It is therefore of utmost

importance to understand how these emotions affect older adult’s memory performance.

The Effects of Anxieg and Depression on Cogm'tion

It is a commonly held belief that our affect inﬂuences our cognitive abilities. The
cognitive sequelae of depression in the elderly have attracted great attention (LaRue,
1992). Depressed patients frequently complain of problems with memory, attention and
concentration, and on neuropsychological testing it is not unusual to ﬁnd evidence of
cognitive impairment associated with depression (for reviews, see Caine, 1986; Johnson

& Magaro, 1987; Marcopulos, 1989; Miler, 1975; Niederehe, 1986; Levy-Cushman &

l3

Abeles, 1998). In addition, even mild depression is associated with an increase in
memory complaints (Levy-Cushman & Abeles, 1998). Eysenck (1979) stated that
anxiety also clearly affects the amount of memory capacity available for task
performance. However, the inﬂuence of anxiety on cognitive performance is
understudied in older adults (Salzman & Lebowitz, 1991). Several investigations have
been conducted on the relationship between anxiety and sustained attention and memory
in younger adults but there is very little research of this type with older populations
(Rankin, Gilner, Gfeller, J. D, Katz, 1994). It was therefore essential that we investigate

the inﬂuence of anxiety on memory performance in older adults.

Eysenck’s Theoretical Overview of Anxieg

Eysenck (1979) combined the views of several theorists to create a more complete
model of how anxiety affects memory and learning that would be consistent with
available empirical evidence. Eysenck ﬁrst stated that there is a distinction between state
and trait anxiety. State anxiety is responsive to situational factors whereas trait anxiety
(being a stable personality characteristic) is not. Therefore, state anxiety is more
predictive of task performance than trait anxiety.

Eysenck also proposed that anxiety is composed of both cognitive and
motivational components. Leibert and Morris (1967) found that test anxiety is comprised
of two conceptually distinctive components: worry (the cognitive component, involving
performance expectations and negative self-appraisal) and emotionality (involving levels
of physiological functioning) (as cited in Eysenck, 1979). This dual nature of anxiety can

be explained by a theory proposed by Schacter and Singer (1962). They argued that all

14

emotions result in highly similar physiological states, and it is the cognitive state that
determines which emotion is experienced.

Eysenck takes the position that worry (the cognitive component of anxiety) and
other task-irrelevant activities associated with anxiety will impair performance. This is
because task-irrelevant information involved in worry and cognitive self-concem
competes with task-relevant information for space in the processing system. Hence,
individuals who are anxious must divide their attention between the task at hand and their
worry; where individuals who are not anxious or less anxious can devote more of their
processing resources to the performance task.

Eysenck further argues that the part of the processing system that is most involved
in the processing of task-irrelevant and task-relevant information is working memory.
The working memory model comprises a limited central processing space resembling
attention, an articulatory loop which holds information in a speech based form, and a
visuo-spacial scratch pad which is specialized for spatial and visual coding (Baddeley &
Hitch, 1974). In working memory information is manipulated in addition to being stored.
Task irrelevant cognitive activities such as worry take up space in working memory and
hence produce decrements in the quality of performance. The degree of the decrement
will depend on the demands placed on working memory by the task-irrelevant
information (e.g., worry).

This view is similar to that of Craik and colleagues. He proposed that memory
performance is the result of an interaction between external and internal factors. The
external factors include the amount of environmental support provided by the encoding

and retrieval situations (e.g., the availability of prior knowledge that might foster rich

15

encoding) and by the form of the task. The major internal factor is the processing
resources a person has available for memory. Hasher and lacks (1979) propose that older
adults have less processing resources such as attention and working memory. Older
adults are less able to complete resource demanding memory tasks than younger adults.
Therefore older adults may be much more susceptible than younger adults to task
irrelevant cognitive activities such as worry take up space in working memory and hence
produce decrements in the quality of performance.

On the other hand, Eysenck proposed that highly anxious individuals will attempt
to compensate for decrements in performance and limits in working memory space by
increasing their effort expenditure. Kahneman (1973) argued that the amount of effort we
expend on a task depends on the individual’s evaluation of the task demands, the more
demanding the task is perceived to be the more effort that will be expended. Since
anxious individuals must process both task-irrelevant and task-relevant information, it
follows that the task demands for anxious individuals are greater than for those who are
not anxious. Therefore anxiety leads to a greater allocation of resources to effort.
Kahneman also noted that as the processing demands increase the discrepancy between
the supply of effort and task demands will grow. Therefore the increased effort by highly
anxious individuals will only partially compensate for the greater processing demands on
working memory by task irrelevant information.

In summary, Eysenck suggests that anxiety activates two competing processes.
First, worry will impair memory performance because it preempts some of the available
space in working memory. On the other hand, the increased task demands cause an

increase in effort, which enhances performance by increasing attentional capacity.

16

The Yerkes-Dodson Law

The Yerkes-Dodson Law (1908) describes the relationship between performance
efﬁcacy, arousal level, and task difﬁculty. The ﬁrst assertion is that there is a curvilinear
relationship between arousal and performance, with the optimal levels of performance
being associated with intermediate levels of arousal. The second assertion is that the
optimal level of arousal varies inversely with the level of task difﬁculty. Therefore
arousal (and perhaps anxiety) will interact with task difﬁculty.

One problem with the Yerkes-Dodson theory is that the concept of task difﬁculty
is ill deﬁned (Eysenck, 1982). Human memory, in particular retrieval, has been studied
experimentally using several memory tasks including ﬁ'ee recall, cued recall, and
recognition (Craik, 1983). Retrieval is initiated by presentation of cues, self-generation of
cues (as in free recall), or by stimuli encoded in normal perception (Craik, Naveh-
Benjamin, Govoni, & Anderson, 1996). When we try to remember something some
attentional resources are necessary to maintain a retrieval mode in the cognitive system.
Additional resource demanding operations are also likely to be necessary because often
the retrieval information presented by itself is inadequate to invoke recollection. These
additional resources are most likely to be required in free recall, then cued recall, and
least likely in recognition (Craik, Naveh-Benjamin, Govoni, & Anderson, 1996). In
addition, Craik (1983) suggested that free recall is more likely to involve conscious
operations (remembering to remember in the absence of cues) than recognition. Craik et
a1. (1996) found that, as the paradigm shifts from recognition to free recall, divided

attention is associated with a drop in memory performance. Hence, based on these

17

ﬁndings, there is evidence that task difﬁculty decreases as the paradigm shifts from ﬁ'ee
recall through cued recall to recognition. Craik (1986) found that age differences in
memory performance are smallest on recognition tasks and largest on free recall tasks.
Furthermore, Craik and McDowd (1987) found that the resource demands of a free recall
task as compared to a recognition task were differentially greater in older adults.
Performance on a secondary task showed that recall was associated with greater resource
costs than was recognition and this effect was ampliﬁed by age. These studies lend
support for assuming that older adults have reduced memory resources and hence it is
possible that older adults may ﬁnd all types of memory tasks more difﬁcult. It is possible
that, although it has been establish that recognition tends to place lower demands on
working memory in younger adults, this may not be the case in older adults. Since the
optimal level of arousal should vary inversely with the level of task difﬁculty, it is
important to determine if recognition is in fact an “easy task” for older adults.

It should also be noted that Yerkes and Dodson did not propose an explanation for
the interaction between arousal level and task difﬁculty. However, using Eysenck’s
model it is possible to assume that difﬁcult processing tasks make greater demands on
working memory than easy tasks. Since more anxious individuals have less available
working memory resources it is logical that the detrimental effects of anxiety will be
more pronounced on difﬁcult tasks such as free recall than on simpler tasks like cued
recall and recognition. Furthermore, since anxiety restricts working memory capacity it
will also tend to reduce the extensiveness of elaboration and encoding which also
requires available working memory capacity (Mueller, 1979). Therefore, it would also be

expected that anxiety will affect strategy use.

18

A Review of Studies of Anxiety on Memogg Performance in the Elderly

A number of studies have examined the effects of anxiety on younger adults’
memory performance (see Eysenck, 1982; and Hockey, 1986, for reviews). The effects
are not uniform but instead depend on factors such as the type of anxiety and the
difﬁculty of the task performed. In general, decrements in performance have been
associated more consistently with high state anxiety than high trait anxiety (Hockey,
1986; Hodges & Durham, 1972). However, the effects of both types of anxiety interact
with task difﬁculty, with the effects of anxiety generally being greater for harder tasks
(Eysenck, 1982). There have been few investigations on how older adult’s affective
processes affect their cognitive performance (Fisk, 1996, Salzrnan & Lebowitz, 1991). In
addition, those studies that do exist have been plagued with inconsistent and
contradictory ﬁndings (Davidson, Dixon, & Hultsh, 1991). A review of the literature
examining the effects of anxiety in the elderly follows.

There have been several studies of the relationship between state anxiety and
memory performance in older adults however; the results have been inconsistent.
Koenders et al. (1993) studied the relation of state anxiety and memory performance in
35 older adults with memory dysfunction. They found that state-anxiety appeared not to
be associated with memory performance or with severity of dementia. Koenders and
colleagues suggested that further study is required to ascertain whether the associations
reported are speciﬁc for anxiety and differ from those of other emotional states such as

depression.

19

Fisk (1996) also examined the inﬂuence of state anxiety and arousal on learning
and memory tasks in younger and older adults; he found that arousal level was negatively
related to learning performance. However, individual differences in anxiety had no effect
on learning. It is notable however, that Fisk developed the scales that were used in this
study. Waldstein et al. (1997) also found that self-reported levels of anxiety did not
predict performance on measure of attention, learning, memory, mental ﬂexibility, and
eye hand coordination in young and middle aged men.

On the other hand, Hill and Vandervoort (1992) examined the relationship
between state anxiety and ﬁ'ee recall performance in 74 elderly community dwelling
adults who completed a memory-training program. They found that state anxiety was
negatively associated with recall performance but was not related to amount of time spent
studying. Rankin and colleagues also examined the inﬂuence of state anxiety on memory
performance in older adults (Rankin, Gilner, Gfeller, & Katz, 1994). Twenty community
dwelling elderly (10 high in state anxiety and 10 low in state anxiety) were administered
the logical memory, visual reproduction, and digit span subtests of the Weschler Memory
Scale-Revised (WMS-R). There was a signiﬁcant difference between the two groups on
digits forward but not digits backward or the other memory tests. They suggested that
anxiety may have an adverse effect on sustained attention, which may not be evident on
verbal or ﬁgural memory tests.

Ross (1968) contrasted old and young subjects on a series of verbal paired-
associate tasks. The tasks were proceeded by one of three instructional conditions:
supportive, neutral, or challenging. Overall, older adults took longer to master the

leaming task than did their younger counterparts. It was postulated that the anxiety or

20

stress associated with the challenging instructions interfered with the older subjects
performance but this was not the case for younger participants. However the relationship
between anxiety and performance was difﬁcult to determine because the participants
reported their level of anxiety during the task anecdotally.

The relationship between MMPI-2 measures of depression, anxiety, and psychotic
thinking and attention span (digit span), verbal list learning, and memory test
performance (Wechsler Memory Scale- Revised) was examined in 48 male patients with
closed head injuries and in 80 males from a psychiatric inpatient sample (Gass, 1996). In
both samples anxiety as measured by the MMPI-2 signiﬁcantly predicted attention span
and memory but was not predictive of learning. MMPI-2 scores correlated with factor
scores derived from the Logical Memory and Visual Reproduction subtests of the WMS-
R. It is possible that MMPI-2 measures are associated with cognitive inefﬁciency
secondary to intrusive thoughts, preoccupation, distractibility, or anxiety over possible
failure. Other cognitive tasks, such as list learning (which involves repeated trials), may
be less vulnerable to adverse cognitive effects of emotional distress and other
psychological systems.

J onker, Smits, and Deeg (1997) found that anxiety subscale of the Metamemory
in Adulthood Questionnaire (MIA) predicted immediate, delayed, and prospective
memory test performance in older adults; higher levels of anxiety leading to decrements
in performance. There is some agreement that the self-efﬁcacy dimension of the MIA,
which measures self-perceived memory capacity, predicts memory performance
(Hertzog, Dixon, & Hultsch, 1990; Cavanugh & Poon, 1989). However, Jonker and

colleagues suggested that anxiety during testing is more important to memory functioning

21

than the self-efﬁcacy dimension of the MIA. Davidson, Dixon and Hultsch (1991)
reported that neither memory anxiety, or anxiety about one’s own memory abilities, as
measured by the MIA or state anxiety had an impact on memory performance in
participants under sixty-nine years of age. However, they reported that anxiety did have a
disruptive effect on memory performance as measured by free recall tasks in the elderly
(those over sixty-nine years of age). On the other hand, earlier studies examining anxiety
found that the Achievement and Anxiety subscale of the MIA was not a predictor of
performance on memory tests (e.g., Cavanaugh & Poon, 1989; Hertzog, Dixon, &
Hultsch, 1990). Neither was there a signiﬁcant relationship between memory-related
anxiety as measured by the MIA Anxiety subscales and actual performance (Dixon &
Hultsch, 1983). Dixon, Hertzog, and Hultsch (1986) explained the absence of a
relationship in these studies by suggesting that MIA Anxiety subscales are related to state
anxiety levels and that the elderly individuals who are concerned about their memory
ability, experience greater state anxiety in anticipation of, or during, an actual memory-
demanding situation regardless of their performance. Cavanaugh and Murphy (1986) also
suggested that not all measures of anxiety are equally powerful.

Cockbum and Smith (1994) examined the effect of anxiety on prospective
memory in older adults and found that the type of errors made is differentially related to
low and high anxiety and to age. Cockbum suggests that errors can occur because of
failure during encoding, either of intention, resulting in there being no record of the
activity that should be performed or because of failure at the retrieval stage. In the
prospective memory tasks used in this study, the Rivermead Behavioral Memory Test,

the participant attempts successively to encode and store the instructions, activate the

22

intention to respond when the appropriate context appears, retrieve an appropriate
response and execute the correct response. More incorrect responses were found for
individuals high and low in anxiety than for those at intermediate levels, suggesting that
performance is curvilinear function of anxiety, similar to those they would be expected
based on Eysenck’s model (see above). It was argued that anxiety activates a checking
procedure and corrects erroneous responses (so more anxious individuals are less likely
to ask for the wrong item). However, maintaining an effective monitoring system via
increased anxiety has its costs, and these are manifested in reduced working memory
capacity. This leads to blocks in the system (failures of retrieval of previously stored
information-- e. g. subjects needing to be prompted). Errors of asking for the wrong item,
or making no response will be associated with low anxiety; errors failing to make a
correct response until prompted, or knowing one had to ask for something without
recalling what it is, will tend to be associated with high anxiety.
Trait Anxiety

Most studies have found decrements in memory span with state anxiety (Mandler,
1979), but the results are somewhat more inconsistent for trait anxiety (Eysenck, 1979).
State and trait anxiety are separate but related constructs. Differences in trait anxiety do
not invariably translated into differences in state anxiety, hence it is not surprising that
the results have been inconsistent for trait anxiety (Eysenck, 1979). Since state anxiety
and trait anxiety are not always related to one another, and individuals who are high in
trait anxiety do not experience increases in state anxiety in all situations, an interesting
question is how these two types of anxiety interact to produce changes in performance.

Trait anxiety refers to enduring anxiety levels. In this way trait anxiety can be viewed as

23

a chronic stress.

One possibility is that increased levels of trait anxiety can heighten the effects of
state anxiety and hence increase the detrimental effects of anxiety on working memory
capacity. Trait anxiety may lead to hyperarousal to environmental demands (e. g.,
memory demanding tasks) and a decreased capacity to adjust to the demand (Fleming,
Baum, Davidson, Rectanus, & McArdle, 1987). It has been found that when the
psychological demands on the body are chronic, like those of trait anxiety, the body will
respond with a persistently high level of arousal (McEwen & Stellar, 1993). This
continuous arousal may lead to an underlying pathological state. For example, Lepore,
Miles, and Levy (1997) found that when college students reported experiencing high
levels of chronic stresses as compared to low levels of chronic stress they experienced
exaggerated physiological responses to acute challenges, slower recovery from acute
challenges, elevated psychological distress, and reported more illness. If the hyperarousal
theory holds true we would expect individuals who have high levels of trait anxiety to
experience exaggerated detriments from state anxiety during memory testing.

On the other hand it has been proposed that exposure to previous or chronic stress
may have a beneﬁcial effect (H. Eysenck, 1983). The inoculation theory proposes that
exposure to a chronic stress (trait anxiety) increases our resistance to subsequent
challenges. Individuals can experience direct tolerance, in which exposure to one stressor
reduces subsequent impact of the same stress or they can experience cross-tolerance
where exposure to one type of stress reduces the impact of another stress on performance.
For example, Norris and Murrell (1988) examined the impact of ﬂoods on older adult’s

level of anxiety and distress. They reported that older adults who had previously

24

experienced ﬂood disasters did not experience an increase in anxiety or distress as a
result of subsequent ﬂood disasters whereas the opposite was true for older adults without
prior experience with ﬂood disasters. If the inoculation hypothesis holds true we would
expect older adults who experience high levels of trait anxiety to be protected from the
detrimental effects of state anxiety during memory testing.

To our knowledge there has been very little research that has examined the effects
of trait anxiety on memory performance. Koenders et al. (1993) studied the relation of
state and trait anxiety on memory performance in 35 older adults with memory
dysfunction. They found that state-anxiety appeared not to be associated with memory
performance or with severity of dementia; however, trait-anxiety was positively
associated with memory performance but negatively associated with the severity of
dementia. This is one of the only studies to examine the inﬂuence of trait anxiety on
memory performance in older adults.

Sorg and Whitney (1992) reported that traditionally, short-term memory capacity
have been assessed with word span or digit span tests. The authors suggest that perhaps
we should look at Baddeley and Hitch’s working memory model where information is
manipulated in addition to being just stored. Sorg and Witney wished to examine how
working memory capacity is affected by trait anxiety and determine how situational stress
interacts with trait anxiety in producing changes in working memory capacity in 30
college students. They predicted that the combined effects of high trait anxiety and
situational stress would reduce memory performance. The high anxiety subjects
performed signiﬁcantly worse under the stress than the no stress condition and the high

anxiety subjects performed signiﬁcantly better than low anxiety subjects under the non-

25

stress condition. In general, their data conform with the curvilinear ﬁmction often
obtained when performance is plotted against arousal (Easterbrook, 1959; Eysenck, 1982;
Yerkes & Dodson, 1908). High trait anxiety subjects suffer from lower working memory
capacity when placed under additional stress. This study lends support to the theory that
trait anxiety may lead to hyperarousal to environmental demands (e. g., memory
demanding tasks) and a decreased capacity to perform well.

There seems to be clear evidence that anxiety affects cognitive performance at
least in younger adults, but it seems unclear as to how and to what extent. This study
examined how both state and trait anxiety affect memory performance in older adults. It
further examined how state and trait anxiety interact to produce changes in memory
performance. Using multiple measures of anxiety and memory tasks of differing
complexities also afforded us the opportunity to test Eysenck’s model of anxiety in older
adults. An understanding of the inﬂuence of affect on cognitive performance is essential
because diagnosis and treatment decisions are in part based upon the amount and

intensity of elders' affective complaints (Zarit, 1980).

Memog; Complaints

An understanding of memory complaints is also essential because diagnosis and
treatment decisions are in part based upon the amount and intensity of elders' memory
complaints (Zarit, 1980). Fifty to eighty percent of older adults have subjective memory
complaints (Lowenthal, Berkman, Buehler, Pierce, Robbinson and Trier, 1967).
However, most empirical work has also demonstrated that memory complaints are not

directly related to memory performance (Kahn, Zarit, Hilbert & Niederehe, 1975;

26

Rosavage, 1990; Zelinski, Gilewski, & Schaie, 1993, Popkin, Gallagher, Thompson &
Moore, 1982; Bolla, Lindgren, Bonaccorsy, et al., 1991). The proposed diagnostic
criteria for Age Related Cognitive Decline are dependent upon the presence of subjective
complaints in addition to memory impairment. The variable relationship between
memory complaints and objective performance has caused some to question this criterion
(Caine, 1993).

Previous studies suggested that depression is an immrtant component of memory
complaints among elderly adults (Kahn, Zarit, Hilbert & Niederehe, 1975; Popkin,
Gallagher, Thompson & Moore, 1982; Dartigues, Mazaux. Dequae, Letenneur, Giroire,
Barberger-Gateau, 1994; Poitrenaud, Malbezin & Guez, 1989, Levy-Cushrnan & Abeles,
1998). However, the inﬂuence of anxiety on memory complaints is not well studied in
older adults (Salzman & Lebowitz, 1991).

Cavanaugh and Poon (1989) examined the relationship between metamemory
(beliefs about one’s memory ability) and memory performance. The anxiety measure of
the Metamemory in Adulthood Questionnaire (MIA) correlated with the Short Inventory
of Memory Experiences for younger adults but not for older adults. However the anxiety
scale of the MIA speciﬁcally measures anxiety about memory performance and not
general anxiety.

McDougall (1994) examined the effects of depression, health status, self efﬁcacy,
and demographic variables on metamemory of older adults. Metamemory was measured
using the Metamemory in Adulthood Questionnaire which is a measure of memory
components of knowledge, beliefs, and affect and is composed of seven subscales.

McDougall reported that there was no relationship between the seven metamemory

27

subscales and age. However, memory efﬁcacy, both level and strength, were signiﬁcantly
correlated with anxiety. Memory efﬁcacy is an individual’s judgment about how well
they can organize and execute a course of memory actions required to deal with
prospective and unpredictable situations. Memory efﬁcacy provides a mechanism to
understand an individual’s inﬂuence over their motivation and their own behavior.

Hultsch, Hertzog and Dixon (1987) found that older adults tend to report less
memory capacity, more decline in memory functioning, and believe they have less
control over their memory ability than young adults. They also reported that women show
greater anxiety about their memory in memory demanding situations. However, there did
not appear to be an age effect on level of anxiety. Thus the ﬁndings with these samples
were inconclusive.

The present study will examine the effect of anxiety on memory complaints in a
sample of able elderly. Anxiety will be measured using the State Trait Anxiety Inventory
(STAI, Spielberger, Gorsuch & Lushene, 197 0) and the Penn State Worry Questionnaire
(Meyer, Miller, Metzger, & Borkovec, 1990) and memory complaints will be measured
using the Memory Assessment Clinics Self-Rating Scale (MAC-S, Revised, Winterling,

Crook, Salama & Grobert, 1986).

In summary, the ﬁrst aim of this study was to differentiate the symptoms of
anxiety and depression in community dwelling older adults. There seems to be clear
evidence that anxiety affects cognitive performance at least in younger adults but it is
unclear as to how and to what extent. This study investigated the effect of anxiety on

memory performance in community dwelling elderly adults. Furthermore, the effects of

28

anxiety on memory performance at different levels of task complexity was examined
(e.g., free recall, cued recall, and recognition). In addition, this study investigated how
anxiety inﬂuences strategy use on memory tasks. The effects of both state and trait
anxiety on memory performance in older adults were studied. How state and trait anxiety
interacted to produce changes in memory performance was examined. Finally, the present

study examined the effect of anxiety on memory complaints in a sample of able elderly.

Hypotheses

1. Symptoms of depression and symptoms of anxiety were expected to emerge as two
separate constructs in community dwelling elderly.

Scores on the Beck Depression Inventory (BDI; Beck et al., 1961) and the
Geriatric Depression Scale (GDS; Yesavage, Brink, Rose et al., 1983) would be highly
correlated. Scores on the State-Trait Anxiety Inventory (STAI, Spielberger, Gorsuch &
Lushene, 1970) and the Penn State Worry Questionnaire (PSWQ, Meyer, Miller,
Metzger, & Borkovec, 1990) would be highly correlated. Correlations between the
depression and anxiety measures would be signiﬁcantly weaker. Further items on the
BDI and GDS were expected load on one construct (depression) whereas items from the

STAI and PSWQ will load onto a second factor (anxiety).

2. There would be a differential relationship of memory performance as a function of
level of anxiety, with the optimal levels of performance being associated with
intermediate levels of state anxiety.

Speciﬁcally, a differential relationship was expected to appear when examining

29

the relationship between scores on the State-Trait Anxiety Inventory - State (STAI,
Spielberger, Gorsuch & Lushene, 197 0) and memory performance as determined by
scores on the California Verbal Learning Test (CVLT, Delis, Kramer, Kaplan & Ober,

1987).

3. There would be a differential relationship of memory performance as a function of
level of state anxiety at different levels of task complexity.

Optimal performance will occur on a task of low complexity (the recognition trial
of the CVLT) whereas state anxiety will hinder performance on more complex tasks (free

recall trial and cued recall trial of the CVLT).

4. It was expected that state anxiety would inﬂuence strategy use, we expect optimal
levels of strategy use being associated with intermediate levels of state anxiety.
Speciﬁcally, a differential relationship was expected to appear when examining
the relationship between scores on the State-Trait Anxiety Inventory (STAI, Spielberger,
Gorsuch & Lushene, 1970) and strategy use as determined by score on the California
Verbal Learning Test semantic and serial cluster ratio scores (CVLT, Delis, Kramer,

Kaplan & Ober, 1987).

5. Trait anxiety and state anxiety were expected to interact to predict memory
performance.
Speciﬁcally, we expected individuals who experience state anxiety, as measured

by the STAI-S at the time of testing and who have high levels of trait anxiety, as

30

measured by the STAI-T and worry as measured by the Penn State Worry Questionnaire,

to experience exaggerated effects of anxiety on memory performance.

6. Anxiety was expected to be positively correlated with memory complaints.

Scores on the State Trait Anxiety Inventory and the Penn State Worry
Questionnaire were expected to be negatively correlated with scores on the Memory
Assessment Clinics, Inc. Self Rating Scale (MAC-S; Revised, Winterling, Crook, Salama

and Grobert, 1986) (low scores on the MAC-S represent more memory complaints).

31

METHOD

Participants

The participants were a subset of community dwelling elderly recruited through
advertisement and through letters sent to retired faculty and staff of MSU for the ongoing
Michigan State University (MSU) Psychological Clinic Aging Research Project. Each
individual was offered two assessments of their mood and memory, as well as a seven-
session workshop designed to teach relaxation and cognitive strategies for the relief of
memory difﬁculties. Individuals were tested just prior to participation in the workshop
(pre-test) and just after completing the workshop (post-test). This study only used data
collected during the pre-test assessment. The sample consisted of 47 participants who
ranged from 56 to 87 years of age M = 71.47; SD = 8.33). Of the 47 participants all were
Caucasian but one; and 19 were male and 28 were female. They had a mean education
level of 15.17 years (S_D = 3.02). Adults with signiﬁcant health problems were excluded
from the sample. Overall this sample exhibited very low levels of depression: scores on
the Beck Depression Inventory ranged from 0 to 19 (M = 5.73) and scores on the
Geriatric Depression Scale ranged from 0 to 16 (M = 4.93). Levels of trait and state
anxiety were much more variable. State anxiety scores as measured by the STAI-S
ranged from 20 (5th percentile) to 50 (99th percentile) (M = 31.38, 48th percentile). Trait
anxiety as measured by the STAI-T ranged from 20 (3rd percentile) to 61 (100th
percentile) (M = 33.15, 53rd percentile). The trait of worry as measured by the Penn State

Worry Questionnaire (PSWQ) ranged ﬁ'om 17 to 65 (M = 35.30). Individuals with

32

severe depression as determined by scores of 30 or higher on the BDI or 20 or higher on
the GDS were referred elsewhere for treatment.
Pilot Participants

To determine if the anxiety instructions impacted the participant self-reported
levels of state anxiety, the instructions were ﬁrst tested on a sample that consisted of 15
undergraduate students form Michigan State University. These students, who were
completing Introduction to Psychology (Psych 101) ranging from 18 to 24 years of age
(M = 19.60; S_D = 1.72). Three were male and 12 were female. They had a mean
education of 13.5 years (S_D = 1.4). Overall, this sample exhibited very low levels of
depression: scores on the Beck Depression Inventory ranged from 0 to 8 ( M = 3.1). State
anxiety as measured by the STAI-S ranged from 20 (2nd percentile) to 43 (77th
percentile) (M = 29.13, 33rd percentile). Trait anxiety as measured by the STAI-T ranged
from 22 (3rd percentile) to 48 (100th percentile) (M = 31.93, 37th percentile).
Procedure

F orty-seven participants were tested individually. After obtaining informed
consent the experimenter administered the Mini-Mental Status Exam (MMSE, F olstein,
Folstein, & McHugh, 1975). The participant then completed the State Trait Anxiety
Inventory-State subscale (STAI-S, Spielberger, Gorsuch & Lushene, 1970) to obtain an
estimate of each participant’s baseline level of state anxiety. The participants then
completed a non-interfering task which took approximately 10 minutes.

Half the subjects were randomly assigned to the relaxation-anxiety condition. In
this condition participants received the relaxation instructions before the ﬁrst half of the

California Verbal Learning Test (CVLT) and then received the anxiety instructions

33

before beginning the second half of the CVLT. The other half were assigned to the
anxiety-relaxation condition. In this condition participants received the anxiety
instructions before the ﬁrst half of the California Verbal Learning Test (CVLT) and then
received the relaxation instructions before beginning the second half of the CVLT. The
instructions were counterbalanced using a within-subj ects design. The experimenter then
administered the ﬁrst manipulation instructions. The participants in the relaxation-
anxiety heard the following relaxation instructions:
This exercise is just for practice. We are trying to determine if this
measure works, so just do you best.
Those individuals in the anxiety-relaxation heard the following anxiety instructions:
This exercise is very important and gives us an accurate indication of your
memory functioning, so do your best.

All participants then completed half (ten) of the items on the STAI-S. The STAI-S
has respectable internal consistency with older adults (STAI-S g =.85). Hence, by
administering half of the items on the scale we should be able to generate a score
approximately equivalent to administering the ﬁrll scale. This also helped avoid practice
effects. The administration of the items on the STAI-S were counterbalanced with half of
the participants in both conditions (relaxation-anxiety and anxiety-relaxation) completing
the ﬁrst ten items of the STAI-S and the other half of the participants completing the last
ten items of the STAI-S. Participants then completed the ﬁrst half of the California
Verbal Learning Test (CVLT; Delis, Kramer, Kaplan & Ober, 1987). This included all
parts before the twenty-minute delay, with particular importance to this study, the short

delay free recall trial, and the short delay cued recall trial. At this time there was a twenty

34

minute delay in the CVLT during which a series of non-interfering tasks were
administered.

After the delay the experimenter administered the second anxiety manipulation.
Those subjects who were in the relaxation-anxiety condition now received the anxiety
instructions (see above), whereas those participants in the anxiety-relaxation condition
now received the relaxation instructions (see above). All participants then received ten
items of the STAI-S. Participants who received the ﬁrst ten items of the STAI—S after the
ﬁrst anxiety manipulation now received the second ten items, whereas those who
completed the second ten items during the ﬁrst manipulation now completed then ﬁrst ten
items. Then all participants then completed the second half of the California Verbal
Learning Test (CVLT; Delis, Kramer, Kaplan & Ober, 1987). This includes all parts after
the twenty-minute delay including the long delay free recall trial, the long delay cued
recall trial, and the recognition trial.

At the completion of testing the participants were asked to complete a self-report
packet. This packet was usually taken home by the participants and returned a few days
later in the interest of time. This packet contained questions about demographic
information, the Beck Depression Inventory (Beck ct al., 1961), the Geriatric Depression
Scale (Y esavage, Brink, Rose et al., 1983), the Penn State Worry Questionnaire (Meyer,
Miller, Metzger, & Borkovec, 1990), the State Trait Anxiety Inventory-Trait subscale
(STAI-T, Spielberger, Gorsuch & Lushene, 1970), and the Memory Assessment Clinics
Self-Rating Scale (MAC-S, Revised, Winterling, Crook, Salama & Grobert, 1986).

The administration of the instruments took approximately two hours.

35

Pilot Study

The anxiety manipulation that was used in this study was ﬁrst pilot tested on
ﬁfteen college students ﬁ'om Michigan State University. The same procedure as above
was used, except the college students were not be asked to complete the self-report
packet. The order of the manipulation was counterbalanced as above.

Measures

a. Beck Depression Inventog (BDI; Beck et al., 1961) Including items addressing mood,
sense of pessimism and guilt, social withdrawal, sleep disturbances, loss of energy, and
weight and appetite, this 21-item 4-point scale addresses the intensity of depressive
symptoms. Beck et a1. (1961) reported high internal consistency (r_ = .93). Good test-
retest reliability has also been reported (r = .74) after a three month interval (Miller &
Seligrnan, 1973). Studies with the elderly show that the BDI has respectable internal
consistency and stability for use in research with this population (Spitzer, Endicott, &
Robins, 1978). It has a high detection rate for major and minor depressive disorders, with
a misclassiﬁcation rate of approximately 16-17% using customary BDI cutoff scores
(Gallagher, Nies, & Thompson, 1982; Gallagher, Breckenridge, Steinmetz & Thompson,
1983). It seems to be a useful instrument for identiﬁcation of depressed elderly.

b. Geriatric Depression Scale (GDS; Yesavage, Brink, Rose et al., 1983)- The GDS
consists of 30 yes-no items. Designed speciﬁcally for use with older patients, this
instrument surveys mood quite extensively with additional items to assess cognitive
complaints and social behavior. Yesavage et a1. (1983) suggested that the GDS is a
reliable measure. Test-retest reliability was calculated by having subjects complete the

questionnaire twice, one week apart. A correlation of .85 was obtained (p <.001).

36

Convergent validity was found between the GDS, the lung Self-Rating Scale of
Depression (;= .86), and the Harniliton Rating Scale for Depression (I = .83).
Furthermore, the GDS appeared to have respectable internal consistency (; = .94) and
stability for use with the elderly. According to Yesavage et a1. (1993), this scale is
especially suitable for detecting depression in the elderly because it avoids classiﬁcation
errors due to poor physical health.

c. Penn SLateWorry Questionnaire (PSWQ, Meyer, Miller, Metzger, & Borkovec, 1990)
The PSWQ was developed to measure the trait of worry that evaluates a) typical tendency
of the individual to worry b) the excessiveness or intensity of wony experience and c) the
tendency to worry in general without restricting the topic to one or a small number of
situations. This scale consists of 16 items all rated on a 5-point Likert scale, ranging from
“not at all typical of me” to “very typical of me.” Meyer et al., 1990 found good
convergent validity with the STAI-Trait, r(389) = .64, this was substantially higher than
with STAI-State, [(395) = .49 or with the BDI {(154) = .36. Meyer et a1. (1990) also
reported substantial test-retest stability r_ = 92, 12 < .001. In another study the PSWQ was
found to be stable over periods of two weeks, [(56) = .75, E < .001, and 4 weeks, [(52) =
.74, P <.001 in college students (Meyer et al., 1990). Furthermore, the PSWQ appeared to
have respectable internal consistency (or = .95). Beck, Stanley, and Zebb (1995)
investigated the psychometric properties of the PSWQ in older adults with Generalized
Anxiety Disorder (n = 47, mean age = 67.9) and normal controls (p = 94, mean age =
67.5). The PSWQ appeared to have good internal consistency in both of these groups
(control er_ == .803, GAD g = .803). Adequate convergent validity was also demonstrated;

in the control sample the PSWQ correlated signiﬁcantly with the STAI-S (.56), the STAI-

37

T (.38), and the BDI (.45). Beck et a1. (1995) stated that the PSWQ displayed adequate
psychometric properties with older adults and suggested that ﬁrrther use of this scale with
geropsychology population is justiﬁed.

d. S_tate Treit Anxiety Inventorv (STAI, Spielberger, Gorsuch & Lushene, 1970)- an
anxiety measure which contains a state scale which assesses transient feelings of anxiety
and a trait scale which assesses more enduring anxiety levels. Stanley, Beck and Zebb
(1996) reported substantial test—retest stability (.62 state, .84 trait, 2 < .001). Furthermore,
the STAI appears to have respectable internal consistency with older adults (STAI-T g =
.79, STAI-S g =.85). Support exists for the use of this questionnaire with older adults
(Himmelfarb & Murrell, 1983; Pattemson, Sullivan, and Spielberger, 1980). The State
Trait Anxiety Inventory for Children also has been shown to be a reliable and valid
instrument to use with older adults; with regard to internal consistency the alpha
coefﬁcient was determined to be 0.86 (Rankin, Gfeller & Gilner, 1993). The STAI-C is
similar to the adult version except the directions are more speciﬁc and each item only has
three choice alternatives as opposed to the original STAI which provides four-choice
alternatives.

e. glifomiﬂerbel Learning Test (CVLT; Delis, Kramer, Kaplan & Ober, 1987)- The
CVLT is a learning task which assesses verbal learning and memory and requires
learning lists of words presented in successive trials. The CVLT was designed to go
beyond traditional memory tests, and quantify additional indices such as the encoding
strategy used (semantic versus serial clustering), the types of errors made, vulnerability of
memory to time and interference conditions, and the degree to which memory

performance improves with the aid of cues. The normative sample found test-retest

38

reliability to be .59, good internal consistency (g = .74), and split half reliability as 0.63.
Rosenbaum (1984) reported that the CVLT is a very sensitive indicator of subtle memory
deﬁcits. Over a one year interval Rosenbaum reported the test-retest reliability to range
from .26 (percent middle recall) to .79 (long-delay cued recall). Delis, Kramer, Kaplan, &
Ober (1987) found that the total immediate recall of list A across the ﬁve trials correlated
highly (r = .66) with the WMS-R General Memory Index as well as with other WMS-R
variables.

e. Memogy Assessment Clinics Self-Rating Scale (MAC-S, Revised, Winterling, Crook,
Salama & Grobert, 1986)- Designed to assess subjective memory complaints this
questionnaire contains 21 Ability items and 24 Frequency of Occurrence of memory
problems items. All rated on a 5-point Likert scale, ranging from very good to very poor.
Crook and Larrabee (1990) factor analyzed the performance of 1,106 subjects for a
normative sample of adults ranging from 18- 92 years. West, Crook, and Larrabee (1991)
showed concurrent validity of this scale. Five Ability factors were identiﬁed and labeled
Remote Personal Memory, Numeric Recall, Everyday Task Oriented Memory, Word
Recall/ Semantic Memory, and Spatial and Topographic Memory. On the Frequency
scale, ﬁve scales were also identiﬁed, and these were labeled Word and Fact Recall,
Attention and Concentration, Everyday Task-Oriented Memory, General Forgetfulness,
and Facial Recognition. Crook and Larrabee (1991) reported substantial test-retest
stability across 3-week intervals Ability Total _r_ = .88 - . 94, Frequency Total 1‘ = .89 -
.92).

f. The Mini-Mental Status Exam (MSE, Folstein, Folstein, & McHugh, 1975) is a 30-

 

item test created to brieﬂy screen for cognitive functioning. It assesses orientation,

39

immediate recall, attention, calculation, language abilities, and constructional dyspraxia.
Scores range from 0, indicating severe impairment to 30, indicating unimpaired mental
status. It was standardized on a sample of healthy older adults (M = 74 years) whose
scores ranges from 24.6 to 27.6 (Folstein, Folstein, & McHugh, 1975). Demented persons
scored between 9.6 and 12.2, with no overlap between these two groups. A cut-off score

of 24 is often used to identify scores suggestive of impairment (Lezak, 1995).

40

RESULTS

Pilot Study

The effect of minprgetion on a_n_xietv

The effects of the manipulation instructions on anxiety were examined separately
for each treatment order. Undergraduate participants in Order 1 heard the anxiety
instruction ﬁrst followed by the relaxation instruction. Participants in Order 2 heard the
relaxation instructions ﬁrst followed by the anxiety instructions.

Order 1 (anxiegg instructions followed by relaxation instructions)

A one-way repeated measures AN OVA was conducted with the factor being
instructions (anxiety vs. relaxation) and the dependent variable being state anxiety. The
mean state anxiety score (measured using 10-items from the STAI-S) after hearing the
anxiety instructions was 13.5. The mean state anxiety score after hearing the relaxation
instructions was 13.6. This difference was not signiﬁcant (Wilks’ lambda = .993, (F 1,7)
= .052, p < .83).

Order 2 (relaxation instructions followed by a_p_xietvin§truction_s)

A one-way repeated measures AN OVA was conducted with the factor being
instructions (anxiety vs. relaxation) and the dependent variable being state anxiety. The
mean state anxiety score (on a 10-item form of the STAI-S) after hearing the relaxation
instructions was 13.9. The mean anxiety score after hearing the anxiety instructions was

14.0. This difference was not signiﬁcant (Wilks’ lambda = .992, (F 1,6) = .045, p_< .84).

Anxiety level measured by a 10-item form of the STAI-S immediately after the

41

ﬁrst manipulation (time 1) did not correlate signiﬁcantly (g = .061) with the level of
manipulation administered (level = anxiety vs. relaxation). Anxiety level as measured by
a 10-item form of the STAI-S immediately after the second manipulation (time 2) also
did not correlate signiﬁcantly (1 = .062) with the level of manipulation administered.
These results suggest that hearing the anxiety instructions did not increase the

undergraduate participants’ level of anxiety. However, it was expected that the older
adults would be more sensitive to these instructions. Older adults participating in this
study had responded to advertisements looking for individuals who have concerns about
their memory. Therefore, they were expected to have signiﬁcantly more anxiety about
their memory performance than would undergraduates completing requirements for their
Psychology 101 class. In an effort to insure a greater impact of the manipulation of
anxiety level in the older adult participants, the relaxation manipulation instructions were
altered slightly. The pilot participants heard the following relaxation instructions:

“This exercise is just for practice. We are trying to determine if this

measure works, so just do you best.”
Our research team felt that the last statement “so do your best,” might be the most salient
piece of the instructions and might promote anxiety. In view of this the relaxation
instructions for the older participants were changed to read:

”This exercise is just for practice. We are just trying to determine if this measure works.

The anxiety instructions for the older participants were not changed.

42

Hypothesis I

It was predicted that symptoms of depression and symptoms of anxiety would
emerge as two separate constructs in community dwelling elderly. Scores on the Beck
Depression Inventory (BDI; Beck et al., 1961) and the Geriatric Depression Scale (GDS;
Yesavage, Brink, Rose et al., 1983) were expected to be highly correlated. Scores on the
State-Trait Anxiety Inventory (STAI, Spielberger, Gorsuch & Lushene, 1970) and the
Penn State Worry Questionnaire (PSWQ, Meyer, Miller, Metzger, & Borkovec, 1990)
were expected to be highly correlated. Correlations between the depression and anxiety
measures were expected to be weaker. This hypothesis was partially conﬁrmed.

The BDI total score correlated signiﬁcantly with the GDS total score (r = .793, p
< .01 ), indicating that the two measures of depression used in this study were correlated.
Total scores on STAI-S (state anxiety) and STAI-T (trait anxiety) were signiﬁcantly
correlated (p: .430, p < .01). The STAI-S total score correlated signiﬁcantly with the
PSWQ total score (r = .392, p <01), and the STAI-T correlated with the PSWQ total
score (;_= .482, p < .01). These ﬁndings indicated that all measures of anxiety used in this
study were signiﬁcantly correlated with one another.

Depression as measured by the BDI total score and the GDS total score did not
signiﬁcantly correlate with state anxiety as measured by the STAI-S total score.
However, depression was found to be signiﬁcantly correlated with trait anxiety as
measured by both the STAI-T and the trait of worry as measured by the PSWQ.
Depression correlated with trait anxiety as measured by the STAI-T total score (BDI ; =

.342, p < .05; GDS ; = .387, p < .01) and worry as measured by the PSWQ total score

43

(BDI ; = .411, p < .01; GDS ; = .458, p < .01). See Table 1 for a summary ofthe
correlations between anxiety and depression.

Further, it was predicted that items from the BDI and GDS would load on one
construct (depression) whereas items from the STAI and PSWQ would load onto a
second distinct factor (anxiety). It was impossible to run a principle components factor
analysis using all items from the BDI, GDS, STAI, PSWQ because many of the items, in
particular items from the BDI and GDS, had no variance. For this reason the total scores
for the GDS, BDI, STAI-S, STAI-T, and PSWQ were entered into a principle
components factor analysis with Varimax rotation. The rotated solution, as shown in
Table 2, yielded two interpretable factors. As predicted the GDS and the BDI loaded
highly onto factor 1 (depression) but not factor 2 (anxiety). The STAI-S, STAI-T and the

PSWQ loaded heavily onto factor 2 (anxiety) but not factor 1 (depression).

The effecte of the mMulation instructions

Recall that before the participants completed the memory task they heard the
manipulation instructions. The instructions were given in an effort to increase anxiety
(anxiety instructions) or decrease anxiety (relaxation instructions) before the participants
completed the memory tasks. After the ﬁrst manipulation instructions (relaxation or
anxiety) were presented, participants completed a lO-item version of the STAI-S and the
CVLT short delay free and cued recall tasks. For convenience of reference this period
will be referred to as Time 1. After the second manipulation instructions (anxiety or
relaxation) were presented, participants completed a lO-item version of the STAI-S, the

CLVT long delay free and cued recall tasks, and the CVLT recognition task. For

convenience of reference this period will be referred to as Time 2.

The effects of manipulations on anxing

 

A Latin Square design of order 2 was used to examine the effects of the
manipulation instruction on anxiety. In this design the SS and df for any one of the three
factors (here Order, Time, and Manipulation) are algebraically (and thus numerically, up
to rounding and truncation error) identical with the SS and df for the interaction of the
other two factors. The results of this analysis indicated a signiﬁcant effect of the
manipulation instructions on state anxiety E (1, 45) = 4.69, p = .036.

Repeated-measures analysis of variance were also performed separately within
Order 1 and within Order 2. The independent variable for each analysis was type of
instruction (anxiety versus relaxation). The three dependant variables, each measured at
Times 1 and 2, were STAI-S state anxiety and CVLT free and cued recall performance
scores. Participants in Order I heard the anxiety instruction ﬁrst followed by the
relaxation instruction. Participants in Order 2 heard the relaxation instructions ﬁrst
followed by the anxiety instructions.

Order 1 (anxieg instructions followed by relaxation instructions)

A one-way repeated measures AN OVA was conducted with the repeated-
measures factor being manipulation instructions (anxiety vs. relaxation) and the
dependent variable being state anxiety. The mean state anxiety score for the anxiety
condition was 19.0. The mean anxiety score for the relaxation condition was 18.1. The
results of the ANOVA (Wilks’ lambda = .952, E = 1.16 (1, 23), p < .29) indicated there
was not a signiﬁcant effect of the manipulation (relaxation vs. anxiety) on state anxiety

when the anxiety instructions were administered ﬁrst followed by the relaxation

45

instructions. Notably, however, directions of the differences are consistent with the view
that participants were more anxious after the anxiety instructions than after the relaxation
instructions.

Order 2 relaxation instructions followed by anxiety instructions)

A one-way repeated measures AN OVA was conducted with the repeated-
measures factor being manipulation instructions (anxiety vs. relaxation) and the
dependent variable being state anxiety. The mean state anxiety score after hearing the
relaxation instructions was 15.8. The mean state anxiety score after hearing the anxiety
instructions was 17.1. The results of the ANOVA indicated a signiﬁcant effect of the
manipulation instructions on state anxiety when the relaxation instructions were
administered ﬁrst followed by the anxiety instructions, Wilks’ lambda = .841, E (1, 22) =
5.04, p = .035.

Correlations between the manipulations and anxith

Anxiety measured immediately following the ﬁrst manipulation correlated
signiﬁcantly with the level of manipulation administered (level = anxiety vs. relax) (t = -
.313, p < .03), suggesting that hearing the manipulation instructions increased the
participants level of anxiety at time 1. However, the manipulation did not signiﬁcantly
correlate with state anxiety at time 2.

Comparisons between anxiety at Time 1 and Time 2

A paired samples t-test was conducted to evaluate whether there was a difference
between anxiety measured by a lO-item version of the STAI-S after the ﬁrst manipulation
instructions and anxiety measured by a 10-item version of the STAI-S after the second

manipulation instructions, overall. The results indicated that anxiety after the ﬁrst

46

manipulation instructions (M = 17.4, S_D = 5.12) was not signiﬁcantly different than

anxiety after the second manipulation instructions M = 17.6, S_D = 4.80), t (46) = -.37, p

= .72. The mean difference between anxiety measured after the ﬁrst manipulation

instructions and anxiety measured after the second manipulation instructions was -.19.
The effect of mMulation on memory performa_nse

A Latin Square design of order 2 was used to examine the effects of the
manipulation instruction on ﬁee recall memory performance. The results of this analysis
indicated a near signiﬁcant effect of the manipulation instructions on free recall memory
performance E (1, 43) = 3.65, p = .063.

Order 1 (anxiety instructions followed by relaxation instructions)

The effect of the manipulation instruction on memory performance was also
examined seperately for each order. A one-way repeated measures AN OVA was
conducted with the repeated-measures factor being manipulation instructions (anxiety vs.
relaxation) and the dependent variable being free recall memory performance. The mean
free recall performance after hearing the anxiety instructions was 8.1 words. The mean
free recall performance after hearing the relaxation instructions was 8.8 words. The
results of the ANOVA indicated a signiﬁcant effect, Wilks’ lambda = .841, E (l , 22) =
4.160, p = .054.

A one-way repeated measures AN OVA was conducted with the repeated-
measures factor being manipulation instructions (anxiety vs. relaxation) and the
dependent variable being cued recall memory performance. The mean cued recall
memory performance after hearing the anxiety instruction was 9.6 words. The mean cued

recall performance after hearing the relaxation instructions was 9.4 words. The results of

47

the ANOVA indicated the manipulation instructions did not have a signiﬁcant effect on
cued recall memory performance when the anxiety instructions were administered ﬁrst
followed by the relaxation instructions.

Order 2 Lr'elasgrtion instructions followed by enxietv instructions)

A one-way repeated measures AN OVA was conducted with the repeated-
measures factor being manipulation instructions (anxiety vs. relaxation) and the
dependent variable being free recall memory performance. The mean free recall
performance after hearing the relaxation instruction was 8.1 words. The mean free recall
memory performance after hearing the anxiety instructions was 7.9 words. The results of
the AN OVA did not indicate a signiﬁcant effect of the manipulation instructions on free
recall memory performance when the relaxation instructions were administered ﬁrst
followed by the anxiety instructions. Notably, this pattern is the same as in the previous
order: individuals who heard the relaxation instructions outperformed individuals who
heard the anxiety instructions.

A one-way repeated measures AN OVA was conducted with the repeated-
measures factor being manipulation instructions (anxiety vs. relaxation) and the
dependent variable being cued recall memory performance. The mean cued recall
memory performance after hearing the anxiety instructions was 8.9 words. The mean
cued recall memory performance after hearing the relaxation instructions was 8.9 words.

Reco gpition memoty performance

Recognition memory performance was measured only at Time 2 and therefore we
could not analyze this data using repeated measures AN OVA. The recognition (hits)

performance after hearing the relaxation instruction was 13.7 words. The mean

48

recognition (hits) performance after hearing the anxiety instructions was 12.9 words.
Notably, this pattern suggests that individuals who heard the relaxation instructions
outperformed individuals who heard the anxiety instructions. See Table 3 for a summary

of CVLT performance means by order and condition.

Hypotheses H & 111

It was expected that there would be a differential relationship of memory
performance as a function of level of anxiety, with the optimal levels of performance
being associated with intermediate levels of state anxiety. Speciﬁcally, it was predicted
that this relationship would appear when examining the relationship between state
anxiety as measured by scores on the State-Trait Anxiety Inventory (STAI) and memory
performance as determined by scores on the California Verbal Learning Test (CVLT).

Further, it was predicted that the relationship between state anxiety and memory
performance would vary as a function of task complexity. It was thought that optimal
performance would occur on tasks of low complexity (the recognition trial of the CVLT)
whereas state anxiety would hinder performance on more complex tasks (e. g., cued recall
trial and the free recall trial of the CVLT).

Comparisons between short delay ﬁ'ee recall and long delay free recall

First it was important to determine if there was an effect of time on memory
performance. To do this a paired samples t-test was conducted to evaluate whether there
was a difference in short delay free recall memory performance and long delay ﬁee recall
memory performance, overall. The results indicated that mean short delay free recall (M

= 8.09, _S_I_) = 2.82) was not signiﬁcantly different than the mean long delay free recall

49

memory performance (M = 8.36, S_D = 2.88), t (44) = -1.l3, p = .266. The mean
difference between short delay ﬁ'ee recall and long delay free recall was .27 out of a
possible score of 16 on both measures.
Comparisons between Mk complexity

Since the optimal level of arousal should vary inversely with the level of task
difﬁculty, it was ﬁrst important to determine if the three memory tasks used in this study
actually differed in level of difﬁculty for older adults. A series of paired samples t-tests
were conducted to evaluate whether there was a difference between the three level of task
difﬁculties: free recall, cued recall and recognition. The results indicated that each level
of task complexity was signiﬁcantly different from the other two with ﬁne recall being
the most complex and recognition being the least complex. long delay line recall (M =
8.36, S_D = 2.88) was signiﬁcantly different than long delay cued recall M = 9.13, S_D =
2.58), t (44) = -3.69, p = .001. The mean difference between long delay ﬁ'ee recall and
long delay cued recall was -.78. Long delay free recall (M = 8.36, S_D = 2.88) was
signiﬁcantly different than recognition hits M = 13.33, S2 = 2.21), t (44) = -10.80, p <
.001. The mean difference between long delay free recall and long delay cued recall was -
4.98. Long delay cued recall (M = 9.13, S_D = 2.58) was signiﬁcantly different than
recognition hits M = 13.33, £12 = 2.21), t (44) = -10.12, p < .001. The mean difference
between long delay free recall and long delay cued recall was -4.20.

Within-subjects analysis

A MAN OVA was conducted to examine the within-subjects effects, with order of

the manipulation instructions as a between subjects factor, time as a within subjects

factor, anxiety as a within-subj ects covariate and memory performance as the dependent

50

variable. There were no signiﬁcant within-subj ects effects for either free recall memory
performance or cued recall memory performance. Since there were no within-subj ects
effects, we examined each time separately to look at the pattern of the between-subj ects
effects.
Between-subiects analysis

a. Free recall perforrnanc - The scatterplots for the relationship between state
anxiety and free recall performance, as shown in ﬁgures 1 and 2, demonstrate that the
predicted curvilinear relationship between these two variables did not occur. The
scatterplots suggested a more linear relationship between anxiety and performance such
that as state anxiety increases free recall memory performance decreases. Linear
regression analyses were conducted to evaluate the relationship between anxiety and free
recall performance. At Time 1 (after the ﬁrst manipulation instructions) anxiety as
measure by an abbreviated form of the STAI-S (IO-items) did not correlate signiﬁcantly
with memory performance as measured by the CVLT short delay free recall trial (t = -
.217). However, the negative correlation suggested that anxiety had a negative impact on
free recall memory performance. State anxiety, as measured by an abbreviated version of
the STAI-S, was not a signiﬁcant predictor of memory performance as measured by the
CVLT short delay free recall trial at the time of the ﬁrst manipulation instuctions. Only
about 5% of the total score variance in free recall memory performance was accounted
for by its linear relationship with state anxiety (R2 = .05).

At time 2 (after the second manipulation instructions) anxiety as measure by an
abbreviated form of the STAI-S (IO-items) was almost signiﬁcantly correlated with

memory performance as measured by the CVLT long delay free recall trial (I = -.27, p <

51

.07). Again, the negative correlation suggested that anxiety had a negative impact on ﬁne
recall memory performance. State anxiety, as measured by an abbreviated version of the
STAI-S, was almost a signiﬁcant predictor of memory performance as measured by the
CVLT long delay free recall at the time of the second manipulation. Approximately 11%
of the variance in free recall memory performance was accounted for by its linear

relationship with state anxiety (R2 = .107, p < .06).

b. Cued recell performence- The scatterplots for the relationship between state
anxiety and cued recall performance, as shown in ﬁgures 3 and 4, also demonstrates that
the predicted curvilinear relationship between these two variables did not occur. The
scatterplots suggested a more linear relationship between anxiety and cued recall
performance such that as state anxiety increases cued recall memory performance
decreases. A linear regression analyses were conducted to evaluate this relationship. At
Time 1 anxiety as measure by an abbreviated for of the STAI-S (lo-items) did not
correlate signiﬁcantly with memory performance as measured by the CVLT short delay
cued recall trial (I = -.11). However, again the negative correlation suggests that anxiety
is having a negative impact on cued recall performance. State anxiety, as measured by an
abbreviated version of the STAI-S, was also not a signiﬁcant predictor of memory
performance as measured by the CVLT short delay cued recall trial at Time 1. Only about
5% of the total score variance in free recall memory performance was accounted for by
its linear relationship with state anxiety (R2 = .046).

At Time 2 anxiety as measured by an abbreviated form of the STAI-S (IO-items)

was not signiﬁcantly correlated with memory performance as measured by the CVLT

52

delay cued recall (r = -. 15 ). However, again the negative correlation suggests that
anxiety had a negative impact on cued recall memory performance. State anxiety, as
measured by an abbreviated version of the STAI-S, was also not a signiﬁcant predictor of
memory performance as measured by the CVLT long delay cued recall at the time of the
second manipulation instructions. Only about 3% of the total score variance in free recall
memory performance was accounted for by its linear relationship with state anxiety (R2 =

.032).

c. Recogpt'tion performanc - The scatterplot for the relationship between state
anxiety and recognition performance, as shown in ﬁgure 5, demonstrates that the
predicted curvilinear relationship between these two variables did not occur. The
scatterplot suggested a more linear relationship between anxiety and recognition memory
performance such that as state anxiety increases recognition memory performance
decreases. A linear regression analysis was conducted to evaluate the relationship
between anxiety and recognition performance. Recognition performance was measured
only after the second manipulation instructions. Anxiety as measure by an abbreviated
form of the STAI-S (IO-items) was not signiﬁcantly correlated with memory
performance as measured by the CVLT recognition hits (I = -.03 7). State anxiety, as
measured by an abbreviated version of the STAI-S, was also not a signiﬁcant predictor of
recognition memory performance as measured by the CVLT recognition trial.
Approximately 4 percent of the variance in recognition memory performance was
accounted for by its linear relationship with state anxiety (R2 = .039).

It is interesting to note that although the correlations between state anxiety and

53

memory performance are not statistically signiﬁcant there is a consistent negative
relationship between anxiety and performance. This suggests that state anxiety hinders
memory performance. Notably, this negative effect is largest on the most complex task
(free recall) and smallest on the least complex task (recognition). See Table 5 for a

summary of the correlations between state anxiety and memory performance.

Hypothesis IV

It was expected that optimal strategy use would occur at intermediate levels of
state anxiety. Speciﬁcally, a differential relationship was expected to appear when
examining the relationship between scores on the State-Trait Anxiety Inventory-State
(STAI-S) and strategy use as determined by the California Verbal Learning Test semantic
and serial cluster ratios scores (CVLT).

The scatterplots for these two variables, as shown in ﬁgures 6 and 7, showed that
the predicted curvilinear relationship between anxiety and strategy use did not occur.
Linear regression analyses were conducted to evaluate if strategy use could be predicted
ﬁ'om state anxiety level. The correlation between state anxiety (as measured by the STAI-
S) and use of semantic cluster strategy use (as measured by the CVLT semantic cluster
ratio) was 0.022. State anxiety did not explain any of the variance in semantic clustering
strategy use. The correlation between state anxiety (as measured by the STAI-S) and
serial cluster strategy use (as measure my the CVLT serial cluster ratio) was -.091. Less
than 1 percent of the variance in serial cluster strategy use was accounted for by its linear

relationship with state anxiety. Overall, level of state anxiety did not affect strategy use.

54

Hypothesis V

Trait anxiety and state anxiety were expected to interact to predict memory
performance. Speciﬁcally, we expected individuals who experience state anxiety, as
measured by the STAI-S at the time of testing and who have high levels of trait anxiety,
as measured by the STAI-T and worry as measured by the Penn State Worry
Questionnaire, to experience exaggerated effects of anxiety on memory performance.

First a linear regression analysis was conducted to evaluate if state anxiety could
be predicted ﬁ'om trait anxiety. As expected, individuals with higher levels of trait
anxiety tended to have higher levels of state anxiety at statistically signiﬁcant levels. The
correlation between state anxiety (as measured by the STAI-S) and trait anxiety (as
measured by the STAI-T) was .55 (p < .001). The correlation between state anxiety (as
measured by the STAI-S) and the trait of worry (as measured by the Penn State Worry
Questionnaire) was .41 (p < .006). Approximately 31 percent of the variance in state
anxiety was accounted for by its linear relationship with trait anxiety as measured by the
STAI-T (R2 = .31). Approximately 17 percent of the variance in state anxiety was
accounted for by its linear relationship with the trait of worry as measured by the PSWQ
(R2 = .17).

Next a multiple linear regression analysis was conducted to evaluate if memory
performance could be predicted from trait anxiety, the trait of worry, and state anxiety.
Contrary to predictions the combined effects of state anxiety, trait anxiety, and worry
were not signiﬁcant predictors of memory performance. For correlations between
anxiety and memory performance, see Table 5. Ten percent of the variance in memory

performance, as measured by the CVLT short delay free recall, was accounted for by its

55

linear relationship with trait anxiety, worry, and state anxiety as measured by the STAI-T,
PSWQ and the 10 items of the STAI-S measured immediately after the ﬁrst manipulation
instructions, respectively (R2 = .10). Approximately 9 percent of the variance in memory
performance, as measured by the CVLT long delay free recall, was accounted for by its
linear relationship with trait anxiety, worry, and state anxiety as measured by the STAI-T,
PSWQ and the 10 items STAI-S measured immediately after the second manipulation (R2
= .09).

Multiple linear regression analyses were conducted to evaluate if memory
performance could be predicted from trait anxiety, state anxiety, and the interaction
between state and trait anxiety. Contrary to predictions the combined effects of state
anxiety, trait anxiety, and the interaction of state and trait anxiety were not signiﬁcant
predictors of memory performance. Approximately 5 percent of the variance in memory
performance, as measured by the CVLT short delay free recall (R2 = .05), and 4 percent
of the variance in memory performance, as measured by the CVLT long delay free recall
(R2 = .04), was accounted for by its linear relationship with trait anxiety, state anxiety and
the interaction between state and trait anxiety.

A multiple linear regression analysis was conducted to evaluate if memory
performance could be predicted ﬁ'om worry (as measured by the PSWQ), state anxiety,
and the interaction between state anxiety and worry. Similarly, the combined effects of
state anxiety, trait anxiety, and the interaction between state anxiety and worry were not
signiﬁcant predictors of memory performance. Approximately 4 percent of the variance
in memory performance, as measured by the CVLT short delay ﬁee recall (R2 = .04), and

2 percent of the variance in memory performance, as measured by the CVLT long delay

56

free recall (R2 = .02), was accounted for by its linear relationship with worry, state

anxiety and the interaction between state anxiety and worry.

Hypothesis VI

Anxiety was expected to be positively correlated with memory complaints.
Speciﬁcally, it was predicted that scores on the State Trait Anxiety Inventory and the
Penn State Worry Questionnaire would be negatively correlated with scores on the
Memory Assessment Clinics, Inc. Self Rating Scale (MAC-S; Revised, Winterling,
Crook, Salama and Grobert, 1986). Recall that low scores on the MAC-S represent more
memory complaints. This hypothesis was only partially conﬁrmed.

Worry as measured by the PSWQ and state anxiety as measured by the STAI-S
did not correlate signiﬁcantly with memory complaints as measured by the MAC-S the
total score, the MAC-S ability subtotal or with the MAC-S frequency subtotal. However,
trait anxiety as measured STAI-T total score correlated signiﬁcantly with the MAC-S
total score (t = -.411, p < .01) and the frequency of memory problems reported on the
MAC-S frequency subtotal (_r_= -.43, p < .01) but did not correlated with the MAC-S
ability subtotal. Trait anxiety as measured by the STAI-T also correlated signiﬁcantly
with many of the factors on the MAC-S including: global memory complaints, facial

recall, forgetfulness, task memory, attention, and spatial memory.

57

DISCUSSION

Anxiety and depression are two of the most common problems facing older adults
today. However, differentiating anxiety and depression at the mood, symptom, and
syndromal levels has been a long-standing problem in psychopathology research (Nelson
& Novy, 1997; Alexopoulos, 1991). In addition, older adults often seek advice and
treatment for cognitive problems (La Rue, 1992). It is a commonly held belief that our
affect inﬂuences our cognitive abilities. The cognitive sequelae of depression in the
elderly have attracted great attention (LaRue, 1992); however, the inﬂuence of anxiety on
cognitive performance and cognitive complaints is extremely understudied in older adults
(Salzman & Lebowitz, 1991). Eysenck (1979) stated that anxiety clearly affects the
amount of memory capacity available for task performance.

The objective of this study was to examine the effect of anxiety on memory
performance in the able elderly. The ﬁrst aim of this study was to differentiate the
symptoms of anxiety and depression in community dwelling older adults. There seems to
be clear evidence that anxiety affects cognitive performance at least in younger adults but
it seems unclear as to how and to what extent. This study investigated the effect of
anxiety on memory performance in community dwelling elderly adults. F urtherrnore, this
study investigated the effects of anxiety on memory performance at different levels of
task complexity (e.g., free recall, cued recall, and recognition). We were also able to
examine if altering the instructions before the memory measures was administered
(anxiety provoking vs. relaxation provoking) affected anxiety and memory performance.

In addition, we examined the impact of anxiety on strategy use. The effects of both state

58

and trait anxiety on memory performance in older adults were also investigated. It was
further examined how state and trait anxieties interact to produce changes in memory
performance. Finally, the present study examined the effect of anxiety on memory

complaints in a sample of able elderly.

The Reletionship Between Anxiety end Depression

Symptoms of depression and anxiety are two of the most common problems
reported in later life. However, differentiating depression from anxiety has been a long-
standing problem. Understanding the relationship between these two conditions is of
utmost importance because the co-presence of anxiety and depression has been found to
increase the risk of suicide over the risk associated with depression alone (Clark et al.,
1995). Increased rates of suicide have been reported when depression and anxiety co-
occur and these increased rates have been found in many groups including the elderly
(Kunik, 1993). Most research differentiating anxiety and depression to date has used
clinical samples (Paykel, Rowan, Roa, & Bhat, 1982; Johnstone et al, 1980; Alexopoulos,
1991; Nelson & Novy, 1997). These samples are not representative of the population, and
so the results might not be generalizable. Thus, it was important that we examined this
relationship in normal healthy older adults. In addition, most research has tried to
distinguish anxiety disorders from depression disorders. As mentioned previously,
symptoms of anxiety and depression that do not meet the criteria for speciﬁc syndromes
occur in much larger percentages of the population (J este et al., 1999). It was therefore
also important to determine if anxiety could be distinguished from depression at the

symptom level.

59

Participants in this study completed two measure of depression, the Beck
Depression Inventory (Beck et al., 1961) and the Geriatric Depression Scale (Y esavage,
Brink, Rose et al., 1983), and two measure of anxiety, the State Trait Anxiety Inventory
(Spielberger, Gorsuch & Lushene, 1970) and the Penn State Worry Questionnaire
(Meyer, Miller, Metzger, & Borkovec, 1990). It was predicted that symptoms of
depression and symptoms of anxiety would emerge as two separate constructs in
community dwelling elderly. This hypothesis was partially conﬁrmed. State Anxiety as
measured by the STAI-S did not correlate signiﬁcant with either depression measure.
However, trait anxiety as measure both the STAI-T and worry as measured by the Penn
State Worry Questionnaire correlated signiﬁcantly with depression. Depression
correlated with trait anxiety as measured by the STAI-T total score (BDI r; = .342 , p <
.05; GDS t = .387, p < .01) and with worry as measured by the PSWQ total score (BDI I
= .411 , p < .01; GDS I = .458, p < .01). Previous studies have found that mood and
symptom measures of depression and anxiety correlate almost as highly between
constructs as within constructs. After reviewing the literature Clark, Steer, and Beck
(1994) concluded that self reported measure of depression and anxiety correlate on
average between .62 - .70. Notably, previous research has been conducted largely on
psychiatric patients and/or college students (e.g., Clark, Steer, & Beck, 1994). Hence the
ﬁndings of this study suggest that the symptoms of depression and anxiety appear to be
more distinguishable in older population than in the cohorts previously studied.

This study also examined if depression and anxiety would emerge as two separate
factors when the items from all scales used were entered into a factor analysis. It was

impossible to run a principle components factor analysis using all items from the BDI,

60

GDS, STAI, PSWQ because many of the items, in particular items from the BDI and
GDS, had no variance. For this reason the total scores (GDS total score, BDI total score,
STAI-S total score, STAI-T total score, PSWQ total score) from these scale were entered
into a principle components factor analysis with Varimax rotation. From this analysis two
factors emerged. As predicted the GDS and the BDI loaded highly onto factor 1 but not
factor 2 and the STAI-S, STAI-T and the PSWQ loaded heavily onto factor 2.

Overall, this study seems to suggest that at the symptom level anxiety and
depression have unique as well as shared characteristics and are “comorbid conditions.”
Feinstein originally coined the term comorbidity in 1970 to refer to “any distinct
additional clinical entity that has existed or that may occur during the clinical course of a
patient who has the index disease under study.” Anxiety and depression are often referred
to as comorbid syndromes, not only because they often co-occur but also because many
of the symptoms that deﬁne these two disorders are shared. In meta-analysis Clark (1989)
found that 57% of depressed patients met criteria for an anxiety disorder; other studies
have found similar rates of comorbidity (e.g., Kessler et al., 1994). Anxiety and
depression share expectations of hopelessness and negative affect. However, anxiety is
associated with physiological hyperarousal whereas adhedonia is thought to uniquely
deﬁne depression. Although, there seems to be substantial evidence that anxiety and
depression have shared and unique characteristics the nature of the relation between these
characteristics is unclear. Several three-factor models have been proposed to explain the
relationship between anxiety and depression.

Alloy et a1. (1990) proposed a model in which the interplay between hopelessness

and helplessness is used to explain the differences and overlap between anxiety and

61

depression. They suggest that individuals who are uncertain about their ability to control
important situations would experience pure anxiety. Individuals who are sure about their
helplessness but uncertain that the negative outcome will occur experience a mixed
syndrome. Individuals who are both sure about their helplessness and are sure that the
negative outcome will occur (hopeless) experience a pure depression with despair, loss of
interest, and suicidality. Hence, in this model depression and anxiety share expectations
of helplessness but differ in expectations of hopelessness.

One of the most well validated three factor models is the tripartite model as
proposed by Clark and Watson (1991). Recall that this model offers a bridge between the
unitary negative affect theory and competing theories that highlight the differences
between depression and anxiety. The tripartite model also suggests that depression and
anxiety have common and unique features. Both anxiety and depression are characterized
by elevated levels of negative affect that is characterized as a tendency to be distressed,
worried, anxious, self-critical and to have a negative view of oneself. However, low
levels of positive affect and a loss of interest in pleasure (anhedonia) are unique to
depression whereas physiological hyperarousal (somatic anxiety) is unique to anxiety.
Signiﬁcant empirical support exists to support this model. In 1995 Watson and colleagues
administered the Mood and Anxiety Questionnaire (MASQ) to ﬁve samples (three
student, one adult, and one patient); three replicable dimensions matching those described
by Watson and Clark in 1991 consistently emerged. Psychophysiological studies have
offered further support to the tripartite model. The three symptoms groups display
distinctive patterns of brain activity; (e.g., anxious arousal is linked to hyperactivation of

the right parietotemporal region whereas low positive affect is associated with

62

hypoactivation of this same region (Mineka, Watson, & Clark, 1998)). Although there is
an impressive array of evidence to support the tripartite model, one problem is that
anxiety disorders are heterogenous and subsume a large variety of symptoms; a single
factor such as anxious arousal does not seem to account for the diversity of symptoms
subsumed by anxiety disorders.

Barlow and his colleagues (1996) proposed a three-factor hierarchical model to
account for the broad array of symptoms seen in anxiety disorders. They suggest that
anxious apprehension, which essentially represents the negative affect component of the
tripartite model, represent the shared component (the higher order factor) of all anxiety
disorders, as well as depressive disorders. In addition to the shared component, each
anxiety disorder contains a speciﬁc characteristic that distinguishes it for the others.

Mineka, Watson and Clark (1998) suggest a comprehensive model that integrates
Clark and Watson’s (1991) tripartite model and Barlow’s hierarchical model of anxiety
disorders. In this model each syndrome has both common and unique components. The
shared component is negative affect or general distress, which is common to both anxiety
and mood disorders and is responsible for the overlap between these disorders. Each
disorder also includes a unique component that distinguishes it ﬁ'om the other disorders.
For example, anhedonia, disinterest, and absence of positive affect represent the unique
components of depressive disorders. Anxious arousal is no longer considered a general
characteristic of anxiety disorders, but rather it assumes a more limited role as a
component of certain anxiety disorders. Mineka and colleagues propose that all of the
anxiety disorders, with the possible exception of Generalized Anxiety Disorder contains a

large amount of general distress variance and have their own unique component.

63

However, for the most part, they have not yet speciﬁed what these speciﬁc components
are for each of the anxiety disorders.

This study seems to lend evidence for a three factor model of the relationship
between depression and anxiety. The ﬁndings of this study suggest that there is a
signiﬁcant overlap between depression and trait anxiety but that state anxiety can be
distinguished ﬁom depression. State anxiety refers to transient feelings of tension,
apprehension, nervousness, worry, and activation or arousal of the autonomic nervous
system. State anxiety seems to deﬁne what Clark and Watson (1991) considers the unique
characteristics of anxiety or what or what Mineka and colleagues (1998) propose as the
speciﬁc component of panic. Trait anxiety refers to more enduring anxiety levels, distress
and anxiety proneness. Hence trait anxiety is more representative of negative affect, what
Clark and Watson and colleagues refer to as the common factors of anxiety and

depression.

The Manipulation

Again, recall that before the participants completed the memory task they heard
the manipulation instructions. These instructions were given in an effort to increase
anxiety (anxiety instructions) or decrease anxiety (relaxation instructions) before the
participants completed the memory tasks. The instructions were counterbalanced using a
within-subjects design (anxiety-relaxation vs. relaxation-anxiety). State anxiety measured
immediately after hearing the anxiety instructions was signiﬁcantly different than state
anxiety measured immediately after hearing the relaxation instructions in the anxiety-

relaxation group (Manxicty = 19.0 vs. Mrelaxaﬁon = 15.8). When the relaxation instruction

were read ﬁrst (relaxation-anxiety group), state anxiety measured immediately after the
hearing anxiety instructions was not statistically signiﬁcantly different than when it was
measured immediately after the relaxation instructions were read Mammy = 18.1 vs.
MJelaxation = 17.1). Notably, however, the pattern suggests that even in the relaxation-
anxiety group, individuals were more anxious after hearing the anxiety instructions than
after hearing the relaxation instructions. Overall, these results suggest that the
manipulation instructions used in this study had an impact on the participant’s levels of
state anxiety such that hearing the anxiety instructions increased state anxiety.

The effects of the manipulation instructions on memory performance were also
examined. When the anxiety instructions were read ﬁrst (anxiety-relaxation group), free
recall memory performance after hearing the anxiety instructions was signiﬁcantly lower
than free recall memory performance after hearing the relaxation instructions (Manxiety =
8.1 vs. Mnlmu-on = 8.8). When the relaxation instruction was read ﬁrst (relaxation-anxiety
group), free recall memory performance after hearing the relaxation instructions was not
statistically signiﬁcantly different than free recall memory performance after hearing the
anxiety instructions (Mrelmm-on = 8.1, Mammy = 7.9). Notably, the pattern is the same as
above, the participants’ free recall memory performance was better after hearing the
relaxation instructions than after hearing the anxiety instructions. Overall, the instructions
the participants heard (anxiety vs. relaxation) had an impact on their free recall memory
performance such that hearing the anxiety instructions hindered memory performance.
The impact of the manipulation instructions had a smaller impact on less complex tasks
(cued recall and recognition).

It is very interesting that minor alterations in the instructions impacted both

65

anxiety levels and memory performance in older adults. Recall that Ross (1968)
contrasted old and young subjects on a series of verbal paired-associate tasks. The tasks
were proceeded by one of three instructional conditions: supportive, neutral, or
challenging. Overall, older adults took longer to master the learning task than did their
younger counterparts with effects being greatest when the task was proceeded by the
challenging instructions. It was postulated that the anxiety or stress associated with the
challenging instructions interfered with the older subjects performance but this was not
the case for younger participants. It is possible that the anxiety provoking instructions
used in this study had a similar affect to those observed by Ross and interfered with
memory performance by increasing anxiety or perhaps evoking some other physiological

or psychological process that hindered performance.

The Effects of State Anxith on Memoty Performance

Eysenck (1979) stated that anxiety clearly affects the amount of memory capacity
available for task performance. However, the inﬂuence of anxiety on cognitive
performance is extremely understudied in older adults (Salzman & Lebowitz, 1991).
Several investigations have been conducted evaluating the relationship between anxiety
and sustained attention and memory in younger adults but there is very little research of
this type with older populations (Rankin, Gilner, Gfeller, J. D, Katz, 1994). It was
therefore essential that we investigate the inﬂuence of anxiety on memory performance in
older adults.

The Yerkes-Dodson Law (1908) describes the relationship between performance

efﬁcacy, arousal level, and task difﬁculty. The ﬁrst part of the law states that there is a

66

curvilinear relationship between arousal and performance, with the optimal levels of
performance being associated with intermediate levels of arousal. The second statement
is that the optimal level of arousal varies inversely with the level of task difﬁculty.
Therefore arousal (and perhaps anxiety) will interact with task difﬁculty. Eysenck
suggested that anxiety activates two competing processes. First, worry will impair
memory performance because it preempts some of the available space in working
memory. On the other hand, the increased task demands cause an increase in effort,
which enhances performance by increasing attentional capacity.

Based on the theories of Yerkes-Dodson and Eysenck it was predicted that a
differential relationship of memory performance as a frmction of level of anxiety would
appear, with the optimal levels of performance being associated with intermediate levels
of state anxiety. Further, it was predicted that there would be a differential relationship of
memory performance as a function of state anxiety at different levels of task complexity
with optimal performance occurring on a task of low complexity (e.g., recognition).
These hypotheses were not conﬁrmed, there were no signiﬁcant within-subjects effects of
anxiety on memory performance. Furthermore, the predicted curvilinear relationship
between anxiety and memory performance did not occur. The scatterplots suggested a
more linear relationship between anxiety and performance such that as state anxiety
increases memory performance decreases. Notably, the Yerkes-Dodson law suggests a
curvilinear relationship between performance and arousal. It is difﬁcult to know if the
measures of anxiety used in this study are valid indicators of autonomic arousal in the
way Yerkes-Dodson deﬁned it. In fact, trait anxiety was reported to only generally

correlate with blood pressure (Chesney et al., 1981). In addition, several studies with

67

older adults have failed to ﬁnd a curvilinear relationship between anxiety and memory
performance in older adults (e.g., Koenders et al., 1993; Fisk, 1996; Hill and
Vandervoort, 1992).

It is interesting to note that, although the correlations between state anxiety and
memory performance and the regression models were not signiﬁcant, there was a
consistent negative relationship between state anxiety and memory performance,
suggesting that state anxiety hinders memory performance. Notably, this negative
relationship was largest on the most complex task (free recall) and smallest on the least
complex task (recognition). See table 2 for a summary of the correlations between state
anxiety and performance.

One possible explanation for these results is that older adults have diminished
efﬁciency of inhibitory processes compared to younger adults (Hasher & Zacks, 1988).
On a variety of cognitive tasks (e.g., reading with distraction, negative priming, directed
forgetting) Hasher, Zacks and colleagues (1988; Hasher, Zacks, & May, 1999) have
shown that older adults are more affected by task irrelevant information; this provides
evidence of a decreased inhibitory efﬁciency. Hasher, Zacks and colleagues have
proposed that inhibitory processes are vital for working memory performance because
they prevent access of goal irrelevant information to working memory, delete no longer
relevant information from working memory, and restrain prominent responses from
dominating working memory. If these inhibitory processes break down, working memory
will become cluttered with information that is not relevant to the task at hand. Anxiety as
measured in this study could have aspects that evoke thoughts that are not germane to the

task, especially thoughts of worry about performance and cognitive deﬁcits. It is possible

68

that older adults are less able to inhibit these thoughts, that may be triggered by state
anxiety, compared to younger adults; thus they perform more poorly overall. This would
be consistent with the negative correlations we observed between anxiety and
performance. Anxiety may have different effects for older adults than for younger adults
because the impairing aspects of anxiety will overwhelm any positive effects (e. g.,
increased effort).

Anxiety had the strongest correlation with free recall memory performance and
the weakest correlation with recognition performance. Recall, that Craik (1986) found
that age differences in memory performance are smallest on recognition tasks and largest
on free recall tasks. Furthermore, Craik and McDowd (1987) found that the resource
demands of a free recall task as compared to a recognition task were differentially greater
in older adults. Performance on a secondary task showed that free recall was associated
with greater resource costs than was recognition and this effect was ampliﬁed by age.
Therefore, goal irrelevant thoughts are more likely to interfere with more complex tasks
like free recall than simpler tasks such as recognition.

Salthouse (1988, 1990) offers another explanation for the declines in older adults
memory performance. He attributes age-related changes in cognitive frmctioning to a
reduction in the quantity of some essential processing resource. Processing resource
deﬁcits have been proposed in attentional capacity, working memory capacity, and speed
of processing. Salthouse (1990) suggests that there are three key aspects to the process
resources hypothesis for declines in cognitive functioning with age. The ﬁrst assumption
is that the resources are important determinant in age differences in cognition. The

second key aspect is that the processing resources are assumed to enable or enhance

69

cognitive ﬁrnctioning. Finally, the resources are thought to be relevant to many types of
cognitive tasks. In many ways, the notion of processing resources is similar to the
concept of g in psychometric intelligence. Both g and processing resources were
proposed to account for commonalties of performance in what appears to be distinct
types of cognitive tasks. The processing resource interpretations of age differences in
cognitive functioning have been investigated using:
a) Secondary task procedures to obtain estimates of the reserve processing
capacity, or residual resources, available to adults of different ages
b) The examination of the relationship between hypothesized resource demands
and task performance in different age groups
c) The use of statistical techniques to examine the effects of controlling an index
of resource quality on the magnitude of age differences in cognitive
performance.
Salthouse (1990) reported that at least some of age related differences in cognitive
functioning could be attributed to a reduction in simultaneous storage and processing of
information. While Salthouse does not address the impact of anxiety on storage and
processing of information, it is certainly possible that reductions in storage and
processing of information is impacted differentially as a consequence of anxiety.
Another possibility for the pattern of results found is that older adults may come
into the study at a motivation level or an anxiety level that prevents any beneﬁcial effect
of anxiety from occurring. In other words, older adults may be performing at or near their
maximum effort and thus anxiety only has a negative effect on their performance.

Participants in this study were expected to be highly motivated because of the way they

70

were recruited and the instructions they were given at the outset of the study. Older adults
that participated in this study were ones that responded to advertisements looking for
those who have concerns about their memory. Therefore, these individuals could already
be highly anxious about their memory performance, so any additional anxiety the testing
situation or manipulation instructions induced would not have a positive effect on
performance. Kahneman (1973) noted that as the processing demands increase the
discrepancy between the supply of effort and task demands will grow. Therefore the
increased effort by highly anxious individuals will only partially compensate for the
greater processing demands on working memory by task irrelevant information. Thus, we
would ﬁnd the trend we observed with consistently negative correlations between anxiety
and memory performance.

Anxiety has several components including feelings of distress, worry,
apprehension, self-criticalness, having a negative self-view of oneself as well as
physiological hyperarousal. Perhaps self reported measures of anxiety only tap into
certain aspects of anxiety. Speciﬁcally, perhaps individuals are not good at self-reporting
the physiological aspects of anxiety. Although the State-Trait Anxiety Inventory is highly
face valid measure and has a high degree of internal consistency it may not be highly
correlated with physiological measures of anxiety (e. g., blood pressure, heart rate,
adrenaline level, and galvanic skin response). The State Trait Anxiety Inventory manual
(Spielberger, Gorsuch & Lushene, 1983) does not provide validity data to suggest that
this measure correlates with physiological measures of anxiety. In fact, trait anxiety was
reported to only generally correlate with blood pressure and was found to have no

systematic relationship with Type-A behavior (Chesney et al., 1981). It is possible that it

71

is the physiological hyperarousal that has the most signiﬁcant impact on memory
performance rather than feelings of worry and distress. In fact, the original Yerkes-
Dodson law describes the relationship between performance efﬁcacy and arousal level. If
self-reported anxiety measures, such as the STAI, are not tapping into the physiological
arousal component of anxiety we would not expect to observe the predicted curvilinear
relationship suggested by Yerkes-Dodson. Future research should use both physiological
and self-report measures of anxiety when investigating the relationship between anxiety

and memory performance.

Strategy Use

Since state anxiety is believed to restrict working memory capacity it was thought
that it would also reduce the extensiveness of elaboration and encoding of information
because these processes require available working memory capacity (Mueller, 1979).
Therefore, it was expected that anxiety would affect strategy use. It was predicted that
optimal levels strategy use would occur at intermediate levels of state anxiety. This
hypothesis was not conﬁrmed; strategy use was not signiﬁcantly correlated with state
anxiety levels. It is thought that perhaps the strategies measured on the CVLT, serial
strategy use, which involves learning the words in order, and semantic strategy use,
which involves learning the words in groups (e.g., fruits), are overlearned. Since these
strategies are fairly easy and have probably been used by the participants all their life,
they may not require additional space in working memory and hence will not be subject

to the effects of state anxiety.

72

Trait Anxiety

Previous studies have found decrements in memory span with state anxiety
(Mandler, 1979), but the results are somewhat more inconsistent for trait anxiety
(Eysenck, 1979). State and trait anxieties are separate but related constructs. Since state
anxiety and trait anxiety are not always related to one another, and individuals who are
high in trait anxiety do not experience increases in state anxiety in all situations, an
interesting question is how these two types of anxiety interact to produce changes in
memory performance. One possibility is that increased levels of trait anxiety can heighten
the effects of state anxiety and hence increase the detrimental effects of anxiety on
working memory capacity. Trait anxiety may lead to hyperarousal to environmental
demands (e. g. memory demanding tasks) and a decreased capacity to adjust to the
demand (Fleming, Batun, Davidson, Rectanus, & McArdle, 1987). On the other hand, it
has been proposed that exposure to previous or chronic stress may have a beneﬁcial
effect (H. Eysenck, 1983). The inoculation theory proposes that exposure to a chronic
stress increases our resistance to subsequent challenges.

This study allowed us to investigate the combined effects of state anxiety, worry
and trait anxiety on performance. We expected trait anxiety and state anxiety to interact
to predict memory performance. Speciﬁcally, we predicted that individuals who
experience state anxiety, as measured by the STAI-S at the time of testing and who have
high levels of trait anxiety, as measured by the STAI-T, to experience exaggerated effects
of anxiety on memory performance. This study also allowed us to investigate the
interaction between the trait of worry, as measured by the Penn State Worry

Questionnaire, and state anxiety.

73

As expected individuals with higher levels of trait anxiety tended to have higher
levels of state anxiety at statistically signiﬁcant levels. Approximately 30 percent of the
variance in state anxiety was accounted for by its linear relationship with trait anxiety as
measured by the STAI-T. However, the combination trait anxiety, worry, and state
anxiety did not signiﬁcantly predict performance and only accounted for 10% of the
variance in short delayed free recall and 9% of the variance in long delay free recall
performance. Notably, the combined effect of trait anxiety, worry, and state anxiety did
explain slightly more of the variance in short delay free recall performance than did state

anxiety alone (10% vs. 6%).

Memoty Complaints

A positive correlation between anxiety and memory complaints was hypothesized.
Previous studies suggested that depression is an important component of memory
complaints among elderly adults (Kahn, Zarit, Hilbert & Niederehe, 1975; Popkin,
Gallagher, Thompson & Moore 1982; Dartigues, Mazaux. Dequae, Letenneur, Giroire,
Barberger-Gateau, 1994; Poitrenaud, Malbezin & Guez, 1989). However, ther is
insufﬁcient research on the inﬂuence of anxiety on memory complaints in older adults
(Salzman & Lebowitz, 1991). In addition, most of these studies examined clinical
populations, such as individuals suffering ﬁ'om depression (Popkin, Gallagher, Thompson
and Moore, 1982) or late onset dementia (Kahn, Zarit, Hilbert & Niederehe, 1975). There
were some doubts that results of these studies should be generalized to the majority of
older adults who are living independently in the community. For instance, it is thought

that clinical patients tend to complain more about memory problems and that those with

74

dementia have inaccurate perceptions of their memory ability. It seemed important to
examine the effect of anxiety on memory complaints in normal healthy older adults. Trait
anxiety as measure by the STAI-T correlated signiﬁcantly with memory complaints;
however, state anxiety and worry (as measured by the PSWQ) did not correlate
signiﬁcantly memory complaints.

On shortcoming of this study is the timing for the administration of the measures.
State anxiety (STAI-S) was measured at the time of testing. In the interest of time,
measures of trait anxiety and worry (STAI-T and the PSWQ) and the memory complaints
measure GVIAC-S) were characteristically sent home with the participant and returned a
few days later. State anxiety is an emotional reaction that exists in a given moment in
times at a particular intensity. Spielberger, Gorsuch and Lushene (1983) reported that the
STAI-S scale has been found to be a sensitive indicator of changes in transitory anxiety.
For example, Spielberger, Gorsuch and Lushene (1970) found that the mean state anxiety
scores were substantially higher for both men and women when they were asked to
respond with how they ‘Vvould feel just prior to the ﬁnal examination in an important
course” than when given the standard instructions. Since state anxiety is highly
responsive to stress, it is unlikely that the participants level of state anxiety was the same
while they were in their own home completing the memory complaints questionnaire as
when they were completing the memory assessment. Given this consideration, this study
does not provide the opportunity to make valid conclusion about the relationship between
state anxiety and memory complaints. Future research examining the relationship
between state anxiety and memory complaints should obtain a measure of both at the

same time.

75

This study also found that the trait of worry as measured by the Penn State Worry
Questionnaire (PSWQ) did not signiﬁcantly correlate with memory complaints but that
trait anxiety as measured by the State-Trait Anxiety Inventory was signiﬁcantly
correlated with memory complaints. One possible explanation for this is the type of items
on these anxiety scales. The PSWQ was designed to identify chronic worriers who tend
to worry in general without restricting the topic (Molina & Borkovec, 1994). Hence, the
items tend to be very general (e.g., my worries overwhelm me, and I worry all the time).
Whereas the State-Trait Anxiety Inventory trait scale addresses many of the symptoms
of anxiety, and in particular, there is a heavy emphasis on the cognitive symptoms of
anxiety (e.g., I have disturbing thoughts, it is easy for me to make decisions, and some
unimportant thoughts run though my mind and it bothers me). It may be that the heavy
loading of cognitive questions on the STAI-T leads to its strong correlation with memory
complaints. In addition, the population sampled tended to have fairly low scores on the
PSWQ M = 35.3). Molina and Borkovec (1994) reported that the mean worry score in a
sample of 1323 individuals who were not chosen on the basis of any criteria to be 47.65.
This suggests that the cohort examined in this study tend to have very low levels of worry
and may not be representative of the population at large. Therefore, it may be beneﬁcial
for future research to examine the relationship between worry and memory complaints in
populations with more average levels of worry before ruling out the possibility that there

is a relationship between these two variables.

Limitations of Study

Some limitations of this study should be noted. First, the population is somewhat

76

exceptional, and the generalizability of the results may be limited accordingly. Our
participants were better educated than the national average and many were drawn from
higher socioeconomic levels, and were generally in good physical health. These
individuals also likely had access to high quality health care. Such factors could play in
the role of preserving memory functioning and decreasing anxiety and depression.
However, it could be argued that that this group actually provides a purer test of the
relationship between anxiety and objective performance because the potential inﬂuences

of low education, chronic health problems, and so forth were reduced.

Future Research

Himmelfarb and Murell (1984) reported that 17.1 % of males and 21.5 % of
females experience symptoms of anxiety severe enough to warrant some form of
therapeutic intervention (Himmelfarb & Murell, 1984). Prevalence rates of Age Related
Cognitive Decline, a condition in which includes both memory dysfunction and
subjective complaints, have been reported to be 34.9% and 55.8% in random samples of
elderly adults (Lane & Snowdon, 1989; Reinikainen et al., 1990). Given the large
prevalence rates of anxiety and memory problems in older adults it seems imperative that
we continue to investigate these two conditions at both a symptom and syndromal level.
It is a commonly held belief that our affect inﬂuences our cognitive abilities and the
inﬂuence of anxiety on cognitive performance and cognitive complaints is extremely
understudied in older adults (Salzman & Lebowitz, 1991). It is important to note that
although anxiety was generally not a signiﬁcant predictor of memory performance in this

study, the pattern of results strongly suggests that anxiety has a deleterious effect on

77

memory performance particularly on more complex tasks. In addition, the ﬁndings
suggest that anxiety invoking instructions hinders performance. There were only 47
participants in this study. Continuing this type of work is crucial because it is thought that
these effects would reach statistical signiﬁcance with a larger sample size. An
understanding of how anxiety effects memory performance is essential because diagnosis
and treatment decisions for dementia are in part based upon the presence of disturbed
affect (Zarit, 1980).

As mentioned early it is difﬁcult to know if self-report measure of anxiety tap into
the physiological arousal component of anxiety. Therefore, future research should use
both physiological and self-report measures of anxiety when investigating the
relationship between anxiety and memory performance. It may also be possible to also
use imaging data as a physiological measure of anxiety. Also, this study did not measure
state anxiety and memory complaints at the same time. Since state anxiety is highly
responsive to stress it is unlikely that the participants’ level of state anxiety was the same
while they were in their own home completing a self report packet questionnaires
(including the memory complaints measure) as when they were completing the memory
assessment. Future research examining the relationship between state anxiety and

memory complaints should obtain a measure of both at the same time.

78

Table 1

Coorelations between anxith and depression
BDITOTAL GDSTOTAL

BDI Total Score 1.000 .793"
GDS Total Score .793“ 1.000
STAI-S .254 .209

STAI-T .342“ .387”
PSWQ .411" .458“

** Correlation is signiﬁcant at the 0.01 level (2-tailed).
* Correlation is signiﬁcant at the 0.05 level (2-tailed).

79

STAI-S
.254
.209

1.000

.430"

.392“

STAI-T
.342*
.387"
.430"
1.000
.482"

PSWQ
.41 1
.458**
.392
.482**
1.000

Table 2

Rotated Component Matrix- Factor Loadings for Depression Md Anxietv Meesures

Component

Depression Anxiety

BDI Total Score .911 .182
GDS Total Score .924 .192
STAI-S Total Score 7.894E-03 .845
STAI-T Total Score .267 .759
PSWQ Total Score .419 .654

 

Notes: Extraction Method: Principle Component Analysis.

Rotation Method: Varimax with Kaiser Normalization.

80

Table 3

CVLT performance means by order and condition

Condition
m Relaxation
Orient Free recall 8.1 8.8
Cued recall 9.6 9.4
Qr_d_e_r_2 Free recall 7.9 8.1
Cued recall 8.9 8.9
Recognition* 12.91 13.74

Notes: Order 1 = anxiety instructions followed by relaxation instructions
Order 2 = relaxation instructions followed by anxiety instructions
* Recognition performance was measured only at Time 2 was subject only to

condition (anxiety vs. relaxation) and not order

81

Figure 1

Relationship between state anxiety and free reﬁll performsnceﬁat the 1st manipulation

 

 

 

 

14
a
12‘ a o a
canon

_ 10* a on a
g o o a a
g Br 0 a o
g o no a can
u“ Gr 0 o a a a
>
g a a
0 4d a a
t:
O
.C
U) 2: o

OI D

-2

0 10 20 30

State Anxiety Time 1

82

Figure 2

Relationship between state anxiety esp ﬁee recall performance at the 2nd manipulation

Long Delay Free Recall

 

16

141

121

 

 

 

State Anxiety Time 2

83

30

 

Figure 3

Relationship between state anxiety and cued recall performance et the lst manipulation

16

 

14! CI 0

12: a no a

101

8: o a a

6'l a a no

Short delay cued recall

4|

2-

 

 

 

State anxiety time 1

 

84

Figure 4

Relationship between state anxith and cued recall performance at the 2nd manipulation

 

 

 

 

16
o
14!
n

12: o a o o
E o o a a
s. 101 D O O
8 a a a a a a o a
a 8: n a r: r: a a
>
L.“ a o o o a
Q)
U 6! a a a o
O)
C
3

41 o

2: Cl

0

0 10 20 30
State anxiety time 2

 

85

Figure 5

Relationship between state anxieg and recogpition performance

Recognition

 

20

18-

16-

14-

12-

10:

8-

 

 

State Anxiety time 2

 

30

 

86

Table 4

Pearson Correlations Between Memog Performance and State Anxieg

Short Delay Free
Recall (SDFR)
Short Delay Cued
Recall (SDCR)
Long Delay Free
Recall (LDFR)
Long Delay Cued
Recall (LDCR)
Recognition Hits (RH)
Anxiety at
manipulation 1 (M1)
Anxiety at

manipulation 2 (M2)

S’JB

l .000

.830M

.845”

.792"

.240

-.217

-.295"'

S_DC_R

.830M

1 .000

.831"

.847"

.264

-.111

-.203

LDFR

.845”

.831M

1 .000

.872"

.282

-.124

-.265

LDCR

.792M

.847“

.872‘Ml

l .000

331*

-.078

-.l46

IE

.240

.264

.282

.331"

1.000

.105

-.037

-.217

-.111

-.124

-.078

.105

l .000

.741"

-.295*

-.203

-.265

-.l46

-.037

.741"

l .000

 

Note. ** Correlation is signiﬁcant at the 0.01 level (2-tailed).

* Correlation is signiﬁcant at the 0.05 level (2-tailed).

87

Figure 6

Relationship between state anxiety Md semantic cluster strategy use

 

3.5
3.0 I
2.5 I
2.0 l

1.51

Semantic Cluster Ratio

1.01

 

 

 

0 Cl 0
D O
D D
D D
D
D
D
0
4'0 60
Percent State Anxiety

88

100

Figure 7

Relationship between state anxiety end seriel cluster strategy use

Serial Cluster Ratio

 

 

 

 

8
O
6.
O
4 " g o a
O O
O
o a D
2 . ° 8
D a 0 Ch 0 a
°aa '3 o
o o a a o
0 - a
-2
0 20 40 60 80
Percent State Anxiety

89

100

Table 5

Pearson Correlations Between Free Recall Memogt Performance and Anxieg

SDFR LDFR STAI-T PSWQ STAI-SI STAI-

Short delay free recall 1.000
(SDFR)

Long delay free recall .845 ** 1.000
(LDFR)

Trait Anxiety (STAI-T) .030 .047

Penn State Worry -.214 -.106
Questionnaire (PSWQ)

State Anxiety 1 -.217 -.124
State Anxiety 2 -.295* -.265

.845 ** .030

-.214 -.217
.047 -. 106 -. 124
1.000 .550 ** .280
.550 ** 1.000 .324*
.280 .324“ 1.000
.167 .202 .741“

_S_2_

-.295*

-.265

.167

.202

.741”

1 .000

 

Note: * Correlation is signiﬁcant at the 0.05 level (2-tailed).

** Correlation is signiﬁcant at the 0.01 level (2-tailed).

90

REFERENCES

91

REFERENCES

Alexopoulos, G. (1991). Anxiety and depression in the elderly in the elderly. In C.
Salzrnan, B. D., Lebowitz (Eds). Anxith in the Elderly: Treatment end Resem. New
York: Springer.

Alloy, L., Kelly, K., Mineka, S., & Clements, C. (1990). Comorbidity in anxiety
and depressive disorders: a helplessness/hopelessness perspective. See Maser &
Cloninger 1990, pp. 499 - 543.

Baddeley, A. D., & Hitch, G. (1974). Working Memory. In G. H. Bower (Ed.),
Recent Advances in Learning and Motivation (V 01. 8). London: Academic Press.

Barlow, D. H., Chorpita, B. F., Turovsky, J. (1996). Fear, panic, anxiety and
disorders of emotion. Nebraska Smposium of Emotiop, 43, 251 -328.

Beck, A. T., Rush, A. J ., Shaw, B. F., & Emery, G. (1979). Cogpitive Ther_apy of
Depression. New York: Guilford Press.

Beck, A. T ., Ward , C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An
inventory of measuring depression. Archives of General Psychiapy, 4, 561-5 71.

Beck, J., Stanley, M., and Zebb, B. (1995). Psychometric properties of the Penn
State Worry Questionnaire in older adults. Journal of Clinical Geropsychology, 1, 33-42.

Blazer, D. G. (1994). Dysthymia in community and clinical samples of older
adults. American Journal of Psychiag, 151,1567-1569.

Blazer, D., George, L., Hughes, D. (1991). The epidemiology of anxiety
disorders: an age comparison. In C. Salzrnan, B. D., Lebowitz (Eds.). Anxiety in the

Elderly: Treatment and Research. New York: Springer.

92

Bolla, K. I., Lindgren, K. N., Bonaccorsy, C. & Bleeker. (1991). Memory
complaints in older adults: Fact or ﬁction? Archives of Neurology, 48, 61-64.

Butler, R., & Lewis, M. (1977). Agjpg and Mental Health. Saint Lewis: The CV.
Mosby Company.

Burns, D. D. & Eidelson (1998). Why are depression and anxiety Correlated? A
test of the tripartite model. J ournsl of Consulting and Clinical Psychology, 66, 461 -47 3.

Caine, E. D. (1986). The neuropsychology of depression: The pseudodementia
syndrome. In 1. Grant & K. M. Adams (Eds.)., Neuropsycholgical Assessment of
Neuropsychotic Disorders (pp.221-243). New York: Oxford University Press.

Cattell, R. B. & Scheier, I. H. (1961). The Meaning of Measurement of
Neuroticism and Anxieg. New York: Roland Press.

Cavanaugh, J. C. & Murphy, N. Z. (1986). Personality and Metarnememoey
correlates of memory performance in younger and older adults. Educational Gerontology,
1_2, 385-394.

Cavanaugh, J. C. & Poon, L. W. (1989). Metamemorial predictors of memory
performance in young and older adults. Psychology and Aging, 4, 365-368.

Chesney, M. A., Black, G. W., Chadwick, J. H., & Rosenman, R. H. (1981)

Psychological correlates of type A behavior pattern. Journal of Behavioral Medicine 4,

 

217 - 229.

Clark, L. A. (1989). The anxiety and depressive disorders: descriptive
psychopathology and differential diagnosis. In Anxiety and Depression: Distinctive and
Overlapping Features, (Eds. PC Kendall & D Watson) PP. 83 - 129. San Diego:

Academic.

93

Clark, D. A., Steer, R. A., & Beck, A. T. (1994). Common and speciﬁc
dimensions of self-reported anxiety and depression: implications for cognitive and
tripartite models. Journal of Abnormal Psychology, 103, 645—654.

Clark, L. A., Watson, D. (1991). Tripartite model of anxiety and depression:
Psychometric evidence and taxonomic implications. Journal of Abnormal Psychology,
M, 316-336.

Clarkin, J. F ., & Kendall, RC. (1992). Comorbidity and treatment planning.
Jom of Clinical and Consulting Psychology, 60, 904-908.

Cockbum, J ., & Smith, P. (1994). Anxiety and errors of prospective memory
amoung older people. British Journal of Psychology, 85, 273-282.

Craik, F. I. M. (1983). On the transfer of infonnationfrom tempory to permanant
memory. Philosophical Transations of the Royal Sgieg of London, Series B302, 341-
359.

Craik, F. l. M., & McDowd, J. M. (1987). Age differences in recall and
recognition. Journal of Ex erimental P cholo : Learnin Memo an o 'tio
474 -479.

Craik, F. I. M., Naveh-Benjamin, M., Govoni, R., & Anderson, N. D. (1996). The
effects of divided attention on ecoding and retrieval processes in human memory. Mum!

of Experimental Psychology, 125, 159-180.

Caine, E. D. (1993). Should aging-associated cognitive decline be included in the

DSM IV? Journal of Neuropsychiag and Clinical Neurosciences, 5, 1-5.

Crook, T., Bartus, R. T., Ferris S. H. (1986). Age-associated memory impairment:

proposed diagnostic criteria and measures of clinical change-report of a National Institute

94

of Mental Health work group. Developmentel Neuropsychologv. 2. 261-276.

Crook, T. & Larrabee, G. J. (1990). A self-rating scale for evaluating memory in
everyday life. Psychology and Aging, 5, 48-57.

Dartigues, Mazaux. Dequae, Letenneur, Giroire, Barberger-Gateau, 1994

Davidson, H. A., Dixon, R. A., & Hultsch, D. F. (1991). Memory anxiety and
memory performance in adulthood. Applied Cogpttive Psychology, 5, 423-434.

Delis, D., Kramer, J ., Kaplan, J ., & Ober, B.(1987). The California Verbal
Learning Test. New York: Psychological Cooperation.

Feinstein, A. R. (1970). The pretherapuetic classiﬁcation of co-morbidity in

chronic disease, The Journsl of Chronic Disease 23, 455 - 468.

 

Jeste, Dilip, V., Alexopoulos, G. S., Bartels, S. J., et a1. (1999) Consensus
statement the upcoming crisis in geriatric mental health, Archives of General Psychiaﬂ,
5_6_, 848 -853..

Dixon, R. A., Hertzog, C., & Hultsch, D. F. (1986). The multiple relationships
amoung metamemory in adulthood (MIA) scales and cognitive abilities in adulthood,
Human Learning, 5, 165-177.

Dixon, R. A. & Hultsch, D. F. (1983). Metamemory and memory for text
relationships in adulthood : A cross validation study. Journal of Gerontology, 38, 682-
686.

Easterbrook, J. A. (1959). The effect of emotion on cue utilization and the
organization of behavior. Psychological Review, 66, 183-201.

Eysenck, H. J. (1983). Stress, desease and personality: The inocculation effect. In

C. L. Copper (Ed.), Stress Research (pp.121-146). New York: John Wiley & Sons.

95

Eysenck, M. W. (1979). Anxiety, learning, and memory: A reconceptualization.
Journal of Research in Personalig), 13, 363-3 85.

Eysenck, M. W. (1982). Attention and arousal: Cognition end Perfonnaice.
Berlin: Springer-Verlag.

Fisk, J. (1996). Age -related impairment in associative learning: The role of
anxiety, arousal and learning self-efﬁcacy. Personelity end Individuﬂifferences. 21.
675-686.

Fleming, 1., Baum, A., Davidson, L. M., Rectanus, E., & McArdle, S. (1987).
Chronic stress as a factor in physilogicalreactivity to challenge. melth Psycholo 6,
221-237.

Flint, A. J. (1994). Epidemiology and comorbidity of anxiety disorders in the

elderly. American Journal of Psychiag, 151: 640-649.

F olstein, M. F ., Folstein, S. E., & McHugh, P. R. (1975). Mini-mental state.
Journal of Psychiatric Research, 12, 189 - 198.

Friedhoff, A. J. (1991). Consensus development statement: Diagnosis and
treatment depression in late life. In L. S. Schneider, C. F. Reynolds, B. D. Lebowitz, &

A. J. Friedhoff (Eds.). Diagposis and treatment of depression in late life: Results of the

NIH Consensus Development Conference (pp. 493 - 511). Washington, D. C.: American
Psychiatric Press.

Gallagher, D., Nies, G. & Thompson, L. (1982). Reliability of the Beck
Depression Inventory with older adults. Journal of Clinical and Consulting Psychology,
ﬂ, 1 52- 1 53.

Gallagher D., Breckenridge, J ., Steinmetz, J. & Thompson, L. W. (1983). The

96

Beck Depression Inventory and Research Diagnostic Criteria: Congruence in an older
population. Journal of Clinical and Consulting Psychology, 51, 945-946.

Gallo, J ., J ., & Lebowitz, B. D. (unpublished manuscript). The epidemiology of
mental disorders in late life. Psychiatric Services.

Gass, C. (1995). MMPI-2 variables in attention and memory test performance.
Psychological Assessment. 8. 135-138.

Gurian, B., & Miner, J. (1991). Clinical presentation of anxiety in the elderly. In
C. Salzrnan, B. D., Lebowitz (Eds). Anxietv in the Elderly: T reetrnent end Resear_c_h.
New York: Springer.

Gurin, G., Veroff, J., & F eld, S. (1983). Americans View of Their Mental Health.
New York: Basic Books.

Hasher, L., & Zacks, R. T. (1979). Automatic and effortﬁrl processes in memory.
Journal of Experimental Psychology: General, 108, 356 - 388.

Hasher, L., & Zacks, R. T. (1988). Working memory, comprehension, and aging:
A review and a new view. In G. H. Bower (Eds), The Psychology of Learning and
Motovation, vol. 22, pp. 193 -225. San Diego: Academic Press.

Hasher, L., Zacks, R.T., & May, C. P. (1999). Inhibitory control, circadian

arousal, and age. In D. Cooper & A. Koriat (Eds) Attention and Performance XVII

Cogtp'tive Regplation of Performance: Integr_etion of Theoty & Applications. Cambridge,

 

MA: MIT Press.
Hertzog, C., Dixon, R. A., & Hultsch, D. F. (1990). Relationships between
metamemory, memory predictions, and memory task performance in adults. Psychology

and Aging, 5, 215-227.

97

Hill, R., D., & Vandervoort, D. (1992). The effects of state anxiety on recall
performance in older learners. Educational Gerontology, 18: 597-605.

Himmelfarb, S., Murell, S. (1984). The prevalence and correlation of anxiety
symptoms in older adults. Journal of Psychiaﬂ, 116, 159-167.

Himmelfarb & Murrell (1983). Reliability and validity of ﬁve mental health
scales in the older persons. Joumel of Gerontology. 38: 619-640.

Hockey, G. R. J (1986). Changes in operator efﬁciency as a function of
environmental stress, fatigue, and circadian rhythms. In K. R. Boff, L. Kaufman & J. P.
Thomas (Eds). Handbook of Perception on Human Performance (V 01. 2). New York:
Wiley.

Hodges, W. F., & Durham, R. L.(l972). Anxiety, ability, and digit span
performance. J ourpel of Personelim en_d Sociel Psychology,ﬁ, 401-406.

Hultsch, D. F., Hertzog, C., & Dixon, R. A. (1987). Age differences in
metamemory: resolving the inconsistencies. Men Joumel of Psychology, 4, 193-208.

Johnson, M. H. & Magaro, P. A. (1987) Effects of mood and severity of memory
processes in depression and mania. Psychological Bulletin, 101, 28- 40.

Johnstone, E. C., Bernhardt, M. R., Delgrado, A., et al. (1980). Neurotic illness
and its response to anxiolytic and antidepressant treatment. Psychological Medicine, 10:
321-328.

Jonker. C., Smits, C., & Deeg, DH. (1997). Affect-related meta-memory and
memory performance in a population based sample of older adults. Educational

Gerontology, 23, 115-120.

Kahn, R. L., Zarit, S. H., Hilbert, N. M. & Niederehe, M. A. (1975). Memory

98

complaints and impairment in the aged: The effects of depression and altered brain
function. Archives of General Psychiat_ry, 117, 326-328.
Kahneman, D. (1973). Attention and effort. Englewood Cliffs, NJ: Prentice Hall.
Katon, W., et a1. (1986). Panic Disorder: Epidemiology in primary care._J_oMal

of Femilv Practice 23, 233-239.

 

Kendall, P. C., Howard, B. L., & Epps, J. (1988). The anxious child: Cognitive

behavioral treatment strategies. Behavior Modiﬁcation 12, 281-310.

 

Kessler R. C., McGonagle, K. A., Zhao, S., Nelson, C. B., Hughes, M. et al.
(1994). Lifetime and 12-month prevalence of DSM-III-R psychiatric disorders in the
United States: results from the National Comorbisity Survey. Archives of General
Psychiat_ry, 51, 8 - 19.

Koenders, M., Passchier, J ., Teuns, G., van Harskamp, F., et al. (1993). Trait-
anxiety and achievement motivation are positively correlated with memory performance
in patients who visit geriatric outpatient clinic with arnnestic symptoms. Psychological
Reports, 73, 1227-1231.

La Rue, A. (1992). Aging and Neuropsychologjcal Assessment. New York:
Plenum Press.

Leibert, R. M. & Morris, L. W. (1967). Cognitive and emotional components of
test anxiety: A distinction and some initial data. Psychological Reports, 20, 975 - 978.

Lepore, S. J ., Miles, H. J ., & Levy, J. (1997). Relation of chronic and eposodic
stressors in psychological distress, reactivity, and health problems. International Journal
of Behavioral Medicine. 4. 39-59.

Levy-Cushman, J. S., & Abeles, N. (1998). A study of memory complaints in the

99

able elderly. Clinical Gerontologist,19, 3 - 24.

Lindesay, J ., Briggs, K., & Murphy, E. (1989). The Guys/Age Concern Survey:
Prevalence rates of cognitive impairment depression and anxiety in an urban elderly
community. British Journal of Psychiat_ry, 155, 317-329.

Lowenthal, M. F., Berkrnan, P. L., Buehler, J. A., Pierce, R. C., Robinson, B. C.
& Trier, M. L.(1967). Aging and mental disorder in San Francisco. Jossey Bass, San
Francisco.

Mandler, G. (1979). Steady state memory: What does the one-shot experiment
assess? Center for Human Information Processing Report, 84, 29.

Marcopulos, B. A. (1989). Pseudodementia, dementia and depression: Test
differentiation. In T. Hunt & C. J. Lindley (Eds). Testing Older Adults (pp. 70-90).
Austin, TX: Pro-ed.

Mash, E. J ., & Barkley, R. A. (1996). Child Psychopathology. Guilford Press:
New York.

McDougall, G. J. (1994). Predictors of Metamemory in older adults. N_ursir_rg

Research 43, 212-218.

 

McEwen, B. S., & Stellar, E. (1993). Stress and the individual: Mechanisms

leading to disease. Archives of Internal Medicine 153 2093-2101.

 

Meyer, T., Miller, M., Metzger, R., & Borkovec, T. (1990). Development and
Validation of the Penn State Worry Questionnaire. Behavior Research and Thegpy, 28,
487-495.

Miler, E. (1975). Impaired recall and memory disturbances in presenile dementia.

British Journal of Social and Clinical Psychology, 14, 73-79.

100

Miller, W. R., & Seligrnan, M. E. P. (1973). Depression and the perceptions of
reinforcement. Journal of Abnormal Psychology, 82, 62-73.
Mineka, S., Watson, D., & Clark, L. A. (1998). Comorbidity of anxiety and

unipolar mood disorders. Annual Review of Psychology, 49, 377 - 412.

 

Molina, S., & Borkovec, T. D. (1994). In G. C. L. Davey and F. Tallis (Eds).

Wome'pg: Perspective on Theogt, Assement and Treatment, Pennsalvania: John Wiley &
Son Ltd.

Mueller, J. H.(1979). Test anxiety and encoding and retrieval of information. In
I.G. Sarason (Ed.) Test anxieg: Theog, Research, and Applications, Hillsdale, NJ:
Erlbaurn.

Myers, J ., Weisman, M., Tischler, G., Holzer, C., Leaf, P., Orvaschel, H.,
Anthony, J ., Boyd, J., Burke, J ., Kramer, M., & Stoltzrnan, R. (1984). Six-month
prevalence of psychiatric disorder in three communities. Archives of General Psychiag,
511, 959-967.

Nelson, D., Novy, D. (1997). Self-report differentiation of anxiety and depression
in chronic pain. Journal of Personalig) Assessment, 69, 392-407.

Niederehe, G. (1986). Depression and memory impairment in the aged. In L. W.
Poon (Ed.), Handbook for Clinical Memory Assessment of Older AM (pp. 226- 237).
Washinton, DC: American Psychological Association.

Norris, F. H., & Murrell, S. A. (1988). Prior experience as a moderator of disaster

impact on anxiety symptoms in the elderly. Journal of Communist Psychology, 16, 665-
683.

Parry, H., Baltes, M., Mellinger, G., Cisin, I., & Manheirner, D. (1973). National

101

patterns of psychotherapeutic drug use. Archives of General Psychiag, 28, 769-783.

Pattemson, R., Sullivan, M., & Spielberger, C. (1980). Measurement of state and
trait anxiety in elderly mental health clients. J ournel of Behevioral Assessment. 2:89-97.

Paykel, E.S., Rowan, P. R., Roa, B. M., & Bhat, A. (1982). Atypical depression:
Nosology and response to antidepressants. In P. Clayton & J. Barrett (Eds), Treatment of
Dept'ession: Old Controversies and New Approm (pp. 237-263). New York: Raven
Press.

Poitrenaud, D. A., Malbezin, M., Guez, D. (1989) Self-rating and psychometric
assessment of age-related changes in memory among young elderly-managers.
Developmental Neuropsychology, 5, 285-294.

Popkin, S. J., Gallagher, D., Thompson, L. W. & Moore, M. (1982). Memory
complaint and performance in normal depressed adults. Experimental Agir_rg Research, 8,
141-145.

Rankin, E., Gfeller, J. D., & Gilner, F. H. (1993). Measuring anxiety states in the
elderly using the state-trait anxiety inventory for children. Journal of Psychiatric
Research, 27, 111-117.

Rankin, E. J ., Gilner, F., Gfeller, J. D., & Katz B. M. (1994). Anxiety states and
sustained attention in a cognitively intact elderly sample: preliminary results.
Psychological Remus, 7;: 1176-1178.

Regier, D. A., Boyd, J. H., Burke, J. D., Rae, D. S., Myers, J. K., Kramer, M.,
Robins, L.N., George, L. K., Kamo, M., & Locke, B. Z. (1988). One month prevalence of

mental disorders in the United States. Based on ﬁve epidemiologic catchment area sites.

Archives of General Psychiatty, 45: 977-986.

102

Robins, L., Helzer, J ., Weissman, M., Orvaschel, H., Gruenberg. E., Burke, J.
(1984). Lifetime prevalence of speciﬁc psychiatric disorders in three sites. Archives of
General Psychiag, 38, 381-389.

Rosavage, A. M. (1990). The able elderly: Memory, health, social support, and
psychological well-being. Unpublished doctoral dissertation, Michigan State University,
East Lansing.

Rosenbaum (1984)

Ross, E. (1968). Effects of challeniging and supportive instructions on verbal
learning in older persons. Journal of Educational Psychology, 59, 261-266.

Roth, M., Gurney, C., Garside, R. F., & Kerr, T. A. (1972). Studies in the
classiﬁcation of affective disorders: The relationship between anxiety states and
depressive illness. British Journal of Psychiat_ry, 121, 147-161.

Roth, M., Mountjoy, C. Q., & Caetano, D. (1982). Further investigations into the
relationship between depressive disorders and anxiety state. Pharmacopychait, 15, 135-
141.

Salthouse, T. (1988). Resource-reduction interpretations of cognitve aging.
Developmental Review, 8, 238 - 272.

Salthouse, T. (1990). Working memory as a processing resouce in cognitve aging.
Developmental Review. 10. 101 - 124.

Salzrnan, C. (1991). Treatment of Anxieg in Clinical Geriatric Psycho-
pharmacology. New York: McGraw-Hill Book Company.

Salzrnan & Lebowitz (1991). Anxieg and the Elderly Treatment and Research.

New York, N.Y.: Springer Publishing Company, Inc.

103

Schacter, S., & Singer, J. E. (1962). Cognitive, social and physiological
determinants of emotional state. Psychological Review. 69, 379-399.

Seligrnan , M. (1993). What You Can Change and Whet You Can’t: The
Complete Guide to Successful Self Improvement. Fawcett Columbine: New York

Shamoian, C. (1991). What is Anxiety in the Elderly. In C. Salzrnan, B. D.,
Lebowitz (Eds). Anxieg in the Elderly: Treatment and Research. New York: Springer.

Smith, S., Sherill, K. & Colenda, C. (1995). Assessing and Treating Anxiety in
Elderly Persons. Psychiatric Services, 46, 36-42.

Sorg, B. A., & Whitney, P. (1992). The effect of trait anxiety and situational

stress on working memory capacity. Journal of Resegch on Personality, 26: 235-241.

 

Spielberger, C., Gorsuch, R., & Lushene, R. (1970). Manual for the State-Trait
Anxiety Inventoty, Consulting Psychologists Press, Palo Alto, CA.

Spielberger, C., Gorsuch, R., & Lushene, R., Vagg, P. R., & Jacobs, G. A. (1983).
Manual for the State-Trait Anxiety Inventory, Mind Garden, Palo Alto, CA.

Spitzer; R. L., Endicott, J. & Robbins, E. (1978). Research Diagnostic Criteria:
Rationale and reliability. Archives of General Psychiat_ry, 35, 773-782.

Stavrakaki, C. & Vargo, B. (1986). The relationship of anxiety and depression: a
review of the literature. British Journal of Psychiagy, 149, 7-16.

Steer, R.A., Beck, A. T., Clark, D. A., & Ranieri, W. C. (1995). Common and
speciﬁc dimensions of self-reported anxiety and depression: A replication. Journal of
Abnormal Psychology, 104, 542-545.

Tranel, D., Benton, A., & Olson, K. (1997). A 10-Year Longitudinal Study of

Cognitive Changes in Elderly Persons. Developmental Neuropsychology, 13, 87 - 96.

104

Waldstein, S. R., Ryan, C, M., Jennings, J. R., & Muldoon, M. F. (1997). Self-
reported levels of anxiety do not predict neuropsychological performance in healthy men.
Archives of Clinical Neuropsychology, 12, 567-574.

Warheit, G. Bell, R., Schwab, J ., Buhl, J. (1986). An epidemiological assessment
of mental health problems in the southeastern United States. In M. W. Weissman, J. K.,
Meyers & C. E. Ross (Eds), Community Survey of Psychiatric Disorders (191-208).
New Brunswick: Rutgers University Press.

Wasylenki, D. A. (1987). Depression. In Wasylenki, Martin, Clarck, Lennox,
Perry, & Harrison (Eds), Psychogeriatrics: A Practical Handbook (41- 58). Toronto:
Gage Educational Publishing Company.

Watson, D. & Clark, L. A. (1995). Depression and the melancholic temperament.
European Journal of Personalig, 9, 351 -366.

Williams, A., Denney, N. W. & Schadler, M. (1983). Elderly adults' perception
into their own cognitive development during the adult years. International Journal of
Aging and Human Development. 16. 147-158.

B.J. Winer, Brown, D. R., Michels, K. M (1991). Statistical principles in
Experimental Design. New York : McGraw-Hill.

Wing, J. K., Copper, J. E., & Sartorius, N. (Eds). (1974). Measurement and
Classiﬁcation of Psychiatric Symptoms: An Instruction Manual for the PSE and Catego
mm. Cambridge: Cambridge University Press.

Winterling, D., Crook, T., Salama, M. & Gobert, J. (1986). A self-rating scale for
assessing memory loss, In A. Bes, J. Cahn, S. Hoyer, J. P. Marc-Vergnes & H. M.

Wisniewski (Eds), Senile Dementias: Early Detection (pp. 482-486). London: John

105

Libbey Eurotext.

Yerkes, R. M., & Dodson, J. D. (1908). The relation of strength of stimulus to
rapidity of habit formation. Journel of Compagtive Neurology of Psychology,_1_8_, 459-
482.

Yesavage, J ., Brink, T., Rose, T., Lum, O., Adey, V., & Leirer, V. (1983).
Development and validation of the geriatric depression screening scale: A preliminary

report. Journ_el of Psychietric Research 17, 37-49.

 

Zarit, S. H. (1980). Aging and mental disorders. Free Press, New York.
Zelinski, E. M., M. J ., Gilewski, & K. W., Schaie.(1993). Individual differences

in cross-sectional and 3-year longitudinal memory performance across the adult life span.

Psychology and Aging, 8, 176-186.

106

"Illlllllllllllll111113