V o ,
. .1
“1:2...

n 1‘
‘ ‘5
Lu

3;

d».
a;

”1:.

..
.1.

p
.1

 

.danBJ‘ml .t’

11‘.
1 _.

v ‘ , $1.9 bit...” 4. "I‘i I In“) - |’
La 4r.s.rkl::s . ‘ V V , . , §%..nw%rww.mfi.§wa? $.95

.!

 

 

This is to certify that the
dissertation entitled

THE EFFECTS OF ANXIETY, DEPRESSION, AND AGE ON
COGNITIVE FUNCTIONS IN OLDER ADULTS: A

PhD.

 

LONGITUDINAL STUDY

presented by

TERRY RANDYL BARCLAY

has been accepted towards fulfillment
of the requirements for the

degree in Clinical Psychology

 

lite. Mk

Major Professor’s Signature

6/ 111qu
Date

MSU is an Affirmative Action/Equal Opportunity Institution

 

. _. -. o--o-n-o-.-a-O-0-0-.-l-o-O-O-n-I-l-O-l-

..-.-.--a---a-n-n-a-q—t~L .

 

LIBRARY

MIChlgan State
University

 

 

 

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
DEC 0 8 2007

 

321993

 

 

 

 

 

 

 

 

 

 

 

 

 

6/01 c:/ClRC/DateDue.p65-p.15

 

. ———____

THE EFFECTS OF ANXIETY, DEPRESSION, AND AGE ON COGNITIVE
FUNCTIONS IN OLDER ADULTS: A LONGITUDINAL STUDY

By

Terry Randyl Barclay

A DISSERTATION

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

DOCTOR OF PHILOSOPHY
Department of Psychology

2004

ABSTRACT

THE EFFECTS OF ANXIETY, DEPRESSION, AND AGE ON COGNITIVE
FUNCTIONS IN OLDER ADULTS: A LONGITUDINAL STUDY

By

Terry Randy] Barclay

During the last decade, the relationship between negative affect states and
cognition in older adults has become a topic of tremendous research and clinical interest.
Despite a steady growth of empirical investigation in this area, however, the long-term
impact of depression and anxiety on neuropsychological function remains poorly
understood. This dissertation investigated the hypothesis that anxiety and depression
contribute to age-related cognitive decline in the areas of memory, executive function,
and processing efficiency. The sample included 57 community-dwelling older adults
between the ages of 57 and 92 (Time 1 M = 69.86, S_D = 6.36 years; Time 2 M = 75.09,
S_D = 6.47 years). All variables were measured using standard neuropsychological tests
of performance and a short demographic questionnaire. Mood was assessed using self-
report measures, including the Beck Depression Inventory (BDI), Geriatric Depression
Scale (GDS), and the State-Trait Anxiety Inventory (STAI). Subjects returned for
follow-up evaluation approximately 5 years after initial participation. Results of repeated
measures ANCOVA analyses revealed that higher levels of depressive symptomatology
were not predictive of cognitive decline above and beyond that already accounted for by
age alone. However, higher levels of trait anxiety were found to predict 9.9% (p<.05) of
the decrement in free recall memory, 8.2% (p<.05) of the decline in semantic clustering,

and 8.4% (p<.05) of the deterioration in executive functioning performance. Trait

anxiety was not found to predict decline in processing efficiency. Overall, these findings
are consistent with a small, but growing, literature demonstrating that trait anxiety is
associated with poorer verbal learning and executive dysfunction. Results suggest that
even subclinical levels of anxiety may interfere with neuropsychological performance
and may exacerbate the cognitive decline associated with normal aging. These findings
are discussed in light of their practical, theoretical, and methodological implications and

in the context of suggestions for future research.

TABLE OF CONTENTS

LIST OF TABLES .............................................................................. viii
INTRODUCTION .............................................................................. 1
Depression in Older Adults ............................................................ 6
Prevalence ...................................................................... 6
Etiology of Late-Life Depression ........................................... 10
Demographic Factors ................................................. 10
Gender ......................................................... 10
Family History ............................................... 12
Financial and Educational Status .......................... 12
Social Factors ......................................................... l3
Illness and Disability ....................................... 13
Life Events ................................................... 15
Bereavement ................................................. 16
Quality of Relationships .................................... 17
Summary of Etiological Factors ................................... 19
Effects of Depression on Cognitive Functioning ......................... 20
Cross-Sectional Studies ............................................. 20
Memory ...................................................... 20
Speed of Processing ........................................ 23
Executive Functioning ..................................... 24
Longitudinal Studies ................................................ 25

iv

Summary of Depression-Related Cognitive Decline ............ 27

Anxiety in Older Adults ............................................................... 28
Prevalence ...................................................................... 28
Etiology of Late-Life Anxiety .............................................. 32
Demographic Factors ......................................................... 33

Gender ....................................................... 33

Race .......................................................... 33

Social Factors ........................................................... 34

Illness and Disability ....................................... 34

Loss and Exposure to Traumatic Events ................ 35

Effects of Anxiety on Cognitive Functioning ............................. 37
Memory .................................................................. 38

Speed of Processing ................................................... 39

Executive Functioning ................................................ 40

Summary of Anxiety-Related Cognitive Decline ........................ 41
Specific Hypotheses to be Tested ................................................... 42
METHOD ....................................................................................... 44

Participants ............................................................................. 44

Procedures .............................................................................. 44

Instruments ............................................................................. 45
Screening Instrument ......................................................... 45

Mini-Mental State Exam (MMSE) ................................ 45
Demographic Questionnaire ................................................ 46

Measures of Mood ............................................................ 47

Beck Depression Inventory (BDI) ................................. 47

Geriatric Depression Scale (GDS) ................................. 48

State-Trait Anxiety Inventory (STAI) ............................ 49

Memory ........................................................................ 50
California Verbal Learning Test (CVLT) ........................ 50

Speed of Processing .......................................................... 51
Symbol Digit Modalities Test (SDMT) ........................... 51

Trail Making Test (TMT), Part A ................................. 52

Executive Functioning ....................................................... 52
Wisconsin Card Sorting Test (WCST) ........................... 52

Trail Making Test (TMT), Part B ................................. 53

Stroop Test (ST) ..................................................... 53

RESULTS ....................................................................................... 55
Descriptive Statistics .................................................................. 55
Power Analysis ......................................................................... 57
Missing Data Analysis ................................................................ 58
Hypothesis 1: Age and Cognitive Decline ......................................... 59
Factor Scores .................................................................. 6O
Hypothesis 2: Age and BDI Depression Predicting Cognitive Decline ....... 61
Free Recall Memory ......................................................... 61
Semantic Clustering .......................................................... 62
Processing Speed ............................................................. 62

vi

Executive Functioning ........................................................ 62

Hypothesis 3: Age and GDS Depression Predicting Cognitive Decline ....... 63

Free Recall Memory .......................................................... 64

Semantic Clustering ........................................................... 64

Processing Speed ............................................................. 64

Executive Functioning ....................................................... 65

Hypothesis 4: Age and Trait Anxiety Predicting Cognitive Decline ........... 65

Free Recall Memory .......................................................... 66

Semantic Clustering ........................................................... 66

Processing Speed .............................................................. 66

Executive Functioning ....................................................... 67

Hypothesis 5: Anxiety, Depression, and Global Cognitive Functioning ...... 67
MMSE and BDI Depression ................................................ 68

MMSE and GDS Depression ................................................ 68

MMSE and Trait Anxiety ................................................... 68

DISCUSSION .................................................................................. 7O
Methodological Considerations ..................................................... 81

Future Directions ...................................................................... 83
Summary .............................................................................. 84
APPENDIX .................................................................................... 86
LIST OF REFERENCES ..................................................................... 98

vii

TABLE 1

TABLE 2

TABLE 3

TABLE 4

TABLE 5

TABLE 6

TABLE 7

TABLE 8

TABLE 9

TABLE 10

TABLE 1 1

TABLE 12

LIST OF TABLES

Sample Characteristics at Time 1 and Time 2 .............................

Means and Standard Deviations for Subjects’ Time 1 Performance
on all Dependent Measures, including the MMSE, CVLT, SDMT,

TMT, WCST, and ST .........................................................

Means and Standard Deviations for Subjects’ Time 2 Performance
on all Dependent Measures, including the MMSE, CVLT, SDMT,

TMT, WCST, and ST .........................................................

Correlation Matrix Between Age and Time 1 Measures of Global
Cognition, Processing Speed, Executive Functioning, and

Memory .........................................................................

Correlation Matrix Between Age and Time 2 Measures of Global
Cognition, Processing Speed, Executive Functioning, and

Memory .........................................................................

Frequency of Depression at Time 1 and Time 2 as Measured by
The BDI and GDS ............................................................

Correlation Matrix Between Age, Executive Functioning (EF)
Factor, and Processing Speed (PS) Factor at Time 1 ....................

Correlation Matrix Between Age, Executive Functioning (EF)

Factor, and Processing Speed (PS) Factor at Time 2 .....................

Repeated Measures ANCOVA Results Showing Covariate Effects
of Age and BDI Depression Scores on Free Recall Memory,

Semantic Clustering, Processing Speed, and Executive Functioning...

Repeated Measures ANCOVA Results Showing Covariate Effects
of Age and GDS Depression Scores on Free Recall Memory,

Semantic Clustering, Processing Speed, and Executive Functioning...

Repeated Measures ANCOVA Results Showing Covariate Effects
of Age and Trait Anxiety on Free Recall Memory, Semantic

Repeated Measures ANCOVA Results Showing Covariate Effects
of Age, BDI Depression, GDS Depression, and Trait Anxiety on

MMSE Scores .................................................................

viii

87

88

89

9O

91

92

93

93

94

95

Clustering, Processing Speed, and Executive Functioning .............. 96

97

INTRODUCTION

During the last decade, the relationship between negative affect states and
cognition in older adults has become a topic of tremendous research and clinical interest.
Despite a steady growth of research in this area, however, important questions remain to
be adequately addressed. Key questions as yet unanswered include (1) What is the
strength of the relationship between symptoms of anxiety and depression and cognitive
decline in old age? (2) Is there a difference in the extent to which mild, moderate, or
severe affective symptoms impact cognitive functioning over time? (3) Which types of
cognitive processes are most likely to be affected by anxiety and depression or, stated
another way, do symptoms of anxiety and depression affect cognitive functioning in a
global manner or are there differential effects of these symptoms on various cognitive
domains? and (4) What is the nature of the interaction between age-related cognitive
decline and the decline associated with symptoms of anxiety or depression? Moreover,
given that the overwhelming majority of published studies in this area have been cross-
sectional in nature, the long-term effects of emotional disturbance on cognition in old age
have been ignored. From a research standpoint, as well as from the clinician’s
perspective, the need for longitudinal work in this line of investigation is clear.

While there has been an increase in the number of studies investigating
depression, anxiety, and cognitive functioning in elderly samples, the conclusions drawn
from recent research on these variables are equivocal. For example, with respect to
empirical work investigating the impact of depression on cognitive functioning in older
adult populations, a handful of recent studies have found significant differences in

cognitive functioning between those with and without depression. Memory, verbal

fluency, and naming tasks were identified as sensitive to the effects of depression in
several of these studies (King, Caine, Conwell, & Cox, 1991a, 1991b; LaRue, 1989;
LaRue, D’Elia, Clark, Spar, & Jarvik, 1986; Williams, Little, Scates, & Blockman, 1987).
In each of these studies, individuals with depression scored significantly lower than
control subjects, but significantly higher than comparison groups of patients with
Alzheimer’s disease. Therefore, the researchers concluded that the effects of depression
on cognition were relatively subtle. Other studies utilizing similar samples, however,
have found no differences between depressives and normals, except for the subjects’
perception of their memory, with depressives claiming more memory problems than
controls (Freehan, Knight, & Partridge, 1991; Rohling & Scogin, 1993). Results from
studies using medical patients with and without depression, and demented patients with
and without depression are equally contradictory. Among medical patients, three robust
studies reported no significant differences between depressives and nondepressives on a
variety of cognitive tasks (Bieliauskas, Boczar, & Lamberty, 1991; Bieliauskas, Costello,
& Terpenning, 1991; Bieliauskas & Glantz, 1989), whereas one longitudinal study found
that those with depression performed more poorly on cognitive screening tasks alone
(Parmelee, Katz, & Lawton, 1992). Similar disagreements can be found in research
involving patients with dementia. In one study, demented individuals with depression
displayed greater cognitive deficits than did the demented subjects without depression
(Pearson, Teri, Reifler, & Raskind, 1989), whereas another study found that the rate of
cognitive decline did not differ between depressed and nondepressed Alzheimer disease
patients (Lopez, Boller, Becker, Miller, & Reynolds, 1990). Making sense of these

conflicting reports is further complicated by the fact that almost all of the published work

has employed cross-sectional designs. Therefore, most researchers claim their results to
be tentative in nature and call for longitudinal studies to confirm their findings; however,
few researchers to date have conducted such studies.

The research examining the effects of anxiety on cognitive abilities is much more
sparse in the literature. Those studies that do exist are, unfortunately, equally
contradictory. Two studies (Costa, Fozard, McCrae, & Bosse, 1976; Schultz, Hoyer, &
Kaye, 1980) have reported no age differences with respect to the effect of trait anxiety on
cognition; however, one investigation (Cohen, Eisdorfer, Vitaliano, & Bloom; 1980)
reported an age by trait anxiety interaction, with elderly subjects’ performance more
adversely affected by high anxiety than that of middle-aged subjects. An interaction
between age and state anxiety has been demonstrated in one study of learning and
memory (Whitbourne, 1986), and inferred in a second investigation (Ross, 1986) where
older participants were more negatively influenced than younger subjects by
“challenging” instructions prior to learning. However, other investigations (Martin,
1984; Monge & Gardner, 1972) have reported that the elderly are not disproportionately
affected by heightened symptoms of anxiety.

The variable nature of the findings described above may be the result of
methodological inconsistencies and deficiencies among the studies. Small sample sizes
represent a major limitation to many of the conclusions drawn from cross-sectional
studies that have been published to date. In fact, many cross-sectional studies have used
fewer than 30 individuals with depressive symptoms in their investigations (Amadeo, &
Self, 1990; O’Boyle, Stoudemire et al., 1991; Pearson et al., 1989; Reynolds et al. 1987;

Rohling & Scogin, 1993) and this is clearly problematic. The wide variety of instruments

used to measure cognition may also help to explain inconsistent results, as some
assessment tools are far more sensitive to cognitive impairments than others. For
instance, a number of studies have used the Mini Mental Status Examination as their
measure of cognitive functioning (O’Boyle et al., 1990; Pearson et al., 1989) even though
that particular measure is known not to be sensitive to more subtle cognitive deficits in
areas such as memory, executive functioning, or speed of processing. While the studies
that use global measures of cognition nevertheless add to our understanding of the
association between these variables, researchers need to know which cognitive processes
are most readily influenced by anxiety and depression and which are relatively
undisturbed. In addition to lack of measurement specificity, the variety of ages used in
recent research may also explain some of the inconsistent results. Sample demographics
vary widely among published articles, with a wide variation in the interpretation of “older
adults”. For example, in their studies of patients with major depression, Williams et al.
(1987) used a sample with a mean age of 53.7 (SD = 11.1) and Rohling and Scogin
(1993) reported that their “old” group had a mean age of 64.9 years (SD = 3.6). In
contrast, Parmelee et al. (1992) used a sample with a mean age of 83.9 years (SD
unreported). As of yet, few researchers have speculated as to how the age of their
samples may be interacting with the cognitive decline they attribute to depression or
anxiety.

There are other problems in this line of research that also need to be addressed.
For example, the great majority of studies investigating depression or anxiety do so by
dividing their samples into dichotomous groupings (i.e. anxious versus nonanxious,

young versus old). The use of these variables as dichotomous has been challenged as

inappropriate, and an argument for using continuous variables has been strongly asserted
(Caine, Lyness, & King, 1993). It has been noted that, by comparing extremes on a
continuum, such as highly anxious versus not at all anxious, the nature of the relationship
between affective states and cognition may be distorted (Pedhazur, 1982). Utilizing
measures of anxiety, depression, and cognition as continuous variables thus may add to
the knowledge of the relationship between these constructs.

The present study seeks to add to the existing literature on the effects of anxiety
and depression on cognitive functioning in older adults by improving upon the
methodological and theoretical shortcomings of earlier studies. The proposed study will
do so in the following important ways: (I) examine these variables in a within-subjects
longitudinal design, thereby allowing for the direct examination of the long-term effects
of anxiety and depression on cognitive functioning (2) using neuropsychological
measures that assess the specific cognitive functions of memory, speed of processing, and
executive functioning as opposed to measures of global, non-specific types of cognition
and by (3) employing statistical analyses using continuous variables of anxiety,
depression, and cognitive functioning in order to avoid the distortion that may occur by
falsely dichotomizing these constructs.

It should be noted that the current study has been designed to examine the
potential impact of depression and anxiety on various neuropsychological fimctions and
therefore rests on the assumption that affective states may play a causal role in cognitive
decline. While it could be argued that the direction of causality is actually reversed in
some cases (i.e., signs of cognitive slippage actually precipitate feelings of

depression/anxiety in some individuals), the present study is limited to the examination of

whether anxiety or depression-proneness is related to eventual cognitive deterioration and
is based on the methodological framework of most recent investigations in the field.
Clearly, a great deal of further empirical work will be needed to more fully elucidate the
complex, bi-directional interactions that potentially exist between negative affect states
and cognition in older adulthood.

Before describing the specific methodology and hypotheses of the present study,
it will be necessary to briefly review the existing literature pertaining to the prevalence
and nature of depressive and anxiety-related symptoms in late life, to articulate the
factors that increase an elder’s risk of deve10ping symptoms of anxiety and depression,
and to then elaborate upon the conflicting evidence linking negative affective states in old
age to cognitive decline.

Depression in Older Adults
Prevalence

Late-life depression is a prevalent and serious condition that causes considerable
suffering for both the afflicted individual and their family. Although symptoms of
depression are common among the elderly, depression is not a normal part of aging and
should not be accepted as such. Depression in the elderly often coexists with other
physical illnesses, as well as anxiety, and thus patients and their primary care physicians
may miss or misinterpret common symptoms of depression. This comorbidity not only
confounds the diagnosis and treatment of depression, but also has been shown to have
detrimental effects on general health status and patient outcomes (Bruce & Hoff, 1994).
Though underdiagnosed and undertreated, it is possible to distinguish late-life depression

from other geriatric illnesses and then treat it accordingly (Katz, 1996). Late-life

depression, just like depression affecting younger individuals, is frequently a recurrent
illness that requires long-term treatment and a number of therapies have been proven safe
and effective (Kupfer, 1991). As the number of people over the age of 65 continues to
increase around the world, so does the importance of diagnosing and treating this
condition appropriately.

Recent research has been somewhat conflicting with regard to the prevalence of
depressive disorders in late life, though most agree that the numbers are significantly
greater than originally hypothesized. In fact, the prevalence of depression among older
adults is now reported to be second only to dementia among persons over the age of 65
(LaRue, 1992). Among community-dwelling residents in this age range, approximately
15% exhibit depressive symptoms, l-2% suffer from major depression, and 2% are living
with dysthymia (Blandchard, 1996). Noteworthy is the fact that much higher rates,
between 13-27%, are recorded for subsyndromal or “sub-clinical” depression among this
group (Johnson, Weissman, & Klerrnan, 1992; Judd et al., 1994). Although
subsyndromal depressions do no meet diagnostic criteria for major depression or
dysthymia (American Psychiatric Association, 1994), the existence of depressive
symptomotology is associated with an increased risk of developing major depression, a
physical disability or medical illness, and an elevated use of health services. As many as
50% of the medically ill elderly who are nursing home residents suffer from
subsyndromal depressions (Koenig & Blazer, 1996; Mossey, Knott, & Craik, 1990). The
course of subsyndromal depression varies, but when accompanied by a severe physical
disability, such as stroke, it may persist for long periods of time (Alexopoulos, Young, &

Shindeckler, 1992). Many of these individuals suffering from various disabilities must

live in nursing home settings and the prevalence of depression among nursing home
residents is even higher than that of the general population. Recent studies report that
approximately 15-25% of nursing home residents are suffering from either minor or
major depression (Lepine & Bouchez, 1998).

Although estimates of the prevalence of late-life depression may vary due to the
size of the study sample, the definition of depression used in the study, and the method of
diagnosis, it is clear that late-life depression is a common condition. If unrecognized and
untreated, depression can interfere with physical and social functioning, as well as
diminish general health and well-being, and possibly interfere with later cognitive
functioning. Of particular concern is the fact that while the elderly comprise but 15% of
the total population, they account for 25% of all suicides (Blazer, 1993). Additionally, it
is known that depression imposes substantial economic and emotional costs on society.
Indeed, the annual cost of depression in the United States is approximately 43.7 billion
(Greenberg, Stiglin, F inkelstein, & Bemdt, 1993).

The difficulty that researchers have encountered in gathering consistent data with
respect to the prevalence of depression among older adults is complicated by the fact that
those over the age of 60 are the least likely of any age group to seek out mental health
services (Blazer & Williams, 1980). This underutilization of available mental health
services by older adults has been attributed to a variety of potential causes, including
knowledge of the service, the individual’s perceived need, available transportation,
affordability, and attitudes toward treatment (Krout, 1983). Most of the literature,
however, has focused on the apparent negative attitudes older adults hold toward mental

health services and service providers. It has been reported that older adults would rather

seek services for psychological problems from physicians and religious authorities than
from mental health professionals (Priddy, Tipton, & Prohaska, 1982). It has also been
suggested that older adults believe general practitioners would be most effective in
treating their problems and that older adults generally lack confidence in mental health
specialists’ abilities to help (Waxman, Carner, Dubin, & Klein, 1982). Despite reports
that the attitudes of older adults toward psychological interventions are not universally
negative (Lasoski & Thelen, 1987), the general belief that older adults prefer medical
interventions still prevails (Adler, 1992).

In addition to these difficulties, a number of reports show that the elderly are not
always aware that they are suffering from depression, as they believe that the
manifestations of the disorder are simply symptoms of aging (Blandchard, 1996;
Shulman, 1989). Complicating matters is the fact that depressive disorders in the elderly
also tend to show an atypical clinical picture. In fact, several studies have suggested that
older people have a different pattern of depressive symptoms to that found in their
younger counterparts. In particular, it has been reported that older people present with a
focus on somatic symptoms while minimizing spontaneous complaints of low mood
(Shulman, 1989). Sleep disturbances are more readily reported, not only in the form of
early awakening, but also as frequent awakenings during the night (Shulman, 1989).
Loss of appetite and fluctuations in weight may also be primary symptoms. Tebbs and
Martin (1987) have reported that the ten most common somatic symptoms reported by
elderly people with depression include asthenia (weakness), headache, palpitations, pain,
dizziness, abdominal pain, dyspnoea (shortness of breath), localized pain, back pain, and

gastrointestinal dysfunction. It is easy to see how depressive disorders among this group

are overlooked given these common physical complaints. Indeed, depression in the
elderly is often hidden behind somatic symptoms, either because of somatization of the
disorder or because of the accentuation of physical symptoms related to comorbid
illnesses.

Etiology of Late-Life Depression

An understanding of the possible etiological factors underlying depression in old
age is not only necessary to explain the relatively high overall prevalence of depressive
symptomatology in this population, but can also be useful in identifying those at
particularly high risk for developing the disorder. Notably, the majority of studies to date
have focused on demographic and social factors, while only a few investigations have
attempted to generate more integrated etiological hypotheses. Therefore, the following
section will summarize the possible roles that demographic and social factors play in
predisposing the elderly toward depression, in precipitating or maintaining their
symptoms, or in protecting this group from affective disturbance altogether.
Demographic Factors

Gender.

Most epidemiological studies have reported a considerably greater prevalence of
depressive symptomatology in elderly females than in elderly male subjects, thereby
suggesting that female gender predisposes one toward depression in old age. The
etiological influence of gender on depression in old age does not, however, appear to be
purely direct. In fact, several studies have suggested that gender is probably best
described as a moderating variable in the relationship between age and the development

of depressive symptomatology. In one such study, Kennedy et a1. (1989) found a highly

10

significant association between age and depression in women (p<0.0002) but no such
relationship in men. Such a finding may reflect real differences in the frequency with
which women and men experience symptoms of depression or, alternatively, it may be
reflective of the fact that men tend to deny symptoms of depression more frequently than
women. It is important to note that subsequent stepwise regression analysis in this
particular study, however, showed the effects of both age and gender to be very small
once health and disability variables had been properly controlled.

In a complex study attempting to tease out the influence of both age and gender
on social risk factors for depression, Lewinsohn et al. (1991) found that correlations
between social variables and depression differed significantly between men and women
in 12 of the 115 variables examined. Depression in women was associated with feelings
of dissatisfaction with important life roles, with experiencing physical disease, and with
marital conflict, to a much greater extent than was the case in men.

A few studies have also examined the influence of gender on the relationship
between medical problems and depression. The relationship between physical illness and
depression was found by Husaini et al. (1991) to be equally evident in males and females
in a sample of 600 African American elderly community residents. In contrast, Cadoret
and Widmer (1988) examined a group of elderly patients with a recent occurance of life-
threatening or severely disabling illness with an age and sex matched control group.
Depressive symptoms increased significantly in the group of ill male patients but not in
the ill female patients. The increase in depression in the ill male group remained evident
when controlling for other depression-linked variables, such as nursing home placement,

prior history of depression, and recent stressful life events.

11

Family history.

There is consistent evidence that a positive family history is an important
predisposing factor for depression in old age. This is particularly true for individuals
with Bipolar Disorder, who suffer from both depression and mania (Shulman, 1992).
Interestingly, the genetic contribution appears to be the least important in those with
unipolar depression who become depressed for the first time in late life (Alexopoulos,
1989).

Financial and educational status.

Several studies have noted a relationship between poverty and depression in old
age. Dean, Kolody, and Wood (1990) noted a significant direct effect of financial strain
on depressive symptoms, even when controlling for the effect of negative life events and
disability. Carpiniello, Carta, and Rudas (1989) found a similar relationship between
poverty and depression; however, this relationship was only evident in urban residents.
Kennedy et al. (1989) found that older subjects earning less than $5,000 a year had more
than three times the risk of being depressed than those with incomes of more than
$15,000 a year. In a follow-up study from the same sample, Kennedy et al. (1990)
demonstrated that low income was significantly and independently associated with the
emergence of new depression (p<0.02). Low income also emerged as a significant and
independent correlate with depressive symptoms in a large community study by Blazer,
Birchett, Service, and George (1991).

Relative to financial status, education has received somewhat less attention as a
possible correlate of depression in old age. Carpiniello et al. (1989) found an association

between low levels of education and depression, but this finding was restricted to women

12

and urban dwellers. More recently, Evans and Katona (1993) found a significant
correlation between premorbid intelligence and self—rated depression scores, but not with
interview-identified cases of depression.

Social Factors

Illness and disability.

An early study done by Linn et al. (1990) was one of the first to examine the link
between depressive symptomatology and physical disability. They found a small but
statistically significant correlation (r = 0.23) between level of disability and degree of
depressive symptomatology. In a similar but more detailed study, Murphy (1992)
compared elderly depressed subjects, the majority of whom were psychiatric outpatient
referrals, with age and sex matched community controls. Severe chronic health
difficulties were present in 39% of the depressed elderly compared with 26% of the
normal elderly subjects.

In a much larger community sample of elderly residents, Kennedy et al. (1989)
found that 30% of those with four or more medical conditions scored above the cutoff
point for depression, compared with only 5% of those with no medical conditions. When
depression scores were regressed against predictor variables, the contribution of health
and disability to variance in depression remained highly significant (p<0.0002).
Similarly, in a primary care based study, Evans and Katona (1993) found that depression
was almost twice as common in those with significant physical health problems than in
those without. Similarly, Blazer et al. (1991) found that disability and chronic illness,
analyzed as separate variables, were both highly significantly associated with depression

rating scores.

13

Several studies have examined the effects of disability as distinct from poor
health. Smallegan (1989) found a significant relationship between level of disability and
severity of depression, despite the fact that a very low overall rate of depression was
found in their sample. In a larger stratified community sample, Dean et al. (1990) found
a similar relationship between impairment in activities of daily living and depression, as
did Phillips and Henderson (1991) in a nursing home cohort. In a study by Bruce and
McNamara (1992), both major depression and dysthymia were more than twice as
common in bed or chair-bound elderly subjects than in the non-home-bound. The
relationship with dysthymia, but not that with major depression, remained statistically
significant after controlling for degree of physical illness. Finally, Oxman et al. (1992)
examined the influence of disability on depression prospectively through follow-up
interview after a three-year interval on an institutionalized sample. Deterioration in
firnctional ability between baseline and follow-up interviews was significantly associated
with increased depression (p<0.0001). Physical illness thus emerges as a major factor
both in precipitating and in maintaining depression in old age.

The studies described above have used cross-sectional data to examine the
relationship between depression and illness and/or disability. This relationship can be
examined more thoroughly with longitudinal follow-up studies involving subjects
initially identified as normal (not depressed). In one such study, Murrel, Meeks, and
Walker (1991) found that deterioration in health was significantly associated with the
development of new symptoms of depression. They were not, however, able to
demonstrate any protective effect of good physical health on depression brought on by

negative life events.

14

Life events.

The potential for recent adverse events to precipitate depression has been well
documented. An early study by Murphy (1982) replicated the studies of Brown and
Harris (1978) with elderly patients and matched controls, finding that 48% of the
patients, as compared with only 23% of controls, had experienced at least one severe
independent event (such as bereavement, life-threatening illness of someone close) in the
year prior to the onset of depression. It should be noted that the single most common life
event in this study was personal physical illness, though the results remained statistically
significant when this was excluded. In a more recent study using similar methodology,
Emmerson et al. (1999) examined severe life events in a shorter three-month time period
prior to onset of illness and found rates of 24% for the depressed patients and 7% for the
controls. In a much larger stratified community sample, Dean et al. (1990) found highly
significant correlations between undesirable life event scores and depression scores.

A number of studies have also examined the influence of more specific life events
on the onset of depression. Linn et a1. (1990) found that deaths or accidents among
relatives and friends, and arguments with family members or close friends, were most
closely associated with depression. In a larger community study in Canada, Stephenson-
Cino et al. (1992) confirmed the overall relationship between life events and depression
and found specific associations with bereavement and with recent illness of a friend or
relative. Very similar associations between depression and serious illness and death
among relatives and friends were reported by Pahkala, Kivela, and Laippala (1991).

The effect of moderating variables in buffering the extent to which life events

trigger depression is also of considerable interest. Murphy (1982) found the association

15

between life events and depression to be maintained only in a subsample lacking an
intimate relationship. In those who had such a relationship, rates of depression did not
differ significantly between those experiencing a severe” life event and those without such
a stress. A similar prospective effect of intimacy against life-event-induced depression
was reported by Evans and Katona (1993).

Bereavement.

Not surprisingly, bereavement is the life event most clearly implicated in the
onset of depressive symptomatology in elderly subjects. Murphy (1982) found that 15%
of her depressed sample, compared with only 7% of controls, had recently experienced
the death of a spouse or child. Even more strikingly, Pahkala et al. (1991) found that
only 2% of non-depressed community subjects, compared with 31% of depressed
subjects, had recently experienced the death of a close friend or relative. Kennedy et a1.
(1989) found that the relative risk of depression was nearly twice as great in subjects who
had experienced the death of a family member in the past six months than in those spared
such an experience. In a multiple regression analysis they found that the combined
variable of family illness and bereavement remained a highly significant (p<0.0001)
associate of depression. They noted, however, that it explained only 2.9% of the
variance, which was much less than the contribution of other variables, such as sleep
disturbance or health and disability.

Prospective studies have also attempted to identify which bereaved subjects have
higher risks of developing depression. In a two-year study of 189 widows and widowers,
Zisook, Shuchter, and Lyons (1997) found that subjects who became depressed tended to

be younger, to have been depressed in the past, and to have experienced sudden and

16

unanticipated bereavement (as opposed to bereavement following the death of a spouse
who had long been ill). In contrast, an earlier study by Hill, Thompson, and Gallagher
(1988) involving 95 widows found no relationship between depression at one year and
the extent to which bereavement had been expected. The discrepancy between these
findings may be due, in part, to the relatively small number of subjects in the Hill et al.
study (1988). In another study, Dimond, Lund, and Caserta (1997) examined the role of
social support in a two-year prospective follow-up study involving a group of elderly
bereaved subjects. By far the strongest predictor of depression at all time points was
severity of depression three weeks after bereavement. Depression at two months also
showed a small but statistically significant negative relationship with age and with length
of marriage, but no such effect was found between either of these variables and later
depression.

Perhaps the single clearest finding from these studies is that although depressive
symptomatology is very common in the weeks immediately after bereavement, most
subjects experience a gradual decrease in depressive symptoms without developing a full-
blown depressive illness. Persistent, clinically significant depression appears to be
associated more with demographic and psychiatric variables (e. g., age, gender, past
history of depression) than with bereavement alone.

Quality of relationships.

Being married appears to protect against the development of depression in old
age, at least for some elderly individuals. For example, Carpiniello et al. (1989) found
that only 8% of their elderly married subjects were depressed, compared with 17% of

single subjects and 21% of those who were widowed. Similarily, Stephenson—Cino et al.

17

(1992) found that only 5.6% of their married subjects scored above the depression cut-off
point on self-report measures, compared with 11.% of widows and 15.4% of separated or
divorced subjects. Smallegan (1989) also found depression to be associated with lack of
a spouse, as well as with the experience of marital change.

Similar relationships have been demonstrated between low levels of social
interaction and a higher risk of depression. Kennedy et al. (1990) found that living alone
was associated with a more than two-fold increase in the relative risk of depression.
Pahkala et al. (1991) found significant relationships between depression and both living
alone and decreased participation in social activities. A trend just failing to reach
significance was found in the same study between depression and lack of intimate friends
(19% versus 9%; p<0.6). Husaini et al. (1991) reported a significant relationship in
women but not in men between depression and reduced frequency of contact with
relatives and friends. Prospective studies by Russel and Cutrona (1991) and by Oxman et
a1. (1992) both suggest that lack of social support predicts the subsequent development of
depression. The latter study also found that worsening levels of social support between
the baseline and follow-up interviews (three years apart) predicted the emergence of
depression.

Some researchers have suggested that quality of relationships may be more
important in determining whether elderly subjects become depressed than “harder”
variables such as marital status, whether subjects live alone, and the number of social
contacts per unit of time. Interestingly, Murphy (1982) found that the absence of any

close relationship was associated with rates of depression that were at least double those

18

found in subjects with some degree of intimate social contact, a finding replicated almost
exactly in a community study by Kennedy et al. (1989).

The type of one’s relationship and the nature of social interaction are clearly also
relevant variables. Dean et al. (1990) found that availability of social support from
spouses, friends, and adult children (in that order) were important predictors of freedom
from depression, but that support from other relatives was not associated with the
presence or absence of depression. In an interesting small study of nursing home
residents, Rotenberg and Hamel (1988) found that depression was more common among
subjects with reciprocally intimate relationships. These authors hypothesized that
whereas chatting protected against depression, more emotionally laden interactions might
actually cause it.

Gender and life events may be important moderating variables between intimacy
and depression. Emerson et al. (1989) found the significant association between
depression and the lack of a confiding relationship to be evident only in men. Fifty-seven
percent of depressed men and only 3% of male controls had no confidant, whereas in
women the figures were 23% and 18%, respectively. Both Murphy (1982) and Evans and
Katona (1993) found that a significant association between life events and depression in
elderly subjects was only present in those without any confiding relationship.

Summary of Etiological Factors

Both demographic and social factors appear to be of considerable importance in
the etiology of depression in old age. A positive family history, poverty, and loneliness
emerge as the most robust predisposing factors. Adverse life events, particularly

bereavement, and deteriorating physical health, appear to be the most powerful

l9

precipitants. Poor physical health also appears to be the single factor most important in
impeding recovery from depression. On a more positive note, good social support,
particularly in the form of an available confidant, may actually protect elderly individuals
from becoming depressed even in the face of adversity.
Eflects of Depression on Cognitive Functioning

Despite an increased interest in the depressive symptomatology in older adult
populations, the neuropsychology of late—life depression is poorly understood. While
numerous studies have examined cognitive functioning in late-life depression, only a
handful have employed a comprehensive assessment of cognitive domains. Considering
the studies available for review, it is clear that the reported results from these
investigations have been contradictory. While a number of authors have reported severe
and global cognitive deterioration related to depression (Savard, Rey, & Post, 1980;
Siegel & Gershon, 1986; Siegel, Gurevich, & Oxenkrug, 1989; Siegfiied, 1985), others
have been unable to document abnormal cognition in this population (Bieliauskas,
Costello, & Terpenning, 1991; Niederehe & Camp, 1985; Popkin, Gallagher, Thompson
& Moore, 1982; Williams, Little, Scates, & Blockman, 1987). The following sections
will briefly review the contradictory results obtained from cross-sectional studies
examining depression and cognition in the areas of memory, speed of processing, and
executive functioning. While the majority of studies in this line of research have been
cross-sectional in nature, there have also been a few longitudinal studies that have been
completed recently. These investigations will also be reviewed briefly.
Cross-Sectional Studies

Memory.

20

Although the effects of depression on several domains of cognitive function have
been investigated in the elderly, the domain of memory has been the most often studied.
An issue that has been the subject of increasing attention among investigators in recent
years is whether depression in late life results in impairments of memory functioning
compounding those observed in normal aging. Although it has been repeatedly shown
that subjective memory complaints are far more common among depressed than among
nondepressed older adults (Feehan, Knight, & Partridge, 1991; Williams, Little, Scates,
& Blockman, 1987), researchers using neuropsychological measures have found a mixed
picture concerning the effects of depression on memory fimctioning in old age. A host of
investigators have concluded that depression can impair memory, broadly construed, in
this population (Backman & F orsell, 1994; Gainotti & Marra, 1994; King, Caine,
Conwell, & Cox, 1991; LaRue, Goodman, & Spar, 1992) and an equally imposing array
of authorities have found little or no effect (Geffen, Bate, Wright, Rozenbilds, & Geffen,
1993; Loewenstein et al., 1991; O’Hara, Hinrichs, Kohout, Wallace, & Lemke, 1986;
Rohling & Scogin, 1993). Still others have found mixed effects; that is, depressed
subjects are impaired on some tasks but not others (Hart, Kwentus, Taylor, & Harkins,
1987; LaRue, D’Elia, Clark, Spar, & Jarvik, 1986).

These discrepant findings may in part be due to differences across studies with
regard to which aspects of memory have been assessed. For example, there is evidence
that depression-related deficits in memory may be pronounced in tasks that involve little
retrieval support, as in free recall, whereas differences between those with depression and
control subjects are reduced or eliminated in tasks involving a high degree of retrieval

support, such as in recognition tasks (Cummings & Benson, 1983; Hart et al., 1987;

21

Weingartner, 1993). In a related vein, depression-related deficits in memory appear most
likely to occur when there are high demands on effortful, elaborative activities at
encoding the information (Cohen, Weingartner, Smallberg, Pickar, & Murphy, 1982;
Weingartner, Cohen, Murphy, Martello, & Gerdt, 1981). For example, depressed patients
have difficulties in spontaneously using an appropriate organizational structure with
material that is presented in a seemingly disorganized fashion (Watts, Dalgleish, Bourke,
& Healy, 1990) and they do particularly poorly when long memory lists are used (Henry,
Weingartner, & Murphy, 1973). By contrast, elderly depressives perform better on those
memory tasks that require less effortful organization during the encoding process (Cohen
etal., 1982; Norton, Houston, Johnson, & Zechmeister, 1988).

A recent meta-analysis of the effects of depression on memory has helped to
clarify the relationship between these variables somewhat. The meta-analysis, published
in 1995 by Burt, Zembar, and Niederehe, was based on 147 studies of depressed patients
involving recall and recognition tasks. Their overall conclusion was that depression and
memory impairments are significantly associated with one another. However, the effect
sizes were small to moderate, depending on task and subject conditions. Average ds,
reflecting the difference between depressed and control subjects, ranged from 0.27 to 0.6.
Although depression was related to memory impairment, the strength of the relationship
was smaller than that seen in degenerative conditions, where Burt and colleagues found
ds ranging from 1.32 to 2.25. Interestingly, depression was more strongly related to
memory impairment among younger than older subjects. The modest effect sizes found
in the depression versus control groups, and the finding that depression has a greater

impact on memory in younger than in older individuals, leaves unanswered the question

22

as to why so many experts believe the relationship between geriatric depression and
memory is robust. Clearly there is more investigative work that needs to be done in this
area.

Speed of processing.

Patients suffering from depression often report the subjective experience of a
slowing in mental speed (O’Connor etal., 1990). It has been recognized for some time,
and makes intuitive sense, that depressed subjects’ performance on cognitive tasks may
be linked to the efficiency with which the task must be completed (Weingartner, 1986).
For example, Hart and Kwentus (1987) found that depressed elderly patients
demonstrated psychomotor slowing on two tasks requiring rapid motor performance, the
Digit Symbol subtest of the Wechsler Adult Intelligence Scale and the Stemberg Memory
Scanning Task. Indeed, cross-sectional studies have generally reported that both motor
and cognitive speed appear to be impaired in depression (Calgiuri & Ellwanger, 2000;
Sobin & Sackheim, 1997), although Elliott et al. (1996) found that middle-aged
depressed patients were impaired on a measure of cognitive speed but not motor speed.
Nevertheless, while numerous studies have demonstrated processing speed deficits in
those with depressive symptomatology, the relationship between these variables has not
been well examined outside of the traditional cross-sectional design. Therefore, the long-
term effects of depression on cognitive slowing are as yet unknown.

One of the most frequently used tests to measure psychomotor slowing is the
Digit Symbol Modalities Test, a very short measure in which subjects have to substitute
symbols for digits according to a key shown at the top of the page. The Digit Symbol

Modalities Test owes its clinical sensitivity to the fact that it makes demands on several

23

processes at once, including perception, working memory, sustained attention, and
visuomotor coordination. For the successful completion of the task, all of the processes
need to be functioning efficiently (Lezak, 1983). A number of studies that have used this
test with depressed subjects have found impaired performance relative to controls (Austin
et al., 1992; Wolkin et al., 1992), however, there is no published data on this measure
involving depressed subjects in a longitudinal design.

Executive functioning.

Of all the cognitive domains investigated with depressed elderly subjects,
executive fitnctioning abilities have been the least studied. The few studies assessing
executive skills in older depressed patients have revealed impairments on at least some
tests, including word generation tasks, the Wisconsin Card Sorting Test, and the Halstead
Category Test (Emery & Breslau, 1989; Geffen et al., 1993; Hart, Kwentus, Taylor, &
Harkins, 1987; Savard et al., 1980). Although only a few neuropsychological studies
have been conducted in this area, limited evidence from neuroimaging studies has
demonstrated that individuals with depressive symptoms have reduced cerebral blood
flow in different regions of the frontal cortex (Bench et al., 1992, 1993; Drevets et al.,
1992), the area of the brain believed to be most associated with higher-order, executive-
type abilities. Moreover, structural scanning using MRI has shown increases in
subcortical white matter changes in elderly depressives (Coffey, Figiel, Djang, & Weiner,
1990; Krishnan et al., 1988), another area of the brain associated with executive abilities.
This area of the brain is commonly referred to when discussing the “fronto-subcortical
loop”, the connections of which are thought to be necessary (that is, intact) for proper

execution of executive abilities. Patients with damage to subcortical structures often

24

demonstrate impairment on measures sensitive to frontal lobe dysfunction, such as
attentional set-shifiing and planning (Owen et al., 1992; Robbins et al., 1994).
Longitudinal Studies

Very few studies have followed older depressed patients across time to document
the frequency of later cognitive deterioration. Those studies that have been conducted in
this area have largely focused on elderly depressives’ relative risk of developing a
dementing illness, such as Alzheimer’s disease (AD), and have not examined the effects
of depression on the cognitive functioning of normal, non-demented individuals. While
few studies of this kind have incorporated the healthy elderly, studies involving dementia
patients nevertheless help to shed light on the possible long-term negative effects of
depressive symptoms on cognition.

One of the first longitudinal studies investigating the link between depression and
the development of dementia was conducted by Kral in 1983. In that study, 22 older
adults diagnosed with pseudodementia (mean age = 76.5) were followed semiannually
over the course of 4 to 18 years. Twenty of the 22 patients (91%) developed Alzheimer’s
disease by the end of the research period. The diagnosis of AD was supported by
neuropathologic evidence obtained postmortem. Of the 22 patients followed, 11 were
alive at the end of the study, 9 of whom were demented and living either in hospitals or in
nursing homes. While these results are quite dramatic, the interpretation of the study is
made difficult because Kral (1983) used the term pseudodementia as a diagnosis and
because there was inadequate description of the subjects and of how diagnoses were
made. However, this study did illuminate the need for investigation into the relationship

between depression and cognitive impairment in older age.

25

Reding, Haycox, and Blass (1985) followed 28 older patients diagnosed with a
nondementing depression per DSM-III-R criteria (APA, 1987) and neurological and
psychiatric evaluations. The study covered 3 years and resulted in 16 of 28 (57%)
developing “frank dementia.” Thirteen of those 16 patients demonstrated some sign,
often subtle, of neurological disease. Reding and colleagues (1985) speculated that
depression may initially present as the first sign of a developing progressive dementia.

A later study by Nussbaum, Kaszniak, Allender, and Rapcsak (1995) followed 35
older adults (70 years of age and older) diagnosed with either major depression or
dysthymia, both without cognitive impairment. Two board-certified specialists, who
were reported to be expert in differential diagnosis with older adults, made diagnoses
independently from a review of medical and neuropsychological data. Repeat
neuropsychological examinations were administered at the end of the study period
(approximately 2 years), and 8 patients (22%) demonstrated progressive dementia
consistent with AD. Predictors of later cognitive decline included the presence of deep
white matter and subcortical lesions as measured by brain CT/MRI and abnormal
electrocardiogram (EKG) conducted at the initial evaluation. These authors asserted that
for older depressed individuals, an abnormal brain CT/MRI in the presence of a normal
cognitive profile may represent a neurobiological marker for the development of a
progressive dementia.

In contrast to the findings reported above, a longitudinal study conducted by
Sachdev, Smith, Lepan, and Rodriguex (1990), found no substantial decline in a sample
of 19 middle-aged adults diagnosed with pseudodementia. No decline was found in the

study across a 12-year time period. However, this finding may be less relevant given the

26

fact that the mean age of the sample at the start of the investigation was 53. Nonetheless,
Sachdev and colleagues (1990) asserted that pseudodementia was not a risk factor for
later cognitive decline. Since the time of the report by Sachdev et al. (1990), at least
three other investigators have also been unable to replicate the finding that depressive
symptomotology increases the risk for onset of dementia (Buntinx, Kester, Berrgers, &
Knottnerus, 1996; Dufouil et al., 1996; Henderson et al., 1997).

Overall, the longitudinal studies reviewed here suggest that depression in some
older adults may represent an early marker for the later development of a progressive
dementia. However, due to the small number of published studies and the obvious
methodological concerns with this body of empirical work, researchers do not yet have
enough evidence to suggest that a strong relationship exists between depression and later
cognitive decline. In addition, as stated above, because these studies have focused on the
risk of developing dementia, little is known about the long-term effects of depressive
symptoms among the healthy elderly. The need for more research in this area is clear.
Summary of Depression-Related Cognitive Decline

While a number of authors have reported severe and global cognitive
deterioration related to depression in old age, others have been unable to document
abnormal cognition in this population. A handful of studies have investigated the impact
of depressive symptomatology on specific cognitive abilities; however, it remains unclear
whether there is a widespread impairment affecting all cognitive processes or whether
separate functions such as memory, speed of processing, and executive functioning, are

affected differentially.

27

Despite previous research, there are a number of important questions that remain
unanswered in the literature. Specifically, researchers do not fully understand the
strength of the relationship between depressive symptoms and cognitive decline, the
degree to which varying levels of depression severity influence cognitive functioning, the
types of cognitive processes most likely to be affected by depression, those least likely to
be affected, or the possible interaction between age-related cognitive decline and
depression-related cognitive decline. F urtherrnore, the vast majority of empirical work
that has been done with these variables has been cross-sectional in nature and thus the
long-term affects of affective disturbance on cognition have been ignored.

Anxiety in Older Adults
Prevalence

The prevalence of anxiety disorders in elderly adults is a topic that has received
little attention, especially given the impressive amount of research on anxiety disorders
over the past several years (Norton, Cox, Asmundson, & Maser, 1995) and the fact that
symptoms of anxiety are fairly common in late life (Krasucki, Howard, & Mann, 1998;
Palmer, Jeste &Sheikh, 1997). A review of the literature on affective symptoms in old
age reveals that much less attention has been focused on anxiety than on depression (Raj,
Corvea, & Dagon, 1993), despite the fact that both types of affective disturbance
represent significant factors associated with increasing age. The serious need for more
research on the impact of anxiety in the elderly population has recently been highlighted
by the National Institute of Mental Health (Pearson, 1998).

Prior to the Epidemiological Catchment Area (ECA) survey (Regier et al., 1988),

there was little information on population prevalence rates for anxiety disorders among

28

the elderly. The ECA survey documented a one-month prevalence rate of 5.5% for all of
the anxiety disorders in individuals aged 65 and older, which is lower than that noted in
younger adults (7.3%). However, the diagnoses of generalized anxiety disorder (GAD)
and posttraumatic stress disorder (PTSD) were not included in this DSM-III-based
(American Psychiatric Association, 1980) study. A later survey by Blazer et al. (1991),
which included GAD but not PTSD, found a 6-month prevalence of 19.7% and a lifetime
prevalence of 34.1% among an elderly community sample.

Although anxiety disorders have generally been found to be less common in the
elderly than in the younger adult population, such disorders nonetheless affect a
significant proportion of older adults. Flint (1994) reviewed eight random-sample
community surveys and found prevalence rates of diagnosed anxiety disorders ranging
from 0.7 to 18.6% among elderly people. When the prevalence rates for specific anxiety
disorders are examined in this population, phobias (agoraphobia, social phobia, and
simple phobia) represent the most common form of the disorder, with a rate of 4.8%
(Regier et al., 198 8). Lower prevalence rates have been reported for obsessive-
compulsive disorder (0.8%) and panic disorder (0.1%), although generalized anxiety
disorder appears to occur more frequently, with a prevalence rate of around 4.6%.
Indeed, it is possible that GAD is even more prevalent than this figure indicates, given
that previous studies using DSM-III-R criteria (American Psychiatric Association, 1987)
did not diagnose GAD if any other disorder was present. Thus, it is possible that even
higher rates of GAD will emerge from studies using current diagnostic practices.

Some experts have also argued that the significance of anxiety among older adults

has been greatly underestimated by studies that have excluded subsyndromal

29

presentations (Himmelfarb & Murrel, 1984). This occurs when subjects, who present
with symptoms of anxiety but fail to meet criteria for a specific disorder, are excluded
from participation. Fortunately, while limited in number, there have been a few studies
that have examined the prevalence of anxiety by focusing on symptoms, rather than
disorders, and these have yielded considerably higher estimates of anxiety in elderly
adults (Himmelfarb & Murrel, 1984; Magni & DeLeo, 1984). Indeed, when significant
anxiety symptoms that do not reach criteria for a specific disorder are considered,
prevalence rates among the elderly increase to approximately 20 to 25% in community
samples (Copeland et al., 1987; Himmelfarb & Murrell, 1984). A crucial difference
between these studies and research designed to assess anxiety disorders is the flexibility
of the diagnostic criteria. In the latter situation, a person either qualifies for participation
in the study or not. In the Magni and DeLeo study (1984), participants’ scores on a self-
report measure of anxiety were rated according to a continuum of categories. The
advantage of this approach is that people who fail to qualify for a diagnosis of high
anxiety were not excluded from the survey, but instead were classified as having
symptoms of anxiety along a continuum from low to high. Well-defined diagnostic
criteria that sharply distinguish between significant and non-significant anxiety are
crucial for some types of research purposes, but it has been suggested that rigid
diagnostic criteria may exclude some of the wide range of expressions of significant
anxiety in elderly adults (Bliwise, McCall, & Swan, 1987).

There have been other factors complicating the assessment of anxiety in older
adult populations, including difficulty with the identification of symptoms, both on the

part of older adults and their treating physicians, lack of consensus regarding the

30

definition of anxiety as applied to the elderly, as well as the comorbidity of physical and
other psychiatric illnesses that are frequently seen with symptoms of anxiety. As stated
previously, mental illness in the elderly often manifests primarily as somatic complaints
(Lurie, 1987) and there is evidence to suggest that anxiety, like depression, may have
such a presentation in the elderly (Sallis & Lichstein, 1982; Turnball, 1989). Since many
physical ailments that commonly afflict older persons have symptoms that are also
considered anxiety symptoms, the problem lies in determining what the symptoms
represent. Furthermore, while late-life anxiety is common and appears to have
potentially serious consequences, most anxious older adults do not seek help
(Himmelfarb & Murrell, 1984). Generally, older adults are believed to underutilize
mental health services (Ettner & Hermann, 1997). Older people may not seek treatment
for anxiety in particular because they may believe it is an inevitable part of the aging
process or because they may interpret somatic anxiety symptoms as manifestations of
physical illness (Turnbull, 1989). Alternatively, older adults may underutilize services
because they grew up in an era when mental health treatment was considered a stigma
and was reserved only for those who were “crazy”. Those older adults who do seek
treatment often receive care exclusively from their primary care physician. For example,
in a study of older adults who received outpatient services reimbursed by Medicare for a
psychiatric diagnosis, only 29% saw a psychologist or psychiatrist (Ettner & Hermann,
1997).

To complicate the problem further, physicians and mental health experts may
have their own biases when it comes to evaluating symptoms of anxiety in the elderly. A

study by Ford and Sbordone (1980) demonstrated that psychiatrists view elderly patients

31

as less likely to respond to treatment than their younger counterparts. These types of
negative attitudes may influence clinical judgment and decision making and lead to
improper identification and treatment of these disorders in older adults. Even when
clinicians are alert to the symptoms of anxiety in this population, the coexistance of
anxiety and depression makes distinguishing between the two quite difficult (Shamoian,
1991; Sheikh, 1991). For example, in a study conducted by Lindesay, Briggs, and
Murphy (1989), up to 39% of phobic elderly patients experienced depression, compared
with only 11% of nonphobic elderly. Similarly, Alexopoulos (1990) found a 38%
prevalence of anxiety disorders in elderly outpatients with major depression. Other
recent reviews have also highlighted the comorbidity of these disorders in older adults
(Flint, 1994). Examination of smaller-scale studies indicates that over 90% of
community-dwelling elders who received a diagnosis of GAD also met criteria for
depression (Lindesay et al., 1989). In this particular sample, close to 40% of individuals
who were diagnosed with a specific phobia also received a diagnosis of depression.
Thus, the rates of diagnostic comorbidity between anxiety and depression appear to be
substantial among the elderly. Obviously this finding has important implications for the
recognition and treatment of these disorders in late life.
Etiology of Late-Life Anxiety

Although much has been written on the etiology of anxiety throughout the
lifespan, relatively little literature exists on the specific causes of anxiety in the elderly.
Although sparse, recent studies have attempted to identify at least some of the important
risk factors associated with anxiety in older adults. These include demographic factors

such as gender and race, and social factors including illness, disability, loss, and exposure

32

to traumatic events. This section will briefly summarize the ways in which these factors
may predispose elderly subjects toward developing symptoms of anxiety, may precipitate
or maintain such symptoms, or alternatively, serve as a protection against their
manifestation.

Demographic Factors

Gender.

Most epidemiological studies have reported a considerably greater prevalence of
anxiety symptomatology in elderly females than in elderly male subjects, thus suggesting
that female gender predisposes one toward anxiety in old age. The ECA study reported
earlier (Regier et al., 1988) demonstrated that older women stand a much greater chance
of experiencing an anxiety disorder, relative to older men, with a documented 2:1 ratio.
This finding has been extended to self-reported anxiety levels as well. More specifically,
women have been documented to report higher levels of trait anxiety (Himmelfarb &
Murrell, 1984), fearfulness (Liddell, Locker, & Burman, 1991), as well as specific
phobias (F redrikson et al., 1996).

Race.

Race has been another demographic characteristic reported to be of importance
with respect to the development of anxiety symptoms in old age. However, despite
claims that race is an important variable, it has generally been an overlooked
characteristic within the literature (Kastenbaum, 1994). One of the first studies to
incorporate race was the ECA study by Regier et al. (1988). However, sampling
problems occurred within some minority groups that unfortunately precluded accurate

prevalence data (Beidel & Turner, 1991). Although the rate of cognitive impairment was

33

higher among elderly African-Americans relative to elderly Caucasians in the Yale site of
the ECA study (the only site that oversampled the elderly), all other DSM-III disorders,
including anxiety disorders, were less prevalent in African-Americans. Unfortunately,
very few reports have appeared since the ECA study was published that have tried to
examine race within specific anxiety-related diagnostic categories. An exception is a
study conducted by Blazer et al. (1991) in which separate prevalence rates were
computed for African-American and non African-American groups by gender. This
analysis indicated the highest prevalence of GAD was found in African-American
women (3.7%), followed by non African-American women (2.7%), non African-
American men (0.7%), and lastly, African-American men (0.3%). Although this study
provides a preliminary examination into the role of race in the development of anxiety in
old age, clearly more empirical work is needed to expand upon these findings.

Social Factors

Illness and disability.

Although a substantial amount of literature has addressed the overlap between
depression and medical conditions, the same attention has not been given to anxiety
disorders in this domain. Because older adults may attribute anxiety-related symptoms,
such as agitation, muscle tension, and sleep problems, to physical causes, this issue is
particularly important in understanding the differential diagnosis of anxiety (Gurian &
Miner, 1991). Additionally, some physical conditions (e.g. cardiovascular disease,
hypothyroidism) create anxiety-like symptoms, suggesting the need for thorough medical

and psychological evaluation (Cohen, 1991). Although the literature is limited, several

34

studies have examined disability and symptoms of anxiety in both community-dwelling
populations as well as in elderly patients presenting to medical clinics and hospitals.

It is no surprise that patients with chronic physical illnesses have been shown to
be particularly vulnerable to the development of psychiatric disorders, and anxiety
disorders in particular. At least one community-based survey found that persons with a
chronic medical condition had a 41% increased risk of developing a psychiatric disorder,
and more than 11% of the individuals with chronic medical conditions had suffered from
a recent anxiety disorder (Wells, Golding, & Burnam, 1988). Another study conducted
by Stein, Heuser, Juncos, and Uhde (1990) found that 38% of patients with Parkinson’s
disease met criteria for an anxiety disorder. Similarly, panic disorder has been diagnosed
in 8% of outpatients with stable chronic obstructive pulmonary disease (Karaj gi, Rifldn,
Diddi, & Kolli, 1990). Of patients with early dementia, 38% have been shown to have a
comorbid anxiety disorder (Wands et al., 1990). In another study by Magni et al. (1988),
involving a group of elderly medical inpatients, 40% of their sample was reported to have
experienced significant symptoms of anxiety. In this study, the severity of anxiety
symptoms was not associated with whether the individual had chronic versus acute
medical problems, but symptoms of phobia were notably higher in patients with disorders
of the central nervous system.

While this brief review cannot be considered an exhaustive list of the available
research conducted in this area, it is clear that the interrelationship between specific
medical problems and symptoms of anxiety is an important one that deserves greater
attention.

Loss and exposure to traumatic events.

35

Anecdotal reports, as well as empirical research on anxiety in late life, suggest a
major etiological role in the life changes associated with old age, many of which involve
some form of loss. This includes loss of, or decline in, both mental and physical abilities,
together with the approaching reality of death, loss of a spouse, loss of health, loss of
independence, and loss of self-esteem upon retiring (Banazak, 1997; O’Brien-Counihan,
1997; Spar & LaRue, 1997). Changing roles within the family and society, financial
constraints associated with a limited income and potential medical costs, and feelings of
helplessness and loss of control are other potential precipitants of anxiety in older adults
(Zisook, Schneider, & Shuchter, 1990).

Anxiety disorders in the elderly are also frequently associated with traumatic
events. Lindesay (1991) reported that the onset of agoraphobia in old age frequently
occurs after a traumatic event such as acute physical illness, falls, or muggings. The
documented presence of anxiety disorders such as PTSD (Sembi et al., 1998),
agoraphobia (Burvill et al., 1995), and GAD (Castillo et al., 1993) after a stroke suggests
that the trauma of this medical event, and/or specific neurobiological changes, may play
an important role in the onset of anxiety disorders in old age. One special population in
this category that has received little attention is individuals at risk for posttraumatic stress
disorder. In particular, individuals exposed to extreme traumas, such as combat, the
Holocaust, or being taken as a prisoner of war, may function adequately until they begin
to age, at which point PTSD symptomatology may begin (Averill & Beck, 2000). Many
explanations exist for this phenomenon, including increased psychological and physical
vulnerability associated with aging, exposure to an event that resembles the initial

trauma, and the reflective nature of many older adults (Averill & Beck, 2000). Thus,

36

older adults who have been exposed to various types of traumatic events may represent
an underserved anxiety-disordered population. These findings, in combination with
evidence from other studies reported here, are important in shedding light on the ways in
which trauma and loss in old age may be associated with the development anxiety-related
symptomatology.
Effects of Anxiety on Cognitive Functioning

Given the lack of attention that past research has given to symptoms of anxiety in
old age, it is not surprising that the neuropsychology of late-life anxiety is poorly
understood. While much research has focused on age-associated cognitive decline in the
older adult population, emphasis has begun to be placed on other factors, such as
heightened anxiety, that may contribute to decrements in an older person’s performance.
Oberleder (1967) recognized the importance of examining affective status in older adults
and summarized this line of reasoning in the following way: “Perhaps the greatest
difference in working with the older subject is that physical, psychosocial, and anxiety
factors increase with age; tests become more threatening generally, and performance
more likely to be affected by non-test factors The effects of both longstanding and
situational anxiety are particularly important” (p. 189). However, despite this early
recognition that emotional factors may play a key role in the cognitive changes associated
with age, researchers have yet to fully explore this hypothesis. Furthermore, research that
has examined these variables has been mostly concerned with the negative effects of
anxiety and depression on the performance of elderly subjects at the time of the
emotional disturbance (cross-sectional designs), and there exists virtually no empirical

investigations on the long-term effects of negative emotional states, particularly that of

37

trait anxiety. With this in mind, the following section will briefly summarize the limited
body of research pertaining to late-life anxiety and cognition in the areas of memory,
speed of processing, and executive functioning.

Memory

Of all the cognitive domains investigated with respect to the effects of anxiety on
performance, that of learning and memory has been the most extensively evaluated. An
early study conducted by Leon (1989) suggests that anxious subjects may generate
internal task—irrelevant information, which may strain memory capacity or otherwise
interfere with the allocation of attentional resources to the task at hand. Indeed, the most
recent studies of anxiety and memory have tended to focus on subjects’ biases with
respect to attentional focus rather than on general deficits in actual memory performance.
For example, Broadbent, Broadbent, and Jones (1986) found that highly anxious subjects
experienced more difficulty in focusing attention during experimental tasks, regardless of
the content of presented material. In a similar study, Mathews et al. (1990) found that
individuals suffering from generalized anxiety disorder had more difficulty detecting a
letter at an unexpected location when words were appearing elsewhere on the screen,
suggesting a general difficulty with attentional control.

In additional to the difficulty that highly anxious older adults may have with
maintaining attentional focus, there is also literature to suggest that working memory and
initial encoding of information are the most affected by anxiety, while more automatic
tasks are the least affected (Bradley, Mogg, & Williams, 1994; Darke, 1988; Hill &
Vandervoort, 1992; MacLeon & Donnellan, 1993). One specific hypothesis that has

gained considerable attention with respect to these findings has been that anxiety biases

38

people away from encoding information in terms of semantic features and makes them
more susceptible to encode information in terms of more superficial features, such as the
temporal order in which information is presented. For example, when a list is constructed
of words that are capable of being grouped into several semantic categories, but are
presented to the subject in random order, normal subjects tend to recall them in semantic
clusters. It seems that this may be disrupted in individuals with high levels of anxiety.
This has been reported in patients with sub-threshold anxiety disorders by Schwartz
(1985), and with highly anxious subjects by Mueller (1986). In general, it appears that
both state and trait anxiety are particularly associated with significantly poorer
performance in verbal learning tasks during late life (Deptula et al., 1993; Patemiti,
Dufouil, Bisserbe, & Alperovitch, 1999; Whitboume, 1976).
Speed of Processing

Unfortunately, there is very little information concerning the effects of anxiety on
processing speed among older adults and the literature concerning the long-term effects
of anxiety on information processing is non-existent. Nevertheless, those studies that
have been conducted have reported results consistent with the idea that anxiety negatively
impacts processing efficiency. For example, King, Hannay, Masek, and Burns (1978)
found a significant inverse correlation between self-reported symptoms of anxiety and
performance on two timed tasks, the finger tapping and form board tests. However, this
finding pertained only to female subjects. Although this result may be related to actual
gender performance, they note that only the female subjects in their study endorsed
significant anxiety symptoms. In another study, Hockey (1986) found that

experimentally-induced anxiety resulted in increased speed of processing performance

39

paired with significantly poorer accuracy. In an elaboration of these findings, some
researchers have proposed that heightened anxiety restricts an individual’s ability to
efficiently process relevant information. This explanation has been proposed in the
processing efficiency theory put forward by Eysenck and Calvo (1992). This theory
predicts that, especially in stressful conditions or in the case of high demands on
information processing, high trait-anxious individuals are at a disadvantage. An excess
of worries and other anxious thoughts preempt part of the processing resources they
would otherwise have available and therefore negatively impact accuracy of
performance. While this represents an interesting theory in the literature, little evidence
has been collected in support of its claims.
Executive Functioning

Given the scant research associated with anxiety and the elderly, it is not
surprising that very little information exists with respect to the specific impact of anxiety
on executive functions within this population. One of the only studies that has
investigated these variables was conducted by Cohen et al. (1980). Cohen and his
colleagues reported that higher levels of trait anxiety were associated with poorer
performance on a test of abstract reasoning in the elderly. Although other evidence in
this domain is lacking, some writers have suggested that the anxious elderly may be at a
disadvantage when it comes to solving abstract, novel problems because of an increase in
cognitive rigidity that may occur (Kuhlen, 1964). Rigidity has been defined by Kuhlen as
the adherence to a previously established behavior pattern in situations in which response
change would be optimal. Flexibility is characterized as the ability to appropriately

modify established behaviors in response to situational demands. Kuhlen (1964)

40

proposed the idea that increasing anxiety leads to cognitive ridigity in the elderly, which
in turn may negatively impact their ability to make use of novel problem-solving skills
when performing tasks of executive functioning. This theory has yet to be thoroughly
investigated.
Summary of Anxiety-Related Cognitive Decline

At present, the effects of anxiety in the aged have not been thoroughly
investigated. In addition, it seems that researchers have not yet well connected the
symptoms of anxiety to specific types of tasks or cognitive processes, such as memory,
speed of processing, or executive functioning. Rather, anxiety is most often discussed in
the literature as a nonspecific mediator of age-related cognitive decline. Presumably the
effects of anxiety are not ubiquitous with respect to cognitive functioning, however little
empirical work has been conducted to make the necessary distinctions between the
domains of functioning most impacted by anxiety-related symptoms in old age.
F urtherrnore, all reports on anxiety and cognition published to date have been limited to
cross-sectional data. The existing literature has so far not presented studies relating to the
relationship between anxiety and normal age-related cognitive decline (Wetherell,
Reynolds, Gatz, & Pedersen, 2002). Therefore, an important question remains as to
whether general anxiety proneness is related to accelerated rates of decline relative to
those who are less anxious. Exploring the relationships between anxiety and cognitive
performance and decline in the older adult population is important for several reasons, as
indicated by previous authors (Wetherell et a1, 2002). First, because neuropsychological
tests are used in diagnosing dementia, there are serious consequences for our deficient

knowledge with respect to which tests may be vulnerable to the effects of anxiety in older

41

adults. Additionally, learning more about the effects of anxiety on cognition across the
lifespan can lead to a better understanding of cognitive processes associated with anxiety
and aging.
Specific Hypotheses to be Tested

The present study seeks to add to the existing literature on the effects of anxiety
and depression on cognitive functioning in older adults by improving upon the
methodological and theoretical shortcomings of earlier studies. This investigation will do
so in the following important ways: (I) examine these variables in a within-subjects
longitudinal design, thereby allowing for the direct examination of the long-term effects
of anxiety and depression on cognitive functioning (2) using neuropsychological
measures that assess the specific cognitive functions of memory, speed of processing, and
executive functioning as opposed to measures of global, non-specific types of cognition
and by (3) employing statistical analyses using continuous variables of anxiety,
depression, and cognitive firnctioning in order to avoid the distortion that may occur by
falsely dichotomizing these constructs.

Based on the literature reviewed above, the specific hypotheses of the current
study include the expectation of the following results:
1. Confirmation of an age-related decline in cognitive functioning between Time 1 and
Time 2 in the areas of memory, speed of processing and executive functioning.
2. Together, age and depression will be more predictive of cognitive decline between
Time 1 and Time 2 than age alone in the following domains:

a. memory (free recall and semantic clustering)

b. speed of processing

42

c. executive functioning
3. The Geriatric Depression Scale (GDS), which was designed specifically for an older
adult population, will be a stronger predictor of the decline in these domains than the
Beck Depression Inventory (BDI), which was not developed exclusively for use with the
elderly.
4. Together, age and trait anxiety will be more predictive of cognitive decline between
Time 1 and Time 2 than age alone in the following domains:

a. memory (free recall and semantic clustering)

b. speed of processing

c. executive functioning
5. Measures of depression and anxiety will ggt be significant predictors of cognitive
decline between Time 1 and Time 2 above and beyond that of age when examined with

the Mini Mental Status Exam, a measure of gross (non-specific) cognitive functioning.

43

METHOD
Participants

The present investigation included 57 community dwelling older adults from the
Lansing, Michigan area, between the ages of 57 and 87 (Time 1 M = 69.86, SD = 6.36
years; Time 2 M = 75.09, $2 = 6.47 years). Public advertisements in local newspapers
and other mailings were used in 1997 and 1998 to solicit willing participants for the study
procedures of Time 1. Subjects who expressed an interest in the study were then
screened using the Mini Mental Status Exam (MMSE). Individuals who demonstrated
cognitive impairment with a score of 23 or below on the MMSE were excluded from the
sample in order to ensure that all participants were healthy and functioning at a level
consistent with normal aging. Those excluded from participation in the study at Time 1
were also referred to other agencies in the community for more comprehensive
assessments. Subjects who met screening requirements and participated in the study
were given the opportunity to attend mood and memory training workshops at Michigan
State University.

In 2002 and 2003, approximately five years after their initial participation,
subjects were sent letters and received telephone calls to solicit their involvement in the
Time 2 portion of the study. Informed consent was obtained from each participant prior
to the completion of the study procedures described below. Participation in the study was
completely voluntary.

Procedures
The procedures and instruments used during Time 1 and Time 2 of the study were

identical in order to assess the cognitive changes that occurred over the 5-year time

44

period. Participants completed detailed questionnaires yielding demographic and life
history information before completing a set of neuropsychological tests. The total length
of testing for each individual was approximately 2 hours. The order of measures was
systematically varied across participants. All testing sessions were conducted
individually. The administered procedures assessed, among other things, the individual’s
mood, memory, executive functioning, speed of processing, and general cognitive ability.
The protocol used in this investigation was formally approved by the Michigan State
University Committee on Research Involving Human Subjects (UCRIHS) under IRB #
03-912.
Instruments
The measures used in this investigation included the following:
Screening Instrument
Mini-Mental State Exam (MMSE)

The MMSE (Folstein & McHugh, 1975) is a 30-item measure used to screen for
cognitive impairment and document changes in intellectual functioning over time. The
measure is easily administered and scored in approximately 5-10 minutes. The MMSE
requires the examiner to ask questions and record responses given by the subject. The
test consists of items that assess orientation to time and place, attention and
concentration, language, constructional ability, and immediate and delayed memory
recall. Each correct response is awarded one point and the total score is the number
correct out of 30. Scores below 24 are considered abnormal when screening for dementia

or delirium and this cut-off is recommended for most populations (Lezak, 1995).

45

The psychometric properties of the MMSE are well documented in the literature.
Inter-rater reliability has been shown to be above .65 (Foster et al., 198 8) and test-retest
reliability estimates for intervals of less than two months generally fall between .80 and
.95 (O’Connor et al., 1989). Most studies report that the MMSE has adequate specificity
and sensitivity for detecting the presence of dementia, particularly in cases where there
exists moderate to severe forms of cognitive impairment (Spreen & Strauss, 1998). The
MMSE also shows modest to high correlations with measures of intelligence, memory,
attention and concentration, and executive functions (Axelrod et al., 1992). In addition,
Crum et al. (1993) has published extensive norms for the MMSE by age (18-85) and
education based on probability sampling of more than 18,000 community-dwelling
individuals.

Demographic Questionnaire

Information on age, gender, level of education, occupation, and health status were
obtained by a self-report style demographic questionnaire. Level of education was
measured in the conventional manner and based on responses to the following questions:
a.) ‘What is the highest grade of school you completed?’ and b.) ‘If you attended
college/graduate school, please specify how many years and what degree(s) you
obtained’. Occupational level was assessed by asking participants, ‘What kind of work
did you do most of your working life?’. Answers to this question were then coded
according to the Eppinger Demographic Formula (Eppinger et al., 1987) in which
unskilled labor = 1; semi-skilled labor = 2; not in labor force (including homemakers) =
3; skilled labor = 4; managerial/official/clerical/sales/ self-employed = 5;

professional/technical/artistic = 6. Finally, health status was measured using the

46

following questions: a.) ‘How would you rate your overall health at the present time?’ in
which excellent = 1; good = 2; fair = 3; poor = 4 and b.) ‘Do your health problems stand
in the way of doing things you want to do?’ in which not at all = 1; a little = 2; a great
deal = 3.

Measures of Mood
Beck Depression Inventory (BDI)

The BDI (Beck, 1987) is a self-report measure consisting of 21 multiple-choice
statements presented on four pages of paper. Each item is concerned with a particular
aspect or symptom of depression, including cognitive symptoms, vegetative signs, mood
changes, as well as somatic complaints, and asks subjects to rate their experience on a
scale of graded severity. The approximate time required to complete the inventory is 5-
10 minutes. A total score is obtained by adding the highest score circled for each of the
21 items. The total score can range from 0-63, with higher scores indicating more severe
levels of depression. The instrument has classified depression severity in the following
manner: 0-9 = normal range; 10-15 = mild depression; 16-19 = mild/moderate
depression; 20-29 = moderate/severe depression ; 30+ = severe depression (Spreen &
Strauss, 1998).

The BDI is a well researched instrument with strong psychometric properties. In
addition, it has been used with a wide range of populations due to the ease of
administration and scoring. In the original norrning sample, Beck (1970) reported that
test-retest reliability was above .90 and that change scores tended to parallel changes in
depression severity for individual subjects. Reynolds and Gould (1981) have reported

Spearman-Brown reliability to be .93 and internal consistency of test items to be .86.

47

Other authors have reported a coefficient alpha of .88 (Steer et al., 1989). Furthermore,
the BDI is known to have good concurrent validity with instruments such as the MMPI

Depression Scale (.75; Reynolds and Gould, 1981) and the Hamilton Rating Scale (.85;
Brown et al., 1995) as well as clinical ratings of depression (.66; Schaefer et al., 1985).

Geriatric Depression Scale (GDS)

The GDS (Brink et al., 1982) is a self-report measure consisting of 30 yes/no
questions pertaining to potential symptoms of depression. The instrument was designed
specifically for the elderly population and is recommended for use by Spreen and Strauss
(1998) due to the simplicity of administration and the fact that it contains somatic items
which are more suitable for older adults than those presented on the BDI. The GDS
requires subjects to read statements and answer each by circling the yes/no response that
most accurately reflects their experience. Alternatively, the items can be read to subjects
if there is any concern regarding their ability to read or respond using paper-and-pencil.
The estimated time required for completion of the GDS is approximately 5-10 minutes.
A total score is obtained on the instrument by adding the point values assigned to each
response (0 or 1). The following cutoff points are used to determine level of depression:
0-9 normal range; 10-19 = mild depression; 20-30 = moderate/severe depression (Spreen
& Strauss, 1998).

The psychometric properties of the GDS warrant the instrument’s wide-spread use
among older adult populations. Brink et al. (1982) reported that in the original norrning
sample, internal consistency (alpha) and split-half reliability were both .94. Retest
reliability over a one-week time span has been reported to be .85 (Koenig et al., 1988).

In addition, concurrent validity has been established with other self-report measures

48

including the BDI (.73; Hyer & Blount, 1984), Hamilton Rating Scale (.83; Yesavage et
al., 1986), MMPI Depression Scale (.72; Bielauskas & Lamberty, 1992), and the DMS-
based Symptom Checklist for Major Depressive Disorders (.77; Bielauskas & Lamberty,
1992). Criterion validity as measured against the Research Diagnostic Criteria has been
reported as .82 (Yesavage & Brink, 1983).

State-Trait Anxiety Inventory (S TAI)

The STAI (Spielberger et al., 1979) is a self-report measure consisting of 40 items
pertaining to potential symptoms of anxiety. The first 20 items measure state anxiety and
subjects are asked to read each question and indicate how they feel at the current moment
according to a scale that includes the responses “Not at All”, “Somewhat”, “Moderately
So”, and “Very Much So”. The second half of the test measures trait anxiety and subjects
are asked to read each question and indicate how they feel generally (not necessarily at
the current moment) according to a similar scale of responses.

A great deal of data has accumulated to attest to the utility of this measure with
older populations. In two recent studies, psychometric properties of the STAI were
examined in samples of participants carefully selected via semi-structured diagnostic
interviews. In one report, the STAI was administered to 50 community dwelling older
adults with GAD diagnosed according to the DSM-IV and 94 normal control participants
with no psychiatric diagnoses (Stanley et al., 1996). Internal consistency for both STA]
subscales was good, as was test-retest reliability of the Trait subscale. Stability of the
State subscale over time was only moderate, as would be expected. High levels of
convergent validity were also evident in the control group. For both GAD patients and

control group participants, the State and Trait subscales correlated strongly (r = .74). In

49

another recent report, Kabacoff, Segal, Hersen, and Van Hasselt (1997) administered the
STAI to over 200 older adult psychiatric outpatients diagnosed according to the SCID.
Similar to findings from Stanley et al. (1996), internal consistency for the measure was
strong, and the State and Trait subscales correlated strongly with one another (r = .72). In
this study, the STAI-Trait significantly discriminated patients with an anxiety disorder
from those without, although the STAI-State did not, as would be expected.
Memory

California Verbal Learning Test (C VLT)

Memory performance is frequently assessed with the California Verbal Learning
Test (CVLT; Delis, Kramer, Kaplan, & Ober, 1987). The CVLT is a measure of verbal
learning and memory consisting of five oral presentations of a list of 16 shopping items
(Monday’s list) followed by periods of free-recall. The words on the list are presented
verbally at the rate of one word per second. The list contains 4 items that belong to 4
different categories, including fruits, clothing, tools, and spices/herbs. After 5 trials of
Monday’s list, the examiner presents a second, interference, list of 16 shopping items
(Tuesday’s list) and a period of free-recall. Immediately after the presentation and free-
recall of Tuesday’s list, the subject is asked to recall all of the items from Monday’s list
in a period of free-recall as well as in a period of cued-category recall (Short Delay).
This same procedure is undertaken 20 minutes later (Long Delay). Finally, a recognition
trial is completed involving the oral presentation of 44 shopping items in which the
subject attempts to identify those that appeared on Monday’s list. There are no time

limits on any of the memory tasks undertaken with this test.

50

 

The authors of this test report split-half reliability coefficients ranging from .77-
.86. Analyses on this tests’s normative sample also revealed a coefficient alpha of .74
(Delis, Kramer, Freeland, & Kaplan, 1988). The correlations between the CVLT and
other tests commonly used to measure the intended construct are said to be quite good
(Lezak, 1995).

Speed of Processing
Symbol Digit Modalities Test (SDMT)

The SDMT (Smith, 1982) is a measure of speed of processing, complex scanning,
and visual tracking that allows comparison between oral and written performance (Lezak,
1995). The oral and written portions of the test each have a time-limit of 90 seconds.
During the first trial, subjects are presented with a sheet of paper that includes a key
displaying pairs of digits and symbols. Beneath the key are rows of double boxes in
which symbols are presented without their corresponding numbers. The participants are
then asked to use the key to determine which number is associated with each symbol and
to write as many numbers as possible in the empty boxes in the 90 second period. During
the second trial, the subject is asked to call out the missing numbers for each symbol,
which are then recorded by the examiner. The score obtained from this test is the number
of correctly placed numbers in each portion of the test with a maximum score of 110 on
each. The norms for the SDMT are well established and reliability is satisfactory (Spreen
& Strauss, 1998). More specifically, test-retest reliability correlations have been reported

to be .80 for the written portion and .76 for the oral portion (Spreen & Strauss, 1998).

51

Trail Making Test (T MT), Part A

The TMT (Reitan, 195 8) consists of two parts, A and B. Trails A requires the
drawing of lines to connect numbered circles in chronological order and is thought to be a
good measure of processing speed and visual scanning (Spreen & Strauss, 1998).
Subjects are asked to connect the dots as rapidly as possible without making any
mistakes, and scores on the TMT are recorded in total seconds to complete the task. On
both parts of this test, mistakes made by the subject concerning the order of circles
connected are corrected by the examiner. The normative data for this test have been well
established. Interrater reliability estimates of the measure have been reported to be .90
(Spreen & Strauss, 1998).

Executive Functioning
Wisconsin Card Sorting Test (W CS T)

The WCST (Heaton, 1993) assesses the ability of an individual to use abstraction
and to shift cognitive strategies in response to changing environmental cues (Lezak,
1995). It is widely accepted as a measure of executive function, as individuals must
employ strategic planning, organized searching, the ability to use environmental cues to
shift cognitive sets, and must act in a way that is goal-oriented (Spreen & Strauss, 1998).
Subjects are asked to match a set of cards, which vary according to the number and color
of different shapes, with one of 4 “key” cards. The individual is asked to complete the
task in the absence of explicit matching guidelines. Instead, they must rely solely on the
feedback given after each trial indicating whether they have made a correct or incorrect
match. In addition, once the subject correctly matches 10 cards consecutively, the sorting

principle is changed. Subjects are not aware of this change and must modify their

52

responses accordingly. A great deal of empirical work on executive functioning has been
done using the WCST as both the construct validity and interrater reliability of the
instrtunent have been well established (Lezak, 1995). In fact, interrater reliability
estimates have been reported to range from .88 to .96 (Lezak, 1995).

Trail Making Test (T M7), Part B

Part B of the TMT (Reitan, 1958) is used to assess aspects of executive
functioning including mental flexibility, divided attention, and sequencing (Spreen &
Strauss, 1998). The measure has enjoyed a long history of use as it is easy to administer
and takes little time to complete (Lezak, 1995). Subjects are asked to connect a series of
circles in a “connect-the-dot” fashion, alternating between a sequence of numbers and
letters. The instructions ask participants to complete the task as fast as they can without
removing the pencil from the paper. Interrater reliability of the test has been reported to
be .90 (Spreen & Strauss, 1998).

Stroop Test (ST)

This test measures components of executive processes including concentration,
the ability to shift attentional resources, and response inhibition (Golden, 1978). There
are 3 trials administered to the subject and each has a 45 second time-limit. During the
first trial, subjects read words aloud (either “red”, “blue”, or “green”) printed in black
ink. Twenty words are printed in each of 5 columns on the page and appear in random
order. In the second trial, these words are replaced with X’s which are printed in
different color inks (either red, blue, or green) and the subject is asked to name these
randomly listed colors as quickly as possible. The final trial combines both of the

preceding stimuli and requires the subject to read the color of ink in which the words

53

“red”, “blue”, and “green” are printed. This last trial is known as the interference
condition because individuals must suppress a more instinctual response (i.e., reading
words) in favor of a less common way of processing information (i.e., naming colors).
Lezak (1995) reports that the ST has satisfactory reliability and more specific test-retest
reliabilities have been reported by Spreen & Strauss (1998) to be .90, .83, and .91 for the

three portions of the test.

54

RESULTS
Descriptive Statistics

In all, a total of 57 of the 81 subjects who participated in Time 1 of the study were
successfully recruited to participate in Time 2. Of the 24 subjects who were lost to
follow-up, 11 could not be reached by mail or by phone, 6 refused participation (3 for
medical/health reasons, 1 due to scheduling complications, and 2 did not comment on the
reason for refusal), 5 had moved out of the Lansing area, and 2 passed away. The sample
retention rate over the 5-year period was therefore 71%. This group of 57 subjects was
designated as the final sample for the present study.

Comparisons between those subjects who were lost to follow-up and those who
completed Time 2 procedures revealed that the two groups did not differ significantly on
any of the key demographic variables, including age, gender, education, occupational
status, marital status, or self-perceived health rating. Cross-tab comparisons involving
dependent variables revealed that subjects who dropped out of the study scored
significantly better on one measure of processing speed, Trails A (M=30.08, p=.047). No
other significant differences between the two groups were found.

Sample characteristics, including age, gender, premorbid IQ, years of education,
occupation, marital status, and health ratings are displayed in Table 1. Twenty-eight of
the subjects were men (49.1%) and 29 were women (50.9%). All participants were
Caucasian. The average age of the subjects at Time 1 was 69.86 (SQ = 6.36 years) and
ranged from 57 to 87. At the time of follow-up, the average age of participants was 75.09
(S_D = 6.47 years), with a range of 61 to 92. Although there was some variability

between participants in the amount of time that elapsed between measurement at Time 1

55

and Time 2, the majority of subjects completed follow-up within 2 months of their five-
year anniversary (M=4.73 weeks, _S_D=2.01). Most subjects were married at the time of
data collection (55.4% at Time 1 and 59.6% at Time 2). The sample was well educated
with an average of 17.22 years (S_D = 2.47) of formal schooling. With respect to
occupational status, 84.2% of the subjects reported occupational attainment at or above
the managerial level, with a modal rating at the rank of professional (52.6%). Finally, on
a scale of 1 to 4, modal self-reported health status was ‘good’ (as opposed to ‘excellent’,
‘fair’, or ‘poor’) at the time of initial data collection (55.4%) and five years later (71.9%).

Means and standard deviations for subjects’ Time 1 performance on all dependent
measures, including the MMSE, CVLT, TMT, SDMT, WCST, and ST, appear in Table
2. Means and standard deviations for the same variables at Time 2 of the investigation
are presented in Table 3. Correlation matrixes between age and all dependent variables
are presented in Tables 4 and 5 for data collection at Time 1 and Time 2, respectively.

Average depression scores as measured by the Beck Depression Inventory (BDI)
were 9.25 (SD = 4.56, range 0-23) at Time 1 and 8.12 (SD = 4.27, range 0-24) at Time 2.
Depression was also measured with the Geriatric Depression Scale (GDS), a tool
designed specifically for use with an older adult population. Average depression scores
as measured by the GDS were 11.69 (SD = 4.98, range 0-26) at Time 1 and 11.02 (SD =
5.86, range 0-24) at the second wave of data collection. Table 6 summarizes the
frequencies at which subjects scored at various levels of depression, (i.e., normal,
mild/moderate, moderate/severe) as defined by cut-off scores for each of the measures. It
is important to note, however, that although this table implies a categorical approach to

the operationalization of depression, these scores were kept as continuous variables

56

during data analysis to improve upon the methodological limitations of earlier
investigations. Trait anxiety scores, as measured by the State-Trait Anxiety Inventory
(STAI) were 36.35 (SD = 19.23, range 20-72) at Time 1 and 35.98 (SD = 19.64, range
20-75) at Time 2. When compared to the normative sample of older adults with similar
levels of education (Spielberger et al., 1979), these scores represent a range from the 3rd
to >99‘h percentile ranking. As expected, trait anxiety remained stable across time points.
The Pearson Product Moment correlation between scores at Time 1 and Time 2 was .94
(p<.01). Interestingly, there were no group differences in rates of depression or anxiety
among men and women or among those with higher or lower levels of education.
Power Analysis

A power analysis was conducted to ensure adequate statistical power in detecting
small, medium, and large effect sizes with the present sample. In this investigation, the
program G-Power (Faul & Erdfelder, 1992) was used to estimate an appropriate sample
size that would allow proper examination of the proposed hypotheses. For significance
tests of Pearson Product Moment Correlations, Cohen (1992) defines small, medium, and
large effect sizes as .10, .30, and .50 respectively. For a desired power of .80 and with
alpha set at .05, G-Power (Faul & Erdfelder, 1992) indicates that 414 individuals are
needed to detect small effect sizes, 48 to detect medium effect sizes, and 21 to detect
large effect sizes of .50 and above. Therefore, the present study has adequate power to
detect medium and large effects for correlational analyses but is inadequate for effects
smaller than .30.

For significance tests involving analysis of variance, Cohen (1992) defines small,

medium, and large effect sizes as .10, .25, and .40 respectively. For a desired power of

57

.80 and with alpha set at .05, G-Power (Faul & Erdfelder, 1992) indicates that 322
individuals are needed to detect small effect sizes, 52 to detect medium effect sizes, and
21 to detect large effect sizes of .40 and above. Therefore, the present study has adequate
power to detect medium and large effects using analysis of variance techniques, but is
inadequate for effect sizes smaller than .10.
Missing Data Analysis

Analyzing data with missing cases can lead to biased results and incorrect
inferences. Although there were very few cases of missing data in the present study, a
procedure for data imputation was used to lay the groundwork for later statistical
analyses. In order to run a missing data analysis, the dataset was first imported into
SYSTAT v. 10.2. Variables that contained missing cases were then analyzed and data
was imputed using the EM estimation method. The expectation maximization (EM)
method uses a two-step process to impute missing data (Little & Schenker, 1995). First,
in the expectation (E) step, expected values are generated based on the algorithym’s “best
guess” as to what a complete dataset would look like given the parameters of existing
data. After filling in missing data with these estimates, the maximization (M) step is
conducted to recalculate the regression coefficients. These new coefficients are then
substituted back into the E step and a new M step is performed. This process is repeated
over and over again until the changing parameter estimates from iteration to iteration
become negligible. Once finished, the process yields a complete dataset that preserves
correlations between the variables.

After completing the data imputation process in SYSTAT, the output was

examined to determine whether missingness was random. Results from the analysis did

58

not generate evidence that missingness followed a systematic pattern and therefore data
was found to be missing in random.
Hypothesis 1: Age and cognitive decline

The first hypothesis of the present study sought a confirmation of an age-related
decline in cognitive functioning between Time 1 and Time 2 on measures of memory,
executive functioning, and speed of processing. As expected, age was significantly
correlated with measures of cognitive functioning at Time 1 and Time 2. With respect to
memory, increasing age was associated with poorer performance on the CVLT in both
the areas of free recall (Time 1 _r=-.432, p<.01; Time 2 r=-.518, p<.01) and semantic
clustering (Time 1 r=-.300, p<.05; Time 2 r=-.399, p<.05). Increasing age was also
shown to be associated with poorer performance on measures of executive functioning,
including the WCST (Time 1 [=.303, p<.05; Time 2 _r=.458, p<.01), the Stroop Test
(Time 1 r=-.475, p<.01; Time 2 _r=-.400, p<.01), and Trails B (Time 1 r=.332, p_<.05;
Time 2 F594, p<.01). Lastly, age was associated with slower processing speed
efficiency on the SDMT (Time 1 r=-.445, p<.01; Time 2 _r=-.489, p<.01) and Trails A
(Time 1 [=.292, p<.05; Time 2 _r=.576, p<.01).

The relationship between age and cognitive decline on these measures over the 5-
year time period was examined using repeated measures analysis of covariance
(ANCOVA). In each analysis, the dependent variables were Time 1 and Time 2 scores
for the cognitive domain in question (e.g., memory, processing efficiency, executive
functioning) and age at Time 1 was entered as a covariate. Results revealed a significant
main effect for time in each area, including free recall (Wilks’ =.817, F(l , 55)=6.659,

p=.018, partial 2=.183), semantic clustering (Wilks’ =.913, F(l, 55)=2.639, p=.041,

59

partial 2=.087), processing speed (Wilks’ =.878, F (1, 55)=6.947, p=.010, partial
2 =. 122), and executive functioning (Wilks’ =.744, F (1, 55)=7.044, p=.012, partial
2 =.256). Similarly, a covariate effect for age was found in each of the models,
including free recall (Wilks’ =.910, F (1, 55)=5.241, p=.025, partial 2=.090), semantic
clustering (Wilks’ =.906, F(l, 55)=2.880, p=.035, partial 2=.094), processing speed
(Wilks’ =.890, F (1, 55)=5.016, p=.019, partial 2=.110), and executive functioning
(Wilks’ =.744, F(1, 55)=7.044, p=.012, partial 2=.256). Overall, these results provide
support for an age-related decline in neuropsychological function across a wide-range of
cognitive processes.
Factor Scores

An advantage of the present study is that multiple measurements were used with
each subject to examine cognitive performance in the areas of processing speed and
executive functioning. As can be seen in Tables 4 and 5, correlations between the two
measures of processing speed, the Symbol Digit Modalities Test (SDMT) and the
Trailmaking Test Part A (TMT), were significant both at Time 1 (r=-.524, p<.01) and
Time 2 (_r=-.549, p<.01). Likewise, the correlations between measures of executive
functioning were significant at each time point, including those between the Wisconsin
Card Sorting Test (WCST) and the Stroop Test (ST) (Time 1 r=-.336, p<.05; Time 2 {=-
.442, p<.05), the WCST and Trails B (Time 1 r=.644, p<.05; Time 2 5.691, p<.05), and
between Trails B and the ST (Time 1 _F-.390, p<.01; Time 2 _r=-.485, p<.01).

In order to combine the information contained in the measurements of these two
constructs (and to avoid overrnodelling the data), factor scores were computed using a

principle components analysis. As can be seen in Tables 7 and 8, these new composite

60

scores correlated significantly with age and were therefore used in subsequent analyses as
overall indicators of processing speed and executive functioning.
Hypothesis 2: Age and BDI depression predicting cognitive decline

Repeated measures analyses of covariance (ANCOVA) were conducted to
determine the effects of age and BDI depression scores on various domains of cognitive
functioning over a 5-year time period. More specifically, it was hypothesized that age
and scores on the Beck Depression Inventory (BDI) would be associated with a decline in
performance on measures of free recall memory, semantic clustering, processing speed,
and executive functioning. Prior to conducting the analyses to test this hypothesis, the
relationships between covariates and dependent variables were examined to determine
whether or not there were any non-linear components (e.g., quadratic functions).
Linearity was analyzed by creating bivariate scatterpots of the variables in question.
Scatterplots revealed that none of the relationships deviated significantly fi'om linearity,
which simplified the analyses.

The dependent variables used in each repeated measures ANCOVA were Time 1
and Time 2 scores for the type of cognitive functioning in question (i.e., free recall
memory, semantic clustering, processing speed, executive functioning). Covariates in
each model were age at Time 1, BDI score at Time 1, and BDI score at Time 2. Results
from the analyses appear in Table 9.

Free Recall Memory

With respect to free recall memory, results indicate a main effect for time (Wilks’

= .803, F(l, 53)=6.580, p=.022, partial 2=. 197) as well as a covariate effect for age

(Wilks’ =.908, F(1, 53)=5.279, p=.026, partial 2=.092). However, the effects of BDI

61

depression scores at Time 1 (Wilks’ =.1.000, F (1, 53)=005, p=.944, partial 2=.000)
and Time 2 (Wilks’ =.988, F(1, 53)=623, p=.434, partial 2=.012) were shown to be
non-significant.
Semantic Clustering

When examining semantic clustering, results indicate a main effect for time
(Wilks’ =.899, F (1, 53)=2.789, p=.037, partial 2=.101) and a covariate effect for age
(Wilks’ =.922, F(1, 53)=2.582, p=.046, partial 2=.07 8). However, BDI depression
scores failed to reach significance in predicting change in semantic clustering
performance between time points. As covariates in the model, BDI depression at Time 1
(Wilks’ =. 1.000, F(l, 53)=.025, p=.876, partial 2=.000) and Time 2 (Wilks’ =.967,
F(l , 53)=1.766, p=.190, partial 2=.03 3) were non-significant.

Processing Speed

When processing speed factor scores were entered as dependent variables in the
repeated measures ANCOVA, results indicate a main effect for time (Wilks’ =.876,
F(1, 53)=6.998, p=.009, partial 2=. 124) and a covariate effect for age (Wilks’ =.883,
F(1, 53)=6.651, p=.012, partial 2=.117). Similar to previous analyses, however, the
effects of BDI depression scores at Time 1 (Wilks’ =.991, F(1, 53)=.467, p=.498,
partial 2=.009) and Time 2 (Wilks’ =.999, F(1, 53)=.044, p=.834, partial 2=.001)
were shown to be non-significant.

Executive Functioning

Finally, with respect to executive functioning, results indicate a main effect for

time (Wilks’ =.752, F(1, 53)=6.021, p=.019, partial 2=.248) as well as a covariate

effect for age (Wilks’ =.805, F(1, 53)=4.192, p=.034, partial 2=.195). However, BDI

62

depression scores failed to reach significance in predicting change in semantic clustering
performance between time points. As covariates in the model, BDI depression at Time 1
(Wilks’ =.989, F(1, 53)=.557, p=.459, partial 2=.011) and Time 2 (Wilks’ =.983,
F (1, 53)=.909, p=.345, partial 2=.017) were non-significant.

Overall, results from repeated measures analysis of variance procedures indicate
that this hypothesis was not supported as BDI depression was not shown to be a
significant predictor of cognitive decline in this sample of older adults.

Hypothesis 3: Age and GDS depression predicting cognitive decline

Repeated measures analyses of covariance (ANCOVA) were conducted to
determine the effects of age and GDS depression scores on various domains of cognitive
functioning over a 5-year time period. More specifically, it was hypothesized that age
and scores on the Geriatric Depression Scale (GDS) would be associated with a decline in
performance on measures of free recall memory, semantic clustering, processing speed,
and executive functioning. Since the GDS was designed specifically for use with an
older adult population, it was anticipated that this measure would be a stronger predictor
of cognitive deterioration than the BDI, which was not developed exclusively for use
with the elderly. Prior to conducting the analyses to test this hypothesis, the relationships
between covariates and dependent variables were examined to determine whether or not
there were any non-linear components (e.g., quadratic functions). As before, linearity
was analyzed by creating bivariate scatterpots of the variables in question. Scatterplots
revealed that the relationships did not deviate significantly from linearity, with the

exception of the GDS score as plotted against free recall memory. Therefore, the natural

63

log transformation of the GDS score was used in the analysis in which free recall
memory was the dependent variable.

The dependent variables used in each repeated measures ANCOVA were Time 1
and Time 2 scores for the type of cognitive functioning in question (i.e., free recall
memory, semantic clustering, processing speed, executive functioning). Covariates in
each model were age at Time 1, GDS score at Time 1, and GDS score at Time 2. Results
from the analyses appear in Table 10. Since the effects of time and age on each of the
dependent variables have been described earlier, only the results pertaining to the
covariates GDS Time 1 and GDS Time 2 will be reported here.

Free Recall Memory

Results indicate that GDS depression scores failed to reach significance in
predicting change in free recall memory performance between time points. As covariates
in the model, GDS depression at Time 1 (Wilks’ =.962, F (1, 53)=2.058, p=.157, partial

2 =.038) and Time 2 (Wilks’ =.951, F(1, 53)=2.666, p=.109, partial 2=.049) were
non-significant.
Semantic Clustering

When examining semantic clustering, the effects of GDS depression scores at
Time 1 (Wilks’ =.978, F(1, 53)=1.181, p=.282, partial 2=.022) and Time 2 (Wilks’

= .957, F(1, 53)=2.355, p=.131, partial 2=.043) were shown to be non-significant.
Processing Speed
When processing speed factor scores were entered as dependent variables in the

repeated measures ANCOVA, results indicate that GDS depression scores at Time 1

64

(Wilks’ =.969, F(1, 53)=1.660, p=.203, partial 2=.031) and Time 2 (Wilks’ =.999,
F (1, 53)=.029, p=.866, partial 2=.001) were non-significant as covariates in the model.
Executive Functioning

Finally, results indicate that GDS depression scores failed to reach significance in
predicting change in executive functioning performance between time points. As
covariates in the model, GDS depression at Time 1 (Wilks’ =.948, F(1, 53)=2.851,
p=.097, partial 2=.052) and Time 2 (Wilks’ =.956, F(1, 53)=2.389, p=.128, partial

2 =.044) were non-significant.

Overall, results from repeated measures analysis of variance procedures indicate
that this hypothesis was not supported as GDS depression scores were not shown to be
significant predictors of cognitive decline in this sample of older adults.

Hypothesis 4: Age and trait anxiety predicting cognitive decline

Repeated measures analyses of covariance (ANCOVA) were conducted to
determine the effects of age and trait anxiety on various domains of cognitive functioning
over a 5-year time period. More specifically, it was hypothesized that age and scores on
the State-Trait Anxiety Inventory (STAI) would be associated with a decline in
performance on measures of free recall memory, semantic clustering, processing speed,
and executive functioning. Prior to conducting the analyses to test this hypothesis, the
relationships between covariates and dependent variables were examined to determine
whether or not there were any non-linear components (e.g., quadratic functions).
Linearity was analyzed by creating bivariate scatterpots of the variables in question.
Scatterplots revealed that none of the relationships deviated significantly from linearity,

which simplified the analyses.

65

The dependent variables used in each repeated measures AN COVA were Time 1
and Time 2 scores for the type of cognitive functioning in question (i.e., free recall
memory, semantic clustering, processing speed, executive functioning). Covariates in
each model were age at Time 1, STAI score at Time 1, and STAI score at Time 2.
Results fi'om the analyses appear in Table 11. Since the effects of time and age on each
of the dependent variables have been described earlier, only the results pertaining to the
covariates STAI Time 1 and STAI Time 2 will be reported here.

Free Recall Memory

Results indicate that trait anxiety scores were significantly related to the change in
free recall memory performance between time points. As covariates in the model, STAI
scores reached significance at both Time 1 (Wilks’ =.901, F (1, 53)=6.271, p=.025,
partial 2=.099) and Time 2 (Wilks’ =.938, F(1, 53)=4.953, p=.039, partial 2=.062).

Semantic Clustering

When semantic clustering scores were entered as dependent variables in the
repeated measures ANCOVA, results indicate that the trait anxiety score at Time 1
(Wilks’ =.918, F (1, 53)=2.019, p=.051, partial 2=.082) was significant as a covariate
in the model. Trait anxiety at Time 2 failed to reach significance, although it came very
close to the .05 level (Wilks’ =.939, F (1, 53)=1.997, p=.057, partial 2=.061).

Processing Speed

With respect to processing speed, results indicate that trait anxiety scores failed to

reach significance in predicting change in performance between time points. As

covariates in the model, trait anxiety at Time 1 (Wilks’ =.986, F( 1, 53)=.727, p=.398,

66

partial 2=.014) and Time 2 (Wilks’ =.991, F(1, 53)=.463, p=.499, partial 2=.009)
were non-significant.
Executive Functioning

Finally, results indicate that trait anxiety scores were significantly related to the
change in executive functioning performance between time points. As covariates, STAI
scores reached significance at both Time 1 (Wilks’ =.916, F(1, 53)=3.682, p=.041,
partial 2=.084) and Time 2 (Wilks’ =.945, F(1, 53)=3.125, p=.052, partial 2=.055).

Overall, results from repeated measures ANCOVA analyses indicate that this
hypothesis was partially supported. Trait anxiety was shown to be a significant predictor
of cognitive decline in the areas of free recall memory, semantic clustering, and executive
functioning. However, trait anxiety was not found to be predictive of processing speed
efficiency in this sample of older adults.

Hypothesis 5: Anxiety, depression, and global cognitive functioning

Repeated measures analyses of variance (ANCOVA) were conducted to test the
hypothesis that depression and anxiety would Qo_t be significant predictors of cognitive
decline between Time 1 and Time 2 above and beyond that of age when examined with
the Mini Mental Status Exam, a measure of gross (non-specific) cognitive functioning.
Prior to conducting the analyses to test this hypothesis, the relationships between
covariates and dependent variables were examined to determine whether or not there
were any non-linear components (e.g., quadratic functions). As in earlier hypotheses,
linearity was analyzed by creating bivariate scatterpots of the variables in question.

Scatterplots revealed that none of the relationships deviated significantly from linearity,

67

which simplified the analyses. The results of repeated measures ANCOVA analyses
appear in Table 12.
MMSE and BDI depression

The dependent variables used in this analysis were Time 1 and Time 2 scores on
the Mini-Mental Status Exam (MMSE). Covariates in the model were age at Time 1,
BDI depression at Time 1, and BDI depression at Time 2. Results indicate a main effect
for time (Wilks’ =.83 8, F(1, 53)=3.105, p=.043, partial 2=.162) as well as a covariate
effect for age (Wilks’ =.851, F(1, 53)=2.649, p=.051, partial 2=. 149). However, as
predicted, the effects of BDI depression scores at Time 1 (Wilks’ =.990, F(1, 53)=.503,
p=.481, partial 2=.010) and Time 2 (Wilks’ =.991, F(1, 53)=.473, p=.495, partial

2 =.009) were shown to be non-significant.
MMSE and GDS depression

MMSE scores at Time 1 and Time 2 were used as the dependent variables in this
analysis. Covariates in the model were age and GDS depression scores at Time 1 and
Time 2. Results indicate that GDS depression failed to reach significance in predicting
change in MMSE scores between time points. As covariates in the model, GDS
depression at Time 1 (Wilks’ =.996, F(1, 53)=.214, p=.646, partial 2=.004) and Time
2 (Wilks’ =.987, F( 1, 53)=.698, p=.407, partial 2=.Ol3) were non-significant.

MMSE and trait anxiety

The dependent variables used in this analysis were Time 1 and Time 2 scores on

the Mini-Mental Status Exam (MMSE). Covariates in the model were age, STAI scores

at Time 1, and STAI scores at Time 2. As predicted, the effects of trait anxiety at Time 1

68

(Wilks’ =.975, F(1, 53)=1.312, p=.257, partial 2=.025) and Time 2 ((Wilks’ =.981,
F (1 , 53)=.998, p=.322, partial 2=.019) were shown to be non-significant.

Overall, results from repeated measures ANCOVA analyses indicate that this
hypothesis was supported, as measures of depression and anxiety were not shown to be
significant predictors of decline when using a global, non-specific measure of cognitive

functioning.

69

DISCUSSION

Hypothesis 1 of the present study predicted that age would be significantly
associated with a decline in measures of processing speed, executive functioning, and
memory. This hypothesis was supported. As predicted, age was correlated with
measures of processing speed, the Symbol Digit Modalities Test (SDMT) and the
Trailmaking Test (TMT) Part A, such that increasing age was associated with slower
performance. This finding is consistent with previous research (Cerella, 1990; Hartley,
1986, 1993; Hultsch, Hertzog, & Dixon, 1990; Kail & Salthouse, 1994; Salthouse 1985a,
1992; Schaie, 1989, 1994; Van Gorp, Satz, & Mitrushina, 1990) and demonstrates that
normal aging is indeed associated with a general slowing of cognitive processing. This
finding lends support to proponents of the general slowing hypothesis (N ettelbeck and
Rabbitt, 1992) who argue that the decrease in performance demonstrated by older adults
on a variety of cognitive tasks is related, in part, to a general decrease in processing rate
with age. It is also important to point out that the size of the correlation between age and
decreased processing speed in the present study is consistent with other research.
Salthouse (1985) has demonstrated that the median correlation between age and measures
of speed across a wide range of behavioral activities is .45. The correlation between age
and processing speed efficiency in the present study was .445 (p<.01) at Time 1 and .489
(p<.01) at Time 2. Furthermore, when entered as a covariate in the repeated measures
ANCOVA analysis, age accounted for 11.0% (p<.05) of the variance in processing speed
decline between Time 1 and Time 2.

As predicted, age was also associated with measures of executive functioning,

including the Wisconsin Card Sorting Test (WCST), the Trail Making Test (TMT) Part

70

B, and the Stroop Test (ST). This is consistent with a vast amount of research indicating
that executive abilities deteriorate with age (Ardila & Rosselli, 1989; Barr & Giambra,
1990; Burke, 1972; Daigneault, Braun, & Whitaker, 1992; Haaland et al, 1987; Salthouse
& Mitchell, 1990; Shimamura & Jurica, 1994; Pierce et al, 1989; Whelihan & Lesher,
1985). When entered as a covariate in the repeated measures ANCOVA analysis, age
accounted for 25.6% (p<.01) of the variance in executive functioning decline between
time points.

Finally, this study was successful in demonstrating a significant association
between age and decline in overall memory functioning. As predicted, age was
associated with a decrement in memory performance on the California Verbal Learning
Test (CVLT) such that increasing age was associated with a decline in both free recall
ability and the use of appropriate encoding strategies to learning new information (i.e.,
reduced semantic clustering). This finding is consistent with an extensive amount of
empirical work suggesting that memory functions decline with age (Backman & Forsell,
1994; Gainotti & Marra, 1994; Geffen et al., 1993; Loewenstein et al., 1991; O’Hara et
al., 1986; Rohling & Scogin, 1993). In the current study, repeated measures AN COVA
analyses revealed that age predicted 9.0% (p<.05) and 9.4% (p<.05) of the variance in
free recall memory and semantic clustering, respectively.

Hypotheses 2 and 3 predicted that, like age, depression would be predictive of
cognitive decline in the areas of processing speed, executive functioning, and memory.
Although there has been very little research on the long-term impact of depression on
cognitive abilities, single episodes of depression have been shown to have negative

effects on many cognitive functions (Backman & Forsell, 1994; Calgiuri & Ellwanger,

71

2000; Emery & Breslau, 1989; Gainotti & Marra, 1994; Geffen et al., 1993; Hart et al.,
1987; LaRue et al., 1992; King, et al., 1991; Sobin & Sackheim, 1997). It stands to
reason that recurrent or persisting depressive symptomatology may have even greater
deleterious effects on processing efficiency and overall cognitive functioning, especially
in an older adult population which has already suffered cognitive decline associated with
increased age. However, results from the current study failed to support such a
hypothesis. More specifically, scores on the BDI and GDS did not result in statistically
significant covariate effects in repeated measures ANCOVA analyses in which
processing speed, executive functioning, or memory were entered as dependent variables.
Therefore, higher levels of depressive symptomatology were not predictive of cognitive
decline above and beyond that already accounted for by age alone. Although it was
hypothesized that depression would be identified as a risk factor for cognitive decline in
old age, these results were not all-together surprising given the conflicting findings with
respect to these variables in the literature (Bieliauskas et al., 1991; Geffen et al., 1993;
Hart et al., 1987; Niederehe & Camp, 1985; Popkin et al., 1982; Williams et al., 1987).
Taking into account results from the current study, it would appear that symptoms of
depression do not, in fact, contribute to age-related cognitive decline in older adults.
Another possible explanation for the lack of association between depression
scores and cognitive decline in the present study may be related to restricted variance
among the independent variables. It is important to keep in mind that the group of
subjects used in this investigation was not a clinical sample. Although a range of
depressive symptoms was detected by both the BDI and GDS for subjects across time

points, the average level of depression as reported by participants was actually quite low

72

overall. More specifically, average depression scores at Time 1 and Time 2 were
consistent with only mild levels of depressive symptomatology according to the
established criteria for each self-report questionnaire. Perhaps if the sample had included
a greater number of individuals suffering from more severe depression, the increased
variability would have made it easier to detect differences in cognitive decline among the
participants. In a similar vein, it is worth noting that the sample as a whole was generally
high functioning and may therefore have encompassed selection biases not well
controlled for in this study. For example, the sample was a fairly homogeneous group of
older adults with higher than average education (M=17.22, SD=2.47), IQ scores
(M=111.95, SD=5.8), and occupational attainment (i.e., 84.2% of subjects were at or
above the managerial level). This lack of variation in the sample may have contributed to
an attenuation of the correlations between the variables, making it difficult to detect
differences in the ANCOVA analyses. Therefore, future research that investigates the
relationship between depression and cognitive functioning in older adults should make
attempts to include a more diverse sample.

A third explanation for the lack of association between depression and cognitive
functioning may have to do with inadequate power. A power analysis was conducted
prior to examining the data. This analyses revealed that the present sample size of 57
subjects was sufficient for detecting medium and large effect sizes, .25 and .40
respectively (Cohen, 1992), but was inadequate for effects smaller than .10. In order to
detect smaller effects, a sample size of 322 subjects would be needed (G-Power; Faul &

Erdfelder, 1992). Therefore, it may be that depression indeed has a significant, but

73

relatively small, impact on cognitive functioning in older adults but the present sample
size was inadequately powerful to detect such an effect.

Hypothesis 4 of the present study stated that trait anxiety would be predictive of
cognitive decline above and beyond that of age alone in the areas of memory, speed of
processing, and executive functioning. This hypothesis was partially supported as trait
anxiety was indeed found to impact the rate of decline in memory and executive
functioning, but was not shown to have an effect on processing efficiency. More
specifically, scores on the State-Trait Anxiety Inventory (STAI) resulted in statistically
significant covariate effects in ANCOVA analyses in which memory and executive
functioning were entered as dependent variables, while scores on the STAI failed to reach
significance in the model in which speed of processing scores were used as dependent
measures. Trait anxiety scores at Time 1 were found to predict 9.9% (p<.05) of the
variance in free recall memory, 8.2% (p<.05) of the variance in semantic clustering, and
8.4% (p<.05) of the variance in executive functioning. At Time 2, anxiety scores
accounted for 6.2% (p<.05) of the variance in free recall memory and 5.5% of the
variance in executive functioning performance. Although Time 2 scores on the STAI
failed to reach significance in predicting decline in semantic clustering, the covariate
effect came very close to the .05 level (Wilks’ =.939, F (1, 53)=1.997, p=.057, partial

2 =.O61). Overall, these findings are consistent with a small, but growing, literature
demonstrating that trait anxiety is associated with poorer verbal learning (Deptula, Singh,
& Pomara, 1993; Patemiti, Dufouil, Bisserbe, & Alperovitch, 1999; Whitboume, 1986)
and executive dysfunction (Cohen et al, 1980). The fact that the current study failed to

show an association between anxiety and speed of processing, while unexpected, is

74

consistent with at least one study currently published in the literature (Schultz et al.,
1980). Taken as a whole, these data suggest a significant relationship between trait
anxiety and cognitive decline and provide support that, like age, anxiety-proneness is a
risk factor for intellectual deterioration. In other words, higher levels of trait anxiety are
related to accelerated rates of cognitive decline relative to those who are less anxious.
To account for these findings, researchers have proposed that heightened anxiety
restricts an individual’s ability to efficiently process relevant information. This
explanation has been articulated most clearly by Eysenck and Calvo (1992) in
descriptions of their processing efficiency theory. This theory predicts that, especially in
stressful conditions or in the case of high demands on information processing, high trait-
anxious individuals are at a disadvantage. An excess of worries and other anxious
thoughts preempt part of the processing resources they would otherwise have available
and therefore negatively impact accuracy of performance. Given that storage and
processing resources are presumed to be limited, the “space” taken up by anxiety results
in reduced capacity for processing new information. Furthermore, because older age is
already associated with reduced working memory and processing resources, this
population may be more vulnerable than younger adults to the effects of anxiety across
multiple cognitive domains. It may be that the elderly have less reserve capacity to
compensate for the adverse effects of anxiety than the young. In particular, older adults
with higher levels of anxiety would be expected to demonstrate poorer performance on
tasks involving memory and executive fimctioning skills, as was shown in the present
investigation. Since neuropsychological tests designed to measure psychomotor speed do

not generally make high demands on information processing (i.e., the tests rely on more

75

familiar, over-learned skills), this theory may help explain why anxiety had no effect on
declines in processing speed in the current sample. For example, because a task like
connecting dots on the Trail Making Test requires very little working memory capacity to
complete, the task can still be done adequately even in cases where anxiety has occupied
much of the “space” required for information processing. Taken together, findings from
the current study are consistent with research suggesting that working memory and initial
encoding of information are the most affected by anxiety, while more automatic tasks are
among the least affected (Bradley, Mogg, & Williams, 1994; Darke, 1988; Hill &
Vandervoort, 1992; MacLeon & Donnellan, 1993).

While the processing efficiency theory (Eysenck and Calvo, 1992) offers an
explanation as to why anxiety may interfere with working memory processes in older
adulthood, there is at least one other theory that has been proposed to specifically address
the negative impact of anxiety on executive functioning skills. Some writers in the field
have suggested that the anxious elderly may be at a particular disadvantage when it
comes to solving abstract, novel problems because of an increase in cognitive rigidity that
may occur (Kuhlen, 1964). Rigidity has been defined by Kuhlen as the adherence to a
previously established behavior pattern in situations in which response change would be
optimal. Flexibility is characterized as the ability to appropriately modify established
behaviors in response to situational demands. Kuhlen (1964) proposed the idea that
increasing anxiety leads to cognitive ridigity in the elderly, which in turn may negatively
impact their ability to make use of novel problem-solving skills when performing tasks of
executive functioning. While this theory has not been well examined in the literature to

date, it may well help to account for the finding that trait anxious individuals suffer a

76

greater loss in executive functioning abilities over time relative to those who are less
anxious.

In addition to the theories set forth by Eysenck and Calvo (1992) and Kuhlen
(1964), it is important to consider that the cognitive impairment resulting from trait
anxiety in some individuals may also be secondary to CNS dysfunction. In other words,
the neuropsychological impairments of highly trait anxious subjects may reflect subtle,
undetected pathological processes that profoundly affect brain integrity. For example, an
emerging body of research in the neuroscience literature describes how the
neuroendocrine system responds to anxiety by stimulating the hypothalamic-pituitary-
adrenal (HPA) axis to release corticosteroids, namely cortisol, which has been known for
some time to affect cognitive function (de Quervain, Roozendaal, & McGaugh, 1998;
Kirschbaum, Wolf, May, Whippich, & Hellhammer, 1996; Newcomer, Craft, Hershey,
Askins, & Bardgett, 1994; Newcomer, Selke, Melson, Hershey, & Craft, 1999). In fact,
Mitchell (1995) in a careful review of 17 studies in which cognitive impairment was
examined in relation to measures of HPA axis activity, concluded that a significant
correlation between these two variables was found consistently. It should be noted that
the specific deficits observed with increased levels of cortisol are quite similar to those
noted in the context of normal aging (Newcomer et al, 1999). That is, subjects appear to
direct more attention to irrelevant stimuli, which results in decreased processing of
relevant information and poorer performance. Increased cortisol levels have been shown
to cause particular deficits in memory performance (Lupien et al., 1994; Seeman,
McEwen, Singer, Albert, & Rowe, 1997; Squire, 1992) and executive functioning

(Carpenter & Gruen, 1982).

77

The precise mechanisms by which corticosteroids may alter cognition are
unknown at this time. However, a few studies have shown that cortisol reduces the
average evoked potential response to relevant, but not irrelevant, stimuli (Kopell, Wittner
& Lunde, 1980; Sarrieau, Dussaillant & Agid, 1986). This in turn causes impairment in
selective attention by increasing the level of inappropriate “noise” in the system (c. g,
irrelevant stimuli). Interestingly, numerous studies have also shown that increased levels
of cortisol may impair cognitive functioning as a result of toxic effects to a region of the
brain that controls memory function, namely the hippocarnpus. The hippocarnpus
appears to be particularly sensitive to the effects of increased cortisol because it contains
the highest concentration of corticosteroid binding sites in the brain (McEwen, Davis &
Parsons, 1997; Sarrieau et al., 1986). Corticosteroids are reportedly damaging to
hippocampal cells because they act to inhibit regional cerebral glucose metabolism (de
Leon, McRae, Rusinek, Convit, & De Santi, 1997) and are associated with hippocampal
atrophy (Lupien et al., 1998; Sapolsky, 1996). By adversely affecting glucose uptake,
cortisol increases the neuronal vulnerability of hippocampal cells to a variety of
neurotoxic insults. In fact, corticosteroid levels appear to predict the magnitude of
hippocampal neuronal loss as well as the extent of cognitive impairments (Landfield,
1991)

Interestingly, the effects of corticosteroids on neuronal activity tend to observe an
inverted U-shaped curve, with extremely low and high cortisol levels suppressing
neuronal activity (Joels & de Kloet, 1994). This seems to fit nicely with the Yerkes-
Dodson theory (1908) which describes the now classic, U-shaped pattern of association

between arousal and cognitive performance. According to this well-known theory, low

78

and high levels of arousal lead to poor performance, whereas mid-range arousal promotes
optimal cognitive functioning. The biological underpinnings of the Yerkes-Dodson
theory may therefore be explained, at least in part, by the effects of corticosteroids. The
research reviewed above suggests that individuals who are anxiety-prone tend to
demonstrate higher levels of arousal, which is accompanied by increased release of
cortisol in the brain. Prolonged cortisol release is in turn associated with neurotoxic
effects, neuronal loss, and cognitive impairment. Clearly, the accumulation of research in
this area strongly suggests that increased HPA activity in the brains of anxiety-prone
individuals is related to accelerated rates of cognitive decline. While cortisol levels were
not measured in the current study, it would make sense that such neuroendocrine
responses to increased levels of anxiety may have contributed to the cognitive
deterioration seen in the areas of memory and executive functioning.

Hypothesis 5 stated that measures of depression and anxiety would be predictive
of cognitive decline when examined in relation to particular domains of functioning (e.g.,
memory, executive functioning, speed of processing), but would n_ot predict decline on a
measure of gross, non-specific cognition such as the MMSE. Repeated measures
ANCOVA analyses revealed that this hypothesis was supported, as anxiety and
depression did not result in statistically significant covariate effects when MMSE scores
were used as dependent measures. This finding is important because it suggests that
measures of global cognitive functioning, like the MMSE, are not sensitive enough to
pick up changes in cognition that are negatively impacted by chronic affective
disturbance. In the present study, trait anxiety was clearly demonstrated to have an

adverse affect on cognition, with higher trait anxiety predicting declines in memory and

79

executive functioning. However, had this investigation not used such domain-specific
measures of neuropsychological functioning, this finding would not have been detected.

A review of the literature suggests that many researchers in the field are still using
the MMSE as the primary measure of cognition in studies examining affective
disturbances among older adults. While studies that use such non-specific measures of
cognitive functioning nevertheless add to our understanding of the association between
these variables, researchers need to know which cognitive processes are most readily
influenced by anxiety and depression and which are left relatively undisturbed. It seems
that so far, investigators have not yet well connected the symptoms of anxiety and
depression to specific types of neuropsychological functioning, such as memory,
processing speed, or executive skills. Rather, these variables are most often discussed in
the literature as nonspecific mediators of age-related cognitive decline. Presumably the
effects of anxiety and depression are not ubiquitous with respect to cognition, however
little empirical work has been conducted to investigate the extent to which these variables
differentially affect specific domains of brain functioning. Findings from the current
study suggest that future research in this area should examine cognition in a way that will
allow these distinctions to be made clear.

While the results from this investigation suggest that depression and anxiety do
not predict decline on measures of gross cognitive functioning such as the MMSE, a few
alternative explanations must also be considered. First, the lack of significance
associated with anxiety/depression and the MMSE may be related to inadequate
statistical power. A power analysis conducted prior to examining the data revealed that

the present sample size of 57 subjects was sufficient for detecting medium and large

80

effect sizes, .25 and .40 respectively (Cohen, 1992), but was inadequate for effects
smaller than .10. In order to detect smaller effects, a sample size of 322 subjects would
have been needed (G-Power; Faul & Erdfelder, 1992). Therefore, it may be that anxiety
and depression indeed have a significant, but relatively small, impact on global cognitive
functioning that could not be detected with a sample size of 57 subjects.

A third explanation for the lack of association between mood measures and global
cognitive functioning may have to do with restricted variance among MMSE scores. In
order to ensure that all participants were healthy and functioning at a level consistent with
normal aging, subjects were excluded from participation in the study if their performance
on the MMSE fell below the 23-point cutoff. Clearly, this procedure acted to restrict
variance on this measure as those who performed the most poorly were not included in
the analyses. It is possible that, if these subjects had been included in the study, the
increased variance may have led to significant results. In a related vein, it should be
noted that the change in MMSE scores between Time 1 and Time 2 was actually quite
small overall (M=1.49 points). It may simply be that this small decrement represents
unsystematic change that could not be predicted in the analyses.

Methodological Considerations

This investigation sought to add to the existing literature on the effects of anxiety
and depression on cognitive functioning in older adults by improving upon the
methodological and theoretical shortcomings of earlier studies. This was accomplished
in the following important ways: (1) by examining the variables in a within-subjects
longitudinal design, thereby allowing for the direct exploration of the long-term effects of

anxiety and depression on cognitive functioning (2) controlling for cohort effects, which

81

has not been accomplished in earlier investigations using cross-sectional data (3) using
neuropsychological measures that assessed the specific cognitive functions of memory,
speed of processing, and executive functioning as opposed to measures of global, non-
specific types of cognition and by (4) employing statistical analyses using continuous
variables of anxiety, depression, and cognitive functioning which avoided the distortion
that can occur by falsely dichotomizing these constructs.

Despite these improvements in methodology, important shortcomings of this
investigation must also be noted. First, operationally defining the constructs of interest
poses an exceptionally troublesome limitation in the research of neuropsychological test
performance. What is the most accurate way of assessing constructs such as executive
functioning or speed of processing? Unfortunately, experts in the field disagree as to
what constitutes an accurate or appropriate definition of such cognitive abilities and often
debate about which instruments should be used to adequately measure them. For
example, the on-going dispute between Parkin (1998) and Baddeley (1998) regarding the
construct of executive functioning poses empirical questions that need to be considered
when designing any study involving the measurement of such abilities. These
disagreements in the field highlight the need to use multiple measures when attempting to
assess constructs of the sort examined in the present study. Indeed, one strength of this
study is that three instruments were used to measure executive functioning, including the
Wisconsin Card Sorting Test (WCST), Trail Making Test (TMT) Part B, and the Stroop
Test (ST). Likewise, two measures were used to assess speed of processing, including
the Symbol Digit Modalities Test (SDMT) and the Trail Making Test (TMT) Part A.

However, it should be noted that a limitation exists with regard to the measurement of

82

memory, as only one instrument was utilized to assess this construct in the present study.
Future research should make every attempt possible to include more than one measure of
each variable. While the use of multiple measures does not completely solve the issue of
construct validity, it can be argued that doing so contributes to a more accurate estimation
of the constructs of interest.

A second limitation of this study is concerned with the generalizability of the
research findings. The sample consisted of fairly well-functioning, normal, older adults
and was not particularly representative of the socioeconomic, educational, or cultural
variation in the general population. While there is undoubtedly interest in understanding
the processes of normal aging as it pertains to such individuals, it is important to point
out that the conclusions drawn from the present study may not generalize to the larger
population of older adults.

Future Directions

As already mentioned, future research would do well to include samples of
individuals that are more heterogeneous with respect to socioeconomic, educational, and
cultural characteristics and therefore more representative of the population as a whole. A
diverse sample not only helps to avoid the problems associated with restricted variance
and attenuated correlations, it also increases the likelihood that results will be
generalizable to a more extensive population of older adults.

In addition, future research would likely benefit from the inclusion of a younger
comparison group. The comparison of extreme age groups, as opposed to examining data
solely from a group of older adults, would be more likely to pick up on and emphasize

age-related decline in cognitive performance. Including a younger comparison group

83

would serve to increase the variation in performance assessed and, in turn, more clearly
highlight age-related differences in the areas of memory, executive functioning, and
speed of processing.

Future research may also consider the inclusion of both normal, healthy adults as
well as those who are functioning at a level inconsistent with normal aging, such as those
with Alzheimer’s disease or other dementing illnesses. This would not only allow an
investigation of the relative impact of anxiety and depression on the processes of normal
aging, but could also assess their impact on the progression of pathological diseases
processes in later life. As the present study points out, anxiety accounts for a significant
amount of variance in cognitive decline among older adults and may therefore contribute
to decline associated with certain types of dementia. Such empirical questions could not
be answered by this study and warrant further investigation.

Finally, results from this study suggest that even subclinical levels of anxiety may
interfere with cognitive functioning in older adults. Clearly, this variable can no longer
be dismissed as unimportant or irrelevant in an older adult population and must be
controlled for in future research. The creation of new diagnostic tools as well as better
treatment methods must also be encouraged. Likewise, more research is needed to
investigate whether or not behavioral and/or pharmacological interventions aimed at
reducing anxiety will be capable of attenuating the negative effects of this variable on
cognitive decline in an older adult population.

Summary
It appears that age is significantly associated with cognitive deterioration in the

areas of memory, executive functioning, and speed of processing. Contrary to

84

expectation, depressive symptomatology did not contribute to degeneration in these
domains and is therefore not considered a risk factor for cognitive decline in old age.
However, trait anxiety was found to predict a significant and unique amount of the
variance in reduction of memory and executive functioning skills. As a result, anxiety-
proneness appears to be related to accelerated rates of cognitive decline. These findings
highlight the need for future research in this area in order to more clearly elucidate the

impact of such variables on the process of normal aging.

85

APPENDIX

86

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Table 1. Sample characteristics at Time 1 and Time 2.
Participant Variable Value N
Valid Missing
Gender 57 0
Male 28 (49.1%)
Female 29 (50.9%)
Estimated Premorbid IQ M=1 11.95 (SD=5.8) 57 0
Years of Education M=17.22 (SD=2.47) 57 0
Primary OccupaLtion 57 0
Not in Labor Force 3 (5.3%)
Unskilled Labor 0
Semi-skilled Labor 2 (3.5%)
Skilled Labor 4 (7.0%)
Managerial 18 (31.6%)
Professional 30 (52.6%)
Ag; 57 0
Time 1 M 69.86 (SD=6.36)
Time 2 M=75.09 (SD=6.47)
Marital Status
Time 1 Never Married 3 (5.4%) 54 3
Married 31 (55.4%)
Widowed 13 (23.2%)
Separated 0
Divorced 7 (12.5%)
Time 2 Never Married 2 (3.5%) 57 0
Married 34 (59.6%)
Widowed 16 (28.1%)
Separated 1 (1.8%)
Divorced 4 (7.0%)
Health Rating
Time 1 Excellent 18 (32.1%) 54 3
Good 31 (55.4%)
Fair 4 (7.1%)
Poor 1 (1.8%)
Time 2 Excellent 7 (12.3%) 57 0
Good 41 (71.9%)
Fair 6 (10.5%)
Poor 3 (5.3%)

 

 

 

 

87

 

Table 2. Means and standard deviations for subjects’ Time 1 performance on all

dependent measures, including the MMSE, CVLT, SDMT, TMT, WCST, and ST.

 

 

 

 

 

 

 

 

 

 

 

Dependent Measure Time 1 N
Mean (SD)
Valid Missing

MMSE Total Score' 28-50 (1-54) 57 0
CVLT Free Recall Totall 4430 (9'83) 54 3

Semantic Clustering Scorel 2'13 (0'79) 54 3
SDMT Total Number Correct, Oral 53'” (12'51) 55 2

Trial'
TMT, Part A Total Time in Seconds2 34°36 (1232) 57 0
WCST Perseverative Errors2 16'” (12'05) 52 5
ST Interference Scorel "4'39 (7'44) 54 3
TMT, Part B Difference Score in Seconds2 53 ’95 (31'08) 57 0

 

 

 

 

 

' Higher score indicates better performance
2 Lower score indicates better performance

88

 

Table 3. Means and standard deviations for subjects’ Time 2 performance on all

dependent measures, including the MMSE, CVLT, SDMT, TMT, WCST, and ST.

 

 

 

 

 

 

 

 

 

 

 

Dependent Measure Time 2 N
Mean (SD)
Valid Missing

MMSE Total Score' 27-01 (115) 56 1
CVLT Free Recall Totall “'04 (1340) 55 2

Semantic Clustering Scorel 1'61 (0'96) 55 2
SDMT Total Number Correct, Oral 49'72 (14'00) 52 5

Trial1
TMT, Part A Total Time in Seconds2 48'” (38:60) 55 2
WCST Perseverative Errors2 1822 (133-2) 54 3
ST Interference Score1 "6'75 (6‘56) 57 0
TMT, Part B Difference Score in Seconds2 65'“ (33 '97) 56 1

 

 

 

 

 

‘ Higher score indicates better performance
2 Lower score indicates better performance

89

 

S. v 83> a .1. ”mo. v 2:9 9 ...

&ng carapace 023%: E $832280 E0802 832m cannon .262

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DOA **mm©. vomr amm. *N_m.- Sir mwfi. wvm. *oomr eta—«50m
.1.me co; *ommr *mwm. wmmr simmvr **th. 1.05m. .Ivaf :58”—
”N14080:
VON- *ommr 004 1.09%- .3440. aim—v. 3.534.- *mmmr *Nmm. Amy P2P
0mm. *mwN. .Ibamr GOA *ommr wmmfi CNN. *ooM. :wbvr AOOM—Hm
*Nfimr mef .140. *ommr co; ca. *awamr *ommr *mom. HmUB
”MCEOmHO—HSHH
0>U300Xm~
elf inmmvr *amfiv. mer SN. co." **VNm.- *oomr 1.05m. At; PEP
mwfi. **th. *hvvr CNN. **wam- **VNm.- co.“ 9%. :awvr Hznm
wvm. 1.05m. *mmmr *Oom. Inwom- *ommr 3%. Go; .130.
*OOM.- **va.- *Nmm. 1.2%.- *moM. 3.05m. fawvr .ImeH 00;
a: g
3235qu 530% PEP moomhm HMO?» HEP PEA-m mum—>22 ow<

 

 

 

 

 

 

 

025088 63% 9:38an .conamoo fine—m mo 85.0.88 _ 082. can own 5253 x59: scum—2.80

 

 

 

 

 

.boEoE was $58325.“

.v 2an

90

S. V 03? a .I. We. v big a ...

&ng acmtamfiou 838m: 3 mcosflotoo 28802 84605 :ofiwom $32

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

cog 1.3.5. .ENE wNfi. Sir moor “bf 02. .52“: eta—«Eom

sigh. GOA *acovr .Ibov. .1034- fwwcvr 35$! inhmm. 1.23:0- =Noom
OED:

in. $84.. 2: 324.- .. so. :85 *3..- *8? team. EC :5

mm“. 1.094. 1.33.- 0o; *var _mNr VWN. .1,— TV. *aoovr LOOM—Pm

Cl: 1.034.- law. .33? co; hwm. 1.594.- *oomr .1.me FWUB
a

moor .Inwovr “1.005 _mNr 5mm. 00; 1.9%.. tamwnr *Nom. 3; PEP

52. .1094. *mmvr VmN. 1.594- 1.3%.- oo._ mvv. 1.9%.-

0:. .1.th *mOmr 12v. *QONK **mwb.- 9%. oo._ :mer Mum—ZS

*oomf **w_m.- sivom. **oov.- .1.me *NQN. **mvv.- **vwm.- CO; ow<
a: ?a
oznafiom 530M HER. mOOMHm Pmua PEP PEA-m mum—22 amt.

 

 

 

 

 

 

 

25388 .3on @5885 .zowmawoo 152m mo 85938 N 08C. can own. 5253 x59: confine—o0

 

 

 

 

 

.boEoE Ea .mficosofia

.m £an

91

Table 6. Frequency of depression at Time 1 and Time 2 as measured by the BDI

 

 

 

 

 

 

 

 

 

 

and GDS.
Level of Depression Time 1 Time 2
Bl
Normal (0-9) 27 29
Mild/Moderate (10-19) 26 23
Moderate/Severe (20-63) 4 5
92;
Normal (0-9) 20 17
Mild/Moderate (10-19) 28 33
Moderate/ Severe (20-3 0) 9 7

 

 

 

 

92

Table 7.

processing speed (PS) factor at Time 1.

Correlation matrix between age, executive functioning (EF) factor, and

 

 

 

 

 

 

 

 

 

Age EF Factor PS Factor
Age 1.0 .483“ .610**
EF Factor .483** 1.0 .573**
PS Factor .610** .573“ 1.0
Table 8. Correlation matrix between age, executive functioning (EF) factor, and

processing speed (PS) factor at Time 2.

 

 

 

 

 

Age EF Factor PS Factor
Age 1.0 .544* .492* *
EF Factor .544“ 1.0 .574”
PS Factor .492” .574” 1.0

 

 

 

 

93

 

 

Table 9.

Repeated measures ANCOVA results showing covariate effects of age and

BDI depression scores on free recall memory, semantic clustering, processing speed, and

executive functioning.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DV Covariate F Wilks’ p Value Partial Eta
Squared
Free Recall Memory 6.580 .803 .022* .197
Age 5.279 .908 .026* .092
BDI Time 1 .005 1.000 .944 .000
BDI Time 2 .623 .988 .434 .012
Semantic Clustering 2.789 .899 .037 * .101
Age 2.582 .922 .046* .078
BDI Time 1 .025 1.000 .876 .000
BDI Time 2 1.766 .967 .190 .033
Processing Speed 6.998 .876 .009M. .124
Age 6.651 .883 .012* .117
BDI Time 1 .467 .991 .498 .009
BDI Time 2 .044 .999 .834 .001
Executive Functioning 6.021 .752 .019* .248
Age 4.192 .805 .034* .195
BDI Time 1 .557 .989 .459 .011
BDI Time 2 .909 .983 .345 .017

 

 

 

 

 

 

Note: *p<.05 **p<.01

94

 

Table 10.

Repeated measures ANCOVA results showing covariate effects of age and

GDS depression scores on free recall memory, semantic clustering, processing speed, and

executive functioning.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DV Covariate F Wilks’ p Value Partial Eta
Squared
Free Recall Memory 6.407 .809 .024* .191
Age 5.222 .911 .027* .089
GDS Time 1 2.058 .962 .157 .038
GDS Time 2 2.666 .951 .109 .049
Semantic Clustering 3.014 .880 .027 * .120
Age 2.792 .915 .036* .085
GDS Time 1 1.181 .978 .282 .022
GDS Time 2 2.355 .957 .131 .043
Processing Speed 3.235 .908 .025* .092
Age 2.858 .926 .037* .074
GDS Time 1 1.660 .969 .203 .031
GDS Time 2 .029 .999 .866 .001
Executive Functioning 5.954 .761 .020* .239
Age 4.330 .789 .031* .202
GDS Time 1 2.851 .948 .097 .052
GDS Time 2 2.389 .956 .128 .044

 

 

 

 

 

 

Note: *p<.05 **p<.01

95

 

Table 11. Repeated measures ANCOVA results showing covariate effects of age and

trait anxiety on free recall memory, semantic clustering, processing speed, and executive

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

functioning.
DV Covariate F Wilks’ p Value Partial Eta
Squared

Free Recall Memory 7.049 .881 .010** .1 19
Age 7.449 .875 .009“ .125
STAI Time 1 6.271 .901 .025* .099
STAI Time 2 4.953 .938 .039* .062

Semantic Clustering 2.615 .901 .040* .099
Age 2.580 .904 .047* .096
STAI Time 1 2.019 .918 .051* .082
STAI Time 2 1.997 .939 .057 .061

Processing Speed 6.944 .882 .011* .118
Age 6.602 .887 .013* .113
STAI Time 1 .727 .986 .398 .014
STAI Time 2 .463 .991 .499 .009

Executive Functioning 4.610 .787 .027* .213
Age 4.534 .791 .028* .209
STAI Time 1 3.682 .916 .041* .084
STAI Time 2 3.125 .945 .052* .055

 

 

 

 

 

 

Note: *p<.05 **p<.01

96

 

Table 12.

Repeated measures ANCOVA results showing covariate effects of age,

BDI depression, GDS depression, and trait anxiety on MMSE scores.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

DV Covariate F Wilks’ p Value Partial Eta
Squared

MMSE 3.105 .838 .043* .162
Age 2.649 .851 .051* .149

BDI Time 1 .503 .990 .481 .010

BDI Time 2 .473 .991 .495 .009

3.619 .819 .036* .181

Age 3.502 .824 .037* .176

GDS Time 1 .214 .996 .646 .004

GDS Time 2 .698 .987 .407 .013

4.973 .709 .023* .291

Age 4.448 .813 .030* .187

STAI Time 1 1.312 .975 .257 .025

STAI Time 2 .998 .981 .322 .019

 

 

 

 

 

 

Note: *p<.05 **p<.01

97

 

LIST OF REFERENCES

98

LIST OF REFERENCES

Alexopoulos, G. S, Young, R. C., & Shindeckler, R. (1992). Brain computed
tomography in geriatric depression and primary degenerative dementia.
Biological Psychiatry, 31, 591-599.

Alexopoulos, G. S. (1989). Biological abnormalities in late life depression. Journal of
Geriatric Psychiatry, 141, 25-34.

Alexopoulos, G. S. (1990). Anxiety-depression syndromes in old age. International
Journal of Geriatric Psychiatry, 5, 351-353.

American Psychiatric Association (1987). Diagnostic and statistical manual of mental
disorders, 3rd edition, revised. Washington, DC: American Psychiatric
Association.

American Psychiatric Association (1994). Diagnostic and statistical manual of mental
disorders, 4th edition. Washington, DC: American Psychiatric Association.

Ardila, A., & Rosselli, M. (1989). Neuropsychological characteristics of normal aging.
Developmental Neuropsychology, 5, 307-320.

Averill, P. M., & Beck, J. G. (2000). Posttraumatic Stress Disorder in older adults: A
conceptual review. Journal of Anxiety Disorders, 14, 133-156.

Axelrod, B. N., Goldman, R. S. & Henry, R. R. (1992). Sensitivity of the Mini-Mental
State Examination to frontal lobe dysfunction in normal aging. Journal of
Clinical Psychology, 48, 68-71.

Backman, L., & F orsell, Y. (1994). Episodic memory functioning in a community-based
sample of old adults with major depression: Utilization of cognitive support.
Journal of A bnormal Psychologi, 103, 361-370.

Baddeley, Alan. (1998). The central executive: A concept and some misconceptions.
Journal of the International Neuropsychological Society, 4, 523-526.

Banazak, D. A. (1997). Anxiety disorders in elderly patients. Journal of the American
Board of Family Practice, 10, 280-289.

Barr, R. A., & Giambra, L. M. (1990). Age-related decrement in auditory selective
attention. Psychology and Aging, 5, 597-599.

Beck, A. T. (1970). Depression: Causes and treatment. Philadelphia: University of
Pennsylvania Press.

99

Beck, A. T. (1987). Beck Depression Inventory. San Antonio, TX: The Psychological
Corporation.

Beidel, D. C., & Turner, S. M. (1991). Anxiety disorders. In Hersen, M. & Turner, S. M.
(Eds), Adult psychopathology and diagnosis (pp. 226-278). New York, NY:
Wiley.

Bench, C. J., Friston, K. J., Brown, R. G., Frackowiak, R. S. J., & Dolan, R. J. (1993).
Regional cerebral blood flow in depression measured by positron emission
tomography: The relationship with clinical dimensions. Psychological Medicine,
23, 579-590.

Bench, C. J ., F riston, K. J ., Brown, R. G., Scott, I. C., Frackowiak, R. S. J., & Dolan, R.
J. (1992). The anatomy of melancholia - focal abnormalities of cerebral blood
flow in major depression. Psychological Medicine, 22, 607-615.

Bielauskas, L. A. & Lamberty, G. J. (1992). Assessment of depression in elderly
patients. Clinical Neuropsychologist, 6, 322.

Bieliauskas, L. A., Costello, 8., &Terpenning, M. (1991). Depression and screening for
cognitive deficit in elderly patients. Journal of Clinical and Experimental
Neuropsychology, 11, 597-604.

Bieliauskas, L., Boczar, J., & Lamberty, G. (1991). Lack of depression effects on
cognitive functions in the elderly. Journal of Clinical and Experimental
Neuropsychology, 13, 433.

Blanchard, M. (1996). Depressive illness. In Ebrahim, S., & Kalache, A. (Eds),
Epidemiology in old age. London, UK: BMA.

Blazer, D. G. (1993). Depression in late life, 2nd ed. St. Louis, MO: Mosby.

Blazer, D. G., & Williams, C. D. (1980). Epidemiology of dysphoria and depression in
an elderly population. American Journal of Psychiatry, 13 7, 439-444.

Blazer, D., Birchett, B., Service, C., & George, L. C. (1991). The association of age and
depression among the elderly: An epidemiological exploration. Journal of
Gerontology, 46, M210-M21 5.

Blazer, D., George, L. K., & Hughes, D. (1991). The epidemiology of anxiety disorders:

An age comparison. In Salzrnan, C. & Lebowitz, B. D. (Eds.), Anxiety in the
elderly: Treatment and research (pp. 17-30). New York, NY: Springer.

100

Bliwise, N., McCall, M. E., & Swan, S. J. (1987). The epidemiology of mental illness in
late life. In E. E. Lurie & J. H. Swan (Eds), Serving the mentally ill elderly:
Problems and perspectives (pp. 3-3 8). Toronto, ON: Lexington.

Bradley, B. P., Mogg, K., & Williams, R. (1994). Implicit and explicit memory for
emotional information in non-clinical subjects. Behavior Research and Therapy,
32, 65-78.

Brink, T. L., Yesavage, J. A., Lum, 0., Heersema, P. H., Adey, M. & Rose, T. S. (1982).
Screening tests for geriatric depression. Clinical Gerontologist, 1, 37-43.

Brown, C., Schulberg, H. C., & Madonia, M. J. (1995). Assessing depression in primary
care practice with the Beck Depression Inventory and the Hamilton Rating Scale
for Depression. Psychological Assessment, 7, 59-65.

Bruce, M. L., & Hoff, R. A. (1994). Social and physical health risk factors for first onset
major depressive disorder in a community sample. Social Psychiatry and

Psychiatric Epidemiology, 29, 165-171.

Bruce, M. L., & McNamara, R. (1992). Psychiatric status among the homebound elderly:
An epidemiological perspective. Journal of the American Geriatric Society, 40,
561-566.

Buntinx, F ., Kester, A., Berrgers, J., & Knottnerus, J. A. (1996). Is depression in elderly
people followed by dementia? A retrospective cohort study based in general
practice. Age and Aging, 25, 231-233.

Burke, H. R. (1972). Raven’s Progressive Matrices: Validity, reliability and norms.
Journal of Psychology, 82, 253-257.

Burt, D. B., Zembar, M. J ., & Niederehe, G. (1995). Depression and memory
impairment: A meta-analysis of the association, its pattern and specificity.
Psychological Bulletin, 11 7, 285-305.

Burvill, P. W., Johnson, G. A., & Jamrozik, K. D. (1995). Anxiety disorders after stroke:
Results from the Perth Community Stroke Study. British Journal of Psychiatry,
166, 328-332.

Cadoret, R. J., & Widmer, R. B. (1988). The development of depressive symptoms in
elderly following onset of severe physical illness. Journal of Family Practice, 27,

71 -76.

Caine, E. D., Lyness, J. M., & King, D. A. (1993). Reconsidering depression in the
elderly. The American Journal of Geriatric Psychiatry, 1 , 4-20.

101

Caligiuri, M. P., & Ellwanger, J. (2000). Motor and cognitive aspects of motor
retardation in depression. Journal of Affective Disorders, 5 7, 83-93.

Carpenter, W. T., & Gruen, P. H. (1982). Cortisol’s effect on human mental functioning.
Journal of Clinical Psychopharmacologv, 2, 91-101.

Carpiniello, B., Carta, M. G. & Rudas, N. (1989). Depression among elderly people: A
psycho-social study of urban and rural populations. Acta Psychiatrica
Scandinavica, 80, 445-450.

Castillo, C. S., Starkstein, S. E., & Fedoroff, J. P. (1993). Generalized anxiety disorder
after stroke. Journal of Nervous Mental Disorders, 181 , 100-106.

Cerella, J. (1990). Aging and information-processing rate. In J. E. Birren & K. W. Shaie
(Eds), Handbook of the Psychology of Aging (3rd ed., pp. 201-221). New York:
Academic.

Coffey, C. E., Figiel, G. S., Djang, W. T., & Weiner, R. D. (1990). Subcortical

hyperintensity on magnetic resonance imaging: A comparison of normal and
depressed elderly subjects. American Journal of Psychiatry, 147, 187-189.

Cohen, D., Eisdorfer, C., Vitaliano, P. P., & Bloom, V. (1980). The relationship between
age, anxiety, and serum immunoglobulins with crystallized and fluid intelligence.
Biological Psychiatry, 15, 699-709.

Cohen, G. D. (1991). Anxiety and general medical disorders. In Salzrnan, C. &
Lebowitz, B. D. (Eds), Anxiety in the elderly: Treatment and research (pp. 47-
62). New York, NY: Springer.

Cohen, J. (1992). A Power Primer. Psychological Bulletin, 112(1), 155-159.

Cohen, R. M., Weingartner, H., Smallberg, S. A., Pickar, D., & Murphy, D. L. (1982).
Effort and cognition in depression. Archives of General Psychiatry, 39, 593-597.

Copeland, J. R. M., Dewey, M. E., Wood, N., Searle, R., Davidson, 1. A., & McWilliam,
C. (1987). Range of mental illness among the elderly in the community:
Prevalence in Liverpool using the GMS-AGECAT package. British Journal of
Psychiatry, 150, 815-823.

Costa, P. T., Fozard, J. L., McCrae, R. R., & Bosse, R. (1976). Relations of age and
personality dimensions to cognitive ability factors. Journal of Gerontology, 31,
663-669.

Crum, R. M., Anthony, J. C., Bassett, S. S. & Folstein, M. F. (1993). Population-based

norms for the Mini-Mental State Examination by age and education level.
Journal of the American Medical Association, 269, 23 86-2391.

102

Cummings, J. L., & Benson, D. F. (1983). Dementia: A clinical approach. Boston, MA:
Butterworth.

Darke, S. (1998). Effects of anxiety on interential reasoning task performance. Journal
of Personality and Social Psychology, 55, 499-505.

Dean, A., Kolody, B., & Wood, P. (1990). Effects of social support from various sources
on depression in elderly persons. Journal of Health and Social Behavior, 31, 148-

161.

de Leon, M. J ., McRae, T., Rusinek, H., Convit, A., De Santi, S., Tarshish, C., Golomb,
J ., Volkow, N., Daisley, K., Orentreich, N., & McEwen, B. (1997). Cortisol
reduces

hippocampal glucose metabolism in normal elderly, but not in Alzheimer’s disease.
Journal of Clinical Endocrinology and Metabolism, 82, 3251-3259.

Deptula, D., Singh, R., & Pomara, N. (1993). Aging, emotional states, and memory.
American Journal of Psychiatry, 150, 429-434.

de Quervain, D. J., Roozendaal, B., & McGaugh, J. L. (1998). Stress and glucocorticoids
impair retrieval of long-term spatial memory. Nature, 394, 787-790.

Diagneault, 8., Braun, C. M. J ., Witaker, H. A. (1992). Early effects of normal aging on
perseverative and non-perseverative prefrontal measures. Developmental
Neuropsychology, 8, _99-1 14.

Dimond, M., Lund, D. A., & Caserta, M. S. (1997). The role of social support in the first
two years of bereavement in an elderly sample. Gerontologist, 27, 599-604.

Drevets, W. C., Videen, T. 0., Price, J. L., Preskom, S. H., Carmichael, S. T., & Raichle,
M. E. (1992). A firnctional anatomical study of unipolar depression. Journal of
Neuroscience, 12, 3628-3641.

Dufouil, C., Fuhrer, R., Dartigues, J. F ., & Alperovitch, A. (1996). Longitudinal analysis
of the association between depressive symptomatology and cognitive
deterioration. American Journal of Epidemiology, 144, 634-641.

Elliott, R., Sahakian, B. J ., McKay, A. P., Herrod, J. J ., Robbins, T. W., & Paykel E. S.
(1996). Neuropsychological impairments in unipolar depression: The influence of
perceived failure on subsequent performance. Psychological Medicine, 26, 975-
989.

Emerson, J. P., Burvill, P.W., Finlay-Jones, R. et al. (1989). Life events, life difficulties

and confiding relationships in the depressed elderly. British Journal of
Psychiatry, 155, 787-792.

103

Emery, O. B., & Breslau, L. D. (1989). Language deficits in depression: Comparisons
with SDAT and normal aging. Journal of Gerontology, 44, M85-M92.

Eppinger, M. G., Craig, P. L., Adams, R. L. & Parsons, O. A. (1987). The WAIS-R
index for estimating premorbid intelligence: Cross-validation and clinical utility.
Journal of Consulting and Clinical Psychology, 55(1), 86-90.

Ettner, S. L., & Hermann, R. C. (1997). Provider specialty choice among Medicare
beneficiaries treated for psychiatric disorders. Health Care Financing Review,
18, 43-59.

Evans, S. & Katona, C. L. E. (1993). The epidemiology of depressive symptoms in
elderly primary care attenders. Dementia, 4, 327-333.

Eysenck, M. W. (1992). Anxiety: The cognitive perspective. Hillsdale, NJ: Erlbaum.

Eysenck, M. W., & Calvo, M. G. (1992). Anxiety and performance: The processing
efficiency theory. Cognition and Emotion, 6, 409-434.

Paul, F., & Erdfelder, E. (1992). GPOWER: Aprior, post-hoe, and compromise power
analyses for MS-DOS [computer program]. Bonn, F RG: Bonn University,
Department of Psychology.

Feehan, M., Knight, R. G., & Partridge, F. M. (1991). Cognitive complaint and test
performance in elderly patients suffering from depression or dementia.
International Journal of Geriatric Psychiatry, 6, 287-293.

Flint, A. J. (1994). Epidemiology and comorbidity of anxiety disorders in the elderly.
American Journal of Psychiatry, 151, 640-649.

Folstein, M. F., Folstein, S. E. & McHugh, P. R. (1975). ‘Mini-Mental State’: A
practical method for grading the cognitive state of patients for the clinician.
Journal of Psychiatric Research, 12, 189-198.

Ford, C. V., & Sbordone, R. J. (1980). Attitudes of psychiatrists toward elderly patients.
American Journal of Psychiatry, 13 7, 571-575.

Foster, J. R., Sclan, S., Welkowitz, J., Boksay, I. & Seeland, I. (1988). Psychiatric
assessment in medical long-term care facilities: Reliability of commonly used
rating scales. International Journal of Geriatric Psychiatry, 3, 229-233.

Fredrikson, M., Annas, P., Fischer, H., & Wik, G. (1996). Gender and age differences in
the prevalence of specific fears and phobias. Behavior Research and Therapy, 34,
33-39.

104

Gainotti, G., & Marra, C. (1994). Some aspects of memory disorders clearly distinguish
dementia of the Alzheimer’s type from depressive pseudo-dementia. Journal of
Clinical and Experimental Neuropsychology, 16, 665-678.

Geffen, G., Bate, A., Wright, M., Rozenbilds, U., & Geffen, L. (1993). A comparison of
cognitive impairments in dementia of the Alzheimer’s type and depression in the
elderly. Dementia, 4, 294-300.

Golden, C. J. (1978). Stroop Color and Word Test. Chicago, IL: Stoelting Press.

Golden, C. J. (1981). Diagnosis and rehabilitation in clinical neuropsychology, 2nd
edition. New York, NY: Wiley.

Greenberg, P. E., Stigline, L. E., F inkelstein, S. N., & Bemdt, E. R. (1993). The
economic burden of depression in 1990. Journal of Clinical Psychiatry, 54, 405-
418.

Gurian, B. S., & Miner, J. H. (1991). Clinical presentation of anxiety in the elderly. In
Salzrnan, C. & Lebowitz, B. D. (Eds), Anxiety in the elderly: Treatment and
research (pp. 31-44). New York, NY: Springer.

Haaland, K. Y., Vranes, L. F., Goodwin, J. S., & Garry, P. J. (1987). Wisconsin Card
Sort test performance in a healthy elderly population. Journal of Gerontology,
42, 345-346.

Hart, R. P., Kwentus, J. A., Taylor, J. R., & Harkins, S. W. (1987). Rate of forgetting in
dementia and depression. Journal of Consulting and Clinical Psychology, 55 ,
101-105.

Hartley, J. T. (1986). Reader and text variables as determinants of discourse memory in
adulthood. Psychology and Aging, 1, 150-158.

Hartley, J. T. (1993). Aging and prose memory: Tests of the resource-deficit hypothesis.
Psychology and Aging, 8, 538-551.

Heaton, R. K., Chelune, G. J ., Talley, J. L., Kay, G. G., & Curtiss, G. (1993). Wisconsin
Card Sorting Test manual: Revised and expanded. Odessa, FL: Psychological
Assessment Resources.

Henderson, A. S., Korten, A. E., Jacob, P. A., Mackinnon, A. J ., Jorrn, A. F., Christensen,
H., & Rogers, B. (1997). The course of depression in the elderly: A longitudinal
community-based study in Australia. Psychological Medicine, 27, 119-129.

Henry, G., Weingartner, H., & Murphy, D. L. (1973). Influence of affective states and

psychoactive drugs on verbal learning and memory. American Journal of
Psychiatry, 130, 966-971.

105

Hill, D. C., Thompson, L. W., & Gallagher, D. (1988). The role of anticipatory
bereavement in older women’s adjustment to widowhood. Gerontologist, 28,
792-796.

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

Himmelfarb, S., & Murrel, S. A. (1984). The prevalence and correlates of anxiety
symptoms in older adults. Journal of Psychology, 116, 159-167.

Hockey, G. R. J. (1986). Changes in operator efficiency as a firnction of environmental
stress, fatigue, and circadian rhythms. In K. R. Boff, L. Kaufinann, & J. P.
Thomas (Eds), Handbook of perception and human performance, Vol. 2
Cognitive processes and performance (pp. 1-49). New York: Wiley.

Hultsch, D. F., Hertzog, C., & Dixon, R. A. (1990). Ability correlates of memory
performance in adulthood and aging. Psychology and Aging, 5, 356-368.

Husaini, B. A., Moore, S. T., Castor, R. S. et al. (1991). Social density, stressors and
depression: Gender differences among the black elderly. Journal of Gerontology,
46, 236-242.

Hyer, L. & Blount, J. (1984). Concurrent and discriminant validities of the GDS with
older psychiatric patients. Psychological Reports, 54, 611-616.

J oels, M., & de Kloet, E. R. (1994). Mineralocorticoid and glucocorticoid receptors in
the brain: Implications for ion permeability and transmitter systems. Progress in
Neurobiology, 43, 1-36.

Johnson, J., Weissman, M. M., & Klerman, G. L. (1992). Service utilization and social
morbidity associated with depressive symptoms in the community. Journal of the
American Geriatrics Society, 121, 89-99.

Judd, L. L., Rapaport, M. H., Paulus, M. P., et al. (1994). Sub-syndromal symptomatic
depression: A new mood disorder. Journal of Clinical Psychiatry, 55 (4), 18-28.

Kabacoff, R. I., Segal, D. L., Hersen, M., & Van Hasselt, V. B. (1997). Psychometric
properties and diagnostic utility of the Beck Anxiety Inventory and the State-Trait
Anxiety Inventory with older adult psychiatric outpatients. Journal of Anxiety
Disorders, 11, 33-47.

Kastenbaum, R. J. (1994). Anxiety in old age. In Wolman, B. B., & Stricker, G. (Eds),
Anxiety and related disorders (pp. 209-225). New York, NY: Wiley.

106

Katz, 1. R. (1996). On the inseparability of mental and physical health in aged persons:
Lessons from depression and medical comorbidity. American Journal of
Geriatric Psychiatry, 4, 1-16.

Kennedy, G. J., Kelman, H. R., Thomas, C. et al. (1989). Hierarchy of characteristics
associated with depressive symptoms in an urban elderly sample. American
Journal of Psychiatry, 146, 220-225.

Kennedy, G. J ., Kelman, H. R., Thomas, C. et al. (1990). The emergence of depressive
symptoms in late life: The importance of declining health and increasing
disability. Journal of Community Health, 15, 93-104.

King, D. A., Caine, E. D., Conwell, Y., & Cox, C. (1991a). The neuropsychology of
depression in the elderly: A comparative study of normal aging and Alzheimer’s
disease. Journal of Neuropsychiatry and Clinical Neurosciences, 3, 163-168.

King, D. A., Caine, E. D., Conwell, Y., & Cox, C. (1991b). Predicting severity of
depression in the elderly at six-month follow-up: A neuropsychological study.
Journal of Neuropsychiatry and Clinical Neurosciences, 3, 64-66.

King, G. D., Hannay, H. J., Masek, B. J., & Burns, J. W. (1978). Effects of anxiety and
sex on neuropsychological tests. Journal of Consulting and Clinical Psychology,
46, 375-376.

Kirschbaum, 0, Wolf, 0. T., May, M., Wippich, W., & Hellhammer, D. H. (1996).
Stress and treatment induced elevations of cortisol levels associated with impaired
declarative memory in healthy adults. Life Sciences, 58, 1475-1483.

Koenig, H. G., & Blazer, D. G. (1996). Minor depression in late life. American Journal
of Geriatric Psychiatry, 4 (1), $14-$21.

Koenig, H. G., Meador, K. G., Cohen, H. J. & Blazer, D. G. (1988). Self-rated
depression scales and screening for major depression in older hospitalized patients
with medical illnesses. Journal of the American Geriatrics Society, 36, 699-796.

Kopell, B. S., Wittner, W. K., & Lunde, D. (1980). Cortisol effects on averaged evoked
potentials, alpha-rhythm, time estimation, and two-flash fusion threshold.
Psychosomatic Medicine, 32, 39-49.

Kral, V. A. (1983). The relationship between senile dementia (DAT) and depression.
Canadian Journal of Psychiatry, 28, 304-306.

Krasucki, C., Howard, K, & Mann, A. (1998). The relationship between anxiety
disorders and age. International Journal of Geriatric Psychiatry, 13, 79-99.

107

Krishnan, K. R. R., Goli, V., Ellinwood, E. H., France, R. D., Blazer, D. G., & Nemeroff,
C. B. (1988). Leukoencephalopathy in patients diagnosed as major depressive.
Biological Psychiatry, 23, 519-522.

Krout, J. A. (1983). Knowledge and use of services by the elderly: A critical review of
the literature. International Journal of A ging and Human Development, 1 7, 153-
167.

Kuhlen, R. G. (1964). Personality change with age. In P. Worchal & D. Byme (Eds),
Personality change. New York: Wiley.

Kupfer, D. J. (1991). Long-term treatment of depression. Journal of Clinical Psychiatry,
52 (5), 28-34.

Landfield, P. W., Baskin, R. W., & Pitler, T. A. (1991). Brain-aging correlates:
Retardation by horrnonal-pharmacological treatments. Science, 214, 581-584.

LaRue, A. (1989). Patterns of performance on the Fuld Object Memory Evaluation in
elderly inpatients with depression or dementia. Journal of Clinical and
Experimental Neuropsychology, 11, 409-422.

LaRue, A. (1992). Depression. Aging and Neuropsychological assessment (pp. 259-
289). New York, NY: Plenum Press.

LaRue, A., D’Elia, L. F., Clark, E. 0., Spar, J. E., & Jarvik, L. F. (1986). Clinical tests of
memory in dementia, depression and healthy aging. Journal of Psychology and
Aging, 1, 69-77.

LaRue, A., Goodman, S., & Spar, J. E. (1992). Risk factors for memory impairment in
geriatric depression. Neuropsychiatry, Neuropsychology, and Behavioral

Neurology, 5, 178-184.

Lasoski, M. C., & Thelen, M. H. (1987). Attitudes of older and middle-aged persons
toward mental health intervention. The Gerontologist, 27, 288-292.

Leon, M. R. (1989). Anxiety and inclusiveness of information processing. Journal of
Research in Personality, 23, 85-98.

Lepine, J. P., & Bouchez, S. (1998). Epidemiology of Depression in the elderly.
International Clinical Psychopharmacology, 13 (5), S7-S12.

Lezak, Muriel D. (1987). Norms for growing older. Developmental Neuropsychology,
3(1), 1-12.

Lezak, Muriel D. (1995). Neuropsychological assessment, 3rd ed. New York: Oxford
University Press.

108

Liddell, A., Locker, D., & Burman, D. (1991). Self-reported fears (FSS-II) of subjects
aged 50 years and older. Behavior Research and Therapy, 29, 105-112.

Lindesay, J. (1991b). Phobic disorders in the elderly. British Journal of Psychiatry, 15 9,
531-541.

Lindesay, J., Briggs, K., & Murphy, E. (1989). The guy’s/age concern survey:
Prevalence rates of cognitive impairment, depression, and anxiety in an urban
elderly community. British Journal of Psychiatry, 155, 317-329.

Little, R., & Rubin, D. (1987). Statistical analysis with missing data. New York: John
Wiley and Sons.

Little, R., & Schenker, N. (1995). Missing data. In Arminger, Clogg, & Sobel (Eds)
Handbook of statistical modeling for the social and behavioral sciences. New
York: Plenum.

Loewenstein, D. A., D’Elia, L., Guterman, A., Eisdorfer, C., Wilkie, F.‘, LaRue, A.,
Mintzer, J ., & Duara, R. (1991). The occurance of different intrusive errors in
patients with Alzheimer’s disease, multiple cerebral infarctions, and major
depression. Brain and Cognition, 16, 104-117.

Lopez, 0. L., Boller, F., Becker, J. T., Miller, M., & Reynolds, C. F. (1990). Alzheimer’s
disease and depression: Neuropsychological impairment and progression of the
illness. American Journal of Psychiatry, 14 7, 855-860.

Lupien, S., Lecours, A. R., Lussier, I., Schwartz, G., Nair, N. P. V., & Meaney, M. J.
(1994). Basal cortisol levels and cognitive deficits in human aging. Journal of
Neuroscience, 14(5), 2893-2903.

Lurie, E. E. (1987). The interrelationship of physical and mental illness in the elderly. In
E. E. Lurie & J. H. Swan (Eds), Serving the mentally ill elderly: Problems and
perspectives (pp. 39-60). Toronto, OH: Lexington.

Macleod, C., & Donnellan, A. M. (1993). Individual differences in anxiety and the
restriction of working memory capacity. Personality and Individual Diflerences,
15, 163-173.

Magni, G., & DeLeo, D. (1984). Anxiety and depression in geriatric and adult medical
patients. Psychological Reports, 55, 607-612.

Magni, G., Schifano, F., deDominicis, M. G., & Belloni, G. (1988). Psychological

distress in geriatric and adult medical inpatients. Archives of Gerontology and
Geriatrics, 7, l 5 1-161 .

109

Martin, G. A. (1984). Selective attention deficits associated with aging: Anxiety in
experimental subjects? Dissertation Abstracts International, 44, 2900-B.

McEwen, B. S., Davis, P. G., & Parsons, B. (1997). The brain as a target for steroid
hormone action. Annuls of Neuroscience, 2, 65-112.

Mitchell, A. J. (1995). The contribution of hypercortisolaemia to the cognitive decline of
geriatric depression. International Journal of Geriatric Psychiatry, 10, 401-409.

Mossey, J. M., Knott, K. A.,& Craik, R. (1990). The effects of persistent depressive
symptoms of hip fracture recovery. Journal of Gerontological Medicine and
Science, 45, M163-Ml68.

Murphy, E. (1992). Social origins of depression in old age. British Journal of
Psychiatry, 141, 135-142.

Murrell, S. A., Meeks, S., & Walker, J. (1991). Protective functions of health and self-
esteem against depression in older adult facing illness or bereavement.
Psychology and Aging, 6, 352-360.

Nettelbeck, Ted & Rabbitt, Patrick M. A. (1992). Aging, Cognitive Performance, and
Mental Speed. Intelligence, 16, 189-205.

Newcomer, J. W., Craft, S., Hershey, T., Askins, K., & Bardgett, M. E. (1994).
Glucocorticoid-induced impairment in declarative memory performance in adult
humans. Journal of Neuroscience, 14, 2047-2053.

Newcomer, J. W., Selke, G., Melson, A. K., Hershey, T., Craft, S., Richards, K., &
Alderson, A. L. (1999). Decreased memory performance in healthy humans
induced by stress-level cortisol treatment. Archives of General Psychiatry, 56,

527-533.

Niederehe, G., & Camp, C. J. (1985). Signal detection analysis of recognition memory in
depressed elderly. Experimental Aging Research, 11, 207-213.

Norton, G. R., Cox, B. J., Asmundson, G. J. G., & Master, J. D. (1995). The growth of
research on anxiety disorders during the 1980’s. Journal of Anxiety Disorders, 9,
75-85.

Norton, N. A., Houston, H. 0., Johnson, J. E., & Zechmeister, E. B (1988). Memory for
event frequency as a function of depression, age, intentionality, and level of
processing. Journal of General Psychology, 115, 369-381.

Nussbaum, P. D., Kaszniak, A. W., Allender, J., & Rapcsak, S. (1995). Cognitive decline
in elderly depressed: A follow-up study. The Clinical Neuropsychologist, 9, 101-
111.

110

O’Boyle, M. O., Amadeo, M., & Self, D. (1990). Cognitive complaints in elderly
depressed and pseudodemented patients. Psychology and Aging, 5, 467-468.

O’Brien-Counihan, U. (1997). Problems of anxiety and sleep disorders in the elderly:
The GP’s perspective. British Journal of Psychiatry, 183, 3-11.

O’Connor, D. W., Pollitt, P. A., Roth, M., Roth, M., Brook, P. B. & Reiss, B. B. (1990).
Memory complaints and impairment in normal, depressed, and demented elderly
persons identified in a community survey. Archives of General Psychiatry, 4 7,
224-227.

O’Connor, D. W., Pollitt, P. A., Treasure, F. P., Brook, C. P. B. & Reiss, B. B. (1989).
The influence of education, social class and sex on Mini-Mental State scores.
Psychological Medicine, 19, 771-776.

O’Hara, M. W., Hinrichs, J. V., Kohout, F. J., Wallace, R. B., & Lemke, J. H. (1986).
Memory complaint and memory performance in the depressed elderly.
Psychology and Aging, 1, 208-214.

Oberleder, M. (1967). Adapting current psychological techniques for use in testing the
aging. The Gerontologist, 19, 188-191.

Owen, A. M., James, M., Leigh, P. N., Summers, B. A., Marsden, C. D., Quinn, N. P.,
Lange, K. W., & Robbins, T. W. (1992). F ronto-striatal cognitive deficits at
different stages of Parkinson’s disease. Brain, 115, 1727-1751.

Oxman, T. E., Berkman, L. F., Kasl, S. et al. (1992). Social support and depressive
symptoms in the elderly. American Journal of Epidemiology, 135, 356-368.

Palmer, B. W., Jeste, D. V., & Sheikh, J. I. (1997). Anxiety disorders in the elderly:
DSM-IV and other barriers to diagnosis and treatment. Journal of Aflective
Disorders, 46, 183-190.

Parkin, A. J. (1998). The central executive does not exist. Journal of the International
Neuropsychological Society, 4, 518-522.

Parmelee, P. A., Katz, 1. R., & Lawton, M. P. (1992). Incidence of depression in long-
terrn care settings. Journal of Gerontology: Medical Sciences, 47, M189-M196.

Patemiti, S., Dufouil, C., Bisserbe, J. C., & Alperovitch, A. (1999). Anxiety, depression,

psychotropic drug use and cognitive impairment. Psychological Medicine, 29,
421-428.

111

Pearson, J. L. (1998). Research in late-life anxiety: Summary of a National Institute of
Mental Health workshop on late-life anxiety. Psychopharmacology Bulletin, 34 ,
127-138.

Pearson, J. L., Teri, L., Reifler, B. V., & Raskind, M. A. (1989) Functional status and
cognitive impairment in Alzheimer’s patients with and without depression.
Journal of the American Geriatrics Society, 3 7, 1 117-1 121.

Pedhazur, E. J. (1982). Multiple regression in behavioral research: Explanation and
prediction 2nd edition. Fort Worth, TX: Holt, Rinehart and Winston.

Phillips, C. J ., & Henderson, A. S. (1991). The prevalence of depression among
Australian nursing home residents: Results using draft ICD-10 and DSM-III-R
criteria. Psychological Medicine, 21, 739-748.

Pierce, T. W., Elias, M. W., Keohane, P. J ., Podraza, A. M., Robbins, M. A., & Schultz,
N. R. (1989). Validity of a short form of the Category Test in relation to age,
education, and gender. Experimental Aging Research, 15, 137-141.

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

Priddy, J. M., Tipton, R. M., & Prohaska, T. R. (1982). Seeking help for emotional
concerns: An examination of age differences. Counseling and Values, 26, 203-
212.

Raj, B. A., Corvea, M. H., & Dagon, E. M. (1993). The clinical characteristics of panic
disorder in the elderly: A retrospective study. Journal of Clinical Psychiatry, 5 4,
15-155.

Reding, M., Haycox, J., & Blass, J. (1985). Depression in patients referred to a dementia
clinic. Archives of Neurology, 42, 894-896.

Regier, D. A., Boyd, J. H., Burke, J. D., Rae, D. S., Myers, J. K., Kramer, M., Robins, L.
N., George, L. K., Karno, M., & Locke, B. Z. (1988). One-mont prevalence of
mental disorders in the United States: Based on five epidemiological catchment
area sites. Archives of General Psychiatry, 45, 977-986.

Reitan, R. M. (195 8). Validity of the Trail Making Test as an indicator of organic brain
damage. Perceptual and Motor Skills, 8, 271 -276.

Reynolds, C. F., Perel, J. M., Kupfer, D. J ., Zimmer, B., Stack, J. A., & Hoch, C. C.

(1987). Open-trial response to antidepressant treatment in elderly patients with
mixed depression and cognitive impairment. Psychiatry Research, 21, 111-122.

112

Reynolds, W. M. & Gould, J. W. (1981). A psychometric investigation of the standard
and short form Beck Depression Inventory. Journal of Consulting and Clinical
Psychology, 49, 306-307.

Robbins, T. W., James, M., Owen, A. M., Sahakian, B. J ., McInnes, L., & Rabbitt, P.
(1994b). Cambridge Neuropsychological Test Automated Battery (CANTAB): A
factor analytic study of a large sample of normal elderly volunteers. Dementia, 5,

266-281 .

Rohling, M. L., & Scogin, F. (1993). Automatic and effortful processes in depressed
persons. Journal of Gerontology: Psychological Sciences, 48, P87-P95.

Ross, E. (1986). Effects of challenging and supportive instructions in verbal learning in
older persons. Journal of Educational Psychology, 59, 261-266.

Rotenberg, K. J., & Hamel, J. (1988). Social interaction and depression in elderly
individuals. International Journal of A ging and Human Development, 2 7(A),
305-318.

Russell, D. W., & Cutrona, C. E. (1991). Social support, stress, and depressive
symptoms among the elderly: Test of a process model. Psychology and Aging, 6,
190-201.

Sachdev, P. S., Smith, J. S., Lepan, H. A., & Rodriguez, P. (1990). Pseudodementia
twelve years on. Journal of Neurology, Neurosurgery, and Psychiatry, 53, 254-
259.

Sallis, J. F ., & Lichstein, K. L. (1982). Analysis and management of geriatric anxiety.
International Journal of Aging and Human Development, 15, 197-21 1.

Salthouse, T. A. (1985). Speed of behavior and its implications for cognition. In J. E.
Birren & K. W. Schaie (Eds), Handbook of the Psychology of Aging (2"‘1 ed., pp.
400-426). New York: Van Nostrand Reinhold.

Salthouse, T. A. (1992). What do adult age differences in the Digit Symbol Substitution
Test reflect? Journal of Gerontology: Psychological Sciences, 4 7, P121 -128.

Salthouse, T. A., & Mitchell, D. R. (1990). Effects of age and naturally occurring
experience on spatial visualization performance. Developmental Psychology, 26,
845-854.

Sapolsky, R. M. (1996). Why stress is bad for your brain. Science, 273, 749-750.
Sarrieau, A., Dussaillant, M., & Agid, F. (1986). Autoradiographic localization of

glucocorticoid and progesterone binding sites in the human post-mortem brain.
Journal of Steroid Biochemistry, 25, 717-721.

113

Savard, R. J ., Rey, A. C., & Post, R. M. (1980). Halstead-Reitan Category Test in bipolar
and unipolar affective disorders. Journal of Nervous and Mental Disease, 168,
297-304.

Schaefer, A., Brown, J. & Watson, C. G. (1985). Comparison of the validities of the
Beck, lung, and MMPI depression scales. Journal of Consulting and Clinical
Psychology, 53 , 415-418.

Schaie, K. W. (1989). Perceptual speed in adulthood: Cross-sectional and longitudinal
studies. Psychology and Aging, 4, 443-453.

Schaie, K. W. (1994). The course of adult intellectual development. American
Psychologist, 49, 304-313.

Schultz, N. R., Hoyer, W. J ., & Kaye, D. B. (1980). Trait anxiety, spontaneous flexibility
and intelligence in young and elderly adults. Journal of Consulting and Clinical
Psychology, 235-258.

Seeman, T. E., McEwen, B. S., Singer, B. H., Albert, M. S., & Rowe, J. W. (1997).
Increase in urinary cortisol excretion and memory declines: MacArthur studies of

successful aging. Journal of Clinical Endocrinology Metabolism, 82(8), 2458-
2465.

Sembi, S., Tarrier, N., & O’Neill, P. (1998). Does post-traumatic stress disorder occur
after stroke: A preliminary study. International Journal of the Geriatrics Society,
13, 315-322.

Sharnoian, C. A. (1991). What is anxiety in the elderly? In C. Saizman and B. D.
Lebowitz (Eds), Anxiety in the elderly: Treatment and research (pp. 3-15). New
York, NY: Springer.

Sheikh, J. L. (1991). Anxiety rating scales for the elderly. In C. Saizman and B. D.
Lebowitz (Eds), Anxiety in the elderly: Treatment and research (pp. 251-265).
New York, NY: Springer.

Shimamura, A. P., & Jurica, P. J. (1994). Memory interference effects and aging:
findings from a test of frontal lobe function. Neuropsychology, 8, 408-412.

Shulman, K. (1989). Conceptual problems in the assessment of depression in old age.
Psychiatry Journal of the University of Ottawa, 14, 364-371.

Shulman, K. I., Tohen, M., & Satlin, A. (1992). Mania revisted. In Arie T. (Ed), Recent
Advances in Psychogeriatrics. London, UK: Churchill Livingstone.

114

Siegel, B., & Gershon, S. (1986). Dementia, depression, and pseudodementia. In H. J.
Altman (Ed), Alzheimer ’s disease (pp. 29-44). New York, NY: Plenum Press.

Siegfried, K. (1985). Cognitive symptoms in late-life depression and their treatment.
Journal of Affective Disorders, 9 (1), $33-$40.

Smallegan, M. (1989). Level of depressive symptoms and life stresses for culturally
diverse older adults. The Gerontologist, 29, 45-50.

Smith, A. (1982). Symbol Digit Modalities Test (SMDT) Manual, Revised. Los Angeles:
Western Psychological Services.

Sobin, C., & Sackeim, H. A. (1997). Psychomotor symptoms of depression. American
Journal of Psychiatry, 154, 4-17.

Spar, J. E., & La Rue, A. (1997). Concise guide to geriatric psychiatry, 2nd edition.
Washington DC: American Psychiatric Press.

Spreen, Otfried & Strauss, Esther. (1998). A compendium of neuropsychological tests:
Administration, norms, and commentary, 2nd edition. Oxford University Press:
New York, NY. Springfield, IL: Charles C. Thomas.

Squire, L. R. (1992). Memory and the hippocarnpus: A synthesis fi'om findings with rats,
monkeys, and humans. Psychology Review, 99, 195-231.

Stanley, M. A., Beck, J. G., & Zebb, B. J. (1996). Psychometric properties of four
anxiety measures in older adults. Behavior Research and Therapy, 34, 827-838.

Stein, M. B., Heuser, I. J ., Juncos, J. L., & Uhde, T. W. (1990). Anxiety disorders in
patients with Parkinson’s disease. American Journal of Psychiatry, 14 7, 217-220.

Stephenson-Gino, P., Steiner, M., Kramer, L. et al. (1992). Depression in elderly persons
and its correlates in family practice: A Canadian study. Psychological Reports,
70, 359-368.

Stoudemire, A., Hill, C. D., Morris, R., Martino-Saltzman, D., Markwalter, H., &
Lewison, B. (1991). Cognitive outcome following tricyclic and electroconvulsive

treatment of major depression in the elderly. American Journal of Psychiatry,
148, 1336-1340.

Tebbs, V. M., & Martin, A. J. (1987). Affective disorders in the elderly: 1000-patient GP
trial on a new drug. Geriatric Medicine, 1 7, 17-21.

Turnbull, J. M. (1989). Anxiety and physical illness in the elderly. Journal of Clinical
Psychiatry, 50, 1140-1145.

115

Van Gorp, Wilfred, G., Satz, Paul, Mitrushina, Maura. (1990). Neuropsychological
processes associated with normal aging. Developmental Neuropsychology, 6(4),
279-290.

Wands, K., Merskey, H., Hachinski, V., F isman, M., Fox, H., & Boniferro, M. (1990). A
questionnaire investigation of anxiety and depression in early dementia. Journal

of the American Geriatrics Society, 38, 535-538.

Watts, F. N., Dalgleish, T., Bourke, P., & Healy, D. (1990). Memory deficit in clinical
depression: Processing resources and the structure of materials. Psychological
Medicine, 20, 345-349.

Waxman, H. M., Carner, E. A., Dubin, W., & Klein, M. (1982). Geriatric psychiatry in
the emergency department: Characteristics of geriatric and nongeriatric
admissions. Journal of the American Geriatrics Society, 30, 427-432.

Weingartner, H., Cohen, R. M., Murphy, D. L., Martello, J., & Gerdt, C. (1981).
Cognitive processes in depression. Archives of General Psychiatry, 38, 42-47.

Weingartner, H., Eckardt, M., Grafrnan, J., Molchan, S., Putnam, K., Rawlings, R., &
Sunderland, T. (1993). The effects of repetition on memory performance in
cognitively impaired patients. Neuropsychologi, 7, 385-395.

Wells, K., Golding, J ., & Burnam, M. (1988). Psychiatric disorder in a sample of the
general population with and without chronic medical conditions. American
Journal of Psychiatry, 145, 976-981.

Wetherell, J. L., Reynolds, C. A., Gatz, M., & Pedersen, N. (2002). Anxiety, cognitive
performance, and cognitive decline in normal aging. Journal of Gerontology:
Psychological Sciences, 5 78(3), P246-P255.

Wheliham, W. M., Lescher, E. L. (1985). Neuropsychological changes in frontal
functions with aging. Developmental Neuropsychology, 1, 371-380.

Whitboume, S. K. (1986). Test anxiety in elderly and young adults. International
Journal of Aging and Human Development, 7, 201-210.

Williams, J. M, Little, M. M., Scates, S., & Blockman, N. (1987). Memory complaints
and abilities among depressed older adults. Journal of Consulting and Clinical
Psychology, 55, 595-598.

Wolkowitz, O. M., & Weingartner, H. (1988). Defining cognitive changes in depression

and anxiety. A psychobiological analysis. Psychiatric Psychobiology, 3, 1315-
1385.

116

Wolkowitz, O. M., Reus, V. 1., Weingartner, H., Thompson, K., Breier, A., Doran, A.,
Rubinow, D., Pickar, D. (1990). Cognitive effects of corticosteroids. American
Journal of Psychiatry, 1 4 7, 1297-1303.

Yerkes, R. M., & Dodson, J. D. (1908). The relation of strength of stimulus to rapidity of
habit formation. Journal of Comparative Neurological Psychology, 18, 459-482.

Yesavage, J. A., Brink, T. L, Rose, T. L. & Adey, M. (1986). The Geriatric Depression
Rating Scale: Comparison with other self-report and psychiatric rating scales. In
L. W. Poon (Ed), Handbook of Clinical Memory Assessment of Older Adults.
Washington, D. C.: American Psychological Association.

Yesavage, J. A., Brink, T. L., Rose, T. L., Lum, 0., Huang V., Adey, M. B. & Leirer, V.
O. (1983). Development and validation of a Geriatric Depression Rating Scale: A
preliminary report. Journal of Psychiatric Research, 1 7, 37-49.

Zisook, S., Schneider, D., & Shuchter, S. R. (1990). Anxiety and bereavement.
Psychiatric Medicine, 8, 83-96.

Zisook, S., Shuchter, S. R., & Lyons, L. E. (1987). Predictors of psychological reactions

during the early stages of widowhood. Psychiatric Clinics of North America, 10,
355-368.

117

IIIIIIIIIIIIIIIIIIIIIIIIIIII

llllflllllllIlllllllllllllllll

1293 027