H

\
‘

104
202

 

THC-m9

TYLIBRARIES

lllllHllUllllllllllllllllilllllllmI 2|| \I

3 1293 008850

 

 

 

 

 

 

 

 

 

 

This is to certify that the

dissertation entitled

Correlates of Dementia and Depression
in the Able Elderly

presented by

Maureen Marks Cash

has been accepted towards fulfillment
of the requirements for

PhD degreein Psychology

MMM

Major professor

Date September 4, 1992

 

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

 

 

 

._—_'——n~_.-—.. _____.._.—

 

F LIBRARY
Michigan State
University.

- J

*—

 

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or before date due.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

MSU Is An Affirmative Action/Equal Opportunity Institution
chimera-94

 

 

CORRELATES OF DEMENTIA AND DEPRESSION IN THE ABLE ELDERLY

BY

Maureen Marks Cash

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

DOCTOR OF PHILOSOPHY

Department of Psychology

1992

ABSTRACT
CORRELATES OF DEMENTIA AND DEPRESSION IN THE ABLE ELDERLY

BY

Maureen Marks Cash

The present study compared three groups of able elderly
(i.e., those who live independently in the community): (1)
mildly demented group, (2) mildly depressed group, and (3)
normal controls. They were compared on the basis of memory
function, memory complaints and reported
psychological/behavioral function. One hundred and forty-
six individuals participated in this project. The test
battery included a structured interview, memory assessment,
and mental health assessment questionnaires. Scores from a
brief battery measuring evidence of dementia (Storandt,
Botwinick, Danziger, Berg, and Hughes, 1984) and two
depression measures (Geriatric Depression Scale and Hamilton
Rating Scale for Depression) were used to assign
participants to the aforementioned groups. They were then
compared based on their scores from the Brief Symptom
Inventory (the psychological/behavioral measure), the
Selective Reminding Test (the memory function measure), and
a portion of the Multilevel Assessment Instrument (the
memory complaints measure). It was found that depressed
participants had significantly more memory complaints than
the control group. The demented subjects were not different

from either of the other groups on this variable. The

depressed group and the demented group differed on only
three symptom dimensions (depression, paranoid ideation, and
psychoticism) while the demented and control group only
differed on two dimensions (somatization and phobic
anxiety). The depressed and control groups significantly
differed on all symptom dimensions. Based on global
symptomatology, it was found that the depressed group
differed from the demented group on number of symptoms
reported and general severity, but not on pure intensity of
symptoms. The demented group differed from the control
group on only the pure intensity measure. The demented
group was found to perform significantly worse on all long
term memory functions than either group, but no differently
on short term recall. There were no memory function
differences found between the depressed and the control
groups. It was found that among the variables, reported
symptomatology and memory complaints predicted most
effectively for depressed group status while the long term

memory scores predicted best for demented group status.

This work is dedicated to my husband, Bret, whose faith
in me never wavered even when mine did, and to my parents,
Allan and Sharon Marks, who always told me I could be

whatever I wanted. I'm glad I listened.

iv

ACKNOWLEDGEMENTS

The author would like to extend heartfelt thanks to Dr.
Norman Abeles for his time and guidance as chairperson of
this dissertation. Thanks are also given to Drs. Dozier
Thornton and Joseph Papsidero for their input on the
conceptual aspects of this project, and especially to Dr.
Michael Lindell for his invaluable help with the statistical

components of this study.

TABLE OF CONTENTS

List Of TableSOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO. Vii

Introduction ........................ ....................

Definitions of Dementia ............................
General Effects on Cognitive Function ..............
General Effects on Memory Function .................
Effects on Psychological State and Behavior ........
Role of Depression in Dementia .....................
Measures of Cognitive Function .....................
Measures of Memory Function ........................
Measures of Psychological State and Behavior .......

Hypotheses O O O I O O O O O O I O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O
Methads O O O O O O O O O O O O O O O O O O O O O O O O O O O O I O O O O ..... O O O O O O O O O O 0
subjects 0 O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O O 0

measures 0.0.0.... ....... 0.00000... ........ 00.00....
Procedures OOOOIOOOOOOOOOOOOOOOOOOOOOOOOOOOOCOOOOOOO

Results 0....OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO
DiscuSSion OOOOOCOOOOOOOOOOOOOOOO OOOOOOOOOOOOOOOOOOOOOOO O
Perce ived memory Funct i on ing 0 O O O O O O O O O O O O O O O O O O O O O 0
Reported Psychological/Behavioral Functioning ......
Actual Memory Functioning ..........................
Predictive Value of the Variables ..................
General Issues 0.0.0.0....OO...OOOOOOOOIOOOOOOOOOOOO
COflClusions 0.0.0.0....0.0.0.000...OOOOOOOOOOOOOOOI.

Appendix: List of Tables ................................

List of References ............ ..................... .....

vi

1

28
31
31
32
36
39
51
51
53
57
59
6O
63
66

85

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

Table

10:

11:

12:

13:

14:

15:

16:

LIST OF TABLES

T-tests of Memory Complaints ..................

Significant Main Effects for Group on
381 Global Indices and Dimensions .............

Means and Standard Deviations for
T-tests of Global Indices and Subscales ...... .

T-tests for BSI Global Indices ......... ..... ..

Main Effects and Correlations for Education
on 881 Global Indices and Dimensions ..........

T—tests for 881 DimeDSions 0.00.00.00.0000000.0

Significant Main Effects for Group
on SRT scores 00.000000000000000000000000000000

Means and Standard Deviations for
memory scores 0000000000000... ........ 0 ........

T—tests for SRT scores 000000.. ....... 0.0.0000.

Significant Main Effects and Correlations
for Covariates on SRT Scores .................

Discriminant Function Coefficients
fromAnaIYSis #1 00000000000000... 000000000000

Discriminant Function Coefficients
fromma1Y818#2 0.000000000000000 00000000000 0

Classification Results of the Two
Discriminant Function Analyses ...............

Standardized Canonical Function
coeffiCients 00.000000000000000 000000000000000

T-test Data by Group for Selection
criteria 00000000000000.000000.000000000000000

T-test Data by Group for Age and Education ...

vii

66

67

68

69

7O

71

72

73

74

75

76

77

78

79

80

81

Table 17: Correlations Within 881 Variables ............ 82
Table 18: Correlations Within SRT Variables ............ 83

Table 19: Correlations Between BSI Variables, SRT
Variables, and Memory Complaints ............. 84

viii

INTRODUCTION

Definitions of Dementia

There have been various attempts to define dementia in
such a way that it can be easily distinguished from other
disorders (e.g. McLean, 1987: Janowsky, 1982: Esser &
Vitaliano, 1988). In a review by McLean (1987), two major
classifications systems were noted. The first was found in
the Research Diagnostic Criteria (RDC: Eisdorfer & Cohen,
1980). In order to diagnose an individual as demented the
following criteria had to be met: "(1) impairment of at
least two of the following abilities: learning, attention,
memory, or orientation. (2) impairment of at least one of
the following cognitive skills: calculation, abstraction and
judgment, or comprehension. (3) problems in at least one of
the following areas: ability to work, function socially, and
relate to family and peers. (4) cerebral dysfunction as
indicated by either cerebral atrophy on CT scan or abnormal
EEG."(p.554) While these criteria delineate to some degree
the areas in which impairment must exist, they do not make
specific the exact nature or extent of these impairments.
They also do not address mood or personality changes that
may also accompany the dementing process and impact on
diagnosis.

The second major system to which McLean refers is the

DSM-III diagnosis of dementia (American Psychiatric

2
Association, 1980). The DSM-III delineates criteria for
this syndrome as follows: "(1) loss of intellectual
abilities of sufficient severity to interfere with social or
occupational functioning. (2) memory impairment. (3) at
least one of the following: impairment of abstract thinking,
impaired judgment, disturbance of higher cortical function,
and personality change. (4) no clouding of consciousness.
(5) other non-organic conditions have been reasonably
excluded and the behavior change represents cognitive
impairment in a variety of areas.”(p.111-112) These
criteria are more specifically defined in that they give
some clue as to the severity of intellectual impairment
necessary for diagnosis. The DSM-III criteria also mention
the importance of personality change and exclusion of
psychological causes. However, this definition is also
somewhat general in scope. What changes in personality
would warrant a diagnosis of dementia? What kind(s) of
memory impairment should clinicians look for to determine
whether their patients have the relevant cognitive deficits?
other criticisms include the lack of empirical support for
these criteria, and the lack of dimensionality to them (Jorm
& Henderson, 1985). Neither of these systems addresses the
problem of levels of severity, particularly for mild
dementia. Jorm and Henderson (1985) point out that this
problem is especially pertinent for studies using community

samples as opposed to clinical populations because it is

3
much more likely that the milder forms of dementia will
predominate in a community population.

To deal with the difficulties involved in
distinguishing memory changes due to early dementia from
those due to normal aging, a new "diagnostic" category has
been suggested: age-associated memory impairment (AAMI:
Crook, Bartus, Ferris, Whitehouse, Cohen, & Gershon, 1986).
The criteria for inclusion are: (1) over 50 years of age,

(2) gradual worsening of memory with complaints of memory
loss in everyday situations, (3) memory test performance one
standard deviation below mean for young adults, (4) evidence
of adequate intellectual function, and (5) absence of
evidence of dementia (i.e., >24 on Mini-Mental State Exam).
These criteria were designed to describe older people who
have no evidence of neurological impairment, but have
complaints about their memory function and show changes in
memory that could be of "functional significance" (Smith,
Ivnek, Petersen, Malec, Kokmen, & Tangalos, 1991).
Unfortunately, as presently proposed these criteria show
poor reliability in samples of able elderly. It has been
found that the rates of AAMI diagnosis are widely variable
depending upon the number and type of memory tests used
(Smith, et al., 1991).

These are some of the major problems involved in making
accurate diagnosis of dementia: even the criteria which are

supposed to operationalize it may be of insufficient use to

4

the practicing clinician. It seems reasonable that more
empirical work needs to be done in order to make more
reliable and valid diagnoses.
GENERAL EFFECTS ON COGNITIVE FUNCTION

Many studies have noted the effects that a dementing
process can have on the cognitive function of the individual
as well as the importance for accurate detection of these
cognitive dysfunctions (e.g., Lazarus, Newton, Cohler,
Lesser, & Schweon, 1987: Vitaliano, Breen, Albert, Russo, &
Prinz, 1984: and Cohen & Dunner, 1980). Lazarus et al.
(1987) found that when comparing demented patients with
healthy individuals, the demented subjects were rated as
significantly more cognitively impaired. Another study
examined the cognitive performance (attention and memory)
patterns of mildly to moderately demented elderly compared
to normal, healthy elderly (Vitaliano et al., 1984). These
investigators found that control subjects performed
significantly better than the mildly demented group in
performance of complex commands, immediate recall, and
calculations. The control subjects were also significantly
better oriented to time and place than the mildly demented
group. Winogrond & Fisk (1983) found similar results in
testing a group of elderly people in a day care setting for
Alzheimer's patients. It was found that on a test of
orientation all patients scored in the moderately to the

severely "defective" range. The overall cognitive

5
assessment showed no patient achieving better than a rating
of moderate dysfunction.

These studies suggest that in comparison to normal
elderly individuals, the demented elderly score
significantly worse on measures of cognitive function, and
can be fairly easily distinguished from a healthy group.
GENERAL EFFECTS ON MEMORY FUNCTION

Along with cognitive difficulties, memory function in
the demented elderly has been shown to be equally affected
by this progressive disorder (McLean, 1987: Harris & Dowson,
1982: Kopelman, 1986: Miller, 1973: Miller, 1975). Janowsky
(1982) reported that in the early stages of dementia, the
first indications of problems almost always include lapses
in memory. He stated that the individual ”cannot remember
specific pieces of information that were once easily
recalled (e.g., names, places, dates, or words)." In many
ways, the early stages of dementia may be hard to
distinguish from what could be viewed as the normal effects
of aging. However, it has been found in longitudinal
studies that once individuals' reach their sixties observed
declines in memory functioning are often a result of changes
in health status as opposed to normal aging (Siegler, 1983).

If one examines specific memory functions, differences
between the effects of dementing and normal processes begin
to emerge. McLean (1987) reported that patients with

primary dementia tend to have poor recent memory in

6
comparison with remote memory and have no gaps in their
memories. It was also noted that patients with dementia
tend to perform worse on tests of general memory than on
tests of concentration. other studies also support the
finding of poor recent memory in the demented through their
investigations of short-term memory capacity (e.g. Miller,
1972: Hughes, Berg, Danziger, Coben, & Martin, 1982: Gershon
& Herman, 1982: Kopelman, 1985: Kopelman, 1986: Miller,
1980). Miller (1972: 1973) found that demented patients do
poorly on word recall tests as compared to normal subjects.
Normal elderly were found to be able to recall a
significantly larger number of words from a given list.
Miller interpreted this to mean that individuals with a
dementing disorder have a reduced capacity in their short-
term memory. In this experiment, lists were repeated over
and over adding one new word each time as long as the
subject could recall the previous words. Because long-term
memory capacity is also needed in order to be successful,
Miller suggested that the demented person's ability to
"establish new material in long-term memory“ was also
impaired. However, in a later study by Miller (1975), it
was found that the demented were able to attain levels of
recall similar to those of normal subjects when an
appropriate cue (initial letter of word) was given. Miller
interpreted this to mean that long-term memory difficulties

may be due to a problem with retrieval rather than with

7

encoding new information into long-term storage. This
interpretation was examined in work done by Davis and
Mumford (1984). Their hypothesis was that if retrieval is
the problem in dementia than demented subjects should get
significantly greater benefit from any kind of cue than
should normal subjects. However, they found that not only
was there similar benefits for both groups from cues in
general, but that only certain kinds of cues (letter rather
than category) aided the demented group. These authors
suggested that there may also be a processing problem
responsible for poorer recall in demented patients: that is,
information can be placed into long-term memory, but is
encoded is such a way that normal retrieval methods are much
less effective. Davis and Mumford came to the conclusion
that demented patients likely have difficulties at both
encoding and retrieval stages of information processing.
EFFECTS ON PSYCHOLOGICAL STATE AND BEHAVIOR

Studies also indicate that dementia is often
accompanied by changes in psychological state and/or
behavior (e.g., Gershon & Herman, 1982: McLean, 1987:
Janowsky, 1982: Miller, 1980: Lazarus et al., 1987). Miller
(1980) notes that "characteristic disturbances" of mood have
been cited throughout the literature on dementia in the
elderly. Several studies cite irritablity, withdrawal,
denial, fear, anxiety, depression, helplessness,

hopelessness, and worthlessness as common mood states for

8
the demented individual (Rubin, Morris, 8 Berg, 1987;
Lazarus et al., 1987: Merriam, Aronson, Gaston, Wey, 8 Katz,
1988: Teri, Larson, 8 Reifler, 1988: Cohen 8 Dunner, 1980).
Gershon and Herman (1982) report that family members of
demented elderly often complain that their relative has
become "restless, agitated, argumentative, and suspicious"
before any truly significant cognitive impairment is
evident. Nott and Fleminger (1975) compared two psychiatric
populations: one with evidence of dementia and another
without. It was found that lability of mood and
irritability were symptoms more often found in the demented
group than in the non-demented group. Similar findings have
been confirmed elsewhere (Knesevich, Martin, Berg, 8
Danziger, 1983: Rabins, Merchant, Nestadt, 1984).

Data from the Seattle Longitudinal Study of Aging has
suggested that in normal aging the individual's behavioral
and attitudinal flexibility changes as s/he advances in
years. It has been found that there is moderately
increasing rigidity in style as participants reached their
sixties (Schaie 8 Willis, 1991). This could have an adverse
effect on the individual's ability to adapt to changes in
his/her mental status.

There has been some debate in the literature as to the
prominence of depressive symptoms in the dementing process.
Some studies have shown that demented individuals show

significantly higher levels of depression than that of

9
normal subjects (e.g., Shuttleworth, Huber, 8 Paulson, 1987:
Lazarus et al., 1987). However, a study by Knesevich et al.
(1983) found that while their demented group did score
significantly higher for depressive symptoms than did a
control group, both groups were well within the nondepressed
range. This has been supported elsewhere as well (Gottlieb,
Gur, 8 Gur, 1988). However, if memory impairment is mild,
then the losses might be made light of or defended against
in some way so depression does not become a major factor
(Salzman 8 Gutfreund, 1986).

Some work has also been done suggesting that depressive
symptoms may be related to the severity of the dementia.
Reifler, Larson, and Hanley (1982) found that RDC depressive
symptoms were typically more common in the beginning stages
of the process, and significantly less in the more severely
impaired patients. Reasons for differing results regarding
the prominence of depression in dementia include the use of
differing criteria to determine depression and the use of
observer vs. participant ratings, but these issues will be
discussed in greater detail later.

Obsessional symptoms and paranoia, which have also been
noted as prominent in the dementing process, may exist as
coping mechanisms (e.g., Rubin et al., 1987: Merriam et al.,
1988: Janowsky, 1982: McLean, 1987). As patients become
aware of increasing cognitive deficits, they may try to

compensate by keeping track of tasks that have been or need

10

to be done by writing copious notes for themselves and
relying on other kinds of "external props” (Janowsky, 1982:
Fopma-Loy, 1986). Studies indicate that subclinical levels
of obsessional symptoms may be common in people experiencing
cognitive deficits (Sher, Frost, Kushner, Crews, 8
Alexander, 1989: Sher, Mann, 8 Frost, 1984: Sher, Frost, 8
Otto, 1983). Individuals who show subclinical obsessive
checking behavior report "higher frequencies of failures of
everyday memory and cognitive functioning,“ and also score
significantly lower on memory tests than noncheckers (Sher
et al., 1984). These results have been found to be true for
psychiatric as well as nonpsychiatric populations of varying
ages (Sher et al., 1984: Sher et al., 1989). Sher et al.
(1989) also found that "checkers" indicated greater general
distress on psychological symptom measures. These findings
can be viewed as further evidence for the reports of
significant obsessive-compulsive symptoms in the demented
elderly.
ROLE OF DEPRESSION IN DEMENTIA

The ability to differentially diagnose between dementia
and depression (sometimes called "pseudodementia”) in the
elderly has long been a thorn in the side of the medical and
psychological professions. There are various reasons why it
has proven so difficult to make accurate, early diagnosis.
Probably the most troublesome one is the existence of

overlapping symptoms that occur in both of these disorders

11
in older adults. These include psychomotor retardation,
fatigue, hopelessness, worthlessness, loss of interests,
withdrawal, irritability, memory loss, and cognitive
impairment (e.g., Esser 8 Vitaliano, 1988: Gershon 8 Herman,
1982: Janowsky, 1982: Jorm, 1986: Kim 8 Hershey, 1988:
Shamoian, 1985: Straker, 1984). Another difficulty in
accurate diagnosis is that although the depressed elderly
may often complain about symptoms seen in depressives of
various ages, there may just as often be stereotypic
symptoms which are conspicuously absent from the picture
(e.g., McLean, 1987: Janowsky, 1982: Gershon 8 Herman, 1982;
Shamoian, 1985). For example, an elderly depressed patient
might not speak directly about feelings of sadness and
despair. They might not talk about guilt feelings, lower
self-esteem, or suicidal thoughts, and often the depression
is expressed in the form of somatic complaints (Janowsky,
1982). Thus evidence of cognitive impairment or complaints
of memory loss might take on greater diagnostic importance
in depression in the elderly.

Cognitive impairment has often been cited as possible
evidence of depression in the elderly, and has led to the
term "pseudodementia" to describe this condition (e.g.,
Esser 8 Vitaliano, 1988: Janowsky, 1982: Gershon 8 Herman,
1982). A number of investigators have found that memory and
cognitive impairment is significantly worse in depressive

elderly than in age-matched controls, and that the level of

12
impairment was associated with the degree of depression
(Whitehead, 1973: Weingartner, Cohen, Murphy, Martello, 8
Gerdt, 1981: Cohen, Weingartner, Smallberg, Pickar, 8
Murphy, 1982). Other studies have confirmed that depressive
elderly subjects do perform statistically worse on cognitive
testing compared to healthy age-matched controls but note
that the degree of impairment was actually fairly small.
Moreover these studies failed to find an association between
degree of impairment and severity of the depression
(Coughlan 8 Hollows, 1984: Kopelman, 1986).

It has also been suggested that depressed elderly
patients tend to complain of memory losses for both recent
and past events even though these were not obvious to the
examiner while truly demented persons try to play down their
memory deficits. These kinds of memory and cognitive
deficits appear to be accentuated by elderly depressed
individuals (Gershon 8 Herman, 1982: Esser 8 Vitaliano,
1988: Fopma-Loy, 1986: McLean, 1987). Miller (1980)
corroborates this finding stating that ”aged individuals
with clinical depression complained about memory as much as
those with altered brain function, even though the latter
showed far more severe impairment on objective tests of
memory." Thus, it may be that in ”pseudodemented”
individuals, cognitive impairment as assessed by self-report

might be more apparent than real.

13

In addition to greater number of memory complaints,
other clinical features that may aid in distinguishing
between "pseudodementia" and primary dementia have been
noted in the literature. Earlier it was noted that demented
individuals tend to have greater memory loss for recent than
for remote events, to have few memory gaps, and perform
consistently poorer on cognitive tasks. By contrast,
pseudodemented individuals tend to have equally severe
losses for recent and remote events, have specific memory
gaps, and show variability in performance on cognitive tasks
(Kim and Hershey, 1988: Kramer, 1982). It has also been
noted that depressed elderly have difficulty recalling new
information, make an increased number of errors of omission,
take fewer guessing risks in tasks, and have decreased
attentional abilities, but are typically without impairment
in other aspects of cognitive function (Salzman 8 Gutfreund,
1986). Weingartner, Cohen, Bunney, Jr., Ebert, and Kaye
(1982) reported a study that found elderly in the early
stages of dementia have trouble organizing, producing, and
making use of verbal material, while depressed elderly
tended to maintain these skills.

With respect to psychological symptoms, demented
individuals tend to show lability and greater irritability
while depressed individuals are more consistently
discouraged and have poorer self-image than demented persons

(Janowsky, 1982). Unfortunately, there are far fewer

14
studies which examine whether there are distinguishable
psychological and behavioral symptom constellations between
these two populations.

While it seems that many authors in this field have
accepted the term "pseudodementia" as appropriate for the
condition described above, there are those who find this
language to be less than precise. Some have stated that
this term is misleading because there may be an actual
organic basis to a dementing disorder which is precipitated
by depression (Folstein 8 McHugh, 1978: McAllister, 1983).
Another major problem seen with this term is that it
apparently denies the possibility that depression and
dementia could co—exist within the same individual and that
this co-existence could be the condition that has been
labelled ”pseudodementia" (Reifler et al., 1982: McAllister
8 Price, 1982).

However, one thing is clear - it is of the utmost
importance to establish an appropriate initial diagnosis.
There is some evidence that persons diagnosed with dementia
live only three years beyond the initial diagnosis while
nondemented individuals of the same age live approximately
six years longer (Steel 8 Feldman, 1979). If depression is
correctly diagnosed, appropriate treatment can be initiated.
If it is not, institutionalization, "progressive
invalidism,” and possibly death can be the result (Kramer,

1982).

15

Although there do seem to be many similarities between
the presentations of demented and depressed elderly people,
it does appear that there may be ways of distinguishing
between them clinically. It has been noted that it is
easier to make differential diagnoses in elderly persons who
are under the age of 80 and are mildly to moderately
depressed than in persons over 80 and/or severely depressed
(Salzman 8 Gutfruend, 1986). However, the literature
suggests that there currently are no tests which can
unequivocally differentially diagnose these disorders
(Shamoian, 1985). Diagnoses at this time tend to be made on
the basis of the individual's medical and psychosocial
history, the physician's clinical observations, and the
person's therapeutic response to anti-depressants.

Unfortunately there is evidence that diagnoses of
dementia made based on medical records can often lead to
misclassification. 0f research done with 27 participants in
the Baltimore Longitudinal Study of Aging who had been
retrospectively classified as demented based on prior
medical records, twelve were later found to perform on
cognitive measures in a manner inconsistent with this
diagnosis (Arenberg, 1990). This finding emphasizes the
importance of developing and using empirical means to

appropriately diagnose individuals.

16
MEASURES OF COGNITIVE FUNCTION

It was noted earlier that it is of great importance for
clinicians to accurately diagnose the presence of depression
or dementia in the elderly. This can be done by providing
evidence that there are qualitative differences in the types
of cognitive deficits involved in each condition (Jorm,
1986). Hughes et a1. (1982) believe this is particularly
important in distinguishing mildly demented persons from
depressed or normal individuals because the use of
neuropathologic studies has proven ineffective in
identifying early dementia.

One of the most widely used measures of cognitive
impairment is the Mini-Mental State Examination (MMS: Breen
et al., 1984: Cohen 8 Dunner, 1980: Esser 8 Vitaliano, 1988:
Heyman, Wilkinson, Hurwitz, Helms, Haynes, Utley, 8 Gwyther,
1987: McLean, 1987: Teri et al., 1988: Vitaliano et al.,
1984:). The MMS is composed of 30 questions which are
divided into two sections: (1) orientation, memory and
attention and (2) object recognition and ability to
understand and accurately follow verbal and written
instructions (Folstein, Folstein, 8 McHugh, 1975). It has
been found to correlate well with the DSM-III criteria for
dementia (Kay, Henderson, Scott, Wilson, Rickwood, 8
Grayson, 1985). However, it can be significantly influenced
by education and age of subjects, thus making matched

samples critical (Folstein et al., 1975). Most studies use

17
a criterion score of less than 24 as a diagnosis for
dementia when the subject is alert and making an effort
(e.g., Breen, et al., 1984: Heyman et al., 1987: Rabins et
al., 1984). One study used 3 ranges to define severity of
dementia on the MMS: scores of 26—30 were normal, 20-25 were
considered mildly demented, 15-19 were considered moderately
demented, and 5-14 were considered severely demented
(Shuttleworth, Huber, 8 Paulson, 1987).

In distinguishing mildly demented elderly from age-
matched controls, it has been found that the demented group
does significantly worse on the three-step command, serial
sevens, and time/place orientation indicating that there are
patterns of attention that can differentiate between groups
(Vitaliano et al., 1984). However Klein, Roca, McArthur,
Vogelsang, Klein, Kirby and Folstein (1985) warn against
giving undue weight to particular items on mental status
exams. These investigators concluded that orientation items
may not be sensitive indicators because 25% of demented
patients are fully oriented to person, place, and time. By
combining items to establish an overall rating, one can
achieve a high degree of reliability, yet still retain
adequate sensitivity and specificity. The MMS has also been
found to discriminate well among demented elderly, depressed
elderly with cognitive impairment, and depressed elderly
without cognitive difficulties. Scores for the MMS are

lowest in the demented as well as the depressed with

18
cognitive deficits, and highest in the depressed without
cognitive impairment (Folstein 8 McHugh, 1978).

Despite the evidence that the deficits of demented
individuals are generally more severe than those of
depressed people, some authors believe that there is no
convincing evidence for qualitative cognitive differences
that can differentiate these groups (Jorm, 1986). Jorm
(1986) believes some studies using mental status exams are
confounded because they compare depressed patients with mild
to moderate cognitive impairments with demented patients who
likely have much more severe impairments. He suggests that
the best way to examine qualitative differences would be to
compare mildly impaired demented people with depressed
patients (who are also typically mildly impaired). In this
way the level of impairment would be controlled for, and the
qualitative aspects of their deficits could be detected more
easily.

other researchers have noted that for a mental status
exam to be an effective measure of cognitive ability it
should be more than a test of whether there has been a loss
of knowledge which shows up as confusion regarding
orientation and language. Specifically, Cohen and Dunner
state that ”primary neuronal degeneration affects the way an
individual thinks and acts, and a test of the presence
and/or extent of a dementing process should be a test of how

well the individual can handle information across the

19
cognitive domain (1980, p.129)." Tests that are
neuropsychologically based may be what is needed to add to
the effectiveness of current mental status measures. McLean
(1987) states that these kinds of tests may be more reliable
ways of assessing dementia in the elderly because they use
sophisticated techniques for statistical analysis and item
selection.

One such method has been developed by Storandt,
Botwinick, Danziger, Berg, and Hughes (1984). This brief
battery for senile dementia of the Alzheimer's type (SDAT)
uses four tests including the logical memory test and mental
control tests from the Wechsler Memory Scale, Form A of the
Trailmaking Test, and word fluency for letters 8 and P. In
this study, it was found that the SDAT battery was able to
classify 100% of their subjects into the appropriate groups
(mildly demented vs. normal control). Tierney, Snow, Reid,
Zorzitto, and Fisher (1987) also found that the SDAT battery
was able to accurately classify demented individuals and
normal controls into the correct groups, but at a somewhat
lower level of accuracy (96%). It was also found, however,
that this battery was unable to distinguish dementia from
other disorders where cognitive impairment was prominent,
and suggested that cognitive tests alone might not be enough

to make this type of distinction accurately.

20

MEASURES OF MEMORY FUNCTION

Cohen and Dunner noted that although there are many
different models of memory function, one common theme is
that memory organizes material in a series of processing
stages. They assert that memory testing "should evaluate
the various stages in this process and thereby show how
efficiently a person assembles, stores, and uses information
(1980, p.130)." Memory research has become increasingly
complex, and several theories about the workings of human
memory have been developed (e.g., Broadbent, 1984: Charness
8 Bosman, 1990: Craik 8 Lockhart, 1972: Hasher 8 Zacks,
1979: Hultsch 8 Dixon, 1990). However, almost every theory
includes implicitly or explicitly the notions of long term
memory and short term memory as well as the storage and
retrieval of information into and out of them. These
general notions of memory seem to be prevalent in the
clinical research where memory variables are used to shed
light on a larger topic (such as depression or dementia)
rather than add to the research on human memory processes
(e.g., Cohen 8 Dunner, 1980: Hart, Kwentus, Taylor, 8 Hamer,
1987: Masur, Fuld, Blau, Thal, Levin, 8 Aronson, 1989:
Sternberg 8 Jarvik, 1976). Since complex memory models are
difficult to test in a clinical setting, it seems reasonable
that these general memory variables are sufficient to add to
the accuracy of clinical diagnosis. So by testing the

effectiveness of each of these general memory components,

21
clinical researchers should be able to identify qualitative
differences between disorders which seem to affect memory
function such as dementia and depression.

Support for this notion of qualitative differences in
memory impairment dependent upon disorder can be found in
the literature concerning the selective reminding procedure
first developed by Buschke (1973). This procedure was
designed to test learning and memory in the individual. The
subject is asked to recall a list of words read aloud to
him/her. After the first recall attempt the subject is
reminded only of those words which were forgotten. The
subject is then asked to recall the entire list again
(reminded and unreminded words). This continues until the
subject recalls the entire list twice consecutively or until
all test trials have been completed. It has been shown to
be one of the most sensitive measures in discriminating
between early dementia and the normal elderly because it
measures different components of general memory functioning
(Masur, et al., 1989). Specifically, measures of short term
recall (STR), long term storage (LTS), retrieval from long
term storage (LTR), and consistent long term retrieval
(CLTR) are obtained from the Selective Reminding Test (SRT:
Buschke 8 Fuld, 1974). STR is determined by the number of
words each trial which have been recalled only after
reminding each time. LTS is defined as the number of words

recalled spontaneously without further reminding after

22
initial recall. A word is considered to be in LTS even if
it is forgotten after this criterion is met (this case is
seen as a retrieval failure as opposed to a storage
failure). Thus all subsequent recalls of words meeting the
LTS criterion are defined as LTR. CLTR is defined as the
number of words that are eeneietentiy recalled from LTS on
at least the last three trials of the test without further
reminding. A sum of recall score is also computed by adding
the STR and LTR scores on each trial. A more detailed
explanation of the Selective Reminding Test can be found
elsewhere (SRT: Buschke, 1973: Buschke 8 Fuld, 1974).

Masur, et al. (1989) found that subjects in the early
stages of dementia scored significantly worse on every
component of the SRT than normal subjects. The scores for
sum of recall (STR plus LTR), long term retrieval (LTR), and
consistent long term retrieval (CLTR) seemed to be the best
at predicting whether a subject was in the demented group or
the normal group and thus can be seen as having the lowest
risk of misclassification of the demented elderly. The
authors went on to suggest that it is long term memory which
is most impaired by dementia, and may be the important
distinguishing memory characteristic among demented and
normal elderly subjects. Not only were LTR and CLTR found
to be predictive, but it was also noted in their analyses

that STR alone had no significant predictive value. Thus

23
the significance of the sum of recall score was attributed
to the influence of the LTR component.

In another study, the SRT was found also to be
sensitive enough to distinguish between demented and
depressed persons. Hart, Kwentus, Hamer, and Taylor (1987)
reported that patients with mild dementia recalled a much
lower proportion of unreminded words from one trial to the
next than did depressed patients or control subjects. They
found that all of the measures obtained from the SRT
distinguished between depressed and demented patients, but
that the measure for long—term retrieval (LTR) seemed to be
the most significant in this study. They found that the
depressed patients did better overall than the demented
patients, but not as well as the control subjects.

The SRT has also provided evidence that difficulties
with memory, especially long term memory, may begin as much
as 2 years before any social or cognitive change warrants a
diagnosis of dementia, and has been cited as very useful as
a marker for the early stages of dementia (Masur, Fuld,
Blau, Crystal, 8 Aronson, 1990: Smith, et al., 1991). In a
longitudinal study, Masur, et al. (1990) found that the SRT
can be used to determine whether patients may be in a
"preclinical phase” of dementia. They found that the sum of
recall score was one of the best at contributing to the
accuracy of prospective detection of dementia. They

contended that the long term memory processes, both LTS and

24

LTR, are likely more sensitive to disruption than the STR
process in the demented. This is supported by an earlier
finding that higher sum of recall scores achieved by
demented patients actually indicate failure to put items
into long term memory (Masur, et al., 1989).
MEASURES OF PSYCHOLOGICAL SYMPTOMS AND BEHAVIOR

There are a variety of ways in which psychological
symptoms and behavior are evaluated in studies of dementia
and depression. One of the more general arguments among
these studies seems to be the appropriateness of objective
vs. subjective ratings for these populations. Miller (1980)
reported that examiner ratings of normal, depressed, and
demented elderly patients showed significantly different
levels of depression among groups: but when subjective
ratings of depression were used, no significant differences
among groups were found. For the most part however, it
appears that when both objective and subjective ratings are
used in a study, the two seem to correlate well (i.e.,
r=.49) for mildly demented groups (Gottlieb et al., 1988:
Knesevich et al., 1983). Another study noted that although
mildly demented patients may "lack insight into many aspects
of their disease," they still seem to be able to describe
their feeling state (Shuttleworth et al., 1987: p. 735). It
is likely that elderly who are in the beginning stages of
dementia are very much aware of their progressive deficits

and the effects these have on their moods.

25

Not enough work has been done to determine whether a
constellation of psychological/behavioral symptoms can be
defined as qualitatively different for demented elderly vs.
depressed elderly (Miller, 1980). However, there do appear
to be self-report scales that might be able to help
differentiate these groups in this way. One of these is the
Brief Symptom Inventory (BSI: Derogatis 8 Melisaratos,
1983). The B81 is a 53 item self report inventory which is
designed to assess the psychological sypmtoms of an
individual. It is an abbreviated version of the Symptom
Checklist-90 (SCL—90: Derogatis, 1977). There are three
global indices of distress unique to the BSI: (1) positive
symptom total (PST) is a count of the symptoms reported
regardless of perceived intensity, (2) positive symptom
distress index (PSDI) yields information about the perceived
intensity of symptoms corrected for symptom total, and (3)
general severity index (GSI) combines information on the
numbers of symptoms and intensity of perceived distress
making it the best single indicator of current distress
levels. In addition to these global indices, through factor
analysis of the 53 items, nine symptom dimensions have been
defined: somatization, obsessive-compulsive, interpersonal
sensitivity, depression, anxiety, hostility, phobic anxiety,
paranoid ideation, and psychoticism. Forty-nine of the 53
items load exclusively on these constructs. The remaining

four items do not load on any particular dimension, but have

26
been left because of they give important vegetative
information. These nine dimensions yield a profile of the
pattern of an individual's behavioral and psychological
symptomatology.

The BSI has been found to have a high degree of
accuracy in classifying rheumatoid arthritis patients into
their appropriate subgroups (Vollhardt, Ackerman, Grayzel, 8
Barland, 1982). The BSI has also been found to have
discriminatory capabilities for levels of drug abuse in
clients of a treatment center, and for distinguishing
anorgasmic women into their appropriate subtypes (Royse 8
Drude, 1984: Derogatis, Schmidt, Fagan, and Wise, 1989). If
the BSI can effectively distinguish such varied groups as
these, it is possible that it would be an effective
discriminator for demented and depressed ("pseudodemented")
elderly. Also, unlike some other self-report inventories,
the BSI has norms specifically for the elderly population
(Hale, Cochran, 8 Hedgepeth, 1984).

In conclusion, it is clear that there has been a great
deal of work done in the past ten to fifteen years in the
field of dementia and depression in the elderly. However,
it also clear that there are aspects which have received
much less attention, but which have the potential to add
much to the current pool of knowledge. An important issue
which deserves more attention is the study of possible

psychological/behavioral concommitants to dementia and

27
"pseudodementia" as a means of differentially diagnosing
between these conditions.

The present study plans to compare three groups of able
elderly (i.e., those who are living independently in the
community): (1) individuals who appear to be in the early
stages of a dementing process without significant
depression, (2) individuals who appear to be mildly
depressed without significant evidence of dementia, and (3)
normal controls. These three groups will be compared on the
basis of memory function, perceived memory/cognitive
function, and psychological/behavioral functioning. It is
expected that examination of the pattern(s) of
psychological/behavioral symptoms for dementia and
depression will significantly add to the accuracy of

differential diagnosis of these disorders.

HYPOTHESES
The mildly demented elderly and the normal control
elderly will have significantly fewer memory complaints
than the mildly depressed elderly. A one way analysis
of covariance will be used to assess any differences
among these groups. Age and education will be used as
the covariates.
The mildly depressed elderly will report significantly
higher levels of global psychological/behavioral
symptomatology than will the mildly demented elderly
and the normal controls. A one way analysis of
covariance will be performed to determine any
differences using age and education as covariates
against the three groups. It is expected that the
mildly depressed elderly will differ more from the
mildly demented and the normal control in terms of the
intensity of perceived distress than in terms of the
total number of symptoms reported or the current level
of distress. An eta squared will be calculated for
these 3 variables to analyze the effect sizes of the F
ratios.
The pattern of psychological/behavioral symptom report
will also be significantly different between the mildly
depressed and mildly demented groups. It is expected
that the mildly depressed will score significantly

higher than the mildly demented and normal controls on

28

29
the dimensions of depression, somatization,
interpersonal sensitivity, and anxiety while the mildly
demented will score significantly higher than the
mildly depressed and normal controls on the dimension
of obsessive-compulsive and hostility. A one way
analysis of covariance will be calculated using age and
education as covariates against the three groups.
On actual memory functioning, here again there will be
significant differences between the control, demented,
and depressed groups. It is expected that the mildly
demented will do poorer on all measures of memory than
the control group and the mildly depressed group. The
depressed group will do significantly poorer than the
control group on the short term memory measures. A one
way analysis of covariance will be calculated to
examine any differences using age and education as
covariates against the three groups. Moreover, the
mildly demented group and the mildy depressed group
will differ more on total recall and long term memory
measures than on short term memory measures. An eta
squared will be calculated to determine the effect
sizes of the F ratios.
While information about perceived memory function,
memory function, and psychological/behavioral symptoms
will be able to discriminate fairly well between the

groups, it is expected that memory function and

30
psychological/behavioral symptoms will be the best
predictors of group status. A discriminant analysis
will be performed using group membership as the
criterion variable and the measures perceived memory
function, memory function, and psychological/behavioral

symptoms as predictor variables.

METHODS

ect

Data was obtained from subjects participating in the
Coping with Aging project (CWA) which is an ongoing research
effort being conducted through the Michigan State University
Psychological Clinic under the direction of Dr. Norman
Abeles. The CWA project focuses on the current mood and
memory functioning of the able elderly in the Lansing
metropolitan area. By participating in the project,
participants are given opportunities for receiving free
feedback on their current functioning as well as taking part
in mood and memory workshops with other older individuals.
Subjects are recruited for this project through newspaper
advertisements, site visit presentations at senior citizen
independent housing, and posted flyers. Flyers were placed
at grocery stores, churches, community centers, and at
locations on Michigan State's campus.

In this study, data from 146 subjects were used for
analysis. Participants ranged in age from 51 to 92 years.
The mean age was found to be 70.34 years old. There were
113 females and 33 males. With respect to religious
preferences, the majority of the sample was Protestant, 102
subjects (69.9%). The rest of the sample broke down as
follows: Catholic, 24 subjects (16.4%), Jewish, 1 subject
(0.7%), other, 6 subjects (4.1%), and none, 13 subjects
(8.9%). The mildly demented group consisted of 52 subjects,

31

32
the mildly depressed group 24 subjects, and the normal
control group 70 subjects.
HQBQQEQQ
MULTILEVEL ASSESSMENT INSTRUMENT (MAI)
(Lawton, M.P., Moss, M., Fulcomer, M., Kleban, M.H., 1982)

The MAI is an assessment instrument capable of
measuring the well-being of the aged in the areas of
behavioral competence, psychological well-being, and
perceived environmental quality. For the purposes of this
study, only the section regarding memory complaints was
utilized.

HAMILTON RATING SCALE FOR DEPRESSION (HRS-D)
(Hamilton, 1960)

This interviewer-rated depression scale has been widely
used for years, and has been demonstrated to have good
internal consistency (.84) and interrater reliability (.85)
(Hamilton, 1960: Lyons, et al., 1989). Because the HRS-D
has a relatively high number of somatic items, there could
be concern in its utility in screening for depression among
the elderly. However, Yesavage, et al. reported that the
somatic items "generally correlated less strongly..." with
the total score ”..than the items measuring loss of
interests, depressed mood and anxiety” (1983, p. 48). It
might be that in a relatively healthy elderly population
somatic items will not overpower the results of a depression

screening measure. A score of 11 or greater on the HRS-D is

33
seen as evidence of depression (Yesavage, Brink, Rose, Lum,
Huang, Adey, 8 Leirer, 1983).
GERIATRIC DEPRESSION SCALE (GDS)
(Yesavage, et al., 1983)

This 30 item self-report depression scale was designed
specifically for use with the elderly. It has been found
that the 608 correlates well with interviewer-rated measures
of depression (e.g., the HRS-D: r=.83) as well as with the
Research Diagnostic Criteria for depression, r=.82
(Yesavage, et al., 1983). In this study, it was found to
correlate fairly well with the HRS-D (r=.68, p$.01).
Internal consistency as measured by coefficient alpha has
been found to range from .88 to .93 (Lyons, Strain, Hammer,
Ackerman, 8 Fulop, 1989). A score of 11 or greater on the
GDS is considered to be indicative of depression (Yesavage,
et al., 1933).

BRIEF SYMPTOM INVENTORY (BSI)
(Derogatis 8 Melisaratos, 1983)

This self report inventory is a 53-item abbreviated
version of the SCL-90 (Derogatis, 1977). It is comprised of
nine dimensions of symptomatology and three global distress
indices. The nine dimensions are somatization, obsessive-
compulsive, interpersonal sensitivity, depression, anxiety,
hostility, phobic anxiety, paranoid ideation, and
psychoticism. Norms for the elderly obtained from Hale, et

a1. (1984) will be used as the baseline on these dimensions.

34
The nine dimensions have been found to have internal
consistency ranging from .71 (psychoticism) to .85
(depression). The global indices of distress on the BSI
(general severity index:GSI, positive symptom distress
index:PSDI, and positive symptom total:PST) were also used.
The 681 is calculated by summing the scores and dividing by
53. The PST is calculated by summing the number of items
which were endorsed. The PSDI is calculated by dividing the
sum of the 53 items by the PST. The normal controls are
expected to score from 0 to .61 on the 681, from .88 to 1.70
on the PSDI and from 2.16 to 20.74 on the PST. Test-retest
reliability has been found to be .87 for PSDI and .90 for
both GSI and PST (Derogatis 8 Melisaratos, 1983).
SELECTIVE REMINDING TEST (SRT)
(Buschke, 1973)

The SRT is a measure of verbal learning and memory
processing. It consists of a list of 12 unrelated words
which the subject is asked to recall. After each trial, the
subject is reminded only of the words which were forgotten,
and is asked to then repeat the entire list if possible.

The test is completed when the subject recalls the list
twice consecutively or when six trials have been attempted.
The SRT has been shown to be an effective measure of a
variety of memory functions including total recall,
cumulative long term storage (CLTS), storage estimate, long

term retrieval (LTR), consistent long term retrieval (CLTR),

35
consistent retrieval (CR) and retrieval from short-term
storage (STR), and has been used to distinguish between
persons with and without dementia (Masur, et al., 1989:
Masur, et al., 1990). Total recall is calculated by summing
the number of words that are recalled for each trial. CLTS
is the sum across trials of the number of words that have
been recalled e; leeet twice in a row (not necessarily
consistently after that) without reminding. Storage
estimate is the sum of the number words occurring in the
lee; trial which have been recalled e; ieeet twice in a row
previously, but not necessarily consistently. Both CLTS and
storage estimate are measures of long term storage. LTR is
the sum of the number of words eetgeiiy recalled on each
trial without reminding. CLTR occurs when a word has been
recalled for a; ieeet the last three trials without
interruption. The number of words that meet this criterion
are totalled for each trial and then summed across trials.
Consistent retrieval (CR) is a measure of the trial to trial
consistency with which a subject recalls the words. It is
scored by summing the number of words recalled on two
consecutive trials for each trial beginning with trial 2.
Short term recall (STR) is measured by the sum of the number
of words that were recalled only with reminding on each
trial. Test-retest reliablity for these variables has been
found to range from .84 to .92 (Masur, et al., 1989). Norms

for the elderly on the SRT obtained from Masur, et al.

36
(1989) will be used as baselines for these memory variables.
SENILE DEMENTIA ALZHEIMER'S TYPE BATTERY (SDAT BATTERY)
(Storandt et al., 1984)

This is a brief battery of four psychological tests
used to assess memory and cognitive functioning. Although
it does not have an official name, it will be referred to in
this study as the SDAT Battery. It consists of the logical
memory and mental control sections of the Wechsler Memory
Scale (Wechsler, 1945), Form A of Trailmaking, and word
fluency for the letters 3 and p. Discriminant analyses were
performed to obtain unstandardized coefficients for each of
these tests to be used in a canonical function. These
coefficients are -0.445 for logical memory, -0.066 for
Trailmaking, -0.036 for word fluency, and +0.130 for mental
control. A constant of +3.588 is used in the computation of
the canonical score. Multiplying a person's raw score by
the coefficient for each test, and then adding the products
with the constant will result in a canonical score. Persons
with canonical scores of 0 or greater are considered to be
experiencing a dementing process. This measure has been
found to accurately classify normal controls and persons
with dementia at a rate of 96%.

RIQQQQEIQI

Each person who volunteered to participate in the

research was assigned to an advanced graduate student for

testing. Testing was conducted either at the MSU

37
Psychological Clinic or in the individual's home, whichever
was preferred by the participant. The test battery took
approximately 1 1/2 hours to administer, and included a
structured interview, memory assessment and mental health
assessment questionnaires. The participants were also asked
to sign a general consent form which informed them of their
right to withdraw from the project at any time.

Each person also received feedback on the results of
his/her testing either by phone or in person. Feedback was
focused on the individual's memory function. However, if
clinical levels of depression were found, this was
communicated and appropriate referrals were made. If
significant memory problems were found, a similar procedure
was followed. All subjects, regardless of testing results,
were given the opportunity to participate in a seven session
workshop which focused on improving current memory and mood
status and/or preventing future difficulties. These
workshops were led by two advanced graduate students, and
are typically composed of six to eight research
participants. Participation in the workshops was also
purely voluntary. All subjects in the Coping With Aging
Project are tested three times at approximately 2 month
intervals. The first testing always occurs prior to any
participation in a workshop. The current study uses data
strictly from this first testing, but not all subjects went

on to join a workshop.

38

In this study, subjects were assigned to one of the
three groups under study based on two criteria: evidence of
dementia (SDAT Battery score) and evidence of depression
(Geriatric Depression Scale score and Hamilton Rating Scale
for Depression score). Subjects assigned to the mildly
demented group had to obtain a canonical score 0 or greater
on the SDAT Battery. Subjects assigned to the mildly
depressed group had to obtain a canonical score of less than
0 on the SDAT Battery and score 11 or greater on either the
GDS or the Hamilton scales for depression. Because the
literature is equivocal as to the the value of self-report
vs. objective reporting for depression, it was decided to
include a subject in the depressed group if a clinically
significant score was obtained on either measure. Seventeen
subjects obtained clinically significant scores on both the
dementia and the depression measures, but this was too small
a sample to form a fourth group (i.e., demented and
depressed). These participants were placed in the mildly
demented group since the literature has suggested that poor
overall performance on objective testing of cognitive
function, such as the SDAT Battery, is more indicative of a
dementing process rather than primary depression (e.g.,
Miller, 1980). Subjects who obtained a canonical score of
less than 0 end scored 10 or less on DELL depression

measures were assigned to the normal control group.

RESULTS

Hypothesis #1 stated that the mildly demented elderly
and the normal control elderly would have significantly
fewer memory complaints than the mildly depressed elderly.
Performance of a oneway analysis of variance (ANOVA) and
follow-up t-tests gave partial support to this hypothesis.
It was found through the ANOVA that there were significant
main effects based on group status for this variable.
However, t-tests showed no significant difference between
the mildly demented group and the mildly depressed group
(see Appendix A: Table 1 for t-test data) and no difference
between the mildly demented and the control groups. The
depressed group was found to have significantly more memory
complaints than the control group, however. A cut—off level
of .01 was used in all t-tests to guard against the effects
of statistical artifacts. There were no significant effects
in the ANOVA for either of the covariates, age and
education.

Hypothesis #2 stated that the depressed participants
would report significantly higher intensity on global
symptoms than would either the demented or control groups as
measured by the Brief Symptom Inventory. Specifically, it
was expected that the depressed group would differ from the
demented and control groups more in terms of intensity of
perceived distress (PSDI) pattern than in terms of total

number of symptoms (PST) or general level of current

39

4O

distress (681). This hypothesis, however, was not
supported. Significant main effects for group were found in
the ANOVA for all three global indices (see Appendix A:
Table 2). However, no significant differences were found
between the mildly demented and the mildly depressed groups
on the pure intensity measure (see Appendix A: Table 3 for
means and standard deviations, and Appendix A: Table 4 for
t-test data). Significant differences between these two
groups gene found though for the other two global indices
with the depressed group scoring significantly higher than
the demented group on each index. Significant differences
were also found between the depressed and the control groups
for all three global indices. Significant differences were
found, though not hypothesized, between the demented and the
control groups on only the PSDI. In addition, on all three
variables there was a significant effect for education (see
Appendix A: Table 5). ‘Education was found to be negatively
correlated to these three global indices of distress (see
Appendix A: Table 5). There was no significant effect found
for age on any of these variables. An eta2 was also
calculated to determine which variable appeared to explain
the greatest amount of variance. The following effect sizes
were found: PSDI=.13, PST=.19, and GSI=.15.

Hypothesis #3 stated that the pattern of symptom report
would also be significantly different between the mildly

depressed and the mildly demented groups. It was expected

41
that the depressed subjects would score significantly higher
than the demented and control participants on the Brief
Symptom Inventory dimensions of depression, somatization,
interpersonal sensitivity, and anxiety while the demented
would score significantly higher than the depressed and
control on the dimensions of obsessive-compulsivity and
hostility. In the ANOVA all symptom dimensions exeept
hostility showed significant main effects for group status
(see Appendix A: Table 2). T-tests (see Appendix A: Table 3
for means and standard deviations and Appendix A: Table 6
for t-test data) revealed significant differences between
the mildly depressed and the mildly demented groups only on
the dimensions of depression, paranoid ideation, and
psychoticism with the depressed group scoring higher than
the demented group on each dimension. The depressed and the
control subjects were found to significantly differ on the
eight symptom dimensions and the additional items factor.
Significant differences between the demented group and the
control group were found only on the dimensions of
somatization and phobic anxiety. When compared with the
norms developed by Hale, et al. (1984), it was found that
the depressed group scored above normal limits on only two
of the dimensions, obsessive-compulsive and paranoid
ideation. The demented and control groups were within
normal limits on all dimensions. With respect to the

covariates, it was found that age was not a significant

42

factor for any of the symptom dimensions. Education,
however, was found to have a significant effect on the
dimensions of somatization, anxiety, and paranoid ideation
(see Appendix A: Table 5). A significant negative
correlation was found between education and somatization,
and education and anxiety. A negative correlation was also
found to exist between education and paranoid ideation, but
this did not reach significance (see Appendix A: Table 5).

Hypothesis #4 stated that there would be significant
differences between the three groups on actual memory
functioning as measured by the Selective Reminding Test. It
was expected that the mildly demented group would do poorer
on all measures of memory than the mildly depressed and the
control groups. This was largely supported with one
exception (short term memory). Significant main effects for
group were found in the ANOVA for all measures of memory
except short term recall (see Appendix A: Table 7). T-tests
revealed that the demented participants did indeed perform
significantly worse than the other groups on all memory
measures that showed main effects (see Appendix A: Table 9
for t-test data). It was also expected that the depressed
group would do significantly poorer than the control group
with respect to the short term memory measure. However,
there were no significant differences found between the
depressed and the control groups on any of the measures of

memory. In addition, it was expected that the demented and

43

the depressed groups would differ most on the total recall
and the long term memory measures rather than on the short
term memory measure. This was certainly found to be true
since short term recall was not found to differentiate
significantly between any of the groups. Eta2 was
calculated for each of the measures to analyze the
significance of the effect sizes. Total recall was found to
have the largest effect size (eta2=.31), and short term
recall the smallest (eta§=.05). The rest were as follows:
long term retrieval (LTR) - eta§=.28, storage estimate -
eta2=.29, cumulative long term storage (CLTS) - eta2=.26,
consistent long term retrieval (CLTR) - eta§=.27, and
consistent retrieval (CR) - eta§=.28. It was also found
that DEED covariates, age and education, had significant
effects on all memory measures except for short term recall
where only age was significant (see Appendix A: Table 10).
Age was found to be negatively correlated to total recall,
LTR, storage estimate, CLTS, CLTR, and CR. Age was
significantly peeieiveiy correlated with short term recall.
Education was found to be positively correlated to total
recall, storage estimate, CLTS, CLTR, and CR (see Appendix
A: Table 10).

Hypothesis #5 stated that while information about
perceived memory function, actual memory function and
reported symptomatology would discriminate between groups,

it was expected that memory function and

44
psychological/behavioral symptoms would be the best
predictors of group status. This hypothesis was supported.
Discriminant function analyses were performed using the nine
symptom dimensions, six of the seven memory measures, and
the perceived memory function variable in one analysis, and
the three global symptom indices in another analysis. The
short term recall measure was left out of the equation
entirely for two reasons. The first was because it did not
produce significant main effects for group. The second
reason is that it is derived by subtracting the long term
retrieval score from the total recall score. Including an
item that is linearly related to other items would make
interpreting its contribution to a discriminant function
analysis impossible. It is for this same reason that the
three global indices were placed in a separate analysis -
they are all linearly related to the nine symptom dimensions
so could not be included in the same analysis.

The results of analysis #1 revealed two functions were
needed to account for the variance in group membership. The
first function in analysis #1 contrasted the mildly demented
and the depressed groups with the control group as indicated
by the group centroids: -1.09, -0.41, and 0.83 respectively.
The second function in analysis #1 contrasted the depressed
group with the demented and control groups as indicated by
the group centroids: 1.79, -0.53, and -0.19 respectively.

It was found that the memory measures were significantly

45
better predictors of group status for the mildly demented
subjects as seen on function #1 (see Appendix A: Table 11).
Hostility was found to be a significant predictor of
demented status, but it cannot be interpreted with any
confidence since it was not shown to have significant main
effects for group in the ANOVA. Interpersonal sensitivity
was found to have significant main effects for group, and
was found to be a significant predictor of demented status.
However, it was the least correlated of all the predictors.
In addition, examination of the cross-loadings between the
functions suggests that these two BSI dimensions were much
less cleanly loaded on either function than were the memory
indices. It is likely that these two dimensions loaded onto
function #1 as a result of a collinearity problem within the
BSI. Hostility and interpersonal sensitivity were also the
only predictors found to be negatively correlated. Again,
it must be noted that both the demented and the depressed
groups were within normal limits on these two dimensions.
With respect to function #2, it was found that the remaining
seven symptom dimensions, the BSI additional items plus
memory complaints were significant predictors of group
status for the mildly depressed subjects (see Appendix A:
Table 11). All predictors were positively correlated.
Examination of the cross-loadings of the BSI dimensions
within this function revealed that depression, anxiety,

psychoticism, and the additional items loaded the most

46
cleanly and were therefore the best of these predictors.
Memory complaints was not found to load cleanly on either
function and was fairly weak as a predictor on function #2.
This lends support to the hypothesis.

The results of analysis #2 found that among the global
indices on the Brief Symptom Inventory (BSI) again two
functions were needed to account for the variance in group
status. The first function contrasted the depressed group
with the demented and the control groups as indicated by the
group centroids: 1.14, 0.29, and —0.53 respectively. The
second function contrasted the demented group with the
depressed and the control groups as indicated by the group
centroids: 0.23, -0.24, and -0.08 respectively. It was
found that within function #1 the positive symptom total
(PST) and the general severity index (GSI) were significant
predictors of depressed status, and within function #2 the
pure symptom distress index (PSDI) was a significant
predictor of demented status (see Appendix A: Table 12).
Again, all predictors were positively correlated. Of the
three variables, the PST and the 681 loaded more cleanly on
their function than did the PSDI on its function.

It was also found that, of the two analyses, the groups
were most accurately classified overall when symptom
dimensions, memory scores, and memory complaints were taken
together (analysis #1) as predictors as opposed to simply

using the BSI global indices (see Appendix A: Table 13).

47
For both analyses, the probability that members would be
placed in their correct group by chance was calculated at
39%. With respect to classification for the individual
groups, it was found that there was not much difference for
either the depressed or control groups between the analyses.
However, the demented group was accurately classified at a
much higher rate in analysis #1 than in analysis #2 (70% vs.
29%).

There is one issue which must be mentioned with respect
to the discriminant function analysis. In examining the
standardized canonical discriminant function coefficients,
it was found that three of the Selective Reminding Test
(SRT) memory scores (long term retrieval, cumulative long
term storage, and consistent retrieval) had coefficients
well above 1.00, and one of the BSI global indices, PSDI,
had a coefficient just slightly over 1.00 (see Table 14).
The PSDI's is very close to 1.00 and is due to its being
mathematically derived using one of the other indices.
Therefore the coefficient is a statistical artifact and does
not pose any real interpretive problem. However, there is
some concern regarding the psychometric properties of the
SRT. These scores are not mathematically derived from each
other or from the other SRT indices but are instead various
manipulations of the same data set. It eennet be said that
they are linearly related, and that the collinearity is

therefore a statistical artifact. The collinearity is a

48
problem that appears to be inherent in this instrument
making interpretation of these particular indices difficult.
However, these indices are all intended to be measures of
long term memory and do follow the same pattern within the
discriminant analysis so it would seem that some general
conclusions along these lines could be discussed even though
individual interpretation of these indices might be
inappropriate.

Although not hypothesized, some interesting results
were found with respect to the characteristics of the
groups. The groups were found to be significantly different
from each other based on the selection criteria (see
Appendix A: Table 15). However, it was found that the means
for the depressed group on the Geriatric Depression Scale
and the Hamilton Rating Scale for Depression were just
barely eiinieeiiy significant (see Appendix A: Table 15).
T-tests on the covariates revealed that the demented group
was significantly older than either of the other two groups,
but no difference in age was found between the depressed and
control groups (see Appendix A: Table 16). The modal age
for the entire sample was sixty-five. With respect to
education, it was found that the depressed and the control
groups had significantly higher levels of schooling than did
the demented group (see Appendix A: Table 16). The
depressed and control participants did not differ from each

other on this variable (using p$.01 as the cut-off). In

49
addition, in order to determine whether there was an
equivalent level of severity between the demented and the
depressed groups, z-scores were computed for the group means
of the four SDAT Battery subtests (the mean of these 2-
scores being used for comparison) within the demented group.
z-scores were then computed for the group means of the
Geriatric Depression Scale and the Hamilton Rating Scale for
Depression (again using the mean of these two z-scores)
within the depressed group. It was found that both groups
fell within the mildly dysfunctional range: that is, the
demented were mildly demented and the depressed mildly
depressed (z=1.20 and z=1.60 respectively).

Again, although not hypothesized, it was found that
there were some interesting correlations within and between
the variables. Using p5.01 as the cut-off level, it was
found that the symptom dimensions on the Brief Symptom
Inventory are all highly correlated with each other ranging
from r=.25 to r=.73 (see Appendix A: Table 17). The indices
on the Selective Reminding Test were found also to be highly
correlated (see Appendix A: Table 18). Leaving out the
short term recall score, the correlations for the long term
memory indices ranged from r=.87 to r=.99. The short term
recall score correlated at r=-.42 to r=-.70. There were
significant positive correlations for memory complaints with
the BSI dimensions of obsessive-compulsivity, depression,

anxiety, hostility, phobic anxiety, psychoticism, and the

50
additional items as well as with the SRT index for short
term recall. It was found that memory complaints was
negatively correlated with all of the rest of the SRT

indices (see Appendix A: Table 19).

DISCUSSION

The purpose of this research was to determine whether
any specific qualitative differences might exist between
three groups of able elderly: mildly demented, mildly
depressed, or normal control (no signs of either dementia or
depression). Participation in the research was purely
voluntary and self-selecting. Anyone who was 55 years or so
of age or older and lived independently in the community was
eligible. Subjects were compared based on perceived memory
functioning, actual memory functioning, and reported
psychological/behavioral functioning.
21I211!!!.!2!2£!.IH£2£12B13§

In this study, mildly demented participants were not
different from either the mildly depressed or the control
subjects with respect to number of memory complaints. It
was expected that the mildly depressed would have
significantly higher numbers of complaints than the
demented, but this was not the case. The mildly depressed
did complain much more often about their memory funtions
than did the control subjects which was expected. While
other studies have noted that memory deficits appear to be
particularly accentuated by elderly depressed individuals
(e.g., Gershon 8 Herman, 1982: Esser 8 Vitaliano, 1988:
Fopma-Loy, 1986), this study only found this to be true when
compared to normal controls but not with the mildly

demented. These findings seem to be in line with Miller's

51

52
(1980) finding that "aged individuals with clinical
depression complained about memory ee mnen ee those with
altered brain function...” A possible explanation for the
current findings, however, is that the depressed in this
sample suffered from only mild dysfunction. Other studies
have used clinical populations, rather than community
samples, where the mean scores on the Hamilton Rating Scale
for Depression have been reported at 16.6 for the mildly
depressed compared to 11.46 for this sample (Miller, 1980).
It should be noted that scores on the Hamilton may have been
low because this instrument was designed for use with a
clinical population, and may not be the most appropriate
measure of depression for a community sample. Thus, it is
difficult to know whether more severe levels of depression
in this group would have increased memory complaints to the
point of differentiating it from the demented group. The
presence of some depressed eng demented subjects in the
demented group might also act to cloud potential differences
between the comparison groups. This is less likely though
due to the lack of difference between the demented and the
control groups on this variable.

The finding that the depressed have significantly
higher numbers of memory complaints than the controls poses
a potential problem for the recently debated "diagnostic"
category of age-associated memory impairment (AAMI: Crook,

et al., 1986). This category was developed to help deal

53

with the difficulties involved in distinguishing memory
changes due to early dementia from those due to normal
aging. The criteria which define this category were
designed to describe older people who have no evidence of
neurological impairment, but have complaints about their
memory function and show some changes in memory that could
be functionally significant (Smith, et al., 1991).
Unfortunately, the evidence here suggests that increase in
memory complaints is strongly related to the presence of
depression, and thus might confound AAMI as a clear
diagnostic category for normal aging. Thus, it might be
wise for practitioners to not only look for evidence of
neurological damage in older individuals, but also for
depression, before assigning AAMI as a diagnosis.

t c v o c

The psychological/behavioral functioning of the three
groups was examined on two levels: global and symptom-
specific. At the global level, distress was measured by
number of symptoms, intensity of symptoms, and a general
severity index which combined both. It was found that
mildly depressed participants obtained higher scores than
the mildly demented on only two of the indices (number of
symptoms and general severity) while obtaining higher scores
‘than the controls on all three. The mildly depressed and
‘the mildly demented did not differ based on intensity of

symptoms. This could have occurred because both groups were

54
of equal severity (both in the mild range according to
previously reported z-scores). In addition, intensity of
symptoms was the only index that differentiated the demented
from the controls. Thus this index was the only one which
could distinguish the two "dysfunctional” groups from the
normal control group. One would expect based on the
selection criteria that the depressed group would score
significantly higher on global psychological/behavioral
symptomatology than the other groups, but the intensity
index did not consistently follow the same trend as the
other two global indices. This could make the intensity
index on the Brief Symptom Inventory a useful method in
future studies for selecting groups based on severity of
dysfunction (e.g., if studying mild, moderate, and severe
depression in the elderly).

Another interesting finding related to the global
indices of symptomatology was the distribution of effect
sizes. It appeared that the number of symptoms had the
largest effect size of the three in that it seems to account
for a larger amount of the variance than the others do. The
intensity index appeared to have the smallest effect size
according to the bivariate results which is striking since
it seemed to be the most powerful with respect to
differentiating among the three groups in the discriminant
analysis.

At the symptom-specific level, it was found that the

55
depressed scored significantly higher than the controls on
all dimensions which showed significant main effects for
group. This of course is not surprising due to the
selection criteria. The depressed reported significantly
higher symptoms than the demented on only three dimensions:
depression, paranoid ideation, and psychoticism. Upon
closer examination of each factor's items, it was noted that
items on the psychoticism dimension could easily fit into a
depressive framework (e.g., ”feeling lonely even when you
are with people,” "the idea that you should be punished for
your sins," and ”never feeling close to another person").
In this study, it was found that the depression and
psychoticism are the most highly correlated of all the
dimensions (r=.73). It was not expected that the depressed
would differ on psychoticism, but this makes more sense
given the underlying relatedness of the items. It was also
not expected that the depressed would be more highly
suspicious than the demented. This is contrary to the
literature which noted that paranoia, as well as obsessional
symptoms, might exist as coping mechanisms in the dementing
process (Rubin, et al., 1987: Merriam, et al., 1988:
Janowsky, 1982). It might be that the dementia found in
this sample is so mild that these symptoms have not become
an issue yet. It was expected that obsessive-compulsivity
and hostility would be the significant dimensions for the

demented, but this was not born out by the data. In fact,

56
the only dimensions on which the demented score highly are
somatization and phobic anxiety on which they are
significantly different from the normal controls. These are
the only areas of overlapping symptoms between the depressed
and demented groups. One could speculate that the seeming
lack of memory complaints for the demented group could be a
function of an increase in somatic symptoms. Rather than
acknowledge any deterioration in memory functioning, the
mildly demented attribute any problems to physical
difficulties or ailments. Examination of the items which
comprise the phobic anxiety factor reveal an underlying
theme of mistrust in one's capacity to take care of oneself
(e.g., ”feeling afraid to travel on buses, subways, or
trains,” "feeling uneasy in crowds," "feeling nervous when
you are left alone"). It might be that the demented
participants are becoming increasingly wary of what have
likely been familiar situations: maybe because they are
vaguely aware that their capacity to handle them is
decreasing. Unfortunately, it is impossible to know whether
this high score on phobic anxiety is a premorbid condition
or not.

It should be noted that of all the symptom dimensions,
hostility was the only one which did not significantly
differentiate between eny of the groups. It could be that
this resulted from a demand characteristic inherent in the

recruitment for the Coping With Aging Project. Since

57

subjects volunteered for this project due to their own
interests in their current mood and memory statuses, and/or
due to their wishes to participate in this research effort,
they might have believed it to be counterproductive to
display or report any hostile feelings or behavior. They
may have been concerned that doing so might curtail their
chances of being accepted into the project even though no
conditions relating to interpersonal functioning were ever
mentioned by project staff. In this context, low frequency
and intensity of reported hostility makes sense for all the
groups.
WW

Based on data from the Selective Reminding Test, it was
found that the demented group performed significantly worse
than either the depressed or the control group on all of the
memory indices except short term recall. There were no
significant performance differences among the groups on this
variable. These findings are similiar to those of Masur, et
al. (1989) who reported that subjects in the early stages of
dementia scored significantly worse on every component of
the SRT than the normal subjects. And although Masur, et
al. (1989) did find short term recall to be a significant
variable, they also found that it was the total recall score
and the long term memory indices that had the best
predictive value as to whether a subject was in the demented

group or not. It should be remembered that total recall is

58
calculated as the sum of long term retrieval and short term
recall. Since short term recall was nonsignificant in the
current study but total recall was, these findings appear to
lend even greater support to the importance of long term
memory indices in the diagnosis of dementia. They also lend
support to Masur, et al.'s (1990) contention that it is long
term memory processes which are more likely sensitive to
disruption than short term in the demented. No significant
differences were found between the depressed group and the
control group on any of the memory indices. This finding is
in contradiction to other studies which have found that
though depressed subjects perform better on memory tasks
than do demented individuals, they still do not perform as
well as control subjects (e.g., Hart, et al., 1987). Again
this could be a result of the mild form of depression that
existed among this study's sample. However, the demented
group was also found to be only mildly demented, and yet
there seemed to be a significant impact on their memory
performances. One explanation might be that depression must
be of at least moderate severity before any evidence of
memory impairment will be seen, while even very mild degrees
of dementia are enough to interfere with memory functioning.
This is supported by findings from a longitudinal study
which found that difficulties with memory, especially long
term memory, may begin as much as two years (the "pre-

clinical phase") before any social or cognitive change might

59

warrant a diagnosis of dementia (Masur, et al., 1990: Smith,
et al., 1991). The findings above related to the relative
lack of major symptomatology for the demented group seem to
bear this out as well. The effect sizes found for the long
term memory indices also supports their importance. These
measures were found to account for the greatest amount of
the variance.
WW

Although not surprising, it was found that overall the
total recall and long term memory indices best predicted for
demented group status, and that the symptom dimensions and
level of memory complaints best predicted for depressed
group status. Of the global indices, it was found that
number of symptoms and general severity best predicted for
depressed group status, and intensity for demented group
status. This too was not surprising since intensity was the
only distinction between demented and normal subjects.
Although there were two symptom dimensions which predicted
for demented group status (hostility and interpersonal
sensitivity), neither one loaded cleanly on either function.
The correlations were probably due to the relatedness among
all the factors within Brief Symptom Inventory. Overall
however, there was a fairly clean breakdown of the variables
which suggests that in the very early (or mild) stages of
these disorders there are clearer distinguishable

differences between them than what might be found in more

60

advanced stages. The issue of collinearity within the
Selective Reminding Test, though it does not condemn the
instrument, does suggest caution in interpreting the indices
much further than the concepts of short term vs. long term
memory.
genersl_Igznns

On several of the analyses, age and education were used
as covariates to determine their impact on the main effects
for group for the dependent variables. With respect to the
global indices of symptomatology, education was found to be
negatively correlated with all of them. This is striking
due to the fact that the depressed group was found to have
significantly more education than the demented group, and
yet still obtained significantly higher scores on the
general severity index and the numbers of symptoms. In
addition, the depressed group reported higher levels of
distress on all three indices than the control, and there
was no difference on education between these two groups. It
would appear then that depression may override the potential
”buffering" effect education might have, and suggests that
perhaps even a little depression in the elderly has powerful
effects. What is not clear is the extent to which this
population may have underreported depression. It would be
interesting to note if any other affective disorders (e.g.,
anxiety disorder) would display the same pattern or if

depression is unique in this way. Education was also found

61

to be negatively correlated with the dimensions of
somatization and anxiety. This might shed some light on the
finding that the demented individuals scored significantly
higher on somatization than the control subjects. The
demented subjects were found to have significantly lower
educational levels than the control group therefore it could
be that the demented may describe and/or attribute more
complex feeling states to physical problems or ailments.
Age, on the other hand, was not found to be a significant
factor for either the global indices or the symptom
dimensions. This is contrary to reports in the literature
which cite that the risk of depression increases with age
(Gershon 8 Herman, 1982). There have also been special
norms developed for the Brief Symptom Inventory especially
for elderly samples because it has been noted that the
elderly tend to obtain higher scores than young adults
(Hale, et al., 1984). However, this study seemed to be
testing the young-old at best, with the modal age being
sixty-five, and therefore this group may not be much
different from the young adults studied in other reports.

With respect to memory functioning, it was found that
education was positively related to all of the memory
indices except short term recall to which it was not related
at all, and age was negatively related to all memory indices
except short term recall to which it was positively related.

This suggests that aging is related to decline in long term

62
memory functions, but to an increase in short term memory
functions. A decrease in attention and concentration
abilities as one ages might explain this decline in long
term memory functioning. It appears that education may tend
to counteract this however. It is possible though that
these correlations may also have occurred because there were
significant differences found between the demented group and
the other two groups on each of the covariates with the
demented group being older and less educated than the
depressed and control groups.

It has been mentioned that among the demented
participants in this sample some also obtained clinically
significant scores for depression. However, the depression
did not appear to be a major contributor considering the
results of the analyses. It has been suggested if the
dementia is very mild (as in this study) then depression may
not become a significant factor (Salzman 8 Gutfreund, 1986).
Reifler, et al., (1982) noted that significant depressive
symptoms were typically more common in the beginning stages
of the dementing process, and significantly less in the more
severely impaired patients. It might be that Reifler, et
al. was referring to the beginning stages of giegneeed
dementia whereas in this study none of the subjects has been
officially diagnosed as demented. Therefore it could be
that at what Masur, et al. (1990) call the "pre-clinical"

phase of dementia, depression has not become a significant

63

part of the clinical picture, but may develop later when the
syndrome progresses enough to warrant a diagnosis. So there
may be a "window of risk" for depression in dementia which
could give diagnosticians an idea of how far along the
disorder is. The issue of which comes first dementia or
depression has been recently discussed in the literature.
Teri and Wagner (1992) suggest that there is evidence on
both sides, and cite studies which support the notion that
later life depression may be a precursor to dementia.
However, they admit that although "the reverse argument is
less well studied", it is "no less compelling” (p.381). It
is this argument, dementia occurring prior to significant
depression, which gains support from the findings here. Of
course, without a longitudinal perspective it is very
difficult to state this more conclusively, and it must be
born in mind that one is not necessarily the precursor to
the other in all situations.
9.2mm

In summary, it would appear that in its mildest forms
dementia and depression in the elderly seem to impact on
very different aspects of individual functioning. The first
aspect affected by a dementing process appears to be long
term memory functioning. However, mild depression appears
to affect solely the psychological/behavioral functioning.
While this finding is by no means startling in the general

population, it is somewhat surprising in an elderly sample

64
where it has been reported in the literature that memory
impairment can be used as a signal for depression (Esser 8
Vitaliano, 1988: Gershon 8 Herman, 1982: Janowsky, 1982).
However, it must again be noted that the sample used in this
study was restricted in range with respect to age of the
participants, likely representing the young-old for the most
part.

Thus it might be easier to distinguish between the
disorders of dementia and depression when each is in the
yery early stages. It is difficult to know what the
depressed in this sample might look like if their
depressions became more severe. Perhaps increased memory
impairment would be evident at that stage making
differential diagnosis more difficult. This makes a
convincing case for "preventive" assessments such as ”mood
and memory check-ups". Such assessments could provide able
elderly with a baseline of their functioning which could
then be reassessed over time. It might give health care
professionals much earlier warning of impending problems, a
better idea of relevant diagnoses, and a head start on
effective intervention, if needed. Several avenues for
future research can be suggested as a result of the findings
in this study. One recommendation would be to examine the
question of which comes first, dementia or depression, posed
by Teri and Wagner (1992). This, of course, would require

more longitudinal studies which are badly needed in this

65

area of research. Another avenue might be to examine the
use of the intensity index on the Brief Symptom Inventory as
a way of "scaling" the severity of depression (or other
affective disorders). It is also apparent from this
research that the impact of age and education needs to be
taken into account in future studies and explored more
carefully. One possible reason that age was not found to be
much of a factor here could be a function of studying the
able elderly who are more likely to be members of the young-
old set. Additionally, although not included in this study,
the issues of gender, health and marital status, and
activity level might be useful variables to address in
future studies. Gender was not explored in this study
because of the overwhelming numbers of females vs. males in
the sample, and the difficulty this might pose in the clear
interpretation of statistical findings. Finally, again
although not a focus of this study, the impact of
medications might be useful to explore when examining the
qualitative differences between dementia and depression.
Quantifying such a variable might prove to be difficult, but
with medications being such a common part of old age in this
country, including it might yield some important findings.

Because there are still many questions to be answered
in this field, it is important that research efforts
continue. With the elderly population increasing at a rapid

pace, the answers will be vital.

APPENDIX

 

APPENDIX

List of Tables

Table 1
- o a
Wings
issuable mm EDP
Memory Complaints
Means 1.44 1.63
SDS 0.50 0.50

MDM: mildly demented group
MDP= mildly depressed group
NC= normal control group

I:Yelns§
Degrees of
13.2;- 8 -Va m
22212212111321
Test 1 -1.49 45
Test 2 3.19 37
Test 3 2.07 102

Test 1= MDM vs. MDP
Test 2= MDP vs. NC
Test 3= MDM vs. NC

66

2-Tailed

p_<_. 14
p5.003

p$.04

Table 2

PSDI

PST

GSI

Depression
Somatization
Interpersonal Sensitivity
Anxiety

Additional Items
Obsessive-Compulsive
Phobic Anxiety
Paranoid Ideation
Psychoticism

67

7.095
15.024
10.548
21.955

9.015

4.020

7.991

8.038

5.529

9.494
11.952
13.763

1°.

2-Tailed

5.001
p$.000
pg.000

5.000
pg.000
p$.02
pg.001
p5.000
ps.005
p$.000
pg.000
p5.000

 

68

Table 3
Y‘tl: 11’. E 8'. ’. PA. .1ti.!:
and.
W
Yériéhlfi MDM
PSDI
Means 1.61
SDS 0.49
PST
Means 16.65
SDS 10.91
GSI
Means 0.54
SDS 0.41
Depression
Means 0.44
SDs 0.52
Somatization
Means 0.54
SDS 0.57
Interpersonal Sensitivity
Means 0.57
SDS 0.71
Anxiety
Means 0.43
SDs 0.47
Obsessive-Compulsive
Means 1.28
SDB 0.87
Phobic Anxiety
Means 0.27
SDS 0.37
Paranoid Ideation
Means 0.43
SDS 0.53
Psychoticism
Means 0.34
SDB 0.41
Additional Items
Means 0.38
SDS 0.49

1.72
0.44

25.71

9.14
0.87
0.43
1.03
0.67

0.77
0.73

0.86
0.78

0.74
0.64

1.64
0.99

0.49
0.65

0.87
0.54

0.73
0.61

0.70
0.54

6].”:

1.34
0.35

12.70

8.20
0.37
0.43
0.26
0.38

0.22
0.33

0.40
0.51

0.31
0.37

0.98
0.73

0.11
0.18

0.29
0.38

0.22
0.29

0.29
0.32

Table 4

Degrees of
Ereegem

50
33
85

53
37
91

43
40
113

T:Iest§_fgr_B§l_§lgbal_Indise§
Zariahle -V n
Probability
PSDI: Test 1 -0.98

Test 2 3.87

Test 3 3.39
PST: Test 1 -3.77

Test 2 6.17

Test 3 2.19
GSI: Test 1 -3.09

Test 2 4.90

Test 3 2.28
Test 1= Demented vs. Depressed
Test 2= Depressed vs. Normal Control
Test 3: Demented vs. Normal Control

2-Tailed

p$.33
pg.000
p$.001

pg.000
pg.000
pg.03

p$.003
p$.000
p$.03

Table 5

GSI

Somatization

Anxiety

Paranoid Ideation

NSz Not Significant

70

2.066
(pg-001)

7.895
(pg-006)

7.362
(p$.008)

17.545
(pg-000)

5.724
(pg-02)

5.455
(pg-02)

(PS-01)

-.21
(p_.05)

-.21
(pg-05)

-031
(PS-01)

-.2o
(pg-05)

—.16
(pg-05)

71

Table 6
I:Te§t§_f2r_B§I_Dimen§igns
Degrees of 2-Tailed

yarishle szalue __Ereedgm_ .2rgbabilitx
Depression

Test 1 -3.83 37 p5.000

Test 2 5.35 28 p_.000

Test 3 2.09 89 p5.04
Somatization

Test 1 -1.34 37 p5.19

Test 2 3.58 26 p5.001

Test 3 3.67 76 5.000
Interpersonal Sensitivity

Test 1 -1.56 41 p5.13

Test 2 2.72 30 p5.01

Test 3 1.48 88 p5.14
Anxiety

Test 1 -2.13 35 p5.04

Test 2 3.13 28 p5.004

Test 3 1.54 93 p5.13
Additional Items

Test 1 -2.47 41 p5.02

Test 2 3.50 29 p5.002

Test 3 1.13 82 p5.26
Obsessive-Compulsive

Test 1 -l.54 40 p5.13

Test 2 2.99 32 p5.005

Test 3 2.00 99 p5.05
Phobic Anxiety

Test 1 -1.58 30 p5.12

Test 2 2.83 24 p5.009

Test 3 2.80 68 p5.007
Paranoid Ideation

Test 1 -3.34 44 p5.002

Test 2 4.83 31 p5.000

Test 3 1.54 88 p5.13
Psychoticism

Test 1 -2.88 33 p5.007

Test 2 4.00 27 p5.000

Test 3 1.86 88 p5.07

Test 1= Demented vs. Depressed
Test 2= Depressed vs. Normal Control
Test 3= Demented vs. Normal Control

72

Table 7
.1 ' ., ,. , - 3 . ._. ., ; .
2-Tailed
yariahle Ezyalse Brebabilitx
Total Recall 11.100 p5.000
LTR 10.015 p5.000
Storage Estimate 10.468 p5.000
CLTS 9.190 p5.000
CLTR 8.942 p5.000

CR 9.696 p5.000

 

Total Recall

Means 24.68

SDs 10.97
Long Term Retrieval

Means 9.00

SDs 10.85
Storage Estimate

Means 2.85

SDs 2.91
CLTS

Means 11.00

SDs 12.32
CLTR

Means 3.73

SDs 8.71
Consistent Retrieval

Means 5.45

SDs 7.61

73

74

Degrees of
Eneedon

Table 9
W
Enable -V n
21053511111
Total Recall
Test 1 -4.23 32
Test 2 -0.94 28
Test 3 -7.25 85
LTR
Test 1 -3.73 25
Test 2 -1.03 28
TEst 3 -7.41 100
Storage Estimate
Test 1 -3.51 27
Test 2 -1.12 26
Test 3 -7.19 92
CLTS
Test 1 -3.54 26
Test 2 -0.94 27
Test 3 -6.93 97
CLTR
Test 1 -3.95 24
Test 2 -0.83 30
Test 3 -7.20 100
CR
Test 1 -3.70 24
Test 2 -0.98 29
Test 3 -7.23 101
Test 1= Demented vs. Depressed
Test 2: Depressed vs. Normal Control
Test 3= Demented vs. Normal Control

2-Tailed

p5.000
p5.36
p5.000

5.001
p5.31
p5.000

p5. 002
p5.27
p5. 000

p5.002
p5.35
p5.000

5.001
p5.4l
p5.000

5.001
p5.33
5.000

Table 10

.011 f :1 91°! ._‘ = oid

§BI_§QQI§§
yariahles
Education
Total Recall
LTR
Storage Estimate
CLTS
CLTR

CR

Age
Total Recall

LTR

STR

Storage Estimate

CLTS

CLTR

CR

75

-V u

8.913
(pg-003)

5.960
(PS-02)

6.061
(pg-02)

4.561
(pg-04)

5.515
(pg-02)

5.709
(95.02)

49.567
(pg-000)

45.102
(pg-000)

10.696
(pg-001)

45.608
(pg-000)

42.272
(pg-000)

43.507
(pg-000)

47.910
(pg-000)

sorrelatign_lrl
+.34
(pg-01)

+.30
(pg-01)

+.30
(PS-01)

+.28
(pg-01)

+.30
(PS-01)

+.30
(pg-01)

-.55
(p_.01)

-.54
(pg-01)

+.30
(PS-01)

-.54
(p_.01)

-.53
(pg-01)

-.53
(p5.01)

-.54
(pg-01)

Table 11

3 = u 9:! 1.1 '1

Total Recall
Storage Estimate
LTR

CR

CLTS

CLTR

Hostility
Interpersonal
Sensitivity

Depression
Somatization
Anxiety
Psychoticism
Additional Items
Phobic Anxiety
Paranoid Ideation
Obsessive-Compulsive
Memory Complaints

76

0.67468
0.65078
0.64837
0.63843
0.61550
0.61458
-0.26863
-0.25597

-0.37318
-0.49106
-0.32640
-0.28464
-0.15881
-0.31750
-0.29294
-0.31205
-0.21184

00000000

0.30702
0.25135
0.25990
0.25015
0.25012
0.27915
0.17560
0.21144

0.58835
0.52318
0.47629
0.39907
0.39669
0.38438
0.36997
0.34863
0.29700

000000000

Table 12

77

0.91067:
0.72092

0.61636

-0.39220
-0.16530

0.73085*

78

Table 13

.»:: ._ o! {‘3_ 1 o - fly. 0

Analyzes
ANALYSIS #1:

0 1a t in! 001

# of Predicted Group Membership
Astual_§r232 gases _HDH__ _HDE__ _n£__
MDM 40 28 6 6
70.0% 15.0% 15.0%
MDP 18 2 14 2
11.1% 77.8% 11.1%
NC 61 11 4 46
18.0% 6.6% 75.4%
Percent of "Grouped" Cases Correctly Classified: 73.95%
ANALYSIS #2:
# of Predicted Group Membership
AQEQQI_§IQEP £Q§2§ _MDM__ _MD£__ _E§__
MDM 51 15 15 21
29.4% 29.4% 41.2%
MDP 24 4 17 3
16.7% 70.8% 12.5%
NC 69 6 11 52
8.7% 15.9% 75.4%
Percent of "Grouped Cases Correctly Classified: 58.33%

Table 14

E! l 3' i : i J E !i : EEI . !

251139.15 WW
Somatization -0.54373 0.36382
Obsessive-Compulsive -0.33879 -0.27073
Interpersonal -0.03021 -0.39935
Sensitivity

Depression -0.08715 0.59138
Anxiety 0.14849 0.31041
Hostility -0.08169 -0.37633
Phobic Anxiety 0.26492 0.34011
Paranoid Ideation -0.36238 0.01164
Additional Items 0.19709 0.16590
LTR 1.65977 1.53709
Total Recall 0.83059 0.64351
Storage Estimate 0.53794 -0.11606
CLTS -l.25459 -0.93399
CLTR 0.39084 0.33608
Consistent Retrieval -1.24447 -0.84025
Memory Complaints 0.10062 0.39945
PSDI 0.43675 1.08109
GSI -0.03077 -0.53100
PST 0.82684 -0.31134

79

Table 15

T-T!l:!li EEJH :.!.

80

WW
Measure Demented Depressed Kernel
mm
GDS
Mean 8.02 14.33 4.29
SD 5.65 5.23 3.08
Hamilton
Mean 6.98 11.46 3.67
SD 4.98 7.25 2.61
SDAT Battery
Mean +1.174 -1.243 -2.090
SD 0.774 1.100 1.979
-V s
Degrees of 2-Tailed
Tzleee I:Yelee _£Ieeeem__
Embsbilitx
GDS
Test 1 -4.75 48 p5.000
Test 2 8.87 29 p5.000
Test 3 4.26 72 p5.000
Hamilton .
Test 1 -2.74 33 p5.01
Test 2 5.15 25 p5.000
Test 3 4.36 72 p5.000
SDAT Battery
Test 1 9.71 34 p5.000
Test 2 2.60 73 p5.01
Test 3 12.57 95 p5.000
Test 1= Demented vs. Depressed
Test 2= Depressed vs. Normal Control
Test 3= Demented vs. Normal Control

81

Table 16
I:Teet_Date_bx_erenn_fer_Age_eng_Eeeeetien
n va 8

yarieble Demented Depressed gentrel
Age

Mean 75.41 65.75 68.21

so 9.96 9.84 8.34
Education

Mean 11.12 13.21 14.36

so 2.56 1.87 3.13

szelnee
Degrees of 2-Tailed

TzTeet Izyelee _Ereeeem__ Erebehilitx
Age

Test 1 3.95 46 p5.000

Test 2 -l.10 35 p5.28

Test 3 4.20 96 p5.000
Education

Test 1 -4.02 60 p5.000

Test 2 -2.15 68 p5.04

Test 3 -6.29 119 p5.000

Test 1= Demented vs. Depressed
Test 2= Depressed vs. Normal Control
Test 3= Demented vs. Normal Control

82
Table 17
92W
Wm

EH 99 I§ 22 AK fl§ 2A 21 2§
Al

SM 1.00 .46 .35 .36 .51 .38 .50 .28 .25
.39

OC .46 1.00 .51 .57 .58 .52 .45 .36 .51
.37
IS .35 .51 1.00 .56 .54 .38 .47 .52 .59
.40
DP .36 .57 .56 1.00 .55 .41 .47 .53 .73
.48
AX .51 .58 .54 .55 1.00 .60 .51 .51 .52
.39
HS .38 .52 .38 .41 .60 1.00 .56 .50 .39
.32
PA .50 .45 .47 .47 .51 .56 1.00 .46 .48
.37
PI .28 .36 .52 .53 .51 .50 .46 1.00 .58
.37
PS .25 .51 .59 .73 .52 .39 .48 .58 1.00
.41
AI .39 .37 .40 .48 .39 .32 .37 .37 .41
1.00
SM= somatization HS= hostility
OC= obsessive-compulsive PA= phobic anxiety
IS= interpersonal sensitivity PI= paranoid ideation
DP= depression PS= psychoticism

= anxiety AI= additional items

All correlations significant at the .01 level.

83

$3 QLIfi QLEB 98

Table 18
eerrelatiens_flithin_§BI_¥arieblee
Seleetixe_3emineing_1eet
TB LIB 513

TR 1.00 .94 -.42 .92
LTR .94 1.00 -.70 .97
STR -.42 -.70 1.00 -.68
SE .92 .97 -.68 1.00
CLTS .93 .99 -.69 .97
CLTR .89 .92 -.62 .87
CR .94 .99 -.69 .94

TR= total recall

storage

LTR: long term retrieval
retrieval

STR= short term recall
SE= storage estimate

.93

.89

.97

.89

.92

-.62

.87

.89

1.00

.96

All correlations significant at the .01 level.

.94

.99

-.69

.94

.97

.96

1.00

CLTS= cumulative long term
CLTR= consistent long term

CR= consistent retrieval

84

Table 19
e '0 V ia 8 RT Var ab 5

W

HQ 13 LIB £13 £3 QLTfi .QLIB 98
SM .17" -.07 -.09 .07 -.09 -.07 -.07 -.09
0c .45“ .02 .01 .01 —.00 .02 .00 .00
IS .11 -.03 -.04 .03 -.01 -.02 -.07 -.06
DP .36“ -.15 -.15 .11 -.13 -.14 -.19* -.18*
AX .22” -.09 -.09 .08 -.10 -.08 -.08 -.11
HS .21“ --.07 -.06 .02 -.05 -.04 -.08 -.08
PA .22” -.20* -.18* .09 -.18 -.16 -.17 -.18*
F: .16 -.02 -.01 .09 .01 .01 -.03 -.02
PS .23” -.12 -.13 .11 -.11 -.11 -.15 -.15
AI .23“ -.09 -.07 .02 -.06 -.06 -.09 -.09
140 1.00 -.26** -.29** .25“ -.26** -.29** -.27**-.29**

' denotes p5.05

MC= memory complaints

TR= total recall
LTR= long term retrieval
STR= short term recall

SE= storage estimate

CLTS= cumulative long term storage
CLTR= consistent long term retrieval
CR= consistent retrieval

wdenotes p5.01

SM= somatization

OC= obsessive-
compulsive

IS= interpersonal
sensitivity

DP= depression

= anxiety

HS= hostility

PA= phobic anxiety

PI= paranoid
ideation

PS= psychoticism

AI= additional items

LIST OF REFERENCES

REFERENCES

American Psychiatric Association (1980). Qiegneeeie_eng
Stetietie_Haneel_ef_nentel_ni§ereere. 3rd ed-. American

Psychiatric Association, Washington, D.C.

Arenberg, D. (1990). Misclassification of "probable senile
dementia-Alzheimer's type" in the Baltimore

Longitudinal study of Aging- leernal_ef_§linieal
Eeieemielegx. 13(1). 105-107-

Buschke H. (1973). Selective reminding for analysis of

memory and learning. ie2rnal_ef_!erbel_Learnins_eng
yerbel_eehexier. 12. 543-550-

Buschke, H. and Fuld, P.A. (1974). Evaluating storage,
retention, and retrieval in disordered memory and

learning. Henreiegy, 25, 1019-1025.

Breen, A.R., Larson, E.B., Reifler, B.V., Vitaliano, P.P.
and Lawrence, G.L. (1984). Cognitive performance and
functional competence in coexisting dementia and
depression. '

Seeietx. 22. 132'137-

Broadbent, D.E. (1984). The maltese cross: A new

simplistic model for memory. The_nehexieral_ene_nrein
Seienees. l. 55-94-

Charness, N. and Bosman, E.A. (1990). Human factors and
design for older adults. In J.E. Birren and K.W. Ane
, Academic Press, Inc., San Diego,
CA, 446-463.

Cohen, D. and Dunner, D. (1980). The assessment of
cognitive dysfunction in dementing illness. In

Pexehenethelegx_in_tne_bsee by Jonathan 0- Cole and
James E. Barrett (Eds.), Raven Press, New York, 123-
135.

Cohen, R., Weingartner, H., Smallberg S., Pickar D. and
Murphy, D. (1982). Effort and cognition in depression.
Arehixee.ef.§enerel_£exehietrx. 32. 593-597.

Coughlan, A. 8 Hollows, S. (1984). Use of memory tests in
differentiating organic disorder from depression.
, 155, 164-167.

Craik, F.I.M. and Lockhart, R.S. (1972). Levels of
processing: A framework for memory research. Jenxnei

ef_Ierhel_Leerning_ene_Behexier. 11. 671-684-

85

“; 1w

86

Crook, T., Bartus, R.T., Ferris, S.H., Whitehouse, P.,
Cohen, G.D., and Gershon, S. (1986). Age associated
memory impairment: Proposed diagnostic criteria and
measures of clinical change-Report of a National
Institute of Mental Health Work Group. Degeiepmeneei

Henrenexehelegx. 2. 261-276.

Davis, .E. and Mumford, S. J. (1984). Cued recall and the
nature of the memory disorder in dementia. fixieien
,155, 383- -386.

Derogatis, L.R. (1977). e C - ua ° o
eQmin1e1re11en_ene_nreeeenres_fer_the_§eL:20

Baltimore, MD: Johns Hopkins University School of
Medicine, Clinical Psychometrics Unit.

Derogatis, L. R. and Melisaratos, N. (1983). The Brief
Symptom Inventory: An introductory report.
Eexehelegieel_nedieine. 12. 595- 605-

Derogatis, L.R., Schmidt, C.W. Fagan, P.J. and Wise, T.N.
(1989). Subtypes of anorgasmia via mathematical

taxonomy- Eexeheeemetiee. 29(2). 166-173-

Eisdorfer, C. and Cohen D. (1980). Diagnostic criteria for
primary neuronal degeneration of the Alzheimer's type.
lenrnal_ef_£emilx_£reetiee. 11. 553- 557-

Esser, S. R. and Vitaliano, P. P. (1988). Depression,
dementia, and social supports. 1n;e;ne§ienei_lenrnei
ef_Ag1ng_ene_nnmen_nexel_pmene. 2_(4). 289- -301-

Folstein, M.F. and McHugh, P. R. (1978). Dementia syndrome
of depression. In R. .Katzman, R. D. Terry, 8 K.L. Bick
(Eds.), ' 3 ° ni D me

Beleted_Diserders_IAging___ll_11. Raven Press. New
York, 87.

Folstein, M.F., Folstein, S.E. and McHugh P.R. (1975).
"Mini-Mental State": A practical method for grading
the cognitive state of patients for the clinician.

lenrnel_ef_£exehietrie_8eseereh. 12. 189-198-

Fopma-Loy, J. (1986). Depression and dementia: Differential

diagnosis. Jonznel ofi geyenoseciai Nuzeing, 21(2), 27-
29.

Gershon, S. and Herman, S. P. (1982). The differential

diagnosis of dementia. leernel_ef_the_Ameriean
seriaeriee_§eeietx. 29(1=SUPp1ement). $58-$66.

Gottlieb, G.L., Gur, R.E. and Gur, R.C. (1988). Reliability

87

of Psychiatric scales in patients with dementia of the

Alzheimer typs- Amerieen_leernel_ef_fiexehietzx.
155(7), 857-860.

Hale, W.D., Cochran, C.D. and Hedgepeth, B.E. (1984). Norms
for the elderly on the Brief Symptom Inventory.

Ieernal_ef_Qeneel11ng_end_elinieel_£§¥ehelegx. 52(2).
321-322.

Hamilton, M. (1960). A rating scale for depression.

lenrnel_ef_neereleg1_end_Henrelegieel_£exehiatrx. 22.
56-61 0

Harris, S. J. and Dowson, J. H. (1982). Recall of a 10-word
list in the assessment of dementia in the elderly.

2ri11eh_lenrnel_ef_£exehietrx.111.524-527.

Hart, R.P., Kwentus, J.A., Hamer, R.M. and Taylor J.R.
(1987). Selective reminding procedure in depression

and dementia. Pexehelegx_end_aging. 2(2). 111-115-

Hasher, L. and Zacks, R.T. (1979). Automatic and effortful
processes in memory. e

Eexeheleg¥1_eeneral. 12§(3). 356-333-

Heyman, A., Wilkinson, W.E., Hurwitz, B.J., Helms, M.J.,
Haynes, C.S., Utley, C.M. and Gwyther, L.P. (1987).
Early on-set Alzheimer's disease: Clinical predictors
of institutionalization and death. fienzeiegy, 11, 980-
984.

Hughes, C.P., Berg, L. Danziger, W.L., Coben L.A. and
Martin, R.L. (1982). A new clinical scale for the

staging of dementia. a1i1ieh_leernel_ef_£exehieerx.
149, 566-572.

Hultsch, D.E. and Dixon, R.A. (1990). Learning and memory
in aging. In J.E. Birren and K.W. Schaie (Eds.),
, Academic Press,
Inc., San Diego, CA, 258-274.

Janowsky, D.S. (1982). Pseudodementia in the elderly:
Differential diagnosis and treatment. lenznei_e£

glinieel_£sxenietrx. 11(9). 19'25-

Jorm, A. F. (1986). Cognitive Deficit in the depressed
elderly: A review of some basic unresolved issues.

 

Jorm, A.F. and Henderson, A.S. (1985). Possible
improvements to the diagnostic criteria for dementia in

88
DSM-III. ' s u a s at , i47, 394-399.

Kay, D.W.K., Henderson, A.S.,Scott, R., Wilson, J., Rickwood
D. and Grayson, D. (1985). Dementia and depression
among the elderly living in the Hobart community.

Pexehelegieel_ued1eine. 15. 771-788-

Kim, K.Y. and Hershey, L.A. (1988). Diagnosis and treatment
of depression in the elderly.

Eexehietrx_1n_nedieine. 12(3). 211-221-

Klein, L.E., Roca, R.P., McArthur, M., Vogelsang, G., Klein,
G.D., Kirby, S.M. and Folstein, M. (1985). Diagnosing
dementia: Uni-variant and multi-variant analysis of
the MSE. c G c ,
11, 483-488.

Knesevich, J.W., Martin, R.L., Berg, L. and Danziger, W.
(1983). Preliminary report on affective symptoms in the
early stages of senile dementia of the Alzheimer's

type. Amer1een_lenrnel_ef_zsxehietrx. 149(2). 233-235-

Kopelman, M.D. (1985). Rates of forgetting in Alzheimer-
type dementia and Korsakoff's syndrome.

Neuronexehelesie. 22(5). 623-638-

Kopelman, M.D. (1986). Clinical tests of memory. fizieien
leernel.ef_£sxehietrx. 112. 517-525-

Kramer, B.A. (1982). Depressive pseudodementia.
, 22(6), 538-544.

Lawton, M.P., Moss, M., Fulcomer, M., Kleban, M.H. (1982).
A research and service oriented multilevel assessment

instrument. lenrnel.ef_eerentelegx. 21(1). 91-99-

Lazarus, L.W., Newton, N., Cohler, B., Lesser, J. and
Schweon, C. (1987). Frequency and presentation of
depressive symptoms in patients with primary
degenerative dementia. J ur o s ,

111(1). 41-45.

Lyons, J.S., Strain, J.J., Hammer, J.S., Ackerman, A.D. and
Fulop, G. (1989). Reliability, validity, and temporal
stability of the geriatric depression scale in

hospitalized elderly. Internetienel_lenrnel_ef
Berenietrx_in_nedieine. 12(2). 203-209-

Masur, D.M., Fuld, P.A., Blau, A.D., Thal, L.J., Levin, H.S.
and Aronson, M.K. (1989). Distinguishing normal and
demented elderly with the Selective Reminding Test.

- 1— 0 9 ._ :10. .10‘ an?! . '_ 008 90 03:1,

89
11(5). 615-630.

Masur, D.M., Fuld, P.A., Blau, A.D., Crystal, H. and
Aronson, M.K. (1990). Predicting development of
dementia in the elderly with the Selective Reminding
Test.

neuronexehelesx. 12(4). 529-533-

McAllister, T. (1983). Overview: Pseudodementia. Anexieen
lenrnel.ef_£exehietrx. 112. 528-533-

McAllister, T. and Price, T. (1982). Severe depressive
pseudodementia with and without dementia. Anerieen

lenrnel_ef_£exehietrx. 122. 626-629-

McLean, S. (1987). Assessing dementia. Part I:
Difficulties, definitions and differential diagnosis.

Anstralien_and_nex_2eelend_lenrnel_ef_ze¥ehietrx. 21.

142-174.

Merriam, A.E., Aronson, M.K., Gaston, P., Wey, S.L. and
Katz, I. (1988). The psychiatric symptoms of
Alzheimer's disease.

Seeietx. 25. 7-12.

Miller, E. (1972). Efficiency of coding and the short-term

memory defect in presenile dementia. EeBIQDEXQDQLQQIB.
19, 133-136.

Miller, E. (1973). Short- and long-term memory in patients
with presenile dementia (Alzheimer's disease).

Pexehelegieel_nedieine. 2. 221-224-

Miller, E. (1975). Impaired recall and the memory

disturbance in presenile dementia. flziiien_lenznei_e£
aeeiel_end_elinieel_£e¥ehel991. 11. 73-79.

Miller, N.E. (1980). The measurement of mood in senile
brain disease: Examiner ratings and self-reports. In
by Jonathan 0. Cole and
James E. Barrett, Raven Press, New York, 97-118.

Nott, P.N. and Fleminger, J.J. (1975). Presenile dementia:

The difficulties of early diagnosis. ne;e_2eyenie§riee
Seendinexiee. 21. 210-217.

Rabins, P.V., Merchant, A. and Nestadt, G. (1984). Criteria
for diagnosing reversible dementia caused by
depression: Validation by 2-year follow-up. fizieien

Journel_ef_zexehietrx. 112. 488-492-
Reifler, B.V., Larson, E. and Hanley, R. (1982).

90

Coexistence of cognitive impairment and depression in

geriatric populations. Amer1een_ieernal_ef_zexehietrx.

122, 623-626.

Royse, D. and Drude, K. (1984). Screening drug abuse
clients with the Brief Symptom Inventory. Ine

Interne11enal_lenrnal_ef_the_Ad§ietiens. 12(8). 849-
857.

Rubin, E.H., Morris, J.C. and Berg, L. (1987). The
progression of personality changes in senile dementia

of the Alzheimer' 8 type. leernal.ef_the_Ameriean
Geriatrie_§eeietx. 22. 721-725.

Salzman, C. and Gutfreund, M. J. (1986). Clinical techniques
and research strategies for studying depression and
memory. In L. W. Poon (Ed. ),

Assessment. 257-257-

Schaie, K.W. and Willis, S.L. (1991). Adult personality and
psychomotor performance: Cross-sectional and

longitudinal analyses. leernal.ef_eerentelegx. 25(6).
p. 275-284.

Shamoian, C.A. (1985). Assessing depression in elderly

patients. Heep11a1_and_eemmnnitx_zexeniatrx. 25(4).

Sher, K.J., Frost, R. and Otto, R. (1983). Cognitive
deficits in compulsive checkers: An exploratory study.
e a , 21. 357-363.

Sher, K. J., Mann, B. and Frost, R. (1984). Cognitive
dysfunction in compulsive checkers: Further

explorations. 2eha212ral_3e_earen__nd_lnerepL .22.
493- 502.

Sher, K.J., Frost, R., Kushner, M., Crews, T. and Alexander,
J. (1989). Memory deficits in compulsive checkers:
Replication and extension in a clinical sample.

Behe21erel_Beeeereh_and_Therepx 21. 55‘ 59-

Shuttleworth, E.C., Huber, S.J. and Paulson, C.W. (1987).
Depression in patients with dementia of Alzheimer type.

Journal.ef_1he_Ha11enel_nedieal_Aeseeietien. 12(7).

733-736.

Siegler, I.C. (1983). Psychological aspects of the Duke
Longitudinal Studies. In K.W. Schaie (Ed.),
Longitudinal_etediee_ef_Adelt_Pexehelegieel
peyeiennen;, The Guilford Press, New York, 136-190.

91

Smith, G., Ivnik, R.J., Petersen, R.C., Malec, J.F., Kokmen,
E. and Tangalos, E. (1991). Age-associated memory
impairment diagnoses: Problems of reliablity and

concerns for terminology. Eexeheleg¥_end_bging. 5(4).
551-558.

Steel, K. and Feldman, R. (1979). Diagnosing dementia and
its treatable causes. genieeziee, 2, 79-89.

Storandt, M., Botwinick, J., Danziger, W.L., Berg, L. and
Hughes, C.P. (1984). Psychometric differentiation of
mild senile dementia of the Alzheimer type. Azeniyee

ef_nenrelegx. 11. 497-499.

Straker, M. (1984). Depressive pseudodementia. EeyenieEIie
Annals. 11(2). 93-99.

Teri, L., Larson, E.B. and Reifler, B.V. (1988). Behavioral
disturbance in dementia of the Alzheimer's type.

lenrnal_ef_tne_Anerisen_§eriatries_§eeietx. 2e. 1-6.

Teri, L., and Wagner, A. (1992). Alzheimer's disease and
depression. '

Psxehelegx. 59(3). 379-391-

Tierney, M.C., Snow, W.G., Reid, D.W., Zorzitto, M.L. and
Fisher, R.H. (1987). Psychometric differentiation of
dementia: Replication and extension of the findings of

Storandt and coworkers. Arenixee_ef_nenrelegx. 11.
720-722.

Vitaliano, P.P., Breen, A.R., Albert, M.S., Russo, J. and
Prinz, P.N. (1984). Memory, attention, and functional
status in community-residing Alzheimer type dementia
patients and optimally healthy aged individuals.

lenrnal.ef_eerentelegx. 22(1). 58-64-

Vollhardt, B.R., Ackerman, S.H., Grayzel, A.I. and Barland,
P. (1982). Psychologically distinguishable groups of
rheumatoid arthritis patients: A controlled, single

blind study. Eexeheeenetie_nedieine. 11(4). 353-362.

Wechsler, D. (1945). A standardized memory scale for
clinical use. lenrnal.ef_zexeneleex. 12. 87-95.

Weingartner, H., Cohen, R.M., Murphy, D.L., Martello, J. and
Gerdt, C. (1981). Cognitive processes in depression.
, 1Q, 42-47.

Weingartner, H., Cohen, R.M., Bunney, Jr., W.E., Ebert, M.H.
and Kaye, W. (1982). Memory learning impairments in

progressive dementia and depression. Ane;ieen_lenznei

92

21.2222811122. 112. 135'135-

Whitehead, A. (1973). Verbal learning and memory in elderly
depressives. ar11ien_lenrnal_ef_Bsyeniatry. 122. 203-
208.

Winogrond, I.R. and Fisk, A.A. (1983). Alzheimer's disease:
Assessment of functional status. genznei_ef_;ne

Anerisen_eeriatriee_eeeiety. 21(12). 780-785.

Yesavage, J.A., Brink, T.L., Rose, T.L., Lum, 0., Huang, V.,
Adey, M. and Leirer, V. O. (1983). Development and
validation of a geriatric depression screening scale:

A preliminary report.1enrnal_ef_Eey_niatrie_Beeearen.

11(1), 37-49.

   

"I71111111111081“