 

missus IllllallllllllllllNIHHIIIUIIIHIHIIHHIUHIIlillHINHll

31293 01570 3188

LIBRARY
Michigan State
University

 

 

 

 

This is to certify that the

thesis entitled

A Study of Memory Complaints in the Able Elderly

presented by

Jodi S. Levy

has been accepted towards fulﬁllment
of the requirements for

M . A . degree in pSAYChO'I 09]

[like Am

Major professor

 

Date é/?/?7

 

0-7639 . MS U is an Afﬁrmative Action/Equal Opportunity Institution

 

 

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 ie An Afﬁrmative Action/Emmi Opportunity inetiiuion
W

new

 

 

A STUDY or MEMORY COMPLAINTS IN THE ABLE ELDERLY
By

Jodi Suzanne Levy

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

MASTER OF ARTS

Department of Psychology

1997

ABSTRACT
A STUDY OF MEMORY COMPLAINTS IN THE ABLE ELDERLY
By

Jodi Suzanne Levy

Many Older adults complain Of memory decline (Lowenthal, Berkman, Buehler,
Pierce, Robbinson & Trier, 1967). Most empirical work, however, has demonstrated that
memory complaints are not directly related to memory performance (e.g., Zelinski,
Gilewski, & Schaie, 1993). This study addressed the inﬂuence Of depression, physical
health complaints, educational level, and premorbid levels of ability on subjective memory
complaints among 130 community dwelling Older adults (Mean age = 67.4). The impact of
certain types Of memory complaints and mood and memory training was also investigated.
Depression and physical health complaints accounted for signiﬁcant (p’s < .001)
proportions of the variance in memory complaints (R2 = .15; R2 = .10, respectively).
Contrary to expectations, level of education did not relate to memory complaints.
Consistent with previous research on general memory complaints, speciﬁc types Of
memory complaints (e.g. memory of names and faces) were not directly predictive of
memory performance. However, when premorbid ability was taken into account a
relationship between memory complaints and Objective memory performance was
demonstrated; diﬁ‘erence scores accounted for more than 5% of the variance in Rivermead
scores (R2= .05, p < .01, Beta = -.23, p < .01). Finally, treatment signiﬁcantly reduced

memory complaints (1(81) = -3.78, p < .001).

To my Mom, Dad, and Matt for all their love and support.

iii

ACKNOWLEDGMENTS

I have been fortunate to have an extraordinarily supportive network of people who
have helped me to grow and develop both personally and professionally. First, I would
like to sincerely thank my mentor and committee chairperson, Dr. Norman Abeles,
without whose patience, insightful comments, and constant support I would have not
completed this work. In addition, Dr. Abeles has generously given of his time, Oﬁ'ered me
many opportunities, and has helped me to develop as both a clinician and a researcher. I
would also like to express thanks to the members of my thesis committee, Dr. John Hurley
and Dr. Ellen Strommen. Dr. Hurley Oﬁ'ered many helpful comments and criticisms
throughout the completion of this thesis, but more importantly he always believed in me
and has helped me to believe in myself. Many thanks also go to Dr. Strommen, for her
insight from a developmental perspective and warm encouragement. I

I would like to express thanks to classmates, in particular Camilla Williams for
sharing the “grad school experience” with me. Your support has kept me going and I will
always be grateﬁrl to you for that. Camilla- I could not have asked for a better ﬁiend. I
would also like to thank my Mom and Dad for their lifelong commitment to my personal
grth and happiness. My achievements are a result Of your constant love and
encouragement.

Finally, I’d like to thank Matt Cushman for being my partner in life. Your personal

strength and commitment inspires me. “You complete me.”

iv

TABLE OF CONTENTS

INTRODUCTION ..................................................................................................... 1
Depression ..................................................................................................... 3
Physical Health ............................................................................................... 6
Level of Education .......................................................................................... 6
Premorbid Ability ............................................................................................ 7
Self-Report Memory Measures ........................................................................ 8
Treatment Eﬁ‘ect ............................................................................................. l3
Hypotheses ..................................................................................................... 15

METHOD .................................................................................................................. l7
Participants and Procedures ............................................................................ 17
Measures ........................................................................................................ 18

RESULTS .................................................................................................................. 22
Hypothesis I .................................................................................................... 22
Hypothesis H .................................................................................................. 23
Hypothesis III ................................................................................................. 25
Hypothesis IV ................................................................................................. 26
Hypothesis V .................................................................................................. 27

DISCUSSION ............................................................................................................ 29
Depression ...................................................................................................... 29
Physical Health ............................................................................................... 31
Level of Education .......................................................................................... 32
Premorbid Functioning .................................................................................... 33
Speciﬁc Types of Complaints .......................................................................... 35
Treatment ....................................................................................................... 37
Limitations of Study ........................................................................................ 38
Broader Implications ....................................................................................... 39
Directions for Future Research ........................................................................ 40

LIST OF REFERENCES ........................................................................................... 42

INTRODUCTION

According to recent population estimates, there are over 51 million people in the
United States (21% of the total population) who are at least 55 years of age (US Bureau
of Census, 1987 as cited in LaRue, 1992). By 2030, it is projected that one in every three
Americans will be 55 years old or older. This age group is by far the fastest growing
segment of the population (Tranel, Benton, & Olson, 1997). “nth the growing number Of
older adults, the importance of understanding aging and common disorders of later life is
diﬁcult to exaggerate.

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

Generally, memory problems becOme more common with advancing age; for
example, sensory memory (a kind of immediate registration like hearing someone’s name

once or looking at a photograph quickly) declines slightly with increasing age; older

2
adults were found to require longer exposure time to adequately register external stimulus

than younger individuals (e.g., Cerella, Poon & Pozard, 1982). Age diﬁ‘erences in long
term memory have also been widely reported (La Rue, 1992). However, this trend does
not generalize to all elderly adults (La Rue, 1992). In fact, many individuals in their sixties
and seventies perform as well on memory measures as younger adults (La Rue, 1992).
Since the ﬁndings are equivocal, it is important to consider individual differences such as
gender, education, health, and aﬁ‘ective status when studying memory functioning in older
adults to evaluate their impact on memory functioning and memory problems.

An understanding of memory complaints is also essential because diagnosis and
treatment decisions are in part based upon the amount and intensity of elders‘ memory
complaints (Zarit, 1980). Fifty to eighty percent of Older adults have subjective memory
complaints (Lowenthal, Berkrnan, Buehler, Pierce, Robbinson and Trier, 1967). However,
empirical work has also demonstrated that memory complaints are not directly related to
memory performance (Kahn, Zarit, Hilbert & Niederehe, 1975; Rosavage, 1990; Zelinski,
Gilewski, & Schaie, 1993, Popkin, Gallagher, Thompson et al., 1982; Bolla, Lindgren,
Bonaccorsy, et al., 1991). The proposed diagnostic criteria for AAMI are dependent upon
the presence of subjective complaints in addition to memory impairment. The variable
relationship between memory complaints and objective performance has caused some to
question this criterion (Caine, 1993).

One aim of this study was to look more closely at memory complaints, speciﬁcally
the inﬂuence of affective status and, in particular depression, physical health, education,

and premorbid functioning on memory complaints. This study also explored speciﬁc types

3
of memory complaints and their relation to objective performance. The Observed weak

relationship between self-reported memory and objective performance might also be a
consequence of the type of questionnaire used. Most empirical work has examined global
types of memory complaints. It has been suggested that a stronger relationship between
self reported memory and Objective performance might be found if speciﬁc types of
memory beliefs; for example, memory of names and faces, were examined individually
(lirﬁschel, 1968). This study investigated the relationship between the various dimensions
of subjective memory (i.e. the ten factors on the MAC-S; Wmterling, Crook, Salama, &
Grobert, 1986) and objective performance. Finally, this study determined if memory
complaints could be decreased through treatment; participants also completed Mood and
Memory training workshop groups. A great majority of research in the area of memory
complaints to date has used clinical samples (Scogin, 1985). These samples are not
representative of the population, and so the results might not be generalized. For example:
depressed patients tend to complain more about memory problems. Thus, it was important
that we examined the variables that inﬂuence memory complaints in normal healthy older
adults in order to make appropriate diagnostic and treatment decisions. To further Show
the rationale underlying this study, we have included a review of the literature of memory
complaints in the areas of depression, level of education, physical health, premorbid
ability, self report measures and treatment.
Depression

Depression appears to play a role in the relationship between memory complaints

and memory performance. Kahn, Zarit, Hilbert and Niederehe (1975) examined a group of

4
elderly adults diagnosed with late onset dementia. They found that individuals with mild

dementia but with high reported levels of depression were more likely to complain about
their memory than individuals who suffered more from severe dementia but were not
depressed. Furthermore, Popkin, Gallagher, Thompson and Moore (1982) found that as
depression remitted, memory complaints were signiﬁcantly reduced. Gagon, Dartigues,
Mazaux, Dequae, Letenneur, Giroire, and Barberger-Gateau (1994) conducted a survey of
self-reported memory complaints and memory performance of 2,726 older adults living in
the south-west of France. Those who scored above a threshold for depressive symptoms
reported more problems with their memory.

Poitrenaud, Malbezin and Guez (1989) conducted a seven year longitudinal study
to determine the relationship between age associated changes in memory test performance
and self-ratings of memory abilities. Results demonstrated an age-associated decrease in
memory test performance, but this objective impairment was not related to self-rating of
memory. They suggested that complaints about memory performance by many aged
participants reﬂect affective status more than actual decline in memory performance.

La Rue, Swan and Carmelli (1995) administered the Center for Epidemiological
Studies Depression scale (CBS-D) to 1,232 older adults. This measure is a 20-item self-
report scale in which individuals are instructed to rate the frequency with which they
experience particular thoughts and feelings. Participants also completed a three hour
cognitive battery to assess level of memory functioning. In contrast to previous ﬁndings,
La Rue et al. (1995) found a modest but consistent relationship between depressive

symptoms and cognitive performance.

5
The results of these studies suggest that depression is an important component of

memory complaints among elderly adults. However, most elderly adults are not clinically
depressed. Pettinati, Magee, and Mathisen (1984) found that non-depressed elders
reported signiﬁcantly more problems in certain types of memory ﬁrnctions than did
depressed elders. In another study White and Cunninghan (1984) found that depression
was related to memory performance, but memory complaints and memory performance
were not signiﬁcantly correlated. Thus the ﬁndings with these samples were inconclusive.
The present study examined the effect of depression on memory complaints in a
sample of able elderly. Depression was measured using the Beck Depression Inventory
(BDI; Beck et al., 1961) and the Geriatric Depression Scale (GDS; Yesavage, Brink, Rose
et al., 1983). Jones (1995) reported that when the GDS was used as a measure of
depression in older adults, the correlation between depression and memory complaints was
lower than the correlations reported in previous studies (e.g. Kahn, Zarit, Hilbert, &
Niederehe, 197 5). He suggested that the use Of a depression measure such as the GDS,
which measures mood related depression without a preponderance of vegetative/somatic
symptom complaints, may have contributed to the smaller correlations between depression
and subjective complaints and he recommended that future investigations utilize both types
of measures of depression. This study used both the BDI (which taps into the
vegetative/somatic symptoms Of depression) and the GDS (which measures only the

aﬁ‘ective component of depression).

Physical Health

Individual diﬁ‘erences in memory complaints can also be a consequence of
diﬁ‘erence in physical health. In a national survey of nearly 15,000 older adults, Cutler and
Grams (1989) found that those reporting memory complaints were more likely to be in
poor physical health and had more hearing and visual impairments than those individuals
without memory complaints. In a study of 403 Older adults, Hanninen, et al. (1994) found
that individuals who had stronger tendencies toward somatic complaining and greater
anxieties about physical health reported more memory complaints regardless of their
Objective performance. Collins and Abeles (1996) found a signiﬁcant relationship between
subjective memory complaints and somatic complaints (e.g., insomnia, pain, physical
problems). The present study also examined the association of physical health, physical
health complaints, and memory complaints.

Level of Education

Intelligence and level of education can also effect the amount and type of memory
complaints experienced by older adults. Derouesne et al. (1989) did not ﬁnd relationship
between severity of memory cOmplaints and educational level, but later studies have
diﬁ’ered. The association between memory complaints and performance on standardized
memory tests was examined by Bolla, Lindgren, Bonaccorsy, and Bleeker (1991). They
found that individuals with higher levels of verbal intelligence reported fewer memory
complaints, less forgetting, and placed less emphasis on forgetting. Bolla et al. (1991) also
found that those with higher verbal intelligence were more likely to be well read and be

better informed about the aging process and age related disorders. This knowledge may

7
help them to recognize everyday memory problems as parts of the normal aging process

rather than signs of dementia. Gagon et al. (1994) examined the relationship between self-
reported memory decline, demographic variables and memory performance in 2,726 adults
living in the south-west of France. They found lower memory performance was related to
both older age and lower levels of education. A signiﬁcant relationship was detected
between memory complaints and education; those reporting more memory complaints
were from the lower educational group. Derouesne, Dealberto, Boyer, and Lubin (1993)
reported similar ﬁndings; low educational status was predictive of cognitive complaints in
a sample of 634 adults ages forty-ﬁve to seventy-ﬁve. Larnberty, Bieliauskas, Chatel, and
Holt (1993) found that education accounted for a signiﬁcant proportion of the variance
(19%) in memory performance of a geriatric clinical sample. Many years of education
might place an individual ahead of one’s cohorts with respect to general cognitive abilities.
They suggested that education might serve as a buffer against the eﬁ‘ects of the dementing
process and hence against decline that might manifest itself in the form of memory
complaints.

This study examined the moderating eﬂ‘ect of level of education on the
relationship between self reported memory complaints and actual performance.
Participants were asked to report the years of formal education that they have received.
Premorbid Ability

The questionable relationship between memory complaints and test performance
might be a function of premorbid ability. Individuals with extremely high levels of

premorbid ﬁmctioning may experience declines that bring their current functioning down

8
to 'normal levels'. This decline may manifest itself in more memory complaints even

though standard tests of memory ability may not reveal any evidence of decline. Measures
of intelligence have been reported to predict memory performance well in elderly subjects
(Rabbitt and Goward, 1986; Schlosser & Ivison, 1989). Christensen (1991) found that
those who complained of memory problems were more likely than non-complainers to
show a discrepancy between their current level of ﬁrnctioning and that which would be
expected on the basis of their premorbid IQ. On the other hand, Jorrn, Christensen,
Henderson, Korten, Mackinnon & Scott (1994) found results that were inconsistent with
those found by Christensen (1991) with a similar population. When premorbid estimates
were factored in correlations between memory complaints and performance did not

increase. Further investigation into this area seemed warranted.

Self-Report Memory Measures

The weak relationship observed between self-reported memory and objective
performance might also be a consequence of the type of self report questionnaire used.
The content and wording of items, the type ofjudgrnent required of the subject, scaling,
and the number of dependent measures in a questionnaire can all eﬁ‘ect the external
validity of self report measures.

Self assessment of memory complaints is important for a number of reasons.
Memory complaints provide valuable information about memory impairments. Reisberg,
Ferris, Borenstein, Sinaiko, de Leon, and Buttinger (1986) reported that individuals with

severe memory impairment tended to overestimate their memory abilities. The discrepancy

9
between patients' self-assessment of their memory ability and their actual performance may

help diﬁ‘erentiate a pathological processes ﬁ'om normal aging. In addition, patients'
memory complaints might be an early warning sign of dementia (Gilewski & Zelinski,
1986). Memory complaints are also part of the diagnostic criteria for Age Associate
Memory Impairment (AAMI).

Self-reported memory might also be useﬁrl in the identiﬁcation of depression in
older adults. Complaints about ability without evidence of performance deﬁcits may be an
indicator of depression. Depressed people are thought to complain more about their
memory ability than those who are not depressed.

Questionnaire assessment of memory complaints can also facilitate the
investigation of memory problems. Research on memory used in everyday life (e. g. ability
to remember names and faces) has been restricted because it seems diﬁcult to conduct
(Herrmann, 1982). For instance, participants often have to be observed for long periods of
time to obtain an accurate estimation of their ability and observations alone do not reveal
the type of memory strategy used. In addition, laboratory tests seem threatening to many
older people, so the use of memory questionnaires in lieu of such tests might increase
cooperation during assessment (Gilewski & Zelinski, 1986). If questionnaire responses
corresponded well with observed performance, researchers would have a method of
memory assessment that would bypass at least some of the problems associated with ﬁeld
and experimental research, so we next report research relating to this issue.

Self-report questionnaires that have been developed usually describe prototypical

memory situations and ask the respondent a number of multiple-choice questions about

10
how their memory ﬁmctions in a speciﬁc situation (Herrmann, 1982). Evaluating the

utility of memory questionnaires requires examining correlations between responses to
those questionnaires and objective memory performance.

Memory questionnaires differ in the kinds of questions they ask. In general, self-
report questionnaires have asked about one or more of the following: forgetting,
remembering, memory quality, memory change, memory use, and attitudes about memory.
The format of the questionnaires also vary. For example, participants may be asked how
often they forget in absolute time (e. g., once a week), in relation to the number of
opportunities (e.g., once out of every ten attempts), or relative to the amount others
forget (e.g., about average compared to my peers). Memory questionnaires also vary in
the amount of detail used in the questionnaire description; some memory questionnaires
have very low question detail whereas other questionnaires have high question detail.

In a review, Herrmann (1982) presented evidence suggesting that self-report
memory questionnaires are not a substitute for measures for memory performance.
Although some questionnaires were shown to relate to performance moderately well (r_>
.5), the evidence is not strong enough to warrant using them as a sole means of
assessment. In fact, some questionnaires yielded negative validity coeﬁicients (e.g.
Herrmann, 1975).

Although the correlations between self-report measures of memory and actual
performance tend to be low, ﬁlrther investigation is still warranted. Experimenter- invoked
laboratory tasks oﬁen differ substantially from memory tasks in everyday life. Herrmann

(1982) states that if a memory test is not reﬂective of the situations asked about on a

ll

questionnaire, the correlations between subjective complaints and objective performance
have limited value. Thus, the ability of self-report memory measures to predict actual
performance may be substantially better than that reported to date.

Self-report questionnaires might better predict actual performance if more
speciﬁc questions were used (Mischel, 1968). The use of questionnaires that focus on
only one type of performance (e. g., memory of names and faces) might also strengthen the
predictive validity of memory questionnaires (Herrmann, 1982). Gilewski and Zelinski
(1986) suggest that questionnaires either focus on one or several dimensions of memory
functioning. Bolla et al. (1991) found that older adults tended to report more forgetting in
speciﬁc situations than younger adults. They suggested that older adults may be more
sensitive to situations in which they forget, whereas younger individuals pay less attention
to the situations in which they experience forgetting. Christensen (1991) found that elderly
individuals who considered their memory functioning to be worse than that of their peers
did, in fact, have poor memory performance. However, those having general memory
complaints did not differ from those without complaints on the clinical memory tasks.
Larrabee, West and Crook (1991) conducted a study addressing the relationship of self-
reported memory using the MAC-S (Memory Assessment Clinics Self-Report; Winterling,
Crook, Salama, and Grobert, 1986) to self-rated depression and actual performance on
computer simulated everyday memory tasks. They found that MAC-S Ability factors
(remote, everyday and semantic memory) and Frequency factors (concentration and
forgetfulness) were associated with objective performance. However other MAC-S

Ability factors (numeric and spatial memory) and Frequency factors (semantic, everyday

12
and facial memory) had primary associations with depression. These studies suggest the

importance of questionnaires which address speciﬁc types of complaints, rather than
global complaints, for use with elderly adults. Herrmann (1982) states that such tests
should include items that could be used to identify participants who are not focusing on
the questions. All of these considerations speak to the importance of selecting an
appropriate instrument to assess memory complaints.

The MAC-S, selected for use in this study, was developed by Winterling, Crook,
Salama, and Grobert in 1986 to address many of the limitations of the existing scales. The
MAC-S includes global items (e.g. "In general, as compared to the average individual,
how would you describe your memory?"), a 21 item subscale examining overall ability to
remember, and a second subscale of 24 items measuring frequency of memory failure
occurrences.

The present study also addressed the association between performance on the
Rivermead Behavioral Memory Test (Wilson, Cockbum, Baddley, & Hiorns, 1989) a
standard measure of everyday memory functioning and the MAC-S self-rating scores in
normal healthy adults. The factor structure of the MAC-S allowed us to examine speciﬁc
types of subjective memory and determine their relationship to objective performance. As
mention earlier Herrmann (1982) suggested that this might increase the predictive validity
of self-report measures. In addition, each factor contains several items, permitting an
internal consistency analysis and facilitates eliminating who are not attending to the

questionnaire.

13

Treatment Effect

As mentioned above, the majority of older adults report subjective memory
complaints (Lowenthal, Berkrnan, Buehler, Pierce, Robbinson & Trier, 1967). Lane and
Snowdon (1989) examined a random sample of 146 older adults with a mean age of 73.6
years and reported the prevalence rate of AAMI (Age Associated Memory Impairment) to
be 34.9%. Because memory decline and memory complaints are so prevalent in older
adults, it is imperative that we develop effective treatments to address these concerns. One
possible way to treat AAMI is through memory training. This study examined the ability
of mood and memory training groups to reduce memory complaints, one of the two major
diagnostic criteria for a diagnosis of AAMI.

Scogin, Storandt, and Lott (1985) examined the effectiveness of a self-taught
program of memory skills for older adults who had complained of memory problems in an
immediate-treatment/waiting-list design. They found a signiﬁcant impact of training on
memory performance but not on memory complaints. A subsequent evaluation of the
waiting-list group replicated these ﬁndings. Scogin et al. reported that self-evaluations of
memory performance were not systematically related to objective measures because older
people seeking treatment for memory problems may have special attitudes and
expectations about their cognitive abilities. Individuals who are more concerned with the
aging process and more sensitive to age related changes might be more critical of changes
in their own performances; though they did not verify this empirically. Low complainers,

on the other hand, may have more accepting attitudes about what they consider to be parts

14
of the normal aging process. They further indicated that treatments should address

attitudes and expectations about the aging process in addition to providing strategies that
facilitate memory performance. Scogin and Bienias (1988) conducted a three year follow-
up study to determine if participation in this memory training by Scogin et al. (1985) had
long term effects. They found that the increased memory performance was temporary and
observed no signiﬁcant changes in memory complaints.

Best, Harnlett and Davis (1992) conducted a study in which one experimental
training group received memory training and another (the expectancy change condition)
was presented with stereotypes concerning aging and then participated in a discussion of
research that contradicted these popular negative beliefs. They found that individuals in
the expectancy change condition experienced a signiﬁcant decline in memory complaints
and those in the memory training group experienced a signiﬁcant improvement in memory
performance but no change in their memory complaints. These ﬁndings indicate that
changing expectations of the elderly concerning their cognitive capabilities effectively
decreases complaints about memory but has little impact on memory performance.

The training program used in the present study taught strategies to facilitate
memory while addressing the elderly expectations about memory performance.
Participants received educational instruction and discussion of memory loss and its
association to aging. The purpose of this part of the training was to discourage negative
stereotypes about aging and provide the participants with a more positive expectancies of
the aging process. Participants were also taught several mnemonic strategies which were

expected to improve performance.

l 5
Hypotheses

1. Depression was expected to be positively correlated with memory complaints.

Scores on the Beck Depression Inventory (BDI; Beck et al., 1961) and the
Geriatric Depression Scale (GDS; Yesavage, Brink, Rose et al., 1983) were expected to
be negatively correlated with scores on the Memory Assessment Clinics, Inc. Self Rating
Scale (MAC-S; Revised, Winterling, Crook, Salama and Grobert, 1986) (low scores on
the MAC-S represent more memory complaints).

2. Additional factors that were expected to inﬂuence memory complaints were physical
health complaints and level of education.
a. individuals with more physical health complaints were expected to report more
memory complaints.

Individuals who complained about their physical health (e.g., poor overall health

at the present time, decline in physical health over the past three years, and

physical health worse than their peers) were expected to have lower MAC-S

scores (Winterling, Crook, Salama & Grobert, 1986).

b. higher levels of education were expected to be correlated with lower levels of
memory complaints.

Individuals with more years of schooling were expected to have lower MAC-S

scores (Winterling, Crook, Salama & Grobert, 1986).

3. When premorbid abilities are taken into account, participants memory complaints were

expected be a more accurate representation of their memory decline.

16
This utilized standard scores on the American Version of the Nelson Adult

Reading Test (AMNART, Schwartz & Saﬁ‘ran, 1987) and the Rivermead Behavioral
Memory Test (Wilson, Cockbum, Baddley, & Hiorns 1989)). The standard score
diﬁ‘erence was used to predict scores on the MAC-S.

4. Certain types of memory complaints were expected to predict performance, while
others are not.

MAC-S (Memory Assessment Clinics Self-Report; Winterling, Crook, Salama, and
Grobert, 1986) Ability factors: remote, everyday and semantic memory and Frequency
factors: concentration and forgetfulness were expected to predict objective performance.
5. Participation in Mood and Memory Workshops were expected to reduce memory
complaints in older adults.

Scores on the MAC-S (Winterling, Crook, Salama & Grobert, 1986) were

expected to increase at both clinical and statistical signiﬁcant levels from time 1 to time 2.

METHOD
Participants and Procedures

From the ongoing Michigan State University (MSU) Psychological Clinic Aging
Research Project, the participants were a subset of community dwelling elderly recruited
through advertisements. Each individual was oﬁ‘ered two assessments of their mood and
memory, as well as a seven-session workshop designed to teach relaxation and cognitive
strategies for the relief of depression and /or memory difﬁculties. This pool of 132
participants contained protocols of older individuals ranging from 47 to 90 years of age
(M = 67.6 ; _S_D = 8.8 ). Of the 132 participants (87 women and 45 men), with a mean
education of 14.8 years (SD = 2.8), 88 completed both pre and post testing. Adults with
signiﬁcant heath problems were excluded in the sample. Individuals with severe
depression, as determined by scores of 30 or higher on the BDI and 20 or higher on the
GDS referred elsewhere for treatment.

Each participant was administered all instruments at two diﬁ‘erent times: once
before a memory training group and once immediately following the memory training
group, with approximately one month between each test. Each testing period took
approximately one hour. During this time demographic information was also collected.

The memory training consisted of mood and memory workshops including
discussions of memory loss and its association to aging. Participants also received
specialized training in mnemonic techniques (e.g., the method of loci). Seven 90 min.
workshops were conducted by a clinical psychology graduate student for 6 to 8

participants. Workshops were based on a manual developed by researchers ﬁ'om the

17

18

Ethel Percy Andrus Gerontology Center and have demonstrated prior utility Lewinsohn,
Antonuccio, Breckenridge, & Teri, 1984). Treatments of this type have been conducted
since 1979; as of 1984 over 300 people had enrolled in the course. Research has shown
that when evaluated as a group, the participants in this course show marked improvement,
which is maintained for at least six months after the treatment (Lewinsohn, Antonuccio,

Breckenridge & Teri, 1984).

Mﬂures

a. Memog Assessment C1in_ics Self-Rating Scale (MAC-S, Revised, Winterling, Crook,
Salama & Grobert, 1986)- Designed to assess subjective memory complaints this
questionnaire contains 21 Ability items and 24 Frequency of Occurrence of memory
problems items. All rated on a 5-point Likert scale, ranging from very good to very poor.
Crook and Larrabee (1990) factor analyzed the performance of 1,106 subjects for a
normative sample of adults ranging from 18- 92 years. West, Crook, and Larrabee (1991)
showed concurrent validity of this scale. Five Ability factors were identiﬁed and labeled
Remote Personal Memory, Numeric Recall, Everyday Task Oriented Memory, Word
Recall/ Semantic Memory, and Spatial and Topographic Memory. On the Frequency scale,
ﬁve scales were also identiﬁed, and these were labeled Word and Fact Recall, Attention
and Concentration, Everyday Task-Oriented Memory, General Forgetﬁrlness, and Facial
Recognition. Crook and Larrabee (1991) reported substantial test-retest stability across 3-

week intervals Ability Total [ = .88 - . 94, Frequency Total r = .89 - .92).

19

b. Am_e_r_ic_an Version of the Nelson A_dult Reading Test (AMNART, Schwartz & Saman,
1987)- This measure of premorbid intelligence includes 46 words whose spelling and
pronunciations are nonstandard in American English (e.g., hiatus). Standardization was
conducted on 109 normal adults aged 40 - 89. In agreement with Nelson (1982) high
correlations between predicted IQ using the AMN ART and various WAIS IQ measures
were found when the inﬂuence of education was partialled out: ,r_= .72, .51, and .72 for
VIQ, PIQ, and FSIQ respectively.

c. BecLDepression Inventory (BDI ; Beck et al., 1961) Including items addressing mood,
sense of pessimism and guilt, social withdrawal, sleep disturbances, loss of energy, and
weight and appetite, this 21-item 4-point scale addresses the intensity of depressive
symptoms. Beck et al. (1961) reported high internal consistency (r_ = .93). Good test-
retest reliability has also been reported (r = .74) after a three month interval (Miller &
Seligrnan, 1973). Studies with the elderly show that the BDI has respectable internal
consistency and stability for use in research with this population (Spitzer, Endicott, &
Robins, 197 8). It has a high detection rate for major and minor depressive disorders, with
a misclassiﬁcation rate of approximately 16-17% using customary BDI cutoﬁ‘ scores
(Gallagher, Nies, & Thompson, 1982; Gallagher, Breckenridge, Steinmetz & Thompson,
1983). It seems to be a useful instrument for identiﬁcation of depressed elderly.

d. Geriatric Depression Scﬂ ( Yesavage, Brink, Rose et al., 1983)- The GDS consists of
30 yes-no items. Designed speciﬁcally for use with older patients, this instrument surveys
mood quite extensively with additional items to assess cognitive complaints and social

behavior. Yesavage et al. (1993) suggested that the GDS is a reliable measure. Test-retest

20

reliability was calculated by having subjects complete the questionnaire twice, one week
apart. A correlation of .85 was obtained (p <.001). Convergent validity was found
between the GDS, the lung Self-Rating Scale of Depression (.86), and the Hamiliton
Rating Scale for Depression (.83). Furthermore, the GDS appeared to have respectable
internal consistency ([ = .94) and stability for use with the elderly. According to Yesavage
et al. (1993), this scale is especially suitable for detecting depression in the elderly because
it avoids classiﬁcation errors due to poor physical health.
e. Rivermead Behaviml Memory Test (Wilson, Cockburn, Baddley, & Hiorns 1989)-
This is a brief test of everyday memory problems. There are twelve components of the
Rivermead, and are as follows: Remembering a name (ﬁrst (a) and last (b)), (c)
Remembering a hidden belonging, (d) Remembering an appointment, (e) Picture
recognition, (f) Remembering a newspaper article (immediate and delayed), (g) Face
recognition, (h) Remembering a new route (immediate), (i) Remembering a new route
(delayed), (j) Delivering a message, (k) Orientation (9 questions in both time and place),
and (1) Date. Each component is allocated 2 points (normal), 1 point (borderline), or 0
points (abnormal) depending on the raw score. The Standardized Proﬁle Score is the sum
of these points across all components. This enables comparison across components
(Wilson, Cockborn, & Baddeley, 1995).

Wilson et al. (1989) found high interrater reliability (100%) between two raters
scoring the proﬁles of 40 subjects (Wilson et al., 1989). High face validity as well as
normative data on normal and impaired participants have also been established. The test-

retest reliability with 118 patients whom were test twice yielded a correlation of .85 for

21

the Proﬁle Score. Additionally, the normative study for elders (mean age 80.5 years)

showed that the test was at the appropriate level of diﬁiculty for this population.

RESULTS

Hypothesis 1.

The hypothesized positive relationship between depression and memory
complaints was supported. The Beck Depression Inventory (BDI; Beck et al., 1961) and
the Geriatric Depression Scale (GDS; Yesavage, Brink, Rose et al., 1983) scores were
expected to correlate negatively with scores on the Memory Assessment Clinics, Inc. Self
Rating Scale (MAC-S; Revised, Winterling, Crook, Salama & Grobert, 1986). Recall
that low scores on the MAC-S represent more memory complaints. This hypothesis was
conﬁrmed.

The BDI total score on the correlated signiﬁcantly with MAC-S the total score (5
= -.38, p < .001), with the ability total (5 = -.333, p < .001), and with the frequency total (r
= -.38, p < .001). The GDI total score also correlated with MAC-S total score (r_= -.39, p_
< .001), with the ability sub-total (r = -.34, p < .001), and with the frequency subtotals ([
= -.37, p < .001).

Depression, as measured by both the BDI and the GDS, was a signiﬁcant
predictor of memory complaints as measured by the MAC-S. The BDI score by itself
accounted for slightly more than 14% of the MAC-S total score variance (R2 = .14, p <
.001), approximately 13% of the MAC-S frequency total score variance (R2 = .13, p <
.001), and 11% Ofthe MAC-S ability total score variance (R2 = .11, p < .001). The GDS

score by itself accounted for slightly more than 15% of the MAC-S total score variance

22

23

(R2 = .15, p < .001), 13% Ofthe MAC-S frequency total score variance (R2 = .13, p <

.001), and 11% Ofthe MAC-S ability total score variance (R2 = .11, p < .001).

Hypothesis II.

It was predicted that physical health complaints and level of education would be
related to an individual’s level of memory complaints.

Individuals with more physical health complaints were expected to report more
memory complaints. Speciﬁcally, individuals who complained about their physical health
were expected to have lower MAC-S scores (Winterling, Crook, Salama & Grobert,
1986). This hypothesis was conﬁrmed.

Physical health complaints scores were determined by the participants responses to
three questions: a) How would you rate your overall health at the present time (1-
excellent, 2-good, 3-fair, 4-fair); b) is your health now better about the same , or not as
good as it was three years ago (l-better, 2-the same, 3-not as good), and e) do your health
problems stand in the way of doing things you want to do (l-not at all, 2-a little, 3-a great
deal). We calculated the participant’s physical health complaint score by summing the
responses to the three items mentioned above. Level of physical health complaints
reported were signiﬁcantly correlated in the predicted negative direction with the total
score on the MAC-S (r = -.3 1, p < .001), the ability total (r = -.25, p < .001), and the
ﬁequency total (r = -.307, p < .001). Physical health complaints were also found to be a
signiﬁcant predictors of memory complaints as measured by the MAC-S. The total

physical health complaint score by itself accounted for 10% of the MAC-S total score

24

variance (R2 = .10, p < .001), approximately 6% of the MAC-S ability total score
variance (Rz= .06, p < .001), and 9% of the MAC-S frequency total score variance (R2 =
.09, p < .001).

Higher levels of education were expected to be signiﬁcantly correlated with lower
levels of memory complaints. Individuals with more years of schooling were expected to
have lower MAC-S scores (W interling, Crook, Salama & Grobert, 1986). This hypothesis
was not conﬁrmed. Level of education was not signiﬁcantly correlated with the total score
on the MAC-S, the ability total, or the ﬁ'equency total. Level of education did not
signiﬁcantly predict memory complaints.

The correlates of actual memory ability, or every day memory functioning, as
determined by the sum of standard scores on the Rivermead Behavioral Memory Test
(Wilson, Cockbum, Baddley, & Hiorns 1989), were also investigated. Actual ability
correlated signiﬁcantly with MAC-S total scores (1' = .19, p < .05), ability total (r = .22, p
< .01), and frequency total (1; = .18, p < .05). These correlations are all positive,
suggesting that a higher memory ability is associated with fewer memory complaints.
Memory ability also signiﬁcantly predicted MAC-S of memory complaints. There are
twelve components on the Rivermead Behavioral Memory Test (RBMS); each component
is allocated 2 points (normal), 1 point (borderline), or 0 points (abnormal) depending on
the raw score. The Standardized Proﬁle Score is the sum of these allocated points across
components. This enables comparison across components (Wilson, Cockborn, &
Baddeley, 1995). The RBMS Standard Score Proﬁle by itself accounted for than 4% of

the MAC-S total score variance (R2 = .04, p < .05), approximately 3% of the MAC-S

25

frequency total score variance (R2 = .03, p < .05), and 5% of the MAC-S ability total
score variance (R2 = .05, p < .01).

Next, all variables (GDS, BDI, Rivermead Behavioral Memory Test, Level of
Education, and Physical Health Complaints) were entered into a regression model as
predictors. The combination of these ﬁve variables accounted for 21% of the MAC-S total
scores (R2 = .21, p < .001); however, none of the variables turned out to be signiﬁcant
predictors of MAC-S total score. The uncorrected correlation between the GDS score and
the BDI score is high (1; = .84, p <.001) resulting in colinearity. For this reason the BDI
score was eliminated from the model. The GDS score was selected as the measure of
depression to be incorporated in the model because it was designed speciﬁcally for use
with older patients and is especially suitable for detecting depression in the elderly because
it avoids classiﬁcation errors due to impaired physical health (Y esavage et al., 1993).
When GDS, Rivermead Behavioral Memory Test, Level of Education, and Physical Health
Complaints were entered into the model as predictors almost 20% of the variance in total
MAC-S is accounted for (R2 = .20, p < .001, GDS Beta = -.298, p < .002, Rivermead

Beta = .162, p < .05, Education Beta = -.075, ns, Health Complaints Beta = -. 171, ns).

Hymthesis HI.

When premorbid abilities are taken into account, participants memory complaints
were expected to be a more accurate representation of their memory decline. Speciﬁcally,
it was predicted that individuals who had larger diﬁ‘erences between their expected

performance as determined by their scores on the American Version of the Nelson Adult

26

Reading Test (AMNART, Schwartz & Saﬁ‘ran, 1987) and their actual performance as
determined by their sum of standard scores on the Rivermead Behavioral Memory Test
(Wilson, Cockbum, Baddley, & Hiorns 1989) would have more memory complaints. This
hypothesis was supported.

Difference scores were calculated by converting the total Rivermead Behavioral
Memory Test Score and the AMNART score to standard units. The standard score on the
Rivermead Behavioral Memory Test was subtracted ﬁom the standard AMNART score.
Sample means were used in the calculation of standard scores. This highly educated (M =
14.8), sample yielded above average scores on these two measures. Established means
therefore were not representative of this sample. The sample mean score on the Rivermead
Behavioral Memory Test is M= 18.4 and on the AMNART is M = 115.1. Diﬁ‘erence
scores accounted for more than 5% of the variance in total MAC-S scores (R2 = .05, p <
.01), 6% in ability scores (R2= .06, p < .001,), and 4% in frequency scores (Rz= .04, p <

.01).

Hypothesis IV.

Certain types of memory complaints were expected to be predictive of memory
performance, while others would not. Speciﬁcally MAC-S Ability factors (remote,
everyday and semantic memory) and Frequency factors ( concentration and forgetﬁrlness)
were expected to predict objective performance. To test this hypothesis all ten factor

scores (as proposed by Winterling et al., 1986) were entered into a regression model. Only

27

complaints of Spatial Memory problems signiﬁcantly predicted objective performance (B =
.34, p < .05). However, overall the regression model did not ﬁt the data.

A principal components factor analysis was run on the MAC-S to determine if the
factors reported by Crook and Larrabee (1990) could be replicated. This factor structure
could not be reproduced since all the items loaded onto one factor, suggesting that this
scale is unidirnensional. For this reason the total score on the MAC-S was used to predict
objective performance as determined by scores on the Rivermead Behavioral Memory
Test. Three percent of the Rivermead total score was explained by memory complaints (R2

= .03, p< .01).

Hmthesis V.

Participation in Mood and Memory Workshops was expected to reduce memory
complaints in older adults. Speciﬁcally, scores on the MAC-S (Winterling, Crook, Salama
& Grobert, 1986) were expected to increase at both clinical and statistical signiﬁcant
levels ﬁ'om time 1 to time 2. This hypothesis was conﬁrmed.

The mean total score on the MAC-S at pre-test M = 153.2, S_D = 23.7) was
lower than that of post-test M = 160.3, S_D = 23.2) suggesting that participation in the
mood and memory workshops reduced overall memory complaints. A paired diﬁ'erence t-
test determined that this difference was statistically signiﬁcant (t(81) = -3.78, p < .001).
This also held true for the ability and ﬁ'equency subtotals on the MAC-S .The mean ability
subtotal score on the MAC-S at pre-test M = 56.1, S_D = 9.8) was lower than that of

post-test M = 58.8, SD = 9.2) suggesting that participation in the mood and memory

28

workshops reduced complaints about memory ability. A paired difference t-test
determined that this difference was statistically signiﬁcant (t(87) = -3.45, p < .001). The
mean frequency subtotal score on the MAC-S at pre-test M = 64.0, S_D = 10.4) was
lower than that of post-test M = 67.1, SD = 10.0). A paired difference t-test also
determined that this diﬁ'erence is statistically signiﬁcant (t(86) = -3.47, p < .001).

The Reliable Change Index (RCI, Jacobson & Truax, 1991) was an indicator used
to determine if the magnitude of change by a given participant was clinically signiﬁcant.
Using this index, 10 participants reported less memory ability complaints (p < .05, two
tailed), and 17 participants reported less frequency of occurrence memory complaints.
Three participants experienced an increase in ability memory complaints and six in

frequency of occurrence memory complaints (2 < .05) by RCI.

DISCUSSION

Most empirical work has demonstrated that memory complaints are not directly
related to memory performance. Diagnostic decisions are often based upon the presence of
subjective complaints in addition to memory impairment. It is thus important to investigate
factors that might inﬂuence the relationship between complaints and performance. The
ﬁrst aim of this study was to look speciﬁcally at the inﬂuence of aﬁ'ective status and, in
particular depression, physical health, education, and premorbid functioning on memory
complaints. This study also explored the impact speciﬁc types of memory complaints
(e.g., memory for names and faces) and their relation to objective performance because
previous work has only investigated global memory complaints. Finally, the study
investigated the effect of treatment on memory complaints. The inﬂuence of these factors
will be discussed in more detail below.
Depression

A positive correlation between depression and memory complaints was
hypothesized. Previous studies suggested that depression is an important component of
memory complaints among elderly adults (Kahn, Zarit, Hilbert & Niederehe, 1975;
Popkin, Gallagher, Thompson & Moore 1982; Dartigues, Mazaux. Dequae, Letenneur,
Giroire, Barberger-Gateau, 1994; Poitrenaud, Malbezin & Guez, 1989). However, most of
these studies examined clinical populations, such as individuals suﬁ‘ering ﬁom depression
(Popkin, Gallagher, Thompson and Moore, 1982) or late onset dementia (Kahn, Zarit,
Hilbert & Niederehe, 1975). There are some doubts that results of these studies should be
generalized to the majority of older adults who are living independently

29

30

in the community. For instance, it is thought that clinically depressed patients tend to
complain more about memory problems and that those with dementia have an inaccurate
perception of their memory ability. It seemed important to examine the eﬁ‘ect of
depression on memory complaints in normal healthy older adults. As predicted, even mild
reports of depression correlated positively with memory complaints. This suggests
perhaps, that some consideration ought to be given to providing assistance for normal,
older adults with memory complaints.

Depression was measured using the Beck Depression Inventory (BDI; Beck et al.,
1961) and the Geriatric Depression Scale (GDS; Yesavage, Brink, Rose et al., 1983).
Jones (1995) reported that when the GDS rather than the BDI was used as a measure of
depression in older adults, the correlation between depression and memory complaints was
below those reported earlier (e. g. Kahn, Zarit, Hilbert, & Niederehe, 1975). He suggested
that the use of a depression measure such as the GDS, which measures only mood related
depression without a preponderance of vegetative/ somatic symptom complaints, may have
contributed to the smaller correlations between depression and subjective complaints. He
recommended that future investigations utilize both types of measures of depression. This
study used both the BDI addressing vegetative/somatic symptoms, and the GDS
addressing primarily affective components. As mentioned above, the BDI accounted for
14% of the variance in total memory complaints and the GDS accounted for 15% of the
variance in total memory complaints. The diﬁ‘erence in variance accounted for by the two

measures was not signiﬁcant. The participants in this study were healthy and ﬁ'om high

31

socio—economic groups; such a population is less likely to report somatic complaints and
hence might have lower BDI scores. The Geriatric Depression Scale is also a better
measure to detect depression in older samples (Y esavage et al., 1993). The uncorrected
correlation between the GDS score and the BDI score is high (5 = .84, p <.001) resulting
in colinearity. For these reason the BDI score was eliminated from the model when all the

variables were entered.

Physical Health

Individuals with more physical health complaints were expected to report more
memory complaints. In a recent longitudinal study Tranel, Beton, and Olson (1997), found
that those who remain in good health show good preservation of memory, perception,
concentration, language, and praxis on the other hand, signiﬁcant health problems were
associated with deterioration in neuropsychological function. This ﬁnding suggests that if
physical health remains good, aging does not necessarily have an adverse eﬁ‘ect on
memory. The present study found that those who report more physical health complaints
also have more memory complaints. It is possible that this ﬁnding could mean that some
individuals tend to complain more overall. On the other hand, it could suggest that
memory complaints are more suggestive of memory decline than previously believed.
Since decline in physical health is signiﬁcantly related to decline in memory functioning
(Beton et al., 1997), it follows that increases in physical health complaints would be

correlated with increased memory complaints, if complaints are an accurate indication of

32

the underlying problems (i.e., poor health and memory). It may also be that physical health

complaints aﬁ’ect attention and concentration and thus impact memory.

Lelelﬁof Education

Higher levels of education were expected to be signiﬁcantly correlated with lower
levels of memory complaints. This hypothesis was not conﬁrmed. This ﬁnding diﬁ‘ers from
the results of previous investigators. Two possible explanations for this diﬁ‘erence include
the sample used and the way intellectual ability was measured. This sample was highly
educated M = 14.8, SD = 2.8), giving us very little opportunity to investigate the
inﬂuence of lower levels of education on memory complaints; previous studies have
demonstrated that lower intellectual ability is predictive of more memory complaints. This
study may not have had enough participants with low levels of education to detect such an
effect. This study also used years of education as a measure of intellectual ability. Some
previous investigations that have reported signiﬁcant eﬁ‘ects used verbal IQ as a measure
of intellectual ability. For example, Bolla, Lindgren, Bonaccorsy, and Bleeker (1991),
found that individuals with higher levels of verbal intelligence reported fewer memory
complaints, less forgetting, and placed less emphasis on forgetting. Bolla et al. (1991) also
found that those with higher verbal intelligence were more likely to be well read and be
better informed about the aging process and age related disorders. This knowledge may
help them to recognize everyday memory problems as parts of the normal aging process

rather than signs of dementia. Bolla-Wilson and Bleeker (1986) suggested that it is more

33

appropriate to use vocabulary level than years of education when studying older adults;
older adults may not have many years of formal education but may have spent many years
self educating. Future studies should include both types of measures of intellectual

ﬁrnctioning.

MorbidﬁFunctionLng

The fourth hypothesis predicted that when premorbid abilities are taken into account,
participant’s memory complaints will be a more accurate representation of their memory
decline. Most empirical work has demonstrated that memory complaints are not directly
related to memory performance (Kahn, Zarit, Hilbert & Niederehe, 1975; Rosavage, 1990;
Zelinski, Gilewski, and Schaie, 1993, Popkin, Gallagher, Thompson et al., 1982; Bolla,
Lindgren, Bonaccorsy, et al., 1991). However, cognitive tests used in previous studies
measured memory functions at only one point in time and were compared to normative
data. Subjective measures of memory such as the MAC-S ask the participant about their
memory decline and general memory functioning (e.g., How would you describe your
memory to the best it has ever been). Therefore, to make accurate assessments we must
also examine decline in objective memory performance when examining the relationship
between complaints and performance. This study used the American National Adult
Reading Test , to provide an estimate of each participant’s premorbid memory ability
(highest level of memory functioning during their lifetime). The discrepancy between how

the participant performed at the time of testing and how they should perform based on

34

their premorbid IQ was calculated. A positive relationship existed between memory
complaints and memory test performance when expected level of performance was
considered. Greater discrepancies between current functioning and expected levels based
on premorbid IQ levels predictive of more memory complaints.

Christensen (1991) examined whether memory complaints were greater in those
with discrepancies between predicted and obtained memory scores compared to those
without discrepancies and found a ‘legitimate’ basis for memory complaints. However,
only six participants had memory scores more than one standard deviation below that
which would be expected based on their premorbid ability. Jorrn et al. (1994) reported that
resulting differences scores did not increase the overall correlations between objective
memory performance and subjective complaints. Because Christensen’s (1991) small
sample and the contradictory ﬁndings of Jorm et al. (1994), it was important to investigate
further the inﬂuence of premorbid functioning.

J orm et al. (1994) suggested that the cognitive test used in their study ( the Mini-
Mental State Examination, MMSE) may be a relatively blunt instrument for detecting early
decline which may be noticeable to the individuals affected. This study used the Rivermead
Behavioral Memory Test. The Rivermead was developed speciﬁcally to detect impairment
in everyday memory functioning by providing analogues to everyday memory situations
(Wilson, Cockbum, & Baddeley, 1985). Hence, the instruments used in this study are
more reﬂective of memory decline in the early stages. Contrary to previous assertions, the

present ﬁndings suggest a relationship between objective test performance and subjective

35

complaints can be demonstrated when a consideration is given to the extent to which

manory has declined over the lifespan.

sm' c Types of Complaints

It was hypothesized that certain types of memory complaints would be predictive
of memory performance, while others would not. This hypothesis was not conﬁrmed.
Larrabee, West and Crook (1991) conducted a study addressing the relationship of self-
reported memory using the MAC-S (Memory Assessment Clinics Self-Report; Winterling,
Crook, Salama, and Grobert, 1986), self-rated depression and actual performance on
computer simulated everyday memory tasks. They found that the MAC-S Ability factors:
remote, everyday and semantic memory and Frequency factors: concentration and
forgetfulness were predictive of objective performance. The present study addressed the
association between performance on the Rivermead Behavioral Memory Test (erson,
Cockburn, Baddley, & Hiorns, 1989) a standard measure of everyday memory ﬁrnctioning
and the MAC-S self-rating scores in normal healthy adults. When all ten factor scores (as
proposed by Winterling et al., 1986) were entered into a regression analysis, only
complaints of Spatial Memory problems signiﬁcantly predictive objective performance and
the overall model did not ﬁt the data.

Larrabee, West and Crook (1991) examined the relationship between self reported
memory complaints and objective performance on a computer-sirnulated memory battery

which was developed by Crook and Larrabee (1988). The computer-sirnulated memory

36

battery was designed to address the problem that standard test procedures bear little
resemblance to self-report questionnaires items which usually evaluate everyday memory
(e.g., remembering names and faces). However, this battery was designed to be
isomorphic to the factors on the MAC-S. Hence, we would expect the two measures to be
highly correlated. It was therefore important to examine the relationship of MAC-S and
the Rivermead Behavioral Memory Test, a standard measure of everyday memory. The 10
individual factors as proposed were not predictive of memory performance on the
Rivermead Behavioral Memory Test.

This difference in ﬁndings between the present study and the one conducted by
Crook et a1. (1991) can be explained by the fact that factor structure proposed by
Winterling et al., (1986) did not hold in this sample. When a principal components factor
analysis was conducted for this sample the items on the MAC-S clustered around one
factor. There are several possible explanations. First, there were some diﬁ‘erences in the
populations sampled. The mean age of the standardization sample was 56.21, with a range
of 18 to 92 years. The sample used in this study was signiﬁcantly older M = 67 .6 ; S_D =
8.8). The standardization sample also contained a subset of participant who were formally
diagnosed with Age Associated Memory Impairment who had both memory complaints
and mild memory decline; all participants in this study were community dwelling able
elderly without remarkable impairment. When Winterling et al. (1986) developed the
factor structure of the MAC-S they eliminated items that were highly loaded on more than

one factor, however they did not eliminate these items from the scale. The items that were

37

eliminated from the factor model should have been eliminated from the scale and the scale

should have been re-normed on a new population.

Treatment

It was predicted that participation in Mood and Memory Workshops would
reduce memory complaints in older adults. This hypothesis was conﬁrmed. Previous
research has suggested that memory training can improve memory performance but not
on memory complaints (e. g., Scogin, Storandt, & Lott, 1985; Scogin & Bienias, 1988).
These studies examined the effectiveness of a self-taught programs for older adults who
had complained of memory problems. The treatment did not try to moderate mood, did
not have a facilitator to increase motivation, and did not occur in a group format. As
suggested by this study and many others, emotional status greatly inﬂuences subjective
memory complaints. Social interaction can also enhance positive mood.

The present training program taught strategies to facilitate memory while
addressing the elderly expectations about memory performance. Participants received
educational instruction and discussion of memory loss and its association to aging. The
purpose of this part of the training was to discourage negative stereotypes about aging and
provide the participants with more positive expectancies of the aging process. Participants
were also taught several mnemonic strategies which are expected to improve performance.
Using this type of training participants experienced both a statistically and clinically

signiﬁcant reduction in memory complaints.

38

These ﬁndings indicate that changing expectations concerning cognitive capabilities
eﬁ‘ectively decreases complaints about memory and suggests that the most effective
memory training will be those that address both memory performance and aﬁ'ect. There
were a few participants in this study who experienced an increase in memory complaints
after participation in the memory training groups. It is possible that these individuals were
experiencing stress from outside factors. This stress might have altered their mood and

subsequently resulted in more memory complaints.

Limitations of Study

Some limitations of this study should be noted. First, the population is somewhat
exceptional, and the generalizability of the results may be limited accordingly. Our
participants were better educated than the national average and many were drawn ﬁ'om
higher socioeconomic levels, and were generally in good physical health. These individuals
also likely had access to high quality health care. Such factors could play in the role of
preserving memory functioning and decreasing memory complaints. However, it could be
argued that that this group actually provides a purer test of the relationship between
memory complaints and objective performance because the potential inﬂuences of low
education, chronic health problems, and so forth were reduced.

This study also lacked a control group. This study lacked participants who
completed both the pre and the post test without participating in the mood and memory

training groups. This questions the effects of participation in the group. The ﬁndings of

39

this study suggest that memory complaints were reduced because of participation in the
memory group. However, we can’t rule out the possibility that other factors inﬂuenced the
change in memory complaints.

Broader Implications

Memory complaints were found to be predictive of objective performance.
However, it is important to note that complaints only accounted for 3 percent of the
variance in objective performance. These ﬁnding support previous ﬁndings: subjective
memory complaints are not good predictors of objective performance. However, the
importance of assessing premorbid ability was demonstrated.

As proposed, diagnostic criteria for Age Related Cognitive Decline are dependent
upon the presence of subjective complaints in addition to memory impairment. The utility
of these criteria have been questioned due to the controversial relationship between
subjective memory complaints and objective performance. Under the criteria proposed it is
assumed that the individual fell within the average range of memory functioning when he
or she was a younger adult. Obviously, 16% of older adults fell above the normal range as
younger adults. Thus an individual with superior ability as a younger adult might have
experience signiﬁcant loss that would not be detected by standard measures, of everyday
memory. This study suggests that when making diagnosis and treatment decisions it is
essential that we take into account premorbid functioning since current objective measure

of memory functioning do not seem to provide accurate estimates of decline.

40

One consideration that must be addressed is if the magnitude of deﬁcit seen in Age
related Cognitive Decline warrants treatment. Prevalence rates of 34.9% and 55.8% have
been reported in random samples of elderly adults (Lane & Snowdon, 1989; Reinikainen
et al., 1990). This suggests that it is “normal” for older adults to experience some decline
in their memory performance and complain about their memory ability. When older adults
appear for treatment and complain that they are no longer able to read because they can
not see, it would not occur to a clinician to inform them that their problems are normal
and no worse than their same aged peers and hence do not warrant treatment (Crook,
1993). Therefore, despite the prevalence and mild severity (as claimed by some) of Age
Associated Cognitive Decline, it is important that health care workers take an active role
in developing treatment. Age Associated Cognitive Decline is associated with both
memory complaints and impaired memory performance. It is important that the treatments
devised address both issues. The results of this study suggest that mood and memory
training groups which include discussions of memory loss and its association to aging
while participants received specialized training in mnemonic techniques can signiﬁcantly
decrease memory complaints. These groups are cost eﬁ‘ective, as one clinician can lead a
group with six to eight participants. Future studies should analysis examine the eﬁ’ect of
these groups on actual performance.

Directions for Future Resggh

Because the ﬁndings of this study suggest that memory complaints are

representative of decline based on premorbid intelligence, it is essential that a longitudinal

study be conducted to track the relationship between objective memory performance and

41

subjective complaints. If a positive relationship is found between memory decline and
memory complaints over time it would further support the importance of using premorbid

estimates of memory ability when making treatment decisions.

LIST OF REFERENCES

LIST OF REFERENCES
Beck, A. T., Ward , C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An
inventory of measuring depression. Achhives of General Psychimi, 561-571.
Best, D. L., Hamlett, K. W. & Davis, S. W. (1992). Memory complaints and
memory performance in the elderly: The eﬁ‘ects of memory-skills training and expectancy

change. Applied Cogm'tive Psychology,_6, 405-416.

Bolla, K. I., Lindgren, K. N., Bonaccorsy, C. & Bleeker. (1991). Memory

 

complaints in older adults: Fact or ﬁction? Archives ofNeurologLﬁ, 61-64.

Bolla-Wilson, K., Bleeker, M. L., (1986). The inﬂuence of verbal-intelligence,
gender, age and education on the Rey Auditory Verbal Learning Test. Neuropgychology,
2, 203-211.

Caine, E. D. (1993). Should aging-associated cognitive decline be included in the
DSM TV? J oumal ofNeuropsychgtry and Clinical NeurosciencesLS, 1-5.

Cerella, J ., Poon, L. W., & Fozard, J. L. (1982) Age and iconic read-out. M
of Gerontology. 37,197- 202.

Christensen, H. (1991). The validity of memory complaints by elderly persons.
International Journal of Geri_alric Psychiatry, 6. 307-312.

Collins, M. W., Abeles, N. (1996). Subjective memory complaints in the able

elderly. Clinical Gerontolagist, 16, 29-54.

42

43

Crook, T., Bartus, R T., Ferris S. H. (1986). Age-associated memory impairment:
proposed diagnostic criteria and measures of clinical change-report of a National Institute
of Mental Health work group. Developmental Neuropgchology, 2, 261-276.

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

Cutler, S. J. & Grams, A. E. (1989). Correlates of self-report everyday memory
problems. Journal of Gerontologyﬂ, 582-590.

Derouesne, C., Dealberto, M. J ., Boyer, P., Lubin, 8., Sauron, B., Piette, F.,
Kohler, F., Alperovitch, A (1993). Empirical evaluation of the cognitive diﬁculties scale
for assessment of memory complaints in general practice: a study of 1628 cognitively
normal subjects aged 45-75 years. International Journal of Geriatric Pﬂchiatry, 8, 599-
607.

Gagon, M., Dartigues, J. F., Mazaux, J. M., Dequae, L., Letenneur, L., Giroire, J.
M., Barberger-Gateau, P. (1994). Self-reported memory complaints and memory
performance in elderly French community residence: Results of the PAQUID research
program. Neuroepidemiology, 13, 145-154.

Gallagher, D., Nies, G. & Thompson, L. (1982). Reliability of the Beck
Depression Inventory with older adults. Journal of Clinical and Cpnsulting Psychology,

§_(_), 152-153.

44

Gallagher D., Breckenridge, J ., Steinmetz, J. & Thompson, L. W. (1983). The
Beck Depression Inventory and Research Diagnostic Criteria: Congruence in an older
population. Journal of Clinical_ and Con_sulting Psychology,_5l, 945-946.

Gilewski, M. J. & Zelinski, E. M.(1986). Questionnaire assessment of memory
complaints. In L. W. Poon , T. Crook. K. L. Davis, C. Eisdorfer, B. J. Garland, A W.
Kasznaik, & L. W. Thompson (Eds), M Memory Assessment of Older Adults (pp.
93 - 107). Washington, DC: American Psychological Association).

Hanninen, T., Reinikainen, K., Helkala, E., Koivisto, K., Mykkanen, L., Laakso,
M., Pyorala, K., Riekkinen, P. (1994). Subjective Memory Complaints and Personality
Traits in Normal Elderly Subjects. Journal of the Amen'ean Gelatric Society. 42:1-4.

Herrmann, D. (1982). Know thy memory: The use of questionnaires to assess and
study memory. Pychological Bulletip, 92, 432-452.

Jones, J. D., A M. (1995). Memory Complaints, depression, and memory
functions in the able elderly. Unpublished master’s thesis, Michigan State University.

Jorrn, A. F ., Christensen, H., Henderson, Korten, A. E., Mackinnon, A J. & Scott,
R (1994). Complaints of cognitive decline in the elderly: A comparison of reports by
subjects and informants in a community survey. Psychological Medicine, 24, 365 -3 74.

Kahn, R. L., Zarit, S. H., Hilbert, N. M. & Niederehe, M. A. (1975). Memory
complaints and impairment in the aged: The effects of depression and altered brain

ﬁrnction. Archives of General Psvchiatg , 117. 326-328.

45

Lamberty, G. J., Bieliauskas, L. A, Chatel, D. M., & Holt, C. S.(l993). Eﬁ‘ects of
demographic factors and mental status on wechsler memory scale performance in a
geriatric clinic sample. Developmental Neuropsychology, 9, 271-281.

Lane. F. & Snowdon. J. (1989). Memory and dementia: A longitudinal survey of
suburban elderly. In: Lovibond. P. & Wilson. P. (Eds), Clinical and Abnormal
Paychology. Elsevier. Amsterdam. 365-376.

La Rue, A. (1992). Agm g and Neuropgchological Assessment. New York:
Plenum Press.

La Rue, A., Swan, G., Carmelli. D. (1995). Cognition and depression in a cohort
of aging men: results ﬁ'om the western collaborative group study. Pyohology and Aging,
10, 30-33.

Lewinsohn, P., Antonuccio, D., Breckenridge, J. S., & Teri, L. (1984). Tag
Coping with Depression Course. Eugene, OR: Castalia Publishing Company.

Larrabee, G., West, R., & Crook, 1991. The Association of memory complaints
with computer-sirnulated everyday memory performance. Journal of CHM
Experimental Neuropychology, 13, 466- 478.

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

Miller, W. R., & Seligrnan, M. E. P. (1973). Depression and the perceptions of

reinforcement. Journal of Abnormal Pachology, 82, 62-73.

46

Mischel, W. (1969). Personality and assessment . New York: Wiley.

Nelson, H. E. (1982). Nelson Adult Reading Test Manual. London: National
Hospital for Nervous Diseases.

Pettinati, H. M., Magee, M., & Mathisen, K. S. (1984). Self-reports about memog
M severity of depression in older adults. Paper presented at the annual convention of the
American Psychological Association, Toronto, Canada.

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

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

Rabbitt, P.& Goward, L. (1986). Effects of age and raw IQ scores on mean
correct and mean error reaction times in serial choice tasks: A reply to Smith and Brewer.
BritiﬁJoumaﬂf Psychology. 77. 69-73.

Reinikainen, K. J ., Koivisto, K., Mykkanen, L., Hanninen, T., Laakso, M., Pyorala,
K., & Riekkinen, P. (1990). Age-Associated Memory Impairment in aged population: An
epidemiological study. Neurology. 40. 177.

Reisberg, Barry, Ferris, Steven, Borenstein, Jemey, Sinaiko, Eli, de Leon, Mony
J ., & Buttinger, Catherine(1986). Questionnaire assessment of memory complaints. In L.

W. Poon , T. Crook. K. L. Davis, C. Eisdorfer, B. J. Gurland, A. W. Kasznaik, & L. W.

47

Thompson (Eds). Assasament of Presenting Syraptoms (pp. 108 - 128). Washington, DC:
American Psychological Association).

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

Schlosser, D. & Ivison, D. (1989). Assessing memory deterioration with the
Wechsler Memory Scale, the National Adult Reading Test, and Schonell Graded Word
Reading Test. Journal of Clinical Experimental Neuropggchology, 11, 785- 792.

Schwartz, M., & Saffran, E. (1987). The American-Nan: Replication and
extension of the British ﬁndings on the persistence of word pronunciation skills in patients
with dementia, unpublished manuscript, Philadelphia, PA

Scogin, F ., & Bienias, J. (1988). A three-year follow-up of older adult participants
in a memory-skills training program. Paychology and Aging, 3, 334-337.

Scogin, F., Storandt, M. & Lott, L. (1985). Memory-skills training, memory
complaints, and depression in older adults. Journal of Gerontology, 40, 562-568.

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

Storandt, M., Botwinick, J. & Danzinger, W. L. (1986). Longitudinal changes:
Patients with mild SDAT and matched with healthy controls. In L. W. Poon (Eds),
Mbook for Cling] Memory Assessment of Older Adults (pp. 277-284). Washington,

DC: American Psychological Association.

48

Tranel, D., Benton, A, & Olson, K. (1997). A lO-Year Longitudinal Study of
Cognitive Changes in Elderly Persons. Developmental Neuropychology, 13, 87 - 96.

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

Wilson, B., Cockburn, J ., Baddley, A & Hiorns, R (1989). The development and
validation of a test battery for detecting and monitoring everyday memory problems.
Journal of Clinica_l__ and Experimental Neuropsychom 855-870.

Winterling, D., Crook, T., Salama, M. & Gobert, J. (1986). A self-rating scale for
assessing memory loss, In A. Bes, J. Cahn, S. Hoyer, J. P. Marc-Vergnes & H. M.
Wisniewski (Eds), Senile Dementias: Early Detection (pp. 482-486). London: John
Libbey Eurotext.

Yesavage, J., Brink, T., Rose, T., Lum, 0., Adey, V., & Leirer, V. (1983).
Development and validation of the geriatric depression screening scale: A preliminary
report. Journal of Psychiatric Research. 17. 37-49.

Zarit, S. H. (1980). Ag'mgand mental disorders. Free Press, New York.

Zelinski, E. M., M. J., Gilewski, & K. W., Schaie.(1993). Individual differences in

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

chhology and Aging, 8, 176-186.

HICHIGQN STRTE UNIV. LIBRRRIES
lllllllllllllllllllllllllllllHWilllllllllllllllllllll
31293015703188