Perceived Stress and Cognitive Decline in Different Cognitive
Domains in a Cohort of Older African Americans
Arlener D. Turner, Ph.D., Bryan D. James, Ph.D., Ana W. Capuano, Ph.D., Neelum T.
Aggarwal, M.D., and Lisa L. Barnes, Ph.D.
Rush Alzheimer's Disease Center, Rush University Medical Center, Chicago, IL, USA
Abstract
Background—Research indicates that stress is linked to cognitive dysfunction. However, few
community-based studies have explored the relationship between perceived stress and cognitive
decline, and fewer still have employed cognitive domains rather than a global measure of
cognition.
Objective—We examined the relation between perceived stress and the rate of decline in
different cognitive domains.
Method—Participants were older African Americans without dementia from the Minority Aging
Research Study (MARS; N=467, mean age: 73 SD=6.1). A battery of 19 cognitive tests was
administered at baseline and at annual intervals for up to 9 years (mean follow up=4 years), from
which composite measures of global cognitive function and five specific cognitive domains were
derived. The 4-item Cohen's Perceived Stress Scale (PSS) was also administered at baseline.
Results—In linear mixed-effects models adjusted for age, sex, education, and vascular risk
factors, higher perceived stress was related to faster declines in global cognition (β=−0.019;
SE=0.008; t
(1951)
=−2.46), episodic memory (β=−0.022; SE=0.011; t
(1954)
=−1.99), and
visuospatial ability (β=−0.021; SE=0.009; t
(1939)
=−2.38) all
p
< .05. Findings were similar in
subsequent models adjusted for demographics, vascular diseases, and depressive symptoms.
Conclusions—Results indicate that older African Americans with higher levels of perceived
stress have more rapid declines in global cognition than those with lower levels, most notably for
episodic memory and visuospatial ability.
Keywords
African Americans; Cognition; Cognitive Domains; Cognitive Decline; Perceived Stress
Corresponding Author: Arlener D. Turner, Ph.D., Rush Alzheimer's Disease Center and Department of Behavioral Sciences, Rush
University Medical Center, 710 S. Paulina, Suite 600, Chicago, IL, 60612, Phone: 312.563.4292, Fax: 312.942.8961,
Arlener_Turner@rush.edu.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our
customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of
the resulting proof before it is published in its final citable form. Please note that during the production process errors may be
discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Conflicts of Interest
There are no relationships, financial or otherwise, that could be considered conflicts of interests affecting this manuscript. No
Disclosures to Report.
HHS Public Access
Author manuscript
Am J Geriatr Psychiatry
. Author manuscript; available in PMC 2018 January 01.
Published in final edited form as:
Am J Geriatr Psychiatry
. 2017 January ; 25(1): 25–34. doi:10.1016/j.jagp.2016.10.003.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
INTRODUCTION
It has been well documented that stress, defined as a consequence of environmental events or
demands (stressors) that exceed an individual's perceived ability to cope
(1)
, can affect a
person's physical and psychological health. Research has described the detrimental effects of
chronic stress on psychological well-being and cognitive functioning, with an emphasis on
the relationship between stress and memory
(2)
. However, much of the evidence regarding the
relation between stress and cognition comes from clinical observations and basic
neuroscience studies involving biological measures of the physiological stress response
(3)
.
Few community-based studies have explored the relationship between the perception of
stress, which precedes the physiological response, and cognition, and fewer still have
examined the full range of cognition rather than memory alone. The focus on memory in
stress and cognition research stems from the fact that both animal and human studies show
that psychosocial stress can lead to a spectrum of cellular changes in the hippocampus, a
part of the limbic system that is important for declarative memory processes
(4)
. The cellular
changes seen in the hippocampus occur, at least in part, because of a high number of
glucocorticoid/cortisol (also known as the stress hormone) receptors in this area
(5-7)
.
However, glucocorticoid/cortisol receptors also exist in the prefrontal cortex, suggesting the
possibility that psychosocial stress could lead to deficits in executive functioning and
attention as well
(8,9)
.
Research also suggests that chronic stress is associated with inflammatory and hormonal
indicators of accelerated aging
(10)
, with reports of greater perceived stress levels increasing
the possibility of brain infarcts, reduced brain volume
(11)
, and elevated risk of stroke
(12,13)
.
Consequently, these same physiological markers of accelerated aging are also associated
with rates of decline in cognition. Taken together, these different lines of research suggest
that perceived stress may be related to cognitive decline by causing cellular changes within
the hippocampus or via an association with inflammatory and hormonal markers of brain
aging. We are aware of only one population-based study that examined perceived stress and
change in cognition
(14)
. While this study confirms a relationship between perceived stress
and cognitive decline, they used only a brief global measure of cognitive function consisting
of four tests. Thus, the potential relationship between perceived stress and
specific
declines
in memory and attention, as suggested by previous literature, is still an unanswered question.
Previous research has suggested that African Americans experience a disproportionate
burden of stressful life experiences linked to poverty, racism, and residential segregation,
due, in part, to their relative low social status in US society
(15-18)
. For example, African
Americans are more likely to experience stressful life conditions such as unfair treatment,
environmental stressors such as impoverished neighborhoods, and limited job
opportunities
(19,20)
. Yet most research on psychosocial stress and health outcomes has been
conducted in the majority white population. In addition, several studies have documented
high rates of cognitive impairment in older African Americans
(21)
. Examination of the link
between perceived stress and cognition in a population characterized by societal inequalities
and high rates of cognitive impairment
(16,18,21)
would represent an important step in
delineating whether psychosocial stress may be a modifiable risk factor for cognitive
decline. Therefore, we examined the relation between perceived stress and cognitive decline
Turner et al.
Page 2
Am J Geriatr Psychiatry
. Author manuscript; available in PMC 2018 January 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
in a sample of community-dwelling older African American adults. We hypothesized that
higher levels of perceived stress would be related to faster rates of cognitive decline,
particularly in the well documented area of hippocampus mediated tasks such as memory (as
measured by episodic memory specifically) and the projected areas of prefrontal cortex
mediated tasks such as executive function and attention (as measured by perceptual speed
specifically).
METHODS
Participants
All the participants were enrolled in the Minority Aging Research Study (MARS) and had
signed an informed consent form in which they agreed to annual clinical evaluations that
included an assessment of lifestyle and psychosocial factors (e.g., depressive symptoms and
perceived stress) and cognitive testing, as previously described
22
. MARS is a longitudinal
epidemiological cohort study of risk factors for cognitive decline and Alzheimer's disease in
African-Americans, approved by the Rush University Medical Center Institutional Review
Board. The participants were older community-dwelling adults, without dementia, recruited
from churches, community-based organizations, senior-subsidized housing facilities in the
Chicagoland area, and through the Clinical Core of the Rush Alzheimer's Disease Center.
Community presentations were held throughout the city on aging and Alzheimer's disease in
African Americans during which a description of the study and eligibility were discussed.
Interested persons were asked to complete a form describing their level of interest in the
study, and were contacted later to have questions answered and obtain further information on
the study. At the time of the analysis, 602 older adults had enrolled and completed a baseline
evaluation. Follow up time ranged from 1-9 years (approximately 44% of individuals had
2-4 years follow up time, 24% had 5-7 years and 32% had 8-9 years) with an average of 4.0
(SD=3.52) years. Because we are interested in cognitive change, we excluded 116
individuals who had not yet completed at least two clinical evaluations. Based on clinical
evaluations conducted at baseline, we excluded another 19 individuals who met criteria for
dementia. The remaining 467 participants were included in these analyses.
Clinical evaluation
All participants agreed to annual uniform structured clinical evaluations. These included a
complete neurological examination, a detailed medical history, and neuropsychological
testing (described below). An experienced clinician classified the participants with respect to
dementia diagnosis. Dementia classifications were based on the criteria of the joint working
group of the National Institute of Neurological and Communicative Disorders and Stroke
and the Alzheimer's Disease and Related Disorders Association
(23)
.
Neuropsychological testing
Annual neuropsychological testing consisted of detailed tests selected to assess a broad
range of cognitive abilities commonly affected in older adults. It included the Mini-Mental
State Examination (MMSE)
(24)
, which was used for descriptive purposes only. As
previously described
(25)
, the remaining tests were used to form summary measures of five
cognitive domains (episodic memory, semantic memory, working memory, perceptual speed,
Turner et al.
Page 3
Am J Geriatr Psychiatry
. Author manuscript; available in PMC 2018 January 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
and visuospatial ability) and global cognition. The measures included in each of the
cognitive domains are as follows: episodic memory includes Word List Memory, Word List
Recall, and Word List Recognition from the procedures established by CERAD
(26)
,
immediate and delayed recall of the East Boston story
(27)
, and, immediate and delayed recall
of Story A from the Logical Memory subtest of the Wechsler Memory Test-R (WMS-R)
(28)
;
semantic memory includes semantic Verbal Fluency from CERAD (animal and fruits/
vegetables)
(26)
and a 15-item version of the Boston Naming Test
(29)
; working memory
includes Digit Span Forward and Backward from the WMS-R
(28)
and Digit Ordering
(30)
;
perceptual speed includes symbol digit modalities
(31)
, number comparison
(32)
, and two
indices from a modified version of the Stroop Neuropsychological Screening Test: the
number of color names correctly read aloud in 30 seconds minus the number of errors and
the number of colors correctly named in 30 seconds minus the number of errors
(33)
;
visuospatial ability includes a 16-tem version of Standard Progressive Matrices
(34)
and a 15-
item version of Judgment of Line Orientation
(35)
. As described previously
(25)
, the composite
scores of the five cognitive domains were created by converting the raw scores of the
measures for each cognitive domain into z scores and then averaging the z scores. The global
cognitive score was derived by averaging the z scores of all measures.
Stress measure
Stress was assessed using Cohen's Perceived Stress Scale (PSS)
(36)
. The PSS is an index of
the degree to which a person finds their lives unpredictable, uncontrollable, and overloading
- characteristics central to the evaluation of stress. Initially Cohen introduced the full length
version (14 items) and a short version (4 items) of the scale, later another short version of
the scale with 10 items was introduced. Validation studies of the English versions of the PSS
conducted by the creators demonstrate acceptable internal consistency across all versions
(Chronbach's alpha = .78, .75, .60 for −14, −10, and −4 item versions respectively)
(37)
. We
used the 4-item version of the measure and asked participants to rate the frequency at which
they perceived these characteristics within the last month. The responses were coded in a
Likert scale format ranging from 0 (never) to 4 (very often), with higher scores indicating
higher levels of perceived stress. Scale items included “In the last month, how often have
you felt.... (1) that you were unable to control the important things in your life?, (2)
confident about your ability to handle your personal problems?, (3) that things were going
your way?, and (4) difficulties were piling up so high that you could not overcome them?”
Covariates
The covariates included age, sex, education (years of formal schooling), a composite score
of vascular risk factors, and depressive symptoms. Vascular risk factors (i.e., the sum of
hypertension, diabetes mellitus, and smoking, resulting in a score from 0 to 3 for each
individual) and vascular disease (i.e., the sum of heart attack, congestive heart failure,
claudication, and stroke resulting in a score from 0 to 4 for each individual) were computed
on the basis of self-report questions and inspection of medications, as previously
described
(38)
. Depressive symptoms were assessed using the 10-item version of the Center
for Epidemiologic Studies Depression Scale (CES-D)
(39)
. The participants were asked
whether they had experienced specific symptoms in the last week. The responses to the
Turner et al.
Page 4
Am J Geriatr Psychiatry
. Author manuscript; available in PMC 2018 January 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript
items on the CES-D are coded in a yes/no format, yielding a summary score ranging from 0–
10.
Data analysis
We computed descriptive statistics for the PSS, including the mean scores of demographic
subgroups and used
t
-tests and one way ANOVAs to compare the mean PSS scores across
the different demographic subgroups. Least square means were compared using Tukey-
Kramer adjustment for multiple comparisons. The relation of stress and cognitive decline
(global cognition and the five separate domains) was examined with linear mixed models
with a random intercept and random slope for time, using restricted maximum likelihood
and an unstructured covariance structure. All the models controlled for age (centered at the
mean), sex, education (centered at the mean), vascular risk factors, and depressive
symptoms. In order to ensure a full examination of vascular effects, subsequent models
adjusted for vascular disease.
All analyses were programmed using SAS/STAT® software version 9.3 for Linux on a
Hewlett Packard DL380 server
(40)
.
RESULTS
Among the 467 African American participants, 75% were women. They had a mean age of
73.4 (SD = 6.12) years, a mean education of 15.0 (SD = 3.46) years, and a mean MMSE
score of 27.9 (SD = 1.91). The mean vascular risk score of the participants was 1.5 (SD =
0.85), and they had a median CES-D score of 1.0 (IQR= 0.0, 2.0) and a mean PSS score of
0.85 (SD = 0.67) (Table 1).
We first examined the mean scores for PSS within demographic subgroups (age: 65-75,
76-85, & 86+; education: <12years, 13-15 years, 16+ years; gender: women, men). In these
bivariate analyses, there was only an association between perceived stress level and years of
education, such that perceived stress was lower in those with higher education. Perceived
stress did not differ as a function of age, or among men and women. (See Table 2).
Next, we constructed mixed-effects models to examine the relation of perceived stress to the
rate of cognitive decline, while adjusting for the effects of age (centered at 73 years), sex,
education (centered at 15 years), vascular risk, and depressive symptoms. Level of perceived
stress was not related to global cognition at baseline, but it was related to global cognitive
decline (Table 3), such that persons who scored higher on the PSS declined in global
cognition significantly faster than those who scored lower on the PSS (See Figure 1). For
example, a typical person with the mean age and education but with a PSS score one point
above the mean was estimated to enter the study with a non-significant 0.005 points lower
cognition and significantly decline by .019 points per year. In a model adding vascular
disease as a covariate the PSS continued to be related to global cognitive decline (data not
shown).
In subsequent mixed-effects models, we examined the relationship between perceived stress
and each of the five cognitive abilities. We repeated the main analysis while adjusting for the
Turner et al.
Page 5
Am J Geriatr Psychiatry
. Author manuscript; available in PMC 2018 January 01.
Author Manuscript Author Manuscript Author Manuscript Author Manuscript