Aging & Mental Health
Vol. 13, No. 5, September 2009, 715–724
Race differences in emotional adaptation of family caregivers
Kimberly A. Skarupskiab*, Judy J. McCannac, Julia L. Bieniasab and Denis A. Evansab
aRush Institute for Healthy Aging, Rush University Medical Center, Chicago, IL, USA; bDepartment of Internal Medicine,
Rush University Medical Center, Chicago, IL, USA; cRush University College of Nursing,
Rush University Medical Center, Chicago, IL, USA
(Received 27 January 2009; final version received 9 February 2009)
Objectives: This study compares black and white caregivers of persons with Alzheimer’s disease (AD) on two
general measures of negative and positive emotion (depressive symptoms, positive mood) and two caregiving
specific measures of negative and positive emotion (caregiver burden, caregiver satisfaction). We hypothesized
that black caregivers would exhibit lower levels of negative emotion and higher levels of positive emotion over
time than whites.
Method: Three hundred ninety-six caregivers were recruited from the Rush Alzheimer’s Disease Center
in Chicago, Illinois, as part of a longitudinal study of persons with AD. The analyses for this report are based
on data from 307 caregivers who were interviewed quarterly over approximately 4 years from 1999 to 2002,
an average of nine observations per person.
Results: The results showed that black caregivers reported fewer depressive symptoms over time than whites
(^¼0.14, p5 0.01) but this finding was only for those caregivers living with the care recipient with AD.
No race differences were found for measures of positive emotion.
Conclusion: Our data add to the growing body of evidence that blacks have better emotional outcomes when
exposed to the stress of providing informal care to a disabled family member.
Keywords: caregiver affect; psychological adjustment; race/ethnicity; Alzheimer’s disease
Introduction
The debilitating effects of Alzheimer’s disease (AD)
and other late-life dementia syndromes eventually
require assistance with most activities of daily living,
supervision, and possible legal guardianship. These
demands are sources of substantial chronic stress
as indicated by the high rate of negative emotional
states and psychiatric morbidity for family caregivers
(Schulz, O’Brien, Bookwala, & Fleissner, 1995). There
is evidence that substantial racial differences may
exist in caregiver adjustment with better adjustment
for blacks. Despite a number of studies that have
shown better psychological adjustments among black
caregivers, it is not well understood what explains this
apparent advantage.
Psychological adaptation to chronic stress has been
modeled in persons providing care for a chronically
disabled family member, particularly for disabilities
produced by a late-life dementia syndrome. These
observational studies have consistently shown that
family dementia caregivers are likely to report clini-
cally significant levels of negative emotion (Haley
et al., 1995, 2004; Lawton, Rajagopal, Brody, &
Kleban, 1992; Miller, Campbell, Farran, Kaufman, &
Davis, 1995; Schulz et al., 1995) and reduced fre-
quency of positive emotion (Haley et al., 2004; Lawton
et al., 1992; Roff et al., 2004; White, Townsend, &
Stephens, 2000).
Among basic demographic variables, race has
emerged as an important predictor of caregiver
psychological adaptation (Connell & Gibson, 1997).
Blacks are less likely than whites to report symptoms
of emotional distress in the context of providing daily
care for a family member with dementia (Farran,
Miller, Kaufman, & Davis, 1997; Fredman, Daly, &
Lazur, 1995; Haley et al., 1995, 1996, 2004; Lawton
et al., 1992; Miller et al., 1995; Roth, Haley, Owen,
Clay, & Goode, 2001; White, Townsend, & Stephens,
2000). Similarly, blacks are more likely than whites to
report positive emotional experiences and perceptions
of caregiving experiences (Haley et al., 2004; Lawton
et al., 1992; Roff et al., 2004). Although the consistency
of better psychological outcomes found in black
caregivers is impressive, much of the evidence is
based on cross-sectional studies.
In one of the few published longitudinal studies of
dementia caregivers, Roth et al. (2001) found evidence
of better psychological adaptation in blacks over time
on measures of positive and negative emotional
function, life satisfaction, and depressive symptoms;
however, the temporal pattern varied between mea-
sures. Racial differences in life satisfaction increased
over a 2-year period, primarily due to a decline
in scores for white caregivers. By contrast, depressive
symptoms remained stable over time in both racial
subgroups, with higher levels reported by white
*Corresponding author. Email: kimberly_skarupski@rush.edu
ISSN 1360–7863 print/ISSN 1364–6915 online
 2009 Taylor & Francis
DOI: 10.1080/13607860902845582
http://www.informaworld.com

caregivers at each of three measurement points. Thus,
positive emotion may provide a more sensitive indica-
tor of racial differences in caregiver psychological
adaptation.
To further evaluate racial differences in caregiver
emotional adaptation, this report will present data
from a 4-year longitudinal study of family caregivers
for persons with AD. A major strength of the study
design was that caregiver’s emotional function was
measured at 3-month intervals over approximately 4
years to characterize individual differences. Previous
longitudinal studies of dementia caregivers have typ-
ically relied on three to five observation points to
estimate rates of change, potentially limiting
precision. AD symptom severity was also sequentially
measured over the course of this study to character-
ize dementia-specific care-related stressors over time.
Based on previous reports, black caregivers were
expected to exhibit better emotional adaptation
across study time than white caregivers.
A secondary objective of this report is to examine
two hypotheses that have been postulated to account
for racial differences in caregiver psychological adap-
tation – higher levels of religiosity and more supportive
social environments in blacks than whites (Dilworth-
Anderson, Williams, & Gibson, 2002). As secondary
objectives, the present study will evaluate religiosity
and three features of the social environment – living
arrangements, quality of the past relationship between
caregiver and AD participant, and quality of caregiver
social support – as moderator variables with respect to
any observed racial differences in caregiver emotional
function.
Methods
Participants
Caregivers in the present study were recruited as part
of a longitudinal study of 396 persons with AD. This
AD cohort was recruited from the Rush Alzheimer’s
Disease Center (June 1999–April 2002) and lived
within the Chicago metropolitan area. Eligibility
required that each participant should complete a
standardized clinical evaluation (Gilley, McCann,
Bienias, & Evans, 2005), meet diagnostic criteria for
possible or probable AD (McKhann et al., 1984), and
reside in a community setting. Community residence
encompassed a wide range of private and public
housing arrangements, excluding only institutional
facilities providing assisted living or nursing care.
All data collection procedures were presented verbally
by study representatives and specified in writing
via consent documents. As an added precaution, AD
participants and a responsible family member jointly
signed consent documents. The Institutional Review
Board of Rush University Medical Center approved
consent procedures for AD and caregiver participants.
Caregiver participants included for the purpose of
this current report met three criteria: (a) identified
himself or herself as the primary family member
responsible for the AD participant’s care; (b) was the
adult child or spouse of the AD participant; and (c)
completed three or more data collection interviews.
Thirty-four caregivers who were not related by blood
or marriage to the AD participant were excluded and
55 caregivers were excluded who had less than three
data collection interviews primarily because of institu-
tionalization (n¼ 26) or death (n¼ 15) of the AD
participant. A total of 307 caregivers met the three
criteria. Of the 307 caregiver participants, 234 lived
with the AD participant, including 138 spouse and
96 adult child caregivers. The remaining 73 caregivers
(all adult children) lived in the area, but not with the
AD participant.
Measures
Caregiver emotional function
Four measures of emotional function were obtained
using structured interviews, conducted either in-person
or by telephone at baseline and at 3-month intervals
during the study. These measures crossed two content
dimensions (general emotion, emotional appraisal of
caregiver experiences) with two valence dimensions
(positive, negative). General measures of emotion
require caregivers to describe current levels of negative
and positive emotion without reference to care-related
stresses. The second set of measures covered negative
and positive appraisals of caregiving-specific experi-
ences and responsibilities. Because there is evidence
in both the general and caregiver populations that
negative- and positive-valence responses are largely
independent patterns of emotion (Lawton, Moss,
Kleban, Glicksman, & Rovine, 1991; Watson, Clark,
& Tellegen, 1988), the valence dimension was system-
atically incorporated into the measurement design.
Negative emotion
General negative emotion was assessed using a mea-
sure of depressive symptoms. Depressive symptoms
have been widely used to document clinically signifi-
cant levels of negative emotion in dementia caregivers
(Schulz et al., 1995). For this study, we used the
10-item version of the Center for Epidemiologic
Studies Depression scale (CES-D) (Kohout,
Berkman, Evans, & Cornoni-Huntley, 1993) to mea-
sure depressive symptoms. Items were rated as present
or absent during the past week. Total scores range
from 0 to 10, with higher scores indicating greater
depressive symptoms. The internal consistency of the
CES-D was examined in the total study population
(Cronbach’s ¼ 0.75).
Negative emotion specific to caregiving experiences
was measured using a 10-item measure of caregiver
burden (Burden) (Lawton et al., 1991). This measure
uses a 4-point ordinal response scale and asks the
respondent to report how often in the past week they
716 K.A. Skarupski et al.