NIH Public Access

Author Manuscript
Am J Geriatr Psychiatry. Author manuscript; available in PMC 2012 November 02.
Published in final edited form as:
Am J Geriatr Psychiatry. 2012 September ; 20(9): 815819. doi:10.1097/JGP.0b013e318235b62f.

$watermark-text

Differences in Depression Severity in Family Caregivers of

Hospitalized Individuals with Dementia and Family Caregivers of
Outpatients with Dementia
Gary Epstein-Lubow, MDa,b, Brandon Gaudiano, PhDa,b, Ellen Darling, BAb, Mathew
Hinckley, BAc, Geoffrey Tremont, PhDa,d, Robert Kohn, MD, MPhila,b, Louis J Marino Jr.,
MDa,b, Stephen Salloway, MD, MSa,b, Rene Grinnell, BAb, and Ivan W Miller, PhDa,b
aThe Warren Alpert Medical School of Brown University; Providence, RI
bButler

Hospital; Providence, RI

cWomen
dRhode

and Infants Hospital; Providence, RI

Island Hospital; Providence, RI

Abstract
$watermark-text

ObjectivesTo determine if family caregivers of hospitalized individuals with dementia exhibit

greater depression severity than caregivers of outpatients.
MethodsCaregivers were recruited during care-recipient treatment. Measures assessed
depression, stress, burden and grief.

$watermark-text

ResultsForty-one caregivers of a hospitalized patient and 44 caregivers of an outpatient (total

N = 85) were recruited. The groups did not differ except caregivers of hospitalized patients were
younger and less likely to reside with the care- recipient. Regarding depression, 63.4% of
caregivers of a hospitalized patient and 43.2% of caregivers of an outpatient scored within the
clinical depressive symptoms range. Independent sample t-tests showed that caregivers of a
hospitalized patient had greater severity of depression, burden, and grief. Caregiving for a
hospitalized person remained a significant predictor of greater depression severity in regression
models.
ConclusionsFamily caregiving for a person hospitalized for psychiatric treatment of dementia
is a risk factor for depression.
Keywords
Caregiver; Depression; Dementia; Hospital

Corresponding Author: Gary Epstein-Lubow, MD, Psychosocial Research Program, Butler Hospital; 345 Blackstone Boulevard,
Providence, RI 02906, Telephone: 401-455-6378, Fax: 401-455-6235, [email protected].
Previous Presentation: These data were presented at the Annual Meeting of the American Psychiatric Association in San Francisco,
CA, May 16 21, 2009.
Conflicts of Interest: During a portion of the data collection for this manuscript, Gary Epstein-Lubow, MD served as an unpaid
consultant to a project supported by Forest Research Institute. Robert Kohn, MD has received honoraria from Pfizer, Novartis and
Forest; and research support from the National Institute on Aging and the Health Resources and Services Administration. Stephen
Salloway, MD, MS has served on the scientific advisory board of Elan, Sanofi-Aventis, Pfizer, Eisai, Astra Zeneca and Bristol-Myers
Squibb; received honorarium from Eisai Inc., Pfizer Inc., Novartis, Forest, Elan and Athena Diagnostics as well as data monitoring
from Merck-Serono and Medivation; also, received research support from Elan, Wyeth, Janssen Immunotherapy, Bristol-Myers
Squibb, Eisai, Pfizer, Medivation, Myriad, GlaxoSmithKline, Neurochem, Cephalon, Bayer, Forest, Alzheimers Disease
Neuroimaging Initiative, Dominantly Inherited Alzheimers Network, The Alzheimers Association, The Normal and Rosalie Fain
Family Foundation, the John and Happy White Foundation and the Champlin Foundation.

Epstein-Lubow et al.

Page 2

Objective

$watermark-text

Family caregivers of ill elderly are vulnerable to physical and mental health problems
including depression.1 There are few published reports regarding family caregiving in the
context of severe behavioral problems that result in hospitalization for dementia. Such
studies are important because increased dementia symptom severity may lead to greater
caregiver burden2 and depression. Similarly, after hospitalization, caregivers of individuals
with cognitive impairment due to dementia or delirium may require more-intensive aftercare
assistance; yet the types and intensity of service needs are still not known.3 In comparison,
there has been relatively more study of family caregiving during and after hospitalization for
illnesses such as acute stroke,4 cancer, and heart failure.

$watermark-text

It has been reported that family caregivers of individuals with dementia are curious for
information about hospital treatment, particularly during end-of-life care. Beyond this,
regarding psychiatric symptoms in caregivers of hospitalized individuals with dementia, we
were able to identify only one publication which psychometrically assessed caregivers
during hospital treatment for a care- recipient with dementia. Liptzin et al5 in 1998
compared the family burden reports of 11 caregivers of individuals undergoing hospital
treatment for dementia to the burden reported by 38 caregivers of elders hospitalized for
depression. In addition to assessing burden, Liptzin assessed caregivers perceptions of
behavioral problems; assessments were conducted prior to admission and at a two-month
follow-up. Caregiver depression was not measured. Caregiver burden did not change over
time for either group, although there was a reduction in burden for the subsample in which
dementia patients were assessed by their caregiver as clinically improved at discharge.5
The objective of the current study was to evaluate the extent to which dementia family
caregivers report depressive symptoms in the inpatient setting. In the absence of any
previous reports, the current study of a small sample is exploratory; the studys hypothesis
was that family caregivers of individuals hospitalized for treatment of dementia will report
greater depression and other psychiatric symptoms compared with caregivers of nonhospitalized individuals with dementia.

Methods
$watermark-text

This investigation was part of a larger project seeking to find methods to easily and routinely
screen family caregivers for health-related risks.6 The study used a set of short validated
measures to collect data regarding family caregivers self-report of symptoms in several
spheres in addition to depression. Candidate covariates including demographic variables,
perceived stress, burden, and grief were selected based on their relevance in prior dementia
caregiving research. Each of these constructs has been previously evaluated in dementia
caregivers in the outpatient setting and the specific measures used in this study were selected
so that the overall battery could remain short while still capturing diverse symptoms.
Recruitment and Procedures
Participants were recruited from an inpatient geriatric psychiatry unit associated with a freestanding psychiatric hospital and outpatient settings including a memory disorders program
and community support groups for dementia caregivers. Inclusion criteria were age greater
than 18, self-identification as a family caregiver, and ability to read English and complete
psychometric questionnaires. In the instructions for this study, caregiving was defined as
unpaid thinking about, planning activities for, feeding, dressing, or providing other care
for someone with dementia. No data were collected regarding the care- recipient with
dementia other than the caregivers written report of the diagnosis of the care- recipient. In

Am J Geriatr Psychiatry. Author manuscript; available in PMC 2012 November 02.

Epstein-Lubow et al.

Page 3

the inpatient setting, a research assistant reviewed the admitting diagnoses for continuous
admissions over an 8-month period and approached the primary family contact for every
patient admitted with a diagnosis of a dementia; the completed measures were received from
participants during care-recipient hospitalization or soon after hospital discharge. To
participate in the outpatient setting, a caregiver responded to a posted sign or invitation from
a clinician. Participants in both the inpatient and outpatient cohorts completed written
informed consent prior to providing written responses to the questionnaires. All procedures
were approved by the designated Institutional Review Board.
Measures

$watermark-text

The Center for Epidemiological Studies Depression Scale, 10-item version (CES-D), is a
shortened form of the original 20-item CES-D. This rapid screen asks ten Yes or No
questions regarding an individuals symptoms over the past week. Higher scores indicate
greater depressive symptoms. This scale has been effectively used with older adults and is
reliable. Individual items are scored on a four-point Likert scale (0 3) and a total score of
10 or greater indicates depressive symptoms suggestive of a clinical syndrome.
The Perceived Stress Scale, 4-item version (PSS4), is a shortened form of the original 14item self-report measure designed to assess an individuals beliefs about stressful
experiences during the past month. Each item is answered according to a five-point Likert
scale (0 4). The PSS4 is recommended and has adequate reliability (Cronbachs internal
consistency = 0.60) in situations requiring a short assessment.7 The mean score (standard
deviation) for a sample of married people (n = 1427) in the U.S. was 4.2 (SD = 2.8).7

$watermark-text

The Rapid Screen for Caregiver Burden (RSCB)8 is a shortened version of the 25-item
Screen for Caregiver Burden. The RSCB is a seven-item scale; each item is answered
according to a five-point Likert scale. This scale has been effectively used with caregivers
and has been found to have good internal consistency ( = 0.88) and a mean (standard
deviation) of 31.8 (7.6).8
The Inventory of Traumatic Grief, pre-loss version (ITG PL), is based upon Prigersons
Inventory of Complicated Grief. The pre-loss version was developed to assess for the
symptom complex associated with grief during terminal illness and has been shown to be
reliable. Each question is answered according to a five-point Likert scale and a score > 25
suggests impaired social, general, mental and physical health functioning.

$watermark-text

Analyses
Demographic information and the group means for measures of depression, stress, burden
and grief were compared by independent samples t-tests and chi square analyses to
determine group differences. Cohens d effect size estimates were calculated for clinical
variables. Hierarchical multiple regression models were conducted to assess whether
membership in the caregiving-to-an-inpatient group was a unique predictor of outcomes
(e.g., depression severity) after controlling for other group differences in the first step of the
model.

Results
Forty-one caregivers of a hospitalized patient and 44 caregivers of an outpatient (total n =
85) were recruited. The full sample was predominantly female (65%) and white (97%) with
no differences between the inpatient and outpatient groups regarding gender, race, or
resources as shown in Table 1. The two groups did show significant differences regarding
caregiver age, caregiving relationship, and co-habitation status; specifically, caregivers of a
hospitalized patient were significantly younger and less likely to be married to and reside
Am J Geriatr Psychiatry. Author manuscript; available in PMC 2012 November 02.

Epstein-Lubow et al.

Page 4

with the care- recipient. Regarding depression, 63.4% of caregivers of inpatients and 43.2%
of caregivers of outpatients scored within the clinical depressive symptoms range (CESD >
9). Independent sample t-tests showed that caregivers of a hospitalized patient had greater
severity of depression, burden, and grief, with Cohens d effect size estimates ranging from
0.31 to 0.58, also shown in Table 1. In hierarchical regression models, after controlling for
age, co-residence status, and spousal status in the first step, caregiving for a hospitalized
patient emerged as a significant predictor of greater depression severity in the second step of
the model (F = 4.322, dfs = 1, 76, p = .041, R2 = .101, R2 = .051). After controlling for
the same background variables, however, caregiving in the hospital setting did not predict
burden or grief.

$watermark-text

Conclusions
To our knowledge, this is the first report of depressive symptoms in the family caregivers of
individuals hospitalized for treatment of behavioral disturbance in dementia. Nearly twothirds of the family caregivers surveyed in the hospital setting reported acute depressive
symptoms at a level of severity signifying risk for a clinical depressive syndrome. The
participants included fewer spousal caregivers in the hospital setting, suggesting greater
involvement from adult children of hospitalized patients. Our interpretation of these data is
that caregiving for a patient hospitalized for psychiatric treatment of dementia is a risk factor
for depression.

$watermark-text

Alternative interpretations of the data are possible. An unmeasured variable unrelated to

caregiving, such as the witnessing of other patients suffering when a family caregiver visits
a hospital, could explain the differences in depression. Also, the results could be explained
by sampling bias. To address how potential recruitment bias may have impacted the results,
we conducted alternative statistical analyses controlling for spousal relationship and
outpatient recruitment site; the significance of the results regarding greater depression
severity in the hospital setting was unchanged. Regarding the demographic differences
between the study cohorts, we think this is likely related to role functioning in each
environment. If a hospitalized individual has both a spouse and an adult -child caregiver, it
may be that the adult child takes a leadership role. Of note, the finding of fewer spousal
caregivers in the hospital setting is consistent with other reports that the absence of a spouse
increases risk for hospitalization.9

$watermark-text

There are several weaknesses in the current study. The sample sizes are relatively small.
More research is needed regarding epidemiological cross-sectional and longitudinal
assessment of caregivers of individuals hospitalized for dementia. Similarly, the current
studys participants include few individuals from racial and ethnic minorities. This is
representative of the background of the patients admitted for inpatient and outpatient
treatment at the study sites; and, this weakness means that the current data may not
generalize to family caregivers who are not white. Another limitation is that the results do
not include symptom severity data regarding dementia symptoms of the care-recipient;
without this information it is not possible to draw conclusions regarding the relationship
between dementia symptoms and caregiver mood. By definition, care- recipients undergoing
hospital treatment for behavioral disturbance demonstrate greater symptoms related to
dementia than outpatients; nevertheless, future studies should include measures of
neuropsychiatric symptoms as assessed by both the family caregiver and a clinician rater.
Finally, these data are cross-sectional. Without longitudinal follow-up, it is not possible to
determine whether depressive symptoms are likely to remit or worsen; it may be that
heightened symptoms are a time-limited response to the acute stressor of having a parent or
spouse in the hospital.

Am J Geriatr Psychiatry. Author manuscript; available in PMC 2012 November 02.

Epstein-Lubow et al.

Page 5

It is clinically important to consider caregivers risk for depression during the hospital
treatment for a care-recipient with dementia. Family members are inextricably part of patient
care services for dementia and a caregivers functioning and decision-making can impact
patient outcome. The results reported here suggest that being the primary caregiver for an
individual who is hospitalized for dementia treatment is a risk factor for depression in the
caregiver. The regression effect size that included a number of covariates is small in
magnitude; however, the overall estimate of effect size according to Cohens d suggests
medium group differences regarding caregiver depression (.58), burden (.49) and grief (.53).

$watermark-text
$watermark-text

This greater severity of depressive symptoms for caregivers of hospitalized patients, if

substantiated in future studies, is likely clinically significant for several reasons. Forty-three
percent of the caregivers of outpatients in this study reported clinical-level depressive
symptoms. For caregivers of outpatients, there are interventions proven to mitigate
depressive symptoms; however, most clinical trials of such programming excluded
participation from individuals with dementia at imminent risk for hospitalization;1 given
this, it remains an empirical question as to whether currently-existing caregiver treatments
may be successfully used to reduce depressive symptoms in caregivers of hospitalized
individuals. Also, one must consider data from other studies showing that hospitalization of
a spouse places a caregiver at subsequent increased risk for depression10 and death. In this
context, our studys results suggest that depressive symptoms may start or increase during a
care- recipients hospital treatment. If future research does demonstrate this risk to be true,
that caregivers who are depressed during psychiatric hospital treatment of a spouse or parent
with dementia are at risk for subsequent major depressive disorder, then new programs with
innovative treatment designs11 may be necessary in order to assist family caregivers
regarding in-hospital and post-hospital caregiving responsibilities while concurrently
treating or preventing the development of depression.

Acknowledgments
This research was supported by Butler Hospital. Research and administrative assistance were provided by Nomanda
Nongauza, BA and Heather Harrell, BA. Ms. Nongauza received compensation related to employment at Butler
Hospital; Ms. Harrell received no compensation regarding her role with this study.

References
$watermark-text

1. Schulz R, Martire LM. Family caregiving of persons with dementia: prevalence, health effects, and
support strategies. American Journal of Geriatrec Psychiatry. 2004; 12:2409.
2. Mohamed S, Rosenheck R, Lyketsos CG, Schneider LS. Caregiver burden in Alzheimer disease:
cross-sectional and longitudinal patient correlates. American Journal of Geriatric Psychiatry. 2010;
18:91727. [PubMed: 20808108]
3. Naylor MD, Hirschman KB, Bowles KH, Bixby MB, Konick-McMahan J, Stephens C. Care
coordination for cognitively impaired older adults and their caregivers. Home Health Care Services
Quarterly. 2007; 26:5778. [PubMed: 18032200]
4. Epstein-Lubow GP, Beevers CG, Bishop DS, Miller IW. Family functioning is associated with
depressive symptoms in caregivers of acute stroke survivors. Archives of Physical Medicine and
Rehabilitation. 2009; 90:94755. [PubMed: 19480870]
5. Liptzin B, Grob MC, Eisen SV. Family burden of demented and depressed elderly psychiatric
inpatients. Gerontologist. 1988; 28:397401. [PubMed: 3396920]
6. Epstein-Lubow G, Gaudiano BA, Hinckley M, Salloway S, Miller IW. Evidence for the validity of
the American Medical Associations caregiver self-assessment questionnaire as a screening measure
for depression. Journal of the American Geriatrics Society. 2010; 58:3878. [PubMed: 20370867]
7. Cohen, S.; Williamson, GM. Perceived Stress in a Probability Sample of the United States. In. In:
Spaccapan, S.; Oskamp, S., editors. The Social Psychology of Health. Newbury Park: Sage; 1988.
p. 31-67.

Am J Geriatr Psychiatry. Author manuscript; available in PMC 2012 November 02.

Epstein-Lubow et al.

Page 6

8. Hirschman KB, Shea JA, Xie SX, Karlawish JH. The development of a rapid screen for caregiver
burden. Journal of the American Geriatrics Society. 2004; 52:17249. [PubMed: 15450052]
9. Woo BK, Golshan S, Allen EC, Daly JW, Jeste DV, Sewell DD. Factors associated with frequent
admissions to an acute geriatric psychiatric inpatient unit. Journal of Geriatric Psychiatry and
Neurology. 2006; 19:22630. [PubMed: 17085762]
10. Zivin K, Christakis NA. The emotional toll of spousal morbidity and mortality. American Journal
of Geriatrec Psychiatry. 2007; 15:7729.
11. Yen YC, Rebok GW, Gallo JJ, Jones RN, Tennstedt SL. Depressive symptoms impair everyday
problem-solving ability through cognitive abilities in late life. American Journal of Geriatric
Psychiatry. 2011; 19:14250. [PubMed: 20808123]

$watermark-text
$watermark-text
$watermark-text
Am J Geriatr Psychiatry. Author manuscript; available in PMC 2012 November 02.

$watermark-text
59
6.2
4.1
1.6
9.4
6.0
10.5
17.0

Mean Income Category

Hours/Week Category

Number who can Help

Depression

Stress

Burden

Grief

% Spousal

67

10.9

6.4

3.5

6.1

1.4

2.2

2.1

13.3

SD

23.5

13.8

7.1

13.2

1.7

3.5

6.5

20

30

65

100

68

57.0

Mean or %

13.7

7.0

3.5

6.9

1.4

2.1

2.5

12.1

SD

Caregivers of a Hospitalized Person (n = 41)

1
1
1
1

2 = 1.91
2 = 3.07
2 = 11.6
2 = 13.9

t = 2.43

t = 2.26

t = 1.46

t = 2.74

t = 0.34

t = 1.14

83

82

82

83

82

83

74

= 0.21

t = 0.57

81

DF

t = 3.17

Test Statistic

0.31
0.49
0.53

.026*
.017*

0.58

.008**
.148

Cohens d

.733

.259

.570

<.001**

<.001**

.080

.167

.651

.002**

Sig. Group Difference

significance < .01

significance < .05

**

This table reports the mean value and standard deviation (SD) (or percent for dichotamous variables) and the significance regarding paired t-tests and chi square tests assessing group differences. Hours /
week category of 4 corresponds with 21 30 hours / week. The test statistic and degrees of freedom (DF) are reported for all tests. Cohens d effect size is reported only for clinical variables. Clinical
measures are described in Methods.

% Co-Habitating

% Married

82

95

% White

64

% Female

65.9

Mean Age

Mean or %

Caregivers of an Outpatient (n = 44)

$watermark-text

Demographic & Clinical Variables: Group Differences

$watermark-text

Table 1
Epstein-Lubow et al.
Page 7

Am J Geriatr Psychiatry. Author manuscript; available in PMC 2012 November 02.