Full text of DOI:10.1093/gerona/56.8.M477

Journal of Gerontology: MEDICAL SCIENCES
2001, Vol. 56A, No. 8, M477–M482
Copyright 2001 by The Gerontological Society of America
Evidence for a Shift in the Th-1 to Th-2 Cytokine
Response Associated With Chronic Stress and Aging
Ronald Glaser,1,2,3 Robert C. MacCallum, Bryon F. Laskowski, William B. Malarkey,
2,4
1
1,2,5
John F. Sheridan,1 and Janice K. Kiecolt-Glaser
,2,6
2,4,7
1
2
Department of Molecular Virology, Immunology and Medical Genetics and Institute for Behavioral Medicine Research, Ohio
State University Medical Center, Columbus.
3
Comprehensive Cancer Center, Ohio State University, Columbus.
Departments of Psychology, Internal Medicine, Oral Biology, and Psychiatry, Ohio State University, Columbus.
4
5
6
7
Background. A number of studies have shown that the chronic stress of caring for persons with dementia can have
significant immunological consequences as demonstrated by the down-regulation/dysregulation of the cellular immune
response.
Methods. Utilizing flow cytometry to measure the percentages and absolute numbers of CD-4ꢀ and CD-8ꢀ T lym-
phocytes producing the cytokines indicative of Th-1, Tc1 and Th-2, and Tc2 cells, we compared spousal caregivers and
control subjects. The expression of interleukin-2 (IL-2), interferon gamma (IFN-ꢁ), and interleukin-10 (IL-10) in the cy-
toplasm of CD-4ꢀ and CD-8ꢀ lymphocytes was assessed.
ꢀ
ꢀ
ꢀ
ꢀ
Results. Neither stress nor age was significantly related to the percentage or number of IFNꢁ /CD-8 , IL-2 /CD-8
cells, or IFNꢁ , IL-2 , CD-4 cells. However, the percentage of IL-10ꢀ cells was higher in lymphocytes obtained from
caregivers than control subjects. In addition, the significant interaction between stress and aging for IL-10 /CD-4 and
IL-10 /CD-8 cells demonstrated that the difference between caregivers and control subjects was age dependent; the
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
difference between caregivers and control subjects was substantially larger in younger individuals than in older individ-
uals.
Conclusions. The data are consistent with previous reports on acute stress and suggest that there may also be a shift
from a Th-1 to a Th-2 response associated with a chronic stressor such as caregiving. This shift could have implications
for an individual’s responses to pathogens.
HE chronic stress of caregiving for a family member
with progressive dementia has significant immuno-
In a further study, caregivers exhibited significant deficits
relative to control subjects in both the antibody and virus-
specific T-cell responses to an influenza virus vaccine (6).
Caregivers were less likely than control subjects to show a
fourfold increase in antibody titers to the trivalent flu vac-
cine 4 weeks post-vaccination, and they also showed a
poorer virus-specific T-cell response in vitro. Additionally,
PBLs from caregivers produced lower levels of IL-1ꢂ when
stimulated with LPS. These data suggest that caregivers are
potentially more vulnerable to the influenza virus and to
other infectious agents than aged-matched control subjects
(7). A recent study confirmed stress-related decrements in
the antibody response to an influenza virus vaccination in
caregivers of dementia patients (8). These vaccine data may
provide a window on the body’s response to pathogens,
such as viruses or bacteria, and support the hypothesis that
the immune modulation induced by psychological stressors
is large enough to have health effects (9).
T
logical consequences. Kiecolt-Glaser and colleagues (1)
found a poorer proliferative response of peripheral blood
leukocytes (PBLs) to mitogens in caregivers compared with
PBLs obtained from noncaregiving control subjects. Con-
comitant with changes in cellular immunity, caregivers re-
ported significantly more days of infectious illness, prima-
rily upper respiratory tract infections, and these reports were
corroborated by physician data (1). McCann (2) reported
that caregivers had poorer responses to delayed hypersensi-
tivity skin testing than noncaregivers. Plasma levels of neu-
ropeptide Y, a sympathetic neurotransmitter that is released
during emotional stress and that may also modulate immu-
nity (3), were significantly elevated in blood samples ob-
tained from caregivers compared with nondepressed control
subjects. We have reported that natural killer (NK) cells ob-
tained from caregivers showed a poorer response to stimula-
tion with recombinant gamma interferon (rIFN-ꢁ) and recom-
binant interleukin-2 (rIL-2) than NK cells obtained from
control subjects (4). Wound healing was also significantly de-
layed in caregivers compared with control subjects, and PBLs
obtained from these subjects produced less interleukin-1 beta
Importantly, preliminary data suggest that cessation of care-
giving does not terminate risk; in the data collected thus far in
our longitudinal study, former spousal caregivers have not
differed significantly from “continuous” caregivers on cellu-
lar immune function after the death of the impaired spouse. In
fact, spousal caregivers have continued to show immunologi-
cal down-regulation for several years after bereavement (4).
(
(
IL-1ꢂ) in vitro after stimulation with lipopolysaccharide
LPS) as compared with PBLs from control subjects (5).
M477

M478
GLASER ET AL.
For example, current and former caregivers did not differ
programs, and governmental caregiver support programs.
At the time of recruitment into the study, all caregivers were
caring for a spouse with Alzheimer’s disease or another pro-
gressive dementia and had to be providing five or more
hours of care per week.
Control subjects were recruited through newspaper ad-
vertisements, senior citizen centers, area newsletters, church
groups, university alumni publications, and referrals from
other participants; potential control subjects who reported
any caregiving activities were excluded. The Ohio State
University Biomedical Research Review Committee ap-
proved the project; all subjects gave written informed con-
sent prior to participation.
The 76 subjects included 16 continuing caregivers, 16
former caregivers, and 44 control group subjects. There
were no significant differences among the three groups in
education (70% had at least a partial college education),
race (82% Caucasian, 18% African American), or gender
(71% were women). There was, however, a significant dif-
ference among groups with respect to age, F(2,73) ꢃ 3.72,
p ꢄ .05; for control subjects, mean age ꢃ 69.89, SD ꢃ 9.26;
for former caregivers, mean age ꢃ 77.00, SD ꢃ 9.51; for
current caregivers, mean age ꢃ 71.69, SD ꢃ 7.25. All sub-
jects were paid $40 for their annual participation in the
study. Subjects were excluded if they reported diabetes, the
use of anti-inflammatory medication or other medications
with obvious immunological consequences, or immunologi-
cally related health problems (e.g., cancer, autoimmune dis-
ease, or recent surgery).
Continuing caregivers were defined as those who were
still actively caregiving in the year these data were col-
lected. Continuing caregivers had been providing care for
an average of 104.11 months (SEM ꢃ 12.31). They reported
spending an average of 7.60 h/d in caregiving activities. For
former caregivers, an average of 2.57 years (SEM ꢃ 0.50)
had elapsed since the death of their spouse, with a range of
several months to 5 years.
from each other in the inability of their NK cells to respond
to the stimulatory effects of rIFN-ꢁ or rIL-2 (4), and both
showed poorer responses to these cytokines than did control
subjects. Similarly, the protective capacity of viral vaccines
is dependent on their ability to induce both humoral and
cell-mediated immune responses (7); both were poorer in
caregivers than in control subjects, and current caregivers
did not differ from former caregivers (10).
In a recent study from our laboratory, we explored a pos-
sible mechanism underlying the stress-associated immune
changes in caregivers by measuring lymphocyte growth
hormone (GH) mRNA levels in lymphocytes. GH is an im-
mune-enhancing hormone that may be important in modu-
lating humoral and cellular immune function (11). Using
RT-PCR to measure GH mRNA levels in T- and B-cell pop-
ulations, the level of GH mRNA was 50% less in cells ob-
tained from caregivers compared with control subjects (11).
These differences in GH mRNA levels in both B and T cells
may be related to the poorer immune response to the influ-
enza virus vaccine already discussed.
ꢀ
Different subpopulations of CD-4 T cells synthesize spe-
cific cytokines and have been designated Th-1 or Th-2 cells.
Whereas there is some overlap between the patterns of cyto-
kines synthesized by both these populations of lymphocytes,
the evidence suggests that there is also specificity. Further-
more, the cytokine products of Th-1 and Th-2 cells tend to be
ꢀ
mutually inhibitory. Additionally, CD-8 T-lymphocytes can
secrete Th-1 as well as Th-2 cytokines; these cells are desig-
nated Tc1 and Tc2. Examples of Th-1- and Tc1-derived cyto-
kines include IL-2 and IFN-ꢁ. Th-2- and Tc2-derived cyto-
kines include IL-4, IL-5, IL-6, and IL-10 (12).
Glucocorticoid hormones can profoundly affect cell-medi-
ated immunity. Glucocorticoid hormones differentially modu-
late the expression of the cytokines IL-2 and IL-4, resulting in
a polarized shift of T-cell responses to the Th-2 subset (13).
The polarization of T-helper cell responses can be driven in
vivo by the stress-induced elevation of endogenous glucocor-
ticoid hormones, and this shift may play an important role in
determining the host’s response to a pathogen (14).
Health-related behaviors were assessed when blood was
drawn (15). Two questions assessed exercise (16). The 36-
item Rand Health Survey (17) provided a nondisease spe-
cific measure of functioning with excellent normative data.
In this study we examined stress-associated immune al-
terations in caregivers by measuring the expression of cy-
tokines synthesized by PBLs that help regulate and balance
the cellular and humoral immune responses. Specifically,
Measurement of Cytoplasmic Cytokines Including IL-2,
IL-10, and IFN-ꢁ
ꢀ
we measured the percentage and total number of CD4 and
Heparinized whole blood samples were obtained from
each subject, and PBLs were obtained using routine proce-
ꢀ
CD8 lymphocytes synthesizing IL-2, IL-10, and IFN-ꢁ by
6
flow cytometry. The data suggest that there may be a shift
toward a Th-2 response associated with chronic stress and
that this shift is related to age.
dures. To detect cells synthesizing IL-2, 4 ꢅ 10 PBLs were
washed in phosphate-buffered saline (PBS). The PBLs were
stimulated with 20 ng/ml phorbol-23 myristate 13-acetate
(
PMA) (Sigma, St. Louis, MO), 250 mg/ml ionomycin
METHODS
(Sigma), and 2 ꢆM monensin (Sigma) in RPMI 1640 me-
dium supplemented with 5% fetal bovine serum (FBS) for 6
hours at 37ꢇC. The cells were then placed in calcium- and
magnesium-free Dulbecco’s PBS (DPBS) containing 1%
heat-inactivated FBS and 0.1% sodium azide, pH adjusted
between 7.4 and 7.6. The PBLs were adsorbed with either
the surface marker IgG1 monoclonal antibody (MAb) con-
jugated to fluorescein isothiocyanate (FITC) (for control),
or CD-4- FITC, CD-8- FITC, or CD-3- FITC MAbs
(Pharmingen, San Diego, CA). The cells were washed and
then resuspended in fixation buffer (4% paraformadehyde
Subject Characteristics
Subjects for this study were part of a longitudinal study
of caregiver stress, health, and immune function (1,4,6).
The longitudinal study includes ongoing recruitment into
the cohort across years, and caregivers are followed after
the death of the spouse. Subjects were recruited from multi-
ple local sources, including dementia evaluation centers in
area hospitals, neurologists’ referrals, the city’s Alzheimer’s
Association support groups and its newsletter, respite care

STRESS-INDUCED DYSREGULATION OF IMMUNITY
M479
in DPBS, with the pH adjusted between 7.4 and 7.6) for 20
minutes at 4ꢇC, then washed twice with calcium- and mag-
nesium-free DPBS, 1% heat-inactivated FBS, 0.1% sodium
azide, 0.1% saponin, with the pH adjusted between 7.4 and
able, and (ii) a dummy variable coded as 0 to represent con-
trol subjects and 1 for caregivers. Regression analyses fol-
lowed methods recommended by Cohen and Cohen (18) to
provide tests of the main effect of each of these predictors as
well as their interaction. Results of each test include an F ra-
tio, which provides a significance test, as well as a measure
of the variance in the dependent variable accounted for
uniquely by the corresponding effect. This approach takes
into account the association between group and age men-
tioned earlier by partialing out each effect from the other.
7
.6. The cells were resuspended in permeabilization buffer
containing 0.25 ꢆg of a rat anti-human IL-2 MAb conju-
gated to phycoerythrin (PE) (Pharmingen). A block control
sample was prepared by adding a rat anti-human IL-2 MAb
with an excess of recombinant human IL-2 (Pharmingen) to
a tube containing PBLs. The tubes were incubated for 30
minutes at 4ꢇC, washed twice with staining buffer, pelleted,
and resuspended in 100 ꢆl staining buffer.
RESULTS
ꢀ
To detect IL-10 T cells, PBLs were incubated with 10
To determine if our conditions allowed us to accurately
determine the percentage of CD-4ꢀ cells versus IL-10ꢀ
cells, we analyzed the patterns to assess the quality of the
data used to calculate the mean values for our analysis. An
example of these scatter plots is shown in Figure 1 and is
representative of the patterns observed for all the samples.
No significant main effects or interactions were found for
analyses of cells expressing IL-2 or IFN-ꢁ. That is, there
were no significant effects of stress or age on the percentage
ng/ml LPS (Sigma) and 1 ꢆg/ml Brefeldin A (Sigma) in
RPMI 1640 medium supplemented with 10% FBS at 37ꢇC
for 24 hours. Following stimulation, the cells were adsorbed
with a surface marker, IgG1-FITC MAb (for control), CD-
4
-FITC, or CD-8-FITC MAbs (all Pharmingen). The cells
were washed and placed in permeabilization buffer contain-
ing 0.125 ꢆg of rat anti-human IL-10-PE MAb (Pharmin-
gen). A blocked control sample was prepared by adding rat
anti-human IL-10-PE MAb with an excess of purified rat
anti-human IL-10 MAb (Pharmingen). The cells were incu-
bated and washed as described above.
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
or number of IFN-ꢁ /CD-8 , IL-2 /CD-8 , IFN-ꢁ /CD-
4
ꢀ
ꢀ
ꢀ
, or IL-2 /CD-4 cells. However, a consistent pattern of
significant effects was obtained for analyses of percentages
6
To measure intercellular IFN-ꢁ, 4 ꢅ 10 PBLs were incu-
ꢀ
ꢀ
and absolute numbers of both IL-10 /CD-4 PBLs and IL-
bated for 6 hours with 20 ng/ml PMA (Sigma), 250 ng/ml
ionomycin (Sigma), and 2 ꢆM monensin (Sigma) in RPMI
ꢀ
ꢀ
1
0 /CD-8 PBLs. Because the same pattern of results was
found for analyses of both percentages and significant num-
bers of these measures, detailed results are presented for
percentages only. In addition, no differences in the mean
number of CD-4ꢀ and CD-8 cells between groups before
stimulation were found (data not shown).
1
640 medium supplement with 10% FBS. Following stimu-
lation, the cells were incubated either with a surface marker,
IgG2a-FITC or CD-8-FITC, or CD-3-FITC MAbs (all
Pharmingen). After washing, the cells were resuspended in
permeabilization buffer containing 0.125 ꢆg of mouse anti-
human IFN-ꢁ-PE MAb (Pharmingen). A block control was
prepared by adding preincubated mouse anti-human IFN-ꢁ-
PE MAb with an excess of purified mouse anti-human IFN-ꢁ
MAb (Pharmingen). The cells were incubated and washed
as previously described. All samples were read on a Profile
II FACS (Coulter Corporation, Hialeah, FL).
ꢀ
ꢀ
ꢀ
ꢀ
For percentage of both IL-10 /CD-4 PBLs and IL-10 /
ꢀ
CD-8 PBLs, a significant effect of group membership was
found: for IL-10 /CD-4 PBLs, F(1,73) ꢃ 5.75, p ꢄ .05;
for IL-10 /CD-8 PBLs, F(1,73) ꢃ 13.97, p ꢄ .001. For
both combinations, the percentage of IL-10 cells was
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
higher in PBLs obtained from caregivers than in cells from
control subjects. Group membership uniquely accounted for
A forward versus side scatter plot was used to set a gate
around the lymphocytes. A minimum of 10,000 gated
events were acquired per sample. Two-parameter quadrant
analyses demonstrating cytokine staining were created us-
ing ELITE software (Coulter), and quadrant markers were
set using the isotype negative control and the blocked cyto-
kine control to include 99% of the cells in the negative quad-
rant. Data are presented as percentages of cytokine-express-
ing cells within the total lymphocyte population gated.
ꢀ
ꢀ
7
1
.30% of the variance in IL-10 /CD-4 PBLs and for
ꢀ
ꢀ
6.10% of the variance in IL-10 /CD-8 PBLs, above and
beyond any effect of age. Group means for these two mea-
sures are shown in Figures 2 and 3.
ꢀ
ꢀ
For IL-10 /CD-8 PBLs, a significant interaction of
stress and age was also obtained, meaning that the influence
of age on this measure was significantly different for care-
givers and control subjects (F(1,72) ꢃ 6.31, p ꢄ .05). The
nature of this interaction is shown in Figure 4, which shows
ꢀ
ꢀ
a substantial negative influence of age on IL-10 /CD-8
Statistical Analyses
PBLs for caregivers and a rather weak influence for control
subjects. From another perspective, the difference between
Statistical analyses assessed the effects of stress and age
on selected Th-1 and Th-2 cytokines measured in PBLs ob-
tained at two occasions 2 weeks apart. Regression analyses
were used to test the effects of two independent variables on
each of six cytokine measures. Dependent variables of inter-
ꢀ
ꢀ
caregivers and control subjects in IL-10 /CD-8 PBLs is
reduced considerably with age. The same trend was ob-
ꢀ
ꢀ
served for IL-10 /CD-4 PBLs, although the effect did not
reach statistical significance (F(1,72) ꢃ 1.58, p ꢃ .21).
ꢀ
ꢀ
est were percent and absolute number of IL-10 /CD-4 , IL-
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
ꢀ
1
4
0 /CD-8 , IFN-ꢁ /CD-4 , IFN-ꢁ /CD-8 and IL-2 /CD-
ꢀ
ꢀ
ꢀ
, and IL-2 /CD-8 cells. For each of these six variables,
Health-Related Behaviors
the mean value of the two separate measures of each variable
was used as the dependent variable in the ANOVA. When
only a single measure was available, that value was used.
The independent variables were (i) age, as a continuous vari-
The majority of older adults take some medication, and
the individual who reported no medications was atypical
among these subjects. In this sample, 24% used estrogen
supplements, 22% took beta blockers or calcium channel

M480
GLASER ET AL.
Figure 1. Flow cytometric detection of interleukin (IL)-10 in CD-4ꢀ and CD-8ꢀ cells. A shows the forward/side scatter distribution and the
gate used to select lymphocytes for analysis. B shows the quadrant marker set using the isotype and IL-10 blocked controls. C and D show cells
in quadrant 2 that are CD-4ꢀ IL-10ꢀ and CD-8ꢀ IL-10 , respectively.
ꢀ
blockers, 24% used prescription diuretics, and 16% took thy-
roid supplements, without any systematic group difference.
Most other health-related behaviors also failed to distin-
guish between caregivers and control subjects. Alcohol con-
sumption was low and did not differ between groups; 92%
were nonsmokers, without significant group differences. The
two groups did not differ in weight change in the prior 2
weeks or hours of sleep within the last 3 days. All subjects
had plasma albumin levels (as a measure of nutrition) within
the normal range (19). Caregivers and control subjects did
show reliable differences in response to a question about
whether they typically engaged in a regular activity long
enough to build up a sweat (16), chi square (2, n ꢃ 74) ꢃ
nificantly associated with the significant immunologic data.
The groups did not differ on any of the health dimensions of
the RAND Health Survey (17); all p ꢈ .28. In sum, there
were no reliable differences in health-related behaviors be-
tween groups that would have accounted for the immuno-
logical differences.
DISCUSSION
In this study we found evidence for a shift from a Th-1,
Tc1 to a Th-2, Tc2 response associated with caregiving of
persons with dementia. Our findings are similar to data re-
ported by Marshall and colleagues (20), who found a similar
shift associated with academic stress (an acute stressor) in
medical students. We extended their study by measuring the
expression of IL-2, IL-10, and IFN-ꢁ in the cytoplasm of
CD-4ꢀ and CD-8ꢀ T cells obtained from older individuals
7
.24, p ꢄ .05, with the majority of control subjects and
former caregivers answering yes and the majority of current
caregivers answering no. However, the variable was not sig-

STRESS-INDUCED DYSREGULATION OF IMMUNITY
M481
Figure 2. Effect of stress on percent interleukin (IL)-10 /CD4ꢀ pe-
ꢀ
ripheral blood leukocytes (PBLs). PBLs were treated with 10 ng/ml
lipopolysaccharide and 1 ꢆg/ml Brefeldin in complete RPMI 1640
Figure 4. Interactive effect of stress and age on percent interleukin
medium for 24 hours at 37ꢇC. The percentage of CD4ꢀ cells express-
(IL)-10
ꢀ
/CD8 cells. A significant interaction of stress and age was
ꢀ
ing cytoplasmic IL-10 was determined using flow cytometry as de-
scribed in Methods.
found between caregivers and control subjects.
and suggests that physiological and psychological conse-
quences of chronic stressors may persist well beyond the
cessation of the actual stressor. Indeed, research with other
populations suggests that such effects are not unique to care-
givers; other data suggest that continued immunologic down-
regulation may be related to persistent rumination about a
past stressful event (23), consistent with rumination data from
former caregivers (24). However, because of age-related im-
munological changes, older adults may be the most vulnera-
ble to actual health changes.
caregiving for a patient with dementia, a chronic stressor,
using flow cytometry. Although caregiving was not associ-
ꢀ
ated with the percentage or total number of either CD-4 or
ꢀ
CD-8 cells expressing IL-2 or IFN-ꢁ, there was a signifi-
ꢀ
cant increase in the percentage and total number of IL-10 /
ꢀ
ꢀ
ꢀ
CD-4 and IL-10 /CD-8 cells. Additionally, this pattern
was significantly stronger in younger individuals when
compared with older subjects, indicating that there may be
an aging component to the Th-1, Tc1 to Th-2, Tc2 shift.
Whereas data from mouse studies suggest that a Th-1 to Th-2
shift is associated with aging, data from human studies are
less well defined (21). The range of the percentages of IL-10ꢀ
cells found in this study is consistent with normal values of
In previous work using an acute academic stress model
with medical students, we found a significant decrease in
the production of IFN-ꢁ by PBLs obtained at the time of ex-
aminations as compared with PBLs obtained at baseline
(25). These data were confirmed by Marshall and col-
leagues (20). In an earlier study, we did not find a difference
in the production of IFN-ꢁ by PBLs stimulated in vitro in
caregivers versus control subjects (unpublished data). In
this study we confirmed these results and found that the per-
ꢀ
IL-10 PBLs (Podma Kodukula, Pharmingen, personal com-
munication, April 2000) as well as previously published
data (22).
The absence of immunologic differences between current
and former caregivers is consistent with prior work (4,10)
ꢀ
ꢀ
centage and total number of CD-4 or CD-8 lymphocytes
expressing IFN-ꢁ were the same for the Alzheimer’s dis-
ease caregivers and control subjects. We do not know if the
differences in the results with regard to IFN-ꢁ and in the
medical student studies versus the study with caregivers are
related to age, the effects of a chronic stressor versus an
acute stressor, or a combination of both. However, consis-
tent with the study by Marshall and colleagues (20), we did
find an increase in the number of CD-4ꢀ and CD-8ꢀ cells
expressing cytoplasmic IL-10. As already discussed, IL-10
will down-regulate IFN-ꢁ production and will tend to
dampen the overall cellular immune response (12).
One possible explanation for a shift from Th-1, Tc1 to
Th-2, Tc2 lymphocytes may be related to changes in cate-
cholamine levels, which are elevated as a result of psycho-
logical stress. Elenkov and Chrousos (26) recently reported
that an increase in IL-10 production was associated with an
increase in catecholamines, and these changes resulted in a
shift in the Th-1 to Th-2 direction. Possible implications for
hypothalamic pituitary adrenal axis activation and a shift
Figure 3. Effect of stress on percent interleukin (IL)-10 /CD8ꢀ pe-
ꢀ
ripheral blood leukocytes (PBLs). PBLs were treated as described for
Figure 1. The percentage of CD8ꢀ cells expressing cytoplasmic IL-10
was determined by flow cytometry as described in Methods.

M482
GLASER ET AL.
from Th-1 to Th-2 cytokine responses has also been demon-
strated (27). Previous data from our laboratory and others
on the impact of dementia caregiving on the cellular im-
mune response and the data obtained in this study are con-
sistent with these reports and support the hypothesis that
stress-associated hormones can induce cytokine dysregula-
tion.
Alzheimer’s dementia is associated with down-regulation of B-lym-
phocyte GH mRNA. J Gerontol Med Sci. 1999;54:M212–M215.
2. Mosmann TR, Sad S. The expanding universe of T-cell subsets: Th1,
Th2 and more. Immunol Today. 1996;17:138–146.
3. Elenkov I, Papanicolaou D, Wilder R, Chrousos G. Modulatory effects
of glucocorticoids and catecholamines on human interleukin-12 and
interleukin-10 production: clinical implications. Proc Assoc Am Physi-
cians. 1996;108:374–381.
1
1
1
4. Dobbs CM, Feng N, Beck FM, Sheridan JF. Neuroendocrine regula-
tion of cytokine production during experimental influenza viral
infection: effects of restraint stress-induced elevation in endogenous
corticosterone. J Immunol. 1996;157:1870–1877.
Acknowledgments
This work was supported in part by National Institutes of Health Grants
K02 MH01467, R37 MH42096, and PO1 AG 11585, The Ohio State Uni-
versity Comprehensive Cancer Center Core Grant CA 16058, and by the
Gilbert and Kathryn Mitchell endowment. We thank Ron Whisler for his
helpful suggestions.
1
1
5. Kiecolt-Glaser JK, Glaser R. Methodological issues in behavioral im-
munology research with humans. Brain Behav Immun. 1988;2:67–78.
6. Washburn RA, Adams LL, Hale GT. Physical activity assessment for
epidemiologic research: the utility of two simplified approaches. Prev
Med. 1987;16:626–646.
Address correspondence to Ronald Glaser, PhD, Department of Molec-
ular Virology, Immunology and Medical Genetics, 2175 Graves Hall, 333
W. 10th Ave., Columbus, OH 43210. E-mail: Glaser.1@osu.edu
1
7. Ware JE, Sherbourne CD. The MOS 36-Item-Short-Form Health Sur-
vey (SF-36): I. Conceptual framework and item selection. Med Care.
1
992;30:473–483.
1
1
2
8. Cohen J, Cohen P. Applied Multiple Regression/Correlation Analysis
for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Erlbaum; 1983.
9. Cooper JK, Gardner C. Effect of aging on serum albumin. J Am Geri-
atr Soc. 1989;37:1039–1042.
0. Marshall GD Jr, Agarwal SK, Lloyd C, Cohen L, Henninger EM,
Morris GJ. Cytokine dysregulation associated with exam stress in
healthy medical students. Brain Behav Immun. 1998;12:297–307.
1. Karanfilov CI, Liu B, Fox CC, Lakshmana R, Whisler RL. Age-related
References
1
2
3
4
5
6
7
. Kiecolt-Glaser JK, Dura JR, Speicher CE, Trask OJ, Glaser R. Spousal
caregivers of dementia victims: longitudinal changes in immunity and
health. Psychosom Med. 1991;53:345–362.
. McCann JJ. Effects of stress on spouse caregivers’ psychological
health and cellular immunity (dissertation). Chicago, IL: Rush Univer-
sity College of Nursing; 1991.
. Irwin M, Smith TL, Christian G. Electroencephalographic sleep and
natural killer activity in depressed patients and control subjects. Psy-
chosom Med. 1992;54:10–21.
. Esterling BA, Kiecolt-Glaser JK, Glaser R. Psychosocial modulation
of cytokine-induced natural killer cell activity in older adults. Psycho-
som Med. 1996;58:264–272.
2
defects in Th1 and Th2 cytokine production by human T cells can be
dissociated from altered frequencies of CD45RAꢀ and CD45RO
ꢀ
T
cell subsets. Mech Ageing Devel. 1999;109:97–112.
2
2. Crucian B, Dunne P, Friedman H, Ragsdale R, Pross S, Widen R. De-
tection of altered T Helper 1 and T Helper 2 cytokine production by
peripheral blood mononuclear cells in patients with multiple sclerosis
utilizing intracellular cytokine detection by flow cytometry and sur-
face marker analysis. Clin Diagn Lab Immunol. 1996;3:411–416.
3. Solomon GF, Segerstrom SC, Grohr P, Kemeny M, Fahey J. Shaking
up immunity: psychological and immunological changes after a natu-
ral disaster. Psychosom Med. 1997;59:114–127.
. Kiecolt-Glaser JK, Marucha PT, Malarkey WB, Mercado AM, Glaser
R. Slowing of wound healing by psychological stress. Lancet. 1995;
2
3
46:1194–1196.
. Kiecolt-Glaser JK, Glaser R, Gravenstein S, Malarkey WB, Sheridan
J. Chronic stress alters the immune response to influenza virus vaccine
in older adults. Proc Natl Acad Sci USA. 1996;93:3043–3047.
. Gravenstein S, Drinka P, Duthie EH, et al. Efficacy of an influenza he-
magglutinin-diphtheria toxoid conjugate vaccine in elderly nursing
home subjects during an influenza outbreak. J Am Geriatr Soc. 1994;
2
4. Bodnar J, Kiecolt-Glaser JK. Caregiver depression after bereavement:
chronic stress isn’t over when it’s over. Psychol Aging. 1994;9:372–380.
5. Glaser R, Rice J, Speicher CE, Stout JC, Kiecolt-Glaser JK. Stress de-
presses interferon production by leukocytes concomitant with a decrease
in natural killer cell activity. Behav Neurosci. 1986;100:675–678.
6. Elenkov IJ, Chrousos GP. Stress hormones, Th1/Th2-patterns, pro/
anti-inflammatory cytokines and susceptibility to disease. Trends En-
docrinol Metab.1999;10:359–368.
2
4
2:245–251.
2
2
8
. Vedhara K, Cox NKM, Wilcock GK, et al. Chronic stress in elderly
carers of dementia patients and antibody response to influenza vacci-
nation. Lancet. 1999;353:627–631.
7. Agarwal SK, Marshall GD Jr. Glucocorticoid induced type-1/type-2
cytokine alterations in humans: a model for stress-related immune dys-
function. J Interferon Cytokine Res. 1998;18:1059–1068.
9
. Glaser R, Rabin B, Chesney M, Cohen S, Natelson B. Stress-induced
immunomodulation: implications for infection diseases? JAMA. 1999;
2
81:2268–2270.
1
0. Glaser R, Kiecolt-Glaser JK, Malarkey WB, Sheridan JF. The influ-
ence of psychological stress on the immune response to vaccines. Ann
NY Acad Sci. 1998;840:656–663.
Received December 2, 1999
1
1. Wu H, Wang J, Cacioppo JT, Glaser R, Kiecolt-Glaser JK, Malarkey
WB. Chronic stress associated with spousal caregiving of patients with
Accepted May 19, 2000
Decision Editor: John E. Morley, MB, BCh