BRAIN,
BEHAVIOR,
and IMMUNITY
Brain, Behavior, and Immunity 18 (2004) 65–75
www.elsevier.com/locate/ybrbi
The relationship between distress and the development of a
primary immune response to a novel antigen
Alison Smith,a,b,* Ute Vollmer-Conna,b Barbara Bennett,b Denis Wakefield,c
Ian Hickie,b and Andrew Lloydc
a
School of Psychology, University of Western Sydney, Locked Bag 1797, Penrith South, DC NSW 1797, Australia
b
School of Psychiatry, University of NSW, Sydney 2052, Australia
c
Inflammation Research Unit, School of Medical Sciences, University of NSW, Sydney 2052, Australia
Received 27 November 2001; received in revised form 6 February 2003; accepted 7 May 2003
Abstract
Forty-five medical students were recruited to examine the effects of distress on the development of an immune response to the
novel antigen, keyhole limpet hemocyanin (KLH). The subjectsÕ level of distress was manipulated by immunizing them either at the
time of an important viva voce examination (n ¼ 22) or during examination-free term time (n ¼ 23). This manipulation increased
variance amongst the subjects, but the emphasis in this research was on individual distress as a predictor of immune function. In the
group as a whole, the likelihood of developing DTH skin responses to KLH was reduced in the more distressed subjects (r ¼ :45;
p ¼ :002), independently of a number of behavioral (e.g., sleep disturbance) and demographic (e.g., sex) variables. Proliferation of T
cells against KLH in vitro and the development of anti-KLH IgG antibodies were not related to levels of distress. Thus, cellular,
rather than humoral, immune responses in vivo appear susceptible to the influence of distress. This immunization model provides
the opportunity to further dissect the basis of these stress–immune pathways.
Ó 2003 Elsevier Inc. All rights reserved.
Keywords: Keyhole limpet haemocyanin; Psychoneuroimmunology; Cellular immunity; Humoral immunity; Primary immune response
1. Introduction
Human research in the field of psychoneuroimmunology has profited in recent years from the development
of models that are increasingly biologically relevant,
potentially of clinical significance, and also permit the
functional assessment of immunity (e.g., Glaser et al.,
1992; Kiecolt-Glaser et al., 1995; Kiecolt-Glaser et al.,
1996). However, much of this research relies upon recall
responses to previously encountered infectious agents
(for example, Herpes simplex viruses) or is confounded
by variable degrees of prior immunity to the immunogen
(e.g., influenza vaccine). This study utilized a standardized method of functional assessment of immunity based
on the immunization of subjects with the novel antigen,
*
Corresponding author. Fax: +612-9772-6736.
E-mail address: [email protected] (A. Smith).
0889-1591/$ - see front matter Ó 2003 Elsevier Inc. All rights reserved.
doi:10.1016/S0889-1591(03)00107-7
keyhole limpet hemocyanin (KLH). This model may be
considered an immunological challenge that broadly
models the events occurring upon host exposure to a
previously unencountered pathogen. Such a response
necessitates processing and presentation of the novel
antigens, followed by the expansion of the effector arm
of the response and generation of immunological
memory.
KLH is a purified protein derived from an inedible
marine mollusc (Megathura crenulata) found off the
coast of Southern California (Snyder et al., 1990). The
advantages of using KLH in psychoneuroimmunology
research is that it allows the comprehensive assessment
of an orchestrated cellular and humoral immune response in vivo, it is simple and safe, and it is readily
applicable to almost all potential subjects. Both cellular
and humoral immunity can be assessed, permitting a
comparison of the relative effects of psychological factors on these two arms of the adaptive immune re-
66
A. Smith et al. / Brain, Behavior, and Immunity 18 (2004) 65–75
sponse. In addition, cellular immunity can be assessed in
vivo, by delayed-type hypersensitivity (DTH) skin tests
(Curtis et al., 1970).
KLH is unlikely to have been seen by the immune
system and is thus expected to promote a primary immune response. Consequently, the model should be unaffected by varying levels of prior immunity against the
immunogen (as is the case for more naturalistic immunization models such as influenza vaccine). However,
past research suggests that some individuals demonstrate
responses to KLH prior to immunization with the antigen. In vitro blast transformation of human leucocytes
from individuals with no known history of exposure to
KLH has been demonstrated (Green and Borella, 1971),
although a failure to find this response has also been
observed in unimmunized individuals (Burke et al.,
1977). In unimmunized human subjects, high avidity IgG
antibodies against KLH have been demonstrated using
radioimmunoassay (Moroz et al., 1973) and by a sensitive haemagglutination technique (Burke et al., 1977). In
unimmunized individuals, the development of DTH responses to an intradermal dose of 100 lg of KLH has
also been demonstrated (Hortobagyi et al., 1981), although classical DTH responses are typically elicited by
considerably smaller doses of antigen.
These baseline responses are likely to be due to the
presence of cross-reacting epitopes, such as those that
have been demonstrated with antigens from phylogenetically close species (Amkraut et al., 1969); mammalian red cell stroma (Frick and Shimbor, 1970); the
Thomson–Friederich antigen present on bladder and
other adenocarcinomas (Wirguin et al., 1995) and a
carbohydrate epitope on the surface of Schistosoma
mansoni larval schistosomes (Yuesheng et al., 1994). In
fact, the high immunogenicity of KLH may in part reflect the existence of these cross-reacting epitopes
(Harris and Markl, 1999). Consequently, an important
part of the research reported in this paper is an examination of the extent to which subjects display significant
responses to KLH prior to immunization.
Before commencement of this research a suitable
immunization protocol with an appropriate immunization dose and timing of follow-up assessment was developed. A response rate of approximately 70% was
deemed appropriate so that both immune enhancing and
immune suppressive effects of distress could be observed. The method and results of this protocol development are reported briefly in this paper, prior to an
examination of the effects of distress on anti-KLH immunity. The relationships between the various measures
of immune function are also examined, since they may
clarify mechanisms underlying psycho-neuro-immunological influences.
The stressor chosen for this study was a medical
student examination, a widely used model in psychoneuroimmunology research (reviewed by Kiecolt-Glaser
and Glaser, 1992). Medical students provide the advantage of generally low rates of pre-morbid psychological dysfunction. Thus, this paradigm facilitates an
evaluation of the effects of a relatively minor and common stressor on anti-KLH immunity in a psychologically and physically healthy group of individuals. It
should be noted, however, that the stressfulness of the
examination, as well as the individualÕs perceived competence in the subject to be examined may vary significantly. Thus, the analyses reported in this paper
emphasize the prediction of immunological outcomes by
the subjective distress levels recorded for each individual, rather than exclusively by the comparison of study
groups (Vedhara and Nott, 1996). Levels of serum
noradrenaline and cortisol were also determined. These
measures generally correlate with distress (Frankenhaeuser, 1983) and provide covert indices of psychological state. They may also be related to changes in
immune function (Black, 1994).
2. Materials and methods
2.1. Protocol development
2.1.1. Subjects
Twenty-one healthy volunteers aged between 18 and
30 years (12 males and nine females) volunteered to take
part in the research. Of these, six participated in an
examination of the kinetics of the immune response and
a further 15 in an examination of the rates of response.
These volunteers were university students experiencing
normal everyday levels of distress since they were recruited during examination free term time.
The research reported in this paper was approved by
the Committee on Experimental Procedures Involving
Human Subjects of the University of New South Wales.
Informed, written consent was obtained from all subjects before the commencement of procedures. Subjects
were excluded from the study if they (i) had a history of
seafood allergy; (ii) were pregnant; (iii) had experienced
a significant illness or operation in the past month; (iv)
had a history of a major immune-related illness (for
example, diabetes or cancer); (v) were taking any immuno-suppressive medication or (vi) were symptomatic
with an intercurrent infective illness at the commencement of the study.
2.2. Immunization dose
KLH (Pierce, Rockford, USA) was adsorbed to the
adjuvant alum (Pierce) according to the manufacturerÕs
instructions. The immunization protocol investigated
(and finally chosen) consisted of 0.1 mg of KLH
(1.25 mg/ml) adsorbed to 0.9 mg of alum (45 mg/ml)
administered into the deltoid muscle.
A. Smith et al. / Brain, Behavior, and Immunity 18 (2004) 65–75
2.3. Assessment of anti-KLH immunity
2.3.1. KLH-induced DTH skin responses
Delayed type hypersensitivity reactions were assessed
by administration of 0.001 mg of KLH in 0.01 ml of
saline solution as an intradermal injection into the volar
aspect of the arm. The response was read 48 h after
administration by measuring induration in millimetres
for two diameters at right angles. The mean of these
measurements was recorded. A positive response was
defined as a mean diameter of greater than 2 mm
(Hortobagyi et al., 1981). Such a classification is regularly used in clinical settings where DTH skin test categories of response are used as an index of health status
(Gm€
ur et al., 1976; Hersh et al., 1971; Poenaru and
Christou, 1991). Classification of individuals as responders or non-responders for DTH (and the other
measures of anti-KLH immunity) also facilitated an
examination of the extent to which individuals demonstrated significant anti-KLH responses at baseline.
2.3.2. KLH stimulation assays
Peripheral blood mononuclear cells (PBMCs) were
cryopreserved until the completion of each experiment
so that samples collected longitudinally from each subject could be assessed in parallel. The blood collection,
as well as freezing and thawing protocols, and proliferation assays were performed under strict endotoxinminimised conditions. Standard cryopreservation and
thawing methodology were employed (Gjerset et al.,
1986) with the exception that cells were resuspended in
supplemented RPMI 1640 (Flow Laboratories, Sydney,
Australia) that contained 20% autologous plasma.
Procedures for the antigen proliferation used standard
methods (Maluish and Strong, 1986; Snyder et al., 1993).
In brief, for the antigen stimulation assays, aliquots of
100 ll of the cell suspensions were added to quadruplicate
wells of round-bottomed microtitre cell culture plates
(Nunc, Roskilde, Denmark) in addition to 100 ll of KLH
in supplemented RPMI 1640 at final concentrations of 0,
0.01, 0.05, and 0.1 mg/ml. Each plate included samples
from all time points of one subject plus cells from a positive control subject (multiply immunized with KLH) to
ensure reliability of the assay. The cultures were incubated at 37 °C and 5% CO2 in air for 5 days before
measurement of incorporation of tritiated thymidine.
Stimulation indices are reported for the highest concentration of KLH in vitro. A stimulation index of greater
than 5.0 was deemed a positive response. Results are reported for the highest concentration of KLH in vitro.
2.3.3. Enzyme-linked immunosorbent assay
Sera were frozen in aliquots at )20 °C until the
completion of the experiment so that all samples from
each subject could be included in the same assay. The
procedures employed standard ELISA methods (Snyder
67
et al., 1990; Voller and Bidwell, 1986). In brief, ELISA
plates (Polysorp, Nunc, Roskilde, Denmark) were
coated with 1 lg/ml of KLH in a carbonate buffer (pH
9.6) and kept at 37 °C for 2 h and then overnight at 4 °C.
The plates were blocked with 1% non-fat dry milk
powder in PBS (200 ll per well). Sera at concentrations
of 1/100 or 1/1000 in blocking buffer plus 0.05% Tween
20 were added to the wells (100 ll per well). The detection antibody (horseradish peroxidase conjugated rabbit
anti-human IgG; Dako, Glostrup, Denmark) was added
at a concentration of 1/500 in blocking buffer plus 0.05%
Tween 20 (100 ll per well). The optical densities (ODs)
of the wells were read using an automated plate reader
(Multiskan PLUS Mk II, Titertek, Sydney, Australia)
with a filter at 405 nm.
Consistent with the widespread adoption of an
ELISA OD criterion for seroconversion (as opposed to
end-point titers) in clinical practice, seroconversion was
defined as an OD level three standard deviations above
the mean of six unimmunized control subjects. These
negative control sera and a positive control serum were
routinely measured with each ELISA run to control for
assay-to-assay variability. Results are reported for sera
concentrations of 1/100.
2.4. Kinetics and rate of response to KLH
To examine the kinetics of the development of immune responses to KLH, six subjects were immunized
with KLH. Blood was drawn for analysis of T-cell
proliferation against KLH and assessment of anti-KLH
IgG antibody responses at the time of immunization
(referred to as ‘‘baseline’’) and then weekly for four
subsequent weeks.
The rate of response to KLH/alum was assessed in a
further 15 subjects. The DTH skin test to KLH was
measured and blood was drawn for in vitro analyses at
immunization and again three weeks later (‘‘follow-up’’).
3. The effects of distress on anti-KLH immunity
3.1. Subjects
To evaluate the sensitivity of the KLH model to
distress, 45 medical students, aged between 21 and 31
years (mean age ¼ 22.4, SD ¼ 1:85) were allocated to a
control (N ¼ 23) or experimental (N ¼ 22) group according to their availability for follow-up assessment.
There were 13 males and 10 females in the control group
and 10 males and 12 females in the experimental group.
3.2. Measures of distress
The main purpose of immunizing subjects at either a
time of low or high stress was to increase variance in the
68
A. Smith et al. / Brain, Behavior, and Immunity 18 (2004) 65–75
sample. The assumption underlying this research was
that it is the effect of stressor exposure (level of distress)
experienced by the participant that determines immune
function, rather than just that exposure. Thus, the profile of mood states (POMS) questionnaire was administered to measure current psychological distress. Total
POMS scores are presented. Subjects were asked to rate
how well 65 individual descriptors of feelings (e.g.,
‘‘tense,’’ ‘‘blue,’’ and ‘‘exhausted’’) applied to them over
the past week, including the current day, on a five-point
rating scale from 0 (‘‘not at all’’) to 4 (‘‘extremely’’). The
reliability and construct validity of this scale have been
demonstrated (McNair et al., 1981).
3.3. Behavioral assessment
Behavioral factors known to affect immune function
were also considered (Kiecolt-Glaser and Glaser, 1988;
Palmblad, 1981). Subjects were asked to report the
number of cigarettes smoked daily, their average weekly
alcohol consumption, amount of aerobic exercise, and
sleep changes in the past week and any weight loss over
the previous month. Female students recorded the stage
of their menstrual cycle. The exclusion criteria ensured
that participants with either acute or chronic illnesses, or
taking immuno-suppressive medication were not included in the study.
3.4. Assessment of anti-KLH immunity
The procedures outlined for the protocol development stage were followed, with the exception that a
further concentration of KLH (0.25 mg/ml) was added
to the in vitro T-cell proliferation assays.
Consequently, experimental subjects were immunized
within 15 min of the completion of a yearly viva voce
examination, which is widely regarded by the students as
being stressful. Control subjects were immunized in
normal examination-free term time.
Prior to the immunization, an intra-dermal dose of
KLH was administered to evaluate DTH responsiveness
and blood was drawn for in vitro assays, as well as
cortisol and noradrenaline assays. Demographic information was collected and the self-report instruments
were completed. Three weeks later, DTH skin test responses to KLH were recorded again and blood was
drawn for the in vitro anti-KLH measures. Responses
to the distress and behavioral questionnaires were
recorded.
As the examination schedule constrained the timing
of immunization, experimental subjects were stratified
as being immunized in the morning (before 1:00 p.m.;
n ¼ 10) or afternoon (after 1:00 p.m.; n ¼ 12) with
control subjects immunized either in the morning
(n ¼ 14) or afternoon (n ¼ 9). The stratification designated by time of collection was maintained for casecontrol pairs in the KLH stimulation assays.
3.7. Statistical analysis
Comparisons of the development of anti-KLH immunity between the control and experimental groups
were performed using t tests or the v2 statistic (for categorical data). The prediction of immune responses
from individual factors was performed using point biserial correlations. The influence of group membership
and behavioral factors on any observed relationships
was assessed using direct logistic regression.
3.5. Hormonal assays
4. Results
Sera for use in the cortisol and noradrenaline assays
were stored at )70 °C. Serum cortisol levels were determined by radioimmunoassay using commercially
available kits (Orion Diagnostica, Espoo, Finland) by
the Department of Endocrinology, The Prince of Wales
Hospital, Sydney. Noradrenaline assays were conducted
using high-performance liquid chromatography by the
Department of Clinical Chemistry, The Prince of Wales
Hospital, Sydney.
3.6. Experimental protocol
The weight of evidence from animal models suggests
that antigen encounter must coincide with, or follow
shortly after the stressor (within approximately 24 h), if
immuno-modulatory effects are to be observed (Blecha
and Kelley, 1981; Cocke et al., 1993; Coe et al., 1987;
Fleshner et al., 1992; Hibma and Griffin, 1994; Irwin,
1993; Okimura and Nigo, 1986; Wood et al., 1993).
4.1. Protocol development
The immunization procedure was well-tolerated by
all subjects. The injection produced minimal, transient
discomfort in all but one of 21 subjects who reported
local tenderness, which lasted for three days. Proliferative responses and an anti-KLH IgG antibody response
were evident in all but one subject by week 3. Based on
these data, it was decided to assess immune function at 3
weeks post-immunization in subsequent experiments.
The response rate using the specified dose of KLH
adsorbed to alum was 73% for both the development of
anti-KLH IgG responses and DTH skin test responses
against KLH. Eighty percent of subjects developed
proliferation responses in vitro of T cells against KLH.
These response rates were deemed appropriate to allow
an examination of the immune enhancing and suppressive effects of psychological variables. In addition, it was
A. Smith et al. / Brain, Behavior, and Immunity 18 (2004) 65–75
69
Fig. 1. Mean induration in response to delayed-type hypersensitivity skin test three weeks post-immunization for control (n ¼ 23) and experimental
(n ¼ 21) subjects.
anticipated that the response rate would not be so great
as to mask the immuno-modulatory effects of these
variables (Vedhara and Nott, 1996).
4.2. The effects of distress on anti-KLH immunity
4.2.1. Group comparisons
POMS. At baseline, subjects in the experimental
group (M ¼ 60:1, SD ¼ 42:9) displayed greater negative
mood than the control group (M ¼ 26:7, SD ¼ 29:3)
tð43Þ ¼ 3:1; p ¼ :004. At follow-up, there were also
significant differences between the control (M ¼ 37:8,
SD ¼ 37:6) and experimental (M ¼ 6:9, SD ¼ 25:2)
subjects on the total POMS score, with the control
subjects reporting higher levels of negative mood than
the experimental subjects tð43Þ ¼ 3:2; p ¼ :002.
DTH responses to KLH. At baseline, no subject displayed a DTH response. Of 45 subjects, 41 reported zero
induration. At follow-up, the mean induration in the
control group (n ¼ 23) was 5.6 mm (SD ¼ 4:6) while in
the experimental group (n ¼ 21) the mean response was
5.4 (SD ¼ 7:8). This difference was not significant
tð42Þ ¼ :088, p ¼ :93. The pattern of responses at follow-up is reported in Fig. 1 at which time 68% (30 of 44)
subjects developed DTH responses. Significantly more
control subjects (19 of 23) developed these responses to
KLH than experimental subjects (11 of 21)
v2 ð1; N ¼ 44Þ ¼ 4:62, p ¼ :03.
Lymphocyte proliferation against KLH. At baseline
the mean stimulation index for the control group
(n ¼ 23) was 3.8 (SD ¼ 1:5) and for the experimental
group (n ¼ 22) it was 3.3 (SD ¼ 1:0). This difference was
not significant tð43Þ ¼ :717, p ¼ :48. At follow-up, the
mean stimulation index for the control group was 5.6
(SD ¼ 3:5) and for the experimental group (n ¼ 22) the
mean index was 8.1 (SD ¼ 6:1). These differences were
not significant tð43Þ ¼ :79, p ¼ :44. The proliferation
responses are presented in Figs. 2 and 3. Considerable
variability of responses is evident, both at baseline and
follow-up. At baseline, seven subjects (five control and
two experimental) had stimulation indices greater than
5. At follow-up, 56% (25 of 45) subjects developed
proliferation responses against KLH. The number of
subjects who developed a proliferative response in the
control (n ¼ 13) and experimental (n ¼ 15) groups did
not differ v2 ð1; N ¼ 45Þ ¼ 1:14, p ¼ :29.
Anti-KLH IgG antibodies. The mean optical densities
in the ELISA at baseline for the control group (n ¼ 23)
was 0.208 (SD ¼ 0:14) and for the experimental group
(n ¼ 22) the value was 0.188 (SD ¼ 0:20). At follow-up
the mean for the control group (n ¼ 23) was 1.061
(SD ¼ 0:88) and for the experimental group (n ¼ 23) it
was 1.151 (SD ¼ 0:82). These differences were not significant at baseline [(tð43Þ ¼ :396, p ¼ :69] or at followup [tð43Þ ¼ :357, p ¼ :72]. The variability of responses at
Fig. 2. Stimulation indices for proliferation of T cells against KLH
in vitro at baseline for control (n ¼ 23) and experimental (n ¼ 22)
subjects.
70
A. Smith et al. / Brain, Behavior, and Immunity 18 (2004) 65–75
veloping antibody responses did not differ between the
control (n ¼ 13) and experimental (n ¼ 12) groups
v2 ð1; N ¼ 45Þ ¼ 0:02, p ¼ :89.
Fig. 3. Stimulation indices for proliferation of T cells against KLH in
vitro at three weeks post-immunization for control (n ¼ 23) and experimental (n ¼ 22) subjects.
baseline and follow-up is illustrated in Figs. 4 and 5.
One subject was seropositive at baseline, although, this
subject did not have a proliferation or DTH response at
that time. In total, 56% (25 of 45) of subjects developed
anti-KLH IgG responses. The number of subjects de-
Fig. 4. Optical densities in ELISAs for anti-KLH IgG levels at baseline
for control (n ¼ 23) and experimental (n ¼ 22) subjects.
Fig. 5. Optical densities in ELISAs for anti-KLH IgG levels at three
weeks post-immunization for control (n ¼ 23) and experimental
(n ¼ 22) subjects.
4.2.2. Prediction of immune function from individual
difference factors
POMS. Table 1 provides correlations between the
POMS scales (at baseline and follow-up) and the indices
of anti-KLH immunity at follow-up. The POMS score
at follow-up is included since there is evidence in animal
models that psychological state at the time of immune
assessment may influence immune competence (Dhabhar and McEwen, 1996). Distress at baseline and the
likelihood of developing a DTH skin test response three
weeks later were negatively correlated (r ¼ :45,
p ¼ :002). There were no correlations between psychological state at baseline and proliferation against KLH
in vitro (r ¼ :02, p ¼ :90) or the development of antibody responses (r ¼ :04, p ¼ :79).
Hormonal measures. The relationship between levels
of noradrenaline at baseline and the development of
cutaneous DTH against KLH three weeks later was
examined using a point biserial correlation. The correlation between noradrenaline levels and the POMS
scores at baseline was .29 (p ¼ :05). There were no other
significant correlations between the physiological indices
of distress (levels of noradrenaline and cortisol) at
baseline and the development of anti-KLH 3 weeks later
(Table 2).
Behavioral and demographic measures. For both the in
vitro proliferation of T cells against KLH and the development of anti-KLH IgG antibodies, there was only
one significant association with the behavioral and
demographic measures. KLH-stimulated proliferation
responses were more likely to develop in females
(n ¼ 16) than in males (n ¼ 9) v2 ð1; N ¼ 45Þ ¼ 5:14,
p ¼ :02. There were no correlations between the DTH
skin test responses against KLH and continuous behavioral variables (Table 3). DTH responses were not
related to sleep disturbance at baseline [v2 ð1; N ¼
44Þ ¼ 0:47, p ¼ :66, FisherÕs exact test] or at follow-up
[v2 ð1; N ¼ 44Þ ¼ 0:09, p ¼ 1:0, FisherÕs exact test]. The
subjectÕs sex [v2 ð1; N ¼ 44Þ ¼ 0:00, p ¼ 1:00] and racial
group membership [v2 ð1; N ¼ 43Þ ¼ 0:01, p ¼ :95] were
unrelated to the development of DTH skin test responses against KLH.
Analyses with subjects positive to KLH at baseline
removed. Seven subjects displayed proliferative responses against KLH and one demonstrated significant
anti-KLH IgG levels at baseline. The major results of
the present study were re-analyzed without these 8
subjects included leaving a total of 37 subjects (18 control and 19 experimental).
Again, there were no significant differences between
the control (n ¼ 18) and experimental (n ¼ 19) subjects in
the development of anti-KLH IgG [v2 ð1; N ¼ 37Þ ¼ 0:04,
A. Smith et al. / Brain, Behavior, and Immunity 18 (2004) 65–75
71
Table 1
Correlations between the POMS at immunization (I) and follow-up (FU), and the categorical indices of anti-KLH immunity
POMS (I)
POMS (FU)
DTH to KLH
Proliferation
IgG antibodies
*
POMS (I)
POMS (FU)
DTH
Proliferation
IgG
—
.32
—
—
).45
).07
—
—
—
.02
.01
).06
—
—
—
—
).04
.01
).02
).01
—
—
—
—
—
Significant at the .05 a-level.
Significant at the .01 a-level.
**
Table 2
Correlations between cortisol (N ¼ 44) and noradrenaline (N ¼ 44) at baseline and categorical anti-KLH measures at follow-up
DTH to KLH
Proliferation
Anti-KLH IgG
Cortisol
Noradrenaline
DTH to KLH
Proliferation in vitro
Anti-KLH IgG
Cortisol
Noradrenaline
—
).06
—
—
).01
).01
—
—
—
.18
.05
).10
—
—
—
—
.31
.22
.09
).01
—
—
—
—
—
Levels of noradrenaline at the time of immunization predicted the development of DTH responses against KLH 3 weeks later (p ¼ :04). There
were no other relationships between the likelihood of developing a DTH response, proliferation against KLH in vitro or seroconversion and levels of
cortisol and noradrenaline.
*
Significant at the .05 a-level.
Table 3
Correlations between health related behaviors at immunization and follow-up and categorical DTH skin test responses against KLH (N ¼ 45)
Variable
DTH
Immunization
Cigarettes
Alcohol
Exercise
Weight loss
Follow-up
Cigarettes
Alcohol
Exercise
Weight loss
DTH
Immunization
Follow-up
Cig
Alc
—
).03
).11
—
—
.02
—
—
—
—
Ex
Wgt
Cig
Alc
Ex
Wgt
.27
.19
).03
).26
.04
.17
—
).05
.16
—
—
—
.00
).13
).23
—
—
—
—
.99
.01
).07
.02
.16
.81
.09
).09
.04
.08
.47
.03
).05
.12
.15
.31
—
—
—
—
—
—
.15
—
—
—
—
—
—
—
.03
.19
—
—
—
—
—
—
—
—
).07
).03
.03
—
—
—
—
—
—
—
—
—
No correlations between behavioral variables and the development of DTH skin test responses against KLH were evident.
p ¼ :84] or in control and experimental subjects in the
proliferation of T cells against KLH in vitro
[v2 ð1; N ¼ 37Þ ¼ 2:17, p ¼ :14] at follow-up. With this
smaller number of subjects, the differences between
control and experimental subjects who demonstrated
DTH responses against KLH at follow-up was significant
[v2 ð1; N ¼ 36Þ ¼ 4:50, p ¼ :03].
There was no relationship between the POMS at
baseline and the levels of anti-KLH antibodies (r ¼ :13,
p ¼ :47) or T-cell proliferation (r ¼ :14, p ¼ :42) at follow-up. The correlation between these POMS scores at
baseline and cutaneous DTH at follow-up was .40
(p ¼ :02).
Logistic regression. The possibility that the relationship between distress and DTH skin test responses was
actually accounted for by behavioral/demographic factors was further explored using direct logistic regression.
Data was available for 45 subjects. Analyses were performed using SPSS 9.0.
Group status was entered in the logistic regression
equation to control for differential treatment of the
subjects in the control and experimental groups. To reduce the number of variables included in the logistic
regression, only behavioral or demographic variables
related to the immune measure at the .1 level of significance were considered for inclusion in the model
72
A. Smith et al. / Brain, Behavior, and Immunity 18 (2004) 65–75
(Table 3). Levels of aerobic exercise at baseline and alcohol consumption at follow-up were significant at this
criterion level. However, only 4.4% (2) of subjects engaged in more than 5 h of aerobic exercise per week so
this variable was not considered further. Alcohol consumption varied considerably (M ¼ 4:2, SD ¼ 10:2) at
follow-up and has been associated with a decrement in
immune function (Glassman et al., 1985; Roselle and
Mendenhall, 1982; Saxena et al., 1986; Stacey, 1984) and
so was included in the logistic regression analysis.
The model was significant [v2 ð3; N ¼ 44Þ ¼ 12:35,
p ¼ :01] and correctly classified 93% of responders and
50% of non-responders (80% overall). A model without
group status included was also considered. This model
was not significantly different from the model that included group status v2 ð1; N ¼ 44Þ ¼ 0:21, p > :05. This
model correctly identified 93% of responders and 57% of
non-responders. The Wald statistic was significant for
the POMS (z ¼ 7:08, p ¼ :01), but alcohol consumption
at follow-up was not a significant predictor of the outcome (z ¼ 1:31, p ¼ :25). However, removal of alcohol
from the regression equation reduced the number of
correctly classified cases to 75%.
4.2.3. Relationships between immune measures
There were no significant associations in the likelihood of responding to KLH between the two measures
of cellular immune function [v2 ð1; N ¼ 44Þ ¼ :17;
p ¼ :68], between anti-KLH IgG levels and KLH-stimulated T-cell proliferation [v2 ð1; N ¼ 45Þ ¼ 0:00;
p ¼ :95] or between the IgG levels and DTH skin test
responses against KLH [v2 ð1; N ¼ 44Þ ¼ 0:00; p ¼ :98].
The correlation between pre- and post-immunization
KLH-stimulated proliferation in vitro was .35 (p ¼ :02)
and between pre- and post-anti-KLH antibody responses it was .53 (p ¼ :00).
5. Discussion
The experimental manipulation was successful in increasing variability in distress amongst the subjects at
baseline. In addition, at follow-up, the experimental, but
not the control, subjects had commenced the end-ofsemester vacations and consequently were considerably
less distressed than the control subjects. This was not an
ideal experimental design, but it did have the advantage
of allowing assessment of a range of distress measures
(very low to high) at different stages in the development
of an immune response. Individuals who were exposed
to an examination were less likely to develop DTH skin
test responses against KLH, although the other measures of anti-KLH immunity were not affected.
However, the emphasis in this research was on the
capacity of individual factors to predict anti-KLH immunity, rather than on group differences. Psychological
distress at immunization affected the in vivo cellular
immune measure (DTH) rather than in vitro T-cell
proliferation or antibody production against KLH at
follow-up. This influence on DTH was independent of
behavioral and demographic factors (Table 3), although
the role of alcohol consumption at follow-up requires
further investigation. The reduction in DTH skin test
responses against KLH suggests that there was a deficit
in production of antigen-specific activated and memory
T cells at baseline that initiate the DTH response three
weeks after immunization (Roitt et al., 1998). There was
no evidence of an effect of psychological variables at the
time of follow-up on anti-KLH immunity (Table 1).
The development of DTH skin test responses against
KLH in vivo and KLH-stimulated T-cell proliferation in
vitro both depend on the presence of memory and/or
effector T cells specific for KLH and may therefore be
expected to be correlated. However, the lack of a relationship between the two indices of cellular immunity
has been demonstrated previously (Curtis et al., 1970)
and may reflect the different compartments of the immune system (skin versus blood) and the different APCs
(principally Langerhans cells in DTH versus monocytes
and B cells in the in vitro stimulation) involved in these
measures.
The failure to find a relationship between the cellular
and humoral arms of the immune response against KLH
may reflect differences in T-cell and B-cell epitopes. This
possibility is supported by the observation that subjects
positive for T-cell proliferation at baseline (probably as
a result of cross-reacting epitopes) were not also positive
for the development of anti-KLH IgG antibodies. In
addition, the concept of a divergence of DTH and humoral immunity is recognised. Notably, the strict definition of the term anergy is that it is a deficit in the
formation of DTH in an individual with an intact antibody response (Dwyer, 1984). In addition, a dose-dependent, inverse relationship between DTH and
vigorous antibody responses (immune deviation) has
been observed (Parish, 1971; Parish, 1996; Parish and
Liew, 1972). Immune deviation has been recently re-interpreted in terms of genetically or environmentally
determined individual differences in the propensity to
develop TH 1 and TH 2 responses. A number of host
and antigen factors influence this propensity (Seder
and Mosmann, 1999). Alternatively, it may be that
correlations between anti-KLH antibodies at follow-up
and levels of distress at the time of immunization
may have been detected if the kinetics of antibody
production was examined over each of the three weeks
post-immunization.
Noradrenaline (but not cortisol) levels at baseline
were correlated with the likelihood of developing DTH
against KLH at follow-up. In addition, these hormone
levels were correlated with levels of distress at baseline,
suggesting a mechanism by which mood may affect
A. Smith et al. / Brain, Behavior, and Immunity 18 (2004) 65–75
immune function in this sample. These findings are
particularly interesting given that it was not possible to
exercise complete control over the time of day that the
subjects were immunized, resulting in a potential influence on the results of diurnal variations in hormonal
levels (K
agedal and Goldstein, 1988; Rosenfeld et al.,
1971). Furthermore, the lack of a correlation between
DTH and cortisol levels may have been due to the occurrence of threshold effects rather than a linear relationship between the variables (Coe et al., 1987). Neither
noradrenaline nor cortisol levels at baseline were correlated with KLH stimulated proliferation or antibody
responses.
Perhaps the most interesting question arising from
this research is the extent to which alterations in DTH in
response to distress reflect an immunological disturbance of clinical significance. Cutaneous DTH is simply
a convenient representation of systemic hypersensitivity
in vivo (Dwyer, 1984) and the likelihood that reductions
in DTH against KLH may be clinically significant is
suggested on a number of grounds. DTH skin tests are
widely used as an indication of the capacity to induce
cell-mediated immunity, and hence relative protection
against tuberculosis (Beck, 1991). These responses have
been of prognostic significance (Dwyer, 1984; Hortobagyi et al., 1981; McLean, 1988; Poenaru and Christou, 1991), although DTH skin test responses are not
always predictive of the outcome of disease states (Golub et al., 1974). Overall, these data emphasise the value
of including an in vivo measure of immune function in
PNI research (Hickie et al., 1990), particularly when the
focus of the research is on the clinical significance of
alterations in immune status in response to psychological perturbations (Vedhara et al., 1999).
The variance accounted for by negative affect at the
time of immunization was 18% (Table 1). Thus, the
magnitude of the effect of distress on the development of
DTH skin test responses to KLH in this sample of
healthy, young students was not large. With more profound or prolonged levels of distress, or when individuals are already immuno-compromised (such as in the
elderly or medically ill) the effect is likely to be more
significant (Cohen et al., 1998). Furthermore, the sample
used in this study was relatively homogeneous medically, psychologically and behaviorally. This homogeneity may have acted against the capacity to detect
distress/immune relationships.
Assessing the level of pre-immunization anti-KLH
immunity is important in determining whether a primary immune response is being observed at follow-up.
Pre-immunization responses to a truly novel antigen are
not expected. However, seven subjects demonstrated
significant proliferation responses against KLH at
baseline and another subject demonstrated significant
levels of anti-KLH antibodies at baseline. There was
also considerable variability in the immune responses of
73
subjects at baseline (Figs. 2 and 4). These data may well
account for the correlations between pre- and post-immunization KLH-stimulated proliferation in vitro
(r ¼ :35; p ¼ :02) and anti-KLH antibody responses
(r ¼ :53; p ¼ :00) by way of a quasi-secondary immune
response. One means by which to deal with these results
in psychoneuroimmunology research would be to eliminate from analyses those subjects developing baseline
responses. In the present study, this procedure produced
results consistent with the larger group findings, indicating that baseline levels of response had little effect on
overall conclusions. However, in research examining
mechanisms of a primary immune response removal
from analyses of subjects positive at baseline would be
essential. The stimulation index adopted in the research
reported here was a very conservative one since the aim
was to detect biologically significant effects of distress on
anti-KLH immunity. More liberal indices may be appropriate with other research questions.
In conclusion, immunization of individuals with
KLH provides information about the capacity to induce
an immune response in vivo. Cellular rather than humoral immune responses were affected by distress levels
and the effect may have been mediated by noradrenaline. The occurrence of distress at the time of sensitization rather than elicitation appears to bring about these
alterations in immune competence. Further studies
to elucidate the cellular and molecular basis of this
distress-induced reduction in immunity in humans are
warranted.
Acknowledgments
This study was supported by the National Health and
Medical Research Council of Australia (Program Grant
No. 953208) and a University of Western Sydney
Foundation Grant.
References
Amkraut, A.A., Malley, A., Begley, D., 1969. Immunogenicity of
hemocyanins and their subunits. J. Immunol. 103 (6), 1301–1310.
Beck, J.S., 1991. Skin changes in the tuberculin test. Tubercle 72, 81–
87.
Black, P.H., 1994. Central nervous system–immune system interactions: psychoneuroendocrinology of stress and its immune consequences. Antimicrob. Agents Chemother. 38 (1), 1–6.
Blecha, F., Kelley, K.W., 1981. Effects of cold and weaning stressors
on the antibody-mediated immune response of pigs. J. Anim. Sci.
53 (2), 439–447.
Burke, G.P., Smith, K.A., Stocking, R.I.G., Ferm, M., McIntyre,
O.R., 1977. Anti-keyhole limpet hemocyanin antibody in normal
unsensitized individuals. J. Allergy Clin. Immun. 59 (4), 309–313.
Cocke, R., Moynihan, J.A., Cohen, N., Grota, L.J., Ader, R., 1993.
Exposure to conspecific alarm chemosignals alters immune responses in BALB/c mice. Brain Behav. Immun. 7, 36–46.
74
A. Smith et al. / Brain, Behavior, and Immunity 18 (2004) 65–75
Coe, C.L., Rosenberg, L.T., Fischer, M., Levine, S., 1987. Psychological factors capable of preventing the inhibition of antibody in
separated infant monkeys. Child Dev. 58, 1420–1430.
Cohen, S., Frank, E., Doyle, W.J., Skoner, D.P., Rabin, B.S.,
Gwaltney, J.M., 1998. Types of stressors that increase susceptibility
to the common cold in healthy adults. Health Psychol. 17 (3), 214–
223.
Curtis, J.E., Hersh, E.M., Harris, J.E., McBride, C., Freireich, E.J.,
1970. The human primary immune response to keyhole limpet
haemocyanin: interrelationships of delayed hypersensitivity, antibody response and in vitro blast transformation. Clin. Exp.
Immunol. 6, 473–491.
Dhabhar, F.S., McEwen, B.S., 1996. Stress-induced enhancement of
antigen-specific cell-mediated immunity. J. Immunol. 156, 2608–
2615.
Dwyer, J.M., 1984. Anergy. Prog. Allergy 35, 15–92.
Fleshner, M., Watkins, L.R., Lockwood, L.L., Bellgrau, D., Laudenslager, M.L., Maier, S.F., 1992. Specific changes in lymphocyte
subpopulations: a potential mechanism for stress-induced immunomodulation. J. Neuroimmunol. 41, 131–142.
Frankenhaeuser, M., 1983. The sympathetic-adrenal response to
challenge: comparison between the sexes. In: Dembroski, T.M.,
Schmidt, T.H., Blumchen, G. (Eds.), Biobehavioural Bases of
Coronary Heart Disease. Karger, Basel, pp. 91–105.
Frick, O.L., Shimbor, C., 1970. Cross reactivity between hemocyanins
and erythrocytes. Fed. Proc. 29, 573Abs.
Gjerset, G., Nelson, K., Strong, D., 1986. Methods for cryopreserving
cells. In: Rose, N., Friedman, H., Fahey, J.L. (Eds.), Manual of
Clinical Laboratory Immunology, third ed. American Society of
Microbiology, Washington, DC, pp. 61–67.
Glaser, R., Kiecolt-Glaser, J.K., Bonneau, R.H., Malarkey, W.,
Kennedy, S., Hughes, J., 1992. Stress-induced modulation of the
immune response to recombinant hepatitis B vaccine. Psychosom.
Med. 54, 22–29.
Glassman, A.R., Bennett, C.E., Randall, C.L., 1985. Effects of ethyl
alcohol on human peripheral lymphocytes. Arch. Pathol. Lab.
Med. 109, 540–542.
Gmur, J.P., Deluigi, G., Sauter, C., Straub, P.W., 1976. Correlation of
skin test reactivity and in vitro lymphocyte stimulation by soluble
antigens in normal subjects and leukemic patients. Eur. J. Cancer
12, 625–631.
Golub, S., OÕConnell, T.X., Morton, D.L., 1974. Correlation of in vivo
and in vitro assays of immunocompetence in cancer patients.
Cancer Res. 34, 1833–1837.
Green, A.A., Borella, L., 1971. In vitro response of human leukocytes
to associated and dissociated hemocyanin. J. Immunol. 107 (1),
293–297.
Harris, J.R., Markl, J., 1999. Keyhole limpet hemocyanin (KLH): a
biomedical review. Micron 30, 597–623.
Hersh, E.M., Whitecar, J.P., McCredie, K.B., Bodey, G.P., Freireich,
E.J., 1971. Chemotherapy, immunocompetence, immunosuppresssion and prognosis in acute leukemia. New Engl. J. Med. 285 (22),
1211–1216.
Hibma, M., Griffin, J.F.T., 1994. The influence of maternal separation
on humoral and cellular immunity in farmed deer. Brain Behav.
Immun. 8, 80–85.
Hickie, I.B., Silove, D., Hickie, C., Wakefield, D., Lloyd, A.R., 1990.
Is there immune dysfunction in depressive disorders? Psychol. Med.
20, 755–761.
Hortobagyi, G.N., Smith, T.L., Swenerton, K.D., Legha, S.S.,
Buzdar, A.U., Blumenschein, G.R., Gutterman, J.U., Hersh,
E.M., 1981. Prognostic value of prechemotherapy skin tests in
patients with metastatic breast carcinoma. Cancer 47 (6), 1369–
1376.
Irwin, M., 1993. Brain corticotropin-releasing hormone- and interleukin-1b-induced suppression of specific antibody production. Endocrinology 133 (3), 1352–1360.
K
agedal, B., Goldstein, D.S., 1988. Catecholamines and their metabolites. J. Chromatogr. 429, 177–233.
Kiecolt-Glaser, J.K., Glaser, R., 1988. Methodological issues in
behavioral immunology research with humans. Brain Behav.
Immun. 2, 67–78.
Kiecolt-Glaser, J.K., Glaser, R., 1992. Psychoimmunology: can
psychological interventions modulate immunity? J. Consult. Clin.
Psychol. 60 (4), 569–575.
Kiecolt-Glaser, J.K., Glaser, R., Gravenstein, S., Malarkey, W.B.,
Sheridan, J., 1996. Chronic stress alters the immune response to
influenza virus vaccine in older adults. Proc. Natl. Acad. Sci. USA
93, 3043–3047.
Kiecolt-Glaser, J.K., Marucha, P.T., Malarkey, W.B., Mercado, A.M.,
Glaser, R., 1995. Slowing of wound healing by psychological stress.
Lancet 346, 1194–1196.
Maluish, A., Strong, D., 1986. Lymphocyte proliferation. In: Rose, N.,
Friedman, H., Fahey, J. (Eds.), Manual of Clinical Laboratory
Immunology, third ed. American Society of Microbiology, Washington, DC, pp. 274–281.
McLean, L.D., 1988. Delayed type hypersensitivity testing in surgical
patients. Surg. Gynecol. Obstet. 166, 284–293.
McNair, D.M., Lorr, M., Droppleman, L.F., 1981. Profile of Mood
States: Manual. Educational and Industrial Testing Service, San
Diego, CA.
Moroz, L.A., Krygier, V., Kotoulas, A.O., 1973. Normal human IgG
and antibody activity for keyhole limpet haemocyanin. Immunology 25, 441–449.
Okimura, T., Nigo, Y., 1986. Stress and immune responses. I.
Supression of T cell function in restraint-stressed mice. Jpn. J.
Pharmacol. 40, 505–511.
Palmblad, J., 1981. Stress and immunologic competence: studies in
man. In: Ader, R., Felten, D., Cohen, N. (Eds.), Psychoneuroimmunology. Academic, New York, pp. 229–257.
Parish, C.R., 1996. Immune deviation: a historical perspective.
Immunol. Cell Biol. 74, 449–456.
Parish, C.R., 1971. Immune response to chemically modified flagellin
II. Evidence for a fundamental relationship between humoral and
cell-mediated immunity. J. Exp. Med. 134, 21–47.
Parish, C.R., Liew, F.Y., 1972. Immune response to chemically
modified flagellin III. Enhanced cell-mediated immunity during
high and low zone antibody tolerance to flagellin. J. Exp. Med. 135,
298–311.
Poenaru, D., Christou, N.V., 1991. Clinical outcome of seriously ill
surgical patients with intra-abdominal infection depends on both
physiologic (APACHE II score) and immunologic (DTH score)
alterations. Ann. Surg. 213 (2), 130–136.
Roitt, I., Brostoff, J., Male, D., 1998. Immunology, fifth ed. Mosby,
London.
Roselle, G.A., Mendenhall, C.L., 1982. Alteration of in vitro human
lymphocyte function by ethanol, acetaldehyde and acetate. J. Clin.
Lab. Immunol. 9, 33–37.
Rosenfeld, R.S., Hellman, L., Roffwarg, H., WFeitzman, E.D.,
Fukushima, D.K., Gallagher, T.F., 1971. Dehydroisoandrosterone
is secreted episodically and synchronously with cortisol by normal
man. J. Clin. Endrocrinol. 33, 87–92.
Saxena, Q.B., Saxena, R.K., Adler, W., 1986. Inhibition of human killer
cell activity by ethanol in vitro. Indian J. Exp. Biol. 24, 561–564.
Seder, R.A., Mosmann, T.M., 1999. Differentiation of effector
phenotypes of CD4þ and CD8þ T cells. In: Paul, W.E. (Ed.),
Fundamental Immunology, fourth ed. Lippincott–Raven, Philadelphia, pp. 879–908.
Snyder, B.K., Roghmann, K.J., Sigal, L.H., 1990. Effect of stress and
other biopsychosocial factors on primary antibody response. J.
Adolesc. Health Care 11, 472–479.
Snyder, B.K., Roghmann, K.J., Sigal, L.H., 1993. Stress and psychosocial factors: effects on primary cellular immune response. J.
Behav. Med. 16 (2), 143–161.
A. Smith et al. / Brain, Behavior, and Immunity 18 (2004) 65–75
Stacey, N., 1984. Inhibition of antibody dependent cell mediated
cytotoxicity by ethanol. Immunopharmacology 8, 155–161.
Vedhara, K., Fox, J.D., Wang, E.C.Y., 1999. The measurement of
stress related immune dysfunction in psychoneuroimmunology.
Neurosci. Biobehav. Rev. 23, 699–715.
Vedhara, K., Nott, K., 1996. The assessment of the emotional and
immunological consequences of examination stress. J. Behav. Med.
19 (5), 467–478.
Voller, A., Bidwell, D., 1986. Enzyme-linked immunosorbent assay.
In: Rose, N.R., Friedman, H., Fahey, J.L. (Eds.), Manual of
Clinical Laboratory Immunology, third ed. American Society of
Microbiology, Washington, DC, pp. 99–109.
75
Wirguin, I., Suturkova-Milosevic, L., Briani, C., Latov, N., 1995.
Keyhole limpet hemocyanin contains Gal(b1-3)-GalNAc determinants that are cross-reactive with the T antigen. Cancer Immunol.
Immunother. 40, 307–310.
Wood, P.G., Karol, M.H., Kusnecov, A.W., Rabin, B.S., 1993.
Enhancement of antigen-specific humoral and cell-mediated immunity by electric footshock stress in rats. Brain Behav. Immun. 7,
121–134.
Yuesheng, L., Rabello, A.L.T., Simpson, A.J.G., Katz, N., 1994. The
serological differentiation of acute and chronic Schistosoma japonicum infection by ELISA using keyhole limpet haemocyanin as
antigen. Trans. R. Soc. Trop. Med. Hyg. 88, 249–251.