Neuroendocrine and psychometric evaluation of a placebo version of the 'Trier Social Stress Test'.
ABSTRACT The "Trier Social Stress Test" (TSST) is one of the most prominent laboratory stress paradigms. It is often used to investigate the effects of stress on cognitive or affective parameters. Such studies need a non-stress control condition. However, control conditions currently employed are often rather ill defined and do not parallel important modulating variables, e.g., physical or cognitive load of the TSST. We here introduce a placebo version of the TSST, which contains a free speech and a simple mental arithmetic task without uncontrollability and social-evaluative threat. In two studies, this control condition was evaluated using salivary markers of stress reactivity (cortisol and alpha-amylase) and a questionnaire for anticipatory cognitive stress appraisal (PASA). In experiment 1 participants who were treated with the placebo condition showed no cortisol response and a small, but significant salivary alpha-amylase (sAA) response. Both responses were significantly smaller than those of TSST-treated participants. The placebo-treated participants also rated the treatment situation as less stressful. In experiment 2 a crossover study with the use of an intercom to instruct the participants and ensure their compliance was conducted. Again there was a strong cortisol response to the TSST, which differed significantly from the cortisol levels observed during the placebo condition. Importantly the cortisol response was not influenced by treatment order (TSST or placebo first). However, in this study we found similar reactions between TSST- and placebo-treated participants with regard to sAA-response. We suggest that the introduced placebo protocol for the TSST is a promising tool for future psychobiological research. The exact procedure for a given experiment should be tailored to the specific needs of the empirical question studied.
- SourceAvailable from: Dirk Hellhammer[show abstract] [hide abstract]
ABSTRACT: The measurement of cortisol in saliva provides the basic scientist as well as the clinician with a reliable tool for investigations of hypothalamus-pituitary-adrenal axis activity. Since saliva samples can be obtained stress-free and independent from medically trained personnel this method may be well suited for use in psychobiological studies. This overview intends to give a comprehensive introduction to the method of salivary cortisol assessment and to briefly discuss its application in different scientific disciplines.Neuropsychobiology 02/1989; 22(3):150-69. · 2.37 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Exposure of human plasma in vitro to gas-phase cigarette smoke (CS) causes a marked modification of plasma proteins as measured by protein carbonyl assay. Aldehydes present in CS may cause this elevation of protein carbonyls by reacting with sulfhydryl groups of proteins. Saliva is the first body fluid to confront the inhaled CS. Thus, in vitro exposure of saliva to nine "puffs" of CS also showed a distinct increase in protein carbonyls. Ascorbate and desferrioxamine mesylate had little effect on protein carbonyl formation, while GSH and N-acetylcysteine considerably inhibited the accumulation of protein carbonyls due to CS exposure. Following the exposure to CS, the activities of several salivary enzymes-amylase, lactic dehydrogenase (LDH), and acid phosphatase-were found to be significantly reduced (34, 57, and 77%, respectively). However, CS had no effect on the activities of aspartate aminotransferase and alkaline phosphatase. Addition of 1 mM of GSH and N-acetylcysteine considerably protected LDH and amylase activities, suggesting that sulfhydryl groups are affected in LDH and amylase. On the other hand, addition of 1 mM ascorbate caused a further loss of LDH and amylase activities, which could be partially prevented by the addition of desferrioxamine mesylate, implicating metal-catalyzed oxidation processes. Finally, loss of acid phosphatase activity was completely unaffected by any of the above antioxidants. It is concluded that the loss of salivary enzyme activities may be due to various agents in the CS that affect the enzyme activities via different mechanisms.Archives of Biochemistry and Biophysics 08/2000; 379(2):229-36. · 3.37 Impact Factor
- [show abstract] [hide abstract]
ABSTRACT: Some studies suggest that stress-induced effects of cortisol on memory are modulated by the valence of the stimuli to be learned and retrieved. The present study investigated the effect of acute stress-induced cortisol secretion on acquisition and retrieval of pleasant, unpleasant and neutral words. Sixty healthy men were randomly assigned to one of the three experimental groups. Participants were either exposed to a standardized laboratory stressor (the Trier Social Stress Test) before learning a wordlist, or before retrieval, or were not stressed. Free recall and recognition were tested 24 h later. Free recall was not affected by stress exposure. For recognition, there was no main effect of the stressor, but a main effect of valence and a valence by group interaction emerged: recognition for positive words was significantly impaired when subjects were stressed before retrieval. In addition, a positive correlation between the cortisol response and errors of commission was found. The results suggest that acute stress impairs memory for positive stimuli and that stress-induced cortisol secretion interferes with accuracy of memory retrieval, i.e. the ability to discriminate true memories from false ones.Stress 10/2004; 7(3):173-81. · 3.25 Impact Factor
This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
Author's personal copy
Neuroendocrine and psychometric evaluation of a
placebo version of the ‘Trier Social Stress Test’
S. Heta,1, N. Rohlederb,1, D. Schoofsa, C. Kirschbaumc, O.T. Wolfa,*
aDepartment of Cognitive Psychology, Ruhr-University Bochum, Germany
bDepartment of Psychology, Brandeis University, Waltham (MA), USA
cDepartment of Psychology, Dresden University of Technology, Dresden, Germany
Received 1 October 2008; received in revised form 13 January 2009; accepted 13 February 2009
Studies on psychological stress effects have used different
types of stress tasks, like emotion induction procedures,
public speaking tasks, cognitive tasks, noise exposure and
tasks which combine public speech and cognitive tasks
(Biondi and Picardi, 1999). Already Cannon (1935) empha-
Psychoneuroendocrinology (2009) 34, 1075—1086
Social stress test
paradigms. It is often used to investigate the effects of stress on cognitive or affective
parameters. Such studies need a non-stress control condition. However, control conditions
currently employed are often rather ill defined and do not parallel important modulating
variables, e.g., physical or cognitive load of the TSST. We here introduce a placebo version of
the TSST, which contains a free speech and a simple mental arithmetic task without uncontroll-
ability and social-evaluative threat. In two studies, this control condition was evaluated using
salivary markers of stress reactivity (cortisol and alpha-amylase) and a questionnaire for
anticipatory cognitive stress appraisal (PASA). In experiment 1 participants who were treated
withtheplacebocondition showed nocortisol response andasmall, butsignificant salivaryalpha-
amylase (sAA) response. Both responses were significantly smaller than those of TSST-treated
participants. The placebo-treated participants also rated the treatment situation as less stress-
ful. In experiment 2 a crossover study with the use of an intercom to instruct the participants and
ensure their compliance was conducted. Again there was a strong cortisol response to the TSST,
which differed significantly from the cortisol levels observed during the placebo condition.
Importantly the cortisol response was not influenced by treatment order (TSSTor placebo first).
with regard to sAA-response. We suggest that the introduced placebo protocol for the TSST is a
promising tool for future psychobiological research. The exact procedure for a given experiment
should be tailored to the specific needs of the empirical question studied.
# 2009 Elsevier Ltd. All rights reserved.
The ‘‘Trier Social Stress Test’’ (TSST) is one of the most prominent laboratory stress
* Corresponding author at: Institute of Cognitive Psychology, Uni-
versity of Bochum, Universitaetsstrasse 150, D-44780 Bochum, Ger-
many. Tel.: +49 234 32 28670; fax: +49 234 32 14308.
E-mail address: email@example.com (O.T. Wolf).
1SH and NR contributed equally to this publication.
available at www.sciencedirect.com
journal homepage: www.elsevier.com/locate/psyneuen
0306-4530/$ — see front matter # 2009 Elsevier Ltd. All rights reserved.
Author's personal copy
sized the importance of psychological and emotional stimuli
in activating the ‘‘fight-or-flight-response’’. Mason (1968a,b)
referred to Cannons concept of stress and assumed that
psychological variables such as novelty, unpredictability,
anticipation of negative outcome and ego-involvement are
factors that most commonly define a stressful situation.
Nearly 40 years later Dickerson and Kemeny (2004) delivered
a quantitative summary of the empirical evidences for
Masons assumptions by meta-analytically reviewing more
than a hundred laboratory stress studies. They found uncon-
trollability and threat to the social self and the self-esteem
to be especially effective for inducing a significant cortisol
responses and being implemented into several psychosocial
laboratory stressors. The combination of an evaluated public
speech and a cognitive task integrates these factors and
reliably stimulates the hypothamalus—pituitary—adrenal
(HPA) axis in the laboratory (Linden et al., 1998; Biondi
and Picardi, 1999; Dickerson and Kemeny, 2004; Kudielka
and Kirschbaum, 2005)
One prominent laboratory stress procedure is the ‘‘Trier
Social Stress Test’’ (TSST), published by Kirschbaum et al.
(1993). This active performance task consists of a public
speech and a mental arithmetic task (see description below).
The participants’ self-esteem is threatened by a committee
that pretends to evaluate the participants’ performance
without any signs of social support. Thus the participant does
not know whether his/her behaviour is accurate, which leads
to feelings of uncontrollability. This procedure was designed
to be in accordance with Mason’s assumptions (1968a,b) and
is quite effective in activating the HPA and the sympathetic
et al., 2003; Kern et al., 2008).
Studies using the TSST can be divided into two experi-
mental approaches. One strategy is to use the TSST to
investigate the neuroendocrine stress response (e.g. cortisol
response) and to compare the stress responsivity between
certain groups of interest, e.g. women versus men (Kirsch-
baum et al., 1999; Kudielka and Kirschbaum, 2005), young
versus old (Kudielka et al., 2004a), ‘normal’ versus diseased
(Stones et al., 1999; Buske-Kirschbaum et al., 2002; Gaab
et al., 2002; De Vente et al., 2003; Rohleder et al., 2003;
Ahrens et al., 2008) or whites versus blacks (Chong et al.,
2008). In these studies the cortisol stress response is the
dependent variable of interest and the stress response is
assessed by comparing poststress cortisol levels with a pre-
stress baseline measure.
Another approach is to use the TSST to induce stress and
to test the effects of stress and its biological responses on
cognitive or affective outcome measures (e.g. Kuhlmann
et al., 2005; Schoofs et al., 2008) or on physiological
measures (e.g. Nater et al., 2006; Rohleder et al.,
2006a). In these studies the TSST is used for the creation
of the independent variable (stress versus no stress). In the
latter design the stress condition has to be contrasted with
a non-stress control condition. Although the TSST has been
widely used, there is a lack of a standardized control
condition or a ‘‘placebo version’’. When researchers make
use of an appropriate control condition the internal valid-
ity and the statistical conclusion validity increases, due to
the exclusion of confounding variables (Cook and Camp-
bell, 1979; Krauth, 2000). Thus an appropriate control
condition helps to eliminate alternative explanations for
a detected causal relationship between an independent
research this is especially important in studies which
investigate the effects of stress on cognitive or affective
variables since these are vulnerable towards subtle
changes in physical and cognitive demands. In the case
of the TSSTa control condition would be needed in order to
demonstrate that observed effects are indeed caused by
the stress response induced by the TSST and are not just
secondary to the physical or cognitive demands of the task
(e.g. giving a speech or calculating).
is identical to the intended treatment, except its specific
psychological or physiological effective factors. So, the
appropriate control situation for the TSST must be equal
to it except of its effective factors, namely the social
evaluative component and the uncontrollability, according
to the theory of Dickerson and Kemeny (2004). However,
typical control conditions used for this treatment in past
studies had often been quiescent and uneventful circum-
stances, in which the participants of the control group
usually stayed alone in a room reading a magazine or com-
pleting questionnaires (e.g. Kirschbaum et al., 1993; Wolf
et al., 2001; Domes et al., 2002, 2004; Nater et al., 2006,
2007a; Rohleder et al., 2006a). These control conditions
differed from the TSST not only in their stressfulness, but
also in the physical and cognitive load they impose on the
affective and cognitive measures taken during or after the
control condition. Body posture may be one factor that
needs to be similar, since an orthostatic response may
influence SNS parameters (Lake, 1979; Januszewicz et al.,
1982; Goldstein, 1987; Carnethon et al., 2002). Thus, an
appropriate placebo version of the TSSTshould require the
participants to stand in an upright posture. Doing so, the
physiological load of the participants of the control group
would be comparable to that of the TSST group. Further-
more, the placebo version of the TSSTshould include tasks
leading to a cognitive load comparable to the TSST, such as
speaking aloud and/or performing mental arithmetic. In
contrast to that, the participants must under no conditions
perceive these tasks as stressful.
In our laboratories we have already started using a stan-
dardized placebo version of the TSST (e.g. Kuhlmann et al.,
2005; Schoofs et al., 2008). We created a condition which is
similar in physical and mental demand (speech and math
task) to the TSST, but in which the stress inducing negative
social evaluation component of the TSST is lacking. The
participants of the placebo group are usually alone in a room
and complete the tasks by themselves (see description
below). Until now there is no neuroendocrine and psycholo-
gical evaluation of the practicability of this standardized
control condition of the TSST. Thus, the aims of the present
set of two studies are as follows:
1. To evaluate a standardized placebo version of the TSST
using neuroendocrine and psychometric stress measures.
2. To find out whether there is a difference in neuroendo-
crine parameters if an intercom is used to control the
participants’ compliance during the placebo version.
3. To test for carryover effects within the use of crossover
design with the TSST and its placebo version.
1076S. Het et al.
Author's personal copy
We predicted that the placebo TSST would not lead to a
HPA stress response. With respect to the SNS (indirectly
assessed via alpha-amylase) we expected that the placebo
TSST might lead to a modest response due to the physiolo-
gical demands of the task. With respect to the intercom we
speculated that it might activate thoughts of being con-
trolled and evaluated by the experimenter. Thus, we pre-
dicted that an intercom might induce some neuroendocrine
stress responses (modest HPA and/or sAA increases). With
regard to carryover effects within the crossover design, we
predicted that there would be no effect of treatment order
(TSST first, placebo on the second trial or vice versa) with
regard to cortisol- and sAA-levels. This would be in line with
results from former studies of our laboratory (e.g. Kuhlmann
et al., 2005). Finally, in accordance to the findings of Gaab
et al. (2005) we predicted that participants treated with the
TSST will show more negative anticipatory cognitive stress
appraisal than participants treated with the placebo version
of the TSST.
2. Experiment 1: neuroendocrine evaluation
of the ‘placebo TSST’ using a between
The aim of experiment 1 was to evaluate the standardized
placebo version of the TSST in general using a randomized,
between subject comparison study design. For this purpose
we compared participants treated with TSST with partici-
pants treated with its placebo version with regard to
salivary cortisol, sAA-levels and anticipatory cognitive
stress appraisal. We predicted that placebo-treated parti-
cipants show lower cortisol and sAA-levels on average and
lower ratings of stress appraisal than participants treated
with the TSST.
Participants were recruited among students of the University
of Bielefeld within the context of a series of studies on the
effect of stress on memory. Results of the male part of the
sample are already published elsewhere (Schoofs et al.,
2008). The total sample consisted of 84 participants, with
40 men and 44 women participating. The participants under-
went a brief medical and psychological examination prior to
testing. The following exclusion criteria were selected: (a)
BMI below 19 or above 26 kg/m2, (b) younger than 18 or older
than 35 years of age (Kudielka et al., 2004a), (c) acute or
chronic somatic and/or psychiatric disease, (d) intake of
medication, (e) current psychotherapy treatment, (f) acute
or chronic self-reported stress, and (g) female participants
with an irregular menstrual cycle or intake of oral contra-
ceptives (Kirschbaum et al., 1996a, 1999). All female parti-
cipants were tested in the luteal phase of their menstrual
cycle (self-report). Participants refrained from smoking,
drinks (e.g. coffee or tea) at least 1 h prior to testing.
Participants received detailed information about the study,
provided written consent and were paid for their participa-
tion. The study protocol was approved by national ethics
committee of the German Psychological Society (DGPs).
2.1.2. Experimental protocol
Single experimental sessions took about 90 min and were
conducted between 1000 h and 1230 h. After arrival at the
laboratory, participants rested for 30 min. The first saliva
sample (baseline) was obtained 1 min prior to treatment
(?1 min). Each participant was assigned randomly to the
TSST or its placebo version. Both treatments lasted 15 min.
At the beginning of the TSST or its placebo version the
participants completed a questionnaire on stress appraisal
(see below). After completion of the TSST or its placebo
version the next saliva sample (+1 min) was obtained. Saliva
sampling was repeated again after 10 and 25 min. At the end
participants were debriefed by the experimenter.
188.8.131.52. TSST. The TSST was performed similarly to the
description provided by Kirschbaum et al. (1993). The parti-
cipant was told to introduce him-/herself to a selection
committee. After an initial preparation period of 5 min dur-
ing which the participants filled out the questionnaire on
stress appraisal (PASA; see below) participants had to give a
free speech. During the speech (5 min duration) he or she had
to convince the committee that he/she was the perfect
applicant for a vacant position (his or her ‘dream job’).
The committee was dressed in white coats and was intro-
duced as consisting of psychologists who are specially trained
to monitor and analyze verbal and nonverbal behaviour.
Furthermore it was announced that the participant’s perfor-
mance was recorded on the video-cassette-recorder to later
analyze the interview and the nonverbal behaviour. If the
participant finished his/her speech in less than 5 min, stan-
dardized questions were used. Then the participant was
asked to serially subtract the number 17 from 2043 as fast
and as accurately as possible within 5 min. On every failure
the psychologist interfered and the participant has to start
again at 2043. Both members of the committee acted in a
very cold and reserved manner.
184.108.40.206. Placebo TSST. The standardized control version for
theTSST(the ‘Placebo TSST’)wasdesigned tobeassimilaras
possible to the TSST without being stressful for the partici-
pants. At first the experimenter led the participant into an
empty room where he/she was asked to talk for 5 min aloud
about a movie, a novel, or a recent holiday trip. He/she was
told that there is a preparation time of 5 min. In this period
he/she should think about the topic he/she will present.
Before preparing the speech the participants were asked to
complete the PASA. After 5 min the experimenter entered
the room and asked the participant to stand up, take position
somewhere in the room and start to talk loudly about the
chosen topic. Five minutes later the experimenter entered
the room again and asked the participant to start adding up
the number 15 starting at 0. This second task also lasted
5 min. After the second task the experimenter led the parti-
cipant back to the waiting room. The experimenter con-
trolled the compliance of the participant by asking the
participant about the number he/she had reached while
counting forwards in steps of 15. The placebo TSST was
performed in the same room as the TSST but all ‘stressing’
elements of the TSST (committee, video camera, and micro-
phone) were removed prior to the start of it. This procedure
was expected to eliminate the main effective factors of the
TSST, namely the social evaluative threat and the uncontroll-
Placebo version of the Trier Social Stress Test1077
Author's personal copy
ability, according to the theory of Dickerson and Kemeny
(2004). It also fits to Mason (1968a,b) assumptions on factors
that define stressful situations. Ego-involvement is mini-
mized by performing a speech on a superficial, not self-
relevant topic. This is also the case for the factor novelty,
because people are rather used to talk on such topics than on
their personality traits. Because of a lack of an audience the
factor unpredictability and anticipation of a negative out-
come is reduced as well.
220.127.116.11. Saliva sampling and biochemical analysis. Saliva
was collected to obtain free cortisol levels (Kirschbaum
and Hellhammer, 1989, 1994) as well as sAA-activity (Rohle-
der and Nater, 2009) as markers of HPA and SNS activity,
respectively. The samples were obtained using Salivette
sampling devices (Sarstedt, Nu ¨mbrecht, Germany) 1 min
before treatment and 1, 10 and 25 min after treatment. Free
cortisol levels were measured using a commercially available
immunoassay with chemiluminescence detection (IBL-Ham-
burg, Germany). sAA-activity was measured by using a quan-
titative enzyme kinetic method, as described elsewhere
(Rohleder and Nater, 2009). Inter- and intra-assay coeffi-
cients of variation were below 15%.
18.104.22.168. Measurement of anticipatory cognitive stress
appraisal. Participants’ anticipatory cognitive stress apprai-
sal at the beginning of the TSST or its placebo version was
measured using the German version of ‘‘Primary Appraisal—
Secondary Appraisal (PASA)’’ questionnaire which was devel-
oped by Gaab et al. (2005). This questionnaire consists of 16
items and was developed in accordance to the model of
Lazarus and Folkman (1984). The questionnaire assesses four
anticipatory cognitive appraisal processes: ‘‘Threat’’, ‘‘Chal-
lenge’’, ‘‘Self Concept of Own Abilities’’ and ‘‘Control Expec-
tancy’’. Each scale of the PASA comprised four items with a 6-
point scale ranging from ‘‘Strongly disagree’’ to ‘‘Strongly
agree’’. All items were directly related to the treatment
situation (TSSTor placebo) before. For the placebo TSST the
items of the scale ‘‘Control Expectancy’’ were altered,
because of inappropriate wording. The word ‘‘interview’’
was replaced by the word ‘‘situation’’ and the description
‘‘experts’ judge’’ was left out. In addition two items were
excluded because of inappropriate content with regard to
the placebo TSST (‘‘It mainly depends on me whether I
manage this situation successfully.’’ and ‘‘My Success in this
situation is a consequence of my effort and personal commit-
ment’’). For each of the four scales sum scores were calcu-
lated and averaged.
2.1.4. Statistical analysis
Demographic and descriptive variables were investigated by
Pearson’s Chi-square-test and Student’s t-test. The results of
the questionnaire were analyzed using Mann—Whitney-U-
test, because of ordinal level of measurement. Pearson’s
correlation was used due to investigate the relationship
between the results of the stressed participants in the
questionnaire and their actual neuroendocrine reaction,
using the method of area under the curve (AUC, Pruessner
et al.,2003). AUC was calculated with respect to ground, as a
measure of the total neuroendocrine reaction, and with
respect to baseline, for the increase of the neuroendocrine
variables, as a measure of their reactivity. Mixed model
analyses of variance (ANOVA) for repeated measures were
effects of time, treatment and sex. The sAA data were tested
for normal distribution with Kolmogorov—Smirnov-test (K—S-
test), because sAA-data are usually positively skewed dis-
tributed andANOVApremisesnormal distribution.In caseofa
significant result the sAA-data were log-transformed and the
ANOVA was performed with the transformed data. Green-
house-Geisser adjusted p values are reported in case of
violated sphericity assumption. Overall level of significance
was defined as p < .05. Differences of interest between the
groups within a particular variable and a specific sampling
time were evaluated post hoc by using Student’s t-test for
independent samples. In this case level of significance was
Bonferroni-adjusted. Statistical analyses were performed by
SPSS 13 for MAC OS X.
2.2.1. Description of the sample
The total sample of experiment 1 consisted of 84 participants
with a mean age of 24.8 yrs (?.43 S.E.M.) and a mean BMI of
22.7 kg/m2(?.27 S.E.M.). The experimental group consisted
of 19 male and 23 female participants. The control group
consisted of 21 women and 21 men. There was no significant
group difference in distribution of sex, age, BMI, or smoking
status (ps > .10). Because of missing data (sample contam-
ination or insufficient saliva volume) one participant of the
control group had to be excluded from the analysis of cortisol
and two participants of the control group from the analysis of
sAA, respectively. See Table 1 in the supplementary material
2.2.2. Salivary cortisol
As expected, the TSSTcaused an activation of the HPA system
as shown in Fig. 1. While placebo-treated participants
showed decreasing cortisol levels. A three-way ANOVA
revealed a significant effect of time (F(2.05/161.99)= 10.99;
p < .01), an interaction between treatment and sex (F(3/
79)= 4.78; p < .01), and between time, treatment and sex
participants exposed to the TSST or the placebo TSST. As
expected participants under the TSSTcondition showed elevated
cortisol concentrations 10 min after the TSST. Participants, who
were treated with the placebo version of the TSST showed a
decline during the course of the experiment.
Salivary cortisol concentrations (mean and S.E.M.) of
1078 S. Het et al.
Author's personal copy
(F(6.15/161.99)= 7.94; p < .01). The cortisol response defined
as the difference between the baseline level and the time of
the expected cortisol peak (+10’) was calculated for both
sexes. Although not significant (t40= 1.04; p > .05), male
participants displayed on average a stronger cortisol increase
(6.3 nmol/l; ?8.93 S.D.) in response to the TSST than female
participants (3.6 nmol/l; ?7.77 S.D.). This difference disap-
peared when participants of the TSST group showing a corti-
sol response of at least 2.5 nmol/l (difference in cortisol
concentration between baseline and time of expected corti-
sol peak) were compared (9.61 nmol/l; ?8.35 S.D. (female)
vs. 9.33 nmol/l; ?5.57 S.D. (male)). There was no difference
between thesexes duringthe placeboTSST (data notshown).
2.2.3. Salivary alpha-amylase
The results for sAA are shown in Fig. 2. K—S-tests indicated
thatsAAdatawereskewed(ps ? 0.009).Thus,sAAdatawere
log-transformed to approach a normal distribution. Three-
way ANOVA revealed a significant effect of time (F(2.4/
191.8)= 28.3; p < .01), treatment (F(1/80)= 4.75; p < .05),
and the interaction between these two factors (F(2.4/
191.8)= 5.1; p < .01). There was no effect of sex on these
results. The groups did not differ at baseline (t81= ?0.88;
p > .05). Placebo-treated participants showed a small, but
significant increase in sAA immediately after the treatment
(t40= ?3.24; p < .01). The stressed participants showed a
nearly twofoldincrease 1 min
(t41= ?5.27; p < .01).
The two groups differed on all four scales of the PASA (data
not shown). Wilcoxon-U-tests revealed significant differ-
ences for threat (p = .01), challenge (p = .007), self-con-
cept of own abilities (p = .007), and control expectancy
(p = .02). Stressed participants felt more threat and chal-
lenge, and rated their own abilities and control expectancy
lower than participants who had been exposed to the pla-
expectancy and the total cortisol levels (AUCground) within
the stressed group. Neither for cortisol nor for sAA did any
additional correlation reach the uncorrected threshold of
p < .05.
2.3. Discussion of experiment 1
The neuroendocrine and psychometric data of experiment 1
suggest that the placebo version of the TSSTis a non-stressful
and thus a well-suited control condition. While the stressed
participants showed the expected increase of cortisol in
response to the TSST with peak levels reached 10 min after
the cessation of the stressor (Kirschbaum et al., 1993, 1999;
Kudielka and Kirschbaum, 2005), the placebo-treated parti-
cipants showed a continuous decline in cortisol. The missing
HPA response in the control condition reflects the fact that
the effective component of the TSST, i.e. uncontrollable,
social evaluative threat (Dickerson and Kemeny, 2004) was
successfully eliminated in the placebo version.
The results for sAA are similar to the results for cortisol.
While the control group showed only a small increase in sAA-
activity, the stressed participants displayed the expected
elevation of sAA-activity as already reported in recent stu-
dies (e.g. Rohleder et al., 2004; Nater et al., 2006). The
slight, although significant increase within the control group
1 min after the treatment probably reflects the physical
activity since participants had to take an upright posture
in the room and had to talk loudly.
In line with the neuroendocrine measures were the PASA
results. Participants of the control group felt less threat
and challenge (primary appraisal), and had a greater self-
concept of positive abilities and control expectancy (sec-
ondary appraisal) than TSST-treated participants. The
negative correlation between control expectancy and
overall cortisol output indicates that reduced control
expectancy is accompanied by a higher overall cortisol
output. This is in accordance with the social self-preserva-
tion theory of Dickerson and Kemeny (2004). The higher
primary and the lower secondary anticipatory cognitive
appraisals in participants immediately prior to stress expo-
sure are in line with previous studies using this question-
naire (Gaab et al., 2005).
In sum, this first experiment indicated that the placebo
TSST appears to be a promising standardized control con-
dition for the TSST. Its lack of uncontrollable socio-eva-
luative threat appears to prevent an HPA response. For
sAA, only a moderate increase occurs, which suggests that
the physical demand and/or the somewhat unusual task of
speaking alone in a room lead to a slight SNS response,
which however is substantially weaker than the sAA
response to the TSST.
3. Experiment 2: neuroendocrine evaluation
of the placebo TSST using a within subjects
The aim of experiment 2 was to find out whether a cross-
over design, which is quite common in psychoneuroendo-
crine research, can be used with the TSST and its placebo
version. In other words, we wanted to test whether initial
exposure to the placebo TSST might influence the response
to the TSST and vice versa. Furthermore we asked if there
is an effect on the reactivity of the HPA and the SNS when
and S.E.M.) of participants exposed to the TSST or the placebo
TSST (not logarithmized data used for display purposes). While
the stressed group showed a strong sAA increase in response to
the TSST, the control group showed only a minor increase in
response to the placebo TSST.
Salivary alpha-amylase (SAA) concentrations (mean
Placebo version of the Trier Social Stress Test1079
Author's personal copy
experimenters use an intercom to instruct the participants
for the placebo version of the TSST and to control their
compliance (to control that they actually spoke aloud and
counted aloud). For this purpose we invited participants
into our laboratory twice. On the first day participants
were treated either with TSST or its placebo version, and
on the second day they received the opposite treatment.
Again salivary cortisol, sAA and the PASA were used to
evaluate the effects. We predicted that there is no effect
of treatment order. Furthermore, we assumed that the
intercom might affect the neuroendocrine stress response,
especially the response of sAA, which appears to be far
more sensitive to subtle emotional challenges (van Stege-
ren et al., 2008).
The participants of this experiment were recruited among
students of the Technical University of Dresden. Similar to
experiment 1, the volunteers underwent a brief medical and
psychological examination prior to testing, using the same
inclusion and exclusion criteria except the menstrual cycle
and intake of oral contraceptives. Female participants were
their cycle phase. Again, all participants refrained from
smoking, physical exercise, meals, alcoholic beverages and
stimulating drinks (e.g. coffee or tea) at least 1 h prior to
testing. Participants received detailed information about the
study, provided written consent and were paid for their
participation. The study protocol was approved by the local
According to its purpose, experiment 2 was based on a
crossover design, with 2 days of experimental investigation
and repeated measures of salivary cortisol, alpha-amylase
and anticipatory cognitive stress appraisal. The assignment
of treatment order (TSST/placebo or placebo/TSST) was
randomized, with similar distribution of both sexes in session
orders. The second trial was conducted at least 14 days after
the first testing.
3.1.2. Experimental protocol
Both experimental sessions took about 90 min and were
conducted between 1400 h and 1800 h. After arrival at the
laboratory, participants were allowed to rest for 20 min. The
TSST was conducted as described in Section 22.214.171.124. The
placeboTSST was similar
126.96.36.199), except, wehereused an intercom which wasturned
on during the whole placebo TSST in order to communicate
with the participants and control their compliance to the
tasks of the placebo TSST. The intercom use was announced
to the participants during the instructions but they were told
not to be recorded or evaluated. Like in experiment 1, the
participants had tocomplete thePASA duringthepreparation
phase of both treatments. Finally, participants were
to experiment1 (Section
188.8.131.52. Saliva sampling and biochemical analysis. Saliva
samples were obtained 1 min before treatment and 1, 15
and 30 min after treatment. Biochemical analyses were con-
ducted using the same assays as described in experiment 1.
184.108.40.206. Measurement of stress appraisal. The anticipatory
cognitive stress appraisal was again assessed by the PASA (see
experiment 1 for description) within the preparation time of
the TSSTand the placebo. For the placebo phase we used the
same modified PASA as in experiment 1.
3.1.4. Statistical analysis
Statistical analyses were performed as similar as described
for experiment 1, except that this time procedures for
repeated measurement designs were applied. Thus paired
t-tests with adjusted a-levels were used to evaluate differ-
ences at particular sampling times. Analyses of variance
(ANOVA) for repeated measures were employed with time
and treatment as within subject factors. Treatment order
and sex were used as between subject factors. Overall level
of significance was defined as p < .05.
3.2.1. Description of the sample
The total sample consisted of 47 participants (28 women and
19 men) with a mean age of 23.2 years (?.42 S.E.M.) and a
mean BMI of 22.5 kg/m2(?.43 S.E.M.). The two groups with
opposite treatment order (TSST first versus placebo first) did
not differ with respect to sex, age, BMI, smoking, intake of
oral contraception and menstrual cycle phase (ps > .10).
Twenty-four participants performed on the TSST first, while
23 participants started with the placebo TSST. Because of
missing data (sample contamination or insufficient saliva
volume), three participants had to be excluded from the
analysis of cortisol and eight participants from the analysis of
sAA, respectively. See supplementary material (Table 1) for
3.2.2. Salivary cortisol
Results for salivary cortisol are presented in Fig. 3.
Fig. 3 (top) displays the results for the entire group, while
middle and bottom display the findings for the two different
treatment orders separately. ANOVA with the factors time,
treatment, treatment order and sex revealed a significant
effect of time (F(1.84/73.74)= 23.97; p < .001), and treatment
(F(1/40)= 10.99; p < .01). Furthermore, there was a signifi-
canttime ? treatment
p < .001). There was no main-effect of sex (F(1/40)= 1.40;
p = .24), but the order of treatment (F(1/40)= 6.85; p = .012)
was significant. Treatment order did not interact with any of
the other factors (time and sex). The average cortisol
response for each group of treatment order was assessed
by computing the differences between baseline and cortisol
levels obtained 15 min after the treatment. While the parti-
cipants with the placebo/TSST treatment order showed on
average lower baseline cortisol levels (Fig. 3 bottom) their
response to the TSST (6.01 nmol/l ? 1.42 S.E.M.) was very
similar to the participants of the TSST/placebo group
(8.04 nmol/l ? 2.05 S.E.M.). Thus, the two treatment orders
did not differ significantly in their cortisol response to the
TSST(t42= .81;p > .10). Thedifferencebetweenthecortisol
levels 1 min prior and 1 min after the placebo treatment was
significant (t44= 3.27; p < .01), reflecting a small but sig-
nificant increase of cortisol level within this treatment con-
1080S. Het et al.
Author's personal copy
3.2.3. Salivary alpha-amylase
The results on the salivary alpha-amylase for each condition
are shown in Fig. 4.
Fig. 4 (top) displays the results for the entire group, while
middle and bottom display the findings for the two different
treatment orders separately. K—S-test revealed a significant
result for three sampling times (ps ? .05). ANOVA with the
factors time, treatment, treatment order and sex revealed
only a significant effect of time (F(2.1/73.54)= 6.44; p < .01).
However, there was neither an interaction with treatment,
treatment order, nor sex. These results indicate that the sAA
response to the TSST and placebo TSST did not differ from
each other. In addition, there was a main effect of sex (F(1/
35)= 4.8; p < .05). Male participants were found to display
higher levels of sAA during both experimental conditions
(data not shown).
Results on the PASA revealed that TSST-treated participants
scored higher on items of primary appraisal (i.e. threat and
exposed to the TSST or its placebo version in a within subject
crossover design. The figure on top shows both treatment orders
combined, the figure in the middle displays participants who
were first exposed to the TSST and the figure on the bottom
displays participants who were first exposed to the placebo TSST.
Irrespective of treatment order participants showed a robust
cortisol increase in response to the TSST, which peak levels
obtained 15 min after stress cessation.
Salivary cortisol concentrations of participants
ticipants exposed to the TSST or its placebo version in a within
orders combined, the figure in the middle displays participants
who were first exposed to the TSSTand the figure on the bottom
displays participants who were first exposed to the placebo TSST.
In each experimental condition participants showed an increase
of sAA in response to the treatment (not logarithmized data used
for display purposes). There was no difference between TSSTand
Salivary alpha-amylase (sAA) concentrations of par-
Placebo version of the Trier Social Stress Test1081
Author's personal copy
challenge), and lower on items of secondary appraisal (i.e.
self-concept and control expectancy) than control partici-
pants (data not shown). Wilcoxon-signed ranks test revealed
significant differences between the experimental conditions
on threat, challenge and self-concept (ps < .0001). How-
expectancy (p = .51). Table 3 of the supplementary material
reports a significant positive correlation between the scale
threat and the reactivity of cortisol (AUCincrease) during the
TSST condition (both treatment orders combined). Neither
for cortisol nor for sAA did any additional correlation reach
the uncorrected threshold of p < .05.
3.3. Discussion of experiment 2
The results of experiment 2 revealed significant differences
in salivary cortisol responses between the TSST and the
placebo condition irrespective of treatment order. Within
the TSSTcondition we observed enhanced cortisol levels with
a maximum at 15 min after the treatment. The significant
differences between the conditions indicate a stronger HPA
response to the TSST compared to the placebo TSST. Again,
the placebo version of the TSSTwas a useful control condition
with respect to the HPA stress response. Within the placebo
condition the participants displayed a small but significant
cortisol increase, this contrasts findings of the first experi-
ment and might, at least in part, reflect the use of the
The intercom might lead to cognitions of being monitored
or evaluated. Due to the spatial separation of the experimen-
ter the participants had no knowledge about her/his beha-
viour. In addition, thePASA resultsfor thisexperiment showed
no difference in control expectancy for the two conditions.
This might be secondary to the use of the intercom, since
participants can not control whether or not the experimenter
is listening to the content of their speech or not. This might
lead to a slight, but significant threat of the drive to preserve
the social self and keep control, due to keep or increase the
2004,2007).Thesecognitionsmighthaveresulted ina slightly
enhanced HPA activity in some individuals.
The differences on the PASA scales threat, challenge and
self-concept of own abilities are in accordance with our
expectations and indicate a higher stress appraisal within
the TSSTcondition in contrast to the placebo TSSTcondition.
The positive correlation between the scale threat and the
cortisol increase due to the TSSTcondition is in line with the
findings of Gaab et al. (2005). It indicates that a higher level
of threat appraisal with regard to a particular situation is
accompanied with a stronger cortisol increase.
We found no evidence for a carryover effect. Independent
of which kind of treatment participants received first, their
cortisol response to the TSSTwas very similar. This is in line
with a former study of our laboratory (Kuhlmann et al., 2005)
in which we used the TSST and its placebo version also in a
crossover manner. However, we found a significant main
effect of treatment order reflecting differences in baseline
levels between the two treatment orders.
For sAA we found no differences between the two treat-
ment conditions. Thus, our results indicate that TSSTas well
as the placebo TSST with the intercom activated the SNS
system to a similar degree. The presence of the intercom
most likely induced an increased emotional arousal. This
effect was obviously strong enough in order to abolish any
differences between the two treatment conditions for sAA.
Future studies should use additional markers of SNS activity
(heart rate, heart rate variability, etc.) in order to monitor
SNS activity during the placebo TSST.
A second finding of interest with respect to sAA is a
significant main effect of sex. Male participants on average
displayed higher sAA-activity throughout the experiment
than women. A similar sex-difference in sAA basal activity
was recently reported in a laboratory study by van Stegeren
et al. (2008). However, no sex differences in sAA were
observed in a circadian day profile study (Nater et al.,
2007b). In the general discussion the issue of sex differences
in sAA-activity will be discussed in some more depth.
4. General discussion
In the present study we evaluated a placebo version of the
TSST. While the TSST is frequently used in psychoneuroendo-
crine research, a standardized control condition is often
missing. Considering the social self-preservation theory of
Dickerson and Kemeny (2004) we developed a placebo ver-
sion of the TSST by eliminating its effective components: the
uncontrollability and socio-evaluative threat to the self-
esteem. In two studies we evaluated this placebo treatment
with regard to neuroendocrine measures (salivary cortisol,
sAA) and subjective ratings. The first study employed a
between subject design. The second study was done to
answer the question whether crossover designs can be used
with the TSSTand its placebo version. In addition we tested
the impact of the use of an intercom for monitoring the
compliance of participants during the placebo version.
4.1. Effects of the placebo TSST on the HPA axis
In both experiments strong differences between the TSSTand
its placebo version were observed for salivary cortisol levels.
it an appropriate control condition in studies interested in the
HPA response to the TSST. The obtained cortisol levels of
control participants were similar to those of former studies
using resting and quiet control conditions (Kirschbaum et al.,
1993, 1999; Rohleder et al., 2001, 2003; Kudielka et al.,
2004b). Unexpectedly we found a slight enhancement of
HPA activity within the control condition of experiment 2 as
This issue will be discussed later. In general our results on the
accord to the results of former studies of our laboratory
(Kuhlmann et al., 2005; Schoofs et al., 2008), which already
used this placebo treatment. Our findings support the conclu-
sionsofDickerson and Kemeny(2004) who observed enhanced
HPA activity only in situations with the experience of uncon-
trollability and social evaluation.
4.2. Effects of the placebo TSST on sAA
In experiment 1 the TSST led to the expected strong rise in
sAA-activity, similarly to previous studies (Rohleder et al.,
1082 S. Het et al.
Author's personal copy
2004; Nater et al., 2006). This rise was significantly stronger
in the TSST condition than in its placebo version. For the
placebo group however we also obtained a more modest, but
significant enhancement of sAA-activity immediately after
upright posture, talking) as already reported elsewhere (e.g.
Goldstein, 1987). Thus while the TSST leads to a stronger sAA
response than the placebo TSST, a sAA increase does still
occur in the placebo TSST. This is in contrast to studies
comparing the TSST with resting control conditions (e.g.
Rohleder et al., 2004; Nater et al., 2006). Thus, the benefit
of controlling for the physical and cognitive demands of the
TSSTwith the placebo TSSTcomes at a certain cost, namely
the modest activation of the SNS system. However, using the
placebo TSSTwithout an intercom enables the researcher to
characterize the specific amount of sAA activation that is
caused by the physical and cognitive demands of the TSST
situation. In other words, the researcher has the opportunity
to characterize the amount of sAA activation that is speci-
fically caused by the psychological factors of the situation.
In experiment 2 we found similar sAA-levels in the TSST
condition and in the placebo condition, which suggests
comparable SNS activities in both treatment conditions.
Experiment 2 differed from experiment 1 in the use of an
intercom, which was introduced to control the participants’
compliance within the placebo treatment and to instruct
them. Thus, we conclude an arousing effect of intercom
usage resulting in equally enhanced sAA-levels in both
conditions. In addition, the intercom led also to a slight,
but significant enhancement of HPA activity within the
control condition of experiment 2 and nearly similar ratings
of control expectancy with regard to the PASA. How this
arousing effect might occur will be discussed in the next
section. The finding that the intercom effect is more
obvious for sAA fits to the general notion that the SNS is
more responsive to emotional arousal and effort, whereas
the HPA only gets activated in uncontrollable situations of
more serious and continuous threats (e.g. distress; see
Frankenhaeuser et al., 1978; Lundberg and Frankenhaeu-
ser, 1980; Wortsman, 2002; Schommer et al., 2003; van
Stegeren et al., 2008).
4.3. Effects of intercom use
The observed intercom effect could be explained by the
social facilitation theory (Zajonc, 1965) as well as the theory
lund, 1979). In terms of the social facilitation theory, the
intercom could lead to arousal, similar to a performance in
front of an audience. In terms of the theory of objective self-
awareness, one might think that our participants were con-
fronted by the intercom with a discrepancy between their
actual performance and their self-expectations. This could
lead to cognitive dissonance which causes additional arousal
within the SNS and HPA activity (e.g. Elkin and Leippe, 1986;
Losch and Cacioppo, 1990; Etgen and Rosen, 1993; Harmon-
Jones and Harmon-Jones, 2007). Alternatively the intercom
effect canbeexplained bythesocialself-preservation theory
of Dickerson and Kemeny (2004) and by Grawe’s consistency
intercom can be seen as an uncontrollable, socio-evaluative
threat to one’s social status. This would lead, in terms of
Grawe (2004, 2007), to a frustration of two basic goals,
namely the need for control and orientation in life and for
self-enhancement and self-esteem. Frustration of at least
one of these goals leads to stress and psychological disorders
(Grawe, 2004, 2007; Grosse Holforth et al., 2006).
Thus, while it is certainly desirable to control the com-
pliance of study participants during the placebo TSST the
usage of an overt intercom abolishes between condition
differences for sAA. It awaits to be shown whether similar
results would be obtained with other SNS markers. Future
studies might want to explore alternative compliance mea-
sures. One possibility is touse a hidden camera (with debrief-
ing afterwards) as was done in a previous stress study (Kern
et al., 2008).
4.4. Sex differences in alpha-amylase
Male participants displayed higher sAA-levels than women at
all sampling times in experiment 2. This finding is in line with
a current study of van Stegeren et al. (2008). Interestingly,
this difference was not seen in experiment 1. These conflict-
ing results are typical for the issue of sex differences in sAA-
levels (e.g. Kivlighan and Granger, 2006; van Stegeren et al.,
2006; Yamaguchi et al., 2006; Nater et al., 2007b). Interest-
ingly, significant sex-differences are reported in studies with
female samples without any restriction on menstrual cycle or
intake of oral contraceptives. In contrast to this, experiment
1 was performed with female participants all being in the
luteal phase of the menstrual cycle. In experiment 2, we
investigated female participants with and without the intake
of oral contraceptives and during different phases of men-
strual cycle. The impact of the menstrual cycle is relatively
well documented for the HPA axis (Kirschbaum et al., 1995a,
1999; Rohleder et al., 2001; Kudielka and Kirschbaum, 2005),
but an in-depth view of similar effects on sAA is still missing.
diseases or the exact sampling procedure need to be eval-
uated (Bosch et al., 1996; Nagler et al., 2000; Nater et al.,
2007b; Rantonen and Meurman, 2000; Kivlighan and Granger,
2006; Rohleder et al., 2006b; Harmon et al., 2008).
4.5. Effects of the placebo TSST on anticipatory
cognitive stress appraisal
With respect to anticipatory cognitive stress appraisal we
found that stressed participants felt more threat and chal-
lenge and estimated their own abilities to get through the
TSST as being lower, indicating a negative anticipatory cog-
nitive stress appraisal. Interestingly, presence of the inter-
com leads to similar ratings of control expectancy. These
findings are in line with previous studies (Gaab et al., 2005;
Wirtz et al., 2006, 2007). Thus, similar to the HPA and SNS
measures the subjective ratings support the general conclu-
sion that the placebo version of the TSST is appropriate as a
tool for psychoneuroendocrine research. Having said this it
must be acknowledged that the PASA was developed for the
assessment of appraisal processes occurring during the pre-
paration period of the TSST. The questionnaire was therefore
not ideally suited for the placebo version and had to be
modified by us, especially the content and meaning of the
scale control expectancy was problematic.
Placebo version of the Trier Social Stress Test1083
Author's personal copy
Even though not the primary goal of our current study we
also searched for associations between the PASA and the
cortisol as well as sAA stress response. For cortisol we found
some associations with the PASA (with the scales control
expectancy and threat), but the strength of these associa-
Wirtz et al., 2006, 2007). For the sAA response no association
were detected suggesting, that the constructs assessed with
the PASA are closer linked to HPA than to SNS activity.
4.6. Carryover effects
The second experiment of this study was performed to
evaluate the use of the TSST and the placebo TSST in a
crossover design, which often is more powerful than a
two-group comparison study. We found no carryover effect
for salivary cortisol, sAA, or anticipatory cognitive stress
appraisal. Treatment order also did not interact with any
of the measures of interest. Thus, a previous experience with
the placebo TSST has no substantial impact on the HPA
response to the TSST, which is in line with a previous study
from our group (Kuhlmann et al., 2005). This is in contrast to
the strong habituation of cortisol stress responses observed if
the TSST is presented a second time to the same participants
in identical settings (Kirschbaum et al., 1995b; Pruessner
et al., 1997; Schommer et al., 2003).
4.7. Limitations and outlook
The two studies presented in this report have several limita-
tions which have to be acknowledged. The studies were
conducted at different times of the day (experiment 1 in
themorning and experiment 2 in theafternoon). Eventhough
some of the differences observed between these experi-
ments are secondary to this factor. In both experiments only
salivary neuroendocrine stress markers were obtained. Espe-
cially in light of the findings for sAA (small increase in
experiment 1 and similar increase as in the TSST in experi-
ment 2) additional psychophysiological measures of SNS
activity should be obtained in future studies. In addition,
the present study only employed a single questionnaire in
both studies. Additional mood measures obtained in parallel
in both studies would have been desirable.
Another limitation is the fact that in both studies only one
baseline (pre treatment) salivary sample was obtained.
Future studies should consider taking more baseline samples
in order to exclude with more power a priori group differ-
Furthermore, it would have been interesting to contrast
the placebo TSST with the naturalistic diurnal course of
cortisol and sAA, obtained by a resting control condition.
This third treatment leg would have allowed us to test
whether or not the cortisol decrease observed during the
P-TSST in study 1 is similar or smaller than the decrease
occurring during a rest condition.
Finally, we want to emphasize that the introduced pla-
cebo version of the TSST is one paradigm among several
possible options. Although not explicitly intended, Dickerson
et al. (2008) showed another possibility how a placebo
treatment for the TSSTcould look like. In this study partici-
pants delivered a speech on why they would be a good job
applicant. In one of the experimental conditions the parti-
cipants had to deliver their speech in the presence of a
research assistant who worked on a computer in the parti-
cipants’ line of vision but who did not look at or acknowl-
edged the participant. It was observed that the mere
presence of another person did not lead to HPA activation
in this public speaking paradigm, which is in line to our
findings reported here. However no SNS markers were
obtained in this study.
Future methodological studies might want to develop
alternative standardized control conditions along some of
the lines suggested in this discussion. In the meantime we
propose that the introduced placebo version of the TSST is a
useful standardized control condition for researchers inter-
ested in controlling some of the non-specific effects of the
TSST (physical and cognitive demand) on their outcome
measure of interest.
Role of the funding sources
The funding source (German Research Foundation; DFG)
had no further role in the design of the study, and in the
collection, analysis and interpretation of the data. In
addition, it had no role in the decision to submit the paper
Conflict of interest
The authors declare that they have no conflict of interest.
This study was supported by grants from the German
Research Foundation (DFG; Ki 537/14-1; Ki 537/21; WO
733/6-2, WO 733/7-1).
Appendix A. Supplementary data
Supplementary data associated with this article can be
found, in the online version, at doi:10.1016/j.psyneuen.
Ahrens, T., Deuschle, M., Krumm, B., van der Pompe, G., den Boer,
J.A., Lederbogen, F., 2008. Pituitary-adrenal and sympathetic
nervous system responses to stress in women remitted from
recurrent major depression. Psychosom Med 70, 461—467.
Biondi, M., Picardi, A., 1999. Psychological stress and neuroendo-
crine function in humans: the last two decades of research.
Psychother Psychosom 68, 114—150.
J., Nieuw Amerongen, A.V., 1996. Psychological stress as a
determinant of protein levels and salivary-induced aggregation
of Streptococcus gordonii in human whole saliva. Psychosom Med
Buske-Kirschbaum, A., Geiben, A., Hollig, H., Morschhauser, E.,
Hellhammer, D., 2002. Altered responsiveness of the hypothala-
mus—pituitary—adrenal axis and the sympathetic adrenomedul-
lary system to stress in patients with atopic dermatitis. J Clin
Endocrinol Metab 87, 4245—4251.
1084 S. Het et al.
Author's personal copy
Cannon, W.B., 1935. Stresses and strains of homeostasis. Am J Med
Sci 189, 1—14.
Carnethon, M.R., Liao, D., Evans, G.W., Cascio, W.E., Chambless,
L.E., Heiss, G., 2002. Correlates of the shift in heart rate varia-
bility with an active postural change in a healthy population
sample: the Atherosclerosis Risk In Communities study. Am Heart
J 143, 808—813.
Chong, R.Y., Uhart, M., McCaul, M.E., Johnson, E., Wand, G.S., 2008.
Whites have a more robust hypothalamic-pituitary—adrenal axis
response to a psychological stressor than blacks. Psychoneuroen-
docrinology 33, 246—254.
Cook,T.D., Campbell,D.T., 1979. Quasi-experimentation:designand
analysis issues for feld settings. Rand McNally College Publishing
De Vente, W., Olff, M., Van Amsterdam, J.G., Kamphuis, J.H.,
Emmelkamp, P.M., 2003. Physiological differences between burn-
out patientsand healthy controls:bloodpressure, heart rate,and
cortisol responses. Occup Environ Med 60 (Suppl 1), i54—61.
Dickerson, S.S., Kemeny, M.E., 2004. Acute stressors and cortisol
responses: a theoretical integration and synthesis of laboratory
research. Psychol Bull 130, 355—391.
Dickerson, S.S., Mycek, P.J., Zaldivar, F., 2008. Negative social
evaluation, but not mere social presence, elicts cortisol
responses to a laboratory stressor task. Health Psychol 27,
Domes, G., Heinrichs, M., Reichwald, U., Hautzinger, M., 2002.
Hypothalamic-pituitary-adrenal axis reactivity to psychological
stress and memory in middle-aged women: high responders exhi-
bit enhanced declarative memory performance. Psychoneuroen-
docrinology 27, 843—853.
Domes, G., Heinrichs, M., Rimmele, U., Reichwald, U., Hautzinger,
M., 2004. Acute stress impairs recognition for positive words– —
association with stress-induced cortisol secretion. Stress 7, 173—
Academic Press, New York.
Elkin, R.A., Leippe, M.R., 1986. Physiological Arousal, Dissonance,
and Attitude Change. J Pers Soc Psychol 51, 55—65.
Etgen, M.P., Rosen, E.F., 1993. Cognitive dissonance: physiological
arousal in the performance expectancy paradigm. Bull Psychon
Soc 31, 229—231.
Frankenhaeuser, M., von Wright, M.R., Collins, A., von Wright, J.,
Sedvall, G., Swahn, C.G., 1978. Sex differences in psychoneur-
oendocrine reactions to examination stress. Psychosom Med 40,
Gaab, J., Huster, D., Peisen, R., Engert, V., Heitz, V., Schad, T.,
Schurmeyer, T.H., Ehlert, U., 2002. Hypothalamic-pituitary—
adrenal axis reactivity in chronic fatigue syndrome and health
under psychological, physiological, and pharmacological stimula-
tion. Psychosom Med 64, 951—962.
Gaab, J., Rohleder, N., Nater, U.M., Ehlert, U., 2005. Psychological
determinants of the cortisol stress response: the role of antici-
patory cognitive appraisal. Psychoneuroendocrinology 30, 599—
Goldstein, D.S., 1987. Stress-induced activation of the sympathetic
nervous system. Baillieres Clin Endocrinol Metab 1, 253—278.
Grawe, K., 2004. Psychological therapy. Hogrefe, Seattle and Tor-
Grawe, K., 2007. Neuropsychotherapy. Lawrence Erlbaum Associ-
ates, Mahwah, NJ.
Grosse Holforth, M., Bents, H., Mauler, B., Grawe, K., 2006. Inter-
personal distress as a mediator between avoidance goals and
satisfaction in psychotherapy inpatients. Clin Psychol Psychother
Harmon,A.G.,Towe-Goodman, N.R.,Fortunato,C.K., Granger, D.A.,
2008. Differences in saliva collection location and disparities in
baseline and diurnal rhythms of alpha-amylase: a preliminary
note of caution. Horm Behav 54, 592—596.
Harmon-Jones, E., Harmon-Jones, C., 2007. Cognitive dissonance
Januszewicz, W., Sznajderman, M., Wocial, B., Chodakowska, J.,
Feltynowski, T., Zukowska-Grojec, Z., 1982. Sympathetic reac-
tivity to upright posture in borderline and established hyperten-
sion. Cor Vasa 24, 429—440.
Kern, S., Oakes, T.R., Stone, C.K., McAuliff, E.M., Kirschbaum, C.,
cortex are associated with HPA axis response to a psychosocial
stressor. Psychoneuroendocrinology 33, 517—529.
Kirschbaum, C., Hellhammer, D.H., 1989. Salivary cortisol in psy-
chobiological research: an overview. Neuropsychobiology 22,
Kirschbaum, C., Hellhammer, D.H., 1994. Salivary cortisol in psy-
choneuroendocrine research: recent developments and applica-
tions. Psychoneuroendocrinology 19, 313—333.
specific effects of social support on cortisol and subjective
Kirschbaum, C., Kudielka, B.M., Gaab, J., Schommer, N.C., Hellham-
mer, D.H., 1999. Impact of gender, menstrual cycle phase, and
oral contraceptives on the activity of the hypothalamus—pitui-
tary—adrenal axis. Psychosom Med 61, 154—162.
Kirschbaum, C., Pirke, K.M., Hellhammer, D.H., 1993. The ‘Trier
responses in a laboratory setting. Neuropsychobiology 28, 76—81.
Kirschbaum, C., Platte, P., Pirke, K.M., Hellhammer, D., 1996a.
Adrenocortical activation following stressful exercise: further
evidence for attenuated free cortisol responses in women using
oral contraceptives. Stress Med 12, 137—143.
Kirschbaum, C., Prussner, J.C., Stone, A.A., Federenko, I., Gaab, J.,
Lintz, D., Schommer, N., Hellhammer, D.H., 1995b. Persistent
high cortisol responses to repeated psychological stress in a
subpopulation of healthy men. Psychosom Med 57, 468—474.
Kivlighan,K.T., Granger, D.A.,2006. Salivaryalpha-amylase response
to competition: relation to gender, previous experience, and
attitudes. Psychoneuroendocrinology 31, 703—714.
Krauth, J., 2000. Experimental design. a handbook and dictonary for
medical and behavioral research. Elsevier, Amsterdam, NL.
Kudielka, B.M., Buske-Kirschbaum, A., Hellhammer, D.H., Kirsch-
baum, C., 2004a. HPA axis responses to laboratory psychosocial
stress in healthy elderly adults, younger adults, and children:
impact of age and gender. Psychoneuroendocrinology 29, 83—98.
Kudielka, B.M., Kirschbaum, C., 2005. Sex differences in HPA axis
responses to stress: a review. Biol Psychol 69, 113—132.
Kudielka, B.M., Schommer, N.C., Hellhammer, D.H., Kirschbaum, C.,
psychosocial stress (TSST) in humans at different times of day.
Psychoneuroendocrinology 29, 983—992.
Kuhlmann, S., Piel, M., Wolf, O.T., 2005. Impaired memory retrieval
after psychosocial stress in healthy young men. J Neurosci 25,
Lake, C.R., 1979. Relationship of sympathetic nervous system tone
and blood pressure. Nephron 23, 84—90.
Lazarus, R.S., Folkman, S., 1984. Stress, appraisal, and coping.
Springer Publishing Company, New York.
Linden, W., Rutledge, T., Con, A., 1998. A case for the usefulness of
laboratory social stressors. Ann Behav Med 20, 310—316.
sympathetic tonus, but attitudes are changed to reduce negative
affect rather than arousal. Journal of Experimental Social Psy-
chology 29, 289—304.
Lundberg, U., Frankenhaeuser, M., 1980. Pituitary—adrenal and
sympathetic-adrenal correlates of distress and effort. J Psycho-
som Res 24, 125—130.
Mason, J.W., 1968a. A review of psychoendocrine research on the
pituitary-adrenal cortical system. Psychosom Med 30 Suppl.,
Placebo version of the Trier Social Stress Test1085
Author's personal copy
Mason, J.W., 1968b. A review of psychoendocrine research on the
Nagler,R., Lischinsky,S., Diamond,E.,Drigues,N., Klein,I.,Reznick,
A.Z., 2000. Effect of cigarette smoke on salivary proteins and
enzyme activities. Arch Biochem Biophys 379, 229—236.
Nater, U.M., La Marca, R., Florin, L., Moses, A., Langhans, W., Koller,
M.M., Ehlert, U., 2006. Stress-induced changes in human salivary
alpha-amylase activity– —associations with adrenergic activity.
Psychoneuroendocrinology 31, 49—58.
Nater, U.M., Moor, C., Okere, U., Stallkamp, R., Martin, M., Ehlert,
is better in high than low cortisol responders to psychosocial
stress. Psychoneuroendocrinology 32, 758—763.
Nater, U.M., Rohleder, N., Schlotz, W., Ehlert, U., Kirschbaum, C.,
2007b. Determinants of the diurnal course of salivary alpha-
amylase. Psychoneuroendocrinology 32, 392—401.
Pruessner, J.C., Gaab, J., Hellhammer, D.H., Lintz, D., Schommer,
N., Kirschbaum, C., 1997. Increasing correlations between per-
sonality traits and cortisol stress responses obtained by data
aggregation. Psychoneuroendocrinology 22, 615—625.
Pruessner, J.C., Kirschbaum, C., Meinlschmid, G., Hellhammer, D.H.,
2003. Two formulas for computation of the area under the curve
represent measures of total hormone concentration versus time-
dependent change. Psychoneuroendocrinology 28, 916—931.
Rantonen, P.J., Meurman, J.H., 2000. Correlations between total
protein, lysozyme, immunoglobulins, amylase, and albumin in
stimulated whole saliva during daytime. Acta Odontol Scand 58,
Rohleder, N., Nater, U., 2009. Determinants of salivary alpha-amy-
lase in humans and methodological considerations. Psychoneur-
oendocrinology [e-pub ahead of print].
Rohleder, N., Nater, U.M., Wolf, J.M., Ehlert, U., Kirschbaum, C.,
2004. Psychosocial stress-induced activation of salivary alpha-
amylase: an indicator of sympathetic activity? Ann N Y Acad Sci
Rohleder, N., Schommer, N.C., Hellhammer, D.H., Engel, R., Kirsch-
baum, C., 2001. Sex differences in glucocorticoid sensitivity of
proinflammatory cytokine production after psychosocial stress.
Psychosom Med 63, 966—972.
Rohleder, N., Wolf, J.M., Herpfer, I., Fiebich, B.L., Kirschbaum, C.,
Lieb, K., 2006a. No response of plasma substance P, but delayed
increase of interleukin-1 receptor antagonist to acute psychoso-
cial stress. Life Sci 78, 3082—3089.
Rohleder, N., Wolf, J.M., Kirschbaum, C., 2003. Glucocorticoid sen-
sitivity in humans—interindividual differences and acute stress
effects. Stress 6, 207—222.
Rohleder, N.,Wolf,J.M., Maldonado, E.F.,Kirschbaum,C.,2006b.The
psychosocial stress-induced increase in salivary alpha-amylase is
independent of saliva flow rate. Psychophysiology 43, 645—652.
Schommer, N.C., Hellhammer, D.H., Kirschbaum, C., 2003. Dissocia-
tion between reactivity of the hypothalamus—pituitary—adrenal
axis and the sympathetic-adrenal-medullary system to repeated
psychosocial stress. Psychosom Med 65, 450—460.
Schoofs, D., Preuss, D., Wolf, O.T., 2008. Psychosocial stress induces
working memory impairments in an n-back paradigm. Psycho-
neuroendocrinology 33, 643—653.
Shapiro, A.K., Morris, L.A., 1978. The placebo effect in medical and
psychological therapies. In: Garfield, S.L., Bergin, A.E. (Eds.),
Stones, A., Groome, D., Perry, D., Hucklebridge, F., Evans, P., 1999.
The effect of stress on salivary cortisol in panic disorder patients.
J Affect Disord 52, 197—201.
alpha amylase as marker for adrenergic activity during stress:
effect of betablockade. Psychoneuroendocrinology 31, 137—141.
van Stegeren, A.H., Wolf, O.T., Kindt, M., 2008. Salivary alpha
amylase and cortisol responses to different stress tasks: impact
of sex. Int J Psychophysiol 69, 33—40.
Wicklund, R.A., 1979. The influence of self-awareness on human
behaviour. Am Sci 67, 187—193.
Wirtz, P.H., Ehlert, U., Emini, L., Rudisuli, K., Groessbauer, S., Gaab,
J., Elsenbruch, S., von Kanel, R., 2006. Anticipatory cognitive
stress appraisal and the acute procoagulant stress response in
men. Psychosom Med 68, 851—858.
Wirtz, P.H., von Kanel, R., Emini, L., Suter, T., Fontana, A., Ehlert,
U., 2007. Variations in anticipatory cognitive stress appraisal and
differential proinflammatory cytokine expression in response to
acute stress. Brain Behav Immun 21, 851—859.
Wolf, O.T., Schommer, N.C., Hellhammer, D.H., McEwen, B.S.,
Kirschbaum, C., 2001. The relationship between stress induced
cortisol levels and memory differs between men and women.
Psychoneuroendocrinology 26, 711—720.
Wortsman, J., 2002. Role of epinephrine in acute stress. Endocrinol
Metab Clin North Am 31, 79—106.
Yamaguchi, M., Deguchi, M., Miyazaki, Y., 2006. The effects of
exercise in forest and urban environments on sympathetic ner-
vous activity of normal young adults. J Int Med Res 34, 152—159.
Zajonc, R.B., 1965. Social facilitation. Science 149, 269—274.
1086S. Het et al.