Embed Size (px)
Citation preview
Cardiovascular and cortisol reactivity and habituationto a virtual reality version of the Trier Social StressTest: A pilot study
Peter Jonsson a,*, Mattias Wallergard b, Kai Osterberg a, Ase Marie Hansen c,Gerd Johansson b, Bjorn Karlson a
aDepartment of Laboratory Medicine, Division of Occupational Health and Environmental Medicine, Behavioural Medicine Section,Lund University, Barngatan 2, Gamla Barnsjukhuset, SE-22185 Lund, Swedenb Ergonomics and Aerosol Technology Design Sciences, Lund University, P.O. Box 118, SE-221 00 Lund, SwedencNational Research Centre for the Working Environment, Lersø Parkalle 105, DK-2100 Copenhagen, Denmark
Received 4 December 2009; received in revised form 8 April 2010; accepted 8 April 2010
Psychoneuroendocrinology (2010) 35, 1397—1403
KEYWORDSTSST;HPA axis;SAM system;Virtual reality;Psychosocial stress;Habituation
Summary The Trier Social Stress Test (TSST) is a widely used protocol to induce stress inlaboratory settings. Briefly, in the TSST, the test participant is asked to hold a speech and to do anarithmetic task in front of an audience. In the present pilot study, we examined endocrine andautonomic reactivity and habituation to repeated stress provocations using a virtual reality (VR)version of TSST. The VR system was a CAVETM system with three rear projected walls (4 m� 3 m),and one floor projection. The system also included a head tracking system and passive stereos-copy. The virtual audience consisted of one woman, and two men. Ten healthy men, mean age28.3 years (24—38 years), were confronted with the test twice (1 week between sessions), duringwhich salivary cortisol, heart rate (HR), high frequency heart rate variability (HF-HRV, parasym-pathetic activity), and T-wave amplitude (TWA, suggested to be related to sympathetic influenceon myocardial performance) were assessed.
Cortisol secretion showed a marked increase (88% vs. baseline) during the first stressprovocation, but habituated in the second session. The magnitude of HR and TWA reactivityduring stress provocation was approximately the same at both sessions, implying a stable increasein sympathetic activity. Heart rate showed amaximum increase of 40% at the first session, and 32%at the second. TWA showed amaximum decrease of 42% at the first session, and 39% at the second.The results resemble those obtained in prior studies using the real-life TSST. If these results can bereplicated with larger samples, VR technology may be used as a simple and standardized tool forsocial stress induction in experimental settings.# 2010 Elsevier Ltd. All rights reserved.
* Corresponding author. Tel.: +46 46 173994; fax: +46 46 177885.E-mail address: [email protected] (P. Jonsson).
ava i lab le at www.sc ienced i rect .com
journa l homepage: www.e l sev ie r.com/locate/psyneuen
0306-4530/$ — see front matter # 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.psyneuen.2010.04.003
1398 P. Jonsson et al.
1. Introduction
The Trier Social Stress Test (TSST) is a widely used protocol toinduce psychosocial stress in laboratory setting (Kirschbaumet al., 1993). Briefly, the participant is asked to deliver aspeech and to perform an arithmetic task in front of anevaluating audience. The audience, consisting of threetrained actors, does not respond emotionally during the test,but maintains a neutral facial expression, which makes thesituation very stressful for the participant. These two tasks,as well as anticipation during preparation of the speech,consistently evoke subjective stress and activation of thehypothalamus—pituitary—adrenal (HPA) axis and the sym-patho—adrenal—medullary (SAM) system (Schommer et al.,2003; Kudielka et al., 2004, 2006; Zgraggen et al., 2005;Nater et al., 2006; Kelly et al., 2008).
The standardized protocol of TSST increases the validity ofcomparisons of results between studies and across labora-tories. Still, the laboratory environments, settings and audi-ence vary. Although the audience consists of actors, it may bedifficult to hold the acting constant across all sessions. Byusing virtual reality (VR) technology, with virtual environ-ments and persons, it should be possible to hold these vari-ables even more constant. Furthermore, replacing actorswith virtual persons reduces the practical problems and costsassociated with hiring professional actors. The present studywas conducted to evaluate a recently developed VR versionof TSST.
Prior studies have shown that it is possible to evokesubjective social stress or anxiety in virtual environmentsusing virtual audiences during public speaking tasks (Pertaubet al., 2002; Slater et al., 2006). For example, during a publicspeaking task, people respond with greater anxiety to ahostile virtual audience than to a neutral or friendly virtualaudience (Pertaub et al., 2002). It has also been shown thatstress provocation in virtual environments results in stress-related endocrine and autonomic reactivity, although with avariety of success. In a study by Kelly et al. (2007), endocrinereactivity was examined in response to TSST with a realaudience and to TSSTwith a virtual audience. Cortisol levelsincreased during stress provocation in both conditions, butwere much more accentuated in response to the real audi-ence (that is, 30% increase to the virtual audience, and 90% tothe real audience). More recently, blood pressure, heartrate, adrenalin, and noradrenalin were shown to increaseduring a speech in front of a virtual audience. Surprisingly, nosignificant cortisol elevation was found in response to thistask (Kotlyar et al., 2008).
The VR studies described above have all utilized head-mounted displays to present the virtual environment. Mostsuch systems suffer from limited resolution and field-of-viewand are therefore not able to present a virtual environmentto the user in a very realistic manner. In the present paper, weintended to replicate the TSST in a virtual environment usinga fully immersive, projector-based VR system, a so-calledCAVETM (see below).
The TSST includes two components that are hypothesizedto be crucial to induce stress in experimental settings, thatis, a social-evaluative context combined with uncontrollableoutcomes (Dickerson and Kemeny, 2004). In two studies, partof this hypothesis was tested. Gruenewald et al. (2004) let
one group of participants complete the standard TSST andanother group deliver the speech alone in an empty room.They found that the group performing in front of an audienceshowed a significant cortisol response, whereas the groupperforming alone did not, suggesting that social evaluationby others is a critical component in eliciting a stress response.However, a recent study using a similar design let one groupdeliver a speech in front of a committee and another groupdeliver a speech alone in front a one-way mirror that washiding a committee. Both conditions permitted verbal com-munication between the participant and the audience. Theresults showed that there was no significant difference incortisol reactivity between two groups; both conditionsshowed the same amount of cortisol reactivity. The authorsconcluded that visual presence of a panel is not a necessarycondition to exert social evaluation, and that confirming thepresence of a panel by verbal communication is sufficient toelicit a significant stress response (Andrews et al., 2007).
Hence, to increase the element of social evaluation, orpresence, we included the possibility of two-way commu-nication between the test participants and the audience inthe VR version of TSST.
TSST is often used to examine cortisol response changesover time. A hallmark of the HPA axis stress response is that itgenerally habituates quickly to repeated exposures to thesame stressful stimulus. It has been demonstrated that thecortisol response, as a measure of HPA axis reactivity, torepeated TSST provocations gradually decreases most nota-bly during the second session (Pruessner et al., 1997; Schom-mer et al., 2003). Regarding autonomic reactivity torepeated TSST, prior studies report on small decreases overtime (Schommer et al., 2003), or no habituation at all (Gerraet al., 2001; von Kanel et al., 2006). To our knowledge, nostudies to date have examined habituation of the stressresponse in a virtual environment. Consequently, in thepresent study, TSSTwas presented twice on 2 different daysto examine the effects of habituation. Autonomic reactivitywas assessed by heart rate, T-wave amplitude (TWA) as ameasure of sympathetic cardiac control (Rau, 1991; Klineet al., 1998), and high frequency heart rate variability (HF-HRV) as a measure of vagal cardiac control (Berntson et al.,1997). Saliva cortisol was collected to assess HPA axis reac-tivity.
Thus, the objectives of the present study were to (a)assess whether TSST in a virtual environment using a fullyimmersive CAVETM system would induce a stress responsecomparable to previous studies using the real-life TSST, and,if so, (b) whether this stress response would habituate torepeated provocations in a fashion similar to findings fromstudies using real-life TSST.
2. Method
2.1. Participants
The ten participants in the present pilot study were employ-ees from various subdivisions of one of the departmentsinvolved in the study. All were healthy men with no historyof cardiovascular or endocrine disorder. One participant hadpsoriasis and used topical treatment. The others were not onany medication. Their mean age was 28.3 years (24—38
Virtual TSST 1399
years). Because the specific phase in themenstrual cycle mayaffect the magnitude of the salivary cortisol response topsychosocial stressors (Kudielka et al., 2009), women wereexcluded from this small sample pilot study to decreasevariance. The study was conducted in accordance with theDeclaration of Helsinki. All participants signed an informedconsent form, which indicated specifically that participationwas voluntary and could be terminated at any time.
The participants all knew of the existence of the VR lab,but were not involved in any projects in the lab and had notechnical knowledge of the specific equipment. Some of theparticipants knew each other. Therefore all were explicitlytold not to tell the others about the details of the experi-mental procedures.
2.2. Virtual TSST
The VR version of the TSSTwas designed so as to resemble thetraditional TSST as closely as possible. The image displayhardware system was a CAVETM (Cave Automatic VirtualEnvironment, Electronic Visualization Laboratory at the Uni-versity of Illinois) system with three rear projected walls(4 m� 3 m), and one floor projection. Passive stereoscopywas used to achieve three-dimensional vision. The systemalso included an InterSense head tracking system (InterSenseInc.) that created amotion parallax effect to further increasethe realism of the VR simulation. Two virtual rooms werecreated with the software Autodesk1 3ds Max1 9 and EONprofessional 5.5 (EON Reality, Inc.): a waiting room includinga table, three pictures on the walls, two chairs and a smalltable to the right, a couch to the left, and a door on theopposite wall–—and behind the door a room in which thecommittee was seated. Three virtual persons designed bythe company aXYZ design# constituted the committee; amiddle-agedman placed in themiddle, a young woman to theleft, and a young man to the right, all sitting down behind atable, facing the participant (see Fig. 1).
Comments and instructions from the committee weregiven using pre-recorded voices, following the standard TSSTprotocol according to Kirschbaum et al. (1993). For example,if the participant stopped talking for a while, the man in themiddle told him that he had time left, or ‘‘please continue, Iwill tell you when your time is up’’. If it appeared that he washaving trouble determining what to speak about, the manasked him if he had any previous experiences of the specific
Figure 1 The figure shows the committee in front of a participastereoscopic projection. The participants, however, experienced a
kind of job, or what he was going to do if he did not get thejob. If he asked the committee anything, the man answered‘‘yes’’, ‘‘no’’, or ‘‘unfortunately I can’t answer that’’. Theman did not move his lips, but nodded his head slightly insynchrony with the comments. The comments were activatedby one of the test leaders with a remote keyboard invisible tothe test participant. A surround sound system was used toplay the pre-recorded voices.
The virtual persons expressed no emotions, showing aneutral facial expression. To enhance the feeling of realism,the committee members made subtle movements, such ascasual nodding of the head, turning the face away from ortowards the participant, or moving a foot.
2.3. Procedure
The sessions started between 0730 h and 1515 h. The parti-cipants were told not to ingest food, caffeine, or tobaccoduring the 2 h before the experiment. Upon arrival to the lab,the test participant (TP) was placed in a comfortable chairand asked to fill in forms covering background data andinformed consent. Then the physiologic recording equipmentwas attached. He was told that the experiment would last forapproximately one and a half hours, and that he was going toperform two tasks in a VR environment. Next, the experimentwas carried out according to the following sequence ofconditions:
1. BASE: The TP entered the VR waiting room and a 5 minbaseline was recorded.
2. The TP was then let into the other VR room, facing thecommittee. He was told that, after preparation, he wasgoing to give a presentation in front of the committee,pretending that he was applying for a specific job. He wasalso told that, after the presentation, the committeewould give him a second task to perform.
3. PREP: The TP was transferred back to the waiting room toprepare the speech for 5 min. He was permitted to takenotes during preparation, but not permitted to use themduring the presentation.
4. SPEECH: The TP again entered the other room and gavehis presentation in front of the committee (5 min).
5. MATH: The TP performed the second task, which con-sisted of counting backwards from 1687 in steps of 13(5 min).
nt. The interior and the persons are a bit blurred due to theclear 3d environment.
1400 P. Jonsson et al.
6. REST: The TP returned to the waiting room and rested for40 min.
During the second session, the job description was slightlychanged (Schommer et al., 2003). Otherwise, the identicalprocedure was used on both days, with the addition that theTP was informed about the experiment after it had beenconcluded on day 2, and was rewarded with two cinematickets.
3. Data collection and reduction
3.1. Cortisol
Saliva cortisol was collected in sampling tubes with cottonswabs (Salivette1; Sarstedt Ltd., Leicester, UK). Saliva sam-ples were collected after BASE, PREP, TSST (SPEECH + MATH),REST + 10 min, +20 min, +30 min, and +40 min, i.e. 7 sam-ples. Estimation of saliva cortisol was performed using acompetitive radioimmunoassy (RIA), designed for quantita-tive in vitro measurement of cortisol in saliva (SpectriaCortisol Coated Tube RIA, Orion Diagnostica, Espoo, Finland),in accordance with the manufacturer’s specifications. For adetailed description of the analytical methods, see Osterberget al. (2009).
For one participant, one saliva sample was missing due toinsufficient amount of saliva (dry swab) in the +30 min con-dition in the first TSST session; this was also the case foranother participant in the +40 min condition in the secondTSST session. These missing values were replaced with esti-mated values, using the following procedure: the meanquotient (slope) between the +20 min and the +30 min con-dition (first session), and between the +30 min and the+40 min condition (second session), of all other participants’values were calculated. These two quotients, representingthe relative mean decrease, were multiplied by the indivi-dual participant’s cortisol values from the preceding relevantcondition.
3.2. Heart rate
ECG and respiration were recorded at 1 kHz using the ML866Power Lab data acquisition system and analysed using itssoftware Chart 5 (ADInstruments Pty Ltd.) and MATLAB (Math-Works, Inc., Natick, MA). ECG was assessed using disposableelectrodes (Lead II Einthoven) and respiration using a straingauge over the chest. Mean HR was analysed for 5 min in eachcondition: BASE, PREP, SPEECH, MATH, and during the fourfollowing resting periods, i.e. 8 conditions. The same appliesto TWA and RSA below.
3.3. T-wave amplitude (TWA)
TWA is generally considered to be an index of cardiacsympathetic activity, although its reliability has been ques-tioned by some researchers (Furedy and Heslegrave, 1983).TWA was computed as the difference in mV between themaximum 100—300 ms after the R-wave peak and the meanof the isoelectric period (40 ms) between the P- and Q-wave (Rau, 1991) for each heart beat and averaged over5 min.
3.4. High frequency heart rate variability (HF-HRV)
ECG and respiration were sampled at 1 kHz. R—R intervalswere transformed to a tachogram (ms) and linearly inter-polated at 4 Hz. The data were further linearly detrendedand high-pass filtered (second order Butterworth filter,0.10 Hz) to eliminate fluctuations below the respiratoryfrequency. For each 5-min sequence, HRV power spectrawere calculated, for 17 segments of 128 points (32 s) with50% overlap, by means of fast Fourier transform (1024 points)following the application of multiple peak matched windows.The peak matched multiple windows (PM MW) method opti-mizes the mean square error of the spectrum estimate whenthe spectrum can be expected to include peaks (Hansson andSalomonsson, 1997; Hansson, 1999). The PM MW method hasbeen shown to give reliable results for the HRV spectrum(Hansson and Jonsson, 2006; Hansson-Sandsten and Jonsson,2007) and has previously been used in psychophysiologicalresearch (Jonsson, 2007; Jonsson and Hansson-Sandsten,2008). The integral of the power spectrum was studied inthe high frequency (HF) region (0.12—0.4 Hz) that is relatedto respiration (Berntson et al., 1997). The data were log-transformed (ln) to approach a normal distribution. Therespiration measures were used to ensure that the respira-tory rate was within the HF range.
3.5. Inventories
The state scale of the Spielberger state and trait anxietyinventory (STAI) (Spielberger, 1983) was used to estimate theparticipants’ experiences of the TSST. The state scale wascompleted after the 40 min rest period.
3.6. Statistics
Repeated measures ANOVA were used in all analyses for thephysiological measures ( p < 0.05), with experimental CON-DITION and DAY as repeated factors. Greenhouse-Geisseradjustments (2) were used to correct for violation of theassumption of sphericity, and were reported together withunadjusted degrees of freedom, adjusted p-values, and h2.Effect sizes, d, for differences between TSST and baselinewere calculated according to Becker (1988), and correctedfor small sample bias (Hedges and Olkins, 1985).
4. Results
4.1. Cortisol
The omnibus repeated ANOVA showed a significant effect ofDAY [F(1, 9) = 8.89, p = .015, h2 = .50]. However, a CONDI-TION � DAY interaction showed that the increase in cortisol inresponse to TSST was clearly pronounced during the firstsession, but habituated and was nearly absent during thesecond session [F(6, 54) = 4.68, p = .03, h2 = .34, 2 = .28],quadratic trend [Fquad(1, 9) = 5.97, p < .037, h2 = .40]; seeFig. 2.
Because cortisol secretion shows diurnal variation, time ofday was included as a covariate to control for its possible
Figure 2 Cortisol, T-wave amplitude (TWA), and heart rate (HR) as a function of experimental condition. Values are means (�SEM).*p < .05.
Virtual TSST 1401
confounding effect. Neither for day 1, nor for day 2, did timeof day interact with CONDITION: [F(6, 48) = 0.73, n.s., andF(6, 48) = 1.16, n.s.].
Two of the participants were not tested at the same timeon the 2 days. When analysing possible confounding effectsof time difference, no significant results were found for thecovariate, and the results above were about the same:effect of DAY [F(1, 8) = 8.38, p = .02, h2 = .51], and effectof CONDITION � DAY [F(6, 48) = 4.20, p = .044, h2 = .34,2 = .28], and quadratic trend [Fquad(1, 8) = 5.34, p = .049,h2 = .40].
4.2. Heart rate
Due to technical problems, ECG data were missing for oneparticipant during the second session. The analyses werebased on 9 participants. HR varied significantly as a func-tion of CONDITION [F(7, 56) = 20.40, p < .0001, h2 = 72,2 = .26]. HR increased during PREPARATION, SPEECH andMATH compared to baseline, and then recovered and sta-bilized during the four succeeding resting conditions: poly-nomial contrasts [F linear (1, 8) = 34.97, p < .0001, h2 = .81;Fquad (1, 8) = 19.09, p = .002, h2 = .71; Fcubic (1, 8) = 34.24,p < .0001, h2 = .81]; see Fig. 2. There was no significanteffect of DAY or a CONDITION � DAY interaction.
4.3. T-wave amplitude
In concert with HR, a main effect of CONDITION showed thatTWA decreased during the three stress conditions, indicating
increased sympathetic activity, and then recovered and sta-bilized [F(7, 56) = 9.88, p < .002, h2 = .55, 2 = .28; polyno-mial contrasts: F linear(1, 8) = 6.59, p = .033, h2 = .45; Fquad(1,8) = 18.79, p < .002, h2 = .70; Fcub(1, 8) = 11.42, p = .010,h2 = .59]; see Fig. 2. There was no significant effect of DAYor a CONDITION � DAY interaction.
4.4. HF-HRV
No significant results were found for HF-HRV.
4.5. Effect sizes for physiological measures
Effect sizes were calculated for the differences betweenbaseline and (1) the speech task, (2) the math task, and(3) for the mean of the 32 s epoch representing the maximumdifference from baseline during the two TSST tasks; seeTable 1. Means and standard deviations for the max differ-ence for HR were 25.4 BPM (SD = 15.5, t = 4.93, p = .001) andfor TWA �0.12 mV (SD = 0.09, t = 4.01, p = .004) during ses-sion 1. Corresponding data during session 2 for HR were 21.5BPM (SD = 12.4, t = 5.21, p = .001) and for TWA �0.12 mV(SD = 0.08, t = 4.58, p = .002).
4.6. State anxiety scale
The test participants experienced the first TSST session asmore anxiety provoking than the second session: (M = 49.2,SD = 10.6) and (M = 36.9, SD = 8.5), respectively, t(9) = 4.32,p < .005.
Table 1 Effect sizes and increases/decreases in percent forthe differences in cortisol, HR, and TWA, between TSST andbaseline.
Measure Session 1 Session 2
d % d %
Cortisol 1.77 +88 .004 .003
HRSPEECH 1.53 +17 1.24 +13MATH 1.98 +22 1.52 +15MAX 3.68 +40 3.17 +32
TWASPEECH 0.64 �22 0.58 �20MATH 0.70 �24 0.64 �22MAX 1.21 �42 1.12 �39
Note: + = increase, � = decrease.
1402 P. Jonsson et al.
5. Discussion
The present pilot study aimed at investigating the reactivityand possible habituation of the HPA axis and the cardiovas-cular system in response to repeated exposures to a VRversion of the TSST, using a fully immersive CAVETM system.During the first session, there was an almost twofold increase(88%) in saliva cortisol after stress provocation compared tobaseline. Thus, the effect of the cortisol response was con-siderably larger than in prior VR studies inducing social stress:30% increase in cortisol in response to a virtual TSST (Kellyet al., 2007), and no significant results during a similar virtualpublic speaking task (Kotlyar et al., 2008). The effect size ofthis increase (d = 1.77) was actually above the mean effectsize reported in a meta-analysis of cortisol reactivity to thereal-life TSST and similar provocation methods (d = 0.87)(Dickerson and Kemeny, 2004). One possible explanationfor the substantially higher cortisol response in the presentstudy compared to previous studies using VR provocationsmay be the use of a fully immersive CAVETM system, whichgives an illusion closer to the real-life TSST, compared to thehead-mounted displays used in previous studies to presentthe virtual environment. Moreover, the possibility to interactverbally with the committee, although restricted, may haveincreased the feeling of social evaluation.
During the speech task, HR increased by 17% and duringthemath task by 22%. The peak HR during TSSTwas 40% highercompared to baseline. These effects are close to the findingsin real-life TSST studies, most showing HR increases in therange of 24—34% (Kirschbaum et al., 1993; Schommer et al.,2003; Kudielka et al., 2004; Nater et al., 2006; von Kanelet al., 2006). The mean HR increase during the speech taskwas, however, smaller than in the VR study conducted byKotlyar et al. (2008), who noticed a HR increase of approxi-mately 30% in a similar task. TWA during the speech taskdecreased by 22%, and by 24% during the math task. Theminimum value of TWA during TSST represented a 40%decrease compared to baseline. No effects were found forHF-HRV, indicating that the HR increase was mainly a result ofincreased sympathetic activity.
Extending previous VR research on social stress induction,the present study also demonstrated the characteristic HPA
response habituation to a repeated stressor; during thesecond exposure to the virtual TSST, the cortisol responsewas clearly reduced.
Although HR and TWA magnitudes differed significantlyduring TSST compared to baseline, there were no signs ofhabituation. However, this finding concurs with other studiesreporting dissociated stress habituation patterns betweenthe SAM and the HPA response systems to repeated TSSTprovocations (Schommer et al., 2003; von Kanel et al., 2006).Frankenhaeuser (1986) suggested that ‘‘distress’’ is a neces-sary condition of an effortful situation and that it serves toactivate both the HPA and the SAM system. An effortfulsituation without distress leads to SAM activation only. Thus,the unchanged reactivity of HR and TWA during the twosessions may reflect the fact that the same amount of effortmobilization was required. The reduced HPA activation dur-ing the second session may reflect the fact that the task wasperceived as less distressing than it was during the firstsession.
There are a number of limitations in this pilot, and theresults raise some questions. Only young men participated,which of course limits the generalizability. Cortisol showsdiurnal variation, and in future studies, it would be desirableto keep the test time constant, preferably in the afternoonwhen cortisol levels are low. The effect of cortisol reactivitywas rather strong during the first session, as was the habitua-tion during the second session, that is, there was almost nocortisol response at all. We find it hard to believe that the VRversion of TSSTwould be more stress provoking than the TSSTwith a real audience. It must also be noted that in habituationstudies using real-life TSSTcortisol responses decreases overrepeated trials but not as much as in this pilot study. Perhaps,the effect size has been overestimated in the first session (orunder estimated in the second session), due to the smallsample size. The non-response pattern in the second sessionmight of course also reflect that the VR version of TSST differpsychologically from the traditional TSST, which conse-quently is a topic for future studies.
The VR equipment used in the study was rather expen-sive (about 400,000 s). However, technical development inthe area is extremely rapid largely thanks to the computergame industry. For example, hardware solutions that givestereoscopic vision and head tracking are already availableas relatively cheap consumer products. In the foreseeablefuture, it will most likely be possible to present the virtualTSST with a small and portable system, at a reasonableprice.
Although the present findings should be interpreted withcaution, and replicated with larger samples before anydefinite conclusions can be drawn, they do suggest thatVR technology may be used to provide a standardized andeasily replicable tool to induce social stress in laboratorysettings.
Role of funding source
Funding for this study was provided by the Swedish Council ofWorking Life and Social Research (FAS, grant 2006-1514). FAShad no further role in study design; in the collection, analysisand interpretation of data; in the writing of the report; and inthe decision to submit the paper for publication.
Virtual TSST 1403
Conflict of interest
All authors declare that they have no conflicts of interest.
Acknowledgements
Funding for this study was provided by the Swedish Council ofWorking Life and Social Research (FAS, grant 2006-1514,Centre for Medicine and Technology for Work Life and Societyat Lund University; METALUND), the Medical Faculty of LundUniversity, and the County Councils of Southern Sweden.
The authors would like to thank four anonymous reviewersfor their helpful suggestions for improving the paper.
References
Andrews, J., Wadiwalla, M., Juster, R.P., Lord, C., Lupien, S.J.,Pruessner, J.C., 2007. Effects of manipulating the amount ofsocial-evaluative threat on the cortisol stress response in younghealthy men. Behav. Neurosci. 121 (5), 871—876.
Becker, B.J., 1988. Synthesizing standardized mean-change mea-sures. Brit. J. Math. Stat. Psychol. 41, 257—278.
Berntson, G.G., Bigger Jr., J.T., Eckberg, D.L., Grossman, P., Kauf-mann, P.G., Malik, M., Nagaraja, H.N., Porges, S.W., Saul, J.P.,Stone, P.H., van der Molen, M.W., 1997. Heart rate variability:origins, methods, and interpretive caveats. Psychophysiology 34(6), 623—648.
Dickerson, S.S., Kemeny, M.E., 2004. Acute stressors and cortisolresponses: a theoretical integration and synthesis of laboratoryresearch. Psychol. Bull. 130 (3), 355—391.
Frankenhaeuser, M., 1986. A psychobiological framework for re-search on human stress and coping. In: Appley, M.H., Trumbull,R. (Eds.), Dynamics of Stress-physiological, Psychological andSocial Perspectives. Plenum, New York, pp. 101—116.
Furedy, J.J., Heslegrave, R.J., 1983. A consideration of recentcriticisms of the T-wave amplitude index of myocardial sympa-thetic activity. Psychophysiology 20 (2), 204—212.
Gerra, G., Zaimovic, A., Mascetti, G.G., Gardini, S., Zambelli, U.,Timpano, M., Raggi, M.A., Brambilla, F., 2001. Neuroendocrineresponses to experimentally-induced psychological stress inhealthy humans. Psychoneuroendocrinology 26 (1), 91—107.
Gruenewald, T.L., Kemeny, M.E., Aziz, N., Fahey, J.L., 2004. Acutethreat to the social self: shame, social self-esteem, and cortisolactivity. Psychosom. Med. 66 (6), 915—924.
Hansson-Sandsten, M., Jonsson, P., 2007. Multiple window correla-tion analysis of hrv power and respiratory frequency. IEEE Trans.Biomed. Eng. 54 (10), 1770—1779.
Hansson, M., 1999. Optimized weighted averaging of peakedmatched multiple window spectrum estimates. IEEE Trans. SignalProcess. 47, 1141—1146.
Hansson, M., Jonsson, P., 2006. Estimation of HRV spectrogram usingmultiple windowmethods focussing on the high frequency power.Med. Eng. Phys. 28 (8), 749—761.
Hansson, M., Salomonsson, G., 1997. A multiple window method forestimation of peaked spectra. IEEE Trans. Signal Process. 45,778—781.
Hedges, L.V., Olkins, I., 1985. Statistical Methods for Meta-analysis.Academic Press, Orlando, FL.
Jonsson, P., 2007. Respiratory sinus arrhythmia as a function of stateanxiety in healthy individuals. Int. J. Psychophysiol. 63 (1), 48—55.
Jonsson, P., Hansson-Sandsten, M., 2008. Respiratory sinus arrhyth-mia in response to fear-relevant and fear-irrelevant stimuli.Scand. J. Psychol. 49 (2), 123—131.
Kelly, M.M., Tyrka, A.R., Anderson, G.M., Price, L.H., Carpenter, L.L.,2008. Sex differences in emotional and physiological responses tothe Trier Social Stress Test. J. Behav. Ther. Exp. Psychiatry 39 (1),87—98.
Kelly, O., Matheson, K., Martinez, A., Merali, Z., Anisman, H., 2007.Psychosocial stress evoked by a virtual audience: relationto neuroendocrine activity. Cyberpsychol. Behav. 10 (5), 655—662.
Kirschbaum, C., Pirke, K.M., Hellhammer, D.H., 1993. The Trier socialstress test–—a tool for investigating psychobiological stressresponses in a laboratory setting. Neuropsychobiology 28(1—2), 76—81.
Kline, K.P., Ginsburg, G.P., Johnston, J.R., 1998. T-wave amplitude:relationships to phasic RSA and heart period changes. Int. J.Psychophysiol. 29 (3), 291—301.
Kotlyar, M., Donahue, C., Thuras, P., Kushner, M.G., O’Gorman, N.,Smith, E.A., Adson, D.E., 2008. Physiological response to a speechstressor presented in a virtual reality environment. Psychophysi-ology 45 (6), 1034—1037.
Kudielka, B.M., Hellhammer, D.H., Wust, S., 2009. Why do werespond so differently? Reviewing determinants of human salivarycortisol responses to challenge. Psychoneuroendocrinology 34(1), 2—18.
Kudielka, B.M., Schommer, N.C., Hellhammer, D.H., Kirschbaum, C.,2004. Acute HPA axis responses, heart rate, and mood changes topsychosocial stress (TSST) in humans at different times of day.Psychoneuroendocrinolohy 29 (8), 983—992.
Kudielka, B.M., von Kanel, R., Preckel, D., Zgraggen, L., Mischler, K.,Fischer, J.E., 2006. Exhaustion is associated with reduced habit-uation of free cortisol responses to repeated acute psychosocialstress. Biol. Psychol. 72 (2), 147—153.
Nater, U.M., La Marca, R., Florin, L., Moses, A., Langhans, W., Koller,M.M., Ehlert, U., 2006. Stress-induced changes in human salivaryalpha-amylase activity-associations with adrenergic activity. Psy-choneuroendocrinology 31 (1), 49—58.
Osterberg, K., Karlson, B., Hansen, A.M., 2009. Cognitive perfor-mance in patients with burnout, in relation to diurnal salivarycortisol. Stress 12 (1), 70—81.
Pertaub, D.P., Slater, M., Barker, C., 2002. An experiment on publicspeaking anxiety in response to three different types of virtualaudience. Presence-Teleop. Virt. 11 (1), 68—78.
Pruessner, J.C., Gaab, J., Hellhammer, D.H., Lintz, D., Schommer,N., Kirschbaum, C., 1997. Increasing correlations betweenpersonality traits and cortisol stress responses obtained bydata aggregation. Psychoneuroendocrinology 22 (8), 615—625.
Rau, H., 1991. Responses of the T-wave amplitude as a function ofactive and passive tasks and beta-adrenergic blockade. Psycho-physiology 28, 231—239.
Schommer, N., Hellhammer, D., Kirschbaum, C., 2003. Dissociationbetween reactivity of the hypothalamus-pituitary-adrenal axisand the sympathetic-adrenal-medullary system to repeated psy-chosocial stress. Psychosom. Med. 65 (3), 450—460.
Slater, M., Pertaub, D.P., Barker, C., Clark, D.M., 2006. An experi-mental study on fear of public speaking using a virtual environ-ment. Cyberpsychol. Behav. 9 (5), 627—633.
Spielberger, C.D., 1983. Manual for the State-Trait Anxiety Inventory,STAI (Form Y). Consulting Psychologists Press Inc., Palo Alto, CA.
von Kanel, R., Kudielka, B.M., Preckel, D., Hanebuth, D., Fischer,J.E., 2006. Delayed response and lack of habituation in plasmainterleukin-6 to acute mental stress in men. Brain Behav. Immun.20 (1), 40—48.
Zgraggen, L., Fischer, J.E., Mischler, K., Preckel, D., Kudielka, B.M.,von Kanel, R., 2005. Relationship between hemoconcentrationand blood coagulation responses to acute mental stress. Thromb.Res. 115 (3), 175—183.
