Response to Stress

Journal Article
of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
  Scandinavian Journal of Psychology, 2006, 47, 137–148  © 2006 The Scandinavian Psychological Associations/Blackwell Publishing Ltd. Published by Blackwell Publishing Ltd., 9600 GarsingtonRoad, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA. ISSN 0036-5564.  BlackwellPublishingLtd  Health and Disability  Repressive coping style and autonomic reactions to two experimental stressors in healthy men and women  MICHAEL MARTINI JØRGENSEN   1  and ROBERT ZACHARIAE   1,2 1  Psychooncology Research Unit, Aarhus University Hospital, Aarhus, Denmark 2  Department of Psychology, Aarhus University, Aarhus, Denmark   Jørgensen, M. M. & Zachariae, R. (2006). Repressive coping style and autonomic reactions to two experimental stressors in healthy men andwomen. Scandinavian Journal of Psychology  , 47   , 137–148.Autonomic and affective responses to two different stress tasks were measured in 45 males and 74 females, categorized as repressive, true low-anxious, true high-anxious, and defensive high-anxious. Electrodermal activity (EDA) was used as a measure of sympathetic activity and thehigh frequency (HF) spectral component of heart rate variability as a measure of parasympathetic activity. Contrary to our predictions, reactivityof repressors did not differ from the reactivity of true low-anxious participants. The results draw attention to previous inconsistent findingswithin the literature on repressive coping style and autonomic nervous system dysregulation. It is suggested that future research could benefitfrom the use of more consistent operationalizations of the repressive coping construct and from comparing alternative measures of repressivecoping within the same study.  Key words:  Autonomic nervous system, electrodermal activity, experimental stress, heart rate variability, repressive coping style.  Michael Martini Jørgensen, PhD, Psychooncology Research Unit, Aarhus University Hospital and Department of Psychology, Nobelparken, JensChr. Skous vej 4, DK-8000 Aarhus C, Denmark. Tel: +45 89 42 66 27; fax: +45 89 42 66 23  ; e-mail:  INTRODUCTIONThe influence of expressed versus non-expressed emotions inhuman health and well-being has long been a central themein psychosomatic and behavioral medicine. The general hypo-thesis that non-expression or inhibition of emotions mayincrease vulnerability to illness and disease, while disclosureand expression of emotions may either prevent or relievesuffering and disease, dominates the literature on emotions andhealth (Pennebaker, 1995, 1997; Singer, 1990; Wickramasekera,1998). As a theoretical framework for psychophysiologicalresearch into emotional inhibition, Schwartz (1983, 1990)has proposed a model based on General Systems Theorythat integrates findings from biofeedback research, neuro-endocrinology, and psychology. In short, the model proposesthat disattention to signals of distress may result in physical  dysregulation  due to cortical overriding of the autonomicnervous system and immunoregulatory homeostatic mecha-nisms. The theory predicts that individuals characterized bya repressive/defensive coping style will, by some mechanismof cerebral disconnection, have a tendency to disattend toimportant negative feedback   , thereby exhibiting a discrepancy  between psychological reactions (e.g. no perception or recallof negative emotions) and physiological responses (e.g. highskin conductance levels) to stressful stimuli. Such a repres-sive reaction pattern may subsequently lead to disattentionto risk-related behaviors (Roth & Cohen, 1986; Schwartz, 1990;Suls & Fletcher, 1985). If the repressive/defensive responsepattern constitutes a relatively stable trait, it may thereforeprevent the individual from coping effectively with differenttypes of stressful situations, leading to both autonomicnervous system (ANS) and immune dysregulation, which inturn may increase the risk of physiological disease.In recent years, several investigators have studied therepressive/expressive aspect of emotional coping using a two-dimensional approach, combining a trait measure of negativeaffectivity (NA), usually trait-anxiety, with the MarloweCrowne Social Desirability Scale (MC-SDS) (Crowne &Marlowe, 1964), thought to measure defensiveness (DEF).This approach, introduced by Weinberger (1990), combineshigh and low scores on each dimension into four categories,resulting in what is usually interpreted as four prototypicalcoping styles: truly low-anxious (low NA/low DEF), repres-sion (low NA/high DEF), truly high-anxious (high NA/lowDEF), and defensive high-anxious (high NA/high DEF).Theoretically, the combination of NA and DEF offers asolution to the problems of interpreting the self-reportedscores of scales such as the Repression-Sensitization (R-S)Scale (Byrne, 1961). The R-S scale, which was developed fromitems showing maximum contrasts between groups clinicallyrated as “defensive” and normals, has shown high correlationswith trait measures of anxiety, suggesting that it is unable todistinguish between repressed anxiety and “truly” reportedlow anxiety. This distinction is the main rationale behind thesrcinal approach of Weinberger (1990), which attempts tocircumvent the ambivalence of the low self-reported anxiety   138  M. M. Jørgensen and R. Zachariae  Scand J Psychol 47 (2006)© 2006 The Scandinavian Psychological Associations/Blackwell Publishing Ltd.  scores by taking into account the individuals’ scores on theMC-SDS, where high scores are interpreted as being associ-ated with an inclination to emotional inhibition.The study of autonomic reactivity of emotionally expressiveversus non-expressive healthy individuals during experimen-tally induced stress has been one commonly used method of investigating the hypothesized association between emotionalnon-expression and autonomic dysregulation. In accordancewith the concept of repressive coping (Schwartz, 1983), arobust measure of repressive coping style would be expectedto identify a subgroup of individuals exhibiting autonomicdysregulation as well as discrepancy between subjective reporteddistress and physiological reactivity (understated report of distress). So far, most studies using the Weinberger approachhave focused on trait anxiety as the NA measure and eithercardiovascular activity or electrodermal activity (EDA) asthe measure of ANS reactivity.EDA is involuntary and is thought to be associated onlywith excitatory sympathetic nerve impulses (Boucsein, 1992).It has been argued that EDA reflects emotional inhibition,with EDA being sensitive to behavioral inhibition and threatbut relatively uncontaminated by motor responses (Barger,Kircher & Croyle, 1997), and this measure has thereforebeen seen as a particularly interesting autonomic indicatorof repressive coping style. When reviewing the literature onEDA and repressive coping, defined as high scores on MC-SDS combined with low scores on an anxiety-measure, we findthat some investigators have reported associations betweenrepressive coping style and increased EDA during differenttypes of experimental stress conditions (Barger et al.  , 1997;Gudjonsson, 1981; Sparks, Pellechia & Irvine, 1999;Weinberger, Schwartz & Davidson, 1979) while others havefound no associations (Houtveen, Rietveld, Schoutrop,Spiering & Brosschot, 2001; Kiecolt-Glaser & Greenberg,1983; Lang, Greenwald, Bradley & Hamm, 1993; Tomaka,Blascovich & Kelsey, 1992; Wallbott & Scherer, 1991). Itcould be that the inconsistent results reflect differences inthe experimental challenges used, and that only emotionallythreatening tasks and/or specific social contexts have thepotential to elicit differences in EDA between the differentcoping styles, as has previously been suggested (Barger et al.  ,1997). The experimental challenges used in the studies withpositive findings, e.g. phrase completion tasks, self-disclosingspeech tasks and emotional film excerpts, are consistent withsuch a hypothesis. However, emotional threatening stimuliwere also used in some of the negative studies. Thus noEDA-differences were found between repressors andnon-repressors in studies where pictures with aggressive andsexual content were used as stressors (Houtveen et al.  , 2001;Lang et al.  , 1993). Taken together, the results concerning therelationship between repressive coping style and EDA are sofar inconclusive.The results of several studies using cardiovascular meas-ures suggest that individuals categorized as repressors  reactwith greater heart rate (HR) reactivity when exposed to variousstressors than the control groups, including the group of   true low-anxious  (Asendorpf & Scherer, 1983; Derakshan &Eysenck, 1997; Kiecolt-Glaser & Greenberg, 1983; Newton& Contrada, 1992; Weinberger et al.  , 1979). One study reportedincreased systolic blood pressure reactivity in the group of repressors following a mental arithmetic task (King, Taylor,Albright & Haskell, 1990). In another study, increaseddiastolic blood pressure was seen in the group of repressorsduring an assertion role-play based on imagery (Kiecolt-Glaser& Greenberg, 1983). Other investigators have been unable toconfirm such findings. In several studies, no differences werefound in HR reactivity in repressors following experimentalstress, when compared to control groups (Barger et al.  , 1997;Barger, Marsland, Bachen & Manuck, 2000; Derakshan &Eysenck, 1999; Houtveen et al.  , 2001; King et al.  , 1990; Lang  et al.  , 1993; Tomaka et al.  , 1992), and in three out of fivestudies (Barger et al.  , 2000; Newton & Contrada, 1992;Warrenburg, Levine, Schwartz, Fontana, Kerns, Delaney &Mattson, 1989) no associations were found between repres-sive coping style and blood pressure. The apparent inconsist-encies between the results of these studies could reflect alimited ability of cardiovascular measures in differentiatingbetween repressive and non-repressive coping styles (Barger  et al.  , 2000). In general, the one-dimensional classification of an individual as low or high in HR reactivity under a givencondition ignores possible differences in the autonomicsrcins of this reactivity. Since HR is determined by both thesympathetic and parasympathetic branches of the ANS, HRreactivity may be better conceptualized as representing atwo-dimensional autonomic plane (sympathetic nervoussystem (SNS) and parasympathetic nervous system (PNS))(Berntson, Bigger, Eckberg et al.  , 1997; Berntson, Cacioppo,Quigley & Fabro, 1994). Thus, an individual’s classificationas high in HR reactivity could srcinate from elevated SNSactivity, decreased PNS activity or reciprocal activation of SNS and PNS. It therefore seems relevant to includemeasures of heart rate variability (HRV), which through   power spectrum analysis  can be analyzed and filtered intohigh frequency (HF) and low frequency (LF) componentswith the HF component being a standard measure of para-sympathetic influences on the heart (Task Force of theEuropean Society of Cardiology and the North AmericanSociety of Pacing and Electrophysiology, 1996). EDA meas-ures could then be included as an indicator of sympatheticactivity. We are aware of only one study of the associationbetween HRV and repressive coping style (Fuller, 1992).Here, no differences were found for PNS activity before andafter a naturally occurring stressor, but repressors and truehigh-anxious individuals did, however, overall differ signifi-cantly from the true low-anxious participants by exhibitingrelatively less parasympathetic influence.In some studies, repressors have been found to report sig-nificantly less emotional distress during experimental stresscompared to one or more of the control groups (Asendorpf & Scherer, 1983; Newton & Contrada, 1992; Weinberger et al.  ,   Scand J Psychol 47 (2006)  Repressive coping style and stress reactivity  139  © 2006 The Scandinavian Psychological Associations/Blackwell Publishing Ltd.  1979). It is, however, only in few of these studies that dif-ferences were found between repressors and low-anxiousparticipants (Asendorpf & Scherer, 1983; Weinberger et al.  ,1979). Moreover, several investigators were unable to detectany differences in self-reported distress or arousal betweenrepressors and any of the remaining three groups of non-repressors (Barger et al.  , 1997; Houtveen et al.  , 2001; Lang  et al.  , 1993; Tomaka et al.  , 1992; Wallbott H.G. et al.  , 1991).However, when calculating a measure of verbal-autonomicdiscrepancy, the results of three out of four studies(Asendorpf & Scherer, 1983; Barger et al.  , 1997; Brosschot& Janssen, 1998; Newton & Contrada, 1992) indicated adegree of discrepancy (i.e. under-reporting) in repressors,which was higher or in the opposite direction of the discrep-ancy seen in one or more of the remaining groups.The present study was designed with the aim of providingadditional data on the influence of repressive coping style,using the Weinberger construct, on autonomic reactivity andpsychophysiological discrepancy following experimentallyinduced stress, while taking as many of the methodologicalissues of the previous studies into consideration as possible,thereby testing the validity of the repressive coping construct.Two different mental stress tasks were included: mentalarithmetic and self-disclosure. Based on predictions derivedfrom a theoretical model of emotional repression, autonomicdysregulation and psychophysiological discrepancy (Schwartz,1983; 1990) as well as on the evidence presented in the liter-ature, we wished to test the following three hypotheses:(1)Repressors will show autonomic dysregulation expressedas greater sympathetic reactivity following a stress taskcompared to low-anxious participants.(2)Repressors will exhibit autonomic dysregulation expressedas either less or greater parasympathetic reactivity (HFpower) following a stress task compared to low-anxiousparticipants.(3)Repressors will (3a) report less emotional distress thantrue low-anxious participants, and (3b) show greaterpsychophysiological discrepancy (i.e. under-reporting of NA relative to sympathetic reactivity) compared to truelow-anxious participants following a stress task.In addition to the testing of these specific hypotheses, wewished to examine possible differences between repressorsand low-anxious participants at baseline and during recovery.We also wanted to explore possible differences at baselineand during the stress tasks between repressors and truelow-anxious participants with respect to other HRV-relatedmeasures. Insofar as the above stated hypotheses were con-firmed, we expected the results to be more clear-cut for self-disclosure than mental arithmetic, based on the assumptionthat an emotional self-disclosure task in front of anotherperson is more likely to elicit a repressive coping behaviorthan “mere” problem solving as in a mental arithmetic task(Barger et al.  , 1997; Weinberger, 1990). Both true and defen-sive high-anxious participants were included in the study tocontrol for the possibility that differences found betweenrepressors and true low-anxious participants could be explainedby main effects of defensiveness or trait anxiety alone.METHODS  Participants  The participants were recruited among graduate and undergradu-ate students, who previously, during their participation in otherquestionnaire-based studies, had indicated an interest in participat-ing in upcoming psychophysiological experiments. Of 214 graduateand undergraduate students, who received letters inviting them toparticipate in “a psychophysiological study of personality andstress”, a total of 119 individuals (mean age (  ±  SD): 24.9 ±  4.0 years,range 20–47 years; 45 men, 74 women) were able to and volunteeredto take part in the present study.  Procedure  On the day of the experiment, the participants were asked to refrainfrom smoking or caffeine intake two hours prior to the experiment.After giving their informed consent, the participants completed aquestionnaire combining the 33-item MC-SDS (Crowne & Marlowe,1960) and the Bendig 20-item version of the Taylor Manifest AnxietyScale (TMAS) (Bendig, 1956). Both questionnaires had been translatedinto Danish using the translation-back-translation method (Bradley,1994) and tested in a student sample. The Danish versions of MC-SDS and TMAS showed acceptable internal consistencies (KR-20alpha coefficient = 0.74 and 0.82; N   = 425) and 10-week test-retestreliabilities of 0.78 and 0.82 (  N   = 107) (Christensen, Jørgensen &Zachariae, 1999). Readministration between 18 and 24 months lateryielded test-retest reliabilities of 0.66 for MC-SDS (  N   = 101) and0.76 for TMAS (  N   = 99).To exclude participants with symptoms or on medication thatcould be hypothesized to confound the physiological measurements,they also completed a brief physical health questionnaire askingabout any chronic disease, smoking, alcohol consumption, use of medications, as well as current symptoms of disease.After having completed the questionnaires, they were asked towash and dry their hands before settling into a comfortable chair, afterwhich physiological monitoring devices for continuous measurementof HR and EDA were attached (see below). The participants were thenexposed to the following experimental conditions: (1) Twenty-minuterest period   : The participants were asked to close their eyes and restcomfortably in the chair while listening to classical music (acousticguitar). The first 10 minutes served as the habituation phase, whilethe last 10 minutes of the resting period served as baseline. Theparticipants then completed an Emotional Stress Response Scale(ESRS-1) (see below). (2) The first 10-minute stress induction  : Thevolunteers were then subjected to either (a) a 10-minute mentalarithmetic stress task or (b) a 10-minute self-disclosure task (phrase-completion). The two stress tasks were presented in counterbalancedorder. Immediately after the stress task, the participants were againasked to complete the ESRS (ESRS-2). (3) Ten-minute recovery fromthe first stress induction  : The participants were then asked to relaxin the chair and listen to the same music as during the initial restperiod for 10 minutes. They were then asked to complete the ESRS-3.(4) Second 10-minute stress induction  : Depending on which stresstask had been presented during the first stress-induction period, theparticipants were then subjected to either (a) a 10-minute mentalarithmetic stress task or (b) a 10-minute self-disclosure task. Theparticipants then completed the ESRS-4. (5) Ten-minute recovery   140  M. M. Jørgensen and R. Zachariae  Scand J Psychol 47 (2006)© 2006 The Scandinavian Psychological Associations/Blackwell Publishing Ltd.   from the second stress induction  : The participants were again askedto relax and listen to music for 10 minutes and then completed theESRS-5. (6) Debriefing   : After the experiment, the electrodes wereremoved. The participants were then debriefed and asked abouttheir experiences during the experiment.  Stress-induction  Mental arithmetic task.  A mental arithmetic task was used as aproblem-solving task. The test consisted of addition, subtraction,multiplication, and division tasks presented with an interval of 8 s.If a task was solved correctly, a flower would appear on the screenas a positive reinforcement and the next task would be presentedimmediately after. If the task was not solved correctly within the 8 s,a loud beep would sound as a negative reinforcement, and the nexttask would appear on the screen. The computer program was designedso that all participants, regardless of their mathematical skills, wouldreceive approximately the same proportion of incorrect answers, byautomatically adjusting the difficulty level in response to the numberof correct or incorrect responses given during the task. The percent-age of incorrect answers were calculated continuously, presented onthe screen, and stored in the computer. A scale in the upper rightcorner of the screen indicated how “skilled” the subject was, com-pared to “the average subject”. Regardless of the difficulty level andthe number of correct answers, the individual feedback from this scalewould overall be similar and most of the time be indicating that theperformance was more or less “below average”. A similar task hadpreviously been found to be a reliable stress-induction procedure(Zachariae, Jørgensen, Ehrnrooth, Svendsen & Bjerring, 1998).  Self-disclosure task.  A timed phrase-completion task, similar to oneused by Weinberger and Davidson (1994), was constructed and usedas a self-disclosure task. Ten phrase stems were designed to direct theparticipants’ attention to negative experiences or aspects of the self (e.g., “I felt ashamed . . .”, “If I was unfaithful to my beloved . . .”).Five positively framed stems designed to direct the participants’attention to positive experiences or aspects of the self (e.g.,“I feltpopular . . .”) were added in a counterbalanced way. All 15 phraseswere presented on a computer screen with an interval of 40 s. Theparticipants were given “expressive” instructions that definednegative as well as positive self-disclosure as desirable and wereinstructed to complete the sentences with a minimum of two words.The self-disclosure task was completed in the presence of a researchassistant and videotaped. The participants were informed that thevideotape of their performance would later be analyzed, discussed,and evaluated by a panel of experts.  Measurements  Psychological stress reactivity.  An Emotional Stress Response Scale(ESRS) was constructed to measure subjective reactions to theexperimental conditions and consisted of five dimensions of negativeaffectivity, each consisting of two items with a seven-point Likert scaleas response format. The five dimensions were: (1) stress (stressed,relaxed), (2) anxiety (nervous, calm), (3) anger (irritable, patient),(4) helplessness (helpless, in control), and (5) embarrassment(embarrassed, ill at ease). A total NA score was calculated as a sumof all ten items. Internal consistency (Cronbach’s alpha) of the totalscale was found to be satisfactory (0.80 at baseline and 0.92 for theself-disclosure task).  Sympathetic activity.  Changes in EDA were recorded using standardAg-AgCl electrodes, 8 mm in diameter (Med-Associates) attachedwith adhesive collars (Med-Associates) to the medial phalanx of theindex and middle finger of the non-dominant hand. Skin conductanceelectrode paste with 0.5% saline in a neutral base (MED Associates)was used as the electrolyte. Skin conductance was measured using a24-bit DC coupled constant voltage (0.5 V) skin conductance coupler(Psylab, Contact Precision Instruments (CPI), model SC5). A PCrunning CPI’s PsyLab Software under Windows 98 controlled theEDA acquisition. For EDA recording, the sample rate was 25 Hz.The PSYLAB software was programmed to calculate the averageskin conductance level (SCL) per minute and to record the (non-specific) skin conductance responses (nSCR) for each minute duringthe different parts of the experiment. A skin conductance responsewas defined as a through-to-peak increase of at least 0.03  µ  S. Thecumulative amplitude of skin conductance responses (SCR   amp  ) wasalso calculated and averaged for each minute. To increase the relia-bility of responses and to simplify statistical analyses, the EDA andHRV responses were averaged over the different 10-minute experi-mental conditions.  Parasympathetic activity.  An electrocardiogram (ECG) was obtainedfrom three disposable Ag-AgCl ECG electrodes, 34 mm in diameter(Medicotest). The electrodes were attached to the right clavicle, tothe left rib cage and to the right waist (reference electrode). TheECG-signal was detected with ElectroCard 2000 (Bang & OlufsenMedicom a/s, version 093). ElectroCard 2000 is a research device,which has been developed for clinical research on HRV, featuring thereal-time HRV analysis in time and frequency domain in referenceto the recommendations from the Task Force of the European Societyof Cardiology and the North American Society of Pacing andElectrophysiology (1996). It collects the ECG signal at a samplingfrequency of 1,000 Hz, 16 bits in one channel, and determines thenormal sinus rhythm of the heart as a sequence of RR-intervals. TheRR-intervals are also stored in memory for transmission to a PC withthe manually marked events, enabling further off-line analysis. TheRR-intervals obtained from the ECG-signal were used for furtheranalysis in the time and frequency domain. The value calculated inthe time domain was the mean HR. Power spectral analysis, using1,024 point Fast Fourier Transformation and a two-minute Hanningwindow, was used for the calculation of LF (0.04 to 0.15 Hz) andHF (0.15 to 0.50 Hz) in absolute units (abs), and normalized units(nu), as well as total power (TP) and the LF/HF ratio. The resultsof normalized LF, HF and the LF/HF ratio basically reflect thesame measure, as LF   nu  and HF   nu  is in fact a simple monotone trans-formation of the LF/HF ratio (Højgaard, Holstein-Rathlou, Agner& Kanters, 1998). Therefore, in the following we only report theresults for normalized HF (HF   nu  ) as a measure of the relativedistribution of HF.All analyses of the raw data were conducted by a person who wasblind to the MC-SDS and TMAS scores of the participants and theexperimental conditions.  Coping style classification  Repressors were defined as individuals scoring below the median of TMAS and above the median of the MC-SDS. True low-anxiousparticipants were defined as having scores below the median on theTMAS and below the median of the MC-SDS; true high-anxiousparticipants were defined as scoring above the median on theTMAS and below the median of the MC-SDS, and defensive-high-anxious participants were defined as scoring above the median onthe TMAS and above the median of MC-SDS. The classificationwas based on the descriptive statistics for the present sample (asopposed to using normative cut-off values) in accordance with theprocedure used in the majority of studies investigating psychophysi-ological reactions of repressors and non-repressors. It was also
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks