| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Centre for Psychological Trauma, Department of Psychiatry (R.J.L.L., E.P.M.v.M., M.J., M.O., B.P.R.G.), de Bascule and Department of Child and Adolescent Psychiatry, Academic Medical Centre (R.J.L.L.), University of Amsterdam, Amsterdam, The Netherlands; Graduate School Neurosciences, Amsterdam, The Netherlands (R.J.L.L.); and Centre for Work-Related Mental Disorders, Altrecht Institute for Mental Health Care, Utrecht, The Netherlands (I.V.E.C.).
Address correspondence and reprint requests to Ramón J. L. Lindauer, MA, MD, Academic Medical Centre, Department of Child and Adolescent Psychiatry, Tafelbergweg 25, 1105 BC Amsterdam, The Netherlands. E-mail: R.J.Lindauer{at}amc.uva.nl; RJL.Lindauer{at}12move.nl
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Methods: Thirty-nine subjects with PTSD (24 civilian outpatients and 15 police officers) and 15 trauma-exposed, non-PTSD control subjects underwent psychophysiological assessment while listening to neutral, stressful, and trauma scripts. Psychophysiological measures were heart rate (HR) and blood pressure in combination with subjective anxiety ratings. In a randomized clinical trial, 20 of the civilians were then assigned to treatment or waitlist groups. Psychophysiological assessment was repeated on them after the treatment stage.
Results: Both civilians and police with PTSD showed significantly higher HR responses to trauma scripts than the control subjects. After successful psychotherapy with the civilians, HR responsivity to the trauma scripts was significantly reduced, and it correlated positively with PTSD clinical symptoms.
Conclusions: We confirmed previous findings of heightened psychophysiological responses in PTSD for two new populations while minimizing comorbidity and medication as confounding factors. Successful psychotherapy normalized HR response to trauma imagery.
Key Words: posttraumatic stress disorder • psychophysiology • script-driven imagery • randomized clinical trial • psychotherapy
Abbreviations: BEP = brief eclectic psychotherapy; DBP = diastolic blood pressure; EMDR = eye movement desensitization and reprocessing; EMG = electromyographic; HR = heart rate; M = mean; MANOVA = multivariate analysis of variance; PLES = Police Life Event Scale; PTSD = posttraumatic stress disorder; SBP = systolic blood pressure; SCID = Structured Clinical Interview for DSM-IV; SD = standard deviation; SI-PTSD = Structured Interview for Posttraumatic Stress Disorder; STAI = State-Trait Anxiety Inventory; TG = treatment group; WG = waitlist group.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Previous studies have been limited by a focus on specific populations such as combat veterans (3–5,7–14), victims of child abuse (15), motor vehicle accidents (16,17), patients with breast cancer (18), or preschool children (19). The results in many studies could also have been affected by high comorbidity rates (10,20–23) or by the use of psychiatric medication (10,22,23). The first aim of our study was to provide a controlled comparison of psychophysiological responsivity in PTSD populations not studied before—a range of civilian outpatients and police officers—while minimizing any confounding effects of comorbidity and psychiatric medication.
If a heightened physiological responsivity is associated with PTSD, one might expect a normalization of this response after successful treatment of PTSD. Some studies that have assessed treatment responses have indeed reported diminished psychophysiological reactions to trauma-related imagery after treatment (17,24–27). These studies, however, were mostly case reports or noncontrolled studies in which habituation to the scripts over time could have also explained the reduction (6). Some studies were randomized trials and have also reported diminished psychophysiological reactions after treatment (28–31). However, these studies are limited because of: 1) no inclusion of a traumatized control group without PTSD at the pretest stage and no use of different scripts with neutral, stressful, and trauma narratives, which is necessary to assess the specific physiological reactions to traumatic reminders; and 2) most of these studies compared different treatments with an overall decreased effect on physiological reactions after treatment and did not include a control or waiting-list condition (28,30,31), except for one study (29) that also found decreased physiological reactions in the control condition. The authors, therefore, concluded that these physiological findings reflected an apparent habituation effect and were not differentially affected by treatment. The second aim of our study was to investigate the effects of psychotherapy on the psychophysiological responsivity in a randomized, controlled trial. We examined whether psychophysiological responses to trauma- specific or other imagery are mitigated in psychotherapy compared with a waiting-list condition and whether any such reductions in response specifically correspond to changes in PTSD symptomatology at the posttest stage, ie, not an effect of habituation.
We hypothesized that 1) both the civilian outpatients and the PTSD police officers would show heightened psychophysiological responses specifically to the trauma scripts in comparison to the controls; 2) psychotherapeutic treatment would normalize such responses and these responses did not only reflect a habituation effect; and 3) posttreatment psychophysiological responses to trauma scripts would correlate positively with the posttreatment PTSD clinical symptom picture.
The first part of our study compares our civilian and police PTSD samples with a non-PTSD, trauma-exposed control group in terms of baseline conditions; psychophysiological responses to neutral, stressful, and trauma-specific scripts; and physiological recovery after the scripts. The second part reports on a randomized, controlled trial involving the civilian outpatients, which assesses the effects of psychotherapy in terms of clinical and psychophysiological variables.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
The diagnosis of PTSD, based on the DSM-IV criteria, was made with the Structured Interview for Posttraumatic Stress Disorder (SI-PTSD) (32). For the Dutch version of the SI-PTSD, we found a Cronbach’s alpha of 0.93 and a Cohen’s kappa of 0.88, which can be considered acceptable (33). Besides assessing PTSD, the SI-PTSD elicits information about the presence or absence of the three symptom clusters (reexperiencing, avoidance, and hyperarousal) and scales their severity in both a current and a lifetime perspective. The Structured Clinical Interview for DSM-IV (SCID) (34) was administered to assess comorbidity, the intensity of any major depression (mild to severe), and the severity of any excessive alcohol intake. Three of the 24 civilians and three of the 15 police officers with PTSD currently exhibited a mild secondary (onset after PTSD) first episode of major depression. Although some overlap existed between the PTSD and depressive symptoms, no antidepressive medication was needed. PTSD was the primary diagnosis on which the psychotherapy for the civilians would focus. All participants also completed the List of Traumatic Events and the Police Life Event Scale (PLES) (trauma history), two semistructured interviews recording numbers and severity of traumatic experiences in the past (35,36). An added question about the perceived adverse effects of each specific event was rated on a five-point scale (1 = no effects; 5 = very strong effects). The events listed satisfied the stressor A(1) criterion for PTSD diagnosis. A background questionnaire was also administered to collect information about education.
Exclusion criteria included any major lifetime or current medical or psychiatric diagnosis: organic mental disorder, head trauma with loss of consciousness, mental retardation, seizures, neurological disorders, schizophrenia, psychotic disorders, bipolar disorder, moderate to severe depressive disorder, panic disorder, phobia, obsessive compulsive disorder, or dissociative disorder. Candidates with lifetime or current alcohol or drug abuse or dependence, or use of psychiatric medication, were also excluded.
Procedure
All selected participants visited the Department of Psychiatry for psychophysiological assessment. The 24 civilian PTSD outpatients were then assigned randomly to a treatment or waitlist group. A colleague who had done no assessments used a computer program to randomly assign the patients to each condition in a block design. The researchers were blind to group assignment. The PTSD police officers and the traumatized non-PTSD control subjects were examined at the pretest stage only. After termination of the treatment stage 4 months later, all the civilians in both the treatment and waitlist groups were tested a second time with the List of Traumatic Events, SI-PTSD, and SCID. Within 1 week afterward, a second psychophysiological assessment was carried out.
Psychophysiological Assessment
We asked all 54 participants to relate two past personal experiences—one stressful but nontraumatic event and one traumatic event that satisfied DSM-IV criterion A for PTSD—and we used these to create scripts as described in the literature (4). For the patients with PTSD, we created the trauma script using the event thought to have caused PTSD; in the traumatized control group, we asked for the traumatic event that had had the greatest impact on their lives. The trauma scripts for most participants, both civilian and police, involved interpersonal violence. The neutral script (brushing your teeth) was the same for every participant, and the stressful script was based on the personally experienced nontraumatic event. During the psychophysiological assessment, the scripts were played back to each participant one at a time in the following order: neutral, stressful, and trauma. They listened to the scripts and were instructed to imagine each event as vividly as possible, as though they were actually reexperiencing it.
Physiological Responsivity
Physiological variables were collected 2 minutes before (baseline phase), during (imagery phase of 2 minutes), and 4 minutes after (recovery phase) the playing of each script. Physiological variables consisted of heart rate (HR), systolic blood pressure (SBP), and diastolic blood pressure (DBP) measured with a Boso Carat (oscillometric) device on the left arm.
Psychological Responsivity
All participants completed a State-Trait Anxiety Inventory (STAI) (37) after listening to each script to determine their subjective psychological anxiety score in relation to that script.
Psychotherapy
Details of our psychotherapeutic procedure have been published previously (38,39) and are described here only briefly. The treatment described in the manual consisted of Brief Eclectic Psychotherapy (BEP) with 16 weekly individual sessions of 45 to 60 minutes each. As Brewin (40) has noted, psychotherapy for PTSD generally requires two elements: 1) detailed, repeated exposure to traumatic information; and 2) modification of maladaptive beliefs about events, behaviors, or symptoms. BEP addresses both these elements.
The psychotherapists working on the present study were clinically experienced psychiatry residents, and they received supervision every 2 weeks from two senior psychiatrists (I.VE.C. and B.P.R.G.). All sessions were audiotaped. A rating system covering the elements of BEP described here was developed to analyze treatment integrity. Treatment integrity was approximately 75%, with a kappa >0.81, indicating good adherence to the protocol.
Statistical Analysis
Statistical analysis was performed with SPSS 11.0 for Windows (SPSS, Chicago, IL). Demographic and clinical PTSD variables, and STAI scores were analyzed to compare the two PTSD groups and the traumatized control group, as well as the treatment and waitlist subgroups, using chi-squared tests for categorical variables and multivariate analysis of variance (MANOVA) for continuous variables. In connection to each script, three values were calculated for each physiological variable: 1) the baseline mean, 2) the mean response score, and 3) the mean recovery score. Response scores were calculated by subtracting the baseline mean from the imagery phase mean and recovery scores by subtracting the recovery phase mean from the imagery phase mean. Multivariate analysis of variance was used to compare the baseline means, response scores, and recovery scores on each script for the civilian and police PTSD groups and the control group, as well as for the civilian treatment and waitlist subgroups, analyzing HR, SBP, and DBP simultaneously as dependent variables. At the posttest stage, after the treatment group received psychotherapy, the effects of the therapy were measured by again analyzing the PTSD clinical variables, STAI scores, and physiological variables with MANOVA using the pretest measurements as covariates. Correlations of the posttreatment PTSD total scores with the HR response and STAI scores in reaction to especially the trauma scripts were obtained using Pearson product-moment correlation coefficients (two-tailed) with a significance level of 5% (
= 0.05).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
|
No statistical differences were found among the two PTSD groups and control group in terms of age, gender, education, or smoking. Consistent with their diagnoses, both PTSD groups had higher PTSD total scores as well as higher reexperiencing, avoidance, and hyperarousal scores than the controls. No statistical differences were found for comorbid mild major depression (see Table 1).
Pretest Psychophysiological Responsivity in Civilians With Posttraumatic Stress Disorder as Compared With Traumatized Controls
Table 2 shows the psychophysiological variables of the two PTSD groups and the control group.
|
Psychological Responsivity
The STAI (subjective anxiety) scores in reaction to each of the three scripts at pretest were significantly higher in the civilians with PTSD than in the traumatized control subjects (multivariate tests: F[1] = 23.81, p < .001; tests of between-subjects effects: neutral: F[1] = 27.17, p < .001; stressful: F[1] = 36.05, p < .001; trauma: F[1] = 53.73, p < .001).
Physiological Responsivity
The multivariate tests results for the physiological variables were F(1) = 2.39, p = .058). The tests of between-subjects effects were as follows: In terms of baseline (prescript) means on the physiological variables, only the HR before the trauma script (F[1] = 4.69, p = .037) was significantly higher in the PTSD group than in the controls. In terms of response scores, the increase in HR during the trauma script (F[1] = 5.64, p = .023) was also significantly greater in the PTSD group, but not during the neutral and stressful scripts. In terms of recovery scores, there were no significant differences on physiological variables between PTSD and control subjects after any of the scripts.
Although three PTSD subjects had a secondary first episode of mild major depression as compared with none of the control subjects, MANOVA revealed no significant differences within the PTSD group between depressed and nondepressed subjects in terms of physiological responsivity.
Psychophysiological Responsivity in Police Officers With Posttraumatic Stress Disorder as Compared With Traumatized Controls
Psychological Responsivity
STAI scores in reaction to each of the three scripts were likewise significantly higher for the police officers with PTSD than for the control subjects (multivariate tests: F[1] = 5.51, p = .005; tests of between-subjects effects: neutral: F[1] = 9.45, p = .005; stressful: F[1] = 8.67, p = .006; trauma: F[1] = 16.38, p < .001) (Table 2).
Physiological Responsivity
The multivariate tests results for the physiological variables were F(1) = 0.70, p = .75. The tests of between-subjects effects were as follows: In terms of the baseline means on the physiological variables, no significant differences were apparent between PTSD police officers and non-PTSD controls before any of the scripts. Response scores for HR (F[1] = 5.44, p = .027), for SBP (F[1] = 4.46, p = .044), and for DBP (F[1] = 5.86, p = .022) were significantly higher in the PTSD group in reaction to the trauma script than for the controls, whereas no significant differences appeared on the neutral and stressful scripts. In terms of recovery scores, no significant physiological differences were found between PTSD and control groups after any of the scripts.
Here, too, three PTSD subjects had secondary first-episode mild major depression as compared with none of the controls, but MANOVA again revealed no significant differences between depressed and nondepressed subjects in the PTSD group in terms of physiological responsivity.
Treatment Effects in the Civilian Outpatients
Demographic and Clinical Variables in the Treatment and Waitlist Groups
Our statistical analysis of the treatment effects was carried out on the net PTSD trial sample (N = 20) comparing the treatment subgroup (N = 9) to the waitlisted subgroup (N = 11). Figure 1 shows a flow diagram of patients with PTSD who were included, randomly assigned, received treatment, and analyzed for the primary outcome.
|
At the pretest stage, no statistical differences were evident between the two subgroups on any of the demographic or clinical variables (not in the table).
After treatment, the condition of the patients in the treatment group had improved on all PTSD clinical variables (multivariate tests: F[1] = 4.79, p = 0.018; tests of between-subjects effects: PTSD total score: F[1] = 13.80, p = .002; reexperiencing: F[1] = 12.33, p = .003; avoidance: F[1] = 7.43, p = .016; hyperarousal: F[1] = 16.47, p = .001) in comparison to the waitlist group.
Psychophysiological Changes After the Treatment Period in the Treatment and Waitlist Groups
Table 3 summarizes the pretest–posttest comparison of the two groups in terms of psychophysiological responses to the scripts (for purposes of clarity, only the STAI and the physiological response scores are shown; statistics refer to the entire set of variables).
|
Psychological Responsivity
At the pretest stage, no significant differences between the treatment and waitlist groups were evident in terms of STAI scores or physiological measures. After treatment, the STAI scores were significantly lower in the treatment group after the stressful and trauma scripts (multivariate tests: F[1] = 1.66, p = .22; tests of between-subjects effects: stressful: F[1] = 4.43, p = .053; trauma: F[1] = 4.82, p = .044).
Physiological Responsivity
The multivariate tests results for the physiological variables after treatment were F(1) = 3.94, p = .38. The tests of between-subjects effects were as follows: Significant reductions relative to the waitlist group occurred in DBP baseline means (stressful: F[1] = 4.59, p = .046). Especially for the trauma scripts, significant reductions relative to the waitlist group also occurred in SBP baseline mean (trauma: F[1] = 5.43, p = .032), mean HR response score (trauma: F[1] = 4.61, p = .046), and mean DBP recovery score (trauma: F[1] = 4.78, p = .042). Figure 2 shows the HR responses to the trauma scripts in the two PTSD groups and the non-PTSD control group at pretest, and the effects of psychotherapy at posttest in the treatment group as compared with the waitlist group.
|
Correlations Between Heart Rate Responses and Posttraumatic Stress Disorder Severity at the Posttest Stage
After the treatment group had received psychotherapy, the PTSD total scores in the combined treatment and waitlist groups were positively and linearly correlated with the HR responses to the trauma script (r = 0.56, p = .01), as well as with the STAI scores after the stressful and trauma scripts (stressful: r = 0.67, p = .001; trauma: r = 0.79, p < .001). No significant correlations were found between PTSD total scores and HR responses to the neutral and stressful scripts.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
Furthermore, we showed that the elevated heart rate response to the trauma script was normalized after successful psychotherapy in the civilian sample. Because all patients in this randomized clinical trial heard the same individual scripts at the pretest and posttest stages, habituation to the scripts probably developed in both the treatment and the waitlist groups. Despite this, HR response to the trauma script significantly decreased in the treatment group relative to the waitlist group. Subjective anxiety scores, which were significantly elevated in PTSD subjects but, unlike HR response, did not differ between scripts at pretest, were now significantly lower on the stressful and trauma scripts in the treatment group. At posttest stage, HR responsivity to the trauma scripts was positively and linearly correlated with the severity of PTSD symptomatology. Several other studies have likewise reported diminished psychophysiological responses to trauma-related imagery after treatment (17,24–27,28–31). In a case report on combat-related PTSD, Keane and Kaloupek (26) found that HR response during recall of the traumatic event was lower after treatment with imaginal flooding than before treatment. Boudewijns and Hyer (24) reported that better adjustment 3 months posttreatment was associated with reduced skin conductance responses during trauma-related imagery. Shalev et al. (27) described in a case report three patients with PTSD treated with systematic desensitization and found a reduction in psychophysiological responses to trauma-related imagery. Cohen et al. (25), although not recording responses to trauma imagery, noted that HR variability at rest was normalized in patients with PTSD treated with a selective serotonin reuptake inhibitor. Because most of these studies were noncontrolled case reports, habituation to the trauma imagery cannot be excluded as an explanation for the reduced heart responsivity after treatment. Our study lends support to their findings and uses a waitlist control group to overcome such habituation effects. Five studies were randomized trials in which psychophysiological reactions were assessed (28–31). Renfrey and Spates (28) investigated three groups who were treated with eye movement desensitization and reprocessing (EMDR) and two modified forms of EMDR, and found an overall effect of decreased HR after treatment. Rogers et al. (30) found a trend toward decreased HR reactivity in both treatment groups in which a single EMDR session was compared with exposure. Wilson et al. (31) described three groups who were treated with EMDR and two modified forms of EMDR and found, in the regular EMDR, an overall slowed HR and the galvanic skin response decreased in a clear "relaxation response." Blanchard et al. (17) showed that cardiovascular reactivity to trauma cues diminished with successful psychosocial treatment of PTSD. Carlson et al. (29) compared EMDR, biofeedback-assisted relaxation, and routine clinical care with each other, and reported that the significant overall pre/post effects obtained in electromyographic (EMG) variables may have reflected a general habituation of arousal responses that was independent of treatment. In our randomized, controlled trial, successful psychotherapy normalized HR responses to trauma imagery, and this finding could not be explained by a habituation effect.
Even when few studies had investigated the effects of treatment for PTSD using psychophysiological outcome measures, hypotheses were formulated on how different treatments might affect those biological correlates. Foa and Kozak (43) suggested that effective exposure-based treatments may intervene to modify a fear network in which associative learning has combined psychophysiological and neuroendocrine responses with situational cues of the traumatic event. Pitman et al. (44) proposed that effective medication may modulate subcortical activation and/or increase the capacity of cortical structures to do so. Our BEP psychotherapy of 16 weekly sessions involved repeated detailed exposure to the traumatic event and subsequent modification of maladaptive beliefs about it, the elements described by Brewin (40). Exposure therapy processes the trauma to modify the affective and cognitive valences associated with the traumatic memories (43,45). This emotional adaptation should lead to decreased autonomic arousal during exposure (43). Our study seems to confirm this.
We should comment on several limitations of our study. Organizational considerations prevented us from examining the PTSD and non-PTSD police officers at the posttest stage. As a consequence, our trial sample of civilians was relatively small, but we nevertheless obtained significant effects in the randomized, controlled trial. Because we did not play the three scripts to the subjects in a pseudorandom order, they may have learned to anticipate them. Had we varied the sequence, however, the startle response induced by the trauma script might have seriously confounded the results. Although it would have also been informative to record physiological measures like facial electromyogram and skin conductance in addition to HR and BP, increased HR has been the most consistent finding in the PTSD literature so far (3,8,46).
Future studies could use psychophysiological assessment to investigate vulnerability factors for the development of PTSD. Such procedures could have practical implications for personnel selection, especially in such high-risk professions as police work. Other important issues that need investigating are the effects of different types of treatments on the biological correlates identified in PTSD and the role that psychophysiological assessment could play in gauging vulnerability to relapse.
The authors thank Ineke Vrijlandt, MD, and Gré Westerveld, MA, both of the Centre for Psychological Trauma, Department of Psychiatry, the former for supervising the treatment and the latter for randomizing the patients.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
This study was presented at Annual Meetings of the International Society for Traumatic Stress Studies (Chicago, November 1, 2003, and New Orleans, November 16, 2004) and at the 12th World Congress of Psychophysiology (Thessaloniki, September 22, 2004).
DOI:10.1097/01.psy.0000188566.35902.e7
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONNOTESREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
