Psychosomatics 42:133-140, April 2001
© 2001 The Academy of Psychosomatic Medicine
Psychophysiologic Assessment of Posttraumatic Stress Disorder in Breast Cancer Patients
Roger K. Pitman, M.D.,
Douglas M. Lanes, M.D.,
Stephanie K. Williston, Ph.D.,
Jeanette L. Guillaume, M.A.,
Linda J. Metzger, Ph.D.,
Gerald M. Gehr, M.D., and
Scott P. Orr, Ph.D.
Received March 3, 2000; revised June 21, 2000; accepted September 6, 2000. From VA Research Service, Manchester, New Hampshire and the Department of Psychiatry, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts. Address correspondence and reprint requests to Dr. Pitman at PTSD Research Laboratory, MGH-East, Building 149, 13th Street, Charlestown, MA 02129; E-mail roger_pitman{at}hms.harvard.edu
|
ABSTRACT
|
The authors performed psychodiagnostic, psychometric, and psychophysiologic evaluations on 37 patients referred by local surgeons approximately 2 years after tissue diagnosis of Stage I to III breast cancer. The Clinician-Administered Posttraumatic Stress Disorder (PTSD) Scale (CAPS) was used to classify patients into the following groups: "Current PTSD" (n=5) "Past PTSD" (n=7), and "Never had PTSD" (n=25). Individualized "scripts" portraying personal life events were tape recorded and played back to the patients in the laboratory. Current PTSD patients showed significantly higher heart rate, skin conductance, and corrugator electromyogram responses during imagery of their personal breast cancer experiences than Past and Never patients. Physiologic responses were significantly and positively correlated with CAPS scores. These results provide psychophysiologic support for the proposition that a diagnosis of with a life-threatening illness can cause PTSD.
Key Words: Posttraumatic Stress Disorder • Breast Cancer • Stress Disorder
|
INTRODUCTION
|
The DSM-IV cites "being diagnosed with a life-threatening illness" as capable of causing posttraumatic stress disorder (PTSD).1 Although it is clear from an abundance of research during the past 2 decades that extreme, acutely stressful events such as military combat, sexual assault, and accidents can and do produce PTSD, the ability of less acute stressors to do so has not received as much research attention. Terr2 suggested that "the traumatic experience must be...surprising, unanticipated, and piercingly intense" to produce PTSD. Although a diagnosis of breast cancer may be as dangerous as a rapist's knife, the threat posed by cancer is less immediate, palpable, and "piercingly intense."
Laboratory measurement of physiologic responses during recollection of traumatic events adds objectivity to the evaluation of PTSD and serves as a useful technique for testing the validity of this diagnosis. Laboratory measurement of physiologic responses also provides a means of testing what kinds of stressors are capable of causing PTSD. "Physiological reactivity on exposure to internal or external cues that symbolize or resemble an aspect of the traumatic event" is now a DSM-IV PTSD diagnostic criterion.1 Heightened physiologic responses during mental imagery of the traumatic event have been reported in military combatants,3–5 victims of terrorist attacks and physical assaults,6 child abuse,7 and motor vehicle accidents.8
Although PTSD resulting from a diagnosis of breast cancer has been reported in descriptive and epidemiologic studies,9–11 this research has relied on self-report for diagnosing PTSD. The question as to whether persons who are diagnosed with a life-threatening illness, such as breast cancer, can develop psychophysiologically responsive PTSD has not been studied. In this study, we evaluated recently diagnosed breast cancer patients for PTSD using a structured psychodiagnostic interview, psychometric testing, and a script-driven imagery technique, which has proven useful in characterizing physiologic responsivity in PTSD stemming from other traumatic events.3–7
|
METHODS
|
Recruitment and Referral
Using a tumor registry, we identified female breast cancer patients that had been diagnosed during 1994–1996 by four local surgeons in private practice. University and local institutional review board (IRB) approval was obtained for the release of identifying information from the registry back to the surgeons' offices, but the local IRB did not permit the investigators to promote recruitment by having access to patients' identities. Local surgeons identified suitable patients from their practices and asked the patients whether they were willing to talk with the researchers about participating in a research study on the emotional reactions of persons with breast cancer. The names and telephone numbers of willing patients were given to the researchers. The inclusion criteria for participation were the following: age 18 or older; diagnosed (via tissue pathology) with breast cancer greater than Stage 0 and less than Stage IV during 1994–1996; and finished with initial treatment; and free of recurrence. We excluded referred patients from participation if they had a current or past organic mental, schizophrenic or other psychotic, bipolar I, or substance dependence disorder. No patients were pregnant.
Psychodiagnostics and Psychometrics
Referred patients were interviewed over the telephone by a psychologist who administered a modified version of the PTSD Checklist (PCL).12 Patients were then invited for a personal interview by a psychiatrist (DML) who administered the Clinician-Administered PTSD Scale-diagnostic version (CAPS),13 with regard to PTSD, and the Structured Clinical Interview for DSM-IV (SCID),14 with regard to other Axis I diagnoses. Patients also completed the civilian version of the Mississippi PTSD Scale,15 the Symptom Checklist 90-Revised (SCL-90-R),16 the State-Trait Anxiety Inventory (STAI),17 and the Marlowe-Crowne Social Desirability Scale (MCSDS).18
Written informed consent was obtained after a full explanation of the study's procedures. For patients interviewed only by telephone, oral informed consent was obtained.
Script Preparation
Patients who completed the CAPS were invited to participate in laboratory testing. The script-driven imagery techniques employed in our study have been presented in detail elsewhere3–7 and are only summarized here. With the assistance of a psychologist (SKW), patients composed narratives portraying five past personal experiences, including their two most stressful experiences with breast cancer, their most stressful other life experience, and a positive and neutral experience. Narratives included self-reported visceral and muscular reactions that accompanied each event. Patients completed the Impact of Event Scale-Revised (IES-R)19 pertaining to each of their breast cancer and other stressful life events.
The psychologist condensed each narrative into a 30-second "script" and recorded it in a neutral voice for playback in the laboratory. The study also employed six standard scripts: two neutral, two fear (public speaking, trapped in a sauna), one positive (sandy beach), and one action (riding a bicycle).
Physiologic Recording
A Coulbourn Modular Instrument System was used to measure heart rate (HR), skin conductance (SC), and electromyograms (EMGs) of the left lateral frontalis (frnt.), and corrugator (corr.) facial muscles. We input an amplified electrocardiogram to a tachometer that provided a voltage output measuring interbeat interval, which was transformed to HR. SC level was obtained through Ag/AgCl electrodes filled with isotonic paste placed on the nondominant hypothenar surface using a constant-voltage technique.20 EMGs were obtained through Ag/AgCl electrodes placed according to published specifications.21 The amplified EMG signals were integrated using a 300-msec. time constant.
Laboratory Procedure
Approximately a week after the script preparation session, patients were familiarized with the laboratory, had electrodes attached, and listened to a 3-min. relaxation instruction tape. Each script presentation consisted of four sequential 30-sec. periods. After a baseline recording period preceding each script (Baseline period), patients were instructed to listen to the script as it was played and imagine it as vividly as possible, as though it were actually occurring (Read period). When the script ended, patients continued to imagine the experience (Imagery period) until signaled by a tone and then relaxed (Recovery period) until signaled by another tone. At this time, patients completed self-reports of arousal, pleasantness, image vividness, and control, and seven discrete emotions (happiness, sadness, fear, anger, disgust, guilt, and surprise)22 on 12-point Likert-type scales. The Baseline period for the next script was initiated after a rest period of 1 min or after HR had returned to within 5% of its value during the previous Baseline period, whichever was longer. This rarely exceeded 3 min.
A personal computer controlled the recorded script presentations, collected patients' emotional ratings, and sampled and stored the digitized physiologic signals at 2 Hz. A standard neutral script was presented first, followed by two blocks of five scripts each. Each block contained one of the patients' two personal breast cancer-related scripts, a neutral script, a positive script, either the personal other stressful script or a standard fear script, and either the action or other standard fear script. No two of the personal breast cancer, other stressful, or fear scripts were presented sequentially. Otherwise, order of script presentation was randomized.
Data Reduction and Analysis
A response score was calculated for each physiologic dependent variable for each script by subtracting the preceding Baseline period mean from the Imagery period mean. Physiologic responses during the two breast cancer scripts were averaged before analysis. Group data were examined via multivariate analyses of variance (MANOVAs) employing the four physiologic responses simultaneously, and analyses of variance (ANOVAs) and analyses of covariance (ANCOVAs) for each of the four physiologic responses separately. ANOVAs with F significant at P<0.05 were followed by Tukey's pairwise comparisons (significance criterion P<0.05.)
The relationship between total CAPS score and physiologic responses was examined via Pearson product-moment zero-order and partial correlations.
An a priori discriminant function was used to classify the present patients as physiologic "responders" or "nonresponders" during their personal breast cancer imagery (insufficient previous data were available for EMG-corr). This discriminant function was derived from the HR, SC, and EMG-frnt responses during personal traumatic imagery of 92 previously studied individuals with current PTSD and 86 individuals who never had PTSD.1–6
|
RESULTS
|
We identified 187 patient candidates from the tumor registry. Of these, 100 were not referred for participation by their surgeon for the following reasons: deceased (n=9), no longer being seen in surgeon's practice (n=52), met study exclusion criteria (n=30), and refused participation (n=9). All of the 87 patients referred for participation completed the telephone PCL. Of these, 50 agreed to come in for the CAPS.
According to the PCL, of the 37 patients who only underwent the telephone interview, 1 patient met DSM-IV criteria for Current PTSD and 5 for Past PTSD related to their experiences with breast cancer. Of the 50 patients who were interviewed in person, 7 met criteria for Current PTSD and 8 for Past PTSD according to the CAPS. The mean±standard deviation time elapsed from tissue diagnosis to interview was 20.5±7.6 (range from 11.2 to 44.2 months). Because agreement was good between the PCL and CAPS (weighted  =0.57)23 in the patients who were administered both instruments, PCL-classified patients were added to CAPS-classified patients to yield estimated incidences of current breast cancer-related PTSD of 9% (8/ 87) and past breast cancer-related PTSD of 15% (13/87).
Group Compositions for Psychophysiologic Testing
Thirty-seven patients who completed the CAPS agreed to participate in the laboratory testing. According to the CAPS, 5 patients had Current PTSD, 7 had Past PTSD, and 25 Never had PTSD. According to the SCID, 4 of the 5 Current patients (80%) had one or more current comorbid mental disorders (including 4 major depressive, 1 panic, 2 social phobic, 2 specific phobic, and 1 body dysmorphic disorders). Three of the 7 Past patients (43%) had a current comorbid mental disorder (including 1 major depressive, 1 panic, and 1 social phobic disorders). Seven of the 25 Never had PTSD patients (28%) had one or more current comorbid mental disorders (including 1 social phobic, 2 specific phobic, 1 obsessive compulsive, 1 generalized anxiety, 1 anxiety not otherwise specified, and 2 adjustment disorders). Some patients had more than one comorbid disorder. Three of the 5 Current (60%), 4 of the 7 Past (57%), and 7 of the 25 Never patients (28%) were using one or more psychotropic or autonomically active substances of various classes (including serotonergic and nonserotonergic antidepressants, anticonvulsants, opioids, anti-hypertensives, and decongestants).
Group mean demographic and psychometric data are presented in Table 1. There were no significant group differences in age, education, time elapsed since tissue diagnosis of breast cancer, or MSCDS score. The Current group scored higher than the Past and Never groups on all measures of PTSD and general psychopathology, including IES-R scores related to both their personal breast cancer and their personal other stressful experiences.
Group Physiologic Responses
Group mean resting physiologic levels and responses are presented in Table 2. There were no significant group effects for resting levels before the script-driven imagery. For physiologic responses, the only scripts for which overall MANOVA yielded a significant group effect were the personal breast cancer scripts (F=2.5, df=8,62, P=0.02) and the personal other stressful scripts (F=2.6, df=8,62, P=0.01). Hence, only the group mean responses during imagery of those scripts are included in Table 2. On ANOVA, Current patients showed significantly larger HR and EMG-corr responses than both Past and Never patients and significantly larger SC responses than Never patients, during personal breast cancer-related imagery. These data are illustrated in Figure 1. Current patients also showed significantly larger EMG-corr responses than Past and Never patients during their personal other stressful imagery.
To control for general responsiveness during imagery of personal stressful events, ANCOVAs were performed for each of the four physiologic response variables during personal breast cancer-related imagery, with the respective response during personal other stressful imagery as the covariate. A significant ANCOVA Group effect was maintained for SC (F=3.6, df=2,33, P=0.04).
Correlations with CAPS Score
Pearson product-moment (zero-order) correlations across all patients between current total CAPS score and physiologic responses during personal breast cancer imagery were significant for all physiologic variables (HR: r=0.38, df=35, P=0.02; SC: r=0.34, df=35, P=0.04; EMG-frnt: r=0.34, df=35, P<0.05; and EMG-corr: r=0.55, df=35, P<0.001). The partial correlation adjusting for the respective response during personal other stressful imagery remained significant for HR (r=0.40, df=34, P=0.02); nearly significant for SC (r=0.31, df=34, P=0.07); but not significant for EMG-frnt (r=0.27, df=34, P=0.12); or EMG-corr (r=0.27, df=34, P=0.12).
Discriminant Function
The a priori physiologic discriminant function classified 2 of the 5 Current (40%), 2 of the 7 Past (29%), and 6 of the 25 Never patients (25%) as physiologic responders ( 2 NS).
Self-Reported Emotional Responses
Group mean self-reported emotional responses during the personal breast cancer and other stressful imagery are presented in Table 3. During the personal breast cancer imagery, Current patients reported feeling less in control than Never patients. Past patients also reported feeling more guilty than Never patients. There were no significant group effects during the other stressful imagery.
|
DISCUSSION
|
Our present findings of significantly elevated physiologic responses during script-driven imagery of personal breast cancer experiences in patients with current PTSD and significant positive correlations between physiologic responses and current CAPS scores provide psychophysiologic support for the proposition that being diagnosed with a life-threatening illness, in this case breast cancer, can be a stressor sufficient to result in PTSD. The Current PTSD patients' mean physiologic responses were elevated to a degree comparable to those of patients with PTSD from other traumatic events studied using the same technique.3–7
The small number of patients studied precluded the exclusion of patients who had other psychiatric comorbidity or who were taking potentially confounding psychotropic or autonomically active substances. Recruiting "pure" PTSD subjects has proven a difficult task in research of this nature. The present patients' comorbidity profiles are similar to those reported in most published studies. However, it is reasonable to assume that the nature of the experimental task (i.e., imagery of personal traumatic events) means that group differences with regard to PTSD will override differences in other comorbidities. The findings that skin conductance response during personal breast cancer-related imagery remained significantly higher in patients with current PTSD after adjusting for SC response during other personal stressful imagery and that the partial correlation between CAPS score and heart rate response during personal breast cancer imagery remained significantly positive after adjusting for heart rate response during other personal stressful imagery support a specific relationship between physiologic responding during recollection of the traumatic event and the PTSD diagnosis. One would not have predicted this result if general anxiety or affective disturbance was solely responsible for the increased physiologic responding observed in the current PTSD patients.
We also doubt that medication usage can explain the elevated physiologic responding during personal breast cancer imagery observed in the current PTSD group, for the following reasons. First, the data reduction technique of subtracting Baseline from Imagery response eliminates any contribution of possible drug influences on tonic physiologic state. Second, the drugs patients were taking would likely dampen, rather than increase, phasic responsivity. Third, using the respective physiologic response during personal, other (non-breast-cancer-related) imagery as a covariate in the analyses helped to control for drug effects on responsivity. Fourth, in unpublished work, we have frequently noted robust physiologic responses during traumatic mental imagery in medicated PTSD subjects.
Our interview results estimate the incidence of breast cancer-related PTSD at 9% current and an additional 15% past during the approximate 2 years following tissue diagnosis. These estimates are somewhat higher than those obtained in previous studies.9–11 Possible explanations for the higher rates found here are our inclusion of later Stage (i.e., Stage III) patients and our use of the (more thorough) CAPS rather than the SCID or PCL to diagnose PTSD. However, epidemiologic inferences from the present data are limited by failure to recruit a substantial portion of patient candidates.
Another limitation of our present study is that the PTSD group's larger mean physiologic responses during breast cancer imagery, which was the study's main finding, was heavily influenced by the responses of two patients. The overall specificity of 40% (2 of 5 Current patients) was lower than the 60%–70% specificities we have observed in script-driven imagery studies of other trauma-exposed populations.3–7 This raises the possibility that fewer breast cancer patients diagnosed with PTSD may be physiologically responsive. However, the sample size is far too small to justify such a conclusion. Further psychophysiologic studies that attempt to place this finding on a broader foundation are indicated. Such studies may need to find ways of over-recruiting breast cancer patients with current PTSD in order to obtain larger sample sizes.
|
ACKNOWLEDGMENTS
|
The authors acknowledge the assistance of the referring surgeons, Drs. William Clutterbuck, Edward Dalton, Amir Khazei, Richard Knab, and Patrick Mahon, their office staffs, as well as Catholic Medical Center, Elliot Hospital, New Hampshire Oncology, and the New Hampshire Breast Cancer Coalition. This research was supported by U.S. Army Grant #DAMD17–94-J-4365.
|
REFERENCES
|
-
American Psychiatric Association: Diagnostic and Statistical Manual for Mental Disorders, 4th Edition, Washington, D.C., American Psychiatric Association, 1994
-
Terr LC: Psychic trauma in children and adolescents. Psychiatr Clin North Am 1985; 8:815–835[Medline]
-
Pitman RK, Orr SP, Forgue DF, et al: Psychophysiology of PTSD imagery in Vietnam combat veterans. Arch Gen Psychiatry 1987; 44:970–975[Abstract]
-
Pitman RK, Orr SP, Forgue DF, et al: Psychophysiologic responses to combat imagery of Vietnam veterans with posttraumatic stress disorder versus other anxiety disorders. J Abnorm Psychol 1990; 99:49–54[CrossRef][Medline]
-
Orr SB, Pitman RK, Lasko NB, et al: Psychophysiologic assessment of posttraumatic stress disorder imagery in World War II and Korean combat veterans. J Abnorm Psychol 1993;102:152–159
-
Shalev AY, Orr SP, Pitman RK: Psychophysiologic assessment of traumatic imagery in Israeli civilian posttraumatic stress disorder patients. Am J Psychiatry 1993; 150:620–624[Abstract/Free Full Text]
-
Orr SP, Lasko NB, Metzger LJ, et al: Psychophysiologic assessment of women with PTSD resulting from childhood sexual abuse. J Consult Clin Psychol 1998; 66:906–913[CrossRef][Medline]
-
Blanchard EB, Hickling EJ: After the Crash: Assessment and Treatment of Motor Vehicle Accident Survivors. Washington, D.C., American Psychological Association, 1997
-
Alter CL, Pelcovitz D, Axelrod A, et al: Identification of PTSD in cancer survivors. Psychosomatics 1996; 37:137–143[Abstract/Free Full Text]
-
Andrykowski MA, Cordova MJ, Studts JL, et al: Posttraumatic stress disorder after treatment for breast cancer: prevalence of diagnosis and use of the PTSD Checklist-Civilian Version (PCL-C) as a screening instrument. J Consult Clin Psychol 1998;66:586–590
-
Green BL, Rowland JH, Krupnick JL, et al: Prevalence of posttraumatic stress disorder in women with breast cancer. Psychosomatics 1998; 39:102–111[Abstract/Free Full Text]
-
Weathers FW, Litz BT, Herman DS, et al: The PTSD Checklist (PCL): reliability, validity, and diagnostic utility. Paper presented at the 1993 Annual Meeting of the International Society for Traumatic Stress Studies, San Antonio, TX
-
Blake DD, Weathers FW, Nagy LN, et al: The development of a clinician-administered PTSD Scale. J Trauma Stress 1995; 8:75–90[CrossRef][Medline]
-
First MB, Spitzer RL, Gibbon M, et al: Structured Clinical Interview for Axis I DSM-IV Disorders. Biometrics Research Department, New York, 1994
-
Keane TM, Caddell JM, Taylor KL: Mississippi Scale for Combat-Related Posttraumatic Stress Disorder: three studies in reliability and validity. J Consult Clin Psychol 1998; 56:85–90
-
Derogatis LR: SCL-90-R: Administration, Scoring and Procedures Manual-II for the Revised Version. Clinical Psychometric Research, Towson, MD, 1983
-
Spielberger CD, Gorsuch RL, Lushene RE: Manual for the State-Trait Anxiety Inventory (Self-Evaluation Questionnaire). Palo Alto, Consulting Psychologists Press, 1990
-
Crowne DP, Marlowe DA: A new scale for social desirability independent of psychopathology. J Consult Psychol 1960; 24:349–354[CrossRef][Medline]
-
Weiss DS, Marmar CR: The Impact of Event Scale-Revised, in Assessing Psychological Trauma and PTSD, edited by Wilson JP, Keane TM. New York, Guilford Press, 1997, p 399–411
-
Fowles DC, Christie MJ, Edelberg R, et al: Publication recommendations for electrodermal measurements. Psychophysiology 1981; 18:232–239[Medline]
-
Fridlund AJ, Cacioppo JT: Guidelines for human electromyographic research. Psychophysiology 1986; 23:567–589[Medline]
-
Izard CE: Patterns of Emotion. New York, Academic Press, 1972
-
Spitzer RL, Cohen J, Fleiss JL, et al: Quantification of agreement in psychiatric diagnosis. Arch Gen Psychiatry 1967;17:83–87
This article has been cited by other articles:
|
|
|
|
 
S. K. Smith, S. Zimmerman, C. S. Williams, J. S. Preisser, and E. C. Clipp
Post-Traumatic Stress Outcomes in Non-Hodgkin's Lymphoma Survivors
J. Clin. Oncol.,
February 20, 2008;
26(6):
934 - 941.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
R. T. L. Lindauer, E. P. M. van Meijel, M. Jalink, M. Olff, I. V. E. Carlier, and B. P. R. Gersons
Heart Rate Responsivity to Script-Driven Imagery in Posttraumatic Stress Disorder: Specificity of Response and Effects of Psychotherapy
Psychosom Med,
January 1, 2006;
68(1):
33 - 40.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
T. Akechi, T. Okuyama, Y. Sugawara, T. Nakano, Y. Shima, and Y. Uchitomi
Major Depression, Adjustment Disorders, and Post-Traumatic Stress Disorder in Terminally Ill Cancer Patients: Associated and Predictive Factors
J. Clin. Oncol.,
May 15, 2004;
22(10):
1957 - 1965.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
S. C. Palmer, A. Kagee, J. C. Coyne, and A. DeMichele
Experience of Trauma, Distress, and Posttraumatic Stress Disorder Among Breast Cancer Patients
Psychosom Med,
March 1, 2004;
66(2):
258 - 264.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
M. Gurevich, G. M. Devins, and G. M. Rodin
Stress Response Syndromes and Cancer: Conceptual and Assessment Issues
Psychosomatics,
August 1, 2002;
43(4):
259 - 281.
[Abstract]
[Full Text]
[PDF]
|
|
|
Get information about faster international access.
Privacy Policy
Copyright © 2001
Academy of Psychosomatic Medicine.
All rights reserved.
Home
| Search
| Current Issue
| Past Issues
| Subscribe
| All APPI Journals
| Help
| Contact Us
|
TOP
ABSTRACT
INTRODUCTION
