Journal of Traumatic Stress February 2014, 27, 108–111
BRIEF REPORT
Externalizing and Internalizing Subtypes of Posttraumatic Psychopathology and Anger Expression Diane T. Castillo,1,2 Jeremy S. Joseph,1,5 Andra T. Tharp,6 Janet C’de Baca,1 Lorraine M. Torres-Sena,1 Clifford Qualls,4 and Mark W. Miller3 1 New Mexico VA Health Care System, Albuquerque, New Mexico, USA Department of Psychiatry, University of New Mexico, Albuquerque, New Mexico, USA 3 National Center for PTSD, VA Boston Healthcare System, Boston, Massachusetts, USA 4 Department of Statistics, University of New Mexico, Albuquerque, New Mexico, USA 5 University of Wyoming 6 Baylor College of Medicine, Houston, Texas, USA
2
Subtypes of posttraumatic psychopathology were replicated and extended in 254 female veterans with posttraumatic stress disorder (PTSD). Cluster analyses on Minnesota Multiphasic Personality Inventory-2 and Personality Psychopathology Five scales (Harkness, McNulty, & Ben-Porath, 1995) yielded internalizing and externalizing psychopathology dimensions, with a third low psychopathology group (simple PTSD). Externalizers were higher than the internalizers and the simple PTSD groups on the antisocial, substance, and aggression scales; internalizers were higher on depression and anxiety scales. Further validation included an independent measure of psychopathology to examine anger (Buss-Durkee Hostility Inventory, [BDHI]; Buss & Durkee, 1957). Externalizers were higher on extreme behavioral anger scales (assault and verbal hostility); and externalizers and internalizers were higher than the simple PTSD subjects on other anger scales. Positive correlations between the BDHI scales and the PTSD symptom of “irritability and anger outbursts” were found across scales in the total sample (range: r = .19–.36), on the assault scale in externalizers (r = .59), and the verbal hostility scale in both internalizers (r = .30) and simple PTSD (r = .37) groups, suggesting the broad utility of the symptom in the diagnosis. The results demonstrate the generalizability of the internalizing/externalizing typology to the female veteran population and highlight clinically relevant distinctions in anger expression within PTSD.
Considerable variability exists within the manifestations of psychiatric disorders, which researchers have accounted for through the examination of symptom clusters categorized within the externalizing and internalizing psychopathology dimensions (e.g., Kendler, Prescott, Myers, & Neale, 2003; Kessler, Chiu, Demler, & Walters, 2005). Across psychiatric diagnoses, results support these two latent dimensions in which externalizers are characterized by disconstraint, aggressiveness, and high-negative emotion, whereas internalizers are characterized by high-negative emotion and low-positive emotion. Furthermore, externalizers are more likely to endorse substance abuse, symptoms of antisocial personality disorder, and conduct disorder whereas internalizers are more likely to endorse comorbidity for anxiety and depression. Recent studies have
found this personality typology relevant to the diagnosis of posttraumatic stress disorder (PTSD), with an added third, lowpathology group classified as simple PTSD (absence of comorbidity) in male combat veterans (Miller, Greif, & Smith, 2003; Miller, Kaloupek, Dillon, & Keane, 2004) and female civilians (Miller & Resick, 2007). One extension of the externalizing/internalizing typology in PTSD is the exploration of anger, as aggression is a key component of the externalizing type. The question becomes whether the PTSD anger symptom is present only in externalizers with PTSD, or if anger is relevant for internalizers and simple PTSD individuals. Anger in PTSD is behaviorally defined as “irritability and anger outbursts” in the Diagnostic and Statistical Manual of Mental Disorders (4th ed.; DSM-IV; American Psychiatric Association, 1994, pg. 428). The examination of anger using this personality typology has the potential to clarify the variability in anger expression among individuals with PTSD. We first hypothesized that the externalizing/internalizing typology would be replicated in a previously unstudied PTSD population—female veterans—and would produce the externalizing, internalizing, and simple PTSD personality types
Correspondence concerning this article should be addressed to Diane T. Castillo, New Mexico VA Health Care System, BHCL (116), 1501 San Pedro SE, Albuquerque, New Mexico, 87108. E-mail: [email protected] C 2014 International Society for Traumatic Stress Studies. View Copyright this article online at wileyonlinelibrary.com DOI: 10.1002/jts.21886
108
109
Externalizing and Internalizing Subtypes
previously identified. Once classified, the characteristics of the groups would be validated by an independent measure. Finally, we hypothesized the groups would differ on a cognitive and behavioral anger measure, with internalizers high on cognitive and externalizers high on behavioral anger scales. The simple PTSD group was expected to score comparatively lower on all anger scales. Method Participants and Procedures Data were collected from a help-seeking sample of 254 female veterans with PTSD in an outpatient women’s PTSD clinic at the New Mexico VA Health Care System. The mean age was 41.36 years (SD = 10.28) and ethnically identified as non-Hispanic White (54.7%), Hispanic (23.6%), other (11.4%), and no response (10.2%). Traumatic events were identified as 50.4% sexual assault, 43.3% multiple traumas (including sexual assault), 2.8% combat, and 3.6% other. Participants referred for PTSD treatment completed various assessments, which included psychological tests and structured interviews. Demographic variables were collected through medical record chart review and study approval was granted by local VA and University of New Mexico Institutional Review Boards. Measures The Minnesota Multiphasic Personality Inventory-2 (MMPI2; Butcher, 1989), a self-report instrument consisting of 567 items, defined the sample (N = 254). Cluster analyses were conducted on three of the Personality Psychopathology Five (Psy5; Harkness et al., 1995) MMPI-2 scales—Disconstraint, Negative Emotionality, and Introversion, as in previous studies (Miller et al., 2003). From the MMPI-2, the Demoralization, Low Positive Emotion, Dysfunctional Negative Emotions, and Antisocial Behavior scales were selected for analysis. The Clinician Administered PTSD Scale (CAPS; Blake et al., 1995) is a structured clinical interview that assesses for frequency and severity of PTSD, current (past month), and lifetime symptoms; 208 participants completed the CAPS. The Millon Clinical Multiaxial Inventory-III (Millon, Millon, Davis, & Grossman, 1994) is a self-report measure of psychiatric conditions consisting of 175 items; 120 participants completed the Millon Clinical Multiaxial Inventory-III. From the Millon Clinical Multiaxial Inventory-III, the Anxiety Disorder, Dysthymic Disorder, Major Depression, Antisocial, Aggressive/Sadistic, Alcohol Dependence, Drug Dependence scales were selected for analysis. The Buss-Durkee Hostility Inventory (BDHI; Buss & Durkee, 1957) is a self-report measure with 75 questions to assess for five behavioral and three cognitive scales of anger; 170 participants completed the BDHI. Variations in numbers completing the CAPS, Millon Clinical Multiaxial Inventory-III, and BDHI reflect inconsistent clinical assessments in this convenience sample, therefore analyses are based on numbers available for each instrument.
Data Analysis A two-stage procedure for lustering was used to analyze the Disconstraint, Negative Emotionality, and Introversion Psy5 scales without a priori group specification (Miller et al., 2003) in the 254 subjects. The data were then submitted to a Kmeans cluster analysis to sort participants. Correlations and mean differences were then examined. Results The a priori analysis using an agglomeration plot to evaluate the scales determined three clusters to be the optimum number, replicating past research (Miller et al., 2003). In the second step, K-means cluster analysis portioned the sample into three groups, which were labeled externalizers (n = 33, 12.9%), internalizers (n = 106, 41.7%), and simple PTSD (n = 115, 45.3%). An analysis of variance (ANOVA) computed on select scales representing internalizing/externalizing characteristics from the MMPI-2 and the Millon Clinical Multiaxial Inventory-III revealed multiple significant group differences (see Table 1). As in previous studies, externalizers scored highest on scales for antisocial behavior, aggression, and alcohol dependence, whereas internalizers scored highest on scales of anxiety, depression, and dysthymia. An ANOVA computed on the combined current and lifetime CAPS scored revealed internalizers (M = 189.73, SD = 39.67) significantly higher (p < .001) than externalizers (M = 170.41, SD = 51.28; d = 0.42) and simple PTSD (M = 162.57, SD = 38.11; d = 0.70) groups. An ANOVA on the BDHI found the externalizers higher than the simple PTSD group on seven of the eight anger scales (Resentment d = 0.74, Suspicion d = 1.19, Guilt d = 0.74, Irritability d = 0.60, Assault d = 1.29, Negativism d = 0.69, and Verbal Hostility d = 1.29), and higher than internalizers on the Assault (d = 0.81) and Verbal Hostility (d = 0.94) scales; internalizers scored higher than the simple PTSD group on Resentment (d = 0.76), Suspicion (d = 1.17), Guilt (d = 0.62), and Irritability (d = 0.75) scales (see Table 2 for full results). Within all three groups, two cognitive scales—Resentment and Suspicion—were the highest of all BDHI scales. Spearman correlations were computed between the BDHI scales and the irritability and anger outbursts D2 symptom of the current CAPS scores (see Table 3) for the total sample and by cluster to determine the types of anger captured in the D2 symptom interview question. In the total sample, six of the eight BDHI scales (behavioral and cognitive) correlated significantly with the D2 PTSD anger symptom (range: r = .19 to .36), with the highest correlation on the Verbal Hostility scale. Within groups, assault was the only significant correlation (r = .59, p = .010) for externalizers, and verbal hostility the only significant correlation for internalizers (r = .30, p = .009) and simple PTSD (r = .37, p = .002) groups. Discussion This study succeeded in applying the externalizing/ internalizing personality characterization to female PTSD
Journal of Traumatic Stress DOI 10.1002/jts. Published on behalf of the International Society for Traumatic Stress Studies.
110
Castillo et al.
Table 1 Means and Standard Deviations for Select MMPI-2 T-score and MCMI-III Base Rate Scale Scores for PTSD Female Veterans by Cluster Cluster
MMPI-2 RC Scalesa Demoralization Low Pos Emotion Dysfunc Neg Emo Antisoc Behavior MCMI-IIIb Anxiety Disorder Dysthymic Dis Major Depression Antisocial Aggr/Sadistic Alcohol Depend Drug Depend
Externalizer (1)
Internalizer (2)
Simple PTSD (3)
M
SD
M
SD
M
SD
F(2, 251)
64.41 57.83 67.69 68.62
8.84 13.00 9.95 8.99
75.92 80.51 72.36 63.99
7.08 12.52 9.45 10.36
63.69 65.62 57.19 57.36
10.63 13.68 9.69 8.64
85.17 62.33 69.17 64.75 67.08 62.42 61.75
12.27 26.61 31.73 12.36 5.89 6.11 6.11
88.52 77.11 80.50 52.72 59.13 55.94 51.93
14.23 18.69 16.72 21.54 15.42 21.97 25.39
74.89 60.11 62.39 46.65 49.24 44.13 44.93
23.95 26.23 29.17 22.89 21.20 26.41 25.34
52.8*** 51.1*** 68.2*** 22.6*** F (2, 117) 7.0*** 7.7*** 7.3*** 3.7* 7.1*** 5.0** 2.8
Pairwise Contrast 2 > 1, 3 2>3>1 2>1>3 1>2>3 2>3 2>3 2>3 1>3 1>3 1>3 ns
Note. MMPI-2 = Minnesota Multiphasic Personality Inventory-2; RC = Restructured, Low Pos Emotion = Low Positive Emotion, Dysfunc Neg Emo = Dysfunctional Negative Emotions, Antisoc Behavior = Antisocial Behavior; MCMI-III = Millon Clinical Multiaxial Inventory-III, Dysthymic Dis = Dysthymic Disorder, Aggr/Sadistic = Aggressive/Sadistic, Alcohol Depend = Alcohol Dependence, Drug Depend = Drug Dependence. a MMPI-2 RC: Externalizers n = 33, Internalizers n = 106, Simple PTSD n = 115. b MCMI-III 2: Externalizers n = 12, Internalizers n = 54, Simple PTSD n = 54. *p < .05. **p < .01.***p < .001.
Additionally, this study provides unique information in understanding the role of anger in PTSD, both generally and within personality clusters, with expected and unexpected results. Although externalizers were predictably higher than the other two groups on the two most extreme behavioral anger scales of the BDHI, our hypothesis that internalizers would be higher on cognitive scales was not supported. In fact, both externalizers and internalizers showed significantly higher cognitive and
veterans, with a third, low-pathology simple PTSD group. As predicted, and consistent with male veterans (Miller et al., 2004) and female civilians (Miller & Resick, 2007), female veterans with PTSD showed externalizers high in aggression, alcohol use, and antisocial personality disorder; and internalizers high in depression and anxiety symptoms. These findings were further validated by an independent personality measure, the Millon Clinical Multiaxial Inventory-III.
Table 2 Means and Standard Deviations for Buss-Durkee Hostility Inventory T Scores by Cluster Cluster
Variable Resentment Suspicion Guilt Irritability Assault Indirect hostility Negativism Verbal hostility
Externalizer (1) n = 18
Internalizer (2) n = 77
Simple PTSD (3) n = 75
M
SD
M
SD
M
SD
F(2, 167)
Pairwise contrast
67.22 70.06 54.78 55.67 62.22 50.17 56.50 59.72
11.51 14.63 8.26 7.42 10.85 10.89 10.89 10.20
67.31 71.29 53.69 57.10 53.56 50.14 51.75 49.36
11.38 10.33 8.34 7.90 10.50 10.26 12.24 11.89
58.91 57.32 48.09 50.83 49.19 46.72 48.49 46.60
10.83 13.45 9.80 8.74 9.28 10.11 12.27 10.11
11.9*** 26.3*** 8.8*** 11.4*** 13.0*** 2.3 3.6* 10.4***
2&1>3 2&1>3 2&1>3 2&1>3 1>2&3 ns 1>3 1>2&3
Journal of Traumatic Stress DOI 10.1002/jts. Published on behalf of the International Society for Traumatic Stress Studies.
111
Externalizing and Internalizing Subtypes
Table 3 Spearman Correlations Between BDHI Scale Scores and CAPS D2 PTSD Irritability/Anger Outbursts by Group
BDHI Scales Assault Indirect hostility Irritability Negativism Resentment Suspicion Verbal hostility Guilt
Simple Total Externalizer Internalizer PTSD sample n = 18 n = 72 n = 67 n = 157 .59** .25 .24 .20 .16 .07 .27 .25
.19 −.11 .13 .12 .12 −.03 .30** −.13
.10 .03 .14 .18 .02 .12 .37** −.16
.27*** .06 .23** .19* .23** .20* .36*** .03
Note. CAPS = Clinician Administered PTSD Scale; BDHI = Buss-Durkee Hostility Inventory; PTSD = Posttraumatic Stress Disorder. Numbers of subjects in each group reflect individuals with both CAPS and BDHI scores. *p < .05. **p < .01. ***p < .001.
behavioral anger than the simple PTSD group, a finding not necessarily driven by psychopathology. As seen in correlations within clusters, the simple PTSD group also showed the highest correlation between verbal hostility and D2. Although D2 correlated singularly with behavioral scales within all three clusters, the total sample showed significant correlations across the full range of behavioral and cognitive scales. A final consideration was that the two highest BDHI scale means within each cluster was not behavioral anger, but rather cognitive anger, specifically resentment and suspicion. This finding is consistent with previous results (Castillo, C’de Baca, Conforti, Qualls, & Fallon, 2002) where female PTSD veterans showed higher cognitive anger than male veterans with other psychiatric diagnoses. Limitations of this study are the use of a convenience sample, with low numbers of externalizers, and higher CAPS scores for internalizers. Although the real-world clinical sample contributed to generalizability of the results, it also led to missing data by inconsistent administration of instruments. As such, a reduction in power among the externalizers could have impaired the ability to identify other group differences. Finally, although higher CAPS scores in internalizers could have confounded results, the externalizers did score significantly higher on some subscales, suggesting that PTSD severity did not impact scale scores differentially. The present study lays the groundwork for future research in the examination of anger in personality subtypes in PTSD. It is the first to examine anger dimensions within subtypes in any sample with PTSD. The findings support the conclusion that anger is a core symptom of PTSD in general, represented by both cognitive and behavioral features, and within psychopathology types. Anger contributes to PTSD onset, maintenance, and treatment noncompliance (Novaco & Chemtob,
1998), and though anger is typically viewed as behavioral aggression, this study provides some initial information on the breadth and types of anger, particularly the role of cognitive anger in PTSD. The current findings provide direction for future research in assessment and treatment of PTSD, both in the categorization of PTSD subtypes in general, and specifically in anger within these subtypes in women with PTSD. References American Psychiatric Association. (1994). Diagnostic and statistical manual for mental disorders (4th ed.). Washington, DC: Author. Blake, D., Weathers, F. W., Nagy, L. M., Kaloupek, D. G., Gusman, F. D., Charney, D. S., & Keane, T. M. (1995). The development of a clinician-administered PTSD scale. Journal of Traumatic Stress, 8, 75–90. doi:10.1002/jts.2490080106 Buss, A. H., & Durkee, A. (1957). An inventory for assessing different kinds of hostility. Journal of Consulting Psychology, 21, 343–349. doi:10.1037/h0046900 Butcher, J. N. (1989). Minnesota Multiphasic Personality Inventory-2, user’s guide, the Minnesota report: Adult clinical system. Minneapolis, MN: National Computer Systems. Castillo, D. T., C’de Baca, J., Conforti, K., Qualls, C., & Fallon, S. K. (2002). Anger in PTSD: General psychiatric and gender differences on the BDHI. Journal of Loss and Trauma, 7, 119–128. doi:10.1080/153250202753472282 Harkness, A. R., McNulty, J. L., & Ben-Porath, Y. S. (1995). The Personality Psychopathology Five (PSY-5): Constructs and MMPI-2 scales. Psychological Assessment, 7, 104–114. doi:10.1037/1040–3590.7.1.104 Kendler, K. S., Prescott, C. A., Myers, J., & Neale, M. C. (2003). The structure of genetic and environmental risk factors for common psychiatric and substance use disorders in men and women. Archives of General Psychiatry, 60, 929–937. Kessler, R. C., Chiu, W., Demler, O., & Walters, E. E. (2005). Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the national comorbidity survey replication. Archives of General Psychiatry, 62, 617– 627. doi:10.1001/archpsyc.62.6.617 Miller, M. W., Greif, J. L., & Smith, A. A. (2003). Multidimensional Personality Questionnaire profiles of veterans with traumatic combat exposure: Externalizing and internalizing subtypes. Psychological Assessment, 15, 205–215. doi:10.1037/1040–3590.15.2.205 Miller, M. W., Kaloupek, D. G., Dillon, A. L., & Keane, T. M. (2004). Externalizing and internalizing subtypes of combat-related PTSD: A replication and extension using the PSY-5 scales. Journal of Abnormal Psychology, 113, 636–645. doi:10.1037/0021–843X.113.4.636 Miller, M. W., & Resick, P. A. (2007). Internalizing and externalizing subtypes in female sexual assault survivors: Implications for the understanding of complex PTSD. Behavior Therapy, 38, 58–71. doi:10.1016/j.beth.2006.04.003 Millon, T., Millon, C., Davis, R., & Grossman, S. (1994). Millon Clinical Multiaxial Inventory III. Minneapolis, MN: National Computer Systems. Novaco, R. W., & Chemtob, C. M. (1998). Anger and trauma: Conceptualization, assessment, and treatment. In V. M. Follette., J. I. Ruzek, & F. R. Abueg (Eds.), Cognitive–behavioral therapies for trauma (pp. 162–190). New York, NY: Guilford Press.
Journal of Traumatic Stress DOI 10.1002/jts. Published on behalf of the International Society for Traumatic Stress Studies.
BRIEF REPORT
Externalizing and Internalizing Subtypes of Posttraumatic Psychopathology and Anger Expression Diane T. Castillo,1,2 Jeremy S. Joseph,1,5 Andra T. Tharp,6 Janet C’de Baca,1 Lorraine M. Torres-Sena,1 Clifford Qualls,4 and Mark W. Miller3 1 New Mexico VA Health Care System, Albuquerque, New Mexico, USA Department of Psychiatry, University of New Mexico, Albuquerque, New Mexico, USA 3 National Center for PTSD, VA Boston Healthcare System, Boston, Massachusetts, USA 4 Department of Statistics, University of New Mexico, Albuquerque, New Mexico, USA 5 University of Wyoming 6 Baylor College of Medicine, Houston, Texas, USA
2
Subtypes of posttraumatic psychopathology were replicated and extended in 254 female veterans with posttraumatic stress disorder (PTSD). Cluster analyses on Minnesota Multiphasic Personality Inventory-2 and Personality Psychopathology Five scales (Harkness, McNulty, & Ben-Porath, 1995) yielded internalizing and externalizing psychopathology dimensions, with a third low psychopathology group (simple PTSD). Externalizers were higher than the internalizers and the simple PTSD groups on the antisocial, substance, and aggression scales; internalizers were higher on depression and anxiety scales. Further validation included an independent measure of psychopathology to examine anger (Buss-Durkee Hostility Inventory, [BDHI]; Buss & Durkee, 1957). Externalizers were higher on extreme behavioral anger scales (assault and verbal hostility); and externalizers and internalizers were higher than the simple PTSD subjects on other anger scales. Positive correlations between the BDHI scales and the PTSD symptom of “irritability and anger outbursts” were found across scales in the total sample (range: r = .19–.36), on the assault scale in externalizers (r = .59), and the verbal hostility scale in both internalizers (r = .30) and simple PTSD (r = .37) groups, suggesting the broad utility of the symptom in the diagnosis. The results demonstrate the generalizability of the internalizing/externalizing typology to the female veteran population and highlight clinically relevant distinctions in anger expression within PTSD.
Considerable variability exists within the manifestations of psychiatric disorders, which researchers have accounted for through the examination of symptom clusters categorized within the externalizing and internalizing psychopathology dimensions (e.g., Kendler, Prescott, Myers, & Neale, 2003; Kessler, Chiu, Demler, & Walters, 2005). Across psychiatric diagnoses, results support these two latent dimensions in which externalizers are characterized by disconstraint, aggressiveness, and high-negative emotion, whereas internalizers are characterized by high-negative emotion and low-positive emotion. Furthermore, externalizers are more likely to endorse substance abuse, symptoms of antisocial personality disorder, and conduct disorder whereas internalizers are more likely to endorse comorbidity for anxiety and depression. Recent studies have
found this personality typology relevant to the diagnosis of posttraumatic stress disorder (PTSD), with an added third, lowpathology group classified as simple PTSD (absence of comorbidity) in male combat veterans (Miller, Greif, & Smith, 2003; Miller, Kaloupek, Dillon, & Keane, 2004) and female civilians (Miller & Resick, 2007). One extension of the externalizing/internalizing typology in PTSD is the exploration of anger, as aggression is a key component of the externalizing type. The question becomes whether the PTSD anger symptom is present only in externalizers with PTSD, or if anger is relevant for internalizers and simple PTSD individuals. Anger in PTSD is behaviorally defined as “irritability and anger outbursts” in the Diagnostic and Statistical Manual of Mental Disorders (4th ed.; DSM-IV; American Psychiatric Association, 1994, pg. 428). The examination of anger using this personality typology has the potential to clarify the variability in anger expression among individuals with PTSD. We first hypothesized that the externalizing/internalizing typology would be replicated in a previously unstudied PTSD population—female veterans—and would produce the externalizing, internalizing, and simple PTSD personality types
Correspondence concerning this article should be addressed to Diane T. Castillo, New Mexico VA Health Care System, BHCL (116), 1501 San Pedro SE, Albuquerque, New Mexico, 87108. E-mail: [email protected] C 2014 International Society for Traumatic Stress Studies. View Copyright this article online at wileyonlinelibrary.com DOI: 10.1002/jts.21886
108
109
Externalizing and Internalizing Subtypes
previously identified. Once classified, the characteristics of the groups would be validated by an independent measure. Finally, we hypothesized the groups would differ on a cognitive and behavioral anger measure, with internalizers high on cognitive and externalizers high on behavioral anger scales. The simple PTSD group was expected to score comparatively lower on all anger scales. Method Participants and Procedures Data were collected from a help-seeking sample of 254 female veterans with PTSD in an outpatient women’s PTSD clinic at the New Mexico VA Health Care System. The mean age was 41.36 years (SD = 10.28) and ethnically identified as non-Hispanic White (54.7%), Hispanic (23.6%), other (11.4%), and no response (10.2%). Traumatic events were identified as 50.4% sexual assault, 43.3% multiple traumas (including sexual assault), 2.8% combat, and 3.6% other. Participants referred for PTSD treatment completed various assessments, which included psychological tests and structured interviews. Demographic variables were collected through medical record chart review and study approval was granted by local VA and University of New Mexico Institutional Review Boards. Measures The Minnesota Multiphasic Personality Inventory-2 (MMPI2; Butcher, 1989), a self-report instrument consisting of 567 items, defined the sample (N = 254). Cluster analyses were conducted on three of the Personality Psychopathology Five (Psy5; Harkness et al., 1995) MMPI-2 scales—Disconstraint, Negative Emotionality, and Introversion, as in previous studies (Miller et al., 2003). From the MMPI-2, the Demoralization, Low Positive Emotion, Dysfunctional Negative Emotions, and Antisocial Behavior scales were selected for analysis. The Clinician Administered PTSD Scale (CAPS; Blake et al., 1995) is a structured clinical interview that assesses for frequency and severity of PTSD, current (past month), and lifetime symptoms; 208 participants completed the CAPS. The Millon Clinical Multiaxial Inventory-III (Millon, Millon, Davis, & Grossman, 1994) is a self-report measure of psychiatric conditions consisting of 175 items; 120 participants completed the Millon Clinical Multiaxial Inventory-III. From the Millon Clinical Multiaxial Inventory-III, the Anxiety Disorder, Dysthymic Disorder, Major Depression, Antisocial, Aggressive/Sadistic, Alcohol Dependence, Drug Dependence scales were selected for analysis. The Buss-Durkee Hostility Inventory (BDHI; Buss & Durkee, 1957) is a self-report measure with 75 questions to assess for five behavioral and three cognitive scales of anger; 170 participants completed the BDHI. Variations in numbers completing the CAPS, Millon Clinical Multiaxial Inventory-III, and BDHI reflect inconsistent clinical assessments in this convenience sample, therefore analyses are based on numbers available for each instrument.
Data Analysis A two-stage procedure for lustering was used to analyze the Disconstraint, Negative Emotionality, and Introversion Psy5 scales without a priori group specification (Miller et al., 2003) in the 254 subjects. The data were then submitted to a Kmeans cluster analysis to sort participants. Correlations and mean differences were then examined. Results The a priori analysis using an agglomeration plot to evaluate the scales determined three clusters to be the optimum number, replicating past research (Miller et al., 2003). In the second step, K-means cluster analysis portioned the sample into three groups, which were labeled externalizers (n = 33, 12.9%), internalizers (n = 106, 41.7%), and simple PTSD (n = 115, 45.3%). An analysis of variance (ANOVA) computed on select scales representing internalizing/externalizing characteristics from the MMPI-2 and the Millon Clinical Multiaxial Inventory-III revealed multiple significant group differences (see Table 1). As in previous studies, externalizers scored highest on scales for antisocial behavior, aggression, and alcohol dependence, whereas internalizers scored highest on scales of anxiety, depression, and dysthymia. An ANOVA computed on the combined current and lifetime CAPS scored revealed internalizers (M = 189.73, SD = 39.67) significantly higher (p < .001) than externalizers (M = 170.41, SD = 51.28; d = 0.42) and simple PTSD (M = 162.57, SD = 38.11; d = 0.70) groups. An ANOVA on the BDHI found the externalizers higher than the simple PTSD group on seven of the eight anger scales (Resentment d = 0.74, Suspicion d = 1.19, Guilt d = 0.74, Irritability d = 0.60, Assault d = 1.29, Negativism d = 0.69, and Verbal Hostility d = 1.29), and higher than internalizers on the Assault (d = 0.81) and Verbal Hostility (d = 0.94) scales; internalizers scored higher than the simple PTSD group on Resentment (d = 0.76), Suspicion (d = 1.17), Guilt (d = 0.62), and Irritability (d = 0.75) scales (see Table 2 for full results). Within all three groups, two cognitive scales—Resentment and Suspicion—were the highest of all BDHI scales. Spearman correlations were computed between the BDHI scales and the irritability and anger outbursts D2 symptom of the current CAPS scores (see Table 3) for the total sample and by cluster to determine the types of anger captured in the D2 symptom interview question. In the total sample, six of the eight BDHI scales (behavioral and cognitive) correlated significantly with the D2 PTSD anger symptom (range: r = .19 to .36), with the highest correlation on the Verbal Hostility scale. Within groups, assault was the only significant correlation (r = .59, p = .010) for externalizers, and verbal hostility the only significant correlation for internalizers (r = .30, p = .009) and simple PTSD (r = .37, p = .002) groups. Discussion This study succeeded in applying the externalizing/ internalizing personality characterization to female PTSD
Journal of Traumatic Stress DOI 10.1002/jts. Published on behalf of the International Society for Traumatic Stress Studies.
110
Castillo et al.
Table 1 Means and Standard Deviations for Select MMPI-2 T-score and MCMI-III Base Rate Scale Scores for PTSD Female Veterans by Cluster Cluster
MMPI-2 RC Scalesa Demoralization Low Pos Emotion Dysfunc Neg Emo Antisoc Behavior MCMI-IIIb Anxiety Disorder Dysthymic Dis Major Depression Antisocial Aggr/Sadistic Alcohol Depend Drug Depend
Externalizer (1)
Internalizer (2)
Simple PTSD (3)
M
SD
M
SD
M
SD
F(2, 251)
64.41 57.83 67.69 68.62
8.84 13.00 9.95 8.99
75.92 80.51 72.36 63.99
7.08 12.52 9.45 10.36
63.69 65.62 57.19 57.36
10.63 13.68 9.69 8.64
85.17 62.33 69.17 64.75 67.08 62.42 61.75
12.27 26.61 31.73 12.36 5.89 6.11 6.11
88.52 77.11 80.50 52.72 59.13 55.94 51.93
14.23 18.69 16.72 21.54 15.42 21.97 25.39
74.89 60.11 62.39 46.65 49.24 44.13 44.93
23.95 26.23 29.17 22.89 21.20 26.41 25.34
52.8*** 51.1*** 68.2*** 22.6*** F (2, 117) 7.0*** 7.7*** 7.3*** 3.7* 7.1*** 5.0** 2.8
Pairwise Contrast 2 > 1, 3 2>3>1 2>1>3 1>2>3 2>3 2>3 2>3 1>3 1>3 1>3 ns
Note. MMPI-2 = Minnesota Multiphasic Personality Inventory-2; RC = Restructured, Low Pos Emotion = Low Positive Emotion, Dysfunc Neg Emo = Dysfunctional Negative Emotions, Antisoc Behavior = Antisocial Behavior; MCMI-III = Millon Clinical Multiaxial Inventory-III, Dysthymic Dis = Dysthymic Disorder, Aggr/Sadistic = Aggressive/Sadistic, Alcohol Depend = Alcohol Dependence, Drug Depend = Drug Dependence. a MMPI-2 RC: Externalizers n = 33, Internalizers n = 106, Simple PTSD n = 115. b MCMI-III 2: Externalizers n = 12, Internalizers n = 54, Simple PTSD n = 54. *p < .05. **p < .01.***p < .001.
Additionally, this study provides unique information in understanding the role of anger in PTSD, both generally and within personality clusters, with expected and unexpected results. Although externalizers were predictably higher than the other two groups on the two most extreme behavioral anger scales of the BDHI, our hypothesis that internalizers would be higher on cognitive scales was not supported. In fact, both externalizers and internalizers showed significantly higher cognitive and
veterans, with a third, low-pathology simple PTSD group. As predicted, and consistent with male veterans (Miller et al., 2004) and female civilians (Miller & Resick, 2007), female veterans with PTSD showed externalizers high in aggression, alcohol use, and antisocial personality disorder; and internalizers high in depression and anxiety symptoms. These findings were further validated by an independent personality measure, the Millon Clinical Multiaxial Inventory-III.
Table 2 Means and Standard Deviations for Buss-Durkee Hostility Inventory T Scores by Cluster Cluster
Variable Resentment Suspicion Guilt Irritability Assault Indirect hostility Negativism Verbal hostility
Externalizer (1) n = 18
Internalizer (2) n = 77
Simple PTSD (3) n = 75
M
SD
M
SD
M
SD
F(2, 167)
Pairwise contrast
67.22 70.06 54.78 55.67 62.22 50.17 56.50 59.72
11.51 14.63 8.26 7.42 10.85 10.89 10.89 10.20
67.31 71.29 53.69 57.10 53.56 50.14 51.75 49.36
11.38 10.33 8.34 7.90 10.50 10.26 12.24 11.89
58.91 57.32 48.09 50.83 49.19 46.72 48.49 46.60
10.83 13.45 9.80 8.74 9.28 10.11 12.27 10.11
11.9*** 26.3*** 8.8*** 11.4*** 13.0*** 2.3 3.6* 10.4***
2&1>3 2&1>3 2&1>3 2&1>3 1>2&3 ns 1>3 1>2&3
Journal of Traumatic Stress DOI 10.1002/jts. Published on behalf of the International Society for Traumatic Stress Studies.
111
Externalizing and Internalizing Subtypes
Table 3 Spearman Correlations Between BDHI Scale Scores and CAPS D2 PTSD Irritability/Anger Outbursts by Group
BDHI Scales Assault Indirect hostility Irritability Negativism Resentment Suspicion Verbal hostility Guilt
Simple Total Externalizer Internalizer PTSD sample n = 18 n = 72 n = 67 n = 157 .59** .25 .24 .20 .16 .07 .27 .25
.19 −.11 .13 .12 .12 −.03 .30** −.13
.10 .03 .14 .18 .02 .12 .37** −.16
.27*** .06 .23** .19* .23** .20* .36*** .03
Note. CAPS = Clinician Administered PTSD Scale; BDHI = Buss-Durkee Hostility Inventory; PTSD = Posttraumatic Stress Disorder. Numbers of subjects in each group reflect individuals with both CAPS and BDHI scores. *p < .05. **p < .01. ***p < .001.
behavioral anger than the simple PTSD group, a finding not necessarily driven by psychopathology. As seen in correlations within clusters, the simple PTSD group also showed the highest correlation between verbal hostility and D2. Although D2 correlated singularly with behavioral scales within all three clusters, the total sample showed significant correlations across the full range of behavioral and cognitive scales. A final consideration was that the two highest BDHI scale means within each cluster was not behavioral anger, but rather cognitive anger, specifically resentment and suspicion. This finding is consistent with previous results (Castillo, C’de Baca, Conforti, Qualls, & Fallon, 2002) where female PTSD veterans showed higher cognitive anger than male veterans with other psychiatric diagnoses. Limitations of this study are the use of a convenience sample, with low numbers of externalizers, and higher CAPS scores for internalizers. Although the real-world clinical sample contributed to generalizability of the results, it also led to missing data by inconsistent administration of instruments. As such, a reduction in power among the externalizers could have impaired the ability to identify other group differences. Finally, although higher CAPS scores in internalizers could have confounded results, the externalizers did score significantly higher on some subscales, suggesting that PTSD severity did not impact scale scores differentially. The present study lays the groundwork for future research in the examination of anger in personality subtypes in PTSD. It is the first to examine anger dimensions within subtypes in any sample with PTSD. The findings support the conclusion that anger is a core symptom of PTSD in general, represented by both cognitive and behavioral features, and within psychopathology types. Anger contributes to PTSD onset, maintenance, and treatment noncompliance (Novaco & Chemtob,
1998), and though anger is typically viewed as behavioral aggression, this study provides some initial information on the breadth and types of anger, particularly the role of cognitive anger in PTSD. The current findings provide direction for future research in assessment and treatment of PTSD, both in the categorization of PTSD subtypes in general, and specifically in anger within these subtypes in women with PTSD. References American Psychiatric Association. (1994). Diagnostic and statistical manual for mental disorders (4th ed.). Washington, DC: Author. Blake, D., Weathers, F. W., Nagy, L. M., Kaloupek, D. G., Gusman, F. D., Charney, D. S., & Keane, T. M. (1995). The development of a clinician-administered PTSD scale. Journal of Traumatic Stress, 8, 75–90. doi:10.1002/jts.2490080106 Buss, A. H., & Durkee, A. (1957). An inventory for assessing different kinds of hostility. Journal of Consulting Psychology, 21, 343–349. doi:10.1037/h0046900 Butcher, J. N. (1989). Minnesota Multiphasic Personality Inventory-2, user’s guide, the Minnesota report: Adult clinical system. Minneapolis, MN: National Computer Systems. Castillo, D. T., C’de Baca, J., Conforti, K., Qualls, C., & Fallon, S. K. (2002). Anger in PTSD: General psychiatric and gender differences on the BDHI. Journal of Loss and Trauma, 7, 119–128. doi:10.1080/153250202753472282 Harkness, A. R., McNulty, J. L., & Ben-Porath, Y. S. (1995). The Personality Psychopathology Five (PSY-5): Constructs and MMPI-2 scales. Psychological Assessment, 7, 104–114. doi:10.1037/1040–3590.7.1.104 Kendler, K. S., Prescott, C. A., Myers, J., & Neale, M. C. (2003). The structure of genetic and environmental risk factors for common psychiatric and substance use disorders in men and women. Archives of General Psychiatry, 60, 929–937. Kessler, R. C., Chiu, W., Demler, O., & Walters, E. E. (2005). Prevalence, severity, and comorbidity of 12-month DSM-IV disorders in the national comorbidity survey replication. Archives of General Psychiatry, 62, 617– 627. doi:10.1001/archpsyc.62.6.617 Miller, M. W., Greif, J. L., & Smith, A. A. (2003). Multidimensional Personality Questionnaire profiles of veterans with traumatic combat exposure: Externalizing and internalizing subtypes. Psychological Assessment, 15, 205–215. doi:10.1037/1040–3590.15.2.205 Miller, M. W., Kaloupek, D. G., Dillon, A. L., & Keane, T. M. (2004). Externalizing and internalizing subtypes of combat-related PTSD: A replication and extension using the PSY-5 scales. Journal of Abnormal Psychology, 113, 636–645. doi:10.1037/0021–843X.113.4.636 Miller, M. W., & Resick, P. A. (2007). Internalizing and externalizing subtypes in female sexual assault survivors: Implications for the understanding of complex PTSD. Behavior Therapy, 38, 58–71. doi:10.1016/j.beth.2006.04.003 Millon, T., Millon, C., Davis, R., & Grossman, S. (1994). Millon Clinical Multiaxial Inventory III. Minneapolis, MN: National Computer Systems. Novaco, R. W., & Chemtob, C. M. (1998). Anger and trauma: Conceptualization, assessment, and treatment. In V. M. Follette., J. I. Ruzek, & F. R. Abueg (Eds.), Cognitive–behavioral therapies for trauma (pp. 162–190). New York, NY: Guilford Press.
Journal of Traumatic Stress DOI 10.1002/jts. Published on behalf of the International Society for Traumatic Stress Studies.
