Personality Disorders in Patients With Panic Disorder: Association With Childhood Anxiety Disorders, Early Trauma, Comorbidity, and Chronicity Mark H. Pollack, Michael W. Otto, Jerrold F. Rosenbaum, and Gary S. Sachs The rates of comorbid personality disorders in patients with panic disorder are reported to be elevated, have an adverse impact on the response to treatment, and increase the likelihood of relapse on treatment discontinuation. We examined the rates of personality disorders in panic disorder patients in a longitudinal, naturalistic study of panic disorder. of 100 panic disorder patients studied, 42 met criteria for at least one personality disorder as determined by the Personality Disorder Questionnaire-Revised (PDQ-R). The presence of a personality disorder as determined by the PDQ-R was associated with a past history of childhood anxiety disorders, comorbidity with other anxiety disorders and depression, and a chronic, unremitting course of panic disorder in adulthood. The
presence of a personality disorder in these patients was not significantly associated with a history of physical or sexual abuse in childhood. Our findings support the notion that an anxiety diathesis, demonstrated by significant difficulties with anxiety in childhood, influences the development of apparent personality dysfunction in panic patients. In other cases, personality pathology may reflect the presence of comorbid anxiety disorders or depression. The association of personality disorder in panic patients with a more unremitting course of illness underscores the importance of axis II pathology in understanding the longitudinal course of panic disorder. Copyright 0 1992 by W.B. Saunders Company
T
HE RELATIONSHIP between personality disorders and panic disorder is an important area of clinical concern; the presence of comorbid personality disorder in panic disorder patients adversely affects response to antipanic treatment’** and increases the likelihood of relapse after treatment discontinuation.3 Personality dysfunction in patients with panic disorder is apparently common, with reported rates ranging from 27% to 58%, depending on the study and diagnostic instrument used to assess personality.4,5 However, the nature of the association between panic and personality is unclear. Patients with recurrent panic attacks and high levels of anticipatory anxiety, phobic fear, and avoidance are often dependent, avoidant, and unassertive; whether these personality traits reflect a premorbid personality disorder or arise from the chronic experience of anxiety remains uncertain. Some personality characteristics, such as dependency, may improve with antipanic treatment, suggesting either that the dependent characteristics are secondary to panic disorder
From rhe Anxiety Clinical Research Unit, Clinical Psychopharacology Unit, Massachusetts General Hospital, Boston, MA. Address reprint requests to Mark H. Pollack, M.D., Anxiety Clinical Research Unit, Clinical Psychopharmacology Unit, Massachusetts General Hospital, ACC-815, 15 Parkman St, Boston, MA 02114. Copyright 0 I992 by W.B. Saunders Company 0010-440X/92/3302-0014$03.00/0 78
or that the treatment has specific effects on dependence. However, other traits, such as histrionic and avoidant, may be more resistant to change, arguing against the proposition that personality measures are simply epiphenomena of the panic disorder.4 Furthermore, no specific link has been demonstrated between particular personality disorders and panic disorder, as elevated rates of personality dysfunction also occur in other anxiety disorders such as social phobia and OCD, as well as in depression.6-8 Mavissakalian has suggested that panic disorder and some personality traits may be separate manifestations of a common diathesis.’ Evidence for a diathesis of risk for anxiety that may first manifest in childhood is supported by findings that offspring of panic disorder patients have higher rates of shyness and behavioral inhibition to the unfamiliar than comparison children.’ In addition, behaviorally inhibited children are at increased risk to develop anxiety disorders.” Furthermore, adult panic disorder patients often report a history of childhood anxiety disorders.“~” Significant difficulties with anxiety in childhood may also be associated with the development of dysfunctional personality characteristics that may interact with developmental and psychosocial protective or risk factors (eg, trauma) and manifest in the emergence of axis II disorders. We undertook the present study to examine whether panic patients with comorbid personal-
Comprehensive Psychiatry, Vol. 33, No. 2 (March/April), 1992: pp 78-83
PERSONALITY
79
DISORDERS IN PANIC PATIENTS
ity disorders showed evidence for an anxiety diathesis as manifested by a history of anxiety disorders in childhood and had experienced higher rates of a known risk factor for personality dysfunction (ie, early abuse). We also sought to assess the association between personality dysfunction and comorbid depression, comorbid anxiety disorders, and chronicity of the panic disorder. HYPOTHESIS
1
Panic Disorder Patients With Personality Disorders Are More Likely to Have a History of Childhood Anxiety Disorders
If panic disorder and its comorbid personality disorders are expressions of a common diathesis first manifest in childhood and influencing development, we would expect that adult panic patients with personality disorders would be more likely to have a history of an anxiety disorder during childhood. HYPOTHESIS
2
Panic Disorder Patients With a Comorbid Personality Disorder Are More Likely to Have Experienced Childhood Trauma
The risk for developing a personality disorder is reportedly higher for individuals exposed to early traumatic life experiences.13 If early trauma, in the form of sexual or physical abuse, influences the development of personality dysfunction, we would also expect to find a history of childhood abuse more commonly in panic patients with a concomitant personality disorder than in panic patients without a personality disorder. HYPOTHESIS
3
Panic Disorder Patients With Comorbid Personality Disorders Will Have Higher Rates of Comorbid Depression and Other Anxiety Disorders
Studies suggest that depression and a general quality of increased “anxiety proneness” are correlates of personality disorder in panic patients.‘4-1hAlthough the direction of the association is unclear (ie, whether a personality disorder predisposes to anxiety or depression or whether a mood or anxiety disorder influences the development of personality disturbance),
we predict personality disordered panic patients to be generally more severely ill as manifested in higher rates of depression and comorbid anxiety disorders. HYPOTHESIS 4
Panic Disorder Patients With a Comorbid Personality Disorder Will Have a More Chronic Course of Illness
In clinical trials of drug treatment of panic disorder, the presence of a personality disorder predicted poorer treatment outcome’.‘; thus, a personality disorder diagnosis may be associated with a lower likelihood of achieving a period of full remission from the panic disorder. METHOD To examine these issues, we assessed data from the first 100 patients entered into a naturalistic study of the longitudinal course of panic disorder. Description of study methods is detailed in a separate report.” Patients met DSMIII-R criteria for panic disorder with or without phobic avoidance.” Baseline assessments included the Structured Clinical Interview for DLY~M-III-R’~ to diagnose anxiety, affective, somatization. and substance use disorders, if present, and a detailed history of course of illness and past and current treatment. History of childhood anxiety disorders was examined using structured interviews including the Schedule for Affective Disorder and Schizophrenia (KSADS) items to assess childhood agoraphobia, overanxious disorders, separation anxiety disorder, and social phobia, and the Diagnostic Interview for Children and Adolescents (DICA-P) for avoidant disorder.“‘” Subjects were drawn from a naturalistic study that enrolled patients entering or already receiving treatment for panic disorder, and who provided informed consent for participation in the study. Failure to achieve at least a 2-month period of remission at some point since the onset of panic disorder was used as a measure of chronicity of illness. Remission was defined as a period of no panic attacks and phobic avoidance and absent or very mild anticipatory anxiety that did not interfere with function. The presence of personality disorders was assessed using the Personality Diagnostic Questionnaire-Revised (PDQR), a self-report questionnaire of 152 true/false questions to assess axis II disorders as determined by DSM-III-R.” Personality disorders in DSM-III-R are divided into three groups: cluster A, the “schizoid” cluster, including paranoid, schizoid, and schizotypal personality disorders; cluster B. the “dramatic” cluster, including antisocial, borderline, histrionic, and narcissistic personality disorders; and cluster C. the “anxious” cluster, including avoidant. dependent. obsessive-compulsive, and passive-aggressive personality disorders.” The PDQ has been found to have adequate test-retest reliability for most disorders (K = .56 or above for paranoid, schizotypal, antisocial. borderline, avoidant. and compulsive personality disorders),” although it may
80
POLLACK ET AL
generate a number of false-positive diagnoses and overestimate the overall prevalence of personality disordersZ3 Hyler et a12’have suggested that the instrument’s high sensitivity, but only modest specificity, supports its use as a screening instrument for the presence of personality disorder. RESULTS
Subjects
Sixty-three of the 100 patients were women and 37 were men, with a mean age of 40.4 years (k11.3). The mean age at onset of panic disorder was 29.8 years (+llS), with a duration of illness of 8.6 years (r8.9; range, 1 to 34) when the current episode of treatment was initiated. Overall Rates of Personality Disorders
Of the 100 patients studied, 42 met criteria for at least one personality disorder as determined by the PDQ-R. The most common personality disorders were histrionic, borderline, avoidant, and compulsive (Table 1). Patients meeting criteria for a personality disorder in one of three diagnostic clusters of the DSMIII-R frequently met criteria for more than one cluster: 19 of the 42 (45%) patients meeting criteria for one personality disorder met criteria for a personality disorder in each of the three clusters, while 10 (24%) met criteria for one cluster diagnosis only. This overlap between personality disorders detected by the PDQ-R in this sample is consistent with the expected high Table 1. Rates of Personality Disorders in Panic Disorder Patients (n = 100) Overall
42%
Cluster A (25%) A only With other clusters Schizoid Schizotypal Paranoid
3% 22% 11% 8% 19%
sensitivity and moderate specificity of this instrumen? and limited the precision by which each cluster could be examined for individual associations with patient variables. Indeed, in the following analyses, when significant associations between the presence of a personality disorder and patient characteristics were obtained, they were generally obtained for each cluster. Thus, we report on the overall association between the presence of any personality disorder and specific patient characteristics; these associations were examined with Pearson chi-square tests. Significance was defined as P I .05 for each predicted association. Hypothesis 1: Panic Disorder and Childhood Anxiety Disorder
Panic disorder patients with a personality disorder according to the PDQ-R were more likely to have a history of a childhood anxiety disorder than those without a personality disorder. Seventy-six percent of patients with a personality disorder had a childhood history of anxiety, compared with 38% of patients without a personality disorder (x’ = 14.36, P < .OOl). Hypothesis 2: Personality Disorder and Past History of Childhood Abuse
The presence of a personality disorder was not significantly associated with a history of childhood sexual abuse or physical abuse. Twenty-four percent of patients with a personality disorder reported being sexually abused as children, compared with 12% of patients without a personality disorder (x’ = 2.56, P < .ll). Nineteen percent of patients with and 10% of those without a personality disorder reported physical abuse during childhood (x’ = 1.53, P < .22).
Cluster B (34%) B only With other clusters
5% 29%
Histrionic
22%
Borderline
23%
Narcissistic
12%
Antisocial
4%
Cluster C (34%) C only With other clusters
2% 32%
Avoidant
22%
Obsessive-compulsive
21%
Passive-aggressive
8% 7%
Hypothesis 3: Personality Disorders and Comorbidity of Other Psychiatric Disorders
Patients with a personality disorder were significantly more likely to have a lifetime history of major depression; 69% of patients with a personality disorder had lifetime history of a depressive disorder, compared with 36% of patients without a personality disorder (x’ = 10.51, P < .Ol). In addition, patients with a personality disorder were more likely to meet criteria for another anxiety disorder (social
PERSONALITY
DISORDERS IN PANIC PATIENTS
phobia, generalized anxiety disorder, or simple phobia); 79% of patients with a personality disorder had another anxiety disorder, as compared with 52% of those without a personality disorder (x’ = 7.53, P < .Ol). Hypothesis 4: Personality Disorder and Chronic@ of Course
Panic patients with a personality disorder were less likely to have achieved even a 2-month period of remission of the panic disorder. Only 29% of patients with a personality disorder had at least a 2-month period of remission, compared with 59% of patients without a personality disorder (x” = 8.86, P < .Ol). DISCUSSION
Forty-two percent of patients with panic disorder in this study met criteria for at least one personality disorder as assessed by the PDQ-R. These findings are consistent with the rates of personality disorders in panic disorder patients previously reported by others.‘.14 Personality disorders identified were from all three personality clusters, with 45% of patients with a personality disorder meeting criteria for a disorder in each of the three clusters. Meeting criteria for more than one personality disorder may be common,‘3,‘4 and reflect a number of influences, including the propensity of the PDQ-R to overestimate the prevalence of personality disorders, the failure of the diagnostic criteria to identify discrete personality syndromes, and the possibility that axis I pathology may truly be associated with multiple dysfunctional personality characteristics. Our findings must be considered in light of certain methodologic limitations. The overlap among diagnostic categories did not permit examination of associations with individual personality diagnoses. Perhaps the most significant limitation of this study is the use of the PDQ-R, a self-report instrument that, though sensitive to the presence of personality dysfunction, may not be adequately specific to permit more refined analysis of specific personality characteristics. However, this limitation may be inherent, to some extent, to the current study of personality disorders; diagnostic instruments and the criteria themselves continue to undergo further empirical refinement and there is as yet no clear
“gold standard” for personality assessment. The PDQ-R, given its ease of administration and reliability, may be useful in generating and examining initial hypotheses. Results of this study should be viewed as preliminary and warrant further evaluation with standardized clinical interview-based instruments. Nevertheless, results from two earlier studies of personality and panic disorder suggest high concordance between the PDQ and the clinician rated Structured Interview for DSM-III Personality Disorder (SIDP)” both for the presence or absence of any personality disorder, as well as for some individual diagnoses.‘4.‘” Thus, our use of the PDQ-R to establish the presence of personality dysfunction may be reasonable. The most frequently detected disorders in our sample were histrionic, borderline, avoidant, and obsessive-compulsive personality disorders. Dependent personality disorder was not as prevalent in our sample compared with previous reports.‘4~‘5 Since most of our patients were receiving treatment at the time of assessment, the lower rates of dependent personality observed may reflect previously reported findings of a decrease in dependent traits with treatment of panic disorder.’ The presence of a personality disorder was significantly associated with a history of an anxiety disorder in childhood. This finding may reflect the development of maladaptive character styles in response to long-standing exposure to anxiety, excessive physiological arousal, and the fearfulness associated with childhood anxiety disorders. Alternatively, panic disorder and personality disorder may be separate outcomes of the same biological or environmental risk factors that predispose to anxiety in childhood. Some children at high risk for the development of panic disorder manifest early shyness, fearfulness, and inhibition to the unfamiliar.’ These traits may be the early expressions of a diathesis for anxiety later expressed in adulthood as panic disorder or agoraphobia. Our finding of an association between personality disorders in adulthood and childhood anxiety disorders, suggests that some aspects of apparent personality dysfunction may reflect the presence of anxiety and a characteristic pattern of response to it, first manifest in childhood and variably expressed over time as axis I and axis II pathology.
82
POLLACK ET AL
This hypothesis might best be examined by longitudinal study of a cohort of behaviorally inhibited children to examine protective and risk factors for the adult expression of these characteristics. Although there were weak trends in our data in the direction of an association between early trauma and the development of adult personality disorder, these did not reach statistical significance and our findings for patients with panic disorder do not strongly support previous descriptions of a role for childhood sexual or physical in the development of adult personality disorders.13 It is possible that patients may have underreported their history of abuse, or that a larger sample may have detected a significant association. However, our findings are consistent with the hypothesis that personality dysfunction in panic patients is related to the axis I pathology, perhaps as another manifestation of an anxiety diathesis particularly in patients evincing early significant anxiety difficulties, and is less directly attributable to physical or sexual trauma than in patients without other extant psychopathology. The presence of a personality disorder appears to increase the risk for a chronic and more symptomatic course of illness.1‘3 Panic patients with personality disorders in our study were more likely to have a chronic, unremitting course of illness; almost three quarters of these patients have never experienced a 2-month period of symptom remission since the onset of panic disorder despite a mean duration of illness of 8% years. In addition, we observed an
increased rate of other (comorbid) anxiety disorders, as well as depression in panic patients with personality dysfunction, consistent with previous reports of increased frequency of personality dysfunction in depressed panic disorder patients.14-16Although the direction of this relationship is unclear (ie, whether having a personality disorder predisposes a patient to anxiety or depression or whether the experience of other axis I pathology leads to the development of apparent personality disturbance), the association has important clinical implications. The presence of comorbid personality dysfunction, depression, and other anxiety disorders has an adverse impact on treatment outcome for panic di~order’~Zl2~26. , in addition, improvement in axis I disorders such as panic, depression and obsessive-compulsive disorder may be accompanied by improvement in personality function.27-29For some patients, an apparent personality disorder may respond to treatment of the axis I disorder. In other cases, additional interventions (ie, behavioral or dynamic psychotherapy) may be necessary to address the personality dysfunction so that the overall treatment response may be optimized. Our findings of an association between personality disorder and childhood anxiety disorders, comorbidity, and chronicity in panic disorder patients indicate the importance of understanding the role of axis II pathology in the longitudinal course of panic disorder and, further, underscore the importance of including assessments of axis II pathology in both acute and long-term studies of panic disorder.
REFERENCES 1. Mavissakalian M, Hamann M. DSM-III personality disorder and agoraphobia II: changeswith treatment. Compr Psychiatry 1987;28:356-361. 2. Reich JH. DSM-III personality disorders and the outcome of treated panic disorder. Am J Psychiatry 1988;145: 1149-1152. 3. Green M, Curtis GC. Personality disorders and panic patients: Response to termination of antipanic medication. J Person Disord 1988;2:303-314. 4. Mavissakalian M, Hamann M. DSM-III personality disorder and agoraphobia. Compr Psychiatry 1986;27:471479. 5. Friedman C, Shear MK, Frances A. DSM-III personality disorders in panic patients. J Person Disord 1987;1:132135. 6. Reich J, Troughton E. Comparison of DSM-III person-
ality disorders in recovered depressed and panic patients. J Nerv Ment Dis 1988;176:300-304. 7. Reich J, Noyes R, Yates W. Alprazolam treatment of avoidant personality traits in social phobic patients. J Clin Psychiatry 1989;50:91-95. 8. Mavissakalian M. The relationship between panic disorder/agoraphobia and personality disorders. Psychiatr Clin North Am 1990;13:661-684. 9. Rosenbaum JF, Biederman J, Gersten M, Hirshfeld DR, Meminger SR, Merman JB, et al. Behavioral inhibition in children with panic disorder and agoraphobia: a controlled study. Arch Gen Psychiatry 1988;45:463-470. 10. Biederman J, Rosenbaum JF, Hirshfeld DR, Faraone SV, Bolduc EA, Gersten M, et al. Psychiatric correlates of behavioral inhibition in young children of parents with and without psychiatric disorders. Arch Gen Psychiatry 1990;47: 21-26.
PERSONALITY
DISORDERS IN PANIC PATIENTS
11. Klein DF, Fink M. Psychiatric reaction patterns to imipramine. Am J Psychiatry 1962;119:432-438. 12. Pollack MH, Otto MW, Rosenbaum JF, Sachs GS, O’Neil C, Asher R, et al. The longitudinal course of panic disorder: findings from the Massachusetts General Hospital naturalistic study. J Clin Psychiatry 1990;51(Suppl A):12-16. 13. Ogata SN. Silk KR, Goodrich S, Lohr NE, Westen D, Hill EM. Childhood sexual and physical abuse in adult patients with borderline personality disorder. Am J Psychiatry 1990:147:1008-1013. 14. Mavissakalian M, Hamann MS. Correlates of DSMIII personality disorder in panic disorder and agoraphobia. Compr Psychiatry 1988;29:535-544. 15. Reich J. Troughton E. Frequency of DSM-III personality disorders in patients with panic disorder: comparison with psychiatric and normal controls. Psychiatry Res 1988;26: 89-100. 16. Klass ET, DiNardo PA, Barlow DH. DSM-IIIR personality diagnoses in anxiety disorder patients. Compr Psychiatry 1989;30:251-258. 17. American Psychiatric Association. Diagnostic and Statistic Manual of Mental Disorders. ed. 3, rev. Washington, DC: American Psychiatric Press, 1987. 18. Spitzer RL, Williams JBW, Gibbon M, et al. Structured Clinical Interview for DSM-III-R. New York, NY: Biometrics Research Department, New York State Psychiatric Institute, 1988. 19. Orvaschel H, Puig-Antich J. Schedule for Affective Disorder and Schizophrenia for School Aged Children (K-SADS-E). ed. 4. Pittsburgh, PA: Western Psychiatric Institute, 1987.
a3
20. Herjanic B, Reich W. Development of a structured psychiatric interview for children: agreement between child and parent. J Abnorm Child Psycho1 1982;10:307-324. 21. Hyler SE, Rieder RO. PDQ-R: Personality Diagnostic Questionnaire-Revised. New York, NY: New York State Psychiatric Institute, 1987. 22. Hurt SW, Hyler SE, Frances A. Clarkin JF, Brent R. Assessing borderline personality disorder with self report. clinical intelview, or semi-structured interview. Am J Psychiatry 1984;141:1228-1231. 23. Hyler SE, Skodol AE, Kellman HD, Oldham JM. Rosnick L. Validity of the Personality Diagnostic Questionnaire-Revised: comparison with two structured interviews. Am J Psychiatry 1990;147:1043-1048. 24. Widiger TA, Roger JH. Prevalence and comorbidity of personality disorders. Psychiatr Ann 1989;19:132-136. 25. Pfohl B, Stangl D, Zimmerman M. Structured Interview for DSM-III Personality Disorders (SID-P). Iowa City. IA: University of Iowa Hospitals and Clinics, 1986. 26. Noyes R, Reich J, Christiansen J, Suelzer M, Pfohl B. Coryell WA. Outcome of panic disorder: Relationship to diagnostic subtypes and comorbidity. Arch Gen Psychiatry 1990;47:809-818. 27. Reich J, Noyes R, Cotyell W, O’Gorman TW. The effect of state anxiety on personality measurement. Am J Psychiatry 1986;143:760-763. 28. Shea MT, Pilkonis P, Beckham E, Collins JF, Elkin I. Sotsky SM, et al. Personality disorders and treatment outcome in the NIMH treatment of depression collaborative research program. Am J Psychiatry 1990;147:711-718. 29. Riccardi J, Baer L, Jenike M. Personality change following treatment of OCD. Unpublished manuscript.
presence of a personality disorder in these patients was not significantly associated with a history of physical or sexual abuse in childhood. Our findings support the notion that an anxiety diathesis, demonstrated by significant difficulties with anxiety in childhood, influences the development of apparent personality dysfunction in panic patients. In other cases, personality pathology may reflect the presence of comorbid anxiety disorders or depression. The association of personality disorder in panic patients with a more unremitting course of illness underscores the importance of axis II pathology in understanding the longitudinal course of panic disorder. Copyright 0 1992 by W.B. Saunders Company
T
HE RELATIONSHIP between personality disorders and panic disorder is an important area of clinical concern; the presence of comorbid personality disorder in panic disorder patients adversely affects response to antipanic treatment’** and increases the likelihood of relapse after treatment discontinuation.3 Personality dysfunction in patients with panic disorder is apparently common, with reported rates ranging from 27% to 58%, depending on the study and diagnostic instrument used to assess personality.4,5 However, the nature of the association between panic and personality is unclear. Patients with recurrent panic attacks and high levels of anticipatory anxiety, phobic fear, and avoidance are often dependent, avoidant, and unassertive; whether these personality traits reflect a premorbid personality disorder or arise from the chronic experience of anxiety remains uncertain. Some personality characteristics, such as dependency, may improve with antipanic treatment, suggesting either that the dependent characteristics are secondary to panic disorder
From rhe Anxiety Clinical Research Unit, Clinical Psychopharacology Unit, Massachusetts General Hospital, Boston, MA. Address reprint requests to Mark H. Pollack, M.D., Anxiety Clinical Research Unit, Clinical Psychopharmacology Unit, Massachusetts General Hospital, ACC-815, 15 Parkman St, Boston, MA 02114. Copyright 0 I992 by W.B. Saunders Company 0010-440X/92/3302-0014$03.00/0 78
or that the treatment has specific effects on dependence. However, other traits, such as histrionic and avoidant, may be more resistant to change, arguing against the proposition that personality measures are simply epiphenomena of the panic disorder.4 Furthermore, no specific link has been demonstrated between particular personality disorders and panic disorder, as elevated rates of personality dysfunction also occur in other anxiety disorders such as social phobia and OCD, as well as in depression.6-8 Mavissakalian has suggested that panic disorder and some personality traits may be separate manifestations of a common diathesis.’ Evidence for a diathesis of risk for anxiety that may first manifest in childhood is supported by findings that offspring of panic disorder patients have higher rates of shyness and behavioral inhibition to the unfamiliar than comparison children.’ In addition, behaviorally inhibited children are at increased risk to develop anxiety disorders.” Furthermore, adult panic disorder patients often report a history of childhood anxiety disorders.“~” Significant difficulties with anxiety in childhood may also be associated with the development of dysfunctional personality characteristics that may interact with developmental and psychosocial protective or risk factors (eg, trauma) and manifest in the emergence of axis II disorders. We undertook the present study to examine whether panic patients with comorbid personal-
Comprehensive Psychiatry, Vol. 33, No. 2 (March/April), 1992: pp 78-83
PERSONALITY
79
DISORDERS IN PANIC PATIENTS
ity disorders showed evidence for an anxiety diathesis as manifested by a history of anxiety disorders in childhood and had experienced higher rates of a known risk factor for personality dysfunction (ie, early abuse). We also sought to assess the association between personality dysfunction and comorbid depression, comorbid anxiety disorders, and chronicity of the panic disorder. HYPOTHESIS
1
Panic Disorder Patients With Personality Disorders Are More Likely to Have a History of Childhood Anxiety Disorders
If panic disorder and its comorbid personality disorders are expressions of a common diathesis first manifest in childhood and influencing development, we would expect that adult panic patients with personality disorders would be more likely to have a history of an anxiety disorder during childhood. HYPOTHESIS
2
Panic Disorder Patients With a Comorbid Personality Disorder Are More Likely to Have Experienced Childhood Trauma
The risk for developing a personality disorder is reportedly higher for individuals exposed to early traumatic life experiences.13 If early trauma, in the form of sexual or physical abuse, influences the development of personality dysfunction, we would also expect to find a history of childhood abuse more commonly in panic patients with a concomitant personality disorder than in panic patients without a personality disorder. HYPOTHESIS
3
Panic Disorder Patients With Comorbid Personality Disorders Will Have Higher Rates of Comorbid Depression and Other Anxiety Disorders
Studies suggest that depression and a general quality of increased “anxiety proneness” are correlates of personality disorder in panic patients.‘4-1hAlthough the direction of the association is unclear (ie, whether a personality disorder predisposes to anxiety or depression or whether a mood or anxiety disorder influences the development of personality disturbance),
we predict personality disordered panic patients to be generally more severely ill as manifested in higher rates of depression and comorbid anxiety disorders. HYPOTHESIS 4
Panic Disorder Patients With a Comorbid Personality Disorder Will Have a More Chronic Course of Illness
In clinical trials of drug treatment of panic disorder, the presence of a personality disorder predicted poorer treatment outcome’.‘; thus, a personality disorder diagnosis may be associated with a lower likelihood of achieving a period of full remission from the panic disorder. METHOD To examine these issues, we assessed data from the first 100 patients entered into a naturalistic study of the longitudinal course of panic disorder. Description of study methods is detailed in a separate report.” Patients met DSMIII-R criteria for panic disorder with or without phobic avoidance.” Baseline assessments included the Structured Clinical Interview for DLY~M-III-R’~ to diagnose anxiety, affective, somatization. and substance use disorders, if present, and a detailed history of course of illness and past and current treatment. History of childhood anxiety disorders was examined using structured interviews including the Schedule for Affective Disorder and Schizophrenia (KSADS) items to assess childhood agoraphobia, overanxious disorders, separation anxiety disorder, and social phobia, and the Diagnostic Interview for Children and Adolescents (DICA-P) for avoidant disorder.“‘” Subjects were drawn from a naturalistic study that enrolled patients entering or already receiving treatment for panic disorder, and who provided informed consent for participation in the study. Failure to achieve at least a 2-month period of remission at some point since the onset of panic disorder was used as a measure of chronicity of illness. Remission was defined as a period of no panic attacks and phobic avoidance and absent or very mild anticipatory anxiety that did not interfere with function. The presence of personality disorders was assessed using the Personality Diagnostic Questionnaire-Revised (PDQR), a self-report questionnaire of 152 true/false questions to assess axis II disorders as determined by DSM-III-R.” Personality disorders in DSM-III-R are divided into three groups: cluster A, the “schizoid” cluster, including paranoid, schizoid, and schizotypal personality disorders; cluster B. the “dramatic” cluster, including antisocial, borderline, histrionic, and narcissistic personality disorders; and cluster C. the “anxious” cluster, including avoidant. dependent. obsessive-compulsive, and passive-aggressive personality disorders.” The PDQ has been found to have adequate test-retest reliability for most disorders (K = .56 or above for paranoid, schizotypal, antisocial. borderline, avoidant. and compulsive personality disorders),” although it may
80
POLLACK ET AL
generate a number of false-positive diagnoses and overestimate the overall prevalence of personality disordersZ3 Hyler et a12’have suggested that the instrument’s high sensitivity, but only modest specificity, supports its use as a screening instrument for the presence of personality disorder. RESULTS
Subjects
Sixty-three of the 100 patients were women and 37 were men, with a mean age of 40.4 years (k11.3). The mean age at onset of panic disorder was 29.8 years (+llS), with a duration of illness of 8.6 years (r8.9; range, 1 to 34) when the current episode of treatment was initiated. Overall Rates of Personality Disorders
Of the 100 patients studied, 42 met criteria for at least one personality disorder as determined by the PDQ-R. The most common personality disorders were histrionic, borderline, avoidant, and compulsive (Table 1). Patients meeting criteria for a personality disorder in one of three diagnostic clusters of the DSMIII-R frequently met criteria for more than one cluster: 19 of the 42 (45%) patients meeting criteria for one personality disorder met criteria for a personality disorder in each of the three clusters, while 10 (24%) met criteria for one cluster diagnosis only. This overlap between personality disorders detected by the PDQ-R in this sample is consistent with the expected high Table 1. Rates of Personality Disorders in Panic Disorder Patients (n = 100) Overall
42%
Cluster A (25%) A only With other clusters Schizoid Schizotypal Paranoid
3% 22% 11% 8% 19%
sensitivity and moderate specificity of this instrumen? and limited the precision by which each cluster could be examined for individual associations with patient variables. Indeed, in the following analyses, when significant associations between the presence of a personality disorder and patient characteristics were obtained, they were generally obtained for each cluster. Thus, we report on the overall association between the presence of any personality disorder and specific patient characteristics; these associations were examined with Pearson chi-square tests. Significance was defined as P I .05 for each predicted association. Hypothesis 1: Panic Disorder and Childhood Anxiety Disorder
Panic disorder patients with a personality disorder according to the PDQ-R were more likely to have a history of a childhood anxiety disorder than those without a personality disorder. Seventy-six percent of patients with a personality disorder had a childhood history of anxiety, compared with 38% of patients without a personality disorder (x’ = 14.36, P < .OOl). Hypothesis 2: Personality Disorder and Past History of Childhood Abuse
The presence of a personality disorder was not significantly associated with a history of childhood sexual abuse or physical abuse. Twenty-four percent of patients with a personality disorder reported being sexually abused as children, compared with 12% of patients without a personality disorder (x’ = 2.56, P < .ll). Nineteen percent of patients with and 10% of those without a personality disorder reported physical abuse during childhood (x’ = 1.53, P < .22).
Cluster B (34%) B only With other clusters
5% 29%
Histrionic
22%
Borderline
23%
Narcissistic
12%
Antisocial
4%
Cluster C (34%) C only With other clusters
2% 32%
Avoidant
22%
Obsessive-compulsive
21%
Passive-aggressive
8% 7%
Hypothesis 3: Personality Disorders and Comorbidity of Other Psychiatric Disorders
Patients with a personality disorder were significantly more likely to have a lifetime history of major depression; 69% of patients with a personality disorder had lifetime history of a depressive disorder, compared with 36% of patients without a personality disorder (x’ = 10.51, P < .Ol). In addition, patients with a personality disorder were more likely to meet criteria for another anxiety disorder (social
PERSONALITY
DISORDERS IN PANIC PATIENTS
phobia, generalized anxiety disorder, or simple phobia); 79% of patients with a personality disorder had another anxiety disorder, as compared with 52% of those without a personality disorder (x’ = 7.53, P < .Ol). Hypothesis 4: Personality Disorder and Chronic@ of Course
Panic patients with a personality disorder were less likely to have achieved even a 2-month period of remission of the panic disorder. Only 29% of patients with a personality disorder had at least a 2-month period of remission, compared with 59% of patients without a personality disorder (x” = 8.86, P < .Ol). DISCUSSION
Forty-two percent of patients with panic disorder in this study met criteria for at least one personality disorder as assessed by the PDQ-R. These findings are consistent with the rates of personality disorders in panic disorder patients previously reported by others.‘.14 Personality disorders identified were from all three personality clusters, with 45% of patients with a personality disorder meeting criteria for a disorder in each of the three clusters. Meeting criteria for more than one personality disorder may be common,‘3,‘4 and reflect a number of influences, including the propensity of the PDQ-R to overestimate the prevalence of personality disorders, the failure of the diagnostic criteria to identify discrete personality syndromes, and the possibility that axis I pathology may truly be associated with multiple dysfunctional personality characteristics. Our findings must be considered in light of certain methodologic limitations. The overlap among diagnostic categories did not permit examination of associations with individual personality diagnoses. Perhaps the most significant limitation of this study is the use of the PDQ-R, a self-report instrument that, though sensitive to the presence of personality dysfunction, may not be adequately specific to permit more refined analysis of specific personality characteristics. However, this limitation may be inherent, to some extent, to the current study of personality disorders; diagnostic instruments and the criteria themselves continue to undergo further empirical refinement and there is as yet no clear
“gold standard” for personality assessment. The PDQ-R, given its ease of administration and reliability, may be useful in generating and examining initial hypotheses. Results of this study should be viewed as preliminary and warrant further evaluation with standardized clinical interview-based instruments. Nevertheless, results from two earlier studies of personality and panic disorder suggest high concordance between the PDQ and the clinician rated Structured Interview for DSM-III Personality Disorder (SIDP)” both for the presence or absence of any personality disorder, as well as for some individual diagnoses.‘4.‘” Thus, our use of the PDQ-R to establish the presence of personality dysfunction may be reasonable. The most frequently detected disorders in our sample were histrionic, borderline, avoidant, and obsessive-compulsive personality disorders. Dependent personality disorder was not as prevalent in our sample compared with previous reports.‘4~‘5 Since most of our patients were receiving treatment at the time of assessment, the lower rates of dependent personality observed may reflect previously reported findings of a decrease in dependent traits with treatment of panic disorder.’ The presence of a personality disorder was significantly associated with a history of an anxiety disorder in childhood. This finding may reflect the development of maladaptive character styles in response to long-standing exposure to anxiety, excessive physiological arousal, and the fearfulness associated with childhood anxiety disorders. Alternatively, panic disorder and personality disorder may be separate outcomes of the same biological or environmental risk factors that predispose to anxiety in childhood. Some children at high risk for the development of panic disorder manifest early shyness, fearfulness, and inhibition to the unfamiliar.’ These traits may be the early expressions of a diathesis for anxiety later expressed in adulthood as panic disorder or agoraphobia. Our finding of an association between personality disorders in adulthood and childhood anxiety disorders, suggests that some aspects of apparent personality dysfunction may reflect the presence of anxiety and a characteristic pattern of response to it, first manifest in childhood and variably expressed over time as axis I and axis II pathology.
82
POLLACK ET AL
This hypothesis might best be examined by longitudinal study of a cohort of behaviorally inhibited children to examine protective and risk factors for the adult expression of these characteristics. Although there were weak trends in our data in the direction of an association between early trauma and the development of adult personality disorder, these did not reach statistical significance and our findings for patients with panic disorder do not strongly support previous descriptions of a role for childhood sexual or physical in the development of adult personality disorders.13 It is possible that patients may have underreported their history of abuse, or that a larger sample may have detected a significant association. However, our findings are consistent with the hypothesis that personality dysfunction in panic patients is related to the axis I pathology, perhaps as another manifestation of an anxiety diathesis particularly in patients evincing early significant anxiety difficulties, and is less directly attributable to physical or sexual trauma than in patients without other extant psychopathology. The presence of a personality disorder appears to increase the risk for a chronic and more symptomatic course of illness.1‘3 Panic patients with personality disorders in our study were more likely to have a chronic, unremitting course of illness; almost three quarters of these patients have never experienced a 2-month period of symptom remission since the onset of panic disorder despite a mean duration of illness of 8% years. In addition, we observed an
increased rate of other (comorbid) anxiety disorders, as well as depression in panic patients with personality dysfunction, consistent with previous reports of increased frequency of personality dysfunction in depressed panic disorder patients.14-16Although the direction of this relationship is unclear (ie, whether having a personality disorder predisposes a patient to anxiety or depression or whether the experience of other axis I pathology leads to the development of apparent personality disturbance), the association has important clinical implications. The presence of comorbid personality dysfunction, depression, and other anxiety disorders has an adverse impact on treatment outcome for panic di~order’~Zl2~26. , in addition, improvement in axis I disorders such as panic, depression and obsessive-compulsive disorder may be accompanied by improvement in personality function.27-29For some patients, an apparent personality disorder may respond to treatment of the axis I disorder. In other cases, additional interventions (ie, behavioral or dynamic psychotherapy) may be necessary to address the personality dysfunction so that the overall treatment response may be optimized. Our findings of an association between personality disorder and childhood anxiety disorders, comorbidity, and chronicity in panic disorder patients indicate the importance of understanding the role of axis II pathology in the longitudinal course of panic disorder and, further, underscore the importance of including assessments of axis II pathology in both acute and long-term studies of panic disorder.
REFERENCES 1. Mavissakalian M, Hamann M. DSM-III personality disorder and agoraphobia II: changeswith treatment. Compr Psychiatry 1987;28:356-361. 2. Reich JH. DSM-III personality disorders and the outcome of treated panic disorder. Am J Psychiatry 1988;145: 1149-1152. 3. Green M, Curtis GC. Personality disorders and panic patients: Response to termination of antipanic medication. J Person Disord 1988;2:303-314. 4. Mavissakalian M, Hamann M. DSM-III personality disorder and agoraphobia. Compr Psychiatry 1986;27:471479. 5. Friedman C, Shear MK, Frances A. DSM-III personality disorders in panic patients. J Person Disord 1987;1:132135. 6. Reich J, Troughton E. Comparison of DSM-III person-
ality disorders in recovered depressed and panic patients. J Nerv Ment Dis 1988;176:300-304. 7. Reich J, Noyes R, Yates W. Alprazolam treatment of avoidant personality traits in social phobic patients. J Clin Psychiatry 1989;50:91-95. 8. Mavissakalian M. The relationship between panic disorder/agoraphobia and personality disorders. Psychiatr Clin North Am 1990;13:661-684. 9. Rosenbaum JF, Biederman J, Gersten M, Hirshfeld DR, Meminger SR, Merman JB, et al. Behavioral inhibition in children with panic disorder and agoraphobia: a controlled study. Arch Gen Psychiatry 1988;45:463-470. 10. Biederman J, Rosenbaum JF, Hirshfeld DR, Faraone SV, Bolduc EA, Gersten M, et al. Psychiatric correlates of behavioral inhibition in young children of parents with and without psychiatric disorders. Arch Gen Psychiatry 1990;47: 21-26.
PERSONALITY
DISORDERS IN PANIC PATIENTS
11. Klein DF, Fink M. Psychiatric reaction patterns to imipramine. Am J Psychiatry 1962;119:432-438. 12. Pollack MH, Otto MW, Rosenbaum JF, Sachs GS, O’Neil C, Asher R, et al. The longitudinal course of panic disorder: findings from the Massachusetts General Hospital naturalistic study. J Clin Psychiatry 1990;51(Suppl A):12-16. 13. Ogata SN. Silk KR, Goodrich S, Lohr NE, Westen D, Hill EM. Childhood sexual and physical abuse in adult patients with borderline personality disorder. Am J Psychiatry 1990:147:1008-1013. 14. Mavissakalian M, Hamann MS. Correlates of DSMIII personality disorder in panic disorder and agoraphobia. Compr Psychiatry 1988;29:535-544. 15. Reich J. Troughton E. Frequency of DSM-III personality disorders in patients with panic disorder: comparison with psychiatric and normal controls. Psychiatry Res 1988;26: 89-100. 16. Klass ET, DiNardo PA, Barlow DH. DSM-IIIR personality diagnoses in anxiety disorder patients. Compr Psychiatry 1989;30:251-258. 17. American Psychiatric Association. Diagnostic and Statistic Manual of Mental Disorders. ed. 3, rev. Washington, DC: American Psychiatric Press, 1987. 18. Spitzer RL, Williams JBW, Gibbon M, et al. Structured Clinical Interview for DSM-III-R. New York, NY: Biometrics Research Department, New York State Psychiatric Institute, 1988. 19. Orvaschel H, Puig-Antich J. Schedule for Affective Disorder and Schizophrenia for School Aged Children (K-SADS-E). ed. 4. Pittsburgh, PA: Western Psychiatric Institute, 1987.
a3
20. Herjanic B, Reich W. Development of a structured psychiatric interview for children: agreement between child and parent. J Abnorm Child Psycho1 1982;10:307-324. 21. Hyler SE, Rieder RO. PDQ-R: Personality Diagnostic Questionnaire-Revised. New York, NY: New York State Psychiatric Institute, 1987. 22. Hurt SW, Hyler SE, Frances A. Clarkin JF, Brent R. Assessing borderline personality disorder with self report. clinical intelview, or semi-structured interview. Am J Psychiatry 1984;141:1228-1231. 23. Hyler SE, Skodol AE, Kellman HD, Oldham JM. Rosnick L. Validity of the Personality Diagnostic Questionnaire-Revised: comparison with two structured interviews. Am J Psychiatry 1990;147:1043-1048. 24. Widiger TA, Roger JH. Prevalence and comorbidity of personality disorders. Psychiatr Ann 1989;19:132-136. 25. Pfohl B, Stangl D, Zimmerman M. Structured Interview for DSM-III Personality Disorders (SID-P). Iowa City. IA: University of Iowa Hospitals and Clinics, 1986. 26. Noyes R, Reich J, Christiansen J, Suelzer M, Pfohl B. Coryell WA. Outcome of panic disorder: Relationship to diagnostic subtypes and comorbidity. Arch Gen Psychiatry 1990;47:809-818. 27. Reich J, Noyes R, Cotyell W, O’Gorman TW. The effect of state anxiety on personality measurement. Am J Psychiatry 1986;143:760-763. 28. Shea MT, Pilkonis P, Beckham E, Collins JF, Elkin I. Sotsky SM, et al. Personality disorders and treatment outcome in the NIMH treatment of depression collaborative research program. Am J Psychiatry 1990;147:711-718. 29. Riccardi J, Baer L, Jenike M. Personality change following treatment of OCD. Unpublished manuscript.
