LE1TERS TO THE EDITOR system. “ Clinical experience ple state/trait dichotomy: terweave with developmental to the simraises two objections 1) traumatic events in early life inforces to form admixtures of with state survivors diagThis often meet criteria for both disorder) and 2) reasonably in setting 3 (10 true positives and Si false positives). Only when the base rate is very high, as in Hyler et al.’s two studies (3, 4), does the magnitude of false positive cases diminish. Finally, Skodol et al. offer a useful suggestion to deal the problem of axis II assessment validity: considering noses as definite whenever two or more methods concur. certainly is agreement clearly with most valid a more when stringent a weaker method, method such is in as extensive follow-up data, as in the earlier report from my colleagues and me (5), which they cite. However, in the case of two methods of unknown measurement validity, the situation is more complicated. For instance, if two cross-sectional interviews are unbiased in their measurement error (i.e., it is random), then the danger that the sample obtained will produce findings that are not generalizable to the population is lessened. However, in the case where both instruments produce measurement whelming I . Hunt C, Andrews G: Measuring personality disorder: the use of self-report questionnaires. J Personality Disorders 1992; 6:125133 2. Rennebeng B, Chambless DL, Dowdall DJ, Fauerbach JA, Gracely EJ: The Structured Axis II and the Millon Clinical validity study of personality tients. J Personality Disorders 3. Hyler SE, Skodol AE, Oldham lidity ofthe Personality lication in an outpatient 77 4. Hyler SE, Skodol Validity comparison of Diagnostic sample. AE, Questionnaire-Revised: Compr Psychiatry Kellman HD, Interview Schedule J Personality SH, Hoke JM, Personality L, O’Connell and DSM-III antisocial Disorders 1987; 1:121-131 J. CHRISTOPHER Borderline Oldham 1992; a rep33:73- Disorder and PERRY, Montreal, who and persistent symp- merely experience chological trauma in adult begs life lead the question: Can to borderline psy- personality disorder? Personality organization sumptions stable, and catastrophic in adults is founded on certain about being reasonably worthwhile, reasonably safe. As Janoff-Bulman events who can shatter had previously such basic basic as- reasonably (2) points out, assumptions, consolidated them. even Psychologi- cal trauma disrupts the epigenetic framework of personality at its base and the ego is fundamentally changed. Some survivors become descriptively borderline in all criteria except age at onset. Development is a lifelong process. It is therefore reasonable to assume that vulnerabilities persist long after adolescence. The usual division between axes I and II may not apply to the character pathology that follows psychological trauma. Nor- mal adults may develop clinical sense, label this new a personality borderline. Some state a personality must, by definition, arbitrary definitions theory development need for consistency. disorder that will argue that disorder because begin in early is, in every you cannot personality life. Nonetheless, may limit both accurate description and as much as or more than they satisfy a J Psychiatry 2. 1993; Janoff-Bulman and conceptual and PTSD. Am 150:19-27 R: The aftermath of shattered assumptions, in Trauma CR. New York, Brunnen/Mazel, victimization: and Its Wake. 1985 rebuilding Edited by Figley L: HAROLD S. KUDLER, Durham, M.D. N.C. ME: The personal- M.P.H., M.D. Que., Canada PTSD Drs. Gunderson To and Sabo Reply We welcome Dr. Kudler’s extension of our interface of borderline personality disorder and PTSD into the boundaries between axis I and axis II. The emerging awareness of the sequelae of prolonged and repeated trauma ( 1 ), the emerging data documenting how sensitization ThE discussion may To ThE EDITOR: Current DSM classification holds that mental disorders represent either an axis I state or an axis II trait and that these two pathologies are fundamentally different. Gunderson and Sabo ( 1 ) follow this reasoning in reporting that posttraumatic stress disorder (PTSD) (state) and borderline personality disorder (trait) can be distinguished from one another by careful attention to longitudinal history. They also point to theoretical problems posed by “the presence of seemingly separate but related categories in our diagnostic 1906 profound 1 . Gunderson JG, Sabo AN: The phenomenological interface between borderline personality disorder N: Va- Rosnick may develop incest borderline personality adults suffering over- REFERENCES Interview for DSM-III-R, Inventory: a concurrent among anxious outpa- I 992; 6:117-124 JM, Kellman HD, Doidge of childhood PTSD and normal the Personality Diagnostic Questionnaire-Revised: with two structured interviews. Am J Psychiatry 1990; 147:1043-1048 S. Perry JC, Lavoni PW, Cooper Diagnostic ity disorder. Clinical Multiaxial disorders events catastrophic disorder REFERENCES (as in adult toms descriptively equivalent to those of borderline patients. All that separates the latter patients from patients with “true” borderline personality disorder is the requirement that these traits be present from an early age. The proposed DSM-IV concept of enduring personality change following er- nor that is biased in unknown directions (e.g., underdiagnosing disorder X or underdiagnosing criteria I and 2 of disorder X), then the sample obtained will be a product of both biases, and hence there is a greater danger in generalizing the findings from the sample. In short, we still need to study the assessment validity of our instruments to determine the generalizability of the samples they obtain. Currently, I am placing my bet for the most likely valid cross-sectional diagnostic method on the development of a guided clinical interview, which I will validate against extensive follow-up data. Yet, the issue that Dr. Skodol et al. raise warrants systematic attention. As they suggest, until the measurement characteristics of our instruments are known, it is useful to compare a number of different strategies. and trait EDITOR: of the reactivate long dormant the evidence that kindling (3) all suggest ways that sonality Dr. Kudler can be unexpectedly that traumatic can even a seemingly altered. overwhelming experiences (2), and induce gene transcription formed and stable per- Moreover, trauma can we agree “shatter . . . with basic assumptions” such that the person’s basic personality organization is altered. Yet, whether the normal adult personality can or will shatter along has borderline personality standing developmental Am J the same fault lines as the disorder as an outcome processes seems unlikely Psychiatry 150:12, December adult who of longand cer- 1993 LETTERS tainly uncertain. personality Moreover, organization if the adult’s of persons borderline personality disorder, and treatment implications? velopmentally based have similar prognostic seems unlikely. This would be important person as having borderline more likely that the person ality disorder-like phenomena “normal” personality personality derline” normal to know before identifying the personality disorder. It seems who develops borderline personas an adult was not previously but was already quite vulnerable organization prior to the trauma. change in adults that became after exposure to trauma, but healthy person who personality disorder fractured with dedoes it This also to that conforms has been converted “in every clinical in intrapsychic We have seen “descriptively we have not borseen a to having borderline sense.” This must be rare. We have also occasionally seen patients identified as borderline who undergo impressive “remissions” and others with PTSD who were mislabeled as having borderline personality disorder (primarily self-mutilators). But what we have most often seen are adult patients with borderline personality disorder who, because of childhood trauma, have been mislabeled as having PTSD. These are the clinical realities that prompted our article. The larger issue raised by Dr. Kudler and by our article is whether and when the division between axis I and axis II is valid and useful. It is our sense that placing personality disordens on a separate axis has served useful purposes by reminding clinicians of the more enduring and developmentally significant precursors to axis I disorders but that, in many instances, this division may well have now outlived its utility. When this issue receives its needed review, Dr. Kudler’s thoughtful observations will be contributory. nitely different. operational symptoms these Typical depressive symptoms, criteria of the authors, occur or change rapidly between mixed states TO are unstable, ThE EDITOR as cited in the together with manic each other. Most of which means the combina- tion of the symptoms is not fixed nor does the duration of mixed symptoms take all or most of the time of the phase (2-4). As cited by McElroy et al., Kraepelin further described stable mixed and mood cal form. Agitated stead, syndromes for most with depression agitation opposite of the phase is not is argued alterations in a similar interpreted of drive psychopathologias a mixed to be a severe form state. In- of the subjective feeling of unrest, while inhibition of drive is still present with the typical symptoms of subjective resistance against intended actions and narrowing of thoughts (5). In our sample of 187 manic or mixed phases in 109 bipolar patients we found dysphoric mania according to the afore- mentioned criteria in only 8% of the phases. Those patients who had one dysphonic phase tended to be dysphonic in the second phase or to show dysphonic components in the depressed states. Mixed states were present in 60% of the phases. Most of them were of short duration (2-3 days) and started after 3-4 weeks of neuroleptic treatment or before the switch to a depressive phase. Ten patients had Schneidenian first-rank symptoms during the mixed states making differential diagnosis regarding schizophrenia difficult. Stable mixed symptoms were present in only three of the phases, and it seems that these states our opinion, high-dose tend to reoccur in the same patients. In lithium has the best therapeutic results. Neuroleptics act more quickly, pressive states often occur. but changes to full-blown de- REFERENCES REFERENCES 1 . Herman JL: Sequelae of prolonged and repeated trauma: evidence for a complex posttraumatic syndrome (DESNOS), in Posttnaumatic Stress Disorder: DSM-IV and Beyond. Edited by Davidson JRT, Foa EB. Washington, DC, American Psychiatric Press, I 993 2. Antelman SM: Time-dependent for a new approach ful stimuli. Drug sensitization to pharmacothenapy: Development Research as the cornerstone drugs as foreign/stress- 1988; 14:1-30 3. Post RM: Transduction of psychosocial stress into the neurobiology of recurrent affective disorder. Am J Psychiatry 1992; 149: 999-1010 I . McElroy SL, Keck Swann AC: Clinical dysphonic or mixed PE Jn, Pope and research mania HG, Hudson implications or hypomania. JI, Faedda GL, of the diagnosis of Am J Psychiatry I 992; 149:1633-1644 2. Specht G: Uber die Kandinalfnage den Paranoia. Zbl Nervenheilk Psychiatr 1908; 3 1:817-883 Bernen P: Psychiatnische Systematik, 3 AufI. Wien, Huben, 1982 3. 4. Burger-Prinz H: Probleme den Phasischen Psychosen. Stuttgart, Enke, 1961 5. Ebert D: Alterations of drive in differential diagnosis of mild depressive genomorphic disorders-evidence affective psychosis. for the spectrum Psychopathology concept of endo1992, 25:23- 28 JOHN G. GUNDERSON, ALEX N. SABO, Belmont, M.D. M.D. Mass. DIETER THOMAS PETER Dysphonic or Mixed pointing out some different interpretations. Differences can be outlined between dysphonic mania and stable or unstable mixed mania. It has already been described in 1908 (2) that dysphonic mania may be a manifestation of pure mania in bipolar disorder. Since then it was defined by increased drive, aggressiveness, and querulousness combined with irritability, displeasure, and constant feelings of discomfort and dissatisfaction. tunes may occur Am J Psychiatry M.D. M.D. MARTUS, PH.D. Erlangen, Germany Mania To ThE EDITOR: In their article on dysphonic or mixed mania, Dr. McElroy and associates ( 1 ) focused on problems of definitions and clinical implications of these disorders. We would like to stress a European view with its psychopathological tra- ditions EBERT, LOEW, Paranoid but euphoria 1 50:1 or is absent. 2, December typical Mixed 1993 depressive states fea- are defi- To mE EDITOR: We read with interest the article by Susan McElroy, M.D., and colleagues in which they present a cornprehensive overview of the clinical status of dysphonic or mixed mania and propose operational criteria for its diagnosis. Mixed states have not received their due share of research inquiry, and this is, in part, because of the nonavailability of comprehensive yet valid diagnostic criteria. We wish to share some of our observations regarding this highly polymorphous and dynamic clinical entity and highlight some of the problems encountered in developing a set of criteria. The terms mixed states and dysphonic mania are used interchangeably. As originally described by Kraepelin (1), depressive on anxious mania was only one of the six different mixed 1907